UMI MICROFILMED 1989 INFORMATION TO USERS The most advanced technology has been used to photo­ graph and reproduce this manuscript 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. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print 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 re­ produced 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. These are also available as one exposure on a standard 35mm slide or as a 17" x 23" black and white photographic print for an additional charge. 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.

University Microfilms International A Bell & Howell Information Com pany 300 North Z eeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600 I Order Number 8013706

Comparative analysis of sporodcrm ultrastructure in fossil and extant lycopods

Taylor, Wilson Anthony, Ph.D. The Ohio State University, 1989

UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106

COMPARATIVE ANALYSIS OF SPORODERM

ULTRASTRUCTURE IN FOSSIL AND EXTANT LYCOPODS

DISSERTATION

Presented in partial fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Wilson Anthony Taylor, B.S., M.S.

The Ohio State University

1989

Dissertation committee: Approved by

T.N. Taylor

V. Raghavan

D.J. Crawford Adviser I F.D. Sack Departmentleeartment of BotanyRotanv* To Dad from Boo

ii ACKNOWLEDGEMENTS

I would like to thank the members of my reading committee, Drs. Daniel Crawford, Valayamghat Raghavan, and

Fred Sack, for their time and effort. I am also grateful to the staff at the Royal Botanic Gardens, Kew, for their support during specimen collection and transport. I am especially indebted to Dr. Thomas N. Taylor for his financial support and scholarly guidance throughout all phases of this investigation. And finally, I recognize the loving suport of my wife, Robin, without whom none of this would have been possible.

iii VITA

June 3, 1960 ...... Born - Indiana, Pennsylvania

1982 ...... B.S., Geoscience Department, Indiana University of Pennsylvania, Indiana, Pennsylvania

1984 ...... M.S., Department of Botany, The Ohio State University, Columbus, Ohio

Field of Study: Botany

PUBLICATIONS

1986a Ultrastructure of Sphenophyllalean spores. Review of Palaeobotany and Palynology 47y 105-128.

1986b Spore wall ultrastructure of Protosalyinia. American Journal of Botany 74? 437-443 (with T.N Taylor).

1987 Subunit construction of the spore wall in fossil and living lycopods. Pollen et Spores 29? 241-248. (with T.N. Taylor)

1988 Ultrastructural analysis of selected Cretaceous megaspores from Argentina. Journal of Micropalaeontology 7? 73-87. (with T.N. Taylor)

1989a Comparative ultrastructural analysis of fossil and living gymnoBperm cuticles. Review of Paoaeobotany and Palynology (with T.N. Taylor)

1989b Comparative ultrastructural analysis of microspores from fossil and living Selaginella. (with T.N. Taylor)

1989c Ultrastructural and developmental analysis of Carboniferous lycopod megaspore walls. Review of Palaeobotany and Palynology. 1984a Ultrastructure of sphenopsid spores. Ohio Journal of Science 84; 7.

1984b Spore wall ultrastructure in the Sphenophyllales. 2nd International Organization of Paleobotany Conference, Edmonton, Alberta. Abstracts of Contributed Papers, Poster Sessions.

1984c An ultrastructural investigation of sphenophyllalean spores. American Association of Stratigraphic Palynologists, Arlington, Virginia. AASP, Inc. Program and Abstracts; 21.

1985a Spore ultrastructure of the enigmatic alga Protosalyinia. Ohio Journal of Science 85; 18. (with T.N. Taylor)

1985b Spore and thallus ultrastructure of the enigmatic alga Protosalvinia. American Journal of Botany 72; 902.

1985c An ultrastructural comparison of Cretaceous and extant ginkgoalean cuticles. Botanical Society of America, Gainesville, Florida.

1986a Spore wall architecture of a megaspore from the Cretaceous of Argentina. Mid-continent Paleobotanical Colloquium IV, Ann Arbor, Michigan.

1986b Ultrastructural stasis within the Selaginellales. American Journal of Botany 73; 712.

