INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

H ie quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, prim bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

A Bell & Howell Information Company 300North Zeeb Road. Ann Arbor. Ml 48106-1346USA 313.'76l-4700 800/521-0600 MORPHOMETRIC AND PALEOBIOGEOGRAPHIC ANALYSES OF

DICROIDIUM FROM THE TRIASSIC OF GONDWANA

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Lisa Diane Boucher, B.S.

The Ohio State University

1995

Dissertation Committee:

Thomas N. Taylor

Edith L. Taylor Advisor Department of Plant Biology Morris G. Cline

Fred D. Sack Co-Advisor Department of Plant Biology UMI Number: 9612154

UMI Microform 9612154 Copyright 1996? by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17? United States Code.

UMI 300 North Zeeb Road Ann Arbor? MI 48103 In Memory of My Grandparents ACKNOWLEDGMENTS

This research was supported in part by the National Science Foundation (OPP 91-

18314), Sigma Xi, The Scientific Research Society (Grant-in-Aid of Research, Graduate

Research Award), Latin American Studies Program, The Ohio State University (Tinker

Foundation, Foreign Field Research Travel Grants), the Graduate School, The Ohio State

University (Presidential Fellowship and Graduate Student Alumni Research Award), and the Department of Plant Biology, The Ohio State University (Graduate Teaching and

Research Associateships). The results of this study were significantly aided by this assistance, which is gratefully acknowledged.

Several people have contributed in a variety of ways to this investigation. I would like to thank Ms. Ann Osterfeld, Dr. Joseph Rosenblatt and Dr. Fred Ruland for technical assistance regarding imaging, Fourier, and statistical analyses. In addition, Dr. John

Mitchell and Mr. David Stutes are gratefully acknowledged for assistance with scanning electron microscopy. Several museums have assisted with the loan and availability of specimens for examination. I would like to thank Dr. Margaret Bradshaw and Dr. Karen

Watson at the Canterbury Museum, Christchurch, New Zealand for their assistance at the museum and for the loan of several Dicroidium specimens. I also thank Dr. William

DiMichele and Mr. James Ferrigno for assistance and loan of material from the

Smithsonian Institution. I am very appreciative of assistance from several scientists at the

Museo Argentino de Ciencias Naturales Bernardino Rivadavia, Buenos Aires, Argentina including, Dr. Sergio Archangelsky, Dr. Georgina Del Fueyo, and Mr. Luis Lezamo, and at PRINGEPA, Corrientes, Argentina, I thank Dr. Raphael Herbst and Dr. Alicia Lutz for

iii their assistance. In addition, I thank Dr. Rubdn Ciineo for Antarctic stratigraphic and paleoecological information. I would also like to thank several past and fellow students in the paleobotanical laboratory for discussion and technical assistance, including Dr. Jeffrey

Osborn, Dr. Xuanli Yao and Mr. Brian Axsmith.

I am especially appreciative of the members of my general exam and advisory committees, including Dr. Thomas Taylor, Dr. Edith Taylor, Dr. Morris Cline and Dr.

Fred Sack. In particular, I am extremely grateful to Drs. Thomas and Edith Taylor, who have greatly supported this study by providing me with the opportunities necessary to conduct this study, and for their guidance and assistance. It has been an honor to work with them.

I will always be grateful to my family and friends for their continued support of my endeavors. This document is dedicated to my grandparents, Mildred and Alton Becker, each of which fostered my interest of nature in their own special way. VITA

July 3, 1968 Bom - Albany, New York

1990 ...... B. S. in Biology, Cornell University, Ithaca, New York

1990 Graduate Research Associate and Graduate Teaching Associate, The Ohio State University, Columbus, Ohio

1995 Presidential Fellow, The Ohio State University

PUBLICATIONS

Boucher, L.D., E.L. Taylor, and T.N. Taylor. Biomechanical stability of Dicroidium pteridosperm foliage. Royal Botanic Gardens, Kew. (in review).

Boucher, L.D., E.L. Taylor, T.N. Taylor, N.R. Cdneo, and J.M. Osborn, (in press). Dicroidium compression floras from southern Victoria Land. Antarctic Journal of the U.S. 1995 Review Issue.

Boucher, L.D., T.N. Taylor, and E.L. Taylor, (in press). An unusual plant organ from the Triassic of Antarctica. Antarctic Journal of the U.S. 1994 Review Issue.

Boucher, L.D. and E.L. Taylor. 1995. Dicroidium species from Antarctica and South America: temporal and paleobiogeographic ranges. 1995 Annual Meeting, Botanical Society of America (San Diego, California). American Journal of Botany 82 (6, Suppl.): 83-84.

Boucher, L.D., E.L. Taylor, and T.N. Taylor. 1995. Trends in Dicroidium leaf architecture. The Evolution of Plant Architecture meeting (London, UK). Royal Botanic Gardens, Kew: 4-5.

Boucher, L.D. 1994. Classification of pteridosperm foliage using elliptic Fourier analysis. 1994 Annual Meeting, Botanical Society of America (Knoxville, Tennessee). American Journal of Botany 81 (6, Suppl.): 88.

Boucher, L.D.*, E.L. Taylor, and T.N. Taylor. 1994. Dicroidium distribution in Antarctica and its role in paleoenvironmental reconstructions. 1994Annual v Meeting, Botanical Society of America (Knoxville, Tennessee). American Journal of Botany 81 (6, Suppl.): 88-89.

Boucher, L.D.*, E.L. Taylor, and T.N.Taylor. 1993. A comparison of cuticular structures from Dicroidium and associated corystosperm reproductive organs. 1993 Annual Meeting, Botanical Society of America (Ames, Iowa). American Journal of Botany 80 (6, Suppl.): 87.

Boucher, L.D., E.L. Taylor, and T.N. Taylor. 1993. Dicroidium from theTriassicof Antarctica. In: S.G. Lucas and M. Morales (eds.), The Nonmarine Triassic. New Mexico Museum of Natural History and Science Bulletin No. 3:39-46.

Taylor, E.L., L.D. Boucher, and T.N. Taylor. 1992. Dicroidium foliage from Mount Falla, central Transantarctic Mountains. Antarctic Journal of the U.S. 27(5): 2-3.

Boucher, L.D.*, E.L. Taylor, and T.N. Taylor. 1992. Distribution of Dicroidium in the Southern hemisphere during the Triassic. 4th International Organization of Paleobotany Conference (Paris, France). OFP Informations N* Special 16-B: 31.

Taylor, E.L.*, T.N. Taylor, N.R. Cdneo, and L.D. Boucher. 1991. Osmundaceous ferns from the Triassic of Antarctica. 1991 Annual Meeting, Botanical Society of America (San Antonio, Texas). American Journal of Botany 78 (6, Suppl.): 127.

HELDS OF STUDY Major Reid: Plant Biology

Studies in Paleobotany TABLE OF CONTENTS

DEDICATION ...... ii

ACKNOWLEDGMENTS ...... iii

VITA ...... v

LIST OF TABLES ...... ix

LIST OF FIGURES ...... x

CHAPTER

I. GENERAL INTRODUCTION ...... 1

The Corystospermales ...... 1 Morphological and Cuticular Features of Dicroidium ...... 5 The Triassic Paleoenvironment of Gondwana ...... 7 Dissertation Objectives ...... 13

II. MATERIALS AND METHODS ...... 15

Fossil Material and Localities ...... 15 Techniques ...... 16

III. MORPHOMETRIC ANALYSES OF DICROIDIUM...... 19

Classification Using Elliptic Fourier Analysis ...... 19 Results ...... 21 Discussion ...... 22

IV. SYSTEMATICS OF DICROIDIUM...... 25

Description of Specimens from Antarctica ...... 25 Description of Cuticular Features ...... 31 Discussion ...... 35

vii V. DICROIDIUM FLORA FROM ANTARCTICA 39

Triassic Stratigraphy and Paleoenvironment ...... 39 Dicroidium Floras Previously Described from Antarctica ...... 43 Description of Dicroidium Assemblages from Antarctica ...... 44 Discussion ...... 50

VI. DICROIDIUM FLORA FROM OTHER GONDWANA LOCALITIES ...... 53

Dicroidium from South America ...... 53 Dicroidium from Australia, Tasmania and New Zealand ...... 54 Dicroidium from Africa and Madagascar ...... 56 Dicroidium from India ...... 57

VII. PALEOBIOGEOGRAPHY OF DICROIDIUM ...... 59

Distribution of the Genus ...... 59 Paleogeographical Analyses ...... 61 Discussion ...... 65

VIII. GENERAL DISCUSSION ...... 72

Morphological Trends Through Time ...... 72 Environmental Correlations ...... 75

LITERATURE CITED ...... 79

APPENDICES ...... 99

A. Tables ...... 100

B. Figures ...... 148

C. Data Relative to Chapter III ...... 208

D. Data Relative to Chapter VII ...... 209 LIST OF TABLES

TABLE

Table 1. Directory of specimens ...... 101

Table2. Dicroidium species from Antarctica ...... I ll

Table 3. Dicroidium species from South America ...... 116

Table 4. Dicroidium species from Australia, Tasmania and New Zealand ...... 124

Table 5. Dicroidium species from Africa and Madagascar ...... 133

Table 6. Dicroidium species from India ...... 138

Table 7. Morphological and cuticular features of Dicroidium from Mt. Falla (collections T-5 and T-7) ...... 140

Table 8. Other corystosperm stem and reproductive organs found in Antarctica during the Triassic ...... 143 Table 9. Dicroidium species from all regions of Gondwana during the Early, Middle, and Late Triassic ...... 144

Table 10. Simpson Indexes for Dicroidium species in each region of Gondwana during Early, Middle, and Late Triassic...... 146

Table 11. Jaccard Indexes for Dicroidium species in each region of Gondwana during Early, Middle, and Late Triassic ...... 147

be LIST OF FIGURES

FIGURE Figure 1. Summary of the steps completed for morphometiic analyses of pinnule outlines ...... 149

Figure 2. Reconstructed pinnule outlines from 1 to 25 harmonics using EFA (specimen AF 24) ...... 150

Figure 3. Reconstructed pinnule outlines from 1,5,10,15,20 and 25 harmonics (specimen T-ll #446b)...... 151

Figure 4. PCA factor scores from Fourier coefficients of 330 pinnules (data set A )... 151

Figure 5. PCA factor scores and cluster analysis results of the pinnule subset B 152

Figure 6. PCA factor scores from Fourier coefficients of 75 pinnules above the dichotomy (data set C) ...... 152

Figure 7. PCA factor scores from Fourier coefficients of 75 pinnules below the dichotomy (data set D)...... 153

Figure 8. PCA factor scores from Fourier coefficients of a pinnule subset E from above the dichotomy ...... 153

Figure 9. Dicroidium crassinervis from Gordon Valley (T-4 #33) ...... 155

Figure 10. D .dubium from Mt. Falla (T-5 #128) ...... 155

Figure 11. D.elongatum from Allan Hills (T-ll #417) ...... 155

Figure 12. D .dutoitii from Allan Hills (T-8 #271) ...... 155

Figure 13. D.feistmantelii from Mt. Falla (Ant 70-8-164) ...... 157

Figure 14. D.lancifolium from Mt. Falla (T-5 #38) ...... 157

Figure 15. Detail of alethopteroid venation from D. lancifolium (T-5 #50)...... 157

Figure 16. D. odontopteroides from Mt. Bumstead (Ant 67-9-65a) ...... 157

Figure 17. D.odontopteroides from Allan Hills (T-ll #769) ...... 159

Figure 18. Odontopteroid venation from D. odontopteroides distal regions with fused pinnules (T-7 #162b) ...... 159 Figure 19. D. spinifolium from Allan Hills (T-ll #435b)...... 159

Figure 20. Venation of D. spinifolium (T-ll #423b)...... 159

Figure 21. D. stelznerianum from Gordon Valley (T-4 #33) ...... 161

Figure 22. Detail of venation from the apex of a D. stelznerianum frond (T-8 #188)... 161

Figure23. D.trilobitum from Allan Hills (T-ll #418a) ...... 161

Figure 24. D. zuberi from Roscolyn Tor, Allan Hills (T-ll #855) ...... 161

Figure 25. Cuticle of D. odontopteroides illustrating the epidermal cell pattern and orientation of stomatal complexes (T-5 #72; Slide no. 72-1) ...... 163

Figure 26. Cuticle with papillae on subsidiary cells surrounding the stoma (T-5 #103; Slide no. 103-6) ...... 163

Figure 27. The outer surface of a cuticle piece in a region with continuous papillae (T-7 #186) ...... 163

Figure 28. Stomatal complex illustrating paracytic arrangement (T-5 #72; Slide no. 72-1) ...... 163

Figure 29. Inner surface of. cuticle showing monocyclic stomatal complex and thickened guard cells (T-5 #65) ...... 165

Figure 30. Inner surface of cuticle with rectangular cells and a stomatal complex oriented parallel to veins (T-5 #65) ...... 165

Figure 31. Outer surface of cuticle with papillae and one stomatal pore opening indicated with an arrow (T-5 #103) ...... 165

Figure 32. Outer surface of cuticle illustrating two pore openings, both indicated with arrows (T-5 #72)...... 165

Figure 33. Inner surface of a stomatal complex with thickened aperture and guard cell thickenings (T-5 #65)...... 167

Figure 34. Inner surface of an open stomatal complex with thickened guard cells (T-5 #103)...... 167

Figure 35. Inner surface of a stomatal complex with striated guard cell thickenings (T-5 #81a)...... 167

Figure 36. Inner surface of a trichome base (T-5 #72) ...... 167

Figure 37. Figure 37. Cuticle of D. lancifolium with two monocyclic stomatal complexes indicated with arrows (T-5 #38; Slide no. 38-4) ...... 169 Figure 38. Epidermal cells and stomatal complex with orientation along axis of veins (T-5 #78; Slide no. 78-5) ...... 169

Figure 39. Inner surface illustrating a stomatal complex (T-5 #38)...... 169

Figure 40. Inner surface of a stomatal complex with guard cell thickenings (T-5 #38)...... 169

Figure 41. Stomatal complexes of D. dubium illustrating two different subsidiary cell arrangements (T-5 #128; Slide no. 128-1) ...... 171

Figure 42. Cuticle with papillae and paracytic stomatal complex indicated by an arrow (T-5 #128; Slide no. 128-1) ...... 171

Figure 43. Inner surface of a stomatal complex and epidermal cells (T-5 #128) ...... 171

Figure 44. Inner surface of a stomatal complex illustrating guard cell and aperture thickenings (T-5 #128) ...... 171

Figure 45. Specimen of D. dutoitii providing cuticular material (USNM 40357) ...... 173

Figure 46. Inner surface of D. dutoitii cuticle showing three stomatal complexes indicated by arrows (USNM 40357)...... 173

Figure 47. Inner surface of a stomatal complex with guard cell thickenings and subsidiary cells (USNM 40357) ...... 173

Figure 48. Stomatal complex with three subsidiary cells and polar extensions of guard cell thickenings (USNM 40357; Slide no. 40357-4) ...... 173

Figure 49. Inner surface of a stomatal complex (USNM 40357) ...... 175

Figure 50. Inner surface of a stomatal complex with striated guard cell thickenings and well-defined polar regions of guard cells (USNM 40357) ...... 175

Figure 51. Specimen of D. dubium providing cuticular material (USNM 458659). .. 175

Figure 52. Cuticular surface illustrating several stomatal complexes (USNM 458659; Slide no. 458659-3) ...... 175

Figure 53. Inner surface with several stomatal complexes (USNM 458659) ...... 177

Figure 54. Inner surface of a paracytic stomatal complex (USNM 458666) ...... 177

Figure 55. Stomatal complex with paracytic subsidiary cells and guard cell thickenings (USNM 458661; Slide no. 458661-1) ...... 177

Figure 56. Stomatal complex with adjacent subsidiary cells surrounded by encircling cells (USNM 458666; Slide no. 458666-2) ...... 177

Figure 57. Cuticle of D. zuberi with sinuous anticlinal walls and monocyclic stomatal complexes (USNM 458665A; Slide no. 458665A-1) ...... 179 Figure 58. Inner surface of cuticle illustrating sinuous anticlinal walls (USNM 458665A) ...... 179

Figure 59. Outer surface of cuticle with prominent papillae (USNM 458665A) ...... 179

Figure 60. Inner surface of cuticle illustrating papillae regions (USNM 458665A). .. 179

Figure 61. Inner surface of a stomatal complex with thickened guard cell walls (USNM 458665A) ...... 179

Figure 62. Cuticle of D. odontopteroides. Inner surface of a stomatal complex with monocyclic subsidiary cell arrangement and thickened guard cells (USNM 458663) ...... 179

Figure 63. Map of Antarctica illustrating localities in the Transantarctic Mountains. .. 180

Figure 64. D. odontopteroides from 135 m above the base of the Mt. Falla type section (T-5 #103)...... 182

Figure 65. D. lancifolium from 135+ m above the base of the type section (T-7 #177) ...... 182

Figure 66. D. odontopteroides from about 120 m above the base of the type section (Ant 70-8-170) ...... 182

Figure 67. D. dubium from approximately 130 m above the base of the type section (Ant 70-8-31) ...... 182

Figure 68. D. zuberi from Mt. Falla (Ant 70-8-52) ...... 184

Figure 69. D. stelznerianum from Gordon Valley (T-4 #29b) ...... 184

Figure 70. Several specimens of D. stelznerianum from Gordon Valley (T-4 #31). .. 184

Figure 71. D. odontopteroides from Fremouw Peak (T-10 # 292) ...... 184

Figure 72. D. dubium from Fremouw Peak (Ant 70-9-143) ...... 186

Figure 73. D. odontopteroides and D. elongatum (arrow, right) from Mt. Bumstead (Ant 67-9-52a) ...... 186

Figure 74. D. stelznerianum from Mt. Bumstead (Ant 67-9-60a) ...... 186

Figure 75. D. zuberi from Mt. Bumstead illustrating venation (Ant 67-9-140) ...... 186

Figure 76. D. crassinervis from Mt. Wisting (Ant 71-1-7) ...... 188

Figure 77. D. odontopteroides from Mt. Wisting (Ant 71-1-21) ...... 188

Figure 78. Two specimens of D. stelznerianum with rhomboid pinnules from Shapeless Mountain (AF 9/1) ...... 188 xiii Figure 79. D. lartcifoliwn from Horseshoe Mountain (AF 315) ...... 188

Figure 80. Locality site in Feather Bay, Allan Hills ...... 189

Figure 81. D. odontopteroides from Level 1, Feather Bay locality (T -ll #490) ...... 191

Figure 82. Several specimens of D. dutoitii from Level 2 (T-8 # 275b) ...... 191

Figure 83. Specimens of D. dutoitii from Level 2 illustrating possible expanding immature foliage (T-8 #255) ...... 191

Figure 84. D. trilobitum and D. elongatum (right) from Level 4 (T -ll# 442b) ...... 191

Figure85. D. spinifolium from Level 4 (T-ll #450a) ...... 193

Figure86. D.elongatum from Level 4 (T-ll #451) ...... 193

Figure 87. D. stelznerianum with rhomboid pinnules from North Roscolyn Tor, Allan Hills (T-ll #783) ...... 193

Figure 88. D. lancifolium from South Roscolyn Tor, Allan Hills (T-l 1 #876) ...... 193

Figure 89. Present-day rectilinear map of the Southern hemisphere generated using the computer program PGIS/Mac™ ...... 194

Figure 90. Paleomap of Gondwana during the Early Triassic ...... 195

Figure 91. Paleomap of Gondwana during the Middle Triassic ...... 196

Figure 92. Paleomap of Gondwana during the Late Triassic ...... 197

Figure 93. Paleomap of Gondwana during the Anisian, summer season, indicating general precipitation amounts ...... 198

Figure 94. Paleomap of Gondwana during the Anisian, summer season, indicating relative temperature levels ...... 199

Figure 95. Paleomap of Gondwana during the Anisian, winter season, indicating general precipitation amounts...... 200

Figure 96. Paleomap of Gondwana during the Anisian, winter season, indicating relative temperature levels ...... 201

Figure 97. Paleomap of Gondwana during the Norian, summer season, indicating general precipitation amounts...... 202

Figure 98. Paleomap of Gondwana during the Norian, summer season, indicating relative temperature levels ...... 203

Figure 99. Paleomap of Gondwana during the Norian, winter season, indicating general precipitation amounts ...... 204

xiv Figure 100. Paleomap of Gondwana during the Norian, winter season, indicating relative temperature levels ...... 205

Figure 101. PAE cladogram results of taxa present during the Early Triassic 206

Figure 102. The two most parsimonious PAE cladgrams for the Middle Triassic data set...... 206 Figure 103. The three most parsimonious PAE cladograms for the Late Triassic data set ...... 207

xv CHAPTER I

GENERAL INTRODUCTION

The Corystospermales

The Triassic period (208-245 m.y.) is characterized by a unique fossil megaflora

including representatives of the Lycophyta, Sphenophyta, Pteridophyta,

Pteridospermophyta, Coniferophyta and various other gymnosperm taxa (Taylor and

Taylor, 1993). Many of the Triassic representatives belonging to these groups,

particularly the seed ferns (Pteridospermophyta), existed only during this time period,

becoming extinct during the Jurassic. The Northern hemisphere floras were dominated by

ferns, cycads and conifers. However, the floras of the Southern hemisphere were

pteridosperm dominated. The contrast in floras is perhaps in part due to the physical

separation of these regions and climatic differences.

The corystosperms are one such group within the Pteridospermophyta that are

generally restricted to the Triassic of the Southern hemisphere, although some genera have

been identified from the Northern hemisphere (Thomas, 1933). Megasporophylls placed

within the Corystospermales are characterized by uniovulate cupules, and include

Umkomasia, Pilophorosperma, Spermatocodon and Karibacarpon (Petriella, 1981). To

date they have only been found as compression fossils. The pollen organs, Pteruchus and

Pteroma, possess bisaccate pollen contained in elongate pollen sacs attached to the abaxial surface of microsporophylls (Townrow, 1962a; Harris, 1964; Yao et al., 1995). These

1 2

genera have been described from compression material, however, the discovery of

permineralized material has provided additional information regarding the anatomy and

homologies of Pteruchus (Yao et al., 1995).

Consistently associated with Umkomasia megasporophylls and Pteruchus

microsporophylls are the vegetative remains of stems and leaves restricted to the Triassic of

the Southern hemisphere that are also assigned to the order Corystospermales.

Permineralized stems known as Rhexoxylon and Kykloxylon have been described from

various Gondwana localities (Archangelsky and Brett, 1961; Herbst and Lutz, 1988;

Taylor, 1992; Meyer-Berthaud et al., 1993). Rhexoxylon possesses a discontinuous

cylinder of secondary xylem divided by sections of parenchyma, similar to the anatomy of

some lianes (Walton, 1923). Kykloxylon has a continuous cylinder of secondary xylem.

However, it possesses several other anatomical features in common with Rhexoxylon

(Meyer-Berthaud et al., 1993). Dicroidium is a form genus of pteridosperm foliage found associated with Rhexoxylon (Archangelsky, 1968). It is characterized by a dichotomously

forking rachis. Reconstructions of the plants that may have borne Dicroidium foliage are

based on these constant associations of organ genera, and evidence for possible stem attachment (Archangelsky, 1968; Crane, 1985; Retallack and Dilcher, 1988; Meyer-

Berthaud et al., 1992). According to these reconstructions Dicroidium may have belonged to a tree with a slender trunk bearing frond-like leaves at its apex or a woodland tree with large branches arranged in non-whorled tiers (Petriella, 1978; Taylor, in press). Leaves have not been found attached to stems, but some compressions from South Africa suggest a fascicle-like arrangement of the fronds (Anderson and Anderson, 1983). Evidence for a deciduous habit is based on the abundance of leaves found at the base of trunks at several localities and periderm formation beneath leaf bases (Thomas, 1955; Meyer-Berthaud et al., 3

1993). Dicroidium is generally known from compressions and/or impressions, but the anatomical structures of some permineralized specimens from Antarctica, Dicroidium fremouwensis, have been described (Pigg, 1990). The vascular tissue arrangement is dorsiventral with an abaxial ring of six to eight bundles and an adaxial linear arrangement of five to eight bundles at the level of and distal to the frond bifurcation. This arrangement corresponds to the leaf trace organization and other anatomical features in Kykloxylon

(Meyer-Berthaud et al., 1993). One anatomical feature linking Pteruchus, Dicroidium and

Kykloxylon is the presence of distinct lacunae in the ground tissue that are interpreted as secretory cavities. The existence of two stem types from South America and Africa in western Gondwana, and Antarctica in southern Gondwana suggests different plant types and/or habits possessing Dicroidium foliage in these regions.

Dicroidium is thought to be related to Thinnfeldia and other late Paleozoic and early

Triassic pteridosperms. In particular, Thinnfeldia callipteroides, a more than twice-forked leaf of the late Permian, has been interpreted as occurring at the base of the phylogenetic line evolving to the genus Dicroidium (Retallack, 1977). However, this interpretation is highly speculative, as it is based only on stratigraphic age and general leaf shape similarity.

No reference is made to associated reproductive structures or verification of transitional assemblages. There is considerable speculation as to the relationships between Paleozoic and Mesozoic pteridosperms (Thomas, 1933). Based on the similarity of Rhexoxylon to

MeduUosa stems, the corystosperms have been generally related to some Paleozoic pteridosperms (Archangelsky and Brett, 1961). However, the paraphyletic relationship of pteridosperms precludes any statement regarding their phylogenetic association. Cladistic analyses using reproductive and vegetative characters support a close relationship among the Corystospermales, Caytoniales and glossopterids. Some authors report that the 4 corystosperms represent the sister group to the Bennettitales and are part of the

"platysperm" clade (Crane, 1985; Doyle and Donoghue, 1992; Rothwell and Serbet,

1994).

Dicroidium compressions are abundant in Triassic rocks of Gondwana. However, there have been many inconsistencies in assigning compression specimens to Dicroidium because of their morphological similarity to other foliage taxa and their high degree of variability. The first published specimen of Dicroidium was identified as Pecopteris odontopteroides (Morris, 1845). Subsequent identifications included the genera

Thinnfeldia and Pachypteris. Dicroidium was first introduced by Gothan (1912) to separate the dichotomously forking rachis of Triassic specimens found in the Southern hemisphere from the unforked fronds, Thinnfeldia, known from the Northern hemisphere.

Other synonyms have included; Johnstonia, Zuberia, Diplasiophyllum, Dicroidiopsis,

Xylopteris, Hoegia, and Tetraptilon (Townrow, 1957; Archangelsky, 1968; Anderson and

Anderson, 1983). However, not all authors agree that these specimens should be placed within Dicroidium, so some taxa are still in use, in particular Johnstonia, Zuberia, and

Xylopteris (Retallack, 1977; Baldoni, 1980; Rigby, 1985; Artabe, 1990). Specimens of

Zuberia that have been identified from Argentina possess intercalary pinnules that are used by some authors as the basis for its separation into a distinct genus (Artabe, 1990).

Specimens assigned to Johnstonia are distinguished by the absence of pinnules, possessing entire or slightly lobed lamina. Xylopteris is characterized by extremely narrow, elongate pinnules. Each of these genera possess the characteristic bifurcating rachis. Separation of these taxa is, in a sense, arbitrary, and thus they will be treated as synonyms of Dicroidium in this study. 5

Morphological and Cuticular Features of Dicroidium

Dicroidium species are principally delimited by morphological characters, such as

pinnule size and shape, number of divisions and extent of pinnule lobing, and venation

patterns. Other characters available include epidermal features preserved on cuticular

remains. Historically, more than thirty species have been identified by various authors.

Subspecies and forma have been recognized (Anderson and Anderson, 1983), as well as

varieties (Retallack, 1977; Holmes, 1982). These may reflect variations in populations,

preservation and/or ecological differences. Species identifications are subjective,

dependent on characters chosen, regional differences, and the practices of "lumpers vs.

splitters." The most common species is D. odontopteroides, which is a pinnate frond with odontopteroid venation. There are several bipinnate species, including D.feistmantelii and

D. zuberi. Frond arrangements vary from simple to tripinnate with the bifurcation in approximately the lower 1/3 of the frond. Frond length averages from 10 to 30 cm. The angle of the forking in the rachis is between 15° and 55°. Pinnule size varies along the position of the fronds, becoming gradually smaller at the tips and inside the fork. Pinnules extend below the bifurcation. Venation varies with the shape of the pinnule, ranging from taeniopteroid in simple fronds to odontopteroid in ovate pinnules, and alethopteroid in elongate pinnules.

The first descriptions of cuticular features from Dicroidium foliage were made by

Gothan (1912). Further analyses were carried out by Antevs (1914), Thomas (1933),

Jacob and Jacob (1950), and Townrow (1957). Cuticular features aid in identification of some species, but are not often used as criteria since the cuticle is usually not well preserved. Type species were mainly established by morphological differences of the 6

pinnules, and these characters continue to dominate classifications. In addition, those

cuticular features identified to date tend to vary within a single specimen, and detailed

features may not be taxon dependent, but due to environmental influences or natural

variation (Stace, 1965; Kerp, 1990). However, some distinct cuticular differences can be

recognized between certain species, such as between D. zuberi (sinuous anticlinal walls)

and D. elongata (smaller stomatal complexes), and this further supports the identification of

two separate species. Many of the Dicroidium specimens identified from Triassic rocks in

India have been named using cuticular features, since the cuticle in these sediments is

sometimes better preserved than the overall frond morphology.

Cuticular features include epidermal cell outlines, stomatal complexes and

trichomes. Epidermal cells are usually isodiametric, but are elongate over the rachis and

major veins. Anticlinal walls are straight to sinuous, and the leaves are amphistomatous.

Stomatal complexes are randomly arranged over the lamina, but are usually oriented with

the long axis of cells over the rachis and veins. Subsidiary cells range from 2-7 cells in a

monocyclic to incomplete dicyclic arrangement Stomatal complexes have been described as

paracytic (haplocheilic). Guard cell walls are thickened with striations. Trichomes are

rare, but when present are surrounded by a ring of 4-8 cells.

Even though Dicroidium is a common Triassic fossil in the Southern hemisphere

and has been used as a biostratigraphic marker, there are serious problems in assigning

specimens to the genus and characterizing form species, as noted above. Previous studies

have failed to resolve these inconsistencies in the nomenclature. In an attempt to resolve

these problems, a more quantitative approach to the characterization of the pinnule shapes and associated cuticular features is necessary. Morphometric analysis is a useful approach

to the quantification of pinnule shape, since it provides an objective and repeatable method 7 with which to classify pinnules (Rohlf and Marcus, 1993). When incorporated with other features, such as qualitative information, morphometric analysis can provide a method to standardize descriptions. One morphometric technique, elliptic Fourier analysis, offers a distinct advantage, since it allows for more flexibility in image acquisition and the calculation of harmonics is relatively fast (Kuhl and Giardina, 1982; White et al., 1988).

Elliptic Fourier analysis is a procedure that fits a closed curve to an ordered set of points in two-dimensional space. The curve is represented as a Fourier series of simple sine and cosine curves of decreasing amplitude and period. Essentially, the curve is decomposed into a number of harmonically related ellipses. When these ellipses are summed, they approximate the original curve. Fourier coefficients, in addition to other quantitative and qualitative features, can then be used in multivariate analyses, such as principal components analysis. Cluster analyses can be performed to summarize relationships among shapes.

Several studies have used this methodology for characterizing shapes, including outlines of mosquito wings, mussel shells, and leaves (Rohlf and Archie, 1984; Ferson et al., 1985;

McLellan, 1993).

The Triassic Paleoenvironment of Gondwana

Dicroidium is found on all of the present continents and islands of the Southern hemisphere, including Australia, Antarctica, South Africa, South America, India, New

Zealand, Tasmania, and Madagascar, indicating that the plants that bore Dicroidium foliage were widely distributed throughout Gondwana during the Triassic. Therefore, these fossils play a significant role in Triassic biostratigraphy and paleoenvironmental reconstructions. They also offer the possibility to examine the biogeography of the taxon.

Compression specimens are abundant in the fossil record, however, little is known about 8

the spatial and temporal range of species within this genus. Some species are found in all

geographic regions, whereas others are restricted to specific localities. This implies that

some species were successful in a variety of environments and/or there was little competition between plants, thereby allowing populations to occupy extremely broad ranges.

