Paleobotany -The Biology and Evolution of Fossil Plants

1,253 Pages • 580,073 Words • PDF • 188 MB
Uploaded at 2021-09-24 06:09

This document was submitted by our user and they confirm that they have the consent to share it. Assuming that you are writer or own the copyright of this document, report to us by using this DMCA report button.


This page intentionally left blank

PALEOBOTANY The Biology and Evolution of Fossil Plants Second Edition


Department of Ecology and Evolutionary Biology and Natural History Museum and Biodiversity Research Center, The University of Kansas, Lawrence, Kansas

EDITH L. TAYLOR Department of Ecology and Evolutionary Biology and Natural History Museum and Biodiversity Research Center, The University of Kansas, Lawrence, Kansas

MICHAEL KRINGS Bayerische Staatssammlung für Paläontologie und Geologie und GeoBio-Center LMU, Munich, Germany


Academic Press is an imprint of Elsevier 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 32 Jamestown Road, London NW 1 7BY, UK 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA 360 Park Avenue South, New York, NY 10010-1710, USA Copyright © 2009, Elsevier Inc. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher. Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (44) (0) 1865 843830; fax (44) (0) 1865 853333; email: [email protected]. Alternatively visit the Science and Technology Books website at for further information. Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-12-373972-8

For information on all Academic Press publications visit our website at

Typeset by Charon Tec Ltd., A Macmillan Company. ( Printed and bound in the USA 09 10 11

10 9 8 7 6 5 4 3 2 1

CONTENTS Chemical Fossils Ancient DNA Mummification Amber Summary Discussion Palynology Geochronology and Biostratigraphy Paleoecology Absolute Dating Geologic Timescale Biological Correlation Systematics and Classification Nomenclature of Fossil Plants Classification of Organisms Background Reading

Preface xv Acknowledgments xvii About the Authors xxi



Introduction to Paleobotany, How Fossil Plants are Formed 1 What Is Paleobotany? The Objectives of Paleobotany Reconstructing the Plants Evolution of Plant Groups Form and Function in Fossil Plants Biostratigraphy and Correlation Paleoecology: Plants in Their Environment Determining Paleoclimate from Fossil Plants Tree Rings Nearest Living Relative Leaf Physiognomy Stomatal Index Summary Preservation: How Plant Fossils are Formed and Preserved Depositional Environments of Fossil Plants Compressions Cuticle Biofilms and Plant Fossil Preservation Electron Microscopy Confocal Microscopy Maceration and Dégagement Other Techniques Coal and Charcoal Impressions Molds and Casts Cellular Preservation Permineralization Peel Technique Coal Balls Other Permineralizations Petrifaction Unaltered Plant Material

1 2 2 3 4 4 5 6 6 6 7 7 7



Precambrian Life


The Origin of Life on Earth Origin of Life: Theory and Biology Earliest Record of Life on Earth Historical Background Earliest Records of Life: Paleoarchean (3.6–3.2 Ga) Geochemistry Microfossils (Body Fossils) Isua Greenstone Belt, Greenland Warrawoona Group, Australia Barberton Greenstone Belt, South Africa Stromatolites Sedimentary Evidence Mesoarchean–Neoarchean Life Conclusions: Archean Life Oxygenation of the Earth (2.45–2.2 Ga) Proterozoic Life Paleoproterozoic Origin of Eukaryotes Mesoproterozoic Earliest Multicellular Life Neoproterozoic Bitter Springs Biota

8 8 10 13 16 17 17 17 18 18 21 22 23 25 25 27 29 30 30


32 33 33 33 34 34 36 37 38 39 40 40 41 42 42

44 46 47 47 47 47 49 49 49 51 52 53 54 55 57 59 59 61 64 64 64 65



Stromatolites Other Microfossils Doushantuo Formation Conclusions CHAPTER

66 67 70 70


Fungi, Bacteria, and Lichens 71 Fungi Earliest Fossil Fungi Systematics of Fungi Chytridiomycota Zygomycota Glomeromycota Ascomycota Basidiomycota Other Fungal Remains Fungal Life-History Strategies Saprotrophism Parasitism Mutualism Fungi–Animal Interactions Geologic Activities of Fungi Epiphyllous Fungi Fungal Spores Fungal-like Organisms Peronosporomycetes (Oomycota) Eubacteria and Archaea Archaea Eubacteria Cyanobacteria Lichens CHAPTER


71 73 77 77 82 84 90 93 97 98 98 99 103 105 107 108 111 112 112 112 113 113 115 117



Chlorophyta (Green Algae) Prasinophyceae Chlorophyceae Volvocales Tetrasporales Chlorococcales Ulvophyceae Dasycladales Receptaculitida and Cyclocrinales Caulerpales Taxa Incertae Sedis Charophyceae Charales Zygnematales Euglenophyta

123 124 126 126 126 127 128 128 130 130 133 133 134 138 138

Dinophyta (Dinoflagellates) Heterokontophyta Bacillariophyceae (Diatoms) Dictyochophyceae (Silicoflagellates) Xanthophyceae (Yellow-Green Algae) Phaeophyceae (Brown Algae) Prymnesiophyta (Haptophytes) Rhodophyta (Red Algae) Solenoporaceans Other Calcified Red Algae Corallinales Uncalcified Red Algae Acritarcha (Acritarchs) CHAPTER


Hornworts and Bryophytes


Early Fossil Evidence Anthocerotophyta (Hornworts) Bryophyta (Bryophytes) Marchantiophytina (Liverworts or Hepatophytes) Bryophytina (Mosses) CHAPTER


The Move to the Land

139 141 141 142 142 143 144 145 146 149 149 150 158

163 165 166 167 174


Enigmatic Organisms Nematophytes Prototaxites Nematothallus Nematoplexus Nematasketum diversiforme Pachytheca Spongiophytaceae Spongiophyton Orestovia Other Enigmatic Organisms Protosalvinia Parka Isolated Fragments: Clues to the Transition to Land? Cuticle and Cuticle-Like Material Spores and Spore Tetrads Tubes Land Plant Ancestors The Transition to Land Anchorage and Water Uptake Structural Support and Water Transport Protection Against Desiccation and Radiation Gas Exchange Reproduction on Land Life History Biology

180 180 180 183 183 183 184 185 185 186 186 186 188 189 189 189 192 193 194 194 195 195 195 196 196


Homologous Theory Antithetic Theory Animals A Fungal Partner Conclusion


196 196 198 198 199


Introduction to Vascular Plant Morphology and Anatomy 201 Plant Organography Cell Types Parenchyma Collenchyma Sclerenchyma Tracheary Elements Tracheids Vessel Elements Sieve Elements Plant Tissues and Primary Growth Xylem Tissue Phloem Tissue Meristems Epidermis Cuticle Stomata Trichomes Anatomy of Stems and Roots Arrangement of Primary Tissues Primary Xylem Maturation Patterns Secondary Development Vascular Cambium Cork Cambium (Phellogen) Secondary Xylem Secondary Phloem Stele Types Primitive Vascular Plants (Vascular Cryptogams) Seed Plants Leaf Morphology and Anatomy Leaf Anatomy Leaf Evolution Further Reading


202 203 203 203 203 204 204 206 206 207 207 207 208 208 209 209 210 210 210 212 212 212 213 214 216 216 216 219 221 221 222 222


Early Land Plants with Conducting Tissue 223 Conducting Elements in Early Land Plants History of Discovery

224 225

Rhyniophytes Rhynie Chert Plants Aglaophyton major Rhynia Gwynne-vaughanii Horneophyton lignieri Asteroxylon mackiei Nothia aphylla Trichopherophyton teuchansii Ventarura lyonii Gametophyte Generation Other Rhyniophytes Discussion: Rhyniophyte Evolution Zosterophyllophytes Zosterophyll Evolution Trimerophytes Trimerophyte Evolution Early Land Plant Evolution CHAPTER



vii 227 228 229 235 237 238 239 241 241 241 246 251 252 259 259 262 263


Evolution of the Microphyll Drepanophycales Protolepidodendrales Lepidodendrales Vegetative Features Stem Surface and Leaf Bases Stem Anatomy Cortical Tissues Stem Development Leaves Underground Organs Development of Underground Organs Reproductive Biology Microsporangiate and Bisporangiate Cones Megasporangiate Cones Gametophytes Sigillariaceae Leaf Bases Leaves Stem Structure Underground Organs Reproductive Biology Other Lepidodendrid Genera Lycopodiales Selaginellales Pleuromeiales Isoetales Putative Lycopsids Conclusions

267 268 271 279 282 282 285 286 287 289 289 293 294 295 297 302 303 304 305 305 306 306 307 310 312 316 320 325 326







Pseudoborniales Sphenophyllales Devonian Sphenophyllales Sphenophyllum Leaves Stem Anatomy Roots Reproductive Biology Other Sphenophyllales Ecology Equisetales Calamitaceae Archaeocalamites Calamites Pith Casts Stem Anatomy Extraxylary Tissues Growth and Development Roots Leaves Other Calamitean Leaves Reproductive Biology Spores Tchernoviaceae and Gondwanostachyaceae Vegetative Body Reproductive Biology Equisetaceae Forms with Uncertain Affinities Sphenophyte Evolution CHAPTER

331 332 333 334 334 335 337 337 338 341 342 343 343 345 349 350 352 352 353 354 357 358 366 368 368 369 371 376 379


Ferns and Early Fernlike Plants 383 Evolution of the Megaphyll Cladoxylopsida Pseudosporochnales Calamophyton Plant Iridopteridales Phylogenetic Position of the Cladoxylopsids Early Fernlike Plants Rhacophytales Rhacophyton Other Taxa Systematics of the Rhacophytales Coenopterid Ferns Stauropteridales Zygopteridales

386 387 388 396 398 400 401 401 402 403 404 405 405 408

Zygopterid Evolution Marattiales Psaroniaceae: Vegetative Features Psaronius Plant Other Stem Taxa Psaroniaceae: Reproductive Features Paleozoic Compression Taxa Mesozoic Marattialeans Marattialean Evolution Ophioglossales Leptosporangiate Ferns Osmundales Paleozoic Stem Taxa Guaireaceae Mesozoic and Cenozoic Stem Taxa Sterile and Fertile Foliage Osmundalean Evolution Botryopteridaceae Vegetative Organs Reproductive Organs Other Genera Anachoropteridaceae Kaplanopteridaceae Psalixochlaenaceae Sermayaceae Tedeleaceae Skaaripteridaceae Tempskyaceae Schizaeaceae Hymenophyllaceae Gleicheniaceae Dicksoniaceae Cyatheaceae Matoniaceae Loxsomataceae Dipteridaceae Polypodiales Salviniales Marsileaceae Salviniaceae Conclusions CHAPTER



417 418 418 418 425 425 431 433 434 435 436 436 437 438 438 440 442 443 443 446 449 449 451 452 453 454 457 457 459 462 462 464 465 466 469 469 470 472 472 473 476


Archaeopteridales Archaeopteris Leaves Archaeopterid Reproduction Callixylon Stems Other Archaeopterids

480 481 483 484 487


Aneurophytales Aneurophyton Tetraxylopteris Triloboxylon Rellimia Other Aneurophytes Protopityales Noeggerathians Progymnosperm Evolution CHAPTER

489 489 489 491 492 494 496 497 501


Origin and Evolution of the Seed Habit 503 Homospory, Heterospory, and the Seed Habit Homospory Heterospory Sporangia Endospory Lycopsid Heterospory Seed Habit Evolution of the Integument Evolution of Pollen Capture Pollen Cupules Cupulate Devonian Seeds Reproductive Biology Carboniferous Seeds Pollen Chamber Function Microgametophytes Diversity of Early Seeds Paleozoic Seeds with Embryos CHAPTER


Paleozoic Seed Ferns

503 503 504 504 507 508 508 509 510 511 511 511 517 518 523 524 525 526


Calamopityales Buteoxylonales Lyginopteridales Lyginopteris Plant Vegetative Organs Reproductive Structures Other Lyginopterids: Vegetative Remains Heterangium Microspermopteris Schopfiastrum Pitys Devonian–Mississippian Taxa Problematic Lyginopterids Other Lyginopterids: Seeds and Cupules Sphaerostoma Salpingostoma

531 539 540 540 540 542 546 547 550 550 551 552 554 555 556 556

Conostoma Coronostoma Physostoma Tyliosperma Calathospermum Gnetopsis Megatheca Other Lyginopterids: Pollen Organs Incertae Sedis Lyginopterid Evolution Medullosales Stems Medullosa Other Stem Taxa Leaves (Fronds) Roots Growth Habit Seeds Pollen organs Pollen Medullosan Evolution Callistophytales Vegetative Organs Reproductive Structures Callistophytalean Evolution Glossopteridales Leaves Glossopteris Gangamopteris Other Leaf Types Stems and Roots Ovulate Reproductive Structures Permineralized Forms Impression–Compression Specimens What is the Glossopterid Ovulate Structure? Pollen Organs Glossopteris Habit and Habitat Phylogenetic Position CHAPTER


Mesozoic Seed Ferns Caytoniales Sagenopteris Caytonanthus Caytonia Ruflorinia and Ktalenia Corystospermales Foliage Stems

ix 556 557 557 558 558 559 559 560 563 565 566 566 566 569 570 572 572 573 581 590 591 593 594 595 598 598 599 599 603 603 605 606 606 609 614 616 618 618

621 622 622 623 624 626 627 627 630



Pollen Organs Ovulate Structures Petriellales Peltaspermales Foliage Reproductive Organs and Whole-Plant Concepts Conclusions CHAPTER

631 634 637 639 639 643 648


Late Paleozoic and Mesozoic Foliage 651 Late Paleozoic Foliage Adiantites Alethopteris Aneimites Aphlebia Alloiopteris Botrychiopsis Callipteridium Cardiopteridium Cardiopteris (Fryopsis) Charliea Cyclopteris Dicksoniites Discopteris Eremopteris Ginkgophytopsis Kankakeea Karinopteris, Mariopteris, and Pseudomariopteris Lesleya Linopteris, Reticulopteris, and Barthelopteris Lobatopteris Lonchopteridium and Lonchopteris Megalopteris Neuropteris sensu lato Laveinopteris Macroneuropteris Margaritopteris Neuralethopteris Neurocallipteris Neurodontopteris Neuropteris sensu stricto Paripteris Sphenoneuropteris Neuropterid Growth Habit Blanzyopteris Nothorhacopteris Odontopteris and Lescuropteris Pecopteris

652 655 656 657 658 658 659 659 660 660 660 661 662 664 664 664 665 665 669 669 671 672 672 673 674 674 674 674 675 675 675 675 676 676 676 677 677 679

Rhodea (Rhodeopteridium) Sphenopteris Sphenopteris sensu stricto Eusphenopteris Spiropteris Taeniopteris Tinsleya Triphyllopteris, Genselia, and Charbeckia Mesozoic Foliage Anomozamites Cladophlebis Coniopteris Ctenis Deltolepis and Cycadolepis Dictyophyllum Dictyozamites Doratophyllum Macrotaeniopteris Matonidium Mesodescolea Nilssonia Nilssoniopteris Otozamites Pachypteris, Komlopteris, and Thinnfeldia Phlebopteris Pseudoctenis Pseudocycas Pterophyllum Ptilophyllum Ptilozamites Ruflorinia Taeniozamites Ticoa Wingatea Yabeiella Zamites CHAPTER



680 680 682 682 683 683 685 685 685 687 687 688 689 689 689 689 690 690 690 690 690 691 693 695 696 696 697 697 698 699 699 700 700 700 700 701


Cycadales Leaves and Petioles Stems Paleozoic Reproductive Structures Triassic Cycads Jurassic Cycads Pollination Biology Discussion: Cycad Evolution Bennettitales Cycadeoidaceae

703 706 707 709 715 718 721 721 722 725


Stem Anatomy Reproductive Structures Development Williamsoniaceae Ovulate Structures Pollen Organs Discussion: Bennettitales CHAPTER

744 747 747 750 750 750 752 752 752 753 754 755


Gymnosperms with Obscure Affinities 757 Gigantopteridales Vegetative Remains Reproductive Organs Vojnovskyales Czekanowskiales Iraniales Pentoxylales Hermanophytales Gnetales Extant Genera Ephedra Gnetum Welwitschia Extant Reproductive Structures Fossil Gnetophyte Pollen Gnetophyte Megafossils Putative Gnetophytes Dirhopalostachyaceae



758 758 762 763 765 768 768 773 775 776 776 776 776 777 777 778 781 785


Vegetative Features Stems Foliage




Paleozoic Record Ginkgophyte Wood Ginkgophyte Foliage Pollen-Producing Structures Ginkgophyte Plants Ginkgoaceae Karkeniaceae Umaltolepidiaceae Yimaiaceae Schmeissneriaceae Taxa with Uncertain Affinities Conclusions


Roots Reproductive Features Reproductive Organs Seeds Angaran Cordaites Phylogenetic Position and Origin of the Cordaites CHAPTER




725 728 730 732 734 738 739

788 788 791

xi 794 795 795 798 801 803


Early Conifers Voltziales Utrechtiaceae Utrechtia Ernestiodendron Ortiseia Otovicia Moyliostrobus Other Taxa Thucydiaceae Emporiaceae Majonicaceae Ullmanniaceae Bartheliaceae Other Voltzialeans Ferugliocladaceae Buriadiaceae Pollen Cones Summary Discussion: Voltzialeans Coniferales Palissyaceae Cheirolepidiaceae Summary Discussion: Cheirolepidiaceae Podocarpaceae Summary Discussion: Podocarpaceae Araucariaceae Summary Discussion: Araucariaceae Cupressaceae Cunninghamioideae Taiwanioideae Athrotaxoideae Sequoioideae Taxodioideae Cupressoideae Cupressaceous Wood Summary Discussion: Cupressaceae Sciadopityaceae Pararaucariaceae Pinaceae Pinoideae

806 807 808 809 809 809 810 811 811 814 815 816 819 820 820 823 826 826 828 830 830 831 837 838 843 843 848 849 850 851 851 852 854 857 859 859 860 861 861 863



Genus Pinus Pinus Wood Larix Piceoideae Abietoideae Summary Discussion: Pinaceae Cephalotaxaceae Taxaceae Summary Discussion: Cephalotaxaceae and Taxaceae Conclusions CHAPTER


Flowering Plants

864 866 866 867 867 868 868 869 869 870


Angiosperm Origins Origin of the Flower Pseudanthial Theory Euanthial Theory Microsporangial Theories Transitional–Combination Theory Habit Ecological Considerations Site of Origin Pre-Cretaceous Fossil Evidence Sanmiguelia Furcula Problematospermum Pre-Cretaceous Pollen Dispersed Pollen Early Angiosperm Evidence Pollen Pollen Evolution Evidence from Leaves Angiosperm Ancestors Caytoniales Czekanowskiales Glossopteridales Bennettitales Pentoxylales Gigantopteridales Phylogenetic Analyses and Angiosperm Origins Selected Angiosperm Families Basal Angiosperms Amborellaceae Hydatellaceae Archaefructaceae Chloranthaceae Nymphaeales

876 877 877 878 878 878 879 879 880 880 881 882 883 883 884 885 885 889 889 893 894 895 895 895 895 895 895 897 898 898 898 898 899 901

Nymphaeaceae Austrobaileyales Austrobaileyaceae Illiciaceae Schisandraceae Ceratophyllales Ceratophyllaceae Magnoliids Canellales Winteraceae Laurales Calycanthaceae Lauraceae Magnoliales Annonaceae Magnoliaceae Myristicaceae Piperales Lactoridaceae Saururaceae Monocotyledons Alismatales Alismataceae Araceae Hydrocharitaceae Zosteraceae (Seagrasses) Asparagales Agapanthaceae Hemerocallidaceae Orchidaceae Dioscoreales Dioscoreaceae Liliales Petermanniaceae Pandanales Pandanaceae Triuridaceae Commelinids Arecales Arecaceae (Palmae) Commelinales Commelinaceae Poales Cyperaceae Poaceae (Gramineae) Zingiberales Musaceae Zingiberaceae Eudicots

901 902 902 902 903 904 904 904 904 904 906 906 906 908 908 909 914 915 915 915 917 917 917 917 917 920 921 921 921 921 922 922 922 922 923 923 923 923 923 923 925 925 925 925 926 928 928 929 929


Buxaceae Trochodendraceae Proteales Nelumbonaceae Proteaceae Platanaceae Ranunculales Berberidaceae Ranunculaceae Core Eudicots Gunnerales Gunneraceae Caryophyllales Phytolaccaceae Saxifragales Cercidiphyllaceae Haloragaceae Hamamelidaceae Iteaceae Saxifragaceae Rosids Vitaceae Myrtales Lythraceae Trapaceae Myrtaceae Onagraceae Eurosids I (Fabids) Fabales Fabaceae (Leguminosae) Fagales Betulaceae Casuarinaceae Fagaceae Juglandaceae Myricaceae Nothofagaceae Malpighiales Clusiaceae Euphorbiaceae Salicaceae Malpighiaceae Oxalidales Cunoniaceae Elaeocarpaceae Rosales Moraceae Rhamnaceae Rosaceae

930 931 933 933 935 937 940 940 940 941 941 941 941 941 942 942 943 945 945 946 946 947 948 948 948 948 950 950 950 950 953 953 955 956 961 966 966 967 967 968 970 970 971 971 971 971 971 971 971

Ulmaceae Eurosids II (Malvids) Brassicales Capparaceae Malvales Tiliaceae Sapindales Anacardiaceae Meliaceae Rutaceae Sapindaceae Asterids Cornales Cornaceae Curtisiaceae Hydrangeaceae Ericales Ebenaceae Ericaceae Theaceae Euasterids I (Lamiids) Icacinaceae Garryales Eucommiaceae Gentianales Gentianaceae Rubiaceae Lamiales Avicenniaceae Byblidaceae Lentibulariaceae Oleaceae Solanales Solanaceae Euasterids II (Campanulids) Bruniaceae Quintiniaceae Apiales Araliaceae Aquifoliales Aquifoliaceae Asterales Asteraceae (Compositae) Menyanthaceae Dipsacales Caprifoliaceae Cenozoic Floras Conclusions

xiii 973 976 976 976 976 976 977 977 978 978 979 981 981 981 984 984 985 985 985 985 986 986 987 987 987 987 987 988 988 988 988 988 988 988 988 988 988 989 989 989 989 990 990 991 991 991 991 996





Interactions Between Plants and Animals 999 Early Terrestrial Ecosystem Associations Animals on Land Early Plant–Animal Associations Herbivory Defenses Against Herbivory Mechanical Protection Chemical Defenses Fossil Evidence of Herbivory Coprolites Gut Contents Marginal Feeding Defoliation Leaf Miners Wound Tissue

1001 1001 1001 1003 1004 1005 1006 1007 1007 1011 1011 1013 1013 1015

Interactions with Vertebrates Herbivory Dentition Coprolites and Stomach Contents Dispersal Plants as Habitat Other Plant–Animal Interactions Mimicry Pollination Conclusions

1016 1016 1018 1018 1018 1019 1021 1021 1022 1024

Appendix 1: Classification of Organisms Glossary References Index

1027 1031 1049 1199

This page intentionally left blank

Preface One of the major challenges faced by paleobotanists is the extraordinary interdisciplinary nature of the science. Most paleobotanists are quick to point out, however, that they were practicing collaborative and interdisciplinary science long before the concepts became fashionable in the research environment of today. We view this book as an up-to-date introduction to the discipline for advanced undergraduate and graduate students, and as a book that is more encyclopedic in organization than other paleobotany textbooks and that can be used as a reference for a number of disciplines that today encompass the biological and geological sciences, whether professional or amateur. Although this book is not technically a second edition, it does include material from The Biology and Evolution of Fossil Plants by Thomas N. Taylor and Edith L. Taylor (1993), which has long been out of print. To make the book usable to a wider range of readers, we begin each chapter with a general introduction that provides the essential characteristics of a particular group, including not only information about fossil members but also, where applicable, living representatives. In addition to a comprehensive table of contents, we have added a table to each chapter that summarizes the higher taxa in the chapter and the geologic range of each group. Chapters are subdivided to make it easier for readers to find information. For the nonbiologist, we have included a discussion of plant structure, tissue systems, and plant organs (Chapter 7) that is supplemented by illustrations and diagrams. To further assist in making the book useful, we have

expanded the glossary from Taylor and Taylor (1993) to more than 900 entries. For easy reference, a chart showing the geologic periods is included inside the front and back covers. With more than 5000 references, this book provides an introduction to the primary literature. For further literature, please see the Bibliography of Paleobotany, http://paleobotany. We have also included more than 2100 illustrations, many in color, and a large number unpublished. We received numerous favorable comments regarding the portraits of distinguished paleobotanists and therefore have included many more in this book. We found the following online sources to be of enormous assistance in writing this book and would like to thank those who maintain these resources: (1) Index Nominum Genericorum,; (2) GBIF portal,; (3) Peter Hoen’s Glossary of Pollen and Spore Terminology 2nd edition, http://www., from the University of Utrecht (4) the International Commission on Stratigraphy site,; and (5) L. Watson and M. J. Dallwitz’s The Families of Flowering Plants, http://delta-intkey. com/angio/. We have followed the 2008 International Commission on Stratigraphy (ICS) conventions on naming geologic time periods (htpp:// and have provided the international name in addition to local stage names throughout.


This page intentionally left blank

Acknowledgments We are truly indebted to a large number of individuals and institutions who eagerly assisted us in the preparation of this book. This includes all of our colleagues both past and present, who readily contributed illustrations as well as the various professional journals, book publishers, and professional organizations that granted permission to use copyrighted material. We are also indebted to numerous colleagues who took the time to read chapters and sections of the book, and freely discussed their ideas so that we could produce a book that is as up-to-date as possible in a field covering as much information as does paleobotany. These include:

J. D. Aitken ● V. M. Albright ● A. C. Allwood ● K. L. Alvin ● H. M. Anderson ● J. M. Anderson ● H. N. Andrews, Jr. ● Anschutz Science Library staff, University of Kansas ● A. Archangelsky ● S. Archangelsky ● S. R. Ash ● K. R. Aulenback ● L. Axe ● B. J. Axsmith ● N. R. Banerjee ● H. P. Banks ● F. Baron ● D. Barr ● M. Barthel ● R. A. Baschnagel ● J. F. Basinger ● D. Bassi ● P. W. Basson ● L. H. Batenburg ● R. W. Baxter ● Bayerische Staatssammlung für Paläontologie und Geologie, Munich (BSPG) ● S. C. Beadle ● C. B. Beck ● J. Bek ● J. M. Benson ● M. L. Berbee ● S. Berger ● M. E. C. Bernardes-de-Oliveria ● C. M. Berry ● J. Bogner ● B. Bomfleur ● P. M. Bonamo ● M. N. Bose ● H. Blütmann ● S. D. Brack (Brack-Hanes) ● D. F. Brauer ● M. Brea ● W. Brenner ● D. W. Brett ● D. E. G. Briggs ● H. K. Brooks ● N. C. Brotzman ● J. T. Brown ● V. I. Burago ● J. Burgess ● N. D. Burgess ● N. J. Butterfield ● R. Butzmann ● L. Calvillo-Canadell ● D. J. Cantrill ● L. M. Carluccio ● A. V. Carozzi ● R. J. Carpenter ● S. N. Césari ● S. R. S. Cevallos-Ferriz ● W. G. Chaloner ● S. Chandra ● D. S. Chaney ● M. Chaphekar ● R. L. Chapman ● I. Chen ● S. Chitaley ● D. C. Christophel ● M. A. Cichan ● J. A. Clement-Westerhof ● J. A. Clendening ● G. T. Cole ● M. Collinson ● N. Combourieu ● M. E. Cook ● B. Cornet ● P. R. Crane ● W. L. Crepet ● A. A. Cridland ● W. N. Croft ● Y. M. Crosbie ● A. T. Cross ● N. R. Cúneo ● R. Daber ● C. P. Daghlian ● V. Daviero-Gomez ● E. Y. Dawson ● A.-L. Decombeix ● T. Delevoryas ● G. del Fueyo ● C. Delwiche ● T. Denk ● R. L. Dennis ● Denver Museum of Nature ● M. E. Dettmann ● M. DeVore ● C. Diéguez ● D. L. Dilcher ● R. M. Dillhoff ● W. A. DiMichele ● M. Dolezych ● S. Doerenkamp ● J. B. Doran ● H. Dörfelt ● N. Dotzler ● J. A. Doyle ● A. N. Drinnan ● X.-M. Du ● J. Dupéron ● M. Dupéron-Laudoueneix ● D. Edwards ● D. S. Edwards ● L. E. Edwards ● D. A. Eggert ● D. L. Eggert ● H. Eklund ● W. E. El-Saadawy ● G. Elliott ● M. S. Engel ● I. Ernstmeier ● J. Erl ● D. M. Erwin ● I. Escapa ● W. R. Evitt ● M. Fairon-Demaret ● M. Feist ● D. K. Ferguson ● P. F. Fields ● T. I. Fine ● T. Fischer ● F. Fleming ● G. L. Floyd ● V. K. Folkman ● W. H. Forbes ● C. B. Foster ● J.-P. Frahm ● F. Franzmeyer ● W. E. Friedman ● E. M. Friis ● W. L. Fry ● G. Fuchs ● J. Galtier ● M. A. Gandolfo ● Z. Gao ● R. A. Gastaldo ● C. T. Gee ● P. G. Gensel ● Geological Society of America ● Geologische Bundesanstalt, Vienna (GBA) ● E. A. George ● P. Gerrienne ● R. W. Gess ● D. E. Giannasi ● W. H. Gillespie ● C. W. Good ● I. Glasspool ● K. D. Gordon-Gray ● R. Gossmann ● K. Goth ● R. E. Gould ● F. Gradstein ● S. R. Gradstein ● A. Graham ● L. E. Graham ● N. GrambastFessard ● L. Grauvogel-Stamm ● J. Gray ● J. D. Grierson ● R. Grolle ● M. A. Haban ● D. W. Haines ● J. W. Hall ● Hancock Museum ● S.-G. Hao ● T. M. Harris ● C. M. Hartman ● H. Hass ● A. R. Hemsley ● P. S. Herendeen ● E. J. Hermsen ● G. R. Hernández-Castillo ● G. Heumann ● F. A. Hibbert ● D. S. Hibbert ● L. J. Hickey ● A. Hill ● R. S. Hill ● S. A. Hill ● N. Hiller ● L. Hillis ● J. M. Hilton ● Hirmer Verlag GmbH, Munich ● H. J. Hofmann ● J. C. Holmes ● W. B. K. Holmes ● H. J. Hoops ● R. C. Hope ● C. A. Hopping ● D. G. Horton ● C. L. Hotton ● J. Hsü ● F. M. Hueber ● A. Iglesias ● I. A. Ignatiev ● W. I. Illman ● Interlibrary Loan staff, University of Kansas ● International Commission on Stratigraphy ● L. C. Ivany ● B. F. Jacobs ● H. Jähnichen ● G. Janssen ● J. A. Janssens ● H.-B. Jansson ● S. Jardiné ● E. A. Jarzembowski ● D. M. Jarzen ● J. R. Jennings ● A. J. Jeram ● K. R. Johnson ● M. E. Johnson ● D. S. Jones ● J. H. Jones ● W. W. Jung ● M. J. Kaever ● E. Karasev ● E. E. Karrfalt ● A. E. Kasper ●




K.-P. Kelber ● P. Kenrick ● E. M. Kern ● H. Kerp ● P. F. Kidwai ● B.-K. Kim ● T. Kimura ● M. Kirchner ● B. L. Kirkland ● S. Kiyokawa ● S. D. Klavins ● M. J. Knaus ● A. H. Knoll ● A. S. Konopka ● E. B. Koppelhus ● W. L. Kovach ● V. A. Krassilov ● J. Kukalová-Peck ● J. Kvacˇek ● Z. Kvacˇek ● C. C. Labandeira ● Laboratory of Palaeobotany and Palynology, Utrecht University (LPPU) ● W. S. Lacey ● W. H. Lang ● C. A. LaPasha ● I. S. Latimer, Jr. ● J.-P. Laveine ● R. L. Leary ● S. Leclercq ● G. A. Leisman ● K. U. Leistikow ● B. A. LePage ● U. Leppig ● C.-S. Li ● H.-L. Li ● J. H. Lipps ● R. J. Litwin ● Ludwig-Maximilians-Universität München ● S. T. LoDuca ● A. G. Long Collection ● C. V. Looy ● T. A. Lott ● D. R. Lowe ● B. Lugardon ● B. Lundblad ● S. D. Lys ● H. K. Maheshwari ● S. H. Mamay ● S. R. Manchester ● S. B. Manum ● G. Mapes ● F. Marsh ● D. M. Martill ● L. C. Matten ● J. D. Mauseth ● Max Kade Center for German-American Studies, University of Kansas ● H. Mayr ● A. M. McClain ● E. E. McIver ● S. McLoughlin ● C. A. McRoberts ● J. Mehl ● K. Meister ● S. V. Meyen ● H. W. Meyer ● B. Meyer-Berthaud ● J. E. Mickle ● M. A. Millay ● C. E. Miller ● C. N. Miller, Jr. ● J. M. Miller ● B. A. R. Mohr ● M. Montenari ● E. D. Morey ● J. Morgan ● J. E. Morris ● Museum für Naturkunde, Berlin (MNB) ● Národní Muzeum (National Museum), Prague ● N. S. Nagalingum ● K. K. Namboodiri ● E. M. V. Nambudiri ● Naturhistorisches Museum Schloss Bertholdsburg Schleusingen (NMS) ● Naturhistorisches Museum, Vienna (NHM) ● Naturhistoriska Riksmuseet, Stockholm (NRM) ● S. V. Naugolnykh ● D. D. Nautiyal ● M. E. Nelson ● K. J. Niklas ● E. Nisbet ● H. Nishida ● M. Nishida ● M. H. Nitecki ● K. C. Nixon ● H. Nøhr-Hansen ● R. Noll ● M. Nose ● L. L. Oestry ● T. Ohana ● Oklahoma Geological Survey ● G. Oleschinski ● J. M. Osborn ● A. D. Pan ● D. D. Pant ● B. C. Parker ● L. R. Parker ● K. Parris ● M. Parrish ● C. S. Pearsall ● K. R. Pedersen ● K. Perch-Nielsen ● C. P. Person ● J. Peters ● B. Petriella ● H. W. Pfefferkorn ● H. Pflug ● R. N. Pheifer ● T. L. Phillips ● C. J. Phipps ● K. B. Pigg ● G. Playford ● D. T. Pocknall ● S. A. J. Pocock ● G. O. Poinar ● D. Pons ● R. J. Poort ● R. W. Portell ● C. Pott ● F. W. Potter ● N. W. Radforth ● C. G. K. Ramanujam ● P. H. Raven ● C. B. Read ● J. D. Reed ● M. A. Reihman ● D. Remy ● R. Remy ● W. Remy ● G. J. Retallack ● R. Riding ● E. Rieber ● J. F. Rigby ● M. O. Rischbieter ● C. R. Robison ● D. M. Rohr ● E. J. Romero ● R. Rössler ● G. W. Rothwell ● N. P. Rowe ● A. C. Rozefelds ● P. E. Ryberg ● C. Rydin ● B. Sahni ● F. Schaarschmidt ● J. T. Schabilion ● S. E. Scheckler ● R. Schmid ● A. Schmidt ● S. Schneider ● J. W. Schopf ● J. M. Schramke ● R. E. Schultes ● S. Schultka ● R. M. Schuster ● H.-J. Schweitzer ● U. Schweitzer ● A. C. Scott ● R. A. Scott ● S. Sekido ● P. A. Selden ● A. Selmeier ● B. S. Serlin ● T. Servais ● R. L. Seymour ● G. L. Shadle ● B. D. Sharma ● W. A. Shear ● M. A. Sherwood-Pike ● C. H. Shute ● M. A. Siders ● J. Silander ● A. D. Simper ● Z. Šim˚unek ● C. A. Sincock ● R. S. Singh ● J. E. Skog ● J. J. Skvarla ● C. J. Smiley ● M. L. So ● I. Sobbe ● S. K. Srivastava ● G. D. Stanley ● P. Steemans ● W. E. Stein, Jr. ● H. Steur ● W. N. Stewart ● B. M. Stidd ● R. A. Stockey ● W. C. Stowe ● M. Streel ● C. A. E. Strömberg ● P. Strother ● S. P. Stubblefield ● T. F. Stuessy ● O. P. Suthar ● K. Sugitani ● K. R. Surange ● N. P. Swanson ● H. A. J. M. Swinkels ● R. E. Taggart ● T. Tanai ● W. R. Tanner ● D. W. Taylor ● G. H. Taylor ● J. W. Taylor ● W. A. Taylor ● G. F. Thayn ● B. A. Thomas ● J. R. Thomasson ● L. Tidwell ● W. D. Tidwell ● A. M. Torres ● J. M. Trappe ● A. Traverse ● N. Trewin ● M. L. Trivett ● G. R. Upchurch ● H. W. J. van Amerom ● M. van Campo ● R. W. J. M. Van der Ham ● D. E. Van Dijk ● J. H. A. Van Konijnenburg-Van Cittert ● G. Vasanthy ● M. Vecoli ● J. C. Vega ● R. Verwer ● Alexander von Humboldt-Stiftung ● L. Voronova ● C. Vozenin-Serra ● C. A. Wagner ● R. H. Wagner ● D.-M. Wang ● H.-S. Wang ● L. Wang ● X. L. Wang ● Y. Wang ● Z.-Q. Wang ● J. V. Ward ● S. Warner ● J. Watson ● J. A. Webb ● C. H. Wellman ● R. Werneburg ● W. Werner ● F. Westall ● E. A. Wheeler ● D. C. White ● J. F. White, Jr. ● M. E. White ● R. Wicander ● D. C. Wight ● P. Wilf ● L. R. Wilson ● R. B. Winston ● J. A. Wolfe ● V. P. Wright ● S. Xiao ● Z. L. Xu ● A. Yabe ● X. Yao ● Z. Yao ● Z.-Q. Yao ● J. R. Young ● Y. D. Zakharov ● R. J. Zakrzewski ● S.-Q. Zan ● M. S. Zavada ● Z. Zhang ● Z. Zhou ● J.-N. Zhu, if we has missed someone we are deeply apologetic. We are especially indebted to Rudolph Serbet and Andrew Schwendemann for their invaluable help and assistance in photographing specimens, adding bar scales to images, and in general overseeing the myriad tasks associated with preparing the illustrations in a volume of this size. Without their attention to detail this volume could not have been completed. Christian Pott and Hans Kerp skillfully took and graciously provided, a number of digital images from paleobotanical collections in Europe, some of which have not been published earlier. Jeannie Houts deserves special mention for her assistance in all phases of the preparation of this book. Her ability to deal simultaneously with the three of us was a challenge that she met with diligence, skill, and efficiency, and which now requires a new definition for the word “patience.”



We are indebted to Frank Baron and the Max Kade Center for German–American Studies at the University of Kansas for making accommodations available when MK visited Lawrence to work on the book, and to Judith Erl, Munich, who displayed special skill and patience in ferreting out literature for the bibliography. Thanks also to J. William Schopf for his assistance with many questions on Precambrian life. We are appreciative of the extraordinary help we received from Pat Gonzalez, Developmental Editor, and Andy Richford, Senior Acquisition Editor, both of Academic Press, for their unusual patience and assistance in bringing this volume to fruition; and to Mani Prabakaran and his staff at Macmillan Publishing Solutions for their attention to detail throughout the editing process. To the many individuals who participated in this enterprise that we never directly interacted with – we sincerely thank you for your professional expertise and assistance along the way. We could not have completed the book without your help. We gratefully acknowledge the National Science Foundation and Alexander von Humboldt-Stiftung for financial assistance that has supported our research programs. This support has provided assistance in many ways that has allowed us to continue our research and scholarly activities and interests, and to actively participate in the discoveries that make the discipline of paleobotany so exciting. To our families, friends, colleagues, and institutions where we are employed—we can never truly express how very much we appreciate your support and patience during the preparation of this book. You all have made sacrifices of untold proportion and exhibited extraordinary patience as we have worked on. This book would not have been completed without your understanding and friendship. There are many individuals who have influenced our careers or have made it possible for us to participate in this writing exercise. They have been there to teach and critique, to support and challenge, and to question and inspire—and their support and enthusiasm has never wavered. George W. Burns, Bill Crepet, Fred Daniëls, Chuck Daghlian, Ted Delevoryas, David Dilcher, Don Eggert, Ray Evert, Hans Kerp, Serge Mamay, Mike Millay, Elisabeth Peveling, Richard Popham, Winfried Remy, Gar Rothwell, Ruth Stockey, and Bill Stewart deserve special mention. Finally, we thank T. Cartel for friendship and support during our careers. As every chapter begins with a quotation, we will end with this famous one from Isaac Newton: “If [we] have seen farther it is by standing on the shoulders of giants.”

About the Authors Thomas N. Taylor is a distinguished Professor in the Department of Ecology and Evolutionary Biology, and Curator of Paleobotany in the Natural History Museum and Biodiversity Research Center at the University of Kansas. He also holds a courtesy appointment in the Department of Geology. He received his Ph.D. in botany from the University of Illinois, and was a National Science Foundation Postdoctoral Fellow at Yale University. He is a member of the National Academy of Sciences. His research interests include Permian and Triassic biotas of Antarctica, early land plant–fungal interactions, the origin and evolution of reproductive systems in early land plants, symbiotic systems through time, and the biology and evolution of fossil microbes.

Edith L. Taylor has been a Professor of Ecology and Evolutionary Biology and Senior Curator of Paleobotany in the Natural History Museum and Biodiversity Research Center at the University of Kansas since 1995, and also serves as a Courtesy Professor of Geology. She received her Ph.D. in paleobotany from the Ohio State University, where she was an American Association of University Women Dissertation Fellow. She is the author of seven books or edited volumes, and more than 140 publications. She was elected a Fellow of the American Association for the Advancement of Science in 1992. Her research interests include fossil wood growth and paleoclimate, Permian and Triassic permineralized plants from Antarctica, distribution and diversity of Permian–Triassic Antarctic floras, and the structure and evolution of fossil phloem.

Michael Krings is Curator for Fossil Plants in the Bavarian State Collection for Palaeontology and Geology (BSPG) at Munich, Germany, and Professor of Plant Paleobiology at the Ludwig-Maximilians-Universität, Munich. He also holds an affiliate faculty position in the Department of Ecology and Evolutionary Biology at the University of Kansas. He received his Ph.D. in botany from the University of Münster, Germany, and was an Alexander von Humboldt Foundation Postdoctoral Fellow at the University of Kansas. His research interests include Carboniferous, Permian, and Triassic seed plants from Europe and North America, and the biology and ecology of microorganisms in late Paleozoic terrestrial ecosystems.


This page intentionally left blank

1 INTRODUCTION TO PALEOBOTANY, HOW FOSSIL PLANTS ARE FORMED WHAT IS PALEOBOTANY? .............................................................1

Cellular Preservation ...........................................................................23

THE OBJECTIVES OF PALEOBOTANY ..................................... 2

Unaltered Plant Material .....................................................................30 Summary Discussion ..........................................................................34

Reconstructing the Plants......................................................................2 Evolution of Plant Groups.....................................................................3

PALYNOLOGY ..................................................................................34

Form and Function in Fossil Plants .......................................................4

Geochronology and Biostratigraphy ...................................................36

Biostratigraphy and Correlation............................................................4

Paleoecology .......................................................................................37

Paleoecology: Plants in Their Environment ..........................................5

ABSOLUTE DATING ......................................................................38

Determining Paleoclimate from Fossil Plants .......................................6 GEOLOGIC TIMESCALE ..............................................................39

Summary ...............................................................................................7

BIOLOGICAL CORRELATION .................................................. 40

PRESERVATION: HOW PLANT FOSSILS ARE FORMED AND PRESERVED ............................................... 8

SYSTEMATICS AND CLASSIFICATION ................................ 40

Depositional Environments of Fossil Plants .........................................8

Nomenclature of Fossil Plants ............................................................41

Compressions ......................................................................................10

Classification of Organisms ................................................................42

Coal and Charcoal ...............................................................................18 BACKGROUND READING.........................................................42

Impressions .........................................................................................21 Molds and Casts ..................................................................................22

The Earth is a vast cemetery where the rocks are tombstones on which the buried dead have written their own epitaphs. Louis Agassiz … intoxicated joy and amazement at the beauty and grandeur of this world, of which man can just form a faint notion. Albert Einstein


no flowering plants, and when the continental land masses were in different positions than they are today. Who has not been captivated by the various forms of life that are recorded in the rocks and the enormous reconstructions of dinosaurs exhibited in various museums? It is natural to wonder about such examples of prehistoric life—how these organisms

Humans are by nature curious, and we are all interested in the Earth on which we live and how various aspects have changed through geologic time. We speculate about what the Earth looked like when there were no trees, when there were



paleobotany: the biology and evolution of fossil plants

lived, what their patterns of behavior were, and even why they became extinct. Although the paleontologist is interested in the geologic history of animals, the paleobotanist is concerned with the plants that inhabited the Earth throughout geologic time (Ward, 1885) (FIG. 1.1). In a general sense, the paleobotanist is a plant historian who attempts to piece together the intricate and complicated picture of the history of the plant kingdom. Although molecular and genetic analyses of living plants have become increasingly important as tools in reconstructing the phylogeny and evolutionary history of plants, the discipline of paleobotany, in all its various forms, remains the only method by which this history can be documented and visualized. Two books that discuss paleobotany from a historical perspective and that capture the excitement of the discipline are The Fossil Hunters—In Search of Ancient Plants (Andrews, 1980) and History of Palaeobotany—Selected Essays (Bowden et al., 2005). These volumes discuss the origins of the field and the scientists who have made the science so exciting and fascinating. Fossil plants and floras from one period of geologic time are different in size and shape, level of complexity, and abundance from those of other time periods. The most logical explanation for these differences is that the types of plants changed, or evolved, through geologic time. Unless one believes that there were an almost infinite number of “special creations,” we must assume that new plant forms were derived from preexisting ones by the processes of evolution. By studying the record of fossil plants, it is possible to assess the time at which various

major groups originated, the time each reached its maximum diversity, and, in the case of certain groups, when they became extinct.

THE OBJECTIVES OF PALEOBOTANY One of the aspects of paleobotany, which makes it unusual and interesting, is that it is inherently interdisciplinary and can be approached from either a biological or a more geological perspective—or both together. Each perspective presents a variety of questions that are unique to that discipline. Today more than ever before, the questions being asked by paleobotanists necessitate that both the botanical and geological perspectives be fully understood. To a large degree, the research questions that paleobotanists ask are influenced by whether their training emphasized a biological or a more geological perspective. RECONSTRUCTING THE PLANTS

Paleobotanists who have been trained primarily as biologists are interested in research directions which include all aspects of the organisms themselves. Because the majority of fossil plants are generally preserved in rocks as disarticulated plant parts (FIG. 1.2), that is leaves (FIG. 1.3), stems, pollen, or reproductive structures, a major aim of paleobotany is to reconstruct the whole plant, that is to say, to put the pieces of the puzzle back together. Once this is accomplished,

Figure 1.2 Impression of angiosperm leaf from the Dakota Figure 1.1

Lester Ward. (Courtesy H.N. Andrews.)

Formation (Cretaceous). Bar  2 cm.

chapter 1 introduction to paleobotany, how fossil plants are formed

the research can turn to other areas, such as determining the group of living plants, if any, to which the fossil is most closely related. Some paleobotanists are interested in aspects of plant life history that can be determined from fossils. For example, how did these plants reproduce, and how and what types of propagules were disseminated? Are their reproductive strategies similar to those of closely related living plants, or have there been major modifications in the reproductive systems of certain types of plants through geologic time? If so, how did this happen and when? What can we determine about the environment in which the plants lived millions of years ago, based on features of the fossil plants? For example, fossil wood collected from the Permian and Triassic of Antarctica (FIG. 1.4) indicates that the climate was quite


favorable for tree growth, based on the analysis of tree rings (FIGS. 1.5, 1.6). General circulation models of Permian paleoclimate, however, have proposed that these high paleolatitudes were very cold and not favorable for plant growth. Some paleobotanists are interested in what strategies these plants, and the animals that lived among them, developed to survive in the extreme seasonality at polar latitudes. EVOLUTION OF PLANT GROUPS

Paleobotanists are also interested in the origin and subsequent evolution of major groups of plants and their interrelationships. When did plants first inhabit the Earth and what did they look like? When did the first representatives of different groups of plants first arise? Other researchers want

Figure 1.4 Permineralized wood extending from paleostream

channel in the Triassic of Antarctica.

Compressed pinna showing detail of pinnule venation (Cretaceous). Bar  2 cm.

Figure 1.3

Figure 1.5 Section of Antarctic wood (Triassic) showing sev-

eral growth rings. Bar  1.4 mm.


paleobotany: the biology and evolution of fossil plants

Figure 1.6 Frost ring in Triassic permineralized wood from

Antarctica. Bar  3 mm.

to know why certain types of plants developed the capacity to produce secondary tissues (such as wood), whereas others have remained small throughout their geologic history. A number of paleobotanists study not only the plants themselves, but also the interactions of the plants with other organisms in the environment, especially the symbiotic interrelationships between plants and other organisms. For example, today almost all terrestrial plants possess mutualistic associations with fungi that inhabit their roots (mycorrhizae). Paleobotanists want to know when and why these associations became established, why they are absent in some groups today, and why only certain types of fungi form these associations. Can the fossil record of plants tell us anything about various biological activities that existed millions of years ago, such as the interactions between plants and certain groups of animals that use plants for food or protection? What types of evidence can be used, and what information does this provide about the interdependence of organisms through time? Can we determine from the fossil record if plants possessed certain features that served to attract pollinators, or produced edible seeds, or whether some plants produced certain chemicals that deterred herbivory? The answer to all these questions is a resounding YES! There is a multitude of information that can be gleaned from careful examination of the plant fossil record, and the types of information that we can obtain are constantly increasing as more and more research is done on fossil plants. FORM AND FUNCTION IN FOSSIL PLANTS

From many plant fossils, it is possible to understand the relationship between form and function in ancient plants, that is, what advantages or limitations are imposed on the growth and development of a plant based on certain biomechanical

properties? For example, are all arborescent (treelike) plants constructed of cells and tissue systems of the same type? If not, in what other ways can plants grow to tower over their neighbors? Studies of this type examine the anatomical and morphological properties of various fossil plants, often using computer simulations to model growth, in an attempt to better understand broad evolutionary patterns of plant growth, as well as changes in growth form through time (Niklas, 1992; Rowe and Speck, 1998; Niklas and Spatz, 2004; Niklas et al., 2006). Biomechanical studies have been especially useful in delimiting adaptations necessary for plants to move onto the land, including upright growth, size limitations, and the nature of the conducting strand (Niklas, 1986), and, once plants became established in terrestrial environments, the influences of gravity and wind on their reproduction (Niklas, 1998), and even aerodynamic features of pollen (Schwendemann et al., 2007). Factors such as plant size and form can also be examined over a broad spectrum of plant morphologies and thus offer insights as to why certain plants and plant groups have developed particular anatomical and morphological characteristics. Examining tree growth and other factors in extant plants has demonstrated that there are a variety of variables in play. Because fossils demonstrate a number of different growth forms that are not seen in modern plants, they offer a unique resource of data to allow paleobotanists to explore a host of intriguing biological questions. Fossil plants can also be used to infer developmental processes (Sanders et al., 2007). For example, Boyce and Knoll (2002) analyze the morphospace of numerous Paleozoic leaves representing various clades and show that leaf evolution follows the identical sequence of morphologies in all groups. Such an approach provides the ability to test hypotheses using living leaf development as a proxy for the leaves seen in various groups in the fossil record. BIOSTRATIGRAPHY AND CORRELATION

Paleobotany has also played a key role in many areas of geology, especially in biostratigraphy—placing rock units in stratigraphic order based on the fossils within them. Pollen grains and spores (one aspect of palynology) have been extensively used as index fossils in biostratigraphy and in the correlation of rock units, as have various forms of algal cells and cysts. In some instances, megafossils, such as leaves and seeds, have also provided a method of correlating rock units which are widely separated geographically. Pollen and spores, as well as megafossils, are especially useful in correlating terrestrial rocks, as these are generally deposited in limited areas (former lakes, ponds, river systems, etc.), making correlation by lithology (i.e., rock characteristics) very difficult.

chapter 1 introduction to paleobotany, how fossil plants are formed

1. Lepidophloios 2. Diaphorodendron 3. Synchysidendron 4. Paralycopodites 5. Sigillaria 6. Pteridosperms 7. Tree ferns 8. Sphenopsids








7 6



Figure 1.7 Transect through a Westphalian D mire showing habitat partitioning among major genera (A) and distribution of other plant

groups (B). (From DiMichele and Phillips, 1994.)


Paleoecology, the study of past environments, is a rapidly changing field that involves the integration and synthesis of both botanical and geological information. In recent years there has been a concerted effort by many paleobotanists to understand the paleoenvironment of fossil land plants more completely. For example, Bateman and Scott (1990) examined the famous late Tournaisian (lowermost Carboniferous) plant-bearing deposits at Oxroad Bay, Scotland, from a number of different perspectives, including an analysis of the geologic history and sedimentology of the site, as well as the paleoenvironment and paleoecology of the plants. Their studies indicate that the Oxroad Bay flora is found at eight levels in five successive facies, and these facies show the increasing influence of nearby volcanoes in the ocean-marginal setting. Details of the depositional environments through time at this site make it possible to understand plant adaptations to a rapidly changing, lowland environment, and to better understand both the biological and evolutionary importance of the floras.

Much paleoecological work initially focused on analyses of the swamp vegetation that contributed to extensive coal deposits in the Midcontinent of North America during the Carboniferous (Phillips and Peppers, 1984). The data used in these early analyses came primarily from the study of Pennsylvanian plants in coal balls—nodular concretions of limestone that contain permineralized peat (FIG. 1.5; see section on “Preservation”), coupled with a precise stratigraphic framework for the coal ball deposits based on palynology and careful field observations and measurements. Through the pioneering efforts of Phillips, DiMichele, and coworkers, we now possess an excellent understanding of many aspects of the paleoecology of coal-swamp vegetation during the Carboniferous (Wagner and Diez, 2007). Analysis of the plants preserved at different levels in these deposits not only documents the partitioning of the habitat among the different plant groups along ecological lines (Fig. 1.7), but also records changes in the depositional environment through time (DiMichele and Phillips, 1994; Wagner and Mayoral, 2007). In one study on the peat flora just above the Mahoning coal


paleobotany: the biology and evolution of fossil plants

in Ohio (Conemaugh Group, Pennsylvanian), DiMichele et al. (1996), utilizing macrofossils, palynomorphs, and coal petrology, concluded that the lepidodendrid trees (lycopsids—see Chapter 9) in this succession were flooded once, recovered, and then finally drowned by another flood event. This type of information can be utilized to recognize regional (Phillips et al., 1985) and global responses of plant communities to climate change (DiMichele et al., 2001a; Wagner and Mayoral, 2007). Understanding the interplay between these swamp-inhabiting plants and a variety of environmental parameters has now made it possible to interpret large-scale ecological shifts (e.g., the role of sea level fluctuations) in the community structure of the swamps, and to examine evolutionary questions within these habitats (DiMichele et al., 1985; DiMichele et al., 1996). These studies in turn have stimulated interdisciplinary research focused on broader questions, for example the evolution of major terrestrial ecosystems through time (Behrensmeyer et al., 1992; DiMichele and Hook, 1992). Paleoecological studies are very important in revealing the diversity of fossil communities inhabiting a geographic area (horizontal variation in floras) at the same time. Wing et al. (1993) examined fossil floras preserved in an ash fall and in fluvial deposits from the Upper Cretaceous (midMaastrichtian) of Wyoming, USA. They found that ferns (49%), along with palmettolike palms (25%), dominated the landscape, and that other angiosperms (Chapter 22), mostly herbaceous dicots, were dominant only in disturbed areas close to streams (fluvial deposits). The flowering plants were more diverse—constituting 61% of the species present—but they represented only 12% of the vegetational cover in this area. This study, and other similar ones (e.g., McElwain et al., 2007), have detailed the difference between the diversity of plants in fossil floras and the dominance of particular taxa within the paleoecosystem. Of course, it is only possible to fully comprehend the fossil assemblages, or taphocoenoses, by comparison with extant plants in various depositional environments (Spicer, 1981; Burnham, 1989, 1997) and by being aware of the taphonomy of fossil plants (Spicer, 1989) (see section on “Preservation”). Understanding and interpreting the sedimentological nature of the fossil assemblage, whether based on megafossils or microfossils, is only one of several aspects required in determining the diversification of plants through geologic time (Wing and DiMichele, 1995; Lupia, 1999). Paleoecologists use many of the same statistical methods used in contemporary ecological studies, including a variety of multivariate methods (Spicer and Hill, 1979; McCune and Grace, 2002). These tools, and many others, now make

it possible to examine the evolutionary and ecological processes that governed the plant communities which we now document as the fossil record (Jackson and Erwin, 2006). For a more in-depth approach to the study and methodologies of plant paleoecology see DiMichele and Wing (1988), Gastaldo (1989), and Jones and Rowe (1999). DETERMINING PALEOCLIMATE FROM FOSSIL PLANTS

Understanding climates of the past has become more and more crucial to appreciating the changes occurring on our warming planet today, and paleobotany is very important in providing baseline data to reconstruct past climates and in calibrating paleoclimate models based on physical parameters (Steppuhn et al., 2007). This area is rapidly expanding, so we will only cover a few of the many ways in which plant fossils can be used to reconstruct paleoclimate: TREE RINGS Data from fossil tree rings (paleodendrology) (FIG. 1.3) represent an important source of paleoclimate information, in some instances with very fine resolution, for example, major atmospheric disturbances (Miller et al., 2006). Although initially used for Holocene climate information (especially dating of archeological sites), some of the techniques used to analyze recent and subfossil wood have been extended to older material (Jefferson, 1982; Creber and Chaloner, 1984a; Creber and Francis, 1999; Taylor and Ryberg, 2007). Based on the changes in radial cell diameter within the tree rings and the variation in ring width (FIG. 1.3), it is possible to extrapolate climate information, which is especially useful when coupled with information from megafossils, microfossils, and the sedimentological record of the site. This approach has been utilized successfully by Parrish and Spicer in their work on Late Cretaceous floras from the North Slope of Alaska (Parrish and Spicer, 1988; Spicer and Parrish, 1990). More recently, Taylor and Ryberg (2007) have examined tree rings in Permian and Triassic woods from Antarctica. Based on their analysis using a variety of techniques, they suggest that the small amount of latewood indicates a very rapid transition to seasonal dormancy in response to decreasing light levels at these high polar latitudes. The mechanisms these plants evolved to cope with life in a polar light regime are of continuing interest in this and other studies based on plants that were once living at very high paleolatitudes. NEAREST LIVING RELATIVE The nearest living relative (NLR) method has been in use since the beginnings of paleobotany, particularly when

chapter 1 introduction to paleobotany, how fossil plants are formed

dealing with late Mesozoic or Cenozoic floras, as these are more likely to have close living relatives. It is based on the premise that climatic tolerances of the fossils are very similar to those of their NLR. The paleobotanist compares as many fossils as possible within a flora to their most closely related extant taxa; the more species in a fossil flora that have NLRs, the more precise the paleoclimate estimate, and the more closely related a fossil taxon is to an extant one, the more precise the method. It depends, therefore, partly on the paleobotanist’s ability to identify the fossils very accurately. The further back in time, the less effective this method is, as more and more extinct species or taxa which have no living relative appear. As a result, NLR has been used to best effect for Cenozoic angiosperm floras (Wolfe, 1995). This method can provide a general estimate of paleoclimate, but is limited by the fact that some fossil taxa do not have the same climatic limitations as their modern counterparts. LEAF PHYSIOGNOMY Leaf physiognomy analysis is a powerful technique that has been widely used in paleobotany to reconstruct Cenozoic paleoclimates. It is based only on angiosperms, however, so its applicability before the Cretaceous is uncertain (but see Glasspool et al., 2004a). Physiognomy is the general appearance of a plant, and it has long been known that plant physiognomy, especially leaf physiognomy, can be related to climate (Bailey and Sinnott, 1916). Physiognomy is primarily independent of taxonomy, for example plants with thick water-storing stems and leaves tend to grow in arid regions of the world, even though they may belong to a number of different families of plants. For fossil floras, this means that leaves do not have to be identified in order to obtain a paleoclimate signal. In his now-classic papers, Jack Wolfe (FIG. 22.276) presented the applications of leaf physiognomy to paleobotany, based on large collections of many leaves from extant floras, which he then was able to compare with Cenozoic angiosperm floras (Wolfe, 1993, 1995). Webb (1959) had previously completed a detailed physiognomic classification of Australian floras, and his definitions of leaf types are generally used in physiognomic methods today. There are presently two methods of leaf physiognomic analysis that are in general use: leaf-area and leaf-margin analysis. Leaf area directly correlates with mean annual precipitation (MAP). CLAMP (Climate-Leaf Analysis Multivariate Program; Wolfe, 1995) measures 31 leaf character states of woody dicots (Chapter 22) and uses multivariate analysis to map leaf shape in two-dimensional space (Wolfe and Spicer, 1999). CLAMP can provide a number of climatic parameters related to precipitation, humidity, and temperature.


Leaf-margin analysis (LMA) relies on the relationship between leaf margin (toothed versus entire) and climate (Greenwood and Wing, 1995; Wilf, 1997). Specifically, the proportion of leaves in the flora with toothed margins can be correlated with mean annual temperature (MAT), as toothed leaves are more abundant in wet environments. Both CLAMP and LMA can provide quantitative reconstructions of past climates, including estimates of MAT and MAP. More recently, paleobotanists have refined physiognomic methods by using computer image analysis to analyze both leaf-shape and leaf-margin morphology (Huff et al., 2003; Royer et al., 2005). Further data on the ecophysiology of modern plants and the function of various leaf shapes (Royer and Wilf, 2006) will no doubt help to refine these methods and improve their accuracy in the fossil record. Both methods are very robust, as both rely on large databases of leaf physiognomy of living leaves from many different sites and habitats. STOMATAL INDEX The stomatal index (the ratio of the number of stomata to the total number of epidermal cells plus stomata within a given leaf area expressed as percentages; see Salisbury, 1927) has been widely used in recent years to reconstruct past ρCO2 levels, as the stomatal index is inversely proportional to atmospheric CO2 levels. Woodward (1987) was one of the first to demonstrate the value of this relationship for ancient climate prediction, based on comparisons of modern leaves with herbarium specimens from preindustrial times. The best results have been obtained from comparisons of the same genus and species in order to control for genetic differences, so younger fossils, such as Holocene plants, have provided reproducible results (Wagner et al., 2004). The technique has been extended further back in time, for example the Cenozoic (Royer et al., 2001), as well as to the Mesozoic and Paleozoic, although there are limitations to the technique, especially with older fossils (Roth-Nebelsick, 2005; Uhl and Kerp, 2005). For studies in deep time, researchers have coupled CO2 estimates from stomatal indices with other proxy records, such as isotope data (Beerling, 2005 and papers cited therein). A summary of the pros and cons of methods to reconstruct past levels of atmospheric CO2 can be found in Royer (2001) and Kerp (2002). Details regarding the stomatal index technique can be found in Beerling (1999) and Poole and Kürschner (1999). SUMMARY

Throughout this book there are numerous examples of many of the biological and geological questions being asked by


paleobotany: the biology and evolution of fossil plants

paleobotanists today, and how the fossil record contributes to answering these questions. The field of paleobotany continues to advance, not only by the discovery of new fossils but also by the use of new methods applied to existing fossils and the application of techniques from other fields to paleobotany. Plant parts preserved in different ways or ones that show additional features are continually being discovered. More sophisticated and improved methods to study the fossils and interpret the results also provide new data which contribute to an enhanced understanding of the plants and communities that existed through geologic time.

PRESERVATION: HOW PLANT FOSSILS ARE FORMED AND PRESERVED A relatively small fraction of the plants and other organisms that live on the Earth at any particular time will ever become fossils. Most dead plant material is decayed by aerobic (oxygen-loving) fungi and bacteria. So, the first requirement for fossilization is that dead plants must be deposited in an environment where air is excluded, that is an anaerobic environment. This usually involves deposition in a body of water (discussed below), but not always. Once deposited, the plants must be buried by sediments so that air is excluded. In addition, these sediments must have enough acidity that anaerobic decay is also reduced. Paleobotanists are often asked the question, where do you look for fossil plants? The answer is that they typically are found in places where the rocks containing them have been exposed in some way (FIG. 1.8); these rocks may be as far away as the Arctic (FIG. 1.4) or the Antarctic. Because streams and rivers cut down through the rocks, exposed strata along waterways are often excellent sites to prospect for fossil plants. Erosion by water in many other places also exposes fossil-bearing rocks. Sometimes it is possible to find plants in eroded cliffs along seashores. In addition to the natural exposure of plant-bearing strata, excavations are frequently the source of many fossils. Road cuts, for example, often reveal fresh surfaces with unweathered rocks that contain wellpreserved fossils. As might be expected, quarries and mines are rich sources of fossil plants, revealing rocks that would otherwise have been inaccessible to paleobotanists. Coal balls (FIG. 1.43), a type of permineralization, are frequently encountered in coal mines, and often the shales immediately above the coal seams in such mines contain abundant fossil plants. Quarries in which clay is being excavated for bricks,

Figure 1.8 Students collecting fossil plants from a narrow lens of fine-grained shale.

tiles, or pottery are sites that often provide fossils. In fact, almost any massive construction site, such as for a dam, a hydroelectric plant, or a building with a deep foundation, can, and has, yielded an abundance of fossil plants. DEPOSITIONAL ENVIRONMENTS OF FOSSIL PLANTS

Fossil plants are found in almost all regions of the Earth, the most notable exceptions being recent volcanic islands or in rocks that have been extensively metamorphosed (FIG. 1.9). Marine plants, such as various forms of algae (Chapter 4), are generally found in rocks deposited in marine environments (e.g., nearshore deposits, carbonate shelves, etc.). Although land plants are occasionally found in marine rocks, generally, wherever terrestrial sedimentary rocks occur, there is a good chance that fossil plants will be found in them. Sedimentary rocks are those formed by the accumulation of rock particles derived from the weathering and mechanical abrasion of existing rocks. The great majority of sedimentary rocks are formed by deposition of particles in water, but wind deposits (e.g., eolian sands, loess) can also form, and rock breakdown can occur by chemical weathering, with rock components being released into solution, later to solidify at some other place. Plant parts are typically fossilized, then, in areas where sediment is accumulating. The delta of a river is just such a depositional environment. As the course of the river constantly shifts, channels are abandoned and new ones initiated; natural levees are destroyed during flooding, and new ones built up later. A meandering river cuts into the bank on the outside of each meander, and deposits sediment on the inside of meanders, often covering plants growing along the water. When the river breaks

chapter 1 introduction to paleobotany, how fossil plants are formed


Figure 1.9 Exposed rocks in Antarctica. Zone of dark red rocks are volcanic and lack fossil plants, including microfossils.

through a levee, a rush of water and sediment, called a splay deposit, can rapidly cover the adjacent floodplain, inundating the plants growing there. Associated with the deltaic system are abandoned stream channels, often referred to as oxbow lakes, and vegetation growing along the banks of these abandoned channels may be undisturbed for some time. If a subsequent flood destroys the levee, knocking down trees and other plants growing on it, these plant parts can be carried to abandoned channels and other places where a high concentration of sediment would bury the plant fragments and fill in the oxbow. As might be expected, plant parts carried for great distances would tend to be fragmented and shredded, and those deposited close to the place of growth would be less distorted and better preserved. Plants that become preserved at the same locality where they were growing are termed autochthonous (e.g., many Pennsylvanian coal ball deposits), whereas those assemblages that have been transported are termed allochthonous. Preservation of whole plants or plant parts (usually stems and roots) in growth position is termed in situ. The plants that once made up a community, together with the other organisms in the ecosystem, are preserved in the Earth’s crust in a variety of ways, and different kinds of physical and chemical processes were involved during the process of preservation (FIG. 1.6). Moreover, various environmental settings and depositional processes also result in fossils that occur in a variety of forms. Taphonomy is the study of all the processes occurring between the time the organism died and its discovery today as a fossil. These include burial by sediments of some type (e.g., sand, fine mud, ash, etc.), and diagenesis, which is defined as all the chemical and physical changes to the sediment (and the fossils within

Figure 1.10 Cuticle preparation of the seed fern Blanzyopteris praedentata showing numerous trichomes extending from the surface. Bar  1 mm.

it) as it is converted into rock (Gastaldo, 1989; Gee and Gastaldo, 2005). Because of the countless ways in which plants are preserved, the paleobotanist must employ different techniques to extract the maximum amount of information from fossils. For example, when a paleobotanist finds a fossil leaf, it would first be compared to modern leaves, based simply on the overall size and shape, that is, the morphology of the leaf, to identify it. This can include describing the shape and distribution of teeth on the margin of the leaf, if present, and the shape of the base of the leaf compared to the tip, as well as the length and shape of the petiole. Next, the discoverer might examine the pattern of veins in the leaf—the venation, followed by a microscopic examination of the types and distribution of stomata (pores for gas exchange) and other structures on the surface, such as hairs (trichomes) (FIGS. 1.10, 1.11), or trichome bases if the hairs themselves are no longer attached to the leaf. Still later, the paleobotanist might study the ultrastructure of the cuticle (the waxy


paleobotany: the biology and evolution of fossil plants

Figure 1.11 Multicellular, uniseriate trichomes on Blanzyopteris praedentata cuticle. Bar  400 μm.

covering on leaves), or the molecules that remain part of the leaf after diagenesis (geochemistry). It is possible to determine the proportion of carbon isotopes (13C versus 12C) in many fossil plants, and to use these to reconstruct paleoenvironment or the type of photosynthetic pathway (C3 versus C4 photosynthesis) employed by the plant. And in the future? Perhaps paleobotanists will be able to extract information that reveals details about biochemical pathways, developmental mechanisms, and families of genes involved in response to parasites or herbivores that attacked the leaf surface, all from a fossil plant leaf! Although there are numerous variations on the ways in which plant fossils are preserved, there are a few basic types which we discuss later. It is important to keep in mind, however, that all preservation types can intergrade, or a fossil plant may be preserved in more than one way, for example, a compressed plant with a stem that is partially petrified. Finally, there may be certain structures that appear to be an organism, but are not of organic origin. One of the most common of these are dendrites (FIG. 1.12).

Figure 1.12 Pseudofossil (dendrites) that look like the leaves

of a plant, but are manganese oxide that has grown on the bedding plane of a limestone (Jurassic). Bar  2 cm. (Courtesy BSPG.)


As sediments accumulate, such as in an oxbow lake, water is squeezed out, so the sediments become much more compact, and plant fragments contained within them become flattened (Rex and Chaloner, 1983; Chaloner, 1999a). Internal structure is usually obliterated as the cells become flattened, and frequently all that is left is a delicate carbonaceous film that conforms to the original outline of the plant part. This type of fossil is called a compression (FIG. 1.13), and it is one of the most common types of plant fossils. As you might predict, if the sediment grains that bury the plant parts are large and angular, such as sand grains, the resulting compression

Figure 1.13 Compression specimen of Osmunda claytoniites

from the Triassic of Antarctica. Bar  2 cm.

shows less detail than if the sediment particles are smooth edged and very fine, such as clay particles (FIG. 1.14). There is a vast range of sediment size and structure; plant parts have even been preserved in conglomerates (rock made up

chapter 1 introduction to paleobotany, how fossil plants are formed


Figure 1.14 Portion of compressed fern frond (Cretaceous).

Bar  2 cm.

of variously sized pebbles), but generally, compressed plants will be better preserved in clays or shales than in sandstone deposits. Compressions are not always formed in deltaic and fluvial (river) systems; they may be formed in lagoons, along meandering rivers (not necessarily near deltas), and in ponds, swamps, or other depositional systems, as well as in wind-blown sediments. Most often a terrestrial ecosystem is involved (as opposed to a marine environment), although there are instances in which terrestrial plants are even preserved in marine limestones. Plant compressions can also be found in consolidated volcanic ash. These fossils represent plants growing near an area where there was volcanic activity that spewed clouds of ash into the air. Often there is severe atmospheric turbulence near a volcano and thunderstorms may develop as a result. The rainwater and the ash make a fine-grained mud which cascades down the hillsides, picking up and burying plant parts as it goes (Burnham and Spicer, 1986). When the mud hardens, it entombs pieces of plant material, in many cases exactly in the position in which they grew (in situ). For example, in the Cretaceous Baqueró Group in Santa Cruz Province,

Figure 1.15 Portion of a fern frond preserved in tuff from

Argentina (Cretaceous). Bar  2 cm.

southern Patagonia, the tuff deposits were laid down so rapidly that it is possible to trace the fern Gleichenites vertically from the rhizome through the sediments to the leaves (FIG. 1.15) (Archangelsky, 2003). So well preserved are some of the compressions from this site that the plant material can be sectioned and examined with the transmission electron microscope (Archangelsky and Villar de Seoane, 2004). Fossil cytoplasm has even been described in one unusual compression specimen (Hall, 1971) (FIG. 1.16).


paleobotany: the biology and evolution of fossil plants

Figure 1.17 Compression of pinnae in which only the axes and

pinnule venation are preserved (Pennsylvanian). Bar  1 cm.

Figure 1.16 Pollen grain Kollospora extrudens with contents

(cytoplasm?) extending from the wall. Bar  35 μm. (From Hall, 1971.)

A very unusual kind of matrix in which compressions occasionally occur is diatomite—a rock formed from the siliceous shells (frustules) of diatoms (Bacillariophyceae; see Chapter 4), which today inhabit both fresh water and marine sites. Diatomite is especially fine grained and preservation of plant remains in it is often superb. Since the diatom frustules are actually the cell walls of these microscopic algae, in this method of preservation one organism is serving as the preserving matrix for another organism. Leaves are some of the most common plant parts preserved as compressions (FIG. 1.13) and, in many instances, they occur in numerous, closely spaced layers within the rock matrix. Often a collector can uncover the leaves by splitting the rock along bedding planes with a knife (if the matrix is clay) or with a hammer and chisel (if the rock is harder), although sometimes the paleobotanist must resort to more energetic measures to uncover fossils, such as using a jackhammer or even dynamite! Many compressions are of value in showing surface details and overall morphology. Experimental evidence suggests, however, that the size and shape of fossils can vary depending on the matrix in which they are preserved (Rex, 1986). Such features as leaf shape, presence or absence of

a petiole (leaf stalk), leaf margin, trichomes, and the pattern of venation (FIGS. 1.17, 1.18) are generally readily discernible. In some cases it is possible to examine the distribution of stomata in the leaf surface. When there is an abundance of leaves presumably from one species of plant, it is possible to determine the degree of variability exhibited. In other cases this is far more difficult, especially as leaves tend to be the most plastic in their morphology of any plant part. For example, it is often difficult to find two leaves that are morphologically identical on some modern plants, such as the common mulberry tree (Morus spp.). Many conifers (Chapter 21) produce juvenile leaves with a different morphology than that of mature leaves. When found as fossils, these may have been described as two different species. For these reasons, paleobotanists must be extraordinary sleuths in uncovering features that will help distinguish variability within a single species (intraspecific variability) from differing leaf forms that reflect different species (interspecific variability) or genera. Compressions preserved as relatively dark carbonaceous films on a lighter colored rock matrix make examination and imaging of the fossil relatively easy. Sometimes, however, the matrix and the fossil have similar color values and imaging is more difficult. In these cases, details can be enhanced by using sidelighting (a light source at an oblique angle to the compression) or a polarized light source, or by submerging the fossil in some liquid, such as water, xylene, or alcohol. In many instances, the use of cross-polarization

chapter 1 introduction to paleobotany, how fossil plants are formed


Figure 1.19 Arctocarpus cuticle showing paracytic stomata

with cuticular ridges (Cretaceous). Bar  4 μm. (Courtesy G. R. Upchurch.)

Figure 1.18 Cuticle preparation showing venation of Barthelopteris germari pinnule. Bar  1 mm.

(polarized light sources together with a polarizing filter over the camera lens) can significantly enhance contrast (Schaarschmidt, 1973; Crabb, 2001). Another method that has been used with compressed animal remains uses backscattered electron imaging to help differentiate among superimposed anatomical features compressed into a single plane (Orr et al., 2002). CUTICLE Although most compressions show only superficial details, in some instances it is possible to learn a great deal about cellular details of the epidermis from preservation of the cuticle (FIG. 1.19). Primary aerial parts of all vascular land plants are covered with a thin film of waxy material, the cuticle, which prevents excess water loss from the surfaces of the plant. Cuticle is continuous over the surface of the plant, except over the stomatal openings; it is a non-cellular, amorphous layer that is deposited on the outside and into the walls of the epidermal cells. It closely conforms to the contours of the surface of the epidermal

cells and may also extend slightly downward between these cells in flanges which are perpendicular to the surface of the plant. Plant cuticles, as well as waxes deposited on top of the cuticle, are important for the plant in controlling transpiration—the movement of water through the plant, from the roots to its eventual evaporation from the leaves. Most water is lost from stomatal openings, but some can evaporate through the cuticle if it is thin enough or its texture allows for cuticle transpiration. Cuticle is also important in control of gas exchange with the environment, in repelling water from the leaf surface, in attenuation of photosynthetically active radiation (PAR), and in blocking ultraviolet (UV) radiation. The cuticle, especially the leaf cuticle, serves as an interface for a host of biotic interactions between plants and other organisms in their environment, including parasites and herbivores (Riederer and Müller, 2006). Because cuticle is inert and resistant to decay, it is widely preserved in the fossil record, and represents a valuable source of paleobotanical information (Mösle et al., 1997). The leaf cuticle is often preserved as an intact envelope which once surrounded the living leaf tissue (FIG. 1.20). Many fossil cuticles are fragile and must be protected prior to transport from the collecting site. One way to do this is to apply a mixture of nitrocellulose to the surface of the fossil (LePage and Basinger, 1993, 1994), thus preventing loss and breakage of specimens as the freshly excavated sediment dries. Common nail polish has also been used in a similar way. Covering the cuticle with a preservative in the field, however, may prevent subsequent geochemical study of the cuticle (Collinson, 1987).


paleobotany: the biology and evolution of fossil plants

Figure 1.20 Cuticle preparation of Odontopteris brardii pinnule. Bar  1.5 mm.

Figure 1.21 Laboratory set up for preparation of cuticle

Figure 1.22 Pinnule of Pseudomariopteris cordato-ovata showing adaxial epidermal anatomy (Pennsylvanian). Bar  1 mm.

mounts from Eocene clay. Note staining jars at left and pieces of cuticle before and after bleaching, right. (Courtesy D. L. Dilcher.)

It is possible to remove the cuticle from many fossil leaf specimens either mechanically, with a needle or brush, or chemically by dissolving away the rock matrix. Pieces of cuticle retrieved in this way can then be bleached and stained with common biological stains, such as safranin (Bartholomew et al., 1970; Dilcher, 1974; Kerp and Krings, 1999; Krings, 2000a) (FIG. 1.21). When mounted on a slide and examined under a microscope, the cuticle or cuticular fragments reveal considerable epidermal detail (Kerp, 1990). Cuticles of epidermal cells are apparent (Fig. 1.22), along

with the structure of the stomatal complex (the cells associated with the pores, i.e., stomata, in the leaf (Fig. 1.23)), the distribution of stomata, presence of hairs or glands, and other distinguishing features. The cuticle in plants is very much like a fingerprint, in that many species have distinctive epidermal features and patterns that can be useful in identification. Furthermore, it is often possible to demonstrate that disarticulated plant parts, such as leaves, stems, flowers, and seeds, actually belong to the same plant because the individual parts have the same complement of cuticle structures.

chapter 1 introduction to paleobotany, how fossil plants are formed

Figure 1.23 Cuticle preparation of Dicksoniites pluckenetii

stomatal apparatus. Bar  20 μm.


Cuticle and epidermal features can be investigated by transferring the compression fossil from the rock matrix in the form of a transparent film, which can then be examined under a microscope. The film can be made by pouring on a liquid-plastic substance (e.g., clear fingernail polish), letting it dry, and then teasing away the film, with the cuticle adhering to it, from the rock matrix. A comparable technique is to embed the surface of the fossil in liquid plastic (such as that used for preparing biological mounts) and then macerating away the rock with an appropriate acid. In some cases, the cuticle can simply be removed from the surface of the rock with a dissecting needle, without the need for transfer film or maceration. Cuticles can also be transferred onto polyester overlays, which reduces the time of preparation and preserves the fossil from which they were taken (Kouwenberg et al., 2006). In addition to preservation of the cuticle (FIG. 1.23), or some chemically altered form of it (see Gupta et al., 2006), some compressions contain cellular remains as part of the carbonaceous layer (Niklas, 1981a). Niklas et al. (1978) embedded and sectioned exquisitely preserved, compressed fossil leaves from the Succor Creek Formation (middle Miocene) of Oregon, USA, for transmission electron microscopy (TEM) and showed that the cellulosic microfibrillar organization of the cell walls could be seen. Even more astounding was the fact that organelles within the mesophyll cells of the leaves, including chloroplasts with grana stacks (FIG. 1.24) and starch deposits, nuclei (FIGS. 1.25,

Figure 1.24 Stacks of grana membranes in the chloroplast of a fossil Betula leaf (Miocene). Bar  100 nm. (From Niklas, 1981a.)


paleobotany: the biology and evolution of fossil plants

Figure 1.25 Lycospora spore with structures interpreted as

possible chromosomes (Pennsylvanian) Bar  12 μm. Figure 1.26 Tetrad of Flemingites spores showing cell contents interpreted as nuclei (Pennsylvanian). Bar  20 μm.

1.26, 10.20) with condensed chromatin, and plasmodesmata, were preserved in these fossils! The possibility always exists, however, that subcellular structures may be contaminants or artifacts formed as a result of the compression and dehydration of other cell components during diagenesis (Niklas, 1982). Chloroplasts have also been reported from Eocene leaves of Metasequoia collected from the Canadian High Arctic (Schoenhut et al., 2004). These authors suggested that the extraordinary preservation may have resulted from high concentrations of tanniniferous cells in the leaves, which may have inhibited microbial degradation and thus left the cell organelles intact. Some compressed Eocene angiosperm leaves from the Geiseltal in Germany and Clarkia beds in Idaho are still green when the rock is split open (FIG. 1.27), which suggests that the chlorophyll is still intact (e.g., Dilcher, 1967).

Figure 1.27 Compressed angiosperm leaves from the Clarkia

beds with chlorophyll preserved (Miocene). Bar  3 cm.

BIOFILMS AND PLANT FOSSIL PRESERVATION The reason that some very delicate structures are preserved is difficult to understand, but in recent years there has been great progress in elucidating the role that various microorganisms, especially those in biofilms, can play in the preservation process (Borkow and Babcock, 2003). Biofilms consist of an aggregation of microorganisms held together in a slimy matrix of extracellular polysaccharides, which are secreted by certain bacteria in the biofilm. We now know that biofilms are ubiquitous on the Earth, and can be found in environments ranging from streams to desert crusts, to hot

springs; the dental plaque on your teeth represents a type of biofilm. One of the first researchers to recognize the importance of biofilms in fossil preservation was Jean-Claude Gall (1990) in his studies of the beautifully preserved, soft-bodied organisms in the Early Triassic Grès à Voltzia Formation (Voltzia Sandstone) in northeastern France. These organisms, both plants and animals, are believed to have been rapidly covered by biofilms which entombed the animals in lowoxygen conditions that inhibited decay. For more information

chapter 1 introduction to paleobotany, how fossil plants are formed


TEM of fossil cuticles has proved useful in detailing the intricate structural organization of the stomatal complex in certain fossil plants and in understanding preservation processes (Villar de Seoane, 2003). Fossil cuticles have also been examined for their chemical constituents (Tegelaar et al., 1991; Mösle et al., 2002; Gupta et al., 2006). Some cuticles are too thin for standard preparation techniques to be effective, or have been fragmented into minute pieces during fossilization so that they cannot be removed from the rock surface. Under these circumstances, incident light, dark field, or epifluorescence microscopy have been useful in revealing certain cuticle and epidermal characters (Kerp and Krings, 1999; Thomas et al., 2004). Figure 1.28 Surface of Pseudofrenelopsis cuticle showing

distribution of stomata (Cretaceous). Bar  15 μm. (Courtesy C. P. Daghlian.)

on this fascinating subject, see the excellent compendium of papers in Krumbein et al. (2003a). Fossil deposits which show a great diversity of organisms preserved, or excellent preservation, or both, are called Lagerstätten (sing. Lagerstätte). Lagerstätten of various ages have provided paleobotanists with a wealth of information on plants of the past (see, e.g., Chapter 6). ELECTRON MICROSCOPY Scanning electron microscopy (SEM) has become a commonly used research tool in paleobotany to illustrate pollen grains and plant cuticles, because of the extensive range of magnifications available (up to 100,000 times) and the extreme depth of focus that can be achieved. In some instances, compressed leaf surfaces and various structures on them (e.g., stomata and trichomes) can be examined directly with the SEM (FIG. 1.28). In other cases, it is necessary to make latex replicas of the plant surfaces in order to interpret complex structural details. TEM has also been employed in the study of fossil plant cuticles. For TEM studies the plant cuticle is embedded in an appropriate embedding medium (e.g., Spurr epoxy resin), sectioned on an ultramicrotome, and stained in much the same way as living plant tissues are prepared for ultrastructural examination. Many fossil cuticles reveal lamellae and delicate structural features similar to those in modern cuticles (Archangelsky and Taylor, 1986; Guignard and Zhou, 2005). Varying patterns in the ultrastructure of cuticle from the same leaves have been documented and suggest that such differences may reflect cuticle from sun and shade leaves (Guignard et al., 2001). In addition,

CONFOCAL MICROSCOPY Recently, three-dimensional confocal laser scanning microscopy (CLSM) has been added to the arsenal of tools used by paleobotanists to extract information from the fossil record (Schopf et al., 2006). This technique utilizes a sequence of closely spaced images that can provide information in three dimensions. The technique is non-destructive and noninvasive, and has been successfully applied to Precambrian microscopic fossils in order to characterize not only morphology, but the nature of the preservation, including possible cell contents. Because the specimens examined must provide an autofluorescent signal, the fossils cannot have been geochemically altered. Like many techniques used in paleobotany, the procedure needs to be investigated on a particular fossil and, if the desired information cannot be obtained, the investigator needs to modify the technique or explore another means of obtaining the information needed. MACERATION AND DÉGAGEMENT Doran (1980) employed a bulk maceration technique to study Devonian compression fossils that were preserved in a silicified tuff matrix. He submerged the rock in hydrofluoric acid (HF) until the fossils were freed from the matrix. This technique provided nearly complete plants, and Doran could thus more accurately reconstruct the complete morphology of the plants. This method is useful because plant axes that extend into the matrix can be totally freed, but it only works on material where enough organic matter is preserved for the plants to remain intact through the maceration process. A more widely used technique to uncover compressed plant parts within the rock matrix is dégagement. This technique was developed primarily by Suzanne Leclercq (1960) and involves removing the rock matrix—often grain by grain— using fine steel needles (FIG. 1.29).


paleobotany: the biology and evolution of fossil plants

valuable for studying specimens preserved in highly metamorphosed shales where much of the specimen is covered by the matrix and where bedding planes are poor. This technique not only provides details that cannot be obtained with conventional techniques but also makes some threedimensional reconstructions possible through the use of stereoscopic X-ray analysis. Uncovering the three-dimensional nature of fossil plant parts that are compressed or otherwise encased in the rock matrix is a goal in many paleobotanical studies. More recent improvements in the capture and analysis of X-ray images suggest that this technique will be more widely used in paleobotany in the future (Dietrich et al., 2000). X-ray computed tomography (CT) scans have been widely used in medicine, and vertebrate paleontologists have adapted these methods, using high-resolution scans (HRXCT), as a non-destructive method to produce three-dimensional images of vertebrate bones (Conroy and Vannier, 1984). Only recently, however, these methods have been used on fossil plants. Devore et al. (2006) used HRXCT to image the morphology and anatomy of pyritized fruits and seeds from early Eocene London Clay Formation. Because fossils preserved the in pyrite are fragile and deteriorate over time if exposed to air, they are conserved by placing them in sealed tubes of silicon oil. Using HRXCT it is possible to examine the fossils without removing them from the vials, thus decreasing the chances of exposing the specimens to air and subsequent deterioration. This technique makes it possible to study type specimens non-destructively and to reexamine characters that were initially used to define taxa and to evaluate forms for which the taxonomic affinities remain equivocal. It also preserves the integrity of the fossils so that they may be utilized in subsequent studies, perhaps when other, newer techniques are developed (Matysová et al., 2008). COAL AND CHARCOAL

Figure 1.29 Impression-compression of stem surface of Colpodexylon deatsii (Devonian). Bar  2 cm.

OTHER TECHNIQUES Although X-ray analysis has been used for many years by paleontologists working with animal fossils, historically this technique has been little used for fossil plants (Stürmer and Schaarschmidt, 1980). X-ray analysis has been especially

Technically, coal (FIG. 1.30) comes under the definition of a compression fossil, since it represents a complex, heterogeneous mixture of macromolecular organic compounds derived from plant material that has been compressed over time (Scott, 1987). In general, the lower the rank of the coal (the degree of coalification), the more details of plant structure one can observe. The higher the rank, the more the coal has been metamorphosed and the higher the carbon content. Ranks from lowest to highest are lignite, subbituminous coal, bituminous coal, and anthracite. Lignite represents an early stage in coal formation, so the plant parts within lignite are not excessively crushed or decayed and are generally easily recognizable. In some instances, SEM has been a

chapter 1 introduction to paleobotany, how fossil plants are formed


Figure 1.30 Exposure of coal in Antarctica (Permian). Ham-

mer  30 cm.

useful tool in identifying plant parts preserved in lignites (Alvin and Muir, 1969). In some lignites it is possible to tease apart the plant fragments and make whole mounts of various structures, for example in the Brandon lignite, a famous early Miocene plant locality in Vermont, USA (Haggard and Tiffney, 1997). Bituminous coal is more metamorphosed and the plant parts are more flattened, but it is still possible to study plant fragments within the coal. Anthracite coal, the most highly metamorphosed type, is altered to such an extent that little of the original plant material is recognizable. Some coals can be thin sectioned for microscopic examination, and pollen grains, spores, and fragments of cuticle can be discerned. In other instances it is possible to coat the polished surface of coal with epoxy resin and etch it in a low-temperature plasma asher (Winston, 1989). Pieces of coal, or peels of the etched surfaces, can then be examined by light microscopy or SEM to determine the biological composition of the coal. This procedure makes it possible to quantify the plants in various types of coals in instances where coal ball permineralizations (see section “Cellular Preservation”) are not available (Winston, 1986). Coal can also be macerated using chemicals that break down the solid coal and release the plant fragments. It is possible to recognize cuticle, pieces of bark, bits of wood, solidified resins, and especially spores and pollen grains in this type of preparation. Examination of these components allows one to determine the kinds of plants that were growing in the ancient swamps where the coal was formed. Application of 13C nuclear magnetic resonance (NMR) and pyrolysis–gas chromatography–mass spectrometry techniques has been used to define stages in the coalification process more accurately (Hatcher et al., 1989). This same technique has also been used for Cenozoic

Figure 1.31

Marie C. Stopes.

leaf tissues and wood to identify various biomolecules (Yang et al., 2005). The components of coals can also be useful in documenting paleoecology (Poole et al., 2006). Macerals are defined as the organic constituents that comprise coal as seen in polished thin sections. The system of maceral types was originally proposed by the paleobotanist Marie C. Stopes (FIG. 1.31) in 1919, expanded in 1935, and is constantly kept up to date (ICCP, 2001). Maceral names end in -inite; for example, funginite is made up of fungal spores and various fungal bodies, secretinite is composed secretory deposits formed by medullosan seed ferns (Lyons, 2000), and sporinite consists of the sporopollenin walls of fossil pollen and spores. More detailed studies of coal composition provide valuable information about environmental parameters. For example, G. Taylor et al. (1989) suggested that the association of alginite and inertodetrinite (redeposited small particles of fusinite) in the Permian coals of Australia indicates a paleoenvironment of wet, cool summers and freezing winters. It is possible to distinguish between angiosperm and gymnosperm woods in some coals using macerals (Sýkorová et al., 2005). Fossil charcoal or fusain (carbonaceous residue that results from the incomplete combustion of organic material;


paleobotany: the biology and evolution of fossil plants

also called fusinite) is also an important source of paleobotanical data (Cope and Chaloner, 1985; Lupia, 1995), with charcoalified plant remains dating back to the earliest land plants (Glasspool et al., 2004b). There are several techniques used to examine fossil charcoal (Sander and Gee, 1990; Guo and Bustin, 1998; Figueiral et al., 2002) which provide information on taphonomy and paleoecology (Scott et al., 2000), including past atmospheric composition (Scott and Glasspool, 2006) and the presence of fire in paleoecosystems (Uhl et al., 2004, 2007a; Collinson et al., 2007). The discovery of beautifully preserved charcoalified flowers in Cretaceous (Tiffney, 1977; Friis and Skarby, 1981) rocks from around the world has contributed large amounts of information to our knowledge of early flowering plants (see Chapter 22). In North America, Carboniferous coals of different ages are typically characterized by the dominance of different types of swamp plants (Cross and Phillips, 1990). Among the Pennsylvanian coal beds for example, lycopsids, tree ferns, calamites, seed ferns, and cordaites constitute the major types of tropical–subtropical arborescent plants that contributed to the peat formation. Although Carboniferous peat swamps have represented the model system in most interpretations of coal-forming ecosystems, the plants lived in atypical terrestrial communities in which the pH and available nutrients were low. It is now becoming apparent that such factors as types of litter accumulation, nature of the biomass, preservational characteristics of certain tissue systems, microbial diversity, biology of the plants, paleoclimate, and paleogeography are but a few of the parameters necessary to understand and properly interpret coal-forming ecosystems through geologic time (Cross and Phillips, 1990; DiMichele et al., 2002, 2007a). An important study by Gastaldo et al. (2004) underscores the fine resolution needed to understand fossil plant community structure, using an in situ three-tiered forest above a Pennsylvanian coal in Alabama, USA, as the data set. Detailed sampling of this fossil forest indicated that the proportion of canopy, understory, and ground-cover plants was variable across the study area, and that wet–dry gradients and/or increasing habitat specialization did not control the distribution of the plants species in this swamp ecosystem. In rare instances, a coal is formed that consists entirely of cuticular fragments and amorphic organic material (DiMichele et al., 1984). The cuticle is so abundant that it can be peeled off in thin layers. This type of coal is termed a paper coal, alluding to its papery appearance, and is known from relatively few localities, some as early as the Devonian, for example the famous Orestovia paper coal from Siberia

(Ergolskaya, 1936a; Krassilov, 1981a). It is a simple matter to isolate these cuticular fragments by using a chemical base such as potassium hydroxide. The cuticle can then be washed, stained in some cases, and mounted directly on slides for examination (DiMichele et al., 1984; Kerp and Barthel, 1993). Lenses of leaf fragments may sometimes be preserved within coals; these apparently formed in small depressions containing acidic water, which inhibited the normal degradation activities of various microorganisms (Gastaldo and Stub, 1999). In other instances, coals have been found to be made up exclusively of algal remains (see Chapter 4), some as early as the Precambrian (Tyler et al., 1957). Kerogen is a type of fossilized insoluble organic matter that is widely found in sedimentary rocks, and is a common component of various paleobotanical preparations. It is the most abundant form of organic carbon on Earth—more even than coal deposits. The presence of kerogen in rocks has been used as evidence of some of the earliest life on Earth (Moreau and Sharp, 2004; see Chapter 2). Understanding the chemical composition and source of kerogen, termed “typing” the kerogen, is especially important, since these factors help to determine the petroleum-generating potential of source rocks. In the past, kerogen and coal were generally analyzed using just thin sections and light microscopy. Today, both substances are also characterized using standard geochemical methods, such as pyrolysis, gas chromatography– mass spectrometry (GC–MS), analysis of carbon isotopes for total organic carbon (TOC), and others. There are a variety of other techniques available today to investigate the nature of the organic matter that remains after fossilization and to compare the carbonaceous residue to determine the chemical-structural characteristics. Raman spectroscopy has been used to characterize carbonaceous matter in highly metamorphosed rocks for some time (Nestler et al., 2003), but has recently been applied to microscopic fossils preserved in chert of varying ages (Schopf et al., 2005). Unlike most geochemical techniques, Raman spectroscopy is a non-destructive means to analyze ancient organic matter. It provides information on the original biochemistry of the organism, and can help resolve the nature of certain ancient fossil-like organisms (see Chapter 2). Another approach that has been used to examine the chemical composition of fossil plant materials involves energydispersive X-ray microanalysis (EDXMA). With this technique the elemental composition and spatial distribution of fossils can be studied without damaging the specimen (Briggs et al., 2000). Other have used EDXMA to map the distribution of elements in fossil cells (Boyce et al., 2001), and cell walls (Boyce et al., 2002).

chapter 1 introduction to paleobotany, how fossil plants are formed

Figure 1.32 Impression specimen of Osmunda claytoniites

from Triassic of Antarctica. Bar  2 cm.

Figure 1.33 Impression of several whorls of Annularia stellata

leaves (Pennsylvanian). Bar  1 cm. IMPRESSIONS

When a paleobotanist splits a rock that contains fossil plant fragments along a bedding plane, it is sometimes possible to see the carbonaceous film of a compression along one face, and a negative imprint of the plant part, with little or no carbon adhering, on the other face (FIG. 1.32); these two faces are called part and counterpart in paleobotany. The fossil with little or no carbonaceous material is called an impression


1.34 Impression of Sigillaria leaf bases showing parichnos scars and position of leaf trace (Pennsylvanian). Bar  1 cm. (Courtesy BSPG.)


(FIGS. 1.33, 1.34). The imprint will show all the surface details of the compression, such as leaf shape and venation, but there is no actual plant material, that is no carbon, preserved. If you have ever seen a leaf imprint in a concrete sidewalk, you have seen an impression. The process involved in the formation of an impression is also analogous to these modern “fossils.” Such imprints are formed when leaves fall and settle into the wet concrete just after it is poured. As the concrete hardens, it conforms to the contours of the lower side of the leaf that rests on it. Eventually, the leaf disintegrates and the pieces are blown away, but a negative replica of the leaf remains on the hardened concrete. If you have ever put your initials in wet concrete, you have formed an impression. The impression of dinosaur footprints represents an excellent example of this type of fossilization process. When several footprints are of the same type or a series of trackways are discovered in close proximity, it may be possible to extrapolate the stride of the organism and, from this, infer something of the biomechanics of the animal. No cellular details can normally be seen on an impression because there is no adhering organic material, but, in some instances, especially where the matrix is exceedingly fine grained, a replica of the impression can be made with latex or similar material. The replica faithfully reproduces whatever surface details were on the original organism when it was impressed into the mud. Examination of part of the replica with the SEM may reveal details with great clarity, such as the pattern of the epidermal cells, hairs, glands, or other surface features. Some impression fossils are covered with mineral encrustations of different composition, for example iron (Spicer, 1977). These deposits may be the result of the


paleobotany: the biology and evolution of fossil plants

Figure 1.35 Cast of large seed fern seed (Pennsylvanian).

Bar  2 cm.

activities of microorganisms during the decay process. Regardless of their origin, however, the mineral crust may provide an excellent replica of the surface of the plant part, and this can be studied using a variety of imaging modes. MOLDS AND CASTS

Figure 1.36 Cast of arborescent lycopod (Protostigmaria egg-

In addition to two-dimensional plant parts, such as leaves, three-dimensional structures, such as stems, seeds, or fruits, can also be carried into sites where sediment is accumulating and buried. During flood events, massive trunks and tree branches can be moved some distance before they are eventually deposited. If these plant parts became crushed over time, they would be preserved as compression or impression fossils. If, however, the sediment surrounding the threedimensional plant parts hardens before the plant fragment is crushed, the sediment will form a three-dimensional negative, or mold, of the plant fragment. As the plant material disintegrates, a hollow remains in the sediment, which can subsequently be filled in with sediment, thus forming a cast inside the mold. The surface of the cast and the mold can often faithfully reproduce the surface features of a particular plant part, such as characteristic leaf bases on the surface of a stem or the ornamentation of seeds (FIG. 1.35) and fruits. The sediment that fills in the cavity of the mold and solidifies becomes a three-dimensional cast of the original plant part

(FIG. 1.36). In almost all molds and casts no actual plant material remains, but the surface contours are the same as those of the original plant part. The formation of fossil molds and casts parallels the method by which a sculptor creates a bronze statue. The sculptor does not carve directly on a block of bronze, but creates a sculpture with some other medium— wood or wax perhaps. A mold is then constructed around the original sculpture and, when the mold is complete, the original is removed in some fashion (disassembling the mold temporarily or melting the wax). When the mold is reassembled, molten bronze is poured into it, and an exact replica of the original sculpture (but one that involves none of the original material in that sculpture) is formed. Rates of sedimentation in certain areas where molds and casts were formed must have been spectacular. As an example, the sea cliffs at Joggins, Nova Scotia, reveal exposed casts of Pennsylvanian tree trunks 3–8 m tall. The trees must have been buried quite

ertina) (arrows) (Mississippian). Hammer  15 cm.

chapter 1 introduction to paleobotany, how fossil plants are formed


1.37 Compressed trunk cast of Eospermopteris (Devonian). (Courtesy W. E. Stein.)


Figure 1.39 Cast of several tracheids showing circular bor-

dered pits (Miocene). Bar  55 μm.

Figure 1.38 Large pith cast of Calamites gigas (Permian).

Bar  20 cm.

rapidly in place. Sediment hardened and the trees subsequently died, leaving hollows (molds) in the hardened rock that were subsequently filled with other sediments (casts). Compressions, and casts (FIG. 1.37) are important in showing the external form of plant parts in a three-dimensional fashion. Root casts of trees can provide important morphological information useful in determining the type of soil formation and soil drainage conditions when the plants were growing (Retallack, 1990). They also may reveal specialized taphonomic processes and how degradation of organic tissues may have proceeded (Driese et al., 1997). A special form of cast is the calamite pith cast or steinkern, which is a common form of preservation of larger calamite stems and branches. Pith casts are casts (FIG. 1.38) of the hollow pith or medullary region in calamites and

preserve an impression of the outside of the pith cavity, which represents the inside of the vascular tissue and cortex (see Chapter 10 for further details). An unusual example of a mold and cast is represented by fossil wood that was exposed to colloidal silica during the diagenesis; the silica permeated the cell cavities, but somehow did not impregnate the cell walls. After precipitation of the silica within the cell cavities, the cell walls (molds) disintegrated and all that is left are casts of the cavities of the wood cells (FIG. 1.39). These cell casts have the negative contours of the insides of the cell walls and show counterparts of specialized wall structures, such as bordered pits (Chapter 7). CELLULAR PRESERVATION

With few exceptions, none of the preceding types of fossil preservation provide the opportunity to examine the internal structure of a plant or plant part. In the case of permineralizations and petrifactions, however, it is possible to study the internal anatomy of ancient plants (Schopf, 1975). In these fossils, one can examine cells and tissue systems within a plant, as well as produce a series of serial sections that can be used to reconstruct the three-dimensional organization of a structure. This type of fossil is called a permineralization or a petrifaction.


paleobotany: the biology and evolution of fossil plants

Figure 1.40 Several silicified logs in the petrified forest of Patagonia, Argentina. S. Archangelsky, left and T. Delevoryas, right.

In both types, the process begins when a plant part becomes immersed in water containing a high concentration of dissolved minerals, the most common being silica (silicon dioxide, SiO2), which is often readily available in areas of volcanic activity. The plant part, for example a log, gets thoroughly waterlogged, with water and dissolved minerals permeating all the cells and tissue systems. The dissolved minerals may be silica compounds (silicification) (FIG. 1.40), carbonates (e.g., calcium carbonate, CaCO3), oxides, pyrites (iron sulfide, FeS2), or some other type of chemical. At this stage it is unclear what happens, but something triggers precipitation of the dissolved mineral (e.g., a change of pH) so that it hardens around and within the plant fragment. The cell walls of the plant itself may serve as nucleation sites for the growth of the mineral crystals. When the mineral is completely solidified, the plant fragment, in effect, is entombed within solid rock. The fossil can now be sectioned by various means and examined under the microscope to see internal details of the plant. Although several authors have attempted experimental silicification of wood in the laboratory (Leo and Barghoorn, 1976; Laroche et al., 1989), the preservation seen in fossils is often much better. Preservation of plants as petrifactions or permineralizations probably involves several stages of mineral growth, with different sizes of crystals involved. Some more recent work (Channing and Edwards, 2004) suggests that colloidal or gel phases of some minerals may be involved in the apparently rapid preservation of minute details. Permineralizations and petrifactions can both be studied by means of thin sections, sometimes called petrographic thin sections (FIG. 1.41) (Hass and Rowe, 1999). The piece of rock containing the fossil is cut and ground thin enough

Prepared commercial thin section by the James Lomax company of Carboniferous coal ball plants.