1987 Evolutionary and developmental significance of megaspore wall ultrastructure. American Journal of Botany 74; 692. (Cookson Award)

1988 Developmental and structural aspects of lycopod megaspore wall ultrastructure. American Journal of Botany 75(6) Part 2; 50-51.

v TABLE OF CONTENTS

ACKNOWLEDGEMENTS...... ili

VITA ...... iv

LIST OF T A B L E S ...... ix LIST OF P L A T E S ...... X

CHAPTER I INTRODUCTION ...... 1 Literature survey ...... 4

CHAPTER II MATERIALS AND METHODS ...... 16 M a t e r i a l s ...... 16 M e t h o d s ...... 17 Terminology ...... 22

CHAPTER III WALL DEVELOPMENT IN FOSSIL SYSTEMS . . . 24 Introduction...... 24 R e s u l t s ...... 28 Mazocarpon - functional megaspores ...... 29 Mazocarpon aborted megaspores ...... 30 Lepidocarpon functional megaspores ...... 36 Lepidocarpon aborted megaspores ...... 36 Lagenoisporites ...... 40 Valvisisporites ...... 44 Discussion...... 46 Mazocarpon...... 46 Valvisisporites ...... 48 Lepidocarpon...... 49 Lagenoisporites ...... 51 Morphogenetic changes through wall development 52 Correlation of ultrastructure with dispersive s t r a t e g y ...... 53

CHAPTER IV WALL ULTRASTRUCTURE IN SELAGINELLA . . . 56 Introduction ...... 56 Terms and concepts...... 59

vi Species descriptions ...... 65 Selaginella flabellata type ...... 65 Selaginella flabellata ...... 65 Selaginella u s t a ...... 66 Selaginella viridangula ...... 66 Selaginella inaequalifolia type ...... 72 Selaginella inaequalifolia ...... 72 Selaginella pallescens ...... 73 Selaginella pulcherrima ...... 74 Selaginella argentea type ...... 82 Selaginella argentea ...... 82 Selaginella ornata type ...... 83 Selaginella ornata ...... 83 Selaginella erythropus type ...... 87 Selaginella erythropus ...... 87 Selaginella pilifera ...... 88 Selaginella brevipes type ...... 92 Selaginella brevipes ...... 92 Selaginella elmeri ...... 93 Selaginella plana ...... 93 Selaginella fissidentoides type ...... 97 Selaginella fissidentoides ...... 97 Selaginella yemensis ...... 100 Selaginella intermedia type ...... 101 Selaginella intermedia ...... 101 Selaginella frondosa type ...... 104 Selaginella frondosa ...... 104 Selaginella lyallii type ...... 108 Selaginella lyallii ...... 108 Selaginella willdenovii ...... 109 D i s c u s s i o n ...... 115 Observations on basic megaspore types .... 117 Internal wall l a y e r s...... 119 Surface layers ...... 120

CHAPTER V WALL DEVELOPMENT IN SELAGINELLA .... 123 Introduction ...... 123 Description...... 128 Selaginella sulcata ...... 128 Selaginella galeottii .... 136 D i s c u s s i o n ...... 157 S. sulcata - wall d e v e l o p m e n t ...... 157 Structural basis for expansion ...... 158 Structural basis for unit enlargement .... 159 S. galeottii - wall d e v e l o p m e n t ...... 160 Published descriptions of mature megaspores of S, galeottii ...... 164 The role of the t a p e t u m ...... 166 The m e s o s p o r e ...... 168 Control of sporoderm patterning - sporophytic or gametophytic ...... 170

vii Concordance with proposed models of sporoderm morphogenesis ...... 170 Correlations with the ordered wall construction t y p e ...... 173 Implications on the nature of sporopollenin . 174

CONCLUSIONS ...... 176

APPENDIX A - GLOSSARY OF T E R M S ...... 182

APPENDIX B - LIST OF SPECIMENS...... 185

LIST OF REFERENCES...... 186

viii LIST OF TABLES

Table 1. Megaspore size, wall and unit thickness, and construction type of nineteen species of Selaginella...... 116

I ix LIST OF PLATES

Plate I. Stereo pairs Mazocarpon functional, fig. 1 Mazocarpon aborted, figs. 2, 3 32