Continental positions and thus global climate were clearly different during the

Triassic compared to today (Smith et al., 1981; Ziegler et al., 1983; Smith et al., 1994).

The present continents and islands of the Southern hemisphere that constituted Gondwana were part of a larger land mass called Pangaea. This association was established by the

Late Carboniferous, when Gondwana collided with Laurasia to form Pangaea, and existed until the Late Jurassic (Parrish, 1990). Gondwana is often treated as a separate region, since it occurred at mid- to high-southern latitudes and is characterized by nonmarine facies distinct from the rest of Pangaea. The biota of Gondwana is different from the rest of

Pangaea and this is believed to be due to the geologic and climatic barriers of the Tethys seaway on the east and a tropical land zone on the west. However, it apparently was not an absolute.barrier, since there is some evidence of north-south faunal migrations and "mixed" floras in the Tethys region (Veevers, 1988). From the Early to Late Triassic, Gondwana steadily shifted north, with the most southerly floras extending to 80°S paleolatitude during the Early Triassic.

Within these paleolatitudes, there was a globally mild climate, even in the extreme

Southern latitudes. For example, although Antarctica was positioned at only slightly lower latitudes during the Triassic, fossils from this region provide evidence of a seasonal temperate flora for these extreme southern latitudes (Spicer and Chapman, 1990; Taylor and Taylor, 1990). In addition, tree ring growth from fossilized trunks provides evidence 9 for temperate climates at high Southern latitudes for the Permian and Triassic (Creber,

1990; Taylor et al., 1992). Since the limiting factor for tree growth today at high latitudes is temperature and precipitation rather than solar radiation, these studies corroborate that temperatures were much warmer at this time (Creber and Chaloner, 1985). The climate in the Southern hemisphere during the Triassic is thought to have been relatively mild and humid based on geological, fossil, and climate modeling studies (Tucker and Benton,

1982; Hallam, 1985; Donn, 1987; Kutzbach and Gallimore, 1989; Parrish, 1990).

Moreover, Tucker and Benton (1982), and Hallam (1985) suggest an increase in aridity for

Gondwana regions during the later stages of the Triassic. The existence of such an extensive landmass is thought to have resulted in climates dominated by strong seasonality and central continental aridity, and this interpretation is confirmed by the presence of extensive red beds in Triassic formations within continental interiors (Crowley and North,

1991). However, at higher latitudes and altitudes the climate may have been humid, with cooler temperatures and lower evaporation rates (Parrish, 1990). Furthermore, climate models predict that the position of the supercontinent influenced the formation and direction of megamonsoons (Kutzbach and Gallimore, 1989; Parrish, 1993).

Even though all the plants that bore Dicroidium foliage have not been reconstructed in their entirety, a study of the leaves can provide information about the nature of the plant, paleoclimate, morphological changes in leaf shape through time, and variation within and between species. The importance of leaf morphology cannot be underestimated. Leaves are critical to the survival of the plant, since they are the major site of photosynthesis, and are especially sensitive to environmental changes. The shape of the leaf has been shown to affect the rate of gas exchange for photosynthesis and water loss by evaporation (Taylor,

1972). As a result, plants evolve leaf adaptations, including specialized shapes, and 10

anatomical and cuticular features, that improve the survival and therefore reproductive

success of the plant (Givnish, 1979; Jones, 1985). Leaf adaptations are particularly

conspicuous in regions with harsh environments, such as regions of high latitude and

altitude. There are many documented studies of ecological trends in leaf form (Givnish,

1987). However, the morphology of a plant is influenced by several factors including:

genetic potential and phylogenetic constraints; functional aspects such as biomechanical constraints; environmental factors including climate, nutrient availability, and herbivory;

biochemical, physiological and developmental constraints; and chance (Raup, 1972; Dodd and Stanton, 1990). One factor, or a combination of several factors, may be involved in controlling the morphology of a leaf during its development, and may be indirectly responsible for the evolution of leaf shape in a particular species. In addition, trends in leaf characters seen in certain plant groups may not reflect an adaptation solely to the present environment, but may also reflect historical factors, adaptation to a past environment, or random genetic drift

In the case of Dicroidium, distinct differences in pinnule morphology are primarily used to define form species. Some of these differences in morphology may reflect ecotypes or intraspecific variation, resulting from phenotypic plasticity. Regardless, if pinnule shape is adaptive, pinnule morphology and cuticular features would be expected to vary in different environments. Leaf adaptations to the environment have been principally studied in gymnosperms and angiosperms. Environmental influences on leaf morphology can result over time in the selection of an adaptive leaf shape in specific populations (Jordan and Hill, 1994). For example, studies have illustrated correlations of latitude and geography with leaf shape (Andersson, 1991). Correlations between the leaf physiognomy of major extant vegetation types and climate have been used to predict past climates (e.g. 11

Spicer, 1990; Woodcock, 1992). For example, relationships between the physiognomy of

extant dicot leaves and climate have been applied to assemblages of dicot fossil leaves to

predict temperature and precipitation increases following the Cretaceous/Tertiary boundary (Wolfe, 1990).

Since no modem analogue is known, leaf shape variation of seed ferns measured

against the environment has not been explored in detail. However, anatomical and cuticular

features of seed fem foliage have been used as evidence of microhabitat adaptations

(Reihman and Schabilion, 1978). If Dicroidium form species could be correlated with

different regions or environments, then it would suggest that these species possess leaf

traits that are adaptive. Initial studies of species distribution suggest that there are trends in

Dicroidium species distributions in regions of Gondwana, perhaps reflecting differences in paleoenvironments.

In order to examine the range of Dicroidium species, data from detailed paleobiogeographical studies must be obtained. Determining the distribution of taxa in regions at specific time intervals is the first step in paleoenvironmental reconstructions.

Paleobiologists have applied the techniques of both ecological and historical biogeography to the fossil record, using provinces to characterize regions (Jablonski et al., 1985).

Paleobiogeographical studies provide a framework for determining the spatial distribution of organisms, which is typically dependent on many factors, including favorable abiotic and biotic components of the environment (Springer and Miller, 1990). Fossil communities provide the necessary information to address ecosystem dynamics within large time spans and address questions of ecological theory (Miller, 1990; Ricklefs, 1990;

DiMichele, 1994). In addition, because of the long time ranges, questions regarding species diversity and macroevolution can also be addressed (Kitchell, 1985; Valentine, 12

1990). Without a doubt, the fossil record provides unique information for

paleobiogeographical studies about organisms through extensive periods of time that cannot

be obtained elsewhere (Valentine and Jablonski, 1993).

Area cladograms of sample localities can be produced directly from geographical distributions. Areas similar in composition can be quantified by calculating various biogeographical indexes, which were reviewed by Shi (1993). Calculations can also be corrected for abundance (Thulbom, 1986). Problems in interpreting paleobiogeographic data are derived from the same types of problems encountered when working with fossils.

For example, the absence of suitable deposits in a region may be interpreted as a disjunct in the taxa or small sample sizes may lead to incorrect results. In addition, taphonomic processes must be evaluated when interpreting the results of paleobiogeographic data. For example, leaf litter of a temperate forest can represent from 45% to 92% of the tree species, depending on the sample size and dominance pattern of the temperate forest (Burnham,

1993).

Biological data are important components of paleoclimatic reconstructions. In particular, high latitude floras are ideal regions in which to study the effects of climate change on biotas, since the floras in these regions are particularly sensitive to changes in the environment (Spicer and Chapman, 1990; Chapin and Komer, 1994). High latitude regions tend to have relatively less diverse biotas, since few taxa can evolve tolerance to such marginal climatic conditions (Stevens, 1989). Abiotic stress in an environment tends to evolve a terrestrial plant community consisting of a limited number of highly specialized species (DiMichele et al., 1987). During the Triassic, those species inhabiting the higher paleolatitudes in the south experienced extended periods of darkness for part of the year and a more humid seasonal temperate climate, thus differing substantially from the more 13 northern areas. In addition, the northward movement of Gondwana during the Triassic caused concurrent changes in the climate. Paleoecological data show that plant taxa respond to climate change at the individual species level, rather than as whole communities

(Graham and Grimm, 1990). Research on tetrapod faunas suggests that Late Triassic patterns of latitudinal variation among these assemblages can be correlated to those seen among terrestrial plants, including Dicroidium (Shubin and Sues, 1991). However, in the case of Dicroidium, a detailed study of the latitudinal variation of species has not been completed.

Dissertation Objectives

A paleobiogeographical study of Dicroidium species will attempt to answer several questions about the spatial and temporal dynamics within the genus. Such a study will determine whether species are correlated with local or global environmental gradients, whether high latitude floras are more sensitive to environmental changes, and whether or not those regions are low in diversity. Mapping the distribution of species at specified time intervals will also determine the diversity of species through time, providing the opportunity to address macroevolutionary questions.

Since Dicroidium specimens have variable features, it is an ideal system for the application of morphometric analyses. In addition, since Dicroidium is well preserved throughout the Triassic from a number of Gondwana localities, it offers a unique opportunity for a biogeographical study through time. The objectives of this study include:

1. Dicroidium specimens from Antarctica and other Gondwana localities will be identified using both morphological and cuticular features. In order to achieve this goal, the cuticle of

Dicroidium will be described in detail from Antarctic specimens. 14

2. The use of morphometric techniques in classifying form species of Dicroidium will be

evaluated. Using Fourier coefficients, and other qualitative and quantitative information

about the frond and cuticle, specimens will be identified to establish ranges of

morphological variation within and among species.

3. The paleobiogeography of Dicroidium species for the Antarctic and entire Gondwana

region will be analyzed using this standardized system. First-hand information and

published data will be incorporated into this analysis. Time equivalent strata will be

compared to determine the changes that the genus has undergone from Early to Late

Triassic. Trends in species diversity will also be evaluated. Paleobiogeographical studies

will provide information to help understand the spatial distribution and paleoecosystem

dynamics of the plants that bore Dicroidium foliage.

4. The distribution of Dicroidium species will be compared to paleoclimate data during the

Triassic in Gondwana. A paleoclimate simulation program will be used, as well as

information from geological and tree ring data. Knowledge of associated plant taxa will

provide additional information regarding habitats. Patterns between species types and diversity, and regions with specific paleoclimates may provide evidence for possible

adaptation to the environment. Integrating this information will contribute to environmental

reconstructions of a globally warm Triassic. CHAPTER II

MATERIALS AND METHODS

Fossil Material and Localities

Specimens of Dicroidium described in this study were collected from several

Antarctic localities, including Gordon Valley, Fremouw Peak, Mt. Falla, Allan Hills,

Kenyon Peaks, Mt. Bumstead and Mt. Wisting (Table 1). Additional Antarctic specimens from Mt. Shapeless and Horseshoe Mt., belonging to the Canterbury Museum,

Christchurch, New Zealand, were borrowed for study and analysis (Table 1). Most

Antarctic specimens consist of compressions and impressions in shale to sandstone matrixes ranging in age from Early to Late Triassic. Specimens from Mt. Falla provided well-preserved cuticle for examination.

Dicroidium specimens from South America and Australia were borrowed from the

National Museum of Natural History, Smithsonian Institution, Washington, D.C. (Table

1). Large collections of Dicroidium specimens from Argentina, Brazil, and other South

American localities were also examined at the Museo Argentino de Ciencias Naturales

Bernardino Rivadavia, Buenos Aires, Ciudad Universitaria, Buenos Aires, and Programa

Investigaciones Geoldgicas y Paleontoldgicas (PRINGEPA)-CONICET, Corrientes,

Argentina (Table 1). Specimens from these localities also consisted of compressions and impressions in shale to sandstone matrixes ranging in age from Early to Late Triassic.

Specimens from Argentina and several Australian localities provided well-preserved cuticle

15 16 for preparation. Other specimens and data from the literature were used for complete morphometric and paleobiogeographic analyses (Tables 2-6). All recorded synonyms of

Dicroidium were included in the tables, with some corrections in parentheses following the original names.

Techniques

Macrophotos of specimens were taken with a Polaroid MP-3 and/or a Nikon FM-2 camera. Specimen cuticle was examined using light microscopy (LM) and scanning electron microscopy (SEM). Cuticle was removed from the rock matrix using a dissecting needle or nail polish peels (Kerp, 1990). Dilute nitric acid or Schulze's solution was used for maceration, with maceration times ranging from 10 minutes to several hours. Cuticle was washed several times with distilled water followed by staining with 1% safranin in water. Cuticular features were also analyzed without staining. Specimens were then dehydrated using an ethanol series, suspended in xylene, and mounted on glass microscope slides using CoverBond™ mounting medium. Transmitted light photomicrographs were taken with a Zeiss Ultraphot. Cuticular measurements were made using a light microscope with attached camera lucida and stored using the computer program, Sigma Scan™ (Jandel Scientific).

Aluminum SEM stubs were either covered with colloidal silver and a piece of one­ sided silver tape, or double-sided tape was used alone. Cuticle was macerated as above and washed with distilled water before being mounted on the tape. Cuticle was mounted on SEM stubs using dissecting needles and/or glass pipettes with a drop of water. SEM stubs with cuticle samples were sputter coated with gold-palladium and observed with an 17

Hitachi S-500 or a JEOL JSM-820 model SEM. Photographs were taken using Polaroid

55 film with negatives.

Morphometric analyses of specimens were restricted to the pinnules, since entire

fronds are often incomplete. Since variation in pinnule shape on one frond is common,

complete pinnules from the same relative position among frond specimens were also

compared. For some analyses, the pinnules in the second position basal and distal to the

dichotomy were chosen. Video images of fossil pinnules from camera lucida drawings

with black outlines were digitized using a Dage CCD-72 camera (8-bit; 640 x 480 frame)

and frame-grabber attached to a computer. Images were saved as TIF files. The program

Optimas™ (BioScan) was used for gathering images (Campus Microscopy and Imaging

Facility). Outlines of specimens were stored as x-y coordinate data using two hundred and

fifty points per outline gathered by using the MAr Sampled Points option on Optimas.

These coordinates were stored in a spreadsheet and used to calculate Fourier coefficients.

Elliptic Fourier coefficients and reconstructed outlines were generated using the computer

program EFA (Elliptic Fourier Analysis) (Kuhl and Giardina, 1982; Rohlf and Archie,

1984). Twenty-five harmonics per outline were selected and normalized to correct for size,

rotation and starting position of the trace. Based on the reconstructed pinnule outlines, coefficients calculated from 10 harmonics were adequate to describe shape. Principal component analysis (PCA) of the 37 coefficients was performed using standard

multivariate statistical software (SYSTAT). A correlation matrix was used and the data were subjected to orthogonal rotation (varimax), tolerance of 0.001, and 25 iterations.

Two to three factors were analyzed and some were graphed on scatter plots. Cluster analyses of PCA factors using k-means were used to assist in elucidating morphologically distinct groups. See Figure 1 for a summary of the steps completed for morphometric 18

analyses. Comparisons within and between taxa were made for Early, Middle, and Late Triassic specimens.

Paleobiogeographical analyses included comparing the presence and absence of taxa

in each of the different Gondwana localities using the Simpson Index (C/Ni) (Simpson,

1960; Thulbom, 1986) and the Jaccard Index (C/N 1 +N2 -C) (Cheetham and Hazel, 1969).

Two indexes were chosen since they quantify similarity between biotas; presence being

much less ambiguous than absence in fossil data. Multiple similarity indices were

calculated as a check to determine the reproducibility and therefore reliability of the data

(Shi, 1993). In addition, parsimony analysis of endemicity (PAE) was generated using the

computer program Phylogenetic Analysis Using Parsimony (PAUP), which produces area cladograms of sample localities directly from geographical distributions (Rosen, 1988;

Swofford, 1993). Heuristic, and Branch and Bound algorithms were used. Species distributions over time were also compared to other plant taxa present in order to obtain an overall impression of the plant communities. Fossil species identified from all localities were plotted on current paleocontinental reconstructions available on the Macintosh computer program PaleoGeographic Information System™ (PGIS/Mac™) (Ross, 1992).

Paleocoordinate readings were obtained from the paleomap reconstructions and were used in comparing species presence at various paleolatitudes. Species distributions were compared to paleoclimatic data available from the computer program Paleoclimate™ (Ross and Scotese, 1992). In addition, tree ring and geological data, and knowledge of associated plant taxa provided additional information regarding habitats. CHAPTER III

MORPHOMETRIC ANALYSES OF DICROIDIUM

Classification Using Elliptic Fourier Analysis

Outline information is useful in circumstances where there are few or no homologous points on a structure, and for research questions where the entire shape is of interest Several methods can be used to quantify outline information, including Fourier or eigenshape analysis (Lohmann, 1983; Rohlf, 1986). Elliptic Fourier analysis is a procedure that fits a closed curve to an ordered set of points. The curve of a polygon can be represented as a Fourier series of simple sine and cosine curves of decreasing amplitude and period. This curve is decomposed into a number of harmonically related ellipses and when these ellipses are summed, they approximate the original curve. The coefficients generated with this analysis correspond to the numerical constants in the equation for each harmonic. Four coefficients are generated per harmonic. Coordinate data from the entire outline are used, which can be normalized to correct for size, orientation, and starting point

(Kuhl and Giardina, 1982). From ten to twenty-five harmonics have been used in previous biological studies (White et al., 1988). Fourier coefficients, which quantify shape, are then used in multivariate analyses, such as principal components analysis. There has been some controversy about using Fourier analysis in systematics (Bookstein et al., 1982), but these arguments have been successfully addressed (Ehrlich et al., 1983). Eigenshape analysis is a related approach that also quantifies outline shapes. The use of eigenshape analysis has

19 20 been criticized by Ehrlich and Full (1986), and because of these criticisms Fourier analysis will be used to quantitatively classify Dicroidium species.

In this study, Five sets of data were analyzed to explore the use of morphometric data in classifying pinnule shape in Dicroidium. Two hundred and fifty x, y coordinates were chosen to represent each pinnule outline. To evaluate how well each harmonic reconstructed the original pinnule shape, several pinnule outlines were regenerated with the

EFA program. Figure 2 illustrates a pinnule outline regenerated from 250 points using 1,

5, 10, 15, 20 and 25 harmonics. Pinnule outlines regenerated from 10, 15, 20 and 25 harmonics were essentially identical. The same result was obtained for a pinnule with a more complex shape (Figure 3). In this case, as few as ten harmonics were adequate to regenerate the original pinnule outline. Ten harmonics were used in all subsequent analyses. Each harmonic generates 4 coefficients, however, when made invariant to size, orientation, and starting point, the first 3 coefficients degenerate. Therefore, 37 coefficients were used in all subsequent multivariate analyses.

Five different sets of pinnules were analyzed in this study (Appendix A). In the first set of data (A), 330 pinnules of Dicroidium from 12 specimens collected at Shapeless and Horseshoe Mt., Southern Victoria Land, Antarctica, were selected (Table 1). On each specimen all complete pinnules were analyzed. The second set of data (B) is a subset of the above 330 pinnules, selecting complete pinnules at the same relative position on the leaf. One pinnule from the fifth position distal to the dichotomy and outside the fork, was selected on each of the twelve specimens.

The third (C) and fourth (D) sets of data included only specimens illustrating the dichotomous fork, with complete basal and distal pinnules. These specimens were from various localities in Antarctica, Australia, and South America. Two pinnules from each of 21

75 specimens were chosen. For data set C, pinnules in the second position distal to the

dichotomy and outside the fork were selected. For data set D, pinnules in the second

position basal to the fork were selected. The fifth set of data (E) is a subset of C, selecting

a group of twenty specimens that illustrate typical pinnule shapes from several different

species.

Results

The factor scores from a principal component analysis on data set A coefficients

revealed no simple structure for grouping different pinnules (Figure 4). The first two

factors accounted for 41% of the variance. However, increasing the analysis to 3 and 4

factors did not clarify the data set. Although there was no simple structure in the graph, the

scores were arranged in a parabolic curve, indicating that there is a continuous variation in

shape. In data set B, five groups were resolved using cluster analysis (Figure 5). Within

the first two factors, 60% of the variance was explained. Pinnules with larger length to

width ratios scored positively on factors 1 and 2. In general, pinnules with smaller length

to width ratios scored more negatively on factor 2. These values depended on the actual

shape of the pinnule. In subset B, the five clusters correspond to the following species; D.

crassinervis, D. lancifolium, D. odontopteroides, D. spinifolium and D. stelznerianum (Figure 5).

For data set C, the first two factors accounted for 44% of the variance. The factor

scores from principal component analysis revealed complex structure (Figure 6). As in data set A, the points were scattered in a parabolic curve. In data set D, the curvature in the scatter of the points was less obvious, indicating less continuous variation in shape (Figure

7). The pinnule outlines from each specimen were similar to the results for set C. However, since pinnules have smaller length to width ratios and different shapes below the

dichotomy, the parabolic structure was skewed. The first two factors accounted for 43%

of the variance. In data setE, eight groups were supported by cluster analysis (Figure 8).

In this subset, the pinnules were more resolved along the factor scores. Factors 1 and 2

accounted for 52% of the variance in the data. In subset E, the eight clusters represent

pinnules from D. crassinervis, D. dubium, D. elongatum, D. lancifolium, . D

odontopteroides, D. spinifolium, D. stelznerianum and D. trilobitum (Figure 8).

Discussion

The results from the data sets suggest that considerable variation in pinnule

morphology on each specimen exists and is dependent on the position of the pinnule on the

frond. This variation is a biological characteristic of Dicroidium. Pinnules gradually

decrease in length, and consequently change in shape, from the base to the tip of the leaf.

In addition, the pinnules on the inside of the fork tend to be shorter. Other factors may also

play a role in variation, such as the position of the leaf on the plant and phenotypic

plasticity. In addition, some of the variation may be due to the quality of the fossil

specimen and/or errors made in the initial graphing of the outlines. Variation due to

phenotypic plasticity, leaf position on the plant, fossil quality and human error is

misleading and should be accounted for when identifying significantly different pinnule

shapes that classify species. Comparing pinnules in the same relative positions controls for within specimen variation and allows for the identification of significantly different pinnule

morphologies. In general, the pinnule shapes above the dichotomy were more characteristically distinct than those below the dichotomy. Therefore, pinnule shapes above 23

the dichotomy were more successful in resolving clusters of pinnules with similar

morphology.

Classification of Dicroidium species is traditionally based on general pinnule

morphology. As this study shows, there is considerable overlap in pinnule shape, which is

reflected in the literature as taxonomic inconsistencies. Using a morphometric technique

assists in removing some of these ambiguities, but it should be used in addition to other quantitative and qualitative characters that are available on the specimens. Characters such

* as pinnule position and arrangement, rachis width, venation and cuticular features also contribute to species determination. Species not tested in this analysis include those specimens with entire laminae, and bipinnate or tripinnate frond arrangements. Species with these arrangements are distinctly different from pinnate fronds.

The results from this study provide additional evidence in support of the division of species types. For subsets B and E, each cluster represented five and eight species, respectively. Each species has a relatively distinct pinnule shape, as determined by morphometric analysis. The most common species, D. odontopteroides, has relatively short and ovate pinnules. Dicroidium stelznerianum has short, ovate pinnules, as in D. odontopteroides, but they are more rhomboid in shape. Dicroidium trUobitum also has short, ovate pinnules, but there are usually three lobes at the apex of the pinnule. In D. crassinervis, the pinnules are rhomboid, medium in length, and are distinctly separate along the frond. Dicroidium kmcifolium has entire lanceolate pinnules. Dicroidium dubium also has lanceolate pinnules, but they are lobed. In D. elongatum, the pinnules are long and very narrow compared with D. lancifolium. Dicroidium spinifolium has long, narrow pinnules with acute lobes at their base. Each of these shapes was recognized as 24 significantly different by EFA and multivariate analysis. The following chapter provides more detailed descriptions and features of these species.

The use of morphometric analysis provides a method to quantitatively classify shape, and identify species by comparing the pinnule shapes of unidentified specimens to known species. This method is advantageous, since it prevents inconsistencies in identifications and is repeatable. Elliptic Fourier analysis is especially favorable, since it employs an algorithm that can correct for size, orientation, and starting point, and generates coefficients very quickly. This technique is applicable in distinguishing form taxa that are primarily based on morphology, including other types of foliage or plant organs, or in virtually any study that relies on resolving shape. Although high variation in pinnule shape within and among fronds was detected, a subset of up to eight different Dicroidium species was recognized from Antarctica, Australia and South America using this technique. CHAPTER IV

SYSTEMATICS OF DICROIDIUM

Description of Specimens from Antarctica

Eleven species of Dicroidium are recognized from several collections of Antarctic

specimens (Table 1). In the following descriptions, the frond axes above the dichotomy

will be referred to as pinnae each bearing rachial pinnules. Those leaf parts distal to the

fork are attached to the petiole of the frond and will also be referred to as pinnules. Since

Dicroidium is placed in the Pteridospermophyta, the term frond has been consistently used

to describe the leaf of this plant, hence the terms frond and leaf will be used

interchangeably.

Division PTERIDOSPERMOPHYTA

Order CORYSTOSPERMALES

Family CORYSTOSPERMACEAE

Form genus Dicroidium Gothan 1912

Dicroidium crassinervis (Geinitz 1876) Anderson and Anderson 1983 (Figure 9)

Localities: Gordon Valley, Mt. Wisting, Mt. Falla, Kenyon Peaks, Mt. Bumstead and

Shapeless Mountain

25 26

Pinnate fronds are estimated to be up to 10 cm in length. The angle of forking is

20-30°. Rachis and petiole widths are 1-2 mm. Pinnule arrangement is sub-opposite to alternate. Pinnule length to width ratio is less than 1.5:1. Pinnules are rhomboid, with an acute apex and decurrent base (Figure 9). Margin is entire to slightly sinuate. Venation is odontopteroid. These specimens are similar to D. crassinervis forma crassinervis

(Anderson and Anderson, 1983) and Dicroidiopsis crassa var. prolungata (Menendez,

1951).

Dicroidium dubium (Feistmantel 1878) Gothan 1912 (Figure 10)

Localities: Mt. Falla, Mt. Bumstead and Fremouw Peak

Pinnate fronds are estimated to be up to 20 cm in length (Table 7). Angle of dichotomy ranges from 35° to 40°. Rachis width is 1.2-3 mm. Pinnule arrangement ranges from opposite to sub-opposite. Pinnule size is 0.8-4.35 cm in length and 0.5-1.2 cm in width. The length to width ratio is between 1.3:1 and 6:1. Pinnule shapes are ovate to lanceolate with lanceolate pinnules having lobed margins (Figure 10). The extent of lobing varies and usually the lobes are acutely rounded. Apices are also acute with normal to decurrent bases. Venation is alethopteroid. These specimens are similar to D. dubium subsp. dubium (Anderson and Anderson, 1983) and D. dubium var. australe (Retallack, 1977).

Dicroidium elongatum (Carruthers 1872) Archangelsky 1968 (Figure 11)

Localities: Allan Hills, Mt. Bumstead and Mt. Falla

Pinnate frond is 9-12 cm in length. Petiole and rachis width is 1.5-2 mm. Pinnule arrangement along the rachis is opposite to subopposite. Pinnules are relatively long and 27 narrow, commonly with a length to width ratio of greater than 2:1. The pinnules inside and below the dichotomy are only slightly smaller in length. Tips of the pinnae end in a long, narrow segment with shorter adjacent pinnules sometimes appearing as a tri-lobed tip

(Figure 11). Venation limited to one major vein that extends to the tip of each pinnule. D. elongatum (elongata) resembles D. elongatum forma elongation (Anderson and Anderson,

1983).

Dicroidium dutoitii (Johnston 1887) Townrow 1967 (Figure 12)

Localities: Allan Hills and ML Bumstead

Frond is simple with entire to slightly sinuate lamina (Figure 12). Leaf length up to

11 cm. Angle of the dichotomy is 20-30“. Rachis width is approximately 1-2.5 mm. The tips of the pinnae are rounded. Venation is taeniopteroid. Specimens are similar to D. coriaceum subsp. dutoitii (Anderson and Anderson, 1983) and Johnstonia coriacea

(Retallack, 1977).

Dicroidiumfeistmantetii (Johnston 1894) Gothan 1912 (Figure 13)

Localities: Mt. Falla and west of Lindsay Peak

Bipinnate fronds are estimated to be greater than 20 cm in length. Rachis widths are 1.5-3.5 mm. Pinnae arrangement along the main rachis is subopposite to alternate with no pinnules attached to the rachis between second order pinnae (Figure 13). Pinnules are ovate to rhomboid in shape. The venation is odontopteroid in each ultimate pinnule. These specimens are similar to D. dubium subsp. switzfolium (Anderson and Anderson, 1983) andD. zuberi var. feistmantelii (Retallack, 1977). 28

Dicroidium lancifolium (Gothan 1912) Retallack 1977 (Figures 14-15)

Localities: Mt. Falla, Horseshoe Mountain, Fremouw Peak, Allan Hills and west of

Lindsay Peak

Pinnate fronds are estimated to be up to 23 cm in length and 14 cm in width (Table

7). Approximately one-third of the total frond is distal to the fork. Angle of the fork extends from 17° to 55°. Petiole and rachis widths are 1-3.5 mm. Pinnule arrangement along the rachis is opposite to sub-opposite and separate (Figure 14). Pinnule size is 0.7-

4.8 cm in length and 0.35-1.1 cm in width. The length to width ratio is usually greater than 3:1. Pinnule shapes are commonly lanceolate, however, pinnules located inside and below the fork are generally shorter. Apices are acute in lanceolate pinnules and are sometimes obtuse in the shorter pinnules. Bases are normal to decurrent. Venation is mainly alethopteroid with odontopteroid venation in the smaller pinnules (Figure 15).

These specimens are similar to D. odontopteroides var. lineatum (Anderson and Anderson,

1983).

Dicroidium odontopteroides (Morris 1845) Gothan 1912 (Figures 16-18)

Localities: Mt. Falla, Fremouw Peak, Kenyon Peak, Mt. Bumstead, Mt. Wisting, Allan

Hills, Horseshoe Mt., Shapeless Mountain, Gordon Valley and west of Lindsay Peak

Pinnate fronds are estimated to be up to 18 cm in length and 9 cm in width based on incomplete fronds (Table 7). Angle of the fork is usually from 30° to 35° and varies according to position of the frond at preservation. Petiole and rachis widths range from

0.5-3 mm with the petiole at 2.5 mm becoming gradually narrower towards the bifurcation; subsequent rachis width is approximately half this size and gradually narrows to the tips. 29

Pinnule arrangement along the rachis is opposite to sub-opposite with pinnules below the fork often arranged alternately. Pinnules are separate to fused at the base (Figure 16).

Pinnule size is 0.3-3 cm in length and 0.35-0.95 cm in width. Pinnules immediately proximal and distal to the fork are smaller; pinnules gradually decrease in length from the bifurcation to the tips. The pinnule length to width ratio is usually less than 3:1. Pinnule shapes are ovate with obtusely rounded apices; bases are obtuse to decurrent. Some pinnules are rhomboid in shape and these tend to have acute apices. Venation is usually odontopteroid (Figure 17). Pinnules gradually decrease in size and fuse towards the tips of the fronds, and here the venation may range from odontopteroid to taeniopteroid (Figure

18). Many of the smaller fragments in this collection may be parts of larger bipinnate fronds, but this cannot be determined. Specimens are similar to D. odontopteroides forma odontopteroides, D. odontopteroides subsp. orbiculoides (Anderson and Anderson, 1983) and D. odontopteroides (Townrow, 1957).

Dicroidium spinifolium (Tenison-Woods 1883) Anderson and Anderson 1970

(Figures 19-20)

Localities: Allan Hills and Shapeless Mountain

Pinnate fronds are estimated to be over 10 cm in length. Pinnules are arranged along the rachis in an opposite to subopposite arrangement (Figure 19). Pinnule size is between 1-2.1 cm in length. The length to width ratio is approximately 2:1 or greater.

Pinnule shape is narrow lanceolate with lobed margins; apices are acute and bases are decurrent. Lobes are inconsistent and have acute apices. One major vein connects each pinnule to the main rachis, which subsequently divides into 2-3 veinlets (Figure 20). D. 30 spinifolium resembles D. elongatum forma spinifolium (Anderson and Anderson, 1983) and Xylopteris spinifolia (Frenguelli, 1943).

Dicroidium stelznerianum (Geinitz 1876) Frenguelli 1941 (Figures 21-22)

Localities: Gordon Valley, Mt. Wisting, Mt. Bumstead and Shapeless Mountain

Pinnate fronds are up to 10 cm in length. Rachis width ranges from 0.5-1.5 mm.