Figure 1.41

to transmit light through the section, essentially, the same technique that geologists use to make petrographic rock sections. Rock saws are available that can cut through most types of rock matrix; most have steel blades with diamond particles embedded in the cutting edges. Oil or some other coolant is used to keep the blade from getting too hot as it slices through the rock. Saw blades covered with or made of particles of silicon carbide or some other abrasive can be used to cut through softer material. The fossil to be studied is cut out with a saw, and the surface of the fossil is polished with an appropriate abrasive (e.g., silicon carbide of various grades) until it is smooth. The surface showing the fossil is then attached to a piece of glass with some type of adhesive. After the adhesive has solidified, the glass slide with the piece of permineralized material is placed back in a saw, now a specialized thin-sectioning saw, and the remainder of the rock is cut away to leave as thin a slice as possible. At this point, the rock is still opaque. The next step involves grinding the surface, either by hand on a lapidary wheel or plate, or on a thin-sectioning machine, so that more and more light can pass through the specimen. Eventually, the sliver

chapter 1 introduction to paleobotany, how fossil plants are formed

of rock is thin enough to be examined with a microscope. Sometimes the piece of glass to which the fossil material is attached is the actual slide used for study. In that case, a permanent adhesive, such as epoxy resin, may be used. Some prefer to transfer the ground specimen to a clean microscope slide. In those instances, a cement that can later be remelted is used, such as Lakeside thermoplastic resin. Before the thin section is transferred, it is coated with a transparent plasticlike material to keep the section intact. This thin slice is then placed on a clean slide with a natural or synthetic mounting medium and covered with a cover glass. Once the medium has hardened, the slide can be examined with a compound microscope. Some paleobotanists use no cover glass, but rather examine the rock surface directly using oil immersion microscope objectives; this method has been applied very successfully with the Early Devonian Rhynie chert (see Chapter 8). PERMINERALIZATION In a permineralization, minerals fill the cell lumina and the intercellular spaces, but do not completely replace the cell walls. The cell walls still consist of organic matter, although they may be chemically altered to various degrees. Chemically, the various layers of the cell wall may still be distinct (Boyce et al., 2002), and the permineralization may faithfully reproduce the microstructure of the wall, for example the position of cellulose microfibrils (Smoot and Taylor, 1984). Cellular contents have even been described from permineralizations! The processes involved in the formation of certain types of permineralization in silica are being studied in modern hot springs ecosystems like Yellowstone National Park, USA (Channing and Edwards, 2004), and in filamentous microbes from similar ecosystems in New Zealand (Renaut and Jones, 2003; Jones et al., 2004; Phoenix et al., 2005). These studies underscore the complexity of the preservation process. In general it involves the formation of opaline silica (opal-A) films that coat structures and colloidal silica that permeates cells and tissue systems. More recent work has shown that bacteria may be involved in or even necessary for many mineralization processes, and the field of geomicrobiology is a rapidly growing area of study. In biomineralization, the bacteria may serve as catalysts for chemical reactions and also as nucleation sites for mineralization (see Konhauser, 2007 for additional information on this topic). An analogy of the process of permineralization is the technique used to embed and section living biological material. For example, a piece of plant is killed and fixed in an appropriate chemical. It is then passed through a series of


alcohols to dehydrate the tissue, and finally transferred to molten paraffin or plastic. When the paraffin is cooled, the plant part is completely embedded in it—paraffin is present within the tissues as well as around them. The entombed specimens can then be serially sectioned to reveal details of the cells and tissue systems. PEEL TECHNIQUE. The peel or acetate peel technique (FIG. 1.42) is a simple and rapid method for preparing sections of permineralized plants (Joy et al., 1956; Galtier and Phillips, 1999). In order to use the peel technique, there must be a certain amount of organic matter still present in the cell walls of the fossil plant. If not, thin sections have to be prepared. The rock containing the fossil is sliced with a rock saw (FIG. 1.43) and the resulting slab is polished (FIG. 1.44), first with a coarse abrasive (100–400 grit size) on a lapidary wheel and finally with abrasive of progressively finer grain sizes (600 grit size). The polished surface is then ready to be etched. If the fossils are entombed in calcium carbonate (see coal balls below), etching is done in a dilute solution (5%) of hydrochloric acid (FIG. 1.45). The acid reacts with the carbonate, but not with the organic remains, so the mineral material (CaCO3 in this case) is slowly etched away, leaving the plant cell walls (and cellular contents, if present) projecting in relief from the surface of the slab (FIG. 1.46); the etched surface should not be touched at this stage as the cell walls are very delicate. After the surface has been rinsed and air dried, it is ready to be peeled. Acetone, which is an organic solvent, is poured on the etched surface and, before it evaporates, a thin sheet of transparent cellulose acetate (or similar plastic) is carefully rolled on the surface (FIG. 1.47). The acetone will partially dissolve the lower surface of the acetate sheet, converting it to a liquid that flows in and around cell cavities and intercellular spaces. Because acetone is quite volatile, it evaporates readily, so the lower surface of the acetate sheet quickly solidifies, embedding the cell walls within it. When the acetate is completely dry, it can be pulled from the surface of the rock, taking with it a thin section of the entombed plant (FIG. 1.48). The greatest advantage of the peel technique is that serial sections can be made quickly down through the rock by polishing, etching, and repeating the process again and again (FIG. 1.49). The peel technique (Stewart and Taylor, 1965) can be used for different types of permineralizations, but when the matrix is something other than a calcium or magnesium carbonate, a different acid or a different concentration of acid must be used. When the peel technique was first devised, preformed sheets of cellulose acetate were not available; rather, a solution of parlodion, butyl acetate, amyl alcohol, xylene,


paleobotany: the biology and evolution of fossil plants

Diagrammatic representations of the steps involved in the preparation of the coal ball peel technique. A. Section of coal ball slab (calcium carbonate matrix) containing plant material (crosshatched); B. coal ball slab after acid etching to partially expose plant material; C. etched coal ball slab surface with cellulose acetate sheet in place; D. cellulose acetate sheet (peel) being pulled from the surface with adhering plant material; and E. coal ball peel containing embedded plant material. (From Taylor and Taylor, 1993.)

Figure 1.42

Figure 1.44 Figure 1.43

Several pieces of coal ball after sectioning.

abrasive powder.

Polishing the coal ball slab on a glass plate using

chapter 1 introduction to paleobotany, how fossil plants are formed

Figure 1.45 Etching the coal ball slab in dilute hydrochloric


Figure 1.48 Removing the peel from the coal ball slab



Figure 1.46 Etched surface of coal ball slab prior to flooding

Figure 1.49 Coal ball peel, left, and coal ball slab at right from which it was removed.

the surface with acetone.

castor oil, and ether was poured on the surface and allowed to dry (Darrah, 1936) (FIG. 1.50). This resulted in peels that were not uniform in thickness and were sometimes difficult to mount on microscope slides. Another drawback was the amount of time required for the poured peels to dry on the coal ball surface, as compared with the approximately 20 min required for cellulose-acetate-sheet peels to dry. Despite these drawbacks, the poured peels may still be useful, especially when examining very delicate structures and surfaces that are irregular.

Rolling the acetate sheet into position on the coal ball slab. Bottle contains acetone.

Figure 1.47

COAL BALLS. We know more about the anatomy, morphology, and biology of Carboniferous coal-swamp plants than those from any other time period, and this is primarily due to coal balls. During the Carboniferous, North America


paleobotany: the biology and evolution of fossil plants

Figure 1.52

Digging coal balls from a stream bank in Illinois,


Figure 1.50 William C. Darrah.

Figure 1.53 Transporting bags of coal balls from a site in

Kentucky, USA.

Figure 1.51 Collecting coal balls at a strip mine in southern


and Europe were close to the equator and contained extensive tropical forests which contributed to the extensive coal deposits characteristic of these areas today. Associated with some of these coal deposits are coal balls (FIGS. 1.51–1.53), variously shaped nodules which occur in bituminous coal seams. Coal balls represent permineralized peat deposits and are composed almost entirely of plant parts preserved in calcium carbonate. Some of the first ones found in England

were nearly spherical, hence the name, coal ball, but they can be irregular in shape and range from a few centimeters across to many meters in thickness. Some of the oldest ones come from the upper Namurian (Upper Mississippian) of Germany and the Czech Republic, but they are also known from Permian coal deposits in China. They can be readily studied by means of the peel technique. The method of formation of coal balls has been examined by a number of paleobotanists (Falcon-Lang, 2008), beginning with Stopes and Watson (1908), but the process is still not fully understood. When fresh or partially decayed, the peat was infiltrated by carbonates (fibrous calcite) before there was extensive compaction of the plants within. Since some coal balls are associated with marine limestones, it has been suggested that the plants were growing in lowlying, swampy areas close to the sea, and this hypothesis fits

chapter 1 introduction to paleobotany, how fossil plants are formed

with the paleogeography of Midcontinent North America during the Carboniferous. During storms or marine transgressions (Mamay and Yochelson, 1962), the coal swamp was inundated by seawater, which provided a source of calcium carbonate for permineralization. This hypothesis explains the mixed nature of some coal balls in which both plant and marine animal remains are preserved. Scott and Rex (1985) suggested that all coal balls are not formed by the same process and put forward a non-marine model of formation in which the permineralizing fluids are derived from percolating groundwater high in carbonates. Scott et al. (1996) examined the origin of Carboniferous and Permian coal balls from Euramerica and China and concluded that several different mechanisms were involved, depending on the region and the location of the coal balls within the coal seam. Based on carbon isotopes, they found that some coal balls involved a mixture of marine and meteoric fresh water percolating through the peat and noted that most coal balls formed in freshwater basins with at least some marine influence. There can be little doubt that the formation of coal balls was a highly specialized process, as none are known after the Carboniferous–Permian. To the coal miner these calcium carbonate coal balls represent impurities in the coal that are often termed “fault,” but to the paleobotanist they provide a source of fascinating information that can be used to investigate the biology of the plants that lived in the peat swamps hundreds of millions of years ago. OTHER PERMINERALIZATIONS. Many permineralizations contain silica as the embedding mineral (FIG. 1.54). In fossil peat from Permian and Triassic rocks (FIGS. 1.55, 1.56) from the central Transantarctic Mountains of Antarctica, the silica is in the form of chalcedony (Schopf, 1971). It is possible to make acetate peels of silica permineralizations; however, they must be etched in concentrated HF. When using HF, precise safety procedures (e.g., etching in a fume hood, proper gloves and other protective clothing, and eyewear) must be employed because of the very dangerous nature of this acid. In some cases, especially in certain Devonian fossils, preservation involves permineralization via pyrite (iron sulfides) or limonite (hydrated iron oxides). Plant parts preserved by these minerals have been difficult to study because the material often breaks apart during grinding and is lost. To eliminate such problems, specimens of this type need to be first embedded in plastic prior to cutting (Stein et al., 1982). For fossils preserved in ironstone (fine-grained sedimentary rock), a useful technique involves selectively macerating the specimen in acid to free the silicified axes and then embedding the


Figure 1.54 In situ stump of Triassic tree in Antarctica. Yellow

pen for scale.

Figure 1.55 Block of Triassic chert (orange color) from


axes in bioplastic to examine internal anatomy (Aulenback and Braman, 1991; Serbet and Rothwell, 2006). The degree of detail that can be preserved by permineralization is truly extraordinary, with such delicate structures


paleobotany: the biology and evolution of fossil plants

Figure 1.58 Spores of Cyathotheca tectata showing dis-

tal (left) and proximal surfaces, and distinct ornamentation (Pennsylvanian). Bar  18 μm. Figure 1.56 Block of permineralized peat from Antarctica showing root of Glossopteris, with wood wedges alternating with lacunae, Vertebraria (Permian). Bar  2 cm.

Figure 1.57 Two animals (Ebullitiocaris oviformis) (arrows)

attached to an Aglaophyton major stem (Devonian). Bar  0.5 mm. (Courtesy H. Kerp.)

as starch grains, nuclei, various types of membranes, tapetal deposits, and cells of seed-plant microgametophytes are known. Spores of lycopsids and microspores of Pentoxylon have been interpreted as containing chromosomes (BrackHanes and Vaughn, 1978; Bonde et al., 2004). The flagellum of a chytrid zoospore (FIG. 3.20) (Taylor et al., 1992) and rotifers (FIG. 1.57) from the Rhynie chert, and sperm within the pollen chamber of a Permian seed (Nishida et al., 2003) have been described from permineralized plant remains. There are numerous examples of exceptionally well-preserved plant structures throughout this book.

In some instances, the matrix of the permineralization is too crumbly to allow preparation of ground thin sections or does not lend itself to the peel technique. In such cases, it may be necessary to examine the cut and polished surface with reflected light. If a series of sections is necessary, one must make a photographic record or a series of drawings, because the specimen will be lost as it is continually ground away, leaving no actual record of each face examined. PETRIFACTION Cellular details can also be observed in a petrifaction. In this case, all of the original organic matter in the plant has been replaced by minerals. Many fossil woods, such as those from the Triassic Petrified Forest in Arizona and the Cerro Cuadrado (Jurassic) Petrified Forest in Patagonia, Argentina (FIG. 1.40), are preserved in this manner. It is necessary to make thin sections to study petrifactions, since the etching involved in producing a peel preparation would completely dissolve the specimen. Other techniques such as cathodeluminescence are providing a new source of information about silicified wood (Matysová et al., 2008). UNALTERED PLANT MATERIAL

Some plant parts are found as fossils in an unaltered form, either as body fossils or as chemical fossils. Pollen grains and spores (FIG. 1.58), diatom frustules, cuticle envelopes, various types of resins, such as amber (FIG. 1.59) and calcium carbonate remains of certain types of algae are all examples of unaltered plant fossils. In some instances even the soft parts are sufficiently preserved so that comparisons can be made at the cytoplasmic and ultrastructural level (Wolfe et al., 2006). Holocene peat is an example of relatively unaltered plant material (Williams and Yavitt, 2003). Plant parts became

chapter 1 introduction to paleobotany, how fossil plants are formed


Figure 1.59 Stamen (arrow) embedded in amber in the proc-

ess of shedding pollen as it was being preserved (Miocene). Bar  1 mm. (Courtesy G. O. Poinar.)

Phytolith of a dicot (Oligocene). Bar  20 μm. (Courtesy C. A. E. Strömberg.)

Figure 1.61 Phytolith (laminated trichome-type) from a dicot (Eocene). Bar  30 μm. (Courtesy C. A. E. Strömberg.)

Figure 1.60

incorporated into peat bogs, and because of the high acidity in the bogs, microbial activity is greatly reduced, so little or no decomposition occurs. The accumulated plant debris may build up to a considerable thickness, and while there is some disassociation of plant parts as well as flattening, the bits and pieces preserved can be handled like the same parts of modern plants. Another excellent example of unaltered plant material is diatomaceous earth. Although the cell contents are no longer present, the silica cell walls remain intact and are preserved in such fine detail that the exquisite sculpturing can be easily detected on the surface. Phytoliths also represent unaltered plant material secreted by the living plant in the form of calcium carbonate or opaline silica (FIGS. 1.60, 1.61). They occur in various types

Figure 1.62 Phytolith (echinate sphere-type) from a palm (Eocene). Bar  15 μm. (Courtesy C. A. E. Strömberg.)

of grasses and some trees and, depending upon the species, vary in morphology and size (FIGS. 1.62, 1.63). Although they have been used for many years to study Holocene or Pleistocene habitats, recent research has utilized these plant markers to study older paleoenvironments. They have been especially useful in the interpretation and reconstruction of grassland ecosystems (Strömberg, 2004). Extracting fossil phytoliths can be accomplished using a variety of techniques (Parr, 2002), but it is important to avoid methodological bias, as discussed in Strömberg (2007, 2008). Spores and pollen grains are represented in the fossil plant record in great abundance because the wall (sporoderm) of


paleobotany: the biology and evolution of fossil plants

Figure 1.63 Phytolith consisting of epidermal long cells with papillae and a row of short, siliceous vertical cross cells from a grass (Cretaceous). Bar  50 μm. (Courtesy C. A. E. Strömberg.)

the spore or pollen grain is composed of an especially resistant material called sporopollenin. As such they also represent unaltered plant material. Spores and pollen grains may be common in certain rocks, even when there is no evidence of any other plant parts. It is possible to extract these pollen grains and spores from the rock and mount them on microscope slides. TEM has provided a wealth of new information about fossil pollen and spore walls that has been used to examine the development of the pollen wall (Taylor and Alvin, 1984). More recently, Scott and Hemsley (1991) have discussed the value of laser scanning and scanning acoustic microscopy in the study of paleobotanical materials. CHEMICAL FOSSILS Chemical fossils can also represent a type of unaltered plant material (Hemsley et al., 1996). Chemical signatures, sometimes called biomarkers or geomolecules, are specific for certain groups of organisms. These biomolecules are transformed through time with lipids perhaps having the best opportunity of being preserved (Itihara et al., 1974), whereas nucleic acids degrade more rapidly (Briggs et al., 2000).

Biomarkers in the form of terpenoids have recently been reported in the fossil wood type Protopodocarpoxylon from the Jurassic of Poland (Marynowski et al., 2007). An example of chemical preservation is the presence of hydrocarbons within Ordovician oil deposits that are known to be produced only by the putative cyanobacterium Gloeocapsomorpha (Hoffman et al., 1987; Foster et al., 1989; 1990; Blokker et al., 2001). Pristanes and phytanes are also a type of biomarker. These molecules are believed to be derived principally from chlorophyll degradation, but can also be produced by non-photosynthetic organisms (Hahn, 1982). A number of new techniques have been added to the arsenal of the paleobotanist, and these promise to provide significant advances in a number of areas. They include elemental analysis, chemolysis, pyrolysis, and lipid analysis, and have been used to study a large number of organic compounds throughout the geologic column. Pyrolysis and chemolysis have been used to screen for the chemical composition of the fossil material. Such procedures and techniques have increasing application in determining the systematic affinities of a particular organism, as well as determining various taphonomic processes that may have altered the fossil. These various paleobiochemical techniques involve the extraction of organic constituents still associated with the fossil or represented as residues in the rock matrix. Classes of chemical compounds, such as sterols, aromatics, carboxylic acids, polysaccharides, various types of lignin, fatty acids, and nalkanes, are but a few of the chemical constituents that have been identified in fossil plants. Organic chemical profiles have been used effectively with extant angiosperms and, at one time, chemosystematics represented a basic and almost routine technique in plant systematics (Crawford, 1990). These techniques have sometimes been applied to the study of fossil plant systematics, but it is important to consider diagenetic changes, especially in older fossils. Niklas et al. (1985) used steroid and other cycloalkane–alkene profiles to show that a Miocene Liriodendron leaf was chemotaxonomically more similar to one particular living species despite the fact that the fossil shares morphological characters with two living species. More recently, Mösle et al. (2002) demonstrated that the biomolecular cuticle signature was more comparable between more closely related plants, such as Cordaites and Walchia, than between these seed plants and certain coeval seed ferns. There are, however, certain limitations to such paleobiochemical approaches. For example, various microbial activities may alter the organic chemicals shortly after the organism dies, or there may be modifications to the chemical constituents as a result of diagenesis. Some organic compounds may have formed abiotically, rather than

chapter 1 introduction to paleobotany, how fossil plants are formed


representing the original chemical constituents of the plants. Still others may have percolated through the surrounding rocks and constitute contaminants, even in rocks as hard as cherts. Thus, for paleobiochemistry to be useful to paleobiologists, analyses must follow strict qualitative and quantitative protocols that can be standardized and repeated (Van Bergen, 1999). ANCIENT DNA During the last two decades, systematists working with extant plants have switched, in part, from secondary metabolites to the use of molecular sequences, including nuclear, chloroplast, and mitochondrial DNA, as the macromolecules of choice in developing phylogenetic hypotheses for plant relationships. Golenberg and colleagues (1990) reported the extraction and amplification of an 820-base-pair DNA fragment from the chloroplast rbcL gene (ribulose-1,5bisphosphate carboxylase–oxygenase or RuBisCO) of a Magnolia latahensis leaf. The leaf was collected from the famous mid-Miocene (Langhian) Clarkia beds of northern Idaho, which are dated at 15–15.5 million years old (Ma). In another study on fossils from the same site, Kim et al. (2004) reported amplified ndhF (NADH dehydrogenase) from the same type of leaf (M. latahensis), and rbcL from a specimen of Persea, further suggesting long-term preservation of ancient plant DNA. Recently, however, these studies and others, including DNA sequences from bacteria in insects in amber, DNA in dinosaur bones, and in salt crystals have been challenged on various grounds. As a result some believe that DNA in excess of 1 million years old is probably an artifact (Pääbo et al., 2004). Others believe the study of ancient DNA holds promise at some level (Gugerli et al., 2005), but that the evidence must be compelling, and from multiple sources. Obviously, additional samples from the same site that demonstrate similar results will help verify such reports, and also a closely followed set of protocols will be especially useful in demonstrating the authenticity of ancient DNA (Gilbert et al., 2005). Although barely classified as ancient, an interesting study has recently reported 1000-year-old DNA from excavated wood samples using a strict set of procedures to insure that the material was not contaminated (Liepelt et al., 2006). MUMMIFICATION When conditions of burial are rapid, and especially in very dry or cold environments, wood or other plant parts may survive for millions of years in a relatively unaltered condition. Such mummified remains have been described from Cenozoic deposits (Basinger et al., 1988; Francis and Hill,

Figure 1.64 Mummified leaf of Cryptocarya (Lauraceae)

(Eocene). Bar  2 cm. (Courtesy D. C. Christophel.)

1996; Fukushima et al., 1996) and represent a special preservation type in which the plant tissue was rapidly dehydrated and buried (FIG. 1.64). So well preserved are the cells and tissue systems of these mummified plants that they can be studied by the same techniques as those used to examine extant tissues. Mummified wood is not mineralized, so it can be sectioned using techniques identical to those utilized by wood anatomists for extant material. AMBER Another example of unaltered plant material is amber (FIG. 1.65), a name typically applied to a wide variety of fossilized plant resins (Rice, 1987). Amber has been found in rocks from the Carboniferous to the Pleistocene, but most deposits have been reported from Cretaceous and Cenozoic strata. In the authoritative text, Plant Resins, Langenheim (2003) restricted the term amber to a lipid-soluble mixture of terpenoid or phenolic compounds, distinguishing it from gums, waxes, mucilage, oils, and latex. Amber is produced by a


paleobotany: the biology and evolution of fossil plants

the original plant part, but may contain some components of the original plant. Certain types of algae are more common as fossils because they precipitate or bind calcium carbonate around or in their cells. This calcium carbonate skeleton can build up to a considerable thickness and provides excellent preservational potential. When the alga dies, the calcium carbonate skeleton persists, often for millions of years. Sectioning of these calcium carbonate residues allows one to reconstruct the three-dimensional appearance of the alga by following the configurations of the hollows within the calcium carbonate sheath. These limestone-precipitating algae play a very important part in the build up of some so-called coral reefs; in these cases, the bulk of the reef is produced by the accumulation of CaCO3 precipitated by algae, rather than by the corals living on the reef (see Chapter 4).

PALYNOLOGY Figure 1.65 Winged angiosperm seeds preserved in amber

(Miocene). Bar  1 cm. (Courtesy G. O. Poinar.)

large number of plants; phytochemistry, infrared spectrophotometry, and X-ray diffraction have proved to be important analytical tools in determining the botanical origins of fossil resins (Langenheim, 2003). Because of its sticky consistency when it was produced by the plant, amber has also served as the fossilizing matrix for other organisms. In addition to pollen grains and other wind-borne microscopic plant parts, small flowers, fungi, a variety of insects (Peñalver et al., 2006), and other organisms are often preserved within pieces of amber. Even something as delicate as oil bodies in cells of liverwort leaves (Grolle and Braune, 1988), plant organelles such as chloroplasts and mitochondria (Poinar et al., 1996; Koller et al., 2005), a strand of spider silk (Zschokke, 2003), and amoebae (Schmidt et al., 2004) have been preserved in amber. Poinar (1992) provided an excellent historical account of amber, and the importance of this plant resin in examining the diversity of life preserved in this unique manner, and Grimaldi and Engel (2005) demonstrated the extraordinary preservation and diversity of insects entombed in amber in their comprehensive work, Evolution of the Insects. SUMMARY DISCUSSION

The preceding provided examples of the most common ways in which plants become fossilized, but there are other forms as well, or combinations of the preservational types just discussed. For example, some stem casts may contain a faint outline of the conducting system in the center of the cast. In this case, the cast is not simply a three-dimensional replica of

The science of palynology or, perhaps in a geologic context, paleopalynology is devoted to the study of pollen grains and spores, and also encompasses the investigation of other organic microfossils, such as chitinozoans, acritarchs (Javaux and Marshall, 2006), scolecodonts, dinoflagellates, certain types of microscopic algae, microforaminifera, rotifers, testate amoebae, chitinous fungal remains, and other forms of organic debris sometimes termed varia. Characteristic features such as grain shape (FIG. 1.58), wall sculpture, presence or absence of pores, ridges, furrows, or other types of structural features make it possible to distinguish among grains of various kinds and in some instances to assign them to certain groups of plants. The discipline of palynology is a critical component of understanding the biodiversity of the present and the past, and the important volumes by Wodehouse (1965), Erdtman (1969) (FIG. 1.66), Faegri et al., (1989) (FIG. 1.67), and Traverse (2007) (FIG. 1.68) provided an excellent historical context to the discipline. Palynology has greatly benefited from the introduction of various SEM techniques (Villar de Seoane and Archangelsky, 2008) that have made it possible to image and interpret complex external features on the grains (FIG. 1.69). There has also been an attempt to automate palynology, using texture analysis of SEM images (Langford et al., 1990; Vezey and Skvarla, 1990). This procedure greatly reduces the labor-intensive aspects of palynology and perhaps offers more rapid results, larger data sets, finer resolution of taxa, and possibly greater objectivity in identification (France et al., 2000). Fossil pollen grains and spores (FIGS. 1.70, 1.71) are now routinely sectioned and examined with the TEM as well.

chapter 1 introduction to paleobotany, how fossil plants are formed


Figure 1.68 Alfred Traverse. (Courtesy M. Streel.) Figure 1.66

Gunnar Erdtman.

Figure 1.67

Knut Faegri (right) and Ove Arbo Høeg.

These studies have provided a wealth of detailed information about features of the exine (the outer spore or pollen wall that is composed of sporopollenin) that have been useful for systematic studies. Some paleobotanists have combined SEM (FIG. 1.72) and TEM studies of dispersed spores to try to better understand the affinities of these propagules (Edwards et al., 1996), and to more accurately interpret features of the wall (Wellman, 2001). Information on pollen and spore ultrastructure is often determined from single sections

Figure 1.69 Cyathotheca tectata spore viewed with the scanning electron microscope (Pennsylvanian). (From Taylor, 1972.) Bar  23 μm.

of grains in which the plane of section is not easily determined, but techniques have been developed so that the same grain may be examined and recorded in transmitted light, and then scanning and TEM (Daghlian, 1982). In addition, it is often important to prepare serial sections of the same grain in order to view features, such as lamellae, that may not be consistently present throughout the entire wall (Johnson and Taylor, 2005).


paleobotany: the biology and evolution of fossil plants

Figure 1.70 Ultrathin section of the spore Horstisporites iri-

dodea showing organization of the sporoderm. Bar  2.5 μm. Figure 1.72 Fractured surface of Cyathotheca tectata spore.

Bar  3 μm. (From Taylor, 1972.)

daunting, and some palynologists have developed automation techniques to assist in the process. For example, image analysis of measurements can be used to quantify shape and ornamentation on SEM images (Treloar et al., 2004). In a companion study, Li et al. (2004) used a neural network to analyze texture and were able to correctly identify four extant pollen taxa. GEOCHRONOLOGY AND BIOSTRATIGRAPHY

Figure 1.71 Ultrathin section of Cyathotheca tectata spore

as viewed with a transmission electron microscope. (From Taylor, 1972). Bar  12 μm.

Fossil pollen and spores that are preserved within the pollen sac or sporangium (in situ) can provide valuable information on developmental patterns in the formation of the pollen grain and spore walls. In certain types of fossils, such as permineralizations, it is possible to extract many pollen grains of the same type from a single sporangium, or from multiple sporangia at different stages of development. Study of these grains can thus offer insights into biological processes that took place millions of years ago (Taylor and Alvin, 1984). The process of identifying numerous palynomorphs, especially those that are dispersed (i.e., not in situ) can be

Perhaps the most widespread application of palynology is in geochronology, the dating of events in the history of the Earth. Palynomorphs and certain microfossils can be used in geochronology, that is, dating rock units. Fossils can only provide a relative date for strata, that is in relation to other units. Absolute dating relies on other methods to give a specific date (see radiometric). Dating with palynomorphs is possible because many change through time or possess unique features that allow them to be distinguished from other types. For example, Upper Cretaceous and lower Paleogene rocks in the Northern Hemisphere contain a unique type of fossil angiosperm pollen termed triprojectates (Farabee, 1990). These grains are unusual in that they possess polar and equatorial projections (FIG. 1.73) and a variety of ornamentation patterns that make them especially good biostratigraphic markers. Although the botanical affinities remain problematic, it has been suggested that at least some triprojectates possess sufficient characters to include

chapter 1 introduction to paleobotany, how fossil plants are formed

Figure 1.73 Aquilapollenites grain (Cretaceous). Bar  10 μm. (From Jarzen, 1977.)

them within the modern flowering plant order Apiales (Farabee, 1991). Closely associated with the dating of rocks based on the presence of certain types of palynomorphs is the correlation of rock units in time and space based on the fossils within them, a discipline termed biostratigraphy (Gray et al., 1985). Various types of fossils can be found in different sedimentary environments, for example terrestrial and marine, and each may have its own biostratigraphic markers. For example, calcareous nannofossils are especially good markers in marine sediments from the Jurassic to the recent; plant spores are useful in terrestrial rocks from the Devonian onwards and diatoms (a type of microfossil) from the Paleogene to the recent. At one time, applied biostratigraphy was the method of choice in petroleum exploration. Although a number of other methods are used today, palynostratigraphic techniques are still used today to correlate strata based on the presence of certain types of microfossils. Today there are a large number of consulting companies that provide services in exploration and interpretation of petroleum and mineral deposits based on biostratigraphy and geochronology. The underpinnings of these commercial firms are based on the types and distribution of microfossils. PALEOECOLOGY

Palynology has also been extensively used as a method of characterizing past depositional systems (paleoenvironments) (Farley and Traverse, 1990). Here palynomorphs play an important role in defining, for example, the extent of a marine or terrestrial environment. In other instances, certain types of


palynomorphs may provide valuable information about water depth, temperature, salinity, and nutrient levels where the organisms once lived. In a few cases where vertebrates and invertebrates are found with palynomorphs and plant megafossils, an even greater degree of paleoecological resolution can be obtained (Westgate and Gee, 1990). Other detailed ecological studies are possible based on the frequency and types of pollen present both geographically and stratigraphically within a confined area (Graham, 1990). The use of pollen data in association with megafossil information has had a profound influence on the interpretation of paleophytogeographic patterns throughout the world (see, e.g., Graham, 2000, 2003, on plant distribution in the Caribbean). Such studies are especially valuable when they incorporate both extant and fossil data and are founded on well-defined geographic regions of the world (Graham, 1972; 1973). Other investigations have utilized paleoecological data to show that the early flowering plants were herbs or small trees living in unstable habitats during the Cretaceous (Wing and Boucher, 1998). Certain climatic parameters can also be defined by the occurrence of certain palynomorphs, since various plants respond to minor environmental fluctuations (Pocknall, 1990). Tracing the appearance and disappearance of various palynomorphs vertically in the geologic column provides a method of tracking certain types of climatic shifts. Pollen analysis is a branch of palynology in which the relative proportions of pollen and spores are mapped vertically and horizontally; these proportions are then used to reconstruct the paleoenvironment by comparison with modern proportions of the same or closely related taxa. Although primarily applied to Quaternary deposits, similar techniques have been used in older sediments. Recovery of DNA from Holocene pollen (Bennett and Parducci, 2006) has potential for more accurate identification of certain pollen types, as well as tracking populations of plants through time. Using modern pollen and spores, Traverse (1990) determined the palynomorph load in various types of bodies of water in the Trinity River of Texas. Understanding the dynamics of a modern model system such as this is important in the interpretation of past vegetation. Pollen extracted from marine sediments, together with stable isotopes and radiolarian microfossils extracted from ocean sediment cores, were used to provide data about ocean variability on millennial timescales (Pisias et al., 2001). This information can then be used to directly compare the climate responses of continental and oceanic systems, and incorporated into broader scale climate models. Palynomorphs or microfossils are preserved from every time period of geologic history and in many types of depositional environments, so they are a valuable source


paleobotany: the biology and evolution of fossil plants

Figure 1.74 Elaterosporites klaszii (Cretaceous). Bar  20 μm. (From Jardiné and Magloire, 1965; courtesy M. S. Zavada.)

Figure 1.75 Elaterocolpites castelaini (Cretaceous). Bar  20 μm. (From Jardiné and Magloire, 1965; courtesy M. S. Zavada.)

of information with which to characterize changes in paleoecosystems at different scales (Taggart and Cross, 1990). Although several books have been written on various aspects of palynology, the three-volume set, Palynology, Principles and Applications, edited by Jansonius and McGregor (1996) and the volume by Traverse (2007), Paleopalynology, provide very comprehensive and up-to-date surveys of the discipline. The recent volume edited by Van Geel (2006) focuses on the importance of various microfossils in the interpretation and reconstruction of Quaternary environments and the Glossary of Pollen and Spore Terminology (Punt et al., 2007) will be helpful in understanding the complex terminology used to described pollen and spores. Even with our current extensive knowledge of pollen and spores, some palynological preparations occasionally contain structures which cannot be identified (FIGS. 1.74, 1.75). Graham et al. (2000) described sinuous to coiled filaments similar to the elaters of Equisetum spores (Chapter 10), secondary wall thickenings of conducting elements, or germinating fungal spores. They showed that these filaments actually represent artifacts. Termed petrofilaments (FIG. 1.76), they form when hydrocarbons (asphaltenes) react with solvents in the mounting medium. Figure 1.76 Petrofilaments. Bar  25 μm. (Courtesy A. Graham.)

ABSOLUTE DATING One of the most frequent questions asked of paleobotanists is, “How do you date fossil plants?” Most paleobotanists are familiar with the various groups of plants that lived at different points in geologic time. Consequently, when encountering

a new assemblage of plant fossils, they usually recognize the general age immediately, but this has not always been the case. Our current understanding of the age of fossil floras is based on a long series of efforts to date the various rocks

chapter 1 introduction to paleobotany, how fossil plants are formed

in which they are found. At the present time, the best absolute dating method involves the use of naturally occurring radioactive isotopes contained in various minerals that make up a rock unit. The inherently unstable radioactive isotopes undergo a series of complex transformations (decay) that lead to stable isotopes and, in the process, release energy. The rate of decay, λ, for a radioactive isotope of a given element, sometimes called the half-life, is constant (t½  0.693/λ). Therefore, by measuring the present amount of the radioactive isotope and the present quantity of the stable product, one can calculate how much time has elapsed since the minerals in the rock formed. For example, it is known that the long-life uranium isotope 238U decays to 206Pb with a halflife of 4.5 billion years. Consequently, by measuring the relative quantities of 238U and 206Pb in a sample, it is possible to determine the length of time the decay has been going on and thus the time of formation of the rock. A widely used technique involves the analysis of a very small amount of the relative quantities of uranium and lead contained within zircon crystals (Harrison et al., 2005). These crystals, which may be 0.1 mm in size, form as molten rock begins to cool and thus lock small amounts of uranium into their crystalline structure. This technique, which utilizes a high-resolution ion microprobe, uses a powerful beam of ions to vaporize a tiny portion (two-billionths of a gram) of a zircon crystal (Davis et al., 2003). The vapor is then passed through a mass spectrometer where the different elements are separated and analyzed. Zircon crystal geochronology has been applied widely across geologic time, including dating major earth events such as the formation of the continental crust (Harrison et al., 2005). Other radioactive isotopes differ in their half-lives, for example, 87Rb (rubidium), 48.5 billion years; 40K (potassium), 1.25 billion years; and 235U (uranium), 0.704 billion years. One difficulty in employing these dating techniques is that radioactive isotopes occur more commonly in igneous and metamorphic rocks, whereas almost all fossils occur in sedimentary deposits. Today direct isotopic dating for sedimentary rocks is possible, but only when they contain minerals that have crystallized in the environment of deposition at or near the time they were deposited. One of these is glauconite, a silicate mineral that contains potassium (Smith et al., 1998). Since the potassium consists in part of 40K, the potassium–argon method can be used. Rubidium–strontium dating of some very fine-grained sedimentary rocks also has been successful, but the procedure is difficult and not routinely applicable. A technique has been developed in which actual fossils can be dated. In the upper atmosphere, cosmic rays bombard 14N (nitrogen) isotopes to form an isotope of carbon (14C) that is radioactive. This carbon unites with oxygen to


produce carbon dioxide (CO2). Plants take in and fix (assimilate) this carbon dioxide along with that containing the more common isotopes of carbon, 12C and 13C. Carbon dioxide is continuously assimilated during the lifetime of a plant. When the plant dies, however, it no longer exchanges carbon dioxide with the atmosphere, and thus the ratio of 14C to 13C or 12 C is fixed at that time. At that point, the 14C begins to decay to 14N with its characteristic decay rate (t½ of 14C is 5730 years). For this reason, the ratio of 14C to 12C or 13C is proportional to the age of the fossil. An age limit of 50,000 years (Balter, 2006) applies to this technique because of the short half-life of 14C. This technique obviously has somewhat limited usefulness in paleobotany because the bulk of the fossil plant record is far older. Human influence on the Earth has even altered the usefulness of the 14C dating method, because combustion of fossil fuels and nuclear testing have artificially altered the 14C content of the total carbon reservoir, and this has caused problems in maintaining reliable modern standard samples of carbon. Loss or addition of 14C to specimens and apparent fluctuations of past atmospheric 14C abundance also impose limitations on this dating method. Analytical techniques have been developed that allow direct detection of 14C atoms using high-energy accelerators. This method is especially important as it requires 1 mg of carbon (as opposed to 1000 mg in the conventional methods), and dates can be determined in a matter of hours rather than days.

GEOLOGIC TIMESCALE Frequent references will be made to the geologic timescale in succeeding chapters and a summary of geologic time is provided on the inside front and back cover of this volume. In many ways, the naming of rock units is similar to the naming of organisms, in that the geologic timescale is not fixed, but is constantly updated and refined by the International Commission on Stratigraphy (ICS), part of the International Union of Geological Sciences (IUGS). At the formation level, there is a type section of that formation, where rocks that are typical for that formation are exposed. Periods are formally defined at the base of each period by a Global Stratotype Section and Point (GSSP); the GSSP is designated on the ICS timescale as a golden spike. The global stratotype section for the base of a period is somewhere in the world where an excellent section of rocks is exposed and is agreed upon by a committee of experts. For more information, please refer to Gradstein et al. (2004) or the ICS web page, There are two types of units in geologic time: rock units and time units. Rock (lithostratigraphic) units refer to the


paleobotany: the biology and evolution of fossil plants

physical rocks themselves and the terms, Lower, Middle, and Upper are used for these, for example this fossil was found in Lower Devonian rocks. Time (chronostratigraphic) units refer to the period of time represented by those rocks, and use time units (i.e., Early, Middle, and Late), for example, plants of the Early Devonian. Throughout this book, you will see the following abbreviations for geologic time: ka, Ma, Ga—These stand for, respectively, thousands of years, millions of years, and billions of years (gigayears) before present. These are used for dates, for example this plant lived 450 Ma (450 million years ago). kyr, myr, byr (sometimes written Gyr)—These stand for intervals of time, for example this group survived for 20 myr.

BIOLOGICAL CORRELATION Because radiometric dates are not available for all sequences of rocks in specific geographic regions, it becomes necessary to position a given rock unit accurately relative to its absolute age, a type of relative dating. One way in which a sequence of sedimentary rocks can be grouped according to age is through the use of index fossils. To be effective, an index fossil should (1) distinguishable from other fossils and easily identifiable, (2) have existed during a relatively short period of geologic time, (3) be abundant, (4) be widely distributed geographically, and (5) have lived in different environments, so that it may be preserved in different types of sedimentary rocks. Obviously, not many fossils fulfill all these requirements, and assemblages of several fossil taxa (assemblage zone) are typically more useful than a single species. Generally, the most useful organisms for correlation from one section of rocks to another are those that lived in ancient seas. Pelagic organisms (those that live in the open sea and not on the bottom) provide the best long-range correlations because of their worldwide distribution, at least within certain climatic zones. These organisms include such planktonic forms as diatoms (Chapter 4), foraminifera, silicoflagellates, and coccoliths (FIG. 4.43). These organisms are especially important because their skeletal remains are so small that a large number can be concentrated in a small sample, such as the cuttings obtained from a well boring. Other organisms, such as those that inhabited the ocean floors (benthic forms), typically have a spatially restricted distribution that enables them to be used effectively in correlations of a more local extent. In terrestrial rocks, some of the best index fossils are pollen grains and spores (Gonzalez et al., 2006; Souza, 2006).

They can be carried long distances by wind and, consequently, can be deposited in a wide variety of sedimentary environments. Palynostratigraphy has been an especially important tool in providing correlation between marine and non-marine rocks and in determining the various ecological conditions under which plants lived (Dimitrova et al., 2005). Plant megafossils have also been useful in biostratigraphy, especially when used in the form of assemblage zones. Such studies extend from some of the earliest land plants (Edwards and Richardson, 2004) to measuring Holocene vegetational changes. Noteworthy among plant megafossils used as index fossils are a variety of Carboniferous foliage types (see Chapter 16), which have proven useful in establishing stratigraphic sequences in certain geographic regions (Diaz, 1983, 1985). For example, late Paleozoic foliage types have been useful in delimiting biostratigraphic zones in North America (Read and Mamay, 1964; Gillespie and Pfefferkorn, 1979), southwestern Germany (Germer, 1971), and northern France (Laveine, 1987). In some cases, megafossils have been more reliable than palynology, as palynomorphs are often difficult to extract from the high-rank coals that comprise some of the stratotype sections. In other instances, the identification of particular taxa has been useful in precisely dating tectonic events, such as the Upper Carboniferous folding phases in northwestern Spain (Wagner, 1966), and in documenting climatic changes (Wagner, 2004). There can be little doubt that as fossil plants are better understood they will become increasingly important as stratigraphic markers in biozonation and correlation.

SYSTEMATICS AND CLASSIFICATION This book emphasizes the origin, evolution, and diversity of the major groups of plants based on the fossil record, and their relationships through geologic time; floristic changes through time are discussed to a lesser extent. To do this, we need to address the systematics of plants. The field of systematics is concerned with classifying, naming, and determining the evolutionary relationships of taxa. Taxon (pl. taxa) is a general name to indicate any level of organization (i.e., a species, a genus, a family, etc.). Within systematics, taxonomy is the process of describing and classifying organisms into natural groups and nomenclature is the process of naming taxa. In this book, we will use Linnean nomenclature, in which each plant has a two-part name, sometimes called a binomial, which consists of a genus name and a species name (specific epithet). The rules of naming plants are complex and are encoded in the International Code of Botanical Nomenclature (ICBN) (McNeill et al., 2006), which is refined every 4 years.

chapter 1 introduction to paleobotany, how fossil plants are formed

In spite of long study and continued refinements, naming plants still represents a highly subjective exercise. Generally, all classification systems are based on the same type of evidence: shared features. Shared features allow one to recognize genera, families, and other higher categories of a classification scheme (Funk and Brooks, 1990). Such features may fall into two general categories. In one group are the primitive features (plesiomorphies) that evolved relatively early in the evolution of a group of organisms, such as the vascular tissue present in most terrestrial plants. Features of this type may be regarded as evolutionary holdovers that have persisted, but tell us little about the relationships among members of the group, because every member of the group has the same feature. The other group of characters is believed to have evolved more recently. These advanced or specialized features (apomorphies) can be used to identify organisms that have a common ancestry. The cladistic or phylogenetic system of systematics has the goal to produce a hierarchical organization of taxa based on shared, derived features (synapomorphies) that reflect the evolution of particular groups of organisms (Duncan and Stuessy, 1984). Classifications that group organisms based on the overall similarity of characters, whether both primitive or derived, are termed phenetic systems. Another system that has been proposed for classifying living organisms, including plants, is the Phylocode (Cantino and de Queiroz, 2006). This system is very controversial, but is meant to reflect phylogenetic systematics more than Linnean taxonomy (Nixon et al., 2003). Monophyletic groups, that is those consisting of a single common ancestor and all descendants of that ancestor (clade), are defined solely by their position on the tree of life. Clades may have any rank, but the rank is added after nomenclature is completed. It will be interesting to see if this system gains recognition within the plant systematic community since many plant taxa are now thought to be paraphyletic (Rieseberg and Brouillet, 1994); paraphyletic groups include the ancestor and some, but not all, of its descendants). It would appear reasonable to assume that a classification scheme like the Phylocode, which includes only taxa that fit into monophyletic groups, will not be an accurate and useful tool for arranging the enormous biological diversity represented in the fossil record (Briggs and Crowther, 2001). Moreover, it is difficult to envision how fossils would be treated in the phylogenetic nomenclature of this classification system.

the plants are represented in the fossil record as disarticulated parts. This has resulted in the establishment of a special system of nomenclature for parts of fossil plants. As in other areas of botany, fossil plants are named according to the rules in the ICBN, but in paleobotany each disarticulated part is given a separate generic and specific name. In the past, paleobotanists had two types of names for parts of fossil plants. An organ genus was designated when there was enough information to assign a plant part to a family. For example, Lepidodendron, Stigmaria, and Lepidostrobus (Brack-Hanes and Thomas, 1983) (FIG. 1.77) are generic names used to designate parts (stem, roots, and cones) belonging to a particular type of Carboniferous lycopsid. The form genus was used for fossil plant parts that could not be assigned to a family, for example a piece of wood that could be assigned to the gymnosperms, but not to any particular group of gymnosperms. Originally, an organ genus was considered to represent a more natural (i.e., phylogenetic) taxon than a form genus, but confusion arose because names have been given to the same plant parts in different states of preservation or development. Today, the term morphotaxon has replaced the designations form and organ genera in paleobotany. A morphotaxon is a fossil taxon which, for nomenclatural purposes, comprises only the parts, life history stages, or preservational states represented by the corresponding nomenclatural type (Chaloner, 2004). The nomenclatural type is the plant fossil on which the name is based. Why do paleobotanists give names to different parts of the same plant? The first reason for naming parts is so that the fossils can be studied and referred to in publications


Historically, paleobotanists have utilized a somewhat artificial classification system, since in almost all instances


Figure 1.77

Sheila Hanes.


paleobotany: the biology and evolution of fossil plants

and discussed with other paleobotanists. The other reason is that some identical plant parts may be attached to different plants, for example the Carboniferous lycopsid rooting organ Stigmaria, a morphogenus, has been found attached to different genera of stems. In a case like this, the name of the part is maintained, even though the entire plant has subsequently been reconstructed. The name is also maintained because fossil Stigmaria is still found unattached, and a name is necessary to describe and study the part. In addition, some fossil plant parts, despite extraordinary preservation, cannot be distinguished as belonging to only one group of plants. For example, some species of the Carboniferous foliage type Sphenopteris were borne by marattialean ferns (Chapter 11), and other species of this morphotaxon were produced by lyginopteridalean seed ferns (Chapter 16). Since most plants are constructed of many parts, referring to the entire plant once it has been reconstructed requires a complex system of nomenclature. In general, three procedures are followed when the entire plant is reconstructed: (1) the entire organism is provided with a new name, (2) the whole organism bears the generic name of the part that has priority, that is the first part given a formal name, or (3) the whole plant is referred to informally, for example the “Lepidodendron” plant. Non-paleobotanists may find the nomenclature used in paleobotany confusing and perhaps cumbersome, but the way that fossil plants are preserved necessitates its use, and it is currently the only system that provides for an orderly arrangement of names and, most importantly, for the retrieval of information on plant parts. Some have suggested that the Linnaean system of nomenclature be abandoned for certain types of fossils, especially the use of names that suggest affinities with extant taxa when the exact affinities are unknown (Spicer, 1986, for Cretaceous and Cenozoic angiosperm leaves). Hughes (1989) championed a system in which pollen and other plant parts were given artificial names, so-called paleotaxa (Chapman and Smellie, 1992 for fossil wood), but the system has never been in wide use among paleobotanists. CLASSIFICATION OF ORGANISMS

Each author has his or her own ideas concerning the way organisms should be organized, or in the case of plants, whether they represent a single kingdom or multiple kingdoms. With this in mind, the classification scheme in Appendix 1 is presented merely as a guide to the groups of

algae, fungi, bryophytes, and vascular plants that are discussed in this book. In the case of some groups, such as the hyperdiverse flowering plants, there are so many families with a meager fossil record, or no fossil record at all, that it would be impossible to include them all, so we have tried to provide a sampling of major groups and interesting examples. For the angiosperms, we have followed the system in Cronquist (1988) for the most part, with attention to the system of the Angiosperm Phylogeny Group (1998, 2003); for the algae (Chapter 4), the system in Lee (1999), for the hornworts and bryophytes (Chapter 5), the system in Frahm (2001a), and for the fungi (Chapter 3), The Mycota, Volumes VIIA and VIIB (McLaughlin et al., 2001a, 2001b). Some readers may wish to adapt the plant groups presented in the following chapters to a system with which they feel more comfortable.

BACKGROUND READING There are many approaches that one might take in the preparation of a volume dealing with fossil plants. Through the years there have been many excellent books on paleobotany that have covered the discipline from many perspectives. We have included a number of these in the bibliography so that the reader may obtain additional information on some of the plant groups presented here, or additional ones not discussed. The following volumes (and references cited therein) will provide supplemental details on many fossil plants: Hirmer, 1927; Arnold, 1947; Darrah, 1960 (FIG. 1.50); Andrews, 1961; Delevoryas, 1962; Mägdefrau, 1968; Archangelsky, 1970; Banks, 1970; Emberger, 1968; Beck, 1976a, 1988; Hughes, 1976 (FIG. 22.15); Remy and Remy, 1977; Taylor, 1981; Thomas, 1981a; Stewart, 1983 (FIG. 14.116); Gensel and Andrews, 1984; Tiffney, 1985; Spicer and Thomas, 1986; Meyen, 1987; Thomas and Spicer, 1987; Friis et al., 1987; Taylor and Taylor, 1993; Kenrick and Crane, 1997a; Stewart and Rothwell, 1993; Jones and Rowe, 1999 (for methods in paleobotany and palynology); Gensel and Edwards, 2001; Willis and McElwain, 2002; Anderson and Anderson, 2003; Kenrick and Davis, 2004; Anderson et al., 2007; and the Traité de Paléobotanique series, published under the direction of E. Boureau (FIG. 10.108) (Boureau, 1964; Boureau et al., 1967; Andrews et al., 1970; Boureau and Doubinger, 1975) (FIG. 15.14).

2 PRECAMBRIAN LIFE THE ORIGIN OF LIFE ON EARTH ......................................... 44

OXYGENATION OF THE EARTH (2.45–2.2 Ga) ................... 57

Origin of Life: Theory and Biology ................................................... 46

PROTEROZOIC LIFE .....................................................................59

EARLIEST RECORD OF LIFE ON EARTH ............................ 47


Historical Background ....................................................................... 47

Origin of Eukaryotes .......................................................................... 61

Earliest Records of Life: Paleoarchean (3.6–3.2 Ga) ......................... 47

Mesoproterozoic ................................................................................ 64

MESOARCHEAN–NEOARCHEAN LIFE .................................54

Neoproterozoic ................................................................................... 64 CONCLUSIONS ............................................................................ 70

CONCLUSIONS: ARCHEAN LIFE ............................................ 55

We find no vestige of a beginning, no prospect of an end James Hutton (1788)

Although Hutton was speaking about the geology of the Earth in this famous quotation, it could equally apply to the presence of life on Earth. The past 10 years have seen an explosion of information on the origin of life and evidence for the earliest life on our planet. On the geologic side, this explosion has benefited from the applications of new or improved geochemical, isotopic, and microscopic techniques to Precambrian rocks. On the biologic side, the discovery of life in the deep subsurface, both in terrestrial habitats and in the deep ocean, has broadened our knowledge of where organisms are able to survive and has also influenced ideas on where life may have begun on Earth. A textbook in paleobotany would not be complete without a discussion of the earliest evidence of life on Earth. Admittedly, the first organisms were neither plant nor animal, but it was from such simple bacterial or archaeal biological systems that more complex types of plants later evolved. In the past, geologic time was divided into two major eons: the Cryptozoic (literally, hidden life), now called the Precambrian, and the Phanerozoic (visible life) (inside front and back covers). Unlike the Phanerozoic timescale, where divisions are based on rock units (see Chapter 1), the divisions of time in the Precambrian (FIG. 2.1) are based on

absolute, that is, radiometric dates. These divisions are the Archean, for rocks older than 2.5 Ga (Giga annum or billion years), and the Proterozoic for rocks dated from 2.5 Ga to the Precambrian–Cambrian boundary, which is currently 542 Ma (Gradstein et al., 2004). The last period within the Proterozoic was recently established as the Ediacaran (Knoll et al., 2006a), and it is the only period within the Precambrian based on chronostratigraphy, that is, there is a series of rocks, called the global stratotype, that serve as the reference point for the Ediacaran. The period of time between 4.5 and 4.0 Ga is termed the Hadean Eon, although this is not a formally accepted name. The age of the Earth itself is currently thought to be somewhere around 4.53 Ga; thus, the Precambrian represents almost 88% of geologic time! Isotopic analyses suggest that continental crust formed soon after accretion of the planet, possibly by 4.5–4.3 Ga (Bizzarro et al., 2003; Harrison et al., 2005), and that oceans existed by 4.4–4.2 Ga (Mojzsis et al., 2001; Wilde et al., 2001). These dates are based on analyses of the radioactive decay of the elements lutetium and hafnium (176Lu decays to 176Hf with a half-life of 370 myr) within detrital zircons (see Chapter 1) from the Jack Hills in Western Australia (Kramers, 2001). Due to the continuing


Age (Ma)

System period

Erathem era

Paleobotany: the biology and evolution of fossil plants

Eonothem eon


542 Ediacaran Neoproterozoic

~635 Cryogenian 850 Tonian


1000 Stenian Mesoproterozoic

1200 Ectasian 1400 Calymmian 1600 Statherian 1800



Orosirian 2050 Rhyacian 2300 Siderian 2500

Neoarchean 2800 Archean

Mesoarchean 3200 Paleoarchean 3600 Eoarchean 4000 Hadean (informal) ~4600

Figure 2.1 International Stratigraphic Chart showing the

Precambrian. (From the International Commission on Stratigraphy; Courtesy F. Gradstein.)

dynamic processes of plate tectonics, new crust is created and old subducted, and there are few rocks still available that formed in the Paleoarchean. Currently the oldest known rocks are from the Acasta gneisses of the Great Slave Lake area in Canada (Northwest Territories), dated at 4.03–4.0 Ga (previously 3.96 Ga; Bowring et al., 1989) based on uraniumlead isotopes (U-Pb) (Bowring and Williams, 1999) (FIG. 2.2). Slightly younger rocks are known from Isua and Akilia, West Greenland, including the oldest known sedimentary rocks, dated to 3.82 Ga (Manning et al., 2006). Although no older rocks have survived, several studies suggest that continental crust was present, based on the occurrence of detrital zircons, which were weathered out of older rock. The zircons have

ages slightly older than 4.0 Ga (Iizuka et al., 2006) suggesting that weathering processes were also active at this time. Geochronology (dating) methods have improved significantly in recent years; however, the controversy over these very old rocks will remain, as they all have experienced a long and complex history of metamorphism (Kamber et al., 2001). Although it was previously thought that the Earth formed over many hundreds of millions of years, the convergence of dates for the oldest continental crust with the age of the Earth itself suggests that continental crust began to form very soon after Earth accreted, perhaps within 30 myr (Boyet and Carlson, 2005). What was the environment of Earth at that time? The composition of Earth’s early atmosphere has long been a topic of debate. Stanley Miller, in his classic experiments in the 1950s (Miller, 1953), suggested that Earth’s early atmosphere was a reducing one (high hydrogen content), composed of methane (CH4), ammonia (NH3), hydrogen (H2), and water (Miller, 1953). When a spark (e.g., from lightning) was introduced into this system in the laboratory, amino acids formed. The reducing environment hypothesis fell out of favor for many years, although it has resurfaced in a model that contains more CO2 than previous models (Tian et al., 2005) and seems to be the currently prevailing model. An important aspect of the presence of CO2 and especially methane in the early atmosphere is that they are both greenhouse gases. The early Sun is thought to have been 30% less bright than today, an idea called the Faint Young Sun hypothesis; thus, some means of global warming was needed in the Archean for Earth to be hospitable to life (Kasting, 2005). Concentrations of greenhouse gases, especially methane, which is a much stronger greenhouse gas than CO2, would have been important to increase global temperature. These gases may have formed a haze-like atmosphere (Kasting, 2005), and it has been suggested that this haze may have protected early life from harmful ultraviolet rays before the ozone layer was formed. Oxygen is believed to have been only a very minor component of the atmosphere before 2.4 Ga. Most recent research suggests that the early Earth before 3.2 Ga was hot, perhaps as hot as 60–73°C (Lowe and Tice, 2007). Whatever the initial composition of the atmosphere on Earth, it must have affected the evolution of life and, in turn, was itself changed by the presence of organisms, as will be discussed in the following sections.

THE ORIGIN OF LIFE ON EARTH As far as we know, the land surface supported no life at this time, but the accumulation of bodies of water eventually provided an environment for life to evolve and flourish. The




Mesozoic 0.251



Oldest traces of invertebrate animals Bitter Springs biota 0.83–0.80 Ga

Oldest sphaeromorphic acritarchs 1.85–1.4 Ga

Gunflint biota 1.9 Ga

Cyanobacterial biomarkers and steranes (eukaryotes?) 2.7–2.6 Ga


Major deposits of banded iron formations


Decline of uraninite

0.542 Chuaria and other megascopic fossils worldwide 1.0 Peak in stromatolite diversity 1.3–1.1 Ga

1.6 Paleoproterozoic


Bangiophyte red alga 1.2 Ga Oldest spiny acritarchs 1.5 Ga


Evidence of atmospheric O2 2.5


Oldest terrestrial rocks (Great Slave Lake, Canada) 4.03–4 Ga

Eoarchean Paleoarchean

Presence of oxygen, 2.45–2.2 Ga 1. Decline of BIFs 2. Disappearance of uraninite 3. Appearance of red beds 4. Change in sulfur isotopes

3.2 Strelley Pool chert—diverse stromatolites 3.430 Ga 3.6

4.0 Hadean

Oldest stromatolites (Warrawoona Group) 3.5–3.4 Ga


Oldest microbial fossils 3.5–3.4 Ga

Red beds increase in abundance 2.0–1.8 Ga Oldest red beds ~2.2 Ga


Barberton Greenstone Belt 3.5–3.2 Ga


AGE (Ga) 0 0.66


Oldest calcareous algae




Last major bombardment phase 4.1–3.8 Ga

Oceans present 4.4–4.2 Ga 4.5 4.6

Continental crust 4.5–4.2 Ga Age of the Earth 4.53 Ga

Figure 2.2 Summary of major events in Precambrian history. (Modified from Schopf et al., 1983; and Taylor and Taylor, 1993.)

proliferation of life is thought to have been hampered by early bombardment of Earth by asteroids, which continued until around 4.1–3.8 Ga, the Late Bombardment phase, although this idea is still controversial (Koeberl, 2006). Most theories on the origin of life on Earth suggest that life began with the synthesis of organic compounds, both in the atmosphere and on the surface of the Earth (Chang et al., 1983). These

became increasingly complex, and eventually molecules arose that had the ability to duplicate themselves (selfreplicating) and to perform other complex syntheses. The reports of organic matter in carbonaceous chondrites, a type of meteorite, confirm that organic synthesis has occurred in our solar system and beyond (Ehrenfreund et al., 2001), and numerous laboratory experiments have been performed


Paleobotany: the biology and evolution of fossil plants

to replicate these early syntheses. Many researchers have suggested that an RNA (ribonucleic acid) world preceded the DNA (deoxy-ribonucleic acid) world that exists today (Gesteland et al., 2006). The RNA world hypothesis is based in part on the fact that RNA is a simpler molecule than DNA. It can also replicate itself, encode and build proteins, and function to catalyze reactions (Joyce, 2002; W. R. Taylor, 2006). Moreover, RNA is the principal component of ribosomes, the intracellular bodies where proteins are synthesized, thus suggesting that RNA, rather than DNA, was the original information-containing molecule governing protein synthesis. Others suggest that a pre-RNA molecule, perhaps a peptide nucleic acid, may have been important in the early development of life (Nelson et al., 2000). This chemical synthesis of life is termed abiogenesis. ORIGIN OF LIFE: THEORY AND BIOLOGY

Darwin hypothesized in the late nineteenth century that life on Earth probably arose in a “warm little pond,” and since then this idea has been widely believed. There are several barriers to the origin of life in shallow pools, however, including the necessity of shielding early life on the surface from damaging ultraviolet rays. There is also the question of what would have happened to early life on the surface during the Late Bombardment phase, although it has been suggested that the craters formed by the bombardment would have provided an excellent environment for early life (Cockell, 2006). One of the problems with the origin of life in a primeval soup of prebiotic, organic molecules is whether a mix of molecules in liquid would be sufficiently concentrated to ensure the reactions necessary to form more complex macromolecules. It has been suggested that organics may have been attached to layers of clay or pyrite, thus providing the close proximity needed to catalyze reactions (Ferris, 2006). If the early Earth was hot, as some researchers suggest (Lowe and Tice, 2007), these surface organisms would most likely have been anoxygenic photosynthetic hyperthermophiles. In contrast to the warm little pond hypothesis, the idea that life may have arisen on the ocean floor is more recent. This hypothesis gained support when entire ecosystems were discovered in the early 1980s surrounding deep-sea hydrothermal vents (Waldrop, 1990). Subsequent work has shown that bacterial and archaeal life exists deep in the terrestrial subsurface as well (Amend and Teske, 2005; Chapter 1). In the past 20 years, the diversity of life in extreme environments (extremophiles) has been widely demonstrated, ranging from Antarctic ice to hot springs (thermophiles). Studies of microbial evolution based on rRNA propose that

a hyperthermophile represents the ancestral condition in the Archaea and possibly in the Eubacteria as well (Woese, 1987). This deep biosphere is chemosynthetic and thus not dependent upon sunlight for energy. Although some modern hyperthermophiles live in an aerobic environment, because of the low solubility of O2 at such high temperatures and the presence of reducing gases such as H2S, these organisms are basically anaerobic (Stetter, 2006). These findings suggest that Archean life could have survived periods of asteroid bombardment deep within the ocean or within rocks in earth’s crust. As early as 1988, Gunter Wächterhäuser proposed that life may have arisen around deep-sea vents. The Iron–Sulphur World hypothesis suggests that life began in association with pyrite crystals (FeS2). This idea holds that carbonate, phosphate, and sulfide ions would be attracted to iron pyrite and would rapidly cover every surface. With prebiotic molecules concentrated in this way, and in the high heat and pressure of hydrothermal vents, which are a ready source of iron and sulfur, reactions could proceed much more rapidly than they could at surface temperatures and pressures (Russell and Hall, 2006). Experimental evidence has shown that pyruvic acid (Cody et al., 2000), acetic acid, and peptides readily form in these conditions (Huber and Wächterhäuser, 1997, 1998). More recently, support for the origin of life near hydrothermal vents has come from an interdisciplinary perspective involving biophysics in combination with molecular biology. In a simulation experiment, Baaske et al. (2007) demonstrated that nucleotides concentrated around hydrothermal pores, such as the pore spaces of rocks found at vents. This concentration depended only on the size of the pores and a hydrothermal gradient. These authors suggest that prebiotic molecules could have been concentrated in this way to form the first protocells. Thus, the Iron– Sulphur World hypothesis appears to be a viable alternative to the warm little pond and provides several advantages over that idea, including survival of life during bombardment and changes in surface conditions on the early Earth (Wächterhäuser, 2000). Although originally considered only in the realm of science fiction, the possibility that the early Earth was seeded with organic matter from comets or asteroids, termed panspermia, has received serious consideration. For example, Anders (1989) calculated that only dust-sized particles would be slowed by the atmosphere sufficiently to prevent destruction of organics on impact. Chyba et al. (1990), however, assumed that early CO2 atmospheres would be denser and estimated that organics could have accumulated at the rate of 106–107 kg/yr from 4.5 to 3.9 Ga during the Late




Bombardment phase. The announcement of evidence of life in a Martian meteorite from the Allan Hills, Antarctica (ALH84001; McKay et al., 1996) seemed to support the idea of panspermia. The evidence has been refuted, however, based on several different types of data (Schopf, 1999; Barber and Scott, 2002). There are a number of excellent reviews on the prebiotic (i.e., pre-cellular) origin of life, as well as the atmosphere and environment of early Earth, including books on Precambrian geology and life (Coward and Ries, 1995; Lazcano and Miller, 1996; Orgel, 1994; Schopf, 2002; Knoll, 2003b; Eriksson et al., 2004; Schoonen et al., 2004; Kesler and Ohmoto, 2006; Reimold and Gibson, 2006; Schopf et al., 2007a).


Before discussing current evidence, it is important to mention historical work on Precambrian life. Although there were some earlier reports of Precambrian life, the work of Stanley Tyler and Elso Barghoorn (FIG. 2.3) on the Gunflint Iron Formation (Gunflint chert) of the Canadian Shield provided the first detailed, irrefutable evidence for life in the Precambrian. These researchers began with a preliminary report on the organisms (Tyler and Barghoorn, 1954) and the description of coal from the same area (Tyler et al., 1957). A detailed morphological analysis of the organisms themselves appeared in 1965 (Barghoorn and Tyler, 1965). At the time, the diversity of organisms they described from this 1.9 Ga site (Paleoproterozoic) must have seemed phenomenal. Their work was quickly followed by other reports of Proterozoic fossil organisms and eventually by the discovery of Archean microorganisms. The Gunflint biota will be discussed in more detail later (see section on “Paleoproterozoic” below). EARLIEST RECORDS OF LIFE: PALEOARCHEAN (3.6–3.2 Ga)

The earliest records of life on Earth are of several types, including geochemical evidence (biomarkers or carbon or sulfur isotopes), microfossils (body fossils of microbes), stromatolites, and other microbially influenced sedimentary features, and indirect evidence from molecular phylogenies of living microbes. Each record is controversial if considered separately, but the totality of evidence from many sources continues to push back the date of earliest life. Each type of

Figure 2.3 Elso S. Barghoorn. (Courtesy Schultes and Knoll,


evidence has its own positive and negative aspects, which we will try to address on a case-by-case basis. The majority of the research on early life comes from three localities: (1) Warrawoona Group in the Pilbara Craton, Western Australia (3.515–3.427 Ga); (2) Barberton Greenstone Belt, Kaapvaal Craton of South Africa and Swaziland (3.55–3.33 Ga); and (3) the Isua Group in southwest Greenland, including the island of Akilia (3.9–3.7 Ga). GEOCHEMISTRY The geochemical evidence for life in the Paleoarchean continues to increase. Although some of the records remain controversial, the increasing number of records from various sources, such as biomarkers, stable isotopes of carbon and sulfur, and metal isotopes, supports the hypothesis that life began on Earth soon after the formation of continental crust and oceans. The major problem with the Archean signature of life, however, is the nature of these very old rocks themselves. Many have undergone tectonic events and all are metamorphosed to some extent, which can have a significant effect on isotopic signatures. In addition, even very hard cherts are porous to some degree, so contamination after deposition must also be taken into consideration. Evidence points to the earliest life, probably consisting of anaerobic, chemosynthetic microorganisms. Free oxygen is not found in appreciable amounts until the Paleoproterozoic, so these


Paleobotany: the biology and evolution of fossil plants

early microorganisms are believed to have lived in a reducing environment. Shen et al. (2001) reported on the presence of microscopic sulfides in 3.47 Ga barites from the Dresser Formation (3.515–3.458 Ga), North Pole, Australia. The δ34S isotopic evidence supports the presence of sulfate-reducing microbes. Based on molecular phylogenetics of extant Archaea and Bacteria, sulfate reduction is thought to be an ancient metabolic process (Shen and Buick, 2004). Although Philippot et al. (2007) disagree that these sulfides were formed by sulfate reducers, they do acknowledge that they are biogenic, that is, produced by organisms. They conclude that the mix of sulfur isotopes points to a microorganism that was metabolizing elemental sulfur rather than sulfates. Another important source of Paleoarchean rocks is the Barberton Greenstone Belt of South Africa and Swaziland. Using a combination of petrographic observations, X-ray mapping of elements, and carbon isotope measurements, Banerjee et al. (2006) described 3.5–3.4 Ga micron-scale tubular structures as evidence of life, although aspects of this research are controversial (Kerr, 2004). The tubes have organic carbon associated with them, as well as low δ13C isotopes indicative of life. The samples come from subaerial volcanic rocks, including pillow lavas—not a typical site in which to look for early life. The tube morphology and texture, however, are almost identical to etching seen in the glassy edges of modern pillow lavas, at least some of which are microbially mediated (see also Furnes et al., 2004). Similar structures have been described from 3.35 Ga rocks from Western Australia (FIG. 2.4) (Banerjee et al., 2007). A range of organisms have been found around modern hydrothermal, deep-sea vents, including members of both the Bacteria and the Archaea, although the Archaea appear to dominate at very high temperatures. Some of these are sulfate reducers and obligate chemoautotrophs (Reysenbach and Shock, 2002), but methanogenic microbes are also found. Studies on modern basaltic glass show that the presence of microbes greatly enhances weathering and may produce different chemical products than abiotic weathering (Staudigel et al., 1998). Kerogen, the insoluble carbonaceous matter in rocks, is another source of geochemical evidence of life (Chapter 1). Kerogen has been reported from Precambrian rocks of various ages, but there is controversy over interpretation of the carbon isotopic data from some of this kerogen (see discussion in Pavlov et al., 2001). Marshall et al. (2007) examined kerogens from the 3.4 Ga Strelley Pool chert (North Pole, Australia), which is also an important source of stromatolites (see below). Using a variety of spectroscopic techniques,

Figure 2.4 Photomicrograph of interpillow hyaloclastite from Euro Basalt showing tubular structures (Paleoarchean). Bar  20 μm. (Courtesy N. R. Banerjee.)

including Fourier transform infrared, Raman, and NMR (nuclear magnetic resonance) spectroscopy, they found similarities between this Archean kerogen and younger, Mesoproterozoic kerogen (1.45 Ga), which is known to be biogenic. Based on these results, Marshall et al. (2007) suggested that the Strelley Pool kerogen is also derived from organic matter that had a biogenic origin. After analyzing hundreds of specimens from the North Pole area using petrography and isotope analyses, Ueno et al. (2004) concluded that the kerogens in their samples could have been produced by anaerobic chemoautotrophs, including methanogenic microbes. Their evidence suggests that the organic matter did not come from aerobic photoautotrophs, such as photosynthetic cyanobacteria (Chapter 3). In another study on fluid inclusions in quartz from the 3.5 Ga Dresser Formation (North Pole), Ueno et al. (2006) report the presence of microbial methane (CH4). The methane is highly depleted in δ13C isotopes, suggesting it was produced by methanogenic microbes. Controversy has surrounded the possible evidence of life in rocks from southwest Greenland ever since graphite and microfossils (see below) were found there in the late 1970s. As newer geochemical techniques have become available, the interpretation of Isua rocks has recently generated a new wave of controversy. Graphite was described from Isua rocks almost 20 years ago (Schidlowski, 1988) and more recently from the island of Akilia (Mojzsis et al., 1996), which is dated as 3.825Ga (Manning et al., 2006). The graphite occurs as inclusions within apatite (a calcium phosphate mineral) and was interpreted as being biogenically produced,


based on the isotopically light (i.e., more negative) carbon isotopes obtained from it. Much of the controversy centers around the exact strata from which samples were obtained, as the rocks are exceptionally complex; additional sampling has yielded no graphite (Lepland et al., 2005). Others have suggested that the carbon entered the deposits at a later time than the formation of the rocks themselves (Fedo and Whitehouse, 2002). McKeegan et al. (2007) used Raman spectroscopy to examine the same samples used by Mojzsis et al. (1996) and concluded that graphite is completely contained within apatite, and that it is isotopically light carbon. This supports the claim that the carbon represents a biomarker for ancient life. For a review of the controversy and the issues involved, see Eiler (2007) and references cited therein. MICROFOSSILS (BODY FOSSILS) Although it is clear that life was present in the Paleoarchean, the fossil evidence for life at this time is difficult to interpret due to the microscopic nature of the organisms, a lack of diagnostic and preservable morphologic features to distinguish them, and the changes that have occurred in these ancient rocks since they were first formed. For these reasons, there have been a number of reports of Archean unicells, filaments, and other growth forms that have later been reinterpreted as representing abiogenic structures (i.e., not formed by living organisms). These structures have been labeled in the literature as either pseudofossils (formed abiotically) or dubiofossils (uncertain origin). ISUA GREENSTONE BELT, GREENLAND. Perhaps one of the most discussed examples of a dubiofossil was the description of Isuasphaera, a yeast-like microorganism found in 3.8–3.7 Ga Isua Greenstone Belt metamorphic rocks from Greenland (Pflug, 1978). Isuasphaera consisted of spherical–elliptical structures, some of which gave the appearance of budding yeasts. The description of these complex, eukaryotic organisms from some of the oldest rocks on Earth generated a great deal of controversy at the time, which extended to analyses of amino acids supposedly preserved in these rocks. Bridgwater et al. (1981) described a broad range of diameters for the spherical bodies in these rocks and suggested that the structures represent limonitestained, fluid-filled inclusions in the metaquartzite rather than microorganisms. Their hypothesis was widely accepted (Bridgwater et al., 1981; Schopf and Walter, 1983). More recently, Appel et al. (2003) laid the matter of Isuasphaera to rest. Appel was the one who originally collected the samples containing the spherical bodies, and he and his team were subsequently able to reexamine the exact site and show that



the rocks in this zone (metamorphosed chert) had undergone extreme stretching deformation. Since the spherical bodies could not have been preserved through this event, these researchers concluded that they were formed as a result of pre-Quaternary weathering. WARRAWOONA GROUP, AUSTRALIA. A similar controversy has developed over the rocks and vestiges of life from the Warrawoona Group in the North Pole Dome area of Western Australia. Currently, the oldest microfossils come from this area, in the Apex chert (3.465 Ga), now part of the Salgash Subgroup of the Warrawoona Group (Van Kranendonk et al., 2002). The Warrawoona Group rocks (3.515–3.427 Ga) occur within the East Pilbara terrane and contain both stromatolitic deposits (with and without microfossils) (Lowe, 1980; Walter et al., 1980; Walter, 1983; Allwood et al., 2006, 2007) and cherts that contain preserved microfossils. The relatively high diversity of the fossil assemblage from the Apex chert (Awramik et al., 1983; Schopf and Walter, 1983; Schopf, 1993 and references cited therein) is noteworthy and provides evidence that life may have arisen earlier than 3.556 billion years ago. This silicified microbiota includes four types of filamentous bacteria, colonial unicells, organic spheroids, and radiating filaments, which are hypothesized to have been living in a shallow-water environment. The spheroids and radiating filaments are now regarded as only possible microfossils or dubiofossils. In total, 11 species of filamentous organisms are known, which range from 0.5 to 19.5 μm in diameter; although the smaller diameter fossils are probably filamentous bacteria, the larger ones are most comparable to cyanobacteria. The four types of simple filaments in this microbiota include (1) very narrow forms, Archaeotrichion, that range from 0.3 to 0.7 μm in diameter and up to 180 μm long; (2) possibly septate filaments, up to 340 μm long (0.8–1.1 μm in diameter), Eoleptonema; (3) large, tubular sheaths 3–9.5 μm wide and up to 600 μm long (Siphonophycus); and (4) large septate filaments, up to 120 μm long, which consist of mostly isodiametric cells 4–6 μm in diameter (Primaevifilum) (FIG. 2.5). Based on morphology alone, these microorganisms can be compared with a large number of living bacteria and cyanobacteria (Chapter 3), including anaerobic, autotrophic, and heterotrophic microorganisms. Perhaps the most interesting fossils in this microbiota are the relatively large (8–20 μm in diameter) unicells enclosed by what are described as lamellated sheaths. Morphologically, these fossils are comparable to extant chroococcaleans (cyanobacteria) and, as such, may represent the first evidence


Paleobotany: the biology and evolution of fossil plants

Figure 2.5 Primaevifilum amoenum (Warrawoona Group).

Bar  10 μm. (Courtesy J. W. Schopf.)

of oxygen-producing, photosynthetic life. There have been several accounts disputing the biotic origin of these fossils (Brasier et al., 2002), followed by a series of papers on the geochemistry of the deposits, including the presence of kerogen in the cherts (Pinti et al., 2001) and the use of various techniques, such as Raman spectroscopy, to study the carbon isotopes in the microstructures themselves and to reveal the microfossils in three dimensions (summarized in Schopf et al., 2007b). Some of the difficulties with previous isotopic results on these ancient structures and ways to address these problems are discussed in Marshall et al. (2007). More recently, analyses have shown that the kerogen in the chert is of biological origin (Derenne et al., 2008). The strongest evidence that at least some of the Apex microfossils represent life is the increasing number of reports of structurally preserved microfossils from rocks of similar or slightly younger age—some 14 rock units containing 40 described morphotypes of microfossils (Schopf et al., 2007b). For example, Rasmussen (2000) described filamentous microfossils from the slightly younger Sulfur Springs Group (3.235 Ga) from the same area of the Pilbara Craton in Australia. These organisms were interpreted as living in the rock pores in the shallow subsurface of the seafloor in a hydrothermal environment. They consist of unbranched filaments, 0.5–2 μm in diameter and up to 300 μm long. The filaments are abundant, occurring in dense groups, with many intertwined. These microfossils are interpreted as being the remains of thermophilic, chemotrophic microorganisms, similar to prokaryotes that occur in those environments today. More recently, Sugitani et al. (2007) described a diverse assemblage of microfossils from black chert in the slightly younger Gorge Creek Group (3.19–2.97 Ga, Warrawoona Group), using a combination of morphology (via thin sections) and geochemistry (isotopic analyses of C, N, H, and S). The fossils are an integral part of the rock (i.e., not contaminants) and include four morphological types including threadlike filaments, small spherical structures (FIG. 2.6), spindle-shaped microfossils, and film-like objects (FIG. 2.7). The fossils are abundant and all are 1 μm in diameter, with some filaments extending up to 100 μm in length.

Figure 2.6 Colony-like aggregation of small spheroidal

microstructures (Warrawoona Group). Bar  10 μm. (Courtesy K. Sugitani.)

Perhaps the most interesting fossils are these film-like structures, some of which have small spheres associated with them. Some sheets are folded or wrinkled and range from 50 to 500 μm in size. These may represent parts of microbial mats or perhaps fossilized biofilms (Gall, 1990), which can be especially important in the preservation of many other fossils (Chapter 1). One of the most interesting recent discoveries is a diverse assemblage of silicified microfossils from the 3.446 Ga Kitty’s Gap chert of the Warrawoona Group (Westall et al., 2006a). This deposit occurs in volcaniclastic rocks and represents a nearshore environment of channel infill, termed channel-and-flat sediments. The biota includes two sizes of coccoid cells (0.4–0.5 μm and 0.75–0.8 μm), rare, rod-shaped microfossils 1 μm long, and small, short filaments about 0.25 μm wide. Carbon isotopes from the same layers are very light (26‰ to 30‰), consistent with the presence of microbial life. The coccoids occur in colonies, and the cells within each colony are of the same size and shape; cell sizes differ slightly between colonies, as would be expected of living organisms. There is evidence of life in the form of dividing cells (FIG. 2.8) and the formation of chains of cells, and the presence of collapsed cells with wrinkled surfaces, suggesting cell death. Most important, extracellular polymeric




Figure 2.9 Parallel and overturned filaments in a microbial

mat (Warrawoona Group). Bar  10 μm. (From Westall et al., 2006b.) Figure 2.7 Film-like microstructures with small sphere (arrow) (Warrawoona Group). Bar  50 μm. (Courtesy K. Sugitani.)


Figure 2.8 Colony of coccoidal microfossils and extracel-

lular polymeric substances (EPS). Arrow indicates small coccoid (Warrawoona Group). Bar  2 μm. (From Westall et al., 2006a.)

substances (EPS) are ubiquitous in these deposits. EPS are produced by most bacteria (e.g., the mucilaginous sheaths of cyanobacteria) and serve to attach the organism to a substrate and to aggregate cells together into biofilms (Chapter 1). In the Kitty’s Gap biota, the EPS either surround individual or dividing coccoid cells, or extend laterally to cover and embed entire colonies of coccoid cells (FIG. 2.9) (Westall et al., 2006b). Westall et al. interpreted the filaments as

possible anoxygenic photosynthetic microbes. A biofilm with coccoid colonies occurs as a coating on volcanic grains and is interpreted as including chemolithotrophic microbes. Thus, the Kitty’s Gap chert biota is not only excellent evidence of a microbial ecosystem that lived at about the same time as the Apex microfossils, but also provides compelling evidence for the presence of ancient microbial biofilms. BARBERTON GREENSTONE BELT, SOUTH AFRICA. Another important source of data on Archean life is the Onverwacht Group (3.55–3.33 Ga) in the eastern Transvaal, South Africa. The younger Fig Tree Group (3.26–3.23 Ga) is from this same area (Lowe and Byerly, 1999; Altermann, 2001). Although a number of spherical unicells have been described from the Onverwacht cherts, many of them are now regarded as questionable microfossils (Schopf and Walter, 1983). Smaller cells, including several thought to be in the process of dividing (Knoll and Barghoorn, 1977), are more likely to represent biogenic remains. Walsh and Lowe (1985) described 3.5 Ga filamentous microfossils from the Hooggenoeg Formation (Onverwacht Group). These consist of threadlike filaments 0.2–2.6 μm in diameter. Filaments from the overlying Kromberg Formation are 0.1–0.6 μm in diameter and 10–150 μm long (Walsh and Lowe, 1985). Stromatolites from the younger Fig Tree Group (3.26– 3.23 Ga) provide evidence for the complexity of Archean life (Byerly et al., 1986). More recent studies on the Barberton rocks include the discovery of tubular structures in basaltic glass (Furnes et al., 2004; Banerjee et al., 2006) (see above), the presence of microbial mats (Westall et al., 2006b),


Paleobotany: the biology and evolution of fossil plants

and the research of Noffke et al. (2006) on sedimentary structures (see section “Sedimentary Evidence”). Tice and Lowe (2004) described laminated, carbonaceous material that is believed to represent the remains of microbial mats in the Buck Reef Chert (3.416 Ga). Possible coccoid and rod-shaped bacteria, replaced by minerals, have also been described from the Onverwacht Group (Westall et al., 2001), although the biogenicity of some of these structures has been questioned (Altermann, 2001). STROMATOLITES Especially compelling evidence of early life occurs in the form of stromatolites, which can be defined as layered (laminated), organosedimentary microbial deposits that form on

Figure 2.10 Modern stromatolites in Shark Bay, Western

Australia. (Courtesy J. W. Schopf.)

the bottom of a body of water (Riding, 1999). Although the term stromatolite has been used for layered structures of abiogenic origin, we will use the definition as discussed in Riding (1999) to include structures that are assumed to be biogenic. Stromatolites were widespread in the Precambrian, beginning at 3.5 Ga (Buick et al., 1981) and continuing through the Proterozoic. They are much less common in the Phanerozoic, although they occur in all periods as local reefs; stromatolites are found today only in relatively restricted environments. Modern stromatolites are formed by aggregations of microorganisms, most commonly cyanobacteria, but also green algae (Chlorophyta) and diatoms (Bacillariophyceae) that trap, bind, or precipitate calcium carbonate (CaCO3) in thin layers. Stromatolites are lithified by calcium carbonate when “living,” but most of the Precambrian fossilized stromatolites were taphonomically replaced with silica (for more information on stromatolites, see Walter, 1976; Riding, 1991; Bertrand-Sarfati and Monty, 1994). Probably the best-known site where stromatolites are actively formed today is Shark Bay (FIG. 2.10), a hypersaline lagoon along the western coast of Australia. The organisms that form these stromatolites are filamentous and coccoid members of the Cyanobacteria (Chapter 3), which live in colonies on the upper surface of the accumulating calcium carbonate. These microorganisms typically have mucilaginous or gelatinous, polysaccharide sheaths (EPS) that serve to trap particles of carbonate in the seawater. As they photosynthesize, the cyanobacteria deplete the carbon dioxide in the immediately surrounding water, which also causes the precipitation of calcium carbonate (FIG. 2.11). The CaCO3 deposition continues in thin layers, or lamellae (FIG. 2.12), as the algal colonies continue to grow on the

Figure 2.11 Modern stromatolites from Laguna Mormona, Mexico. (Courtesy J. W. Schopf.)


upper surface of the columnar calcium carbonate. When the colony dies, the calcareous structure may persist as evidence of the organisms that formed it. Stromatolites were widespread and important during the Precambrian, as it was a microbially dominated world until the latest Proterozoic. Some Proterozoic oil deposits found associated with stromatolites are thought to be microbially produced. Some fossil stromatolites contain remains of the microorganisms that formed them, but more often only the layered structure is preserved. If the organisms themselves are preserved, it is usually the result of secondary silicification of the stromatolite. Unlike later stromatolites, the Paleoarchean ones were probably produced by anoxygenic (non-oxygen producing) microorganisms, a hypothesis that is supported by isotopic studies of Archean rocks and by modern experimental work (Bosak et al., 2007). In a recent review of Archean life, 48 sites were listed that contain Archean stromatolites believed to be biogenic (Schopf et al., 2007b); a decade ago there were only a handful. A much more detailed picture is beginning to emerge of the environment during deposition of these rocks. Van Kranendonk (2006) provided evidence that the Warrawoona Group began as a continental volcanic platform. The Pilbara Supergroup (FIG. 2.13) includes autochthonous deposition of volcanics with numerous, interbedded chert layers over 565 myr. The Strelley Pool Chert (SPC) (Kelley Subgroup, Van Kranendonk et al., 2002) contains the most widespread stromatolites (FIG. 2.14). Allwood et al. (2006) described seven different morphologies (FIG. 2.15) of stromatolites

Figure 2.12 Modern stromatolite from Laguna Mormona,

Mexico, showing lamellae of cyanobacterial colonies. (Courtesy J. W. Schopf.)



over a distance of several kilometers in the SPC (3.430 Ga). From a detailed sedimentologic analysis, these authors suggested that the stromatolites were best developed in relatively restricted parts of a peritidal carbonate platform (Allwood et al., 2007). The morphologies range from typical, laminated dome-like stromatolites to conical forms and wavy or bumpy (egg carton) types (FIG. 2.16). A biogenic origin of these stromatolites is supported by the fact that all have morphologies similar to known microbialites (microbially induced sedimentary structures) and none can be explained by abiogenic development. The diversity of forms present and their extended distribution suggest an entire ecosystem composed of microbial communities. SEDIMENTARY EVIDENCE A number of other sedimentary and organosedimentary structures are often used as evidence of past microbial life. The general term for such structures is a microbialite, which is defined as an organosedimentary structure formed by the interaction of microbial communities with sediment. This interaction can include trapping or binding sediment, or biologically mediated calcification (Burne and Moore, 1987; Riding, 2006b). Stromatolites are thus a type of microbialite—one that has a finely laminated internal structure. Thrombolites are microbialites with a clotted internal texture. Microbially induced sedimentary structures (also called MISS) also provide evidence of past life (Noffke et al., 2001). Noffke et al. (2006) describe evidence of microbial mats in the form of wrinkle structures, desiccation cracks, and roll-up structures in 3.2 Ga sandstone rocks from South Africa. Similar structures have also been described from the Mesoarchean (2.9 Ga) of South Africa (Noffke et al., 2008) and from various localities throughout the Precambrian (Simonson and Carney, 1999; Schieber, 2004). These Archean sedimentary features are comparable to those observed in similar modern environments and form via the stabilization of sediment by overlying microbial mats. The Archean structures formed in a tidal flat setting and resemble modern mats in cross section, showing a laminated structure; carbon isotope values are also consistent with a biogenic origin (Noffke et al., 2006). Cyanobacteria are the most common organisms involved in the formation of similar modern sedimentary features, but whether these Archean mats were formed by cyanobacteria, or whether they were oxygenic or anoxygenic cyanobacteria, is not yet known. If the mats were formed by oxygenic photosynthesizers, then they would represent an early source of atmospheric oxygen on Earth (Noffke et al., 2006).


Paleobotany: the biology and evolution of fossil plants

Cherts with putative biosignatures 3.240 Ga (Kangaroo Caves Fm.)



Gorge Creek Group

Sulfur Springs Group

V 3.325–3.315 Ga (Wyman Fm.)





3.346 Ga




Kelly Group





3.35 Ga (Euro Basalt) Strelley Pool Chert “Kitty’s Gap” chert


3.458–3.426 Ga (Panorama Fm.)



V Apex chert

Warrawoona Group

3.471–3.463 Ga (Duffer Fm.)









Shale 3.477 Ga (McPhee Fm.)


Dresser Fm.


Banded iron formation



3.496 Ga




Felsic volcanic rocks

V 3.508 Ga


V 3.515 Ga (Coucal Fm.)



Basaltic/komatiitic rocks

 Granitoid rocks 




V Basalt

U-Pb or Pb-Pb age (formation dated)

Figure 2.13 Generalized stratigraphic column of the Pilbara Supergroup showing position of various putative fossil-bearing chert formations. (From Allwood et al., 2007.)

MESOARCHEAN–NEOARCHEAN LIFE The record of life on Earth becomes more widespread and more diverse during the Mesoarchean (3.2–2.8 Ga) and

Neoarchean (2.8–2.5 Ga). Filamentous microorganisms have been described from rocks of the Fortescue Group in Western Australia (2.768 Ga) (Schopf and Walter, 1982). Although poorly preserved, these appear as two types of




Beukes and Lowe (1989) reported on 3 Ga stromatolites from South Africa (Pongola Supergroup) that they believe were formed in a range of shallow depositional settings, including intertidal mud flats and subtidal channel environments. Noffke et al. (2008) reported microbially influenced sedimentary structures (MISS) from ancient tidal flat facies of the 2.9 Ga Pongola Supergroup, South Africa, which are similar to the Paleoarchean ones described earlier (Noffke et al., 2006). They delimit four levels of microbial mats, which can be distinguished by their specific MISS, and which also occur in modern tidal flat settings. Although the analogous modern features are formed by benthic cyanobacteria, what types of microorganisms formed the fossil features remain unknown.


Figure 2.14 Stromatolite from SPC showing concentric lami-

nations (Pilbara Supergroup). Bar  20 cm. (Courtesy J. W. Schopf.)

filaments: narrow threads (1 μm wide) consisting of diskshaped cells and larger filaments (10 μm in diameter) made up of barrel-shaped cells enclosed by a multilamellated sheath. From 3.2-billion-year-old Mesoarchean rocks of the Cleaverville Group, Pilbara Craton, Kiyokawa et al. (2006) described an extensive stromatolitic layer, which extends for 1 km and contains wavy, colloform laminations similar to those seen in columnar stromatolites. Within the same carbonaceous Black Chert Member, a diverse microbiota is preserved in silica, including six morphological types: (1) spiraled structures 50–150 μm long and 10 μm wide; (2) rodshaped filaments 50–80 μm long, with cell walls 2–5 μm thick; (3) delicate, short filaments, which are compared to Primaevifilum from the Apex chert; (4) branching, dendritic filaments (FIG. 2.17) 100 μm long and 1 μm in diameter, which appear to be made up of small rod-shaped units; (5) carbonaceous, mat-like material that may represent part of a biofilm; and (6) spherical masses of carbonaceous material. The putative microbial fossils are syngenetic, that is, they were there at the time the rock formed and are preserved in three dimensions in the fine chert.

Despite the fact that there are doubters for almost all of the evidence for life in the Archean, evidence for early life has expanded tremendously in the past 20 years, from body fossils to isotopic and geochemical studies. These data are supported in most cases by detailed stratigraphic and sedimentologic evidence of the environments of early life, and by knowledge of the depositional and metamorphic history of the rocks that contain ancient carbonaceous matter. Nisbet (2000) provided an excellent summary of the variety of environments for Archean life, including deep-sea hydrothermal vents, open ocean, lacustrine environments, hydrothermal sites around active volcanoes, and anywhere that microbial mats occur today, for example, in coastal sediments (FIG. 2.18). Perhaps more than in any other area of paleobiological research, the multidisciplinary approach has been used to address questions and provide answers relating to the earliest life on Earth. The advances made in phylogenetic and evolutionary microbiology have also contributed to an increased understanding of early life. For example, based on studies of modern taxa and on geochemical studies, it now appears likely that the first organisms were not photosynthetic cyanobacteria, but perhaps chemosynthetic organisms (e.g., chemolithotrophs) that lived around deep-sea vents, or anoxygenic photosynthesizers which could live closer to the surface (Nisbet and Fowler, 1999). Both Archaea and Eubacteria are known to live in these habitats today, and there is isotopic evidence that points to the presence of sulfate-reducing bacteria and methanogenic Archaea around 3.5 Ga. The knowledge that earliest life did not necessarily have to arise in a “warm little pond” has, in some ways,


Paleobotany: the biology and evolution of fossil plants

Encrusting/domical laminites Large complex cones

Egg carton laminite

Cuspate swales

Iron-rich laminite

Small crested/conical laminite

Wavy laminite

Figure 2.15 Synoptic profiles of seven stromatolite facies from the SPC. (From Allwood et al., 2006.)

Figure 2.16 Example of bumpy stromatolite morphology

Figure 2.17 Dendritic carbonaceous material (Cleaverville

(Pilbara Supergroup). Bar  10 cm. (Courtesy J. W. Schopf.)

Group). Bar  100 μm. (From Kiyokawa et al., 2006.)


Hydrothermal communities around andesite volcanoes

Lake communities

Coastal sediment S-microbial mats


Mid-ocean ridge chemotrophic community

Hydrothermal systems around komatiite shields

Mg, SO4


Light Organic debris S cycle and methanogens



Fe, Mn, S, CH4,H2


Ni, Co, Fe, S, Mg

Cu, Mo Zn, S


P cya lankto n o b n ic act e r ia

Stromatolites Lava

Hydrothermal supply of metals and reductant in deeper water

Open ocean


More oxidized More reduced

Figure 2.18 Diagrammatic representation showing numerous sites where early life may have flourished. (From Nisbet, 2000.)

revolutionized the picture of early life on Earth. Nisbet and Fowler (1999) noted that anaerobic photosynthesizers, similar to modern green gliding bacteria, could have colonized shallow-water habitats. Bacterial sulfur metabolizers would have colonized anaerobic sites, perhaps at the bottom of microbial mats, where they would live off of decaying matter from the autotrophic organisms above them. Organisms similar to modern purple bacteria (Proteobacteria), which are anoxygenic photosynthesizers, would occur in more microaerophilic areas within the same microbial mat community, perhaps similar to some of those described by Allwood et al. (2006) from 3.4 Ga rocks. Cyanobacteria would be present on the top surface of such a mat. Similar microstrata occur in modern microbial mats. As oxygenic cyanobacteria diversified and their biomass and distribution increased, the world began to change.

OXYGENATION OF THE EARTH (2.45–2.2 Ga) Around the Archean–Proterozoic boundary, Earth began a transition phase in a number of different areas. The two supercontinents that came together in the Neoarchean began to break apart in the Paleoproterozoic, and a major glaciation (the Huronian) occurred from 2.45 to 2.22 Ga. There were large eruptions of flood basalts and an increase in

atmospheric oxygen, eventually leading to the oxygen-rich environment we live in today (Melezhik et al., 2005). Some researchers have suggested that the rise in O2 triggered the glaciation, as methane in the atmosphere was oxidized to CO2, thereby reducing greenhouse gases (Kopp et al., 2005). Banded iron formations (BIFs) also reached their greatest extent during this time. As a result of all of these interrelated events, especially the rise of free oxygen, large transformations in biogeochemical cycles must also have taken place (Konhauser, 2007). Although the upper parts of the ocean gradually became enriched with oxygen, it is hypothesized that the deeper parts of the ocean remained anaerobic and were rich in hydrogen sulfide (H2S). The origin of oxygenic photosynthesizers (those that produce oxygen) represents an important benchmark in biological evolution (Schopf et al., 1983). Among the prokaryotes, cyanobacteria are the major organisms that fill this ecologic role. They were no doubt present in the Archean, but perhaps not in sufficient numbers to make a difference on a global scale. In the Paleoproterozoic, however, this condition began to change and the increase in numbers of these organisms and their distribution on Earth eventually resulted in an atmosphere enriched in oxygen. The presence of free atmospheric oxygen subsequently resulted in the ozone layer, thus making it possible for organisms to live on the land without being destroyed by ultraviolet radiation (Chapter 6). The presence of adequate oxygen was also necessary for


Paleobotany: the biology and evolution of fossil plants

the evolution of multicellular animals (Berner et al., 2007). The evolution of cyanobacteria thus signaled a revolution in life on Earth. These organisms were not only able to utilize the most abundant form of energy on Earth—sunlight—to split water molecules, but were able to do it more efficiently. Since cyanobacteria utilize both photosystems I and II (as do all green plants), their appearance in the fossil record also signals the evolution of an advanced biochemical system compared to the single system in other photosynthetic bacteria. It has also been hypothesized that the appearance of cyanobacteria and oxygenic photosynthesis resulted in a rapid colonization of shallow-water habitats on Earth (Nisbet and Fowler, 1999). There has been a continued controversy over the timing of the first appearance of oxygen-producing photosynthesizers (Towe, 1990), that is, cyanobacteria. Hofmann and Schopf (1983) suggested that some of the organisms found in Paleoproterozoic biotas represented aerobic photoautotrophs, cyanobacteria, based on the presence of morphologies comparable to extant cyanobacteria. Stromatolitic evidence of early photosynthesizers comes from Neoarchean (2.7 Ga) lacustrine deposits from Western Australia (Fortescue Group) (Buick, 1992). In this study, a combination of morphologic, geochemical, and sedimentologic data were used to propose that the stromatolites were probably formed by filamentous, phototropic bacteria. Based on the complex trophic system in these lakes, Buick concluded that O2-producing photosynthesizers must have been present, since a deficiency of sulfates in the system would have excluded anaerobic photosynthesis. Summons et al. (1999) reported the occurrence of 2-methylhopanoids or their derivatives, which represent a biomarker for cyanobacterial photosynthesis, in rocks as old as 2.5 Ga. They also noted that the abundance of these biomarkers at this time supports microfossil evidence that cyanobacteria arose before 2.5 Ga (Brocks et al., 2003). In the past, the occurrence of BIFs (FIG. 2.19) in Proterozoic rocks has been used as evidence of early oxygenproducing photosynthesis. BIFs are laminated units containing iron oxides, sometimes in very fine laminae, separated by silica-rich layers. Some of the layers are laterally quite extensive, for example, a layer just a few centimeters thick can be followed over 50,000 km2 in the Hamersley Basin of Western Australia. BIFs were widespread from 2.2 to 1.9 Ga and rare afterwards. It was once thought that they formed abiotically, by the oxidation of ferrous iron in the presence of free oxygen. More recent work, however, suggests that only biotic interactions could have precipitated such large quantities of iron (Konhauser et al., 2002). The source of the dissolved ferrous iron is thought to be hydrothermal activity at

Figure 2.19 Banded iron formation (Negaunee Formation). (Courtesy J. W. Schopf.)

mid-ocean ridges. Several studies have suggested that ironoxidizing bacteria (chemolithotrophic organisms) could have precipitated the amount of iron present in the Hamersley Group of Western Australia, even at cell densities lower than those seen in modern coastal waters (Konhauser et al., 2002; Kappler et al., 2005). Thus, it can no longer be assumed that the presence of BIFs represents proxy evidence for early oxygenic photosynthesizers. The distribution of other types of rocks, however, does appear to be important in dating the origin of an oxidizing atmosphere. Detrital uraninite, siderite, and pyrite occur in 3.25–2.75 Ga fluvial deposits from Australia, but are extremely rare after 2 Ga (Rasmussen and Buick, 1999). There is evidence that these minerals had been transported in water that was well aerated. All of these compounds are known to be unstable under oxidizing conditions (J. Walker et al., 1983). In addition, red beds are known to be rare or absent in the Archean. Red beds obtain their characteristic color from the presence of ferric oxides, which form as a result of subaerial oxidation, usually in fluvial rock sequences. Despite the occurrence of suitable rock sequences on the early Earth, red beds are far more abundant in the Proterozoic than in the Archean, and this change is generally attributed to an increase in atmospheric oxygen around 2.3 Ga. Another method to extrapolate paleoatmospheric O2 levels involves the fractionation of sulfur isotopes (Pavlov and Kasting, 2002). Using this method, Bekker et al. (2004) suggested that the atmosphere already included pO2 levels 105 of the present atmospheric oxygen levels (PAL) by 2.32 Ga. The timing for the disappearance of uraninite deposits, appearance of red beds, decline of BIFs, and shift in sulfur isotopes approximately coincides at 2.3–2.2 Ga,


and this time no doubt represents evidence of global atmospheric oxygen, even though the levels were lower than today. The only viable source for this oxygen is the product of photosynthesis by cyanobacteria, since they represent the earliest oxygen-producing photosynthetic organisms in the fossil record. Oxygenic photosynthesis certainly must have evolved earlier than this time, however, in order to build up enough oxygen in the atmosphere and ocean to leave a chemical signature in the various rocks. Tomitani et al. (2006) used molecular phylogeny of extant cyanobacteria, within the context of geochemical evidence, to suggest that oxygen levels rose rapidly between 2.3 and 1.9 Ga. These results coincide with the disappearance of BIFs and suggest that oxygen levels were probably 10% PAL by 1.9 Ga and 100% PAL by the Neoproterozoic. Fossil evidence for these oxygen levels, however, is indirect. Some cyanobacteria form specialized cells called heterocysts, where nitrogen fixation occurs in an anaerobic environment. The evolution of heterocysts is thought to have occurred in response to increasing oxygen levels, but the only Precambrian evidence of heterocysts is indirect. Although heterocysts have been reported from the Gunflint Chert (discussed later), these structures could also be interpreted as the result of diagenesis. Golubic et al. (1995) proposed that the Mesoproterozoic rod-shaped microfossil Archaeoellipsoides represents the akinetes of a cyst-forming cyanobacterium and attribute this fossil to the Nostocales, a group of heterocystous cyanobacteria with living members (Chapter 3). Based on the monophyly of the extant heterocystous forms (Tomitani et al., 2006), Archaeoellipsoides is hypothesized to be the earliest evidence of heterocyst-forming cyanobacteria and used as a benchmark for paleo-oxygen levels (Tomitani et al., 2006). More recent evidence, however, suggests that atmospheric oxygen was present by 2.5 Ga, in the latest Archean. Both sulfur isotopes (Kaufman et al., 2007) and the metals molybdenum and rhenium (Anbar et al., 2007) from the Mount MacRae Shale in Western Australia show a shift in isotopic values at 2.5 Ga years. These authors interpret these shifts to show the presence of oxygen, which correlates with previous data from carbon isotopes. Kaufman et al. noted that equivalent strata in South Africa show the same shift in sulfur isotopes, suggesting that widespread oxygenation of the ocean was present around 2.5 Ga, prior to oxygenation of the atmosphere at 2.45 Ga. The time period for evolution of oxygenic photosynthesis is further constrained by the absence of evidence for oxygen in a 2.7 Ga paleosol (Yang et al., 2002). Brocks et al. (1999) reported on lipids characteristic of cyanobacterial photosynthesis in Neoarchean rocks (2.7–2.6 Ga) from the lowermost Hamersley Group



(Pilbara Craton). Additional data from carbon isotopes also support widespread oxygenic photosynthesis in the ocean before the land (Eigenbrode and Freeman, 2006).

PROTEROZOIC LIFE If the Archean was a prokaryotic world, the Proterozoic was a time of transition to a eukaryotic and eventually a multicellular world. Microfossils of eukaryotes appeared around the Paleoproterozoic–Mesoproterozoic boundary, and by the late Mesoproterozoic—early Neoproterozoic, the major algal clades were present. Eukaryotes (based on acritarch diversity) appear to have gradually increased in diversity until the midNeoproterozoic (700 Ma), the time of the first (Sturtian) Neoproterozoic glaciation. After the second glaciation (630 Ma), their diversity increased rapidly until the latest Neoproterozoic, when they underwent an extinction event. PALEOPROTEROZOIC

There is abundant evidence of diverse life forms in the Proterozoic. In their review of Paleoproterozoic microfossils, Hofmann and Schopf (1983) listed 122 taxa (including 23 dubiofossils and pseudofossils) in 40 different biotas (2.5–1.6 Ga); this number does not include microorganisms that were not named. They classify these genera into five morphological categories: (1) coccoid unicells; (2) septate, unbranched, filaments; (3) tubular, unbranched forms; (4) branched filaments; and (5) bizarre or unusual forms, that is, those with unusual morphologies and uncertain affinities. Coccoid forms dominate most Paleoproterozoic assemblages, both in terms of diversity and abundance. In contrast to later Proterozoic and early Paleozoic forms, they generally exhibit a simple and unornamented morphology and are relatively small (25 μm; most are 2–7 μm). Included in this classification would be coccoid forms that occur in colonies which are usually surrounded by a sheath-like structure. Septate filamentous forms are also usually small and simple, ranging from 1 to 2.5 μm in diameter and generally not surrounded by a sheath. Tubular microfossil forms are less commonly found and many appear to represent the remains of microbial sheaths, similar to those in modern filamentous cyanobacteria. Branched filaments and unusual forms are relatively rare in most biotas. Based on their size and simple organization, Hofmann and Schopf (1983) considered all of these forms as representing prokaryotes. Thus, it would appear that Paleoproterozoic biotas were composed for the most part, if not entirely, of primitive prokaryotic organisms.


Paleobotany: the biology and evolution of fossil plants

Figure 2.20 Stromatolites from the Gunflint Formation.

(Courtesy J. W. Schopf.)

Figure 2.22 Filaments and spheres from the Gunflint Form-

ation. Bar  10 μm. (Courtesy J. W. Schopf.)

Figure 2.21 Thin section from the Gunflint Formation showing stromatolitic laminae. (Courtesy J. W. Schopf.) Figure 2.23 Eosphaera tyleri (Gunflint Formation). Bar 

One of the most extensively studied Paleoproterozoic biotas comes from the Gunflint Formation of southern Ontario (1.9 Ga) and, as noted earlier, it includes the first well-preserved Precambrian organisms described (Tyler and Barghoorn, 1954; Barghoorn and Tyler, 1965). See also Cloud (1965), which helped to give credibility to this work at a time when most believed that the Precambrian was devoid of life. The organisms from this formation are structurally preserved in stromatolitic (FIGS. 2.20, 2.21) and non-stromatolitic cherts and can be divided into four basic morphologic types: coccoid forms, (Huroniospora) (FIG. 2.22); septate, filamentous forms (Gunflintia); tubular,

10 μm. (Courtesy J. W. Schopf.)

unbranched forms (Animikiea); and unusual or bizarre forms (Archaeorestis, Kakabekia, Eoastrion, Eosphaera) (FIG. 2.23). Huroniospora is a simple, spherical—ellipsoidal form originally described by Barghoorn, that has since been found at several localities with budlike outgrowths attached to the cells (Barghoorn and Tyler, 1965). Gunflintia is a narrow filament (1–4 μm in diameter) composed of a single row of cylindrical cells, whereas Animikiea is a broader (6–12 μm) tube that sometimes contains a septate filament.




Figure 2.25 Kakabekia umbellata (Gunflint Formation). Bar 

5 μm. (Courtesy J. W. Schopf.)

Figure 2.24 Archaeorestis sp. (Gunflint Formation). Bar 

10 μm. (Courtesy J. W. Schopf.)