Plate II. Mazocarpon functional 34

Plate III. Mazocarpon aborted, figs. 10, 11 Lepidocarpon aborted, figs. 12-14 Lepidocarpon functional, figs. 15, 16 38

Plate IV. Lagenoisporites, figs. 17-19, 23 Valvisisporites, figs. 20-22, 24 . . 42

Plate V. Diagrammatic representations of basic megaspore wall construction types . 63

Plate VI. Selaginella flabellata 68

Plate VII. Selaginella usta, figs. 30-32 Selaginella viridangula, figs. 33-35 . 70

Plate VIII. Selaginella inaequalifolia, figs. 36-39 Selaginella pallescens, figs. 40, 41 76

Plate IX. Selaginella pallescens, figs. 42-44 Selaginella pulcherrima, fig. 45 . . . 78

Plate X. Selaginella pulcherrima, figs. 46-48 Selaginella argentea, figs. 49-51 . . 80

Plate XI. Selaginella ornata, figs. 52-56 Selaginella erythropus, figs 57, 58 85

Plate XII. Selaginella erythropus, figs. 59, 60 Selaginella pilifera, figs. 62-64 Selaginella brevipes, figs. 65-68 . . 90

Plate XIII. Selaginella elmeri, figs. 69-71 Selaginella plana, figs. 72-24 .... 95

Plate XIV. Selaginella fissidentoides, figs. 75-78 Selaginella yemensis, figs. 79-81 . . 98 Plate XV. Selaginella intermedia, figs. 82-85 Selaginella frondosa, fig. 86 . . . . 102

Plate XVI. Selaginella frondosa, figs. 87-89 Selaginella lyallii, figs. 90-92 . . . 106

Plate XVII. Selaginella lyallii, fig. 93 Selaginella willdenovii, figs. 94-97 . 111

Plate XVIII. Selaginella willdenovii ...... 113

Plate XIX. Selaginella sulcata ...... 133

Plate XX. Selaginella galeottii, early stages 139

Plate XXI. Selaginella galeottii, stage 4 . . . 145

Plate XXII. Selaginella galeottii, stage 5 . . . 150

Plate XXIII. Selaginella galeottii, stage 5, fig. 143; stage 6, figs. 144-150 . . . 155 CHAPTER I

INTRODUCTION

For many years, the study of fossil plants has dealt largely with the description and identification of individual plant organs. One of the major revolutions in paleobotany involved the reconstruction of whole plants from isolated parts based on associated characteristic features (e.g., Oliver & Scott, 1904). Thus began a more synthetic approach to the study of fossil plants which incorporated a consideration of the fossils as living organisms functioning within the framework of a dynamic ecosystem. This allowed aspects of plant biology, such as pollination biology (e.g., Crepet, 1979), paleoecology

(e.g., DiMichele & Phillips, 1985), and plant-animal interaction (e.g., Scott & Taylor, 1983) to be addressed.

All of these areas provide valuable insight to their respective areas of ultimate (evolutionary) causation, but the proximate causative factors (e.g. physiology, development) have not been widely investigated to date.

A wealth of developmental information can be gathered by studying the preservable products which are secreted or deposited by cytoplasmic components which are not usually

1 2 preserved in the fossil record. This approach has proven effective in dealing with ephemeral tissue systems and their derivative tissues (e.g., vascular cambium; Cichan,

1985a, 1985b; phloem; Smoot, 1984a, 1984b). The area with perhaps the broadest potential applicability in this regard involves sporogenesis in fossil plants. Here the preservable derivative is sporopollenin, an oxidative polymer of carotenoids and/or carotenoid esters (Brooks &

Shaw, 1968a, 1968b), which makes up the resistant walls of pollen and spores. Under either of the following two conditions, developmental information can be acquired directly from fossil plant material; 1) fertile areas which possess sequentially developing sporangia (Chapter III,

Lagenoisporites) . and 2) sporangia which possess abortive meiotic products, where the undeveloped member may record an early ontogenetic stage (Chapter III, Mazocarpon.

Lepidocarpon1.