Pinnule arrangement is subopposite (Figure 21). Pinnules are separate in the basal part of the leaf and become progressively fused towards the tips (Figure 22). The shape of the pinnules is ovate to rhomboid with an angled apex and decurrent base. The widest part of the pinnule ranges from 0.8 to 0.95 cm, base to apex. Venation is odontopteroid, with reduced dichotomous branching of the vein in the pinnule lamina (Figure 22). This species is similar to D. crassinervis forma stelznerianum (Anderson and Anderson, 1983) and Johnstoniastelzneriana var. serrata (Retallack, 1977).

Dicroidium trilobitum (Johnston 1886) Antevs 1914 (Figure 23)

Localities: Allan Hills and Shapeless Mountain

Pinnate frond is up to 8 cm in length. The rachis and petiole widths are 0.6-1.3 mm, respectively. The pinnules are usually suboppositely arranged. Pinnules are rhomboid with distinct lobing at the apex (Figure 23). The tips of most pinnules are tri- lobed. This species is similar to D. spinifolium, however, the pinnules are shorter and generally tri-lobed. Pinnules are approximately 0.5 to 0.8 cm in length. The venation is odontopteroid, with limited vein branching. Specimens are similar to Johnstoniatrilobita

(Townrow, 1967; Retallack, 1977). 31

Dicroidium zuberi (Szajnocha 1888) Archangelsky 1968 (Figure 24)

Localities: Mt. Falla, Fremouw Peak, Allan Hills and Mt Bumstead

Bipinnate frond estimated to be greater than 25 cm in length. Second order pinnae are opposite to subopposite. The rachis ranges from 2-4.1 mm in width. Ultimate pinnules are rhomboid to square in shape (Figure 24). Pinnule dimensions are approximately 0.4 to 0.6 cm. Some pinnules are attached to the main rachis between pinnae. The venation pattern is odontopteroid in individual pinnules. D. zuberi is similar to D. zuberi var. zuberi (Retallack, 1977) and Zuberi zuberi (Azcdrate and Fasola, 1970).

Description of Cuticular Features

Some Dicroidium specimens collected from Mt. Falla possess well-preserved cuticle which has not previously been reported from this region. Cuticle was prepared from three species, D. odontopteroides, D. lancifolium, and D. dubium, collected at approximately 135 m above the base of the Falla Formation type section at Mt. Falla (Table 1) (Barrett, 1969).

In D. odontopteroides, the rachis thickness is 2.4-5.0pm , with pinnule thickness on the adaxial and abaxial surfaces 1.5-4.3 pm and 1.7-4.4 pm , respectively (Table 7).

Epidermal cell shapes are isodiametric to rectangular over venation (Figure 25). Epidermal cell size varies for the different regions of the frond, but range in average from 52.6-101.5 pm in length to 24.6-37.6 pm in width. Epidermal cells are randomly arranged over the lamina. Over venation and rachis regions, the cells are oriented with their long axis parallel with the veins (Figure 25). Anticlinal cell walls are straight. Papillae and/or ridges adorn the periclinal cell walls, often on subsidiary cells (Figure 26). The papillae are distinct and extend over stomatal openings, particularly on the lower surface. The presence and extent 32 of papillae are variable on different specimens (Figures 26-27). Thickenings on the periclinal walls are also common in the center of epidermal cells. Stomatal distribution is amphistomatic, with stomatal complexes oriented randomly or arranged with their long axis parallel to veins. The stomatal index ranges from 1.4-3.9 on the adaxial surface, and 7.9-

9.5 on the abaxial surface. The apparatus width ranges from 49.4-88.9 pim. Stomatal arrangement is paracytic, with two lateral subsidiary cells and/or monocyclic with 4-6 subsidiary cells (Figures 28-29). The guard cells and aperture are thickened and slightly sunken (Figures 30-33). Thickenings extend over the polar cell margins (Figure 34).

Guard cell striations are visible on some specimens (Figures 33, 35). Subsidiary cell shapes are isodiametric to rectangular, with the two lateral subsidiary cells distinct (Figure

28). Trichome bases are rare; base cell arrangement consists of 6-8 surrounding cells

(Figure 36).

D. lancifolium rachis thickness is 2.0-5.5 /

(Figure 38). Epidermal cell size varies from 69.7-121.9 pm x 28.5-66.4 pm . Anticlinal cell walls are straight, and papillae and/or ridges are present on the periclinal cell wall.

Stomatal complexes are amphistomatic, with a greater number of complexes on the lower surface (Table 7). Stomatal complex orientation is random over the lamina, but over veins the long axis of the guard cells is parallel to the long axis of the epidermal cells (Figure 38).

Subsidiary cells commonly range from 2-6 in number, and are either encircling the guard cells in a monocyclic arrangement or consist of two lateral cells flanking the guard cells

(Figures 37- 39). The lateral subsidiary cells are distinctly shaped with the adjacent cell 33

walls shorter than the opposite side. Guard cells are slightly sunken, and guard cell walls

and aperture are thickened (Figure 40). Trichome bases are rare.

In the species D. dubium, cuticle thickness ranged from 2.7-5.1 pm over the rachis

and from 1.5-3.4 p m and 1.0-2.6 pm over the upper and lower surfaces of pinnules,

respectively (Table 7). Epidermal cell shape is isodiametric to rectangular. Epidermal cell

size varies according to region and surface on the frond, but in general the length to width

average is approximately 86 pm x 37 pm (Table 7). The cell arrangement is random, but in

regions over major veins the cells are oriented with their long axis parallel to the veins.

Anticlinal cell walls are straight, and periclinal cell walls include papillae and/or ridges

(Figures 41-42). Stomatal distribution is amphistomatic, with more stomatal complexes on

the abaxial surface. Orientation is random, however, over major veins they are oriented

with the long axes of guard cells parallel to the veins. Apparatus width ranges from 61.0 to

84.8 pm ,measuring across the two guard cells. Guard cells are slightly sunken and guard cell walls, as well as the pore opening, are thickened (Figures 43-44). The thickenings on

the guard cells project over the polar cell margins (Figure 44). Subsidiary cell arrangement

is paracytic with lateral subsidiary cells extending the length of the guard cell, and these are sometimes divided into two to form 4 cells surrounding the stoma (Figure 41). Subsidiary cell ornamentation includes papillae and thickenings, with some papillae extending over the

pore opening. Trichome bases are rare and randomly arranged; when present they are surrounded by 4-8 cells.

Cuticle was also prepared and examined from several Dicroidium specimens from

Cacheuta, in Mendoza, Argentina and Mona Vale and North Avalon in New South Wales,

Australia from collections at the Smithsonian Institution (Table 1). Cuticle was prepared from D. dutoitii (Johnstoniacoriacea), D. dubium, D. zuberi. and D. odontopteroides. 34

In D. dutoitii, the cuticle was very well preserved (Figure 45). Epidermal cell

shape is isodiametric to rectangular (Figure 46). Cell arrangement is random, but in

regions over major veins the cells are oriented with their long axis parallel to the veins.

Epidermal cell size averages between 84-132 ftm x 28-40 pim. Anticlinal cell walls are

straight and periclinal walls lack papillae and ridges. Stomatal complexes are

amphistomatic. Orientation is random, however, the long axis of the guard cells is

generally parallel to the venation. The guard cells are slightly sunken with thickenings extending around the polar region of the guard cells (Figures 47-49). Striations are also visible on guard cell thickenings (Figure 50). Subsidiary cells range in number from 2 to

4. Stomatal arrangement is paracytic, but sometimes one lateral subsidiary cell is divided.

Trichome bases were not observed.

In D. dubium from North Avalon, epidermal cell shape is mainly isodiametric to rectangular and randomly oriented, except over veins where the long axis is parallel to the vein (Figures 51-52). Epidermal cell size is generally smaller than that from Antarctic specimens (Figure 53). The anticlinal walls are straight Papillae and thickenings occur on the periclinal walls of both leaf surfaces, with more extensive ornamentation on the abaxial surface. Stomatal complexes are amphistomatic, with random orientation. The guard cells do not appear to be sunken. Guard cell thickenings are present, but are not as extensive as seen in other specimens (Figure 54). The subsidiary cells are paracytic, with additional encircling cells (Figures 54-56). There are 2-4 subsidiary cells. The two lateral subsidiary cells are less cutinized and stain lighter than other epidermal cells (Figures 52, 56).

Trichome bases were not seen.

In D. zuberi, the epidermal cells are defined by sinuous anticlinal walls (Figures 57-

58). Epidermal cell shapes are isodiametric to rectangular over veins. Cell size averages 35

38-64 ptm x 25-35 pm . Cells are randomly oriented, but are parallel over major veins.

Papillae occur on the periclinal walls in some regions (Figures 59-60). The stomatal

complexes are amphistomatic with random orientation. The guard cells are only slightly

sunken with thickened walls (Figure 61). The number of subsidiary cells varies from 3 to

7. Subsidiary cells adjacent to the guard cells often have straight anticlinal walls, and are

arranged in a monocyclic to incomplete dicyclic pattern (Figure 57).

The epidermal features of D. odontopteroides from Australia are consistent with

those from Antarctic specimens. Epidermal cell shape is isodiametric to rectangular over

venation. Anticlinal walls are straight; periclinal walls have papillae and ridges in some

regions. Stomatal complexes are amphistomatic with random orientation over the lamina.

Guard cells are slightly sunken and the walls are thickened (Figure 62). Subsidiary cells number 2 to 6, and they are arranged in a paracytic to monocyclic pattern.

Discussion Dicroidium is a form genus that exhibits several taxonomic difficulties due to character variability and our lack of knowledge regarding the entire plant. Previous research regarding the genus has resulted in the identification of various species, subspecies, varieties, and forms that have not been used consistently. Many of the problems are due to the overlapping (i.e., continuous) nature of features and extensive morphological variation that results in somewhat subjective nomenclature. The subtle differences among some species, subspecies, varieties and forms may reflect environmental differences, growth habits, position on the plant, preservational differences, normal variation within species, population differences, phenotypic plasticity, or may truly reflect different species. Dicroidium Jremouwensis has morphologically variable pinnules and 36

provides evidence for natural variation in pinnule morphology. Since the anatomy of these

fronds is consistent, they provide no anatomical evidence for separating the taxon into

multiple species (Pigg, 1990). Variation in pinnule shape along the frond, as well as other

quantitative and qualitative characters, such as venation and epidermal features, should be

used when available in order to identify form species.

Eleven species of Dicroidium were identified from several stratigraphic levels and

sites in Antarctica. These include the eight species that were resolved using morphometric

analyses, and one species that is not divided into pinnules, and two that are bipinnate. All

of these species have been previously described in the literature from various Gondwana

localities, including Antarctica. Some of the difficulties in determining identifications are

due to the existence of several species names and genera that are synonyms, and others that

are not well defined. D. lancifolium was synonymized with D. odontopteroides by

Townrow (1957). However, the species name has been reinstated, using a revised

description (Retallack, 1977). The type material by Gothan (1912) is D. odontopteroides,

but other specimens with narrow, lanceolate pinnules are designated as separate species.

Furthermore, differences in epidermal features support the continued use of D. lancifolium.

Cuticular features contribute some usefulness in separating species, however,

cuticle is not always preserved. Because of variability in preservation, the cuticle has not

been well studied from all the species identified to date. Cuticular features tend to vary on

one specimen, so for identification purposes the cuticle must be sampled from all parts of

the frond, but this is not always possible due to fragmented preservation. In addition,

although some epidermal features may be influenced by the environment, certain characters

appear to generally be reliable (Stace, 1965; Barthlott, 1981; Kerp, 1990). For

Dicroidium, it appears that the most characteristic epidermal features are the differences in 37 anticlinal wall margin, subsidiary cell arrangement and number, and the extent of periclinal wall ornamentation. Some of the more variable features, such as extensive papillae and thicker cuticle, may indicate regional differences. Reihman and Schabilion (1978) suggested that among the Pennsylvanian pteridosperms, different species and genera occupied different habitats and may serve as paleoecological indicators based in part on their foliar anatomy. To what extent this applies to other groups of pteridosperms, including Dicroidium, still remains to be determined. Paleoecological studies by Retallack

(1977) suggest that communities of Dicroidium species occupied distinct regional settings at different times in eastern Australia. However, the classification scheme used and lack of quantitative sampling limit the reliability of these results.

The stomatal complexes of Dicroidium are typically gymnospermous. Jacob and

Jacob (1950) noted the similarity of the stomatal complexes to bennettitalean stomata. In general, the use of developmental terms should not be used exclusively to describe the morphology of mature stomatal complexes (Rasmussen, 1981). However, in some circumstances the ontogeny is obvious from the morphology. In the case of Dicroidium,

Townrow (1957) classified stomatal complexes as haplocheilic in origin based on the inconsistency of the lateral subsidiary cell length. Considerable variation in the stomatal subsidiary cell arrangements occurs on individual specimens. For example, in D. dubium two major types of stomatal complexes have been found, one with lateral subsidiary cells extending the length of the guard cell, and another with lateral subsidiary cells divided in half to form 4 cells adjacent to the stoma (Figure 41).

Of the five different species with preserved cuticle detailed here, D. zuberi epidermal features were the most distinct with sinuous anticlinal wall margins. The stomatal complexes in D. dubium were paracytic or the lateral subsidiary cells were divided 38 in half. For D. odontopteroides, stomatal complexes and epidermal cells were generally smaller and monocyclic stomatal arrangements were most common. The stomatal complexes in D. lancifolium were typically made up of two lateral subsidiary cells that were significantly longer than the guard cell thickenings. In D. dutoitii, the stomatal complexes typically consisted of three subsidiary cells and rectangular epidermal cells were common.

Some of the differences in epidermal features noted for the five species support their separation based on morphological characteristics. The epidermal features of species from Australia are similar to those from Antarctica. Dicroidium odontopteroides from

Australia has cuticular features that are consistent with those described from Antarctica.

The specimens of D. dubium from Australia have smaller epidermal cells than in the

Antarctic specimens and epidermal cells that stain differently from lateral subsidiary cells.

These differences are not significant enough to separate the specimens into different species, and may be due to leaf position on the plant, age of the frond or environmental factors. CHAPTER V

DICROIDIUM FLORA FROM ANTARCTICA

Triassic Stratigraphy and Paleoenvironment

Triassic exposures containing plant fossils in Antarctica include those regions along

the central Transantarctic Mountains, Victoria Land, King George V Land and the Prince

Charles Mountains in east Antarctica (Figure 63). The Beacon Supergroup includes strata

dated from the Devonian to the Early Jurassic (Barrett, 1991). Triassic sequences are

included within the Victoria Group, which is Late Carboniferous to Early Jurassic in age

(Kyle and Schopf, 1982). In the Beardmore and Shackleton Glacier regions, central

Transantarctic Mountains, Triassic strata include the Fremouw and Falla Formations

(Barrett, 1969; Collinson and Elliot, 1984a; Barrett et al., 1986). The Fremouw

Formation, composed of three members, is approximately 650 meters (m) thick and disconformably overlies the Permian Buckley Formation at the type section on Fremouw

Peak (Barrett, 1991). The lowermost member is 50 m thick with cyclic sandstone and mudstone sequences of Early Triassic age (Kitching et al., 1972). The other members are each approximately 300 m thick and are Middle Triassic in age. The middle member is also cyclic, but includes a higher amount of mudstone and a thinner layer of volcanogenic

sandstones. The upper member consists of mostly sandstone with some silicified logs and siltstone (Collinson and Isbell, 1986). Both anatomically preserved Dicroidium and compressions occur mainly in the upper member of the Fremouw Formation in the

39 40

Beardmore and Shackleton Glacier regions (Collinson and Elliot, 1984a; Pigg, 1990;

Taylor et al., 1991). At Fremouw Peak, Dicroidium has been identified at 30 m, 80 m, and

100 m below the top of the formation (Rigby and Schopf, 1969; Collinson and Elliot,

1984b).

The Falla Formation overlies the Fremouw Formation, but most exposures are

restricted to sites in the central and southern Queen Alexandra Range. It has also been

identified at limited localities across the Beardmore Glacier and at the head of the

Shackleton Glacier (Collinson and Elliot, 1984a; Barrett et al., 1986). The Falla Formation is approximately 530 m thick at the type section (F2) and composed of a lower and upper section (Barrett, 1991). The lower section is characterized by a cyclic sequence of bluff- forming quartzose sandstone and shale. The upper part of the formation is dominated by volcanic ash. The vitric tuff provides an age boundary of 186±9 m.y. (Faure and Hill,

1973; Barrett et al., 1986). This radiometric data and additional palynological information establish a Late Triassic age for the lower and possibly some of the upper Falla Formation

(Kyle and Schopf, 1982; Farabee et al., 1989). The Falla Formation contains carbonaceous shales with plant fossil remains including well-preserved cuticle. In particular, Dicroidium compressions from approximately 140 m above the base of the type section possess cuticle with well-preserved epidermal features (Barrett et al., 1986;

Boucher et al., 1993). In addition, Dicroidium has been found at approximately 80 m and

30 m above the base of the formation (Barrett, 1969; Barrett et al., 1986).

In southern Victoria Land, equivalent Triassic strata include the Feather

Conglomerate Fleming Member and the Lashly Formation. The Fleming Member is Early

Triassic in age, with a cyclic sequence of bluff-forming sandstone (Kyle and Schopf, 1982;

Barrett, 1991). The Lashly Formation, which is Middle to Late Triassic in age, has been 41 divided into four members (Member A-D), each up to 90± m thick (Collinson et al., 1983;

Barrett, 1991). Member A is represented by alternating volcaniclastic sandstone, siltstone and mudstone, and member B consists of massive sandstone with abundant fossil logs.

Member C includes cyclic sandstone, carbonaceous siltstone and mudstone, and coal.

Member D consists of coarse-grained quartzose sandstone, siltstone and shale. At its greatest exposure the Lashly Formation may extend up to 520+ m (Barrett et al., 1971;

Barrett, 1991). Although the formation is subdivided differently and is thinner, members can be lithologically correlated with Triassic sequences of the Fremouw and Falla

Formations in the central Transantarctic Mountains. Plant fossils, including Dicroidium, have been identified from all members in the Lashly Formation at various localities in southern Victoria Land (Collinson et al., 1983; Gabites, 1985).

In northern Victoria Land, Triassic strata include the Section Peak Formation. The

Section Peak Formation is approximately 50-22Q± m thick, varying in thickness at different localities (Collinson et al., 1986; Barrett, 1991). In most regions, it unconformably overlies early Paleozoic granite. The formation is mainly composed of cyclic sandstone sequences characteristic of Beacon Supergroup strata. A Late Triassic age has been proposed for the Section Peak Formation based on the similarity of its sequences to those in the upper Lashly and Falla Formations (Collinson et al., 1986). Dicroidium specimens have been identified from Vulcan Hills from plant-bearing black mudstones underlying sandstone with fossil wood debris (Tessensohn and Madler, 1987). The formation was not identified, but is most likely equivalent to the Section Peak Formation based on its sandstone sequence (Collinson et al., 1986).

In King George V Land, Triassic Beacon Supergroup sequences have been identified at Horn Bluff (68'22'S, 150‘00'E) (Mawson, 1940). These sequences of 80± m 42 are capped by dolerite and include sandstone, conglomerate and thin lenses of coal (Ravich et al., 1968). Small plant debris is abundant in the sediments, but has not been identified. Microfloras indicate a Triassic age for this part of the sequence (Ravich et al., 1968).

In the Prince Charles Mountains, Triassic strata are included within the Ragstone

Bench Formation (Webb and Relding, 1993a). The formation is 500+ m thick, consisting mainly of sandstones that extend from Permian to the Triassic (Barrett, 1991; Webb and

Fielding, 1993a). On Jetty Peninsula, above the coarse-grained sandstone, a section of approximately 60 m consists of interbedded sandstone, siltstone and paleosols. This distinct section is the Jetty Member. Above the top of the Jetty Member, an upper section contains rare, dark gray siltstone beds containing plant fossils, including Dicroidium

(Webb and Fielding, 1993a). The upper part of the Ragstone Bench Formation has been recently dated as Late Triassic based on the associated microflora (Webb and Fielding,

1993b).

The cyclic nature of the sequences within the Victoria Group at all localities suggests that the Triassic sediments are mainly fluvial in origin. In the central

Transantarctic Mountains, most strata were deposited by low sinuosity braided to anastomosing rivers, permitting the accumulation of extensive overbank deposits on the surrounding flood plain (Barrett, 1969; Collinson and Isbell, 1986; Taylor et al., 1989). In southern Victoria Land, sedimentological evidence within the Lashly Formation of the

Allan Hills also provides evidence of a fluvial depositional setting, including both braided and meandering low-sinuosity stream environments (Ballance, 1977; Gabites, 1985; Isbell et al., 1990). At many sites and stratigraphic levels, foliage specimens are usually complete, indicating that, although allochthonous, remains may have traveled only a relatively short distance. During the Early Triassic, floras extended at least as high as 78°S paleolatitude in the Antarctic region. From the Early to Late Triassic, Gondwana steadily shifted northward. The fauna and flora provide evidence for a relatively mild paleoclimate from the Early to Late Triassic. Assorted amphibian and reptilian remains in the lower Fremouw

Formation indicate a mild Early Triassic (Kitching et al., 1972). A diverse flora from the

Middle and Late Triassic in the central Transantarctic Mountains and southern Victoria

Land also suggests a temperate climate (Gabites, 1985; Taylor et al., 1986). In addition, tree ring studies of fossil logs at Fremouw Peak and Gordon Valley (Fremouw Formation) possess rings from 0.3-5.3 mm wide (Jefferson and Taylor, 1983; Taylor, 1989; Taylor et al., 1991). In southern Victoria Land, fossil logs (Lashly Formation) have rings up to 3 mm in width (Gabites, 1985). In general, growth rings of this size provide additional evidence fora seasonal climate, and indicate potentially high growth rates. Polar regions should be able to support rapid tree growth, as long as the summer light levels are adequate, shading is limited, water levels are favorable, and temperatures are high enough to sustain metabolic processes (Creber and Chaloner, 1985).

Dicroidium Floras Previously Described from Antarctica

Dicroidium was first collected in Antarctica from Triassic exposures on Mt.

Fleming and Shapeless Mountain, Southern Victoria Land (Gunn and Warren, 1962).

Those specimens, which were poorly preserved, were described and illustrated by

Plumstead (1962). Approximately fifteen different species have been identified from the central Transantarctic Mountains, Victoria Land, and east Antarctica (Townrow, 1967;

Rigby and Schopf, 1969; Boucher et al., 1993; Webb and Fielding, 1993a) (Table 2). To date the compression species of Dicroidium from Antarctica resemble those found in other 44

Gondwana Triassic localities. The previously described Antarctic Dicroidium compression

flora includes; D. acutum (Diplasiophyllum acutum), D.coriacea (Johnstoniacoriacea), D.

crassinervis, D. dubium, D.dutoitii, D.elongata (Xylopteris elongata), D.feistmanteUi, D.

lancifolium (D.lanceolatum), D. odontopteroides, D. pinnis-distantibus, D. rhomboidalis,

D. spinifolium, D. slelzneriana (Johnstoniastelzneriana), D. trilobita {Johnstonia trilobita),

and D. zuberi. Four of these species have been synonymized with other species:

Dicroidium acutum has been renamed to D. lancifolium, some specimens of D. coriacea to

D. dutoitii and others to D. odontopteroides, D. pinnis-distantibus to D. lancifolium, and

D. rhomboidalis to D. stelznerianum.

Dicroidium at these localities is commonly associated with various types of fern

and gymnosperm foliage, assorted stem compressions, sphenophytes, gymnosperm

cones, various seeds, and corystosperm reproductive organs. At several localities in

Antarctica, corystosperm stem and reproductive organs in the Fremouw, Falla and Lashly

Formations have been found in rocks containing Dicroidium (Table 8). Dicroidium floras

have been collected but remain undescribed from Robison Peak (77°13'S, 160°18'E),

Skew Peak (77° 13.5'S, 160°40'E), Mt. Fleming (77’33’S, 160°05'E), Mt. Kirkpatrick

(84* 18'S, 166° 16'E) and Schroeder Hill (85C23'S, 175°12'W) (Plumstead, 1962; Pinet,

1965; Barrett, 1969; Rigby and Schopf, 1969; Barrett et al., 1973; Collinson and Elliot,

1984a).

Description of D icro id iu m Assemblages from Antarctica A. Mt. Falla

Dicroidium compressions and impressions in shale were identified from Mount

Falla located in the Queen Alexandra Range. The descriptions are based on over two 45 hundred foliage specimens from at least four different stratigraphic levels within the Falla

Formation (Table 1). Associated compression fossils at various levels include fern foliage, sphenophytes, gymnosperm foliage and cones, seeds, and Pteruchus and Umkomasia-like reproductive organs. The specimens from collections T-5 and T-7 consist of compressions and impressions in gray shale (Falla Formation) collected 135-151 m above the base of the type section, and are Camian-Norian in age. At this level, cuticular material is well preserved. Specimens are identified as Dicroidium odontopteroides (77 %), followed by

D. lancifolium (21 %), and D. dubium (2 %) (Figures 64-65). The morphological and cuticular features for these species at this location are summarized in Table 4.

Dicroidium crassinervis, D. elongatum, D. odontopteroides, and D. zuberi are found in brown-gray shale about 120 m above the base of the type section (Figure 66).

Specimens from other levels include D. dubium, D.feistmanteli, D. lancifolium, and D. odontopteroides, near the quartz pebble sandstone, approximately 130 m above the base of the type section in thin gray shale fragments (Figure 67). Dicroidium zuberi was identified just below quartz pebble sandstone at 110 m above the base (Figure 68).

B. Gordon Valley

Dicroidium occurs in light gray siltstone at Gordon Valley (Fremouw Formation) in the Queen Alexandra Range. The specimens from collections T-4, T-8 and T-9 consist of compressions and impressions collected from the bedding plane in which several in situ trunks occur (Table 1) (Taylor et al., 1991). Species include D. crassinervis, D. odontopteroides and D. stelznerianum (Figures 69-70). This site is within the upper part of the Fremouw Formation and is believed to be Middle Triassic in age. Cuticular preparations from the Gordon Valley specimens yielded indiscernible epidermal features. 46

C. Fremouw Peak

Dicroidium compressions and impressions in shale/mudstone were identified from

Fremouw Peak located in the Queen Alexandra Range. The descriptions are based on over

seventy foliage specimens from four different stratigraphic levels within the Fremouw and

Falla Formations (Table 1). Collection T-10 consists of D. odontopteroides impressions in

dark gray shale believed to be near 512 m above the base of the type section of the

Fremouw Formation (Figure 71). Another level containing Dicroidium occurs in the upper

Fremouw Formation, below the siliceous horizon and above coal at 518 m above the base

(Figure 72). Dicroidium odontopteroides and D . dubium are present as

compression/impressions in light to medium gray shale. Specimens from other levels

include D. odontopteroides from below the siliceous material of the upper Fremouw

Formation in a light gray shale/siltstone. Within the Falla Formation at 84 m above the

base of the section, D. lancifolium, D. odontopteroides, and D. zuberi are present in a light

gray mudstone.

D. Kenyon Peaks

Specimens found in moraine material from the upper east slopes of the eastern peak

include D. odontopteroides and D. crassinervis in a dark gray shale/sandstone. Detailed stratigraphic information is lacking, however, these specimens most likely correlate to the plant zone in the upper Fremouw and lower Falla Formations (Table 1). 47

E. west of Lindsay Peak, Marshall Mountains

Approximately thirty specimens from this region were collected 76 m above the lower Falla Formation boundary. D.feistmanteli, D. lancifolium and D. odontopteroides are present in an outcrop of dark gray shale that is regarded as Camian-Norian in age (Table 1).

F. Mt. Bumstead

Dicroidium compressions in light to medium gray siltstone/mudstone are found at

Mt. Bumstead located near the Shackleton Glacier. The identifications are based on over one hundred and fifty foliage specimens from one stratigraphic level within the Fremouw

Formation (Table 1). Associated compression fossils include Taeniopteris, Yabeiella,

Czechanomkial, Sphenopteris, ginkgo foliage, and coniferous leaves and shoots. The presence of Taeniopteris indicates an Early to Middle Triassic age (Townrow, 1967;

Gabites, 1985). Specimens are identified as D. crassinervis, D. dubium, D. dutoitii, D, elongatum, D. odontopteroides, D. stelznerianum, and D. zuberi (Figures 73-75).

G. Mt. Wisting

A small number of specimens from Mt. Wisting have been collected from light to medium gray shale/siltstone that is believed to be part of the Fremouw Formation (Table

1). Dicroidium odontopteroides, D. crassinervis, and D. stelznerianum have been identified (Figures 76-77). The associated flora includes sphenophyte stems and foliage, fern foliage, and various stem compressions. 48

H. Shapeless Mountain

Nearly fifty specimens from this locality were examined from the Antarctic plant

fossil collection at the Canterbury Museum, Christchurch, New Zealand. These specimens

are from two different stratigraphic units on the southwest ridge just west of the Shapeless

Mountain summit. Approximately half of the specimens are within Unit 4 of the Lashly

Formation, 77 m above the base of Section S4 (Barrett et al., 1973). Specimens are well

preserved in light gray siltstone and include D. odontopteroides, D. stelznerianum, D.

trilobitum, D. crassinervis, and D. spinifolium (Figure 78). Within Unit 6 of Section S4,

108.5 m above the base of the section, specimens are preserved in dark gray laminated

siltstone and include D. odontopteroides.

I. Horseshoe Mountain

Approximately ten specimens were examined from collections at the Canterbury

Museum, Christchurch, New Zealand (Table 1). The Triassic locality is on a nunatak west

of Horseshoe Mountain in southern Victoria Land. The stratigraphic level is Unit 8 of

Section H3, approximately 45 m above the base of the section (Askin et al., 1971).

Specimens occur in a light to medium gray siltstone and include D. lancifolium (Figure 79)

and D. odontopteroides.

J. Allan Hills

Triassic exposures in southern Victoria Land are within the Lashly Formation,

which extends from Middle to Late Triassic. Dicroidium compressions and impressions

from several different levels and locations in the Allan Hills have been identified. In collections 70-3 and SR 139, specimens occur in medium gray to black, flaky, micaceous 49

shale. Species include D. lancifolium, D. odontopteroides, and D. zuberi. These

specimens were collected from about 46 m below the summit sandstone at the southern end

of the eastern leg of the Allan Hills and are most likely Late Triassic in age (Member C).

Dicroidium specimens in the more recent collections (T-8 and T-l 1) from the Allan

Hills were collected from several levels within the Lashly Formation, spanning the Middle

to Late Triassic. At the Feather Bay site, four levels within Member C (Late Triassic) are identified that are about 1 m in thickness (Plant levels 1-4) (Figure 80). The levels are approximately 1 to several meters apart. The plant fossils occur in light-medium gray mudstone and siltstone. When cuticle is present, it is poorly preserved. Floral assemblages at each level vary in species diversity and plant types. All levels, except Level

3, are dominated by Dicroidium foliage. These levels are also characterized by the presence of corystosperm reproductive organs. In Level 1 (the oldest level), the dominant foliage is

Dicroidium odontopteroides (Figure 81). The flora also includes D. dutoitii, dispersed seeds, and corystosperm reproductive structures including Pteruchus and cupules.

Higher in the section at Level 2, Dicroidium dutoitii is dominant, with D. odontopteroides also present (Figures 82-83). The assemblage at this site also includes corystosperm reproductive organs, sphenophytes, stem compressions, and conifer cones and foliage. There is a distinct change in the flora at Level 3, which is dominated by conifer cones and foliage, osmundaceous ferns, sphenophytes, and various seeds.

Dicroidium dutoitii is a rare occurrence at this level. t In Level 4, several species of Dicroidium are present including D. elongatum, D. trilobitum, D. odontopteroides and D. spinifolium (Figures 84-86). Also present are corystosperm reproductive structures and seeds, osmundaceous ferns, and conifer foliage. 50

At Roscolyn Tor, four plant levels have also been identified corresponding to those stratigraphic sequences found in Feather Bay. In the northern section of Roscolyn Tor,

Level 2 contains Dicroidium odontopteroides, D. stelznerianum,. dutoitii D (Figure 87).