Some of the most interesting forms in the Gunflint microbiota are those that have uncertain taxonomic affinities. Eoastrion is the name given to a star-shaped group of radiating, filamentous structures considered by most to represent some type of bacterium. Archaeorestis is a non-septate, irregularly branched organism with filaments that range from 2 to 10 μm in diameter and up to 200 μm long (FIG. 2.24). Its affinities are uncertain, but it has been interpreted as a budding bacterium (Awramik and Barghoorn, 1977) and a possible Kakabekia by Hofmann in Hofmann and Schopf (1983). Kakabekia umbellata exhibits a tripartite organization, consisting of a bulb-like base, a so-called stipe, and an umbrellalike crown (FIG. 2.25); some specimens extend up to 30 μm in length. The affinities of these fossils continue to remain obscure. Some of the structures in the Gunflint chert may represent taxa that are not related to any known forms, and it is difficult to assign these fossils to any group of organisms

with certainty. Although there is noticeable morphological similarity between some of the fossil organisms and certain modern bacteria, it would be almost impossible to categorize many of these fossil organisms without a knowledge of their biochemistry and molecular biology. Information on cell morphology alone is not enough to make classification possible. Overall, the types of organisms present in Paleoproterozoic biotas are similar in their morphology and occurrence to both older (Neoarchean) and younger (Mesoproterozoic) types. They differ from those of the Archean by their greater diversity and by the presence of both benthic and planktonic forms. What can be generalized is that there is a trend toward increasing diversity and increasing cell and filament size throughout the Proterozoic. ORIGIN OF EUKARYOTES

Almost 40 years after Schopf (1968) originally reported presumed green algae from the Bitter Springs Formation in the Amadeus Basin (Australia), it is difficult to appreciate the controversy this discovery caused at the time. Originally thought to be 1–0.9 Ga in age, the Bitter Springs is now considered to be 830–800 Ma and is discussed in more detail later (see Neoproterozoic). The deposit includes fossilized microbial mats with microfossils preserved in chert. The microfossils Caryosphaeroides and Glenobotrydion


Paleobotany: the biology and evolution of fossil plants

are spherical cells with opaque material inside the cell walls (Schopf, 1968; Schopf and Oehler, 1976), and it was suggested that these microfossils could be assigned to the Chlorophyta (green algae; Chapter 4). Cells of C. pristina average 13 μm in diameter and those of Glenobotrydion, 9 μm; by comparison, the filamentous and spherical bacterial cells in this same deposit are generally 10 μm in diameter. Both of these organisms contained opaque material, which was interpreted as the remains of nuclei in Caryosphaeroides and the remains of a pyrenoid-like body in Glenobotrydion (Oehler, 1977). Knoll and Barghoorn (1975), however, examined stale cultures of living cyanobacteria (Chroococcus) and concluded that all of the fossil intracellular objects could be explained as artifacts resulting from the coagulation of cytoplasm in a prokaryotic cell. Despite the fact that interpretation of these structures remains controversial, there is no doubt that the presence of intracellular material in Precambrian cells, whether cytoplasm or nuclei, represents a remarkable case of fossil preservation. Eotetrahedrion is another important (and controversial) fossil from the Bitter Springs biota; it consists of a tetrahedral tetrad of spherical cells surrounded by a sheath-like structure (FIG. 2.26) (Schopf and Blacic, 1971). Cells average 9.4 μm in diameter and many include a triradiate mark on the surface. What is especially interesting is that this type of mark is commonly seen on spores of land plants after meiosis has occurred (e.g., see Chapters 5, 6). Such spores are produced during sexual reproduction by the process of reduction division, in which a single cell (spore mother cell) with a diploid complement of chromosomes (referred to as 2n) divides to produce four identical products (spores), each with one half the chromosome complement of the mother cell (n). Many of these spores are produced in a tetrahedral tetrad (4 spores), and the triradiate or trilete mark on the spores denotes the contact face where each spore in the tetrad is in contact with the other three. The presence of a triradiate mark on spores is generally assumed to represent

evidence of meiosis. Thus, it was initially suggested that the marks on Eotetrahedrion indicated the occurrence of meiosis and, therefore, of sexual reproduction during the Proterozoic. Some green algae and cyanobacteria, however, are known to produce mitotically derived tetrahedral tetrads of cells or spores, so this arrangement of cells cannot necessarily be considered definitive evidence of meiosis or sexual reproduction in the fossil record (see Chapter 6). Work completed since Schopf’s (1968), and Schopf and Blacic’s (1971) research on the Bitter Springs chert has demonstrated that eukaryotes had definitely evolved by 830–800 Ma; in fact, they were present far earlier. Cloud (1976) suggested that the earliest eukaryotic cells occur in the 1.3 Ga Beck Spring Dolomite of California, based on the occurrence of filamentous forms resembling siphonaceous green algae and unicells that range from 40 to 62 μm in diameter (Cloud et al., 1969). Similar to the cells described by Schopf from the Bitter Springs microbiota, nearly all of these forms include a dark body within the cell. Because of the diversity of forms present in this biota, Cloud (1983) suggested that the eukaryotes may have arisen anywhere from 2 to 1.3 Ga. Further support for this assumption is the report by Schopf (1977) of an increase in cell size of both filaments and unicells starting at 1.4 Ga. More recently, several authors have suggested specific criteria for identifying early eukaryotes, some of which are similar to standards used to identify the earliest life. Cells presumed to be eukaryotes must have a morphology comparable to known organisms, and this morphology should have a definite range of variability. Knoll et al. (2006b) suggested three distinct criteria for recognizing early eukaryotes, but in most cases these will only apply to acritarchs. Acritarchs (Chapter 4) are generally unicellular microfossils, many of unknown affinity, although some have been attributed to planktonic, cyst-forming algae; the vast majority are assumed to represent eukaryotes (Martin, 1993). Many acritarchs have a complex ornamentation consisting of elongate processes on

Figure 2.26 Different focal planes of Eotetrahedrion sp. (Bitter Springs Formation). Bar  10 μm. (Courtesy J. W. Schopf.)


the surface of the cell wall (FIG. 2.27). Knoll et al. (2006b) proposed that to be definitively classified as eukaryotes, microfossils must (1) be large, (2) have a preservable wall, and (3) have processes (elaborations of the external wall). Prokaryotic cells, with a few notable exceptions, are generally 10 μm in diameter, whereas eukaryotic cells can range from 10 μm to hundreds of microns in diameter. Javaux et al. (2001, 2003) described well-preserved microfossils of Tappania and other acritarchs from coastal facies of the early Mesoproterozoic (1.5 Ga) Roper Formation of north central Australia. Various acritarchs are found in a range of rocks representing marginal marine to basinal settings; the diversity and abundance of eukaryotic fossils decreases from onshore to offshore. Tappania extends up to 160 μm in diameter and is characterized by hollow, cylindrical processes with expanded tips (FIG. 3.9). Z. Zhang (1997) described sphaeromorphic acritarchs from the Changzhougou Formation in northern China (1.85 Ga). Although somewhat poorly preserved, the assemblage contained a large number of smooth-walled acritarchs whose diameters clustered around 60 μm and extended up to 238 μm. From the Meso-Neoproterozoic Ruyang Group (1150–950 Ma), L. Yin (1997) described two acanthomorphic acritarchs, Tappania and Shuiyousphaeridium. The latter is spherical, from 110 to 250 μm in diameter and bears elongate processes on the exterior. Tappania is vase shaped and small, 45–60 μm long. Meng et al. (2005) suggested that Shuiyousphaeridium represents an early dinoflagellate, and support this conclusion

Figure 2.27 Acritarch (Trachyhystrichospahaea) from Siberia

Russia (Lakhanda Formation). Bar  50 μm. (Courtesy J. W. Schopf.)



with evidence of the biomarker, dinosterane, in the same rocks as the fossils in the Mesoproterozoic Beidajian Formation (see also Chapter 4). Other evidence, however, suggests that eukaryotes may have arisen much earlier. Brocks et al. (1999) found steranes in the same Neoarchean rocks (2.7–2.6 Ga) from which they extracted cyanobacterial biomarkers as evidence of oxygenic photosynthesis. Steranes are only produced by eukaryotes, but Summons et al. (2006) have shown that previous reports of steranes in cyanobacteria can be attributed to eukaryote contamination of cultures. In addition, sterane biosynthesis requires oxygen, so this research suggests that both eukaryotes and oxygenic photosynthesis were present 2.6 Ga (Brocks et al., 1999). To date, no generally accepted microfossil evidence has been found to corroborate these geochemical hypotheses. Several authors have suggested that the spiraled compression fossil, Grypania, may represent the earliest eukaryote. Han and Runnegar (1992) described ribbon-shaped, spiraled carbonaceous films from the Negaunee Iron Formation (2.1 Ga) as Grypania, and noted that the thallus could reach 0.5 m in length, although it was only 0.5 mm in diameter. Samuelsson and Butterfield (2001) suggested that the Grypania specimens from the Negaunee Iron Formation are unlike those from younger rocks (which are clearly eukaryotic) and should not be considered the same organism. A number of similar carbonaceous films have been described from Proterozoic rocks, and many of these have been shown to be pseudofossils (Lamb et al., 2007). Possible evidence for the origin of eukaryotic organisms around 1 Ga comes from reports of similar megascopic fossils from various localities around the world (Walter et al., 1976; Hofmann, 1985b). It has been suggested that some of these megascopic fossils represent eukaryotic algae. Chuaria is one of the most widespread genera of this type and has been described from the Little Dal Group in northwest Canada (1080–780 Ma) as black, circular compressions ranging in size from microscopic up to 4.6 mm in diameter (Hofmann and Aitken, 1979). Concentric wrinkles on the surface are present on many specimens. Chuaria occurs worldwide and was originally thought to be restricted to rocks 1.1–0.6 Ga, where it was used as an index fossil. In the Little Dal Group, Chuaria occurs with ribbonlike, often curved, sometimes tubular compressions several centimeters long that are given the name Tawuia. This biota also includes filamentous bacteria, cyanobacterial sheaths, and possible acritarchs. Chuaria was originally interpreted as a planktonic eukaryotic alga and Tawuia as a probable alga, although metazoan affinities could not be completely discounted for


Paleobotany: the biology and evolution of fossil plants

the latter. Better-preserved specimens of both taxa from northern China (Sun, 1987) suggest that at least some of these compressions may represent prokaryotic aggregations. These fossils were studied by means of cellulose acetate peels and bioplastic transfers of the rock surface, techniques which revealed that both taxa represent colonies of filamentous cyanobacteria; Sun compares them to the living genus Nostoc. This interpretation helps to explain the variable size and morphology of these unusual remains. Other authors, however, regard Chuaria and Tawuia as developmental stages of the same organism, which is thought to be some type of benthic, tubular macroalgae (Xiao and Dong, 2006). Similar fossils from Paleoproterozoic rocks were described as seaweed-like algae (W. Zhu and Chen, 1995; Yan and Liu, 1997), but these specimens are now considered to have uncertain affinities or represent aggregations of prokaryotes (Butterfield, 2000; Knoll et al., 2006b). Grazhdankin and Gerdes (2007) examined discoidal compressions made up of concentric rings. These had previously been identified as metazoans, but these authors compare them to ring structures in modern microbial mats and conclude that they are formed by either bacteria, fungi, or protists. The earliest fossil evidence for eukaryotic life occurs in the Paleoproterozoic in the form of acritarchs. Although there is geochemical evidence in the latest Archean, there is yet no widely accepted fossil evidence to support this finding. There is excellent evidence, however, that eukaryotes diversified into multicellular forms in the Mesoproterozoic. MESOPROTEROZOIC

During the Mesoproterozoic (1.6–1 Ga), the supercontinent Rodinia was formed (around 1.1 Ga), atmospheric oxygen levels continued to rise, and stromatolites formed large and widespread reefs. Unicellular, eukaryotic organisms probably arose either in the late Paleoproterozoic or early Mesoproterozoic, and continued to diversify throughout this time period. Coccoid microfossils show an increase in size during the Mesoproterozoic and biotas exhibit an increasingly diverse assemblage of cyanobacterial remains, both filamentous and coccoid. As noted earlier, cyst-forming, planktonic, eukaryotic algae (acritarchs) are believed to have emerged in the Mesoproterozoic, 1.4 Ga ago (Knoll, 1985a), with the earliest evidence for multicellularity not long thereafter. Oehler (1977) described a deep-water biota from the Mesoproterozoic of Australia. The assemblage was dominated by filamentous bacteria, unlike those known from shallow-water deposits or stromatolitic assemblages. Oehler suggested that these organisms lived below the photic zone under anoxic conditions.

EARLIEST MULTICELLULAR LIFE Many textbooks, journal papers, and online sites identify the origin of multicellular life as the famous 610 Ma Ediacaran biota—an assemblage of strange, soft-bodied metazoans that are preserved in only a limited number of sites around the world. These reports tend to forget the fact that irrefutable evidence for multicellular algae existed long before the Ediacaran. In 1990, Butterfield et al. described a permineralized red alga similar to living members of the Bangiophyceae (Rhodophyta) from the Hunting Formation of arctic Canada. This formation consists of shallow-water carbonates and includes stromatolites, as well as cyanobacterial fossils; the rocks that contain the algae are dated at 1200 Ma (late Mesoproterozoic) (Butterfield, 2000). The fossils of Bangiomorpha pubescens (FIGS. 4.43, 4.44) consist of filaments made up of stacked, discoidal cells enclosed within a sheath; many have a holdfast at their base. The filaments, which are up to 2 mm long and range from 15 to 45 μm in diameter, show radial cell division within the filament, a feature characteristic of bangiophytes. As Butterfield (2000) noted in the formal description of B. pubescens, the fossil is so similar to the modern red alga Bangia as to be “indistinguishable.” Based on the different morphology of filaments and circular objects within them, Butterfield suggests that B. pubescens fossils also show the earliest fossil evidence of sexual reproduction. The shallow-water carbonates that include these red algae contain a total of four fossil assemblages, three of which are dominated by mat-forming, photosynthetic prokaryotes (Butterfield, 2001). Each assemblage appears to occupy a specific niche within a larger ecosystem. In the Bangiomorpha assemblage, however, matforming prokaryotes are excluded, perhaps because of the vertical growth of the algae. Butterfield (2001) suggested that this biota documents the oldest case of competitive exclusion of prokaryotes by multicellular eukaryotes. He hypothesized that such exclusion may explain the relatively rapid, even explosive, diversification of eukaryotes in the Neoproterozoic. NEOPROTEROZOIC

A greater diversity of eukaryotic algae begins to appear around the Mesoproterozoic—Neoproterozoic boundary (1 Ga), and the divergence of the major clades of eukaryotes is also believed to have occurred about this time (Knoll et al., 2006b). There is an increasing diversity of ornamented acritarchs (Butterfield, 2000). Although the diversity of taxa and number of clades increase, it is not until the middle Neoproterozoic that the primary radiation of eukaryotes occurred (Porter, 2004), so that many early Neoproterozoic biotas still contain diverse and abundant



Figure 2.28 Cephalophytarion grande (Bitter Formation). Bar  10 μm. (Courtesy J. W. Schopf.)



cyanobacteria, as well as other bacteria. Porter (2004) suggested that eukaryotic diversification occurred in response to selective pressures, which included the appearance of microbial predators, for example, testate amoebae (Porter and Knoll, 2000), and changes in seawater chemistry. Certainly, there were extensive environmental changes during the Neoproterozoic. The supercontinent Rodinia, which formed in the Mesoproterozoic, broke up in the early–middle Neoproterozoic, and two major glaciations, the Sturtian (750–700 Ma) and the Marinoan (635–624 Ma; the so-called snowball Earth glaciation), occurred in the Neoproterozoic (Bodiseltisch et al., 2005). Canfield (1998) proposed that the deep ocean did not become oxygenated until 1–0.54 Ga, based on sulfur isotopes. However, Butterfield (2004) suggested that there has been taxonomic inflation of eukaryotic taxa in the Proterozoic and that eukaryotes evolved at a much slower rate in the Proterozoic than they did in the Phanerozoic. A number of Neoproterozoic fossils can definitely be attributed to eukaryotic clades. Butterfield (2004) described fossil filaments of Jacutianema and attributed them to the Vaucheriales, an order within the Xanthophyceae (Chapter 4), or yellow-green algae. The fossils come from the middle Neoproterozoic, Svanbergfjellet Formation of Spitsbergen and are attributed to the Vaucheriales based on unique constrictions within the filaments. This discovery gives additional credence to an earlier report of a vaucheriacean, Palaeovaucheria, from the 1 Ga Lakhanda Formation of Siberia (Hermann, 1981). Butterfield and Rainbird (1998) described a diverse acritarch biota from the 1077–723 Ma Wynniatt Formation, arctic Canada. Three of the taxa resemble dinoflagellate cysts and may represent an early microfossil record of this group.

2.29 Oscillatoria amena (Extant). Bar  10 μm. (Courtesy J. W. Schopf.)


Figure 2.30 Filiconstrictosus cf. extant Oscillatoria amena (Bitter Springs Formation). Bar  10 μm. (Courtesy J. W. Schopf.)

Figure 2.31 Oscillatoriopsis sp. (Bitter Springs Formation).

BITTER SPRINGS BIOTA One of the better-known Proterozoic biotas comes from the Bitter Springs Formation of central Australia (Schopf, 1968; Schopf and Blacic, 1971; Oehler, 1976). This diverse biota (830–800 Ma; A. Hill, 2005) includes an abundance of filaments (FIGS. 2.28, 2.30, 2.31) and spherical unicells. Some filaments have been given names such as Palaeolyngbya, Oscillatoriopsis (FIG. 2.31), and Palaeoanacystis, and bear a striking morphologic resemblance to various extant cyanobacteria (FIG. 2.29). Palaeolyngbya is a filament with evenly spaced cross walls that are surrounded by a sheath up to 1 μm thick. Cells toward the center of the filament are usually rectangular, 2.6–3.1 μm  8.3–10.9 μm in diameter. The terminal cell of the filament is characteristically rounded. Morphologically, Palaeolyngbya is closely comparable to modern Lyngbya.

Bar  10 μm. (Courtesy J. W. Schopf.)

Some filaments show a narrowing of cells at the extremities of the filaments so that their tips appear pointed. Most of the filaments are 5 μm in diameter, although some reach up to 15 μm. Present-day cyanobacteria may be just as slender, but a proportionately larger number are more robust. The diversity of the Bitter Springs biota is illustrated by the fact that five families of extant cyanobacteria representing two major orders (Chroococcales and Nostocales) can be identified (Schopf and Blacic, 1971). Some of the cyanobacteria in the Bitter Springs Formation were compared with modern, cyst-forming filamentous forms (e.g., in the Nostocales) and these similarities were used to suggest that heterocysts had evolved by this time (Schopf


Paleobotany: the biology and evolution of fossil plants

Figure 2.32 Eomycetopsis robusta (Bitter Springs Formation).

Bar  10 μm. (Courtesy J. W. Schopf.)

and Blacic, 1971). As noted earlier, heterocysts are the site of nitrogen fixation in extant cyanobacteria. Since N2 fixation is inhibited by oxygen, the cyanobacterium maintains an anaerobic environment within each heterocyst; the evolution of heterocysts is presumed to indicate high oxygen levels, possibly comparable to modern levels (Tomitani et al., 2006). As originally described, the Bitter Springs biota included numerous species of cyanobacteria, both filamentous and unicellular, several eubacteria, fungus-like filaments, possible dinoflagellates (Dinophyta), and spheroidal green algae. The filaments that were originally described as fungi, Eomycetopsis (FIG. 2.32), have since been reinterpreted as cyanobacterial sheaths. STROMATOLITES Stromatolites occur beginning in the Paleoarchean, but are particularly widespread and diverse in the Proterozoic. One interesting stromatolite that appears to have a living counterpart growing in hot springs environments is the genus Conophyton. This stromatolite differs from others in that it has acute conical laminations and a distinct axial zone. Columns of Conophyton range from 5 to 10 mm in diameter. In one form from the Chichkan Formation of southern Kazakstan (650 Ma), both cyanobacteria (FIGS. 2.33, 2.34) and eukaryotic algae were identified in the silicified lamellae (Schopf and Sovietov, 1976). It has long been suggested that stromatolites reached their peak in diversity and abundance around 1250 Ma, after which they began to experience a decline (Walter and Heys, 1985). They never again reach the levels of diversity

Figure 2.33 Paleopleurocapsa reniforma (Chichkan Forma-

tion). Bar  10 μm. (Courtesy J. W. Schopf.)

Figure 2.34 Myxococcoides inornata (Chichkan Formation). Bar  10 μm. (Courtesy J. W. Schopf.)

or abundance that they exhibited during the Proterozoic, although there are reefs found in every period of the Phanerozoic which are dominated by stromatolites (Pratt, 1982). Garrett (1970) and Awramik (1971) suggested that this decline could be correlated with the rise of metazoans that grazed on or burrowed into the stromatolites. This theory would seem to be supported by the restricted occurrence of living stromatolites, either in areas of high salinity (e.g., Shark Bay, Australia) or in zones with high sediment influx (e.g., Highborne Cay, Bahamas; Andres and Reid, 2006). There is no fossil evidence, however, for metazoan grazing or damage on Proterozoic stromatolites. Awramik and Sprinkle (1999) examined changes in stromatolite taxonomic diversity for 1187 forms and found that stromatolites reached their peak 1350–1000 Ma,




Proterozoic peak Late Proterozoic decline

Cambrian– Early Ordovician resurgence % 100

Archaean– Proterozoic increase 200


Reefal microbial carbonates

Stromatolite taxa per 50 MYR interval


Ordovician– present-day episodic decline


0 3000







Age in millions of years

Figure 2.35 Distribution of stromatolite taxa through time. (From Riding, 2006a, b.)

after which they declined. Riding (2006a) combined stromatolite diversity data (FIG. 2.35) (Awramik and Sprinkle, 1999) with data on the abundance of reefal microbial carbonates (Kiessling, 2002) and metazoan diversity (Sepkoski, 1997). He found that stromatolites increased in abundance until 2250 Ma, remained at that level until 1450 Ma, and then markedly increased in abundance until 1350 Ma. After 1100 Ma, they began an irregular decline until 550 Ma, the latest Ediacaran. Stromatolites and other microbial carbonates showed an increase in the Cambrian and Early Ordovician (Riding, 2006a, b), and there is little to no correlation with marine metazoan diversity during the Phanerozoic, suggesting that metazoan grazing was not the primary reason for Proterozoic stromatolite decline. If the rise of grazing metazoans is not the primary cause of stromatolite decline, what is? Competitive exclusion by eukaryotic algae has also been implicated. Butterfield (2001) suggested competitive exclusion to explain the monotypic Bangiomorpha community in the Mesoproterozoic Hunting Formation (see above). Andres and Reid (2006) studied modern stromatolites in Highborne Cay, the Bahamas, and found that stromatolites persist in areas where sediment influx is high, because metazoans and macroalgae are excluded. If sedimentation patterns change, both metazoan borers and macroalgae colonize the substrate, outcompeting and eventually excluding the microbial stromatolites. Climate change in the Proterozoic, including two major glaciations, is also hypothesized to have contributed to stromatolite

decline. Riding’s (2006a) analysis showed that the decline itself is more complex than previously thought and cannot be explained by metazoan grazing alone. Further fossil studies are clearly needed, especially on substrates and modes of growth of stromatolites, if we are to fully understand this important Proterozoic event. OTHER MICROFOSSILS Like stromatolites, Neoproterozoic non-stromatolitic microfossils are diverse and are found in varied habitats, ranging from supratidal to open-shelf environments (Knoll, 1985a). Cyanobacteria are particularly diverse; for example, 11 of the 13 genera in the family Chroococcaceae (FIG. 2.36) are found as Proterozoic fossils. Most cyanobacteria occur in stromatolites, shallow marine muds, or tidal pools. The Entophysalidaceae are represented by mat-building organisms in intertidal zones. Pleurocapsales are an important part of tidal flat communities. Filamentous cyanobacteria are well represented as mat builders in a variety of environments. Knoll noted that Proterozoic cyanobacteria are very similar in their morphology and life-history patterns to those found in similar environments today, and that these patterns were established early. Throughout the Meso- and Neoproterozoic (2.0–1.4 Ga), intertidal zone, mat-building assemblages are dominated by Eoentophysalis-type microfossils (FIGS. 2.37, 2.38). Knoll concluded that, by the middle of the Neoproterozoic, cyanobacteria were essentially modern in their morphology and, by inference, in their physiology. In fact, most of the major evolutionary


Paleobotany: the biology and evolution of fossil plants

Figure 2.36 Sphaerophycus medium (Draken Formation).

Bar  50 μm. (Courtesy A. H. Knoll.)

Figure 2.37 Eoentophysalis sp. (Sukhaya Tunguska Form-

ation). Bar  10 μm. (Courtesy J. W. Schopf.)

Figure 2.38 Eoentophysalis belcherensis (Kasegalik Formation). Bar  10 μm. (Courtesy J. W. Schopf.)

changes in the cyanobacteria appear to have been established by the Neoarchean–Paleoproterozoic (Knoll, 1985b). The same cannot be said for the eukaryotes, however. Eukaryotic diversity, as measured by morphotaxa of acritarchs, especially large complex acritarchs (FIGS. 2.39, 4.75) underwent a number of changes in the Neoproterozoic. These planktonic microfossils show a gradually increasing morphologic diversity beginning in the Mesoproterozoic and extending into the mid-Neoproterozoic (1200–700 Ma) (Knoll, 1994; Porter, 2004) (FIG. 2.39), followed by a decline around the time of the Sturtian glaciation (750– 700 Ma). Toward the end of the Marinoan glaciation (625 Ma), acritarchs exhibit a rapid increase in diversity (Zang and Walter, 1989) followed by a major extinction in the latest Neoproterozoic (Vidal and Knoll, 1982; Porter, 2004). Acritarchs again underwent a diversification event in the Early Cambrian concomitant with the radiation of shelly animals. A very diverse assemblage of acritarchs in the Pertatataka Formation of central Australia (650–600 Ma) includes about 25 genera of complex acritarchs, much larger than those found in younger rocks (Zang and Walter, 1989). The lower part of the formation contains simple spheroidal acritarchs, whereas the upper part contains as many as 40 taxa of acritarchs, including many spheroidal forms, but also large, ornamented types. The most abundant taxon in the upper assemblage is Cymatiosphaeroides, which is 180 μm in diameter and characterized by numerous thin spines on the main vesicle. Zang and Walter compared these forms to younger microfossils, but noted that this assemblage is unique. Up to half of the assemblage (10–15 genera) occur only in this formation; a third represent taxa previously known from other Proterozoic assemblages, and only a few taxa are closely related to Paleozoic forms (Knoll and Butterfield, 1989). Perhaps most interesting is that these forms with Paleozoic affinities are consistently two to three times larger than their younger relatives. Zang and Walter suggested that this type of assemblage had not previously been recognized in the Ediacaran because it represents an environmental setting (offshore marine) that is poorly represented in rocks of this age. Extinction of these forms occurred in the latest Neoproterozoic. Whether or not eukaryotic diversity and extinction are related to the Neoproterozoic glaciations is still uncertain. The original idea of a “snowball” Earth included global glaciation and massive extinctions, but the microfossil record does not support this hypothesis (Corsetti et al., 2006). Corsetti et al. (2003) reported a diverse microbial biota containing two types of stromatolites, prokaryotes, and both


Late Paleoproterozoic Early Mesoproterozoic (ca 1800–1300 Myr)

MidLate NeoprotMesoproterozoic Early erozoic Neoproterozoic (ca 850– (ca 1300–850 Myr) 720 Myr)



Ediacaran Early (632–ca 560 Myr)

Late (560– 542 Myr)

Macroscopic Microscopic multicellular


Vase-shaped microfossils 60

 Acritarchs, symmetrical processes

Total taxa per assemblage

Acritarchs, asymmetrical processes 50

Ornamented acritarchs Unornamented acritarchs

40 Global glaciation 30




Figure 2.39 Composition and taxonomic richness of non-metazoan eukaryotic morphospecies for selected Proterozoic to Early

Cambrian fossil assemblages. (Modified from Knoll et al., 2006.)

heterotrophic and autotrophic eukaryotes preserved in chert and carbonate from the Kingston Peak Formation in the western US. The biota occurs in Sturtian glacial deposits but is comparable to the biota in underlying preglacial rocks. This is a particularly interesting deposit, in that the diversity within the biota does not suggest merely a few “disaster” taxa but rather a complex community with several trophic levels. Z. Zhou et al. (2007) examined the stratigraphic distribution of the acritarch biotas that are typical of both the Pertatataka Formation in Australia and the Doushantuo Formation in South China. Specifically, using chemostratigraphy and biostratigraphy, they detailed first appearances of this assemblage with regard to Neoproterozoic glaciations. Their research suggests that this biota appeared immediately after glaciation in the East Yangtze Gorges area; its appearance at Weng’an, however, cannot be correlated with a glaciation event. Interestingly, acritarch biozonation in Australia places the first appearance of acanthomorphic species immediately

after the Acraman impact (580 Ma), and this has led some to suggest that this event may have triggered subsequent eukaryotic diversification (Grey et al., 2003; Willman et al., 2006). It appears that more data are needed from both paleobotany and geochronology before the evolution of these acritarch biotas can be confidently correlated with glaciations or other environmental perturbations. The fossil record from the Neoproterozoic is the most complete in the Precambrian, and correlations between particular biotas and their paleoenvironments can be discerned with much higher resolution (Knoll et al., 1989). Planktonic microfossils show an inshore—offshore pattern, as they do today, with biotas from inshore deposits typically exhibiting low diversity and dominance by one or two taxa. Open-shelf assemblages, however, are much more diverse and the individuals are more complex (Vidal and Knoll, 1983). Three different assemblages of microorganisms were described from the Ryssö Formation of Svalbard, Spitsbergen (Knoll and Calder, 1983). One assemblage consists of typical


Paleobotany: the biology and evolution of fossil plants

stromatolitic microorganisms. Another represents an open coastal environment and is dominated by planktonic forms such as acritarchs, and the third contains a large number of vase-shaped protists. The occurrence of these three microbiotas within a single formation provides a unique opportunity to evaluate organisms from three very different habitats and to develop a more complete picture of life from 800 to 700 Ma. Although much of our information on Precambrian organisms has come from structurally preserved fossils in chert deposits, more and more data are available on compressed organisms from other paleoenvironments than those represented by chert biotas. Knoll and Swett (1985) described a series of biotas from Spitsbergen which include unicellular and filamentous prokaryotes preserved in shales. They are able to compare these organisms to those found in the Bitter Springs chert and to slightly younger, silicified biotas from Svalbard. Butterfield et al. (1988) described a well-preserved fossil biota recovered from subtidal marine shales of the Neoproterozoic of Spitsbergen (800–700 Ma). This biota includes sphaeromorphic acritarchs, eukaryotic multicellular algae comparable to extant Ulvophyceae, cyanobacterial sheaths, rod-shaped and filamentous forms similar to heterotrophic bacteria, and what appear to be germinating zoospores of filamentous protists. Other forms that were present in this biota have been interpreted as allochthonous (i.e., not preserved in situ). These include Chuaria, Tawuia, and a number of morphologically complex acritarchs. Some of the more interesting forms include large vesicles (150– 250 μm in diameter) with long, terminally flared processes. These structures have been interpreted as encystment structures. This interpretation is supported by their occasional occurrence within even larger forms (300 μm in diameter). DOUSHANTUO FORMATION Our knowledge of Neoproterozoic life and the evolution of multicellular life has been greatly enriched by fossils from the Doushantuo Formation (590–555 Ma) of south central China. The formation consists of a series of finely laminated carbonaceous and siliceous shales that were deposited in a quiet, subtidal environment, probably in a restricted basin (Xiao et al., 2002). Fossil animals, algae, and cyanobacteria occur in cherts, phosphorites, and black shales (Xiao and Knoll, 1999). In a reexamination of the compressed Miaohe biota, which includes compressed carbonaceous macrofossils, from the uppermost black shales in the Doushantuo,

Xiao et al. (2002) proposed that there are only 20 taxa present, although more than 100 have been described. Of these, eight taxa are definitely eukaryotic algae, nine represent possible algae, and two taxa are cyanobacteria. The biota includes coenocytic green algae and thalli comparable to members of the Rhodophyta (red algae) and Phaeophyceae (brown algae; Chapter 4). This biota occurs immediately after the Marinoan glaciation and just prior to the appearance of the metazoan Ediacaran fauna, but there appear to be no metazoans in this compressed biota. Algae, including acritarchs, are also preserved as three-dimensional phosphatized remains in the Doushantuo. Xiao et al. (2002, 2004) reported on exceptionally well-preserved, multicellular and pseudoparenchymatous red algae, as well as cell division in algal cells. These, along with other Neoproterozoic algae, are described in more detail in Chapter 4.

CONCLUSIONS In the past, paleobiologists viewed the Precambrian as a long interval of time in which very little biological evolution occurred. Now however, now we recognize that this interval of geologic time was one of the most exciting in terms of biological change and innovation. The complete Precambrian record provides some of the best evidence for evolution through geologic time, beginning with the simplest, unicellular prokaryotic microorganisms and progressing through the earliest, unicellular eukaryotes into more complex colonial types, and finally, multicellular forms. Several authors have suggested that all the important biochemical and cellular evolution took place in the Precambrian and the various Phanerozoic life simply represents the morphological and physiological elaboration of forms that evolved in the Precambrian. The volume and importance of the work done in this area in the past 15–20 years is extraordinary. Perhaps nowhere else in the geologic record has such a diversity of evidence been applied to answer evolutionary questions, including biomarkers, geochronology, isotope chemistry, molecular phylogenetics of extant organisms, and microfossils. At this point in time, we are beginning to gain a clearer picture of the most ancient life—not just the types of organisms present, but also the environments in which they lived, their nutritional modes, and their interactions with each other and the abiotic world. As research on Precambrian life continues, one thing stands out, as it does in our modern world—the ubiquity and versatility of life on Earth.

3 FUNGI, BACTERIA, AND LICHENS Fungi ................................................................................................. 71

Fungal Spores....................................................................................111

Earliest Fossil Fungi............................................................................73

Fungal-like Organisms ......................................................................112

Systematics of Fungi ...........................................................................77

Eubacteria and archaea ................................................ 112

Fungal Life-History Strategies............................................................98 Archaea .............................................................................................113

Fungi–Animal Interactions ...............................................................105

Eubacteria .........................................................................................113

Geologic Activities of Fungi .............................................................107

Lichens ...........................................................................................117

Epiphyllous Fungi .............................................................................108

Rain, and then the cool pursed lips of the wind draw them out of the ground Mary Oliver, Mushrooms


Higher taxa in this chapter:

Kingdom Fungi Chytridiomycota Zygomycota Zygomycetes Trichomycetes Glomeromycota Ascomycota Pezizomycotina Basidiomycota Agaricomycotina Fungal-like organisms Peronosporomycetes (Oomycota) Kingdom Archaea Kingdom Bacteria Cyanobacteria Lichens

Fungi are primarily terrestrial, achlorophyllous, eukaryotic organisms that were at one time grouped with plants. Today, however, fungi are regarded as a monophyletic group more closely related to animals than plants and they occupy their own Kingdom (FIG. 3.1). Plesiomorphies shared with animals include the presence of chitin, food stored as glycogen, and, in the mitochondrial RNA, the bases uracil–guanine– adenine (UGA) code for the amino acid tryptophan (plants use UGG to code for tryptophan). The multicellular fungal body consists of filaments, called hyphae (sing. hypha), which together make up the mycelium or vegetative body of the fungus. From this simple organization, however, fungi can form many types of complex sexual and asexual reproductive structures, and they have developed adaptations that allow them to live in every habitat on Earth. Approximately 100,000 species of fungi have been described, but estimates


Paleobotany: the biology and evolution of fossil plants

Laboulbeniomycetes Lecanoromycetes Eurotiomycetes Sordariomycetes (Xylariales, Meliolales)


Leotiomycetes Dothideomycetes Pezizomycetes







Figure 3.1 Relationships of major fungal groups. (Modified from J. Taylor and Berbee, 2006.)

are that more than 1.5 million species are yet to be discovered (Hawksworth et al., 1995). Other organisms that were historically included with the fungi, but are now placed in the Kingdom Chromista, include the Oomycetes (water molds), Hyphochytridiomycetes, and Labyrinthulomycetes. Of these three clades, only the water molds will be discussed here. Slime molds were also once considered to be Fungi, but are now included in the Kingdom Protista. The fossil record of slime molds (FIGS. 3.2–3.4), once called the myxomycetes, is meager, restricted only to several found in Eocene amber (Dörfelt and Schmidt, 2006), so they will not be covered in this book. Together with a few other groups of heterotrophic organisms (e.g., bacteria), the fungi are the principal decomposers in the biosphere, releasing CO2 into the atmosphere and nitrogenous compounds into the soil. They also function as bioweathering agents and transformers of minerals and rocks (Burford et al., 2003). Fungi are involved in numerous types of associations with other organisms, ranging from those that produce diseases in plants and animals to a variety of beneficial, symbiotic relationships (e.g., mycorrhizae in the roots of most vascular plants) (Newsharn et al., 1995). Despite the fact that fungi, as a

group, have a long and interesting geologic history (Tiffney and Barghoorn, 1974), only relatively recently have they been studied in any detail. As a result, their importance in the evolution of past biotas and our knowledge of the evolutionary history of the major fungal groups through time have been minimized. There are several reasons why our understanding of fossil fungi has lagged behind our knowledge of many other fossil organisms. One of these is the long-held belief that fungi are too fragile to be adequately preserved in the fossil record (FIG. 3.5). Additionally, the study of fossil fungi may have been avoided due to difficulties in recognizing and interpreting them (FIGS. 3.6, 3.7). This is especially true in situations where fungi have been found within permineralized vascular plants (FIG. 3.8), which typically represent the principal focus of the research (LePage et al., 1994). In addition, when collecting fossil plants, paleobotanists may introduce their own inherent bias by focusing on collecting the best specimens of a taxon (e.g., those that provide the most informative characters of the organism). As a result, specimens that might show the effects of fungal activities, such as degraded tissue or necrotic areas, are simply not brought back to the laboratory for study. Some of this bias has been overcome by the modern use of

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.3 Arcyria sulcata (slime molds), capillitium com-

posed of capilitial threads (see FIG. 3.2) (Eocene). Bar  300 μm. (Courtesy A. Schmidt.)

Figure 3.2 Arcyria sulcata (slime molds), sporocarp composed

of a stalk and cupuliform base of the pteridium (cup) (Eocene). Bar  100 μm. (Courtesy A. Schmidt.)

quantitative field techniques, in which all available material within a certain area is examined or collected. Finally, the environment of deposition and fossilization may affect preservation. For example, Carboniferous swamp plants often occur in peat deposits, and the chemistry in the swamp may have discouraged extensive fungal activity. Nevertheless, despite all of these limitations, the literature on fossil fungi and their biotic and abiotic interactions is rapidly increasing. As might be expected, some of the details, especially those that characterize levels of fungal interaction in ecosystems (Taylor, 1990; Taylor et al., 2004) are more difficult to determine from fossils. Earliest Fossil Fungi

Although fungi were probably present during the Precambrian, the earliest fossil record of them has been difficult to interpret (Brunel et al., 1984). There are spore-like bodies and filaments in Precambrian rocks that may represent the remains of fungi, but the affinities of some of them remain equivocal. For example, although initially described as

Protophysarum balticum showing three sporocarps emerging from a plant fragment (Eocene). Bar  100 μm. (Courtesy A. Schmidt.)

Figure 3.4

septate fungal filaments, Eomycetopsis robusta (Schopf, 1968), (FIG. 2.32) from the Neoproterozoic Bitter Springs Formation (830–800 Ma) of Australia is now regarded as a cyanobacterial sheath. Tappania (FIG. 3.9) is a Neoproterozoic acritarch (1.43 Ga) from marine carbonates and shales that has been suggested to be fungal (Butterfield, 2005). Extending from the spherical main body of this organism is what are interpreted as septate hyphae, some


Paleobotany: the biology and evolution of fossil plants

Figure 3.5 Zoosporangium with discharge opening found in

phloem cells of a Carboniferous fern. Bar  20 μm.

Figure 3.7 Chlamydsospore with chytrids on the surface (Devo-

nian). Bar  300 μm.

Figure 3.6 Tracheid from Sphenophyllum containing fungi (Pennsylvanian). Bar  500 μm.

of which anastomose. Other organisms have been described from late Mesoproterozoic rocks as possibly fungal in origin (Hermann, 1979). Convincing evidence that any of these are some type of reproductive organ from a marine fungus, however, has not yet been forthcoming. Several structures regarded as fungal have been reported from the Lakhanda Series of Siberia dated at 1 Ga (Hermann and Podkovyrov, 2006). Although the specimens are hypothesized to be possible zygomycetes, their preservation on organic sapropelic films makes assignment difficult. An even earlier putative fungus from Kola Peninsula (northwestern Russia) is dated at 2 Ga (Belova and Akhmedov, 2006). Petsamomyces varies in morphology and bears what are termed hyphal-like

Figure 3.8 Cortical cell filled with fungal hyphae (Pennsylvanian).

Bar  10 μm.

appendages. Like many of these organic-walled microfossils, the affinities of these structures remain problematic. Despite these uncertainties, the presence of fungi in the Proterozoic has been indirectly inferred on the basis of divergence times using molecular clock assumptions, but see Taylor and Berbee (2006) on problems associated with many of

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.9 Tappania (Neoproterozoic). Bar  100 μm. (From Butterfield, 2005.)

Branching hypha (Silurian). Bar  25 μm. (From Sherwood-Pike and Gray, 1985.)

Figure 3.10

these estimates. Based on these calibrations it is suggested that fungi may have diverged from metazoans about 1.5 Ga (Hedges et al., 2004; Taylor and Berbee, 2006). If these estimates are even close to being accurate, then it would be expected that more fossil fungal remains will be described from these ancient sediments in the years ahead. Some of the oldest fossil remains that are convincingly fungal in origin occur in Early Silurian (Llandoverian) rocks of Virginia, USA (Pratt et al., 1978). The bulk maceration of these terrestrial rocks produced small (6 μm wide), septate, and branched filaments (FIG. 3.10). Some filaments had specialized cells morphologically identical to those of fungi that produce endogenously formed chains of conidia (FIG. 3.11). To date, the single most important source of ancient fungi is the Early Devonian (Pragian—earlist Emsian) Rhynie chert from Aberdeenshire, Scotland. This Lagerstätte represents an entire ecosystem that is petrified in silica—plants, animals, and microbes are all present. Numerous spores (FIG. 3.12), hyphal filaments (FIG. 3.13), and sporocarps have been described from the silicified matrix and from tissues of the land plants Asteroxylon mackiei, Rhynia gwynne-vaughanii, Aglaophyton, Nothia, and Horneophyton. Rhynie chert fungi, such as Palaeomyces (FIG. 3.14) (Kidston and Lang, 1921a),

Septate conidium (Silurian). Bar  15 μm. (From Sherwood-Pike and Gray, 1985.)

Figure 3.11


Paleobotany: the biology and evolution of fossil plants

Figure 3.12 Section of Aglaophyton major axis with two clusters of fungal spores (Devonian). Bar  300 μm. Figure 3.14 Palaeomyces sp. spores within large resting spore

(Devonian). Bar  75 μm.

Figure 3.13 Hyphae and chlamydospores in Rhynie chert plant tissue (Devonian). Bar  120 μm.

Palaeoblastocladia (Remy et al., 1994), and Glomites (Taylor et al., 1995), are but a few of the morphotaxa known from the Rhynie chert ecosystem. These are described later in the sections “Glomeromycota” and “Chytridiomycota” (Palaeoblastocladia). Other fungi in the Rhynie chert cannot yet be assigned to major groups. They consist of non-septate hyphae that branch at irregular intervals, as well as hyphae that are distinctly septate, with the central region of the septum slightly thickened (Kidston and Lang, 1921a). At various points along the hyphae, ovoid to pear-shaped vesicles occur which are believed to have developed into large (250 μm), thickwalled sporangia. In other thin-section preparations, especially those of the chert matrix, larger sporangia occur, but these have not been found attached to hyphae. Those with stratified walls were named Palaeomyces gordonii var. major and are now thought to be members of the Glomeromycota (discussed later). Other sporangia contained a variety of thick-walled structures termed resting spores or resting sporangia. There is no doubt that several different natural forms are represented by these Rhynie chert fungi, but they cannot yet be assigned to a particular clade with certainty, as important parts of their life cycles have not yet been discovered. Recent work indicates that there is not only a considerable diversity of fungi within the Rhynie chert but also that these specimens offer the opportunity to examine the life history

CHAPTER 3 fungi, bacteria, and lichens

of some of these early terrestrial microorganisms as well as their biological interactions with other components of the ecosystem. The biological relationships of the Rhynie chert fungi, especially those preserved within or on land plants in the chert, have been variously interpreted historically. Kidston and Lang (1921a) suggested that the nutritional mode of most of these fungi was saprophytic, but that some may have represented symbionts. Other workers (Boullard and Lemoigne, 1971) hypothesized that some of the Rhynie chert fungi were parasitic. Further work has shown that some of the Rhynie plants exhibit host response features and these can provide clues to the nutritional mode of certain fossil fungi. The presence of fungi in two different taxa from this site, Rhynia gwynne-vaughanii and Aglaophyton major (formerly Rhynia major), was at one time used to support the suggestion that these two different plants represented independent phases of the same life cycle. As will be discussed in Chapter 8, we now

Multicelled fungal spore (Cretaceous). Bar  10 μm. (Courtesy J. M. Osborn.)

Figure 3.15


know that R. gwynne-vaughanii and A. major are not only different organisms, but probably belong to different clades. Fungal spores represent one of the most common examples of fungi in the fossil record, and are found in a variety of facies from the Paleozoic (Pirozynski, 1976a; Ediger and Alisan, 1989) to the recent (see section “Fungal Spores”). The identification of fungal palynomorphs (FIG. 3.15) is difficult, but there are now several glossaries of descriptive terms relating to spore and thallus morphology and structure (Elsik et al., 1983; Kalgutkar and Jansonius, 2000). Systematics of Fungi

In the last edition of this book (Taylor and Taylor, 1993), fungi and fungal-like organisms were described within a stratigraphic framework, and when possible, comments were offered on where they might fit within a modern classification of the Kingdom Fungi. With more information now available about fungal diversity through time, and many more specimens known in greater detail, many of the fossil taxa can now be discussed within the context of modern fungal groups. Phylogeny of the fungi was once based on morphology and, in some cases, characteristics in laboratory cultures. Today, the fungal tree (FIG. 3.1) of life is constantly refined using various molecular sequences. Most analyses include five Phyla within the fungi, three of which, the Glomeromycota, Ascomycota, and Basidiomycota, are considered monophyletic. The other fungi, which are considered to be the earliest-diverging fungi based on molecular phylogenies, include the paraphyletic groups, Chytridiomycota and Zygomycota. See Blackwell et al. (2006) for a summary of progress on the fungal tree of life, as well as the special issue of Mycologia, A Phylogeny for Kingdom Fungi (98(6), December 2006). CHYTRIDIOMYCOTA Members of this phylum are the only fungi that produce motile spores (zoospores) at some stage in their life cycle. Today chytrids are found in both soil and freshwater, with many functioning as saprotrophs and some as parasites (FIG. 3.16). Phylogenetic studies based on molecular markers view the chytrids as an early diverging, probably paraphyletic group within the Kingdom Fungi and the sister group of the remaining non-flagellated fungi (Zygomycota, Glomeromycota, Ascomycota, Basidiomycota) (James et al., 2006a). Increased resolution within the classification of living members of the Chytridiomycota is occurring at several levels (Letcher et al., 2005; James et al., 2006b). For


Paleobotany: the biology and evolution of fossil plants

Figure 3.16 Chytrid (arrow) with discharge papillae attached

to pollen grain (Extant). Bar  25 μm. (Courtesy D. Barr.)

Figure 3.19 Chlamydospore with mycoparasite developing

between wall layers (arrow) (Devonian). Bar  20 μm.

Figure 3.17 Chytrid zoosporangia embedded in outer wall of

Palaeonitella cranii cell. Bar  20 μm.

Figure 3.20 Several chytrid zoospores. Arrow indicates flagellum (Devonian). Bar  10 μm.

Figure 3.21 Several chytrid zoosporangia on the surface of a Figure 3.18 Detail of chlamydospore wall showing mycopara-

site developing between wall layers (Devonian). Bar  12 μm.

land plant spore. Arrow shows zoosporangium exit site (Devonian). Bar  10 μm.

CHAPTER 3 fungi, bacteria, and lichens

example, at the molecular level, the relationship between members of the Chytridiomycota and Glomeromycota is strengthened by the presence of certain tubulin genes in both groups (Corradi et al., 2004). Although there have been some putative chytrids described from Cambrian rocks (see Butterfield, 2005 for a review), the best-preserved forms are from the Lower Devonian Rhynie chert (FIG. 3.17) (Illman, 1984; Taylor et al., 1992). These fossils possess thalli and discharge tubes (FIGS. 3.18 and 3.19) similar to those in certain modern chytrids, and the preservation is so detailed that it is possible to demonstrate the presence of a single flagellum (FIG. 3.20) on a fossil zoospore (Taylor et al., 1992). These fossil chytrids possess thalli that are epi- and endobiontic (living on or within other organisms), and are associated with land plants, spores (FIG. 3.21), and algae (FIG. 3.22), as well as occurring isolated in the matrix. One especially interesting form is Palaeoblastocladia (Remy et al., 1994), a fungus that shares many features with members of the extant Blastocladiales

(FIG. 3.23). Palaeoblastocladia milleri consists of thalli that were produced beneath the cuticle of stems of Aglaophyton and extended out from the surface about 0.5 mm (FIG. 3.24). Some thalli had terminal zoosporangia (FIG. 3.25), whereas others produced pairs of terminal gametangia (FIG. 3.26). This complement of characters suggests that P. milleri possessed an isomorphic alternation of generations with sexual reproduction, a combination of features that is very rare in modern fungi. It is suggested that P. milleri was a saprotrophic, but other Rhynie chert chytrids are believed to have been parasites. Despite the small size of chytrids, their ubiquity in the Rhynie chert provides the opportunity to study their life history biology (FIG. 3.27), since various developmental stages are preserved. Geologically younger fossil chytrids are also known (FIG. 3.28). Some of the first late Paleozoic chytrids to


Figure 3.23

Figure 3.22 Chytrid zoosporangium. Arrow indicates position

of discharge tube (Devonian). Bar  20 μm.



Life history of Palaeoblastocladia milleri.

Figure 3.24 Two tufts (arrows) of Palaeoblastocladia milleri extending from epidermis of Aglaophyton axis (Devonian). Bar  400 μm.


Paleobotany: the biology and evolution of fossil plants

Figure 3.26 Two pairs of Palaeoblastocladia milleri gametangia (Devonian). Bar  30 μm.

Figure 3.25 Zoosporangium of Palaeoblastocladia milleri

showing zoospores with dark central body (Devonian). Bar  15 μm.

Mature holocarpic thallus Zoospore discharge Immature thallus

Zoospore Protoplast Penetration tube

Figure 3.27

Life history of a Devonian chytrid.

CHAPTER 3 fungi, bacteria, and lichens


be accurately identified were Grilletia spherospermii and Oochytrium lepidodendri (FIG. 3.29) which occur in Carboniferous gymnosperm seeds and anatomically preserved tissues of the lycopsid Lepidodendron sp. (Renault and Bertrand, 1885a (FIG. 3.32); Renault, 1895, 1896a). These authors not only related these fossils to modern chytrid genera but also suggested that they represented parasitic fungi. Other microfungal remains associated with permineralized Lepidodendron tissues are more difficult to interpret (FIGS. 3.30, 3.31). Fossil chytrids (FIG. 3.33),

Figure 3.28 Thick-walled lycopsid spore containing fungal

spores (chytrids?) (Pennsylvanian). Bar  225 μm.

Arthroon rochei, thick-walled spore in periderm of Lepidodendron (Mississippian). Bar  100 μm. (Courtesy N. Dotzler.)

Figure 3.30

Figure 3.29 Oochytrium lepidodendri in Lepidodendron tracheids (Mississippian). Bar  20 μm. (Courtesy N. Dotzler.)

Figure 3.31 Palaeomyces gracilis. (Mississippian). Bar  35 μm. (Courtesy N. Dotzler.)


Paleobotany: the biology and evolution of fossil plants

Figure 3.34 Several chytrids attached to substrate (Extant). Bar  50 μm. (Courtesy D. Barr.)

Figure 3.32 Paul Bertrand.

Figure 3.33 Numerous chytrids on surface of fungal spores

(Devonian). Bar  20 μm.

like their modern relatives, are often found on spores and pollen grains (FIG. 3.34) in bodies of water. Millay and Taylor (1978a) described epibiotic and endobiotic chytrid thalli (FIGS. 3.35, 3.36) in association with Pennsylvanian cordaitean pollen grains (Millay and Taylor, 1978a). Some of the chytrid zoosporangia recovered from the Eocene Green River Formation, which is known for its excellent

Saccate pollen grain with two chytrid zoosporangia (arrows) attached (Pennsylvanian). Bar  25 μm. Figure 3.35

preservation, are so nearly identical to those of modern forms that the fossils are assigned to living genera (Bradley, 1964, 1967). ZYGOMYCOTA Zygomycetous fungi are distinguished by the production of thick-walled zygospores (non-flagellated) that form in a

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.36 Epibiontic chytrid zoosporangium extending from

corpus wall of pollen grain (Pennsylvanian). Bar  10 μm.

Figure 3.38 Dubiocarpon (Pennsylvanian). Bar  175 μm.

Figure 3.37 Suggested reconstruction of Endochaetophora antarctica thought to represent an ascomycete sporocarp (Triassic).

special sporangium, the zygosporangium, following gametangial fusion. Hyphae are generally aseptate and asexual reproduction occurs by the formation of internally produced spores. The group is highly diverse and includes saprotrophs (e.g., black bread mold) and certain pathogens, including some that infect other fungi, called mycoparasites. Some forms live as obligate symbionts within the gut of various arthropods (Lichtwardt et al., 1999; 2001). The Zygomycota are currently interpreted as paraphyletic and are believed

to have diverged from the chytrids before the colonization of the land. Two classes are currently recognized (White et al., 2006): Zygomycetes and Trichomycetes. The only fossil record for the Trichomycetes will be discussed under the section “Fungi–Animal Interactions”. Various spore-like bodies that are now interpreted as zygomycetous fungi occur in permineralized peat (FIG. 3.37), especially in Carboniferous coal balls. The discovery of these unique structures, termed sporocarps, is no doubt directly related to the long history of the study of vascular plants in coal balls. The fungal affinities of these fossils are based on the structure of the wall, which consists of aseptate, interlaced hyphae. Some of the most common forms are spherical and 1 mm in diameter. The most common morphogenera include Sporocarpon, Dubiocarpon (FIG. 3.38), Mycocarpon, Coleocarpon, and Traquairia (FIG. 3.39) (Stubblefield et al., 1983). Mycocarpon, a common Middle Pennsylvanian form, consists of a central spore-like structure 550 μm in diameter surrounded by a wall of interlaced, hypha-like cells four layers thick. Inside is an amorphous, cuticle-like membrane. In some species, for example M. bimuratus, the central cavity is filled with small spores. In Sporocarpon, the sporocarp is smaller (200 μm) and constructed of a pseudoparenchymatous tissue,


Paleobotany: the biology and evolution of fossil plants

Figure 3.39 Traquairia williamsonii sporocarp (Carboniferous).

Bar  100 μm.

which extends outward into numerous narrow, conical processes, each 6–7 cells high and 1–3 cells wide. Another form, Dubiocarpon, is distinguished by radially oriented, elongate cells that extend out from the sporocarp wall as spines, some with bifid tips. The most ornate form is Traquairia, a fungus that was initially described from the Lower Coal Measures of Great Britain by Carruthers (1872a). Since that time, numerous specimens have been reported from many localities in Pennsylvanian rocks (Stubblefield and Taylor, 1983). Individual specimens are roughly spherical and up to 1 mm in diameter. The wall is complex, with the outer portion constructed of branching hyphae, some of which are organized into hollow spines. In Roannaisia, from the Visean of central France, the hyphae branch and the lumen of the sporocarp contains a single, multilayered spore (Taylor et al., 1994b). As noted earlier, some of these sporocarps contain one to several spherical structures in the central lumen (FIG. 3.40). Initially these fungi were interpreted as ascomycetes, and the sporocarps as closed ascocarps (cleistothecia). Larger spherical structures inside would represent asci (sporangia), and smaller ones, ascospores (Stubblefield and Taylor, 1988). These sporocarps are now interpreted as zygomycetes, however, and the large, inner spore-like body represents a zygospore, similar to those produced in mycelial sporocarps of modern Mucorales (White and Taylor, 1989a). The smaller internal spores reported in some sporocarps (formerly interpreted as ascospores) are now regarded as mycoparasites, most likely some type of chytrid.

Figure 3.40 Sporocarp of Dubiocarpon containing several larger spores (Pennsylvanian). Bar  225 μm.

An interesting fossil believed to be a zygomycete is Protoascon missouriensis, a fungus found within a seed-like structure in a Pennsylvanian coal ball from Missouri, USA (Batra et al., 1964). The description is based on multiple specimens of a bulb-like structure 150 μm diameter (FIG. 3.41). At one end is a whorl of 12 elongate, aseptate appendages (FIG. 3.42). The appendages are curved to form a loose, basket-like structure which surrounds a highly ornamented sporangium containing a single, thick-walled spore. As the name suggests, this fossil was initially thought to be an ascomycete, but later reinterpreted as a chytrid (Baxter, 1975), and most recently as a zygomycete (Taylor et al., 2005a). Based on reexamination of the type material and additional specimens, Taylor et al. (2005a) concluded that the aseptate appendages (or suspensors) partially enclose a sporangium, which is either an azygosporangium (asexual) or a zygosporangium, containing a single, thick-walled spore (FIG. 3.43). GLOMEROMYCOTA The Glomeromycota are a clade that was instituted based on rDNA phylogenies of living members (Schüssler et al., 2001; Redecker and Raab, 2006). The phylum, which includes the arbuscular mycorrhizal (AM) fungi, was formerly included

CHAPTER 3 fungi, bacteria, and lichens





Figure 3.41 Protoascon missouriensis showing suspensor (S)

with appendages and (a)zygosporangium (AZ) (Pennsylvanian). Bar  50 μm. Figure 3.43 Protoascon missouriensis thick-walled (a)zygosporangium (AZ) (Pennsylvanian). Bar  50 μm.

Figure 3.42 Suspensor appendages of the mucoralean fungus Protoascon missouriensis (Pennsylvanian). Bar  50 μm.

within the Zygomycota, and is now considered to be the sister group of the clade formed by the Ascomycota  Basidiomycota, based on molecular phylogenies (Blackwell et al., 2006; Redecker and Raab, 2006). Extant Glomeromycota are comprised of obligate symbionts that may form arbuscules in plant roots; they produce large (40–800 μm),

multilayered spores which are attached to non-septate hyphae. More than 90% of extant land plants have a symbiotic (mutualistic) relationship with mycorrhizal fungi in their roots. There are two basic types of extant mycorrhizae: ecto- and endomycorrhizae. Endomycorrhizae are formed by members of the glomeromycetes and are the most common form today. The fungal hyphae grow within the host root, and although they penetrate the host cell walls, they do not penetrate the plasma membranes. Most produce arbuscules, highly branched hyphal structures that provide for exchange between the fungal symbiont and its host. Some also produce storage organs called vesicles (the vesicular-arbuscular mycorrhizae). Ectomycorrhizae are formed by members of the Basidiomycota and a few ascomycetes; they are less common today and occur primarily in woody plants of the temperate zone, including many conifers (Chapter 21). In this case, the fungal hyphae form a net around the outside of the plant root, which penetrates between the cells of the root itself. Although there are many hypotheses for the establishment of this fungal–land plant association (e.g., resistance against drought, defense against root herbivory, etc.), the mutualistic association provides for increased mineral nutrient uptake by the plant in exchange for a source of carbon for the fungus.


Paleobotany: the biology and evolution of fossil plants

Figure 3.45 Intercalary glomalean chlamydospore (Devonian).

Bar  70 μm.

Figure 3.44

Bernard Renault.

The fossil record of Glomeromycota is believed to be ancient, extending well back into the Paleozoic. Spores and hyphae of a glomeromycotan type have been reported from rocks as old as the Cambrian (Pirozynski and Daplé, 1989), and from 460–455 Ma Upper Ordovician rocks (Redecker et al., 2000). Palaeoglomus grayi has aseptate (coenocytic) hyphae and spores that resemble living Glomus spores (Redecker et al., 2002). In these reports, however, there is no association with land plant remains, and thus the symbiotic association of these fungi remains unresolved. In addition, Butterfield (2005) noted that the very simplicity of these organisms does not provide enough diagnostic characters to separate them from other protists or parasites. Palaeomyces is a name first used by Renault (1896a) (FIG. 3.44) and later by Kidston and Lang (1921a) and others to describe large, isolated spores associated with Paleozoic plant remains; some of these are now included in the morphotaxon Glomites (FIG. 3.45). Although originally described from the Rhynie chert, they are also known from other sites. As well as being a source for beautifully preserved fossil fungi, the Rhynie chert ecosystem provides some of the best fossil evidence of fungi and land plants interacting in a symbiotic association. Glomites rhyniensis

Figure 3.46 Detail of mycorrhizal arbuscule zone (dark band) in Aglaophyton (Devonian). Bar  325 μm.

consists of aseptate hyphae and globose, multilayered spores, which occur in the tissues of several macroplants from this Devonian site (Taylor et al., 1995). Of special significance is the discovery of highly branched, intercellular arbuscules of this species within the cells of the land plant Aglaophyton major. In extant plants with AM fungi, arbuscules are

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.48 Section of Aglaophyton major axis showing cortical cells and hyphae (arrows) of Glomites rhyniensis in intercellular spaces. Bar  32 μm.

Figure 3.47 Aglaophyton major showing arbuscule trunk hyphae (arrows) and opaque structures representing arbuscules (Devonian). Bar  80 μm. (Courtesy H. Kerp.)

confined to cells of the root, but in A. major the arbuscules occur in a narrow zone of cells inside the cortex, termed the mycorrhizal arbuscule zone (FIGS. 3.46, 3.47), which extends throughout the rhizome and proximal portions of the aerial axes (FIG. 3.48). More recently another land plant from the Rhynie chert, Rhynia gwynne-vaughanii, was described with Glomites fungi in the cortical tissues (Karatygin et al., 2006). In this taxon, G. sporocarpoides, arbuscules were not identified, but there were large, glomoid sporocarps containing numerous spores (20–24 μm in diameter) in the tissues. Rhizomatous and upright axes of Nothia aphylla, another land plant from the Rhynie chert, host a glomeromycotan fungus that closely resembles G. rhyniensis. Glomites rhyniensis is an intercellular endophyte, however, that becomes intracellular only within a well-defined region of the cortex where it forms arbuscules. The fungus in N. aphylla is initially intracellular, but later becomes intercellular in the cortex where it forms vesicles (FIG. 3.49) and thick-walled spores (Krings et al., 2007b). If this fungus is

Figure 3.49 Vesicles and hyphae in Nothia aphylla axes.

Bar  100 μm.

functioning as an endomycorrhiza, N. aphylla displays an alternative mode of colonization by endomycorrhizal fungi, which may be related to the peculiar internal anatomy of the rhizomatous axis. In this part of the axis of N. aphylla, the cells are arranged in radial rows with virtually no intercellular spaces, so perhaps there is no intercellular infection pathway into the cortex (Krings et al., 2007a, b) this part of the axis of (see Chapter 8 for additional data on N. aphylla).


Paleobotany: the biology and evolution of fossil plants

Figure 3.51 Cross section of Psilophyton dawsonii axis contain-

ing chlamydospores (Devonian). Bar  750 μm. (From Stubblefield and Banks, 1983.)

Figure 3.50 Arbuscules (arrows) in gametophyte of Lyonophyton rhyniensis (Devonian). Bar  35 μm.

We now know that several of the gametophytes of the Rhynie chert plants were also endomycorrhizal (FIG. 3.50). Although the Rhynie chert does not represent the oldest terrestrial ecosystem, the extraordinary preservation at this site does indicate that the association of plants and certain types of fungi is ancient and, in fact, lends support to the hypothesis that fungi may have been critical in the early colonization of land by plants (see Chapter 6; Pirozynski and Malloch, 1975). In addition to the fungal remains associated in the Rhynie chert plants, large spores like those associated with mycorrhizal symbiosis in the Rhynie chert plants have been found in many other Paleozoic plants. Within the tissues of Psilophyton dawsonii (Late Devonian) are spherical spores up to 175 μm in diameter (FIG. 3.51), which appear similar to the chlamydospores (thick-walled, asexual resting spores) of certain modern fungi (Stubblefield and Banks, 1983). Such chlamydospores (FIG. 3.52) are also relatively common in the tissues of a large number of Carboniferous plants preserved in coal balls (Wagner and Taylor, 1982). In some instances they occur within the tissues of underground

Figure 3.52 Lycopod megaspore containing chlamydospores (arrows) (Pennsylvanian). Bar  150 μm. (From Stubblefield and Taylor, 1988.)

organs such as Stigmaria, the underground organ of the arborescent lycopsids; in other cases they are solitary within the matrix. Spores of this type range from 100–400 μm in diameter and possess a multilayered wall. Many are preserved attached to the hyphae that produced them, and, in some specimens, the attachment area is structurally identical to those seen in modern endomycorrhizal fungi like Glomus. Dotzler et al. (2006) described thick-walled spores from the Rhynie chert that are unusual because they possess a germination shield (FIGS. 3.53, 3.54), a structural feature associated with spore germination. The shield is a lobed structure that occurs between the inner and outer walls of the spore. What is especially intriguing is that typical germination shields only occur in a few glomeromycotan genera. To date the function of the germination shield remains unknown. Although molecular studies of the Glomeromycota have not resolved whether the germination shield is a derived or

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.55 Arbuscule in root of Antarcticycas (Triassic). Bar 

25 μm.

Figure 3.53 Scutellosporites devonicus showing germination shield in cross section (arrow) (Devonian). Bar  50 μm.

Figure 3.54 Tongue-shaped germination shield of a glomero-

mycotan fungus (Devonian). Bar  100 μm (Courtesy H. Kerp and H. Hass.)

pleisomorphic character, its discovery in the Rhynie chert is important as a direct marker for calibrating molecular clock hypotheses on endomycorrhizal fungi. In spite of the morphological similarities between many of the fossil spores and the thick-walled spores of the modern endomycorrhizal fungi, the presence of spores does

not conclusively demonstrate that all of the plants were mycorrhizal. Thick-walled spores are also produced by several groups of non-mycorrhizal fungi and thought to represent resting propagules. It is the distribution of the fungal hyphae within the tissues of the host plant and the presence of arbuscules, the specialized highly branched structure of the fungus that forms within cells of the host, that are the most conclusive pieces of evidence for endomycorrhizal associations. Although arbuscules have been reported from as early as the Paleozoic in permineralized plants, with the exception of those in the Rhynie chert, all have subsequently been reinterpreted as produced by nonmycorrhizal fungi or as some form of coalesced cell contents. One example of a Mesozoic host plant with true arbuscules comes from silicified Middle Triassic (Anisian) peat collected in Antarctica (Stubblefield et al., 1987a). Within the roots of the cycad Antarcticycas schopfii are large spores, hyphae (FIG. 3.55), and highly branched structures that are most certainly arbuscules. Like the arbuscules of modern AM fungi, the Triassic specimens possess hyphae that arise from a single point on the cell wall and repeatedly branch to fill the host cell lumen. Additional evidence supporting the existence of AM fungi in this Triassic ecosystem is the presence of the thallus of another mycorrhizal fungus, Sclerocystis. This fossil consists of an aggregation of nonseptate hyphae that give rise to 30 terminally produced thick-walled spores (FIG. 3.56) (Stubblefield et al., 1987b). The fungus appears most similar to the living species S. rubiformis. Stockey et al. (2001) reported coiled hyphae and arbuscules in the roots of a middle Eocene taxodiaceous conifer, Metasequoia milleri. Hopefully, these various examples of arbuscules will prompt the reexamination of underground organs of other fossil plants, not only to substantiate the existence of obligate biotrophs but also to offer


Paleobotany: the biology and evolution of fossil plants

Figure 3.56 Chlamydospores of Sclerocystis sporocarp (Triassic). Bar  15 μm.

information that may be useful in tracing the evolution of the important nutrient-exchange structure, the arbuscule.

ASCOMYCOTA The ascomycetes, or sac fungi, are the most diverse clade of living fungi, with more than 40,000 species. In recent phylogenies, the group includes three subclasses: the Taphrinomycotina, an early-diverging clade, the Saccharomycotina (the yeasts), and the Pezizomycotina, which includes the vast majority of the multicellular ascomycetes. The clade includes both singlecelled (yeasts) and filamentous forms, and the hyphae of the latter are septate. Sexual reproduction is characterized by an ascus (pl. asci), a sac-like sporangium in which nuclear fusion (karyogamy), followed by meiosis, take place, leading typically to the production of 4, 8, or 16 ascospores (FIG. 3.57). Asexual reproduction occurs by fission, fragmentation, or the formation of asexual spores. Ascomycetes are worldwide in distribution and can be found in a wide variety of habitats, today and in the past (White and Taylor, 1988). Several important pathogens today are ascomycetes (e.g., chestnut blight, powdery mildews, ergot), as well as the edible truffle. Some ascomycetes form symbiotic associations with certain algae and/or cyanobacteria to form lichens. The Ascomycota are regarded as sister to the Basidiomycota based on morphological and molecular phylogenies (Blackwell et al., 2006). Morphological features that ally them with the basidiomycetes include the segmentation of the hyphae in the form of cross walls (septa, sing. septum),

pairs of unfused nuclei after mating and before nuclear fusion, and structures in both groups that coordinate simultaneous mitosis of the two nuclei. Molecular clock analyses (Taylor and Berbee, 2006) and recent fossil evidence indicate that the Ascomycota are far older than once believed. The earliest fossils attributed to the Ascomycota are specialized cells (phyllides) macerated from middle Silurian rocks of Gotland, Sweden (Sherwood-Pike and Gray, 1985). These remains consist of tubular filaments with perforate septa; on some filaments the surface is ornamented by short branches (FIG. 3.10). Multiseptate spores with 1–9 cross walls occur in the macerates as well (FIG. 3.11). Evidence of ascomycetous fungi has also been reported on the cuticle of the Devonian enigmatic thalloid plant Orestovia devonica from the Early Devonian of Siberia (Krassilov, 1981a), including haustoria, hooked hyphae, and asci. One of the best-preserved examples of an early ascomycete is Paleopyrenomycites from the Rhynie chert (Taylor et al., 2005b). Specimens of P. devonicus occur as perithecial ascocarps just beneath the epidermis of the vascular plant Asteroxylon mackiei (FIG. 3.58). Perithecia are spherical with a short ostiolate neck (FIG. 3.59), and asci and paraphyses (sterile, hair-like filaments) line the inner surface of the ascocarp. Each ascus produces up to 16 ascospores (FIG. 3.60) with 1–5 cells each. In addition to the presence of the sexual (telemorphic) stage, P. devonicus is also known from tufts of conidiophores that are interpreted as the asexual (anamorphic) component of the life history. The presence of both phases makes P. devonicus one of the most completely known fossil fungi. Taylor et al. (2005b) noted that the features of this fungus compared with those in several extant groups (e.g., Xylariales in the Sordariomycetes); a cladistic analysis was not able to clarify relationships. The extraordinary preservation of this Early Devonian ascomycete should help to define character evolution in the ascomycetes, as well as serving as a benchmark in tracing the evolution of major lineages within the Ascomycota (Taylor and Berbee, 2006). Mycokidstonia sphaerialoides is another Rhynie chert fungus interpreted as an ascomycete (Pons and Locquin, 1981). They describe ascomata (fruiting bodies that give rise to asci and ascospores), which are spherical and 175 μm in diameter, but these structures share features with chytrid zoosporangia, so the attribution to the Ascomycota is not certain. Palaeosclerotium pusillum represents a sporocarp found among plant debris in Carboniferous coal balls and is believed to have some affinities with the Ascomycota (Rothwell, 1972). Each sporocarp consists of an ovoid cleistothecium 1.2 mm in diameter with a two-parted wall. The wall consists of a central zone of branched, septate hyphae,

CHAPTER 3 fungi, bacteria, and lichens

Asexual reproduction by spores (conidia)


Ascogonium Antheridium

Ascus Germinating ascospores (n)


Ascospores (n)


Mitotic division Ascospore formation Developing ascogenous hyphae (nn)

Second meiotic division

Dikaryotic ascogenous hyphae (nn)

First meiotic division Young ascus (2n) Karyogamy (nn)

Ascocarp (Apothecium)

Sterile hyphae (n)

Figure 3.57 Life history of an ascomycete. (From Taylor and Taylor, 1993.)

Figure 3.58 Section of Asteroxylon mackiei enations show-

ing position of several Paleopyrenomycites devonicus perithecia (Devonian). Bar  10 μm.

Figure 3.59 Section through ostiole (arrow) of perithecium of Paleopyrenomycites devonicus and numerous ascospores in cavity (Devonian). Bar  60 μm.


Paleobotany: the biology and evolution of fossil plants

3.60 Several asci containing ascospores Paleopyrenomycites devonicus (Devonian). Bar  15 μm.



surrounded by a zone of pseudoparenchymatous tissue (FIG. 3.61). What is especially interesting about P. pusillum is that it appears to exhibit a combination of characteristics of several extant fungal groups. Within the cleistothecium are what have been interpreted as asci (FIG. 3.62) containing a variable number of spores (4–8). Hyphae possess dolipore-like septa and clamp connections, features typical of basidiomycetous fungi, but also found in other groups (McLaughlin, 1976). The taxonomic affinities and nutritional role of this interesting fossil have been variously interpreted (Dennis, 1976). Some regard it as a fungus intermediate between Ascomycota and Basidiomycota (McLaughlin, 1976), whereas others see it as an ascomycetous fruiting structure closely related to the Eurotiales, which has been parasitized by the mycelium of a basidiomycete (Singer, 1977). The Eurotiales are an order of ascomycetes that includes the sexual stages of Penicillium and Aspergillus, with the latter found in Baltic amber from Russia (FIGS. 3.63–3.65). In another interpretation, Palaeosclerotium is used as an example of a fungus that links the Basidiomycota symbiotically with possible nematophytes (Chapter 6), an extinct group of early land organisms (Pirozynski and Weresub, 1979). Other fossil evidence of Paleozoic members of the sac fungi include a variety of multicelled fungal spores with thick septa, which have been interpreted as ascomycetes (Kalgutkar and Jansonius, 2000). Another group of ascomycetes that are represented in the fossil record are the sooty molds belonging to the Capnodiaceae family (Dothideomycetes), a group that today produces colonies of dark hyphae on living plants often associated with aphid infestations. The aphids parasitize the plants, which, in turn, exude a sap that becomes the substrate for fungal growth. Fragments of Bitterfeld (24–22 Ma,

Figure 3.61 Palaeosclerotium pusillum showing outer, pseu-

doparenchymatous zone and central region of branched and septate hyphae (Pennsylvanian). Bar  500 μm.

Figure 3.62 Palaeosclerotium pusillum with central region containing spore-like structures (Pennsylvanian). Bar  500 μm.

Oligocene–Miocene) and Baltic (55–35 Ma, Eocene) amber contain brown hyphae with tapering distal ends that are identical to the extant sooty mold Metacapnodium (Rikkinen et al., 2003). Today some regard the sooty molds as monophyletic based on molecular studies (Reynolds, 1998; Schoch et al., 2006).

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.63 Springtail overgrown by Aspergillus collembolorum (Eocene). Bar  0.5 mm. (Courtesy A. Schmidt.)

Figure 3.65 Aspergillus collembolorum, conidial head with radial chains of conidia (see FIG. 3.64) (Eocene). Bar  25 μm. (Courtesy A. Schmidt.)

Figure 3.64 Aspergillus collembolorum, sporulating conidio-

phores (see FIG. 3.63) (Eocene). Bar  50 μm. (Courtesy A. Schmidt.)

Stigmatomyces succini is a parasitic ascomycete that has been found attached to the thorax of a stalk-eyed fly preserved in Baltic amber (Rossi et al., 2005). It represents the earliest fossil account for the ascomycete order Laboulbeniales, an enigmatic group which is now considered to represent ascomycetes (Weir and Blackwell, 2001). BASIDIOMYCOTA The Basidiomycota, a monophyletic sister group to the Ascomycota, includes 30,000 extant species divided into three major lineages (subphyla): the rusts (Puccinomycotina), smuts (Ustilagomycotina), and mushrooms (Agaricomycotina). They are known from both terrestrial and aquatic habitats around the world and include important plant pathogens (e.g., wheat rust, corn smut), as

well as the edible mushrooms. The most diagnostic feature of the basidiomycetes is the basidium (pl. basidia), a generally club-shaped cell where nuclear fusion (karyogamy) takes place and the structure upon which the sexual spores (basidiospores) are produced. Some basidia are borne on complex, multicellular fruiting bodies, for example the mushrooms (FIG. 3.66). Other basidiomycete features include hyphal outgrowths termed clamp connections, and the presence of a dikaryon phase in the life cycle, a condition in which each cell in the thallus contains two nuclei. Some basidiomycetes are involved in ectomycorrhizal associations, whereas others are symbiotically associated with various insects, for example with leaf-cutter ants and termites. One example of a fungus–termite interaction is the Miocene–Pliocene termite trace fossil Microfavichnus; this ichnogenus is thought to represent fungus combs of fungus-growing termites (Duringer et al., 2007). The combs consist of alveolar masses in which the walls have a pelletal structure. The Basidiomycota today play an important role in the carbon cycle by decaying organic matter, including wood;


Paleobotany: the biology and evolution of fossil plants

Diploid nucleus undergoes meiosis

Pair of nuclei fuse

2n Fertilization

Meiosis nn

n Basidiospores



Basidium Cap Portion of gill

Gill Developing basiodiocarp


Monokaryotic mycelium





Dikaryotic mycelium

Monokaryotic mycelium

Figure 3.66 Life history of basidiomycete fungus. (From Taylor and Taylor, 1993.)

presumed basidiomycetous fungi have been found as early as the Middle Devonian, not long after land plants first began producing secondary growth. There are a number of Carboniferous fossils that superficially resemble modern basidiomycetous fruiting bodies, or basidiocarps (Lindley and Hutton, 1831–1837; Lesquereux, 1877; Herzer, 1893; Hollick, 1910) (FIG. 3.67). Almost all of these reports, however, were later questioned and the specimens reinterpreted as non-fungal (Pirozynski, 1976b). The fact that there are so few reports of basidiomycetes associated with fossil wood is perplexing, since today they are the major decomposers of cellulose and lignin, the major components of plant cell walls. One well-documented basidiomycete from the Pennsylvanian is Palaeoancistrus martinii, a fungus found in the tracheids of the fern Zygopteris (Dennis, 1970). There are several features of this fungus that suggest affinities with living saprotrophic members of the Basidiomycota. One of these is the presence of smooth, narrow, septate hyphae

(4.8 μm in diameter) that follow a straight course within the tracheids. Associated with the hyphae are both terminal and intercalary chlamydospores. Some hyphae possess incomplete clamp connections in which the hook of the clamp does not form a complete union with the hypha, whereas in others the clamp connections are well developed. Another basidiomycete has been described in the secondary xylem of several Paleozoic and Mesozoic woods from Gondwana (FIG. 3.68), but in this case it is the symptoms caused by the fungus, in association with the fungus itself that provides the identification. The activities of this fungus result in the formation of numerous longitudinally oriented, spindle-shaped pockets of decay in the secondary xylem (FIG. 3.69), called pocket rot (Stubblefield and Taylor, 1986). In other regions of the secondary xylem, for example in the root Vertebraria and stems assigned to the morphogenus Araucarioxylon, septate hyphae with simple and medallion clamp connections are present (FIG. 3.70). The elongate,

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.69 Fractured surface of extant wood showing elon-

gate spindles (white) of white pocket rot. Bar  1 cm.

Figure 3.67 Arthur Hollick.

3.70 Medallion clamp connection (arrow) in Araucarioxylon wood infested by pocket rot (Triassic). Bar  20 μm. (From Stubblefield and Taylor, 1986.) Figure

Figure 3.68 Cross section of Araucarioxylon wood showing symptom (white areas) of white pocket rot (Triassic). Bar  2 cm.

spindle-shaped areas in the fossil are identical to the symptoms caused by modern white pocket rot fungi (Blanchette, 1980). Other structural features in the fossil woods indicate the sequential delignification of the secondary cell wall (FIG. 3.71), for example the loss of the middle lamella between the wood cells, and these are also characteristic of modern white rot fungi. Palaeofibulus is another fossil fungus with clamp connections (FIG. 3.72) and thick-walled

spores, known from Middle Triassic permineralized peat of Antarctica (Osborn et al., 1989). Investigations of woodrotting fossil fungi offer the potential to indirectly examine the biochemical evolution of fungi, based on patterns and features associated with delignification and removal of cellulose from the host cell wall. Some of the most common members of the Basidiomycota are the mushrooms (Hibbett, 2006). To date the oldest gilled mushrooms (Agaricales or agarics) come from amber. The oldest of these is Archaeomarasmius leggeti (FIG. 3.73) entombed in a piece of Late Cretaceous (Turonian; 94– 90 Ma) amber from the Raritan Formation of New Jersey, USA (Hibbett et al., 1997). The pileus, or cap of the mushroom, ranges from 3–6 mm in diameter and contains elliptical


Paleobotany: the biology and evolution of fossil plants

Separation of wood cell wall layers due to fungal degradation (Triassic). Bar  55 μm.

Figure 3.71

Figure 3.73 Archaeomarasmius leggeti (Cretaceous). (Courtesy

D. Hibbert.)

Figure 3.72 Palaeofibulus showing hyphae, spores, and clamp

connection (arrow) (Triassic). Bar  35 μm.

basidiospores up to 8 μm long. Another gilled form is Protomycena electra from the younger Dominican amber (Miocene; 20–15 Ma) (Hibbett et al., 1997); this mushroom is similar to the extant leaf-litter and wood-rotting genus Mycena. Coprinites dominicana is another gilled mushroom found in Dominican amber. It has a cap 3.5 mm in diameter with 28 gills extending from the lower surface (Poinar and Singer, 1990). The most recent homobasidiomycete reported from Dominican amber is Aureofungus yaniguaensis (Hibbett et al., 2003). It is similar to the other fossil agarics in amber, and suggests that among certain homobasidiomycete lineages there has been relatively little morphological change since the Cenozoic, at least in those features that can be compared within the amber matrix. Basidiomycetes that lack gills but possess large basidiocarps with poroid hymenophores (spore producing layers in the fruiting bodies) have been described from the

fossil record as bracket, shelf, or polypore fungi (surveyed in Fleischmann et al., 2007). Members of this group are generally saprotrophs involved in the degradation of cellulose and lignin, but some are also parasites of woody plants. Quatsinoporites cranhamii is a fragment of an Early Cretaceous (Barremian) basidiocarp described from permineralized marine calcareous concretions of British Columbia, Canada (Smith et al., 2004). The hymenophore consists of numerous parallel tubes, each up to 540 μm in diameter. Appianoporites is also a poroid (bracket) fungus constructed of smaller tubes, from younger Eocene rocks from the eastern side of Vancouver Island, British Columbia. Both specimens possess septate hyphae; the basidiocarps are interpreted as persistent and placed in the Hymenochaetales. Another fossil polyporous fungus is Ganodermites libycus (FIG. 3.74) from the lower–middle Miocene (Neogene) of North Africa (Fleischmann et al., 2007). The basidia are clavate and produce elliptical basidiospores, each up to 6.5 μm long with two wall layers. The basidiocarp shows evidence of incremental growth and is placed in the extant family Ganodermataceae, a predominantly tropical group of fungi that are characterized by basidiospores with a double (so-called ganodermatoid) wall. The presence of tunnels containing fecal pellets in the basidiocarp indicates that this bracket fungus was visited by fungivores.

CHAPTER 3 fungi, bacteria, and lichens



Figure 3.74 Ganodermites libycus, longitudinal thick section (polished surface) through the basidiocarp showing hymenophoral strata (H) (Miocene). Bar  2 cm. (Courtesy BSPG.)

Basidiomycetes also form a variety of symbiotic associations with other organisms, including many families of temperate forest trees, for example Fagaceae and Pinaceae. Distinctive among these are ectomycorrhizae, estimated to occur on 10% of plants. Ectomycorrhizae are characterized by intercellular hyphae that often form a loose aggregation, called a Hartig net, around the root tip. In addition, the fungal association can also cause a change in root morphology, making the roots shorter, wider, and more branched. The oldest evidence of ectomycorrhizae to date comes from the middle Eocene (50 Ma) Princeton chert of the Allenby Formation, British Columbia, an extraordinary site in which many whole plants are permineralized by silica (LePage et al., 1997). Ectomycorrhizal roots of Pinus from this site contain a dense aggregation of small septate hyphae that extend into the cortex of the roots and represent the Hartig net. This discovery represents another example of the long standing symbiotic relationship in a particular group of seed plants. It is unclear why the majority of living ectomycorrhizae coevolved with woody rather than herbaceous plant hosts. Perhaps the structure and organization of the root, soil type, microbial community, or some other combination of biotic and abiotic factors favored ectomycorrhizal fungi over endomycorrhizal forms in certain hosts. A striking example of the diversity of fossil basidiomycetes is seen in a fossil specimen assignable to the Gasteromycetes, the group that contains the puffballs and some of the earthstars. Geastrum tepexensis (FIG. 3.75) is a compressed late Eocene basidiocarp (called a peridium in this group) approximately 2.5 cm in diameter from the Coatzingo Formation in Puebla, Mexico (Magallón-Puebla and Cevallos-Ferriz, 1993). The endoperidium (inner layer of the peridium) is circular and characterized by a small ostiole; ornamented spores 7 μm in diameter were also

Figure 3.75 Geastrum tepexensis showing central endoperidium surrounded by triangular-shaped segments of the exoperidium (Cenozoic). Bar  7.5 mm. (Courtesy S. R. S. Cevallos-Ferriz.)

found in the structure. Although the morphological features of the fossil make it difficult to place within a modern genus, this discovery does expand the geographic distribution of the Lycoperdales into the tropics during the Cenozoic. Extending the range further back into the Cretaceous is Geastroidea lobata, a compressed earthstar from Mongolia (Krassilov and Makulbekov, 2003). OTHER FUNGAL REMAINS Another group of fossil fungal remains includes the asexual stages (anamorphs) of Ascomycota and Basidiomycota; these have historically been called deuteromycetes, fungi imperfecti, conidial, or mitosporic fungi. Today most of these organisms are saprotrophs or weak parasites of terrestrial plants and a few aquatics. Since information on their sexual reproduction is incomplete, they are therefore placed in artificial groups, which are based on conidial characters, that is, asexual reproduction. Many of the small spores recovered from palynological samples represent conidiospores of these fungi. The Coelomycetes are an artificial group known from temperate and tropical regions today. One form from the Cretaceous of Japan is Archephoma cycadeoidellae (Watanabe et al., 1999). It consists of mature pycnidia up to 250 μm wide that contain numerous smooth, aseptate conidia. In another form the conidia are elliptical and divided by a dark septum.


Paleobotany: the biology and evolution of fossil plants

Other coelomycetes have been reported as epiphytes on dicot and grass leaves preserved in amber (Poinar, 2003). Although paleomycology represents a very old subdiscipline of paleobotany, for reasons noted earlier almost all of the studies to date have focused on the description and identification of the fungi. The study of fossil fungi does, however, provide the opportunity to examine several levels of biotic and abiotic interactions that extend beyond the description and classification of the organisms (Stubblefield and Taylor, 1988). Studies of fossil fungi may examine symptoms of the host as well as the fungi themselves, and can reveal different levels of interaction, as well as providing indirect evidence of coevolutionary relationships between fungi and land plants. In some instances these may be beneficial levels of biological cooperation (e.g., mycorrhizae, lichens), whereas in others they indicate parasitic and pathogenic associations. Following are several examples of such interactions that have been determined based on the fossil record of fungi. Fungal Life-History Strategies

SAPROTROPHISM One of the most obvious and essential activities of fungi today is the degradation of plant and animal tissue. Without this activity, life on Earth would cease. As a result of the carbon cycle, atmospheric CO2 is fixed into organic molecules in plants via photosynthesis. Then after the plant dies, it is ultimately degraded by fungi (and bacteria); this process in turn releases the CO2 back into the atmosphere. One of the oldest examples of saprophytism comes from Late Devonian specimens of the progymnosperm wood morphotaxon Callixylon (Stubblefield et al., 1985a). Inside many of the secondary xylem tracheids are numerous hyphae (FIG. 3.76), some branching and exhibiting intercalary swellings (FIG. 3.77). The surface of some filaments is smooth, whereas others have numerous rounded knobs. In addition to these specimens in the wood, some of the cells of the vascular rays contain spherical structures, which were interpreted as fungal resting spores or ergastic deposits. In some areas of the wood, the tracheid walls are characterized by erosion troughs, which represent areas that have been delignified by the enzymatic activities of the fungus. The patterns produced on the tracheid walls in Callixylon are similar to those formed by living basidiomycetous fungi responsible for white rot (Otjen and Blanchette, 1986). In this particular example, the fossil fungi cannot be identified with certainty, but features of the degradation process provide information about the level of fungus– plant interaction. The production of lignified, secondary tissues (wood) by vascular plants had evolved in some groups by the

Figure 3.76 Callixylon tracheid containing fungal hyphae

(Devonian). Bar  35 μm.

Givetian (Middle Devonian). Today, fungi continue to be the primary decomposers of lignin in the ecosystem. The presence of wood-rotting fungi in Callixylon, one of the oldest known trees (Meyer-Berthaud et al., 1999), provides compelling evidence that this important saprophytic association between fungi and vascular plants arose around the same time that wood development began. Fossil evidence of this association also provides a proxy record of the type of decay process and the basidiomycetes that were responsible for it. For example, white rot (degradation of both cellulose and lignin) and brown rot (degradation of only the cellulosic part of the wall) can be distinguished by observing tracheid cell walls in fossil wood. As noted earlier, wood-rotting fungi in several Paleozoic and Mesozoic woods from Gondwana produced symptoms of white pocket rot similar to those seen in extant wood. One of the interesting aspects of the fungi responsible for white pocket rot is their apparently long geologic history. The woody plants they infected in the Paleozoic and Mesozoic are long extinct, but the fungi appear to have produced the same features in

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.78 Seedling of Picea baltica with Gonatobotryum piceae (arrow) at the base of the unfolding cotyledons (Eocene). Bar  1 mm. (Courtesy A. Schmidt.)

Figure 3.77 Septate hypha with terminal chlamydospore in

Callixylon wood (Devonian). Bar  12 μm.

wood for several hundred million years. Does this mean that the relationships between certain types of fungi and woody plants coevolved as the fungi continued to adapt to new plant hosts? Perhaps the enzymatic system responsible for selective wood degradation evolved several times or in multiple groups of fungi, or perhaps the microenvironment of wood is so similar in all gymnosperms that fungi, once adapted to this habitat, showed little change. It may be possible to answer questions such as these in the near future by using molecular or isotopic techniques applied to modern and fossil wood-rotting fungi. PARASITISM Simply stated, parasites live at the expense of other organisms (FIGS. 3.78, 3.79). Living fungi demonstrate a variety of interactions with their hosts, including those that obtain nourishment without causing death (biotrophs), those that absorb nutrients from dead tissues (necrotrophs), those that cause disease (pathogens), those in which the partners provide mutual benefits (mutualists), and a variety of intermediate levels of interaction. Of all the potential levels of

Figure 3.79 Gonatobotryum piceae, young conidiophores

showing apical conidiogenesis (Eocene). Bar  50 μm. (Courtesy A. Schmidt.)

interaction between fungi and other organisms, parasitism is perhaps the most difficult to demonstrate and distinguish from saprotrophism in the fossil record. Without clear evidence of some host response to the infection, this type of interaction appears similar to saprophytism and even mutualism when examined in fossils. In some instances, the plant host may show several different responses to fungal invasion, further complicating our understanding of specific interactions (Oliver and Ipcho, 2004). One of the best examples of host response in living plants involves the production of swellings or appositions on the inner surface of cell walls (FIG. 3.80). In extant plants, such structures form in response to the invasion of fungi and are regarded as a mechanism by the host to isolate the fungus from uninfected cells nearby. Examples of cell wall appositions in tissues heavily infected with fungal hyphae are known from Late Pennsylvanian coal ball material of Illinois (Stubblefield et al., 1984a). In the Rhynie chert plant, Nothia aphylla, three different fungal endophytes were found in the plant axes (FIGS. 3.81, 3.82)


Paleobotany: the biology and evolution of fossil plants

Figure 3.80 Section of Araucarioxylon wood showing wall appositions (arrows) (Triassic). Bar  50 μm.

Figure 3.81 Cortical cell in rhizoidal ridge of Nothia aphylla containing numerous fungal spores (Devonian). Bar  30 μm.

(Krings et al., 2007a). Parts of the rhizome have hypodermal cells with thickened walls (FIG. 3.83) (Krings et al., 2007a). Other parts show areas that are devoid of cells (FIGS. 3.84, 3.85), or include degraded cells, suggesting that the host may have responded to infection through programmed cell death. Both responses are seen in living plants and function to prevent further spreading of the parasite. The production of resinous material (FIG. 3.86) is another response to fungal invasion of living plant tissue. This type of host response has been noted in wood of Callixylon newberryi from the Upper Devonian and may be more widespread in fossil woods, but as yet underreported. Pathogenic fungi are believed to be responsible for the symptoms found in Late Triassic tree trunks in the Petrified Forest of Arizona (Creber and Ash, 1990) (FIG. 3.87). In cross section the fossil wood shows numerous tubes associated with

Figure 3.82 Nothia aphylla rhizoid with swollen area contain-

ing endophytic fungi (Devonian). Bar  60 μm.

Figure 3.83 Partial section of Nothia aphylla rhizoidal ridge and host response to fungal attack in the form of zigzag line (arrows) of secondarily thickened cell walls (Devonian). Bar  500 μm.

Figure 3.84 Nothia aphylla rhizoidal ridge axis showing rhizoids and space where tissue is lacking (Devonian). Bar  60 μm.

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.85 Section of Nothia aphylla rhizoidal ridge showing

void in response to fungal attack (Devonian). Bar  100 μm.

Figure 3.87 Geoffrey Creber (left) and Sidney R. Ash.

Figure 3.86 Ray parenchyma cells of Callixylon newberryi containing globules (Devonian). Bar  25 μm. (From Stubblefield et al., 1985a.)

areas of disrupted cells. Similar symptoms are known in living trees and are associated with the extant fungi Heterobasidion and Armillaria. The large number of trees with similar symptoms at the same stratigraphic level suggests that perhaps the forest was attacked on a large scale, much like the action of modern Dutch Elm disease (Creber and Ash, 1990).

Another interesting host response which has been observed in fossils is hypertrophy, an abnormal increase in cell size. A charophyte alga (Chapter 4) from the Rhynie chert, Palaeonitella cranii, shows greatly enlarged cells in response to infection by the fossil chytrid Krispiromyces discoides (FIGS. 3.88, 3.89) (Taylor et al., 1992). The consistent presence of this chytrid embedded in the walls of the hypertrophied algal cells indicates that the alga was alive when infection took place, and that the abnormal increase in cell size represents a host response (FIG. 3.90). Interestingly, identical examples of hypertrophy have been reported in the extant charophyte Chara (Karling, 1928). Other forms of chytrid parasites are reported in and on various other fungi in the Rhynie chert (FIG. 3.91) (Hass et al., 1994). For example, chlamydospores associated with the AM fungus Glomites rhyniensis contain papillae extending from the inner surface of the spore wall (FIG. 3.92) which are identical to those produced in living glomeromycotan (FIG. 3.93) spores attacked by chytrids (Boyetchko and Tewari, 1991). Other spores have minute holes like those produced by certain actinomycetes (Lee and Koske, 1994). These examples of mycoparasitism (FIG. 3.94) underscore just a few of the microbial dynamics that occurred in and around the freshwater pools of the Rhynie ecosystem in the Early Devonian.


Paleobotany: the biology and evolution of fossil plants

Figure 3.88 Section of hypertrophoid cell of Palaeonitella with chytrid (Krispiromyces discoides) (arrow) penetrating cell wall (Devonian). Bar  80 μm.

Figure 3.90 Two hypertrophoid cells of Palaeonitella cranii. Arrow indicates normal cell size of axis (Devonian). Bar  275 μm.

3.89 Krispiromyces (Devonian). Bar  25 μm.






Fungal parasitism of fossil plants can also be inferred based on particular characteristics of the fungus. For example, in certain epiphyllous fungi growing on middle Eocene (Paleogene) angiosperm leaves from the Claiborne Formation of Tennessee, haustoria and haustorial pores were found in many of the leaves (Dilcher, 1965). Although the possibility exists that these fungi obtained their nourishment

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.91 Spore tightly packed with coenocytic hyphae.

Arrow indicates possible discharge papilla of fungus (Devonian). Bar  10 μm. (From Hass et al., 1994.) Figure 3.93 Chlamydospore of extant glomeromycotan fungus

showing host response in the form of wall papillae. Bar  100 μm.

Figure 3.92 Chytrid (arrow) on surface of chlamydospore.

Note host response in form of papilla inside spore wall (Devonian). Bar  20 μm.

from some other source (e.g., host leaf excretions or exudates from animals), the presence of haustoria like those of modern parasitic fungi, and the reaction of the leaf to this pattern of penetration, suggests that these fungi parasitized the angiosperms on which they grew. MUTUALISM Mutualistic interactions are those symbioses in which both partners benefit from the relationship. Most of those that have been reported in the fossil record involve mycorrhizal fungi

Figure 3.94 Parasitic or saprotrophic fungi inside glomeromy-

cotan spore (Devonian). Bar  80 μm.

and vascular plants; lichens are another example of a mutualistic association (discussed later). For endomycorrhizal fungi, the existence of such interrelationships in the fossil record has been based on the presence of non-septate hyphae


Paleobotany: the biology and evolution of fossil plants

and various forms of vesicles and chlamydospores within the underground rhizomes or aboveground prostrate axes of permineralized vascular plants. Since their initial description in tissues of the Rhynie chert plants, the occurrence of variously shaped spores and hyphae in these plants has been used as the basis for establishing the early occurrence of endomycorrhizal associations. Based on the presence of such structures, Pirozynski and Malloch (1975) hypothesized that such fungal–plant interactions were necessary for the establishment of plants on the land (Chapters 6, 8). According to these authors, the fungi in this mutualistic association would be provided with a carbon source, whereas the land plants would benefit from increased nutrients and more efficient water uptake from the substrate. Some of the chlamydospores and hyphae that are so common in the Rhynie chert plants may also represent the remains of saprophytic fungi (Taylor and White, 1989). Others, however, confirm the existence of arbuscular mycorrhizae based on well-defined arbuscules in a restricted zone of the cortical tissues (Taylor et al., 2005c) in both sporophytes and gametophytes (FIG. 3.95). Although a morphological structure

cannot be used to conclusively demonstrate a physiological function in a fossil plant, the direct correspondence between arbuscule formation in extant and fossil plants, including location within the host plant and morphology of the arbuscules (FIG. 3.96), is striking. Other forms of evidence, such as molecular sequence data calibrated to molecular clock assumptions, also support the existence of endomycorrhizal symbioses by the Early Devonian (Simon et al., 1993). It is hypothesized that one selective advantage of mycorrhizae is the ability to increase the plant’s uptake of phosphorus via the extended hyphal network. This certainly may have been an important attribute in what must have been a nutrientpoor substrate during the Early Devonian of the Rhynie site. Unfortunately the earliest land plants (Cooksonia-type organisms) are preserved as impressions or compressions, and so their mycorrhizal status is unknown. Molecular biology does suggest, however, that the Glomeromycota extend well back into the Paleozoic (Berbee and Taylor, 2001; Taylor and Berbee, 2006), and even into the Precambrian based on some accounts (Heckman et al., 2001). By conservative estimates this would mean that by the time of the Early Devonian Rhynie ecosystem, fungal relationships with plants were well established, and this is borne out by morphological and structural features of both the host and the fungus. Generally the underground organs of fossil plants, unless they are petrified or permineralized, are not studied in great detail. There is sufficient permineralized plant material available today, including roots, however, from many different geologic horizons and representing most of the major groups of plants, that a systematic study looking for endomycorrhizae

Figure 3.95 Hypha (arrow) extending through gametangiophore stalk of Lyonophyton rhyniensis (Devonian). Bar  35 μm.

Figure 3.96 Detail of arbuscules in Aglaophyton major corti-

cal cells (Devonian). Bar  30 μm.

CHAPTER 3 fungi, bacteria, and lichens

could be successfully undertaken. Although it is estimated that 90% of all living plants enter into some type of mycorrhizal symbiosis, there remains little information about the spatial and temporal distribution of mycorrhizae in geologic time. For example, we know a great deal about Carboniferous coal swamp plants. Were they all mycorrhizal, and if not, why not? What is the distribution of mycorrhizae in major groups of plants which are anatomically preserved, that is either as petrifactions or as permineralizations? These and many other questions can have important implications in deciphering the evolution of certain types of paleoecosystems, as well as tracking fungal-plant interactions through time.


1982, 1984). From Cenozoic amber collected in Mexico, fungi have even been described inside fossil nematodes (FIG. 3.98) (Poinar, 1984; Jansson and Poinar, 1986). Schmidt et al. (2007; 2008) have reported carnivorous fungi preserved in Cretaceous amber that possess hyphal rings

Fungi–Animal Interactions

Despite the obvious role fungi play in modern ecosystems as decomposers, parasites, and mutualists, the interactions between fungi and animals are not extensively documented in the fossil record. A common fungus–animal interaction known from the Paleozoic to the recent is evidence of borings made by endolithic fungi (also algae and bacteria) in calcium carbonate skeletons of marine invertebrates (Gatrall and Golubic, 1970; Grahn, 1981). This represents a trace fossil (or ichnofossil), in that no organic material of the fungus remains, only trace evidence in the form of damage to shells, and so on. Another early example of fungus–animal interactions involves middle Silurian fungal remains from Sweden (Sherwood-Pike and Gray, 1985). These rocks contain spindle-shaped aggregates of hyphae associated with amorphous material up to 260 μm long. The fungal remains are interpreted as frass (fecal pellets) produced by a microarthropod. Another explanation is that the arthropod was using the fecal remains as their primary food source. This example demonstrates the difficulty in determining nutritional modes and interactions in paleoecosystems. In some cases, the morphological similarity between a fossil and extant fungus can be used to infer the nutritional mode and degree of interaction. Geotrichites glaesarius represents a conidial fungus, that is, a fungus that forms external, asexual spores of a particular type (Stubblefield et al., 1985b). It was found on the surface of a partially decomposed abdomen of a spider (FIG. 3.97) preserved in late Oligocene–early Miocene amber from the Dominican Republic (Stubblefield et al., 1985b). The fact that the fungus had not invaded the body cavity of the spider and the pattern of conidial formation suggest that this interaction was saprotrophic. Two other reports of entomophthoralean (Zygomycota) fungi from the Dominican amber include a parasitic infection of a winged termite and a fungus, found on a fossil ant, which resembles modern members of an insect pathogen (Poinar and Thomas,

Figure 3.97 Chains of conidia attached to a spider leg preserved in Dominican amber (Miocene). Bar  50 μm.

3.98 Nematode containing fungus (Oligocene). Bar  15 μm. (From Jansson and Poinar, 1986.)



Paleobotany: the biology and evolution of fossil plants

that serve to trap nematodes (FIGS. 3.99–3.101). Associated with the fungi are small nematodes that probably represented the prey. Modern carnivorous fungi are found in the Zygomycota, Ascomycota, and Basidiomycota, and the presence of these forms in amber indicates that complex devices to trap motile organisms had evolved by the early Cretaceous. Although these reports are important in documenting cases of specific interactions in the fossil record, there are still too few reports currently available from older rocks to make any substantive comment regarding the evolution of these complex interactions. The associations between fungi and animals are perhaps nowhere more unusual than those known from the

Trichomycetes. Today, trichomycetes inhabit the lower digestive tracts of various types of insects and other arthropods, and based on molecular sequences, it is suggested that many groups are polyphyletic (White, 2006). These endosymbiotic microfungi live in freshwater and include more than 130 species; however, this probably represents a fraction of the total number of living forms. In this obligate mutualistic association, the fungi are not capable of existing outside the host gut. The only fossil trichomycete known to date is a specimen from the Triassic of Antarctica (White and Taylor, 1989b). It consists of a small fragment of presumed arthropod cuticle to which are attached numerous, elongate thalli (FIG. 3.102), each anchored by a holdfast cell. At the distal end of each

Figure 3.101 Reconstruction of the carnivorous (nematode

trapping) fungus (Cretaceous). (From Schmidt et al., 2007.) Figure 3.99 Carnivorous fungus, trapping ring (Cretaceous).

Bar  10 μm. (Courtesy A. Schmidt.)

Figure 3.100 Carnivorous fungus, yeast-like growth forming along a hypha (Cretaceous). Bar  10 μm. (Courtesy A. Schmidt.)

Figure 3.102 Palisade organization of thalli on the inner sur-

face of putative arthropod cuticle (Triassic). Bar  100 μm.

CHAPTER 3 fungi, bacteria, and lichens

thallus is a small plug and numerous spores. Although the diagnostic features of the cuticle necessary to identify the host are not present, the occurrence of numerous fecal pellets associated with plant tissues from the same site substantiates the existence of arthropods in this Triassic ecosystem (Kellogg and E. Taylor, 2004). Fungal remains have also been documented in a variety of coprolites (FIG. 3.103) and have been especially useful in Quaternary palynology (Davis, 2006 and references therein). In some examples the producers of the coprolites are believed to have been saprobes dwelling on all sorts of organic particles, whereas in other instances a high percentage of the coprolites contain fungal remains, suggesting that the producers were true fungivores (Pratt et al., 1978). Fungi have been reported in various types of dinosaur coprolites, and the presence of certain types of leaf-borne fungi has been used to infer a foliage diet for these animals (Sharma et al., 2005). An interesting report of partially decayed wood found in coprolites of herbivorous dinosaurs (Chin, 2007) suggested that the animals may have eaten wood that had been partially rotted by fungi when there were few other food sources available. The wood would only be useful as a nutrient source after fungi had begun to break down the lignin and cellulose within it. The report and description of fungi associated in other biotic interactions (e.g., additional types of coprolites and a variety of different types of borings) in Miocene wood is an important contribution that will help to underscore the geological history of fungal-animal interactions (Sutherland, 2003). An unusual interaction between fungi and animals has been documented from the Jurassic (Martill, 1989). On the surface of fossil fish teeth are a series of meandering

Figure 3.103 Coprolite composed of hyphae (Devonian).