An alternative and highly promising avenue of investigation involves correlating a fossil with an extant system, and, in turn, gathering developmental and functional information. The ultimate goal of this approach is to use the information from the modern analog to infer the morphogenetic processes which may have been operative in the fossil system. Carefully choosing a dispersed spore taxon with a characteristic organization also present in a modern analog allows accurate assessment of the processes 3 under investigation. This, in turn, allows consideration of meaningful biological questions previously unapproachable due to the .lack of information available on the parent plant. One such system exists in both a dispersed megaspore type from Cretaceous sediments and modern Selaginella. The megaspore wall construction type present in both is sufficiently distinctive to allow accurate correlation of processes which can be investigated in the living system. A detailed analysis of wall development in this megaspore type is presented (Chapter V;

Sj. galeottii) in order to address various questions pertaining to the mechanics of this developmental system.

In addition, a developmental analysis of the megaspore wall of Selaginella sulcata (Chapter V) was undertaken as an assessment of the generality of the processes operative in

S*. galeottii. It was also decided that a survey of wall construction types present in the genus Selaginella

(Chapter IV) would prove useful as a setting in which to develop a set of concepts and terms which can be used to articulate the sometimes subtle differences which distinguish the various types of "spongy" walls.

Therefore, the objectives of this investigation are;

1) to detail ultrastructural changes which take place in lycopod megaspore walls from the time the strata are initiated through megaspore maturity, 2) to survey a number of different megaspore wall types present within the genus 4 Selaginella in an attempt to establish if the investigated processes are consistent throughout the genus, and 3) to sample the structural types present in Pennsylvanian lycopod systems and see what can be surmised regarding their developmental processes and evolutionary controls.

Literature survey

The surface morphology of pteridophyte spores has a genetic component which has proven to be an effective gauge of taxonomic relationship. This fact has been exploited at the level of the light microscope. Historically, the majority of treatments were aimed at producing a general overview of pteridop.hyte spores, most commonly the ferns, of a particular geographic region, with or without testing existing taxonomic schemes (Greguss, 1941; Lugardon, 1963;

Tardieu-Blot, 1963a, 1963b, 1964, 1965, 1966; Nayar, et al,, 1964; Nayar, 1964; Nayar & Lata, 1965; Nayar & Devi,

1963, 1964a, 1964b, 1964c, 1965, 1966, 1967; Harris, 1955;

Kawasaki, 1970; Knox, 1951; Sorsa, 1964; Anonymous, 1976;

Huang, 1981; Chen & Huang, 1974; Murillo & Bless, 1974,

1978) . The comprehensive survey provided by Nayar and his colleagues waB incorporated into a taxonomic scheme of the ferns (Nayar, 1970). Others have provided an analysis of particular genera (Cystopteris, Hagenau, 1961; Dryopteris,

Britton, 1968; Diellia, Wagner, 1952), particular families

(e.g., Osmundaceae, Peabody, 1964), or an overview of the 5 pteridophytes as a whole (Erdtman, 1957; Kremp & Kawasaki,

1972; Knox, 1938; Bir, 1977).

The full taxonomic potential of pteridophyte spore morphology was not realized until the application of scanning electron microscopy permitted a more detailed analysis of the taxonomic relationships of selected fern genera and species (Gvmnocarpium. Sorsa, 1980; Cvstopteris.

Pearman, 1976; Bolbitis. Hennipman, 1977; Bommeria. Haufler

& Gastony, 1978; Gvmnogramme and Scolopendrium. LeCoq, et al., 1973; Thelvpteris palustris. Tryon, 1971; Tryon, et al., 1980; Eellaea atroourpurea, Tryon, 1972; Trichipteris.

Gastony, 1979; Esteves, et al., 1985; Drvopterls. Britton,

1968; Blechnum. Morbelli, 1974; Grammitis. Wagner, 1985;

Platvzoma, Wollenweber, et al., 1987; Cvclodium. Smith,

1986; Elophoglossum, Mickel & Atehortua G., 1980; Sadlerla.