Also present are corystosperm reproductive organs, seeds, osmundaceous fern foliage, sphenophyte stems and foliage, stem compressions and a rare appearance of conifer foliage. Upper Member C in south Roscolyn Tor is a dark gray shale and Dicroidium foliage is abundant, including D. lancifolium and D. odontopteroides (Figure 88).

Discussion

Fifteen species of Dicroidium are recorded in the literature from thirteen different localities in Antarctica ranging from Early to Late Triassic in age. In this study, eleven species from ten localities were examined and described. There are at least five other

Triassic localities with Dicroidium, however, their species remain unidentified. Those specimens identified as Dicroidium from Williams Point, South Shetland Islands (Orlando,

1968; Lacey and Lucas, 1981; Lemoigne, 1987; Baneiji and Lemoigne, 1987) were not included in the Antarctic specimen data, since the fragmentary nature of the specimens does not support a positive identification, and subsequent evidence supports a Cretaceous age for this region (Rees and Smellie, 1989).

From the Late Permian to the Early Triassic, the Antarctic flora included a mixture of Glossopteris, Taeniopteris, and Dicroidium foliage compressions from the Allan Hills,

Mt. Bumstead and other localities. From the late Scythian to the late Norian, Dicroidium foliage was abundant in the Southern hemisphere. Sphenophytes, pteridophytes and other gymnosperms, including ginkgophytes and conifers, also occurred with Dicroidium. By 51 the late Mesozoic, the dominant role that pteridosperms played in the ecosystem was diminished and they were subsequently replaced by conifers and Bennettitales.

Through the Early to Late Triassic, some of the different regions within Antarctica are similar in species compositions. For example, the lists of species found at localities in

Southern Victoria Land tend to be more similar to each other than those found in equivalent strata in Mt. Falla, central Transantarctic Mountains. This implies that species populations were not homogeneous across the Antarctic craton, but may have been limited by spatial or environmental barriers.

At several sites, different species occur at various stratigraphic levels through the

Triassic. However, at some levels the same assemblages of species tend to occur at several different localities. For example, localities containing D. elongation, D. spinifolium, D. trilobita and/or D. stelznerianum usually occur at the same stratigraphic level during the

Late Triassic. D. lancifolium is often in the same assemblages as D. zuberi, D. odontopteroides and D. dubium during the Middle or Late Triassic. In addition, D. crassinervis and D. stelznerianum are often in the same assemblages. At some of these levels, the presence and/or abundance of other plant types changes. For example, in the

Allan Hills within Late Triassic strata of the Lashly Formation, osmundaceous foliage, sphenophytes and conifer remains vary in abundance at different levels, correlating with changes in species assemblages. There may be a biological explanation for this association, such as adaptation to similar environments i.e., a community of species occupying different niches. Some species are spatially and temporally ubiquitous, such as

D. odontopteroides.

Changes in floral assemblages and Dicroidium species at these localities during the

Late Triassic may reflect several factors, including changes in the alluvial environment due 52 to flooding, successional changes in the region, seasonality and/or climate changes.

Further resolution of the stratigraphy is dependent on additional geological data, especially detailed palynology. If inferences are to be made regarding the temporal and spatial trends in the genus, then independent means of correlating strata are required. In general, the species from Antarctica are most similar to the Australian and South American taxa during the Late Triassic (Jones and DeJersey, 1947; Boucher et al., 1993). In a subsequent chapter, theDicroidium flora of Antarctica will be compared to floras from other regions of

Gondwana to assess their similarities and detect spatial and temporal changes in the genus. CHAPTER VI

DICROIDIUM FLORA FROM OTHER GONDWANA LOCALITIES

Dicroidium from South America Over 25 species of Dicroidium have been identified from South America, including localities in Brazil, Chile and Argentina (Table 3). Triassic exposures containing plant fossils in these regions include the Parand Basin in southern Brazil, Juan de Morales and

Pacific Basins (Vallenar-Los Vilos and Talca-Bfo Bfo Sub-basins) in Chile, and

Ischigualasto-Villa Union, Marayes-Desaguadero, Cuyo, MalargUe, North-Patagonian, and Deseado Basins in Argentina (Bigarella and Salamuni, 1967; Cecioni and

Westermann, 1968; Stipanfcic, 1983). In the Parana Basin, the Santa Maria Formation is the only known Triassic (Ladinian-Camian) formation that contains Dicroidium megafossils (Bigarella and Salamuni, 1967; Bortoluzzi et al., 1983; L6pez-Gamundf et al.,

1994). In Chile, the formations containing Dicroidium are mainly Upper Triassic

(Stipanfcic, 1983). A rich Dicroidium flora of late Middle to early Late Triassic is characteristic of fossil-bearing formations in Argentina (Stipanfcic, 1983; L6pez-Gamundf et al., 1994).

Lower Triassic rocks are essentially absent from South American sedimentary basins (Rocha-Campos, 1973). An extended period of erosion during the late Paleozoic and Mesozoic is represented by an unconformity separating rocks of these ages. As a result, Dicroidium is unknown in South America during this time period. During the

Middle and Late Triassic, a rich flora was preserved including typical Gondwanan 53 54

elements, such as Dicroidium and other corystosperms, Ginkgoales, cycads, sphenophytes

and pteridophytes (Kurtz, 1921; Bonetti, 1966; Stipanfcic, 1983). The rise of the

Dicroidium-dominated flora may have influenced faunal composition, as supported by a

synchronous radiation and increase in varieties of therapsids and archosaurs (Bonaparte,

1982). Geological and paleontological evidence supports a warm oxidizing climate for the

lower Middle Triassic, with a change to more humid conditions in the later Middle and

early Late Triassic. The uppermost Triassic red beds signify a return to a seasonally warm

climate with wet and dry seasons (Rocha-Campos, 1973; Stipanfcic, 1983).

Dicroidium from Australia, Tasmania and New Zealand

More than 30 Dicroidium species have been identified from Australia, Tasmania

and New Zealand (Table 4). Triassic exposures containing plant fossils in Australia

include the Canning Basin in Western Australia, the Bowen, Esk, and Ipswich Basins in

Queensland, the Clarence-Moreton, Lome, Great Artesian, Springfield and Sydney

Basins, the New England Fold Belt in New South Wales, and the Tasmania Basin

(Townrow, 1962c; Dickins, 1973; Collinson et al., 1987; Playford and Rigby, 1988;

Veevers et al., 1994b). In New Zealand, Triassic outcrops occur along the middle and southern part of South Island (Retallack, 1987).

Dicroidium specimens have been identified from rocks that are Scythian to Norian

in age from sedimentary basins in Australia and New Zealand. In Tasmania, the Triassic coal measures extend from Middle to Late Triassic (Collinson et al., 1987). As with

Triassic sedimentary basins in South America, the preservation of non-marine macrofossils

in Lower Triassic rocks is limited in these regions (Balme and Helby, 1973; Dickins,

1973). However, an Early Triassic record of Dicroidium is present in the Lome and 55

Sydney Basins in eastern Australia, and from the Murihiku Supergroup in southern New

Zealand (Balme and Helby, 1973; Waterhouse, 1973; Retallack, 1987; Retallack, 1995).

Although an Early Triassic record exists for Dicroidium, Australian floras

dominated by this genus did not attain high diversity and stability until the Middle Triassic

(Balme and Helby, 1973; Retallack, 1995). At the Permian-Triassic boundary, the low

diversity flora consists of conifers, lycopods, and Dicroidium callipteroides (Retallack,

1995). A typical Gondwanan flora was well-developed by the Middle and Late Triassic of

Australia and Tasmania, including Dicroidium and other pteridosperm organs,

ginkgophytes ( Ginkgoites), cycadophytes ( Taeniopteris, Linguifolium, Pseudoctenis),

sphenophytes (Equisetum stems, Neocalamites) and pteridophytes ( Cladophlebis,

Dictyophyllum) (Townrow, 1962c; Balme and Helby, 1973; Playford and Rigby, 1988).

In New Zealand, similar associations of plant groups dominated by Dicroidium are present

and their composition is dependent on their location relative to the paleocoast (Retallack, 1987).

In the Early Triassic, the formation of red beds in eastern Australia and Tasmania

supports a warm, wet-dry seasonal climate for this region (Dickins, 1973). By Middle

Triassic, a shift to more humid conditions is supported by widespread preservation of coal

(Collinson et al., 1987). In New Zealand, a humid, seasonally cool temperate climate has

been proposed for the Middle Triassic based on fossil soils, coal formation, and fossil tree

rings (Waterhouse, 1973; Retallack, 1987). 56

Dicroidium from Africa and Madagascar

Less than 20 Dicroidium species have been identified from Africa and Madagascar

(Table 5). Triassic exposures containing plant fossils in these regions include the

Mourzouk Basin in Libya, the Luangwa and Luano Valley basins in Zambia, the Karoo

Basin extending from Zimbabwe to South Africa, and the Morondava Basin in Madagascar

(Besairie, 1972; Brenon, 1972; Lejal-Nicol, 1979; Dingle et al., 1983; Veevers et al.,

1994a). In the Mourzouk Basin, the Unar Formation is the only known Triassic formation that contains Dicroidium megafossils (Lejal-Nicol, 1979). In Zambia, the formations containing Dicroidium flora are Middle Triassic in age (Lacey and Lucas, 1984; Veevers et al., 1994a). A diverse, Late Triassic Dicroidium flora is characteristic of the Molteno

Formation of the Karoo Basin in Zimbabwe and South Africa (Lacey, 1970; Anderson and

Anderson, 1984). In Madagascar, Dicroidium has only been identified from the Sakamena

Formation (Early Triassic) (Boureau, 1949; Plumstead, 1962).

Dicroidium species have been identified from Early to Late Triassic rocks in the regions of Africa and Madagascar. In the Karoo Basin, the Burgersdorp Formation

(Scythian) contains a flora of low diversity including Dicroidium (DuToit, 1927). Three species of Dicroidium have been identified from Anisian to Ladinian rocks of Zambia

(Lacey and Lucas, 1984). However, during the Middle to Late Triassic, a rich flora was preserved, including Dicroidium and other pteridosperms, ginkgophytes, cycads, conifers, sphenophytes and pteridophytes (Anderson and Anderson, 1984; Anderson, 1990). Lejal-

Nicol (1979) reports a flora with mixed Northern and Southern hemisphere elements in

Libya. During the late Paleozoic and early Mesozoic in Madagascar, marine transgressions formed marine deposits. Within these deposits plant remains are present, indicating that these localities were close to the shoreline (Brenon, 1972). The Early Triassic flora 57

consists of Equisetales, Filicales, including Cladophlebis, pteridosperms, including

Dicroidium, Ginkgoales, and Coniferales (Carpentier, 1935). Later Triassic continental

facies (Isalo I) from this region are distinguished by frequent marine invasions marked by

marine fossils and occasional plant remains of silicified wood (Brenon, 1972).

Geological and paleontological evidence supports a relatively warm, arid climate

for the Early Triassic. More humid conditions prevailed in the later Middle and Upper

Triassic, supported by the presence of coal measures (Veevers et al., 1994a). Since it is

farther north in the interior of the Gondwana land mass, a seasonally warm climate with

greater extremes in temperature, and wet and dry seasons is predicted for this time

(Anderson and Anderson, 1983).

Dicroidium from India

Approximately 10 Dicroidium species have been identified from India (Table 6).

Triassic exposures containing Dicroidium and other plant fossils in this region include the

South Rewa Basin in Madhya Pradesh, the Damodar Basin in Bihar, and the Godavari

Basin in Andhya Pradesh (Roy Chowdhury et al., 1975). The most extensive preservation

. occurs in the South Rewa Basin within Middle to Late Triassic rocks of the Nidpur,

Parsora and Tiki Formations (Srivastava, 1987; Bose et al., 1990). The only Early

Triassic (Panchet) record of Dicroidium has been tentatively reported from the Damodar

Basin (Baneiji and Bose, 1977). In the Godavari Basin, Dicroidium has been identified from the Maleri Stage (Late Triassic) (Baksi, 1967; Lakhanpal et al., 1976).

The study of Gondwana floras from this region is impeded by the often fragmentary condition of plant macrofossils and lack of stratigraphic resolution. Further complications are due to the fact that the Indian flora contains characteristic Permian genera 58 that extend into the Early Triassic, as well as taxa usually considered characteristic of the

Northern hemisphere (Lele, 1964; Srivastava, 1992). Dicroidium from Lower Triassic rocks is scarce (Lele, 1976a) and the presence of pteridosperms increased during the later part of the Triassic (Srivastava, 1974). During the Middle to Late Triassic, a rich flora developed including typical Gondwanan forms, such as Dicroidium and other corystosperms, Ginkgoales, cycads and sphenophytes (Roy Chowdhury et al., 1975;

Maheshwari, 1977; Maheshwari, 1992). Pteridophyte and conifer remains are rare in some formations, but well represented in others (Roy Chowdhury et al., 1975; Bose et al.,

1990).

The presence of red beds supports a warm, semi-arid climate for the Early to

Middle Triassic, with a change to more humid conditions in the later Middle and early Late

Triassic (Lele, 1976b). However, coal deposits are generally absent. The uppermost

Triassic returns to a seasonally warm climate with wet and dry seasons (Lele, 1964). CHAPTER VII

PALEOBIOGEOGRAPHY OF DICROIDIUM

Distribution of the Genus

Using information from Tables 2-6, the spatial distribution of Dicroidium was mapped using the computer program PaleoGeographical Information Systems

(PGIS/Mac™) (Ross, 1992). Figure 89 is a present-day map of the Southern hemisphere that summarizes all the known Triassic basin localities where Dicroidium species have been collected and identified. All maps are rectilinear projections using 90°S latitude centering.

Paleomaps were also produced, selecting for the "world" plate rotation model containing recent information on plate tectonic positions for the Mesozoic. Using these maps, localities were mapped for the Early, Middle, and Late Triassic (Figures 90-92).

As Figure 89 illustrates, Dicroidium has been found at several Triassic sites in the

Southern hemisphere. For the Early Triassic, specimens have been located in southern

Africa, Madagascar, India, Antarctica, Australia and New Zealand (Figure 90). Several more regions with exposed sites of Middle Triassic age contain specimens of Dicroidium, including Africa, South America, Antarctica, India, New Zealand, Australia, and Tasmania

(Figure 91). Finally, Dicroidium has also been found at Late Triassic sites from these regions (Figure 92).

Using the parametric climate modeler, Paleoclimate™, the results of climate simulations were illustrated on paleomaps of Gondwana along with taxa distributions

59 60

(Ross and Scotese, 1992). Climate simulations for the Middle (Anisian) and Late (Norian)

Triassic were completed (Figures 93-100). Simulation information for the Early Triassic was not available. The program was set at global simulation, grid spacing of 4°, time period (Anisian-237 my, Norian-216 my), tilt of the earth (±23*) (summer and winter seasons), and average temperature gradients over land (±34°C) and water (0-34°C).

Relative highs and lows for precipitation and temperature in different regions are indicated on the paleomaps by the symbols "H" for high, and "L" for low.

During the Anisian in the summer, precipitation was relatively high along the boundary of the Gondwana land mass (Figure 93). South America, Antarctica and

Australia have Dicroidium localities along their boundary that are indicated as high precipitation regions. Interior sites in India, Africa and Australia were relatively dry.

Temperatures for localities in the interior of Gondwana were elevated. Along the southern boundaries and the exterior southern coast of Australia and South America, temperatures were proportionately low for this time period and season (Figure 94). During the winter in the Anisian, precipitation was comparably low along the boundary of southern Gondwana and some internal regions, such as India, the tip of South America, and part of northeastern Australia (Figure 95). Higher precipitation is indicated in the interior regions of Africa, South America and Australia. Temperatures were relatively low for the entire region of Gondwana during the winter months (Figure 96). The coldest regions are estimated at -53"C, however, most areas with Dicroidium averaged around 0°C.

Paleogeography and paleoclimate changed during the transition to the Norian. In the summer, precipitation was relatively high along the southern boundary of Gondwana, the interior of India, part of the western coast of South America, and Tasmania (Figure

97). Interior regions were comparably dry, especially Africa, Australia, part of Antarctica 61 and eastern South America. Some regions of dryness are also indicated along the western coast and tip region of South America, part of the Antarctica peninsula and near Tasmania.

Temperatures were relatively high (15-39°C) along the northern part of Gondwana, interior of southern Africa, part of the interior of southern South America, and part of Australia (Figure 98). Lower temperatures were distributed along the southern part of Gondwana, and some areas along eastern Australia, southern Africa, and southern South America.

During the winter in the Norian, precipitation was comparably high within interior regions, including part of southern Africa, part of India, Antarctica, Australia, and central southern

South America (Figure 99). Precipitation was also higher along part of the coasts of

Australia and South America. Lower levels of precipitation were restricted to southern

Gondwana, the tip of southern Africa and South America, part of India, and part of eastern

Australia. Relatively low temperatures were prevalent in almost all regions of Gondwana, except for an area along the coast of central South America (Figure 100).

Paleobiogeographical Analyses

Using Tables 2-6, and field and lab observations by the author, a list of 27

Dicroidium form species and their temporal and spatial distributions were summarized in

Table 9. Tables 2-6 include the records of species assigned by the author(s), however, in many cases these identifications are out-dated and incorrect. Invalid and inaccurate names were corrected to complete the table. The Triassic was split into Early, Middle and Late, and Antarctica, South America, Australia (including Tasmania and New Zealand), Africa

(including Madagascar), and India were the five major regions considered.

In general, Dicroidium species totals are lower for the Early Triassic in all regions

(Table 9). In Antarctica, South America and Africa, species totals increase from the Early 62 to Late Triassic. Species diversity is greatest during the Middle Triassic for Australia, and remains the same for India from the Middle to Late Triassic. Australia has the highest number of reported species for the Early Triassic. South America appears to have the greatest diversity of species for both the Middle and Late Triassic intervals, with up to twenty species identified. For the Middle Triassic, Australia has the second largest number of species, followed by Antarctica, India and Africa, with only three reported identifications. For the Late Triassic, species abundance in South America is followed by

Africa, Australia, Antarctica and India, with five species.

Temporal and spatial distributions of taxa were analyzed using two different similarity indices and by area cladograms. Simpson and Jaccard coefficients were calculated for the five regions, using established formulae (Simpson, 1960; Cheetham and

Hazel, 1969). The results are expressed as percentages of similarity in Tables 10 and 11.

These similarity indices have been tested using several criteria, and were found among the most suitable for quantifying similarity according to the recent analyses of Shi (1993).

Two indices were chosen for comparison and reproducibility.

During the Early Triassic, the Simpson Index indicates that Australia and Africa were 100% similar with regard to Dicroidium floral composition, compared to 67% for

Antarctica (Table 10). The lack of species identifications for the Early Triassic of South

America and India precludes comparison with other regions. Results for the Jaccard

Index, in general, reveal lower percentages of similarity, primarily since the dynamics of the calculation rely on the total numbers of species for both regions. The similarity of

Dicroidium flora between Australia and Africa decreased to 43% using the Jaccard formula, since Australia has greater than twice the number of species (Table 11). The flora of Antarctica and Australia decreased to 25% similarity, and Antarctica and Africa to 50%. 63

During the Middle Triassic, the floras between Antarctica and South America, and

Antarctica and Australia were identical according to the Simpson Index (Table 10). The same results were found for Africa and South America, and Africa and Australia. The similarity decreased to 67% for Australia and South America, and Antarctica and Africa.

Dicroidium species composition in India was most similar to South America at 80%. This similarity declined to 60% when the flora was compared to Australia, 40% with Antarctica, and 33% with Africa. Using the Jaccard Index, the highest similarity was between

Antarctica and Australia at 50% (Table 11). The South American flora was 47% similar to

Antarctica, and 48% similar to Australia. The similarity of the African and Indian floras decreases to 20% or less between all other regions during the Middle Triassic.

During the Late Triassic, the highest similarity of species composition was between

South America and Africa at 93% according to the Simpson Index (Table 10). The similarities between the floras of Antarctica and South America, and Antarctica and Africa were calculated to be 91%. Antarctica and Australia had a lower similarity of 82%.

Australia and South America, and Australia and Africa were both 75% similar. The Indian flora was most similar to South America and Africa at 80%. The greatest differences in floras were between India and Australia. Using the Jaccard Index, these similarity percentages decreased to a high of 67% in the species flora between Antarctica and Africa

(Table 11). The similarity between Antarctica and Australia was 64%, and between

Antarctica and South America was 48%. Africa and South America were 62% similar.

Australia was 53% and 39% similar to Africa and South America, respectively. As for the

Middle Triassic, the Indian flora was different from all other regions.

Another method for comparing the distribution of taxa is to employ parsimony analysis to cluster regions with related sets of taxa. This method is known as parsimony 64 analysis of endemicity (PAE) (Rosen, 1988). Taxa that are endemic to some but not all localities can be categorized as geographical synapomorphies, as in traditional cladistic analyses of taxa. However, in PAE the taxa are actually the characters scored as present or absent in each region, and the relationships of the localities are based on this information.

Any taxon that is common to all localities (i.e. plesiomorphy) or occurs in one locality only

(i.e. autapomorphy) is uninformative to the analysis and deleted from the data matrix. The results of PAE are summarized in cladograms consisting of the most parsimonious arrangement of taxa from the different localities.

For analyzing Dicroidium species distributions, area cladograms were generated for

Early, Middle and Late Triassic using the computer program Phylogenetic Analysis Using

Parsimony (PAUP™) (Swofford, 1993). Data matrixes were developed directly from

Table 9, excluding taxa common to all localities and taxa unique to individual localities

(Appendix B). Both Heuristic and Branch and Bound algorithms were used, and an outgroup of a hypothetical locality that has no taxa was selected (Lundberg rooting). Both algorithms produced the same most parsimonious cladograms for all data. For the Early

Triassic, the shortest single most parsimonious tree was 4 steps (Figure 101). For the

Middle Triassic, two trees were 16 steps (Figure 102). For the Late Triassic, the lengths of the three shortest trees were all 16 steps (Figure 103).

During the Early Triassic, only three regions were used in the analysis, since no species were identified in India and South America. Antarctica is positioned most closely to the outgroup (Figure 101). Australia and Africa are clustered, since they share more taxa between themselves than with Antarctica. This result is comparable to the Simpson

Index similarities for these regions. For the Middle Triassic, two cladograms were generated with the shortest number of steps (Figure 102). The topology of the cladograms 65

is the same, except that India and Africa switch positions towards of the base of the tree.

The top part of the tree remains the same, with South America and Australia nested within

Antarctica. South America and Australia share the most endemic taxa, followed by

Antarctica. These results are similar to those from the Jaccard index. Africa and India

have the lowest similarities, and they emerge at the base of the tree. Antarctica, South

America and Australia have similar percentages, however, Antarctica has fewer unique

shared taxa, so is basal to the other two regions. During the Late Triassic, this

arrangement changes. India is adjacent to the outgroup in all three parsimonious

cladograms (Figure 103). South America and Africa are grouped together in all trees,

since they have the largest number of species in common. Australia and Antarctica are

unresolved, since all three cladograms involve variation with the positions of these regions

relative to South America and Africa. For the most part, these relationships are comparable

to similarity indices for this time period. South America and Africa have the highest

Simpson coefficient and occur with one another in each cladogram. The lower similarity

values are between India and other regions, as reflected in the cladogram. The details of

the relationships among Australia and Antarctica relative to other regions for this time are

less clear.

Discussion

As with any paleobiogeographical analyses, some of the data may be biased in

terms of number and quality of localities, number of publications and research effort per

site, taxon naming, sampling and preservation problems. It is important to consider these

biases when interpreting paleontological data. Absolute and relative measures of fossil record quality are available (Benton, 1994). However, it is difficult to quantitatively 66 correct for the effect of biases, especially when using data from the literature. A method to modify the Simpson Index to account for the relative abundance of taxa has been developed by Thulbom (1986). For species of Dicroidium, the relative abundance of taxa per locality is often not recorded in publication, and even when it is mentioned, it is usually referred to qualitatively. Furthermore, measurements taken in the field and counts of museum specimens are also sensitive to sampling and preservational biases, so it is unclear in practice whether correcting for relative abundance aids in the analysis. Henderson and

Heron (1977) developed a technique to account for the probability that samples reflect population size, however, it requires an initial estimation of total population size. Another factor that biases records of taxa is the concept of form species in paleobotany i.e., sampling usually represents only part of the whole plant. For analyses of Dicroidium, biases were acknowledged when interpreting the data, but no procedures were available to quantitatively correct analyses.

Dicroidium in most Gondwana regions has been well sampled. The best known records of Dicroidium are from several sites in Australia and South America for the Middle and Late Triassic. Africa has fewer sites, but those localities have extremely well preserved Late Triassic specimens (Anderson and Anderson, 1983). Although it is logistically more difficult to collect in Antarctica, this fact is not reflected in the number of localities and publications, since there has been a greater collecting effort in recent years

(Table 2). India has the lowest number of publications and reports, and often poorly preserved specimens (Table 6). This lack of sampling may be influencing the results from paleobiogeographical analyses. However, Dicroidium specimens that have been identified in India are sufficiently different from those found in other regions, as indicated by the number of endemic species. In addition, the Indian flora includes both Gondwana and 67

Northern hemisphere floral elements, and some traditionally Permian taxa have extended

stratigraphic ranges into the Triassic for this area, implying that this is truly a unique floral

region in Gondwana (Lele, 1964; Srivastava, 1992).

The distribution of Dicroidium across Gondwana also reflects regions with

depositional environments favorable for fossilization. During the Early Triassic, the lack

of species diversity may be due to biological, environmental or unrelated preservational

factors. Biological reasons for low diversity in species during the Early Triassic could

include recovery from the Permian-Triassic extinction and/or lack of high speciation rates

near the origin of the genus. Species are recorded for Africa, Australia and Antarctica at

this time. During the Middle to Late Triassic species numbers increase for most regions.

Consequently, there appears to be an increase in Dicroidium species diversity from the

Early to the Late Triassic.

Twenty-seven species have been used in this analysis (Table 9). Morphometric

analysis, identification, and renaming out-dated species in the literature were used to

correct inconsistencies in naming. Whether or not these form species are indeed true

species is not known, but it nevertheless serves as a measure of difference in leaf types

during this time, which is reflective of morphological diversity and adaptation. Dicroidium

odontopteroides, D. feistmantelii, and D. zuberi are ubiquitous in all regions. Several

species ranges extend from the Early to Late Triassic (Table 9). A few species have

regions restricted to particular areas, such as D. argentinum and D. eskensis. Australia,

South America and India have endemic species that only occur in these regions.

Paleobiogeographical patterns of species distributions may be influenced by several

factors, including the paleoenvironment, biological interactions, paleoclimate, historical, and geographical barriers (Newton, 1990). For Dicroidium, endemic species in some 68 regions occurred because the population was restricted by biological or environmental barriers. Geographical barriers probably did not play a major role in the distribution of

Dicroidium within Gondwana, since the continents were part of a continuous land mass.

Competition between taxa and/or paleoclimate may have played a greater role in controlling species distribution. The species that are cosmopolitan tend to exist for longer time spans in the Triassic, implying that they are competitive, unspecialized and adaptable taxa. Large populations have a greater chance at avoiding background extinction, since there is a greater chance that part of the population will not be affected by adverse conditions. In addition, maintaining a large population prevents evolutionary bottlenecks.

In order to determine if patterns in species distribution exist, quantitative analyses were employed. The Simpson and Jaccard indices measure relative similarity. The

Jaccard Index is based on the total number of species in both areas, so the similarity percentages from the Jaccard Index are usually lower than the Simpson Index. It is necessary to use more than one index in order to help avoid biases inherent in the calculations. For the analyses of Dicroidium, clearly the Indian flora is significantly different from the species assemblages in other regions for the Middle and Late Triassic

(Tables 10-11). The relationships among these major regions can be analyzed using PAE

(Rosen, 1988). Area cladograms were generated in this study for the Early, Middle and

Late Triassic in order to determine the relationships among regions in Gondwana based on species composition. The results imply that environmental factors played a greater role than geographical position in influencing species composition, since many areas that were closely related were not necessarily geographically adjacent (Figures 101-103).

The position and extent of the Gondwana land mass influenced the climate.

Gondwana spanned the Southern hemisphere and from the Early to the Late Triassic 69

shifted northward. In general, the interior of Gondwana was dry, and further south and

along the boundaries were regions that experienced seasonally wet periods (Hallam, 1985;

Parrish, 1990). Paleoclimate simulations also illustrate these general trends (Figures 93-

100). Since Dicroidium localities extended to almost 80* S paleolatitude during the Early

Triassic to about 70’ S paleolatitude by the Late Triassic, solar radiation in these regions

was restricted during the winter. Periods of winter darkness influenced the lower winter

temperatures for both the Middle and Late Triassic, when Dicroidium and other plant life

were most likely dormant. During the summer, the growing season further south would

have been relatively short based on lower temperatures. Summer temperatures are higher

for regions further north. Geological evidence indicates that Dicroidium occurred mainly

on lowlands, including floodbanks, and along river beds (Retallack, 1977; Ziegler et al.,

1983).

Although climate simulations for the Early Triassic were not available, geological

information for this time suggests a relatively warm climate in the Gondwana region

(Parrish, 1993). The highest species diversity occurs in Australia, implying that the

seasonally wet and dry climate may have been favorable for species diversity. Three of

those species also occur in Africa, however, species diversity is much lower there.

Perhaps this is because Africa was more arid. The presence of Dicroidium as far south as

Antarctica during the Early Triassic indicates that solar radiation did not play a role in

limiting the distribution of the genus. However, the low species diversity for this region

indicates that other environmental factors may have been limiting.

For the Anisian, those sites with Dicroidium along the coast of South America, eastern Australia, and the central Transantarctic Mountains had precipitation that was relatively high during the summer and temperatures that were relatively low (Figures 93- 70

94). These regions have a large number of taxa in common (Figure 102). This appears to

have been the most favorable climate for this genus, as species diversity is high. India and

Africa were drier and hotter than other regions of Gondwana at this time and these regions

are lower in species diversity. During the winter, Dicroidium was most likely seasonally

deciduous because of the light levels and temperatures during the winter in the higher

paleolatitudes. Biological evidence for a deciduous habit includes the abundance of leaves

found at the base of trunks at several localities and periderm formation beneath leaf bases

(Thomas, 1955; Meyer-Berthaudetal., 1993).

During the Norian, Dicroidium in India was most likely subjected to a hot and wet

climate (Figures 97-98). The number of species remained the same, but there was a

replacement of one taxon. Sites in Africa and eastern Australia had lower precipitation and

a mixture of higher and lower temperatures. Species diversity increased for Africa, but decreased slightly for Australia. Antarctica and South America have areas of high and low

precipitation and temperature. Species number increased for both regions. Africa and

South America share a number of taxa, although their paleoclimates are slightly different

(Figure 103). India has the least number of taxa in common with the other regions, perhaps due to the differences in climate.

As for the range of individual species, regions including Australia, South America and India have many endemic species. Dicroidium odontopteroides, D.feistmantelii, and

D. zuberi are present in all regions. These cosmopolitan species extend from the Early to

Late Triassic, implying that large populations have a more competitive advantage. Subtle morphological differences in species from various regions are recognized. Taxa with bipinnate fronds, such as D. zuberi and D. brownii, are more abundant in the Middle

Triassic. 71

Overall, the results from paleobiogeographical studies indicate that species diversity in Dicroidium increased for most regions from the Early to the Late Triassic, except in

Australia and India. Species assemblages change at different stratigraphic levels within regions, usually with changes in associated taxa. Regions that were geographically close were not necessarily the most similar in species composition, indicating that environmental barriers other than geography may have played a role in the distribution of species. The

Dicroidium flora of India is significantly different from other regions. Correlation between relative temperature levels and precipitation amounts, and species diversity exist. Areas that are predicted to be more humid and seasonally warm had greater species diversity.

Regions that had more taxa in common, generally were similar in their paleoclimates, implying that species distributions were related to environmental factors. CHAPTER VIII

GENERAL DISCUSSION

Morphological Trends Through Time

Classification of Dicroidium species is traditionally based on general pinnule

morphology. A study of the trends in Dicroidium though time and space depends on an

adequate system for identifying species. Regional differences in species, phenotypic

plasticity and normal variation are problematic when identifying species qualitatively. The

results from morphometric analyses suggest that considerable variation in pinnule

morphology within a single specimen exists, and is dependent on the position of the

pinnule on the frond. Variations due to phenotypic plasticity, leaf position on the plant,

fossil quality and human error are negligible when determining significantly different

pinnule shapes. Pinnule shapes above the dichotomy were more successful in resolving clusters of pinnules with similar morphology and therefore should be more useful in classifying species.