Bar  50 μm.


borings that are restricted to the surface enamel. These have been given the name Mycelites enameloides. Similar borings have been reported on teeth, scales, and bone as early as the Devonian (Gouget and Locquin, 1979). The report of fungal–animal interactions and associations in the fossil record is a very important area in paleomycology. Although many of the associations may be the activities of fungal saprotrophs, others may represent early stages of symbiotic interrelationships that are widespread in modern ecosystems, and which represent one of the cornerstones of biodiversity today (Zook, 2002). Geologic Activities of Fungi

Throughout the geologic past, fungi have played an important role in their interactions with the depositional environment, and together with cyanobacteria, are today receiving increasing attention because of their role in shaping the geosphere. The discipline of geomicrobiology involves the study of organisms, their interactions, and the materials that they colonize (Konhauser, 2007). Certain fungi are known to degrade and liquefy coal (Sterflinger, 2000), and they are the only organisms that can completely break down lignin. Fungi in mycorrhizal and lichen symbioses are also involved in weathering rock, a form of biological weathering (Landeweert et al., 2001; Hoffland et al., 2004). Fungi in very cold or desert environments appear to be important as rock colonizers (Selbmann et al., 2005). Fungi are involved in the alteration of several minerals, such as carbonates and silicates; for example, fungi can be involved in both the formation and destruction of carbonate deposits (Sterflinger, 2000). The roles of fungi in geological activities are only beginning to be fully appreciated as the focus in the past has been mostly on bacteria (Gadd, 2007; Perri and Tucker, 2007). One type of fungal activity seen in the fossil record includes borings in rock. Since there are several groups of organisms, however, that are borers, including cyanobacteria, algae, lichens, sponges, bryozoans, gastropods, and several other invertebrate groups, distinguishing the organism responsible for the boring is often very difficult (Elias and Lee, 1993). In one study on endolithic fungi from Jurassic rocks, scanning electron microscopy was used after resins casts were made of the molds (Gatrall and Golubic, 1970). Based on comparison with extant forms, it was determined that these borings were the result of fungi, although many earlier contributions had regarded such trace fossils as the result of algal borers. Fungal and algal borers may also be involved in a process termed micritization, in which carbonate particles are reduced in size and combined with chemically precipitated


Paleobotany: the biology and evolution of fossil plants

carbonate to produce a finely crystalline form of carbonate, also called calcite mud. Biogenic micritization, which is mediated by organisms, has been noted in rocks dating back to the Paleozoic; in a few cases the filaments of the endolithic borers are preserved in the rocks, but in most cases their presence can only be inferred (Harris et al., 1997). Endolithic filaments are known from the Upper Silurian of Morocco (Barbieri et al., 2004). Although fungi are involved in some of these structures, many are also mediated by various types of microorganisms (cyanobacteria and algae), which produce micrites as byproducts of their metabolism. In some instances these structures demonstrate a faint banding pattern that suggests they may have been produced during different seasons. Fungi have also been important in the geologic record through their activities in biomineralization—the synthesis of minerals from simple compounds by organisms. For example, the occurrence of needle-fiber calcite in Carboniferous paleosols (fossil soils) has been regarded as evidence of ectomycorrhizal associations in the ecosystem (Wright, 1986). Certain extant vascular plant roots accumulate secondary soil carbonate through biomineralization. Some of these roots have larger diameters because of calcification of cortical parenchyma cells. It is suggested that carbonate biomineralization and acid extrusion in these root cells may represent a mechanism used by plants to acquire soluble nutrients from the rhizosphere, or a method to live in soils with excessive calcium (Kosir, 2006). As paleobotanical data are further integrated with sedimentology, it may be possible to document such an adaptive feature in the roots of certain fossil plants.

been overlooked. If this is not the case, perhaps the plants possessed chemical substances that deterred epiphyllous fungi, or fungi simply may not have evolved an epiphyllous habitat by the Paleozoic. Stomiopeltites is perhaps the oldest (Early Cretaceous), well-documented member of the Dothideales (Pezizomycotina, Ascomycota), a group that includes numerous epiphyllous taxa (Alvin and Muir, 1970). It consists of dome-shaped plectenchymatous thyrothecia (a type of ascocarp with a palisadelike layer of hyphae), each bearing a small ostiole. Hyphae are small (1.7–3 μm in diameter) and form the layered wall of the thyrothecia. Pycnidia (asexual reproductive structures) are common and appear to have been borne directly over a stoma on the leaf surface. No spores were described in these specimens. This fungus is unique, not only because of its age but also because it is one of the few fossil epiphyllous fungi known to occur on conifer leaves (Frenelopsis). Stomiopeltites (FIG. 3.104) is also known from the Upper Cretaceous (Cenomanian) of France, also on Frenelopsis leaves (Pons and Boureau, 1977) (FIG. 3.105), and from the Miocene Clarkia locality of Idaho, USA, on leaves of three angiosperm genera (Phipps and Rember, 2004). Numerous structurally preserved ascomycetes have been described from Cretaceous conifer leaves of the Brachyphyllum-type (Van der Ham and Dortangs, 2005). The fossil fungi (FIGS. 3.106, 3.107) are thought to be

Epiphyllous Fungi

Mesozoic and younger rocks contain numerous examples of ascomycetous fungi (Taylor, 1994), many of them preserved on compressed leaves. Epiphyllous fungi are commonly encountered on many forms of vegetation as early as the Cretaceous, and probably far earlier, although few good examples exist. Despite the tremendous diversity of foliage types during the Carboniferous, for example, there are very few reports of fungi associated with the leaves. One example is reported in cuticle preparations of rachides and pinnules of Callipteridium, a pteridosperm from the Stephanian of France (Krings, 2001). The fungus consists of septate hyphae, but no reproductive structures were found. In preservation of this type, it is especially difficult to determine if the fungus was associated with the plant when it was alive, or simply represents a saprotroph that colonized decaying plant remains. As is true for many facets of paleomycology, epiphyllous fungi in the Paleozoic may exist, but have simply

Figure 3.104 Epiphyllous fungus Stomiopeltites amorphos (Miocene). Bar  50 μm. (Courtesy of C. J. Phipps.)

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.107 Pteropus brachyphylli hypostroma (Cretaceous).

Bar  40 μm. (From Van der Ham, 2005.)

Figure 3.105

Denise Pons.

Figure 3.106 Pteropus brachyphylli fruiting body (Cretaceous).

Bar  100 μm. (From Van der Ham et al., 2005.)

closely related to the extant Phaeocryptopus, which infects the leaves of a number of extant conifers, thus suggesting specific parasitic relationships with certain conifers since the Late Cretaceous. Trichopeltinites is a microthyriaceous (Dothideales) fungus that can be traced back to the latest Cretaceous (Maastrichtian). Specimens from rocks of the Cretaceous–Paleogene boundary in Alberta, Canada, suggest that this fungus can be correlated with the accumulation of organic matter (coal) under wet conditions, rather than with a particular paleolatitude (Sweet and Kalgutkar, 1989). Two interesting ascomycetes were found within the permineralized leaf of an extinct palm from the Eocene Princeton chert (Currah et al., 1998). Palaeoserenomyces allenbyensis is compared to modern Serenomyces, which forms leaf spots (so-called tar spots) on palms. Within the locules of the stromata of Palaeoserenomyces are globose ascomata of a mycoparasite, Cryptodidymosphaerites princetonensis. This mycoparasite appears similar to the extant form, Didymosphaeria, which today is a mycoparasite on ascomycetes that form stromata (a type of pseudoparenchymatous fruiting body). The Vizellaceae (Dothideales) is a small family of epiphyllous ascomycetous fungi that can be traced into the Eocene. It is one of only a few families of fungi that can be identified in the fossil record solely on the basis of non-reproductive (vegetative) features, as it has distinctive septate hyphae with regularly alternating dark and hyaline segments (Phipps, 2007). One of the most comprehensive studies of fossil epiphyllous fungi is the contribution by Dilcher (1965), who described a large number of fungi found in association with


Paleobotany: the biology and evolution of fossil plants

middle Eocene angiosperm leaves, including several members of the Vizellaceae. In several instances it was possible to reconstruct the entire life history of the fungus, including both the sexual (telemorphic) and asexual (anamorphic) phases, thereby making assignment to modern groups even more reliable. One of the more common forms in this biota is Vizella memorabilis, which is characterized by hyphae with alternating short and long cells (Selkirk, 1972). In the Eocene clay pit (Claiborne Formation) where this fungus was collected, it is most commonly found on the upper surface of angiosperm leaves of the genus Sapindus. During the growth of the fungus the hyphae dichotomize to form colonies that may be up to 450 μm in diameter. Irregular proliferations of cells form a short, lateral branch that produces two types of fruiting bodies. The larger fruiting bodies (ascocarps) produce two-celled spores, whereas the smaller ones (pycnidia) produce single-celled spores. Up to 60 ascospores are produced in each ascocarp, which are 10  14 μm. Pycnidiospores are oval, unicellular, and oriented in chains. Upon germination, the pycnidiospores form small germinal tubes that penetrate the leaf epidermis. Fossil fungi of this general type, but which lack spores, are often assigned to the genus Entopeltacites (FIG. 3.108) (Phipps, 2007). Epiphyllous fungi of the Vizella type are morphologically identical with the modern genus Manginula, a form that was included within the Microthyriales. In addition to determining stages in the life history of several of these Eocene epiphyllous fungi, Dilcher was able to document various levels of host specificity. For example, V. memorabilis was found on a number of different types of leaves, whereas others, such as Meliola anfracta, appeared consistently only on certain leaf genera. In other instances, fungi appeared to be specific for only one surface of the leaf. Determining host specificity for particular types of fossil fungi is a difficult and long-term endeavor that will require careful documentation of both the fungus and host plant from many localities and stratigraphic levels (Venkatachala and Kar, 1968). Nevertheless, it represents an important line of research that can provide new data on the evolution and variety of fungal interactions in the fossil record. Identification of the spores produced by certain types of Eocene fungi is also important, as it provides a method of correlating sediments where fruiting bodies are not preserved (Sheffy and Dilcher, 1971). A rich assortment of fossil fungi in a flora, such as the ones described from the Eocene of Tennessee, represents an important potential source of information on the paleoclimate of a region (SherwoodPike, 1988). Modern microthyriaceous fungi are principally tropical in distribution. The most important limiting factor

Figure 3.108 Epiphyllous fungal hyphae of Entopeltacites

remberi (Miocene). Bar  50 μm. (Courtesy C. J. Phipps.)

in their geographic distribution, however, appears to be precipitation, not temperature, and these fungi can be correlated with regions where mean annual precipitation reaches 100 cm/yr (Elsik, 1978). On the basis of leaf types present at the Tennessee site, it has been suggested that the region was subtropical during the middle Eocene. A comparison of the fossil fungi with the ranges of comparable modern forms (the Nearest Living Relative method applied to fungi) supports the theory that this area was a low-lying, coastal region characterized by a moist, subtropical climate. It may be possible in other fossil floras to utilize various forms of fungi to provide more accurate information on paleoclimatic regimes (Lange, 1978), especially in those instances where modern fungi can be used as analogs for microhabitat determination. Another fungus that has been used to infer paleoclimatic information is Meliolinites dilcheri (Daghlian, 1978a). This form, from the early Eocene of Texas (Rockdale Formation), consists of colonies 2 mm in diameter (FIG. 3.109). The hyphae (FIG. 3.109) produced numerous short, lateral, twocelled branches consisting of a short stalk and a capitate head. The fossil fungus is included within the Meliolaceae, the modern members of which are parasitic and common in warm, humid, forested tropical areas today. Phragmothyrites (FIGS. 3.110, 3.111) is another fossil epiphyllous fungus; it consists of ascomata with radiating, pseudoparenchymatous hyphae (Phipps and Rember, 2004). Although there are numerous reports of fossil epiphyllous fungi on Cenozoic leaves, the geologic antiquity of this plant-fungal interaction remains relatively unexplored. The increasing attention to documenting fossil fungi over wide geographic areas, however, is a positive step (Herbst and Lutz, 2001). It has been proposed that leaf-associated

CHAPTER 3 fungi, bacteria, and lichens


Figure 3.109 Colony of the epiphyllous fungus Meliolinites dilcheri with radiating hyphae and hyphopodia on cuticle of Eocene angiosperm leaf. Bar  50 μm. (Courtesy C. P. Daghlian.)

Figure 3.111 Detail of epiphyllous fungus Phragmothyrites

concentricus (Miocene). Bar  50 μm. (Courtesy C. J. Phipps.)

Another interesting approach that has been used to identify the presence of epiphyllous fungi in fossils is based on the analysis (gas chromatography–mass spectrometry) of lipid fractions extracted from uninfected and infected conifer twigs of Cretaceous age (Nguyen et al., 2000). This study suggests that it is possible to determine lignin degradation products in the fossils produced by the fungi even when the fungi themselves are not present. Fungal Spores Figure 3.110 Several Phragmothyrites concentricus thalli

(Miocene). Bar  25 μm. (Courtesy C. J. Phipps.)

fungi did not arise until the Cretaceous, perhaps in association with the early radiation of flowering plants (Pirozynski, 1976b). The presence of laminar leaves as early as the Middle Devonian, however, indicates that there was a phylloplane microhabitat available for fungi much earlier. In addition, the discoveries of ascomycete-like fungi as early as the Silurian (Sherwood-Pike and Gray, 1985) provide evidence that this group may have evolved far earlier than we previously thought. Perhaps the early relatives of leafinhabiting fungi occupied other types of habitats. It is also possible that fungi became associated with leaf surfaces far earlier than the Cretaceous, but they have not been found. There is some evidence that the methods used to separate fossil leaf cuticles from the rock and prepare them for study may be destroying epiphyllous fungi.

Although fungal spores are known throughout geologic time, it is only beginning in the Late Jurassic that they constitute a significant and important fraction of the palynomorphs recovered from most rocks (Elsik, 1976). Spores of obvious fungal origin, however, are known as early as the Late Silurian. There is an extensive terminology applied to fungal palynomorphs based on living fungi (Elsik et al., 1983). Spores are often named based on the concept of morphogenera (Elsik, 1989), in which various morphologic features (e.g., size, shape, number of septations or pores, and type of ornament) form the basis of the generic concept (e.g., Fusiformisporites). In addition to spores, fungal fragments recovered in macerations may include various forms of fructifications, hyphae, and mycelia. Funginite is the name applied to sedimentary organic matter (macerals) composed chiefly of fungal material (see Chapter 1). The study of spores and other fungal remains represents an important, yet still largely untapped, paleomycological


Paleobotany: the biology and evolution of fossil plants

and paleobotanical data source. Although fungal spores have been utilized in some stratigraphic applications in Cenozoic rocks, especially around the Cretaceous–Paleogene boundary (Vajda and McLoughlin, 2004), perhaps their greatest untapped potential lies in the areas of biogeography, paleoecology, and paleoclimatology (Sherwood-Pike, 1988). Jarzen and Elsik (1986) examined fungal palynomorphs from a modern, subtropical environment of open savanna and riparian forest in eastern Zambia. The results of that study provide a basis for quantifying the various fungal remains preserved in modern fluvial systems, with applications to similar environments in the fossil record. It also suggests another method for inferring paleoecological parameters when other microfossils and megafossils are lacking.

Figure 3.112 Albugo-like oogonia in Pennsylvanian seed tissues. Bar  100 μm.

Fungal-Like Organisms

There are a number of organisms that historically were classified with the fungi, but as a result of subsequent research, and especially molecular systematic approaches to modern forms, have been removed from the Kingdom Fungi. Nevertheless, each of these organisms has a phylogenetic history, and some are preserved as fossils. PERONOSPOROMYCETES (OOMYCOTA) The oomycetes, or water molds, are organisms that are now included within the Kingdom Straminipila; some authors refer to this group as the Peronosporomycetes. Most forms are filamentous, possess cell walls of cellulose rather than chitin, and lack cross walls, except where reproductive cells are produced. Asexual reproduction occurs by biflagellated zoospores. Sexual reproduction includes the production of sperm in antheridia which fuse with one to several eggs produced in an oogonium. Although there are several pre-Cenozoic reports of peronosporomycetes in association with fossil plants, many of these have been discounted because of the absence of a complete suite of diagnostic features (Johnson et al., 2002), whereas others regard at least some of these fossils as authentic (Blackwell and Powell, 2000). One interesting fossil that perhaps provides the best evidence of peronosporomycetes is Hassiella monospora (T. Taylor et al., 2006). This Rhynie chert organism consists of aseptate hyphae that randomly branch to form terminal oogonia, each approximately 30 μm in diameter. At the base of the oogonium is a funnelshaped structure interpreted as the antheridium. Intracellular peronosporomycetes (Combresomyces cornifer) in lycopod periderm have also been described and illustrated from the upper Visean cherts of central France (Krings et al., 2007e; Dotzler et al., 2008). Another possible peronosporomycete is

found in a seed-like structure from the Carboniferous (Stidd and Cosentino, 1975). Each spherical oogonium (100 μm in diameter) contains a single oosphere (FIG. 3.112); however, no well-defined antheridia are preserved. What makes this specimen so interesting and increases the chances that it represents a peronosporomycete is that the disruption of the tissues in the seed are identical to those found when the extant peronosporomycete Albugo infects flowering plants.

Eubacteria and archaea Despite the microscopic size of the cells (only a few micrometers long), in terms of numbers of individuals and metabolic diversity, bacteria are the dominant organisms of the biosphere and, geologically, the oldest organisms on Earth (Chapter 2). They live in every possible habitat—soil, water, and even the deep subsurface of the continents and oceans; they can even be found in radioactive waste. They are critical in nutrient recycling (Nardi et al., 2002), including the fixation of nitrogen from the atmosphere into a form that is usable by plants. It has been estimated that there are from 40 million to 2 billion bacteria in a single gram of soil (Whitman et al., 1998). Bacteria are classified today based on genetic sequence data. In older classification systems, five kingdoms of organisms were recognized: Protista, Monera (all prokaryotes), Fungi, Plantae, and Animalia. Although modern phylogenetic work has long since shown that this system is out of date for the prokaryotes, it is still reproduced in some textbooks. The most widely used classification of organisms today includes three domains (a level above the level of

CHAPTER 3 fungi, bacteria, and lichens

Kingdom): Eubacteria, Archaea, and Eucarya (Woese and Fox, 1977; Woese et al., 1990). Within the Domain Eubacteria are the Cyanobacteria (including the Chloroxybacteria), and other bacteria (e.g., purple, green sulfur, Gram-positive, spirochetes), with each group occupying the level of kingdom. All possess prokaryotic cells, which differ from the eukaryotic cells of plants, animals, and fungi. Prokaryotic cells are small (0.5–10 μm), and lack a nucleus or other membrane-bound organelles. Neither meiosis nor mitosis occurs; prokaryotes reproduce asexually by means of binary fission. All the nutritional modes used by eukaryotic cells are also found among the prokaryotes, plus several that are unique to the group. Despite the fact that bacteria are very small (0.5–5 μm) and delicate, there is an excellent fossil record of the group dating back to the Archean, with structurally preserved specimens reported as old as 3.7 Ga. All morphologic forms (coccoid, bacilloid or rod-shaped, spiral, and filamentous) have been recognized in a variety of configurations that correspond to those of existing types. They occur as structurally preserved and compressed forms within most types of common mineral matrices and have also been found in fossil vertebrates, invertebrates, plant tissues, and coprolites (fossil fecal material). Archaea

The application of molecular tools has been useful in establishing relationships and the divergence of major groups within the prokaryotes. Using 16S rRNA phylogenetic trees, Woese and Fox (1977) identified a group of prokaryotes called the Archaea. Although they had many features in common with other prokaryotes, they differed in certain genetic processes, and are generally considered to be more closely related to the Eucarya than to the Eubacteria. Today, many archeans live in some of the most severe environments on Earth and are termed extremophiles. They can be found in hot springs, deep within the ocean, in the digestive tracts of animals, and a variety of other harsh environments. Archeans include halophiles (organisms that live in high concentrations of salinity), thermophiles (live in hot and acidic environments), and methanogens (utilize hydrogen [H2] to reduce CO2 to methane). Many believe that the Archaea evolved from single-celled organisms 4 billion years ago. Many of the environments they inhabit today are believed to be the types of conditions present in the first billion years of Earth’s history, and it has been suggested that life may have arisen in these extreme habitats (Leach et al., 2006) (Chapter 2). As a result of ribosomal sequences of modern halobacteria, some suggest that these are the most primitive of the


archean group. Although no body fossils of Archaea have been confirmed to date, there is evidence of biomarkers from archeans in the fossil record. Perhaps the largest obstacle is distinguishing fossil from living Archaea, since members of these group are widespread in the deep biosphere. Eubacteria

It is currently not possible to determine which of the bacterial groups is the oldest geologically. Of the earliest microfossils known to date, all are morphologically simple and at least some were probably phototactic. Early microbiotas contained both anaerobes and photoautotrophs (Knoll, 1985a). In 1966, Barghoorn and Schopf described small rod-shaped cells from the Onverwacht Group of South Africa (3.5 Ga ago) under the name Eobacterium. Although originally thought to represent rod-shaped bacteria based on transmission electron microscopy (TEM), these structures are now regarded as non-fossil or modern bacterial contaminants (Schopf and Walter, 1983). Research with modern bacteria growing in hot springs environments is providing information about how bacteria are preserved, which in turn will aid paleobotanists in their interpretation of fossil microbes (Phoenix et al., 2005), as many ancient hot springs environments resulted in structurally preserved bacterial fossils. Bacteria from the Early Devonian Rhynie chert hot springs environment have been described as both spherical clusters and as unicells. Some of these structures appear similar to bacteria-like organisms (BLOs) that have been described inside the spores of some modern mycorrhizal fungi (Bonfante, 2003; Taylor et al., 2004; Duponnois and Kisa, 2006). Bacteria have also been reported in permineralized plants and coprolites of Carboniferous and Permian age (Renault, 1901) as circularto-elongate structures 4 μm in diameter (FIG. 3.113). These were named Micrococcus and Bacillus, but some of them represent inorganic particles of various types (Taylor and Krings, 2005). Fossil iron bacteria have been found in pyrite from Middle Pennsylvanian rocks and examined by means of replication techniques for TEM (Schopf et al., 1965). The morphological similarity between the fossil and extant types resulted in some of these fossils being assigned to modern taxa. Bacteria have also been suggested as the causal agent for various borings in vesicles of Ordovician chitinozoans (Grahn, 1981). In modern ecosystems, a number of forms of saprophytism involve bacterial decay in association with the activities of fungi (Daniel et al., 1987). Earlier in this chapter, it was noted that saprotrophic fungi have been identified as early as the Devonian, and perhaps earlier. To date the bacteria that are typically associated with certain types of


Paleobotany: the biology and evolution of fossil plants

Figure 3.114 Annella capitata (white spheres) on surface of pollen grain. Arrow indicates a cavity caused by a bacterium (Pennsylvanian). Bar  5 μm. (From Srivastava, 1976a.)

Bacterial colony. (Mississippian). Bar  10 μm. (Courtesy N. Dotzler.)

Figure 3.113

wood rot have not been identified either directly or indirectly in these fossil woods. Fossil woods, however, offer potential for future research on both the bacteria and their interrelationships with fungi during the decay process. Actinomycetes are a type of filamentous, Gram-positive bacteria. Actinomycetes have been reported from Eocene amber (Waggoner, 1994), a Carboniferous seed plant cell (Smoot and Taylor, 1983a), and Eocene dicot wood (Wilkinson, 2003). Although some of these reports suggest that the fossils are actinomycetes based on morphological features, some may represent modern contaminants or coagulated cytoplasm. One method that has been used to indicate that these organisms are fossils is the presence of the filaments growing through the crystals in the wood cells (Wilkinson, 2003). Leptotrichites is a sheathed bacterium from Cretaceous amber that resembles the modern genus Leptothrix (Schmidt and Schäfer, 2005). The fossil forms are interpreted as living in ponds in forest woodlands and became entrapped in amber flowing from the trees. Various forms of bacteria, including forms suggestive of sulfate-reducing bacteria and several types associated with decomposition, have been reported from the Eocene Green River Formation (Mason, 2005). Another source of information about fossil bacteria includes indirect evidence based on the presence of characteristic patterns or scars on sporomorphs. Some of the

microbes were extremely specific, attacking only certain layers of the pollen grain wall (Elsik, 1970). Such occurrences have been documented as early as the Carboniferous, where coccoid and bacilloid microbes were found on the surface of miospores (FIG. 3.114). Instances of fossil bacterial degradation patterns are probably far more common than reported, since palynologists typically search out only well-preserved sporomorphs showing a complement of diagnostic features. This inherent bias toward well-preserved specimens probably eliminates the documentation of more examples of microbial degradation. However, with organisms that are as small as prokaryotes, there is always the possibility of ascribing biotic properties to abiotic artifacts. A case in point involves the description of unmineralized fossil (Cretaceous) bacteria associated with scraps of organic material in lake muds. Upon reexamination, these bacteria turned out to be fluorite artifacts formed during the maceration process (Bradley, 1968). There are other potential sources of confusion regarding the authenticity of fossil bacteria, for example tapetal cell components, such as orbicules produced during microsporogenesis in certain plants, have been mistaken for coccoid bacteria. Due to the ubiquity of bacteria in the modern world, there is always the problem of specimen contamination. For example, it has been reported that fossil bacteria could be isolated and grown from Permian salt deposits (250 Ma) (Dombrowski, 1963; Vreeland et al., 2000; Satterfield et al., 2005). Others suggest that the organisms represent modern contaminants formed during the recrystallization of the

CHAPTER 3 fungi, bacteria, and lichens

salt. It has been noted by several authors that DNA decays quickly, and that no metabolic processes of any organisms could survive for so long. Several analyses have suggested that these bacteria are too similar genetically to modern taxa, so evolution would have to be extremely slow in these organisms (Nickle et al., 2002) and that bacterial spores, which can survive for long periods of time, generally have no DNA repair enzymes within them, so DNA this ancient would have been very fragmented (Graur and Pupko, 2001). The problem of modern bacterial contamination and the formation of bacteria-like artifacts during certain types of paleobotanical techniques is discussed by Edwards et al. (2006a). These studies elegantly underscore that, in dealing with fossil organisms, the pervasive distribution of microbes now and in the past can impact research results at several levels of inquiry. The complexity of the fossilization process and subsequent diagenesis, the procedures used in sample extraction and preparation, and the establishment of biogenicity all must be considered in the analysis of microbial fossils. For example, rod-shaped bacteria, together with certain types of organic compounds (polycyclic aromatic hydrocarbons), were described from a Martian meteorite collected in Antarctica (McKay et al., 1996). Subsequent research has suggested that the “microbes” are in fact the result of certain preparation techniques, and/or contaminants from melt water. What does the future hold for the study of extraterrestrial life based on fossil evidence? One approach will require the cataloging of various types of biosignatures in all forms of sediments, including how modern microbes can change the potential signature (Cady et al., 2003). CYANOBACTERIA The cyanobacteria are the most common and widespread group of photosynthetic bacteria today, and are the primary producers and initial source of free atmospheric oxygen. Their fossil record is among the oldest for any group of organisms, and can be traced back to the Archean (Golubic and Seong-Joo, 1999; Chapter 2). Combining fossil evidence with molecular data of living cyanobacteria has made it possible to hypothesize that cell differentiation (e.g., heterocysts and akinetes) occurred in this group as early as 2.450 Ga (Tomitani et al., 2006). Heterocysts are thick-walled cells where anaerobic nitrogen fixation occurs and akinetes are resting spores. What is especially interesting is that, when modern and fossil cyanobacterial communities are compared, the data suggest that there has been relatively little morphological and probably biochemical change during more than 2 billion years of Earth history (Knoll, 1985a; Sergeev et al., 2002).


Among the other bacteria, the Cyanobacteria appear to be most closely related to Gram-positive bacteria, based on molecular signatures. Cyanobacteria obtain their energy through photosynthesis and occur as unicellular, colonial and filamentous forms. They all contain chlorophyll a and accessory pigments in the form of phycobilins. Most cyanobacteria possess a mucilaginous sheath, which may be variously pigmented. Many of these ancient organisms have the ability to fix atmospheric nitrogen and, combined with their photosynthetic abilities, are thus the most nutritionally autonomous organisms on the Earth. Some extensive petroleum deposits in the world were produced by the decay and accumulation of various cyanobacteria such as Gloeocapsomorpha, a form responsible for some of the Middle Ordovician oil shales of Estonia (Foster et al., 1990) and the central United States (Iowa) (Jacobson et al., 1988). Although present throughout geologic time, cyanobacteria are a dominant component of many Precambrian biotas discussed in Chapter 2; a few additional examples of geologically younger cyanobacteria will be presented here. One indirect method of determining the activities of bacteria in the geologic record involves the identification of certain bacteriogenic isotopic sulfides (δ34S). The geochemical analysis of certain Archean rocks indicates that sulfate has been continuously present since at least 3.5 Ga (Schidlowski, 1989). The identification of the heavy sulfur isotope has also been proposed as a useful tool in determining past environmental conditions and as a stratigraphic marker at the Neoproterozoic–Cambrian boundary (Schröder et al., 2004). However, bacteriogenic sulfide patterns are also regarded as difficult to identify earlier than about 2.8 Ga. As a result, it is currently impossible to determine whether some of the geologically early sulfides were the result of sulfatereducing organisms or were inorganically formed, that is they are abiotic in origin. Activities of cyanobacteria in the fossil record can also be examined by the presence of various biomarkers in the sediments. Molecular fossils in the form of biological lipids have been recovered from 2.7 Ga shales in Australia (Brocks et al., 1999), and in slightly younger rocks from Canada. Akinetes have been reported from 2.1 Ga cherts from Gabon, Africa (Amard and Bertrand-Sarfati, 1997), and slightly younger ones from Siberian cherts that are morphologically identical to those produced by the living cyanobacterium Anabaena. Well-preserved cyanobacteria have also been described from the Lower Devonian Rhynie chert, both coccoid, for example Rhyniococcus (Edwards and Lyon, 1983), and filamentous forms, including the morphogenera Archaeothrix, Kidstoniella, Langiella, Rhyniella (Croft and George, 1958),


Paleobotany: the biology and evolution of fossil plants

Figure 3.115 Elongated tuft-like colonies of Croftalania venusta (Devonian). Bar  500 μm.

Figure 3.117 Spiny spheres (probably peronosporomycete

oogonia) that are associated with the microbial mats formed of Croftalania venusta. Bar  25 μm.

Figure 3.116 Microbial mat formed by Croftalania venusta

(Devonian). Bar  500 μm.

and Croftalania (Krings et al., 2007c). Langiella scourfieldi is a heterotrichous form consisting of a horseshoe-shaped basal portion that produced a number of uniseriate branches. At the base of each filament is a heterocyst which is typically smaller than the other cells of the sheath; at the distal end of each filament is a hair-like extension. Kidstoniella fritschii, from the same locality, contains highly branched filaments. Croftalania venusta is a sessile filamentous form that grows on sediment and submerged plant parts (FIG. 3.115). It is associated with the formation of microbial mats (FIG. 3.116), but may also occur in structured colonies where the individual filaments are aligned into flat, irregular stands or united radially into hemispherical aggregates; it may also form fan-shaped tufts. Associated with some of the mats

are spiny spheres that may represent peronosporomycete oogonia or some other microbial reproductive structure (FIG. 3.117). The Rhynie chert also provides the earliest fossil evidence for endophytic cyanobacteria in land plants in the form of Archaeothrix-type filaments that colonize prostrate axes of the land plant Aglaophyton major (Taylor and Krings, 2005). The cyanobacteria enter the axes through stomata and initially colonize the substomatal chambers, some also occur in voids of degrading sporangia of Aglaophyton. From the substomatal chambers they spread through the outer cortex and individual filaments or groups of filaments penetrate cortical cells to form coils (FIG. 3.118). Both transmission and scanning electron microscopy have provided valuable information about the morphology and ultrastructure of several fossil cyanobacteria, as well as suggesting a basis for determining stages in the growth of some fossil bacteria. Sphaerocongregus is the generic name applied to three morphologically distinct cell types collected from uppermost Proterozoic siliceous shales from southwestern Alberta, Canada (Moorman, 1974). One cell type, assigned to S. variabilis, includes coccoid cells 3–5 μm in diameter; another cell type is larger (5–6 μm) and often arranged in chains. The third morphologic category consists of globose masses 5–20 μm in diameter that are composed of small coccoid subunits. Surrounding these larger cell masses

CHAPTER 3 fungi, bacteria, and lichens

Figure 3.118 Coils of cyanobacteria inside Aglaophyton cell (Devonian). Bar  50 μm.

is a delicate envelope. Comparisons with living cyanobacteria suggest that the various morphologic types assigned to S. variabilis represent different stages in the life cycle of a single organism. Accordingly, the larger masses represent endosporangia that contain endospores, and some of the other cells represent stages in the vegetative plant body.

Lichens Lichens are unique, double organisms that consist of two unrelated components, an alga and/or cyanobacterium (photobiont) and a fungus (mycobiont). The organisms that make up the lichen live in a close symbiotic relationship in which the photobiont gains mechanical protection, increased water availability, reduced desiccation, and an improved ability to obtain nutrients from the mycelium of the fungus. The fungus, in turn, gains organic nutrients synthesized by the photobiont(s) that is, a source of carbohydrates for growth.


If an alga and a cyanobacterium are both involved, then the alga also gains a source of nitrogen. Recent research with modern lichens suggests that the relationship between the partners, although stable, may be highly variable, ranging from mutualism to parasitism. Although the thallus organization of most lichens suggests that they would be easily preserved, there are relatively few substantiated reports of fossil lichens of any antiquity. Hallbauer and van Warmelo (1974) described a putative fossil lichen from the Precambrian of South Africa under the name Thuchomyces lichenoides. It consisted of a horizontal thallus with a cortex of erect columns 5 μm high. The cortex included several zones of branched, septate hyphae suggestive of certain types of lichens. A more recent interpretation is that T. lichenoides was a parasitic, filamentous microorganism (Hallbauer et al., 1977). The occurrence of these structures in rocks that were no doubt strongly heated during diagenesis (a gold-bearing, uranium–lead-oxide conglomerate) makes their assignment as lichens perhaps less convincing. Interestingly, similar filamentous structures have been obtained in the laboratory abiotically using comparable preparation techniques (Cloud, 1976). One structure that may represent some type of lichen symbiosis is a fossil termed a biodictyon (Krumbein et al., 2003b). A biodictyon is part of a biofilm that penetrates the surface on which it grows and forms a three-dimensional net-like structure. Certain types of biofilms have been reported from as early as the Precambrian (Barghoorn and Tyler, 1965). Net-like structures have been described from phosphorites from the famous Neoproterozoic (551–635 Ma) Doushantuo Formation Lagerstätte in south China (Yuan et al., 2005). What makes these fossils so lichen-like is the presence of groups of coccoid cells (cyanobacteria or algae) within the spaces of what appears to be a net of fungal mycelia (FIG. 3.119). The specimens were deposited in a shallow subtidal environment and the site has yielded abundant algal fossils. Although the association of the coccoid cells and the hyphae of the net cannot unequivocally demonstrate a lichen symbiosis, the ordered arrangement of the cells of the two symbionts in the fossil indicates a regular and close physiological relationship between the two organisms. Molecular clock estimates from living lichens have also suggested that lichen symbioses may have existed during the Precambrian. The best known Paleozoic lichen comes from the Rhynie chert. Winfrenatia reticulata, named for Winfried and Renate Remy, is constructed of superimposed layers of aseptate hyphae that form numerous shallow depressions (FIG. 3.120) on the upper surface of the thallus (Taylor et al., 1997). Extending from the sides of the depressions are hyphae organized into a three-dimensional net (FIG. 3.121),


Paleobotany: the biology and evolution of fossil plants

Figure 3.119 Hyphal net-like structure containing what

is interpreted as cyanobacteria (Precambrian). Bar  20 μm. (Courtesy S. Xiao.)

Figure 3.121 Hyphal net of Winfrenatia reticulata enclosing cyanobacterial unicells (Devonian). Bar  60 μm.

Figure 3.120 Section of thallus of Winfrenatia reticulata showing hyphal pockets (arrows) with cyanobacteria inside (Devonian). Bar  800 μm.

with each net space filled by a coccoid, cyanobacterial cell surrounded by a thick sheath (FIG. 3.122). Each depression of the thallus shows various stages of cyanobiont cell division (FIG. 3.123); in some depressions the cells of the cyanobiont are moribund, suggesting that in these regions the cyanobacteria have become fully parasitized by the fungus. The affinities of the fungus are not known, and the cyanobacterium could be any one of a number of coccoid forms. It has been suggested that this fossil represents a colony of cyanobacteria parasitized by a fungus (Poinar et al., 2000). Although the thallus organization of W. reticulata is unlike that of modern lichens, the consistent association of coccoid cells and aseptate hyphae in this Devonian fossil satisfies the organismal and functional components of a lichen symbiosis, and the morphology of W. reticulata is similar

Figure 3.122 Cyanobacterial cell from Winfrenatia reticulata

showing thickened sheath (Devonian). Bar  10 μm.

CHAPTER 3 fungi, bacteria, and lichens

Figure 3.123 Eight-celled stage of Winfrenatia reticulata

cyanobacterium (Devonian). Bar  10 μm.

to the Proterozoic fossils of Yuan et al. (2005). In addition, lichen symbioses represent a type of controlled parasitism, and lichens are believed to have evolved from true parasitism to the more mutualistic relationships seen in many modern taxa. Flabellitha is a leaf-like film interpreted as a lichen from the Devonian of Kazakhstan. The fossil consists of septate hyphae and slightly sunken apothecia with asci; each ascus contains two bicellular spores (Jurina and Krassilov, 2002). Pelicothallos, found on the angiosperm leaf Chrysobalanus sp. from the Eocene of Tennessee, was originally described as an epiphyllous fungus (Dilcher, 1965) and later reinterpreted as an alga (Reynolds and Dilcher, 1984). The thallus bears setae and several dark fruiting bodies. On the basis of the spores and spore-bearing structures, Sherwood-Pike (1985) suggested that the fossil is similar to Cephaleuros (Trentepohliaceae), a genus of green algae that includes forms parasitic on land plants (Joubert and Rijkenberg, 1971), as well as those that function as lichen phycobionts (e.g., in the lichen Strigula). Indirect evidence of a cyanobacterial or algal-fungus symbiosis to form a lichen has been reported from Eocene amber collected from the Baltic region (Garty et al., 1982). This fossil is a multibranched thallus 2 cm wide, with the individual branches 1 mm in diameter. On the surface of the thallus are several structures thought to represent aeration pores, apothecia, and spores. The fossil is encased in amber, and elemental analysis suggests the presence of sulfur, calcium, iron, silicon, and aluminum. The presence of certain elements in the fossil is suggested as demonstrating


that the lichen accumulated iron and sulfur from the Eocene atmosphere prior to fossilization and thus can be regarded as a bioindicator of air pollution at that time. Other lichens have been described from Dominican amber (Rikkinen and Poinar, 2008), including a foliose thallus that morphologically appears similar to modern forms in the Parmeliaceae (Poinar et al., 2000). Additional indirect evidence of fossil lichens comes from the suggestion of Klappa (1979) that some laminar calcretes were formed by so-called lichen stromatolites. It has been known for some time that laminar calcretes (caliche), a type of calcareous soil deposit, contain algal filaments, fungal hyphae, and layers of organic-rich and organic-poor material, but the formation of these structures was not well understood. Lichens are known to be primary colonizers of rock surfaces, and the changes that they cause in their substrate represent the beginnings of soil formation. Klappa proposes that laminar calcretes are formed by a cycle of lichen colonization, followed by hardening of these biologically formed surfaces, followed by further lichen colonization, and so on. Over time, a layered structure, the so-called lichen stromatolite is formed. It is surprising that more lichens have not been described from the fossil record, since many of them, such as the foliose forms, have tissues capable of preservation. The recognition of well-preserved photobionts and mycobionts represents a major problem in the identification of these organisms since both are relatively fragile and may be difficult to recognize in permineralizations. In addition to a possible inability to recognize fossil lichens, another problem may simply be that the total number of lichens throughout geologic time was relatively small. Although many modern species reproduce asexually, which rapidly increases their distribution, the ancestral forms of lichens may have relied exclusively on the sexual mechanism of the fungus, a relatively slow way to distribute new individuals into the environment (Bowler and Rundel, 1975). In addition, many extant lichens grow in dry, exposed habitats, such as on bare rock surfaces, where the chance of fossilization is greatly diminished. Molecular sequence data from living ascomycetes involved in lichen symbioses suggest that not only have lichens evolved multiple times, but that lichens are far older than we once believed (Lutzoni et al., 2001). As more is learned about fossil fungi, algae, and cyanobacteria, obstacles to recognizing lichens in the fossil record and their geologic history may be substantially decreased. This is turn provides an opportunity to discuss not only the timing of lichenization through time but also the evolution of various structures unique to this mutualistic association.

This page intentionally left blank

4 Algae CHLOROPHYTA (GREEN ALGAE) ......................................... 123

Dictyochophyceae (Silicoflagellates) ............................................... 142

Prasinophyceae................................................................................. 124

Xanthophyceae (Yellow-Green Algae) ............................................. 142

Chlorophyceae ................................................................................. 126

Phaeophyceae (Brown Algae) .......................................................... 143

Ulvophyceae..................................................................................... 128

PRYMNESIOPHYTA (HAPTOPHYTES) ............................... 144

Charophyceae ................................................................................... 133 RHODOPHYTA (RED ALGAE) ................................................ 145 EUGLENOPHYTA ......................................................................... 138

Solenoporaceans .............................................................................. 146

DINOPHYTA (DINOFLAGELLATES) ..................................... 139

Other Calcified Red Algae ............................................................... 149

HETEROKONTOPHYTA ............................................................ 141

Uncalcified Red Algae ..................................................................... 150

Bacillariophyceae (Diatoms)............................................................ 141

ACRITARCHA (ACRITARCHS)................................................158

We have lingered in the chambers of the sea By sea-girls wreathed with seaweed red and brown Till human voices wake us, and we drown. T.S. Eliot, The Love Song of J. Alfred Prufrock

The algae are a large informal grouping of heterogeneous, polyphyletic or paraphyletic groups of primarily aquatic organisms ranging from tiny, flagellated unicells only a few microns in diameter to multicellular organisms up to 80 m long, such as the giant kelps (Graham and Wilcox, 2000). Unlike vascular plants, the algal body (thallus) lacks organ differentiation, although some forms have developed structures functionally similar to roots, shoot axes, and leaves. Most algae are photoautotrophic; some forms, however, are mixotrophic and derive energy both from photosynthesis and uptake of organic carbon by osmotrophy, myzotrophy, or phagotrophy. A few forms have reduced or lost their photosynthetic capacities and are entirely heterotrophic. Extant

algae are classified on the basis of a complement of features that do not normally lend themselves to fossil preservation. These include the type of pigments present, storage products, type of flagellation if present, and degree of cellularization. Moreover, molecular, biochemical, and ultrastructural characteristics are increasingly important in algal systematics and phylogeny (Brodie and Lewis, 2007). Although algae thrive in a spectrum of habitats, including Antarctic ice, rock and tree surfaces, animal fur, human and animal skin, and desert sand, most forms are aquatic. Algae are critical in modern aquatic ecosystems, not only in producing oxygen for other aquatic life, but also in serving as primary producers of organic matter at the base of the food



paleobotany: the biology and evolution of fossil plants

web (Round, 1981). Stands of larger algae (e.g., kelp forests) are used by animals as shelters and nursing grounds; other algae are pivotal in the physiology of some aquatic animals and involved in various vital processes (e.g., symbioses with corals [cnidarians]). Still other algae are structural contributors to the formation of reefs. With this recognition of the ecological significance of extant algae, a sound record of the evolutionary history of these organisms, including the roles they played in biological and ecological processes in the past, is critical in understanding the complexity and evolution of both ancient and modern aquatic ecosystems. Molecular clock hypotheses suggest that the first algae occurred on Earth during the late Paleoproterozoic or early Mesoproterozoic, some 1.5 Ga or even earlier (Yoon et al., 2004). This date corresponds with the long fossil record of these organisms, which can be traced into the Precambrian. Despite the often fragile nature of the plant body of many algae, there are numerous algal remains throughout the fossil record. Some algae have contributed to the formation of petroleum and thus are chemical fossils, whereas others are represented by thousands of feet of accumulated siliceous diatom shells. Algae that precipitate or deposit calcium carbonate, CaCO3, are often common in the fossil record. The actual algal thallus may not be preserved, but in many instances, the calcareous structures can form extensive deposits (Coniglio and James, 1985). Calcareous algae first appear in the late Neoproterozoic and become widespread during the Cambrian (Riding and Voronova, 1985). Among extant algae, 10% of benthic multicellular forms are calcified, with about 90 genera known for the red (Rhodophyta), 10 for the green (Chlorophyta), and 2 for the brown algae (Phaeophyceae) (Lüning, 1990; Kraft et al., 2004). Various calcareous algae play a major role in the formation of modern coral reefs (FIG. 4.1) (Wood, 1998), especially some coralline red algae (see section “Rhodophyta”) and the green alga, Halimeda (see section “Ulvophyceae”). Although sections of the fossils rarely indicate any of the cellular details of the original organism, the existence of calcified algae is determined by the presence of calcium carbonate that accumulated around the thallus. Cellular details of fossil algae can be preserved in certain depositional environments, for example, various Proterozoic organisms, including various algae, have been preserved in cherts. In contrast to many groups of microalgae and calcareous macroalgae, the fossil record of uncalcified macroscopic algae is meager, and hence the evolutionary history of these plants remains poorly understood. The failure to more fully

Figure 4.1

Coral (Extant). Bar  2 cm.

document uncalcified macroalgae throughout geological history is probably because algal thalli, with no real strengthening tissue, are not easily preserved (Tappan, 1980). Many marine forms may live in the intertidal zone, where they are easily destroyed after death. In addition, they do not produce characteristic skeletons that can be used for comparison with modern forms. As a result, hypotheses on the affinities of uncalcified fossil macroalgae, especially those preserved as impressions and compressions, are often formulated solely on basic morphological comparisons with modern forms. These scenarios thus remain speculative and place serious constraints on the interpretation of the roles these plants played in ancient ecosystems. Numerous impression and compression fossils were described, especially in the nineteenth and early twentieth centuries, that superficially resembled algae; these fossils, however, lacked certain distinctive features, so it was difficult to establish their biological affinities. To classify such forms, a system of artificial morphotaxa based on a few characteristic macromorphological traits was instituted. To a certain degree this taxonomy is still used today. For example, within the older literature there are numerous references to the morphogenus Fucoides (Harlan, 1830; Taylor, 1834; Lesquereux, 1869). This taxon was instituted by Brongniart (1822), for fossils which he interpreted as

chapter 4

similar to the extant brown alga Fucus and similar forms in the order Fucales (Phaeophyceae). The algal nature of most of these structures, some of which are as old as the Cambrian, however, has been challenged and many of them today are thought to be ichnofossils (Jensen and Bergstrom, 1995), graptolites (Bulman, 1963), or different sedimentary structures. The morphogenus Algites was established by Seward (1894) for fossils that he thought were probably algal, but which, because they lacked reproductive or internal characteristics, could not be assigned to any modern group. Over the years, Algites has become a “wastebasket taxon” for all sorts of enigmatic fossils that are similar to algae in basic organization. One species, A. enteromorphoides from the Devonian of Missouri, is composed of narrow, branching filaments and has been suggested to be a member of the Chlorophyta (Basson and Wood, 1970). Thallites is another morphogenus used for fossil thalloid plants that may have been either algae or bryophytes. Webb and Holmes (1982) suggested that these fossils be referred to as intermediate thalloid fossils. Thallites dichopleurus (Middle Pennsylvanian) is a dorsiventral thallus with a well-developed midrib (DiMichele and Phillips, 1976). The undulate lamina is slightly less than a centimeter wide, and the entire thallus dichotomizes several times. Latex peels made from the surface of the specimen and examined in the scanning electron microscope show epidermal cells of the midrib and thallus to be of different sizes; no epidermal appendages or openings are present. Other species of Thallites from Mesozoic rocks have been delimited on the basis of size, character of the margin, and nature of dichotomies. Since uncalcified macroalgae are not easily preserved, one of the most important discoveries in paleoalgology is numerous, uncalcified macroalgal floras that are well preserved as carbonaceous fossils in Precambrian and Cambrian marine rocks from North America, Europe, and especially China. Over the last few decades, paleobotanists working on the material from China have been particularly productive in deciphering these organisms (Cao and Zhao, 1978; Du and Tian, 1985, 1986; Duan et al., 1985; Y. Zhang, 1989; Zhang and Yuan, 1992; M. Chen et al., 1994; Steiner, 1994; R. Yang et al., 1999). Today many forms can be assigned to one of the major groups of algae with some confidence, based on structural similarities to extant forms, although the affinities of others still remain inconclusive. We have included several examples of fossil microalgae, as well as calcified and uncalcified macroalgae within the major groups:



Higher taxa in this chapter (based largely on the classification system of Lee, 1999):

Phylum Chlorophyta (see text for detail) (green algae) Prasinophyceae Chlorophyceae Ulvophyceae Charophyceae Phylum Euglenophyta (Middle Ordovician–recent) Phylum Dinophyta (dinoflagellates) (?Silurian–recent) Calciodinellaceae Phylum Heterokontophyta (heterokonts) Bacillariophyceae Dictyochophyceae Dictyochales Xanthophyceae (yellow-green algae) Vaucheriales Phaeophyceae (brown algae) Ectocarpales Ectocarpaceae Laminariales Fucales Cystoseiraceae Phylum Prymnesiophyta (haptophytes) Coccolithophorales Discoasters Phylum Rhodophyta (red algae) Solenoporaceans (artificial group) Rhodophyceae Bangiales Corallinales Graticulaceae, Sporolithaceae, Corallinaceae Nemaliales Ceramiales Delesseriaceae Gigartinales Acritarcha (artificial group)

CHLOROPHYTA (GREEN ALGAE) The Chlorophyta (FIG. 4.2), or green algae, are the most diverse group of algae in the world today in terms of number of species (at least 7000 species), organization of the plant body (unicellular to multicellular), and habitat (from the surface of snow to a variety of symbiotic relationships) (Graham and Wilcox, 2000). The green algae are generally assumed to include the ancestral group that has given rise


paleobotany: the biology and evolution of fossil plants

 Gametophyte (n)

 Spore (n)

 Gametangia (n)

Germinating  spore

Germinating  spore  Spore (n)  Gametangia (n)

Spores (n) Meiosis  Gametophyte (n)

 Gamete (n)  Gamete (n)

Sporangia (2n) Sporophyte (2n) Germinating zygote

Zygote (2n)


Fusing gametes

Figure 4.2 Life history of the green alga Ulva showing an isomorphic alternation of generations. (From Taylor and Taylor, 1993.)

to the embryophytes (land plants). The green algae and the embryophytes are today interpreted as a monophyletic group, the Viridiplantae, consisting of two lineages, Chlorophyta and Streptophyta, the latter of which includes all embryophytes and the green algal class Charophyceae. Although the fossil record of the green algae is extensive, the fossil record has provided almost no information about the evolutionary steps involved in this transition (see Chapter 6). Instead, the phylogenetic relationships between these two groups have been hypothesized based on the study of biochemical, cytological, and ultrastructural features of living taxa (Lewis and McCourt, 2004) and these have been confirmed by molecular phylogenetic studies (Karol et al., 2001; Turmel et al., 2007). There are several late Neoproterozoic microfossils that morphologically resemble members of the Chlorophyta. One of these is Caryosphaeroides, a spheroidal unicell from the Bitter Springs Formation (800–830 Ma) of Australia (Schopf, 1968). The cells range from 6 to 15 μm in diameter and lack an outer sheath; a few specimens appear to have been contained within an amorphous organic matrix. Morphologically, these microfossils appear similar to the living green algae Chlorococcum and Chlorella (Chlorococcales). Neoproterozoic (700–800 Ma) shales from northeastern Spitsbergen have yielded a rich, well-preserved microfossil assemblage, including several

types morphologically similar to members of the Chlorophyta (Butterfield et al., 1988). One of these is a complex branched structure 10–50 μm in diameter and 1 mm long. It has been suggested that these structures are similar to the rhizoids of certain members of the Chaetophorales, a group of green algae that possess a prostrate basal system from which arise erect, branching filaments. Also included in this shale assemblage are branching, filamentous thalli constructed of cylindrical cells that range from 50 to 800 μm in diameter. Some fragments of this alga were a centimeter long. Certain extant members of the cladophoralean green algae (Cladophorales) are morphologically similar to these fossils. PRASINOPHYCEAE

The Prasinophyceae is a large, poly- or paraphyletic group of single-celled, flagellate green algae hypothesized to have diverged early in chlorophyte phylogeny (Lewis and McCourt, 2004). The group consists of four (Nakayama et al., 1998) or six to seven (Teyssèdre, 2006) distinct clades, one of which is the Pyramimonadales. The life history of several species of Pyramimonadales includes a unique non-motile stage characterized by the formation of organic-walled cyst-like structures termed phycomata (sing. phycoma). The alga remains metabolically active within the phycoma and undergoes

chapter 4



Higher taxa of green algae in this chapter (based largely on the classification system of Lee, 1999):

Class Prasinophyceae Pyramimonadales Cymatiosphaeraceae, Tasmanaceae Class Chlorophyceae Chlorococcales Hydrodictyaceae, Scenedesmaceae, Chlorococcaceae Tetrasporales Botryococcaceae, Palmellaceae Chaetophorales Volvocales Lageniastraceae Class Ulvophyceae Cladophorales Dasycladales (Cambrian–recent) Families: See Table 4.1 Receptaculitida (artificial group, may not be algae) Cyclocrinales (Cambrian–Devonian) Caulerpales Codiaceae, Caulerpaceae Ulvales Class Charophyceae Charales Eocharaceae, Calvatoraceae Moellerinales Moellerinaceae (Silurian–Permian) Sycidiales Sycidiaceae, Trochiliscaceae, Chovanellaceae, Pinnoputamenaceae Zygnematales Coleochaetales (see Chapter 6)

vegetative reproduction; as a result, the phycoma increases in size over time. Prasinophycean phycomata have decayresistant walls and thus have preservation potential as fossils (Quintavalle and Playford, 2006) (FIG. 4.3); the oldest persuasive specimens come from the Precambrian and are more than 1.2 Ga old (reviewed in Teyssèdre, 2006). The majority of fossil phycomata have been described from strata deposited in marine or brackish water, but there are also several reports of these structures from freshwater deposits (Dotzler et al., 2007). Phycomata assigned to the genus Cymatiosphaera (200 species currently recognized) are globular (spherical– ellipsoidal) vesicles, up to 100 μm in diameter, with the external surface divided into a polygonal reticulum by

Figure 4.3 Geoffrey Playford.

membranous expansions or muri, which are perpendicular to the surface (FIG. 4.4). Cymatiosphaera fossils are structurally similar to the reticulate forms within the extant prasinophycean genus Pterosperma (Colbath and Grenfell, 1995). Phycomata with a surface reticulum (Cymatiosphaeraceae) are present at least by the Ordovician (Tappan, 1980). They diversified extensively in the Devonian and later Paleozoic but are rare in Mesozoic and Cenozoic rocks. Because prasinophytes are most abundant in the absence of other phytoplankton, they became less important when dinoflagellates and other algal groups diversified during these periods. One hypothesis suggests that blooms of prasinophytes and acritarchs observed across the Triassic–Jurassic boundary are in response to rapid CO2 increases which acidified the upper ocean and thus reduced the ability of other organisms to calcify (van de Schootbrugge et al., 2007). Tasmanites (Tasmanaceae) is another type of algal microfossil which occurs in many marine facies from the Cambrian to the Miocene (Martín-Closas, 2003). Specimens are commonly found in the maceral type (Chapter 1) called tasmanite, as well as in certain oil shales. At least one species of this family, Pleurozonaria maedleri, however, has been recorded from Late Pennsylvanian–Early Permian non-marine deposits from France (Doubinger, 1967). The fossils are


paleobotany: the biology and evolution of fossil plants

Figure 4.4 Phycoma of the prasinophyte Cymatiosphaera

(Devonian). Bar  20 μm. (Courtesy N. Dotzler.)

preserved as compressed cysts, usually ranging from 100 to 600 μm in diameter. Haptotypic features are absent, but the surface is covered by numerous, regularly spaced punctae or small pits. Ultrathin sections of the wall show concentric banding and two types of pores which traverse the wall in a radiating pattern (Jux, 1968). The affinities of Tasmanites have remained in doubt since the original description by Newton in 1875. Most researchers now believe, however, that these microfossils represent the phycomata of planktonic prasinophytes. Guy-Ohlson (1988), using a variety of techniques, was able to demonstrate the life history of Tasmanites based on specimens of Toarcian (Early Jurassic) age from southern Sweden. Her investigation substantiates the affinities of Tasmanites with the modern prasinophycean algae Pachysphaera, Halosphaera, and Pterosphaera. CHLOROPHYCEAE

VOLVOCALES The Chlorophyceae encompass the widest range of morphologies in the green algae. The colonial Volvocaceae (Volvocales) and their unicellular relative Chlamydomonas reinhardtii (Chlamydomonaceae) have frequently been used as a model in studies addressing the evolutionary pathways leading from unicellularity to multicellularity, including a division of labor within the algal thallus (Kirk, 1998, 1999). Molecular evidence suggests a minimum age of 400–500 Ma for a few Chlamydomonas species (Van den Hoek et al., 1988).

Nevertheless, the fossil record of Chlamydomonaceae is virtually nonexistent, and that of Volvocaceae is meager, perhaps because colonies (coenobia) disintegrate almost immediately upon death (Tappan, 1980). Fossil unicellular algae suggestive of Chlamydomonas are preserved in Cenomanian (Late Cretaceous) amber from southern Germany (Schönborn et al., 1999). The algal cells are thick-walled, oval in lateral view, up to 10 μm long and display a single, cup-shaped chloroplast, which is characteristic of extant Chlamydomonas; flagella are not recognizable. A rare microfossil that has been interpreted as a volvocacean alga is Eovolvox silesiensis from the Devonian of Poland (Kaz´mierczak, 1975, 1981). This fossil consists of hollow spherules with a surface layer composed of closely spaced, ovoid, pyriform, or spindle-shaped isomorphic cells. It is similar in basic structure to Symphysosphaera radialis from the Lower Cambrian of China (Yin, 1992). Affinities with the Volvocaceae have also been suggested for a few other Paleozoic and Mesozoic microfossils (reviewed in Kaz´mierczak, 1981). Few of these records, however, can be regarded as unequivocal (Kirk, 1998). The most biologically interesting fossil with possible affinities to the Volvocales is the endophyte Lageniastrum macrosporae (Lageniastraceae) from the Lower Carboniferous (Viséan) of France (Renault, 1896a; Krings et al., 2005a). This alga occurs inside lycopsid megaspores in the form of dome-shaped, three-dimensional colonies composed of up to 500 lens- to pear-shaped cells arranged in a single layer and bounded by a transparent membrane. Lageniastrum macrosporae colonies display a striking similarity in organization to certain extant species of Volvox, including the presence of radiating protoplasmic strands that interconnect adjacent cells in the colony (FIG. 4.5). TETRASPORALES Several groups of chlorophycean green microalgae have been important geologically. Some are responsible for the formation of certain coals and also may have contributed to the formation of petroleum in oil shales (Wolf and Cox, 1981). These deposits contain irregularly shaped, yellow bodies that were formed by the hydrocarbon-producing alga Botryococcus braunii. This taxon, the only member in the family Botryococcaceae, is a living planktonic colonial green alga in the order Tetrasporales that is known from both temperate and tropical climates throughout the world. The fossil colonies consist of pear-shaped cells arranged in radial rows and surrounded by a mucilage layer which has been described as a cuticularized layer. In the fossils, the yellow bodies are thought to represent paraffins and fatty acids secreted by the cells and bound together by the mucilage-like

chapter 4



Figure 4.5 Lageniastrum macrosporae colony showing proto-

plasmic strands interconnecting cells. Bar  50 μm.

substance of the sheath. Botryococcus colonies are known from every geologic period beginning in the Ordovician (Martín-Closas, 2003) and extending to the present day, and they can be the dominant plankton of certain freshwater ecosystems (Batten and Grenfell, 1996), for example, in the Rotliegend (Late Pennsylvanian–Early Permian) of the Saar-Nahe Basin in Germany (Clausing, 1999). Fossil Botryococcus has been shown to represent valuable proxy indicator for certain paleoenvironments (Guy-Olsen, 1998). Modern Botryococcus colonies live in standing bodies of both fresh and brackish water (Colbath and Grenfell, 1995). The alga has recently attracted attention in biotechnology since it represents an unusually rich, renewable source of hydrocarbons and other chemicals (Banerjee et al., 2002). CHLOROCOCCALES Chlorophycean green algae belonging to the order Chlorococcales have been reported from the Messel Oil Shale (middle Eocene) of Germany (Goth et al., 1988). These kerogen-rich sediments have a very high percentage of organic matter made up of unicellular algae that are morphologically identical to the modern species Tetraedron minimum (Chlorococcaceae). The rectangular unicells range from 5 to 20 μm long. Geochemical analyses of these cells indicate the presence of an insoluble, non-hydrolyzable, highly aliphatic biopolymer that is believed to represent a significant precursor in the formation of n-alkanes in crude oils. Laminar concentrations of fossil T. minimum have been described from lower Miocene lacustrine sediments near Hausen in the Rhön Mountains, central Germany (Goth and Schiller, 1994). The genus Pediastrum (Hydrodictyaceae) consists of diskshaped coenobia or colonies composed of a variable number

Figure 4.6 Coenobia of Plaesiodictyon decussatus (Triassic). (From Brenner and Foster, 1994.)

of cells (Gray, 1960). Cells are arranged in a concentric pattern, with each cell of the outer ring containing one to three spines. The geological range of Pediastrum remains uncertain. Stanevich et al. (2007) suggested that the Neoproterozoic acritarch Dictyotidium minor from the Chencha Formation of eastern Siberia is structurally similar to the extant P. boryanum, and thus may represent an early relative of that genus. Fossils that appear similar to Pediastrum have also been found in Silurian (Deflandrastrum; Combaz, 1962) and Triassic rocks (Plaesiodictyon (FIG. 4.6); Brenner and Foster, 1994; Wood and Benson, 2000). The first unequivocal representatives of Pediastrum come from the Early Cretaceous of Britain and North America (Batten, 1996); others have been reported from the Upper Cretaceous–Neogene of southern South America (reviewed in Zamaloa and Tell, 2005). The genus is also known from the Miocene of Oregon and the Eocene of southern Sumatra. The Miocene Pediastrum fossils from southern South America are 8- to 32-celled coenobia


paleobotany: the biology and evolution of fossil plants

Figure 4.7 Coelosphaeridium cyclocrinophilum (Ordovician). Bar  1 cm. (Courtesy BSPG.)

with the marginal cells each bearing two spines (Tell and Zamaloa, 2004). Today the genus lives exclusively in freshwater, and its presence in Cretaceous marine rocks suggests a difference in the physiologic tolerance of the taxon, or that the genus is not as good a proxy record for freshwater environments as has been suggested (Evitt, 1963a). Another extant chlorococcalean green alga that is also known from the fossil record is Scenedesmus. It consists of cylindrical cells with rounded or pointed ends that are joined laterally into 4- to 16-celled coenobia. Two fossil species (S. hanleyi and S. tschudyi) have been reported from the Upper Cretaceous and lower Paleocene of Colorado and Mexico (Fleming, 1989). Coenobia of S. tschudyi consist of four or eight cells, with the terminal cells possessing elongate extensions, whereas coenobia of S. hanleyi have four oval cells. Both Pediastrum and Scenedesmus are green algae almost exclusively restricted to freshwater habitats and important non-marine paleoecological indicators of the presence of lacustrine environments (but see above). ULVOPHYCEAE

DASYCLADALES Among the most commonly encountered members of fossil green macroalgae are those forms assignable or structurally similar to the Dasycladales (FIG. 4.7). There are 180 genera currently recognized within this order, of which only 11 still exist today. Berger and Kaever (1992) classified the 180 genera into five families, that is, the Seletonellaceae (Cambrian– Cretaceous), the Diploporaceae (Devonian–Triassic), the

Figure 4.8 Suggested reconstruction of Triploporella remesii. (From Pia in Hirmer, 1927.)

Triploporellaceae (Ordovician–Eocene) (FIG. 4.8), the Dasycladaceae (Jurassic–recent), and the Acetabulariaceae (Carboniferous–recent) (FIG. 4.9). In addition, a sixth family, the Beresellaceae (Late Devonian–Permian), has traditionally been placed in the Dasycladales (Deloffre, 1988), but Berger and Kaever (1992) regarded this family as a heterogeneous group of organisms that does not belong to the order Dasycladales and may not even represent algae (Table 4.1). Adams and Al-Zahrani (2000), however, retained the Beresellaceae in the order Dasycladales and reported on Kamaena khuraisensis from Saudi Arabia, a form that extends the fossil record of the family into the Late Jurassic (Kimmeridgian). Morphologically, fossil dasycladalean algae are radially symmetrical, with a central axis that produces whorls of lateral appendages, some of which are branched. Reproduction includes the formation of operculate cysts containing isogametes. Most members of the Dasycladales secrete lime around the thallus, and this greatly increases their potential for preservation. Uncalcified dasycladaleans are comparatively rare as fossils; an example of one is Chaetocladus (LoDuca, 1997; Kenrick and Vinther, 2006). The morphologically simple, bottle-brush-shaped thalli of Chaetocladus (Ordovician– Devonian) range from 2 to 6 cm high and 0.5 to 1.5 cm

chapter 4



Figure 4.9 Acetabularia acetabulum gametangium-bearing

cup (Extant). Bar  1 mm. (From Berger and Kaever, 1992.) Table 4.1 Geologic range of dasycladalean families.


Geologic range

Seletonellaceae Diploporaceae Triploporellaceae Dasycladaceae Acetabulariaceae Beresellaceae?

Cambrian–Cretaceous Devonian–Triassic Ordovician–Eocene Jurassic–recent Carboniferous–recent Late Devonian–Permian (?Late Jurassic)

Source: From Berger and Kaever (1992).

wide and are composed of an unbranched, parallel-sided central axis surrounded by various projections. The slender, unbranched projections terminate sharply, without rounding or tapering, and are arranged as discrete whorls around the axis. Other Paleozoic genera interpreted as uncalcified dasycladaleans include the Silurian genus Heterocladus (FIG. 4.10) (LoDuca et al., 2003), Medusaegraptus, which ranges from the Ordovician to the Silurian (LoDuca, 1990), and the Devonian Uncatoella (Kenrick and Li, 1998). Among the oldest fossil calcified dasycladalean algae may be Amgaella from the Middle Cambrian of the Amga River (Russia) and Mejerella and Seletonella from the Upper Cambrian of Kazakhstan; the precise affinities of these organisms, however, remain uncertain (Riding, 2001). Among the earliest records of bona fide calcified dasycladaleans are fossils assigned to Dasyporella, Moniliporella,

Figure 4.10 Thallus of Heterocladus waukeshaensis (Silurian).

Bar  2.0 cm. (From LoDuca et al., 2003.)

and Plexa from the Ordovician of China (Riding and Fan, 2001), and Rhabdoporella of probable Late Ordovician age (Riding, 2001). Studies by Elliott (1978) suggest that the fossil dasyclads occupied environments similar to their modern counterparts and that their adaptations to specific microenvironments (e.g., salinity, bottom sediment, attachments, and water energies) are essentially similar to those of the extant forms. In general, calcareous algae are scarce in the Early Ordovician, becoming diversified from the Middle Ordovician onward. Since they are associated with warm-climate carbonate sedimentation, none are known from Gondwana during the Ordovician, except in Australia, which occupied a low paleolatitude at that time (Poncet and Roux, 1990). In the Paleozoic and Mesozoic dasyclads, the primary whorled branches are irregularly positioned. These, in turn, bear secondary and tertiary laterals. Most were cylindrical plants, although some were club-shaped, spherical, or shaped like a string of beads (Bassoullet et al., 1977). The thallus was attached to the substrate by simple rhizoids. In the fossils, the encrusting calcium carbonate is sometimes so thick that


paleobotany: the biology and evolution of fossil plants

several orders of branches are covered. A slightly different morphology is apparent in Primicorallina (Seletonellaceae; Ordovician). Specimens of this dasyclad have a remarkably narrow central axis with three orders of laterals in lax arrangement (Berger and Kaever, 1992). Another form that extended from the Ordovician into the Lower Carboniferous is Rhabdoporella (Seletonellaceae), a type that produced only primary laterals, each of which were slightly clavate or terminally expanded. This genus, along with Dasyporella and Vermiporella, has been suggested as representing a primitive form or early evolutionary stage in the dasyclad algae (Herak et al., 1977). More highly evolved members are interpreted as containing a larger number of whorls and more orders of branches. Pianella is a small (3 mm long) dasyclad with funnel-shaped branches arranged in alternating whorls. This taxon is common in the Early Cretaceous of the Middle East, Italy, and the former Yugoslavia. Mizzia is a Permian dasyclad in the family Triploporellaceae that is found shelfward of the Capitan reef complex of southeastern New Mexico (Kirkland and Chapman, 1990). The modern analog of Mizzia is Cymopolia, which extends from the Cretaceous to the recent. Lithologic evidence suggests that Mizzia grew in water that was shallow, warm, and only slightly agitated, subsequently becoming hypersaline. The study by Kirkland and Chapman (1990), involving the distribution of Mizzia in time and space, uses this genus as an indicator of paleoenvironment. More recently, long-term patterns of dasycladalean biodiversity over the past 350 myr (Carboniferous–Pliocene) were compared to global fluctuations in temperature and sea level. It appears that, in general, diversity peaks occurred when warm, shallow seas were most extensively developed on the Earth (FIG. 4.11) (Aguirre and Riding, 2005). RECEPTACULITIDA AND CYCLOCRINALES Two other groups of Paleozoic calcium carbonate–depositing organisms that are known from fossil examples are the receptaculitids and cyclocrinaleans, which are often considered to be related to one another because they display comparable morphological architectures. Both groups have been classified among various higher-level taxa, including corals, bryozoans, and particularly sponges and algae. The latest interpretation is that receptaculitids are neither sponges nor algae (Riding, 2004), while cyclocrinitids (FIG. 4.7) represent a problematic group of algae, possibly a sister group to the Dasycladales (Nitecki et al., 1999). Members of the order Receptaculitida occur in the fossil record from the Ordovician to the Carboniferous (FIGS. 4.13; 4.14), and perhaps into the Permian (Nitecki, 1972; Nitecki et al., 2004). Although

they are typically found associated with coral reefs, they are not regarded as reef builders. One of the most common representatives of the receptaculitids is the Ordovician Fisherites reticulatus (Finney et al., 1993), sometimes called the sunflower coral (FIG. 4.12). This organism is spherical or cup shaped, 30 cm in diameter, and composed of a central axis from which radiate numerous, spindle-shaped lateral branches (meroms or meromes). Each merome bears a terminal rhomboidal element (FIG. 4.12), which together form a surface pattern of facets that resembles the arrangement of ripe achenes in a sunflower capitulum (inflorescence). The order Cyclocrinales displays a narrower stratigraphic range, extending from the upper Middle Cambrian to the Lower Devonian (Nitecki et al., 2004). Cyclocrinitids existed in communities at depths of 100 m of water. Cyclocrinites is a common Silurian form with a club-shaped central stem and radiating primary branches (FIG. 4.15). At the end of each primary branch shaft are bowl-shaped cortical cells with flattened tops that in turn merge to form a reticulum of hexagonal plates on the outer surface perforated by regularly spaced holes. CAULERPALES Halimeda (Caulerpaceae) is a calcified green alga that is widespread today and important as a structural component of many Cenozoic–recent reefs; the genus is estimated to be responsible for 25–30% of the CaCO3 in Neogene fossil reefs (Stanley and Hardie, 1998). The thallus of Halimeda is constructed entirely of branching filaments that are matted together to form a plant body, which in some species may reach over a meter in length. Since no cross walls are produced in any of the filaments, the plant may be regarded as one giant, multinucleated cell. Despite its dominance today, the fossil record of Halimeda extends back only to the Cretaceous and possibly to the Triassic (Hillis, 2001; Dragastan and Soliman, 2002), although taxa with halimediform morphology are known from the Permian (Poncet, 1989). Several authors (Hardie, 1996; Stanley and Hardie, 1998; Knoll, 2003a) have correlated the rise and fall of certain calcareous organisms, including some green algae, with changes in seawater chemistry through time, especially the ratio of magnesium to calcium, Mg/Ca. Those organisms which deposit massive calcium carbonate skeletons (FIG. 4.16) and have weak physiological control over mineralization are most strongly affected by changes in seawater chemistry. This theory would explain the lack of Halimeda-type algae prior to the formation of “aragonitic” seas in the Late Jurassic–Cretaceous (Ries, 2005). In fact, the dasycladalean algae are the group which contributed most to carbonate rock formation in the Triassic.

chapter 4


50 60

Ol. Mioc.



Pa. Eoc.



P io

10 20



120 130

Present sea level


Present sea level

90 100



150 Age (Ma)

160 170 180

Mesozoic Jurassic


190 200 220 230




270 280 290 300 310 320 330 340

Paleozoic Carboniferous






Haq et al. (1987)



Hallam (1992)

350 0


40 60 80 n° dasyclad species


Figure 4.11 Comparative plots of Carboniferous–Pliocene dasycladalean species richness measured against sea level estimates. Brackets at right indicate periods of maximum diversity. (Redrawn from Aguirre and Riding, 2005.)

The Paleozoic calcified marine green algae also include species assigned to the Codiaceae that lived in dense colonies in warm, relatively shallow waters (Torres, 1995, 1999; Torres et al., 2003). The vegetative thallus of Ivanovia tebagaensis, a Permian form, is coenocytic or siphonous, with a cyathiform (cup-shaped) membrane-like thallus, 1.5–2 cm wide and up to 3 cm tall. The membrane forming the

thallus had inner and outer cortices composed of utricles surrounding a central medulla (FIG. 4.17). The alga apparently reproduced asexually by forming outgrowths (buds) on a parental thallus; the mode of sexual reproduction in this species is unknown. Ivanovia is also known from Triassic rocks of the Yukon Stikine Terrane in Canada (Reid, 1986; Torres, 2003), and thus this genus was among the relatively few


paleobotany: the biology and evolution of fossil plants

Figure 4.12 Fisherites reticulatus (Ordovician). Bar  5 cm. (Courtesy BSPG.) Figure 4.14 Ischadites murchinsonii (Ordovician). Bar  1 cm. (Courtesy BSPG.)

Figure 4.13 Ischadites sp. showing surface pattern. Bar  2 cm.


Figure 4.15 Diagrammatic reconstruction of Cyclocrinites dactyloides (Silurian). (From Taylor and Taylor, 1993.)

chapter 4



algal survivors of the massive Permian–Triassic extinction. The species from the Yukon, I. triassica, is interesting because it displays sexual reproductive structures in the form of stalked spherical outgrowths from the thallus that are interpreted as oogonia, and dome-shaped protuberances thought to be male gametangia (Torres, 2003). Another uncalcified green alga, which morphologically resembles the existing genus Caulerpa, has been reported to be preserved in Miocene diatomite from California (Parker and Dawson, 1965). This fossil, Caulerpites denticulata, consists of a diffusely branched thallus with broadened, ultimate segments.

Figure 4.16 Calcareous alga Lithothamnium fornicatum (Extant). Bar  2 cm.

TAXA INCERTAE SEDIS Early fossils interpreted as uncalcified green macroalgae include Yuknessia and Margaretia from the Middle Cambrian Burgess shale of Canada and the Marjum and Wheeler Formations in Utah (Walcott, 1919; Conway and Robison, 1988). Yuknessia simplex formed delicate, centimeter-sized thalli composed of unbranched or (repeatedly) dichotomizing erect cylindrical branches which arose from what appears to be a central holdfast. Thalli of M. dorus were threadlike, rarely branched structures up to 2 cm in diameter and probably up to 1 m long. Courvoisiella (FIG. 4.18) is a Devonian uncalcified green macroalga characterized by a comb-shaped thallus constructed of non-septate, branching tubes (Niklas, 1976a). Along some tubes are spherical structures about 200 μm in diameter that have been interpreted as some type of gametangium. Chemical analyses of the fossils confirm the presence of cellulose. These findings, along with morphological data, were used to suggest affinities with the green algae. Impression fossils of relatively large ( 15 cm long), flat, sometimes lobed thalli have been described from the Mississippian Bear Gulch Limestone in Montana (Grogan and Lund, 2002), and the Namurian of Hagen-Vorhalle in Germany (Krings, 2005). Although it is difficult to determine the affinities of these fossils because they lack characteristic features, such as reproductive structures, they have been suggested as superficially resembling the thalli of the modern green alga Ulva lactuca (Ulvales) and the red alga Porphyra umbilicalis (Bangiales). Bubnoffphycos rhombeum, described by Daber (1960) from the Permian of Germany, may also represent an algal thallus comparable to these extant forms. CHAROPHYCEAE

Figure 4.17 Cross section of Ivanovia tebagaensis show-

ing relationship between tissues A. and well-preserved membrane B. Bar  0.5 mm. (Courtesy A. Torres.)

The Charophyceae are generally considered to be the ancestral group within the Chlorophyta that gave rise to the land plants (McCourt et al., 2004). Initially, this relationship


paleobotany: the biology and evolution of fossil plants

Figure 4.19 Carbonaceous tuff consisting of Chara sp. plants (Extant). Bar  2 cm. (Courtesy BSPG.)

Figure 4.18 Portion of Courvoisiella ctenomorpha showing projecting tubules (Devonian). (From Taylor and Taylor, 1993.)

was based on the complex morphology of the charophytes, including differentiation into organs and the presence of enclosing structures around the egg cells (oogonia); molecular phylogenies have confirmed this relationship (Sanders et al., 2003; Turmel et al., 2007). Included in this group are four important orders, Klebsormidiales, Zygnematales, Coleochaetales, and Charales (Lee, 1999). We will discuss the Charales in greater detail and briefly address the Zygnematales. For a comprehensive survey of morphology, paleoecology, stratigraphic distribution, phylogeny, and classification of the charophytes, see Feist et al. (2005b). CHARALES The Charales, commonly known as stoneworts or brittleworts (FIG. 4.19), include six living genera that have been assigned to two tribes within the family Characeae: Chareae, including Chara (FIG. 4.20), Lamprothamnium, Nitellopsis, and Lychnothamnus, and Nitelleae, with Nitella and Tolypella (McCourt et al., 1996). They inhabit freshwater and brackish environments worldwide, where some may exceed 30 cm in length. The thallus is characterized

Figure 4.20 Stand of Chara vulgaris showing oogonia (yel-

low spots) (Extant). (Courtesy M. Feist.)

by distinct nodes and internodes, with whorls of laterals borne at the nodes (FIG. 4.21). Charales reproduce both asexually and sexually; male and female reproductive organs are produced on short branches and the female

chapter 4



Figure 4.22 Polar view of gyrogonite showing spiral cell (Triassic). Bar  500 μm.

Figure 4.21 Portion of Chara axis with two oogonia. (From

Taylor and Taylor, 1993.)

organs of extant forms (oogonia) consist of five spirally arranged tubes or elements (FIG. 4.22) that surround the egg. A fossilized oogonium is termed a gyrogonite (FIG. 4.23) and is the principal fossil evidence of the order (FIGS. 4.24–4.26). The oldest charophyte gyrogonites are known from the Upper Silurian (Feist et al., 2005a). Gyrogonites assigned to the genus Trochiliscus (Devonian, Trochiliscaceae) are characterized by more than five enveloping elements that are dextrally coiled (twisted to the right). The gyrogonites are 0.5 mm in diameter and slightly longer than wide; the base is rounded and the distal end is elongated. Inside a gyrogonite of T. podolicus is a thin (1 μm), continuous membrane that has been interpreted as a remnant of the original oospore. In some sections, patches of disorganized cells are found associated with the membrane. The trochiliscs were initially regarded as marine organisms; now they are believed to have inhabited fresh or brackish water habitats. In contrast with the trochiliscs, the younger Devonian and Carboniferous gyrogonites demonstrate greater structural variability, with at least six elements surrounding the egg cavity, and an open pore at the apex (Peck, 1957).

Figure 4.23 Gyrogonite showing spiral cells in side view (Triassic). Bar  500 μm.

Near the close of the Devonian, the pattern of twisting in gyrogonites changed from dextral to sinistral (to the left). Beginning with Eochara (Middle Devonian, Eocharaceae), believed to be the ancestral form leading to modern taxa, there was a progressive reduction in the number of sinistrally


paleobotany: the biology and evolution of fossil plants

Figure 4.24 Lateral view of Perimheste horrida (Jurassic).

Bar  500 μm. (Courtesy M. Feist.)

Figure 4.26 Lateral view of Atopochara triquetra (Cretaceous). Bar  500 μm. (Courtesy M. Feist.)

Figure 4.25 Lateral view of Flabellochara grovesi (Cretaceous).

Bar  500 μm. (Courtesy M. Feist.)

coiled elements, with five (the common number in living species) established by the Pennsylvanian (Peck and Eyer, 1963). The evolution of modern forms has led to the elimination of the apical pore as a result of the tighter association of the spiral cells at the apex (Grambast, 1974) (FIG. 4.27). The reproductive structures in the Paleozoic charophycean families Sycidiaceae (Silurian–Carboniferous) (FIG. 4.28), Trochiliscaceae (Devonian), Chovanellaceae (Devonian– Carboniferous), and Pinnoputamenaceae (Devonian) include a utricle, a calcified supplementary vegetative cover that is believed to protect the zygote against desiccation and which surrounds the gyrogonite. These fossil charophyte families

Figure 4.27 Louis Grambast. (Courtesy J. Galtier.)

have been placed within a single order, the Sycidiales (Feist et al., 2005a) (FIG. 4.29). In other Paleozoic families, including the Eocharaceae (Middle Devonian–?Triassic), fructifications are more similar to those seen in modern taxa, which do not produce utricles; these families are included in the

chapter 4

sb sc

t pb pc v

Figure 4.28 Suggested reconstruction of Sycidium xizangense

utricle showing vesicle (v), thallus (t), primary (pb), secondary (sb) branches, and primary (pc) and secondary canals (sc) (Devonian). (Modified from Feist et al., 2005a.)

et al., 2005a.)


Charales. The only Paleozoic family containing utricleproducing and utricle-free species is the Moellerinaceae (Silurian–Permian), which has been interpreted as occupying a central position in the early phylogeny of the group. In addition to the large number of fossil gyrogonites that have been described and used as index fossils, the vegetative parts of many fossil charophytes are also known. Two genera have been reported from the Upper Devonian of South Africa (Gess and Hiller, 1995). In Hexachara (FIG. 4.30A) each node produces a whorl of six laterals, and oogonia are produced on each lateral, whereas in Octochara (FIG. 4.30B) a whorl of eight laterals are borne at each node; each lateral is branched and produces an oogonium. Palaeonitella cranii (FIG. 4.31) is a relatively small, anatomically preserved charophyte initially described by Kidston and Lang (1921a) from the Early Devonian Rhynie chert of Aberdeenshire, Scotland. The thallus consists of branched, septate filaments that possess a nodal organization. Associated with some of the filaments are long tubular cells, separated from one another by an enlarged node of


Figure 4.29 Suggested charophyte phylogeny. (From Feist



Figure 4.30 Octochara crassa A. and Hexachara setacea B. (Modified from Gess and Hiller, 1995.)


paleobotany: the biology and evolution of fossil plants

The Clavatoraceae is a large group of exclusively Mesozoic Charales that has frequently been used in biostratigraphy of continental facies (Martín-Closas, 1996). They are known from the Oxfordian (Upper Jurassic) through Cretaceous of all continents except Australia and Antarctica, based on both fructifications and vegetative parts, often in organic connection. Clavatoracean fructifications are composed of an oogonium surrounded by a calcified utricle. Utricles are important as characters in species identification as they underscore morphological variability. The Clavatoraceae shows development of utricles similar to those in Paleozoic sycidialean families, and this is interpreted as being a result of similar external constraints, rather than expressing true phylogenetic relationships (Feist et al., 2005a). Fossils of the clavatoracean genus Clavator (Jurassic– Cretaceous) consist of strongly calcified stems with narrow internodes and six lateral branches in each whorl. Oogonia are located in a single vertical row on the adaxial side of a branch, one per node. The vegetative parts of another genus in the Clavatoraceae, Echinochara from the Morrison Formation (Jurassic) of North America, are known in some detail. The plant body consists of 12 dextrally spiraled, cortical tubes constructed of elongate cells arranged in a linear series. At the distal end of each cortical cell are five long spines. Oogonia were produced in whorls of six. cranii showing three nodes (Devonian). Bar  200 μm. (Courtesy W. Remy and H. Hass.) Figure

4.31 Palaeonitella

small cells. These tubular cells are similar to rhizoids in living Charales. Uncalcified oogonia found associated with P. cranii axes (but not in organic connection) are composed of six sinistrally spiraled cells with an equal number of coronula cells arranged in a single layer around an apical pore (Kelman et al., 2004). The shape of the oogonium is reminiscent of the extant Chareae, whereas the morphology of the thallus is similar to that of the Nitelleae. Another species, P. tarafiyensis, comes from the Upper Permian of Saudi Arabia (Hill and El-Khayal, 1983), and P. vermicularis has been reported from the Lower Cretaceous of Spain, along with three other charalean fossils (Martín-Closas and Diéguez, 1998). One of these, Charaxis spicatus, closely resembles members in the extant genus Chara, whereas the other two have been assigned to Clavatoraxis, a morphogenus created for sterile, verticillated clavatoracean (family Clavatoraceae) vegetative remains that cannot be attributed to any species of gyrogonite.

ZYGNEMATALES Thin sections of chert from the Middle Devonian of New York revealed both marine and freshwater algae (Baschnagel, 1966), including a representative of the Zygnematales (or conjugate algae). Paleoclosterium leptum is formed of solitary, elongate, lunate cells 46 μm long and 5 μm wide that appear morphologically similar to species in the extant genus Closterium. Another fossil member of the conjugates is Palaeozygnema spiralis, which occurs in Cretaceous amber from southern Germany (Dörfelt and Schäfer, 2000). This fossil has unbranched chains of cells, each 20 μm long by 14 μm wide, in which chloroplasts and zygotes are exquisitely preserved. Palaeozygnema (FIGS. 4.32, 4.33) is similar to the modern genus Zygnema, although the process of gametogenesis is apparently different in the fossil. Zygnematacean zygospores reported in palynological samples have been useful in assessing changing depositional environments (Zavattieri and Prámparo, 2006).

EUGLENOPHYTA The Euglenophyta is a diverse group of naked, motile, unicellular organisms characterized by a pellicle composed of

chapter 4



Figure 4.33 Filament of Palaeozygnema spiralis with intact

cells showing disk-like axial plastids (arrows) (Cretaceous). Bar  10 μm. (Courtesy A. Schmidt and H. Dörfelt.)

Figure 4.32 Palaeozygnema spiralis, hypnozygote with heli-

cal ornamentation (arrow) (Cretaceous). Bar  10 μm. (Courtesy A. Schmidt and H. Dörfelt.)

helically arranged, interlocking proteinaceous strips (Walne and Kivic, 1989). The number of flagella is two, but in some species, several may be present. Within the group nutrition ranges from autotrophic to heterotrophic. The group is cosmopolitan and may be found in almost all habitats. Modern euglenoids include 43 genera and 600–800 species. They have historically been classified either as plants (Bold and Wynne, 1985) or as protozoans (Leedale, 1985). Molecular data (small subunit rRNA gene sequences) suggest that the euglenoids diverged far earlier than the radiation responsible for green plants (Sogin et al., 1986). Until recently the fossil record of the group included just a few reports of specimens from Cenozoic rocks believed to represent members of the extant genera Phacus or Lepocinclis (Bradley, 1929) and Trachelomonas (Deflandre and LeNoble, 1948). Moyeria is a Middle Ordovician–Silurian acritarch that is believed to represent a fossil euglenoid (Colbath and Grenfell, 1995). The specimens are spindle shaped, range up to 40 μm long, and display bihelical symmetry. The cells are

characterized by a series of longitudinal spiral strips that mirror the pellicle morphology of Euglena (Gray and Boucot, 1989). The fossils are abundant as early as the Middle Ordovician in non-marine, nearshore environments (Gray, 1988a). The most persuasive fossil euglenoids discovered to date are preserved in amber. Schönborn et al. (1999) reported on a diverse microcoenosis in Mesozoic (Late Cretaceous; Schmidt et al., 2001) amber from southern Germany that contains several types of protozoans, including two euglenoids. Recently, a similarly diverse but slightly older (Early Cretaceous) microcoenosis containing the colorless euglenoid Astasia was discovered in amber from Álva in northern Spain (Ascaso et al., 2005).

DINOPHYTA (DINOFLAGELLATES) In some older treatments, the organisms included within this group are classified within the Pyrrhophyta or fire algae (Chapman and Chapman, 1973). Although most dinoflagellates occur in marine waters, freshwater forms are also known. Most extant dinoflagellates are free-living components of the oceanic plankton, but saprotrophic, parasitic, symbiotic, and holozoic (heterotrophic) forms are present


paleobotany: the biology and evolution of fossil plants

as well. They range from 5 μm to 2 mm (e.g., Noctiluca) in diameter and are characterized by two flagella of unequal length (heteromorphic) and different beating patterns (heterodynamic) (Hausmann et al., 2003). One flagellum extends outward, but the other wraps around the equator of the cell (FIG. 4.37) and contributes to their unique mode of locomotion, by spiraling through the water. The adaptation of these organisms to a wide variety of environments is reflected by a tremendous diversity in form and nutrition, as well as an extensive fossil record (Hackett et al., 2004). Some dinoflagellates produce extensive blooms, called red tides. The toxic by-products of these organisms cause death to fish and shellfish, as well as higher organisms in the food web. Extensive fossil blooms of certain dinoflagellates were no doubt responsible for the formation of some oil deposits in the world, based on a comparison of 4-methylsteroidal hydrocarbons in petroleum deposits with 4-methylsterols in modern dinoflagellates (Robinson et al., 1984). Freshwater dinoflagellates have also been suggested as the source for some oils (Goodwin et al., 1988). Fossil evidence for dinoflagellates is based on organicwalled cysts (FIGS. 4.34–4.36), sometimes termed dinocysts, which are variable in size (25–250 μm in diameter) and shape. Since only about 15% of extant dinoflagellate species produce fossilizable cysts, it is highly probable that the fossil record represents only a small segment of the actual diversity of these organisms through time (Head, 1996). The wall of most dinoflagellate cysts consists of a substance similar to sporopollenin, which accounts for their excellent preservation as fossils. Other dinoflagellate cysts have calcified walls (see below). Cysts are subdivided by a transverse furrow (paracingulum) into an anterior epicyst and a posterior hypocyst (Evitt, 1985; Evitt et al., 1977). On the surface are numerous, polygonal paraplates separated from one another by parasutures; the plates and sutures are given numbers for descriptive purposes. Some cysts are relatively smooth, whereas others possess various forms of ornament, ranging from spinelike processes (FIG. 4.38) to elaborate horns. The earliest probable fossil dinoflagellate is Arpylorus antiquus from the Upper Silurian of Tunisia (Sarjeant, 1978); the oldest undisputed fossil member of this group occurs in the Early Triassic (Fensome et al., 1999). Certain acanthomorphic acritarchs from the Mesoproterozoic Beidajian Formation in North China, however, have been reported to display morphological and ultrastructural features similar to those seen in living dinoflagellates (Meng et al., 2005; Yin et al., 2005), and dinosteroids were isolated from the rock matrix containing the fossils. This suggests that these fossils may represent the oldest dinoflagellates. The Precambrian

Figure 4.34 Hapsidopalla exornata (Devonian). Bar  20 μm. (From Playford, 1977.)

Figure 4.35 Dinoflagellate Invertocysta tabulata (Miocene). Bar  40 μm. (Courtesy L. E. Edwards.)

origin of dinoflagellates is substantiated by geochemical analyses that show a nearly continuous dinosterane record in Precambrian to Cenozoic organic-rich marine rocks (Moldowan and Talyzina, 1998; Moldowan et al., 2001).

chapter 4



20 μm

4.38 Florentinia, a dinoflagellate Bar  20 μm. (Courtesy M. S. Zavada.)

Figure Figure 4.36 Dinoflagellate Wilsonidium tabulatus (Eocene). Bar  25 μm. (Courtesy L. E. Edwards.) Apex Plates




in younger Mesozoic shelf and slope sedimentary environments throughout the world (Keupp, 1992), and have been used as valuable proxy indicators for (paleo-)environmental conditions (Zonneveld et al., 1999). Important features in the classification of these fossils include ultrastructure of the calcitic cyst walls, orientation of the crystallographic c-axes of the wall crystals (Young et al., 1997; Kohring et al., 2005), and position and configuration of the excystment aperture or archeopyle (Streng et al., 2004).





Figure 4.37 Structure and morphology of a dinoflagellate cell. (From Evitt, 1985.)

The hydrocarbon dinosterane is concentrated in and nearly exclusive to the dinoflagellates, so this chemical evidence strongly indicates that dinoflagellate lineages must have existed as early as the Precambrian. The small-sized (10 or 75 μm in diameter) cysts of calcareous dinoflagellates (Calciodinellaceae sensu lato; Gottschling et al., 2005), which are sometimes also termed calcispheres, initially occur as fossils in the Late Triassic, becoming highly diverse in the Cretaceous and throughout the Cenozoic (Kohring et al., 2005). They occur in rock-forming abundance

This large and diverse group of phototrophic and heterotrophic organisms is characterized by motile cells that typically have two unequal flagella, that is, a forwardly directed tinsel flagellum and posterior directed whiplash flagellum. The tinsel flagellum is covered with lateral bristles (mastigonemes), whereas the posterior flagellum is smooth and usually shorter or sometimes reduced to a basal body (Lee, 1999). All are golden or brown in color and several classes are generally recognized, including the Synurophyceae, Pelagophyceae, Raphidiophyceae (chloromonads), Eustigmatophyceae, Chrysophyceae (golden algae), Bacillariophyceae (diatoms), Dictyochophyceae (silicoflagellates), Xanthophyceae (yellow-green algae), and Phaeophyceae (brown algae). We will cover only the last four groups here. BACILLARIOPHYCEAE (DIATOMS)

The diatoms today include more than 100,000 extant species (Round et al., 1990). Most diatoms are unicellular organisms (FIG. 4.39) (a few are filamentous) that are ubiquitous in water,


paleobotany: the biology and evolution of fossil plants

4.39 Diatom Cyclotella meneghiniana (Extant). Bar  10 μm. (From Hoops and Floyd, 1979.)


occurring in environments that include freshwater, brackish, terrestrial, subaerial, and marine habitats (Round et al., 1990; Hausmann et al., 2003). They are an important part of the marine phytoplankton and, as primary producers, are estimated to fix at least one-quarter of the inorganic carbon fixed in the ocean (Granum et al., 2005). Many are planktonic, but they also occur as epiphytes on, or endophytes in, other organisms. Diatoms are constructed of a cell enclosed by a rigid, cell wall or shell (FIG. 4.39) composed of opaline silica (silicon dioxide, SiO2) coated by an organic material. This shell or frustule is constructed much like a Petri dish or a box in that one valve fits inside the other valve, and it is the reason that the fossil record of diatoms is so excellent. When a diatom cell dies, the frustule separates into two parts which settle to the bottom. Deposits of fossil diatoms, termed diatomaceous earth, may be more than a thousand feet thick in some areas of the world, for example, in Lompoc, California, where diatomaceous earth is mined and has several commercial uses, as an abrasive or in filtration. Frustule shape and elaborate ornamentation are two of the important characters used to classify fossil diatoms. Two of the most common shapes are centric (radial) and pennate. The fossil record of the diatoms is extensive and can be traced into the Mesozoic (Sims et al., 2006). The earliest unambiguous marine diatoms are known from the Lower Jurassic (Rothpletz, 1896; Barron, 1987), and the earliest diverse diatom assemblages occur in the Lower Cretaceous. Molecular clock hypotheses, however, suggest an earlier origin of the group (Kooistra et al., 2003). Medlin et al. (1997) suggested that the origin of diatoms may be related

to the end-Permian (250 Ma) extinction event, after which there were presumably many marine niches available. The oldest fossil freshwater diatoms to date are latest Cretaceous (Maastrichtian) and have been described from several different localities, including the Tarahumara Formation in northern Mexico (Chacón-Baca et al., 2002) and the Deccan Intertrappean beds and Lameta Formation of India (Ambwani et al., 2003). Diatoms, including exceptionally well-preserved organelles (e.g., internal membranes, lamellate plastid fragments, extracellular mucilage bodies), have been reported from Eocene terrestrial sediments (A. Wolfe et al., 2006). Because of their elaborate ornamentation and excellent preservation, diatoms are important in biostratigraphy. They have also been widely used in paleoecology as indicators of past environments. Although the use of molecular data has proved useful in determining the relationships among various groups of extant diatoms, evaluating morphological characters and determining synapomorphies in groups of fossil diatoms have not rapidly moved forward (Williams, 2007). DICTYOCHOPHYCEAE (SILICOFLAGELLATES)

The silicoflagellates comprise a small group of autotrophic marine, planktonic organisms that range from 20 to 50 μm in length and, as the name implies, possess a siliceous skeleton composed of opaline rods fused together to form a network (Preisig, 1994; Desikachary and Prema, 1996). Some classify them with the diatoms. They are first encountered in Early Cretaceous sediments (McCartney et al., 1990), and their peak abundance is reduced in the Cenozoic (Hausmann et al., 2003) so that they are represented by a single family today. Silicoflagellates have not been used extensively as biostratigraphic markers because of the relatively slow rate of evolution within group. Their value as fossils lies in their apparent sensitivity to temperature, and they are useful for biostratigraphy at higher latitudes and in deeper water, where calcareous microfossils are less common, as well as paleoenvironmental indicators. They are potentially useful in the study of paleoclimatology and perhaps in determining productivity levels in ancient ecosystems. XANTHOPHYCEAE (YELLOW-GREEN ALGAE)

The yellow-green algae are a group of heterokonts composed primarily of freshwater forms and a few marine representatives. Many species are single-celled organisms, whereas others are colonial, living as naked cells in a gelatinous envelope, or produce long filaments of cells. The group also includes a number of coenocytic forms such as the water felt Vaucheria (Vaucheriales). Molecular data have shown that the

chapter 4



Xanthophyceae are most closely related to the Phaeophyceae (Potter et al., 1997). The fossil record of the Xanthophyceae is scanty. One fossil representative is Palaeovaucheria from the one-billion-year-old Lakhanda Formation in Siberia. This fossil alga displays morphological traits characteristic of vaucherian xanthophytes, including branching at right angles, two sizes of filaments on the same individual, and terminal pores and septae at filament ends (Javaux et al., 2003). Other vaucherian algae have been reported from middle Neoproterozoic shales in Spitsbergen (Butterfield, 2004, 2007). PHAEOPHYCEAE (BROWN ALGAE)

The brown algae consist of 1500 extant species in more than 250 genera (Norton et al., 1996). Phaeophycean thalli range in size from microscopic uniseriate filaments, for example, in the Ectocarpaceae, to complex plant bodies with significant organ, tissue, and cellular specialization (Graham and Wilcox, 2000). In some species of the giant kelp Macrocystis (Laminariales), the thallus is more than 70 m long. With two exceptions (the genera Newhousia and Padina; Kraft et al., 2004), all brown algae are uncalcified, which has contributed to the absence of a well-defined fossil record for the group. Brown algae have large amounts of the carotenoid fucoxanthin which gives them their characteristic color, and stored food is in the form of laminarin. Modern representatives typically inhabit colder waters in the northern hemisphere; in the tropics they occur in large numbers in the Sargasso Sea region of the Atlantic Ocean. One major problem in identifying fossil brown algae is their morphologic similarity to some members of the Rhodophyta. Several of the impression–compression thalloid taxa mentioned in this chapter (Rhodophyta, Chlorophyta) may belong to the brown algae, as Leary (1986) noted in his description of three uncalcified algal impressions from Mississippian strata. One of these, Phascolophyllaphycus, has elongate blades with rounded apices and tapered bases attached to the stipe in a helical arrangement. Several pneumatocysts, air-filled sacs that are characteristic of the brown algae, occur at the base of the blades; each is 1 mm in diameter. Early megafossils interpreted as phaeophycean algae come from the Lower Cambrian of southwestern China (Xu, 2001a). Thalli of Punctariopsis latifolia are up to 20 mm high and consist of single or clumped unbranched, leaflike blades, each of which is attached to the substrate by a short, narrow stipe and a globose or irregularly shaped holdfast (FIG. 4.40). In P. simplex, the blades occur singly and have a globose base. Vendotaenia antiqua was first reported from the Neoproterozoic (late Vendian) of the former Soviet Union (Gnilovskaja, 1971) but is also known from the Lower

Figure 4.40 Reconstruction of Punctariopsis latifolia. (From

Xu, 2001.)

Cambrian of China (Xu, 2001a). The thallus is composed of ribbonlike, unbranched or occasionally branched blades up to 85 mm long, which bear sporangia. Several algal morphotypes have been described from rocks of Late Ordovician age in Canada (Fry, 1983). The preservation of the carbonaceous specimens and the sedimentology of the deposits suggest that these algae probably grew in relatively shallow water. Winnipegia has an axis bearing wedge-shaped appendages, each 3.5 cm long and lacking surface features. The thousands of specimens encountered in the rocks suggest that the plants were not transported a great distance and were probably buried close to the site in which they grew. Thalassocystis is a compressed middle Silurian alga that is thought to be either a member of the red or brown algae (Taggart and Parker, 1976). The thallus is branched, with each branch terminating in an inflated bladder about 2.5 cm long. These algae were deposited in a shallow-water marine environment, but the actual habitat of the plants is not known. Another thalloid fossil with possible affinities in the Phaeophyceae is Yeaia africana from the Late Devonian Witpoort Formation in South Africa (Hiller and Gess, 1996), a form that resembles Y. flexuosa from the Upper Silurian Baragwanathia flora (Chapter 8) of central Victoria (Douglas, 1983). The thallus of Y. africana consists of straplike, repeatedly dichotomizing blades that are ornamented with tiny spots. Brown algae resembling members of the Laminariales are known from Miocene rocks of the Monterey Formation in California (Parker and Dawson, 1965). One of the most interesting forms is Julescraneia, a large compound thallus


paleobotany: the biology and evolution of fossil plants

Figure 4.42 Emiliania huxleyi (Quaternary). Bar  1 μm. (Courtesy J. R. Young.)

Figure 4.41 Cystoseirites altoaustriacus thallus fragments.

partschii from Romania (Givulescu, 1975) and the Ukraine (Molhanov, 2004).

Bar  1 cm. (Miocene). (Courtesy BSPG.)

PRYMNESIOPHYTA (HAPTOPHYTES) with lateral branches arising from a central pneumatocyst 16 cm in diameter. Lateral branches extend up to 3.6 cm in diameter and are believed to have terminated in a blade much like that in the extant brown algae Pelagophycus and Nereocystis. Also present in the diatomaceous sediments are a large number of brown algae that can be directly compared to members of the extant family Cystoseiraceae (Fucales). One of these, Paleohalidrys, is a flat, linear thallus 40 cm in length and often demonstrating pinnate branching of the laterals. In many modern members of this family, the blades can become disassociated seasonally, and it is believed that this phenomenon is responsible for the large number of specimens found in the Miocene deposit. Other fossils interpreted as belonging to the Cystoseiraceae have been reported from the Cenozoic of central Europe (reviewed in Kovar, 1982). Among these is Cystoseirites altoaustriacus, a repeatedly branched thallus composed of slender, cylindrical branch segments and spherical or egg-shaped air vesicles (aerocysts), each 2–4 mm in diameter (FIG. 4.41). Another Cenozoic form closely resembling extant members of the Cystoseiraceae is Cystoseirites (or Cystoseira)

The Prymnesiophyta, also known as Haptophyta, is a group of autotrophic, planktonic uninucleate flagellates characterized by the presence of a haptonema (a filamentous, microtubulesupported appendage) that lies between two smooth, approximately equal flagella (Lee, 1999; Andersen, 2004). The group includes at least 500 extant and many more fossil species. Geologically important members of the Prymnesiophyta are certain calcareous nannofossils termed coccolithophores or coccolithophorids (FIG. 4.42). Surrounding the living cell of these organisms are small (20 μm in diameter), calcified scales termed coccoliths which demonstrate a complex morphology and structure. Coccolithophores and coccoliths are valuable biostratigraphic markers, as well as indicators of paleoclimate (Wise, 1988). The living counterparts of these unicellular organisms are included within the Coccolithophorales (Jordan and Chamberlain, 1997), a group that is principally marine, and currently makes up 45% of the total phytoplankton in middle latitudes. Coccolithophores have a significant impact on their environment since they are the major primary producers that convert dissolved CO2 in the ocean to calcium carbonate (CaCO3) (Rost and Riebesell,

chapter 4

2004; Baumann et al., 2005) and therefore influence global biogeochemical cycles. Moreover, they have the ability to increase the albedo of the Earth by reflecting light from their coccolith-covered surfaces and by producing dimethylsulfide, a gas that contributes to the formation of aerosols which enhance cloud formation in the atmosphere (Brand, 1994). Coccolithophores live in the upper 200 m of the water column and, therefore, fossil assemblages can be used to delineate former oceanographic and climatic conditions of the surface waters. Coccoliths almost always occur in the fossil record as isolated entities; rarely are the scales found attached to the nannoplankton that produced them. Because of the small size of the scales, electron microscopy has become an important research tool to study coccoliths. Based on the structure and morphology of the scales, several artificial families have been established (Bown and Young, 1997; Young and Bown, 1997a, b). Coccoliths are commonly divided into two major groups: heterococcoliths are constructed of crystal elements that differ in size and shape, whereas the crystal elements making up holococcoliths are essentially identical in size and shape (Siesser and Winter, 1994). In addition, a third category of similar status, nannoliths, occur as fossils and are most commonly defined as calcareous nannofossils of uncertain affinity, but probably related to the coccolithophores (Young and Bown, 1997b; Young et al., 1997). Nannoliths first occur in the Carnian (Late Triassic); coccoliths appear somewhat later, during the Norian, and are particularly abundant in the younger Mesozoic and Cenozoic. Coccolithophores appear to have been at their zenith during the Late Cretaceous (PerchNielsen, 1985; Bown et al., 2004), where their coccoliths often form thick deposits; the famous White Cliffs of Dover, England, consist largely of coccoliths. Their amazing abundance during the Cretaceous has been attributed to the chemistry of seawater at the time—a low Mg/Ca ratio and high Ca concentration (Stanley, 2006). In contrast, modern seawater has low Ca concentration and a high Mg/Ca ratio, which apparently limits coccolithophore population growth today. Several ideas have been advanced regarding the function of the scales in coccolithophores (Young, 1994). One suggests that the scales function to shield the cell from excessive light, although perhaps the more popular corollary argues that the convexo-concave surface of the scales actually focuses light into the cell. Their small size, along with their abundance in younger Mesozoic and Cenozoic rocks and typically restricted stratigraphic range, has made coccoliths important index fossils and biostratigraphic markers. An analysis of their distribution through time indicates that a major extinction event occurred during the latest Cretaceous,



followed by a recovery and radiation during the early Paleocene. This recovery, however, did not result in species richness similar to that seen in the Late Cretaceous (Bown et al., 2004). The discoasters are a distinct type of calcareous nannofossil believed to be related to the coccolithophores; in some treatments, they are included within the nannoliths (Bown et al., 2004). The tiny skeletons of discoasters appear as stars or rosettes, between 10 and 35 μm in diameter. They lack modern analogs but were a conspicuous component of the nannoplankton during most of the Cenozoic (Aubry, 1984). Discoasters became extinct at the end of the Pliocene (Chepstow-Lusty and Chapman, 1995; Kitaeva et al., 1997). They differ from coccoliths in being composed of tubular forms of calcite, whereas the coccoliths are formed of rhombohedral and hexagonal calcite crystals.

RHODOPHYTA (RED ALGAE) Rhodophyta, or red algae, are distinguished from other algal groups by the presence of chlorophylls a and d in combination with certain accessory pigments (phycobiliproteins), non-aggregated photosynthetic lamellae in the chloroplasts, specialized food reserves, unique sexual reproduction, and the absence of flagellation in all phases of the life cycle. The 4000–6000 extant species are primarily marine, mostly inhabiting warm tropical waters (Graham and Wilcox, 2000; Sounders and Hommersand, 2004). Phylogenies based on molecular data suggest the group is monophyletic (Le Gall and Saunders, 2007). The red algae are almost all multicellular and structurally more complex than other algae, with specialized pit connections between cells and a complicated mode of reproduction. Various red algae are commonly preserved as fossils, because they possess calcified skeletons that form as a result of calcium carbonate precipitation within the cell walls. In this feature, they differ from other lime-precipitating algae that deposit calcium carbonate only on the thallus. In the reds, the calcite is typically deposited in a grid-like pattern. Red algae are widespread in the fossil record, extending back to the late Mesoproterozoic. To date, the oldest red alga is a member of the Bangiales, Bangiomorpha pubescens (FIGS. 4.43, 4.44), from the Hunting Formation (1.2 Ga) of Somerset Island, arctic Canada (Butterfield et al., 1990; see Chapter 2). Not only is B. pubescens the oldest taxonomically resolved eukaryote on record, but it also exhibits the oldest example of eukaryotic sex and complex multicellularity (Butterfield, 2000, 2001).


paleobotany: the biology and evolution of fossil plants

Figure 4.44 Bangiomorpha pubescens (Mesoproterozoic).

Bar  50 μm. (Courtesy N. J. Butterfield.)

these reproductive propagules are impossible to distinguish from vegetative cells. Figure 4.43 Bangiomorpha pubescens (Mesoproterozoic). Bar  50 μm. (Courtesy N. J. Butterfield.)