Lloyd, 1976; Anemia. Hill, 1977, 1979; Polvstichium. Devi,

1977; Pvrrosia. Uffelen, 1985; Uffelen & Hennipman, 1985), familial (Thelypteridaceae, Wood, 1973; Cyatheaceae,

Gastony, 1974; Gastony & Tryon, 1976; Dicksoniaceae,

Gastony, 1981, 1982; Oleandraceae, Liew, 1977; cheilanthoid ferns, Tryon & Tryon, 1973; myrmecophytic ferns, Tryon,

1985; Polopodiaceae, Mitui, 1977) and regional surveys

(Mitui, 1982), and general overviews (Erdtman & Sorsa,

1971; Tryon & Tryon, 1982; Wagner, 1974). Wagner (1974) discussed his views on the role of spore structure in interpreting fern phylogeny. 6

Information on spore morphology of extant sphenophytes is dispersed. The development and morphology of spores of

Eauisetum have received some attention (light microscopy; elater activity, Pringsheim, 1853; Sanio, 1856, 1857;

Sitte, 1963; morphology, Knox, 1938; development, Beer,

1909; size range, Duckett, 1970; electron microscopy; spore germination, Gullvag, 1968a; surface morphology, Guervin, et al., 1972; sporoderm ultrastructure, Lugardon, 1969b,

1971a, 1976; Saxena, 1980; Kurmann & Taylor, 1984; Tryon,

1986), but a survey of the genus utilizing modern techniques remains undone.

A similar condition exists with reference to extant homosporous lycopods (Lvcopodium & Phvlloqlossuml which have been broadly surveyed (Knox, 1938, 1950; Wilson,

1936), but the majority of taxa have been illustrated only by line drawings. Scanning electron microscope treatments of selected species include those by Tryon & Tryon (1982),

Wilce (1972), and Breckon & Falk (1974; Phvlloqlossuml . A few species have also been examined with transmission electron microscopy (Lugardon, 1976, 1986; Pettitt, 1966a).

Similar treatments of spore morphology have been undertaken with heterosporous pteridophytes at the level of the light microscope (Selaginella, Reeve, 1935, megaspores

(Ms) & microspores (ms); Knox, 1950, Ms & ms; Tryon, 1949,

Ms; Isoetes. Pfeiffer, 1922, Ms), and with the electron microscope

Isoetes. Berthet & Lecocg, 1977, Ms & ms; Fuchs-Eckert,

• • 1981, Ms; Kott & Britton, 1983, Ms; Hickey, 1986, Ms;

Taylor, et al., 1975, Ms; Tryon, 1986, Ms & ms), but a slightly different emphasis has been given to the heterosporous condition and associated morphogenetic phenomena regarding: (light microscope) features associated with sporangial differentiation (Selaginella.

Horner, 1966; Horner & Arnott, 1963; Marsilea. Feller,

1953; Boterburg, 1956; Pilularia. Bonnet, 1955), cellular differentiation fSelaginella. Horner & Beltz, 1970; Denke,

1902; Reanellidium. Chrysler & Johnson, 1939), spore wall structure and development (Selaginella & Isoetes. Fitting,

1900; Selaginella. Lyon, 1905; Hannig, 1911; Fowler

& Stennett-Willson, 1978), spore and gametophyte development (Selaginella. Lyon, 1901; Duthie, 1923; Slagg,

1932), spore size in relation to attainment of heterospory

(Marsilea, Shattuck, 1910), and attainment of sexuality at an earlier phase of the plant life cycle than homosporous pteridophytes (Sussex, 1966).

Resolution in these areas has been greatly enhanced by the use of transmission electron microscopy. The walls of pteridophyte spores were among the first biological specimens to be examined at the fine structural level

(Afzelius et al., 1954; Lvconodium clavaturn. Selaginella selaginoides). While this instrument has seen limited use 8 with regard to taxonomic problems (Hennipman, 1970), the broader scope of application of this instrument warrents some discussion, particularly in the area of spore wall ultrastructure. At the present time, most major groups of pteridophytes have received at least cursory attention with regard to sporoderm ultrastructure (ferns & Psllotum:

Blechnum spicant. Lugardon, 1965, 1966, 1971a, 1976, 1986;

Tryon, 1986; Osmunda regalis. Lugardon, 1966, 1969a;

Pphioglossum vulgatum. Lugardon, 1971a, 1986; Anglopterls langifelifl, Agplenlum adiantum-nlgrum. Drvopteris filix- mas, Glelchenia sp., Leptopterls hvmenophvlloides. 1 ^

SUESIba, Marsilea pubescens. Hatteuccia struthiopteris.