The use of morphometric analysis provides a method to quantitatively classify shape and identify species by comparing the pinnule shapes of unidentified specimens to known species. Although high variation in pinnule morphology was detected, eight different Dicroidium species were recognized from Antarctica, Australia and South

America using this technique. These species included D. crassinervis, D. dubium, D. elongatum, D. lancifolium, D. odontopteroides, D. spinifolium, D. stelznerianum and D.

72 73 trilobitum. Using morphometric analysis assists in removing some inconsistencies in naming, but should be used in addition to other quantitative and qualitative characters that are available such as pinnule position and arrangement, rachis width, venation, and cuticular features, some of which are distinct enough to aid in identifying species.

Cuticle from five different species, including D. odontopteroides, D. lancifolium,

D. dubium, D. dutoitii and D. zuberi, was analyzed from Antarctic, Australian and South

American species. For Dicroidium, it appears that the epidermal features useful for species determination are the anticlinal wall margin, subsidiary cell arrangement and number, and the extent of periclinal wall ornamentation. Some of the differences in epidermal features noted for the five species support their separation based on morphological characteristics.

Some of the more variable features, such as extensive papillae and thicker cuticle, may indicate environmental differences. The epidermal features of the same species from two different localities, Australia and Antarctica, are similar. Considerable variation in the stomatal subsidiary cell arrangements was found on individual specimens. For example, in

D. dubium two major types of stomatal complexes were recognized.

In this study, eleven species of Dicroidium were identified from ten localities and several stratigraphic levels in Antarctica. These include the eight species that were resolved with the aid of morphometric analyses, one species with entire laminae, and two bipinnate species. Using this information and species recorded in the literature, the temporal and spatial trends of Dicroidium species from Antarctica were compared to other regions of

Gondwana.

The first unequivocal record of Dicroidium is in Early Triassic rocks of Australia,

Africa and Antarctica. Possible origins of the genus include Supaia and other Permian pteridosperms with dichotomously forked fronds, and other major groups of the Paleozoic 74 from the Northern hemisphere, such asThinnfeldia (Schopf, 1973). From the Late

Permian to the Early Triassic, the Antarctic flora included a mixture of Gbssopteris,

Taenbpteris, and Dicroidium foliage compressions in the Allan Hills, Mt. Bumstead and other localities. Species diversity is low during the Early Triassic. From the late Scythian to the late Norian, Dicroidium foliage was abundant in the Southern hemisphere. Several species ranges extend from the Early to Late Triassic. Sphenophytes, pteridophytes and other gymnosperms also occur with Dicroidium. The last record of Dicroidium is in the upper Late Triassic of several Gondwana localities. By the late Mesozoic, the dominant role that pteridosperms played in the ecosystem was diminished and they were subsequently replaced by other gymnosperms including Bennettitales, conifers and cycads.

At several sites, different Dicroidium species occur at various stratigraphic levels throughout the Triassic. Sampling at several localities illustrates changes in species assemblages through time. These are found at the Sydney Basin in Australia, several sites in South America and Antarctica. Changes in floral assemblages and Dicroidium species at these localities during the Late Triassic may reflect several factors, including changes in the alluvial environment due to flooding, successional changes in the region, seasonality and/or climate changes. Changes in species assemblages usually occur with modifications in associated plant genera, indicating possible environmental changes.

Other morphological trends through time include species diversity increases for most regions of Gondwana from the Early to Late Triassic. During the Early Triassic, the lack of species diversity may be due to biological, environmental or unrelated preservational factors. Biological reasons for this low diversity in species include recovery from the Permian-Triassic extinction and/or lack of high speciation rates associated with the origin of the genus. Species are recorded for Africa, Australia and Antarctica at this 75 time. During the Middle to Late Triassic species numbers increase for most regions.

Consequently, there appears to be an increase in Dicroidium species diversity from the Early to the Late Triassic.

Morphological changes over time within species types include subtle changes in pinnule morphology. Taxa with bipinnate fronds are more abundant in the Middle

Triassic. Some species are spatially and temporally ubiquitous, such as D. odontopteroides and D. feismantelii. Regional differences in species are recognized through time. For example, Dicroidium foliage in Antarctica is generally smaller than other regions, perhaps due to climate or other environmental factors.

Environmental Correlations

Species distributions are determined by interactions with the environment.

Speciation is influenced by heterogeneous environments (Valentine and Moores, 1973).

Broadly adapted, opportunistic species, such as D. odontopteroides, D. feismantelii and D. zuberi, are present in all geographic regions. Several additional species are endemic to particular regions, indicating possible paleoenvironmental or geographical separation.

Dicroidium is found at sites within the lowland regions of Gondwana, along floodplains and riverbanks. Although these areas are favorable depositional environments for plant fossil preservation and thus may indicate bias, the preservation of specimens at some sites implies little transport, providing supporting evidence that these were areas whereDicroidium flourished.

Through the Early to Late Triassic, some of the different regions within Antarctica are similar in species compositions, implying that species populations were not homogeneous across the Antarctic craton, but may have been limited by spatial or 76 environmental barriers. In general, the species from Antarctica are most similar to the

Australian and South American specimens during the Late Triassic. A few species are restricted to particular areas, such as D. argentinum and D. eskensis. Australia, South

America and India have endemic species that only occur in these regions.

Regions that were geographically close were not necessarily the most similar in species composition, indicating that environmental barriers other than geography may have played a role in the distribution of species. The Dicroidium flora of India is significantly different from other regions. Correlations between paleoclimate and species diversity exist. Areas that are predicted to be more humid and seasonally warm had greater species diversity. Regions that had more taxa in common generally were similar in their paleociimates, implying that species distributions were related to environmental factors.

This study illustrates that there are some general correlations of species composition with regional paleoclimate. The interior of Gondwana was generally dry and hot in summer months, other regions were seasonally wet. Although some species of

Dicroidium were cosmopolitan taxa, others may have been specialized for particular climates (e.g. D. remotum and D. sahnii). Although theDicroidium flora occurs on a regional land mass with no major geographic barriers and a relatively uniform paleoclimate compared to today, the present study suggests that differences in species assemblages do exist between areas, and implies that Gondwana biotas may be more heterogeneous than previous interpretations.

Dicroidium from Antarctica, eastern Australia, and southern South America was at high paleolatitudes. The species diversity of these regions was not significantly less than other regions of Gondwana. The availability of light was more limited in these areas, but its affect was not reflected in species composition. Perhaps the limiting factor was 77 temperature or other aspects of the climate rather than solar radiation, as in polar tree growth today, corroborating that temperatures must have been on average warmer during the Triassic.

This study has provided several new findings regarding Dicroidium and the application of new techniques to the field of paleobotany. It is the first application of morphometric techniques in classifying fossil foliage. This technique successfully separated different pinnule shapes from Antarctica, Australia, and South American specimens corresponding to eight species. Antarctic fossil material from several new localities and additional stratigraphic levels has been described. Some of these specimens had well-preserved cuticle, providing material for the first description of Dicroidium cuticle from Antarctic compression specimens. In addition, cuticle from other species and localities were described and compared to the Antarctic specimens. The study of

Dicroidium from Antarctica provided the basis for a detailed summary of all species found in Gondwana to date. Paleobiogeographical analyses of the species revealed trends in temporal and spatial diversity. Analyses of species compositions in different regions resulted in revealing biogeographical trends between regions, and positive correlations of species diversity with paleoclimate. In addition, this study provides the first base level study of high latitude pteridosperm foliage that can be compared with lower latitude floras, possibly providing additional insight into the evolutionary adaptation of high versus low latitude floras.

Potential future studies of Dicroidium include further study of the cuticle, focusing on comparing cuticular features from different species and positions on the foliage. In addition, analyzing stomatal density through time may provide evidence to corroborate environmental changes. In terms of changes in the genus and species assemblages, more 78 detailed macroevolutionary studies need to focus on thorough sampling and taxonomic consistency from several stratigraphic levels at one site. These results can then be compared to data from additional sites in different regions.

Examining the spatial and temporal dynamics of organisms provides information on how ecosystems and individual taxa respond to and recover from major crises, including environmental changes, over long time periods. The knowledge generated through these studies may provide the clues to understanding how biotas will manage in the future. LITERATURE CITED

Amtsberg, H. 1969. A contribution to the mesophytic flora of south Australia (Springfield and Leigh Creek). Transactions of the Royal Society of South Australia 93:79-86, pis. 1-4.

Anderson, H.M. 1990. The gymnosperms of the Triassic austral (Gondwana) flora, pp. 17-25. In: Proceedings of the 3rd IOP Conference, (Melbourne, 1988).

and J.M. Anderson. 1970. A preliminary review of the biostratigraphy of the uppermost Permian, Triassic and lowermost Jurassic of Gondwanaland. Palaeontologia Africana 13:1-22.

Anderson, J.M. and H.M. Anderson. 1983. Palaeoflora of Southern Africa. Molteno Formation (Triassic). Vol. I (Part 1, Introduction; Part 2, Dicroidium). A.A. Balkema, Rotterdam. 227 pp.

and . 1984. The fossil content of the Upper Triassic Molteno Formation, South Africa. Palaeontologia Africana 25:39-59.

Andersson, S. 1991. Geographical variation and genetic analysis of leaf shape in Crepis tectorum (Asteraceae). Plant Systematics and Evolution 178: 247-258.

Antevs, E. 1914. Die Gattungen Thinnfeldia Ett. und Dicroidium Goth. Kungl. Svenska Vetenskapsakademiens Handingar. 51(6): 1-71.

Arber, E.A.N. 1917. The earlier Mesozoic floras of New Zealand. New Zealand Geological Survey, Palaeontological Bulletin 6: 1-80, pis. I-XIII.

Archangelsky, S. 1968. Studies on Triassic fossil plants from Argentia. IV. The leaf genus Dicroidium and its possible relation to Rhexoxylon stems. Palaeontology 11(4): 500- 512.

and D.W. Brett. 1961. Studies on Triassic fossil plants form Argentina. I. Rhexoxylon from the Ischigualasto Formation. Philosophical Transactions of the Royal Society of London 244B: 1-19.

Arrondo, O.G. 1972. Estudio geologico y paleontologico en la zona de la estancia la juanita y alrededores, Provincia de Santa Cruz, Argentina. Revista del Museo de la Plata, Paleontologfa (n.s.), Tomo 7(43): 1-194. L&ninas I-VIII.

79 80

, E.M. Morel, D.G. Ganuza, F. Hervd, M.A. Parada, and W. Muehlebach. 1988. La formacion Panguipulli y su contenido paleofloristico Chile. V Congreso Geologico Chileno (Departmento de Geologia y Geofisica, Universidad de Chile) Tomo I: 137-145.

Artabe, A.E. 1985. Estudio sistematico de la tafoflora Triasica de Los Menucos, Provincia de Rio Negro, Argentina. Parte I. Sphenophyta, Filicophyta, Pteridospermophyta. Ameghiniana 22(1-2): 3-22.

. 1990. Revalidacion del generoTriasico Zuberia (Frenguelli), familia Corystospermaceae. Revista del Museo de la Plata, Paleontologia, 9: 145-57.

Askin, R. A., Barrett, P.J., Kohn, B.P., McPherson, J.G. 1971. Stratigraphic sections of the Beacon Supergroup (Devonian and older (?) to Jurassic) in South Victoria Land. Antarctic Data Series No. 2: 1-88.

Azcdrate, V.M. and A.R. Fasola. 1970. Sobre formas nuevas para la flora tridsica de Los Molles. Boletin Museo Nacional Historia Natural 29(14): 249-269.

Baksi, S.K. 1967. Fossil plants from Raghavapuram mudstone, West Godavari District, A.P., India. The Palaeobotanist 16(3): 206-215, pis. 1-3.

Baldoni, A.M. 1980. Revision de las especies del genero Xylopteris (Corystospermaceae) en el Triasico de Argentina, Australia y Sudafrica. Ameghiniana 17(2): 135-155.

Ballance, P.F. 1977. The Beacon Supergroup in the Allan Hills, central Victoria Land, Antarctica. New Zealand Journal of Geology and Geophysics 20(6): 1003-1016.

Balme, B.E. and R.J. Helby. 1973. Floral modifications at the Permian-Triassic boundary in Australia, pp. 433-444. In: A. Logan and L.V. Hills (eds.), The Permian and Triassic Systems and their Mutual Boundary. Canadian Society of Petroleum Geologists, Calgary.

Baneiji, J., and M.N. Bose. 1977. Some Lower Triassic plant remains from Asansol region, India. The Palaeobotanist 24(3): 202-210, pis. 1-2.

and Y. Lemoigne. 1987. Significant additions to the Upper Triassic flora of Williams Point, Livingston Island, South Shetlands (Antarctica). Geobios 20 (4): 469-87.

Barrett, P.J. 1969. Stratigraphy and petrology of the mainly fluviatile Permian and Triassic Beacon rocks, Beardmore Glacier area, Antarctica. Institute of Polar Studies, Ohio State University, Columbus. Report 34. 132 pp.

. 1991. The Devonian to Jurassic Beacon Supergroup of the Transan tare tic Mountains and correlatives in other parts of Antarctica, pp. 120-152. In: R. J. Tingley (ed.), The Geology of Antarctica. Oxford Monographs on Geology and Geophysics, 17. Clarendon Press, New York. 81

, D.H. Elliot, and J.F. Lindsay. 1986. The Beacon Supergroup (Devonian-Triassic) and Ferrar Group (Jurassic) in the Bear dm ore Glacier area, Antarctica, pp. 339-428. In: M.D. Turner and J.F. Splettstoesser (eds.), Geology of the central Transantarctic Mountains. Antarctic Research Series, Vol. 36. American Geophysical Union, Washington, D.C.

, B.P. Kohn, R.A. Askin, and J.G. McPherson. 1971. Preliminary report on Beacon Supergroup studies between the Hatherton and Mackay Glaciers, Antarctica. New Zealand Journal of Geology and Geophysics. 14(3): 605-614.

, Kyle, R.A., Bamford, D.A., Bright, D.V., Chinn, A.N. 1973. Stratigraphic sections of the Beacon Supergroup (Devonian and older (?) to Jurassic) in South Victoria Land. Part III. Stratigraphic sections from the Shapeless Mountain area and around the Skelton Neve. Antarctic Data Series No. 3: 77-165.

Barthlott, W. 1981. Epidermal and seed surface characters of plants: systematic applicability and some evolutionary aspects. Nord. J. Bot. 1: 345-355.

Benton, M. J. 1994. Palaeontological data and identifying mass extinctions. Trends in Ecology and Evolution 9(5): 181-185.

Besairie, H. 1972. Geologie de Madagascar. Annales Gdologiques de Madagascar 35: 1-463.

Bigarella, J.J. and R. Salamuni. 1967. Some palaeogeographic and palaeotectonic features of the Parana Basin, pp. 235-301. In: Bigarella, J.J. (ed.), Problems in Brazilian Gondwana Geology, Curitiba.

Bonaparte, Jos6 F. 1982. Faunal replacement in the Triassic of South America. Journal of Vertebrate Paleontology 2(3): 362-371.

Bond, G. 1965. Some new fossil localities in the Karroo System of Rhodesia. Amoldia 2(11): 1-4.

Bonetti, M.I.R. 1964. Breve noticia sobre los conocimentos actuales de la flora tridsica de El Tranquilo, Provinciade Santa Cruz. Resumen. Asociacidn Paleontoldgica Argentina, Ameghiniana 3 (6): 182. i . 1966. Consideraciones sobre algunos representantes de la familia "Corystospermaceae". Ameghiniana Tomo 4(10): 389-395.

. 1968. Las especies del genero Pseudoctenis en la flora Triasica de Barreal (San Juan). Ameghiniana 5(10): 433-446.

Bookstein, F. L., Strauss, R. E., Humphries, J. M., Chemoff, B., Elder, R. L., Smith, G. R. 1982. A comment upon the uses of Fourier methods in systematics. Systematic Zoology 31(1): 85-92. 82

Bortoluzzi, C.A., M. Guerra-Sommer, and M. Cazzulo-Klepzig. 1983. Tafoflora tridssica da Forma^do Santa Maria, Rio Grande do Sul, Brasil-II. Representantes de Pteridospermopsida e Pteridophylla. pp. 105-116. In: Paleobotanica e Palinologia na America do Sul: Contribuicoes a IV Reuniao de Paleobotdnicos e Palin61ogos Instituto de Geocidncias Universidade de Sao Paulo, 7-10 Dezembro de 1983.

Bose, M.N., and S.C. Srivastava. 1971. The genus Dicroidium from the Triassic of Nidpur, Madhya Pradesh, India. The Palaeobotanist 19:41-51, pis. 1-4.

, E.L. Taylor, and T.N. Taylor. 1990. Gondwana floras of India and Antarctica: a survey and appraisal, pp. 118-148. In: Taylor, T. N. and Taylor, E. L. (eds.), Antarctic Paleobiology: Its Role in the Reconstruction of Gondwana. Springer-Verlag, New York. 261 p.

Boucher, L.D., E.L. Taylor, and T.N. Taylor. 1993. Dicroidium from the Triassic of Antarctica, pp. 39-46. In: S.G. Lucas and M. Morales (eds.), The Nonmarine Triassic. New Mexico Museum of Natural History and Science Bulletin No. 3.

Boureau, E. 1949. Paleophyogeographie de Madagascar. Mdmoires de l'lnstitut Scientifique de Madagascar sdr. D, 1(2): 81-96.

Bourke, D.J., R.E. Gould, R. Helby, R. Morgan, and G.J. Retallack. 1977. Floral evidence for a Middle Triassic age of the Gunnee Beds and Gragin conglomerate, near Delungra, New South Wales. Journal of the Royal Society of New South Wales 110: 33-40.

Brenon, P. 1972. The geology of Madagascar, pp. 27-86. In: Battistini, R. and Richard- Vindard, G. (eds.), Biogeography and Ecology in Madagascar. Dr. W. Junk B.V Publishers, The Hague.

Burnham, R.J. 1993. Reconstructing richness in the plant fossil record. Palaios 8: 376- 384.

Carpentier, A. 1935. fitudes paliobotaniques sur le Groupe de la Sakoa et le Groupe de la Sakamena (Madagascar). Annales Gdologiques du Service des Mines (Madagascar) 5: 1-32, pis. I-VI.

. 1936. Additions a l'dtude de la flore du Groupe de la Sakamena (Madagascar). Annales Gdologique du Service des Mines (Madagascar) 6:35-40, pi. V.

Carruthers, W. 1872. Notes on fossil plants from Queensland, Australia. Quarterly Journal of the Geological Society of London 28:350-359.

Cecioni, G. and G.E.G. Westermann. 1968. The Triassic/Jurassic marine transition of coastal central Chile. Pacific Geology 1:41-75.

Chapin, F.S. and C. Komer. 1994. Arctic and alpine biodiversity: patterns, causes and ecosystem consequences. Trends in Ecology and Evolution 9(2): 45-47. 83

Cheetham, A.H. and J.E. Hazel. 1969. Binaiy (presence-absence) similarity coefficients. Journal of Palaeontology 43:1130-1136.

Collinson, J.W., and D.H. Elliot 1984a. Triassic stratigraphy of the Shackleton Glacier Area. pp. 103-117. In: M.D. Turner and J.F. Splettstoesser (eds.), Geology of the central Transantarctic Mountains. Antarctic Research Series, Vol. 36. American Geophysical Union, Washington, D.C.

and . 1984b. Geology of Coal sack Bluff, Antarctica, pp. 97-102. In: M.D. Turner and J.F. Splettstoesser (eds.), Geology of the central Transantarctic Mountains. Antarctic Research Series, Vol. 36. American Geophysical Union, Washington, D.C.

and J.L. Isbell. 1986. Permian-Triassic sedimentology of the Beardmore Glacier region. Antarctic Journal of the U.S. 21(5): 29-30.

, D.C. Pennington, and N.R. Kemp. 1983. Sedimentary petrology of Permian- Triassic fluvial rocks in Allan Hills, central Victoria Land. Antarctic Journal of the U.S. 18(5): 20-22.

, _____, and . 1986. Stratigraphy and petrology of Permian and Triassic fluvial deposits in northern Victoria Land, Antarctica, pp. 211-242. In: Geological Investigations in Northern Victoria Land. Antarctic Research Series, Vol. 46. American Geophysical Union, Washington, D.C.

, N.R. Kemp, and J.T. Eggert. 1987. Comparison of the Triassic Gondwana sequences in the Transantarctic Mountains and Tasmania. Geophysical Monograph 41: 51-61. G.D. McKenzie (ed.), Gondwana Six: Stratigraphy, Sedimentology, and Paleontology, American Geophysical Union, Washington, D.C.

Crane, P.R. 1985. Phylogenetic analysis of seed plants and the origin of angiosperms. Annals of the Missouri Botanical Garden 72:716-793.

Creber, G.T. 1990. The south polar forest ecosystem, pp. 37-41. In: T.N. Taylor and E.L. Taylor (eds.), Antarctic Paleobiology: its role in the reconstruction of Gondwana. Springer-Verlag, New York.

and W.G. Chaloner. 1985. Tree growth in the Mesozoic and early Tertiary and the reconstruction of palaeoclimates. Palaeogeography, Palaeoclimatology, Palaeoecology 52: 35-60.

Crowley, T.J. and G.R. North. 1991. Paleoclimatology. Oxford University Press, New York. 339 pp.

Dickins, J.M. 1973. The geological sequence and the Permian-Triassic boundary in Australia and eastern New Guinea, pp. 425-432. In: In: A. Logan and L.V. Hills (eds.), The Permian and Triassic Systems and their Mutual Boundary. Canadian Society of Petroleum Geologists, Calgary. 84

DiMichele, W.A. 1994. Ecological patterns in time and space. Paleobiology 20(2): 89-92.

, T.L. Phillips, and R.G. Olmstead. 1987. Opportunistic evolution: abiotic environmental stress and the fossil record of plants. Review of Palaeobotany and Palynology 50: 151-178. Dingle, R.V., W.G. Siesser, and A.R. Newton. 1983. Mesozoic and Tertiary Geology of Southern Africa. A.A. Balkema, Rotterdam. 375 p.

Dodd, J.R. and R.J. Stanton, Jr. 1990. Paleoecology: concepts and applications. (2nd ed.) John Wiley and Sons, Inc., New York.

Donn W.L. 1987. Terrestrial climate change from the Triassic to recent, pp. 343-352. In: M.R. Rampino, J.E. Sanders, W.S. Newnan and L.K. Konigsson (eds.), Climate: history, periodicity, and predictability. Van Nostrand Reinhold Co., New York.

Doyle, J. A. and M. J. Donoghue. 1992. Fossils and seed plant phytogeny reanalyzed. Brittonia44(2): 89-106.

Du Toit, A.L. 1927. The fossil flora of the Upper Karroo Beds. Annals of the South African Museum XXII: 289-420, pis. 16-32.

Ehrlich, R. and W.E Full. 1986. Comments on "Relationships among eigenshape analysis, Fourier analysis, and analysis of coordinates" by F. James Rohlf. Mathematical Geology 18(8): 855-857.

, Pharr, R. Baxter, Jr., and N. Healy-Williams. 1983. Comments on the validity of Fourier descriptors in systematics: A reply to Bookstein et al. Systematic Zoology 32(2): 202-206.

Farabee, M.J., T.N. Taylor, and E.L. Taylor. 1989. Pollen and spore assemblages from the Falla Formation (Upper Triassic), central Transantarctic Mountains, Antarctica. Review of Palaeobotany and Palynology 61: 101-138.

Faure, G. and R.L. Hill. 1973. Age of the Falla Formation (Triassic), Queen Alexandra Range. Antarctic Journal of die U.S. 8(5): 264-266.

Feistmantel, O. 1878. Palaeozoische and mesozoische flora des ostlische Australien. Pal aeon tographi ca Cassel, Suppl. 3(3): 55-130.

Ferson, S., F.J. Rohlf, and R.K. Koehn. 1985. Measuring shape variation of two- dimensional outlines. Systematic Zoology 34(1): 59-68.

Flint, J.C.E. and R.E. Gould. 1975. A note on the fossil megafloras of the Nymboida and Red Cliff Coal Measures, southern Clarence-Moreton Basin, N.S.W. Journal and Proceedings of the Royal Society of New South Wales 108: 70-74, pis. I-III. 85

Forsyth, S.M. 1989. Upper Parmeener Supergroup, pp. 309-333. In: C.F. Burrett and E.L. Martin (eds.), Geology and Mineral Resources of Tasmania: a Bicentennial Volume. Geological Society of Australia, Brisbane.

Frenguelli, J. 1937. La florula Jurdsica de Paso Flores en el Neuqudn, con referenda a la de Piedra Pintada y otras floras Jurdsicas Argentinas. Revista del Museo de La Plata (Nueva Series) Paleontologia. Tomo I (3): 67-108.

. 1941. Dicroidiumstelzjnerianum (Gein.) n. comb. Notas del Museo de La Plata (Paleontologia) 6(33): 393-403.

. 1943. Reseiia critica de los gdneros atribufdos a la 'Serie de Thinnfeldia'. Revista del Museo de La Plata (n.ser.) 2: 225-342.

Gabites, H.I. 1985. Triassic paleoecology of the Lashly Formation, Transantarctic Mountains, Antarctica. Victoria University of Wellington, M.Sc. thesis (unpublished), 148 pp.

Geinitz, H.B. 1876. Ueber rhatischen Pflanzen und Thierreste in den Argentinischen Provinzen. La Rioja, San Juan, und Mendoza. Palaeontographica (Suppl.) 3: 1-14.

Givnish, T. 1979. On the adaptive significance of leaf form. pp. 375-407. In: O.T. Solbrig, S. Jain, G.B. Johnson, and P.H. Raven (eds.), Topics in Plant Population Biology. Columbia University Press, New York.

. 1987. Comparative studies of leaf form: assessing the relative roles of selective pressures and phylogenetic constraints. New Phytologist 106 (Suppl.): 131-160.

Gordon, M., and R.W. Brown. 1952. Plantas tridssicas do sul do Brasil (1). Notas Preliminares e Estudos (Rio de Janeiro) no. 54:1-7, Plate I.

Gothan, W. 1912. Uber die Gattung Thinnfeldia Ettingshausen. Abhandlungen der Naturhistorischen Gesellschaft zu Nlimberg 19(3): 67-80. Tafel 13-16.

Graham, R.W. and E.C. Grimm. 1990. Effects of global climate change on the patterns of terrestrial biological communities. Trends in Ecology and Evolution 5(9): 289-292.

Gunn, B.M., and G. Warren. 1962. Geology of Victoria Land between the Mawson and Mulock Glaciers, Antarctica. New Zealand Geological Survey Bulletin n.s. 71: 1-157.

Hallam, A. 1985. A review of Mesozoic climates. Journal of the Geological Society of London 142: 433-445.

Harris, T.M. 1964. The Yorkshire Jurassic Flora. II. Caytoniales, Cycadales and Pteridosperms. British Museum of Natural History, London. 191 pp.

Henderson, R. A. and M. L. Heron. 1977. A probabilistic method of paleobiogeographic analysis. Lethaia 10: 1-15. 86

Herbst, R. and A.I. Lutz. 1988. Rhexoxylon brasiliensis n. sp. (Corystospermaceae, Pteridospermales) from the Upper Triassic Caturrita Formation, Brazil, with comments on biology and environment. Mededelingen Rijks Geologische Dienst42: 21-30, 1 pi.

Hill, D., G. Playford, and J.T. Woods (eds.). 1965. Triassic fossils of Queensland. Queensland Palaeontographical Society, Brisbane. 32 p.

Holmes, W.B.K. 1982. The Middle Triassic flora from Benolong, near Dubbo, central- western New South Wales. Alcheringa 6: 1-33.

and S.R. Ash. 1979. An Early Triassic megafossil flora from the Lome Basin, New South Wales. Proceedings of the Linnean Society of New South Wales 103(1): 47-70.

Isbell, John L., T.N. Taylor, E.L. Taylor, N.R. Cuneo, and B. Meyer-Berthaud. 1990. Depositional setting of Permian and Triassic fossil plants in the Allan Hills, southern Victoria Land. Antarctic Journal of the United States 25(5): 22.

Jablonski, D., K.W. Flessa, and J.W. Valentine. 1985. Biogeography and paleobiology. Paleobiology 11(1): 75-90.

Jacob, K., and C. Jacob. 1950. A preliminary account of the structure of the cuticles of Dicroidium (Thinnfeldia) fronds from the Mesozoic of Australia. Proceedings of the National Institute of Science of India, Calcutta, Vol. 16(2): 101-126.

Jain, R.K., and T. Delevoryas. 1967. A Middle Triassic flora from the Cacheuta Formation, Minas de Petroleo, Argentina. Palaeontology 10(4): 564-589.

Jefferson, T.H. and T.N.Taylor. 1983. Permian and Triassic woods from the Transantarctic Mountains: paleoenvironmental indicators. Antarctic Journal of the U.S. 18(5): 55-57.

Johnston, R.M. 1886. General observations regarding the classification of the upper Palaeozoic and Mesozoic rocks of Tasmania, together with a full description of all the known Tasmanian coal plants, including a considerable number of new species. Papers of the Proceedings of the Royal Society of Tasmania (1885): 343-387.

. 1887. Fresh contributions to our knowledge of the plants of Mesozoic age in Tasmania Papers of the Proceedings of the Royal Society of Tasmania (1886): 160- 182.

. 1894. Further contributions to the fossil flora of Tasmania Part 1. Papers of the Proceedings of the Royal Society of Tasmania (1893): 170-178.

Jones, H.G. 1985. Adaptive significance of leaf development and structural responses to environment, pp. 155-173. In: N.R. Baker, W.J. Davies, and C.K. Ong (eds.), Control of Leaf Growth. Cambridge University Press, Cambridge. 87

Jones, O.A., and N.J. de Jersey. 1947. The flora of the Ipswich Coal Measures- morphology and floral succession. University of Queensland Papers, Vol. 3 (3): 1-88, plates I-X.

Jordan, G.J. and R.S. Hill. 1994. Past and present variability in leaf length of evergreen members of Nothofagus subgenus Lophozonia related to ecology and population dynamics. New Phytologist 127: 377-390.

Kerp, H. 1990. The study of fossil gymnosperms by means of cuticular analysis. Palaios 5: 548-569.

Kitchell, J.A. 1985. Evolutionary paleoecology: recent contributions to evolutionary theory. Paleobiology 11(1): 91-104.

Kitching, James W., J.W. Collinson, D.H. Elliot, and E.H. Colbert. 1972. Lystrosaurus zone (Triassic) fauna from Antarctica. Science 175:524-527.

Kuhl, F.P. and C.R. Giardina. 1982. Elliptic Fourier features of a closed contour. Computer Graphics and Image Processing 18: 236-258.

Kurtz, F. 1921. Atlas de plantas fdsiles de la Reptiblica Argentina Actas de la Academia National de Ciencias (Cdrdoba) 7: 129-158.

Kutzbach, J.E. and R.G. Gallimore. 1989. Pangaean climates: megamonsoons of the megacontinent. Journal of Geophysical Research 94(D3): 3341-3357.

Kyle, R.A. and J.M. Schopf. 1982. Permian and Triassic palynostratigraphy of the Victoria Group, Transantarctic Mountains, pp. 649-659. In: Craddock, C. (ed.), Antarctic Geoscience-symposium on Antarctic geology and geophysics (Madison, Wisconsin; August 22-27,1977). University of Wisconsin Press, Madison.

Lacey, W.S. 1961. Studies in the Karroo floras of Rhodesia and Nyasaland. Part 1. A geological account of the plant-bearing deposits. Proceedings and Transactions of the Rhodesia Scientific Association 49(1): 26-53.

. 1970. Some new records of fossil plants in the Molteno Stage of Rhodesia Amoldia (Rhodesia) 5(4): 1-4.

. 1974. Some new African Gondwana plants. Birbal Sahnii Institute of Palaeobotany Special Publication 2:34-41, pi. 1.