Most Paleozoic calcareous red algae grew in open-marine carbonate shelf environments, although, as a group, they tolerated a variety of environments. These wide environmental variations suggest that individual taxa may provide important clues relative to ancient environments. Far more difficult, however, is the problem of relating Paleozoic red algae to living groups, because the taxonomy of the fossil forms often remains uncertain. This is due in part to the fact that reproductive structures have rarely been described. Their absence has suggested to some that the reproductive organs were externally produced and not calcified. Others suggest, however, that spores were produced within cells and that


One of the families traditionally placed within the Rhodophyta is the Solenoporaceae. Although several genera were initially assigned to the animal kingdom (e.g., as tabulate corals, bryozoans, or sponges; see Cózar and Vachard, 2006), solenoporaceans were later generally interpreted as calcified red algae (Pia, 1927) (FIG. 4.45). Today, however, the Solenoporaceae is known to represent a heterogeneous group that includes a variety of animals, red algae, and cyanobacteria, and, as a result, it is no longer possible to support the concept of the Solenoporaceae (Ordovician– Miocene) as a coherent family (Riding, 2001). Solenoporaceans were nodular or encrusting marine organisms composed of closely packed, radially or vertically divergent rows of tubes. Their diameters were almost

chapter 4

always larger than those of living coralline red algae (see section “Other Calcified Red Algae”), which are believed to be related structurally. One common representative is Solenopora (FIGS. 4.46–4.48), established for irregularly lobed, calcium carbonate masses composed of radiating, juxtaposed tubes with shared walls. The type species, S. spongioides from the Upper Ordovician of Estonia, was initially interpreted as a chaetetid sponge, but later transferred to the red algae. Reexamination of the original illustrations and new material from the type locality, however,

Figure 4.45

Julius Pia.



indicates that the initial interpretation was accurate and thus removes Solenopora from the algae (Riding, 2004). Another form initially placed in Solenopora is S. gotlandica from the Silurian of Sweden and Wales. This species, which represents a true red alga, was transferred to the genus Graticula (Brooke and Riding, 1998, 2000). Skeletons of G. gotlandica (FIGS. 4.49, 4.50) may be free nodular or massive encrusting and are composed of laterally joined columns, branches, or pillars, which are mushroom- or umbrella-like, and occasionally up to 10 cm high (Nose et al., 2006). The columns consist of erect-to-radiating, juxtaposed filaments, which are rounded to polygonal or irregular in cross section, and which share adjacent walls; cross partitions (cross walls) in adjacent filaments are sometimes aligned. The arrangement of cross partitions, along with the presence of sporangia in sporangial compartments arranged in irregular sori, separates Graticula from other forms traditionally placed in the Solenoporaceae. Based on these features, Brooke and Riding (1998, 2000) established the new family Graticulaceae, which they assign to the Corallinales. The Paleozoic Graticulaceae are structurally similar to the Sporolithaceae (Corallinales); the type species G. gotlandica closely resembles members in the earliest recorded modern Corallinales, whose fossil record only extends back into the Early Cretaceous (Braga and Bassi, 2007; Tomás et al., 2007). The solenoporacean genus Parachaetetes is found as early as the Late Carboniferous, and Cenozoic forms are especially common in shallow-water reef facies (Wray, 1977). It occurs as bluntly lobed growths up to several centimeters in diameter. In vertical thin sections, typical forms exhibit a tightly packed mass of elongate cells arranged in curved radial lines. In cross section, these cells are circular and variable in length. At least some species within the genus Parachaetetes

Figure 4.46 Solenopora sp. (Jurassic). Bar  2 cm. (Courtesy M. Nose.)


paleobotany: the biology and evolution of fossil plants

4.49 Red algal–calcimicrobe boundstone with Graticula gotlandica (Silurian). Bar  5 cm (Courtesy M. Nose.) Figure

Figure 4.47 Solenopora condensata (Jurassic). Bar  2 cm

(Courtesy M. Nose.)

Figure 4.48 Solenopora sp. (Jurassic). Bar  2 cm (Courtesy


seem to represent true red algae (Aguirre and Barattolo, 2001), whereas others, again, are interpreted as chaetetid sponges today (Riding, 2004). The cells of Solenomeris are polygonal in outline, with those of adjacent rows forming a zigzag configuration. Formerly considered to be a red alga, today Solenomeris belongs with the incrusting foraminifera (Bassi, 2003). Marinella lugeoni (Late Jurassic–Oligocene) was originally described as a cyanobacterium but has more recently been included in the Codiaceae (Chlorophyta) and the Solenoporaceae. This organism forms either encrusting thalli, several centimeters in diameter, or erect, digitiform, and branching thalli with branches up to 9 mm high that are

Figure 4.50 Graticula gotandica framestone from Gotland with laterally linked red algal pillars (Silurian). Bar  2 cm (Courtesy M. Nose.)

attached to the substrate by a narrow base. The internal tissue is composed of radially oriented and densely packed filaments. Specimens of M. lugeoni from the Upper Jurassic of Portugal imply a close relationship with the corallinaceans, but structural similarities to the solenoporaceans have also been noted (Leinfelder and Werner, 1993). What were initially thought to be oval, aggregate reproductive structures in the

chapter 4

Miocene genus Neosolenopora (Mastrorilli, 1955) have been reinterpreted as an unusual bryozoan (Tillier, 1975). Several other genera traditionally referred to the Solenoporaceae, such as Metasolenopora, Petrophyton, and Solenoporella may belong, or be related to the Corallinaceae (Riding, 2004). The identification of various organisms traditionally included in the Solenoporaceae has been useful in understanding how skeletal organisms and microbes interacted to produce certain reef limestones (Adachi et al., 2007). OTHER CALCIFIED RED ALGAE

Fossil red algae, like many of those today, were important in the formation of reef communities as early as the Silurian. The Graticulaceae (see above) were important structural components of Silurian (Nose et al., 2006) and Devonian (Adachi et al., 2007) reefs. Another reef-building red alga from the Devonian is Archaeoamphiroa, which occurs along a 48-km outcrop within the Alexandra Reef Complex (Frasnian) in the Northwest Territories, Canada (Magathan, 1985). This fossil closely resembles the extant coralline red alga Amphiroa (Corallinales) and consists of alternating long and short rows of cells. The lithofacies suggest that this alga grew in a shallow, surf-swept reef in association with other filamentous algae. Masloviporidium is a cosmopolitan calcareous red alga from the Carboniferous, ranging from Russia to central Texas (Groves and Mamet, 1985). The thallus is sheetlike and consists of rows of wedge-shaped cells surrounded by calcified partitions. Each cell is connected to cells above and below by pores in the partitions. Nothing is known about the mode of growth or reproduction. Another Carboniferous (Pennsylvanian) red alga from marine rocks is Litostroma (Mamay, 1959a). This small alga is formed of irregularly shaped, thalloid platelets that are one cell thick and up to 6 mm in diameter. Filaments arise from the surface of the thallus, and irregularly shaped perforations are present along the margin. Many of the cells contain shrunken cell contents or nuclei. In addition to being assigned to the red algae, Litostroma has also been suggested as a member of either the green or brown algae. CORALLINALES One of the geologically most important orders of marine calcified red algae is the Corallinales. The crown group Corallinales includes two living families, the Corallinaceae and the Sporolithaceae, both of which have been documented from the Mesozoic and Cenozoic (Aguirre et al., 2000). Both families are characterized by macroscopic crustose or erect and branched thalli differentiated into two distinct histologic zones. The hypothallus, or medulla, which forms the basal



part of the crustose plants and central part of the erect forms, is constructed of loosely arranged, relatively large cells. The perithallus (cortex), or second zone, which constitutes the major portion of the thallus, is located above the hypothallus in crustose forms and outside of the hypothallus in branched forms. It is characterized by smaller cells (Lee, 1999). In addition, a distinct surface layer, the epithallus, may be present and this is composed of one to a few layers of small, thin-walled cells (Xiao et al., 2004). Sporolithaceae and Corallinaceae are indistinguishable in vegetative anatomy but differ in sporangial structure: the former family is characterized by sporangial chambers grouped in sori, whereas the latter produces sporangial conceptacles with one (uniporate; Lithophylloideae and Mastophoroideae) or several (multiporate; Melobesioideae) small pores for the dispersal of spores (Aguirre et al., 2000). A common member of the Sporolithaceae is Lithothamnion (FIG. 4.16), which can be traced from Cretaceous tropical seas to present-day temperate and polar areas (Johnson,

Figure 4.51

J. Harlan Johnson.


paleobotany: the biology and evolution of fossil plants

1962) (FIG. 4.51). Earlier (Late Jurassic) records of the genus remain questionable with regard to both age attribution and taxonomic circumscription (Aguirre et al., 2000). Taxa demonstrate a variety of growth forms, ranging from tiny crusts to large aggregations 30 cm in diameter. Based on a study of Miocene non-geniculate coralline red algae from Crete (Greece), Kroeger (2007) suggested that various species of Lithothamnion, along with other forms, such as Sporolithon, can be used as valuable proxy indicators for paleoenvironmental studies in Cenozoic sediments, because these algae appear to form specific and consistent associations depending on water depth and water temperature. Early Cretaceous coralline algae from India have also been used as a proxy indicators for water depth and water temperature (Misra et al., 2006). Another widely cited genus is Archaeolithothamnion (Johnson, 1963), which has been regarded as a junior synonym of Sporolithon (Moussavian and Kuss, 1990; Tomás et al., 2007). One Early Cretaceous (Barremian–Albian) species originally assigned to Archaeolithothamnion is Parakymalithon phylloideum (Moussavian, 1987). The thallus of this alga is composed of a relatively thick hypothallus and only a few distinct horizontal rows of cells in the perithallus. Asexual reproduction occurs in the form of ovalshaped or fusiform sporangia arranged in a nemathecium-like receptacle that lies cushioned above the perithallial tissue. Parakymalithon is believed to be phylogenetically intermediate between the Sporolithaceae and Corallinaceae. Recently, however, the separate status of Parakymalithon as a genus has been questioned (Tomás et al., 2007). Karpathia (FIG. 4.52) is an example of a Cenozoic (Paleocene) form assigned to the Corallinaceae (Bassi et al., 2005). The hypothallus of this encrusting alga is composed of straight, loosely arranged filaments of irregularly shaped, large cells, which resulted from cell fusion. The perithallus is formed of densely arranged filaments or a series of smaller, thick-walled cells. Sporangial conceptacles are uniporate. UNCALCIFIED RED ALGAE

The most exquisitely preserved, early uncalcified red algae, some of which may be distantly related to the Corallinales, were discovered in the Neoproterozoic (600 Ma) Doushantuo Formation at Weng’an, South China (Xiao et al., 2004, and references therein). The fossils are phosphatized and thus possess cellular preservation. They are particularly interesting because they help bridge the stratigraphic and evolutionary gap between the earliest rhodophyte fossils and the calcified red algae described from younger strata. The most distinctive feature of the phosphatized Doushantuo algae described by Xiao et al. (2004) is their filamentous construction or cell fountain

Figure 4.52 Section of Karpathia sphaerocellulosa thallus (Paleocene). Bar  300 μm. (From Bassi et al., 2005.)

architecture. Thalli of Wengania globosa are broadly spherical, nodular, or irregular, 70–750 μm in diameter in thin section, and display a simple pseudoparenchymatous construction (FIG. 4.53), in which cortex and medulla are not differentiated. The thalli consist of cuboidal cells that form regular files, radiating outward and branching toward the thallus margin. In W. exquisita, cell files are less regularly arranged, whereas in the third species, W. minuta, regular cell files are absent (Xiao, 2004). Similar simple thalli are formed by Thallophycoides pholeatus and Gremiphyca corymbiata (FIG. 4.54). These thalli may be lobed, however, with the most profound lobation occurring in G. corymbiata. Thalli of Thallophyca ramosa and T. corrugata (FIG. 4.55) display a complex pseudoparenchymatous construction, in which there is a clear differentiation into an inner medulla and outer cortex (FIG. 4.56), with cortical cells either smaller than the medullary cells or arranged differently, and clustered cell islands interpreted as reproductive cells (Y. Zhang et al., 1998). The pseudoparenchyma is arrayed as upward diverging splays of filaments or cell fountains. Diverging filaments form fan-shaped lobes separated by deep invaginations. Particularly striking are cylindrical, cell-lined invaginations that resemble conceptacles. Thalli of Paramecia incognata are millimeter-sized nodules characterized by cortex– medulla differentiation, thallus compartmentalization (FIG. 4.57), and absence of well-developed invaginations (FIG. 4.58). Xiao et al. (2004) stated that the complex pseudoparenchymatous thalli of Thallophyca and Paramecia display features resembling those seen in some Paleozoic members of the Corallinales, and thus these forms may be interpreted as stem group corallinaceans. They are clearly different from the crown group corallines, however, in that they are not calcified in life (Xiao and Knoll, 1999).

chapter 4

4.53 Simple pseudoparenchymatous thallus of Wengania globosa (Neoproterozoic). Bar  50 μm. (From Xiao et al., 2004; courtesy S. Xiao.)




Figure 4.55 Complex pseudoparenchymatous thallus of Thallophyca corrugata (Neoproterozoic). Bar  100 μm. (From Xiao et al., 2004; courtesy S. Xiao.)

Figure 4.56 Detail of complex pseudoparenchymatous thallus Figure 4.54 Gremiphyca corymbiata showing simple pseu-

doparenchymatous, lobed thallus (Neoproterozoic) Bar  100 μm. (From Xiao et al., 2004; courtesy S. Xiao.)

Another genus of anatomically preserved multicellular algae from the Doushantuo Formation is Sarcinophycus (Xiao and Knoll, 1999; Xiao, 2004). It differs from the previously detailed Doushantuo algae in that the thalli lack the cell fountain architecture and possess marginal protuberances

of Thallophyca corrugata (Neoproterozoic). Bar  100 μm. (From Xiao et al., 2004; courtesy S. Xiao.)

(FIGS. 4.59, 4.60). The systematic affinities of Sarcinophycus remain uncertain. Paratetraphycus giganteus (Z. Zhang, 1985) is yet another multicellular alga-like organism from the Doushantuo Formation. Although initially assigned to the cyanobacteria (Chroococcaceae-like coccoids; Z. Zhang, 1985), certain features have been noted that are reminiscent of extant bangiomorphic red algae (Y. Zhang et al., 1998;


paleobotany: the biology and evolution of fossil plants

Figure 4.58 Complex parenchymatous thallus of Paramecia incognata (Neoproterozoic). Bar  100 μm. (From Xiao et al., 2004; courtesy S. Xiao.)

Figure 4.57 Detail of thallus of Paramecia incognata showing

compartmentalized clusters (arrow) of larger, probably reproductive cells (Neoproterozoic). Bar  100 μm. (From Xiao et al., 2004; courtesy S. Xiao.)

Saunders and Hommersand, 2004). Paratetraphycus giganteus has been reported from the Meso-Neoproterozoic (middle Riphean) in the southern Urals (Sergeev and Lee, 2006). Some of the earliest compression fossils indicative of diverse, marine uncalcified macroalgal floras come from the Neoproterozoic of South China. These morphotypes, such as Doushantuophyton (centimeter-sized thallus fragments composed of erect, repeatedly forking branches) (FIG. 4.61), Konglingiphyton (centimeter-sized dendritic thalli), and Miaohephyton (millimeter-sized dichotomously branched thallus fragments characterized by horizontal constrictions), suggest affinities with the Rhodophyta. Other genera, such as Enteromorphites (centimeter-sized, branched thalli

attached to the substrate by a holdfast) (FIG. 4.62), Gesinella (centimeter-sized, lanceolate or strap-shaped thalli with a basal holdfast), and Yemaomianiphyton (centimeter-sized tuft-like thalli composed of a strong holdfast and forked or unforked erect branches), have been interpreted as either red, green, or brown algae (Steiner, 1994). It has also been interpreted as fragmentary specimens of Enteromorphites (Xiao et al., 2002). A detailed study on the holdfast structures formed by some of the Neoproterozoic macroalgae from China and their paleoenvironmental implications has recently been published by Wang and Wang (2006). The assignment of Miaohephyton to the red algae has been questioned by Xiao et al. (1998), who suggested that this form may represent a brown alga with possible affinities in the order Fucales. Enteromorphites siniansis is an organism composed of several forked hollow tubes 17 mm long that arise from a basal holdfast. It was initially interpreted as being structurally similar to the extant green alga Enteromorpha (Ulvales) (Zhu and Chen, 1984), but Steiner

chapter 4



Figure 4.60 Detail of Sarcinophycus radiatus cell packets

(Neoproterozoic) Bar  25 μm. (From Xiao and Knoll, 1999; courtesy S. Xiao.) Figure 4.59 Sarcinophycus radiatus showing radiating packets of cells (Neoproterozoic) Bar  100 μm. (From Xiao and Knoll, 1999; courtesy S. Xiao.)

Figure 4.61 Doushantuophyton lineare. (Redrawn from Steiner, 1994.)


paleobotany: the biology and evolution of fossil plants

Figure 4.62 Enteromorphites intestinalis thallus. (From Xu,

Figure 4.63 Reconstruction of Paradelesseria sanguinea.


(From Xu, 2004.)

(1994) noted that this structure may also resemble various red and brown algae. Enteromorphites intestinalis from the Early Cambrian Chengjiang Biota in southeastern China is a considerably larger (7 cm high) form, composed of hollow unbranched tubes that were attached to the substrate by a basal rhizoidal cell or tubular proliferations (Xu, 2001b). Paradelesseria sanguinea is a Delesseria- or Phycodryslike red alga from the Chengjiang Biota (Xu, 2004). The thallus is composed of leaf-like, lanceolate, or oblanceolate and petiolate blades 90 mm long that are attached to a subcylindrical, nodose stipe (FIG. 4.63). Two other enigmatic organisms from the Chengjiang Biota that have been variously interpreted as algae are Longfengshania cordata, a small thallus (2 cm high) composed of a bladelike distal portion and a hollow, stalk-like proximal portion (FIG. 4.64), and Plantulaformis sinensis (1 cm high), a cotyledon-like thallus composed of a two-parted distal blade and hollow, stalk-like proximal portion (FIG. 4.65) (Xu, 2002). Hofmann (1985a) suggested that Longfengshania may represent algae with possible affinities in the red or brown algae. Z. Zhang (1988), however, interpreted Longfengshania as an early

bryophyte. Another diverse early algal flora has been reported from the Middle Cambrian Kaili Biota in Guizhou Province, China (Yang et al., 2001). This flora consists of more than 20 genera, 5 of which have tentatively been assigned to the red algae, Palaeocodium, Paraamphiroa, Wahpia, Dalyia, and Bosworthia. Thalli of Paraamphiroa siniansis are 1.5 cm high and composed of a main branch that distally produces a cluster of second-order, bi- or trifurcating branches, each consisting of uncalcified joints and calcified, cylindrical segments. This alga is believed to represent the earliest fossil evidence for a calcified red alga (Yang and Zhao, 2000). Dalyia has also been reported from the Middle Cambrian Conasauga Formation in northwestern Georgia (USA) where it co-occurs with several putative green algae (Schwimmer and Montante, 2007), and from the Middle Cambrian Burgess Shale in British Columbia, Canada (Walcott, 1919), along with a second red alga, Waputikia ramosa (FIG. 4.66). The latter differs markedly from other Precambrian and Cambrian algae as it displays a more complex morphology in the form of a central axis interrupted by large branches,

chapter 4



Figure 4.64 Reconstruction of Longfengshania cordata. (From

Xu, 2002.) Figure 4.65 Reconstruction of Plantulaformis sinensis. (From

Xu, 2002.)

Figure 4.66 Suggested reconstruction of Waputikia ramosa. (From Briggs et al., 1994.)

which in turn bear a profusion of smaller branches and terminal filaments (Briggs et al., 1994). Several extant red algae possess morphologies that are similar to the Late Ordovician alga Manitobia patula. In this form, the laminar thallus divides in a single plane, with each segment further divided into three, nearly uniform segments. The margin of the thallus is entire, and the apex of each segment typically truncated (Fry, 1983). Three macroscopic Devonian algae occur in marine rocks in New York (Fry and Banks, 1955). Drydenia (FIG. 4.67) consists of elliptical laminae (8.5 cm long) that are basally attached to a narrow stipe terminating in a branching holdfast. In Hungerfordia (FIG. 4.68), also recorded from the Late Devonian of South Africa (Hiller and Gess, 1996), the lamina is highly dichotomous with the distal segments lobed. Specimens of Enfieldia are circular (5.0 cm in diameter), with the outer margin lobed, and characterized by distinct reticulations. Both Drydenia and Hungerfordia have been compared with existing red and brown algae; Enfieldia is more difficult to position systematically, perhaps representing a thalloid liverwort.


paleobotany: the biology and evolution of fossil plants

Figure 4.68 Dichotomous thallus of Hungerfordia dichotoma

(Devonian). Bar  2.5 cm. (From Fry and Banks, 1955.)

Figure 4.67 Drydenia foliata (Devonian). Bar  4 cm. (From

Fry and Banks, 1955.)

The genus Perissothallus (FIG. 4.69) comprises several types of uncalcified macroalgae from Late Pennsylvanian and Early Permian freshwater environments of North America and Europe. The thallus consists of repeatedly dichotomizing, erect cylindrical branches that radiate from a small holdfast (Krings et al., 2007d). Vegetative reproduction occurs in the form of secondary thalli produced on prostrate branches. Striking similarities in basic structure exist between Perissothallus and members of the extant marine red algal genus Scinaia (Nemaliales), but the fossils superficially resemble species of the extant genera Codium (green algae) and Dictyota (brown algae). Another putative late Paleozoic red alga consists of a central axis with nodes of lateral appendages (FIG. 4.70). Bassonia hakelensis is a Cenomanian (Late Cretaceous) marine fossil from the Haqel fish beds in Lebanon that

Figure 4.69 Perissothallus showing branches radiating from holdfast (Pennsylvanian). Bar  1 cm.

appears as a compressed, irregularly branched thallus, 20.0 cm long (Basson, 1972; Krings and Mayr, 2004). The thallus is more or less monopodially organized and attached to the substrate by a circular holdfast (FIG. 4.71). At irregular intervals, the main long shoot produces second-order long shoots. Numerous determinate short shoots bearing spine-like outgrowths (?trichoblasts) extend from the long shoots and give the whole thallus a spiny appearance. The basic morphology of the spine-bearing short shoots of B. hakelensis closely resembles the extant red alga Pithyopsis tasmanica (Ceramiales) that occurs along the coasts of southern

chapter 4

Figure 4.70 Grateloupia sp. (Permian). Bar  2 cm.

Australia and Tasmania (Krings and Mayr, 2004). Another red alga from the Haqel fish beds is Delesserites lebanensis (Basson, 1981). This compressed, Delesseria-like thallus is 9.5 cm long and includes several blades that radiate from a common stipe (FIG. 4.72). Nothing is known about the holdfast or reproductive organs of this alga. The early Oligocene Haeringiella multifidiformis (FIG. 4.73) is characterized by a bladelike thallus attached to the substrate by a disk-like holdfast (Krings and Butzmann, 2005). The thallus is formed of a central axis that produces irregularly shaped lateral branches. Branches in the proximal portion of the thallus are closely spaced or clustered



Figure 4.71 Bassonia hakelensis, young thallus with circular

basal holdfast (Cretaceous). Bar  3 cm. (Courtesy BSPG.)

and relatively undifferentiated, whereas those in the distal portion are irregularly forked and give off secondorder branches. Branch tips may be fringed. The most similar modern red alga is perhaps the gametophytic thallus of Sphaerococcus coronopifolius (Gigartinales). Various morphotypes believed to represent uncalcified red algae have been reported from middle and upper Miocene rocks from the Monterey Formation in California (Parker and Dawson, 1965). In general, they consist of planated axes with equally spaced laterals.


paleobotany: the biology and evolution of fossil plants


Figure 4.72 Thallus of Delesserites lebanensis (Cretaceous). Bar  2.5 cm. (From Basson 1981.)

At one time, spiny vesicular microfossils were often termed hystrichosphaerids (a designation that alludes to their morphology) regardless of their biological affinities. In 1961, Evitt conclusively demonstrated that many hystrichosphaerids were the cysts of dinoflagellates. Consequently, the remaining “hystrichosphaerids” whose affinities remain uncertain or unknown are placed in the artificial group Acritarcha (Evitt, 1963 b,c). As a result the acritarchs represent a highly heterogeneous group of organic-walled vesicular microfossils (FIGS. 4.74–4.75) interpreted as (cysts of) protists of different biological affinities (Mendelson, 1993; Colbath and Grenfell, 1995; Strother, 1996; Montenari and Leppig, 2003), including the cysts of some naked dinoflagellates (Tappan, 1980). Other acritarchs, however, have been interpreted as multicellular (green) algae (Butterfield, 2004; Stanevich et al., 2007), algal (zygnematacean) spores

Figure 4.73 Haeringiella multifidiformis. Bar 1 cm. (From Krings and Butzmann, 2005.)

Figure 4.74 Dicrodiacrodium sp. (Ordovician). Bar  10 μm. (Courtesy M. Vecoli and T. Servais.)

chapter 4

Figure 4.75 Evittia sp. (Silurian). Bar  10 μm. (Courtesy M. Vecoli and T. Servais.)

(Grenfell, 1995), and fungi-like organisms (Butterfield, 2005); other possible origins outside of the phytoplankton have been reviewed by Colbath and Grenfell (1995). Microfossils assignable to the Acritarcha are among the first eukaryotes preserved in the fossil record (Huntley et al., 2006); they first occur in the Paleoproterozoic (Vidal and Moczydłowska-Vidal, 1997; Huntley et al., 2006) and extend to the Holocene (Mendelson, 1987). Acritarchs dominate the microfossil record in Proterozoic and Cambrian rocks. Some of the earliest accounts come from the Paleoproterozoic of China (Z. Zhang, 1997; S. Sun and Zhu, 2000). Other early acritarchs include the forms described from the Mesoproterozoic Roper (Javaux et al., 2001) and Bangemall (Buick and Knoll, 1999) Groups in Australia, and the Billyakh Group in northeastern Siberia (Sergeev et al., 1995). Paleo- and Mesoproterozoic acritarch assemblages are characterized by rather simple forms, whereas more complex forms typify Neoproterozoic assemblages (Sergeev et al., 1995). Slightly younger Proterozoic acritarchs have been described from shales of the Meso-Neoproterozoic Ruyang Group in China (L. Yin, 1998) and the Neoproterozoic Doushantuo Formation in South China (C. Zhou et al., 2001; Xiao, 2004). Acritarchs are variable in size (on average 5–200 μm) and shape (FIG. 4.78); the vesicle (body) of an acritarch ranges from oval to triangular in outline and may possess various forms of projections. They are classified in an artificial system of morphotaxa based on a complement of morphological characters, like those of spores (FIG. 4.76), including


Figure 4.76 Emphanisporites annulatus Bar  10 μm. (Courtesy M. Vecoli and T. Servais.)



Figure 4.77 Stelliferidium sp. (Ordovician). Bar  10 μm.

(Courtesy M. Vecoli and T. Servais.)

size and shape of the vesicle, number and form of projections, number of symmetry levels, and form of the exit rupture (Williams et al., 2000; Montenari and Leppig, 2003). Regardless of their natural affinities, the widespread occurrence and complex morphology of acritarchs, as well as their rapid rate of evolution, have made them extremely valuable in long-range correlation and biostratigraphic zonation


paleobotany: the biology and evolution of fossil plants

Type genera

Possible biological relationships

Acanthomorphitae (akantha  spine)

Microhystridium Multiplicisphaeridium Vulcanisphaera


Polygonomorphitae (Poly  gonia  polygonous)

Diexallophasis Goniosphaeridium Veryhachium


Sphaeromorphitae (sphaira  sphere)

Leiosphaeridia Synsphaeridium Trachysphaeridium

Netromorphitae (netron  spindle)

Deunffia Domasia Leiofusa

Herkomorphitae (herkos  fence)

Cymatiogalea Cymatiosphaera Dictyotidium

Diacromorphitae (di  akron  two tops)

Acanthodiacrodium Dasydiacrodium

Pteromorphitae (pteros  wing)

Duvernaysphaera Fimbriaglomerella Pterospermella


Prismatomorphitae (primsa  prismatic)

Marrocanium Octoedryxium Polyedryxium


Oömorpitae (oön  egg-shaped)

Aranidium Nothooidium Ooidium








Figure 4.78 Basic acritarch morphology. (Modified from Montenari and Leppig, 2003.)

(Smelror, 1987; Montenari and Leppig, 2003) and as indicators of climatic change (Mohr, 1990). In addition, acritarchs have been useful as paleoecological indicators of specific marine environments. For example, Staplin (1961) used several acritarch associations to predict distances from reefs in the Upper Devonian of Canada.

5 Hornworts and Bryophytes Early Fossil Evidence ........................................................163

Marchantiophytina (Liverworts or Hepatophytes) .......................... 167

Anthocerotophyta (Hornworts) .......................... 165

Bryophytina (Mosses) ..................................................................... 174

Bryophyta (Bryophytes) ...................................................166

These tiny bryophytes reveal their beauty slowly and up close, as do good friends. John Caddy, Morning Earth Poems

Prior to the development of the first efficient microscopes in the late eighteenth century, most people, including scientists, regarded hornworts, liverworts, and mosses as tiny flowering plants. The distinctness of these organisms first became widely acknowledged in the decades following the documentation of the bryophyte life cycle (FIG. 5.1) by Wilhelm Hofmeister (FIG. 5.2) in 1851. Beginning in the twentieth century, scientists became increasingly interested in the origin and evolution (including the fossil record) of the hornworts and bryophytes, and their relationships to other groups of fossil and modern plants. Bryophytes and hornworts are unique among extant embryophytes in that they have a gametophyte-dominant life cycle; the sporophyte usually is relatively short lived and permanently dependent upon the gametophyte. Physiologically, they are poikilohydric, meaning that they cannot control water loss; when the environment dries out, bryophytes also desiccate. Some are reported with endomycorrhizae (e.g., Y. Zhang and Guo, 2007). When moisture is available, they rehydrate; in other words, desiccation tolerance is physiological and not structural, as it is in most vascular plants. The plant body in these groups is a thallus—a simple plant body which is not differentiated into stems, roots, and leaves. Although they have no vascular tissue, some bryophytes have conducting cells in the form of hydroids and leptoids.

Hornworts and liverworts have been interpreted as occupying a position intermediate between the green algae and vascular plants (Smith, 1938). Others suggested that the bryophytes in general represented examples of evolutionary failures; perhaps they originated from early vascular plants, such as the rhyniophytes (Chapter 8), and then, as the group continued to evolve, vascular tissue was lost. Today, however, these views have changed on the basis of a variety of ultrastructural, biochemical, and molecular data (Duckett and Renzaglia, 1988) suggesting that the principal bryophyte groups had separate origins, and that hornworts, liverworts, and mosses represent the earliest divergent lineages of extant land plants, although the specific order of their divergence still remains unresolved (Friedman et al., 2004; Shaw and Renzaglia, 2004). Phylogenetic analysis strongly supports the liverworts as sister to all other land plants, and provides moderate to strong support for hornworts as the sister group to the vascular plants (Nickrent et al., 2000; Qui et al., 2006). The earliest recognizable bryophytes in the macrofossil record include liverworts that appear to have their closest affinities with the Metzgeriales (Krassilov and Schuster, 1984; Frahm, 2001a). True mosses are first encountered in the Mississippian, and appear to be well established by Permian time, including some modern orders (Neuburg, 1960a (FIG. 5.3); Jovet-Ast, 1967). Mosses easily referable to



Paleobotany: the biology and evolution of fossil plants

Sporangium with spore mother cells Sporophyte

Operculum Peristome



Calyptra Spore mother cells undergo meiosis, producing


2n Meiosis

Fertilization n

Spores Female gametophyte




Archegonium Egg


Male gametophyte “Leaf” Rhizoids

Figure 5.1 Life history of typical moss. (From Taylor and Taylor, 1993.)

modern genera do not appear until the Cretaceous. Hornworts are not encountered until the Cretaceous, and their occurrence is based almost exclusively on dispersed spores (N. Miller, 1980). The majority of Cenozoic bryophytes are assignable to modern genera, which may indicate that most of the modern families arose during the Cretaceous. Frey (1990) suggested that most of the leafy liverwort taxa as well as some mosses may have evolved in southern Gondwana and migrated north into tropical and Laurasian regions during the Cretaceous and Cenozoic.

Higher taxa in this chapter (based on the classification system of Frahm, 2001a):

Phylum Anthocerotophyta (hornworts); Dendrocerotaceae Phylum Bryophyta (bryophytes) Marchantiophytina (liverworts or hepatophytes) Treubiopsida Marchantiopsida (thalloid liverworts) Marchantiales (Continued)


Sphaerocarpales Calobryales Ricciales Marchantiaceae, Ricciaceae Jungermanniopsida (leafy liverworts) Jungermanniales Lophoziaceae, Scapaniaceae, Frullaniaceae, Porellaceae, Lejeuneaceae Bryophytina (Mosses sensu lato) Sphagnopsida (Sphagnum mosses) *Protosphagnales Takakiopsida Bryopsida (Mosses sensu stricto; true mosses) Bryales Dicranales Dicranaceae Pottiales Leucodontales Hypnales Polytricales Polytricaceae, Dawsoniaceae

Hornworts and Bryophytes

Figure 5.2 Wilhelm Hofmeister.

Early Fossil Evidence The origin and early evolution of the bryophytes is a complex and perplexing problem (Goffinet, 2000). Phylogenetic evidence suggests that bryophytes in general, and liverwort-like plants in particular, should have been important components of early terrestrial floras (Bateman et al., 1998; Renzaglia et al., 2007). The fossil record of bryophytes, however, especially of early bryophytes, is meager and those known from fossils appear comparable in many ways to extant taxa. Even the earliest bryophyte-like fossils have the basic thallus organization also seen in many living forms. Based on this evidence, it is possible that the bryophytes evolved far earlier than the fossil record suggests, and fossils of the earliest bryophytes have not been found to date. However, it is possible that paleobotanists simply may not recognize the earliest bryophytes, because morphologically they do not resemble modern forms. One interesting hypothesis suggests that several of the enigmatic Cambrian to Devonian fossils traditionally included in the nematophytes (see Chapter 6) may represent remains of ancient liverworts which shared certain features with modern marchantioids (Graham et al., 2004). Some of these fossils consist of sheets of (pseudo-)cells (so-called cuticle) and tubes of some resistant material. Cosmochlaina (Silurian–Devonian) was originally considered to be of

Figure 5.3 Maria Neuburg. (Courtesy H. N. Andrews.)



Paleobotany: the biology and evolution of fossil plants

Figure 5.4 Cuticle of Cosmochlaina verrucosa (Devonian).

Bar  20 μm. (From Edwards, 1986.)

uncertain affinity, but has been interpreted as the lower epidermal surface of a marchantioid liverwort (FIG. 5.4). At least some of the tubular aggregations assigned to nematophytes (and possibly to Nematothallus) have been reinterpreted as masses of resistant liverwort rhizoids, and are sometimes attached to fragments of lower epidermal tissue. It has also been suggested that some of the dispersed Early Silurian cell sheets and tubes may represent fragmentary remains of bryophyte sporangia (Graham and Gray, 2001). One of the oldest reports of bryophyte-like megafossils is Parafunaria sinensis, a compression from the Early–Middle Cambrian Kaili Formation of China (Yang et al., 2004). The specimen is up to 2 cm long and comprises 4–5 leaves (each 5–15 mm long by 5 mm wide) that are densely borne around a short stem. The basal portion of the stem consists of a footlike structure, whereas distally it bears what is interpreted as a short seta with capsule. The authors suggest that the arrangement of leaves is similar to that in the extant moss Funaria hygrometrica (Bryopsida). Due to the age of these fossils, additional supporting evidence will be important in confirming their assignment to the Bryopsida. The study of dispersed spores (sporae dispersae) is an important source of data on the composition of the earliest land floras (e.g., Wellman and Gray, 2000; Steemans and Wellman, 2004). Although in many instances it is difficult to distinguish dispersed bryophyte spores from those of vascular plants, there is increasing support for the suggestion that some of the spores in Ordovician and Silurian rocks resemble those of modern liverworts. The spores most often suggested as bryophytic occur in permanent (sometimes envelope-enclosed) tetrads in these assemblages; it is hypothesized that such tetrads came from plants at a bryophytic, most likely a liverwort, grade of organization (Gray, 1985). Additional support for this hypothesis comes in the form of tiny spore-containing plant fragments from Ordovician rocks of Oman (Wellman et al.,

Figure 5.5 Reconstruction of Sporogonites exuberans. (From Taylor and Taylor, 1993.)

2003). These fossils indicate that the spore producers, although diminutive in size, were true land plants, which produced sporangia containing large numbers of spores—a minimum of 7450 tetrads in some specimens. Ultrastructural features of the spore wall also suggest affinities with the liverworts. In addition to microfossil remains, there are several types of meso- and macrofossils that have been considered important in understanding the early history of bryophytes (Edwards, 2000), including isolated sporangia with in situ spores and axes with conducting elements similar to those in extant bryophytes. The Early Devonian compression fossil Sporogonites (Halle, 1916a, 1936) morphologically resembles a bryophyte. This plant was originally found in Norway and consists of stalks 5 cm long which terminate in elongate capsules (FIG. 5.5). Several longitudinal furrows ornament the base of the sporangium and extend onto the stalk. The sporangium is multilayered, and there is some suggestion that it may contain a central columella-like projection. Inside the sporangia are trilete spores that range up to 30 μm in diameter. Many specimens of S. exuberans are preserved with the sporangial stalks in a more or less parallel orientation (Andrews, 1960), suggesting that they were produced from a common thallus; some stalks appear to be attached at the base to an irregularly shaped, carbonaceous film 15 cm long (FIG. 5.5). The fact that not all stalks bear


Hornworts and Bryophytes


sporangia as well as the subtending axis are characteristically twisted (FIG. 5.6). Sporangial dehiscence is unknown, but suggested to have occurred along a preformed line that split the sporangium into two vertical valves. The morphology and spiraled architecture of the axis and sporangium closely resemble those in the living moss, Takakia ceratophylla (Takakiopsida) (Renzaglia et al., 1997). Morphological correspondence with members of the modern liverwort genus Pellia (Jungermanniopsida), however, have also been noted. Gerrienne (1997) pointed out that Tortilicaulis shares a number of important features with trimerophytes, and suggested that this taxon might therefore be ancestral to the Trimerophytina (Chapter 8). Although the affinities of many of these early fossils will continue to be controversial, fossil organisms that possess bryophytic characters (dominant gametophyte and parasitic, inconspicuous sporophyte) will play an increasingly important role, as paleobotanists continue to decipher steps leading to the colonization of the terrestrial realm.

Anthocerotophyta (Hornworts) Figure 5.6 Tortilicaulis transwalliensis sporangium with twisted stalk (Devonian). (From Taylor and Taylor, 1993.)

sporangia suggests that they were perhaps easily detached or abscised once the spores were mature. Although the compressed nature of the specimens of Sporogonites reveals little about the internal organization of the tissues, it appears that vascular elements are lacking, thus supporting its classification with the bryophytes. It is suggested that Sporogonites may represent a compressed gametophytic thallus bearing upright sporophytes of a primitive moss or perhaps an early hornwort (Poli et al., 2003). Both interpretations are supported by the suggestion of a columella-like projection within the sporangium, because this structure is exclusively known in hornworts and the putative primitive lineages of extant mosses (Goffinet, 2000). Crandall-Stotler (1984) suggested that the basal thallus may represent a persistent protonema, or even a small leafy gametophore, a hypothesis consistent with affinities to the mosses. Tortilicaulis is a Late Silurian–Early Devonian fossil that shares morphological features with Sporogonites and certain extant bryophytes (Edwards, 1979, 1996; Edwards et al., 1994). Specimens consist of unbranched or isotomously branched axis fragments that terminate in solitary or branched, elongate sporangia (FIG. 5.6) containing trilete spores. The

The hornworts include 300 living species (e.g., Duff et al., 2007) and differ from mosses and liverworts in the nature of their spore dispersal, which is accomplished by a longitudinal splitting of the capsule into several valves along with the action of pseudoelaters. The capsules (sporangia) characteristically have a central column of sterile tissue, called a columella. Hornworts produce symmetrical spermatozoids, and on the ventral surface of the gametophytic thallus are specialized, apically derived mucilage clefts surrounded by two cells resembling guard cells. Some authors have considered these homologous to the stomata which occur on the sporophytes (Renzaglia and Vaughn, 2000). In most genera, cells typically have a single, cup-shaped chloroplast (with pyrenoids) that is reminiscent of the chloroplasts seen in some green algae (Frahm, 2001a). There are no reliable reports of fossil hornworts prior to the Cretaceous, although they have been variously implicated as the oldest extant lineage of land plants (e.g., Renzaglia and Vaughn, 2000). The earliest macrofossil that bears some resemblance to a modern hornwort is perhaps Dendroceros victoriensis from the Lower Cretaceous Koonwarra Fossil Bed in Australia (Drinnan and Chambers, 1986). Morphologically it appears to be a sporophyte arising from a thalloid gametophyte. Another fossil that has been assigned to the hornworts is Notothylacites filiformis from the


Paleobotany: the biology and evolution of fossil plants

uppermost Cretaceous of Bohemia (Nˇemejc and Pacltova, 1974). Frahm (2005) noted, however, that this fossil displays a midrib, which does not occur in modern Notothylas, and thus may represent a Riccia-like thallus instead. Additional evidence of Cretaceous hornworts occurs in the form of spores. Three types of spores have been described from Maastrichtian (latest Cretaceous) rocks and compared to the extant genus Anthoceros (Phaeoceros) (Jarzen, 1979). All are trilete, circular, and possess a distinct cingulum; ornamentation is variable among all the three types. Other spores referred to the hornworts come from the Neogene of Hungary (Nagy, 1968) and Cenozoic from elsewhere in Europe and other sites (e.g., Lacey, 1967); these have been assigned to Phaeocerosporites, Rudolphisporis, and Saxosporis, morphogenera defined to accommodate spores that appear similar to those of extant hornworts. Several macrofossils assignable to the Anthocerotophyta have been recorded for the Cenozoic. The most complete representative (Frahm, 2005) is preserved in Dominican amber of early–middle Miocene age (Iturralde-Vinent and MacPhee, 1996). The specimen consists of a thallus segment (3 mm  3.5 mm long) bearing several cylindrical, hornlike sporophytes, which are similar to those in the Dendrocerotaceae (e.g., Dendroceros or Megaceros). Another fossil with possible affinities to the hornworts is Shuklanites deccanii from the uppermost Cretaceous Deccan Intertrappean beds exposed at Mohgaon Kalan village, Madhya Pradesh, India (Singhai, 1973). This specimen represents an isolated, 1.5 mm long, pear-shaped sporogonium (the sporophyte generation in bryophytes and hornworts), which consists of a short, bulbous foot and an elongate capsule containing thin-walled, trilete spores and abundant filamentous structures interpreted as pseudoelaters. A central columella, however, which is characteristic of most extant hornwort capsules, is not present. Two isolated sporogonia reminiscent of those in the extant hornwort Notothylas were also described from the same locality (Gupta, 1956; Chitaley and Yawale, 1980). Another interesting bryophyte from this site is Krempogonium mohgaoensis, an enigmatic fossil (5.5 mm long) composed of a twisted stalk and an oval capsule (FIG. 5.7), containing 10 vertically oriented spore sacs that are separated from each another by parenchymatous septa (Nambudiri et al., 2003). The spore sacs contain numerous small (25–30 μm in diameter) spores and pseudoelaters. The distal half of the capsule is covered by a smooth calyptra-like structure. Krempogonium mohgaoensis is interpreted as a bryophytic sporophyte that contains a mosaic of features seen in extant hornwort, liverwort, and moss sporophytes.

Figure 5.7 Reconstruction of Krempogonium mohgaoensis sporangium (Cretaceous). (From Nambudiri et al., 2003.)

Bryophyta (Bryophytes) Living bryophytes are represented by 900 genera and nearly 24,000 species. They are not a conspicuous portion of the Earth’s flora, although they may dominate the vegetation in certain special environments, for example Sphagnum in certain types of bogs. Most bryophytes are small plants, many 2 cm long. The largest forms rarely exceed 60 cm in length (e.g., species in the genus Dawsonia). In general, bryophytes are most abundant in relatively moist areas. They range throughout the world and can even be found in coastal areas of Antarctica. Bryophytes differ from true vascular plants by the absence of vascular tissue and by the presence of a nutritionally independent gametophyte generation in their life cycle (FIG. 5.1). Although bryophytes do not contain true vascular tissue, some have specialized conducting elements (Hébant, 1977) (FIG. 5.8). In certain mosses, the stem of the gametophyte and the seta (stalk) of the sporophyte contain elongate, non-lignified water-conducting cells called hydroids (analogous to xylem in vascular plants). Surrounding the hydroids are assimilate-conducting cells


Hornworts and Bryophytes


Figure 5.9 Tetrapterites visensis (Mississippian). (From Taylor and Taylor, 1993.)

Figure 5.8 Charles Hébant. (Courtesy J. Galtier.)

termed leptoids, which are comparable to sieve elements in the phloem of higher plants. The fossil record of the Bryophyta is based on both spores and macrofossil remains. Numerous spores have been described as bryophytic, but most of these are sporae dispersae, which limits their taxonomic usefulness and may obscure their biological affinities. One of these is Tetrapterites visensis, an unusual dispersed structure isolated from Mississippian rocks (Hibbert, 1967). It consists of a tetrahedral, non-cellular membrane with winglike ridges; each ridge is attached to a single, trilete spore (FIG. 5.9). Sullivan and Hibbert (1964) compared Tetrapterites to the persistent tetrads of some liverworts, including Sphaerocarpus. The Paleozoic macrofossil record of bryophytes is poor. The earliest accepted specimens come from the Carboniferous (e.g., Walton, 1925, 1928; Oostendorp, 1987), and a few examples have been found in rocks as old as the Devonian (discussed below). Just why there is no extensive record of bryophytes in the Carboniferous is debatable, since the coal-swamp forests of the Pennsylvanian should have provided a wealth of suitable habitats for bryophytes. In addition, the preservation of other fossils from these paleoenvironments is often excellent (e.g., in coal balls). Some believe that the preservational potential of bryophytes is so poor that they are simply not preserved in sufficient numbers to be accurately recorded. Experiments conducted

by Hemsley (2001), however, show that the preservational potential of bryophytic plant material is similar to that of vascular plants, which suggests that bryophytes were indeed rare elements of the Carboniferous coal-swamp ecosystems. Until approximately 25 years ago, the fossil record of the Bryophyta consisted principally of vegetative remains of the gametophyte; with few exceptions, the sporophyte generation remained unknown, and almost no fossil bryophytes were known with identifiable sex organs. This situation has changed, however, because a number of well-preserved fossils of moss and liverwort sporophytes have been described from Mesozoic rocks (e.g., Konopka et al., 1997, 1998) and Cenozoic amber (e.g., Grolle, 1998; Frahm, 1999b, 2001b; Grolle and Schmidt, 2001) (Table 5.1; FIGS. 5.10–5.16). In several of the liverworts in amber, even the androecium, perianth, and gynoecium are preserved (Grolle, 1990, 1998). Although bryophyte remains in amber were noted as early as the nineteenth and early twentieth centuries (e.g., Göppert, 1853; Caspary, 1887; Dixon, 1922), only recently have these fossils received wider scholarly attention. Today, amber fossils represent the single most important source of evidence for the evolutionary history and biodiversity of bryophytes in the Cenozoic. Marchantiophytina (liverworts or hepatophytes)

Recent divergence-time estimates of the origin of the liverworts obtained using penalized likelihood suggest a Late Ordovician divergence of the liverworts (Heinrichs et al., 2007), based on a maximum age from Wellman et al. (2003) for the oldest fossils generally accepted as land plants, and Kenrick and Crane (1997a) for the oldest split of vascular plants. The earliest liverwort in the fossil record is


Paleobotany: the biology and evolution of fossil plants

Table 5.1 Bryophyte genera known from amber.





Baltic (Eocene) and Bitterfeld (Eocene or Oligocene–lowermost Miocene)

Jungermanniales (leafy liverworts)

Bazzania, Calypogeia, Cheilolejeunea, Cylindrocolea, Frullania (FIG. 5.10), Jungermannia, Lophozia, Mastigolejeunea, Metacalypogeia, Nipponolejeunea, Notoscyphus, Plagiochila, Porella, Ptilidium (FIG. 5.11), Radula, Scapania, Spruceanthus

Surveyed in Grolle and Meister (2004a)

Dominican Republic (early–middle Miocene)


Archilejeunea, Bazzania, Blepharolejeunea, Bryopteris, Caratolejeunea, Cyclolejeunea, Cyrtolejeunea, Drepanolejeunea, Leucolejeunea, Lopholejeunea, Marchesinia, Mastigolejeunea (FIG. 5.12), Neurolejeunea, Radula, Stictolejeunea

Grolle (1984a, 1987, 1990, 1993), Gradstein (1993)

Baltic, Bitterfeld

Mosses (Bryophytina)

Aptychella, Atrichum, Barbella, Barbula, Bartramia, Bescherellea, Boulaya, Brachythecium, Brotherella, Calomnion, Campylium, Campylopodiella, Campylopus, Ctenidium, Dichodontium, Dicranum, Dicranites, Eurohypnum, Fabronia, Grimmia, Haplocladium, Hymenostomum, Hypnodontopsis, Hypnum, Mastopoma, Merilliobryum, Muscites, Phascum, Polytrichum, Rhizogonium, Rhytidiadelphus, Sematophyllites, Symphyodon, Trachycystis, Trichostomum, Tristichella

Frahm (1996a, b, 1999a, b, 2000, 2001b, 2004a, b, 2006b)

Dominican Republic

Mosses (Bryophytina)

Acroporiites, Adelothecium, Calymperes (FIG. 5.13), Calyptothecium (FIG. 5.14), Caribaeohypnum, Clastobryum, Entodon, Homalia, Hypnum, Leucobryum, Mittenothamnium, Mniomallia, Octoblepharum (FIG. 5.15), Orthostichella, Orthostichopsis, Plagiomnium, Porotrichum, Syrrhopodon (FIG. 5.16), Thuidium

Frahm and Newton (2005), Frahm (2006a)

Dominican Republic

Anthocerotophyta (hornworts)

Dendroceros or Megaceros

Frahm (2005)

Citations for ages: Dominican (Iturralde-Vinent and MacPhee, 1996); Baltic (Weitschat and Wichard, 1998; Knuth et al., 2002, see discussion in Schmidt and Dörfelt, 2007); Bitterfeld (see discussion in Dunlop and Giribet, 2003 and references cited therein).

Figure 5.10 Frullania schumannii (Eocene). Bar  4 mm. (Courtesy J.-P. Frahm.)

Figure 5.11 Ptilidium sp. (From Grolle and Meister, 2004b.)


Hornworts and Bryophytes


Figure 5.14 Calyptothecium duplicatum (Miocene). Bar  6 mm. (Courtesy J.-P. Frahm.)

Figure 5.12 Mastigolejeunea bidentula (Eocene). Bar  1 mm. (Courtesy S. R. Gradstein.)

Figure 5.13 Calymperes palisoltii (Miocene). Bar  5 mm. (Courtesy J.-P. Frahm.)

Figure 5.15 Octoblepharum cylindricum (Miocene). Bar  3 mm. (Courtesy J.-P. Frahm.)


Paleobotany: the biology and evolution of fossil plants

Figure 5.16 Syrrhopodon incompletus (Miocene). Bar  3 mm. (Courtesy J.-P. Frahm.)

Metzgeriothallus sharonae from Givetian (upper Middle Devonian) shales and siltstones from New York (Van Aller Hernick et al., 2008). The fossils are preserved as carbonaceous films, and display dorsiventral thalli up to 32 mm long and 1.5 mm wide, that consist of a median costa and entire-margined wings. What appear to be unicellular, ribbon-like rhizoids extend from beneath the costa. Associated with the gametophytic thalli is an elongate sporophyte capsule with four valves. Another slightly younger liverwort is Pallaviciniites (Hepaticites) devonicus from the Upper Devonian (Frasnian) of New York (Hueber, 1961). The specimens consist of compressions preserved in a fine-grained shale together with numerous other plant remains. The liverworts were removed by bulk maceration (see Chapter 1) of the shale in concentrated hydrofluoric acid (HF); the carbonaceous, thalloid specimens were then floated onto microscope slides for examination. Pallaviciniites

devonicus is a simple, two-parted, flattened thallus with a central midrib and marginal lamellae or wings. The thallus is dichotomously branched, and along the margin of the wings are closely spaced teeth. The rhizomatous portion of the plant shows outlines of elongate parenchymatous cells, some bearing non-septate rhizoids. No reproductive structures are known. Various species of Pallaviciniites have been described from the Carboniferous to the Pleistocene, and they have been compared with such living genera as Pallavicinia, Metzgeria, Treubia, and Fossombronia (Schuster, 1966). Other late Paleozoic liverwort thalli come from the Carboniferous and have been assigned to morphogenera such as Blasiites, Metzgeriothallus, and Treubiites (He-Nygrén et al., 2006). Treubiites kidstonii from Scotland was initially believed to be similar to the extant Treubia, but later was shown to closely resemble extant Blasia because of its ventral scales (Krassilov and Schuster, 1984). Naiadita is a Triassic liverwort that was preserved in large numbers and many different stages of development. As a result, a great deal of information is known about the total biology of this bryophyte. The most comprehensive and detailed treatment of Naiadita is that of Harris (1938), based on specimens collected in Worcestershire and Warwickshire, England (Late Triassic). The plant is small (rarely exceeding 3 cm) and consists of an unbranched stem with helically arranged, lanceolate leaves (FIG. 5.17). Individual leaves are rounded at the apex (FIG. 5.18) and generally 1–5 mm long. Near the base of the stem are numerous, unbranched, non-septate rhizoids. Located along the stem are gemmae cups, which are specialized vegetative reproductive structures. They produced oval (500 μm in diameter) gemmae, which represent one of the most common components of the Naiadita fossiliferous beds. Gemmae are small pieces of thallus tissue, which can grow into a new plant. Some specimens possess stem-borne archegonia, which are 300 μm long and surrounded by a “perianth” of leaflike lobes. Although antheridia are not known, numerous stages in the development of the embryo and sporophyte are preserved. The fossil sporophyte of N. lanceolata consists of a short foot, slender stalk, and bulbous sporangium. The capsules are about 1.2 mm in diameter and contain spores in tetrahedral tetrads. The spores (100 μm in diameter) are lens shaped with an equatorial flange. On the proximal surface are numerous small, pointed spines; the distal surface bears larger, irregular projections. Dispersed spores with the same complement of morphologic features from rocks of equivalent age, referred to the genus Naiaditaspora, were examined at the fine-structural level (Hemsley, 1989a). The exine is organized into five distinct zones in which the inner


Figure 5.17 Reconstruction of Naiadita lanceolata (Triassic). (From Taylor and Taylor, 1993.)

Figure 5.18 Leaf of Naiadita lanceolata (Triassic). (From Taylor and Taylor, 1993.)

regions contain numerous lamellae; the outermost region is granular. Based on a comparison with extant liverwort spores, the spores of Naiadita are most similar to members of the Marchantiales and Sphaerocarpales. The type of

Hornworts and Bryophytes


spores, presence of unicellular rhizoids, and organization of archegonia and capsules suggested to Harris (1938) that Naiadita represented a liverwort similar to extant Riella (Sphaerocarpales). Naiadita also shares some vegetative features with certain modern liverworts included in the Calobryales (Schuster, 1966). Late Paleozoic, Mesozoic, and Cenozoic impression and compression fossils of liverworts or liverwort-like thalli have been assigned to various morphogenera. These include: Thallites, for thalloid fossils that may represent liverworts or algae (Chapter 4), Hepaticites, for thalli that can confidently be assigned to the liverworts, but cannot be classified further, and Jungermannites, Metzgeriites, or Marchantites, for thalli that can be classified to the ordinal level within the hepatophytes (Cantrill, 1997b). Liverwort thalli in general are relatively rare as fossils, but there are several reports from the Mesozoic, in which bedding planes, which sometimes extend for several square meters, are covered with densely spaced thalli (e.g., Banerji, 1989; Pole and Raine, 1994). In many of the liverwort-rich beds, the thalli are preserved in situ, and therefore are interpreted as colonization horizons of freshly deposited sediment (Cantrill, 1997b). Beautifully preserved, compression fossils assignable to the Marchantiopsida occur in dense mats on bedding planes in the Aptian (Lower Cretaceous) of Spain (Diéguez et al., 2007b). These fossils consist of small, rosette-forming dichotomously branched thalli (FIG. 5.19). Marchantites cyatheoides and M. tennantii are impression fossils of thalloid liverworts referred to the Marchantiales from the Upper Triassic Molteno Formation in South Africa (Anderson, 1976). Marchantites tennantii has dichotomizing thalli (FIG. 5.20), in which individual branches range from 2.5 to 4 mm wide, each with a prominent midrib 1 mm wide. The lateral regions of branches have a surface pattern of polygonal areas between 0.75 and 1.5 mm wide that are arranged in rows arching away from the midrib. Similar patterns of regularly arranged polygonal fields are common in modern Marchantiaceae, where they represent the surface expressions of the subsurface air chambers. Four species of Marchantites were described from the Lower Cretaceous of Alexander Island, Antarctic Peninsula (Cantrill, 1997b), where they functioned as colonizers of fresh sediment near rivers and as an important part of the understory in both fern thickets and conifer forests. The taxa are distinguished based on thallus form, size, and the presence of features such as arcuate ribbing and air pores. Thalli of M. pinnatus are pinnate with short lateral branches ( 10 mm long), which have a prominent midrib and numerous rhizoids arising from the midrib region on the ventral


Paleobotany: the biology and evolution of fossil plants

Figure 5.19 Rosette-like dichotomizing thalli of marchantioid liverwort (Cretaceous). Bar  5 mm. (Courtesy C. Diéguez.)

Figure 5.20 Suggested reconstruction tennantii. (From Anderson, 1976.)



side. Marchantites rosulatus is a rosette-like thallus with individual thallus branches 3–5 mm wide. Thalli of M. taenioides are ribbon-like, up to 60 mm long, and only sparsely branched. On the dorsal surface are numerous circularelliptical pores; on the ventral side rhizoids are borne along the midrib. The fourth species, M. arcuatus, has thalli that display open branching. The affinities of these taxa within the Marchantiales are based on thallus morphology and the presence of structures resembling air pores on the thalli. Other liverwort thalli from Alexander Island were assigned

to the morphogenera Hepaticites and Thallites. Cantrill (1997b) noted that the liverworts of these late Albian floras were both diverse and abundant, and appeared to occupy a number of different ecological niches in this high latitude site. Based on an analysis of the δ13C of the Cretaceous liverwort thalli from Alexander Island, and subsequent comparisons of the results with modern analogs, Fletcher et al. (2005) showed that fossilized bryophytes can be used to gather information on paleoatmospheric CO2 concentrations, and thus offer new methods and insights into paleoclimatic reconstructions. Marchantiolites is an Early Cretaceous liverwort from central Montana (Brown and Robison, 1976). One specimen is 4 cm long and contains a prominent midrib. On the ventral surface are numerous rhizoids, whereas the dorsal surface displays air pores surrounded by specialized subsidiary cells. A slightly different air-pore frequency and morphology is present in the Rhaeto–Liassic (Late Triassic–Early Jurassic) species M. porosus from Scania, Sweden (Lundblad, 1954). As the name suggests, Marchantiolites has been included in the Marchantiales based principally on the general organization of the air pores. Lundblad (1954), however, noted that living species of Marchantia exhibit compound pores, whereas the fossil forms contain mostly simple ones. In general, the record of thalloid fossils interpreted as members of the Ricciaceae ranges from the Pennsylvanian (Walton, 1949a) to the Quaternary (Jovet-Ast, 1967); however, the affinities of most pre-Jurassic forms are still uncertain (Oostendorp, 1987). Fossil Ricciaceae are usually placed in the morphogenus Ricciopsis, but some have also been assigned to the modern genus Riccia (e.g., Sheikh and Kapgate, 1982). Ricciopsis florinii (Ricciaceae) is a rosette-shaped thallus from the Late Triassic–Early Jurassic of Sweden (Lundblad, 1954); it has four main branches, each of which dichotomizes twice. Although most records of fossilized ricciacean thalli come from Europe and Asia, to date only a single species has been described from North America, Ricciopsis speirsae from the Paleocene of Alberta, Canada (Hoffman and Stockey, 1997). This form differs from all other living and fossil Ricciaceae by displaying occasional constrictions and dilations of the repeatedly dichotomizing thallus. These fossils occur in what is interpreted as an oxbow lake deposit, along with lemnaceous angiosperms. Extensive occurrences of fossil Ricciaceae (e.g., Ricciopsis algoaensis) are known from the Lower Cretaceous (Berriasian–Valanginian) of South Africa (Anderson and Anderson, 1985), and dispersed ricciacean spores of Paleocene age have been reported as rare


elements in the Sonda coal deposits in Pakistan (Leghari et al., 2001). Another interesting fossil liverwort is Diettertia, a Cretaceous form from Montana initially described as a moss (Brown and Robison, 1974). Based on additional material, Diettertia is now regarded as a bilaterally symmetrical, leafy liverwort with affinities to the Jungermanniales (Schuster and Janssens, 1989). The gametophyte consists of unistratose, bifid leaves inserted in two ranks on stems approximately 0.5 mm in diameter. Rhizoids are long and slender, non-septate, and up to 25 μm in diameter. Many features of Diettertia suggest that it represents a highly specialized member of the Jungermanniales. As a result of this fossil and other evidence, Schuster and Janssens (1989) suggested that the order probably evolved much earlier, perhaps in the late Paleozoic. Amber represents a valuable source of information about the Cenozoic biodiversity of the liverworts. The richest ambers containing liverwort remains come from the Baltic (Eocene), Bitterfeld in Saxony, Germany (Sachsen-Anhalt), dated as Eocene or Oligocene to lowermost Miocene, and the Dominican Republic (early–middle Miocene; IturraldeVinent and MacPhee, 1996). Amber is often very difficult to date. The Bitterfeld amber occurs in lower Miocene sediments (Barthel and Hetzer, 1982), but some authors have suggested that it is equivalent to the Baltic amber, based on the similarity of some faunal elements (see discussion in Dunlop and Giribet, 2003). Several dozen beautifully preserved specimens, representing more than 50 species, have been documented in recent years (e.g., Grolle, 1983, 1984a, b, 1987, 1993; Grolle and Braune, 1988; Gradstein, 1993; Grolle and Heinrichs, 2003; Grolle and Meister, 2004a, b). The liverworts in Baltic and Bitterfeld amber are almost exclusively leafy liverworts (Jungermanniales, Table 5.1) assignable to 17 extant genera (reviewed in Grolle and Meister, 2004b). Included are several specimens that provide insights into the reproductive biology of Cenozoic liverworts. For example, vegetative reproduction in the form of angular gemmae produced in globules at the tips of leaf lobes (FIG. 5.21) has been reported for Lophozia kutscheri (Lophoziaceae) from Bitterfeld amber (Grolle and Meister, 2004a). In a specimen of Scapania hoffeinsiana (Scapaniaceae) also from the Bitterfeld amber, both a cyathiform perianth (surrounded by involucral and subinvolucral leaves) and capsule (split to the base into four narrow valves) on a seta are preserved (Grolle and Schmidt, 2001). Frullania baltica (Frullaniaceae) from Baltic amber shows a capitate androecium that is positioned on a very short branch (Grolle, 1998). Leaves arise in six leaf circles below a clavate, beaked perianth, which is

Hornworts and Bryophytes




Reconstruction of Lophozia kutscheri showing dorsal surface (left) and gemmae at tips of leaves. (Modified from Grolle and Meister, 2004a.)

Figure 5.21

Figure 5.22 Porella subgrandiloba branch; ventral (left) and dorsal view (Eocene). (From Grolle and So, 2004.)

positioned terminally on the main axis and surrounded by involucral and subinvolucral leaves. At the tip of one perianth beak, the globose capsule of a developing sporophyte is visible. Archegonial branches (FIG. 5.22) with dentate bracts and bracteoles of Porella subgrandiloba (Porellaceae) are also known from Baltic amber (Grolle and So, 2004). Antheridial and archegonial reproductive structures and/or “perianth parts” have been recorded for Leucolejeunea antiqua (Grolle, 1990), Drepanolejeunea eogena (Grolle, 1993), and Mastigolejeunea auriculata (Gradstein, 1993) from Dominican amber. Although the former two taxa are only known as fossils, the latter, along with the amber fossils of Marchesinia brachiata and Stictolejeunea squamata (Lejeuneaceae) from the Dominican Republic, represent


Paleobotany: the biology and evolution of fossil plants

fossils of species that still exist today, and thus indicate that these species are archaic and already existed at least by the early Neogene, 20 Ma (Gradstein, 1993). Bryophytina (mosses)

The fossil record of the mosses is less complete than that of the liverworts. Krassilov and Schuster (1984) suggested that perhaps the earliest mosses evolved rapidly into droughttolerant forms that occupied sites where fossilization was unlikely. Nevertheless, there is some record of mosses as early as the Carboniferous (e.g., Walton, 1928). One of these is Muscites plumatus, an impression of a small leafy shoot in rocks of Mississippian age (Thomas, 1972). The axis is covered with helically arranged leaves with clasping bases. Each leaf is about 7.5 mm long and terminates in an elongate tip. Sex organs, sporophyte capsule, or rhizoids were not present. Additional species of Muscites have been described from the Pennsylvanian of France (Renault and Zeiller, 1885, 1888) and the Triassic of Africa (Townrow, 1959; Anderson, 1976). Although there are only a few records of mosses from the Carboniferous, numerous species have been described from Permian rocks, including an extensive moss flora from Siberia (Neuburg, 1960a). From this flora of well-preserved vegetative shoots, Neuburg described six genera that she assigned to the Bryales and three to a new order, the Protosphagnales. Protosphagnum (FIG. 5.23) has leaves similar to those of the extant genus Sphagnum, except for the presence of a midrib. Ignatov (1990) has also described a diverse flora of well-preserved mosses from the Upper Permian of the Russian Platform. The dozen specimens represent remains of the gametophyte generation and include forms that are referable to the extant orders Dicranales, Pottiales, Funariales, Leucodontales, and Hypnales. To date the best documented permineralized late Paleozoic moss comes from the Permian of Antarctica (Smoot and Taylor, 1986a). Merceria augustica consists of delicate axes 1 mm in diameter to which are attached helically arranged leaves and numerous rhizoids (FIG. 5.24). Each leaf is unistratose and contains a thickened midrib (FIG. 5.25). Although no evidence of reproductive organs or sporophytes was found associated with the specimens, the shape of the leaf cells, structure of the leaf margin, and anatomy of the axes suggest affinities within the Bryales. Megafossils of Triassic and Jurassic true mosses are relatively rare and consist almost entirely of compression specimens, although these plants are often represented in palynofloras (e.g., Zavattieri and Volkheimer, 2003).

Figure 5.23 Reconstruction of Protosphagnum nervatum (Permian). (From Taylor and Taylor, 1993.)

Figure 5.24 Cross section of Merceria augustica showing

leaves near apex (Permian). Bar  110 μm.

Sphagnophyllites triassicus consists of isolated Sphagnum-like leaves from the Triassic of India (Pant and Basu, 1978). Another Mesozoic moss is Tricostium, a small Late Jurassic– Early Cretaceous specimen from Siberia, which consists of a stem densely covered with imbricate leaves, each approximately 1.2 mm long (Krassilov, 1973a); on some leaves are elongate pores. In another form from the same area, Yorekiella, the awl-shaped leaves are two or three ranked and lack midribs.


Hornworts and Bryophytes


Figure 5.25 Detail of Merceria augustica leaf (Permian).

Bar  110 μm.

Figure 5.27 Sporophyte of Campylopodium allonense showing a lateral view of the capsule and partially attached calyptra (Cretaceous). Bar  100 μm. (From Konopka et al., 1998; courtesy P. S. Herendeen.)

Figure 5.26 Diagrammatic view of capsule of Eopolytrichum

antiquum. (From Konopka et al., 1997.)

Some of the most exquisitely preserved younger Mesozoic mosses (sporophytes and gametophytes) come from the Late Cretaceous (Santonian) of North America. Gametophytes of Eopolytrichum antiquum (Polytrichaceae) (Konopka et al., 1997) have broadly lamellate leaves densely arranged around the stem. Some specimens have male “inflorescences” in the form of conspicuous rosettes of overlapping perigonal bracts, antheridia, and clavate paraphyses. Female gametophytes have not been found. Sporophytes (FIG. 5.26) consist of oblong capsules, which are terete in cross section and somewhat flattened dorsiventrally. The capsule wall consists of bulging-mammillose exothecial cells with abruptly thinned outer periclinal walls. The operculum is rounded, and an annulus is missing. The peristome consists of a short

Figure 5.28 Lateral view of peristome of Campylopodium allonense showing teeth with divided tips (Cretaceous). Bar  100 μm. (From Konopka et al., 1998; courtesy P. S. Herendeen.)

peristomal membrane, which originates just within the rim of the capsule. Peristomal teeth are apparently lacking, but cells in the circumference of the peristomal membrane suggest that, if this species had developed teeth, there would have been 32 peristome teeth. Stomata are restricted to the apophysis and resemble those seen in extant Polytrichaceae. Campylopodium allonense (Dicranaceae) (Konopka et al., 1998) has oblong, curved, and nodding capsules (FIG. 5.27) that display a distinct basal stomatiferous swelling. The operculum is obliquely rostrate (beaked) with a smooth, cucullate (hood-shaped), or cone-shaped calyptra. Beneath the operculum is a peristome composed of a single cycle of 16 bifid teeth (FIG. 5.28). The spores are spherical, 10–12 μm


Paleobotany: the biology and evolution of fossil plants

Figure 5.29 Hypnum lycopodioides (Oligocene). Bar  1 cm. (Courtesy C. Gee and G. Oleschinski.)

in diameter, and display a delicate rugose sculpture. Associated sterile gametophytes consist of leafy stem portions; fertile gametophytes have not been found. The Cenozoic record of mosses is considerably more extensive than that of the Paleozoic and Mesozoic, and includes impression–compression specimens (FIG. 5.29– 5.31) and amber fossils (N. Miller, 1980, 1984) (Table 5.1). Aulacomnium heterostichoides from deep-water varved clays (Eocene) of a freshwater lake in British Columbia (Janssens et al., 1979) is an example of a well-preserved compression of a moss gametophyte. The plant is irregularly branched (FIG. 5.30) and contains helically arranged, elliptical leaves with multicellular teeth along the upper half of the margin. The cells of the upper laminal surface are regularly isodiametric and unipapillose, whereas those of the basal surface are rectangular with slightly thicker walls. Based on an analysis of characters in existing populations of Aulacomnium, the fossil is believed to be most closely related to A. heterostichum, a living species found in eastern North America and eastern Asia. Numerous Cenozoic mosses are preserved in amber from the Baltic, Bitterfeld (Germany), and the Dominican Republic (e.g., Frahm, 1993, 1994, 1996a, b, c, 1999a, b, 2000, 2001b, 2004a, b, 2006a, b; Frahm and Reese, 1998; Frahm and Newton, 2005) (Table 5.1). The mosses present in the amber can either be assigned to modern species or represent fossil species belonging to modern genera; others cannot be referred to any modern species, and are therefore assigned to morphogenera such as the compression genera Dicranites (FIG. 5.31), Hypnites, and Muscites. At least 30 genera of mosses have been recorded from Baltic and Saxon amber. Several of the moss species preserved in Baltic amber (e.g., Trachycystis flagellaris) still exist today, but their occurrence is restricted to eastern and/ or southeastern Asia. This suggests that these forms became

Figure 5.30 Several leafy branches of Aulacomnium heterosti-

choides (Eocene). Bar  2 mm. (From Taylor and Taylor, 1993.)

Figure 5.31 Dicranites rottensis (Oligocene). Bar  3 mm.

(Courtesy C. Gee and G. Oleschinski.)

extinct in Europe as a result of the Quaternary climatic changes, but persisted in Asia (Frahm, 2001a). Other mosses from Baltic amber (e.g., Haplocladium angustifolium) represent species that today still occur in Europe, but are restricted to the Southern Alps as Cenozoic relicts. The inventory of mosses preserved in Dominican amber includes representatives of almost 20 genera (Frahm and Newton, 2005; Frahm,


2006a). Most of the species are also known today from the Neotropics, which suggests that the foundation of the neotropical moss flora was in existence as early as the Paleogene (45–25 Ma) (Frahm, 2001a). Most mosses in amber occur as sterile gametophytes or gametophyte fragments; fertile gametophytes and sporophytes are comparatively rare. One fossil example of a gametophyte with attached sporophyte is Dicranites grollei from Eocene Baltic amber (Frahm, 1999b). This fossil consists of the distal portion of a stem with 14 linear leaves, each up to 1.8 mm long. The stem apex bears a twisted seta that terminates in a round, distally constricted capsule, which displays 6 (probably originally 16) short, lancet-like, undivided peristome teeth. Another example that includes gametophyte and sporophyte in organic connection is Hypnodontopsis conferta (Baltic amber) (Frahm, 2001b, 2004a). One of the specimens is a complete plant, consisting of a stem bearing numerous linear leaves and a complete sporophyte (FIG. 5.32). The seta is 1.5–2 mm long, twisted, and cygneous (shaped like the neck of a swan). It terminates in a short, oval capsule with 16 longitudinal ribs. The capsule is open and displays the peristome, where the 16 peristomal teeth are fused into eight pairs. Complete plants displaying gametophyte and sporophyte in organic connection have also been reported from Saxonian amber (e.g., Campylopodiella cf. himalayana; Frahm, 1996b).

Hornworts and Bryophytes


Figure 5.32 Hypnodontopsis conferta (Eocene). Bar  1.5 mm. (Courtesy J.-P. Frahm.)

This page intentionally left blank

6 The move to the land Enigmatic organisms .......................................................180

The transition to land ................................................194


Anchorage and Water Uptake ......................................................... 194

Spongiophytaceae .............................................................................185

Structural Support and Water Transport.......................................... 195

Other Enigmatic Organisms ..............................................................186

Protection Against Desiccation and Radiation ................................ 195

Isolated fragments: clues to the transition

Gas Exchange.................................................................................. 195

to land? ....................................................................................... 189

Reproduction on Land..................................................................... 196 Life History Biology ....................................................................... 196

Cuticle and Cuticle-like Material ......................................................189

Animals ............................................................................................198

Spores and Spore Tetrads ..................................................................189

A Fungal Partner ..............................................................................198

Tubes .................................................................................................192

Conclusion.............................................................................. 199

Land plant ancestors ...................................................... 193

All theory, dear friend, is grey, but the golden tree of life springs ever green. Johann Wolfgang von Goethe, Faust: The Tragedy, Part I Although land surfaces must have been available for colonization soon after life evolved in the Precambrian, the earliest record of terrestrial animals begins around 425 Ma (Ward et al., 2006) and that of plants somewhat earlier, with microfossil evidence from the Ordovician (discussed below). This delay in colonization of terrestrial habitats has been related to oxygen levels in the paleoatmosphere (Berner et al., 2007) and specifically to the lack of a sufficient ozone shield to protect terrestrial organisms from ultraviolet (UV) radiation, which acts as a strong mutagen in organisms. Because UV radiation can penetrate only a few centimeters into water, algae and other forms of marine life were protected from its effects. Ozone (O3) is formed in the stratosphere by a reaction between free oxygen (O2) and UV radiation, so oxygen levels needed to reach a certain level for an ozone layer to develop. Once formed, the ozone layer would mitigate the lethal effects of UV radiation and allow for colonization of the land. Since plants are rooted in place, they would have had to develop mechanisms to shield their cells from UV

radiation very early in evolution. Today plants have a number of adaptations that help to filter UV radiation, including pigmentation, secondary compounds, and cuticle. It is not known precisely when photosynthetic organisms began to exploit terrestrial habitats. Some have suggested that cyanobacteria existed in shallow pools on the earth’s surface relatively early in geologic time (Westall et al., 2006b; see Chapter 2). We do know, however, that beginning in the Ordovician there is evidence of organisms with characters suggesting that they lived in a desiccating environment. For many of these organisms, it is not known whether they existed for all or part of their life history on the land, but many demonstrate features that are today found only in land plants. Although the primary aim of paleobotany is to identify plants in the fossil record and to understand their phylogeny, biological activities, and paleoenvironment as much as possible, in some instances it is impossible even to place organisms in a major hierarchical category. This is true for a number of these late Paleozoic (mainly Silurian and Devonian)



PALEOBOTANY: the biology and evolution of fossil plants

fossils that exhibit some but not all features of land plants, (Edwards et al., 1998a) even though some are well preserved and have been intensively studied. Some of these organisms have been compared to vascular land plants, whereas others have a morphology or internal organization different from any other living or fossil group (Edwards, 1982). We discuss several of these unusual organisms in the section that follows.

Enigmatic organisms

internal organization constructed of longitudinally oriented tubes (hyphae, according to Hueber, 2001) of three different types (FIGS. 6.2, 6.3). Adopting the terminology used for hyphal types in trimitic extant basidiomycetes, that is those have three types of hyphae, Hueber (2001) distinguished (1) skeletal hyphae, which are thick-walled, large, long, straight or flexuous, aseptate, and unbranched; (2) generative hyphae, which are large and thin-walled, septate with an open or occluded pore, and profusely branched (FIG. 6.2); and (3) binding hyphae, which are small, thin-walled with a pore in the septum, and profusely branched. Arrangement of the


The nematophytes or Nematophyta (Strother, 1993) is an informal grouping of enigmatic Silurian–Devonian organisms that range from tiny millimeter-sized structures to severalmeter-long “logs.” Their plant bodies are constructed entirely of variously sized and shaped tubes. Although nematophytes have been studied intensively for more than 150 years, little is known about their systematic affinities, biology, and ecology. PROTOTAXITES This impressive organism is the largest representative of the Nematophyta known to date. It occurs in Silurian and Devonian rocks in the form of compressed or silicified axes (sometimes also termed pseudostems), some of which are up to 1.25 m in diameter (FIG. 6.1) and more than 8 m long (Arnold, 1952b; Jonker, 1979; Bahafzallah et al., 1981; Schweitzer, 1983, 2000; Chitaley, 1992; Hueber, 2001). Charcoalified pieces of Prototaxites have been reported from the Lower Devonian of the Welsh borderland and interpreted as products of wildfire activity (Glasspool et al., 2006). The outer surface of Prototaxites is smooth or mildly ribbed. Thin sections show a pseudoparenchymatous or plectenchymatous

Figure 6.1 Cross section of Prototaxites showing eccentric incremental growth (Devonian). Bar  3 cm.




Figure 6.2 Longitudinal section of Prototaxites southworthii

showing three types of tubes: A. binding, B. generative, and C. skeletal (Devonian). Bar  50 μm.

Figure 6.3 Section of Prototaxites southworthii showing large skeletal and narrow binding tubes/hyphe. (Devonian). Bar  50 μm.


“tissue” made up of these hyphae, along with the presence of well-defined borders of growth increments (FIG. 6.1) marked by increased tissue density, suggests some type of periodicity in growth. In addition to the tubes, numerous lacunae are scattered throughout the pseudoparenchyma. These spaces often contain disintegrated tissue and occasionally tubes, and are interpreted as formed as a result of some type of parasite. When this gigantic nematophyte was initially discovered, the specimens were believed to represent some type of fossilized conifer wood, and the organism was named Prototaxites (Dawson, 1859). After closer examination, however, the fossils were reinterpreted as algal like, transferred to another genus (Nematophycus), and formally classified with the Codiaceae (green algae). Prototaxites has also been compared to the brown alga Lessonia (Carruthers, 1872b; Kräusel, 1936). When tubes of P. southworthii were examined ultrastructurally (Schmid, 1976), however, distinct cross walls (septa) were identified, each with a centrally located elliptical aperture or pore (FIG. 6.4). These pores are superficially similar to the pores and pit connections found in certain red algae, but corresponding structures also occur in fungi (e.g., basidiomycetes) in the form of doliporous septa. In some of the tubes or hyphae, small outgrowths occur close to the septa that remotely resemble basidiomycetous clamp connections (Hueber, 2001). Both mycologists and phycologists have debated the affinities of Prototaxites for many years. Early chemosystematic data suggested an affinity with the algae, whereas the presence of cutin and suberin in the samples implies that the organisms may have been terrestrial. Evidence from other biomarkers has been used to hypothesize that

Figure 6.4 Ultrathin section of Prototaxites tube showing sep-

tal pore (arrow) (Devonian). Bar  2 μm. (Courtesy R. Schmid.)

The move to the land


Prototaxites represents a failed experiment during terrestrialization (Abbott et al., 1998), as Lang (1937) originally suggested, or that it belongs to one of several algal groups which were in the process of adapting to a terrestrial habitat during the Devonian, but failed to survive to the present time (Niklas and Smocovitis, 1983). Another interpretation of Prototaxites has been offered by Schweitzer (1983, 1990) based on a reinterpretation of Mosellophyton hefteri (FIG. 6.5) a large, irregularly branched axis from the Lower Devonian of the Mosel Valley in Germany (Wehrmann et al., 2005), which was originally interpreted as a tracheophyte (Schaarschmidt, 1974). Schweitzer (1983) combined M. hefteri with Prototaxites, and provided a reconstruction (FIG. 6.5) that depicts a kelp-like alga consisting of a monopodial trunk (Prototaxites) with a distal, multibranched

Figure 6.5 Suggested reconstruction of Prototaxites based on isolated “stems” and Mosellophyton hefteri. (From Schweitzer, 1983; Courtesy U. Schweitzer and R. Gossmann.)


PALEOBOTANY: the biology and evolution of fossil plants

Figure 6.7

Francis M. Hueber.

Figure 6.8 Hypothetical reconstruction showing the habit and

habitat of Prototaxites. (By M. Parrish; from Hueber, 2001.) Figure 6.6 Basal holdfast of Prototaxites hefteri (Devonian).

Bar  2.5 cm. (From Schweitzer, 2000; courtesy U. Schweitzer and R. Gossmann.)

crown (Mosellophyton). The alga is believed to have inhabited shallow, tidally influenced coastal marine waters where it was attached by a root-like holdfast. More recently, a branched Prototaxites specimen (FIG. 6.6) was discovered

in the Waxweiler quarry in Germany, and interpreted as a portion of the basal holdfast of this organism (Schweitzer, 2000). It should be noted, however, that M. helteri and Prototaxites have never been found in organic connection, and it remains questionable as to whether they actually belong to a single organism.


The most recent interpretation of Prototaxites is the comprehensive paper on P. loganii by Hueber (2001) (FIG. 6.7), who hypothesizes that the large “logs” represent a giant (FIG. 6.8), terrestrial, saprotrophic organism that may belong with the basidiomycetes (Chapter 3). Thus far, however, there are no easily identified structures that would support the assignment to the basidiomycetes or any other fungal group. Carbon isotope analyses of Prototaxites and land plants that lived in the same environment indicate that Prototaxites has a much wider variation in its C12/C13 ratio than would be expected in any plant (i.e., autotroph), indicating that Prototaxites was a heterotrophic organism (Boyce et al., 2007). Although a heterotroph the size of Prototaxites remains a possibility, the structure of the typical Late Silurian–Early Devonian ecosystem, which consisted of a sparse vegetation of exceedingly small plants and probably algal and cyanobacterial growths, brings into question whether there would be a sufficient source of carbon in the ecosystem to support a heterotrophic organism more than 8 m tall. Still another suggestion is that Prototaxites is an example of an ancient mutualistic association of two (or more) different organisms that combined both heterotrophy and some level of a lichen-like nutritional mode (Selosse, 2002). Although there is no direct evidence of an associated photobiont within the permineralized specimens, the hypothesis has some merit. Perhaps the photobiont consisted of cyanobacteria that were associated with the outer surface of the organism and that have either not been preserved or simply unidentified to date. Regardless of whether Prototaxites was an alga, a fungus, a lichen-like association of several different organisms, or some other type of life form not remotely related to any modern organism, it must have presented an imposing structure. As increased attention is directed at the evolution of early terrestrial ecosystems, it will be interesting to see where the affinities of this unique Paleozoic organism eventually reside. NEMATOTHALLUS Perhaps the earliest detailed account of an enigmatic organism initially believed to represent an early land plant was Lang’s (1937) description of Nematothallus from the Lower Devonian of the Welsh borderland. Specimens of Nematothallus are flat, thallus- or leaf-like, and only a few centimeters in diameter. Like Prototaxites, they are composed of a system of interlacing tubes, often of two distinct orders of size (some with internal thickenings), and usually covered by a cuticle with a pseudocellular pattern (Strother, 1988, 1993). Isolated cuticle sheets of this type are known from the Ordovician to the Early Devonian, and have flanges on the inner surface that would be interpreted as marking

The move to the land


the outlines of epidermal cells if found on a vascular plant. Scattered among the tubes and within the cuticle are spores of various sizes (Strother, 1993). It is not entirely clear, however, whether the spores, tubes, and cuticle-like layers actually represent parts of the same organism. Perforations up to 100 μm in diameter are present on some of the cuticles and these possess various types of thickenings near the surface. Like the pores found in other enigmatic organisms and modern liverworts, those of Nematothallus may have functioned in some form of gas exchange. Edwards and Rose (1984) noted, however, that similar pores have not been found in all Nematothallus cuticles. Other possible functions for the cuticle pores in Nematothallus, as well as those found in some other Paleozoic enigmatic organisms of this type, include sites of gamete liberation or a response to wounding. Corsin (1945), Jonker (1979), and others have suggested that Nematothallus represents the remains of leaf-like structures that were produced on Prototaxites-like axes. Although this is an interesting hypothesis, no specimens in organic connection are currently known that might support any of these hypotheses. Strother (1993), after examining a large assemblage of Nematothallus-like organisms, suggested that the construction of these organisms may be indicative of a subaerial habitat. NEMATOPLEXUS Nematoplexus rhyniensis is a silicified, thalloid fossil known only from incomplete, partially decayed, and generally fragmentary remains from the Rhynie chert (Lower Devonian). The overall morphology and size of the organism remain unknown. It consists of a meshwork of interlaced aseptate and septate tubes of two or three different sizes (Lyon, 1962). Most of the larger, aseptate tubes have annular or helical thickenings; some of the narrow, smooth-walled tubes display distinct septa with a centrally located elliptical aperture or pore (FIG. 6.9). These pores are similar, if not identical, to the pores found in Prototaxites. Nematoplexus tubes apparently do not branch, except for certain areas termed branch knots in which aggregations of tubes form dense, multibranched clusters (FIG. 6.10). The biological significance of the branch knots remains unknown. It has been suggested that N. rhyniensis may represent the permineralized equivalent of Nematothallus. A second, even less well-documented and understood nematophyte from the Rhynie chert has been given the name Nematophyton taiti (Kidston and Lang, 1921a). NEMATASKETUM DIVERSIFORME This Devonian organism is structurally similar to Prototaxites (Burgess and Edwards, 1988) and is based on coalified, three-dimensionally preserved specimens. It also


PALEOBOTANY: the biology and evolution of fossil plants

Figure 6.9 Nematoplexus rhyniensis wide tube with helical

thickenings and narrow tube with septal pore (arrow) (Devonian). Bar  100 μm. (Courtesy H. Hass.)

Figure 6.11 Fractured surface of Nematasketum diversiforme showing large tubes and smaller interwoven filaments (Devonian). Bar  100 μm. (From Burgess and Edwards, 1988.)

Figure 6.10 Nematoplexus rhyniensis with branched knot

(Devonian). Bar  45 μm. (Courtesy H. Hass.)

consists of tubes of two sizes (FIG. 6.11), including some with internal thickenings. In attempting to decipher the affinities of both Nematasketum and Prototaxites, Burgess and Edwards (1988) proposed either fungal or algal affinities,

but offer equally plausible reasons to refute inclusion in either group. Instead, they place both genera in the Nematophytales, an order erected by Lang (1937). Burgess and Edwards (1991) later developed an artificial classification system, based on one used for dispersed palynomorphs, to aid in description and comparison of some of these organisms constructed of various types of tubes.


The move to the land


Figure 6.12 Diagrammatic view of Pachytheca showing two-

layered wall. (From T. Taylor, 1988a.)

PACHYTHECA This genus is known from the Late Silurian to the Middle Devonian (Barber, 1889, 1891; Stockmans and Willière, 1938; Lang, 1945; Schmidt, 1958), and is often cited in discussions of possible land plant origins. Pachytheca specimens are small (0.1–0.6 cm in diameter), spherical, and organized into two distinct zones (Niklas, 1976c; Gerrienne, 1991). The inner zone (medulla) consists of densely spaced, intertwined tubes (FIG. 6.12), whereas the outer layer (cortex) is constructed of tubes that are distinctly radially aligned. Between cortex and medulla is a narrow area in which the tubes seem to change from random to radial alignment. In some specimens, a narrow canal is visible that appears to connect the medulla with the exterior. Pachytheca has repeatedly been interpreted as a dispersal unit (cystocarp) of Prototaxites (Schmidt, 1958; Jonker, 1979; Schweitzer, 1983); however, fossils in organic connection have not been discovered to date. Niklas (1976c) suggested that Pachytheca represents the juvenile form of Parka decipiens (discussed below), and thus the two taxa are merely different ontogenetic stages of a single organism. In an interesting set of experiments, Graham et al. (2004) observed extant specimens of the liverworts Marchantia and Conocephalum at various stages of controlled tissue degradation. Based on the results of these experiments, they suggest that some of the enigmatic fossils commonly referred to as nematophytes may represent remains of ancient liverworts in various stages of decay and that some of these plants shared certain features with modern marchantioids. Although these authors do not suggest that extant liverworts existed in the Devonian, they discuss the possibility that some of the cuticle sheets and tubes of these extinct forms may represent characters that have persisted during the evolution of the group (see Chapter 5).

Figure 6.13 Airpores (arrows) in Marchantia thallus. (Extant)

Bar  100 μm.


The Spongiophytaceae an artificial taxon used to include a variety of enigmatic early organisms, some of which may represent or be related to vascular land plants (Kräusel, 1954; Kräusel and Venkatachala, 1966). Some organisms included in this family show features of vascular plants, but not all; as a result, they are sometimes interpreted as depicting transitional stages in the evolution of true vascular land plants. SPONGIOPHYTON Spongiophyton is a thalloid fossil that has been reported from numerous Devonian localities throughout the world (Kräusel, 1954; Kräusel and Venkatachala, 1966; Boureau and Pons, 1973; Chaloner et al., 1974; Zdebska, 1978; Gensel et al., 1991; Guerra-Sommer, 1993; Griffing et al., 2000), and has variously been interpreted as a colonial animal, alga, vascular plant, and bryophyte. Specimens are small (25 mm wide by 2.5 cm long), branched axes with rounded tips (FIG. 6.14). It has been suggested that Spongiophyton grew in a desiccating environment, based on the presence on one surface of the thallus of circular pores ranging 200–300 μm in diameter; these have been compared to the pores in certain modern liverworts like Marchantia (FIG. 6.13). The pores are believed to have functioned in gas exchange, and perhaps represent a transitional stage in the evolution of stomata (Chaloner et al., 1974). Additional support for life in a desiccating environment is the presence of a cuticle (or cuticlelike surface layer) covering the Spongiophyton thallus. On the thallus surface which contains the pores, the cuticle is approximately three times thicker than on the non-poral


PALEOBOTANY: the biology and evolution of fossil plants

surface (FIG. 6.14). Nothing is known about the internal tissues nor the reproductive organs of this organism, although it has been suggested that perhaps there was a hyphal network like that found in certain lichens (Stein et al., 1993; Taylor, W. A. et al., 2004). The lichen hypothesis was initially supported by carbon isotope ratios from Spongiophyton (Jahren et al., 2003). Fletcher et al. (2004), however, found that the range of δ13C values in Spongiophyton is not significantly different from those of lichens, liverworts, or mosses, so carbon isotopes are not useful in identifying this fossil. As a result, the affinities of Spongiophyton remain conjectural. Continued research is needed to clarify whether this organism possesses a complement of features that are intermediate between algae and land plants. ORESTOVIA Another Early–Middle Devonian organism included in the Spongiophytaceae by Kräusel and Venkatachala (1966) is Orestovia. This interesting, compressed organism has been isolated from paper coals of western Siberia and consists of naked, unbranched cutinized axes up to 20 cm long and 2 cm wide that taper distally (Ergolskaya, 1936; Ishchenko and Ishchenko, 1980; Krassilov, 1981a). Most specimens are preserved as hollow cuticular envelopes, often displaying an epidermis-like (?pseudo-)cellular pattern. Extending the length of each axis is a delicate strand of elongate, tracheidlike tubes with annular to reticulate thickenings on the internal wall. Many tubes contain a central core that is thought to represent resin. On the outer surface of the axes are irregular swellings thought to have had a secretory function. Small, sunken pores are randomly distributed on the axes and these are surrounded by several rings of (?pseudo-)cells; these structures have been interpreted as stomatal complexes with sunken guard cells and subsidiary cells arranged in several

rings. In a few specimens, spores occur in the cortex of the axes and range from 150 to 190 μm in diameter. Fungi have been reported on cuticles of Orestovia, but the nature of these remains equivocal. Further analysis of the structure of the cuticle and stomata prompted Gensel and Johnson (1994) to suggest that Orestovia is a tracheophyte of uncertain affinity. Snigirevskaya and Nadler (1994) reconstructed Orestovia as a semiaquatic plant resembling the extant marsileaceous fern Pilularia globulifera in overall appearance, with creeping rhizomes and naked orthotropous axes that display circinate vernation when immature. Aculeophyton is a third genus assigned to the Spongiophytaceae by Kräusel and Venkatachala (1966). It comes from the Lower Devonian of Siberia and differs from Orestovia in that the surface of the axes is covered by massive, conical papillae. Other Enigmatic Organisms

PROTOSALVINIA Although a great deal of information is known about Protosalvinia (initially named Foerstia), the biological affinities of this Devonian organism remain problematic (Niklas and Phillips, 1976; Niklas et al., 1976; Schopf, 1978b; Gray and Boucot, 1979). Like Spongiophyton, it has been interpreted as a fern, alga, bryophyte, or some form of semiaquatic organism. It is typically found compressed (FIG. 6.15), but may assume a variety of morphologic shapes, ranging from nearly circular to clavate. Some of the largest specimens approach 1 cm in diameter. At least three taxa, P. arnoldii, P. ravenna, and P. furcata (FIG. 6.16), have been suggested as representing different growth forms of a single biological species. According to this developmental chronology, the thallus becomes more clavate and develops depressions

Figure 6.14 Diagrammatic reconstruction of Spongiophyton

nanum thallus showing pores and cuticular flanges. (From T. Taylor, 1988a.)

Figure 6.15 Cleared tip of Protosalvinia thallus (Devonian).

Bar  1 mm.


The move to the land


Figure 6.16 Protosalvinia furcata apical notch (Devonian).

Bar  450 μm. (From Niklas et al., 1976.)

Figure 6.18 Protosalvinia braziliensis showing position of tetrads (Devonian). Bar  475 μm. (From Niklas et al., 1976.)

Figure 6.17 Protosalvinia ravenna showing conceptacles at

apical end (Devonian). Bar  275 μm. (From Niklas et al., 1976.)

containing spores as development progresses (Phillips et al., 1972). On the surface of the thallus are cell patterns that suggest that thalli dichotomized, and specimens with an apical notch are known. At the tips of the thallus are slight depressions (FIG. 6.17), termed hypodermal conceptacles, which contain tetrads (FIGS. 6.18, 6.19) of large (200 μm), thick-walled spores (Niklas and Phillips, 1976). Each conceptacle is slightly 0.5 mm in diameter and constructed of two distinct cell layers. Spores are large and thick-walled (FIG. 6.20). In the Curiri Formation (Famennian) of the Amazon Basin, Melo

Figure 6.19 Tetrad of spores in depression in thallus of Protosalvinia ravenna (Devonian). Bar  125 μm. (From W. Taylor, and T. Taylor, 1987.)

and Loboziak (2003) correlated spores of the sporae dispersae genus Retusotriletes with Protosalvinia, where they characterized a distinct biostratigraphic unit throughout the basin. It has been debated whether the spores represent meiotic


PALEOBOTANY: the biology and evolution of fossil plants

Figure 6.21 Suggested reconstruction of Parka decipiens

showing lower (left) and upper surfaces. (From T. Taylor, 1988a.)

Figure 6.20 Protosalvinia spore (Devonian). Bar  200 μm.

or mitotic products (FIG. 6.20), and thus how they may have functioned in the life history of the organism. The spore wall ultrastructure suggests that they are the result of meiosis, even though they do not appear to have been constructed of sporopollenin (W. Taylor and T. Taylor, 987). The presence of tetrads of spores does not unequivocally establish Protosalvinia as a land plant, since some red and brown algae also produce spores and eggs that are morphologically similar to the spores recovered from Protosalvinia, but the similarity does not extend to the ultrastructural level. The discovery of lignin-like compounds in the fossils suggests land plant affinities for Protosalvinia (Romankiw et al., 1988), but it has also been interpreted as related to the brown algae based on biochemical evidence (Niklas, 1976b). Others consider Protosalvinia inhabited to be an alga that shows some land-adapted features, which may result from convergent evolution at a time when land plants were first becoming established. Gutschick and Sandberg (1991) suggested that Protosalvinia inhabited brackish coastal swamps. Despite all of the tools available to paleobotanists to examine the nature of the organic matter in Protosalvinia, the biomolecular signature is neither distinctly marine nor terrestrial (Mastalerz et al., 1998). Some Protosalvinia specimens have been suggested to represent parts of the cephalopod Sidetes (Hannibal, 1994). Although the biological affinities of Protosalvinia may remain unclear, it has served as a significant index fossil, that is the Protosalvinia Zone, within Upper Devonian black shale sequences of the eastern United States (Schopf and Schwietering, 1970; Murphy, 1973; Schwietering and Neal, 1978). It has also been useful for stratigraphic correlation

among Devonian rocks of the Michigan, Illinois, and Appalachian Basins (Matthews, 1983). More recent studies, however, have shown that the age of the Protosalvinia Zone locally varies from middle to late Famennian, which reduces its value as an index fossil and correlation aid. As noted earlier, a Protosalvinia Zone has also been documented in the Amazon Basin of northern Brazil (Grahn, 1992; Loboziak et al., 1997; Melo and Loboziak, 2003). PARKA Another Late Silurian–Early Devonian thalloid organism of uncertain affinity is Parka decipiens (Fleming, 1831; Reid et al., 1897; Don and Hickling, 1917; Neuber, 1979). This fossil attracted the attention of biologists because of its morphological similarity to some members in the Coleochaetales, an extant order of green algae that many believe is significant in deciphering the ancestry and origin of land plants (discussed below). Specimens of P. decipiens are 7 cm in diameter with an oval outline and slightly undulating margin (FIG. 6.21). Impressions of cells on the surface of some compressed specimens suggest that the underlying tissue was pseudoparenchymatous, and growth simulation models suggest that the thallus grew by means of both apical and anticlinal intercalary growth (Niklas, 1976c). On the surface of the thallus are disk-shaped structures which have been interpreted as some type of sporangium (FIG. 6.22); many contain small (25–45 μm) compressed bodies thought to represent spores. None possess haptotypic features such as a trilete mark. At the ultrastructural level, the wall consists of lamellae of various thicknesses (Hemsley, 1989b) similar to those in the spores of certain extant liverworts (Hemsley, 1990). Parka has also been suggested as a potential ancestor to Coleochaete in the transition to a terrestrial habitat (Niklas, 1976c). Although this hypothesis is intriguing, major obstacles remain, including the time gap of more than


Figure 6.22 Diagrammatic reconstruction of Parka decipiens thallus with structures interpreted as sporangia. (From T. Taylor, 1988a.)

400 million years between the two taxa, together with uncertainties regarding the life history of the fossil. Moreover, antheridia have not been found in P. decipiens, and if these two taxa are similar, the presumed spores in Parka would actually represent zygotes.

Isolated fragments: clues to the transition to land? Isolated cuticle fragments, spores, and tubes are found in Lower to Middle Ordovician rocks, and extend through the Silurian into the Lower Devonian (Edwards et al., 1979; Wellman, 2001; Edwards, 1986). Although even less is known about these fragments than the enigmatic fossils described earlier, they constitute an important source of paleobiological data because they occur at a point in geologic time that coincides with the origin of a terrestrial biota (T. Taylor, 1982a, 1988a; Kenrick, 2003; Steemans and Wellman, 2004). Cuticle and Cuticle-Like Material

Isolated sheets of resistant material in the fossil record are routinely referred to as cuticle or cuticle-like sheets; early records of this type of microfossil include specimens from the Early Ordovician of Tunisia (Combaz, 1967) and early Middle Ordovician of Saudi Arabia (Le Hérissé et al., 2007). The affinities of the organisms that produced the cuticle were probably very diverse and included not only plants, but animals, fungi, and perhaps lichen-like associations. Although most cuticle sheets lack distinctive ornamentation, similar to those of Nematothallus (Edwards and Rose, 1984), a few have pores or projections. Most of these cuticle sheets have been obtained from bulk maceration of rock in acid (usually

The move to the land


hydrofluoric acid) and subsequent oxidation of the sample. Using such techniques, Edwards (1986) recovered numerous cuticles from Lower Devonian shales in Great Britain. On the inner surface of some specimens are inwardly directed flanges that may have marked the position of interior tissues. On the outer surface are projections of low relief. Because the biological affinities of these cuticles remain unknown, an artificial classification was created so that future workers would be able to correlate the various types of cuticles geographically and stratigraphically. These data may ultimately make it possible to trace the organisms in time and space once the questions of biological affinities are resolved. As paleobiologists learn more about the spatial and temporal distribution of cuticle fragments, and their macromolecular composition, the possibility of identifying which organisms produced the cuticle sheets also increases, and perhaps what evolutionary adaptations they represent. Spores and Spore Tetrads

In recent years, isolated spores and spore tetrads of Ordovician and Silurian age have received considerable attention as potential proxy records for early land plants (Wellman and Gray, 1988b, 2000; Edwards and Wellman, 2001; Steemans and Wellman, 2004) (FIG. 6.23). Although spore-like microfossils have been reported from as early as the Middle Cambrian (Strother, 2000; Strother and Beck, 2000), to date the oldest spore assemblages believed to have been produced by some land-inhabiting plant occur in lower Middle Ordovician (Llanvirn) rocks (Vavrdová, 1984, 1990; Gray, 1985; Strother et al., 1996; Steemans, 1999; Le Hérissé et al., 2007). Nearly identical spore assemblages occur from the Llanvirn to the late Llandovery (Early Silurian) (Richardson, 1988; Steemans et al., 1996; Rubinstein and Vaccari, 2004; Richardson and Ausich, 2007), suggesting a period of relative evolutionary stasis 40 myr in duration (Wellman and Gray, 2000; Steemans, 2001). Commonly referred to as cryptospores (Richardson, 1996), these microfossils occur as monads, dyads, and tetrads (FIGS. 6.24, 6.25), some surrounded by a smooth, loosely attached membrane (FIG. 6.26) (Strother, 2000). Strother and Beck (2000) categorized cryptospores (FIG. 6.27) as a class of organic-walled microfossils of probable terrestrial origin, whereas Steemans (2000) restricted the term to spore-like microfossils that are unambiguously attributable to the Embryophyta (Le Hérissé et al., 2007). Individual spores show no clearly defined exit site and are termed inaperturate. Other spores that first appear in Silurian rocks are solitary and possess a distinct trilete laesura (FIG. 6.28) (Steemans, 2000). In younger rocks there is a


PALEOBOTANY: the biology and evolution of fossil plants

40 30

Number of miospore species








6a 6b 6c Ashgill









3a 3b 4a 4b




naked laevigate monads naked monads with pseudo-trilete mark (i.e., Imperfectotriletes) naked fused dyads (i.e., Pseudodyadospora) naked unfused dyads (i.e., Dyadospora) naked fused tetrads (i.e., Tetrahedraletes) naked unfused tetrads (i.e., Stegambiquadrella) smooth envelope enclosed dyads (i.e., Segestrespora) envelope enclosed ornamented tetrads ornamented envelope enclosed dyads (i.e., Segestrespora) smooth envelope enclosed tetrads (i.e., Velatitetras) ornamented envelope enclosed tetrads (i.e., Velatitetras) envelope enclosed monads (i.e., Sphaerasaccus) Trilete spores (Ambitisporites)

Stratigraphic ranges of numerous miospore morphologies showing evolution of biodiversity from the Ordovician into the Silurian. (From Steemans and Wellman, 2004.)

Figure 6.23

Figure 6.24 Tetrahedraletes medinensis tetrad (Silurian). Bar  15 μm. (Courtesy P. Strother.)

6.25 Quasiplanar cryptospore tetrad (Cambrian). Bar  5 μm. (Courtesy P. Strother.)



Figure 6.26 Quasiplanar cryptospore tetrad with thin enclosing wall (Cambrian). Bar  10 μm. (Courtesy P. Strother.)

The move to the land


Ambitisporites dilutus (Silurian). Bar  25 μm. (Courtesy P. Strother.) Figure 6.28

Figure 6.29 Bifurcating tip of Grisellatheca salopensis

(Devonian). Bar  500 μm. (From Edwards et al., 1999.)

6.27 Dicryptosporites radiatus monad (Silurian). Bar  20 μm. (Courtesy P. Strother.)


pronounced increase in the diversity of spore types, as well as in the complexity of ornamentation (Gray et al., 1974; Wood, 1978; Beck and Strother, 2001), suggesting a major radiation of land plants during this period. Based on comparisons of Late Silurian plant microfossils from China with assemblages reported from other parts of the world, Wang et al. (2005) hypothesized that Late Silurian floras were cosmopolitan and exhibited little paleogeographic differentiation.

Although a few fragments of sporangia (FIG. 6.29) have been found containing these early spores (Wellman et al., 1998, 2003), their biological affinity and significance has continued to be challenged. Some solitary spores are believed to have been produced by algae, whereas the dyads and tetrads are thought to have represented a liverwort grade of land plant, or perhaps even vascular plants (Steemans and Wellman, 2004). The presence of land plants with a bryophytic level of organization as early as the Ordovician is supported by both paleobotanical and molecular evidence (Bateman et al., 1998; Qiu et al., 2006; Renzaglia et al., 2007). The ultrastructure (FIGS. 6.30, 6.31) of some of these spores is also expanding the database of wall organization patterns (Taylor, 2002, 2003) and, as a result, is opening new lines of evidence directed at determining the affinities of


PALEOBOTANY: the biology and evolution of fossil plants

Figure 6.30 Ultrathin section of scabrate spore walls (Cambrian). Bar  500 nm. (Courtesy W. Taylor.)