■0lK>Clga struthiopteris. Lugardon, 1971a; Botrvchium ltinarifl, LgRfcSPteclP lEflSSEi, Marattia fraxinea. SPhioctlQgpuTn lusitanicum. Lugardon, 1971a, 1972b; Adiantum gftPiilHg-ygngrlSf Angmia Phvllltidis. Asplenlum ferjchpmangs, Athvrlum alpestre. flllx-femina.

Ceratopteris cornuta. Cheilanthes odora, cibotium qlaucum. gryptp.grflJima crispa. Cvathea cooper 1, Cj_ medullaris. gygfcpp^erip Iraailis, Davallia canariensis. XL. divaricata.

Penpgtagdtia binlnnata, Bicksonia antarctica.

S£ll<2ldana# PrVQPterjs borreri. Gleichenia linearis. ossanis&t Lvgodium iaponicum. Platvcerlum alcicorne.

Polvpodium serratum. Polvstichium setiferum. Pteridium asuiliimis, Efcerlg lonoifolia. scolopendrium vulgare..

Lugardon, 1971a, 1974; Angiooteris hvooleuca. Lugardon, 9

1971a, 1972b, 1976; Glelchenia bankroBtli. Lugardon, 1971a,

1974, 1976; Hvmenophvllum caplllare. Dlcksonla squarrosa,

Lugardon, 1986; Arthrooterls monocarpa. Bolbitis pqr.tprricengjlPr g.teniEis pulverulenta. villosa. Davallla

H*. frcichPlimnqldgg, Elaphoglossum forneranum. lingua, Humaifl fisrrafea, Leucosteqia pulchra. Lomariopsis ianumiais, Nephrolepis hirsutiUa, Oleandra wernerll.

EPlyfeStria alirsdii, Hoy, 1988; Asplenium adiantum.

Psilotum nudum. Pettitt, 1966a; Psilotum trl cruet rum.

Lugardon, 1971a, 1973a, 1979, 1986; homosporous lycopods;

LVPPPPfllmn clavatum. Afzelius, et al., 1954; la. selacro. Pettitt, 1966a; la. flnnfitintim# lu. Phlecrmarla. la. cernuum. la. contextum. Lugardon, 1986; water ferns; Pllularla globullfera ms, Tralau, 1969; Lugardon, 1971a, 1976;

Lugardon & Husson, 1982; E*. mlnuta ms, Lugardon & Husson,

1982; MflEgilea auadrlfolia Ms, Pettitt, 1966a; tL_ drummondii. Ms, Pettitt, 1966a, Ms & ms, Pettitt, 1971a; 1L_ fimbriata ms, &. qlomerata ms, Pettitt, 1979; £L. pubescens ms, Lugardon & Husson, 1982; Azolla pinnata Ms, Kempf,

1969a; hx. filiculoides ms, hx. africana ms, Lugardon &

Husson, 1982; Salvinia natans ms, Lugardon & Husson, 1982; heterosporous lycopods; Isoetes sp. ms, Lugardon, 1971a? I. echinosoora ms, Pettitt, 1966a; Lugardon, 1973; I. englemannii Ms, Pettitt, 1971; Isoetes sp. ms, Lugardon,

1971a; isoetes brochoni ms, Lugardon, 1973b; X*. durieui. I. setaceum ms, Lugardon, 1973b, 1976, 1986? 1^ setaceum Ms, 10

Lugardon, 1986; Selaglnella selaginoldes Ms, Afzelius, et al., 1954; Kempf, 1970; Lugardon, 1986; ms, Robert, 1971b; Lugardon, 1972a, 1976; s. kraussiana ms, Robert, 1971a; Lugardon, 1986; gx. galeottll Ms, Kempf, 1970; Ms & ms,