. 1976. Further observations on the Molteno flora of Rhodesia Amoldia (Rhodesia) 7(36): 1-14.

and C.S. Smith. 1972. Karroo floras from the Upper Luangwa Valley, Zambia 88

pp. 571-574. In: S.H. Haughton (ed.), I.U.G.S. Proceedings and Papers of the Second Gondwana Symposium, (South Africa, 1970). Council for Scientific and Industrial Research, Scientia, Pretoria, South Africa.

and R.C. Lucas. 1981. The Triassic flora of Livingston Island, South Shetland Islands. British Antarctic Survey Bulletin, No. 53: 157-173.

and . 1984. Fossil plants from the Luangwa and Luano valleys of Zambia and their bearing on age determination. Evolutionary Botany and Biostratigraphy: 451-462.

Lakhanpal, R.N., H.K. Maheshwari, and N. Awasthi. 1976. A catalogue of Indian fossil plants. Birbal Sahni Institute of Palaeobotany, Lucknow. 318 p.

Lejal-Nicol, A. 1979. Sur la flore fossile de la Formation Unar dans le Djebel Ben Ghnema (Libye). 104th Congrfcs national des Socidtds savantes, sciences, 1: 103-119.

Lele, K. M. 1956. Plant fossils from Parsora in the South Rewa Gondwana Basin, India. The Palaeobotanist 4: 23-34, pis. 1-2.

. 1962. Studies in the Indian Middle Gondwana flora: 1. On Dicroidium from the South Rewa Gondwana Basin. The Palaeobotanist 10 (1-2): 48-68, pis. 1-4.

. 1964. The problem of Middle Gondwana in India. Proceedings of the 22nd International Geological Congress, New Delhi, Section 9 (Gondwanas): 181-202.

. 1976a. Late Palaeozoic and Triassic floras of India and their relation to the floras of northern and southern hemispheres. The Palaeobotanist 23(2): 89-115.

. 1976b. Paleoclimatic implications of Gondwana floras. Geophytology 6(2): 207- 229.

Lemoigne, Yves. 1987. Confirmation de 1'existence d'une flore triasique dans l'ile Livingston des Shetland du sud (Ouest Antarctique). Comptes Rendus Acaddmie Science (Paris) s

Lohmann, G. P. 1983. Eigenshape analysis of microfossils: A general morphometric procedure for describing changes in shape. Mathematical Geology 15(6): 659-672.

Ldpez-Gamundf, O.R., I.S. Espejo, P.J. Conaghan, and C.M. Powell. 1994. Southern South America, pp. 281-329. In: Veevers, J.J. and C.M. Powell (eds.), Permian- Triassic Pangean Basins and Foldbelts along the Panthalassan Margin of Gondwanaland. Geological Society of America Memoir No. 184, Boulder, CO.

Maheshwari, H. K. 1977. Floristics of the Permian and Triassic Gondwanas of India. The Palaeobotanist 23(3): 145-160.

. 1992. Provincialism in Gondwana floras. The Palaeobotanist 40: 101-127. 89

Mawson, D. 1940. Sedimentary Rocks. Australian Antarctic Expedition 1911-14: Scientific Reports, Series A, Vol. IV, Geology, Part 2: 348-367. 5 pis.

McLellan, T. 1993. The roles of heterochrony and heteroblasty in the diversification of leaf shapes in Begonia dregei (Begoniaceae). American Journal of Botany 80(7): 796-804.

Menlndez, C.A. 1951. La flora Mesozoica de la formacion Llantenes (Provincia de Mendoza). Revista del Instituto Nacional de investigacion de las ciencias naturales. Museo Argentino de Ciencias Naturales Bernardino Rivadavia 2(3): 147-261, pis. 1- 15.

. 1968. Datos palinologicos de las floras preterciarias de la Argentina. Revista del Museo Argentino de Ciencias Naturales Bernardino Rivadavia. Paleontologia 1 (7): 231-242.

Meyer-Berthaud, B., E.L. Taylor, and T.N. Taylor. 1992. Reconstructing the Gondwana seed fern Dicroidium: evidence from the Triassic of Antarctica. Geobios 25(3): 341- 344.

, T.N. Taylor, and E.L. Taylor. 1993. Petrified stems bearing Dicroidium leaves from the Triassic of Antarctica Palaeontology 36(2): 337-356,4 pis.

Miller, W., III. 1990. Hierarchy, individuality and paleoecosystems. pp. 31-47. In: Miller, W., Ill (ed.), Paleocommunity Temporal Dynamics: the long-term development of multispecies assemblies. The Paleontological Society Special Publication No. 5. Lawrence, Kansas.

Mohr, B. and F. Schoner. 1985. Eine obertriassische ~ Dicroidium-Flora siidostlich Alto del Carmen, Regidn de Atacama (Chile). N. Jb. Geol. Palaont. Mh. 1885, H. 6: 368- 379.

Morel, E.M. 1994. El Triasico del Cerro Cacheuta, Mendoza (Argentina). Parte I: Geologia, contenido paleofloristico y cronoestratigrafia. Ameghiniana31(2): 161-176.

, L.A. Spalletti, O.G. Arrondo, and D. G. Ganuza. 1992. Los estratos plantiferos de la formacion Paso Flores. Triasico superior de las Lomas y Canadon de Ranquel Huao. Provincia del Neuquen, Argentina. Revista del Museo de la Plata (Nueva Series) Paleontologia. Tomo IX(58): 199-221.

Morris, J. 1845. Fossil flora, pp. 245-254. In: P. E. De Strezelecki (ed.), Physical Description of New South Wales and Van Diemen's Land. Longman, Brown, Green and Longmans, London.

Newton, C. R. 1990. Palaeobiogeography. pp. 452-460. In: D. E. G. Briggs and P.R. Crowther (eds.), Palaeobiology: A Synthesis. Blackwell Scientific Publishers, Oxford. 90

Nishida, M. 1970. On some fossil plants from Chile, South America. Annual Report of the Foreign Students' College of Chiba University. 5: 12-18.

Orlando, H.A. 1968. A new Triassic flora from Livingston Island, South Shetland Islands. British Antarctic Survey. Bulletin 16: 1-13.

Pal, P.K. 1984. Triassic plant megafossils from the Tiki Formation, South Rewa Gondwana Basin, India. The Palaeobotanist 32(3): 253-309.

Parrish, J.T. 1990. Gondwanan paleogeography and paleoclimatology. pp. 15-26. In: T.N. Taylor and E.L. Taylor (eds.), Antarctic Paleobiology: its role in the reconstruction of Gondwana. Springer-Verlag, New York.

. 1993. Climate of the supercontinent Pangea. The Journal of Geology 101: 215- 233.

Petriella, B. 1978. La reconstruccion de Dicroidium (Pteridospermopsida, Corystospermaceae). Obra del Centenario del Museo de La Plata 5: 107-110.

. 1979. Sinopsis de las Corystospermaceae (Corystospermales, Pteridospermophyta) de Argentina. I. Hojas. Ameghiniana 16(1-2): 81-102.

. 1981. Sistematica y vinculaciones de las Corystospermaceae H. Thomas. Ameghiniana 18(3-4): 221-234.

Pigg, K.B. 1990. Anatomically preserved Dicroidium foliage from the central Transantarctic Mountains. Review of Palaeobotany and Palynology 66: 129-145.

Pinet, P.R. 1965. Petrology of the Upper Division of the Beacon Sandstone, Dry Valley Region, Southern Victoria Land, Antarctica. M.Sc. thesis, University of Massachusetts. 113 pp.

Playford, G. and J.F. Rigby. 1988. Some Lower Mesozoic and Lower Tertiary plant megafossil occurrences in the Ipswich area, southeast Queensland, pp. 119-125. In: L.H. Hamilton (ed.), Field Excursions Handbook for the Ninth Australian Geological Convention. Geological Society of Australia, Queensland Division, Brisbane.

, _____, and D.C. Archibald. 1982. A Middle Triassic flora from the Moolayember Formation, Bowen Basin, Queensland. Geological Survey of Queensland, Publication 380: 1-52.

Plumstead, E.P. 1962. Fossil floras of Antarctica. Trans-Antarctic Expedition 1955-1958: Scientific Reports 9: 1-154.

Pole, M.S. and J.I. Raine. 1994. Triassic plant fossils from Pollock Road, Southland, New Zealand. Alcheringa 18: 147-159. 91

Rao, A.R. and K.M. Lele. 1963. On the cuticle of Dicroidium (Thinnfeldia) safutii (Seward) with some observations on the genera Thinnfeldia and Dicroidium. The Palaeobotanist 11(1-2): 7-12, pi. 1.

Rasmussen, H. 1981. Terminology and classification of stomata and stomatal development—a critical survey. Botanical Journal of the Linnean Society 83: 199-212.

Raup, D.M. 1972. Approaches to morphologic analysis, pp. 28-44. In: T.J.M. Schopf (ed.), Models in Paleobiology. Freeman, Cooper and Co., San Francisco, CA.

Ravich, M.G. L.V. Klimov, and D.S. Solov'ev. 1968. The Pre-Cambrian of East Antarctica. Israel Program for Scientific Translations: Jerusalem. 475 pp.

Rees, P.M. and J.L. Smellie. 1989. Cretaceous angiosperms from an allegedly Triassic flora at Williams Point, Livingston Is., South Shetland Islands. Antarctic Science 1(3): 239-48.

Reihman, M.A. and J.T. Schabilion. 1978. Petrified neuropterid foliage from a Middle Pennsylvanian coal ball. American Journal of Botany 65(8): 834-844.

Retallack, G.J. 1977. Reconstructing Triassic vegetation of eastern Australasia: a new approach for the biostratigraphy of Gondwanaland. Alcheringa 1,247-277. Microfiche supplement: Alcheringa-fiche 1.G1-J17.

. 1979. Middle Triassic coastal outwash plain deposits in Tank Gully, Canterbury, New Zealand. Journal of the Royal Society of New Zealand 9(4): 397-414.

. 1980a. Late Carboniferous to Middle Triassic megafossil floras from the Sydney Basin, pp. 385-429. In: Chris Herbert and Robin Helby (eds.), A Guide to the Sydney Basin. Department of Mineral Resources, Geological Survey of New South Wales, Bulletin No. 26.

. 1980b. Middle Triassic megafossil plants and trace fossils from Tank Gully, Canterbury, New Zealand. Journal of the Royal Society of New Zealand 10(1): 31-63.

. 1981. Middle Triassic megafossil plants from Long Gully, near Otematata, north Otago, New Zealand. Journal of the Royal Society of New Zealand 11(3): 167-200.

. 1983. Middle Triassic megafossil marine algae and land plants from near Benmore Dam, southern Canterbury, New Zealand. Journal of the Royal Society of New Zealand 13(3): 129-154.

. 1985. Triassic fossil plant fragments from shallow marine rocks of the Murihiku Supergroup, New Zealand. Journal of the Royal Society of New Zealand 15(1): 1-26.

. 1987. Triassic vegetation and geography of the New Zealand Formation of the Gondwana supercontinent. Geophysical Monograph 41: 29-39. G.D. McKenzie 92

(ed.), Gondwana Six: Stratigraphy, Sedimentology, and Paleontology, American Geophysical Union, Washington, D.C.

. 1995. Permian-Triassic life crisis on land. Science 267:77-80.

and D.L. Dilcher. 1988. Reconstructions of selected seed ferns. Annals of the Missouri Botanical Garden 75: 1010-1057.

, R.E. Gould, and B. Runnegar. 1977. Isotopic dating of a Middle Triassic megafossil flora from near Nymboida, northeastern New South Wales. Proceedings of the Linnean Society of New South Wales 101(2): 77-113.

Ricklefs, R.E. 1990. Long-term development of biological communties. pp. 1-12. In: Miller, W., Ill (ed.), Paleocommunity Temporal Dynamics: the long-term development of multispecies assemblies. The Paleontological Society Special Publication No. 5. Lawrence, Kansas.

Rigby, J.F. 1977. New collections of Triassic plants from the Esk Formation, southeast Queensland. Queensland Government Mining Journal 78:320-325.

. 1985. Some Triassic (Middle Gondwana) floras from South Victoria Land, Antarctica. New Zealand Geological Survey; Homibrook Symposium, Christchurch, New Zealand, Abstract 78-79.

and J.M. Schopf. 1969. Stratigraphic implications of Antarctic paleobotanical studies, pp. 91-106. In: Amos, A.J. (ed.), Gondwana Stratigraphy (International Union of Geological Sciences Symposium, 1967). Earth Sciences, Vol. 2. UNESCO, Paris.

Rocha-Campos, A.C. 1973. Upper Paleozoic and Lower Mesozoic paleogeography, and paleoclimatological and tectonic events in South America, pp. 398-424. In: Logan, A. and L.V. Hills (eds.), The Permian and Triassic Systems and their Mutual Boundary. Canadian Society of Petroleum Geologists, Calgary.

Rohlf, F.J. 1986. Relationships among eigenshape analysis, Fourier analysis, and analysis of coordinates. Mathematical Geology 18(8): 845-857.

and J. W. Archie. 1984. A comparison of Fourier methods for the description of wing shape in mosquitoes (Diptera: Culicidae). Systematic Zoology 33(3): 302-317.

and L.F. Marcus. 1993. A revolution in morphometries. Trends in Ecology and Evolution 8(4): 129-132.

Rosen, B.R. 1988. From fossils to earth history: applied historical biogeography, pp. 455- 481. In: A.A. Myers and P.S. Giller (eds.), Analytical Biogeography: an integrated approach to the study of animal and plant distributions. Chapman and Hall, London. 93

Ross, M.I. 1992. PaleoGeographic Information System/Mac™ Earth in Motion Technologies. Houston, TX.

and C.R. Scotese. 1992. Paleoclimate™ Earth in Motion Technologies. Houston, TX.

Rothwell, G. W. and R. Serbet. 1994. Lignophyte phylogeny and the evolution of spermatophytes: A numerical cladistic analysis. Systematic Botany 19(3): 443-482.

Roy Chowdhury, M.K., M.V.A. Sastry, S.C. Shah, G. Singh, and S.C. Ghosh. 1975. Triassic floral succession in the Gondwana of peninsular India, pp. 149-158. In: Campbell, K.S.W. (ed.) Gondwana Geology. (3rd Gondwana Symposium; Canberra, Australia, 1973). Australian National University Press, Canberra.

Sahni, M.R. and C.N. Rao. 1958. A note on the correlation of the Parsora and Tiki beds of Vindhya Pradesh. Congreso Geologico Intemacional XX Sesidn-Comision para la correlacion del sistema Karroo (Gondwana): 95-103.

Saksena, S. 1962. On some fossil plants from Karkati, Kamtadand and Parsora, in the South Rewa Gondwana Basin, Central India. The Palaeobotanist 10(1-2): 91-96, pis. 1- 2 .

Schopf, J. M. 1973. The contrasting plant assemblages from Permian and Triassic deposits in Southern continents, pp. 379-397. In: A. Logan and L.V. Hills (eds.), The Permian and Triassic Systems and their Mutual Boundary. Canadian Society of Petroleum Geologists, Calgary.

Seward, A. C. 1903. Fossils floras of Cape Colony. Annals of the South African Museum 4(1): 1-122.

. 1908. On a collection of fossil plants from South Africa. Quarterly Journal of the Geological Society of London 64(253): 83-108, pis. II-VIII.

. 1914. Antarctic fossil plants. British Museum (Natural History), British Antarctic (Terra Nova) Expedition (Geology) 1: 1-49.

and R.E. Holttum. 1921. On a collection of fossil plants from southern Rhodesia. Bulletin of the Geological Survey of South Rhodesia 8:39-45.

Shi, G.R. 1993. Multivariate data analysis in palaeoecology and palaeobiogeography-a review. Palaeogeography, Palaeoclimatology, and Palaeoecology 105: 199-234.

Shirley, J. 1898. Additions to the fossil flora of Queensland. Queensland Department of Mines. Bulletin of the Geological Survey of Queensland 7: 1-25, pis. IOOCVII.

Shubin, N.H. and H-D. Sues. 1991. Biogeography of early Mesozoic continental tetrapods: patterns and implications. Paleobiology 17(3): 214-230. 94

Simpson, G.G. 1960. Notes on the measurement of faunal resemblance. American Journal of Science 258A: 300-311.

Smith, A.G., A.M. Hurley, and J.C. Briden. 1981. Phanerozoic Paleocontinental World Maps. Cambridge Earth Science Series. Cambridge University Press, Cambridge.

, D.G. Smith, and B.M. Funnell. 1994. Atlas of Mesozoic and Cenozoic coastlines. Cambridge University Press, Cambridge.

Solms-Laubach, H. and G. Steinmann. 1899. Beitrage zur geologie und palaeontologie von Siidamerika. VII. Das auftreten und die flora der rhatischen Kohlenschichten von La Temera (Chile). Neues Jahrbuch flir Mineralogie, Geologie und Palaeontologie, Beilage-Band XII: 581-609. pis. XIII-XIV.

Spicer, R.A. 1990. Climate from plants, pp. 401-403. In: Briggs, D.E.G. and P.R. Crowther (eds.), Paleobiology: a synthesis. Blackwell Scientific Publishing, Oxford.

and J.L. Chapman. 1990. Climate change and the evolution of high-latitude terrestrial vegetation and floras. Trends in Ecology and Evolution 5(9): 279-284.

Springer, D.A. and A.I. Miller. 1990. Levels of spatial variability: the "community'1 problem, pp. 13-30. In: Miller, W., Ill (ed.), Paleocommunity Temporal Dynamics: the long-term development of multispecies assemblies. The Paleontological Society Special Publication No. 5. Lawrence, Kansas.

Srivastava, A.K. 1992. Alien elements in the Gondwana flora of India. The Palaeobotanist 40: 147-156.

Srivastava, S.C. 1974. Mesozoic pteridosperms. pp. 179-188. In: Surange, K.R., R.N. Lakhanpal and D.C. Bharadwaj (eds.), Aspects and Appraisal of India Palaeobotany. Birbal Sahni Institute of Palaeobotany, Lucknow.

. 1979. The Triassic flora of Nidpur, India, pp. 105-108. In: B. Laskar and C.S. Raja Rao (eds.), IV International Gondwana Symposium, Calcutta (1977), Papers Volume I (2). Hindustan Publishing Corporation, Delhi.

. 1987. Stratigraphic position and age of plant bearing Nidpur beds. The Palaeobotanist 36: 154-160.

and P.K. Pal. 1983. Upper Triassic fossil plants from Son River section near Giar, Shahdol District, Madhya Pradesh, India. Geophytology 13(2): 238, pi. 1.

Stace, C.A. 1965. Cuticular studies as an aid to plant taxonomy. Bulletin of the British Museum of Natural History (Botany) 4(1): 1-78.

Steinmann, G. 1921. Rhhtische floren und landverbindungen auf der SUdhalbkugel. Geologische Rundschau 11: 350-354. 95

Stevens, G.C. 1989. The latitudinal gradient in geographical range: how so many species coexist in the tropics. 133(2): 240-256.

Stipanfcic, P.N. 1983. The Triassic of Argentina and Chile, pp. 181-99. In: Moullade, M., and A.E.M. Naim (eds.) The Phanerozoic Geology of the World II: The Mesozoic. Elsevier, Amsterdam.

and M.I.R. Bonetti. 1965. Las especies del gdnero Saportaea del Tridsicodel Barreal (San Juan). Revista del Museo Argentino de Ciencias Naturales Bernardino Rivadavia. Paleontologia 1(4): 81-114. Ldminal-IV.

and M.I.R. Bonetti. 1969. Consideraciones sobre la cronologia de los terrenos tridsicos argentinos. Gondwana Stratigraphy. IUGS Symposium. Buenos Aires (1967). UNESCO, Earth Sciences 2: 1081-1119.

Swofford, D.L. 1993. PAUP: Phylogenetic analysis using parsimony, version 3.1.1. Illinois Natural History Survey, Champaign.

Szajnocha, L. 1888. Uber fossile Pflanzenreste aus Cacheuta in der Argentinischen Republik. Sber. Akad. Wiss. Wien. Mathematisch - naturwissenschaftliche klasse 97: 219-245.

Tavera J. J. 1960. El Tridsico del valle inferior del Rfo Bfo-Bfo. Anales Facultad de Ciencias Fisicas y Matematicas (Chile Universidad, Santiago, Chile) 20:321-345.

Taylor, E. L. 1989. Tree-ring structure in woody axes from the central Transantarctic Mountains, Antarctica, pp. 109-113. In: Guo Kun (ed.), Proceedings of the International Symposium on Antarctic Research, Chinese Committee on Antarctic Research. China Ocean Press, Tianjin.

. 1992. The occurrence of a Rhexoxylon-like stem in Antarctica. Courier Forschungsinstitut Senckenberg 147: 183-189.

. (in press). Enigmatic gymnosperms? Structurally preserved Permian and Triassic seed ferns from Antarctica. Review of Palaeobotany and Palynology.

and T.N. Taylor. 1988. Late Triassic flora from Mount Falla, Queen Alexandra Range. Antarctic Journal of the U.S. 23(5): 2-3.

and T. N. Taylor. 1990. Structurally preserved Permian and Triassic floras from Antarctica, pp. 149-163. In: Taylor, T. N. and Taylor, E. L. (eds.), Antarctic Paleobiology: Its Role in the Reconstruction of Gondwana. Springer-Verlag, New York.

, N.R. Cuneo, and T.N. Taylor. 1991. Permian and Triassic fossil forests from the central Transantarctic Mountains. Antarctic Journal of the U.S. 26(5): 23-24. 96

, T.N. Taylor, and J.W. Collinson. 1989. Depositional setting and paleobotany of Permian and Triassic permineralized peat from die central Transantarctic Mountains, Antarctica. International Journal of Coal Geology 12:657-679.

T.N. Taylor, and N.R. Cuneo. 1992. The present is not the key to the past: a polar forest from the Permian of Antarctica. Science 257: 1675-1677.

, T.N. Taylor, B. Meyer-Berthaud, J. Isbell, and N.R. Ciineo. 1990. A Late Triassic flora from the Allan Hills, Southern Victoria Land. Antarctic Journal of the U.S. 25(5): 20-21.

Taylor, S.E. 1972. Optimal leaf form. pp. 73-86. In: D.M Gates and R.B Schmerl (eds.), Perspectives in Biophysical Ecology. Springer-Verlag, New York.

Taylor, T.N. and E.L. Taylor. 1993. The Biology and Evolution of Fossil Plants. Prentice Hall, New Jersey.

, E.L. Taylor, and J.W. Collinson. 1986. Paleoenvironment of Lower Triassic plants from the Fremouw Formation. Antarctic Journal of the U.S. 21(5): 26-27.

Tenison-Woods, J.E. 1883. On the fossil flora of the coal deposits of Australia. Proceedings of the Linnean Society of New South Wales 8: 1-131.

Tessensohn, F., and K. Madler. 1987. Triassic plant fossils from North Victoria Land, Antarctica. Geol. Jb. B 66: 187-201.

Thomas, H.H. 1933. On some pteridospermous plants from the Mesozoic rocks of South Africa. Philosophical Transactions of the Royal Society of London B222: 193-265, 2 pis.

. 1955. Mesozoic pteridosperms. Phytomorphology 5(2-3): 177-185.

Thulbom, R.A. 1986. Early Triassic tetrapod faunas of southeastern Gondwana. Alcheringa 10: 297-313.

Townrow, J.A. 1957. On Dicroidium, probably a pteridospermous leaf, and other leaves now removed from this genus. Transactions of the Geological Society of South Africa 60: 21-60.

. 1962a. On Pteruchus a microsporophyll of the Corystospermaceae. Bulletin of the British Museum of Natural History, Geology, 6(2): 289-320,3 pis.

. 1962b. Note on the type material of Xylopteris elongate (Carruthers) Frenguelli. Proceedings of the Royal Society of Queensland 72(10): 123-127.

. 1962c. Palaeontology. Journal of the Geological Society of Australia 9(2): 224- 226. 97

. 1966. On Dicroidium odontopteroides andZX obtusifolium in Tasmania, pp. 128- 136. In: Symposium on floristics and stratigraphy of Gondwanaland. Birbal Sahni Institute of Palaeobotany, Lucknow.

. 1967. Fossil plants from Allan and Carapace Nunataks, and from the Upper Mill and Shackleton Glaciers, Antarctica. New Zealand Journal of Geology and Geophysics 10(2): 456-473.

Tucker, M.E. and M.J. Benton. 1982. Triassic environments, climates and reptile evolution. Palaeogeography, Palaeoclimatology, Palaeoecology 40:361-379.

Valentine, J.W. 1990. The fossil record: a sampler of life's diversity. Philosophical Transactions of the Royal Society of London B 330: 261-268.

and D. Jablonski. 1993. Fossil communities: compositional variation at many time scales, pp. 341-349. In: Ricklefs, R.E. and D. Schluter (eds.), Species Diversity in Ecological Communities: historical and geographical perspectives. University of Chicago Press, Chicago.

and E.M. Moores. 1973. Provinciality and diversity across the Permian-Triassic boundary, pp. 759-766. In: A. Logan and L.V. Hills (eds.), The Permian and Triassic Systems and their Mutual Boundary. Canadian Society of Petroleum Geologists, Calgary.

Veevers, J.J. 1988. Gondwana facies started when Gondwanaland merged in Pangea. Geology 16: 732-734.

, D.I. Cole, and E.J. Cowan. 1994a. Southern Africa: Karoo Basin and Cape Fold Belt. pp. 223-279. In: J.J. Veevers and C.M. Powell (eds.), Permian-Triassic Pangean Basins and Foldbelts along the Panthalassan Margin of Gondwanaland. Geological Society of America Memoir 184, Boulder, Colorado.

, P.J. Conaghan, and C.M. Powell. 1994b. Eastern Australia, pp. 11-171. In: J.J. Veevers and C.M.Powell (eds.), Permian-Triassic Pangean Basins and Foldbelts along the Panthalassan Margin of Gondwanaland. Geological Society of America Memoir 184, Boulder, Colorado.

Walkom, A.B. 1917. Mesozoic floras of Queensland. Part I. The flora of the Ipswich and Walloon Series. Queensland Geological Survey 257: 1-66, pis. 1-10.

. 1921. Mesozoic floras of New South Wales. Part I. Fossil plants from Cockabutta Mountain and Talbragar. Memoirs of the Geological Survey of New South Wales (Palaeontology) 12: 1-21, pis. I-VI.

. 1925a. Fossil plants from the Narrabeen Stage of the Hawkesbury Series. Proceedings of the Linnean Society of New South Wales 50: 214-224, pis. XXIV- XXXI. 98

. 1925b. Notes on some Tasmanian Mesozoic plants. Part 1. Papers and Proceedings of the Royal Society of Tasmania: 73-89.

Walton, J. 1923. On Rhexoxylon Bancroft - a Triassic genus of plants exhibiting a liane- type of vascular organization. Philosophical Transactions of the Royal Society of London 212B: 79-109.

Waterhouse, J.B. 1973. The Permian-Triassic boundary in New Zealand and New Caldonia and its relationship to world climatic changes and extinction of Permian life, pp. 445-464. In: A. Logan and L.V. Hills (eds.), The Permian and Triassic Systems and their Mutual Boundary. Canadian Society of Petroleum Geologists, Calgary.

Webb, J.A. and C.R. Fielding. 1993a. Permo-Triassic sedimentation within the Lambert Graben, northern Prince Charles Mountains, East Antarctica, pp. 357-369. In: R.H. Findlay, R. Unrug, M.R. Banks, and J.J. Veevers (eds.), Gondwana Eight: Assembly, Evolution and Dispersal. A.A. Balkema, Rotterdam.

and . 1993b. Revised stratigraphical nomenclature for the Permo-Triassic Flagstone Bench Formation, northern Prince Charles Mountains, East Antarctica. Antarctic Science 5(4): 409-410.

White, M.E. 1961. Plant fossils from the Canning Basin, western Australia, pp. 291-320. In: Veevers, J. J. and A. T. Wells (eds.), The Geology of the Canning Basin, Western Australia. Bureau of Mineral Resources, Geology, and Geophysics Bulletin 60, Halstead Press, Sydney.

. 1990. The Flowering of Gondwana. Princeton University Press, New Jersey.

White, R.J., H.C. Prentice, and T. Verwijst. 1988. Automated image acquisition and morphometric description. Canadian Journal of Botany 66:450-459.

Wolfe, J.A. 1990. Palaeobotanical evidence for a marked temperature increase following the Cretaceous/Tertiary boundary. Nature 343:153-156.

Woodcock, D.W. 1992. Climate reconstruction based on biological indicators. Quarterly Review of Biology 67(4): 457-477.

Yao, X., T.N. Taylor, and E.L. Taylor. 1995. The corystosperm pollen organ Pteruchus from the Triassic of Antarctica. American Journal of Botany 82(4): 535-546.

Ziegler, A.M., C.R. Scotese, and S.F. Barrett. 1983. Mesozoic and Cenozoic Paleogeographic Maps. pp. 240-252. In: Brosche/Siindermann (ed.), Tidal Friction and the Earth's Rotation II. Springer-Verlag, Berlin. APPENDICES

99 APPENDIX A

TABLES

100 Table 1. Directory of specimens.

Macrofloras

Specimen No. Collection Locality Formation Age

Ant 71-1-1 to 27 University of Kansas, Mt. Wisting, Fremouw Anisian-Ladinian Museum of Natural History, Transantarctic Mts., Lawrence, Kansas Antarctica and Byrd Polar Research (86°27'S, 165°30'W) Center, Columbus, Ohio

Ant 67-9-31 to 182, Mt. Bumstead, Fremouw Anisian-Ladinian 67-9-203 to 222 Transantarctic Mts., Antarctica (85°39'S, 174°10'E)

Ant 67-9-1 to 30 west of Lindsay Peak, Falla Camian-Norian Marshall Mts., T ransantarctic Mts, Antarctica (84°36'S, 163°15'E)

Ant70-17B-1 to 7 Kenyon Peaks, Fremouw/Falla Ladinian-Norian Transantarctic Mts., Antarctica (84°33'S, 163°36'E)

T-4 #26-37; Gordon Valley, Fremouw Anisian-Ladinian T-8#188-192; Transantarctic Mts., T-9 #280-285 Antarctica (84°23'S, 164°60’E) Table 1 (continued)

Ant 70-8-7 to 48; M t Falla, Transantarctic Falla Carnian-Norian Ant 70-8-51 to 52; Mts., Antarctica Ant 70-8-79 to 179; (84°21'S, 164°42'E) Ant 67-8-17 to 44; 67-8-60 to 78; T-5 #38-146; T-7 #161-187, 194- 200

Ant 67-8-2 to 14; Fremouw Peak, Fremouw/Falla Ladinian-Norian Ant67-8-79 to 103; Transantarctic Mts., Ant70-9-139 to 164; Antarctica T-10 #286-304 (84° 16'S, 164e21'E)

SR 139-1 to 62; Allan Hills, Southern Lashly Anisian-Norian Ant 70-3-1 to 85; Victoria Land, Antarctica (Member B, C) T-8 #201-279; (76°42'S, 159°42'E) T-11 #305-547, 556-667,675-892

T-3 #7, 8, 17, 19, Newport and North Avalon, Newport Scythian-Ladinian 22 New South Wales, Australia (33°40'S, 151°E)

AF 9-34,38-58, 199 Canterbury Museum, Shapeless Mt., Southern Lashly Anisian-Norian Christchurch, New Zealand Victoria Land, Antarctica (77°25'S, 160°20'E) Table 1 (continued)

A F305-315 " Horseshoe Mt., Southern Lashly Anisian-Norian Victoria Land, Antarctica (77°34'S, 159‘57'E)

USNM 458658- Natural Museum of Natural Sydney region (Mona Vale, Newport, Narrabeen Scythian-Ladinian 458667 History, Smithsonian North Avalon, Long Reef), Group; Bald Hill Institution, Washington, New South Wales, Australia Claystone D.C. (33°40'S, 151°E)

USNM 42825- Passo das Tropas, Santa Santa Maria Ladinian-Norian 42827; Maria, Rio Grande do Sul, 253022-253035, Brazil 253038; (29°40'S, 53°48'W) 341337-341440

USNM 40352, Cacheuta, Mendoza, Cacheuta Camian 40355-40359; Argentina 253039-253042 (32°20'S-33°S, 68“45'W-69° 15'W)

USNM 40362- Barreal, San Juan, Barreal Ladinian-Camian 40363; 42178 Argentina (31°33'S, 69°28'W)

USNM 40353- Tierra del Caballo Anca, San — Carnian-Norian? 40354 Juan, Argentina Table 1 (continued)

3935-3994,3995- Museo Argentino de Barreal, San Juan, Barreal and Ladinian-Camian 4006,4105-4138, Ciencias Naturales Argentina Cortaderita 4145-6,4178-4193, Bernardino Rivadavia, (31 °33'S, 69°28'W) (NF 1, 2, 3) 4198-4212,4235, Buenos Aires, Argentina 4257, 4267, 4295, 4298,4312, 4320- 4327, 6057-6100, 6457, 8129-8140

4587, 4787-8,4800- Arroyo Llantenes, Mendoza, Chihuiu, Llantenes Camian-Norian 18, 4829, 4873, Argentina Group 4885-6, 5041-5097, (35‘40'S, 69°40'W) 5250-5323

6345, 6357, 9444- Hilaro, San Juan, Argentina Barreal Ladinian 9550, 9578-9582 (31°33'S, 69°28'W)

10631 Cacheuta, Mendoza, Cacheuta Camian Argentina (32*20'S-33°S, 68°45,W-69e 15'W)

7522-7546 Paso Flores, Neuqudn, Paso Flores Norian Argentina (40’30'S, 70"30'W) Table 1 (continued)

1, 18,31,33-34, " El Tranquilo, Santa Cruz, El Tranquilo Ladinian-Camian 38-39, 43, 54-55, Argentina 64, 80-84, 86, 88- (48°S, 69°W) 89, 94, 101, 103- 104, 138-140, 142, 146-147, 149-150, 162- 163, 167-168

13-15 Department of Botany, Ciudad Universitaria, Buenos Aires, Argentina

16 Barreal, San Juan, Barreal Ladinian-Camian Argentina (31°33'S, 69°28'W)

17-20 Cacheuta, Mendoza, Cacheuta Camian Argentina (32°20'S-33°S, 68°45’W-69B15,W)

10-12 Dinmore, near Brisbane, Ipwich Anisian-Norian Queensland, Australia (27‘24'S, 153’09'E)

578-579,709 Department of Geology, Barreal, San Juan, Barreal Ladinian-Camian Ciudad Universitaria, Argentina Buenos Aires, Argentina (31 °33'S, 69’28'W) Table 1 (continued)

2 Cacheuta, Mendoza, Potrerillos Ladinian-Camian Argentina (32°20'S-33°S, 68*45^-69*15^)

7441 Paso Flores, Neuquln, Paso Flores Norian Argentina (40°30'S, 70°30'W)

7435 Quebrada de la Pena, San Barreal? Ladinian-Camian Juan, Argentina (30*-31°S, 67*-69*W)

10632 El Chifldn, La Rioja, Los Rastros? Ladinian-Camian? Argentina (29°15,-30°30,S, 67*30'- 68*3 O'W)

2304-2305, 2313, PRINGEPA-CONICET, El Tranquilo, Santa Cruz, El Tranquilo Ladinian-Camian 2344, 2357, 2359, Corrientes, Argentina Argentina 2405, 6244, 6246, (48"S, 69*W) 6252, 6292,6314- 16, 6324, 6358-60, 6364-65, 6373-74, 6475, 6490-91, 6504-5, 6513-14, 6776, 6792-93, 6802-3, 8106, 8120, 8124-25, 8130, Table 1 (continued)

8173, 8179-80, 8182, 8149,8183, 8210-11,8217, 8220-21, 8223-30, 8235-37, 8241, 8243, 8251-54, 8262-63, 8265-69, 8288-91, 8293-94, 8330, 8360

246-249 " Rio Chifldn, La Rioja, Los Rastros? Ladinian-Camian? Argentina (approx. 29°15'-30o30'S, 67°30'-68e30'W)

261-286,290 " Quebradade la Cortaderita, Barreal Ladinian Barreal, San Juan, (NF1) Argentina (31C33'S, 69°28'W)

287-289, 291-358, Cortaderita Ladinian-Camian 420 (NF3)

359-364 Cortaderita Ladinian-Camian (NF2) 365-390 Neuqudn, Argentina?