(FIG. 6.24), some of which possess an outer membrane. Smooth-walled, solitary, trilete spores mark the second zone, which extends from the Lower to Middle Silurian. The third zone (Middle and Upper Silurian) contains spores with various types of external ornament, suggesting increasing levels of diversity. Tetrads of spores are also found throughout this last zone, but without enclosing membranes. Despite the uncertainties regarding the affinities of various cryptospores, their presence, together with other types of fragmentary debris as early as the Middle Cambrian, indicates the presence of some type of subaerial photosynthetic cover during this time period (Strother et al., 2004). These authors also noted that at least some of the cryptospores are no doubt homologous with younger Ordoviocian and Silurian forms, but that the record to date suggests that during the Cambrian there was a diverse mesoflora of eukaryotic photoautotrophs that were derived from chlorophytes and/or charophytes, together with various thalloid organisms constructed of filaments. Tubes

Figure 6.31 Ultrathin section of cryptospore wall showing lamellae (Silurian). Bar  500 nm. (Courtesy W. Taylor.)

major plant groups. These studies will also serve to distinguish phylogenetically important features from those that are developmental in scope (Wellman, 2004). Despite the fact that most early spore types cannot be traced with certainty to major groups of plants, they are still a significant source of information with regard to biostratigraphy, paleophytogeography, and paleoecology (Steemans et al., 2007). An excellent synthesis is the study by Gray (1985) who identified three microfossil zones based on dispersed spores. The first zone extends from the Middle Ordovician to the Lower Silurian, and is dominated by spores in tetrads

Isolated tubes represent the third type of fragmentary fossil found as early as the Late Ordovician (Vavrdová, 1988). One study reports two types of tubes with annular–helical thickenings from the Mesoproterozoic (1–1.6 Ga) of northern China (Yin et al., 2004). Most tubes can be separated into two groups. In one group are narrow tubes, 8–20 μm in diameter and about 50 μm long, which are generally smooth, aseptate, and unbranched. The second group has tubes that are longer (200 μm long) with annular–helical thickenings on the inner surface. Some have a papilliform or bulb-shaped tip. Burgess and Edwards (1991) developed an artificial classification system for isolated tubes and filaments found in Ordovician–Lower Devonian rocks, and suggested that many belong to the nematophytes (Lang, 1937). The attempt to define the stratigraphic and morphological extent of these tubes may ultimately prove an important component in understanding their biology. Some tubes are like those found in certain green algae, especially members of the Dasycladales; others suggest the tubes were produced by a variety of different organisms. Although the biological affinities of the cuticle-like sheets, isolated spores, and tubes continue to remain speculative, they all suggest features that are analogous to those of the land plants found in slightly younger rocks (Edwards et al., 1998a). The presence of cuticles on aerial plant parts and spores enclosed in a wall made of sporopollenin are primary antidesiccation features found in vascular plants and a few bryophytes today. Although the Ordovician–Silurian tubes


The move to the land


do not precisely fit the definition of a tracheid or any other known type of putative water-conducting cell, they nevertheless show structural and functional equivalents in organisms that were adapting to a desiccating environment, a major feature in the transition to the land (discussed below).

Land plant ancestors For many years the green algae (Chlorophyta) were the group thought to be most likely to have given rise to the land plants. Today most regard the green algae and embryophytes together as a monophyletic group, the Viridiplantae, which consists of two monophyletic lineages, the Chlorophyta and the Streptophyta. Included in the Streptophyta are all embryophytes, that is bryophytes and vascular plants, and a distinct group of green algae traditionally known as the Charophyceae, which includes the Charales, Coleochaete, and the Zygnematales, among other taxa (Simon et al., 2006) (Chapter 4). Although the fossil record of the green algae in general is extensive (Chapter 4), fossils have thus far provided relatively little information about the steps involved in the transition to a land habitat. In contrast, studies of living green algae have produced an array of molecular and morphological data that underscore the phylogenetic relationships between these two groups. Based on multiple characters, the Charophyceae are regarded as the green algal lineage most closely related to land plants (Huss and Kranz, 1997; Nishiyama, 2007). This evidence was initially biochemical and ultrastructural (Mattox and Stewart, 1984), but more recently has included molecular sequence data (Kranz et al., 1995; Karol et al., 2001; Lewis and McCourt, 2004; McCourt et al., 2004; Qiu, 2008). The question remains, however, as to which of the orders within the Charophyceae—Charales or Coleochaetales—is more closely aligned to the land plants (Chapman and Waters, 2002), as several molecular analyses provide conflicting information (Karol et al., 2001; Delwiche et al., 2002; Lewis and McCourt, 2004). Based on comparative genomic analyses, Turmel et al. (2007) suggested that the Charales are sister to a clade which consists of the Coleochaetales, Zygnematales, and land plants. They note that the question of which particular group of charophycean algae is most closely related to the land plants is a complex one for which we still do not have a clear answer. Much has been written about Coleochaete (FIG. 6.32) as a model organism in the study of land plant ancestors (Graham, 1996). The morphology of this genus has been used to suggest a way in which a filamentous thallus might have given rise to a parenchymatous land plant body

Figure 6.32 Diagrammatic view of Coleochaete with setae.

(From T. Taylor, 1988a.)

Figure 6.33 Coleochaete pulvinata showing several zygotes in various stages of cortication (Extant). Bar  120 μm. (Courtesy C. F. Delwiche.)

(Graham, 1982, 1984), and the apically biflagellate male gamete is also consistent with the sperm of land plants. The non-motile female gamete in Coleochaete and oogamous reproduction in embryophytes (Blackwell, 2003) have also been compared. Moreover, Coleochaete is the only living green alga in which some species possess zygotes (FIG. 6.33) that are retained on the maternal plant and corticated by a layer of sterile gametophytic cells (Graham, 1984;


PALEOBOTANY: the biology and evolution of fossil plants

McCourt et al., 2004). In at least one species, the zygote receives nourishment from placental transfer cells with wall ingrowths that increase the area in contact with the zygote (Graham and Wilcox, 1983). This pattern is similar to the archegonial venter cells of lower embryophytes. The larger number of meiotic products (8–32) in Coleochaete has been suggested as an adaptation in competing for available substrate space (Hopkins and McBride, 1976). Sporopollenin in the inner wall of the zygote of Coleochaete and lignin in the thallus, perhaps functioning in an antimicrobial manner, add additional characters that have been used to suggest a relationship between this charophycean alga and early land plants (Delwiche et al., 1989).

The transition to land The transition from an aquatic habitat to life on land was a major evolutionary event in the history of photosynthetic organisms, which involved a number of important physiological changes and structural modifications to the plant body (T. Taylor, 1982a, 1988a; Graham et al., 2000). When plants first moved onto the land, the Earth’s surface was probably already inhabited by various cyanobacteria, algae, fungi, and perhaps lichens (Taylor et al., 1997). As in arid regions today, microbial mats and biological crusts consisting of communities of microorganisms were probably important in soil formation, and there is evidence of fossil soils (paleosols) as early as the Proterozoic (Hasiotis, 2002; Prave, 2002). Molecular data obtained from the scaly green flagellate Mesostigma viride (Streptophyta) suggest that several major physiological changes, for example, in the regulation of photosynthesis and photorespiration, took place early during the evolution of the streptophytes, that is before the transition to land (Simon et al., 2006). The transition to a desiccating terrestrial habitat probably occurred sometime in the Ordovician or earliest Silurian. By the Late Silurian–Early Devonian, there is evidence of a radiation of land plants and a number of unique adaptations are found in several groups of organisms (Bateman et al., 1998). In some ways, the move of plants onto the land is similar to the evolution of the soft-bodied, late Neoproterozoic marine faunas—the Ediacaran faunas. In both cases, there were apparently a number of open niches and the plants that filled these niches exhibited a number of unusual morphologies, many of which did not survive beyond the Devonian. Although perhaps overly simplistic, it is now apparent that the ability to exist on land is the result of numerous complex interactions that involved the interplay between structural and physiological adaptations in the plants themselves,

symbiotic interactions at several levels, and physical and chemical changes in the environment. In the following sections, we will consider the major adaptations that are necessary for life on the land, including anchorage and water uptake, support for the upright plant body, water movement through the plant, desiccation prevention, some mechanism for gas exchange, reproduction in a terrestrial environment, and life history strategies. All of these features are interrelated and some structures perform multiple roles. Anchorage and Water Uptake

Long before there was any real appreciation of the diversity of fossil plants, the French botanist Octave Lignier (1908) advanced a theory about the morphological changes necessary during the move onto the land and the evolution of roots. His hypothesis used an algal ancestor with a threedimensional, dichotomously branched system that was periodically desiccated during fluctuations of available water. According to Lignier’s scenario, one segment of the branching system became covered with substrate and over time assumed the function of an anchoring and absorbing organ, much like a root. Thus, the “root” of this early land plant would be homologous with an aerial branch system, differing only in function. Although Lignier’s scenario began with an alga that we now know is not closely related to land plants, his hypothesis of a morphological model is strengthened by the occurrence of some early land plants with no organ differentiation between aerial stem and prostrate rhizome (e.g., Aglaophyton major, see Chapter 8). These permineralized axes possess the same complement of cells and tissue systems in both the aboveground aerial axes and the rhizomes. Additionally, the rhizomes produced tufts of rhizoids only on those portions in direct contact with the substrate and stomata can also be found on the rhizome, indicating that these parts of the plant may have been photosynthetic. Lignier’s model addresses a fundamental need for land plants—some way to anchor themselves to the substrate. In addition, unless a plant has a flat, thalloid plant body, the underground portions must also serve to anchor upright axes. The early land plants and the modern vascular cryptogams have no true roots, that is, with the specialized anatomy and morphology of roots (Chapter 7). Their anchoring organ is generally a rhizome, a horizontal stem which is either below ground level or on the surface of the substrate. These early plants produced rhizoids, small, usually unicellular hair-like structures, on the rhizomes, which absorbed water and minerals from the substrate. Once an early land plant was anchored in the substrate, some method of moving water and nutrients from the


substrate to the rest of the plant was needed. The ancestral aquatic alga would have been suspended in water and water could easily enter the organism by diffusion and osmosis; the terrestrial realm, however, was a hostile and desiccating environment. One of the most important structural adaptations of the plant body was the evolution of mechanisms and structures to both obtain and conserve water. A plant growing on land has some water in the surrounding air, for example as rain, fog, or dew, but potentially more water in the substrate in which it is anchored. Thus, early land plants had to have a system to absorb water from the air and/or the substrate, as well as effective mechanisms to prevent water loss (discussed later) in order to survive periods of drought. Extant land plants overcome these obstacles in two different ways. Bryophytes are poikilohydric, that is they have no specialized mechanism to prevent desiccation, but many can tolerate desiccation and rehydrate later. Vascular plants, however, have evolved homoiohydry, that is the capacity to remain hydrated internally. This adaptation, however, except in a few rare cases (e.g., Selaginella lepidophylla, the resurrection plant), is coupled with vegetative intolerance of desiccation (Proctor and Tuba, 2002; Proctor et al., 2007). Water uptake by bryophytes occurs in the form of simple diffusion and osmosis, either internally or externally. Some extant liverworts, for example certain members of the Calobryales and Metzgeriales, contain endohydric conduits (those on the inside of the thallus), and some mosses, for example in the Bryales and Polytrichales, possess hydroids and leptoids that are functionally equivalent to the xylem and phloem of vascular plants, although the hydroids are structurally very different from tracheids (Hébant, 1977; Ligrone et al., 2000). In contrast, vascular land plants use a variety of specialized subterranean and aerial absorbing structures, which allow them to live in almost any environment on Earth. Structural Support and Water Transport

As noted earlier, a successful transition to land required structural modification for upright support of the plant. It is impossible to separate a discussion of support in early, upright land plants from water transport, since in most living vascular plants, the vascular tissues (xylem and phloem) are involved in both support and conduction (Chapter 7). In many early land plants, however, it appears that the central strand initially functioned primarily in conduction and that the plant stood erect as a result of turgor pressure in the parenchymatous cells of the axis (Speck and Vogellehner, 1988; Niklas, 1990). As land plants continued to evolve and grew larger, vascular tissue also took over the role of support, as it does in modern vascular plants. Tracheary elements in

The move to the land


the xylem and fibers in the phloem (Chapter 7) have secondary cell walls impregnated with lignin, a polyphenolic polymer which provides structural support and flexural stiffness to the plant organ. Many of the early land plants, however, exhibit a central strand of conducting elements, but these do not have the secondary wall thickenings that are characteristic of xylem tracheids (Kenrick and Crane, 1991). Instead, the central strand is made up of a series of tubes, some with internal or external bands that superficially resemble tracheid thickenings (Chapter 8). These bands, however, are made of primary cell wall material and do not appear to be lignified, but based on their location and structure they must have served in conducting water throughout the plant. The discovery of these interesting cell types in the central strand of plants that were once thought to contain vascular tissue necessitates a continued reexamination of all early land plant cell types, especially those involved in conduction. Protection against Desiccation and Radiation

In addition to structural support and conduction, evolving land plants also required a method to retain water in a desiccating environment. It is believed that these organisms existed for at least a portion of their life history in a terrestrial, desiccating environment where uncontrolled transpiration, and thus water loss, presented a major physiological problem. Many of the enigmatic organisms discussed earlier in this chapter possess a non-cellular outer envelope that has been referred to as a cuticle or cuticle-like layer, and may have been effective as a boundary layer against excessive transpiration. The presence of sporopollenin in the spore wall represents a similar adaptation to prevent desiccation of reproductive propagules. At the same time, the cuticle or cuticle-like layer may have been effective in the attenuation of UV radiation, including the especially dangerous UV-B (Raven, 2000). As noted earlier, this would have been a particularly important function of the cuticle in terrestrial habitats. It is interesting to note that many of the enigmatic Silurian–Devonian organisms, such as Orestovia and some early embryophytes, are characterized by a massive cuticle, which exceeds in thickness that of most plants found in geologically younger rocks. Gas Exchange

Once early terrestrial plants had developed a cuticle, they would then need some means for gas exchange (Raven, 2002), as the cuticle is only very weakly permeable to gases. Although the process of photosynthesis functions in both aquatic algae and terrestrial plants, the source of carbon dioxide for


PALEOBOTANY: the biology and evolution of fossil plants

each system is quite different. In algae, carbon dioxide dissolved in the water is available to chloroplast-containing cells through osmosis, whereas in most land plants carbon dioxide enters the plant through specialized openings termed stomata (Chapter 7). The regulation of these openings provides for a physiological balance within the plant (Hetherington and Woodward, 2003), a system that is regulated by other means in algae. Stomata have been identified on the naked aerial axes of many early land plants (Edwards et al., 1998a; Habgood et al., 2002); they have also been reported as occurring on the axes of the free-living gametophyte generation of several Rhynie chert plants (Kerp et al., 2004) (Chapter 8). Less specialized pores in the cuticle (FIG. 6.14) occur on other presumably terrestrial organisms, such as Spongiophyton and Orestovia, suggesting that these plants also may have carried out some level of gas exchange (Chaloner et al., 1974). Like the pores in extant liverworts, however, these unspecialized pores do not appear to include a mechanism to regulate their opening and closing, as do the stomata of vascular plants. Reproduction on Land

A primitive land plant requires several adaptations in order to reproduce on land. One of these is a method to move gametes from one gametophyte to another in order to effect fertilization, and the other is some method of spore dispersal in which the dispersal units are protected from the desiccating environment. In addition, it is important to move the sporophytic reproductive units up off the substrate, both to prevent infection by microbes and presumably to disperse spores further, in order to colonize new substrates. Ancestral algae produce motile gametes and spores in an aqueous environment. In the most primitive land plants, gamete transfer is still dependent on water, as it was in the algae, but spores of even the earliest land plants are already protected by a wall of sporopollenin, which persists through all the subsequent land plant lineages. Thus, land plant sporophytes produce spores or seeds that resist desiccation and can be transported great distances by abiotic or biotic vectors. The simplest method to raise dispersal units off the substrate is a sporangium positioned terminally on upright axes, and numerous examples of the earliest bona fide land plants show terminal sporangia on tiny unbranched or branched naked axes (Edwards and Wellman, 2001). As will be seen in later chapters, the vascular plant sporophyte in higher plants has become so dominant that the gametophyte phase is completely dependent on the sporophyte for its survival. In the more specialized vascular plants, that is the seed plants, water as a medium for fertilization is no longer necessary. Whether we are speaking

of a cycad, club moss, lily, or beech tree, each has basically the same life cycle pattern: a dominant, diploid sporophyte capable of producing spores as a result of meiosis, with each spore germinating to produce a haploid gametophyte (or gametophytes) that produce two types of gametes—egg and sperm. The life cycle is complete when the two gametes fuse to initiate the diploid phase of the new sporophyte (Graham, 1985). Life History Biology

There remain a number of significant gaps in our understanding of the transitional steps required to move from a charophycean algal ancestor to a land plant. One of these involves significant differences in the life histories of these two groups of organisms (Nishiyama, 2007). In the haplobiontic life cycle of a charophycean alga, zygotic meiosis leads to the formation of haploid zoospores, each of which develops into a mature haploid organism. In this life history the only cell that is diploid is the zygote. This is in marked contrast to the life history displayed by vascular plants, in which the multicellular sporophyte, that is the diploid organism, is the dominant phase (Graham et al., 2000), or the life cycle of bryophytes, in which a multicellular sporophyte is produced which is dependent on the gametophyte (Chapter 5). Historically there have been two different theories on the evolution of the alternation of generations in land plants (reviewed in Blackwell, 2003; Haig, 2008). HOMOLOGOUS THEORY According to the homologous or transformation theory, both the gametophyte and sporophyte phases of land plant ancestors were morphologically identical (FIG. 6.34), that is they possessed an isomorphic alternation of generations. These two phases differed only in that the gametophyte was haploid and produced gametes, whereas the sporophyte was diploid and produced spores. The presence of extant green algae with isomorphic alternation of generations, and some bryophytes with photosynthetic sporophytes, has been used to support the homologous theory. The proponents of this idea suggest that, during the course of land plant evolution, in particular the evolution of vascular plants, the sporophyte phase eventually evolved as the dominant, nutritionally independent phase. ANTITHETIC THEORY The second idea, termed the antithetic or interpolation theory, begins with the premise that the gametophyte phase was primitive and that the sporophyte phase was later added to the life cycle as a result of a delay in zygotic meiosis (FIG. 6.35). This theory, which was championed by the eminent

Sperm Egg Zygote (2n) Gametangia

Spore producing unit

Isomorphic alteration of generations

Multicellular gametophyte (n)

Heterotrichous habit Sporangium

Multicellular sporophyte (2n)


Meiosis Heteromorphic alternation of generations (land plants)



Figure 6.34 Hypothesized stages in the origin of the alternation of generations according to the homologous theory. Beginning with a green alga with an isomorphic alternation of generations, the sporophyte becomes structurally and physiologically more complex, whereas the gamete producing phase becomes reduced. (From Taylor and Taylor, 1993.)

Heterotrichous habit

Sperm Delay meiosis

Egg Zygote (2n) Gametangia

Multicellular gametophyte (n)

Haplontic life history


(New) Multicellular sporophyte (2n)



Spores Heteromorphic alternation of generations (land plants)


Figure 6.35 Hypothesized stages in the origin of the alternation of generations according to the antithetic theory. Beginning with a

haplontic green alga with zygotic meiosis, a delay in meiosis results in the interpolation of a new multicellular sporophyte that becomes heterotrichous. Meiosis occurs and the spores give rise to a multicellular gametophyte. (From Taylor and Taylor, 1993.)


PALEOBOTANY: the biology and evolution of fossil plants

Figure 6.36

Frederick Orpen Bower. (Courtesy A. C. Scott.)

pteridologist F. O. Bower (1908) (FIG. 6.36), views the sporophyte as gradually evolving away from a parasitic dependence on the gametophyte to become a physiologically independent, photosynthesizing organism. The fossil record provides no direct evidence as to which of these ideas is correct as it is not possible to know the sequence of events that took place after the spores of early land plants germinated. Both ideas do, however, allow for subsequent patterns of plant evolution regardless of which theory may be correct. Those supporting the antithetic theory suggest that the diploid phase was initially better adapted to life on land than the haploid phase (Keddy, 1981). As a result the diploid phase continued to evolve adaptations that contributed to increasing organ differentiation, whereas the haploid phase remained small. Proponents of the homologous theory argue that, although both phases were initially identical, environmental influences provided selective pressures, so that each phase continued to evolve different levels of specialization and reproductive strategy. Why is the diploid phase better adapted to the terrestrial environment? One answer is that water is necessary for fertilization in the most primitive land plants. Sperm cells of primitive extant plants are motile (perhaps a retention of an algal characteristic) and require the presence of water to reach the egg cell. Water may be in the form of a film of dew, a swampy environment, or splashing raindrops. Whatever the source, water is essential to fertilization. Once the new

sporophyte (diploid plant) has developed, spores are produced, released, and carried away by air currents. Even though only a small proportion of the diploid plant body is used for spore production, the absolute number of spores produced can be quite large. Spores may be distributed randomly over wide areas, with each capable of producing a haploid gametophyte. In general, the larger the number of spores produced, the greater the probability that some will settle in places suitable for germination and the eventual production of gametes. It should be pointed out, however, that in some extant plants not all spores produced are viable, and this no doubt happened in the fossil record as well. Our understanding of early land plants is based primarily on features of the sporophyte and dispersed spores. The discovery of gametophytes from the Lower Devonian Rhynie chert has provided important details about the gametophyte phase of some early land plants and these will be discussed in Chapter 8 (Remy, 1982). In these plants the gametophytes are all free-living, autotrophic organisms that differ morphologically from the sporophyte. It is not known, for example, whether any of the early land plants possessed a life history in which the gametophyte phase existed in an aquatic environment, whereas the sporophyte occupied a terrestrial habitat. Perhaps a separation of the two life history phases in the early evolution of land plants is one reason why the sporophyte has become the dominant phase in vascular plants today (but see Bennici, 2005). Animals

The earliest evidence of terrestrial animals comes from Cambrian–Ordovician trace fossils (MacNaughton et al., 2002), but the oldest body fossil to date is from the Early Silurian (Llandovery; Wilson and Anderson, 2004). Age estimates based on molecular clock assumptions, however, suggest a much earlier occurrence date (Pisani et al., 2004). As is the case for plant fossils, there appears to have been a radiation of terrestrial animals in the Late Silurian–Early Devonian when the fossil record of terrestrial animals becomes more abundant and diverse. It is interesting that the arthropods from the Siluro–Devonian (e.g., millipedes, centipedes, arachnids), including those known from the famous Gilboa, New York site (Shear et al., 1984; see Chapter 23), are all believed to have been predators (Shear and Selden, 2001). This underscores the probability that a complex terrestrial ecosystem existed prior to the Late Silurian (Jeram et al., 1990). A Fungal Partner

Another important component of the successful colonization and exploitation of the terrestrial realm by plants may


have involved mutualistic associations with certain fungi (Pirozynski and Malloch, 1975). These symbioses are ubiquitous today (Chapter 3) and may have provided early land plants with an increased ability to obtain nutrients as a result of the extensive hyphal network of the fungus. In exchange for an increased ability to scavenge nutrients, the fungus gained access to a stable source of carbon. The fact that several Early Devonian land plants display well-established endomycorrhizae in both the sporophyte and gametophyte phases (T. Taylor et al., 1995, 2005c) adds credibility to this scenario. To further test this hypothesis we need to either find additional structurally preserved land plants with mycorrhizal fungi in their tissues, or develop techniques to detect the presence of specific mycorrhizal fungal biomarkers in compression fossils of early land plants.

The move to the land


Conclusion The scenario presented earlier hypothesizes that land plants evolved from a green algal ancestor(s) that became adapted to the desiccating environment on land through various physiological, structural, and functional modifications. Some of the transitional phases necessary for the successful colonization of the terrestrial realm are currently impossible to evaluate from the fossil record. Others, however, can be discussed and assessed based on the available fossil evidence. Some of these examples are discussed in Chapter 8.

This page intentionally left blank

7 INTRODUCTION TO VASCULAR PLANT MORPHOLOGY AND ANATOMY PLANT ORGANOGRAPHY .......................................................202

Epidermis ..........................................................................................208

CELL TYPES .................................................................................... 203

ANATOMY OF STEMS AND ROOTS ..................................... 210

Parenchyma .......................................................................................203

Arrangement of Primary Tissues ......................................................210

Collenchyma .....................................................................................203

Primary Xylem Maturation Patterns .................................................212

Sclerenchyma ....................................................................................203

Secondary Development ...................................................................212

Sieve Elements ..................................................................................206

Stele Types ........................................................................................216 LEAF MORPHOLOGY AND ANATOMY ............................... 221

PLANT TISSUES AND PRIMARY GROWTH..................... 207

Leaf Anatomy ....................................................................................221

Xylem Tissue.....................................................................................207

Leaf Evolution...................................................................................222

Phloem Tissue ...................................................................................207

FURTHER READING ...................................................................222


After all, I guess it doesn’t matter whether you look down [through a microscope] or up [through a telescope] — as long as you look. John Steinbeck, Sweet Thursday It is the presence of green vegetation on the surface of the Earth that makes it a pleasant and interesting place to live. Frequently, we take this green mantle for granted, forgetting that for most of earth history the landscape was barren. Cyanobacteria, algae, and algal-like organisms must have lived in terrestrial habitats before true land plants evolved, but surely they did not have the same effect on the appearance of the Earth as true land plants do. We now know that the early land flora included both vascular and thalloid forms (Chapter 6), as does the terrestrial flora today. This chapter will be restricted, however, to a discussion of land plants with vascular tissues. In some systems of classification, vascular plants are placed in a formal division, the Tracheophyta, and molecular data support monophyly for the vascular plants (Nickrent et al., 2000). In this book, the groups of vascular plants are elevated to phylum level (see Chapter 1), in part

reflecting the fossil record of several of these groups, especially the early evolution of vascular plants (Chapter 8). One of the perplexing problems in the history of plant life has been the long interval between the appearance of green, photosynthetic organisms and the evolution of vascular land plants. There is compelling evidence that autotrophic organisms existed at least 2.5 billion years ago (see Chapter 2). There is fossil evidence that plants with conducting systems which functioned like modern vascular plants existed in the Late Silurian, a little more than 400 million years ago. There is also fossil evidence that land organisms were around earlier, perhaps during the Ordovician, based on the spore and microfossil record. As you have seen in Chapter 6, some of these, no doubt, represented relatively short-lived attempts to colonize the land surface. We now know that some of the plants that have traditionally been regarded as



PaleOBotany: the biology and evolution of fossil plants

Figure 7.2 Longitudinal section of Coleus stem showing apex and axillary buds (arrows) (Extant). Bar  600 μm.

Figure 7.1 Cross section of Aglaophyton major stem showing

conducting strand (Devonian). Bar  500 μm. (Courtesy H. Kerp.)

the earliest vascular land plants, for example, Cooksonia and Aglaophyton (FIG. 7.1), have conducting elements that are different from those of true vascular plants and which may have had a different evolutionary origin (see Chapter 8). Since vascular plants contain rigid, resistant conducting tissues and a waterproofing cuticle, they are much more frequently found fossilized than algae, fungi, and bryophytes. This chapter is a brief review of vascular plant morphology and anatomy for those who have not studied these subjects previously, including a brief outline of the principal cell types, tissue systems, and structures found in vascular plants. It is a general introduction to the subject and does not take into account the many exceptions to the definitions and principles listed in the following sections. For more detailed information on morphology or anatomy, we have listed some additional sources at the end of this chapter.

PLANT ORGANOGRAPHY The vascular plant sporophyte consists of a shoot axis stem, roots, and laterals. The primary shoot (stem and leaves) and the root arise in the embryo and are responsible for elongation

growth, that is, vertical growth, in the plant. Laterals (leaves and branches) are borne at nodes on the stem separated by internodes (FIG. 7.2). Roots have no nodes or internodes. Although all extant plants bear laterals of some sort, some of the earliest land plants (Chapter 8) were leafless, photosynthetic axes. Branches are produced either at the apex, by division of the apex, or in seed plants, they can also be produced from axillary buds. See Chapter 8 for the different types of apical branching that occur in early vascular plants. Axillary buds develop at nodes in the angle (the axil) where the leaf meets the stem (FIG. 7.2). Most vascular cryptogams (vascular plants that do not reproduce with seeds) do not have axillary branching, although there are exceptions to this rule in the fossil record. Many plants also produce adventitious organs, the most common being adventitious roots. Adventitious organs are those that develop somewhere on the plant where they do not normally arise. Since a primary root usually arises from the base of the primary embryonic axis, roots are commonly termed adventitious if they arise anywhere else on the plant, for example, from the nodes of the stem and leaf (Barlow, 1994). The basic organs of plants thus consist of stems, roots, and leaves, although the stem and leaf may be treated together as the shoot. Reproductive structures, such as cones and flowers, represent combinations and modifications of these basic organ types, for example, cones or strobili consist of a central axis (stem) with helical or whorled laterals. The laterals are most commonly sporophylls—modified leaves that bear



sporangia. Floral parts in the angiosperms are also believed to represent evolutionarily modified leaves (see Chapter 22), and the flower itself consists of a shoot bearing floral parts. In some cases, the fossil record provides evidence of the evolution of these various modifications of plant organs, but in other cases, the evidence is less clear. Most plant anatomy texts concentrate on the anatomy of the flowering plants. Details that differ from angiosperms in various groups will be covered in those particular chapters, for example, ferns (Chapter 11), and seed plants (Chapters 14–22). In addition, details on seed anatomy and evolution are included in Chapter 13, so they will not be repeated here.



Plant organs are made up of cells. The most basic cell type, which makes up the ground tissue in plants, is the parenchyma cell (FIG. 7.3). Although all tissue types contain parenchyma, certain tissues are predominantly parenchyma, including the cortex and pith in stems and roots, and the mesophyll in leaves. Parenchyma cells are alive at maturity, have primary walls that are relatively thin, and can vary in their shape, from elaborately branched to almost isodiametric. Because they contain the full complement of cellular organelles, parenchyma cells have the potential to become meristematic and are totipotent, that is, they contain all the genetic material to develop an entire plant. They are a general-purpose cell and function in photosynthesis, so they may contain chloroplasts, and in storage of water, photosynthates (reserve foods), and many other compounds. Because of their thin walls (FIG. 7.3) and usually high water content, parenchymatous tissue is not generally an important structural component of plants, except in some of the earliest land plants (see Chapter 8). Parenchyma cells are less commonly preserved in fossils than some other cell types, especially sclerenchyma. COLLENCHYMA

Collenchyma cells are similar to parenchyma in which they are alive at maturity and can be isodiametric in shape. They have irregularly thickened primary cell walls (FIG. 7.3) and function as support in plants either during elongation growth or in plants without much secondary growth. They are most often found just beneath the epidermis in some stems and often contain chloroplasts. SCLERENCHYMA

Sclerenchyma cells are characterized by relatively thick, lignified secondary cell walls. All plant cells initially have only

Figure 7.3 Cross section of Apium sp. petiole showing paren-

chyma (arrow), epidermal, and collenchyma cells (C). Bar 100 μm.

Figure 7.4 Astrosclereid (arrow) in Castalia sp. leaf (Extant).

Bar  150 μm.

a primary wall made predominantly of cellulose. As sclerenchyma develops, a secondary wall with a high proportion of lignin is deposited inside the primary wall. The protoplast usually dies during development so that the typical sclerenchyma cell is dead at maturity. The lignified wall gives sclerenchyma cells their rigidity, and they function primarily in mechanical support and water conduction. They also make up most rigid parts of the plant (e.g., seed coats and some fruit walls) and are often positioned so that they provide mechanical protection for softer plant parts. There are three basic types of sclerenchyma cells: sclereids (FIG. 7.4), fibers, and tracheary elements, although there are intergradations


PaleOBotany: the biology and evolution of fossil plants




X Figure 7.5 Cluster of brachysclereids (arrows) in Pyrus sp.

fruit (Extant). Bar  250 μm.

among these types. Sclereids and fibers function solely in support, whereas tracheary elements function both in support and water conduction. Sclereids are variously shaped, from isodiametric to elongate and branched. They are characterized by a very thick wall with simple pits, that is, there is no special ornamentation associated with the pits. Generally, sclereids are shorter than fibers. Fibers have very thick secondary walls like sclereids (FIG. 7.5) but are elongated, spindle-shaped cells with long, narrow cell lumens; the lignified wall generally contains simple pits with slit-like apertures. They form supportive structures in tissues after the elongation growth has ceased and occur in many plant parts; many fibers, such as flax and jute, are of economic importance. Fibers commonly grow by intrusive growth, that is, the cells elongate and grow between other cells until they reach their mature length. They are common in both the primary and secondary xylem and phloem, especially in woody angiosperms, and can make up a considerable proportion of some of these tissues. Conifers have no fibers in their xylem but may have some in the secondary phloem. In some angiosperms, fibers are commonly found as a “cap” (FIG. 7.6) on the outer surface of the vascular bundle or as a band near the periphery of the stem. Fibers are classified according to their position in the plant into xylary (fibers in the xylem) and extraxylary (fibers elsewhere) fibers, for example, in the phloem or cortex. TRACHEARY ELEMENTS The term tracheary elements includes the two basic types of water-conducting cells in the xylem of vascular plants: tracheids and vessel elements. Tracheids differ from vessel elements; vessel elements have perforated end walls, whereas

Figure 7.6 Cross section of Helianthus sp. stem vascular bundle showing prominent bundle cap (B), phloem (P), cambium (C), and xylem (X) (Extant). Bar  250 μm.

tracheids have primary wall material present on their end walls; both have lignified secondary walls and both can occur in primary and secondary xylem. Both cell types also may have various types of secondary wall patterns on their side walls. Vessel elements are more efficient at water conduction, as there is no barrier to water movement from cell to cell vertically, whereas water must diffuse through primary wall at the end of each tracheid. Although tracheids are often narrower and more elongated than vessel elements, this is not always the case. TRACHEIDS. The tracheid is the basic cell in the xylem, that is, all plants have tracheids, but not the more highly evolved vessel elements. Tracheids are generally spindle shaped, very elongate, and have tapered ends. Tracheids have a dual function of support and water conduction, whereas vessel elements, except perhaps for some primitive types, function in conduction only. Both cell types are readily preserved in fossils and are easily recognized by their secondary wall thickenings. During development, the secondary wall is deposited in various patterns on top of (inside) the continuous primary wall, including rings (annular FIG. 7.14A), helical bands (FIG. 7.14B), ladderlike transverse bars (scalariform) (FIG. 7.7), or continuous except for pits. Pitted tracheids and vessels may have simple pits (no border) or pits that are surrounded by a thickened rim of wall



Figure 7.7 Tracheid of Pteridium sp. showing scalariform

secondary wall thickenings (Extant). Bar  40 μm.

material—bordered pits (FIG. 7.8). The border is a dome-shaped structure, made of secondary wall, that surrounds and arches over the opening in the secondary wall, that is, the pit. There is an opening in the center of the “dome,” the aperture (FIG. 7.9). It is important to remember that the pit itself is not a “hole” in a tracheid or vessel wall, it is merely an area where there is only primary wall and middle lamella (the area between two cells) present. This area is called the pit membrane. Neighboring tracheary elements often develop bordered pits, a pit pair, at the same location on adjacent cell walls. In conifers and a few angiosperms, pit pairs can be very elaborate. In the center of the pit membrane is a thickened area, the torus, which is slightly larger than the pit aperture. Around the torus some of the primary wall and middle lamella are partially dissolved, so this area, the margo, is thinner and very porous, consisting only of strands of cellulose microfibrils. The permeable margo allows for more efficient water conduction in these pit pairs. The margo is somewhat flexible and under water stress, it can move to one side of the pit pair and seal it off. When small bubbles of air form within the water column in the xylem, they can restrict water flow (embolism), or in some cases, the water column can suddenly collapse (cavitation). Severe cavitation can cause the collapse of the tracheary element. Both drought and the freeze-thaw cycle can precipitate cavitation in the xylem.

Figure 7.8 Pinus sp. secondary xylem tracheids with circular

bordered pits (Extant). Bar  50 μm.

Figure 7.9 Circular bordered pits on secondary xylem tracheids of Sequoiadendron giganteum. Note uniseriate and biseriate pattern (Extant). Bar  100 μm.


PaleOBotany: the biology and evolution of fossil plants

can vary from simple perforation plates, with a single large hole, to those with scalariform or reticulate openings. The gnetophytes (Chapter 19) have a unique type—foraminate perforation plates. The most specialized vessel elements are relatively short with horizontal end walls and simple perforation plates, whereas some woods contain vessel members that are more elongate, with oblique end walls (Bailey and Tupper, 1918). Tracheids and vessels occur in both primary and secondary xylem tissue. Like tracheids, vessels also exhibit various secondary wall thickenings on their side walls which range from annular to pitted. Although vessel elements have sometimes been thought to be a synapomorphy of the flowering plants, they apparently arose independently in a number of plant groups (Bailey, 1944). The anatomical evidence for independent origin has been well illustrated in the detailed studies of Schneider and Carlquist in homosporous (Schneider and Carlquist, 1998) and heterosporous (Schneider and Carlquist, 2000c) ferns, and in certain lycopsids (Schneider and Carlquist, 2000a,b). Vessel elements are also known in several gymnospermous groups, including the gnetophytes (Carlquist, 1996; see Chapter 19) and the enigmatic fossil group, the gigantopterids (Li et al., 1996; see Chapter 19). Figure 7.10 Cross section of Cucurbita sp. vascular bun-

dle showing larger diameter vessel elements (center), many with tyloses in them (Extant). Bar  650 μm.

Tracheids and vessel elements with annular or helical thickenings are extensible, so they are most often found in the earliest matured primary xylem (protoxylem, see below), since they can stretch somewhat as the axis continues to elongate. Scalariform thickenings are variable; depending on the amount of wall material deposited, it may grade between helical and pitted. In a tracheary element with pits, most of the primary wall is covered by secondary wall, so these elements are not extensible and occur in primary xylem that has matured after the axis has ceased elongation growth; this part of the primary xylem is called metaxylem. Secondary xylem is made up predominantly of pitted tracheids, although some plant groups also have scalariform secondary xylem tracheids. VESSEL ELEMENTS Vessel elements represent a more specialized type of tracheary element (FIG. 7.10). They are usually shorter than tracheids and have perforated end walls called perforation plates. Individual vessel elements are connected end to end in vertical rows to form vessels; each vessel is a continuous tube with little or no obstruction to water flow, depending on the size and type of perforation plates in the end walls. These

Sieve Elements

Sieve elements are thin-walled cells that are alive at maturity, although the protoplast is greatly changed, and they generally lack nuclei. Sieve elements are elongated and function as the basic photosynthate-conducting cell type in the phloem of vascular plants. The walls of sieve elements contain sieve areas, circular-to-elliptical parts of the wall that are thinner. Each sieve area (FIG. 7.11) includes a number of sieve pores, which allow for transport from one sieve element to the next (FIG. 7.12). Sieve pores are not actual holes in the wall as perforation plates are. Rather, they are protoplasmic connections between two living cells and are lined with a plasma membrane. In addition to the basic sieve element, there are two more specialized types of sieve elements: sieve cells, which occur in conifers, and sieve tube elements, which are a synapomorphy for the angiosperms. Sieve cells are generally long, narrow, and tapered at the ends, whereas sieve tube members are shorter and wider with more horizontal end walls. Sieve tube elements, like vessel elements, are connected end to end in vertical rows to form sieve tubes. Sieve plates occur on the end walls of sieve tube elements; these are groups of sieve areas, usually with larger pores than those on the lateral walls of the cell. Since sieve elements are under great hydrostatic pressure while functioning, they often collapse after death. Thus, preservation of phloem tissue in fossils is relatively rare.




Figure 7.13 Transverse section of Lepidodendron stem showing extensive periderm (P) (Pennsylvanian). Bar  1 mm.

Figure 7.11 Sieve areas (arrows) on phloem cell of fos-

sil Cycadeoidea sp. stem (Cretaceous). Bar  10 μm. (Courtesy P. Ryberg.)

contain one cell type and complex tissues, more than one type. Under the system of Sachs (1875), there are three tissue systems: the dermal, vascular, and ground (or fundamental) tissues. Dermal tissue is the outer covering of the plant, consisting of epidermal cells and cuticle in young plants, and periderm (FIG. 7.13), in plants that produce extensive secondary growth. Vascular tissue is the conducting tissue in the plant and consists of xylem (water conduction) and phloem (food conduction) tissue. Ground tissue is basically everything else in the plant. It can include simple tissues like parenchyma or complex tissues that include parenchyma and other cell types, for example, sclerenchyma. XYLEM TISSUE

Figure 7.12 Sieve areas (arrows) on phloem cells of Sequoia sp. (Extant). Bar  100 μm.

PLANT TISSUES AND PRIMARY GROWTH A plant tissue is a group of cells having a similar origin, structure, or major function. Plant tissues contain a characteristic complement of one or more types of cells; simple tissues

Of all the components in a typical plant axis, xylem elements are found most frequently in the fossil record. Xylem is a complex tissue made up of several types of cells. As noted earlier, all xylem tissue contains tracheids and some plants have both tracheids and vessels in their xylem. All xylem also contains living cells, xylem parenchyma, and sometimes fibers (xylary fibers). PHLOEM TISSUE

In addition to conducting cells, phloem tissue also contains unspecialized parenchyma, the phloem parenchyma, and often strengthening cells such as phloem fibers. When sieve tube members are present, they are always accompanied by companion cells (FIG. 7.15), specialized parenchyma cells that develop from the same procambial initial (mother cell) as the associated sieve tube member.


PaleOBotany: the biology and evolution of fossil plants

sieve tube member. The corresponding cell in gymnosperm phloem is called an albuminous cell, but it does not develop from the same initial cell as does the sieve cell. MERISTEMS

Sieve elements (S) and companion cells (C) in Cucurbita sp. Note conspicuous sieve plate (Extant). Bar  50 μm.

Although all parenchyma cells in a plant have the potential for growth and production of new cells, unless a plant has been wounded, cell production normally occurs in meristematic tissue, which consists of parenchyma cells that remain capable of dividing and producing daughter cells throughout the life of the plant. Because of these meristems, vegetative growth in plants is indeterminate, which means that the plant body is not fixed in its development but is potentially capable of continuous growth. There are two basic types of meristems, apical meristems and lateral ones, and one more specialized type, an intercalary meristem. Apical meristems are responsible for growth in length (height), or primary growth, and lateral meristems or cambia (sing. cambium) for growth in diameter (width) (see section “Secondary Development”). As the name implies, apical meristems occur at the apex of every stem and twig and the apex of every root in a plant. The cells that make up the apical meristem undergo repeated divisions to produce daughter cells; a short distance back from the growing tip, the daughter cells begin to differentiate into xylem, phloem, sclerenchyma, etc. In stems, the apical meristem produces cells and tissues that form the stem, leaves, and axillary buds, or lateral branches. In roots, the apical meristem produces only the cells that make up the root; laterals are produced further back from the meristem itself. Vascular tissue first appears as procambial strands, which will develop into mature xylem and phloem. The first xylem to mature is termed protoxylem and, as noted earlier, usually consists of extensible tracheary elements with helical or annular secondary wall thickenings (FIG. 7.14). Although protoxylem cells are usually smaller in diameter than metaxylem cells, the nature of the wall thickenings is a more exacting way to distinguish these cells in fossil plants. Further back from the meristem, after the stem has for the most part ceased to elongate, the metaxylem matures. It generally consists of non-extensible tracheary elements, such as pitted or scalariform (FIG. 7.7) tracheids. Both proto- and metaxylem are primary xylem, that is, they are produced by the apical meristem. Similar terms are used for phloem cells, that is, protophloem and metaphloem.

At maturity companion cells exhibit dense cytoplasm and have numerous protoplasmic connections with their associated sieve tube member. Physiologically, companion cells are thought to control movement of solutes into and out of the

The epidermis is the layer of cells on the outside of all primary parts of plants (FIGS. 7.3, 7.16); it functions in protection against water loss and infection, gas exchange with the

Figure 7.14 Longitudinal section of Zea mays xylem showing

annular (A) and helical thickenings (B) (Extant). Bar  60 μm.


S C e S



Figure 7.15






Figure 7.16 Section of Podocarpus urbanii stem showing thick

cuticle (arrows), hypodermal sclereids (S), and stomatal chamber (C) (Extant). Bar  50 μm.

Figure 7.18 Abaxial epidermal cells of Kalanchoe sp. with stomata (Extant). Bar  350 μm.

Figure 7.17 Section of Yucca sp. leaf showing thick cuticle

(arrow) (Extant). Bar  50 μm.

atmosphere, and sometimes photosynthesis. If a plant does not exhibit secondary growth, the epidermis may be retained throughout the life of the plant. In plants with secondary growth, the periderm will take over these functions as development continues (see section “Secondary Growth”). The epidermis is most often one cell layer thick, although there are exceptions. Epidermal cells are primarily parenchymatous, as they are alive at maturity, and most are compactly arranged with very few or no intercellular spaces. CUTICLE Covering the epidermis is the cuticle, a non-cellular layer which covers all the aerial parts of plants (FIG. 7.16); epicuticular waxes occur on the outer surface of the cuticle. The cuticle impregnates the outer cell wall of the epidermal cells, may form flanges between epidermal cells, and is also deposited on the surface of the cell (FIG. 7.17). It consists of varying proportions of two lipid polymers, cutin and cutan. Cutan is resistant to decay and was thought to account for the widespread preservation of cuticle in the fossil record (Chapter 1), but Gupta et al. (2006) have recently questioned this assumption after finding little cutan in some fossil leaves.


Figure 7.19 Cross section of Cycas revoluta leaf showing papilla (P) and sunken guard cells (arrows) (Extant). Bar  50 μm.

They suggested that a combination of cutin, waxes, and lipids from the interior of the plant may combine during diagenesis to form the so-called cuticle found on many fossils. STOMATA The epidermis also contains stomata (sing. stoma), the openings in aerial plant parts that allow for exchange of O2, CO2, and water vapor during photosynthesis. The stomatal complex (FIG. 7.18) consists of the stoma itself (FIG. 7.19), which is the pore or opening in the epidermis, two guard cells, and subsidiary cells, if present. Subsidiary cells are epidermal cells that surround the guard cells and differ in function and often also in morphology from other epidermal cells. In the ontogeny of the stomatal complex, the guard cells and subsidiary cells may develop from the same initial, a pattern called mesogenous, or syndetocheilic. If the guard cells and subsidiary cells develop from two different initials, this pattern is called perigenous or haplocheilic. These two types,


PaleOBotany: the biology and evolution of fossil plants

glandular and secrete substances that are important in deterring herbivory. They are known throughout the fossil record and range from multicellular glandular types (Krings et al., 2003a) to simple one-celled forms. Trichome bases may be preserved on fossil leaf cuticles, even when the trichome itself is no longer present, and have been useful in paleoecological studies. The type and arrangement of trichomes can also provide systematic information.

ANATOMY OF STEMS AND ROOTS Figure 7.20 Glandular trichome on Coleus sp. epidermis

(Extant). Bar  50 μm.

along with other epidermal characteristics, have been used to classify fossil foliage in the cycadophytes (Chapter 16). The basic types of stomatal arrangements include: (1) anomocytic (irregular), in which subsidiary cells are lacking, that is, the cells around the guard cells are not distinguishable from other epidermal cells; (2) paracytic (parallel), in which elongated subsidiary cells flank the guard cells; (3) anisocytic (unequal), which has three subsidiary cells, with one substantially smaller than the other two; (4) actinocytic (ringlike), with a circle of subsidiary cells around the guard cells; and (5) diacytic (cross-celled), in which there are two subsidiary cells, and the wall that these two cells share is at right angles to the guard cells. There are many variations on these types and gradations between them; different types can even occur side by side on the same organ (Kothari and Shah, 1975). For further information on these and additional types, see Van Cotthem (1970), Leaf Architecture Working Group (1999), Beck (2005), Carpenter (2005), and Evert (2006). Stomata may be on level with the surface of the leaf or may be sunken, which is common in plants that live in arid environments. Guard cells are usually bean shaped and covered with cuticle. They allow for the controlled opening and closing of the stomatal pore through changes in turgor pressure within the cells. Guard cells are known to respond to various exogenous environmental stimuli, including light (especially UV-B radiation) and CO2 concentration. Certain plant hormones, especially abscisic acid, mediate the internal reactions and ionic changes that control stomatal opening and closing through guard cell water potentials. TRICHOMES Certain epidermal cells produce outgrowths (FIG. 7.20) which develop into trichomes (hairs). Trichomes are common, especially on leaves, and have a variety of functions, including additional protection against water loss. Some of them are


The arrangement of tissues in stems and roots is similar, although there are differences between monocots and dicots, as well as between seed plants and non-seed plants. These will be discussed later under section “Stele Types.” In gymnosperm and dicot stems, the center of the axis is occupied by parenchymatous pith tissue, which functions as a storage tissue and sometimes a water source. Surrounding the pith is a ring of vascular bundles, each containing xylem and phloem. The most common type of bundle is collateral, with xylem toward the pith and phloem immediately outside the xylem. Outside of the vascular tissue is the cortex, with the epidermis external to it. Both the pith and cortex are primarily parenchymatous, although sclereids, fibers, secretory cells or canals, and other specialized cells may occur in either tissue. In the stems of monocotyledons, the vascular bundles are not arranged in a ring but are scattered through the stem. In this case, it is impossible to distinguish pith and cortex, so the tissue in which the bundles are embedded is simply called ground tissue. The vascular tissue in stems and roots, together with associated parenchyma, including the pith if present, is termed the stele. In the roots of most plants, there is no pith, and the xylem occupies the center of the axis; monocot roots have a central pith with vascular tissue in a ring. The xylem is ribbed and the phloem occupies the area between the ribs. In cross section, these ribs resemble two, three, or four arms extending from the center of the root (FIG. 7.21). A fundamental difference between stem and root anatomy is the position of the phloem with respect to the xylem. In stems, the phloem is immediately outside the xylem, but in roots, its position alternates with that of the xylem, so that phloem and xylem are situated on different radii in a root, but the same radius in a stem. This concept is important in classifying fossil axes as roots or stems when there are no laterals attached. In roots and in stems of certain primitive plants, the peri-cycle is the tissue immediately outside the vascular tissue; it may be one or more layers thick. The pericycle is



Figure 7.21 Cross section of Ranunculus sp. root show-

ing tetrarch xylem strand with phloem between arms (arrows). Note starch grains (stained pink) in cortical cells (Extant). Bar  200 μm.


Figure 7.22 Longitudinal section of Lupinus sp. root showing

origin of lateral root from pericycle (P) (Extant). Bar  500 μm.

part of the stele, as it arises from the same group of cells in the apex that produce vascular tissue. The pericycle is an important tissue in roots, as this is the site where branch or secondary roots arise (FIG. 7.22). Cells in the pericycle become meristematic and differentiate into a lateral root, which, as it grows, pushes its way out through the cortex (FIG. 7.23) and epidermis to emerge on the outside of the

Figure 7.23 Cross section of branch root arising from pericycle of Salix sp. root (Extant). Bar  250 μm.

root. As noted earlier, the laterals produced by stems, that is, leaves and axillary buds, are produced by the apical meristem early in primary growth, but this is not the case in roots. Because leaves arise from the outer tissues of the stem, they are said to have an exogenous origin. Since lateral roots arise deep in the root tissue in the pericycle, they have an endogenous origin. In addition to stelar differences, the origin of laterals is another fundamental difference between roots and stems and is often an important anatomical character used to identify isolated axes found in the fossil record. Immediately outside the pericycle is the endodermis (FIG. 7.24), a single layer of cells lacking intercellular spaces that can be recognized by characteristic thickenings, Casparian strips, in its transverse and radial walls. The endodermis represents the innermost layer of the cortex and functions as a physiological barrier that regulates movement of solutes into and out of the vascular tissue in the stele. Casparian strips surround the sides, top, and bottom of each cell, but do not cover the inner and outer tangential walls; they are rich in suberin, making them hydrophobic. These strips effectively seal off the tiny canals and spaces within the walls, or apoplast, of the endodermis, and thus force diffusion of solutes through the plasma membrane and cell lumen of the endodermal cells, that is, materials must move through the symplast or the living parts of the cells, thus providing some physiological control over lateral movement into and out of the stele. In a young stem or root, there is a parenchymatous cortex surrounding the stele; in some stems this tissue may contain


PaleOBotany: the biology and evolution of fossil plants

Figure 7.24 Transverse section of Pinus leaf showing endodermis and thickened casparian strips (arrows) (Extant). Bar 75  m.

is situated toward the outside of the stem, with metaxylem toward the center, the maturation, or development, of the xylem is described as being centripetal (from the outside in) or exarch. The exarch condition is found in almost all roots and in many primitive plants, such as the lycopsids. In axes with a pith, when the protoxylem is next to the pith, that is, closest to the center of the axis, and the metaxylem develops outside of it, the maturation pattern is endarch (centrifugal development of xylem). This is the typical maturation pattern found in the primary xylem of most seed plants. In axes with a solid core of xylem, called a protostele (see section “Stele Types”), if the protoxylem occupies the center of the protostele, this type of maturation is called centrarch. Centrarch xylem maturation is relatively common in several groups of Devonian plants, for example, the Rhyniophyta and Trimerophytophyta (Chapter 8). In axes that have a pith, if the protoxylem occupies the center of the xylem and the metaxylem develops on both the outside and the inside of the protoxylem, this pattern is called mesarch (development is both centrifugal and centripetal). Mesarchy occurs in many ferns. These primary xylem configurations can be seen in anatomically preserved fossils and have been used when attempting to determine relationships among groups. While it is not possible to observe the actual sequence of development in fossils, it is often possible to infer the maturation pattern based on the relative cell size and secondary wall patterns of protoxylem and metaxylem elements. SECONDARY DEVELOPMENT

Figure 7.25 Cross section of Elodea sp. stem showing aeren-

chymatous cortex (Extant). Bar  650 μm.

large intercellular spaces (FIG. 7.25). The cortex is principally a storage tissue, although some of the outer cells in young stems may be photosynthetic. The periphery of the cortex may contain collenchyma in young axes, but sclerenchyma is more common in older stems and roots; epidermis covers the outside of the plant. If a plant does not have secondary growth, then maturation of the primary tissues completes growth. See section “Secondary Development.” PRIMARY XYLEM MATURATION PATTERNS

As mentioned earlier, protoxylem consists of those tracheary elements (usually smaller in diameter) that are the first to mature. Xylem maturation patterns describe the location of the protoxylem in relation to the metaxylem; these terms are used only for primary xylem. When the protoxylem

Secondary growth is responsible for increase in diameter in roots and stems; secondary tissues are produced by lateral meristems or cambia. There are two basic types of cambia, a vascular cambium, which produces vascular tissue, that is, secondary xylem and phloem, and a cork cambium or phellogen, which produces tissues to replace the epidermis (see the following sections). VASCULAR CAMBIUM The vascular cambium arises between the primary xylem and phloem of a young stem or root. Parenchymatous cells become meristematic and begin to produce secondary xylem or wood toward the inside of the cambium and secondary phloem toward the outside of the cambium. The cambium itself remains meristematic, except in some unusual cases, for example, in the Carboniferous arborescent lycopsids (Chapter 9) and may range from a single layer to several layers of meristematic cells (FIG. 7.26). If the primary xylem is a solid core, as in some fossils,



Figure 7.26 Cross section of Pinus sp. stem showing radial files of vascular cambium initials (C) (Extant). Bar  100 μm.

the cambium begins development as a complete cylinder (a ring, as seen in cross section) between the primary xylem and phloem. If the primary vascular tissue occurs in bundles, as is the case in woody dicots and gymnosperms, the cambium begins development within the bundle—the fascicular cambium. Then, parenchyma cells between the bundles become meristematic—the interfascicular cambium—and connect the fascicular cambia together so that the cambium eventually forms a complete ring around the axis, between the primary xylem and phloem. Cambial cells or initials divide primarily by periclinal divisions (parallel to the surface of the axis) on their inner and outer faces, producing files of cells along the radii of the axis. The presence of these orderly files is one way to distinguish secondary growth in fossil axes. Cambial initials must also divide anticlinally (perpendicular to the surface) to produce more cambial cells as the circumference of the axis continues to increase due to the production of secondary tissue. There are two types of initial cells in the vascular cambium. Fusiform initials are elongate cells that produce the conducting cells in both the secondary xylem and secondary phloem and the other cells in the axial system. Ray initials are shorter, generally rectangular cells, which give rise to cells in the ray system (see section “Secondary Xylem”). Generally, many more secondary xylem cells are produced than secondary phloem; indeed, in most living trees the bulk of the trunk represents secondary xylem or wood. The vascular cambium in roots arises in the same place as in stems, that is, between the primary xylem and phloem, but since the primary xylem in many roots is lobed or furrowed, the cambium initially also has this shape. As the root continues to develop, however, more secondary xylem is produced in the furrows so that the cambium eventually has a cylindrical shape, just as it does in stems. See section “Secondary Xylem” and “Phloem” (later) for the cell types produced by the vascular cambium.


CORK CAMBIUM (PHELLOGEN) As the vascular cambium continues to produce cells, the stem or root increases in diameter and the peripheral portion of the cortex and epidermis, which are not meristematic, would eventually be split apart. In older axes, therefore, periderm tissue performs the function of the primary epidermis, that is, to protect the plant from infection and desiccation. The periderm includes the phellogen or cork cambium, cork cells (phellem), and sometimes phelloderm. Like the vascular cambium, the cork cambium produces cells to the inside (phelloderm) and the outside (cork). Also like the vascular cambium, the production of cells is not equal on the two faces, but, in this case, more cells are usually produced on the outside (cork) than on the inside, with the exception of some members of the Lepidodendrales (Chapter 9), which produce more phelloderm than cork. Phelloderm cells are parenchymatous, but cork cells are non-living at maturity and their walls are impregnated with suberin; they thus prevent water loss and also provide a barrier to infection by fungi and bacteria. The cork cambium can arise close to the outside of the stem, that is, subepidermally, or deeper within the cortex or in the secondary phloem. It can even arise in the epidermis itself. The process of development is the same as for the vascular cambium which parenchyma cells become meristematic and produce files of cells by periclinal divisions of the cork cambial initial cells. The cork cambium also undergoes anticlinal divisions to increase in circumference. The cork cambium may initially arise in certain areas of the axis but eventually becomes continuous around the stem or root. As the stem continues to increase in diameter, the older (i.e., outermost) phellem ruptures and may be sloughed off the outside of the stem. Newer cork cambia then differentiate inward of the original cork cambium, initially within the primary cortex but later within the secondary phloem. It is the arrangement of these subsequent cork cambia and the amount of cork they produce that gives the outer bark, or rhytidome, of particular species its characteristic appearance. Smooth bark (e.g., in some species of Betula) forms where there is little cork produced, whereas rough, fissured bark (e.g., in Quercus) results from extensive cork production. The fossil aquatic angiosperm Decodon allenbyensis, from the Eocene of British Columbia, has a very complex rhytidome, and the same structure does not occur in living species of this genus (Little and Stockey, 2006). Roots of D. allenbyensis produce a lacunate phellem, with alternating elongate and isodiametric cells. Bark is a non-technical term that includes all the tissues outside the vascular cambium. If the axis is young, the bark may include, from the cambium outward, secondary phloem,


PaleOBotany: the biology and evolution of fossil plants

Figure 7.27 Cross section of Pinus sp. stem showing 3 years of

growth and resin canals in wood (arrows) (Extant). Bar  250 μm.

primary phloem, primary cortex, phelloderm (if present), cork cambium, and phellem (cork). In older trees, the bark may consist only of secondary phloem, cork cambium, and phellem. In roots, the cork cambium may also arise near the surface of the axis but most commonly arises in the pericycle. SECONDARY XYLEM Secondary xylem (wood) is a much more complex tissue than primary xylem and consists of a number of different cell types arranged in specific ways. Wood includes an axial system, which moves water and minerals up the stem, and a ray system, which runs horizontally through the stem, that is, in a radial direction. The axial system contains the vascular tissue, tracheary elements (tracheids and/or vessels), and axial parenchyma (vertical strands of parenchyma). In certain angiosperms (hardwoods), the axial system may also contain support cells such as fiber-tracheids or libriform fibers, a type of xylary fiber. Gymnosperms do not have fibers in their wood (although fiber-tracheids may be present) and are called softwoods by foresters. Some conifer wood contains resin ducts (FIG. 7.27) or canals in both the axial and ray system, that is, they are oriented both vertically and horizontally. Resin ducts form by the separation of parenchyma cells during development; at maturity, they are hollow tubes which are lined with an epithelial layer, whose cells produce the resin. Resin ducts also form in many conifers as a response to wounding or infection by various pathogens. The ray system extends at right angles to the tracheary elements and is involved in conducting water and nutrients in a radial direction in the mature axis, as well as storage

Figure 7.28 Tangential (A) and radial (B) section of Pinus strobus wood showing vascular rays (arrows) (Extant). Bar  360 μm.

in the older secondary xylem. It consists of vascular rays, which are principally composed of parenchyma cells (homocellular rays). Some conifers have ray tracheids in their rays (heterocellular rays); these are shaped like parenchyma cells but have pitted walls and are non-living at maturity. Vascular rays in conifers are usually uniseriate or biseriate, that is, from one to two cells wide (FIG. 7.28), and can range from one to usually 20 cells high (Evert, 2006). In gymnosperms (e.g., Ephedra) and angiosperms rays range from uni- to multiseriate in width and from one to many cells high, up to several centimeters (FIG. 7.29) (Evert, 2006). Both ray and axial parenchyma cells in the wood may form tyloses where they border a tracheary element. The wall of the parenchyma cell extends through the pit cavity and balloons out into the lumen of the neighboring vessel or tracheid. Tyloses (FIGS. 7.10, 12.40) most often appear in xylem that is no longer functional and are thought to function as a means of sealing off tracheary elements, or perhaps as a host response to infection. Because of the complexity of secondary xylem, three different planes of sections are needed to fully characterize the anatomy of the wood and, in many cases, to classify isolated pieces of wood (e.g., fossils) to genus. A cross, or transverse, section is made at right angles to the axis of a stem or root; this is the section exposed when a tree is cut down. If the tree exhibits growth rings (tree rings), they will be visible in a cross section. A radial section is a longitudinal section which is cut along the radius of the axis. A tangential section is also a longitudinal section but is cut perpendicular to a radial section on a tangent to the surface of the stem.






Figure 7.29 Tangential section of Ephedra trifurca wood showing multiseriate rays (Extant). Bar  650 μm.

A cross section shows the tracheids, fibers, vessels and other cells of the axial system in cross section, and the cells appear as squares, rectangles, or polygons. The vascular ray cells in this section are in longitudinal section, since they are elongated along the radius of the axis. In a cross section, you can measure the width of tracheary elements, but not their length; conversely, it is possible to measure the length of ray cells, but one cannot determine their height. In a radial section, tracheids and fibers appear in longitudinal section as very elongate cells with tapered end walls. Secondary wall patterns can best be seen in a radial section, as pits are more common on the radial faces of the cells. Vessels, if present, are revealed as elongated series of cells, one above the other, and perforation plates in their end walls appear in side view. Vascular rays also appear in side view and the cells that make up the ray look like the face of a brick wall that is many bricks (or cells) long and many bricks (cells) high. From this view, we get no idea of the number of cells that make up the thickness of the ray (or, to continue the analogy, the thickness of the brick wall). It is also very difficult to determine the height of rays in a radial section, as the cut would need to be exactly through the middle of a ray to reveal the full height. In a tangential section, the axial system, that is, tracheids, fibers, vessels, and xylem parenchyma, all look more or less as they do in a radial section, except that bordered pits are not always seen. Vascular rays can be seen in cross section

Cross section of Pinus sp. early wood (E) and late wood (L) transition (Extant). Bar  150 μm.

Figure 7.30

in this view or, in other words, at right angles to the ray’s length. It is now possible to see the height and width (thickness) of the ray and the ray cells but not the length of the ray. To continue the brick wall analogy, in a tangential section you see the ray head on, like looking at the end of a brick wall, but you cannot determine its length. Growth rings or tree rings occur in many woody plants (FIG. 7.30). Those that grow in temperate zones usually, but not always, produce a single ring each year and this can be counted to determine the age of the tree. Many tropical trees also produce rings, however, and they often correspond to wet and dry seasons. In some areas of the temperate zone, trees can produce multiple rings per year due to seasonal precipitation. Tree rings are made up of earlywood and latewood, sometimes called spring wood and summer wood. In the spring, the apical meristems and young growing leaves of the plant produce the plant hormone, auxin, which is integral to the functioning of the vascular cambium. In the spring when the stems and roots are still elongating, auxin levels are higher, and the cambium produces earlywood—larger diameter tracheids with relatively thin walls. As elongation slows down and eventually stops, less auxin moves down the axis (or up in a root), and cambial production switches to producing latewood, which consists of smaller diameter


PaleOBotany: the biology and evolution of fossil plants

tracheids with thicker walls. When the tree goes dormant in the fall, either due to seasonal deciduousness (leaf drop) or to cold temperatures if it is evergreen, the cambial cells cease to divide. The following spring, the cambium begins producing earlywood again. The contrast between the latewood of the previous growing season and the earlywood of the present season is the ring boundary, which can be very sharp and visible to the naked eye. SECONDARY PHLOEM Secondary phloem, the tissue produced to the outside of the vascular cambium, is also a complex tissue that includes an axial and a ray system. Like the xylem, the axial system in secondary phloem includes conducting cells, either sieve cells in conifers or sieve tube members in the angiosperms, which conduct solutes from the sites of photosynthesis to other parts of the plant. Phloem parenchyma occurs in the axial system, as well as companion cells (angiosperms) and albuminous cells (conifers). Fibers are very common in the secondary phloem of both conifers and angiosperms (FIG. 7.31), and the pattern of fiber production by the cambium can sometimes be used to identify secondary phloem and bark tissue taxonomically. Although some conifers can produce regular, repeating bands of sieve cells, fibers, and parenchyma, they do not seem to produce these on an annual cycle, so it is not possible to determine the age of bark as it is to date wood by counting the tree rings. Usually only a narrow band of phloem close to the cambium is actively involved in conduction—the functional phloem or inner bark. As the older phloem becomes nonfunctional, there are many histological changes in the tissue, including the collapse of sieve elements, the development of sclereids from parenchyma cells, and/or the deposition of ergastic substances in parenchyma cells. These changes have also been identified in fossil phloem (Smoot, 1984c). It is in the nonfunctional phloem that subsequent cork cambia may arise in older axes. Vascular rays in the secondary phloem are continuous from the secondary xylem into the secondary phloem and consist only of parenchymatous ray cells. In some plants, the secondary phloem increases tangentially as the stem increases in diameter. This increase can occur by a tangential elongation of either axial or ray parenchyma cells. Some parenchyma cells, especially ray cells, may become meristematic and divide radially to produce additional cells. This process is called dilatation growth and can substantially increase the width of phloem rays. Secondary phloem rays are also important in ethylene signaling during plant responses to wounding and pathogens (Hudgins and Franceschi, 2004).



Figure 7.31 Cross section of Tilia sp. stem showing secondary xylem (X), phloem (P), and dilating vascular rays (V) (Extant). Bar  650 μm.


As noted earlier, the stele is defined as all tissues inside of, but not including, a distinct physiological barrier or boundary layer such as the endodermis (including the conducting tissue), after a concept called the stelar theory, which was initially developed by Van Tieghem and Douliot (1886a, b). The stelar theory is not used today by most botanists working with living plants as it is sometimes difficult to recognize the outer boundary of the stele as originally defined, and because stelar configuration can vary at different developmental stages of the plant or at different levels within a single axis. Nevertheless, the concept has been useful in comparative and phylogenetic studies of fossil vascular plants. In a general sense, stele types become progressively more complex in the fossil record and certain plant groups are characterized by particular types, so a knowledge of stele types is useful in paleobotany. For a treatment of stelar terminology and classification, see Schmid (1982) or Brebner (1902). PRIMITIVE VASCULAR PLANTS (VASCULAR CRYPTOGAMS) The simplest type of stele is a protostele, which consists of a solid core of xylem (no pith) in the center of the axis. Stems of many primitive plants and most roots are protostelic. There are three basic types of protostele: haplostele (FIG. 7.32), actinostele, and plectostele (FIG. 7.33). In a haplostele, the xylem is circular in cross section or cylindrical in three dimensions; phloem is immediately outside the xylem. An actinostele exhibits armlike projections of the xylem in cross section or ridges in three dimensions, with the phloem in


Figure 7.32 Cross section of Gleichenia sp. rhizome showing

haplostele (Extant). Bar  650 μm.

Figure 7.33 Cross section of Lycopodium serratum stem

showing actinostele with exarch xylem development (Extant). Bar  200 μm.

the furrows. Many roots are simple actinosteles, as they are usually diarch (two arms), triarch, or tetrarch. A plectostele exhibits many lobes in cross section, and it may appear as if the xylem is in separate plates (FIG. 7.34); phloem occurs between the plates. A three-dimensional view of a plectostele, however, indicates that the plates are interconnected. Both plectosteles and actinosteles occur in the extant Lycopodiales. A siphonostele occurs in vascular cryptogams that have a pith, with the xylem and phloem forming a continuous cylinder around the pith (FIG. 7.35). A solenostele, or amphiphloic siphonostele (FIG. 7.36), has phloem on both the outside and inside of the xylem (FIG. 7.37); an


Cross section of Lycopodium sp. stem showing plectostele (Extant). Bar  300 μm.

Figure 7.34

Figure 7.35 Cross section of Helmenthostachys siphonostele showing leaf trace (arrow) (Extant). Bar  780 μm.

ectophloic siphonostele has phloem only on the outside of the xylem. Both ectophloic and amphiphloic siphonosteles are found in the ferns. When leaves are produced, a leaf trace is given off from the stem stele and it supplies the leaf with xylem and phloem. In a series of cross sections through a fern stem in the region of leaf attachment, parenchyma cells appear in the siphono-stele at the point of leaf trace emission and there is continuity between the pith and cortex at that level. At higher levels, the leaf trace extends upward and outward into the base of the leaf. This interruption of the stelar cylinder is called a leaf gap, even though it is not actually a space as the name implies, but rather parenchymatous tissue. At higher


PaleOBotany: the biology and evolution of fossil plants

Figure 7.38 Cross section of Pteridium aquilinum rhizome showing dictyostele (Extant). Bar  2 mm.

Figure 7.36 Cross section of Marsilea quadrifolia rhizome showing amphiphloic siphonostele (Extant). Bar  300 μm.

Figure 7.39 Cross section of Osmunda sp. rhizome showing dissected stele and leaf traces (arrows) (Extant). Bar  2 mm.

Figure 7.37 Cross section of Psaronius vascular tissue showing preservation of phloem (arrows) (Pennsylvanian). Bar  1mm.

levels, the xylem and phloem at the edges of the gap appear closer together and eventually the interruption is no longer present. In some taxa, many leaf gaps are present in a single cross section and the stele appears to be dissected into

segments. Such a dissected siphonostele is called a dictyostele (FIG. 7.38). Dictyosteles are typical of many extant and fossil ferns (FIG. 7.39). Since the simplest and oldest stelar type is a protostele, it is hypothesized that plants with siphonosteles evolved from protostelic ancestors, and there are examples in the fossil record that support this hypothesis. Two principal theories have historically been advanced to explain the origin of the siphonostele from the protostele (Ogura, 1972). The intra-stelar origin theory (FIG. 7.40) suggests that, during the course of stelar evolution in some plant groups, cells in the center of the protostele did not mature into tracheids (FIG. 7.41B). The resulting medullated protostele would represent an intermediate







Figure 7.40 Diagrammatic stages in the evolution of the siphonostele according to the intrastelar theory: A. haplostele; B. medullated protostele; C. siphonostele with the beginning of a leaf trace; D. siphonostele with C-shaped leaf trace. (From Taylor and Talyor, 1993.)





Figure 7.41 Diagrammatic stages in the evolution of the siphonostele according to the extrastelar theory: A. haplostele; B. departure of leaf trace causing an interruption in the vascular cylinder; C. vascular cylinder closed at arrow; D. siphonostele with C-shaped vascular trace. (From Taylor and Taylor, 1993.)

stage between a protostele and a siphonostele, in which the central region contains both tracheids and parenchyma cells. According to this theory, the siphono-stele has evolved by the failure of certain procambial cells to develop into tracheids. Fossil evidence to support this theory can be found within the lycopsids, where protostelic forms occur early in the history of the group, followed by plants with medullated protosteles, and finally by those with siphonosteles (Chapter 9). The second hypothesis on the origin of the siphonostele, often termed the extrastelar theory (FIG. 7.41), views the siphonostele as evolving by the continued expansion of cortical parenchyma toward the stem center during the production of leaf traces from the surface of a protostele. In this scenario, cortical parenchyma became “trapped” as the xylem became continuous after trace departure (FIG. 7.41C). The production of a large number of leaf traces from a protostele would

eventually result in a stele in which the center contains parenchymatous pith. Some of the Paleozoic ferns best illustrate this pattern of stelar evolution, for example, the early protostelic botryopterid ferns and Grammatopteris, considered to be a progenitor of the osmundaceous ferns (Chapter 11). SEED PLANTS The primary vascular tissue in seed plants is arranged in a fundamentally different way from non-seed plants. Xylem and phloem occur in distinct strands called sympodial bundles or sympodial strands (vascular bundles), which are embedded in parenchymatous ground tissue. This stelar type is characteristic of seed plants and is called a eustele. The vascular strands are arranged either in a ring around the central pith, as in gymnosperms and dicotyledonous angiosperms, or scattered throughout the ground tissue (atactostele)


PaleOBotany: the biology and evolution of fossil plants

(FIG. 7.42), as in monocots. In a single cross section, a eustele may look like a dictyostele, in that the cylinder of vascular tissue appears dissected. The sympodial strands in seed plants, however, are discrete and continue throughout the stem (FIG. 7.43C, D). The ground tissue is also continuous from the pith to the cortex, that is, around the sympodial bundles. When a leaf trace is produced from a eustele, a stelar

Figure 7.42 Cross section of Zea stem atactostele showing

scattered vascular bundles (Extant). Bar  2 mm.



bundle divides in to two, with one part of the bundle separating tangentially to supply the leaf trace and the other part remaining as a sympodial strand (FIG. 7.43C, D) (Namboodiri and Beck, 1968a, b). Vascular bundles in seed plants are most commonly collateral, with primary xylem on the inside and primary phloem on the outside. Bicollateral bundles also occur; these have phloem both internal and external to the xylem. It was once believed that seed plants had their origin within the ferns (Jeffrey, 1917) and that the eustele evolved by continued dissection of a siphonostele, in part because of the similarity of stelar anatomy in cross section. We now know, based on many lines of evidence, including a better knowledge of early land plant lineages, that the ferns and seed plants had separate evolutionary histories, and the siphonostele and eustele did not have a shared evolutionary history. Progymnosperms and early members of the seed ferns provide evidence of the evolution of the eustele. Namboodiri and Beck (1968c) proposed that the eustele evolved through the continued longitudinal dissection of a protostele (FIG. 7.43B), that is, the intercalation of parenchyma into the vascular tissue. When leaf traces were produced, no gap was formed in the stele. Several Devonian taxa, including the progymnosperm Aneurophyton (Chapter 12), had ribbed protosteles and could serve as model starting points in this progression (FIG. 7.43A). Continued medullation of a ribbed protostele resulted in a three-stranded vascular system (FIG. 7.43B), similar to that seen in several species of Stenomyelon, and finally in a central pith with bundles around the periphery, a stele type seen in Archaeopteris. Finally, a



Figure 7.43 Suggested stages in the evolution of the gymnosperm eustele. A. Lobed protostele, for example, Stenomyelon primaevum (Chapter 14). B. Longitudinal dissection of protostele to form pith, for example, S. tuedianum. C. Continued dissection giving rise to discrete sympodial bundles. Trace formation via tangential division of sympodia, e.g., Calamopitys. D. Trace formation via radial division of sympodial bundles, resulting in the formation of a primary vascular system like that in most gymnosperms. (Redrawn from Namboodiri and Beck, 1986; in Taylor and Taylor, 1993.)


change in the production of leaf traces (FIG. 7.43D), so that sympodial bundles divided tangentially to produce traces, is illustrated by several Carboniferous seed ferns, such as Lyginopteris (Chapter 14). Some modern conifers (Chapter 21) have sympodial strands that undulate through the ground tissue.

LEAF MORPHOLOGY AND ANATOMY Leaves demonstrate the greatest morphological variability of any plant organ. They are also known for their plasticity, that is, difference in leaf form within a single species. Except in the earliest vascular plants, modern Equisetum, Psilotum, and Ephedra, and a few specialized angiosperms such as stemsucculent cacti, which have highly reduced leaves or are leafless, leaves function as the primary photosynthetic units in plants. Cotyledons, or seed leaves, represent the first leaves produced during embryonic development and they function as storage organs, providing food for the developing plant until the first true leaves appear and begin to photosynthesize. In many plants that live in arid environments, leaves may be fleshy and function in water storage. Modified leaves, which may be non-photosynthetic, are important as protective structures, for example, bud scales, and as parts of reproductive structures, such as flower petals and sepals or floral bracts, where they may be brightly colored and serve as signals for pollinators. Although leaf form is highly variable, most consist of a flattened blade borne on a narrow, elongate axis—the petiole. At the point of attachment to the stem (the axil of the leaf), many leaves have an abscission layer, which functions in separating the leaf from the stem and preventing water loss at the same time. Morphologically, leaves are classified as simple or compound; compound leaves are made up of leaflets. If the leaflets are attached at a single point, the leaf is palmately compound, for example, Sagenopteris, a leaf type found in the Mesozoic Caytoniales (Chapter 15). If the leaflets are attached along the petiole, the leaf is pinnately compound. This leaf morphology is very common in ferns, where multiple levels of leaflets may occur (Chapter 11). In a pinnately compound leaf, the term petiole on stipe used below the level of leaflet attachment. The region where the leaflets are attached is called the rachis (pl. rachides). The morphology of fossil angiosperm leaves, including overall shape and features of the margin, are widely used by paleobotanists to reconstruct paleoclimates using leaf physiognomy (see Chapter 1). Leaf venation is highly variable among vascular plants and has evolved differently in different groups of plants at various points in geologic time. The non-seed plants tend


to have venation patterns in which the vascular bundles dichotomize to fill the leaf, and veins often end at the leaf margin (open venation). Seed plants have more complex venation. If they exhibit dichotomous venation, the veins may also anastomose (join together) and then dichotomize again, forming enclosed meshes. This type of venation is seen in the Permian pteridosperm, Glossopteris (Chapter 14) and several other fossil seed plants. Veins often do not end at the leaf margin but bend back to fuse with other veins (closed venation). Some fossil gymnosperms exhibit relatively complex venation with multiple orders of veins (primary, secondary, tertiary, etc.), but the flowering plants have evolved the most diverse and complex pattern of venation. They include multiple, distinct orders of venation, and the ultimate veinlets enclose small patches of the leaf, termed areoles. Doyle and Hickey (1976) were able to trace the progressive changes in venation patterns in some of the earliest angiosperm leaf fossils and demonstrated a progressive increase in the organization of veins and the number of orders of veins (see Chapter 22) in successively younger rocks. Venation patterns in fossil leaves are an important systematic character, especially when coupled with leaf morphology and epidermal anatomy. LEAF ANATOMY

Leaves are composed of three principal tissue systems: epidermis, mesophyll, and vascular tissue, but the organization and extent of each of these systems are almost as variable as leaf morphology. Because a leaf is typically a dorsiventral structure, the epidermis of the abaxial surface is often different from that on the adaxial side. When stomata occur on both the adaxial (upper) and abaxial (lower) surface, the leaf is amphistomatic, on only the upper surface, epistomatic, and on only the lower surface, hypostomatic. The epidermis of leaves often bears trichomes that help to decrease water loss or glandular trichomes, which may help to discourage herbivores. The adaxial epidermis often has a thicker cuticle and fewer stomata than the abaxial side. On some leaves a hypodermis of thick-walled cells may be present immediately internal to the epidermis; this layer is believed to provide mechanical support for the leaf. The mesophyll tissue makes up the major part of the leaf and often consists of two types of cells. Palisade parenchyma consists of thin-walled, columnar cells with numerous chloroplasts (FIG. 7.44). These chlorenchyma cells are typically arranged in rows and are the principal photosynthesizing cells in most leaves. Beneath the palisade parenchyma (PP) is the spongy mesophyll (SM), which is characterized by thin-walled cells that are widely separated by lacunae or


PaleOBotany: the biology and evolution of fossil plants



may extend for several orders of branching (primary, secondary, etc.), depending upon the particular leaf type or plant group. The leaves of gymnosperms (FIG. 7.45) possess a slightly different tissue organization than that found in a typical angiosperm leaf. They often possess greater amounts of sclerenchyma and, in many instances, the vascular bundles are surrounded by a transfusion tissue composed of parenchyma and short tracheids. This tissue is believed to assist in conduction of materials between the vascular bundles and the mesophyll. Some leaves produced secondary vascular tissue, and stomata may be scattered on both surfaces. Many conifers, however, produce stomata in rows, termed stomatiferous bands. LEAF EVOLUTION

Figure 7.44 Section of Nymphaea leaf showing (palisade

parenchyma) and SM (spongy mesophyll). Arrow indicates a sclereid with simple pits on the wall (palisode parenchyma) (Extant). Bar  200 μm.

True leaves appear to have evolved at least twice in plant evolution. Microphylls are believed to have evolved from enations (Chapter 8) and represent a synapomorphy for the Lycophyta. They are generally, but not always, small, have a single vascular bundle, and a leaf trace that leaves no gap when it departs the stele. The evolution of microphylls is discussed in more detail in Chapter 9. Megaphylls evolved from branching systems generally have more complex vasculature and produce a leaf gap when the leaf traces departs the stele. Contrary to the name, they are not all large in size. All groups of vascular plants except the lycopsids possess megaphylls, and the evolution of this leaf type will be discussed in Chapter 11.


Figure 7.45 Cross section of Pinus sp. leaf with two vascular

bundles and resin canals (arrows) in cortex (Extant). Bar  350 μm.

intercellular spaces (FIG. 7.44). This tissue is ideally suited for the circulation of carbon dioxide and also provides a degree of flexibility to the leaf. Spongy mesophyll cells also contain chloroplasts, but they are generally not as densely packed as in palisade cells. The third component of a leaf is the vascular system or vein system. A transverse section of a leaf reveals many vascular bundles, each surrounded by a layer of cells termed a bundle sheath. Vascular bundles in a leaf are usually collateral, but bicollateral and concentric ones occur in some groups. When collateral, the xylem is located toward the abaxial surface with the phloem below it; secondary tissues may be present as well. The veins

Obviously, all facets of the morphology and anatomy of vascular plants cannot be covered in this brief summary. A comprehensive general botany textbook is a good source for more information about the structure and diversity of vascular plants (Mauseth, 2003; Raven et al., 2005; Graham et al., 2006; Stern, 2006). For additional details on plant morphology, see the classic three-volume textbook by Karl Goebel (1928–1933), or Bierhorst (1971) or Gifford and Foster (1989). There are a number of excellent plant anatomy texts, including Esau (1965, 1977), Fahn, (1990), Mauseth (1988), Dickison (2000), Beck (2005), Evert (2006), and Cutler et al. (2007), as well as those that concentrate on angiosperm anatomy, for example, Rudall (2007), or the two series of volumes, Anatomy of the Dicotyledons (Metcalfe and Chalk, 1979, 1983; Metcalfe, 1987) and Anatomy of the Monocotyledons (currently nine volumes), both series edited by anatomists at Kew Gardens.



Discussion: Rhyniophyte Evolution ................................................. 251

HISTORY OF DISCOVERY ...................................................... 225

ZOSTEROPHYLLOPHYTES...................................................... 252

RHYNIOPHYTES......................................................................... 227

Zosterophyll Evolution..................................................................... 259

Rhynie Chert Plants ..........................................................................228

TRIMEROPHYTES ........................................................................259

Gametophyte Generation ................................................................. 241

Trimerophyte Evolution ................................................................... 262

Other Rhyniophytes ......................................................................... 246

EARLY LAND PLANT EVOLUTION ..................................... 263

I would be met and meet you so, In a green airy space, not locked in. Denise Levertov, About Marriage Many paleobotanists regard the upper part of the Silurian as the point in geologic time when the first plants with organized conducting tissue appear. Others have suggested that vascular plants occur in strata as old as the Cambrian (Kryshtofovich, 1953). Some of these pre-Devonian fossils have subsequently been demonstrated to be the remains of nonvascular plants or even animals (Theron et al., 1990). In other instances, reinterpretation of the age of the rocks containing the fossils has negated reports of early vascular plants. The early lycopsid, Baragwanathia, is an unusual case in this regard. (Garratt, 1978) It was initially described from Upper Silurian compressions from Australia (Lang and Cookson, 1935) and the age was based on the occurrence of the graptolite Monograptus. Baragwanathia (FIG. 8.1) will be discussed in more detail in Chapter 9, but it is most certainly a vascular plant, with annular-helical tracheids forming the conducting strand. The age of this plant has been debated in the literature ever since as to whether the rocks are truly Upper Silurian or Lower Devonian. Subsequent studies have confirmed the Late Silurian (Ludlow) age (Rickards, 2000). The age has

been hotly debated (Thomas, 1984) because Baragwanathia represents a relatively complex vascular plant at a point in geologic time when all other vascular plants were comparatively simple. This suggests that either vascular plants evolved far earlier or that tracheids perhaps evolved more than once during the terrestrialization of the earth. As we learn more about early land plants, including those with well-defined conducting tissues, it is becoming clear that a number of these early plants did not possess vascular tissue like that in extant vascular plants. Although identifying plants with vascular tissue early in the geologic record is important, what is perhaps equally significant is understanding the nature of these cells.


Higher taxa in this chapter:

Rhyniophytes (Silurian–Devonian) Zosterophyllophytes (Silurian–Devonian) Trimerophytes (Silurian–Devonian)


Paleobotany: the biology and evolution of fossil plants

Figure 8.1 Baragwanathia longifolia (Silurian). Bar  2 cm.

Figure 8.2 Oblique view of Gosslingia breconensis conducting

elements (Devonian). Bar  10 μm. (From Edwards and Kenrick, 1988a.)

CONDUCTING ELEMENTS IN EARLY LAND PLANTS Kidston and Lang (1920a), in their detailed descriptions of plants from the Rhynie chert (see below), noted that they were unable to resolve secondary wall thickenings in the central strand of Rhynia major (now Aglaophyton) (FIG. 7.1). When David S. Edwards (1986) demonstrated that these conducting elements were structurally different from tracheids, he provided the first evidence that Aglaophyton was not a true vascular plant. Additional detailed studies of water-conducting elements in other early land plants (Kenrick and Crane, 1991) indicate a level of structural complexity in these elements, and demonstrate that they are quite different from the tracheids of vascular plants. Kenrick and Crane (1991) delimit three types of water-conducting cells in Early Devonian plants: (1) unornamented, elongate cells, found in the Rhynie chert taxa Aglaophyton major, Nothia aphylla, and the gametophytes Lyonophyton rhyniensis and Kidstonophyton discoides (Edwards, 2003). These are comparable in structure to the hydroids of mosses; (2) S-type, and (3) G-type conducting cells. Kenrick et al. (1991a) reported S-type water-conducting cells in the central strand of Sennicaulis, an Early Devonian rhyniophyte from Wales (Edwards, 1981). Each cell in the central strand has simple helical thickenings, but at the ultrastructural level the wall is unlike that of vascular plant tracheids. In S. hippocrepiformis, the cell wall consists of two principal layers. The outer layer makes up the helical thickenings, which possess a spongy internal organization. This is bounded on the lumen side by a thin microporate

layer. In addition to Sennicaulis, S-type conducting cells are characteristic of various rhyniophytes, including Huvenia kleui, Stockmansella langii, and S. remyi (Fairon-Demaret, 1985, 1986; Kenrick et al., 1991b; Schultka and Hass, 1997). G-type elements were first described in the Early Devonian zosterophyll Gosslingia breconensis (FIGS. 8.2, 8.3) (Kenrick and Edwards, 1988a), but also occur in other zosterophyllophytes, Hsüa (Li, 1992), Barinophyton (Brauer, 1980), and various early lycophytes (e.g., Asteroxylon, Drepanophycus; Chapter 9) (Berry and Fairon-Demaret, 2001). G-type cells have a two-layered wall. The inner, decay-resistant layer makes up a series of annular thickenings, but also forms a continuous wall area between the thickenings, where it contains small pits. The outer layer is non-resistant and may be mineralized. Following Kenrick and Crane’s (1991) delimitation of these original three types, three more forms of water-conducting cells have been recognized (Edwards, 2003) (FIG. 8.4): the P-, C-, and I-types. Species of Psilophyton are representative of the P-type. These cells exhibit scalariform bars, but the bars are narrowly attached to the cell wall, giving the appearance of scalariform bordered pitting. They have a decayresistant inner layer with decay-resistant material covering the pit apertures, forming strands or holes within the scalariform bars (Hartman and Banks, 1980). Structurally similar cells occur in later Paleozoic lycophytes, including the late Middle Devonian herbaceous form Minarodendron (Li, 1990). The C-type is known exclusively in Cooksonia pertonii (Edwards et al., 1992). These cells resemble conventional annular or



with the hypothesis that turgor pressure in parenchyma cells was the primary means of maintaining rigidity in these early land plants (Speck and Vogellehner, 1988). Although some authors call these cells “tracheids,” we have chosen to use the term conducting cells, since their homologies with either true tracheids or bryophyte hydroids remain uncertain. These cell types do, however, provide a functional perspective that can be used to test hypotheses about water conduction in early terrestrial ecosystems (Edwards et al., 1998b; Cook and Friedman, 1998; Friedman and Cook, 2000) (FIG. 8.5).


Figure 8.3 Diagrammatic section of G-type conducting ele-

ment of Gosslingia breconensis. (Courtesy P. Kenrick.)

spiral tracheids, except that the imperforate lateral walls are thicker than the primary wall of protoxylem elements (Edwards, 2003). Indeterminate stomatiferous axes illustrate the I-type. In this bilayered wall, the outer layer is imperforate and fused with that of adjacent cells, whereas the inner layer has rounded perforations. In section, the perforations are slightly wider toward the middle lamella and so superficially resemble bordered pits (Chapter 7) (Edwards and Axe, 2000; Edwards et al., 2003). Such internal wall thickenings may represent structural features of the cells which help to reduce embolism in early homoiohydric plants. This hypothesis is supported by the presence of G-type cells in the main axes of Sawdonia ornata from Röragen, Norway, but not in the laterals (Edwards et al., 2006b). The spongy organization of the wall may represent the template where lignin was deposited once this biochemical pathway evolved. It is not known whether the central strand cells in many early plants contained sufficient amounts of lignin to be useful in support. The small size of the central strand compared to the diameter of the stem, however, is more consistent

Our understanding of early vascular plants has an interesting history that, to a large degree, has greatly influenced many areas of paleobotany. In 1859, the Canadian geologist and paleobotanist Sir John William Dawson (FIG. 8.6) published a report on a Devonian vascular plant collected from the Gaspé region of Nova Scotia. His reconstruction showed a horizontal rhizome bearing upright, leafless, dichotomizing axes, to which were attached pairs of sporangia. Dawson named this interesting plant Psilophyton princeps (FIG. 8.7). Dawson’s scientific colleagues virtually ignored this important discovery, however, perhaps because the plant he reconstructed looked so unusual and certainly because of its age. Several years later (Dawson, 1871) he described additional specimens, but, again, these were not seriously considered by the scientific community of the day. In the years that followed, other discoveries were made on plants with obvious vascular tissue, and gradually Dawson’s initial report of Devonian vascular plants gained acceptance (Dawson, 1888). One of the most spectacular discoveries in paleobotany finally proved beyond any doubt that vascular plants existed by the Early Devonian. Beginning in 1917, Robert Kidston (FIG. 8.8) and William Lang published a series of papers detailing some exquisitely preserved vascular plants collected near the village of Rhynie, in Aberdeenshire, Scotland. This fossil-bearing rock (FIG. 8.9) consists of a fine-grained chert that is now regarded as coming from the upper part of the Lower Devonian, and dated at approximately 400 Ma (Rice et al., 1995). Recent palynological studies suggest a Pragian–?earliest Emsian age for the deposits (Wellman et al., 2006; and Wellman, 2007). Most of the fossils from the Rhynie locality showed that these plants did in fact consist of dichotomizing and, in general, leafless aerial stems (FIG. 8.10) arising from a horizontal aboveground or subterranean rhizomatous system. At the ends of some axes were terminal sporangia.


Paleobotany: the biology and evolution of fossil plants




















Famennian 367 Frasnian 377 Middle

Givetian 381 Eifelian 386 Emsian 390





Pragian (Siegenian) 396 Lochkovian (Gedinnian) 409


Pridoli 411




421 Wenlock 430 Llandovery 439



Ashgill 570

Caradoc Llandeilo/ Llanvirn Arenig



Known range Possible range Probable range but derivation different Probable range based on megafossils


Figure 8.4 Stratigraphic ranges of microfossils, megafossils and various anatomical features. Numbers across the top refer to (1) Dyads and tetrads of possible bryophytic origin. (2) Single spores with trilete marks. (3) Cooksonia megafossils. (4) Bifurcating axes of putative vascular plants. (5) Nematophyte cuticle. (6) Higher plan cuticle. (7) Stomata on axial fossils and lycophytes. (8) Banded tubes. (9) C-type conducting elements. (10) G-type conducting elements. (11) Zosterophylls. (12) Baragwanthia, Drepanophycus, lycopsids. (13) S-type conducting elements. (14) P-type conducting elements. (15) Trimerophytes. (From Edwards, 2003.)

Since the discovery and publications by Kidston and Lang (1917–1921), additional simple land plants have been described from Devonian and pre-Devonian rocks. For many years most of these early plants were placed in a single order, Psilophytales. These fossils, along with two living

plants, Psilotum and Tmesipteris, made up a separate subdivision of vascular plants, Psilopsida (or Psilophyta). Today this classification is rarely used. Some include Psilotum and Tmesipteris in their own division (Gifford and Foster, 1989), whereas others regard the two extant genera as having their


Resistant layer

Resistant layer

Spongy layer

Degradationprone layer

Primary cell wall

S-type conducting cells (Rhyniopsida)

Primary cell wall

G-type conducting cells (early Lycophytina)


Resistant layer

Resistant layer Degradationprone layer Primary cell wall

Primary cell wall

P-type conducting cells (early Euphyllophytina)

Seed plant tracheids (recent Euphyllophytina)

Figure 8.5 Longitudinal section of cell wall thickenings in fossil S-, G-, and P-type conducting elements and tracheids of extant seed plants showing primary cell wall components. (Modified from Cook and Friedman, 1998.)

Figure 8.6

John W. Dawson. (Courtesy H. N. Andrews.)

closest affinities with certain ferns (Bierhorst, 1968; 1971), a placement that is supported by molecular studies (Manhart, 1995). As additional Silurian–Devonian plants were discovered and carefully evaluated (Andrews et al., 1977), it became apparent that there were suites of characters that might be used to define larger taxonomic groups among the fossils (Høeg, 1954). Harlan Banks (1975) (FIG. 8.11) was the first to propose abandonment of the Psilophytales, which had become a repository for all types of unrelated early plants. In its

place he established three subdivisions—the Rhyniophytina, Zosterophyllophytina, and Trimerophytina. In the earliest cladistic analysis of early land plants (Kenrick and Crane, 1997a, b) the rhyniophytes and trimerophytes were not considered monophyletic, whereas the zosterophyllophytes were similar to the Zosterophyllophytina of Banks, with the inclusion of several other taxa. These fossils are now included in the polysporangiates, a clade of all land plants that bear multiple sporangia in the sporophyte phase, which includes both vascular plants and nonvascular plants (e.g., Aglaophyton). The Eutracheophytes contain all extant vascular plants and most vascular plant fossils, and are further subdivided into the Euphyllophytina and Lycophytina. The characters that are interpreted as plesiomorphic or derived (apomorphic) will be continually debated in the case of certain fossil plants. It is especially difficult to characterize basal groups, such as the earliest land plants, as by definition they will contain multiple plesiomorphies that are shared by all later-evolving land plants. With those limitations in mind, this chapter will discuss the Late Silurian–Early Devonian record of “vascular” land plants by reference to Banks’ original three groups.

RHYNIOPHYTES Some of the plants included in this group were previously included in the Rhyniophyta (Rhyniophytina of Banks, 1975). They can be characterized by dichotomously branched, naked aerial axes with terminal sporangia. The aerial axes arise from horizontal, dichotomizing rhizomes that bear rhizoids; no true roots are known. Sporangial shape


Paleobotany: the biology and evolution of fossil plants

Figure 8.8 Robert Kidston. (Courtesy University of Aberdeen.)

Figure 8.7 Suggested reconstruction of Psilophyton princeps.

Figure 8.9 Trench exposing Rhynie chert bed. Hagen Hass,

(From Taylor and Taylor, 1993.)

far right.

varies from ellipsoidal to branched, and some rhyniophyte sporangia appear to have an opening at the tip; others have been described with an abscission layer at the base of the sporangium. When axes are structurally preserved, they contain a small, terete, centrarch conducting strand of S-type conducting elements. Spores are all of the same morphological type, and hence the plants are considered homosporous. Cooksonia (FIG. 8.12), considered by many to represent the first vascular plant, is typical of this group (see section “Other Rhyniophytes”). We will begin, however, with the

plants from the famous Rhynie chert Lagerstätte, as they represent the best-known of the early land plants and have provided so much information about early land plants in their ecosystems, even though a number of them are now placed within the zosterophylls (i.e., Ventarura, Trichopherophyton, early lycophytes (Asteroxylon). RHYNIE CHERT PLANTS

There can be little doubt that the Rhynie chert organisms have had a profound influence on many of our hypotheses



Figure 8.10 Section of Rhynie chert showing closely spaced

upright axes and matrix (Devonian). Bar  600 μm.

Figure 8.12 Suggested reconstruction of Cooksonia caledonica.

(From Taylor and Taylor, 1993.)

lignieri, and Asteroxylon mackiei. Since the original description, additional land plants have been described from the chert lenses, including Nothia aphylla (Lyon, 1964; ElSaadawy and Lacey, 1979) and Trichopherophyton teuchansii (Lyon and Edwards, 1991), and from coeval deposits at the nearby Windyfield chert site, that is, Ventarura lyonii (Powell et al., 2000).

Figure 8.11

Harlan P. Banks. (Courtesy H. N. Andrews.)

about the early evolution of land plants. In recent years, however, some of these ideas have been challenged in light of new discoveries from the chert beds and the reexamination and reinterpretation of some of the plants. In their pioneering series of papers (1917–1921), Kidston and Lang described four plant taxa: Rhynia major, R. gwynne-vaughanii, Hornea

AGLAOPHYTON MAJOR The best known plant from the Rhynie chert is Aglaophyton major (FIG. 8.13), a macroplant originally described as Rhynia major, but transferred to a new genus by Edwards based on a reexamination of the original Kidston and Lang slides and the discovery of new specimens (D.S. Edwards, 1986). In a very real sense, A. major is the Arabidopsis of the Devonian, not because its genome is well known, but because the plant and its life history are known in such detail (FIG. 8.14). Aglaophyton major is now reconstructed as a plant 18.0 cm tall that consists of a system of naked, stomatiferous (FIG. 8.15), more or less cylindrical, and sinuous prostrate axes, which are loosely lying on the substrate surface and function as rhizomes. The prostrate axes dichotomize repeatedly, periodically turning upwards and passing


Paleobotany: the biology and evolution of fossil plants

Figure 8.13 Model of Aglaophyton major. (Courtesy N. Trewin.)

8.15 Stoma of Aglaophyton major (Devonian). Bar  25 μm. (Courtesy H. Kerp.)


Figure 8.14 Suggested life history of Aglaophyton major/

Lyonophyton rhyniensis showing stages in the development of the dimorphic gametophytes. Mature sporophyte (lower left, not to scale) bears sporangia with spores of two types. Blue spores (color is arbitrary) develop into antheridiophores; orange spores develop into archegoniophores (From Taylor et al., 2005c.)

into upright, fertile axes 6.0 mm in diameter that bear terminal sporangia (FIG. 8.16) with trilete spores (Wellman et al., 2006). There is some suggestion that the apex twisted during the development of the sporangia or that torsion of the sporangium was somehow involved in spore release. This sporangial feature is also found in the Early Devonian compression genus Tortilicaulis and has been compared with the morphology of certain bryophyte sporangia (Chapter 5).

The anatomy of the prostrate and upright axes is simple (FIG. 8.17); most of the axis consists of a parenchymatous cortex (FIG. 8.18), which is subdivided into inner and outer zones. The outer cortex is composed of densely packed elongated cells with narrow intercellular spaces. The cells of the inner cortex are more loosely spaced and exhibit a well-developed intercellular system (FIG. 8.19). The outer cortex is surrounded by hypodermal tissues and epidermis. Between the outer and inner cortex, there is a well-defined region of cortical tissue in which an endomycorrhizal fungus, Glomites rhyniensis, forms intracellular arbuscules (Chapter 3). The most significant component of Edwards’ (1986) work, however, and the reason that a new genus was created, involves the nature of the central conducting strand. As mentioned earlier, although Kidston and Lang (1920a) originally described A. major as a vascular plant, they were never able to discern secondary wall thickenings on the conducting cells. They ascribed the absence of thickenings to poor preservation, but David Edwards showed that the conducting strand of A. major was actually something quite different. He found that the central strand consisted of three regions. The outer zone, topographically in the position of the phloem, was made up of



Figure 8.18 Transverse section of Aglaophyton major axis showing central strand and dark cell zone (arrow) containing arbuscules (Devonian). Bar  650 μm.

Figure 8.16 Longitudinal section of Aglaophyton major sporangium filled with spores (Devonian). Bar  4 mm. (Courtesy H. Kerp.)

Figure 8.19 Transverse section of Aglaophyton major axis showing thin-walled cells of inner cortical zone surrounding central strand (Devonian). Bar  110 μm.

Figure 8.17 Cross section of chert block in FIG. 8.10 showing numerous axes (Aglaophyton major) in transverse section (Devonian). Bar  1 cm.

thin-walled, elongated cells with oblique, S-shaped end walls. The remainder of the central strand consisted of two zones of opaque cells that were initially thought to represent tracheids. The cells in the outer zone are circular in transverse section and 50 μm in diameter. In longitudinal section these cells exhibit a reticulate patterning on their walls, but this appears to be the result of cell wall degradation or silica crystallization. Cells in the center of the strand are uniformly thin walled and more angular in cross section. In longitudinal section, the conducting elements in A. major have S-shaped walls with partly helical


Paleobotany: the biology and evolution of fossil plants

Figure 8.20 Tetrads of spores (Devonian). Bar  65 μm.

Figure 8.21 Germinating spore of Aglaophyton major (Devonian). Bar  50 μm. (Courtesy H. Kerp.)

Figure 8.22 Germinating spore (Devonian). Bar  70 μm. (Courtesy H. Kerp.)

thickenings, consisting of a spongy outer layer and inner, delicate decay-resistant component. Edwards concluded that the cells of the conducting strand of Aglaophyton are more similar to the leptoids and hydroids found in certain bryophytes than to the sieve cells and tracheids of vascular plants. It is clear from this important work that this plant combines features of both bryophytes and vascular plants, and thus it is placed within an informal group of early land plants, the cooksonioids (rhyniophytoids), which includes dichotomizing axes for which there is no information about the presence or absence of a conducting strand. One of the most influential reports involving the Rhynie chert plants was the discovery of free-living gametophytes in the chert lenses (Remy and Remy, 1980a, b). Since the initial description of Lyonophyton rhyniensis, which is the gametophyte of A. major, additional details have been added, (Fig. 8.14) including a sequence of stages leading from spore germination (FIGS. 8.20–8.28) to the development of



Figure 8.23 Germinating Aglaophyton major spore showing initial division (arrow) (Devonian). Bar  30 μm. (Courtesy H. Kerp.)

Figure 8.25 Spore (arrow) germinating and giving rise to

multicellular gametophyte Lyonophyton rhyniensis (Devonian). Bar  20 μm. (Courtesy H. Kerp.)

Figure 8.24 Germinating spore of A. major showing stages of

cell division (Devonian). Bar  50 μm. (Courtesy H. Kerp.)

mature, unisexual antheridiophores—gametophytes that bear antheridia (FIG. 8.29) and archegoniophores—gametophytes that bear archegonia (FIG. 8.30) (Remy and Hass, 1996). Clusters of spores or spore balls are typically found in the

chert matrix, suggesting that the spores may have been shed en masse (FIG. 8.31). The gametangiophores (gametophytes) are fragmentary and there is still some question as to whether they arise from a common thallus (i.e., they are bisexual), or whether each arises from a single spore that only produces antheridiophores or archegoniophores, as has been hypothesized (Taylor et al., 2005c). The same type of gametangia-bearing structures occur in other Rhynie chert sporophyte–gametophyte associations, including Rhynie gwynne-vaughanii–Remyophyton delicatum (Kerp et al., 2004), Nothia aphylla–Kidstonophyton discoides (Remy and


Paleobotany: the biology and evolution of fossil plants

Figure 8.28 Section of Lyonophyton rhyniensis protocorm (Devonian). Bar  400 μm. (Courtesy H. Kerp.)

Figure 8.26 Multicelled gametophyte extending from ruptured spore wall (arrow) (Devonian). Bar  30 μm. (Courtesy H. Kerp.)

Figure 8.27 Multicellular gametophyte of Lyonophyton rhyniensis. Arrow indicates possible pyramid-shaped apical cell (Devonian). Bar  40 μm. (Courtesy H. Kerp.)

Hass, 1991b), and Horneophyton lignieri–Langiophyton mackiei (FIG. 8.32) (Remy and Hass, 1991c), perhaps adding support to the hypothesis that the Aglaophyton–Lyonophyton gametangiophores (FIG. 8.33) were unisexual. All of the Rhynie chert gametophytes contain conducting cells (FIG. 8.34) and are mycorrhizal (see below).

Figure 8.29 Antheridium with escaping sperm (Devonian). Bar  30 μm. (Courtesy H. Kerp.)

All of the Rhynie chert plants studied in detail suggest that various forms of vegetative reproduction (FIG. 8.35) were widespread. The presence of reduced branches, bulbils (FIG. 8.36), and bulges with disorganized vascular tissue on



Figure 8.30 Transverse section of Lyonophyton rhyniensis

archegonium neck showing eight neck Bar  30 μm. (From Remy and Hass, 1991c.)



Figure 8.32 Suggested reconstruction of distal end of Langiophyton mackiei archegoniophore (Devonian). (From Taylor and Taylor, 1993.)


8.31 Spore

ball in Rhynie chert (Devonian).

Bar  1 mm.

axes and various types of fragmentation suggest that these plants were clonal and possessed multiple reproductive strategies. The environment in which they lived has been reconstructed as a freshwater ecosystem with volcanic influence including hot springs (Trewin et al., 2003). Thus, it seems likely that the various forms of asexual reproduction may have allowed rapid colonization in a variable environment. RHYNIA GWYNNE-VAUGHANII This species is reconstructed as a small plant, 18.0 cm tall, with upright, dichotomizing axes that arise from a rhizome

Figure 8.33 Longitudinal section through the antheridio-

phore of Lyonophyton rhyniensis showing two antheridia (arrows) (Devonian). Bar  4 mm. (Courtesy H. Kerp.)


Paleobotany: the biology and evolution of fossil plants

Figure 8.34 Longitudinal view of conducting elements in Lyonophyton rhyniensis (Devonian). Bar  12 μm. (Courtesy H. Kerp.)

Figure 8.36 Bulbil (arrow) extending from vegetative axis of

Aglaophyton (Devonian). Bar  1 mm. (Courtesy H. Kerp.)