Tryon & Lugardon, 1978; Tryon, 1986; g^. mvosurus Ms, Martens, 1960a, 1960b; Ms & ms, Stanier, 1966; s. denticulata ms, Lugardon, 1971a, 1972a, 1976; g_». splnosa ms, Lugardon, 1971a; g*. usta Ms, Kempf, 1970; Qj. martensli

Ms & ms, Tryon & Lugardon, 1978; gj. pulcherrlma ms & Ms,

Pettitt, 1966a; g^ grand Is ms, gj. broadwavli ms & Ms. s. vltlculosa Ms, Pettitt, 1971a; gj. plana ms, Pettitt, 1971a,

1979; gj_ sulcata ms & Ms, Pettitt, 1979; g^ Helvetica Ms,

Sievers & Buchen, 1970; g^ kunzeana. ms, gj. marolnata. ms,

S. ludoviniana. ms, Tryon, 1986).

Aside from the presentation of Pettitt, in which a general overview of sporoderm morphogenesis was assembled from stages of a number of different pteridophytes

(LygPPQflium gnidioides. Ispfete? englemannii Ms, Selaglnella sulcata Ms & ms, gj. broadwavli Ms & ms, s*. grandis ms, Sj. plana ms, g*. viticulosa ms, Pettitt, 1971a), few other treatments have examined sporogenesis in non-fern groups

(Pettitt, 1971b, 1977, 1979; Buchen & Sievers, 1978a,

1978b; g*. selaginoides ms, Robert, 1970, 1971b). Among the ferns (including Psilotum), several taxa have been the subject of detailed developmental analysis (Blechnum spicant, Lugardon, 1966, 1971a, 1976; Osmunda regalis. 11

Lugardon, 1969a, 1971a, 1972b, 1976? Qphioqlossum vulaatum.

Lugardon, 1971a, 1971b, 1972b, 1976; Psllotum trlouetrum.

Lugardon, 1979; Botrvchlura lunaria, Pettitt, 1979; Marsilea

drummondii. Pettitt, 1971a).

Ultrastructural investigations of sporogenesis in

pteridophytes have touched on early phases, including pre-

and immediately post meiosis (e.g., Pteridium aqullinum.

Sheffield & Bell, 1979? Drvopteris borreri. Sheffield et

al., 1983; Selaglnella caulescens Ms, helvetica Ms,

Buchen & Sievers, 1978a; S^. brooksii ms, grandis ms,

Pettitt, 1974), but in the context of the present

investigation, those studies which address some aspect of

sporoderm formation are more relevant. One group of

studies concentrated on initial patterning of sporoderm

form and implicated the tetrad wall which comes to surround

each spore following meiosis, as playing a major role in

control of final form (Pettitt, 1971a, 1979). In addition,

the proposed fundamental structural unit of sporopollenin

deposition for seed plant pollen grains (the tripartite

"white line"? Rowley & Southworth, 1967; Dickinson &

Heslop-Harrison, 1968) appears to be present in

pteridophytes as well (Sievers & Buchen, 1971? Buchen &

Sievers, 1978b). This fact, in combination with the primexine-like functioning of the tetrad wall caused one group of workers to stress the homologies between these two major plant groups, a philosophy which is reflected in 12 their use of pollen derived terminology (nexine/sexine & ectexine/endexine) for pteridophyte spore walls (Sievers &

Buchen, 1970; Buchen & Sievers, 1984; Pettitt, 1979)

The alternate viewpoint arises from an analysis of the ontogenetic differences which become apparent at later stages of sporoderm morphogenesis. The timing, direction, and sequence of layer initiation with respect to the spore cytoplasm, as well as the variety of types of sporoderm layers which have no apparent counterpart in pollen grains, warrants, homologies notwithstanding, a different set of terminology for pollen and pteridophyte walls (Lugardon,

1975, 1976, 1978b, 1986). This view, based on late developmental studies of ferns and an extensive survey of mature pteridophyte spores, is summarized by Lugardon

(1978a).