391-419, 421-489, Quebrada de Hilario, Barreal, Cortaderita Ladinian-Camian 508-524 Barreal, San Juan, Argentina (31033'S, 69°28'W) Table 1 (continued)

4139-4480 Brisbane region (Ipswich, Ipswich, Esk Anisian-Norian Esk, Dinmore, Goodna, Mt. Gravatt, Shomecliffe, Wiwenhoe, Moombre, Bryden, Ti-Tree Gully, Sheep Creek, Denmark Hill) Queensland, Australia (27°24'S, 153°09'E)

4620-4711 Sydney region (S. Avalon, Newport?, Anisian-Norian? Turimetta Head, Long Nanabeen Group ReeQ, New South Wales, Australia (33°40'S, 151°E)

4481-4483 Nymboida, New South Nymboida Coal Anisian-Ladinian? Wales, Australia Measures (30°S, 152°E)

Cuticle Specimen and Collection Locality Formation Age Slide No.

38-1 to 38-6, University of Kansas, ML Falla, Transantarctic Falla Camian-Norian 57A/B-1 to 10, Museum of Natural History, Mts., Antarctica 65-1 to 65-10, Lawrence, Kansas (84‘21'S, 164°42'E) 72/194B-1 to 7, 74-1, T-5andT-7 75/177-1 to 7, 76-1 to 76-4, Table 1 (continued)

78-1 to 78-5, 79-1 to 79-5, 81A-1 to 81A-2, 81B-1 to 81B-3, 85A-1 to 85A-3, 98/110-1 to 11, 103-1 to 103-7, 128-1 to 128-10, 133-1 to 133-3, 145-1 to 145-6, 167-1 to 167-6, 170-1, 181B-1, 186-1, 194A-1 to 194A-2

28 " Gordon Valley, Fremouw Anisian Transantarctic Mts., T-4 Antarctica (84°23'S, 164’60'E)

253 " Allan Hills, Southern Lashly Anisian-Norian Victoria Land, Antarctica (Member B, C) T-8 (77° S, 160°E)

USNM 458659-1 to Natural Museum of Natural Sydney region (Mona Vale, Newport, Narrabeen Scythian-Ladinian 458659-5 History, Smithsonian North Avalon, Long Reef), Group; Bald Hill 458660-1 to 3 Institution, Washington, New South Wales, Australia Claystone 458661-1 to 2 D.C, (33°40'S, 151°E) 458662-1 to 2 Table 1 (continued)

USNM 458663-1 to 458663-4 458664-1 to 3 458665A-1 to 5 458665B-1 458666-1 to 8 458667-1

USNM 40357-1 to 4 " Cacheuta, Mendoza, Cacheuta Camian Argentina (32°20'S-33°S, 68°45'W-69° 15'W) Table 2. Dicroidium species from Antarctica.

Locality Coordinates Formation Age Species Reference

Transantarctic Mountains

Mt. Bumstead, 85°39'S, 174°10'E Fremouw/Falla Ladinian-Camian? Dicroidium dutoitii Townrow, 1967 Mill Glacier D.feistmantelii D. odontopteroides Xylopteris elongata cf. Diplasiophyllum acutum

Misery Peak, 85°31'S, 178°16'W Fremouw/Falla Ladinian-Camian? D. dutoitii Townrow, 1967 Shackleton Glacier

Shackleton Glacier 85° S, 175° W Fremouw Anisian-Ladinian D. cf. Rigby and Schopf, region odontopteroides 1969

Beardmore Glacier 84°30’S, 165° E Fremouw Anisian-Ladinian D.feismantelii Rigby and Schopf, region D. odontopteroides1969 Mt. Falla, 84°21'S, 164°42'E Falla Camian-Norian D. dubium Boucher et al., 1993 Beardmore Glacier D. lancifolium D. odontopteroides

Fremouw Peak, 84°16'S, 164°21'E Fremouw Anisian D. fremouwensis Pigg, 1990 Beardmore Glacier

Southern Victoria Land

Portal Mt. 78°06'S, 159°10'E LashlyA Anisian-Ladinian D. lanceolatum var. Gabites,’ 1985 I lanceolatwn 1 Table 2 (continued)

D. lanceolatwn var. lineatum D. odontopteroides var. odontopteroides X. elongata var. rigida

Horseshoe Mt. 77°34'S, 159°57'E Lashly Anisian-Norian D. dubium Rigby, 1985 D. odontopteroides D. zuberi X. elongata

ML Fleming 77°33'S, 160°05'E Fleming Scythian D. sp. Plumstead, 1962

Shapeless Mt. 77°25'S, 160°20'E Lashly Anisian-Norian D. dubium Rigby, 1985 D. odontopteroides D. zuberi X. elongata

Lashly Anisian-Norian D. odontopteroidesPlumstead, 1962 Fleming & Lashly A Scythian-Ladinian D. lanceolatum var. Gabites, 1985 lineatum D. odontopteroides var. moltense D. odontopteroides var. odontopteroides Lashly A Anisian-Ladinian D. pinna-distantibus Fleming Scythian J. coriacea ML Bastion 77°19'S, 160°29'E Lashly B, C Anisian-Norian D. lanceolatum var. Gabites, 1985 lineatum Table 2 (continued)

D. lanceolatum var. lanceolatum Lashly A Anisian-Ladinian D. odontopteroides var. crassum Lashly B, C Anisian-Norian D. odontopteroides var. moltense Lashly C Camian-Norian D. odontopteroides var. obtusifolium Lashly A, C Anisian-Norian D. odontopteroides var. odontopteroides Lashly C Camian-Norian D. odontopteroides var. remotum D. rhomboidalis D. spinifolium Lashly B Anisian-Norian D. zuberi Lashly C Camian-Norian X. elongata var. elongata Lashly B Anisian-Norian X. elongata var. rigida Lashly A Anisian-Ladinian J.coriacea

Robison Peak 77°13'S, 160°18'E Fleming Scythian D. sp. Pinet, 1965

Allan Hills 76°42'S, 159°42'E Lashly Anisian-Ladinian D.feistmantelii Rigby and Schopf, D. odontopteroides1969

Lashly Ladinian-Camian? D. dutoitii Townrow, 1967 D. odontopteroides X. elongata cf. Diplasiophyllum acutum J. trilobita Table 2 (continued)

Lashly B, C Anisian-Norian D. lanceolatum var. Gabites, 1985 lanceolatum Lashly C Camian-Norian D. odontopteroides var. argenteum D. odontopteroides var. crassum D odontopteroides var. moltense Lashly A, B, C Anisian-Norian D. odontopteroides var. odontopteroides Lashly A Anisian-Ladinian D. zuberi var. sahnii Lashly C Camian-Norian X. elongata var. elongata Lashly A, C Anisian-Norian X. elongata var. rigida Lashly C Camian-Norian J.coriacea Lashly A Anisian-Ladinian J. stelzneriana var. serrata Lashly C Camian-Norian J. stelzneriana var. stelzneriana J. trilobita Lashly C Camian-Norian D. cf. dutoitii Taylor etal., 1990 Lashly B Anisian-Norian D. odontopteroides Lashly C Camian-Norian cf. J. trilobita Lashly C Camian-Norian cf. Diplasiophyllum acutum

Northern Victoria Land

Vulcan Hills 73°40'S, 163°37'E Beacon Rhaetian D. odontopteroides Tessensohn and Madler, 1987 Table 2 (continued)

East Antarctica

Jetty Peninsula, 70°45'S, 68°45'E Flagstone Bench Camian-Norian D. crassinervis Webb and Fielding, Prince Charles Mts. foTma.stelznerianum 1993a D. zuberi Table 3. Dicroidium species from South America.

Locality Coordinates Formation Age Species Reference

Brazil

Santa Maria, 29°40'S, 53°48'W Santa Maria Ladinian-Camian D. odontopteroides Gordon and Brown, Rio Grande do Sul Zuberia sp. 1952

Camian D. acutum Bortoluzzi et al., 1983 D. argentinum D. stelzneriana D. elongatum

Chile near Iquique, 20° 10'S, 69°20'W Juan de Morales Camian D. odontopteroidesStipanfcic, 1983 Juan de Morales Basin

LaTemera, Vallenar- 27’10'S, 69°45'W LaTemera Camian D. lancifolium Solms-Laubach and Los Vilos Basin Dicroidiopsis dubia Steinmann, 1899; Thinnfeldia cf. incisa Frenguelli, 1943; Thinnfeldia cf. Stipanfcic, 1983 lancifolia

Alto del Carmen, 28°30'S, 70° 50'W — — D. lancifolium Nishida, 1970; Coquimbo Province (40 km SE Vallenar) D. (Thinnfeldia) Mohr and Schoner, odontopteroides 1985 Vicuna, Vallenar- 30°S, 70°45'W LaBreas Camian D. lancifolium Stipanfcic, 1983 _ Los Vilos Basin o Table 3 (continued)

Los Vilos, Vallenar- 32o10,S, 71°30'W Pichidangui Camian-Norian D. odontopteroidesStipanfcic, 1983 Los Vilos Basin El Puqudn Norian D. odontopteroides D. zuberi

Los Molles, ~32°10'S, 71“30'W LosMolles Norian D. lancifolium Azcdrate and Fasola, Aconcagua Province D. odontopteroides 1970 Johnstonia coriacea Zuberi zuberi

BfoBfo,Talca-Bfo 37°S, 72°50'W Quilacoya Camian D. odontopteroidesTavera, 1960; Basin Stipanfcic, 19<83 Unihue Camian D. lancifolium Talcamdvida-Gomero Norian D. lancifolium D. odontopteroides

Thinnfeldia Steinmann, 1921 odontopteroides Thinnfeldia cf. rhomboidalis

Lago Panguipulli (southern Chile?) Panguipulli Ladinian-Norian? D. zuberi var. zuberi Arrondo et al., 1988

Mdrgen Norte de (southern Chile?) Panguipulli Ladinian-Norian D. lancifolium var, Arrondo et al., 1988 Lago Calafquen lancifolium D. zuberi cf. var. sahnii Table 3 (continued)

Argentina

Villa Union Basin, 29°15,-30°30,S, Ischichuca Ladinian D. stelznerianum Stipanfcic, 1983 La Rioja/San Juan 67’30,-68°30'W D. zuberi

Los Rastros Ladinian-Camian D. coriaceum Stipanfcic and Bonetti, D. crassum 1969; Arrondo, 1972; D. dubium Petriella, 1979; D.feistmantelii Stipanfcic, 1983 D. intermedium D. lancifolium D. odontopteroides D. stelznerianum D. zuberi Xylopteris argentina X. densifolia X. elongata var. rigida

Ischigualasto Camian D. coriaceum Frenguelli, 1943; D. lancifolium Archangelsky, 1968; D. odontopteroides Stipanfcic, 1983 D. stelznerianum D. zuberi X. argentina X. densifolia X. elongata

Marayes-Desaguadero 31 °25'S, 67°20'W Quebrada del la Mina Ladinian-Camian D. coriaceum Stipanfcic, 1983 Basin, San Juan and Carrizal D. lancifolium D. odontopteroides D. stelznerianium oo Table 3 (continued)

San Luis/Alto 33°40'S, 66’40'W; (unnamed) Camian? D. odontopteroidesStipanfcic, 1983 Pencoso, San Luis 33 °S, 67°W

Hilaro-Barreal, San 31 °33’S, 69028’W Barreal and Ladinian-Camian D. coriaceum Stipanfcic, 1983 Juan, Cuyean Basin Cortaderita D. lancifolium D. odontopteroides D. stelznerianum D. zuberi X. argentina X. elongata var. rigida Thinnfeldia rhomboidalis

Barreal Ladinian D. brownii Stipanfcic and Bonetti, D. coriaceum 1965; Bonetti, 1966; (NFI) D. (Zuberia) Bonetti, 1968; feistmantelii Petriella, 1979; D. groeberi Artabe, 1990 D. odontopteroides D. stelznerianum D. (Zuberia) zuberi Zuberia barrealensis Z papillata Z. sahnii X. argentina X. elongata var. rigida

Cortaderita Ladinian-Camian D. acutum Bonetti, 1968; (NFII, III) D. coriaceum Stipanfcic and Bonetti, D. dubium 1969; Petriella, 1979 D.feistmantelii D. groeberi D. cf. hughesi vo T able 3 (continued)

D. intermedium D, lancifolium D. narrabeenensis (?) D. odontopteroides D. remotum D. stelznerianum D. zuberi X. argentina X. elegans X. elongata var. rigida

Potrerillos-Cacheuta, 32°20'S-33°S, Las Cabras Ladinian D.acutum Stipanfcic and Bonetti, Mendoza 68°45'W-69°15'W D. dubium 1969; Stipanfcic, 1983 D.feistmantelii D. remotum D. stelznerianum D. zuberi X. spinifolia

Potrerillos Ladinian-Camian D. acutum Frenguelli, 1943; D. coriaceum Stipanfcic and Bonetti, D. crassum 1969; Stipanfcic, D. intermedium 1983; Morel, 1994 D. lancifolium D. odontopteroides D. stelznerianum Diplasiophyllum hughesi X. argentina X. densifolia X. elongata var. rigida Zuberia zuberi to o Table 3 (continued)

Cacheuta Camian D. cacheutense Kurtz, 1921; D. coriaceum Frenguelli, 1943; D intermedium Arrondo, 1972; D. lancifolium Petriella, 1979; D. odontopteroides Baldoni, 1980; D. pinnis-distantibus Stipanfcic, 1983; D. stelznerianum Morel, 1994 Diplasiophyllum hughesi X. argentina X. elongata var. rigida Z. feistmantelii Z. zuberi

Minas de Petroleo, 33°03'S, 69°06'W Cacheuta Camian? D. coriaceum Jain and Delevoryas, Mendoza D. feistmantelii 1967 D. odontopteroides X. rigida

Malargiie, Mendoza 35°40'S, 69°40'W Chihufu Camian D. argentinum Stipanfcic and Bonetti, Malargiie Basin D.feistmantelii 1969; Arrondo, 1972; D. stelznerianum Petriella, 1979 D. zuberi X. (D.) elongata

Tronquimalal Camian-Norian D. acutum Stipanfcic and Bonetti, (Llantenes Group) D. crassum 1969; Petriella, 1979; D. feistmantelii Stipanfcic, 1983 D. incissum D. intermedium D. lancifolium D. odontopteroides to I

T able 3 (continued)

D. pinnis-distantibus D. zuberi X. argentina X. elongata var. rigida X. elongata var. irregularis

Paso Flores, 40°30'S, 70°30'W Paso Flores Norian D. incissum Frenguelli, 1937; Neuqudn D. lancifolium Stipanfcic and Bonetti, D. odontopteroides 1969; Stipanfcic, X. argentina 1983; Morel etal., 1992

Los Menucos, 40’50'S, 68°10'W (unnamed) Ladinian-Camian D. lancifolium Stipanfcic, 1983 Rio Negro D. zuberi

Los Menucos, 40-41 °S,68°W Los Menucos Ladinian-Camian D. (Zuberia) brownii Stipanfcic and Bonetti, Rio Negro var. brownii 1969; Artabe, 1985; D. crassum Artabe, 1990 D. dubium var. australe D. dubium var. tasmaniense D. (Z.) feistmantelii D. incisum D. lancifolium var. lancifolium D. odontopteroides var. moltenense D. odontopteroides var. odontopteroides D. odontopteroides to var. remotum to Table 3 (continued)

D. (Z.) zuberi var. feistmantelii D. (Z.) zuberi var. papiUatum D. (Z.) zuberi var. sahnii D. (Z.) zuberi var. zuberi

El Tranquilo, 48°S, 69°W El Tranquilo Ladinian-Camian D. cacheutense Bonetti, 1964; Santa Cruz D. lancifolium Mendndez, 1968; D. odontopteroides Stipanfcic and Bonetti, D. cf. hugkesi 1969; Arrondo, 1972; D. narrabeenensis? Baldoni, 1980; D. cf. zuberi Stipanfcic, 1983 X. argentina X. elongata X. spinifolia Table 4. Dicroidium species from Australia, Tasmania and New Zealand.

Locality Coordinates Formation Age Species Reference

Western Australia Cornish and M t near 18°53'S, Culvida Sandstone Anisian-Ladinian D.feistmanteli White, 1961 Bannerman, Canning 126°14'E D. odontopteroides Basin

Queensland

nearlnjune, 25°15'S, 148°45'E Moolayember Anisian D.feistmanteUi Playford et al., 1982 Carnarvon Range, D. lancifolium Bowen Basin D. odontopteroides D. superbum D. sp.

Northbrook Bridge, 27°22'S, 152°36'E Esk Anisian-Ladinian D. eskensis Rigby, 1977 Coominya Road, Esk D.feistmanteUi Trough D. odontopteroides

Ipswich/Brisbane near 27'30'S, Esk Anisian-Ladinian D. dubium Hill et al., 1965; region (including 152°40'E D. eskensis Rigby, 1977 Esk, Wivenhoe, D.feistmanteUi Slacks Creek), Esk D. odontopteroides Trough D. superbum T able 4 (continued)

Ipswich/Brisbane near 27°35'S, Ipswich Coal Camian-Norian Thinnfeldia acuta Shirley, 1898; region (including 152"45'E Measures (including T. eskensis Walkom, 1917; Jones Denmark Hill, Tivoli Blackstone, Tivoli and T. feistmanteli and de Jersey, 1947; Coal Mine, Slacks Tingalpa Formations) T. lancifolia Townrow, 1962b; Creek, etc.), Ipswich (=T. indica) Hill etal., 1965 Basin, Clarence- T. narrabeenensis Moreton Basin T. odontopteroides T. talbragarensis (-D . dubium) Johnstoniacoriacea (=D. stelznerianum) J. dentata (~D. stelznerianum) J. trilobita Xylopteris elongata X. spinifolia X. tripinnata

New South Wales near Delungra, New 29°39'S, 150°51'E GunneeBeds Anisian-Ladinian D. dubium Bourke et al., 1977 England Fold Belt D. lancifolium D. odontopteroides D. zuberi Jofmstoniacoriacea (=D. dutoitii) J. stelzneriana Xylopteris elongata 'Dicroidiopsis' sp. (-D . gouldii) ls> U\ Table 4 (continued)

Red Cliff, Clarence- 29°40'S, 153°20'E Red Cliff Coal Camian-Norian D. dentatum Flint and Gould, 1975 Moreton Basin Measures (-D . stelznerianum) D.feistmanteUi D. odontopteroides Xylopteris elongata

Nymboida, Clarence- 29°58'S, 152°45'E Basin Creek and Anisian-Ladinian D. dubium Flint and Gould, 1975; Moreton Basin Cloughers Creek (var.tasmaniense) Retallacket al., 1977 Formations D. eskense (Nymboida Coal D.feistmanteUi Measures) D. incisum D. lancifolium D. narrabeenense D. odontopteroides (var. remotum) D. superbum D. zuberi near Port Macquarie, near 31 °25'S, Camden Haven Scythian D. dubium var. Holmes and Ash, Lome Basin 152°25'E Group australe 1979 D. zuberi var. feistmanteUi D. voiseyi (=D. lancifolium) near Dubbo, Great near 32° 1 l'S, Wallingarah Anisian D. coriaceum var. Holmes, 1982 Artesian Basin 148°35'E coriaceum D. dubium var. dubium D. elongatum var. rigidum D. eskense T able 4 (continued)

D. odontopteroides var. moltenense D. natalense D. prolungatum D, shirleyi D. spinifolium D. superbum D. cf. voiseyi (=D. lancifolium) D. zuberi

Talbragar, Great 32'12'S, 148°37'E Wallingarah Anisian Thinnfeldia Walkom, 1921 Artesian Basin (Talbragar) feistmanteli T. pinnata (=D. odontopteroides) T. talbragarensis (=D. dubium)

Quom, Springfield 32°25'S, 138°05'E ? Camian-Norian D. acuta Amtsberg, 1969 Basin D. feistmanteli D. odontopteroides Xylopteris elongata

Sydney region near 33°53'S, Coalcliff and Scythian D.callipteroides Retallack, 1980a; (including Turrimetta 15riO 'E Wombarra White, 1990 Head, Benelong, and Narrabeen), Sydney Basin Table 4 (continued)

Gosford, Bald Hill Scythian-Anisian D. dubium var. Retallack, 1977; Claystone australe Retallack, 1980a D. lancifolium var. lancifolium D. narrabeenensis D. zuberi var. feistmantelii

Newport and Anisian D. acuta Walkom, 1925a; Hawkesbury, D. australis (=D. Jacob and Jacob, Narrabeen Group dubium) 1950; D. brownii var. Retallack, 1977; barrealense Retallack, 1980a D. brownii var. brownii D. dubium (var. australe) D. eskensis D.feistmanteUi D. lancifolium (var. lineatum) D. narrabeenensis (var. bursellii, var. narrabeenense) (-D . lancifolium, D. dubium) D. odontopteroides D. pinnis-distantibus cf. D. talbragarensis D. townrovii D. walkomi (=D. dubium) T able 4 (continued)

D. zuberi (var. feistmantelii, var. papillatum,) Xylopteris elongata (var. rigida) X. tripinnata

Bringelly Shale, Anisian-Ladinian D. dubium var. Jacob and Jacob, Wianamatta Group dubium 1950; Retallack, 1977 D. feistmanteli D. odontopteroides var. moltenense D. zuberi var. feistmantelii D. zuberi var. sahnii

Tasmania

St. Mary's region near 41°35'S, Tiers, Mt. Nicholas Anisian-Norian D. dubium Walkom, 1925b; (including ML 148’11'E and other coal T. feistmanteli Townrow, 1966; Nicholas, Fingal, measures T. lancifolia Forsyth, 1989 Langloh, Dalmayne, (Upper Parmeener D. obtusifolium Berriedale, Cornwall Supergroup) D. odontopteroides Mine, Barbers Mine Johnstonia coriaceum and Seymour Mine (=D. dutoitii, D. localities), stelznerianum) Tasmania Basin J.dentata J. trilobita X. elongata X. spinifolia T able 4 (continued)

Ben Lomond region near 41 °38'S, Tiers, Triassic coal Anisian-Norian D. obtusifolium Townow, 1966 (including Ml 147°42'E measures D. odontopteroides Christie, Stanhope, (Upper Parmeener and Storey's Creek Supergroup) localities)

Campania, Coal R. 42°36'20nS, Tiers, Triassic coal Anisian-Norian D. odontopteroidesTownrow, 1966 147°2030"E measures (Upper Parmeener Supergroup)

Hobart region near 42°50'S, Tiers, New Town Anisian-Norian D. obtusifolium Walkom, 1925b; (including New 147821’E and other coal D. odontopteroidesTownrow, 1966; Town, Lords Hill, measures (Upper Forsyth, 1989 and Hobart) Parmeener Supergroup)

Catamaran 43°30'S, 146"50'E Tiers, Triassic coal Anisian-Norian D. obtusifolium Townrow, 1966 measures (Upper Parmeener Supergroup)

New Zealand

Mt. Potts, 43°30'S, 170°50'E Tank Gully Coal Anisian-Ladinian D. crassum Arber, 1917; Tank Gully Measures, D. dubium var. Retallack, 1979; Torlesse Supergroup dubium Retallack, 1980b; D. dubium var. Retallack, 1987 tasmaniensis D. elongatum var. elongatum D. incisum Table 4 (continued)

D. odontopteroides var. lancifolium D. odontopteroides var. odontopteroides D. odontopteroides var. moltenense D.prolungatum D. stelznerianum var. stelznerianum D. zuberi var. papillatum D. zuberi var. sahnii near Benmore Dam, near 45°01'S, Torlesse Supergroup Anisian-Ladinian D. odontopteroides Retallack, 1981; Long Gully 169°50'E var. moltenense Retallack, 1983; Retallack, 1987

Gore region near 46'05'S, Murihiku Scythian-Anisian D. dubium var. Arber, 1917; (including Kaka 168°58'E Supergroup tasmaniensis Retallack, 1985; Point, Castle Downs middle-late Norian D. stelznerianum var. Retallack, 1987 Swamp, Wairaki Hut, stelznerianum Kaihiku Gorge, and early Anisian D. zuberi var. Ben Callum localities) feistmantelii Scythian-Anisian D. sp.

Mataura Is. 460H'S, I68*51'E Murihiku Scythian-Ladinian D. sp. Retallack, 1985 Supergroup Retallack, 1987

Glenham, near near 46°24'S, Murihiku Norian cf. D. dubium var. Pole and Raine, 1994 Invercargill 168°24'E Supergroup dubium Table 4 (continued)

Owaka Cr. 46°27'S, 169°40'E Murihiku Scythian-Anisian Thinnfeldia Arber, 1917 Supergroup feistmantelii T. lancifolia T. odontopteroides Table 5. Dicroidium species from Africa and Madagascar.

Locality Coordinates Formation Age Species Reference

Libya

Djebel Ben Ghnema near 26°20'N, Unar Triassic D. nidpurensis Lejal-Nicol, 1979 15*50'E D. odontopteroides Thinnfeldia decurrens T. rhomboidalis T. sp.

Zambia

Upper Luangwa 10°30'S; 33“E Ntawere Anisian cf. D. Lacey and Smith, Valley narrabeenensis 1972; Lacey, 1974 near junction of 14°42'S, 29"38’E upper Karroo Anisian-Ladinian D. cf. Lacey and Lucas, Ntimba R. and odontopteroides 1984 Lunsemfwa R., D. cf. zuberi Luano Valley near Kabesha, 14*43'S, 29‘37'E upper Karro Anisian-Ladinian D. cf. zuberi Lacey and Lucas, Luano Valley 1984

Zimbabwe

Sengwa 17°08'S, 28"07'E Molteno Camian-Norian D. lancifolium Lacey, 1970; Lacey, D. narrabeenensis 1976 D. odontopteroides u> D. sp. w T able 5 (continued)

Sinamwenda 17°10'S, 27°47'E Molteno Camian-Norian D. lancifolium Lacey, 1970; Lacey, (D. bursellii) 1976

Sebungwe 17°28'S, 27°56IE Molteno Camian-Norian D. spp. Lacey, 1961

Luzulukulu R. 17°32'S, 27°51'E Molteno Camian-Norian D. lancifolium Bond, 1965; Lacey, (Ruzuruhuru R.) D. odontopteroides1970

Locality no. 2, 17°33'S, 27°49'E Molteno Camian-Norian D. cf. feistmanteli Lacey, 1970 Ruzuruhuru R., D. lancifolium D. cf. lancifolium D. sp.

Locality no. 1, 17°35'S, 27°43'E Molteno Camian-Norian D. lancifolium Lacey, 1970 Ruzuruhuru R. D. cf. lancifolium

Locality no. 3, 17°35'S, 27045'E Molteno Camian-Norian D. sp. Lacey, 1970 Ruzuruhuru R.

Somabula near 18°S, 28°E Molteno Camian-Norian D. feistmanteli Seward and Holttum, D. lancifolium 1921; DuToit, 1927; D. odontopteroidesLacey, 1961

South Africa

Natal near 29°30'S, Molteno Camian-Norian Thinnfeldia acuta DuToit, 1927 30°30’E T. narrabeenensis T. rhomboidalis Stenopteris elongata (=D. spinifolium) Table 5 (continued)

Umkomaas near 30°13'S, Molteno Camian-Norian D. feistmanteli DuToit, 1927; 30°48,E D. odontopteroides Thomas, 1933; (D. lancifolium) Townrow, 1957 Stenopteris densifolia (=D. spinifolium, D. elongata) Johns tonia coriacea (=D.coriacium)

Umkomaas Valley 29-°32°S, 26°-30°E Molteno Camian-Norian D. coriacium Anderson and (39 localities, Karoo D. crassinervis Anderson, 1983 Basin) D. dubium D. elongatum (D. feismanteli) D. hughesii (D. lancifolium) D. narrabeenense D. odonopteroides D. superbum D. zuberi

Matatiele 30’20'S, 28°49'E Molteno Camian-Norian T. odontopteroidesSeward, 1903

Lady Grey 30°43'S, 27" 13'E Molteno Camian-Norian Thinnfeldia sp. Seward, 1908 (-D . stelznerianum)

Cala R., near Cala 31°00'S, 26°20'E Burgersdorp Scythian Thinnfeldia acuta Seward, 1908; Du (-D . lancifolium) Toit, 1927 T. lancifolia T. sphenopteroides (=D. lancifolia) T able 5 (continued)

Kenigha R., W. of 31°02'S, 28°23'E Molteno Camian-Norian Thinnfeldia lancifolia Seward, 1903; Du Maclear T. odontopteroides Toit, 1927 Stenopteris elongata (=D.elongatum)

Jamestown 31o06'S, 26°45'E Molteno Camian-Norian Thinnfeldia Seward, 1908; Du feistmanteli Toit, 1927 T. odontopteroides

Stormberg 31°16'S, 26° 17'E Molteno Camian-Norian Thinnfeldia Seward, 1903; Du rhomboidalis Toit, 1927 (=D.lancifolia)

Elliot 31°22'S, 27°48'E Molteno Camian-Norian Thinnfeldia acuta Seward, 1908; Du T. feistmanteli Toit, 1927 T. odontopteroides (D.feistmanteli) T. sphenopteroides (=D.lancifolia) Stenopteris densifolia (=D.elongatum)

Queenstown 31°52'S, 26°52'E Molteno Camian-Norian Thinnfeldia acuta DuToit, 1927

Madagascar near Ankavandra, 18°3rs, 45‘29'E Sakamena Scythian D. callipteroides Carpentier, 1935; Morondava Basin Carpentier, 1936

u> ON T able 5 (continued)

Tambohohazo, 20*20'S, 45°36'E Sakamena Scythian D. sp. Carpentier, 1935 near Malaimbandy, ( =D. odontopteroides) Morondava Basin Table 6. Dicroidium species from India.