Figure 8.35 Diagrammatic population of Aglaophyton major plants showing single individual (yellow box), association of plants by endomycorrhizae (brown filament), and fragmentation (red lines).

bearing delicate, threadlike rhizoids. The stems are 2.0– 3.0 mm in diameter with a narrow central strand. Surrounding the central conducting strand are some poorly preserved, thin-walled cells that have been suggested to be phloem (FIG. 8.37); some of these cells have been described as containing sievelike thin areas on their walls (Satterthwait and Schopf, 1972). The remainder of the axis consists of a twoparted, parenchymatous cortex. The inner cortex consists

Figure 8.37 Detail of Rhynia gwynne-vaughanii axis showing

histologic differences in cells (Devonian). Bar  125 μm.

of cells with large intercellular spaces and the narrow outer cortex of more tightly packed cells. A thin cuticle is present on the epidermis, as well as stomata, each with two simple, kidney-shaped guard cells. Sections of the rhizome show the



Figure 8.38 In situ stand of Remyophyton delicatum axes

(Devonian). Bar  6 mm. (Courtesy H. Kerp.)

same complement of tissues as the aerial axes. On the tips of some aerial axes are ellipsoidal, thick-walled sporangia up to 3.0 mm long. The sporangial wall is constructed of a zone of outer palisade cells with thickened lateral walls termed a cohesion tissue. Rhynia gwynne-vaughanii was homosporous and produced spores in tetrahedral tetrads; each spore is 40 μm in diameter and ornamented by closely spaced spines. The gametophyte phase is called Remyophyton delicatum (Kerp et al., 2004) and consists of a dense cluster of 200 unbranched axes, 0.2–0.7 mm in diameter (FIG. 8.38), with the basal region represented by globular prothalli (protocorms) bearing rhizoids. Gametophytes are unisexual—either archegoniophores or antheridiophores. Archegoniophores are larger (10–15 mm long) and bear massive archegonia with long protruding necks. Antheridiophores are smaller (4–8 mm long), with stalked antheridia arising from the flattened upper surface. Conducting tissue is made up of S-type elements. A globular mass slightly larger than the gametophyte axis has been described extending from the archegonium in one specimen and this has been interpreted as a putative sporophyte (Kerp et al., 2004). In his reexamination of some of the Rhynie chert plants, David S. Edwards (1980) provided additional information about R. gwynne-vaughanii. His studies confirmed the existence of hemispherical projections and numerous short adventitious branches along the stem. Some of the latter structures are interpreted as underdeveloped lateral branches (FIG. 8.39), whereas others may represent asexual reproductive structures, which presumably broke off the parent plant. The hemispherical projections may produce rhizoids when they are in contact with the substrate. It has also been suggested that these projections may represent possible secretory structures or hydathodes. By documenting the presence

Figure 8.39 Rhynia gwynne-vaughanii showing possible early

stage of adventitious branch (arrow) (Devonian). Bar  650 μm.

of the adventitious branches, Edwards was able to show that R. gwynne-vaughanii was more monopodial in growth architecture than Aglaophyton. In addition, he suggested that sporangia abscised after the release of spores, and in some instances, new branches formed distal to the abscission zone (Edwards, 1980). HORNEOPHYTON LIGNIERI This plant was originally described by Kidston and Lang (1920a) as Hornea, a name that was previously occupied by a flowering plant. Some features of Horneophyton lignieri are of special interest because they differ from those of other Rhynie chert plants. The plant consists of naked and dichotomously branched aerial axes (FIG. 8.40) up to 20 cm high and 2.0 mm in diameter, which are attached to a lobed, cormlike structure that bears numerous rhizoids on the lower surface. Although the structural features of the aerial axes are similar to those of Rhynia, with a central conducting strand, the basal corm lacks any evidence of vascularization. Recent studies indicate that in the basipetal regions of the aerial stem the phloem cells become indistinct, with conducting elements losing their characteristic features in more basal sections. At the transition between stem and corm, opaque parenchyma cells replace the conducting elements and these are distinguishable in the corm proper. Sporangia of H. lignieri are borne terminally at the tips of some of the branches. Each sporangium is branched, consisting of two to four lobes of varying length. Sporangial lobes tend to be ellipsoidal–cylindrical in shape, with the distal end truncated. It has been suggested that sporangial


Paleobotany: the biology and evolution of fossil plants

The sporangium in Horneophyton is unique among plants, both living and fossil, in that it consists of a branched fertile unit of four lobes that resulted from dichotomies of the stem apex (Eggert, 1974). Consequently, each fertile lobe was produced by its own apex, and these apices must have remained meristematic for a short time, as spores within the sporangium show evidence of acropetal maturation. The lobed Horneophyton sporangium also has been interpreted as transitional, leading to a synangium in which some differentiation and partitioning of sporogenous tissue took place.

Figure 8.40 Suggested reconstruction of Horneophyton lig-

nieri (Devonian). (From Taylor and Taylor, 1993.)

dehiscence took place through an apical pore. Extending into the sporangial cavity is a central column of sterile tissue around which a continuous zone of sporogenous tissue, or spores, developed. Both isobilateral and tetrahedral tetrads of spores occur in Horneophyton sporangia (Sharma and Bohra, 1985). The spores (see Wellman et al., 2004 for details) are radial, trilete, and irregularly ornamented by cavities that may represent some type of exine degradation. They range from 39–49 μm in diameter and are most similar to the sporae dispersae taxon Emphanisporites decoratus (Wellman et al., 2004). Some have been described as containing multicellular gametophytes (Bhutta, 1973). The plant is believed to have been homosporous. The gametophytes are described under the generic name Langiophyton mackiei (Fig. 8.32) (see below).

ASTEROXYLON MACKIEI The most complex element in the Rhynie chert flora is Asteroxylon mackiei. It is a particularly interesting plant that is sometimes included in the Drepanophycales (Chapter 9), although it lacks true leaves. Unlike the other Rhynie taxa, A. mackiei is characterized by numerous small flaps of tissue (leaflike appendages or enations) that cover the aerial stems, as well as a more complex central strand and conducting system. Asteroxylon mackiei was homosporous like the other Rhynie chert plants, but the sporangia were located laterally along the stem, not apically, as in the other taxa. In the original description of this plant, Kidston and Lang (1920b) described narrow and naked distal branches with terminal sporangia that, although not actually attached, were thought to be the fertile portion of the plant. The presumed terminal position of sporangia in A. mackiei was considered to be evidence that all the Rhynie plants were closely related. Subsequent studies of chert blocks containing A. mackiei axes, however, indicate that the sporangia were borne laterally on the stems, near the axils of leaflike appendages, instead of in a terminal position (Lyon, 1964). This important discovery greatly altered the taxonomic position of the genus. Asteroxylon mackiei was probably up to 50.0 cm tall and consisted of upright, monopodial axes supported by a horizontal, subterranean rhizome (Fig. 8.41), which may be up to 4.2 mm in diameter (Kidston and Lang, 1920b). From the principal aerial branches arose secondary branches that were regularly dichotomous. In contrast to other Rhynie plants, the aerial stems of Asteroxylon were densely covered by numerous leaflike flaps of tissue, each up to 5.0 mm long, and the taxon is known from compressions (Fig. 8.41) as well. These structures have been called leaves, scalelike leaves, leaflike scales or appendages, or enations. They are not true leaves (Chapter 7), as they contain no vascular tissue and are not produced in a regular pattern (phyllotaxy) from nodes, as true leaves are. The enations in Asteroxylon have been hypothesized to represent the precursors to microphylls, the leaf type which is unique to the Lycophyta (see Chapter 9 for a discussion). It is not certain that stomata present on the stems also existed on the enations.



Figure 8.43 Transverse section of Asteroxylon mackiei showing actinostele and highly lacunate cortex characteristic of mature axes. Arrows indicate several traces in the cortex (Devonian). Bar  4 mm.

Figure 8.41 Suggested reconstruction of Asteroxylon mackiei


Figure 8.42 Cross section of Asteroxylon mackiei stele

(Devonian). Bar  1 mm.

In the rhizomatous portion of Asteroxylon the conducting strand has a central core of thick-walled elements, while the strand is stellate or star shaped in cross section in the aerial axes (Fig. 8.42), the feature from which the generic name is derived. The stele of Asteroxylon has been interpreted as an Asteroxylon-type protostele (Ogura, 1972) or an actinostele. Primary xylem is slightly mesarch, with the protoxylem elements situated near the edges of the xylem ridges. Xylem elements have annular and helical secondary thickenings. Thin-walled cells in the furrows between the xylem ridges are thought to represent phloem. In a transverse section of an aerial axis, numerous small strands of vascular tissue can be seen in the cortex (FIG. 8.43). These traces originate at the outer edges of the xylem ridges and extend through the cortex, ending abruptly near the periphery of the stem at the bases of the enations; they do not pass into the enations themselves. Reniform sporangia, with dehiscence along the distal edge, are produced on short, vascularized pedicels scattered among the enations, apparently in no particular relation to the enations themselves. Sporangia are up to 7.0 mm long; spores are 40–60 μm in diameter and ornamented by closely spaced spines on the distal surface. NOTHIA APHYLLA Kidston and Lang (1920b) initially described naked axes bearing pear-shaped sporangia as the sporangia of Asteroxylon. Today these sporangia are recognized as belonging to N. aphylla (Lyon, 1964; El-Saadawy and Lacey, 1979). The


Paleobotany: the biology and evolution of fossil plants

Figure 8.44 Section of Nothia aphylla rhizome (Devonian). Bar  1.0 mm (Courtesy H. Kerp.) Figure 8.45 Suggested reconstruction of Nothia aphylla

N. aphylla plant consists of an aerial system of upright, dichotomously branching axes that arise from subterranean rhizomes 2.5 mm in diameter (FIG. 8.44); the entire plant was 20 cm tall (Kerp et al., 2001). The aerial axes are covered by slightly elongated emergences (FIG. 8.45), each of which bears a single stoma. While the aerial axis is anatomically similar to that of other Rhynie chert plants, the anatomy of the prostrate rhizome is unique. In cross section, the rhizome is bilaterally symmetrical with a median ridge (rhizoidal ridge) that extends the length of the axis on the lower surface; extending from the ridge are unicellular rhizoids up to 1.5 mm long (Fig. 8.44). The rhizome anatomy includes a central stele with a strand of fibrous conducting cells surrounded by a narrow zone of phloem-like tissue. Outside the phloem is a parenchymatous cortex, hypodermal tissues, and epidermis. The internal anatomy of the ventral rhizoidal ridge includes a rhizoid-bearing epidermis, several hypodermal layers of relatively large and radially arranged cells, files of thin-walled parenchymatous cells that connect to the stelar body, and individual, extrastelar conducting elements (xylematic elements of Kerp et al., 2001).

(Devonian). (From Daviero-Gomez et al., 2004.)

Sporangial position in N. aphylla is highly variable, with individual sporangia borne on adaxially recurved stalks whose attachment may range from helical to almost whorled. Sporangia are reniform (3.0 mm wide by 2.0 mm long) with dehiscence through a long, apical transverse slit; spores are radial (65 μm) and trilete. The gametophytes are assigned to Kidstonophyton discoides and are unisexual (Remy and Hass, 1991b). Like many, and perhaps all of the Rhynie chert plants, N. aphylla is highly clonal, with numerous plantlets borne on the rhizomes (Daviero-Gomez et al., 2005). These appear as lateral buds on the rhizome, and in sections of the chert in which the plants are preserved, it is possible to see several levels of rhizomes in section view. In certain anatomical features, N. aphylla may be compared with members of the old Rhyniophyta, whereas the laterally borne, stalked, reniform sporangia are features that suggest affinities with the Zosterophyllophyta.


TRICHOPHEROPHYTON TEUCHANSII This plant from the Rhynie chert is very rare. The aerial axes are up to 2.5 mm in diameter, branch dichotomously and pseudomonopodially, and exhibit circinate vernation. They are covered with small, unicellular spinelike projections. The central strand has exarch maturation with annular–helical secondary wall thickenings on the conducting elements (Lyon and Edwards, 1991) (FIG. 8.46). Sporangia are reniform and attached laterally to the axes by a small stalk, characters which identify T. teuchansii as a zosterophyll. The sporangia also bore unicellular spines and produced smooth, trilete spores. VENTARURA LYONII To date, this taxon is known only from the Windyfield chert site, which is just a short distance from the original Rhynie chert locality, and of the same age (Fayer and Trewin, 2004). The naked aerial axes of V. lyonii are larger, up to 7.2 mm in diameter, branch dichotomously, and contain a conducting strand which consists of a cluster of thin-walled cells with uneven bands like those seen in G-type cells (Powell et al., 2000); tips of axes are circinately coiled. The middle cortex contains a zone of thick-walled cells interpreted as sclerenchymatous. Like Trichopherophyton, sporangia are reniform and attached laterally to the axes by small stalks. They appear in a row along the axis, suggesting that perhaps they were produced in a loose

Figure 8.46

Geoffrey Lyon. (Courtesy N. Trewin.)


spore-producing structure, such as a strobilus. The wall of the sporangium is multilayered and trilete spores 75 um in diameter were produced by this homosporous plant. The shape and position of the sporangia and the coiled apices suggest that V. lyonii is a zosterophyll. Fayers and Trewin (2004) describe Ventarura growing in association with Asteroxylon and Nothia and suggest that this plant community colonized organic-rich, sandy substrates. To date nothing is known about the gametophyte generation of either Trichopherophyton or Ventarura. GAMETOPHYTE GENERATION

In this chapter, in fact throughout this book, the primary focus is on the sporophyte generation of vascular plants. This is not by choice, but rather reflects the general paucity of information about the gametophyte phase in most fossil plants. Although there is some information about the gametophytes in certain seed plants (Millay and Eggert, 1974; Chapter 14), information about fossil vascular plant gametophytes is generally rare. Until recently this has been especially true in discussions of early land plants and has resulted in a variety of theories to explain both the origin of the dominant sporophyte phase and the absence of gametophytes in the fossil record (Chapter 6). Some of these ideas have influenced the interpretation of the Rhynie chert sporophytes. One of the early theories suggested that Aglaophyton major represented the sporophyte and Rhynia gwynne-vaughanii, the free-living gametophyte phase of a single plant (Pant, 1962). The idea gained support from the homologous theory of the alternation of generations (Chapter 6), since there were no specimens of R. gwynne-vaughanii known to possess sporangia at that time. This theory has been abandoned since R. gwynne-vaughanii was subsequently shown to possess terminal sporangia (Edwards, 1980) and A. major is now known to be structurally somewhat different from a vascular plant. Another theory suggested that the small protrusions near the base of R. gwynne-vaughanii axes represented archegonia (Lemoigne, 1968). According to this hypothesis, the rhizomatous portion of the plant represented the gametophyte phase and the aerial upright axes bearing sporangia constituted the sporophyte. Such a situation might be viewed as evidence for the antithetic theory and parallels the life history in modern bryophytes, where the sporophyte phase is permanently parasitic on the gametophyte. Although this hypothesis cannot be totally discounted, to date there is little structural evidence in the form of a transitional zone or foot-like structure, which should be present at the base of all R. gwynne-vaughanii plants. One fossil that was thought to support the hypothesis of a bryophyte-like gametophyte is Horneophyton lignieri.


Paleobotany: the biology and evolution of fossil plants

Several features of this genus are similar to those found in bryophytes, especially the columellate sporangia and cormlike base. Some suggested that the relatively simple modern hornwort Anthoceros could represent an early morphological type that would lead to the more complex sporophyte of H. lignieri. It is important to understand that these ideas were advanced at a time when many hypothesized that the bryophytes represented an intermediate stage in the evolution of vascular plants from the green algae. Today the hornworts and bryophytes are mostly considered to be paraphyletic (Chapter 5). There is continuing disagreement, however, as to whether the liverworts or the hornworts represent the basal group among the bryophytes (Nickrent et al., 2000; Friedman et al., 2004; Shaw and Renzaglia, 2004). We are still a long way from understanding the evolution of the sporophyte of vascular plants, as well as the nature of the plant body in a number of fossils that exhibit a bryophytic level of organization. Some exceptionally well-preserved fossils that do contribute to a better understanding of the sporophyte–gametophyte relationship in early land plants have been described by the Remys (FIG. 8.47) from the Rhynie chert (Remy and Remy, 1980a, b; Remy et al., 1993). One of these is Lyonophyton rhyniensis (FIG. 8.48), a small axis that terminates in a shallow, bowl-shaped structure (5.0 mm in diameter), described as a gametangiophore (gametophyte) (Remy and Hass, 1991a). Antheridia are distributed over the entire upper surface. Each antheridium is characterized by a central mass of sterile tissue (columella), and many contain exceptionally well-preserved, coiled spermatozoids (FIG. 8.49). Stomata (FIG. 8.50) occur on the lower surface of the gametangiophore bowl. In the center of the stalk are dark, elongated conducting cells which do not extend into the bowl itself.

Figure 8.47 Renate and Winfried Remy. (Courtesy D. Remy.)

Figure 8.48 Longitudinal section of Lyonophyton rhynien-

sis antheridiophore showing central conducting strand (arrow) and antheridia in cup-shaped distal end (Devonian). Bar  2 mm. (Courtesy H. Kerp.)

The similarity of epidermal cells and conducting elements has been used to demonstrate that Lyonophyton rhyniensis is the gametophyte of Aglaophyton major. Kidstonophyton discoides (FIG. 8.51) is another gametophyte discovered in the Rhynie chert that resembles Lyonophyton (Remy and Hass, 1991b). It consists of a stalk terminating in a shallow, cup-shaped structure (antheridiophore) (FIG. 8.52) that contains numerous antheridia interspersed among ridges of sterile tissue. Spermatozoids have also been described. The sporophyte in the Kidstonophyton life cycle is Nothia aphylla. Associated with these two taxa in the Rhynie chert is another gametophyte that Remy and Hass (1991c) named Langiophyton mackiei (Fig. 8.32). This plant consists of upright axes 6.0 mm long; each terminates in a flattened, peltate structure from which arise numerous (about 30) archegonia. The upper surface of the peltate disk is irregular, and each archegonium is characterized by an elongated neck and deep-seated venter. Antheridiophores are bowl-shaped


Figure 8.49 Section of antheridium showing coiled sperm (Devonian). Bar  30 μm. (Courtesy H. Kerp.)

and contain up to 50 antheridia. Langiophyton mackiei is the gametophyte of Horneophyton lignieri. The most recent gametophyte described from the Rhynie chert is Remyophyton delicatum (FIGS. 8.53, 8.54) (Kerp et al., 2004). Unlike the other Rhynie chert gametophytes, those of R. delicatum are attached to rhizoid-bearing protocorms (Fig. 8.38). The upright axes possess S-type conducting elements like those of the sporophyte R. gwynne-vaughanii. Gametangiophores are unisexual with the larger ones bearing archegonia. All of the gametophytes known to date from the Rhynie chert are morphologically quite similar to one another, each consisting of an elongated stalk bearing a flattened or saucer-shaped head that contains antheridia or archegonia (FIG. 8.55). These gametophytes appear similar in organization to those found in some members of the hepatic order Marchantiales (Chapter 5), which produce upright antheridiophores and archegoniophores from a prostrate thallus. In the Rhynie chert gametophytes, at least one taxon, Remyophyton delicatum, appears to arise from a protocorm-like structure, not a thallus. All possess stomata and conducting elements, and some have been demonstrated


Figure 8.50 Paradermal section of Lyonophyton rhynien-

sis gametangiophore showing stoma (Devonian). Bar  80 μm. (Courtesy H. Kerp.)

8.51 Diagrammatic section of distal end of Kidstonophyton discoides antheridiophore (Devonian). (From Taylor and Taylor, 1993.)



Paleobotany: the biology and evolution of fossil plants

to be endomycorrhizal (FIG. 8.56) (Taylor et al., 2005c). As more gametophytes are discovered it will be interesting to see if any of the antheridiophores and archegoniophores are found attached to some type of thallus-like structure.

Figure 8.52 Longitudinal section of distal end of Kidstonophyton discoides antheridiophore showing several antheridia (arrows). Compare with FIG. 8.51. (Bar  1 mm).

Morphologically, Lyonophyton, Langiophyton, Kidstonophyton, and Remyophyton show some similarity to the Early Devonian compression genus Sciadophyton (FIG. 8.57), which also consists of radiating axes terminating in cuplike structures (FIG. 8.58) (Remy et al., 1980a, b; Schweitzer, 1980a). A few elongated cells in the center of the axis are interpreted as conducting elements, but because the specimens are known only as compressions, these features remain equivocal. Although initially regarded as a sporophyte, the presence of oval bodies on the inner surface of the cup, similar in size to those in the Rhynie chert gametophytes, are suggestive of gametangia (Kenrick and Crane, 1997a). The presence of Sciadophyton in the same rocks with several morphological forms considered to be developmental stages prompted Schweitzer (1981) to regard Sciadophyton as the gametophyte of Stockmansella langii. Kenrick et al. (1991b), however, based on mineral casts of the conducting elements, interpret Sciadophyton as the gametophyte of either Stockmansella or Huvenia. As suggested by Remy et al. (1992), it is highly probable that the

Figure 8.54 Distal end of Remyophyton delicatum gametangiophore showing archegonium (arrow) (Devonian). Bar  300 μm. (Courtesy H. Kerp.)

Figure 8.53 Several Remyophyton delicatum gametangiophores with antheridia (arrows) (Devonian). Bar  2 mm. (Courtesy H. Kerp.)

Figure 8.55 Diagrammatic representation of a population of Lyonophyton gametangiophores associated with mycorrhizae.



morphotaxon Sciadophyton represents a gametophyte stage of several early land plants. The anatomical and morphological similarity of the gametophytes described from the Rhynie chert to date would appear to strengthen Remy’s hypothesis.

Figure 8.56 Arbuscules (arrows) in gametangiophore of

Lyonophyton rhyniensis (Devonian) Bar  36 μm (Courtesy of H. Kerp.)

Figure 8.57 Suggested reconstruction of Sciadophyton sp.

(Devonian). (From Taylor and Taylor, 1993.)

Figure 8.58 Sciadophyton steinmannii, a gametophyte believed to belong to the sporophyte Zosterophyllum rhenanum (Devonian). Bar  2 cm. (Courtesy BSPG.)


Paleobotany: the biology and evolution of fossil plants

Calyculiphyton is an Emsian (Lower Devonian) compression fossil that may also represent a gametophyte (Remy et al., 1991). In this type, elongate axes terminate in cupshaped structures that morphologically look much like those of Sciadophyton. The anatomically preserved gametophytes from the Rhynie chert provide the first real fossil evidence that pertains to the evolution of alternation of generations. The nature of the conducting strand and relationship to the sporophyte phase of Rhynie chert plants substantiate the existence of free-living, terrestrial gametophytes as early as the Early Devonian. The conducting strand of Aglaophyton major and several other early land plants provided the impetus to reevaluate the conducting elements of many early land plants and to use this feature to establish the biological affinities of the sporophyte and gametophyte phase. To date there is little evidence of post-fertilization stages or the new sporophyte (embryo) in any early land plants. Kerp et al. (2004) described and illustrated some tissue extending from an archegonial neck of Remyophyton delicatum that might represent the gametophyte–sporophyte junction and development of the new sporophyte generation of R. gwynne-vaughanii. OTHER RHYNIOPHYTES

As mentioned earlier, Cooksonia, a plant that has been historically regarded as the oldest vascular plant, is included in this group. Compressed specimens have been described from localities all over the world, including North America, Great Britain, North Africa, Europe, Siberia, and South America (e.g., Fig. 8.59). Some Cooksonia specimens discovered in Wales are known from deposits as old as the Ludlovian; other specimens suggest that the taxon extended into the Early Devonian (Emsian). Sporangia thought to belong to Cooksonia have been described from Wenlock (mid-Silurian) rocks (Edwards and Feehan, 1980). Cooksonia hemisphaerica, from the Upper Silurian of Wales consists of dichotomous branches up to 6.5 cm long with axes 1.5 mm wide. Stomata occur along the aerial axes (Edwards, 1979). Sporangia are terminal and vary from hemispherical to spherical; in C. pertoni and C. downtonensis, sporangia are wider and longer. In C. caledonica, the shape of the sporangium is highly variable. None of the specimens shows a distinct dehiscence mechanism and all plants were apparently homosporous. One might suggest that all of the Cooksonia specimens described to date merely represent the distal branches of a much larger plant. The uniform sizes of the many specimens described thus far, however, favor the interpretation of Cooksonia as a small plant. More recently, a larger specimen interpreted as Cooksonia was described

from the early Lochkovian (Upper Devonian) of Brazil which includes five orders of dichotomous branching (Gerrienne et al., 2006). Several hypotheses are offered as to whether the specimen represents a sporophyte, a gametophyte, or a sporophyte arising from a thalloid gametophyte (prothallium). Although Cooksonia is considered by many to represent the oldest vascular plant, the evidence for this assumption is still equivocal. When the genus was originally described by Lang (1937), none of the axes with terminal sporangia in the original specimens displayed conducting elements; the presence of a central strand composed of specialized (perhaps conducting) cells was only observed in isolated axes lacking sporangia. Axes containing terminal sporangia of the Cooksonia type have been found to possess thick-walled cells of a sterome (Edwards et al., 1986), but no vascular elements (Edwards et al., 1992). In addition to axes lacking clearly defined vascular tissue, specimens of Cooksonia have also been reported with a variety of sporangial morphologies (Edwards et al., 2004), variously ornamented spores, axes with and without stomata, and various forms of branching. Thus, it appears that the more we learn about Cooksonia, the more difficult it is to interpret precisely what the genus Cooksonia represents. Edwards and Edwards (1986) place Cooksonia in a group they refer to as rhyniophytoids, whereas Taylor (1988a) terms this group cooksonioids, defined as small plants with terminal sporangia borne on narrow axes that lack true tracheids. Other authors simply include the Cooksonia-like axes in the Eutracheophytes (Kenrick and Crane, 1997a). All of these interpretations underscore that the cooksonioids as presently

Figure 8.59 Several Cooksonia specimens from Brazil (Devonian). Bar  5 mm. (Courtesy P. Gerrienne.)


understood represent a highly artificial group of plants that existed during the Late Silurian–Early Devonian, and that may include forms that are ancestral to either bryophytes or vascular plants, or possibly both. Uskiella (FIG. 8.60) is used for both permineralized and compressed cooksonioids from the Lower Devonian of southern Wales. They have naked, simple isotomous branching and ellipsoidal sporangia (Shute and Edwards, 1989). The sporangial wall has several cell layers thick, with a longitudinal row of thin-walled cells along which the sporangium splits (Fanning et al., 1992). The spores have been described as alete and range from 28–42 μm in diameter; they possess a two-layered sporoderm.

Figure 8.60 Suggested reconstruction of Uskiella spargens (Devonian). (From Taylor and Taylor, 1993.)


Another genus that morphologically resembles Cooksonia and Uskiella is Dutoitea (Rayner, 1988). Several species have been described from Lower Devonian (Lochkovian?) compressions of Cape Province, South Africa. Some axes show a thin median line that may represent a conducting strand, perhaps even that of a bryophytic level of organization. In at least one specimen, multicellular spines are present on the axes. Nothing is known about the spores. In the cladistic analysis of Kenrick and Crane (1997a), Hsüa (FIG. 8.61), from the Middle Devonian of Yunnan, China, is included among the zosterophyllophytes. This simple plant consists of main axes 1.0 cm wide that divide to produce lateral branches, some of which terminate in reniform sporangia (C.-S. Li, 1982). Stomata occur along the stems in H. robusta together with tubercles. In the center of each axis in H. deflexa (FIG. 8.62) is a terete, centrarch protostele with G-type conducting cells (D.-M. Wang et al., 2003a). Spores range from 18–36 μm and are trilete. In H. deflexa from South China, spines are present along the axes (D.-M. Wang et al., 2003b). Although features of Hsüa suggest affinities with the cooksonioids, the pattern of branching and features of the sporangia are strikingly similar to that found in some zosterophylls. Another plant at the cooksonioid level of organization is the Early Silurian genus Steganotheca (Edwards, 1970a). Specimens preserved as compressions are 5.0 cm tall and contain several orders of branching, each bearing a terminal sporangium. Although no conducting elements have been identified from the specimens, each axis contains a centrally located striation that may represent some form of conducting strand or

Figure 8.61 Suggested reconstruction of Hsüa robusta (Devonian). (From Kenrick and Crane, 1997a.)


Paleobotany: the biology and evolution of fossil plants

Figure 8.62 Suggested reconstruction of Hsüa deflexa (Devonian). (From D.-M. Wang et al., 2003b.)

sterome. Sporangia are elongated (2.5 mm) and the surface is striated. Nothing is known about the sporangial contents. Hedeia corymbosa consists of dichotomizing axes, each terminated by an elongated sporangium (Cookson, 1935). The specimens, which are Early Devonian in age, are not known in sufficient detail, so features of the sporangium and spores have not been described. Specimens of Yarravia also consist of the distal ends of dichotomizing branches, and may represent a preservational state of Hedeia (Hueber, 1983). In Hedeia the fertile axes show some similarity to that found in the trimerophytes. Another cooksonioid is Salopella (Edwards and Richardson, 1974). It consists of compressed, naked, dichotomously branched axes up to 2.0 mm wide which bear terminal sporangia (Larsen et al., 1987). The trilete spores are all of the same morphological type and described as azonate. At the present time the spores are sufficiently different to allow Salopella to be distinguished from other rhyniophyte taxa (Edwards and Fanning, 1985). Eogaspesiea gracilis is the name given to tufted, dichotomously branched axes up to 10.0 cm long (Daber, 1960a). This Early Devonian taxon includes a tangled mass of axes believed to have been attached to a rhizome. At the end of some of the axes are elongate sporangia, each up to 2.5 mm long and containing thin-walled, perhaps alete spores. Not all of the early land plants had cylindrical axes. Some taxa, such as Taeniocrada, which ranges from the Lower to the Upper Devonian, included plants with flattened axes and dichotomous branching (FIG. 8.63). Sporangia are

Figure 8.63 Suggested reconstruction of Taeniocrada deche-

niana based on Kräusel and Weyland, 1930. (Courtesy H. Kerp.)

typically terminal on specialized fertile, repeatedly forked branches and range from 3.0–7.0 mm long; a few sporangia appear in a lateral position (FIG. 8.64). The occurrence of this fossil in dense mats and the apparent lack of stomata on the stem surfaces prompted some to suggest that it may have been aquatic or semi-aquatic. Taeniocrada stilesvillensis (Upper Devonian of New York) consists of axes that branch dichotomously or pseudomonopodially, with hairlike projections arising from ridges along the axes. This taxon may also be distinguished by exarch maturation of the probable G-type conducting elements (Taylor, 1986). The genus Stockmansella has been instituted by Fairon-Demaret (1985, 1986) for forms previously assigned to Taeniocrada



Figure 8.64 Fertile portion of Taeniocrada sp. (Devonian). Bar  4 mm. (Courtesy H. Kerp.)

that bear single sporangia in a lateral position on the axes. Stockmansella remyi from the Eifelian of northwestern Germany consists of a system of prostrate, aboveground axes up to 10 cm long and 3.1 mm wide which repeatedly bifurcate and are characterized by a central xylem strand consisting of S-type cells (Schultka and Hass, 1997). These basal or rhizomatic axes produce narrow laterals at right angles; these may turn upward and become smooth, erect axes. The rhizomatic axes also produce arrested laterals positioned in the axils of bifurcations and irregularly distributed sporangia arising from nonvascularized pads (sporangiophores) that are attached laterally to the main axes. Sporangia are elongate or ovoid, up to 2.2 mm long and 1.0–2.2 mm wide, and dehisce by one to several longitudinal fissures. Prostrate axes, whether main axes or laterals, produce scattered rhizoid-bearing bulges on all sides. Taeniocrada dubia is a Devonian plant that was historically included in the rhyniophytes, but Hueber (1982) has suggested that it contains a central strand composed of tubes

8.65 Suggested reconstruction of Huia gracilis (Devonian). (From Wang and Hao, 2001.)


of varying diameters. On the inner surface of each tube is a series of helical thickenings that represent a component of the primary wall rather than being secondarily deposited. The wall in these tubes has three parts, consisting of an inner microporate layer, a middle spongy zone that constitutes the majority of the cell wall, including the helical thickenings, and an outer fibrillar layer. Huia gracilis is a permineralized Early Devonian (Yunnan Province, China) plant with oval to reniform sporangia (FIG. 8.65) (Wang and Hao, 2001). The presence of G-type conducting elements in a centrarch strand, K- and


Paleobotany: the biology and evolution of fossil plants

Figure 8.66 Sporangia of Huvenia kleui showing twisted organization of the wall. Sporangium at right shows attachment region (Devonian). (From Kenrick and Crane, 1997a.)

H-type branching, and sporangia borne on elongate stalks are features found among more than a single group of early land plants. Whether H. gracilis represents a plant which is transitional between the rhyniophytes and zosterophyllophytes or is intermediate in a line leading to the trimerophytes is yet to be determined. The Early Devonian (Pragian) plant Huvenia also has what are interpreted as flattened axes (Hass and Remy, 1991; Schultka, 1991). The axes often bifurcate and bear small protrusions interpreted as rhizophores. The conducting stand contains S-type elements. Twisted fusiform sporangia (FIG. 8.66) are attached to stout sporangiophores and borne on the primary axes near branches. In a specimen of Huvenia sp. from Gaspé, Hotton et al. (2001) describe sporangia attached to small branches. Small disks associated with the specimen are interpreted as vegetative reproductive units. It is unknown whether the twisted configuration of the sporangium reflects a postmortem change, the actual morphology, or structural variation perhaps related to periodic drying. Spores of Huvenia from Gaspé are compared to the sporae dispersae taxa Retusotriletes or Calamospora (Hotton et al., 2001). One plant that combines features of two early vascular plant groups is Renalia (FIG. 8.67) (Gensel, 1976). The characteristics of Renalia further underscore the inherent problems in classifying many of the Devonian and pre-Devonian land plants. Specimens of R. hueberi occur as compressions in the Battery Point Formation from the famous Gaspé region of Québec (Lower Devonian). It is estimated that the plant was up

Figure 8.67 Suggested reconstruction of Renalia hueberi (Devonian). (From Taylor and Taylor, 1993.)

to 30.0 cm tall, and consisted of a pseudomonopodial main axis to which were attached lateral branches dichotomizing several times and terminating in reniform sporangia. The spores measure 46–70 μm in diameter and are trilete. Although a few helical–scalariform conducting elements have been recovered in macerates, virtually nothing is known about the conducting system. Terminal sporangia and pseudomonopodial branching are features that suggest affinities with the rhyniophytes, but the large, reniform-shaped sporangia (with dehiscence along the distal margin) are characteristics common to members of the zosterophyllophytes. Pinnatiramosus quianensis was originally described from the Lower Silurian (Llandovery) of China as the oldest vascular



the endogenous origin of the laterals (Edwards et al., 2007). Detailed analysis of the matrix suggests that although the specimens are fossil, they probably represent roots of a geologically younger (?Permian) extinct plant that grew into the Silurian rocks below and subsequently became fossilized. DISCUSSION: RHYNIOPHYTE EVOLUTION

Figure 8.68 Pinnatiramosus Bar  1 cm. (Courtesy D. Edwards.)



Figure 8.69 Detail of Pinnatiramosus quianensis show-

ing interdigitating laterals (Silurian). Bar  5 mm. (Courtesy D. Edwards.)

plant (Geng, 1986; Cai et al., 1996). The plant includes naked, compressed branching systems, some up to 40 cm long, that produced closely spaced, interdigitating laterals (FIGS. 8.68, 8.69). Cells macerated from the main axis show tracheid lumen casts with circular-bordered pits. A more recent reexamination of this fossil within its depositional environment concludes that it represents a rooting system, rather than aerial axes, based on the two dimensionality of the branching systems and

The rhyniophytes have traditionally occupied the position of the oldest and simplest vascular plants. Many authors, including Banks (1975), have suggested that the rhyniophytes gave rise to the trimerophytes. As new information becomes available, however, it is clear that some of the so-called rhyniophytes are not true vascular plants, but share features of both bryophytes and vascular plants; these plants have been accommodated in an artificial group, the cooksonioids, until we know more about some of these early forms. From the foregoing section, it should be apparent that although these Late Silurian–Devonian plants share many features, they include so many plesiomorphic characters that it is difficult to include them in higher taxonomic categories. In addition, the information on conducting elements from exceptionally well-preserved specimens like Aglaophyton major suggests that some of our concepts about early land plants with conducting tissues may be in need of modification. Now that gametophytes are known in at least some of the rhyniophytes, their structure and morphology may help to define broader whole-plant concepts leading to a more robust classification of many of these plants. Even in specimens lacking anatomy, some morphological characters suggest a bryophytic level of organization; it is important to keep in mind, however, that tissue systems, like various organs (e.g., leaves and roots) were evolving as plants adapted to a new, terrestrial environment. The fossil record suggests that conducting elements may have evolved in more than one group and at different times during the early colonization of the earth (Taylor, 1986; Kenrick et al., 1991a). Not all of these elements were true tracheids, but they apparently functioned like tracheids in that they not only had to provide support (mechanical stability), but also function in translocation, an idea suggested many years ago by F. O. Bower and later by H. P. Banks. Some of these plants may represent the ancestral stock of certain bryophyte lineages and others may be true vascular plants, but the majority may simply represent failed attempts in the colonization of the land. As additional specimens are discovered and new information evaluated, there are certain to be additional modifications and refinements of our understanding of these plants. Rhyniophytes are certainly some of the simplest, upright land plants, but are they the oldest? As you will see later, zosterophyll and even lycopsid megafossils are now known from the Upper Silurian in some diversity. At the


Paleobotany: the biology and evolution of fossil plants

present time, the oldest megafossil evidence demonstrates parallel evolution of rhyniophytes and zosterophylls; unfortunately, the spore record currently provides few characters that would enable us to distinguish these early plants from one another, as all contain simple trilete spores. In addition, few are known with spores in situ, which would provide for correlation of sporae dispersae with parent plants and demonstrate plant diversity prior to the Late Silurian.







ZOSTEROPHYLLOPHYTES The zosterophyllophytes (or zosterophylls) are the second major group of vascular plants established by Banks (1975), as the Zosterophyllophytina. They range from Late Silurian to Late Devonian and represent some of the most interesting early vascular plants, in part because there is a considerable amount of anatomical and morphological detail known about them. They demonstrate diversity as early at the Ludlovian (Late Silurian) (Kotyk et al., 2002) and were widespread geographically by the Early Devonian, including Gondwana (Zhu and Kenrick, 1999). As a group they share many features with the Lycophyta and may have given rise to the lycopsids; others interpret them as paraphyletic (Crane, 1990), or as the sister group to the lycopsids (Gensel, 1992). Most zosterophylls exhibit dichotomous branching, although in some genera there is a tendency toward a pseudomonopodial habit; the branching is generally planar (Kenrick and Crane, 1997a) and ultimate branches exhibit circinate vernation (development by means of uncoiling). When specimens are found permineralized, the exarch protostele is more robust than in the rhyniophytes and often elliptical in transverse section. The synapomorphy that distinguishes zosterophyllophytes from other early vascular plants is the presence of sporangia that are borne laterally along the stem (FIG. 8.70); they may either be sessile or attached by short branches. In many taxa, sporangia are aggregated often into terminal clusters or cone-like structures. Sporangial shape varies from globose to reniform, with dehiscence typically occurring along the distal edge and separating the sporangium into two valves. Most zosterophyll sporangia also exhibit a thickened zone bordering the dehiscence line. All zosterophyllophytes are homosporous, although the size range of the spores can be rather extensive in some Devonian forms. In other plants that possess zosterophyllophyte features (e.g., Barinophyton and Protobarinophyton), each sporangium contains both large and small spores (Taylor and Brauer, 1983; Chapter 9). One plant that has an interesting taxonomic history and is now included among the zosterophyllophytes is Sawdonia. As mentioned earlier, Dawson described a plant from the





Figure 8.70 Morphology and arrangement of zosterophyll sporangia. A,H. Gosslingia breconensis; B,I. Zosterophyllum myretonianum; C, J. Zosterophyllum fertile; D. Zosterophyllum spectabile; E. Sawdonia acanthotheca; F. Konioria andrychoviensis; G. Psilophyton princeps (Devonian). (From Kenrick and Edwards, 1988b.)

Devonian of the Gaspé Peninsula in 1859 as Psilophyton princeps. In his initial reconstruction, Dawson illustrated a dichotomizing plant that terminated in straight or circinately curved branches. He later modified his interpretation of P. princeps and described the reproductive units as sporangia borne at the ends of slender branches (Dawson, 1870). This was followed a year later by an emended diagnosis based on additional specimens, some naked and some with helically arranged spines (Dawson, 1871). At this time, the spiny fossils were referred to the taxon P. princeps var. ornatum. Subsequent workers were able to identify globose sporangia in lateral positions on the axes of some specimens of P. princeps var. ornatum. The literature at the time now contained a contradiction relative to Psilophyton. The P. princeps plant originally described by Dawson consisted of spiny axes with terminal sporangia, and P. princeps variety ornatum, which was regarded as the type of the species, represented a very different plant with lateral sporangia. Today we know that Dawson combined two



Deheubarthia splendens axis with axillary tubercle branch (Devonian). (From Kenrick and Crane, 1997a.) Figure 8.72

Figure 8.71 Suggested reconstruction of Deheubarthia splend-

ens (Devonian). (From Kenrick and Crane. 1997a.)

separate plants in his early reconstruction. Psilophyton princeps is used for plants with naked or spiny axes and terminal sporangia. Those specimens with spiny axes and lateral, globose sporangia (Fig. 8.70E) are referred to as Sawdonia ornata (Hueber, 1971). A detailed account of the historical development of the problem and its solution can be found in Banks et al. (1975). It may be appropriate that Dawson’s original description and reconstruction of Psilophyton, which was ignored for so many years, involved numerous researchers and took more than a century to fully understand! Sawdonia ornata is thought to have been 30.0 cm tall and constructed of pseudomonopodial axes that arose from a rhizome. Lateral axes branch dichotomously and are characterized by circinate tips. Branches are covered by numerous tapered spines, which typically show dark tips (Edwards et al., 1989). The sporangia seem to be confined to the distal ends of branches, where they form loosely aggregated spikes. Individual sporangia are reniform and borne on short

stalks in two vertical rows. Dehiscence occurred along the convex margin dividing the sporangium into two equal valves. Sawdonia ornata apparently was a homosporous plant with sporangia containing round–subtriangular trilete spores, each 64 μm in diameter. In structurally preserved specimens, the stele has a solid core of conducting elements with annular secondary wall thickenings that exhibit interconnecting bars suggestive of G-type cells (Rayner, 1983). Epidermal cells are unusual in S. ornata. Some bear papillae, whereas others consist of a central cell surrounded by elongate radiating cells that may represent some type of hair base (Wang and Hao, 1996). Similar cells have been noted in Oricilla, another Early Devonian zosterophyllophyte (Gensel, 1982a). Their function is not known, but secretion, aeration, and storage functions have been suggested (Gensel and Andrews, 1984). Stomata are also present on the epidermis of the stem of Sawdonia, but not on the spines. Deheubarthia (FIG. 8.71) has been proposed for certain Early Devonian specimens previously placed in Sawdonia (Edwards et al., 1989). Specimens of D. splendens consist of spiny axes 30.0 cm tall organized in a planar, pseudomonopodial branching system. Spines without dark tips and the occurrence of subaxillary branches (FIG. 8.72)


Paleobotany: the biology and evolution of fossil plants

distinguish Deheubarthia from Sawdonia. Conducting cells are of the G-type and some epidermal cells possess papillae. Another plant that closely resembles Sawdonia is Discalis longistipa (FIG. 8.73), a zosterophyll from the Lower Devonian of Yunnan, China (Hao, 1989). This plant had H- and K-shaped branching with fertile axes bearing large (3.7 mm in diameter) sporangia organized in loose spikes. Multicellular spines occur on all axes, as well as on the sporangia. Based on the presence of radial, trilete spores 30–50 μm in diameter, D. longistipa is thought to have been homosporous. Kidston and Lang (1923) described a plant from the Devonian Old Red Sandstone Series that they characterized as having a tufted growth habit with predominantly dichotomous branches. Hicklingia attained a height of 17.0 cm and is known from compressed specimens. Because the

sporangia appeared to have been terminal, Kidston and Lang regarded it as similar to Rhynia, and subsequent authors placed the genus among the rhyniophytes. A reexamination of the original specimen by Edwards (1976) indicates that the sporangia were, in fact, borne laterally on the axes, suggesting assignment to the Zosterophyllophyta. Spores are up to 50 μm in diameter and trilete. The Early Devonian genus Zosterophyllum includes a number of species. Zosterophyllum was a leafless and smooth, dichotomously branched plant that produced lateral sporangia on short, delicate stalks. In permineralized specimens of Z. llanoveranum from the Lower Devonian of Britain, the axes are 1.5 mm in diameter and contain an elliptical strand of scalariform conducting elements (Edwards, 1969). Surrounding the stele is a cortex of three zones distinguished by differences in cell size and cell-wall thickness. Sporangia

Figure 8.73 Suggested reconstruction of Discalis longistipa. Inset shows portion of fertile axis with spines. (From Taylor and Taylor, 1993.)



occur in either one or two rows at the distal ends of branches; they vary from circular to reniform (FIG. 8.74), and each is borne on a small stalk that departs from the axis at an abrupt angle. Vascular tissue has not been identified in the sporangial stalk. On either side of the distal line of dehiscence is a band of elongated, thick-walled cells, grading proximally into smaller, thinner-walled cells proximally. Spores are ovoid and average 45 μm in diameter. In Z. ramosum the fertile axes branch many times, each terminating in a spike of 8–15 sporangia; trilete spores are triangular with long laesurae (Hao and Wang, 2000). Zosterophyll features are present in Macivera gracilis from the Upper Silurian (Ludlovian) of Bathurst Island, Canadian Arctic (Kotyk et al., 2002). In this leafless plant the sporangia are sessile, borne in small clusters at the distal ends of axes, and are not in rows. Hueber (1972) suggested that Zosterophyllum subgenus Platyzosterophyllum be used to accommodate species with sporangia in two rows (thus forming dorsiventral spikes) and the subgenus Zosterophyllum be reserved for species with helically arranged sporangia. In several species, such as Z. myretonianum (FIG. 8.70B,I, 8.75) and Z. divaricatum

(FIG. 8.76), the basal region of the plant is characterized by K- and H-type branching patterns. These branching patterns are apparently the result of successive, close order dichotomies. Branching of this type has been described in a number of zosterophylls and is also known to have occurred in some species of the drepanophycalean genus Drepanophycus (Chapter 9). In Z. divaricatum the fertile axes have circinately coiled apices, and the reniform sporangia contain smooth, trilete spores 50–90 μm in diameter (Gensel, 1982b). In Z. deciduum, a compression form from the Lower Devonian (Emsian) of southern Belgium (Gerrienne, 1988), sporangia appear to have been shed at maturity. Recurved branch tips are also present on the Early Devonian (Gedinnian of Germany) Anisophyton (Remy et al., 1986). Spines with truncated apices cover the axes and the sporangia are typically borne on only one side of the axis. Another Early Devonian plant that has features of the zosterophylls is Guangnania cuneata (FIG. 8.77) (D.-M. Wang and Hao, 2002; D.-M. Wang et al., 2002) from Yunnan Province in southwestern China. Elongate sporangia are borne on upright stalks, are slightly curved, and dehisce into two, unequal valves.

Figure 8.74 Structure and morphology of Zosterophyllum

Figure 8.75 Suggested reconstruction of Zosterophyllum myretonianum (Devonian). (From Taylor and Taylor, 1993.)

llanoveranum sporangia (Devonian). (From Edwards, 1969.)


Paleobotany: the biology and evolution of fossil plants

Figure 8.76 Suggested reconstruction of Zosterophyllum divaricatum (Devonian). (From Kenrick and Crane, 1997a.)

Lateral stalked sporangia are also found in Danziella artesiana from the Lower Devonian of France (Edwards, 2006). In Serrulacaulis furcatus (FIG. 8.78) (Upper Devonian), axes are ornamented by two rows of opposite emergences (Hueber and Banks, 1979). The triangular shape of the projections gives the axes a sawtooth appearance. These are now interpreted as triangular, prism-shaped structures that are arranged in two rows in a steplike manner on opposite sides of the axes (Berry and Edwards, 1994). Some axes bear stalked sporangia between the rows of projections on one side of the axis. Sporangia are reniform and borne on short stalks; dehiscence splits the sporangia into two unequal halves. Spores are trilete and 60 μm in diameter. Conducting cells are of the G-type (Berry and Edwards, 1994). Gosslingia breconensis is a zosterophyll that was 50.0 cm tall and is known from the Lower Devonian

Figure 8.77 Suggested reconstruction of Guangnania cuneata

fertile axis. (From D.-M. Wang and Hao, 2002.)

of Wales (Edwards, 1970b). It consists of dichotomizing axes which are up to 4.0 mm wide distally. The base is presumed to have consisted of a rhizome bearing rhizoids, although organic attachment has not been demonstrated. Like Sawdonia, Gosslingia exhibits distal tips that are circinately coiled. The aerial stems are leafless, although some specimens show small protuberances that extend a few hundred micrometers from the surface and larger tubercles



Figure 8.78 Suggested reconstruction of Serrulacaulis furca-

tus (Devonian). (From Taylor and Taylor, 1993.) Figure 8.79 Suggested reconstruction of Thrinkophyton for-

mosum (Devonian). (From Kenrick and Crane, 1997a.)

that are often identified only by their scars. These latter structures, which have been termed axillary tubercles, contain conducting elements in the form of a terete trace and extend from one side of the stem just below the point of a dichotomy. Pyritized axes reveal that the conducting strand is elliptical in outline, with annular protoxylem elements located near the periphery. Conducting elements are of the G-type. The cortex is constructed of thick-walled cells and what appear to be stomata are scattered along the stem. In iron sulfide permineralizations of G. breconensis, all the conducting elements are believed to be metaxylem tracheids, the protoxylem being either lost during fossilization or at some stage in the ontogeny of the stele (Kenrick and Edwards, 1988a). The presence of pyrite in various forms is discussed in relation to the loss of cellulose and lignin during fossilization and the formation of artifacts in the cells. This type of detailed study represents an important starting point in the accurate characterization of the vascular elements and supporting cells in all early land

plants. The sporangia in Gosslingia occur in definite aggregations in the distal regions of the plant. They are variable in shape, ranging from globose to reniform (Fig. 8.70A,H), and are attached to the stems by slender stalks. Little detailed information is known about the histology of the sporangial wall, although some spores have been recovered. These range from 36–50 μm in diameter and are ornamented by small spines. Tarella is an interesting zosterophyll in which the stalked, reniform sporangia split into two equal valves much like those of Gosslingia (Edwards and Kenrick, 1986). The sporangia occur in opposite rows and are not present near the circinate stem tips. Projections, some with hooked apices, are present on both the fertile and sterile axes. Spores are trilete and 40 μm in diameter. Another zosterophyll from the Lower Devonian of Wales is Thrinkophyton formosum (FIG. 8.79) (Kenrick and Edwards, 1988b). It has pseudomonopodial and isotomous branching and is 9.0 cm tall. The stem tips are circinate with projections borne just below a branch dichotomy. Sporangia


Paleobotany: the biology and evolution of fossil plants

are arranged in one or two rows arising from short stalks, and are reniform. The central strand is described as consisting of xylem elements with annular to helical thickenings; xylem maturation is exarch. Konioria is an Early Devonian plant from Poland that has branch tips curved to form hooklike ends (Zdebska, 1982). On the surface of the axes are irregularly positioned spines of various lengths, the longest ones (4.0 mm) bearing delicate teeth. The stems also contain winglike outgrowths that are extended longitudinally. In the center of the axis is an exarch protostele. Sporangia were produced on stalks that extend at right angles from the stem just below the last dichotomy. They are typically 2.5 mm wide and covered by minute spines. To date nothing is known about the spores. Crenaticaulis is a zosterophyll that is known in some detail and thus provides important information about the conducting system and sporangia of this group. Crenaticaulis verruculosus was described from both compressed and structurally preserved specimens collected from Lower Devonian rocks in the Gaspé region (Banks and Davis, 1969). The largest specimen consists of a 22.0-cm-long axis that shows both pseudomonopodial and dichotomous branching. The dichotomies occur at short intervals, and the distal stem tips are slightly coiled. An unusual feature of C. verruculosus is the presence of two rows of multicellular, toothlike protuberances that are nearly oppositely arranged along the surface of the stems. These teeth are triangular and present on the circinately coiled stem apices as well as on the stalks that bear the sporangia. Epidermal cells of the stem are of two types, elongate and papillate. On some specimens, subaxillary tubercles are present; on others their position is indicated by an elliptical scar. The exarch strand is elliptical in cross section and composed of G-type conducting elements. The sporangia of C. verruculosus are clustered and occur in opposite–subopposite groups on the distal branches. They are pedicellate and nearly spherical in outline. Sporangial dehiscence has been termed distal—beginning on one side just above the attachment to the stalk and arching over the adaxial face to the opposite surface. Dehisced sporangia consist of a large abaxial and a small adaxial segment; nothing is known about the spores. Rebuchia is an Early Devonian plant from the Beartooth Formation of Wyoming (Hueber, 1972), with naked, dichotomously branched axes that gradually taper into blunt points. Sporangia are confined to distal branches and occur as spikes of up to 20 sporangia (FIG. 8.80). Individual sporangia are arranged oppositely to suboppositely and borne on short, curved stalks, so that all the sporangia point are essentially in the same direction. Dehisced sporangia have equal valves,

Figure 8.80 Suggested reconstruction of Rebuchia ovata

(Devonian). (From Taylor and Taylor, 1993.)

indicating a basipetal form of dehiscence. Rebuchia ovata was probably homosporous; all the spores recovered are unornamented, 68–75 μm in diameter, and of the Retusotriletes type. Another plant that has aggregations of sporangia in what might be termed strobili is Bracteophyton variatum from the Lower Devonian of China (Wang and Hao, 2004). Individual sporangia are adaxial and associated with pairs of bracts, or with a single bract that bifurcates at the tip. Sporangial dehiscence is distal. Axes are isotomous, naked and up to 4.7 mm in diameter. Nothing is known about the anatomy. Kaulangiophyton is an Early Devonian plant consisting of a horizontal branching system that bears numerous, irregularly spaced, short spines (Gensel et al., 1969). The spines are slightly curved and up to 2.0 mm long, with decurrent bases. Sporangia are borne on short stalks along the stem and appear to be interspersed among the spines, giving this


plant the appearance of modern species of Lycopodium. No spores have been recovered, nor is anything known about the vascular system. In many features, K. akantha is similar to Drepanophycus (Chapter 9). ZOSTEROPHYLL EVOLUTION

There is convincing evidence that supports the lycopsids as most closely related to the zosterophylls. Both groups are characterized by exarch protosteles and laterally produced reniform sporangia. In Crenaticaulis and Gosslingia, axillary tubercles are thought to represent rhizophore-like branches, similar to those of extant species of Selaginella. The absence of leaves in the zosterophylls, although an obvious difference between the two groups, can be explained in terms of the transitional status of the group. Although they do not have true leaves, the zosterophylls are known to bear various types of laterals. These range from unicellular to multicellular spines to multicellular teeth; the latter may have functioned to increase the photosynthetic surface of the stems. Generally, these appendages were randomly scattered over the stem surfaces, although in Crenaticaulis the large, toothlike outgrowths are arranged in two rows along the stems. In the putative early lycopsid Asteroxylon, traces extend through the cortex but do not enter the base of the enations. This tendency toward a definite arrangement and vascularization may constitute the initial stages in the subsequent evolution of the microphyll (Chapter 9). The zosterophylls also demonstrate several stages in the evolution and organization of sporangia. These include forms such as Kaulangiophyton, in which sporangia are apparently helically arranged over the stem surface, to intermediate forms such as Gosslingia, in which the helically arranged sporangia are aggregated into definite spikes. A further modification might result in the arrangement seen in Rebuchia, where sporangia are organized into definite spikes, with the individual sporangia dorsiventral in arrangement (present on only one side of the axis). Niklas and Banks (1990) suggested that the zosterophylls can be separated into two groups based on the presence or absence of terminally located fertile units, as well as on the symmetry of the sporangial aggregation. These two groups can also be distinguished based on the presence of small flaps of tissue along the stems (enations), circinate tips on the axes, and, when preserved, the nature of the conducting strand. Based on their analysis, Niklas and Banks suggest that the lycopsids arose from a zosterophyllophyte-like group of plants, perhaps with a level of organization similar to Asteroxylon or Drepanophycus (Chapter 9). Sporangial dehiscence within the zosterophyllophytes typically divides the sporangium into valves of equal or nearly equal size. As additional zosterophyllophyte taxa are


described, a new perspective is emerging as to the diversity and spatial distribution of these interesting Paleozoic plants (Hao and Gensel, 2001; Kotyk et al., 2002; Wang and Hao, 2002). New taxa are providing an increasing data set of sporangial characters relating to dehiscence (e.g., in Crenaticaulis), where dehiscence results in two unequal valves. In addition, the discovery of new taxa is expanding our understanding of the diversity of sporangial aggregation, position and length of stalks, and pattern and orientation on the fertile axes. These characters may become more important as additional information on features of the fertile parts of these interesting plants become more fully known, and will no doubt result in an increased level of resolution relating to the phylogenetic position of these plants.

TRIMEROPHYTES The third major group that was culled from the original Psilophytales by Banks is the Trimerophytophyta. Trimerophytes1 were generally more complex than either the rhyniophytes (from which they are thought to have descended) or the zosterophyllophytes. Trimerophytes demonstrate monopodial branching of the main axes, with lateral axes showing either dichotomous or trifurcate branching. As with rhyniophytes, the sporangia are terminal, although typically they are fusiform to elongate and aggregated or clustered on fertile branches. Internally, the members of this group also exhibit greater complexity, in the form of a relatively large (compared to stem diameter), centrarch stele. Tracheid wall patterns vary from scalariform-bordered to circular-bordered pitting. Trimerophytes are considered to have given rise to all the other vascular plants, with the exception of the lycopsids. As indicated in the preceding section, Psilophyton specimens with lateral sporangia were transferred to the zosterophyllophyte genus Sawdonia. The remaining Psilophyton specimens, that is, those with terminal sporangia, are retained in the Trimerophytophyta. Psilophyton dawsonii is one of the most completely known members of the group. Both compressed and structurally preserved specimens have been described from several Lower Devonian localities (Banks et al., 1975). A reconstruction of P. dawsonii shows a highly branched plant in which the fertile lateral branches are borne alternately and distichously


Based on the scientific name of this group, the common name should probably be the trimerophytophytes. However, for the sake of simplicity, we have chosen to use trimerophytes.


Paleobotany: the biology and evolution of fossil plants

(in two opposite, vertical rows) along the main axes. Vegetative branches are naked and dichotomize at right angles, terminating in slender, blunt tips. Fertile branches typically branch up to six times before terminating in clusters of 32 sporangia. Individual sporangia are 5.0 mm long, with dehiscence occurring along the lateral surface. The trilete spores vary from 40 to 75 μm in diameter with a smooth spore wall, although in some specimens, a thin layer often appears separated from the spore body. Fine structural details of the spores have been provided by Edwards et al. (1996). If found as dispersed grains, such spores would most closely approximate the sporae dispersae taxa Retusotriletes or Phyllothecotriletes. The conducting tissue of P. dawsonii consists of a centrarch protostele that accounts for approximately one-quarter of the stem diameter. In more basal regions of the plant, the strand has several enlarged protoxylem zones and numerous, radially aligned conducting elements of the P-type (Kenrick and Crane, 1997a). They have bordered pits with secondary wall material that extends across the pit apertures (Hartman and Banks, 1980). Traces supplying the fertile branches are initially terete, becoming more rectangular in outline at higher levels. Surrounding the central strand is a multilayered cortex consisting of collenchymalike cells and, toward the periphery, substomatal chambers. In P. princeps, the stem surfaces are ornamented by a number of cup-tipped spines. The larger size of the sporangia and the nature of the vegetative branching further distinguish this species from P. dawsonii. The conducting strand of P. princeps consists of a solid strand of elements that is either mesarch or centrarch. Psilophyton crenulatum (FIG. 8.81) a species from the Lower–Middle Devonian of New Brunswick, Canada, has spiny branches that dichotomize several times and terminate in slender, recurved tips. On the surface of the axes are multicellular spines up to 6.0 mm long; some have forked tips, whereas others are trifurcate. The vascular strand is described as centrarch and constructed of elements with scalariform–circular bordered pits. The fertile units (clusters of sporangia) in P. crenulatum (FIG. 8.81) (Doran, 1980) are alternate and distichous, or may be helically arranged. In the distal regions they are covered by semicircular crenulations. The fusiform sporangia are up to 5.0 mm long and twisted; spores range from 48 to 102 μm in diameter and conform to the genus Apiculiretusispora. Short recurved lateral branches and elongate pairs of sporangia that are of unequal length are characteristic of Aarabia, an early Emsian (Early Devonian) plant from Morocco (Meyer-Berthaud and Gerrienne, 2001). Psilophyton forbesii was one of the largest species of Psilophyton, estimated as 60.0 cm tall (FIG. 8.82) (Gensel, 1979). The growth habit was either monopodial or

Figure 8.81 Psilophyton crenulatum. Bar  2 mm. (Courtesy

J. B. Doran.)

pseudomonopodial and the naked stems were marked by longitudinal striations. Ellipsoidal sporangia were produced in pairs on fertile lateral branches and spores ranged 53–96 μm in diameter. In P. dapsile (FIG. 8.83) (Kasper et al., 1974), known from the Middle Devonian of Maine, the axes measure 2.0 mm in width and the erect plant is thought to have been about 30.0 cm tall. The stems are smooth and branched dichotomously. Sporangia are small (2.0 mm long) and borne in dense clusters at the ends of closely spaced, distal dichotomies. One species of Psilophyton, P. hedei, which was described from Silurian rocks of Sweden, is now thought to represent some type of invertebrate similar to a graptolite or a pterobranch. In this instance, the branched axis may represent a colony of numerous individuals that share a system of internal tubes and the characteristic stem spines probably represent free zooidal tubes (Lundblad, 1972).



Figure 8.83 Suggested reconstruction of Psilophyton dapsile

(Devonian). (From Taylor and Taylor, 1993.)

Figure 8.82 Suggested reconstruction of Psilophyton forbesii (Devonian). (From Taylor and Taylor, 1993.)

Pertica quadrifaria (FIG. 8.84) is another Devonian plant that can be assigned to the trimerophytes on the basis of its branching pattern and the nature of its fertile region (Kasper and Andrews, 1972). The genus is known from compression remains and has been described from at least two localities of Early Devonian age. The size of some specimens suggests that the plant may have exceeded 1.0 m in height. Numerous, dichotomously branched laterals are borne on the main axis in a tetrastichous pattern (i.e., in four rows). The distal ends repeatedly dichotomize at right angles to each other. Like all trimerophytes, the lateral branches were either completely sterile or completely fertile. The fertile ones consist of closely spaced dichotomies that terminate in masses of sporangia borne on short stalks. Sporangia are elliptical in outline and lack any histologic evidence of a dehiscence mechanism. Poorly preserved spores that are about 64 μm in diameter suggest that P. quadrifaria was homosporous. Pertica varia was much larger than P. quadrifaria, reaching a height of nearly 3.0 m (Granoff et al., 1976). The primary

branches of P. varia are arranged in subopposite pairs (with the successive laterals decussate. Paired sporangia are erect, but the number of sporangia per cluster is smaller than in P. quadrifaria. Spores of P. varia are subcircular–subtriangular and 90 μm in diameter. In P. dalhousii the lateral branches are spirally arranged and divided pseudomonopodially (Doran et al., 1978). Trimerophyton robustius, the type species of the group, was initially described by Dawson (1859) as Psilophyton robustius from some fragmentary specimens collected from the shore of Gaspé Bay. The generic name Trimerophyton was introduced many years later on the basis of a single specimen, also from the Gaspé (Hopping, 1956). The main stem of T. robustius is 1.0 cm wide and consists of numerous helically arranged, trifurcate, lateral branches. The primary and secondary branching patterns of the laterals are trichotomous, with further subdivisions of the dichotomous type. All axes are smooth, with the exception of some that are longitudinally striated. Ultimate branches bear erect sporangia in clusters of three (FIG. 8.85). It has been suggested that Trimerophyton may represent the distal parts of Pertica (Gensel and Andrews, 1984).


Paleobotany: the biology and evolution of fossil plants

Figure 8.85 Suggested reconstruction if Trimerophyton robus-

tius (Devonian). (From Taylor and Taylor, 1993.)

Figure 8.84 Branch of Pertica quadrifaria (Devonian).

Bar  2.0 cm.

Yunia is a Siegenian (Pragian, Lower Devonian) plant with spiny axes that shares some features with Pertica (Hao and Beck, 1991). The axes dichotomize in a cruciate arrangement and each contains a conducting strand with one or two protoxylem strands, depending on the level of the axis. Conducting elements have annual scalariform sculpture patterns. Closely associated sporangia are elongate and bear trilete spores with a relatively smooth sculptural pattern. Dawsonites is a Devonian morphogenus that has been used by some workers for sporangia that are not organically connected to an axis or aggregated into large clusters (Halle, 1916b). The sporangia are fusiform and typically measure 3.0–5.0 mm long. The genus is now restricted to terminal sporangia that are believed to have been produced by Psilophyton. As spores are recovered from additional fertile specimens so that comparisons can be made, many of the Dawsonites species no doubt will be placed in synonymy. Another morphogenus that has been used for naked, dichotomously branched axes of Siluro–Devonian age is Hostinella. Structurally preserved specimens consist of

small fragments up to 3.0 mm wide. The conducting strand is protostelic, centrarch, and composed of tracheids with scalariform to bordered pits. Some specimens of Hostinella are thought to represent an isolated branching fragment of Margophyton (Psilophyton) goldschmidtii, now believed to be a Lower Devonian member of the zosterophylls (Zakharova, 1981). A Late Silurian (Prˇídolí) specimen from New York State includes preserved apices and a rootlike structure (Edwards et al., 2004). TRIMEROPHYTE EVOLUTION

The trimerophytes demonstrate more complex morphology and anatomy than rhyniophytes, their presumed ancestors, although both groups are coeval. In the trimerophytes, plant architecture is monopodial or pseudomonopodial. Laterals are produced in a variety of patterns, including helical (Psilophyton sterile branches), alternately and distichous (Psilophyton fertile branches), tristichous (Trimerophyton), and tetrastichous (Pertica). In Pertica the ultimate branchlets consist of slender, three-dimensional dichotomizing structures. It has been suggested that the planation of these lateral branches would provide the morphologic equivalent of a megaphyllous leaf and that Pertica may be used as a transitional morphotype in the evolution of a frond or a leaf. In another group of Devonian plants, the Aneurophytales


(progymnosperms), some taxa possess planated laterals, whereas in others the branching systems are more three dimensional (Chapter 12). Trimerophytes also demonstrate various patterns of sporangial attachment. In rhyniophytes, sporangia are terminal at the ends of dichotomizing axes. In some species of Psilophyton the number of sporangia is small, while in others (e.g., P. dapsile) numerous small sporangia are clustered together. One possible transformational series might involve Pertica, with its massive clusters of densely packed sporangia, leading to some Carboniferous ferns, such as some species of Botryopteris. Another line might lead to the progymnosperms through such a plant as Tetraxylopteris. One Middle Devonian plant that might bridge the evolutionary gap between trimerophytes and progymnosperms is Oocampsa (Andrews et al., 1975). Specimens of O. catheta consist of closely spaced, helically arranged branches up to 7.0 cm long that were produced from a primary axis (FIG. 8.86). The lateral branches divide pseudomonopodially and dichotomously and terminate in elongate, erect sporangia. Sporangia dehisce longitudinally and contain


large (96–120 μm), trilete miospores. The spores are interesting in that there appears to be some space between the wall layers, suggesting a pseudosaccate morphologic type. Oocampsa, with erect sporangia borne on helically arranged primary and secondary branches, may be transitional between certain trimerophytes, for example, Trimerophyton and Pertica, and Tetraxylopteris, a progymnosperm with pinnate arrangement of ultimate segments. Such a series is congruent with the stratigraphic occurrences of the taxa listed. There are relatively few trimerophytes that are structurally preserved. In Psilophyton dawsonii and P. princeps, there is a simple conducting strand, but one that is more massive than any known for the rhyniophytes. Conducting elements in the rhyniophytes are of the S-type, while Psilophyton contains P-type elements in which secondary wall material is deposited between the scalariform bars. Although the trimerophytes appear to be less diverse than other early land plants, there are some apparent evolutionary trends within the group, including stages in the evolution of a particular type of leaf, modification of conducting element pitting toward the circular-bordered type, and some suggestion of an early stage in the evolution of spores with an increased surface area as a result of the separation of wall layers. As additional specimens are discovered and described from Devonian or possibly even Mississippian rocks, it is obvious that this group will play an increasingly important role in our understanding of levels of specialization in early land plants and their role in the diversification of laterappearing groups.


Figure 8.86 Suggested reconstruction of Oocampsa catheta (Devonian). (From Taylor and Taylor, 1993.)

During the evolution of land plants, the sporophyte generation has become increasingly more complex, both physiologically and morphologically. The gametophytes from the Rhynie chert demonstrate that during the Early Devonian gametophytes and sporophytes were more similar to each other than these generations are today. Some authors working with extant plant development suggest that the Rhynie chert gametophytes provide evidence of developmental genes being reassembled from the gametophyte phase and incorporated into the developmental pathways of the sporophyte (Floyd and Bowman, 2007). The comparison of developmental mechanisms based on various gene families in bryophytes, lycopsids, and seed plant lineages has already revealed some commonalities (e.g., among the MADS-box genes), as well as differences related to various plant structures. These morphological expressions can then be compared to structures


Paleobotany: the biology and evolution of fossil plants

that are present in fossils. This is another way that evidence from the fossil record can be used in understanding the morphological evolution that has led to the complexities in multicellular plants that dominate the Earth today. While cladistic analyses suggest that the land plants are monophyletic based on the presence of multiple sporangia (the polysporangiophytes), the evidence from the fossil record may be interpreted in several ways. The conducting elements in many of these early plants have been interpreted as representing a protracheophyte grade of evolution. The diversity of these elements, however, brings up the question of whether or not conducting elements evolved more than once as a response to the selective pressures of a desiccating environment. Researchers on extant algae are still debating the relationships of the charophycean algae and which group is most closely related to land plants. Many systematists that work only with extant plants believe the fossil record is too

incomplete to provide any information on the origin of land plants, but new discoveries continue to provide more data. Based on the diversity of conducting elements, the fossil record could be interpreted to suggest that land plants (hornworts, bryophytes, and vascular plants) became established in terrestrial environments several times and perhaps originated from a variety of charophycean green algae. The fossil evidence for the lycopsid lineage shows that these plants were clearly differentiated from the other vascular plants (and remarkably modern!) as early as the Late Silurian (Chapter 9). Information on the earliest land plants and the earliest vascular plants continues to be amassed. Regardless of the degree of interrelatedness of these early land plants, how and when the necessary steps took place in the transition from an aquatic to terrestrial habitat will continue to be challenging questions for paleobotanists and neontologists alike.

9 Lycophyta Evolution of the microphyll ...................................267


Drepanophycales ................................................................. 268

Selaginellales.......................................................................... 312

Protolepidodendrales ................................................... 271

Pleuromeiales ..........................................................................316

Lepidodendrales ...................................................................279

Isoetales ...................................................................................... 320

Vegetative Features............................................................................282

Putative lycopsids ...............................................................325

Reproductive Biology .......................................................................294 Conclusions ........................................................................... 326

Sigillariaceae .....................................................................................303 Other Lepidodendrid Genera ............................................................307

And, again, it is an error to imagine that evolution signifies a constant tendency to increased perfection. Thomas Henry Huxley The lycopsids have an extensive geologic history, extending back into the Late Silurian (Kotyk et al., 2002). They were widespread during the Late Mississippian and most of the Pennsylvanian, representing the dominant group of plants in most of the vast Euramerican paleoequatorial coal-swamp ecosystems. The widespread coal-mining operations that have uncovered Carboniferous rocks have been responsible, in large part, for the abundant, well-preserved fossil specimens of this group. The paleoecology of these plants has been extensively studied, and many constitute the focal point of ancient landscape reconstructions in museums around the world. The comprehensive geologic history of the lycopsids and numerous, exquisitely preserved specimens have provided paleobotanists with the opportunity not only to trace the evolution of the group but also to investigate some basic facets of their biology. The Lycophyta are monophyletic and basal within the vascular plants (Bateman, 1996b); together with the Zosterophyllophyta (Chapter 8), they comprise a clade (Lycophytina of Kenrick and Crane, 1997a). Both lycopsids and zosterophylls occupied the same habitats (floodplain and channel margins) during the Devonian (Gensel, 1992). The common names of this group have become somewhat confusing in recent years. Among

paleobotanists, lycopods has been the traditional name for all of the Lycophyta (without the zosterophylls), and you will find this name in the older literature; some neontologists, however, use this name to refer only to members of the Lycopodiales. More recently, lycopsids has been widely used. The term lycophytes has been used to refer not only to the Lycophyta but also to the clade composed of the lycopsids and the zosterophylls together. Kenrick and Crane (1997a) classified the lycophytes as the subphylum (subdivision) Lycophytina, containing two classes, Lycopsida and Zosterophyllopsida. We will use the name lycopsids to refer to the plants discussed in this chapter. According to Gensel and Berry (2001), the lycophytes represent a broader concept, which includes traditional lycopsids as well as the zosterophylls and transitional taxa such as Asteroxylon and Drepanophycus. Synapomorphies of the lycopsids include helically arranged microphylls and sporangia borne in the axil or on the adaxial (upper) surface of sporophylls. Most have exarch primary xylem maturation and, like the zosterophylls, lycopsid sporangia dehisce into two valves. In addition, metaxylem tracheids contain Williamson striations (discussed below). Currently available evidence suggests that the lycopsids originated from the zosterophylls, although these groups were coeval in the Siluro–Devonian (Gensel and Berry, 2001;




Kotyk et al., 2002). From these beginnings, the lycopsids radiated extensively during the Carboniferous, in terms of both diversity and distribution, and then began to diminish in numbers of taxa and abundance toward the end of the Carboniferous, as the climate began to change and the extensive, equatorial peat swamps diminished (DiMichele et al., 2001a). Today the clade is represented by only seven genera in most treatments, of which three genera, Huperzia, Lycopodium (club mosses), and Selaginella (spike mosses), contain most of the 1600 species ( The group is characterized by either dichotomous branching or a combination of dichotomous and monopodial branching. Stems are densely covered with true leaves termed microphylls. Microphylls are generally small (except in some of the extinct arborescent lycopsids), helically arranged, and vascularized by a single bundle that does not produce an interruption (leaf gap) in the stele of the stem when it separates. The roots in the lycopsids, as in other vascular cryptogams, are primarily adventitious. One of the most diagnostic features of the group is the position of the sporangium, which is borne in the axil or on the upper surface of a modified leaf or sporophyll. In some members sporangia are borne on short stalks. Sporophylls may be interspersed among photosynthetic microphylls, or they may be non-photosynthetic and aggregated into loosely constructed strobili or more consolidated cones. In most species the sporophyte produces only one type of spore and these plants are therefore regarded as homosporous. In some lycopsids, however, two types of spores are produced, and they not only look different but also function differently in the life history of the plant. Small spores (microspores, developed in microsporangia on microsporophylls) produce the male gametophyte, whereas the mega- or macrospores (developed in megasporangia on megasporophylls) germinate to produce the female gametophyte. Living heterosporous members of the Lycophyta produce both types of spores in the same strobilus (bisporangiate); in the heterosporous fossil representatives, microspores and megaspores were produced either in the same or in different cones (monosporangiate). All extant lycopsids are herbaceous and do not produce secondary vascular tissue, although many fossil forms are known to have been arborescent. Despite the large size of some of the arborescent members (40 m tall), the amount of secondary vascular tissue is small compared with the total stem diameter. Maturation of the primary xylem is exarch in most lycopsids with scalariform wall thickenings, the most common type of secondary wall pattern on the tracheids. Vascular organization ranges from protostelic to siphonostelic.

Although leaf gaps are not produced, interruption in the vascular cylinder occurs when branch traces are produced. The classification of lycopsids, similar to many groups of fossil plants, is a difficult task, in part because of the large number of fossil taxa for which there is a limited amount of information, especially reproductive characters. Thomas and Brack-Hanes (1984) have suggested the formation of what they term satellite taxa to accommodate various plant organs that cannot be accurately placed in well-defined families, after the initial concept of satellite genera was proposed by Meyen (1978). Their classification is similar to the one used in this volume, with the exception that we prefer the order Lepidodendrales rather than the Lepidocarpales, and we continue to include Miadesmia within the Selaginellales. In some phylogenies, heterospory and the presence of a ligule are used to group certain taxa; in others herbaceous versus arborescent habit has been used to define hierarchy. DiMichele and Bateman (1996) included all the rhizomorphic lycopsids (Lepidodendrales, Isoetales) in a single order, the Isoetales. The rhizomorphic lycopsids are those with a stigmarian-type rooting system, which can be either laterally extensive or small and lobed (discussed below). As research continues with the lycopsids, it is increasingly clear that delineation into major clades is not easily resolved, and that transformation series leading to the origin of some modern forms that at one time seemed well defined, are today more difficult to resolve (Gensel and Berry, 2001). The following traditional classification of the lycopsids into seven orders is intended to provide a framework for discussion of this group of plants. As is true of most groups with fossil members, there are several enigmatic forms that do not fit precisely into this classification. Higher taxa in this chapter:

Lycophyta Lycopsida Drepanophycales (Devonian) Protolepidodendrales (Devonian–Mississippian) Lepidodendrales (Devonian–Permian) Lepidodendraceae, Diaphorodendraceae, Sigillariaceae Lycopodiales (Pennsylvanian–recent) Lycopodiaceae Selaginellales (Pennsylvanian–recent) Selaginellaceae Pleuromeiales (Triassic–Cretaceous) Isoetales (Upper Devonian–recent) Isoetaceae, Chaloneriaceae


Evolution of the microphyll One of the synapomorphies of the lycophytes is the presence of microphylls. Microphylls are true leaves, and as such, are borne in a definite pattern (phyllotaxy) on the stem, but they have an evolutionary history separate from the leaves of other vascular plants (megaphylls, see discussion in Chapter 11). Extant microphylls are small, although not all fossil microphylls are small. They usually contain a single vascular bundle (vein) and there is no leaf gap formed in the stele during the production of leaf traces (Chapter 7). Microphylls are believed to have evolved from enations, which as noted in the previous chapter are small, unvascularized flaps of tissue that do not have a phyllotaxy. There are several fossils that could be used to illustrate the stages in the evolution of microphylls. Beginning with a naked, vascularized axis (FIG. 9.1A), the first laterals produced were small, scattered spines, as seen in many zosterophylls, for example Sawdonia (FIG. 9.1B). Asteroxylon illustrates an intermediate stage (FIG. 9.1C), in which the laterals (enations) are more leaf-like, but vascular tissue extends only to the base of each enation. Finally, the laterals



(microphylls) become vascularized and are borne in a regular pattern on the axis, for example as seen in Leclercqia (FIG. 9.1D). There are several fossil plants that represent intermediates between stages B, C, and D, and some of these are included in the Drepanophycales. Kenrick and Crane (1997a) suggested several hypotheses on the origin of the lycophyte sporophyll, including reduction from a lateral branch (FIG. 9.2A), fusion of a sporangium to an enation– microphyll (FIG. 9.2B), and sterilization of a second sporangium (FIG. 9.2C). A study of stelar and microphyll vasculature in two extant Lycopodium species (Gola et al., 2007) confirms the unique nature of microphylls, especially when compared to megaphylls, the leaf type found in other vascular plants (see section “Evolution of Megaphyll” in Chapter 11). This study reveals that the production of microphylls and the architecture of the stele are not closely connected developmentally, as they are in seed plants. Based on these data, the authors concluded that the origin of the microphyll and the development of vasculature to supply the leaf in the lycopsids may have evolved independently, a theory which is supported by many of the plants discussed in the next section.






(D) (C)

Figure 9.1 Hypothesized stages in the evolution of the microphyll.

A. Vascularized axis (shaded area) lacking epidermal appendages (e.g., Rhynia). B. Axis with epidermal appendages in the form of spines or enations (e.g., Sawdonia). C. Axis with vascular traces extending to the base of the enation (e.g., Asteroxylon). D. Axis with vascularized microphylls (e.g., Leclercqia). (From Taylor and Taylor, 1993.)

Suggested transformational stages in the evolution of the microphyll with sporangium in lycopsids. A. Reduction from a lateral branch system. B. Origin from a sterile stem appendage (enation theory). C. Origin from a sterilized sporangium. (Modified from Kenrick and Crane, 1997a.)

Figure 9.2



Drepanophycales This group includes the oldest known lycophytes (Banks, 1960). In some treatments these Devonian plants are regarded as prelycopsids and classified based on their stratigraphic position and lack of some typical lycopsid characters (Gensel and Andrews, 1984), whereas in others they are considered as transitional between the zosterophylls and the true lycopsids. They represent a diverse collection of presumably herbaceous plants, although anatomical details and complete specimens are not known for all taxa. The upright axes have exarch primary xylem maturation and bear helically arranged appendages that sometimes appear to be in a near-whorled or pseudowhorled pattern. These appendages are either non-vascularized or partially vascularized, that is, the vein runs into the proximal portion of the appendage, but does not extend through the entire structure, so in this way they are similar to microphylls, rather than being true leaves. The appendages do not dichotomize at their tips as do microphylls in the Protolepidodendrales, but this may not be a sufficiently well-defined character, as it can vary with preservation. Fertile specimens suggest that sporangia are not aggregated into cones. In contrast to the true lycopsids, the sporangia of most Drepanophycales are not borne adaxially on sporophylls, but rather arise directly from the axis shortly above an appendage, that is, in an axillary position. The lack of definitive information on this feature, however, has been used to suggest that many of the fossils are better placed in the Lycopodiales (Bateman, 1996b). All members are homosporous. One of the oldest lycopsids is Baragwanathia (FIG. 8.1), known from the Siluro-Devonian rocks of Australia and the lower Middle Devonian of Canada (Lang and Cookson, 1935; Hueber, 1983; Garratt et al., 1984; Rickards, 2000). Baragwanathia longifolia is similar to a modern Lycopodium or Huperzia in that it had dichotomous branches that bore small, closely spaced, helically arranged appendages. The plants are much more robust than those of Lycopodium, however, with stems reaching 6.5 cm in diameter and appendages up to 4 cm long. Associated with some of the appendages are reniform sporangia that produced trilete spores 50 μm in diameter. The compressed nature of the fossil material makes it difficult to determine whether the sporangia are attached to the upper surfaces of the appendages or borne on short stalks in the axils of appendages. The stele is stellate in cross-sectional outline, and the exarch xylem has annular tracheids. Baragwanathia abitibiensis is based on fossils from the Emsian (Lower Devonian) of Canada and consists of axes up

to 3.2 cm in diameter with helically arranged, microphyll-like appendages that bend downward (Hueber, 1983). Anomocytic stomata are randomly scattered on the appendages, but are far more common on the stems. The exact configuration of the xylem cylinder could not be determined, but the stele is constructed of tracheids with both annular and helical wall thickenings that closely conform to the G-type wall thickening. Sporangia were not found on any of the specimens. One interesting feature of B. abitibiensis is the truncated end of the mature appendages. It is believed that this unusual tip morphology is the result of postmortem changes. A similar cause may be used to explain the truncated or cup-tipped spines that have been described in some zosterophylls and trimerophytes. Drepanophycus spinaeformis (FIG. 9.3) was once regarded as a good index fossil of Lower Devonian strata; however, specimens have now been discovered in Middle and Upper Devonian rocks as well. Although traditionally placed with the lycopsids, Schweitzer (1980b) included the genus in the

Figure 9.3 Suggested reconstruction of Drepanophycus spinaeformis (Devonian). (Courtesy D. A. Eggert.)


zosterophyll complex. Aerial axes were probably produced from horizontal rhizomes (Banks and Grierson, 1968) (FIG. 9.4). Branching in D. spinaeformis is in an H or K configuration (as in zosterophylls). This pattern forms when a branch departs from a rhizome at a right angle and then dichotomizes at right angles to produce two stems that parallel the primary axis (Li, 1995). Some specimens from the Emsian of Scotland possess axes interpreted as endogenously formed roots (Rayner, 1984). Drepanophycus spinaeformis axis fragments are up to 27 cm long and 4.2 cm wide. On some specimens from the Canadian Arctic buds occur in the position of branches (Gensel et al., 2001). The surface of the stem in D. spinaeformis is covered with raised mounds (FIG. 9.5) (leaf cushions), representing bases of microphyll-like appendages that have broken off. Other specimens have been preserved so that the inner surface of the outer part of the stem is exposed. In these, the appendages are surrounded by the matrix under the specimen and, instead of mounds, one sees


depressed horizontal or circular areas that denote the position of the appendages still in place. Microphyll-like appendages up to 2 cm long were borne in a shallow helix. Stomata have been described as occurring randomly among elongate, polygonal epidermal cells. The stomatal apparatus of D. spinaeformis is the paracytic type, consisting of two guard cells and two reniform subsidiary cells surrounded by a ring of epidermal cells that vary in number. Permineralized axes of D. spinaeformis reveal a lobed, exarch protostele 2 mm in diameter. Individual tracheids are 70 μm in diameter and 1 mm long. Secondary wall thickenings are annular–helical and possess a perforated reticulum in the position of the middle lamella (Hartman, 1981), now interpreted as G-type thickenings. Within the genus, sporangia are borne in either an axillary position or adaxially on appendages. Little is known about the spores, although the plants are regarded as being homosporous. Another species, D. gaspianus, had more robust axes characterized by rhombic leaf bases that bore microphylllike appendages with broad bases and recurved tips. Appendages were produced in a flat helix that contained 18–22 rows. Specimens from the Lower Devonian of New York

Figure 9.5 Figure 9.4 James D. Grierson. (Courtesy M. A. Millay.)


2 cm.

Drepanophycus spinaeformis (Devonian). Bar 



include a lobed, stellate protostele with annular tracheids and a perforate reticulum similar to that in D. spinaeformis (Grierson and Hueber, 1968). Drepanophycus gujingensis is a species from the Emsian of Yunnan Province in China (C.-S. Li and Edwards, 1995). It has sporangia attached to the stems by short stalks and adventitious roots attached to both the fertile and sterile axes. The stalked, adaxial or axillary sporangia are considered to represent an intermediate feature between plants like Asteroxylon (discussed below) and later lycopsids with sporangia borne directly on sporophylls. Another early lycopsid that is morphologically similar to Drepanophycus is Halleophyton (Early Devonian of China). It has rhomboidal to hexagonal swollen leaf bases and sporangia that split into two equal valves (C.-S. Li and Edwards, 1997). Haskinsia is a herbaceous lycopsid that was once regarded as a species of Drepanophycus (Grierson and Banks, 1983; Xu and Berry, 2008). The helically arranged appendages are falcate and 3 mm long (FIG. 9.6). Metaxylem tracheids are characterized by various patterns of bordered pits. In H. hastata appendages are arranged in a pseudowhorl and are 5 mm long (Berry and Edwards, 1996a, b); sporangia are globose (Yang et al., 2008). These authors placed the genus with the Protolepidodendrales based on the presence of deltoid-shaped sporophylls. Another presumably herbaceous Devonian lycopsid is Haplostigma. In H. baldisii the

microphyll-like appendages are simple and possess subhexagonal bases (Gutiérrez and Archangelsky, 1997). Asteroxylon mackiei (FIG. 8.41) (Chapter 8) is sometimes included in the Drepanophycales, or within a clade that includes Baragwanathia and Drepanophycus as a sister group to all other lycopsids (Kenrick and Crane, 1997a). As noted in Chapter 8, Kidston and Lang originally described terminal sporangia that were thought to belong to this taxon, but subsequent studies showed that the sporangia were borne laterally on the stems near the axils of microphyll-like appendages and not in a terminal position (Lyon, 1964). Whereas A. mackiei is known exclusively from the Lower Devonian Rhynie chert, a second species, A. elberfeldense (FIG. 9.7), is based on impressions with partial anatomical preservation from the Middle Devonian of Germany, Scotland, and

6 cm (A)


Figure 9.7 Asteroxylon elberfeldense, proximal and distal and Figure 9.6 A. Haskinsia sagittata and B. H. hastata (Devo-

nian). (From Berry and Edwards, 1996a.)

narrow, naked axes fragments of Stolbergia spiralis (Devonian). Bar  6 cm. (Courtesy BSPG.)


Norway. The best-preserved specimens come from Kirberg near Elberfeld, western Germany (Kräusel and Weyland, 1926; Weyland et al., 1969). Axes of A. elberfeldense (originally named Thursophyton milleri) are dichotomous to sympodial, up to 1 m long and 0.5 cm in diameter. They bear



helically arranged and densely spaced, short microphyll-like appendages. Often found in association with the spiny A. elberfeldense axes are narrower, dichotomously branched, naked axes of the Hostinella hostimensis type (sometimes misspelled as Hostimella), which were originally believed to represent the terminal portion of A. elberfeldense (Kräusel and Weyland, 1926). Fairon (1967), however, demonstrated that these axes do not belong to A. elberfeldense, but rather represent a different plant, for which she introduced the name Stolbergia spiralis. Although superficially similar to A. mackiei, the systematic affinities of A. elberfeldense remain unclear. Anatomically preserved specimens from the Middle Devonian of the Aachen region in Germany suggest that A. elberfeldense, as well as S. spiralis, belong to the lycophytes (Fairon, 1967). Hestia eremosa is a putatively herbaceous and phylogenetically primitive lycopsid with uncertain affinities that has been described based on isolated stems from a Mississippian sequence of tuffs and lacustrine deposits at Oxroad Bay, East Lothian, Scotland (Bateman et al., 2007). Stems are characterized by the combination of a stellate stele, scalariform xylem pits, and perforate sheets of wall material partially infilling the pits, a complement of features largely consistent with that seen in Huperzia, the most plesiomorphic extant genus of Lycopodiaceae. The limited number of characters preserved in the fossils, however, cannot preclude placement of H. eremosa within the Drepanophycales. An excellent example of a lycopsid that possesses characters of several groups is Smeadia (FIG. 9.8) from the Upper Devonian Cleveland Shale of Ohio (Chitaley and Li, 2004). This plant was herbaceous with a siphonostelic stem (FIG. 9.9) and helically arranged leaves. At the distal end was an erect strobilus (FIG. 9.8) that contained trilete spores 40–80 μm in diameter.


Figure 9.8 Distal end of Smeadia clevelandensis showing ter-

minal cone (Devonian). Bar  5 mm. (From Chitaley and Li, 2004.)

Members of the Protolepidodendrales extend from the Devonian into the Mississippian (Lower Carboniferous) and fit the definition of a true lycopsid. They were either herbaceous or subwoody plants or small trees, and possessed small, helically arranged microphylls that branched at their tips and were vascularized by a single vein (Berry, 1996). Although one genus has been found to bear ligules (discussed below), they have not been found in any other taxa. The generic name Protolepidodendron has been used for a variety of Middle Devonian, dichotomously branched lycopsids (FIG. 9.10) (Grierson and Banks, 1963). Their stems are up to



9.9 Transverse section of Smeadia clevelandensis showing pith, the ring of vascular tissue and cortex of stem (Devonian). Bar 1 mm. (From Chitaley and Li, 2004.)


Figure 9.11 Suggested reconstruction of fertile leaf of

Minarodendron cathaysiense (Devonian). (From Taylor and Taylor, 1993.)

Figure 9.10 Suggested reconstruction of Protolepidodendron (Devonian). (Courtesy D. A. Eggert.)

2 cm in diameter and covered with helically arranged microphylls that were typically bifurcated at the tip. Schweitzer and Cai (1987) described P. cathaysiense from the Middle Devonian of southern China, which they believe is identical to Leclercqia (discussed below), based on the presence of highly bifurcated microphylls. C.-S. Li (1990) combined P. cathaysiense with P. scharyanum and erected a new genus, Minarodendron. Specimens of M. cathaysiense (Givetian) are 3–4 mm wide and generally exhibit a longitudinal series of elongate cushions on which the leaves are borne (FIG. 9.11). The apex of each microphyll is trifurcate, with two tips directed up and a single, median one pointing down. The stems contain an exarch or mesarch strand of primary xylem that is toothed or triangular in cross section. Tracheids range from annular to bordered pitted, and the sporangia are globose–reniform and borne on the adaxial surfaces of unmodified sporophylls. Some Early Devonian specimens that were initially described as species of Protolepidodendron are now called Estinnophyton (Fairon-Demaret, 1978, 1979) (FIG. 9.12). Like other protolepidodendrids, E. gracile was a small, herbaceous plant with axes up to 4 mm wide. Leaves were helically arranged and up to 7 mm long. Each fertile leaf bore two pairs of sporangia, each attached by a small stalk, at a short distance from the tip (FIG. 9.13), although in E. yunnanense (Lower Devonian of China) each fertile leaf contains two




Figure 9.14 Suggested reconstruction of Colpodexylon deatsii

(Devonian). (Courtesy D. A. Eggert.)

Figure 9.12

Muriel Fairon-Demaret.

Figure 9.13 Diagrammatic reconstruction of Estinnophyton

yunnanense. (From Hao et al., 2004.)

single stalked sporangia (Hao et al., 2004). This species also contains poorly preserved trilete spores and tracheids with annular to helical wall thickenings. Estinnophyton is included here with Protolepidodendrales principally on the bifurcate nature of the leaves. However, the paired sporangia suggest affinities with the trimerophytes, whereas their recurved organization is similar to that of some Devonian sphenophytes (Hao et al., 2004).

One of the better-preserved members of the Protolepidodendrales known from Middle and Upper Devonian rocks in New York is Colpodexylon deatsii (Banks, 1944). The dichotomously branched stems are up to 2.5 cm wide and reveal elliptical leaf bases arranged in a low helix or appearing as a pseudowhorl (FIG. 9.14). The characteristic feature of this fossil is the presence of trifurcate leaves, which reached 3 cm in length (FIG. 9.15). The primary xylem strand is lobed in cross section, with exarch– mesarch maturation and annular tracheids. Sporangia are borne on the upper surfaces of unmodified leaves that are scattered along the stem surface. In C. trifurcatum (FIG. 9.16), a Middle Devonian taxon, the trifurcating leaves are 2 mm wide. The primary difference between the two species is the larger leaf base in C. trifurcatum, which may extend up to 5 mm long. In C. camptophyllum (FIG. 9.16) from the Devonian of Venezuela, the tips of the leaves are described as being shorter (Berry and Edwards, 1995). Clwydia (formerly Archaeosigillaria) is a small, dichotomously branched, herbaceous lycopsid that extends from the Devonian into the Carboniferous (Lacey, 1962). The leaf bases range from fusiform on smaller axes (FIG. 9.17) to hexagonal on larger stems. Despite the helical arrangement of the small leaves, they appear decussate and organized into vertical ranks. Some specimens superficially resemble Lycopodites, differing only in the possession of decussate, needlelike leaves (FIG. 9.18). In other species, such as C. vanuxemii, the leaves are deltoid in outline and possess a toothed margin; extending from the apex of the leaf is an elongate hair (Fairon-Demaret and Banks, 1978). The leaves of Clwydia probably did not abscise. The vascular system consists of a lobed protostele with exarch primary xylem and scalariform metaxylem tracheids.









Figure 9.16 Morphologies of various Colpodexylon leaves:

A. C. trifurcatum; B. C. deatsii; C. C. camptophyllum; D. C. cachiriense; E. C. variabile; and F. C. coloradense (From Berry and Edwards, 1995.)

Figure 9.15 Leaf of Colpodexylon trifurcatum (Devonian). (From Berry and Edwards, 1995.)

Phytokneme is a petrified lycopsid axis about 3 cm in diameter discovered in an Upper Devonian phosphatic nodule in the Chattanooga Shale of Kentucky (Andrews et al., 1971). The specimen is exquisitely preserved, with all cells and tissue systems intact (FIG. 9.19). In P. rhodona, the middle cortical zone, which is typically poorly preserved in fossil

lycopsids, contains a network of radially aligned, ray-like strands; these appear similar to cells that characterize the axis of the Carboniferous cone Lepidostrobus kentuckiensis. Roy and Matten (1989) suggested that Phytokneme may have affinities within the Lepidodendrales. A number of lycopsids have been described from the Devonian–Mississippian New Albany Shale of Kentucky and Indiana (Roy and Matten, 1989). One of these is Fodiodendron defractus, characterized by an exarch protostele with scalariform-reticulate metaxylem tracheids. The presence of two




Figure 9.17 Leafy axis of Clwydia (Archaeosigillaria) (Devonian). Bar  1 cm.

strands of included phloem in the leaf traces distinguishes Fodiodendron from other permineralized lycopsids. Probably the most completely known member of the Protolepidodendrales is Leclercqia (FIG. 9.20), a slender, herbaceous plant known from the Early–Middle Devonian of Australia (Fairon-Demaret, 1974; Meyer-Berthaud et al., 2003), North America (Banks et al., 1972; Gensel and Kasper, 2005; Gensel and Albright, 2006), South America (Berry, 1994), and Europe (Fairon-Demaret, 1981).

Figure 9.18 Axis of Clwydia (Archaeosigillaria) showing

prominent leaves (Devonian). Bar  2 cm.

Specimens of L. complexa are up to 46 cm in length and vary from 3.5 to 7 mm in diameter. The dichotomously or pseudomonopodially branched axes are covered by microphylls that attained lengths of 6.5 mm (FIG. 9.21). The leaves are



Figure 9.19 Partial cross section of Phytokneme rhodona

(Devonian). Bar  3 mm. (From Andrews et al., 1971.)

unusual in that they exhibit a pair of lateral divisions about half way up the lamina (Bonamo et al., 1988). Each of these divides into two acuminate tips, whereas the central portion of the leaf gradually tapers and recurves abaxially (FIG. 9.22). The closely spaced leaves, each with five slender tips, must have given the plant an unusual appearance. Stomata of the anomocytic type (Gensel and Albright, 2006) are present on the microphylls and stems, and a few have been observed on the wall of a sporangium. Ligules (FIG. 9.23) have also been reported on Leclercqia microphylls (Grierson and Bonamo, 1979). The function of these small flaps of tissue which occur only in lycopsids has been of historical interest dating back to Hofmeister (1851). Some of the many suggestions as to the function of ligules include secretion and accumulation of water, mucilage, enzymes, and/or nutrients, or superficial conduction of water (Pant et al., 2000). The presence or absence of ligules has been used to define some groups of lycopsids.

Figure 9.20 Leclercqia complexa axis (Devonian). Bar  2 cm.

Some Leclercqia specimens are known in which the vascular cylinder has been preserved as a pyrite petrifaction. In cross section, the stele is circular with up to 18 external protoxylem points and exarch primary xylem maturation.





Figure 9.21 Suggested reconstruction of Leclercqia complexa

(Devonian). (From Kenrick and Crane, 1997a.)

Metaxylem tracheids are scalariform or have oval pits on their walls. Outside the stele is a narrow, parenchymatous cortex with scattered cells with thickened walls, which may be sclerenchyma. Each leaf is vascularized by a single trace that originates from one of the protoxylem ridges on the stele. Sporangia are attached to the adaxial surfaces of sporophylls by a small pad of tissue just proximal to the lateral segments of the leaf. The distribution of sporophylls is similar to that in many species of Hupezia, in which fertile and sterile leaves are almost indistinguishable from one another and interspersed along the stems. Immature spores of Leclercqia are preserved in tetrads. At maturity, the spores are trilete, 60–85 μm in diameter, and ornamented with numerous, closely spaced spines with expanded bases (FIG. 9.24); in situ spores of L. complexa from eastern New York State have been assigned to the dispersed spore genus Aneurospora (Streel, 1972), whereas others have been compared to Actinosporites lindlarensis (Richardson et al., 1993; Gensel and Albright,


Figure 9.22 A. Vegetative and B. fertile leaf of Leclercqia complexa (Devonian). (From Kenrick and Crane, 1997a.)

Figure 9.23 Leaf of Leclercqia complexa with distal branched tips. Arrow indicates position of ligule (Devonian). Bar  625 μm. (From Grierson and Bonamo, 1979.)



Figure 9.25 Leaf of Cervicornus wenshanensis with eight segments (Devonian). Bar  0.5 mm. (From C.-S. Li and Hueber, 2000.)

Figure 9.24 Tetrad of Leclercqia complexa spores with one

spore removed from the tetrad (Devonian). Bar  10 μm. (From Gensel and Albright, 2006.)

2006). Details of the spores, including those in tetrads, have been reported by Gensel and Albright (2006). The large number of fertile specimens with morphologically identical spores has been regarded as evidence that Leclercqia was homosporous. Cervicornus is a small herbaceous lycopsid with helically arranged leaves divided into eight segments (FIG. 9.25) that superficially appear similar to those of Leclercqia (C.-S. Li and Hueber, 2000). Hubeiia dicrofollia is a herbaceous member of the Protolepidodendrales from the Upper Devonian Xiejingsi Formation of Hubei Province in China (Xue et al., 2005). Protoxylem occurs in ridges on the outer surface of the stele and consists of annular tracheids; metaxylem tracheids are scalariform with Williamson striations, delicate, vertical strands of secondary wall material that extend between the scalariform bars, also called fimbrils. These are characteristic of the tracheids in the Lepidodendrales (discussed below). Primary phloem forms a narrow band surrounding the stele. The cortex is relatively thick. Mesarch leaf traces depart directly from the protoxylem. Leaf bases are circular or slightly elliptical in shape and arranged in low helices or in pseudowhorls with the leaf bases alternating. Leaves are

subdivided into four segments as a result of two successive dichotomies. Another interesting lycopsid from China is Wuxia bistrobilata from the Upper Devonian (Berry at al., 2003). Compressed specimens of branches are up to 1.4 cm wide and possess sterile leaves inserted in whorls of six. Megasporangiate structures occur at the dichotomies of the axes and consist of elongate megasporophylls, each with a prominent midrib. Spines occur at irregular intervals along the megasporophyll. Megaspores are up to 4 mm in diameter. Morphologically W. bistrobilata shares most features with Minarodendron cathaysiense from the Middle Devonian (C.-S. Li, 1990). Another Chinese lycopsid of Devonian age with megasporophylls, each bearing 4–6 megasporangia, is Chamaedendron multisporangiatum (Schweitzer and Li, 1996). Microsporangia are stalked and contain spores of the Longhuashanispora type. Chamaedendron is reconstructed as a narrow tree-like plant lacking secondary xylem. Longostachys latisporophyllus from the Middle Devonian of China is reconstructed based on numerous specimens from the same site. It bears elongate megasporophylls with the distal lamina recurved upward and trichome-like appendages arising from the margin and is believed to represent another small tree-like lycopsid (Cai and Chen, 1996). The stele changes from a protostele, surrounded by secondary xylem near the base, to a medullated stele that lacks secondary tissues at higher levels. It is suggested that L. latisporophyllus is intermediate between the herbaceous Protolepidodendrales and the arborescent lepidodendrids. Monilistrobus yixingensis (FIG. 9.26) is a species from the Late Devonian Wutung Formation (Famennian) of Jiangsu, China, that is similar in overall structure to Longostachys and Chamaedendron, but can be


Figure 9.26 Monilistrobus yixingensis showing partial reconstruction of the fertile axis (Devonian). (From Y. Wang and Berry, 2003.)

distinguished from these two taxa (and from all other fossil and most living lycopsids) based on the occurrence of sporangia on modified, proximally widened sporophylls which are compactly arranged into distinct, cone-like fertile zones separated by vegetative regions with more lax microphylls (Y. Wang and Berry, 2003). Another plant believed to represent a transitional stage between the Protolepidodendrales and arborescent lycopsids of the Late Devonian and Carboniferous is Zhenglia radiata from the Lower Devonian of southeastern Yunnan, China (Hao et al., 2006). This plant is characterized by undivided microphylls, sporophylls arranged helically to form a compact area resembling a cone, and ovoid-elongate sporangia positioned adaxially on the widened proximal portion of the sporophyll. Leaf scar arrangement is similar to that seen in Lepidodendrales. Another arborescent lycopsid assigned to the Protolepidodendrales is Protolepidodendropsis pulchra from the Middle and Upper Devonian of Spitsbergen (Høeg, 1942; Schweitzer, 1965, 1999). The largest compressions discovered to date indicate that the stems of this plant, which were originally described as Bergeria mimerensis by Høeg (1942), reach a thickness of 10 cm and a



height of 1.5–2 m. Stem surfaces are covered with helically arranged, broad rhombic leaf cushions that vary considerably in size, depending on the position in the stem, and reach 1.5  1.2 cm in the proximal portions of the stem. In the middle of each of the leaf cushions is a prominent leaf scar; parichnos scars and ligule pit are lacking. Distally the stems dichotomize to form two branches that, in turn, dichotomize up to five times to form a loose crown. Branch fossils do not display leaf cushions, but rather show only leaf scars, which are spindle shaped, that is pointed at both ends, with a thin ridge from the lower scar tip to the upper tip of the scar below (Schweitzer, 1999). Leaves are small and simple (Høeg, 1942). Protolepidodendropsis frickei, a second species in that genus, has been described from the Upper Devonian of Bögendorf-Liebichau near Waldenburg (today Walbryzch) in Silesia (Poland) by Gothan and Zimmermann (1937). In contrast to P. pulchra, stems of P. frickei remain unknown to date, and Banks (1960) suggested that this plant was herbaceous. As we learn more about Devonian and Mississippian lycopsids, it is becoming increasingly clear that not all of them fit into present classification schemes. Such forms as Linietta and Lycopogenia from the Famennian–Tournaisian of North America (Roy and Matten, 1989), and Trabicaulis and Landeyrodendron from the Montagne Noire (Tournaisian) of France (Meyer-Berthaud, 1984a) underscore the necessity of continuing to reevaluate early lycopsids (Meyer-Berthaud, 1984a).

Lepidodendrales The Lepidodendrales includes the arborescent lycopsids that were the most conspicuous elements of the Carboniferous landscape. Members of this group are responsible, to a large extent, for the extensive quantities of plant material that resulted in the formation of Carboniferous coal seams around the world. It is estimated that up to 70% of the biomass in the extensive Westphalian coal-swamp forests of Euramerica was produced by members of the Lepidodendrales (DiMichele et al., 1985). Toward the end of the Westphalian, however, their numbers were in decline and, in the Stephanian epicontinental swamp forest ecosystems, these plants are responsible for merely 5% of biomass production (see Kerp, 2000 and references therein). At the end of the Carboniferous, most arborescent lycopsids become extinct in Europe and North America, and are replaced by tree ferns that, for the first time in geologic history, formed a relatively closed forest canopy; in China, however, the arborescent lycopsids persist



into the Middle Permian. It is suggested that the disappearance of the Euramerican Lepidodendrales toward the end of the Carboniferous was due to climate change (DiMichele and Phillips, 1996a). Others hypothesize that the Variscan tectonic activity considerably reduced the size of the coalswamp ecosystems, which led to more dynamic environments and unstable conditions that were no longer suitable for the lepidodendrids (Kerp, 1996, 2000). Some authors have combined these individual assumptions and suggested that climatic changes were caused by dramatic changes in the floral composition that, in turn, were triggered by tectonic activity (Cleal and Thomas, 2005). Because of the large number of specimens collected, and the variety of ways in which they were preserved (FIG. 9.27), members of the Lepidodendrales are easily the best understood of all fossil lycopsids. The occurrence of structurally preserved members of the order, as well as the extensive stratigraphic distribution of the group, have provided paleobotanists with material to address a variety of geologic and biological aspects of these coal-swamp giants. The great diversity of fossil material has provided exceptionally detailed knowledge not only of the reproductive biology and developmental stages of the vegetative parts of these plants, but also of the role of these plants within the paleoecosystem and the structure of the coal-swamp ecosystem through time. Members of the Lepidodendrales are ligulate and characterized by the presence of secondary xylem, extensive periderm development, a three-zoned cortex, spirally arranged rootlike appendages (stigmarian rootlets) with a monarch vascular strand, and a single functional megaspore per megasporangium, which germinates within the sporangium (DiMichele and Bateman, 1996). Historically specimens preserved in many different modes, for example impressions, compressions, and structurally preserved fossils, were included in Lepidodendron, a genus defined principally on features of leaf cushion external morphology. To alleviate ambiguity and provide better resolution of the systematics and diversity of the arborescent lycopsids, DiMichele (1985) established Diaphorodendron for some structurally preserved specimens previously placed in Lepidodendron, including D. vasculare, D. scleroticum, and D. phillipsii. Later, Diaphorodendron was divided into two genera, Diaphorodendron and Synchysidendron, and placed in its own family, the Diaphorodendraceae (DiMichele and Bateman, 1992). Synchysidendron includes two species, S. dicentricum (formerly Lepidodendron) and S. resinosum (DiMichele and Bateman, 1992). According to DiMichele and Bateman (1996), synapomorphies of the family include a medullated

Figure 9.27 Base of a stem of Tomiodendron peruvianum in growth position (Mississippian). (Courtesy H. W. Pfefferkorn.)

protostele (siphonostele) and a megasporangium that is dorsiventrally flattened with proximal dehiscence; the megaspore has a massa. Synapomorphies of the Lepidodendraceae (Lepidodendron, Lepidophloios, and Hizemodendron) include a bilaterally flattened megasporangium with distal dehiscence and infrafoliar parichnos that extends below the leaf scar (discussed later), among other characters (DiMichele and Bateman, 1996). Today, the generic names Diaphorodendron and Lepidodendron are not only used for stem segments exhibiting cellular preservation but also to encompass the concept of entire plants (FIG. 9.28), including leaves, underground parts, reproductive organs, and a variety of other parts (e.g., decortication stages, spores, isolated sporophylls). Each of these plant organs is assigned both generic and specific names, as relatively few of them have been found organically attached to each other. The generic name Lepidodendron is retained for impression–compression specimens that possess a particular type of leaf cushion morphology in addition to




Figure 9.28 Suggested reconstruction of an arborescent lycopsid (e.g., Lepidodendron). (Courtesy D. A. Eggert.)

certain anatomical features. Specimens of Lepidodendron have been discovered that indicate some trees attained heights in excess of 40 m and were at least 2 m in diameter at the base (Thomas and Watson, 1976). In the midcontinent of North America, the genus appears to have reached its zenith at the end of the Middle Pennsylvanian (Phillips et al., 1974). The massive, erect trunks of some Lepidodendron species branched profusely to produce large crowns of leafy twigs (FIGS. 9.29, 9.30). Some leaves were at least 1 m long (most were much shorter) and, when they dropped from the branches, conspicuous leaf bases were left on the stem surface.