Locality Coordinates Formation Age Species Reference

Madhya Pradesh

Nidpur, South Rewa 24"07'N, 81054'E Nidpur Anisian-Ladinian D. gopadensis Bose and Srivastava, Basin D. nidpurensis 1971;Srivastava, D. papillosum 1979; Bose et al., 1990

Giar, South Rewa 23°49'58"N, Tiki Norian D. coriaceum Srivastava and Pal, Basin 81°16'14"E D. giarensis 1983; Pal, 1984; cf. D. Boseetal., 1990 odontopteroides D. zuberi

Harai, South Rewa 23“40'53"N, Tiki Camian D. hughesi Sahni and Rao, Basin 8ri2'40"E D. odontopteroides 1958; Pal, 1984; D. sp. Bose et al., 1990 D. zuberi Xylopteris sp.

Chicharia, South 23°30'N, 81e10'E Parsora Anisian-Norian D. hughesi Sahni and Rao, 1958; Rewa Basin D. feistmanteli Lele, 1962; Rao and D. odontopteroides Lele, 1963; Lakhanpal et D. sahnii al., 1976

Bhaursen Hill, 23°27'N, 81°07'E Parsora Anisian-Norian D. feistmanteli Lele, 1962; Lakhanpal et South Rewa Basin al., 1976

u> 00 T able 6 (continued)

Pali, Parsora, South 23°26'N, 81°5'30"E Parsora Anisian-Norian D.feistmanteli Lele, 1956; Sahni and Rewa Basin D. hughesi Rao, 1958; Lele, 1962; D. odontopteroidesLakhanpal etal., 1976; Bose et al., 1990

Kamtadand, South 23'25'N, 81°06'E Parsora Anisian-Norian D. hughesi Saksena, 1962 Rewa Basin

Dhaurai, South 23C23'N, 81°02'E Parsora Anisian-Norian D. hughesi Sahni and Rao, 1958 Rewa Basin D. odontopteroides

Bihar east of Kumarpur, near 23°40'N, Panchet Scythian IDicroidium Baneiji and Bose, 1977 Asansol region, 87°01'E Damodar Basin

Andhya Pradesh

Eddla Gattu, 17°02'N, 81°19'42"E Maleri Camian-Norian D. odontopteroides Baksi, 1967; Lakhanpal Raghavapuram, D. sp. etal., 1976 Godavari Basin

u> vo Table 7. Morphological and cuticular features of Dicroidium from Mt. Falla (collections T-5 and T-7).

Feature D. odontopteroides D. lancifolium D. dubium

Foliage morphology frond arrangement pinnate pinnate pinnate total size (1 x w)l 13-18 cm x 7-9 cm 13-23 cm x 6-14 cm >6->12 cm x >6->9 cm angle of forking 20°-55° (ave.=35°) 17°-55° (ave.= 25°) 35°-40° rachis width 0.5-3 mm 1-3.5 mm 1.2-3 mm pinnule size (1 x w) 0.3-3 cm x 0.35-0.95 cm 0.7-4.8 cm x 0.35-1.1 cm 0.8-4.35 cm x 0.5-1.2 cm pinnule 1: w ratio 0.6-3.5: 1 0.75-7.25: 1 1.3-6: 1 pinnule shape ovate-rhomboid ovate-lanceolate ovate-lanceolate arrangement opposite-alternate opposite-subopposite opposite-subopposite apex obtuse-acute acute-obtuse acute base decurrent-obtuse normal-decurrent normal-decurrent margin entire entire lobed venation odontopteroid-taeniopteroid alethopteroid-odontopteroid alethopteroid

Cuticular features^ rachis thickness 2.4-5.0 pm 2.0-5.5 pm 2.7-5.1 pm pinnule thickness (U/L) 1.5-4.3/1.7-4.4 pm 2.7-5.9/0.8-4.7 pm 1.5-3.4/1.0-2.6 pm epidermal cell shape isodiametric-rectangular isodiametric-rectangular isodiametric-rectangular Table 7 (continued) cell size-laminar region (U) length (j* m) 56.2 ± 7.3 (43.7-73.9) 95.2 ± 20.0 (63.3-138.2) 87.7 ± 19.3 (60.9-135.1) width (/«n) 37.6 ± 8.3 (24.4-60.6) 66.4 ± 12.6 (45.6-95.8) 45.6 ± 8.3 (30.9-60.5) cell size-laminar region (L) length (j4m) 52.6 ± 11.6 (34.9-77.9) 69.7 ± 13.3 (42.2-97.6) 70.3 ± 11.1 (46.5-96.0) width (pm) 33.0 ± 5.9 (25.9-53.6) 47.5 ± 9.9 (27.1-68.6) 42.0 ± 7.8 (29.9-52.9) cell size-over veins length (/

guard cell size length ( j a m) 27.3 ± 5.6 (18.8-36.0) 33.4 ±8.1 (19.7-54.1) 33.1 ± 5.4 (25.7-41.4) width (/

1 Abbreviations: Stomatal index (SI)=S/(E + S) x 100, where S=number of stomata per unit area, E=number of epidermal cells per unit area; l=length; w=width; U=upper surface (adaxial); L=lower surface (abaxial); ave.=average.

^Cuticle measurements based on specimen and slide numbers: 38-3,38-4,38-5; 57A-1,57A-2, 57B-7, 57B-8; 65-1,65-2,65-3, 65-9; 75-1,75-2, 177-2, 177-3; 128-1, 128-2, 128-4. 143

Table 8. Other corystosperm stem and reproductive organs found in Antarctica during the Triassic.

Locality Form ation Age Organ genera Reference

ML Falla Falla Camian-Norian Umkomasia-likc Taylor and organ Taylor, 1988 Pteruchus sp.

Fremouw Peak Fremouw Anisian Rhexoxylon-Wse Taylor, 1992 stem Kykloxylon Meyer-Berthaud fremouwensis etal., 1993 Pteruchus Yaoetal., 1995

Allan Hills Lashly C Camian-Norian Pteruchus-hke Taylor etal., organ 1990 Umkomasia

Priestley Glacier Upper Norian/Rhaetian? Rhexoxylon Seward, 1914; Beacon priestleyi Walton, 1923 Table 9. Dicroidiwn species from all regions of Gondwana during the Early (E), Middle (M), and Late (L) Triassic. Australia includes the regions of Tasmania and New Zealand, and Africa includes the Madagascar region.

SPECIES ANTARCTICA SOUTH AMERICA AUSTRALIA AFRICA INDIA

E M L E M L E M L E M L E M L

D. argentinum XX D. brownii XX X D. cacheutense XX D.callipteroides XX D. crassinervis XXX X X X D. densifolium X XX D. dubium X XX XXX X X D. dutoitii XX XX XXX X D. elongatum XXX XXX X D. eskensis X X D.feismantelii XXX XXX X X D. gouldii X D. groeberi XX D. hughesi XX X D. incisum XX D. intermedium X X

D. lancifolium X XXXX XX XX X 144 Table 9 (continued)

D. narrabeenensis X X X X X X X D. odontopteroides XXX XX X X X X X X X X D. remotum X X D. sahnii X X D. spinifolium X X X X X X D. superbum X X D.stelznerianum X X X X X X X D.tripinnata X X D. trilobitum X X X D. zuberi X X X X X X X X

SPECIES TOTAL 3 9 11 0 19 20 18 12 14 0

& Table 10. Simpson Indices* for Dicroidium species in each region of Gondwana during Early, Middle, and LateTriassic.

South America Australia Africa India

EMLEMLEMLEML

Antarctica 100 91 67 100 82 67 67 91 — 40 60

South America — 67 75 — 100 93 — 80 80

Australia 100 100 75 60 40

Africa — 33 80

*Simpson Index=C/Ni x 100, where C=number of species in common to both regions, Ni=total number of species in region with the smaller number of species. Table 11. Jaccard Indices* for Dicroidium species in each region of Gondwana during Early, Middle, and LateTriassic.

South America Australia Africa India

E M L E M L E M L E M L

Antarctica 47 48 25 50 64 50 20 67 — 17 23

South America — 48 39 — 16 62 — 20 19

Australia 43 17 53 — 15 13

Africa — 14 27

* Jaccard Index=C/(Ni + N 2 - C) x 100, where C=numberof species in common to both regions, Ni=total number of species in region with the smaller number of species, N 2 =total number of species in the second region. APPENDIX B

FIGURES

148 Image acquisition Shape description — — — — — — —— Multivariate analysis Outline Measurement (PCA and Digitization tracing (EFA) cluster analysis) Pinnule >- Digitized Stored >■ Fourier Results outline image outline coefficients 1 Shape display and retrieval (x, y coordinates)

Figure 1. Summary of the steps completed for morphometric analyses of pinnule outlines (modified from White et al., 1988). 1 harmonic 5 harmonics

V,-' 10 harmonics 15 harmonics

20 harmonics 25 harmonics

Figure 2. Reconstructed pinnule outlines from 1 to 25 harmonics using EFA (specimen AF 24). 115.000

120.000 10-25

115.000

110.000

105.000

100.000 40.000 45.000 55.000 60.000

X

Figure 3. Reconstructed pinnule outlines from 1,5,10,15,20 and 25 harmonics (specimen T-l 1 #446b).

2.000 r

1.000 J&SsV . •-..7 “hViST* 0.000 * •. ■ jT.I

• 1.000

- 2.000

-3.000

-4.000 •3.000 - 2.000 • 1.000 0.000 1.000 2.000

FACTOR 1

Figure 4. PCA factor scores from Fourier coefficients of 330 pinnules (data set A). Figure 6. PCA factor scores from Fourier coefficients of 75 pinnules pinnules 75 of coefficients Fourier from scores factor PCA 6. Figure Figure 5. PCA factor scores and cluster analysis results of pinnule subset B. subset pinnule of results analysis cluster and scores factor PCA 5. Figure a g g FACTOR 2 « above the dichotomy (data set C). set (data the dichotomy above -4.000 ■3.000 ~D.odontopteroides 2.000 000 0 .0 1 0.000 2.000 1.000 • 2.000 D. crassinervis - 1.000 •u ATR 1 FACTOR ATR 1 FACTOR M 0.000 1.000 D. lancifoliumD. D. stelmerianum 2.000 152 153 3.ooo r

2.000

t.0 0 0

• 2.000

-3.000 1------1 1 ■■ — ------* -3X00 -2.000 -1.000 0.000 1.000 2.000

FACTOR 1

Figure 7. PCA factor scores from Fourier coefficients of 75 pinnules below the dichotomy (data set D).

3.000 A D. spinifolium

D. lancifolium D. dubium 1.000 ] D.trilobitum

D. stelznerianum 0.000

D. elongation - 1.000

D. odontopteroides • 2.000 • 2.000 • 1.000 0.000 1.000 2X00

FACTOR 1

Figure 8. PCA factor scores from Fourier coefficients of pinnule subset E from above the dichotomy. Circled regions represent clusters of similar shapes supported by cluster analysis. Figure 9. Dicroidiumcrassinervis from Gordon Valley (T-4 #33) (scale bar=l cm).

Figure 10. D. dubium from Mt. Falla (T-5 #128) (scale bai^l.25 cm).

Figure 11. D. elongatum from Allan Hills (T-l 1 #417) (scale ban=l cm).

Figure 12. D. dutoitii from Allan Hills (T-8#271) (scale bar=l cm).

154 155

Figure 9. Figure 10.

Figure 11. Figure 12. Figure 13. D.feistmantelii from Mt. Falla (Ant 70-8-164) (scale bai=0.56 cm).

Figure 14. D. lancifolium from M t Falla (T-5 #38) (scale bar=2 cm).

Figure 15. Detail of alethopteroid venation from D. lancifolium (T-5 #50) (scale bar=0.75 cm).

Figure 16. D. odontopteroides from Mt. Bumstead (Ant 67-9-65a) (scale bar=2 cm).

156

Figure 17. D. odontopteroides from Allan Hills (T-l 1 #769) (scale bar=0.5 cm).

Figure 18. Odontopteroid venation from D. odontopteroides distal regions with fused pinnules (T-7 #162b) (scale bar=0.35 cm).

Figure 19. D. spinifolium from Allan Hills (T -ll #435b) (scale bar=l cm).

Figure 20. Venation of D. spinifolium (T -ll #423b) (scale bar=0.5 cm).

158 159

V'-"''

Figure 17. Figure 18.

Figure 19. Figure 20. Figure 21. D. stelznerianum from Gordon Valley (T-4 #33) (scale ban=l cm).

Figure 22. Detail of venation from the apex of a D. stelznerianum frond (T-8 #188) (scale bar=0.6 cm).

Figure 23. D. trilobitum from Allan Hills (T-l 1 #418a) (scale bar=l cm).

Figure 24. D. zuberi from RoscolynTor, Allan Hills (T -ll #855) (scale bar=3.4cm).

160 161

Figure 21. Figure 22. Figure 23. Figure 24. Figures 25-28. Cuticle of D. odontopteroides. Figure 25. Cuticle illustrating the epidermal cell pattern and orientation of stomatal complexes. Three complexes are indicated with arrows (T-5 #72; Slide no. 72-1) (scale bar=100 pm ).

Figure 26. Cuticle with papillae on subsidiary cells surrounding the stoma. Two stomatal complexes are indicated with arrows (T-5 #103; Slide no. 103-6) (scale bar=70 pm ).

Figure 27. The outer surface of a cuticle piece in a region with continuous papillae (T-7 #186) (scale bai^70 pm ).

Figure 28. Stomatal complex illustrating paracytic arrangement (T-5 #72; Slide no. 72-1) (scale ban=35 pm ).

162 163

Figure 25. Figure 26.

Figure 27. Figure 28. Figures 29-32. Cuticle of D. odontopteroides. Figure 29. Inner surface of cuticle showing monocyclic stomatal complex and thickened guard cells (T-5 #65) (scale bar=35 pm).

Figure 30. Inner surface of cuticle with rectangular cells and a stomatal complex oriented parallel to veins (T-5 #65) (scale bar=35 pm ).

Figure 31. Outer surface of cuticle with papillae and one stomatal pore opening indicated with an arrow (T-5 #103) (scale bar=35pm ).

Figure 32. Outer surface of cuticle illustrating two pore openings, both indicated with arrows (T-5 #72) (scale ban=35 pm ).

164 Figure 29. Figure 30.

Figure 31. Figure 32. Figures 33-36. Cuticle of D. odontopteroides. Figure 33. Inner surface of a stomatal complex with thickened aperture and guard cell thickenings (T-5 #65) (scale bar=10 pm ).

Figure 34. Inner surface of an open stomatal complex with thickened guard cells (T-5 #103) (scale bar=10pm ).

Figure 35. Inner surface of a stomatal complex with striated guard cell thickenings (T-5 #81a) (scale bar=10 pm ).

Figure 36. Inner surface of a trichome base (T-5 #72) (scale bar=20 pm ).

166 Figure 33. Figure 34.

Figure 35. Figure 36. Figures 37-40. Cuticle of D. lancifolium. Figure 37. Cuticle with two monocyclic stomatal complexes indicated with arrows (T-5 #38; Slide no. 38-4) (scale ban=50 pirn).

Figure 38. Epidermal cells and stomatal complex with orientation along axis of veins (T-5 #78; Slide no. 78-5) (scale bar=24 f*m).

Figure 39. Inner surface illustrating a stomatal complex (T-5 #38) (scale bar=35 ptm).

Figure 40. Inner surface of a stomatal complex with guard cell thickenings (T-5 #38) (scale bar=10 pim).

168 Figure 37. Figure 38.

Figure 39. Figure 40. Figures 41-44. Cuticle of D. dubium. Figure 41. Stomatal complexes illustrating two different subsidiary cell arrangements. The complex on the upper left (see arrow) has two divided lateral subsidiary cells, a second complex has undivided lateral cells (see arrow, lower right) (T-5 #128; Slide no. 128-1) (scale bar=35 /«n).

Figure 42. Cuticle with papillae and paracytic stomatal complex indicated by an arrow (T-5 #128; Slide no. 128-1) (scale bar=35 /

Figure 43. Inner surface of a stomatal complex and epidermal cells (T-5 #128) (scale bar=35 /

Figure 44. Inner surface of a stomatal complex illustrating guard cell and aperture thickenings (T-5 #128) (scale bar=10 /mi).

170 171

Figure 41. Figure 42.

Figure 43. Figure 44. Figures 45-48. Cuticle from D. dutoitii specimens from Argentina. Figure 45. Specimen of D. dutoitii providing cuticular material (USNM 40357) (scale bar=l cm).

Figure 46. Inner surface of cuticle showing three stomatal complexes indicated by arrows (USNM 40357) (scale bar=60 }4 m).

Figure 47. Inner surface of a stomatal complex with guard cell thickenings and subsidiary cells (USNM 40357) (scale bar=20 ptm).

Figure 48. Stomatal complex with three subsidiary cells and polar extensions of guard cell thickenings (USNM 40357; Slide no. 40357-4) (scale bar=20 ptm).

172 173

Figure 45. Figure 46.

Figure 47. Figure 48. Figures 49-50. Cuticle of D. dutoitii. Figure 49. Inner surface of a stomatal complex (USNM 40357) (scale bar=20 ptm).

Figure 50. Inner surface of a stomatal complex with striated guard cell thickenings and well-defined polar regions of guard cells (USNM 40357) (scale bar=10 ptm).

Figures 51-52. Cuticle of D. dubium from Australia. Figure 51. Specimen of D. dubium providing cuticular material (USNM 458659) (scale bar=1.5 cm).

Figure 52. Cuticular surface illustrating several stomatal complexes (USNM 458659; Slide no. 458659-3) (scale bar=100 ptm).

174 175

Tgure 49.

Figure 51. Figure 52. Figures 53-56. Cuticle of D. dubium. Figure 53. Inner surface with several stomatal complexes (USNM 458659) (scale bar=60 pm ).

Figure 54. Inner surface of a paracytic stomatal complex (USNM 458666) (scale bar=20 m ) - Figure 55. Stomatal complex with paracytic subsidiary cells and guard cell thickenings (USNM 458661; Slide no. 458661-1) (scale bar=20 jam).

Figure 56. Stomatal complex with adjacent subsidiary cells surrounded by encircling cells (USNM 458666; Slide no. 458666-2) (scale ban=20 pm ).

176 177

Figure 53. Figure 54.

Figure 55. Figure 56. Figures 57-61. Cuticle of D. zuberi. Figure 57. Cuticle with sinuous anticlinal walls and monocyclic stomatal complexes. Three complexes show a ring of thickened cuticle around the pore opening (see arrows) (USNM 458665A; Slide no. 458665A-1) (scale bar=50 pim).

Figure 58. Inner surface of cuticle illustrating sinuous anticlinal walls (USNM 458665A) (scale bar=30 pim).

Figure 59. Outer surface of cuticle with prominent papillae (USNM 458665A) (scale bar=30 /mi).

Figure 60. Inner surface of cuticle illustrating papillae regions (USNM 458665A) (scale bar=30 pim).

Figure 61. Inner surface of a stomatal complex with thickened guard cell walls (USNM 458665A) (scale ban=10 pim).

Figure 62. Cuticle of D. odontopteroides. Inner surface of a stomatal complex with monocyclic subsidiary cell arrangement and thickened guard cells (USNM 458663) (scale bar=20 pim).

178 Figure 57. Figure 58. Figure 59. Figure 61. Figure 60. Figure 62. 180

Weddell Sea

Ronne WF ANTARCTIC -90*W 9 0 ? E Ice CRATON Shelf %6o*w 8 0 °S

Pensacola Mtns

Ellsworth Mtns 90°W -(- eo°s

0 " 'a rc,/ • Nilsen • Plateau Shackleton Glacier 80 °S l2 0 «w >V Beardmare Glacier % I20«E

Nimrod Glacier Byrd Glacier

Shelf IS 0- 6 'U 7 O £ O Mockoy Glacier yX AAllon Hills ^ Victoria Ross Sea Da rid Glacier ^ Land 01 Priestley Glacier

Rennick Glacier

Figure 63. Map of Antarctica illustrating localities in the Transantarctic Mountains (modified from Collinson et a!., 1986). Figures 64-67. Dicroidium specimens from Mt. Falla, central Transantarctic Mountains. Figure 64. D. odontopteroides from 135 m above the base of the type section (T-5 #103) (scale bar=2 cm).

Figure 65. D. lancifolium from 135+ m above the base of the type section (T-7 #177) (scale bar=3 cm).

Figure 66. D. odontopteroides from about 120 m above the base of the type section (Ant 70-8-170) (scale bar=2.5 cm).

Figure 67. D. dubium from approximately 130 m above the base of the type section (Ant 70-8-31) (scale bar=1.2 cm).

181 182

Figure 64. Figure 65. Figure 66. Figure 67. Figure 68. D. zuberi from Mt. Falla (Ant 70-8-52) (scale ban=3.75 cm).

Figure 69. D. stelznerianum from Gordon Valley (T-4 #29b) (scale bar=l cm).

Figure 70. Several specimens of D. stelznerianum from Gordon Valley (T-4 #31) (scale bar=1.25 cm).

Figure 71. D. odontopteroides from Fremouw Peak (T-10 # 292) (scale bar=l cm).

183 184

Figure 68. Figure 69. Figure 70. Figure 71. Figure 72. D. dubium from Fremouw Peak (Ant 70-9-143) (scale bar=l cm).

Figure 73. D. odontopteroides and D. elongatum (arrow, right) from Mt. Bumstead (Ant 67-9-52a) (scale bar=1.5 cm).

Figure 74. D. stelznerianum from Mt. Bumstead (Ant 67-9-60a) (scale bar=2 cm).

Figure 75. D. zuberi from Mt. Bumstead illustrating venation (Ant 67-9-140) (scale bar=1.25 cm).

185 186

Figure 72. Figure 73. Figure 74. Figure 75. Figure 76. D. crassinervis from Mt. Wisting (Ant71-l-7) (scale bar=l cm).

Figure 77. D. odontopteroides from Mt. Wisting (Ant 71-1-21) (scale bar=1.5 cm).

Figure 78. Two specimens of D. stelznerianum with rhomboid pinnules from Shapeless Mountain (AF 9/1) (scale bar=2 cm).

Figure 79. D. lancifolium from Horseshoe Mountain (AF 315) (scale bar=2 cm).

187 188

Figure 76. Figure 77. Figure 78. Figure 79. 189

Figure 80. Locality site in Feather Bay, Allan Hills. Plant levels 1-4 are located between the arrows (approximately 20 m) within the Lashly Formation, Member C. Figures 81-84. Dicroidium specimens from Allan Hills, southern Victoria Land. Figure 81. D. odontopteroides from Level 1 (T-11 #490) (scale bar=l cm).

Figure 82. Several specimens of D. dutoitii from Level 2 (T-8 # 275b) (scale bar=2 cm).

Figure 83. Specimens of D. dutoitii from Level 2 illustrating possible expanding immature foliage (T-8 #255) (scale bar=l cm).

Figure 84. D. trilobitum and D. elongatum (right) from Level 4 (T-11# 442b) (scale bar=l cm).

190 191

Figure 81. Figure 82.

Figure 83. Figure 84. Figures 85-88. Dicroidium specimens from Allan Hills, southern Victoria Land. Figure 85. D. spinifotium from Level 4 (T-l 1 #450a) (scale bar=l cm).

Figure 86. D. elongatum from Level 4 (T-l 1 #451) (scale bar=1.25 cm).

Figure 87. D. stelznerianum with rhomboid pinnules from North Roscolyn Tor, Allan Hills (T-l 1 #783) (scale bar=2 cm).

Figure 88. D. lancifolium from South Roscolyn Tor, Allan Hills (T-l 1 #876) (scale bar=2 cm).

192 193

Figure 85. Figure 86. Figure 87. Figure 88.

■*S>

■2000 km at Equator

Figure 90. Paleomap of Gondwana during the Early Triassic generated using the program PGIS/Mac™ (Scythian, 243 mya). Early Triassic Dicroidium specimens have been found at the localities indicated by points. ■2000 km at Equator

Figure 91. Paleomap of Gondwana during the Middle Triassic generated using the program PGIS/Mac™ (Anisian, 235 mya). Middle Triassic Dicroidium specimens have been found at the localities indicated by points. ~ 2 0 0 0 ton a t Equator

Figure 92. Paleomap of Gondwana during the Late Triassic generated using the program PGIS/Mac™ (Norian, 220 mya). Late Triassic Dicroidium specimens have been found at die localities indicated by points. Figure 93. Paleomap of Gondwana during the Anisian, summer season, indicating general precipitation amounts, generated using PGIS/Mac™ and Paleoclimate™. Areas of high precipitation are redicted in regions marked "H", areas of low precipitation are indicated in regions marked "L". Points represent localities where Dicroidium has been found. ^ I

"■3000 km at Equator

Figure 94. PaleomapPaleomap of of Gondwana Gondwana during during the Anisian,the Anisian, summersummer season,season indicatingl» indicating relative relativetemperaturetemperature levels.levels, generated using‘ PGIS/Mac™ PGIS/Mac™ and andPaleoclimate™. Paleoclimate™ Areas of high temperature arepredictediredicted in f*J*mr\ne mortrArl nraor /%f Inm tA m nam hira o v a in<4«/inf/u4 rnniM n«1rA /4 (IT ft represent localities where Dicroidium has been found. 199 H HHH

~3000 km at Equator

Figure 95. urine the Anisian, winter season, indicating genera! precipitation [S/Mac™ and Paleoclimate™. Areas of high precipitation are iredicted in regions marked "I H", areas of low precipitation are indicateain regions marked "L". ’oints represent localities wher t Dicroidium has been found. -3000 km at Equator

Figure 96. Paleomap of Gondwana duringthe Anisian, winter season, indicating relative temperature levels, generated using PGIS/Mac™ and Paleoclimate™. Areas of high temperature are predicted m regions marked "H , areas of low temperature are indicated in regions marked "L". Points represent localities where Dicroidium has been found. ------T ' > i

( ■ --- / ■ ■■ J h //LL V l f J H /

LL p L L l U - t S t "** \ f L V | | LLLL LL .. L L L L L L L jM .L L [ * L L y T i i v V 1 f t . L L f L N ■> L / i f L • / • 0 • L L f \ * r f •I i j 11 m • A ^ A 11/ H i / HHHHHH V k L ^ 6 HHHHh HH HHHHHH ^ ^ H H H H H I - HH

H L -3000 km at Equator

Figure 97. Paleomap of Gondwana during the Norian, summer season, indicating general precipitation amounts, generated using PGIS/Mac™ and Paleoclimate™. Areas of high precipitation are redicted in regions marked "H", areas of low precipitation are indicated in regions marked "L". Points represent localities where Dicroidium has been found. 9905^0 \ H H4 HHHHHH HIp \ . H 4 HHHHhl HbTjM 4HHHH ( / h h h m u HHHH knIHHHHH a M 1 HHH ^ h 4 w HHI- IHHHHV : • H H l f e + 4 )j • j V • • i • \ H M i \ * • \ * •L / H / it* H l l t j j j ' 1 | I /

• • ' ] |||1 f • A Ml • X

V V 1 •

[

•>3000 km at Equator 5< L tUK i

Figure 98. Paleomap of Gondwana duringlhe Norian, summer season, indicating relative temperature levels, generated using PGIS/Mac™ and Paleoclimate™. Areas of high temperature arepredicted in regions marked "H , areas of low temperature are indicated in regions marked "L". Points represent localities where Dicroidium has been found. L LLLi LLLLLL, ^LLl \H L (. >r /LLL >LL7. L I/ 1 ' ^ NT 4 j/! H H H jp \ H \ hhhi-H / HHH ' ¥ h • ■ i - m • j L Li", \Ll \ * I «C^/

VVi • ^ f [ -3000 km at Equator Lr

Figure 99. Paleomap of Gondwana during the Norian, winter season, indicating general precipitation amounts, generated using PGIS/Mac™ and Paleoclimate™. Areas of high precipitation are redicted in regions marked "H", areas of low precipitation are indicatedm regions marked "L" Points represent localities where Dicroidium has been found. -3000 km at Equator

Figure 100.100 . PaleomapPaleomap of of Gondwana Gondwana during during the the Norian,Norian, winter■» winter season,season, indicatingindicating relative relative temperature temperature levels.Ia v a Ic ootiAratAH ncinoncino PTJIQ/K/fijp™ PftT.Q/Mnn™PGIS/Mac™ anHan, and PalpnTlim Paleoclimate™ atp™ . AAreas rpac nof f highhi oh ff»mnpratiirptemperature ; are nredir predicted represent localities where Dicroidium has been found. — i ii ——Antarctica 206

i——— — Australia

I— — — ^ Africa

'■"■■■" Outgroup

Figure 101. PAE cladogram results of taxa present during the Early Triassic.

Antarctica South America Australia India Africa Outgroup

Antarctica South America Australia Africa India Outgroup

Figure 102. The two most parsimonious PAE cladgrams for the Middle Triassic data set Antarctica Australia South America Africa India Outgroup *

Antarctica South America Africa Australia India Outgroup

Antarctica South America Africa Australia India Outgroup

Figure 103. The three most parsimonious PAE cladograms for the Late Triassic data set. APPENDIX C

DATA RELATIVETOCHAPTERIII

The pinnules used in morphometric analyses were from the following specimens;

Data set A: AF 16a, 16b, 24, 26, 29a, 29b, 309, 30a, 30b, 311a, 311b, 311c, 315, 9a and 9b;

Data set B; AF 16b, 24, 26, 29a, 29b, 309,30b, 311a, 311b, 315, 9a and 9b;

Data set C and D: T -ll #312a, 418,442, 446b, 451,467a, 479a, 490, 491, 493, 699a, 722a, 749a, 752,782, 783, 851; T-3#7; T-5 #38,72, 103, 128; T-7#177; Ant 70-8-143,70-8-170; T-10 #292,302; T-8#189; T-4 #29b, 31-1, 31-2,31-3,33a, 33b; AF 16b, 24, 26, 29b, 309, 311a, 9a, 9b; USNM 33024,40a, 40b, 40c, 40354,40355-1,40355-2; Museo Argentino de Ciendas Naturales Bernardino Rivadavia #3960,4118, 4138, 4180, 4183,43-14, 43-24, 5046, 5056, 5075, 5081, 54, 7522, 7536; Department of Geology, Ciudad Universitaria, Buenos Aires #7435; PRINGEPA, Corrientes #2313,341,35-10, 352, 39-21, 39-8, 4330, 8179, 8221, 8288, and 8290; and Data set E: T-ll #451, 493, 699a, 722a, 749a, 752, 783, 851; T-5 #38; T-10 #292,302; T-8#189; T-4#29b, 31,33b; USNM 40a, 40c, 40354; AF 26, and 311a.

208 APPENDIX D

DATA RELATI VETO CHAPTER VII

Data matrices for Early, Middle and Late Triassic PAE analyses are as follows: for the Early Triassic; dimensions ntax=4 nchar=4; matrix Antarctica 0110 Australia 1111 Africa 1111 Outgroup 0000 for the Middle Triassic; dimensions ntax=6 nchar=12; matrix Antarctica 011111010011 South America 111111111111 Australia 111111011111 Africa 000000001001 India 100001100000 Outgroup 000000000000 and for the Late Triassic; dimensions ntax=6 nchar=12; matrix Antarctica 101110101111 SouthAmerica 111111111101 Australia 001110111110 Africa 111111111101 India 000001000001 Outgroup 000000000000.

209