Figure 9.29 Leafy lycopsid twig (Pennsylvanian). Bar  2 cm.





9.30 Compression of Lepidodendron lycopodites (Pennsylvanian). Bar  2.5 cm.


Strobili were borne at the tips of distal branches or in a zone at the top of the main trunk. The underground portions of the Lepidodendrales (sometimes called a stigmarian rhizomorph) consisted of dichotomizing axes that bore helically arranged, lateral appendages that presumably functioned as roots.

Figure 9.31 Lepidodendron leaf base. L: ligule; P: parichnos;

VB: vascular-bundle scar. (From Taylor and Taylor, 1993.)

Vegetative Features

STEM SURFACE AND LEAF BASES Some of the most commonly encountered fossils assignable to the lepidodendrids are compressions of stem surfaces marked by persistent, somewhat asymmetric, more or less rhomboidal leaf cushions (FIG. 9.31). The leaf cushion actually represents the expanded leaf base left behind after the leaf dropped off (FIG. 9.31), since abscission of the leaf did not occur flush with the stem surface. The top and bottom of the cushions, which are also called leaf bolsters, generally form acute angles; the sides are more rounded. The actual scar left by the abscised leaf is slightly above the midpoint of the cushion and is generally elliptical or rhombic in outline (FIG. 9.32). On the surface of the leaf scar are three small, pit-like impressions. The central one represents the single vascular-bundle trace that extended into the leaf. The other two

scars represent the position of channels of loosely arranged parenchyma tissue, termed parichnos (FIG. 9.31); this tissue originates in the cortex and extends through two grooves on the abaxial surface of the leaf. On Lepidodendron stem surfaces, two additional parichnos channels can be identified at a short distance beneath the leaf scar; these do not occur in the Diaphorodendraceae, where parichnos is confined to the foliar scar only. Parichnos is a system of aerating tissues within the stem. A vertical line extends from the leaf scar proper to the lower limit of the leaf base. In many specimens, lateral wrinkles cut across this line. Initially, it was thought that these wrinkles were of systematic value, but it is now understood that they are the result of the growth of secondary tissues in the stem. Just above the leaf scar is a mark that




Figure 9.32 Several Lepidodendron leaf cushions preserved in

a Mazon Creek nodule (Pennsylvanian). Bar  3.5 mm.

represents the former position of a ligule. Additional markings occur on the leaf base in the form of vertical lines that result from lateral expansion of the stem. Rarely is preservation so good that all these features can be observed in a single leaf base. In Synchysidendron, the leaf cushion includes a groove that is formed by folding of the cushion tissue immediately below the leaf scar proper. Some compression specimens of arborescent lycopsid stems (FIG. 9.33) also provide information about the epidermis of these plants. A waxy cuticle covers the stem surface, including the leaf cushions, but is thought to be absent on the leaf scar itself (Thomas, 1966). The epidermis is simple and lacks such specialized cells as hairs and glands. Stomata are common and sunken in shallow pits (Thomas, 1974). Another tree-sized lycopsid, Lepidophloios (Q. Wang, 2007) (originally spelled Lepidofloyos; Sternberg, 1825), occurred in the Carboniferous coal swamps along with Lepidodendron (DiMichele, 1979a). Lepidophloios (Lepidodendraceae) was probably slightly smaller in stature (FIG. 9.34), but in general its features are quite similar to Lepidodendron. One notable difference between the two is the arrangement and organization of leaf bases (FIG. 9.35). In Lepidophloios, the leaves are arranged in a shallow helix, as in other lepidodendrids, but the leaf bases are flattened and wider than they are tall (FIG. 9.35). They are directed downward on the stem and overlap the bases below, much like shingles on a roof. When attached, the leaves bend upward abruptly from the leaf bases; when a leaf abscised, it left a scar on the bottom third of the base

Figure 9.33 Lepidodendron leaf bases (Pennsylvanian). Bar 

1 cm. (Courtesy BSPG.)

(FIG. 9.36). Parichnos and vascular-bundle scars on the leaf scar are like those of Lepidodendron, but parichnos scars are not present on the base itself. Lepidophloios was ligulate, with the ligule attached just above the position of the leaf scar. Many Lepidophloios stems exhibit large, circular to elliptical scars on the stem surface (FIG. 9.37), some up to several centimeters in diameter. The origin of these scars has been debated for many years. Some workers regard them as former sites of vegetative branches that abscised during the normal growth of the plant, whereas others suggest that they represent former positions of specialized branches that bore clusters of strobili. Historically, stems with helically arranged scars of this type have been given the generic name Halonia, whereas those with oppositely arranged scars are



called Ulodendron. Jonker (1976) suggested that Ulodendron scars on the axes of Lepidodendron, Lepidophloios, and a related genus, Bothrodendron, represent the former positions of branches that abscised in a manner similar to that of some existing gymnosperms and angiosperms. Thomas (1967) regarded Ulodendron as a natural genus, basing this hypothesis on the persistent leaves, shallow ligule pits, and rhomboidal leaf bases, whereas DiMichele (1980) suggested it was congeneric with Paralycopodites.




Figure 9.36 Diagrammatic representation of Lepidophloios leaf base showing ligule (L), parichnos (P), and vascular-bundle scar (VB) (Pennsylvanian). (From Taylor and Taylor, 1993.)

Figure 9.34 Suggested reconstruction of distal region of

crown branches of Lepidophloios hallii (Pennsylvanian). (From DiMichele, 1979a.)

Figure 9.35 Paradermal section of Lepidophloios leaf bases

Figure 9.37 Three prominent oval branch scars along a

showing four branch scars (Pennsylvanian). Bar  1 cm.

Lepidodendron axis (Pennsylvanian). Bar  4 cm.


STEM ANATOMY Species of Diaphorodendron have been delimited on the basis of internal stem organization. The configuration of the stele has been used as a taxonomic character as well as to suggest evolutionary changes that may have taken place within the arborescent forms. Knowledge of changes in stelar configuration at various levels of the stem has clearly demonstrated that many of the so-called species of Lepidodendron and Diaphorodendron as well as some of the species of Lepidophloios (FIG. 9.38) may represent different developmental stages of a single species. These studies have provided us with a basic understanding of just how these giant Carboniferous plants actually grew (FIG. 9.39). Before discussing the ontogenetic changes that occur during the growth of arborescent lycopsids, it is important to detail the internal structure of a typical lepidodendralean stem. Stems of D. scleroticum, D. phillipsii, and D. vasculare are relatively abundant in coal-ball deposits from the eastern interior basin of North America (DiMichele, 1979b) and the description that follows will be a composite picture of the internal organization of a typical stem. DiMichele suggests that the principal feature



of lepidodendrids, that is, leaf cushions that are longer than they are wide, masks the extensive diversity present within the genus, but there are several anatomical differences as well. Stems may be protostelic (Diaphorodendron), have a mixed pith, or be siphonostelic (Diaphorodendron and Lepidodendron) (FIG. 9.40). Protostelic stems such as L. pettycurense from the Mississippian of Scotland (see Kidston, 1907) consist of a central strand of primary xylem surrounded by a narrow ring of protoxylem that in turn is surrounded by secondary xylem. In stems with a mixed pith, parenchyma cells are interspersed with tracheids in the center, but the tracheids are shaped more like short, squat parenchyma cells rather than


Figure 9.39 Arborescent lycopsid (Lepidodendron) showing

branch trace (B) (Pennsylvanian). Bar  1 cm.



Figure 9.38 Several steles of the Arran Tree (Lepidophloios

wuenschianus) (Pennsylvanian). Bar  5 cm.

Figure 9.40 Transverse section of Diaphorodendron stem showing small vascular cylinder (arrow), conspicuous cortex (C), and prominent leaf bases (L) (Pennsylvanian). Bar  5 mm.



long, slender cells. This condition is often cited as evidence that the pith in lycopsids originated when immature parenchymatous cells in the center of stems failed to differentiate into tracheids. Surrounding the mixed pith is a narrow band of metaxylem tracheids with scalariform wall thickenings; protoxylem occurs at the periphery of the metaxylem. In cross section, the outer margin of the primary xylem appears fluted due to the large number of protoxylem ridges. As in most members of this division, maturation of the primary xylem is exarch. Surrounding the primary xylem in the arborescent lycopsids may be a zone of secondary xylem, which reaches a maximum thickness of several centimeters. Unlike woody trees today, secondary xylem in the arborescent lycopsids accounts for only a small proportion of the diameter of the stem; rather, the extensive development of periderm is primarily responsible for their massive trunks. Both primary and secondary xylem tracheids are scalariform and have Williamson striations (fimbrils) extending between the scalariform bars. These fimbrils are characteristic of the wood in the arborescent lycopsids, but similar fimbrils also occur in modern club mosses and spike mosses (Wilder, 1970; Schneider and Carlquist, 2000a, b), and are regarded as a shared character for the Lycopsida as a whole (DiMichele and Bateman, 1996). Numerous vascular rays radiate through the secondary xylem; these are generally a single cell wide and up to several cells high. Due to the presence of scalariform thickenings on the walls of the ray cells, it has been suggested that xylem rays in this group evolved from tracheids. Immediately outside the secondary xylem is a zone of thinwalled cells that represent the vascular cambium. Unlike the bifacial cambium of typical seed plants (Chapter 7), however, the vascular cambium in the lepidodendrids is unifacial, producing only secondary xylem on the inner face (Eggert and Kanemoto, 1977). The unusual manner in which this secondary vascular tissue was produced was determined by examination of stems with and without secondary vascular tissue, and by comparing the distribution of cell types within the stems. In Diaphorodendron and Lepidodendron, the phloem zone is separated from the secondary xylem by a band of thin-walled cells termed the parenchyma sheath. Immediately outside the parenchyma sheath are sieve elements with large, elliptical sieve areas on their walls, interspersed with strands of phloem parenchyma. In some specimens a zone of radially seriate, thin-walled cells is in contact with the secondary xylem cylinder. Although some believe that this tissue resulted from cambial activity, others suggest it was formed of living primary-sheath cells that were capable of becoming meristematic. Current evidence seems to suggest that no secondary phloem was produced within the arborescent lycopsids.

CORTICAL TISSUES The cortex of the lepidodendrids is usually subdivided into three general zones—the inner, middle, and outer cortex (FIG. 9.41), which have been defined on the basis of cell types. The inner cortex is the narrowest of the cortical zones and is constructed of small, isodiametric parenchyma cells. Aggregations of cells with dark contents, presumed to be secretory cells, in addition to lacunae and various types of sclerotic cells, also occur in this zone. The middle cortex is more extensive and constructed of larger parenchyma cells. In young stems this zone is characterized by lacunae that extend radially; in older stems the middle cortex is generally not preserved, except for a few parenchyma cells along the inner and outer edges. Cells of the outer cortex show no definite arrangement. They have slightly thicker walls and superficially resemble collenchyma cells. In some species, this zone may be distinguished by longitudinally oriented, anastomosing bands of fibers. Secondary cortical tissue, or periderm, is produced in the outer cortex and, judging from the extensive blocks of this tissue in coal balls, it is this tissue that contributed to most of the trunk diameter in these plants. In Diaphorodendron, the periderm is bizonate, with a differentiation into phelloderm (inner zone) and phellem (outer zone). The inner zone consists of alternating thick- and thin-walled cells, whereas the outer zone is homogeneous and contains cells with dark contents, described as appearing resinous. Lepidodendron has a massive periderm (FIG. 7.13), which may be homogeneous or bizonate, and there is no clear differentiation into phellem and phelloderm (DiMichele, 1985); cells may be tangentially expanded in the outer part of the cortex. Periderm production originates

Figure 9.41 Transverse section of Diaphorodendron stem with

branch trace (arrow) (Pennsylvanian). Bar  1 cm.


from meristematic parenchyma cells (phellogen) in the outer cortex just beneath the leaf bases. The phellogen in the arborescent lycopsids produced a relatively small amount of phellem toward the outside of the stem, and a much larger amount of phelloderm toward the inside, the opposite of most extant seed plants (see Chapter 7). Cells of the periderm are radially aligned and often show a storied arrangement, that is, in longitudinal section, the end walls of the cells are lined up horizontally. In some lepidodendrids, the periderm is more complex and consists of three kinds of cells: (1) thickwalled, axially elongated fibers; (2) radial rows of chambered cells, conspicuous because they are divided by transverse walls; and (3) secretory cells aligned in sinuous bands that give the appearance of growth rings in cross sections. It has been suggested that periderm in the arborescent lycopsids formed in several ways (Eggert, 1961). Some axes show successive tangential bands of meristematic tissue at varying depths in the cortex. In this pattern of periderm production, outer cortical layers become meristematic and produce radially oriented files of cells for a period of time. As each group of meristematic cells ceases dividing, successively deeper cortical cells become meristematic and repeat the process of periderm production. The presence of short files of periderm cells in both stems and underground axes supports such a pattern of development. In some cases, thickwalled cells are produced in the outer cortex. The outer surface of arborescent lycopsid stems is not the only surface to have been preserved as fossils. Both the loose construction of the cortex and the production of large amounts of relatively thin-walled periderm contributed to the sloughing off of stem layers and tissues, either as a feature of normal development or as a result of mechanical separation during the fossilization process. As a result, a variety of presumed external stem and trunk features are represented in fossils that are unlike the leaf bases and have resulted from levels of decortication of the axis. Various generic names have been applied to these decorticated states, but the names do not conform to the concepts of morphogenera and are therefore of little value in systematic studies. Nevertheless, these various fossil forms do provide information that can be used to reconstruct developmental changes in the component tissue systems of an axis. One of the more common decortication stages is Knorria, a name used for stems in which almost all the tissues external to the xylem (and sometimes even the stele) have been lost; Knorria represents a mold–cast type of preservation and is characterized by irregular, longitudinally oriented ridges. These ridges represent sediment infillings of the parichnos strands that accompany the vascular bundles through



the stem (Hirmer, 1927). Knorria casts are known not only from Lepidodendron stems (FIG. 9.42), but have also been reported from Bothrodendron and Jurinodendron (FIG. 9.43) (formerly Cyclostigma) stems. STEM DEVELOPMENT One of the outstanding contributions to modern paleobotany is Eggert’s detailed investigation of development in arborescent lycopsids (Eggert, 1961) (FIG. 9.44). Studies of this type have not only contributed to our understanding of growth processes in fossil groups, but have also made it possible for paleobotanists to distinguish developmental differences represented by fossils from features that are useful in lycopsid taxonomy (Delevoryas, 1964a). In his analysis, Eggert utilized a large number of permineralized axes with varying diameters; the axes included different amounts of primary and secondary tissues. Using this approach, he was able to reconstruct the pattern of growth in these plants and demonstrate the successful application of this technique in the analysis of other groups of plants. In these arborescent forms, the upper portion of the main axis contains a large siphonostele that is characterized by

Figure 9.42 Lepidodendron stem or branch in Knorria preser-

vation (Pennsylvanian). Bar  1 cm. (Courtesy BSPG.)



a wide pith surrounded by a thick zone of primary xylem (FIG. 9.45). As very young specimens exhibit only a small protostele, there must have been a change in stelar configuration from the protostelic sporeling stage to the siphonostele present in the main trunk. This initial expansion of the primary body during early growth is termed epidogenesis. When the plant is relatively immature, the cortex is extensive and the outer surface of the trunk is covered with numerous rows of raised leaf bases (FIG. 9.45). As the tree continues to grow, secondary xylem and periderm are added to the stem

Figure 9.44 Donald A. Eggert.

kiltorkense in Knorria-type preservation (Mississippian). Bar  5 cm. Figure

9.43 Jurinodendron

Figure 9.45 Longitudinal and cross sections showing the distribution of primary (solid) and secondary (radiating lines) xylem in an arborescent lycopsid. (After Eggert, 1961.)


as a result of the vascular cambium and phellogen. The increase in stem diameter results in the sloughing off of the outer cortical tissues (FIG. 9.45), including the leaf bases, so that in older parts of the plant (e.g., at the base), the outer surface of the trunk is protected by periderm. Many of the older reconstructions of Lepidodendron in museums and drawings often err in showing leaf bases extending all the way to the ground on old trunks. At higher levels in the tree, the branches have smaller steles and fewer rows of smaller leaves on the surface. Sections of stems at these levels indicate that less secondary xylem and periderm are produced. A reduction in stele size and tissue production continues until the most distal branches, which contain a tiny protostele with only a few small tracheids, no secondary xylem or periderm, and just a few rows of leaves. This stage in development, in which the plant literally grows itself out, has been termed apoxogenesis. In other words, the small, distal twigs of these arborescent lycopsids do not have the potential of developing into larger branches with time. This type of growth pattern is called determinate and contrasts with indeterminate growth, which is typical of vegetative development in most living woody plants. Paleobotanists must continually devise new methods of investigating the biology of the organisms they study. Eggert’s elegant analysis of growth in the arborescent lycopsids is one such approach. In another, the focus of the study is the nature of the unifacial vascular cambium in two Carboniferous lycopsid morphogenera, Stigmaria and Paralycopodites (Cichan, 1985a). Cichan (FIG. 9.46) prepared serial tangential sections of the secondary xylem in order to determine the pattern of production of cambial derivatives and the method of circumferential increase in the cambium. Cichan’s studies indicate that cambial activity in these plants was also determinate. Circumferential increase took place by the enlargement of fusiform initials, rather than by anticlinal divisions of existing initials, as it does in seed plants. This type of growth would result in a cambium that was limited in its capacity for radial expansion. As secondary growth ceased in the plant, fusiform initials ceased to be meristematic and matured into a cylinder of parenchyma. LEAVES The leaves of arborescent lycopsids are linear and some were up to 1 m long (FIG. 9.47). Chaloner and MeyerBerthaud (1983) demonstrated that stems with the largest diameters have the longest leaves, a feature they correlate with the determinate growth of the plants. Many of the different species established for detached leaves were probably



produced by the same kind of plant and only differed in size, shape, and anatomy because of their position on the plant. The generic name Lepidophyllum was initially used for both structurally preserved and compressed lepidodendrid leaves, but because this name had been used earlier for a flowering plant, the name Lepidophylloides was proposed in its place (Snigirevskaya, 1958). A single vascular bundle, flanked by two shallow grooves on the abaxial surface, extends the entire length of the lamina in Lepidophylloides (FIG. 9.48). Stomata occur on the abaxial surface aligned in rows that parallel the grooves and sunken in shallow pits. A well-developed hypodermal zone of fibers surrounds the mesophyll parenchyma and vascular bundle of the leaf; no palisade parenchyma has been reported. In L. sclereticum from Permian coal balls, the vascular bundle is convex in transverse section and surrounded by transfusion tracheids (S. J. Wang et al., 2002). UNDERGROUND ORGANS Underground axes of the Lepidodendrales are given the name Stigmaria. These dichotomizing structures represent one of the most common lycopsid fossils and constitute the

Figure 9.46 Michael A. Cichan.



Figure 9.48 Cross section of Lepidodendron leaf. Note

two abaxial furrows (arrows) where stomata are located (Pennsylvanian). Bar  1.5 mm. (Courtesy BSPG.)

Figure 9.47 Lepidophylloides in Mazon Creek nodule (Pennsylvanian). Bar  2 cm.

principal organ found in the clay layer or underclay immediately beneath most Carboniferous coal deposits. The underclay represents the soil layer (paleosol) in which these plants were rooted in the coal swamps (Mosseichik et al., 2003). Extensive specimens of Stigmaria have been uncovered in growth position, some with rootlike structures, commonly called stigmarian appendages, still attached (FIG. 9.49). Although there are several species of Stigmaria, our knowledge of the anatomy of these underground systems is based principally on the species Stigmaria ficoides (FIG. 9.50) (Williamson, 1887b). The stigmarian system arises from the base of the trunk as four primary axes, each of which extends out horizontally, so that the rooting system is relatively shallow. Helically arranged lateral appendages were attached to each axis. These appendages abscised during the growth of the plant, leaving characteristic circular scars (FIG. 9.51) on the main axis, and these can be seen on a variety of casts, compressions, and impressions of Stigmaria. The lateral appendages are sometimes called stigmarian rootlets (FIG. 9.52), although their helical arrangement (i.e., phyllotaxy) and abscission are characteristic of leaves rather than lateral roots (see Chapter 7). The primary axes in Stigmaria dichotomize repeatedly to form an extensive subterranean system that may have radiated up to 15 m from the trunk. Primary axes of Stigmaria have a parenchymatous pith that may include scattered tracheids at more distal levels. Primary xylem is endarch and arranged in a series of dissected bands, which are, in turn, surrounded by a vascular cambium. What is interpreted as secondary xylem is distinctive because the wide vascular rays give the wood a segmented appearance (FIG. 9.53). Secondary xylem tracheids are aligned in radial files and possess scalariform wall




Figure 9.49 Stigmaria ficoides compression showing laterally attached rootlets (Pennsylvanian). Bar  2 cm.

Figure 9.50 Cross section of stigmarian axis showing cylinder of secondary xylem and several rootlets (arrows) attached to the outer surface of the cortex (Pennsylvanian). Bar  2 cm.

Figure 9.51 Cast of Stigmaria sp. showing pattern of rootlet scars (Pennsylvanian). Bar  1 cm.

Figure 9.52 Cross section of stigmarian rootlet showing monarch collateral bundle (Pennsylvanian). Bar  2 mm.

thickenings with fimbrils identical to those of the aerial parts. Occasional imperfections in the secondary xylem appear to be the result of a temporary cessation of cambial activity, which caused an abrupt change in the diameter and distribution of the tracheids in the wood. It has been suggested that this erratic cambial activity was a result of some abrupt environmental change. No secondary phloem has been identified in Stigmaria; the vascular cambium was apparently unifacial and translocation effected by the primary phloem. Rothwell



Figure 9.53 Cross section of the secondary xylem cylinder of Stigmaria ficoides (Pennsylvanian). Bar  4 mm.

Both primary and secondary cortical tissues of Stigmaria ficoides are complex, consisting of numerous cell and tissue systems, some of which are difficult to trace developmentally. In general, however, the production of secondary cortical tissues in the underground parts resulted in a narrow zone of periderm that is histologically similar to that of the aerial stems. A cast of the apex of Stigmaria is known from the Middle Pennsylvanian of Iowa (FIG. 9.54) (Rothwell, 1984). The specimen is 8.5 cm long and contains helically arranged rootlet scars that surround a raised rim that is believed to correspond to the shape of the meristem. Morphologically, the specimen is nearly identical to the juvenile stage of the underground rooting structure of Nathorstiana, a Mesozoic lycopsid (Karrfalt, 1984). The lateral appendages, sometimes termed rootlets, produced by the stigmarian axes are up to 40 cm long and 0.5–1 cm in diameter, usually unbranched (some may dichotomize once), and they gradually taper distally. Each rootlet has a small monarch vascular strand surrounded by a compact inner cortex. External to this is a hollow, middle cortical zone and a thin outer cortex (FIG. 9.52). At some levels, a connective extends from the outer cortex to the inner cortex (Weiss, 1902) (FIG. 9.55). Stigmaria stellata, from the Upper Mississippian Chester Series of Illinois, exhibits radiating ridges on the surface around the lateral appendage scars in both casts and impressions (Jennings, 1973). Structurally preserved axes suggest a close relationship to S. ficoides, but S. stellata differs in several anatomical features, including the absence of a connective in the lateral appendages.

Figure 9.54 Apex of Stigmaria ficoides (Pennsylvanian).

Bar  1.5 cm. (From Rothwell, 1984.)

and Pryor (1991) combined observations on mold–cast specimens with permineralized axes and hypothesized that the radially aligned tracheids in the steles of many stigmarian axes were produced by a primary thickening meristem rather than a vascular cambium.

Figure 9.55 Frederick E. Weiss.




Figure 9.56 Detail of Lepidodendropsis sp. leaf bases (Mississippian). Bar  5 cm.

Figure 9.57 Axis of Lepidodendropsis sp. showing leaf bases

Not all stigmarian underground parts are extensive, dichotomously branched systems. Some, such as the Early Mississippian genus Protostigmaria (Jennings, 1975a), consist of cormlike axes bearing helically arranged laterals similar to those of Stigmaria. In P. eggertiana from the Price Formation of Virginia (Mississippian), both underground parts and some aerial axes are preserved in growth position (Jennings et al., 1983). Axes are up to 32 cm in diameter and preserved as impression–compression and mold–cast specimens (FIG. 1.36). The root-bearing portion is divided into lobes by a system of furrows, with additional furrows added as the plants grew; the largest number of lobes recorded in P. eggertiana is 13. Roots were arranged on each lobe in a helical pattern; each of the scars is circular and up to 8 mm in diameter. Features of the stem surface (FIG. 9.56) and its association in the same rocks strongly suggest that Protostigmaria was the underground portion of Lepidodendropsis (FIG. 9.57), a Mississippian arborescent lycopsid. One feature of Protostigmaria that deserves additional comment is the ability of the plant to maintain an upright position despite the relatively small anchoring surface of the lobed base. Except for the larger size and larger number of lobes, Protostigmaria is morphologically identical with the root-bearing structure of the extant lycopsid Isoetes, and this

(Mississippian). Bar  2 cm.

further strengthens the homologous nature of the root-bearing organ of lycopsids. Experimental studies suggest that during the development of Isoetes (Karrfalt, 1977), non-contractile roots in the furrows of the cormlike base move laterally as tissues are added, resulting in the plant being pulled farther down into the substrate. It has been suggested that such a mechanism may have been operative in Protostigmaria, as it lacked the extensive, dichotomizing anchoring system of some of the other arborescent lycopsids (Jennings et al., 1983). Research with Protostigmaria also illustrates that developmental data can be determined from fossils that are preserved in ways other than permineralizations. DEVELOPMENT OF UNDERGROUND ORGANS Development of the underground parts of the arborescent lycopsids was probably quite similar to the epidogenic and apoxogenic stages described for the aerial stems. Despite the large number of Stigmaria specimens that have been collected, several features of these organs remain to be determined and interpreted. For example, the helical arrangement



of the lateral appendages is unlike the irregular arrangement of roots in most living plants. No root hairs have been identified on any specimens; perhaps fungi in some of the cortical parenchyma cells functioned as mycorrhizae. The monarch vascular bundle in these so-called rootlets is bilaterally symmetrical, that is, a collateral bundle. Typically roots have radially symmetrical vascular tissue in cross section, that is the phloem surrounds the xylem, while the vascular bundles in leaves are collateral. Finally, the lateral appendages apparently abscised from the parent axis in a regular manner, perhaps similar to the process of foliar abscission in many plants. Programmed abscission of laterals is unknown in the roots of living plants and has not been observed in any non-lycopsid fossil plants. For these reasons, it has been suggested that the rootlets of Stigmaria are actually homologous with leaves that have been modified for the functions of anchorage and absorption. This interpretation implies that the underground portions of these plants arose by evolutionary modification of aerial axes (Frankenberg and Eggert, 1969). Although some of the coal-swamp lycopsids in the Pennsylvanian grew to 40 m high, the stigmarian underground system was relatively shallow, and questions have arisen as to how much support it could have provided for these towering trees. Many of these plants grew in what must have been a supersaturated soil that also provided little stability. It may be, however, that the extensive horizontal development of the underground systems was sufficient to allow these plants to remain upright. Studies of extant trees suggest that the nature of the wood (e.g., strong, dense wood) and the density of the crown can have a pronounced effect on tree uprooting (Niklas, 1992). The arborescent lycopsids had very little secondary xylem, and this may have been an advantageous mechanical property in remaining upright in heavy wind or rain. In addition, they had somewhat bushy crowns. It has been suggested that the crowns of adjacent trees became entangled and thus provided mutual support for these trees in the Carboniferous swamps.

organization of lepidodendrid cones consists of a central axis with helically arranged sporophylls (FIG. 9.59). Sporangia are located on the adaxial (upper) surface of sporophylls which are upturned at their distal ends so that they overlap the sporophylls above. A portion of the lower surface of the sporophyll often extends downward to form a heel or distal extension. A ligule is present in a small pit just distal to the sporangium. Lepidostrobus is the most common designation for lycopsid cones of this type (FIGS. 9.60, 9.61). The name has been used, however, for cones demonstrating all forms of preservation, and for both monosporangiate (having only

Reproductive Biology

The reproductive units of the lepidodendrids consist of strobili or cones borne on distal branches in the crown of the tree. In Synchysidendron cones occur on late-formed crown branches, whereas in Diaphorodendron they are borne on deciduous lateral branches (DiMichele and Bateman, 1992). Lepidodendrid cones could reach considerable size (FIG. 9.58), for example cones assignable to Lepidostrobus goldenbergii could be more than 50 cm long. The basic

Figure 9.58 Compressed Lepidostrobus cone (Pennsylvanian).

Bar  2 cm.


one type of spore) and bisporangiate (having two types of spores) forms, so that taxonomic problems within the genus are considerable.



MICROSPORANGIATE AND BISPORANGIATE CONES In an attempt to better define the taxonomic limits of arborescent lycopsid cones, Brack-Hanes and Thomas (1983)

suggested that the name Lepidostrobus be used for cones containing only small spores (monosporangiate) and Flemingites for bisporangiate strobili containing both microspores and megaspores (FIG. 9.62). Other authors have used cone morphology and in situ spores to more accurately circumscribe individual species within Lepidostrobus (Bek and Opluštil, 2004, 2006).

9.59 Compression specimen showing axis of Lepidostrobus cone with helically attached sporophylls (Pennsylvanian). Bar  1 cm.

Figure 9.60 Longitudinal section of distal end of permineralized Lepidostrobus cone with distal sporangia containing spores (Pennsylvanian). Bar  3 cm.




Figure 9.61 Transverse section of permineralized Lepidostrobus cone showing central axis and helically arranged microsporophylls (Pennsylvanian). Bar  2 cm.

Figure 9.62 Transverse section of Flemingites schopfii cone showing microsporangia (arrows) intermixed with megasporangia (Pennsylvanian). Bar  5 mm.

One of the oldest lepidodendrid cones is Lepidostrobus oldhamius, known from Lower and Middle Pennsylvanian deposits in both North America and Great Britain. Specimens range from 2–6 cm in diameter and exceed 30 cm in length. Sporangia are massive and, depending on the stage of development, may have an irregularly shaped pad of sterile tissue extending from the sporophyll into the lumen of the sporangium. All of the sporangia contain small (20–30 μm) spores that have a trilete suture on the proximal surface and delicate spines on the distal face. Dispersed spores of Lepidostrobus are often preserved, and one of the common generic names applied to these grains is Lycospora (Willard, 1989). Lepidostrobus shanxiense is a slightly smaller cone from the Xishan Coal Field in China (Wang et al., 1995). An interesting Lepidostrobus cone, L. xinjiangensis, also from China, occurs in Upper Devonian rocks (Wang, Q. et al., 2003a). The discovery of this species suggests that segregation into micro- and megasporangiate cones had already taken place in the lepidodendrids by the Late Devonian. Flemingites schopfii is a permineralized bisporangiate cone that resembles a massive modern Selaginella strobilus in general organization (Brack, 1970). Specimens are up to 8 cm long and 1.3 cm in diameter. The arrangement of parts and cone anatomy are identical with those of L. oldhamius, except for the presence of two types of spores in F. schopfii. Distal sporangia (FIG. 9.63) contain a large number of Lycospora-like microspores (FIG. 9.64), some still in tetrahedral tetrads; more basal sporangia contain 12–29 trilete megaspores (Brack-Hanes, 1978). The megaspores range from 700–1250 μm in diameter and are marked by an elongation of the proximal surface into an apical prominence or gula. Dispersed megaspores of this type are called Valvisisporites. Both compressed and structurally preserved lepidostroboid cones are known from the Fayetteville Shale of Arkansas (Upper Mississippian). These cones, which all appear to have been monosporangiate, may have attained lengths of 22.5 cm. Spores extracted from sporangia of both cone types provide a means of comparing different preservational types. In Lepidostrobus fayettevillense, the small, trilete miospores are characterized by a perforated flange (cingulum), that encircles the spore at the equator (FIG. 9.64) (Taylor and Eggert, 1968). It is probable that apparently monosporangiate cones such as L. oldhamius and L. fayettevillense represent the microsporangiate cones of heterosporous plants. In other instances, however, monosporangiate cones may have been produced by homosporous plants reproductively similar to Lycopodium, in which the spores germinate into free-living gametophytes. In this regard, the neutral term miospore




Figure 9.63 Suggested reconstruction of sporangium of

Flemingites schopfii (Pennsylvanian). Figure 9.65 Megasporangium containing numerous Triletes sp. spores (Pennsylvanian). Bar  5 mm.

might be more appropriate than microspore, as the exact function of these spores in the life history of the plant is currently unknown. Mixostrobus givetensis, from the Middle Devonian of Kazakhstan, is a bisporangiate cone with an irregular arrangement of micro- and megasporangia. In some strobili, one type of sporangium is the dominant type, but this does not depend on the size of the cone (Senkevitsch et al., 1993). Cones assigned to the genus Kladnostrobus from the Pennsylvanian of the Kladno-Rakovník Basin, Czech Republic, differ from other lycopsid reproductive structures in that the in situ spores are reticulate and resemble the dispersed spore genera Convolutispora, Camptotriletes, Reticulatisporites, and Dictyotriletes (Libertín et al., 2005). Figure 9.64 Proximal surface of Lycospora-type spore show-

ing surface ornament, trilete suture, and equatorial cingulum (Pennsylvanian). Bar  7 μm.

MEGASPORANGIATE CONES Some lepidodendrid cones are monosporangiate, but produce only megaspores (FIGS. 9.65, 9.66). Possibly the



Figure 9.66 Megaspore of the Triletes-type showing proximal triradiate ridge and ornamentation (Pennsylvanian). Bar  1 mm.

Figure 9.67 Partial transverse section of Lepidocarpon palm-

erensis cone (Pennsylvanian). Bar  1 cm.

most common of these Lepidocarpon (FIGS 9.67, 9.68), which was borne on Lepidophloios stems. This cone type is regarded as the most highly evolved reproductive structure in the lycopsids, because the arrangement of the sporophylls closely approximates the function of integuments in seed plants. In Lepidocarpon, sporangia are borne on the adaxial surface of the sporophyll (FIG. 9.69), which consists of two lateral laminae (FIG. 9.70) and a distal extension or heel (Balbach, 1962, 1965) (FIG. 9.71) (as in Lepidostrobus and Flemingites) (Thomas, 1978). The lateral laminae partially envelop the sporangium with only a slit-like opening on the top. Within the sporangium is one large, functional, trilete

Figure 9.68 Compressed Lepidocarpon cone (Pennsylvanian).

Bar  1 cm.

megaspore and three aborted spores. The wall of the megaspore (called Cystosporites when found dispersed) is unique in that it is constructed of loosely arranged strands of sporopollenin. In some specimens, cellular megagametophytes are preserved, a few containing archegonia. Embryos have been described from Lepidocarpon cones and they are ellipsoidal, unvascularized, and characterized by an isoclinally folded epidermis (Phillips et al., 1975). As the embryo develops, there is a dichotomy of the vascularized axis; one branch of the dichotomy develops into the stigmarian rooting system and the other into the aerial stem. In more mature embryos, what has been termed secondary


Figure 9.71



Margaret K. Balbach.

Figure 9.69 Longitudinal section of Lepidocarpon sporangium (below) and transverse section (above). Arrows indicate lateral laminae (Pennsylvanian). Bar  5 mm.

Figure 9.70 Cross section of Lepidocarpon lomaxi megasporangium showing lateral laminae, megasporangium, and multicellular megagametophyte (From Hirmer, 1927.)

tissues are present. When only the lepidostroboid sporophyll is preserved in the compressed state, the generic name Lepidostrobophyllum is sometimes used (FIGS 9.72, 9.73). Embryos have also been found in the cone Bothrodendrostrobus (FIG. 9.74) (Stubblefield and Rothwell, 1981). In the earliest stage of development, the embryo consists of an unvascularized globular structure embedded within megagametophyte tissue. In more mature specimens, two vascularized appendages extend through the trilete suture, one representing the first shoot, the other the first root. It was initially suggested that embryology in Lepidocarpon and Bothrodendrostrobus was sufficiently different to demonstrate two evolutionary paths within the lycopsids (Stubblefield and Rothwell, 1981). What was originally interpreted as bipolar development in Bothrodendrostrobus is now considered to be a derived condition of the unipolar developmental pattern demonstrated in Lepidocarpon (Rothwell and Erwin, 1985). Another megasporangiate cone that had an unusually ornamented megaspore is Caudatocorpus. This cone type was apparently monosporangiate, with helically arranged megasporophylls lacking lateral laminae (BrackHanes, 1981). The sporangium is large and the wall is constructed of columnar cells. Each sporangium contains a tetrad of megaspores with one large (FIG. 9.75), presumably



functional spore ( 4 mm long) and three smaller (200– 500 μm) aborted spores. If dispersed, these spores would be included in the genus Lagenicula (Scott and Hemsley, 1993). The sporoderm of the functional megaspore is 10 μm thick and has two layers. The outer surface is covered with numerous spines about 50 μm long; on the proximal surface is a conspicuous apical prominence. The four spores are enclosed in a granulose spongy structure that represents a

Figure 9.72 Detail of Lepidostrobophyllum showing sporangium attachment scar (arrow) (Pennsylvanian). Bar  1 cm.

Figure 9.74 Bothrodendrostrobus embryo showing two appendages extending from spore wall. Arrow indicates vascular element (Pennsylvanian). Bar  150 μm. (From Stubblefield and Rothwell, 1981.)

9.73 Compressed Lepidostrobophyllum showing the point of sporangium attachment (arrow) (Pennsylvanian). Bar  1 cm.


Figure 9.75 Section of Caudatocorpus arnoldii megaspor-

angium showing spiny megaspores of the Lagenicula type (Pennsylvanian). Bar  400 μm. (From Brack-Hanes, 1981.)


distal, winglike attachment to the large functional spore. This extension may have been involved somehow in dispersal, or it may have functioned to orient the apical prominence of the trilete suture so as to enhance the possibility of fertilization. Achlamydocarpon is a monosporangiate cone with reduced lateral laminae and a single large, functional megaspore in each sporangium. Species of both Lepidodendron and Diaphorodendron are known to have borne this cone type (DiMichele, 1985). In Achlamydocarpon, the orientation of the trilete suture is toward the cone axis rather than away, as in Lepidocarpon. The suture of the functional spore in Achlamydocarpon is covered by a massa (FIG. 9.76) or cap of sporopollenin that may have functioned to protect the developing gametophyte and perhaps to help retain moisture in the region of the suture (Taylor and Brack-Hanes, 1976). In Lepidocarpon cones, this protection could have been provided by the conspicuous lateral laminae of the sporophyll, whereas in Achlamydocarpon, the developing megagametophyte may have been protected by both the reverse orientation of the proximal suture and the sporopollenin cap on the megaspore. Microsporangiate cones are assigned to A. varius on the basis of similarities in epidermal structure, pedicel alations, and other histologic details (Leisman and Phillips, 1979). The trilete microspores average 64 μm in diameter and exhibit scattered papillae over their distal surfaces, which may represent tapetal residues in the form of orbicules. If found dispersed, such grains would be included in the genus Cappasporites (Ravn et al., 1986). Achlamydocarpon pingquanensis is a megasporangium–sporophyll unit from the Lower Permian of China (Y. L. Zhou et al., 2006). The structure is 1.6 cm long and contains a large, presumably functional megaspore of the Cystosporites type, and three

Figure 9.76 Aborted megaspore of Achlamydocarpon varius with proximal massa (arrow) (Pennsylvanian). Bar  250 μm.



smaller abortive spores, each with a massa near the proximal suture. Based on the organization of A. pingquanensis, it is suggested that this disseminule adds additional support to the idea that there is a group of arborescent lepidodendrids from China that is distinct from those in Euramerican Carboniferous deposits. A single large megaspore within each sporangium is also a feature of the cone Suavitas imbricata collected from Upper Pennsylvanian marine deposits of Texas, USA (Rice et al., 1996). The cone is permineralized and the sporangium is located at the distal end of the sporophyll. Although the affinities remain conjectural, the analysis of characters suggests some relationship with the rhizomorphic lycopsids. Achlamydocarpon is believed to represent the cone of several different species of Diaphorodendron (DiMichele, 1981, 1985), based on an analysis of several hundred permineralized specimens of Diaphorodendron from Lower and Middle Pennsylvanian rocks. DiMichele recognizes different morphological groups among the Euramerican forms of lepidodendrids, each distinct relative to reproduction, habitat, and evolutionary history. One group, consisting of D. vasculare, D. scleroticum, and D. phillipsii, included trees 8–20 m tall (FIG. 9.77) that had deciduous lateral branches bearing cones of the A. varius type. Synchysidendron trees were smaller (from 10 to 15 m tall) and produced A. varius cones, but these occurred in large numbers near branch tips toward the end of the growing season. DiMichele (1981) suggested that the coal-swamp environments may have acted as evolutionary refugia for some of the arborescent lycopsids, such as Diaphorodendron and Lepidophloios, a habitat preference no doubt dictated by their reproductive biology, which was well adapted for aquatic dispersal. DiMichele (1980) also speculated that speciation in this group may have taken place outside the swamp habitat. Arborescent lycopsid megaspores of several types have been examined at the fine-structural level in an attempt to determine the affinities of the spores and also to investigate the development of the spore wall (T. Taylor, 1974). Wilson Taylor (1990) was able to correlate megaspore ultrastructure with the dispersal strategy in these lycopsids. His study indicates that some of the Carboniferous megaspores share both developmental and dispersal features with some modern species of Selaginella. Others appear to possess a uniquely organized sporoderm pattern (FIG. 9.78) that reflects the degree to which megaspores enlarge within the sporangium. W. Taylor, (1989) distinguished three basic types of construction (laminar, laterally fused spherules, and ordered units) in the walls of Selaginella. All megaspores possess an



Figure 9.78 Detail of sporopollenin units forming the wall of a functional Lepidocarpon megaspore (Pennsylvanian). Bar  3 μm.


Figure 9.77 Suggested reconstructions of several arborescent lycopsids (left to right): Diaphorodendron vasculare, D. scleroticum, D. phillipsii, and Synchysidendron dicentricum. (From DiMichele, 1981.)

inner separable layer, which may be involved in regulating water balance. It is also important to characterize megaspore wall development at the ultrastructural level for its systematic value (Hemsley and Scott, 1989; Hemsley and Galtier, 1991). This type of study is useful in identifying the parent plants of dispersed spores, and the floral composition of particular assemblages where megafossils are absent or poorly preserved. Ultrastructural studies also provide important information about the development and evolution of lycopsid spore walls (Glasspool et al., 2000).

Figure 9.79 Polar view of Flemingites schopfii archegonium (arrow) showing four neck cells (N) (Pennsylvanian). Bar  100 μm.

GAMETOPHYTES Knowledge about the gametophyte generation of the lepidodendrid arborescent lycopsids is generally meager, and is based on only a few specimens (Renault, 1893; Gordon, 1908, 1910; MacLean, 1912). One interesting feature of Flemingites schopfii cones (discussed earlier) is the exquisite preservation of both the micro- and megagametophyte phases (Brack-Hanes, 1978). Within some of the megaspores near the trilete suture is a parenchymatous, cellular megagametophyte (FIG. 9.79). Some of the surface cells of the megagametophyte bear elongated tufts of rhizoids that extend from the trilete suture and actually penetrate the sporangium




Figure 9.80 Polar view of several archegonia inside arbores-

cent lycopsid megaspore (Mississippian). Bar  80 μm. (Courtesy J. Galtier.)

wall. Other megagametophytes possessed archegonia at the time of fossilization, and several archegonial necks have been described interspersed among the rhizoids. Archegonia have from one to three tiers of neck cells (FIG. 9.82) and, in a few specimens, an enlarged cell, suggestive of an archegonial venter, occurs beneath the neck cells. Some microspores in the distal sporangia in F. schopfii reveal stages in the development of the microgametophyte, including partitions suggestive of the antheridial initial and prothallial cells. Some contain material that morphologically resembles chromosomes (Brack-Hanes and Vaughn, 1978) (FIG. 1.25). When compared with the gametophytes of existing lycopsids, F. schopfii has microgametophytes that are more similar to those of extant Selaginella, whereas the structure of the megagametophytes more closely corresponds to that of Isoetes. Other well-preserved lepidodendrid megagametophytes (FIGS. 9.80–9.82) have been discovered in spores assignable to Lepidodendron esnostense or L. rhodumnense that occur in late Viséan (uppermost Mississippian) cherts from central France (Galtier, 1964b, 1970a, b); these occur in dispersed spores as well as in specimens still within the megasporangium (FIG. 9.81). The megagametophytes are multicellular structures that develop inside the megaspore wall (endosporic gametophyte development). As they mature, however, they protrude from the spore in the region of the trilete suture, where they form a mass of tissue, within which several archegonia are produced (FIG. 9.82); rhizoids appear to be lacking. Archegonia are embedded in gametophytic tissue, with only the uppermost ring of neck canal cells protruding from the surface. The microgametophyte of this taxon remains unknown to date.

Figure 9.81 Gametophyte tissue inside arborescent lycopsid megaspore (arrows) still inside megasporangium (white arrow) (Mississippian). Bar  330 μm. (Courtesy J. Galtier.)


9.82 Megagametophyte tissue containing several archegonia (arrows) rupturing lycopsid megaspore wall (S). Bar  100 μm. (Courtesy J. Galtier.)



Sigillaria is another important arborescent Carboniferous– Permian lycopsid, which did not branch profusely and was



Figure 9.84 Sigillaria mammillaris, stem surface with leaf

bases (Pennsylvanian). Bar  2 cm. (Courtesy BSPG.)

not as large as the lepidodendrids. Although some specimens have been reported to be 30 m tall (FIG. 9.83), it is probable that most sigillarians were 20 m tall. The absence of extensive branching and the structure of the leaf bases are the principal features that distinguish Sigillaria from other arborescent lycopsids. Sigillarian leaf bases are typically hexagonal in outline (FIG. 9.84), although some may be elliptical. Although they are helically arranged, they often appear to be aligned in vertical rows. The actual leaf scars are generally elliptical (FIG. 9.85), with a central leaf-trace scar (vascular-bundle scar) flanked by two large parichnos scars. The vascular bundle is v-shaped and may separate into two strands. A ligule scar is present above the leaf scar.

Figure 9.83 Suggested reconstruction of Sigillaria tree (Pennsylvanian). (Courtesy D. A. Eggert.)

LEAF BASES Subgenera of Sigillaria have been established to encompass the various configurations and arrangements of the leaf bases on the surface of the stems. Sigillaria subg. Eusigillaria includes forms with ribbed stem surfaces, and these have been divided into two sections. In the section Rhytidolepis, leaf bases and ribs are separated and the furrows between adjacent ribs are straight or nearly so. In the section Favularia, leaf bases and ribs are close together and the furrows have a zigzag configuration. Forms included in Sigillaria subg. Subsigillaria lack ribs. There are also two sections within this subgenus: Leiodermaria has widely






Figure 9.85 Diagram of Sigillaria leaf bases showing posi-

tion of ligule (L), parichnos (P), and vascular bundle (VB) (Pennsylvanian). (From Taylor and Taylor, 1993.)

Figure 9.86 Sigillaria brardii, transverse section through a permineralized stem (Permian). Bar  5 cm. (Courtesy BSPG.)

separated vertical rows of leaf scars with no raised cushions, whereas leaf bases in Clathraria are closely arranged. LEAVES Most sigillarians must have been rather unusual looking plants with little distal branching and a large number of closely spaced, elongated leaves arising from the top of the trunk. Sigillariophyllum and Sigillariopsis leaves are similar to those of Lepidophylloides except that they may be vascularized by two laterally flattened strands instead of one. On the lower surface are two longitudinal grooves lined with conspicuous trichomes. Stomata are arranged in rows, and the guard cells are sunken. Cyperites is a morphogenus used for isolated linear leaves, usually 1 cm wide, which are butterflyor X-shaped in transverse section (Rex, 1983) and generally thought to represent compressed sigillarian leaves (Doubinger et al., 1995). Although the exact reason for the configuration in cross section is not clear, it is suggested that the two abaxial grooves may have contributed to this morphology. STEM STRUCTURE Although compressed remains of sigillarians are relatively common in Carboniferous rocks, structurally preserved stems are rare (FIG. 9.86). In Sigillaria approximata (FIG. 9.87) (Upper Pennsylvanian), the central portion of the stem consists of a parenchymatous pith surrounded by a continuous band of primary xylem (Delevoryas, 1957; Guo and Tian, 1994). In cross section, the outer edge of the exarch primary xylem appears sinuous, with leaf traces originating from the furrows. Metaxylem tracheids possess fimbrils between the scalariform bars. Relatively little secondary

Figure 9.87 Transverse section of Sigillaria approximata stem

(Pennsylvanian). Bar  1 cm.

xylem is produced, and it consists of scalariform tracheids and narrow vascular rays. Distribution of cortical tissues is similar to that described for Diaphorodendron, and tangentially banded periderm is common in these plants (FIG. 9.88); the periderm often contains concentric bands of presumably secretory cells. In addition, tangentially expanded cells form distinct clusters that appear spindle shaped in transverse section. Extending radially through the periderm are pairs of cylindrical to laterally flattened strands of parichnos tissue that can be related to the parichnos scars on the leaf bases and that may have functioned as an aeration tissue.



Figure 9.88 Periderm of Sigillaria in coal (Pennsylvanian).

Bar  1 mm. (From Winston, 1989.)

Sometimes Sigillaria is found in a partially decorticated state called Syringodendron. The outer surface exhibits vertical rows of large, often double elliptical scars that resemble rabbit tracks. These scars represent the parichnos strands as seen in tangential section, at a level of stem decortication. UNDERGROUND ORGANS Underground organs of Sigillaria are similar to the stigmarian system of the other lepidodendrids, but there are a few anatomic differences (Eggert, 1972). In stigmarian axes of Sigillaria, the pith is relatively narrow in proportion to the diameter of the stele, and consists of an outer zone of mixed tracheids and parenchyma and a central zone of parenchyma. The cortex is also narrow and consists of two primary zones. Secondary cortical development involves concentric rings of meristematic cells in the outer cortex, whereas in the underground parts of Diaphorodendron, periderm was produced from a single, central meristematic layer. In the lateral appendages (rootlets), the connective is continuous, unlike the organization in Stigmaria. The genus Stigmariopsis is known to have been the underground part of some subsigillarians and is distinguished from Stigmaria principally on the basis of unequal branching in which the smaller branch is directed downward (Eggert, 1972). One might be inclined to view the subtle differences between the underground parts of the sigillarians and Stigmaria as more apparent than real. To correlate underground parts with particular stem taxa, it is necessary to collect at sites where not all lycopsid taxa are present. Frankenberg and Eggert (1969) were able to characterize Stigmaria from several Middle Pennsylvanian coal-ball localities where Sigillaria was

Figure 9.89 Longitudinal section of Mazocarpon oedipternum cone. Arrow indicates megasporangium with megaspores (Pennsylvanian). Bar  5 mm.

absent. Eggert (1972), in turn, detailed the nature of the sigillarian underground system from an Upper Pennsylvanian site in which Sigillaria was present, whereas the other arborescent lepidodendrids were not, thus making it possible to correlate the aboveground and subterranean parts of the same plant even though they were not organically attached. REPRODUCTIVE BIOLOGY Sigillaria was heterosporous and produced monosporangiate cones. Evidence of the attachment of the cones to the parent plant comes not only from compressed specimens but also from structurally preserved stems that bear persistent cone peduncles interspersed among the leaf bases (Delevoryas, 1957). Mazocarpon oedipternum is a common Late Pennsylvanian taxon in North America (FIG. 9.89) (Schopf, 1941). The cones are 1.2 cm in diameter and frequently reach 10 cm in length. Sporophylls are arranged in a low helix or pseudowhorl, with the laminae forming conspicuous dorsal heels. The distal ends of sporophylls are relatively short. Megasporangia are roughly triangular as seen in a radial


section and contain a central pad of parenchymatous tissue (subarchesporial pad) with eight megaspores around the periphery (FIG. 9.90). Megaspores are large and trilete, and have been described with short archegonial necks extending from the proximal suture. Microsporangiate cones (FIG. 9.91), also called Mazocarpon, contain trilete spores that average 60 μm in diameter. Mazocarpon villosum is a Late Pennsylvanian species that is 2.2 cm in diameter (Pigg,



Figure 9.90 Transverse section of Mazocarpon oedipternum

megasporangiate cone showing parenchymatous subarchesporial pad (S) in two sporangia. Note megaspores (arrows) between pad and sporangium wall (Pennsylvanian). Bar  3 mm.



1983). In this species, only immature megaspores have been found in apical sporangia. Megagametophytes have been described in several species of Mazocarpon; these consist of prothallial tissue with rhizoids and archegonia up to 65 μm long, each with three tiers of neck cells. Reproductively, the sigillarians are far more complex than originally thought. Pigg (1983) suggested three different types of megaspore dispersal in the genus. In one type, exemplified by Mazocarpon villosum, spores are believed to have developed rapidly and then been shed from the sporangium. In another type, for example M. oedipternum, the sporophylls are retained on the cone at maturity and the megaspores were dispersed (some with megagametophytes) when the sporangial wall broke down. This contrasts with M. pettycurense and M. cashii, which are hypothesized to have dispersed megaspores with their attached sporophylls by fragmentation of the cones. The presence of megaspores containing embryos embedded within the intersporangial tissue of the megasporangium has been used as evidence to suggest that apomixis occurred in this group (DiMichele and Phillips, 1985). A small amount of sterile tissue has been described in the microsporangia of M. bensonii (Feng and Rothwell, 1989). In this species the trilete microspores range from 48–54 μm and the sporoderm is constructed of a dense reticulum of interconnected rodlets. This sporoderm organization is similar to the fine structure of the megaspore wall reported in M. oedipternum (W. Taylor, 1990). Sigillariostrobus is a compressed cone believed to have been produced by sigillarian lycopsids. Many of the species described to date may represent preservational states of the Mazocarpon-type cone and have been correlated with the latter on the basis of size, organization of the sporophylls, and morphology of the spores. Some specimens of Sigillariostrobus may attain lengths of 30 cm. Despite the advances in recent years, there is still much to be learned about the biology of sigillarian plants. They represent one of the arborescent lycopsid groups that flourished during the Late Pennsylvanian. The sigillarians are believed to have inhabited near-swamp environments that were slightly drier than those dominated by the other arborescent lycopsids (DiMichele and Phillips, 1985). Other Lepidodendrid Genera

Figure 9.91 Transverse section of Mazocarpon oedipternum microsporangiate cone (Pennsylvanian). Bar  5 mm.

Some lycopsid stems have leaf bases that are inconspicuous or lacking, and these are sometimes difficult to distinguish from decortication stages. For example, in Bothrodendron (FIG. 9.92), an arborescent lycopsid that was 10 m tall, the leaf scars are not on raised cushions, but are flush with the surface of the stem (Wnuk, 1989). In some specimens,



Figure 9.92 Bothrodendron minutifolium, branched twigs (partially foliated) and a cone (arrow) (Pennsylvanian). Bar  2 cm. (Courtesy BSPG.)

Figure 9.93 Cross section of Paralycopodites (Anabathra) showing primary and secondary xylem, and departing branch trace (arrow) (Pennsylvanian). Bar  6.5 mm. (Courtesy W. A. DiMichele.)

the scars are arranged in an almost whorled pattern; in others, they are borne in a low helix. The ligulate microphylls are narrow and measure up to 25 cm in length. Cones, known from both compressions and petrifactions, consist of sporangia borne adaxially on sporophylls whose distal laminae are greatly elongated. The cones (Bothrodendrostrobus) are bisporangiate, with trilete microspores toward the apex and megasporangia toward the base, each with up to 20 megaspores. Megaspores of Bothrodendron have numerous spines that cover their surfaces. Paralycopodites (FIG. 9.93) differs from both Diaphorodendron and Lepidophloios in having reduced, persistent scalelike leaves that appear morphologically intermediate between these two genera (DiMichele, 1980). Stratigraphically, the genus extends from the Lower Carboniferous to the Middle Pennsylvanian (Westphalian D).

Paralycopodites has been reconstructed as a small tree that produced deciduous lateral branches. The diameter of the siphonostele in P. breviformis was 3 cm; the most distal branches were protostelic. Vegetative remains are consistently associated with strobili of the Flemingites type. Pearson (1986) has suggested that Paralycopodites is the same plant that was initially described by Witham (1833) as Anabathra. Another common lycopsid believed to be a progenitor of, or have affinities with, the lepidodendrids is Sublepidodendron songziense from the Upper Devonian of China (Q. Wang et al., 2002, 2003b). It has small, vertically oriented leaf bases arranged in a tight spiral arising from a stigmarian rhizomorph; branch scars are of the Ulodendron type. The stem has an ectophloic siphonostele with fibers in the pith. Strobili are of the Lepidostrobus type and borne at the tips of lateral branches (Q. Wang et al., 2003a). A second species in that genus, for which stems and branches (described as S. wusihense), and strobili (described as L. grabaui) have been reassembled from isolated parts, is Sublepidodendron grabaui from the Upper Devonian of the Wutung Formation (Famennian) of China (Y. Wang and Xu, 2005). A cladistic analysis of Sublepidodendron and related taxa suggests that members of this genus possess many synapomorphies with phylogenetically more advanced genera in the families Sigillariaceae, Lepidodendraceae, and Diaphorodendraceae, rather than the order Protolepidodendrales as previously thought (Q. Wang et al., 2003b). An interesting strobilar organization occurs in the Late Devonian genus Bisporangiostrobus. In this taxon, a branch dichotomizes and each dichotomy terminates in an eligulate, bisporangiate cone (Chitaley and McGregor, 1988). Sporangia are attached to the adaxial surface of sporophylls which are helically arranged on the cone axis. The cone is a diminutive Flemingites with apical microsporangia and basal megasporangia. Megaspores are 1 mm in diameter and of the Duosporites type, whereas the smaller microspores conform to the generic diagnosis of Geminospora. The affinities of this interesting cone are suggested as being close to Jurinodendron (Cyclostigma; see Doweld, 2001), a Late Devonian arborescent lycopsid that was up to 8 m tall. Jurinodendron stems often occur as decorticated axes of the Knorria preservation type (FIG. 9.43), and the plant is characterized by a distal crown of dichotomizing leafy branches. Underground organs are Stigmaria-like branch systems with appendages (Schweitzer, 1990). Sporophylls are arranged into distinct cones positioned terminally on ultimate branches (Chaloner, 1968). Jurinodendron appears to have been a widespread taxon, known from Upper Devonian and




Figure 9.94 Suggested reconstruction of the distal branches

of Valmeyerodendron triangularifolium (Mississippian). (Based on Jennings, 1972; reproduced from Taylor and Taylor, 1993.)

Early Mississippian deposits throughout the world, including Germany (Mägdefrau, 1936), Bear Island, Spitsbergen (Schweitzer, 1969, 2006; Murašov and Mokin, 1979), China (Feng et al., 2004), Japan (Kimura et al., 1986), and the Himalayas (Pal, 1977). Leptophloeum rhombicum is an arborescent lycopsid which was distributed worldwide during the Late Devonian (see references in Q. Wang et al., 2005). Based on literature data and material from the Frasnian (Upper Devonian) Huangchiateng Formation of Hubei, China, Q. Wang et al. (2005) provided a reconstruction that depicts the plant as a monopodial tree, 10–25 m tall and 0.3–0.4 m wide at the base, which produced lateral branching systems by pseudomonopodial branching of the trunk, rather than by equal ramifications as formerly thought. Branches grew by means of isotomous dichotomies. The lateral underground portion is believed to have been Stigmaria like. Valmeyerodendron (FIG. 9.94) is a Mississippian form which is interpreted as transitional between the Devonian lycopsids, which generally lack leaf bases, and the

Figure 9.95 Suggested reconstruction of Eskdalia fimbriophylla leaf cushion (Mississippian). (From Taylor and Taylor, 1993.)

Carboniferous ones, which are ligulate and possess parichnos scars (Jennings, 1972). Only compressed specimens are known, and they include stems up to 3 cm in diameter. Helically arranged, quadrangular–hexagonal leaf cushions cover the stems. Each cushion bears a rhombic leaf scar at its apex; ligule and parichnos scars are absent. Unlike the majority of arborescent lycopsids, which possess narrow, linear leaves, Valmeyerodendron has leaves that are nearly triangular in outline with a constricted base and attenuated apex. Eskdalia is a Mississippian ligulate lycopsid with expanded leaf cushions (FIG. 9.95), each with a conspicuous keel (FIG. 9.95) (Thomas and Meyen, 1984a). Rowe (1988a) indicated that stomata are absent from the cushions in E. variabilis and suggested that they functioned in support, rather than as photosynthetic organs, as has been documented for certain arborescent lycopsids (Chaloner and Meyer-Berthaud, 1983).



Lycopodiales The Lycopodiales includes homosporous, eligulate, usually dichotomously branched herbaceous plants. The group is represented today by four genera, Lycopodium (with 476 species; sometimes subdivided into two to several genera), Huperzia (439 species), Lycopodiella (41 species), and Phylloglossum (1 species). These are generally small herbaceous plants covered with scalelike microphylls; they have true roots, which arise adventitiously from a horizontal rhizome. All are homosporous and today the order is cosmopolitan in distribution. The fossil members are not well understood, and although the group itself is ancient, there is some ambiguity as to whether some extant members are of recent or ancient origin (Wikström and Kenrick, 2001). The generic name Lycopodites was first used to describe some Cenozoic axes bearing small, scalelike leaves. The fossils were later determined to be fragments of conifer shoots. Today the morphogenus includes axes with helically arranged or pseudowhorled scale leaves and, if present, sporangia on the adaxial leaf surface, in the axil of foliage leaves, or in monosporangiate strobili. Specimens of Lycopodites (FIG. 9.96) have been described from rocks ranging from Devonian to Pleistocene (Kräusel and Weyland, 1937; Harris, 1976a; Krassilov, 1978; Rigby, 1978b), and they include forms that are both isophyllous (one type of leaf) and anisophyllous (two types of leaves). There is a report of Lycopodites from the Paraná Basin in South America (Ricardi-Branco and Bernardes-de-Oliveira, 2002; Jasper et al., 2006) and if accurate, it represents the only occurrence of the genus in Gondwana, with the possible exception of the enigmatic Lycopodites amazonica that has been described from the Middle Devonian of Brazil (Dolianiti, 1967). One difficulty in dealing with fossils of the Lycopodites type is in distinguishing these remains from the distal twigs of members of Lepidodendrales. When sporangia are scattered along the stem in association with leaves resembling vegetative leaves, identification is easy, but most known specimens of Lycopodites consist only of vegetative remains. The absence of ligules is another feature that can be used to distinguish lepidodendrids, but even in exceptionally well-preserved specimens, these structures are often difficult to identify. In addition, as some extant species have strobili, it is conceivable that some of the small, apparently microsporangiate Lepidostrobus species may, in fact, represent a fossil herbaceous plant with a Lycopodium-type cone. The potential confusion in delimiting members of this group from the lepidodendrids will no doubt continue until structurally preserved specimens can be correlated with compression fossils

Figure 9.96

Lycopodites sp. (Pennsylvanian). Bar  2 cm.

or until more complete plants can be reconstructed based on compression specimens. Ultimately, epidermal features, including the distribution and type of stomata, may be useful in separating distal twigs of the arborescent lycopsids from axes of the herbaceous forms. Oxroadia gracilis includes small, dichotomously branched lycopsid axes that lack distinct leaf cushions but possess decurrent leaf bases (Alvin, 1965; Bateman, 1992). The genus is based on structurally preserved specimens from the Mississippian (Calciferous Sandstone Series, Scotland) and is thought to have represented an herbaceous lycopsid. The stem contains an exarch protostele with mesarch traces arranged in a helical manner. Microphylls are eligulate and vascularized by a single strand; parichnos is not present. Similar histology and association in the same block of material are used as the basis for assigning a small (4 cm long) cone to the stem remains. Sporangia are elongate and borne on sporophylls that have downward-projecting heels. A massive parenchymatous pad of tissue extends from the surface


Figure 9.97 Synlycostrobus tyrmensis cone showing subtending bracts (Jurassic). Bar  5 mm. (Courtesy V. A. Krassilov.)

of the sporophyll and partially fills the sporangium cavity. Nothing is known about the spores. The genus is regarded as a herbaceous lycopsid, rather than the distal branches of an arborescent form, because there are no secondary tissues in the vascular system and cortex. Based on our current knowledge of stem development in the arborescent lycopsids, however, this distinction may be less certain. Synlycostrobus tyrmensis is an interesting lycopsid from the Late Jurassic to the Early Cretaceous of the Bureja Basin (Siberia) that has an unusual arrangement of cones (Krassilov, 1978). It is thought to have been a creeping plant, probably not unlike modern Lycopodium. The ligulate leaves are scalelike and anisophyllous. Cones are borne on what have been termed fertile shoots (FIG. 9.97), each located in the axil of a scalelike leaf or bract. The cones are small (5 mm long) and consist of 20 helically arranged sporophylls. Each sporophyll has a conspicuous distal lamina and a downward-projecting heel that partially covers the sporangium below. Only sporangia containing radial, trilete spores (20–22 μm in diameter) have been recovered from the cones, although a single megasporangium with four spores was isolated from the same matrix. Superficially, the fertile branches of Synlycostrobus resemble the primary axis and dwarf shoots (cones) of the



Cordaitales (Chapter 20). Associated in the same rocks are well-preserved vegetative axes that are placed in the genus Lycopodites. Although morphologic and cuticular features suggest affinities with that taxon, the presence of a ligule associated with each leaf may warrant inclusion within the Selaginellales (Skog and Hill, 1992). Onychiopsis psilotoides is a fossil from the Lower Cretaceous (Wealden) of England with a complex nomenclatural history. Originally thought to have affinities within the polypodiaceous ferns, Skog (1986) reinterpreted some specimens as herbaceous lycopsids and transferred these to a new genus, Tanydorus (Skog, 1986). She later noted, however (Skog, 1990), that the type specimen for O. psilotoides was missing and thus was unable to confirm that it conformed to the genus Tanydorus. Skog (1990) therefore designated a new type and named the lycopsid taxon Wathenia. This genus is characterized by helically arranged, decurrent, simple leaves, each with a single vein. Leaves have acute tips and sporangia occur in the axil of the leaves. Spores are trilete and foveolate (ornamented with small pits). Onychiopsis remains a genus for fossil ferns and is discussed in Chapter 11. Small (3 mm in diameter), eligulate permineralized cones from the Middle Pennsylvanian of Kansas are also thought to have been produced by herbaceous lycopsids (Baxter, 1971a). Carinostrobus has an exarch protostele and helically arranged sporophylls. The sporangium was attached to the adaxial surface by a delicate pedicel; spores were small (20– 22 μm), trilete, and covered by minute spines. Another permineralized Carbonifereous cone in which the sporangium is attached by a short stalk is Spencerites (Berridge, 1905) (FIG. 9.98). Spores of S. moorei are triangular in outline and characterized by an equatorial bladder that is also triangular in outline (Leisman, 1962; Leisman and Stidd, 1967). It is difficult to determine whether cones with Spencersporites spores are mono- or bisporangiate. Several types of fossil lycopsids are known from Gondwana. One of the better known taxa is Bumbudendron, a small, eligulate lycopsid with stems up to 3.5 cm wide (Archangelsky et al., 1981a). The specimens come from the Paganzo Basin in west-central Argentina and are Pennsylvanian in age. Leaf cushions are helically arranged with the leaf trace in the upper third of the cushion. Beneath the trace is an elongate structure that represents an infrafoliar bladder, a small depression or elongate mound just beneath the vascular-bundle scar. Topographically, it occupies the position of the two parichnos scars that characterize Diaphorodendron leaf cushions. The infrafoliar bladder is present on eligulate upper Paleozoic lycopsids and has been used as a systematic character in Thomas and Meyen’s



Figure 9.98 Emily M. Berridge.

(1984b) classification of impression–compression lycopsid stem genera. Fertile branches in the same rocks that contain Bumbudendron consist of reflexed, keeled sporophylls. From the same rocks these authors also described Malanzania. The stems of this lycopsid are narrow (1.3 cm wide) and contain widely spaced spines. Brasilodendron is an arborescent Permian lycopsid bearing persistent leaves up to 4 cm long (Chaloner et al., 1979; Jasper et al., 2006). On the abaxial surface of the leaf are two stomatal bands containing numerous, sunken stomata. Leaf bases are fusiform and no ligule has been observed. Several megaspores were recovered from the same matrix as the vegetative stems. They range from 800–1340 μm in diameter and are most similar to Lagenoisporites brasiliensis. Based on vegetative remains, Brasilodendron appears most similar to Ulodendron. As no ligule or ligule scar has been observed, Brasilodendron is discussed here in the Lycopodiales rather than with the other arborescent forms in the Lepidodendrales. Absence is not definitive evidence, however, as ligules may not have been seen due to preservation or to the persistent nature of the leaves. Brasilodendron is also comparable to Azaniadendron, an Early Permian, presumably eligulate, lycopsid from South Africa (Rayner, 1986). The vegetative axes in this taxon exhibit circular–oval leaf cushions with an elongate vascular-bundle scar. Cones are bisporangiate; each sporangium contains a single tetrad of megaspores (2 mm in diameter) of the Triletes type and microspores assignable to Zinjisporites.

Figure 9.99 Barry A. Thomas.

One of the few permineralized lycopsids known from late Paleozoic rocks of Gondwana is Eligodendron (Archangelsky and de la Sota, 1966). The single specimen comes from the Permian of Bolivia and possesses a parenchymatous pith and exarch primary xylem. The cortex is three-parted with large lacunae in the inner zone. The specimen lacks evidence of ligules and parichnos tissue.

Selaginellales Another herbaceous group that coexisted with the Carboniferous arborescent lycopsids is the Selaginellales (Carboniferous evidence reviewed in Thomas, 1997) (FIG. 9.99), today represented by a single genus (Selaginella) that includes around 500 species. Plants assigned to this order are herbaceous, ligulate, and heterosporous (FIGS. 9.100, 9.101), and are characterized by a creeping to erect habit. Modern members generally show leaves in four ranks, with two ranks of smaller leaves (anisophylly). Megagametophytes exhibit endosporic development (FIG. 9.102). Some extant species are capable of surviving extended periods of drought, for example S. lepidophylla,


Figure 9.100 Selaginella selaginoides (Extant). (Courtesy

H. Bültmann.)

Figure 9.101 Partial longitudinal section of Selaginella cone showing microsporangium (left) and megasporangium (right) (Extant). Bar  1 mm.



the so-called resurrection plant. Recent studies based on gene sequence data suggest that the xeric and woodland species evolved from ancestors in the humid tropics, but today there is less resolution in classification schemes than there was earlier understood (Korall et al., 1999; Korall and Kenrick, 2004). The best-known fossil member, Selaginella fraipontii, represents an excellent example of the reconstruction of a complete plant based on isolated organs. For many years the generic name Paurodendron has been used for small (4 mm in diameter), anatomically preserved stems that are relatively common in coal balls from certain localities (Fry, 1954). The axes bear helically arranged, ligulate microphylls and are characterized by exarch protosteles that are stellate in cross section. The underground portion was subsequently discovered attached to a Paurodendron axis (Phillips and Leisman, 1966). It consists of an unbranched and unlobed, clavate rhizophore (root-bearing organ) from which helically arranged, monarch roots arose (FIG. 9.103). Despite the small size of the rhizophore stele, some secondary xylem is noted (FIG. 9.104). Reproductive parts of the plant are bisporangiate cones that were initially described under the binomial Selaginellites crassicinctus (FIG. 9.105). Cones of this type are 1.2 cm long and 5 mm in diameter. Sporophylls are ligulate and attached to the axis in alternating verticils (whorls), each with four sporophylls. Megasporangia are restricted to the basal region of the cone, with each sporangium containing four or occasionally up to seven megaspores of the Triletes type. Microspores are assignable to the sporae dispersae genus Cirratriradites. Ultrastructural features of the megaspores are shared by modern members of both the Selaginellales and Isoetales (Taylor and Taylor, 1988, 1990). There have been a number of studies that have focused on determining the developmental processes involved in the formation of the complex spore wall (Hemsley et al., 1994). Demonstration of organic attachment of Selaginellites cones to Paurodendron axes has made it possible to reconstruct the entire plant, which is now referred to as S. fraipontii (Schlanker and Leisman, 1969). The plant is reconstructed as herbaceous, sprawling, and sparsely branched; it produced cones terminally. There is some suggestion that the S. fraipontii was determinate in growth, much like arborescent lycopsids. The species is known throughout the Carboniferous, and is almost identical morphologically with many of the Selaginella species that inhabit relatively moist environments today. Extant Selaginella and the fossil differ, however, in the organization of their underground parts. In living Selaginella, roots are primarily adventitious, whereas in S. fraipontii, they are formed between adjacent older roots, resulting in a specific pattern of root formation. The



Strobilus Microspore mother cells undergo meiosis, producing Ligule Young sporophyte

Microsporangium Sporophyte Megaspore mother cells undergo meiosis, producing Zygote 2n


Meiosis n Megaspore Megasporophyll Megaspores Archegonium containing egg

Sperm Microsporophyll

Female gametophyte


Microspore Male gametophyte

Figure 9.102 Life history of Selaginella (Extant). (From Taylor and Taylor, 1993.)

organized production of laterals is a feature commonly associated with stems, not roots. The occurrence of this feature in S. fraipontii suggests homologies with the lobed rooting systems of a number of Paleozoic and Mesozoic lycopsids and may perhaps indicate a closer affinity with the Isoetales (Rothwell and Erwin, 1985; Bateman et al., 1992), or an intermediate status between the Selaginellales and Isoetales (Bateman, 1990). Modern-appearing Selaginella-like axes from the Rhaetian (Late Triassic) of Scania have been described as Selaginellites (Lundblad, 1950a). There are also numerous impression–compression specimens of anisophyllous lycopsids from Carboniferous rocks that have been described as either Selaginellites or placed in the extant genus Selaginella (Thomas 1992, 1997). Some of these, such as Selaginellites gutbieri, are represented as exquisite compressions that show the attachment of bisporangiate cones and details of

the microphylls (Rössler and Buschmann, 1994). In this species, there appear to be at least six to seven megaspores of the Triangulatisporites type in each megasporangium; microspores can be assigned to Cirratriradites. Isolated S. gutbieri bisporangiate cones have also been reported from Late Pennsylvanian deposits in the Czech Republic containing these same dispersed spore taxa (Bek et al., 2001). In Selaginellites primaevus (Selaginella primaeva of Thomas, 1997), from the roof shales in the Saar Coalfield, each megasporangium contains four megaspores of the Triangulatisporites type. Several small lycopsid cones from the Famennian (Upper Devonian) of Belgium exhibit an interesting collection of features unlike those of other known Devonian lycopsid cones (Fairon-Demaret, 1977). Barsostrobus cones are up to 14 cm long and bear helically arranged sporophylls and stalked sporangia. The sporophyll margins are evenly toothed, with the margins slightly enveloping the sporangium. The vascular






Figure 9.103 Suggested reconstruction of Selaginella fraipontii (Pennsylvanian). (Based on Phillips and Leisman, 1966; reproduced from Taylor and Taylor, 1993.)

system is that typical of lycopsid cones, and the traces to the sporophylls have centrifugal and centripetal metaxylem. Spores are 240–320 μm in diameter, trilete, and evenly ornamented. The cones are thought to have been heterosporous, although no microspores have been discovered. Features of this cone suggest affinities with members of the Lycopodiales or Selaginellales; preservation prevents the recognition of ligules. The presence of Williamson striations between the scalariform bars of the metaxylem tracheids resembles some species of Drepanophycus; the Drepanophycales, however, lack heterospory and a strobilar organization of sporangia. Minostrobus chaohuensis is a Late Devonian cone from China that is believed to have been borne by a herbaceous lycopsid. There are four megaspores per sporangium with spores of the Lagenicula type (Y. Wang, 2001). Yuguangia ordinata is a ligulate lycopsid from the Givetian of China with bisporangiate terminal cones (Hao et al., 2007). Leaves of this permineralized form are spiny and arranged in pseudowhorls; megasporophylls contain four megaspores of the Triletes type, whereas the microspores are similar to Acinosporites. The occurrence of Y. ordinata suggests that small heterosporous, ligulate lycopsids had diverged from

Figure 9.104 Longitudinal section of Selaginella fraipontii

(Paurodendron) showing rhizomorph (R), secondary xylem (X) and numerous roots (Pennsylvanian). Bar  3 mm.

the homosporous ligulate grade by the Middle Devonian (Hao et al., 2007). Miadesmia membranacea (FIG. 9.106) is used for isolated cones known only from the Carboniferous of Europe. In some classifications of lycopsids, Miadesmia is included in its own order, Miadesmiales (Thomas and Brack-Hanes, 1984); but as nothing is known about the plant that bore this cone type, we will continue to include it within the Selaginellales until more information is available. The cones contain only megasporophylls which are attached to the axis at right angles (Benson, 1908). Each megasporophyll is 3 mm long and bears a megasporangium that is attached near the proximal end of the sporophyll. Lateral laminae completely envelop the sporangium, except in the distal region. The enveloping sporophyll is divided into elongate,




tentacle-like extensions that project beyond the distal opening. The sporangium is somewhat flattened on the sporophyll so that the opening is directed away from the cone axis. A large ligule is present just distal to each megasporangium. Miadesmia is interesting in that the sporangium contains one large, functional megaspore and some are known with a cellular megagametophyte. In the original description, it was noted that M. membranacea occurs in the same coal balls as specimens of lepidodendrids. The small size of the sporophyll units has suggested to some that Miadesmia may represent a ligulate, heterosporous cone type within the Selaginellales that parallels the highly developed heterospory in the Lepidodendrales. This theory is based on the presence of one functional megaspore per megasporangium in each group and the integument-like morphology of the lateral laminae.



Figure 9.105 Longitudinal section of Selaginellites crassicinc-

tus showing both microsporangia (MI) and megasporangia (ME) (Pennsylvanian). Bar  2 cm.

Figure 9.106 Longitudinal section of Miadesmia membranacea (Pennsylvanian). Bar  1 cm.

The Pleuromeiales are an interesting group of lycopsids, and were originally thought to represent an evolutionary transition from fossil arborescent forms to herbaceous, cormose forms such as extant Isoetes. This hypothesis has been questioned, however, based on presumed Isoetes fossils in the Triassic (Retallack, 1997) and the discovery of cormose forms in the Paleozoic. The pleuromeialeans were smaller than the Lepidodendrales and believed to have been herbaceous or pseudoherbaceous. Plants with a similar growth habit have been described from the Paleozoic, for example Chaloneria (discussed below). Members of the Pleuromeiales have been included in the Lepidodendrales as well as the Isoetales, as they share features with both groups. They are generally unbranched and bear one or more terminal cones; their rooting structures are lobed, cormose, and bear stigmarian-type appendages. Some bear bisporangiate cones, whereas only one size of spore is known from other cones. Sporangia are generally somewhat sunken into the sporophyll. Most have trilete megaspores and monolete microspores like the isoetaleans, although Neuberg (1960b) showed that Pleuromeia rossica contained trilete microspores (Pigg, 1992). Pleuromeia is an exclusively Triassic genus known from localities around the world, including Germany, France, Spain, Russia, China, Japan, Argentina, and Australia. The discovery of Pleuromeia at many different localities throughout the world suggests that the plant may have


inhabited varying habitats. For example, numerous specimens of Pleuromeia occur in Lower Triassic beds north of Sydney, Australia, and Retallack (1975) suggested, based on a detailed analysis of the lithology of the fossil beds, that the plant grew as a coastal halophyte. Specimens from China are believed to have grown in more xeric, inland sites, perhaps near desert oases (Z.-Q.Wang and Wang, 1982), and still others from Australia in habitats that were seasonally wet (Cantrill and Webb, 1998). From the Buntsandstein (Lower Triassic) of the Eifel region in Germany, Fuchs et al. (1991) described and illustrated spectacular large slabs with numerous densely spaced P. sternbergii stems in situ, which indicate that Pleuromeia formed extensive bitypic stands with the fern Anomopteris mougeotii. The seemingly sudden abundance of Pleuromeia in the Early Triassic is remarkable. Looy et al. (1999) suggested that the hypothesized dieback of woody vegetation at the very end of the Permian dramatically affected terrestrial ecosystems, and that lycopsids such as Pleuromeia played a central role in repopulating certain landscapes after the mass extinction. Pleuromeia has an unbranched, erect trunk, up to 2 m tall, and a four-lobed base (rhizomorph) from which helically arranged roots arise (FIG. 9.107). At the apex of P. longicaulis is a crown of elongate ligulate leaves, each with two vascular bundles (Retallack, 1975). Slightly below the attached leaves is a zone of persistent leaf bases that grades into an area of widely separated leaf scars. Decortication stages suggest that there was some secondary tissue production in Pleuromeia, although the absence of well-preserved petrified specimens makes it impossible to determine whether these tissues were vascular or cortical in origin. Reports by Mägdefrau (1931), Hirmer (1933b), and Grauvogel-Stamm (1993, 1999), among others, indicated that the axes of P. sternbergii, the type species of the genus from the Lower Triassic of Germany, were covered in leaf bases to near the base of the trunk (FIG. 9.124). This plant bore leaves of two types in a lax (subhorizontal) position. The rhizomorph is lobed and is characterized by a bilateral furrow system that divides the base typically into four lobes. The pattern of development of the rootlets appears to be like that in Isoetes. A species similar to P. sternbergii, P. obrutschewii (FIG. 9.108), has been reported from the Lower Triassic of the Russian Far East (Krassilov and Zakharov, 1975). At the apex of Pleuromeia is a single, relatively large cone, although it is possible that some species could have produced more than one cone. Support for this theory comes from the occurrence of large numbers of the small cones



Suggested reconstruction of Pleuromeia longicaulis (Triassic). (From Retallack, 1975.)

Figure 9.107

called Cylostrobus (FIG. 9.109) (sometimes misspelled as Cyclostrobus), which is thought to have been produced by Pleuromeia, at the same locality (Helby and Martin, 1965). Earlier reports, based on fragments of cones, suggested that Pleuromeia was a dioecious plant in which microsporangiate and megasporangiate cones were produced on different



9.108 Megasporophyll (arrow) and megasporangium of Pleuromeia obrutschewii containing casts of megaspores (Triassic). Bar  2 cm. (From Krassilov and Zakharov, 1975.)


Figure 9.109 Cylostrobus sp. (Triassic). Bar  1 cm. (Courtesy

D. Cantrill.)

plants. It is now known that some species of Cylostrobus were bisporangiate, with the megasporangia located in the basal portion of the cone (Helby and Martin, 1965). The sporophylls are circular, imbricate, and lack any downward extension in the form of a heel. The large megaspores ( 700 μm) are trilete and ornamented with numerous elongate spines; microspores are monolete and 30 μm in diameter. One bisporangiate cone is C. clavatus from the Early Triassic of Australia (Cantrill and Webb, 1998). Microspores are of the Lundbladispora–Aratrisporites type. Krassilov and Zakharov (1975) have suggested that megasporophylls

Figure 9.110 Compressed cone of Pleuromeia epicharis

(Triassic). Bar  2 cm. (From Z.-Q. Wang and Wang, 1990.)

of pleuromeids were dispersed by water, based on their shape and their abundance in the rocks. This hypothesis parallels that of Phillips (1979) for the dispersal of megasporophylls of many Carboniferous lycopsids (see section “Conclusions”).




Figure 9.111 Fractured cross section of Pleuromeia epicharis

cone (Triassic). Bar  2 cm. (Courtesy Z. Wang.) Figure 9.112 Compressed cone of Skilliostrobus (Triassic).

Bar  1 mm. (From Ash, 1979.)

Other species of Pleuromeia are smaller (Meng, 1996). Pleuromeia jiaochengensis was 50 cm tall (Z.-Q. Wang and Wang, 1982). Specimens include compressions that come from the Early Triassic of Shanxi Province, China. On the unbranched stem are leaf scars to which were attached awl-shaped leaves 3 mm long. At the distal end of the stem is a large cone constructed of obovate sporophylls bearing discoid sporangia. Only megaspores are recorded, and these range up to 500 μm in diameter. Pleuromeia epicharis (FIG. 9.110) is one of the better-known species and is based on material from the Shiqianfeng Group in north China (Wang and Wang, 1990). The specimens include stems, leaves, basal parts, and micro- and megasporangiate cones (FIG. 9.111). Megaspores are of the Banksisporites type. Specimens of P. rossica were transferred to the new genus, Lycomeia (Dobruskina, 1985a). Microspores and megaspores of L. rossica reveal (ultra-)structural characteristics regarded as distinctive of the spores of Isoetales, which adds support to the hypothesis suggesting a close relationship between Pleuromeia and Isoetes (Lugardon et al., 1999, 2000). More recently, Grauvogel-Stamm and Lugardon (2004) have also demonstrated that the spores of P. sternbergii have a number of features characteristic of isoetalean spores. Some workers have included Pleuromeia longicaulis in a new genus, Cylomeia (White, 1981a). It is believed that this plant produced terminal cones of the Skilliostrobus type (FIG. 9.112) (Ash, 1979). This Early Triassic, pedunculate, bisporangiate cone is known from Australia and Tasmania.

It consists of helically arranged, wedge-shaped sporophylls with an adaxial groove containing obovate sporangia. The cone is up to 8 cm in diameter and only about half as long. Microspores are monolete (40 μm) and most similar to Aratrisporites, whereas the trilete megaspores range up to 1.1 mm and are most similar to Horstisporites. Petrified Triassic stems of Chinlea from North America were initially thought to represent osmundaceous ferns, but they are now regarded as lycopsids, possibly related to Pleuromeia (Miller, 1968). The stems contain an ectophloic siphonostele with a distinct perimedullary zone of thinwalled parenchyma. Leaf traces are numerous (up to 165 in one transverse section) and collateral. Pleurocaulis rewanense is a protostelic stem with circular to oval scars that may represent a decorticated stem stage (Cantrill and Webb, 1998). Ferganodendron is a Triassic genus that resembles Pleuromeia and Sigillaria in many respects (Dobruskina, 1974). The trunk of the plant varies from 20 to 30 cm in diameter and is covered with numerous, elliptical–rhombohedral leaf bases that are helically arranged. The leaves are small and are found only on the more distal portions of the plant. Nothing is known about the internal structure or reproductive parts. The genus name Lycostrobus is used for isolated, bisporangiate cones which are known from several Triassic deposits and thought to be allied with the Pleuromeiales, based partly on the presence of monolete microspores. The basic construction of these cones is similar to that of a bisporangiate Flemingites with helically arranged sporophylls



and adaxial sporangia. In L. scottii, the microspores occur in groups, and it has been suggested that the grouping may have been the result of sporangial trabeculae (partitions) that were not preserved. Lycostrobus chinleana, from the Triassic of Arizona, is now considered to be a member of the Equisetales (see Chapter 10). Annalepis is another Triassic lycopsid that was initially described based on isolated sporophylls (FIG. 9.113), but articulated cones upto 10 cm in diameter have more recently been reported. Specimens are known from China (Meng, 1998), and several sites in Europe (Kelber and Hansch, 1995; Kustatscher et al., 2004). Sporangia were adaxial and there is some suggestion that lateral laminae may have partially enclosed the sporangia (Grauvogel-Stamm and Lugardon, 2001). Microspores contained in A. zeilleri sporophylls are assignable to the dispersed spore genera Aratrisporites and Tenellisporites (Grauvogel-Stamm and Duringer, 1983). Although the parts are disarticulated,

Figure 9.113 Annalepis zeilleri, sporophyll with megaspores

(Triassic). Bar  1 cm. (Courtesy K.-P. Kelber.)

they suggest that Annalepis had a similar growth habit as Pleuromeia. Isolated lycopsid-like sporophylls from the Voltzia Sandstone (Middle Triassic) are called Bustia ludovici (FIG. 9.125) (Grauvogel-Stamm, 1991). The distal lamina is narrow and up to 3.5 cm long and vascularized by two vascular bundles. Sporangia were large and characterized by internal trabeculae. Spores are of the Aulisporites type. Austrostrobus ornatum (Triassic of Argentina) is a large, structurally preserved lycopsid cone that is believed to represent a petrified Cylostrobus (Morbelli and Petriella, 1973). The two taxa differ only in the size of the megaspores, and that difference may simply represent a combination of preservational phenomena and cone development.

Isoetales Isoetaleans are ligulate, heterosporous, and mostly herbaceous in habit. One or two genera of living plants are included in the Isoetales: Isoetes, which has an extensive distribution of almost 200 species ranging from the tropics to the sub-Arctic, and Stylites, which includes two species restricted to the high Andes of Peru. Most authorities regard Stylites as simply another morphologic form of Isoetes. Isoetes is characterized by a short, squat stem (usually less than a few centimeters long) that produces helically arranged, monarch roots from the lower surface and elongated, ligulate leaves in a dense rosette from the upper portion. Both micro- and megasporangia are produced on the same plant. Microspores of Isoetes are bilateral and monolete; megaspores are radial and trilete, although many of the fossils placed in this order have trilete megaspores and microspores; sporangia have trabeculae (sing. trabecula), sterile plates of tissue that extend into the sporangium. The Isoetales is now recognized as having an extensive fossil history, probably dating back to the Devonian (Pigg, 1992, 2001); the earliest forms with a morphology similar to that seen in the extant Isoetes have been reported from the Triassic (reviewed in Skog and Hill, 1992; Srivastava et al., 2004). Fossils from the Jurassic (e.g., Isoetites rolandii; see Ash and Pigg, 1991) are morphologically similar to extant forms in the presence of a bilaterally symmetrical rhizomorph. The modern genus exhibits great morphological and genetic uniformity (Schuettpelz and Hoot, 2006). Several Late Devonian and Early Mississippian forms have been interpreted as representing early members of the isoetalean lineage. These include Clevelandodendron


ohioensis, a compressed, almost entire lycopsid plant, 1.25 m tall, from the Cleveland Shale member of the Upper Devonian Ohio Shale (Chitaley and Pigg, 1996). This plant consists of an unbranched, slender, monopodial axis, up to 2 cm wide, arising from a base bearing thick appendages. Terminally the axis bears a bisporangiate strobilus (FIG. 9.114), 9 cm long and 6 cm wide. The decorticated stem surface shows helically arranged elongate leaf traces and laterally compressed, slender leaves along the stem margin. Megaspores obtained from the cone are trilete and laevigate, and lack a gula; microspores are trilete, indistinctly punctate, and may be assignable to the dispersed spore genera Calamospora or Punctatisporites. Clevelandodendron demonstrates that lycopsids with a habit similar to the



Carboniferous genera Chaloneria and Sporangiostrobus (discussed below), and the Triassic Pleuromeia and related forms (see above), were present as early as the Late Devonian. Meyen (1987) suggested that several, of what he termed satellite genera, should be included in the Isoetales. One of these is Tomiodendron (FIGS. 9.27, 9.115) an unbranched, protostelic plant that may have reached 30 cm in diameter and had elongate leaf cushions on the stem surface. A second lycopsid in this group is Wexfordia hookense from the uppermost Famennian (Upper Devonian) type locality at Sandeel Bay, County Wexford, in southeastern Ireland. This lycopsid has a forked axis (3 cm in diameter) (Matten, 1989). The permineralized specimens consist of axes with medullated steles; tracheids with scalariform secondary wall thickenings contain fimbrils between the bars. Subsequent research on W. hookense indicates that secondary xylem with uni- to biseriate vascular rays is present in mature axes of Wexfordia (Klavins, 2004). Leaf bases are oval and the crowded leaves are each about 4 mm long. Although initial reports suggested that Wexfordia hookense shared features with anatomically preserved axes of other lycopsids, perhaps including the Protolepidodendrales and Carboniferous Lepidodendrales, the study by Klavins

Figure 9.114 Strobilus of Clevelandodendron ohioensis show-

ing sporangia and attachment of sporophylls to cone axis (Devonian). Bar  2 mm. (From Chitaley and Pigg, 1996.)

Figure 9.115 Stem of Tomiodendron peruvianum showing leaf cushions (Mississippian). Bar  1.5 cm. (Courtesy H. W. Pfefferkorn.)



hypothesizes that Wexfordia was a small tree with anatomical features closer to members of the isoetalean lineage. A third plant with possible affinities in the order Isoetales is Otzinachsonia beerboweri from the Famennian (Upper Devonian) of north-central Pennsylvania (Cressler and Pfefferkorn, 2005). Stems are up to 10 cm wide and the stem base forms a characteristic, four-lobed cormose rhizomorph with rootlets and circular rootlet scars arranged in orthostichies. A transitional zone between the rhizomorph (basal portion of the stem) and the distal part of the stem lacks evidence of rootlet scars. Distal stem portions display spirally arranged leaf scars in parastichies, but no leaf cushions. Carboniferous lycopsids with suggested affinities in the isoetalean lineage include Chaloneria cormosa, one of the most completely known fossil lycopsids within the group. This plant is known from both vegetative and fertile remains from the Upper Pennsylvanian of North America (Pigg and Rothwell, 1979, 1983a, b). The unbranched plant was 2 m tall, with small, ligulate leaves helically arranged about the stem. Leaf cushions were not produced. The base of the plant was cormlike. A limited amount of secondary xylem surrounds an exarch protostele, and in some specimens a thin band of periderm is produced (Pigg and Rothwell, 1985). These authors have been able to correlate various levels of decortication in C. cormosa with axis surface features of Knorria (FIG. 9.42), Asolanus, Bothrodendron, Pinakodendron, Jurinodendron (FIG. 9.43), and Stigmaria. A compressed, cormose lycopsid base of Middle Pennsylvanian age named Cormophyton also exhibits some of the same features as Chaloneria (Pigg and Taylor, 1985). Pigg and Rothwell (1983a,b) placed Chaloneria in its own family, Chaloneriaceae, in which they also included Sporangiostrobus and Polysporia. Chaloneria cormosa is heterosporous, with the fertile region (10 cm long) consisting of alternating megaand microsporophylls. Sporangia contain trabeculae. Microspores are monosaccate and of the Endosporites type, whereas the megaspores can be compared to Valvisisporites. Megagametophytes of Chaloneria, some with archegonia, have also been described (Pigg and Rothwell, 1983b). Their structure and embryology are similar to those known in Bothrodendrostrobus, suggesting perhaps that the latter should be allied with the Isoetales rather than the Lepidodendrales (Stubblefield and Rothwell, 1981). A Pennsylvanian fructification that may be related to Chaloneria is Porostrobus. This bisporangiate cone (2.5 cm long) possesses megasporophylls with hairlike tips, and produced Setosisporites-type megaspores (750–1150 μm in diameter) and Densosporites-type microspores (Leary and

Figure 9.116 Stem surface of Bodeodendron hispanicum

showing leaf cushions (Pennsylvanian). Bar  1 cm. (Courtesy R. H. Wagner.)

Mickle, 1989). The distal end of the megaspore is extended into a structure termed a gula that has been interpreted as a germ tube (Jha and Tewari, 2006). Polysporia includes compressed cones with the same types of spores as those in Chaloneria (Grauvogel-Stamm and Langiaux, 1995), but this genus has also been reported as a permineralized specimen (DiMichele et al., 1979). Sporangiostrobus encompasses relatively large, mono- and bisporangiate lycopsid cones, known from both compression and anatomically preserved specimens (FIG. 9.117); they are characterized by a massive axis occupying more than half of the total cone diameter (Bode, 1928; Chaloner, 1956, 1962; Remy and Remy, 1975a; Bek, 1996). The basic organization of this cone is similar to that of other arborescent lycopsid fructifications, but differs in having large, triradiate megaspores characterized by a broad equatorial flange formed by fused and anastomosing hairs (Zonalesporites type) and triradiate microspores (FIG. 9.118) that exhibit an extended range of morphologic variability that includes such dispersed spore taxa as Densosporites, Radiizonates, Cingulizonates, and Vallatisporites, sometimes found in




Figure 9.117 Cross section of microsporangium of Sporangiostrobus kansanensis showing parenchyma tissue (Pennsylvanian). Bar  1.5 mm. (From Leisman, 1970.) Figure 9.119 Sporangiostrobus axis showing sporophyll bases (Pennsylvanian). Bar  2 cm.

Figure 9.118 Several microspores macerated from the sporan-

gium of Sporangiostrobus cone (Pennsylvanian). Bar  50 μm.

the same sporangium (Bek and Straková, 1996; Bek and Opluštil, 1998). One species from the Middle Pennsylvanian of Kansas, S. kansanensis, measures 16 cm long and is nearly 12 cm in width (Leisman, 1970). Wagner and Spinner (1976) have demonstrated that Sporangiostrobus was the cone of Bodeodendron (FIG. 9.116) based on their constant association and similar morphology. Exceptionally well-preserved specimens of Sporangiostrobus of late Stephanian age from Puertollano, Spain (FIG. 9.119) occur in large numbers in a tuff band at the base of a coal seam (Wagner, 1989) (FIG. 9.120). Some cones exceed 30 cm in length and are hypothesized to have been produced at the tips of vegetative branches. Microsporangia appear to have been intermingled with megasporangia in the cones. Sporangiostrobus has also been correlated with vegetative remains of Omphalophloios (Brousmiche-Delcambre et al., 1995). Mesozoic representatives of the isoetalean lineage include Nathorstiana, a Cretaceous genus that possesses the same root arrangement and stelar morphology as living members



Figure 9.120

Robert H. Wagner. Figure 9.121 Suggested reconstruction of Nathorstiana arborea (Cretaceous). (From Delevoryas, 1962.)

of Isoetes (Richter, 1909, 1910; Karrfalt, 1984). The genus is known from numerous specimens representing various developmental stages of the plant. Mägdefrau’s (1932) reconstruction of Nathorstiana arborea depicts the plant as being 20 cm tall, with elongate, grasslike leaves attached to the stem apex on conspicuous bases (FIG. 9.121). The underground parts were bulbous and produced helically arranged roots at the lobed bottom of the plant. Although nothing is known about the internal structure of the plant or the method of reproduction, Karrfalt (1984) hypothesized the developmental pattern of the fossil based on his studies of living Isoetes (Karrfalt and Eggert, 1977a, b, 1978). A detailed morphological study of this type in which fossils are used in association with a living developmental model (Isoetes) represents an excellent way to infer homologies among the various lycopsid rooting structures and thus more accurately determine phylogenies within the group. Nathorstianella is another Mesozoic fossil that has historically been included in this order based on cast specimens from the Lower Cretaceous of Australia (Glaessner and Rao, 1955). Although it differs in size, several features suggest a close correspondence with Isoetes. Karrfalt (1986) confirmed the isoetalean relationship, indicating that N. babbagensis was divided into five basal lobes, each bearing roots arranged

like those in extant Isoetes. He believes that the basal corm supported an arborescent aerial axis. Several fossils that morphologically resemble modern Isoetes have been described from rocks as early as the Triassic, and the generic name Isoetites has been used for many of these (Pigg, 2001), whereas others have been assigned to the extant genus Isoetes (see Srivastava et al., 2004). Isoetites serratifolius is the name used for compressed sporophylls from the Triassic of India (Bose and Roy, 1964); they extend up to 6.6 cm in length and are characterized by a serrate margin. A single vascular bundle extends the length of the sporophyll. The position of the sporangium is indicated by an elongated impression near the base of the sporophyll. In another species of the same age, I. indicus, sporophylls with megasporangia were preserved. The sporophylls are slightly smaller than those of I. serratifolius. Megasporangia contain up to 1500 trilete spores ranging 285–430 μm in diameter. Isoetites rolandii occurs in the Middle Jurassic of Oregon and Idaho (Ash and Pigg, 1991). Because the specimens are preserved as both mold–casts and impression–compressions they offer an excellent example of how different preservation types




can be used to reconstruct an entire plant. The plants are 10 cm tall and consisted of linear leaves, up to 8 cm long. Casts of sporophyll bases contain structures interpreted as megaspores. What is especially noteworthy is that this discovery represents the oldest occurrence of Isoetes-like plants in western North America. The authors also noted that the occurrence of Isoetites provided the first evidence for an aquatic to semiaquatic environment within the Coon Hollow Formation. Isoetites serratus is a well-known form from Upper Cretaceous rocks (Frontier Formation) in Wyoming (Brown, 1939). The plant consists of rosettes of narrow, spathulate leaves with serrate margins; the leaves arise from the edge of a round corm that is 1.3 cm in diameter. The upper surface of the sporophyll has two rows of rectangular cavities interpreted as either wrinkles or the remains of collapsed internal air sacs. At the base of each sporophyll are compressed elliptical sporangia that contain impressions of either megaspores or microspores. Dichotomously branched roots are attached to the corm base beneath the outer rosette of sporophylls. Isoetes ermayinensis is a Triassic isoetalean from China that includes both vegetative and reproductive material (Z.-Q. Wang, 1991). The plant is small, with leaves up to 7 cm long, each with an elongate, lacunate band on either side of the vascular bundle; the corm is unknown. Megaspores are of the Dijkstraisporites type (50–200 megaspores per sporangium), whereas the microspores are assignable to Aratrisporites. Mature megaspores in stages of germination morphologically resemble Laevigatisporites.

Putative lycopsids There are always fossils that remain impossible to place systematically. The enigmatic Devonian–Carboniferous plant Barinophyton is one of these. Members of the genus consist of alternately arranged naked branches with sporangia organized in two rows on laterally born (FIGS 9.122, 9.123), spikelike fructifications. Structurally preserved specimens of B. citrulliforme from the Devonian of New York include an exarch protostele (Brauer, 1980). The tracheids have a continuous secondary wall that is plicated into the cell lumen, simulating annular secondary wall thickenings. Between the plications, the wall contains numerous delicate pits, each with a recognizable border. The sporangiferous appendages are alternate and two-ranked; each appendage is recurved and bears one large sporangium on the concave surface within the curve. An unusual feature of Barinophyton is that several thousand microspores and 30 megaspores

Figure 9.122 Suggested reconstruction of Barinophyton citrulliforme showing two rows of sporangia in a strobilus-like organization. (From Kenrick and Crane, 1997a.)

occur together in the same sporangium. A similar condition exists in Protobarinophyton pennsylvanicum described from the Upper Devonian of New York (Brauer, 1981). The small spores range from 30–42 μm in diameter, whereas the large spores extend from 410 to 560 μm. In both Barinophyton (Taylor and Brauer, 1983) and Protobarinophyton (Cichan et al., 1984) the wall structure of the large and small spores is different. This indicates that the small spores do not merely represent aborted megaspores and that these two plants were, in fact, heterosporous. Spores of two different sizes have also been reported in the same sporangium of the progymnosperm, Archaeopteris (Medyanik, 1982) (Chapter 12). In A. latifolia the number of megaspores per sporangium is highly variable, ranging 8–20 (Chaloner and Pettitt, 1987).




S Figure 9.123 Portion of Barinophyton citrulliforme axis with

appendages (A) and sporangia (S) (Devonian). Bar  1.5 cm. (Courtesy D. F. Brauer.)

Conclusions Despite the enormous amount of information that has been accumulated about fossil lycopods, there are still gaps in our understanding of their evolution. During the Carboniferous in Euramerica the lycopsids were well represented by at least four major orders. These included the large arborescent forms (Lepidodendrales) (e.g., Diaphorodendron, Lepidodendron, and Lepidophloios) with stigmarian rooting structures; the smaller woody types (Isoetales) (e.g., Chaloneria) with cormose rooting organs; and the herbaceous, ligulate or eligulate taxa included in the Selaginellales and Lycopodiales, respectively. It is becoming increasingly clear that there were other Carboniferous lycopsids, such as Hizemodendron serratum,

that were pseudoherbaceous, but produced mono-sporangiate cones like those of the Lepidodendrales, suggesting that perhaps this type of morphology evolved via reduction of the arborescent habit (Bateman and DiMichele, 1991). Historically, many comparative morphologists and paleobotanists believed that the fossil record, although admittedly incomplete, provided sufficient evidence to suggest that Isoetes represented the end of a transformational series beginning in the Carboniferous. According to this idea, the extant plant Isoetes represented a Flemingites-type cone seated on a stunted, stigmarian base (Stewart, 1947). Proponents of this concept, called the “lycopsid reduction series” (originally “lycopod”), suggested that a sparsely branched, heterosporous lycopsid like Sigillaria was the starting point of the series (Potonié, 1894). With a reduction in the number of aerial branches, a contraction of the branched, stigmarian rooting system to a cormlike base, and a reduction in the overall size of the plant, a plant like the Triassic Pleuromeia (FIG. 9.124) might represent an intermediate stage. Continued reduction of the main axis would produce a plant similar to Nathorstiana and eventually, Isoetes. The discovery of Carboniferous lycopsids with cormlike bases, such as Chaloneria, suggests that the cormlike lycopsids did not evolve from the arborescent forms by reduction, but existed at the same time. Additional specimens of rooting organs from which developmental data can be inferred, together with information on developmental stages of the embryos of two permineralized taxa, also support the idea that the lycopsid reduction series is too simple an explanation of the complexity and diversity of the fossil lycopsids. The discovery of lycopsids with cormlike bases older than Sigillaria, however, provides evidence that stigmarian rooting structures and lobed, cormlike bases coexisted. For example, Protostigmaria (Mississippian) possesses a cormlike base, as does the Pennsylvanian taxon Chaloneria, as well as several Mesozoic lycopsids (e.g., Pleuromeia and Nathorstiana) and a Late Devonian compression form, Otzinachsonia (Cressler and Pfefferkorn,