Other lines of investigation involving pteridophyte spores have included; l) pteridophyte spores (especially ferns) as experimental systems (e.g., Raghavan, 1980

[review); Miller et al., 1983; Wayne & Hepler, 1984;), 2) spore size as a potential indicator of ploidy level (Bir,

1967; Vida, 1974; Pettitt, 1979), 3) spore viability as an indicator of apogamy (Kanamori, 1969), or evidence for hybridization (Graustein, 1930), 4) variation in megaspore size and number (Shattuck, 1910; Duerden, 1929; Bell,

1979), physiological parameters (Brooks, 1973), and ultrastructural features (Pettitt, 1971b, 1974, 1976a, 13

1976b, 1977; Bell, 1979) associated with the cellular basis for the attainment of heterospory, 5) spore germination patterns (Huckaby, et al., 1981), and development of young gametophytes (Gantt & Arnott, 1965, LM), 6) ultrastructure of spore cytoplasm (Gullvag, 1968b, 1971; Fraser & Smith,

1974), germinating spores (Beisvag, 1970), storage products

(Gullvag, 1969; DeMaggio & stetler, 1980), and surface coats (Pettitt & Jenny, 1974), 7) chemical composition of spores utilizing consumptive techniques (Sosa & Sosa-

Bourdouil, 1940; Wayne & Hepler, 1985), energy dispersive x-ray analysis (Tryon & Lugardon, 1978), or microprobe

(Robert, et al., 1973).

Ultrastructural analysis of fossil pteridophyte spores is not significantly younger than that of extant spores, but has been limited to the efforts of a comparatively small number of workers and their colleagues. The pioneer study in this area was produced by Pettitt (1966a), and incorporated both living and fossil material. There have been a number of fine structural studies involving dispersed pteridophyte spores such as; Nikitinsporites canadensis (Taylor, et al., 1980), and Archaeotriletes sp.

(Pettitt, 1966a) from Devonian sediments, Laeviaatosporites glabratus (Pettitt, 1966a) from Pennsylvanian sediments, and from Cretaceous sediments, Mlnerisporites spfrcrapsvgqp, Horstisporites harrisii. Erlansonisporites spinosis, and Ricinospora crvntoreticulatar (Bergad, 1978), 14 ffcJLlet.eg gamerroi. Paxillitrlletes menendezll. and Minerisporites labiosus. (Baldoni & Taylor, 1985),

EglanspnAPRfigiAgP smrassis, Bacutrlletes trlanaulatus.

HQrg.tiep.Qiriteg Izldodsa, HughesisPorltes pataaonlcus. and

Bacutrlletes sp. (Taylor & Taylor, 1988), Cabochonicus carbunculus (Batten & Ferguson, 1987), Horstisporites g.effllreticulatUP (Kempf, 1971a), Tasmanltriletes pedlnacron (Jux & Kempf, 1971), Banksisporites pinquls.

NathorBtlsporlteB hoplitlcus. Marqarltatisporltes t-UtfeaDflgfQOllg,Horstisporites kendalll. and Istlsporites murrayl (Kempf, 1972a), Arcellites Plicatus. Ax vimlnensls. and Crvbelosporites minor. (Li & Batten, 1986), and

Erlansonisporites sp. cf. erlansonii. Ricinospora lflgyjLqata>Sepisporites annulatus. HerbosIsporites pilosus. (Li, et al., 1987).

In some cases, dispersed spores are sufficiently characteristic to allow taxonomic assignment based solely on the morphology of the spore; examples include representatives of the Selaginellales (Taylor & Taylor,

1987, 1988), Isoetales (Kovach & Dilcher, 1985), and

Filicopsida fLophosorla. Kurmann & Taylor, 1987), all from

Cretaceous sediments. The spore apparatus of the heterosporous ferns are sufficiently complex and characteristic to allow valid taxa to be described based solely on the morphology of megaspore floats and microsporangial massulae (Kempf, 1969a, 1969b, 1971b, 1973; 15

Fowler, 1975, Hall, 1974, 1975).

Investigations of In situ spores deserve special emphasis since they often provide biologically meaningful information on their parent plant. In many cases, information can be obtained on groups of plants which have no living representatives. Major groups of pteridophytes which have been studied include: arborescent sphenophytes of Pennsylvanian age with spores referrable to Elaterites triferens (Kurmann & Taylor, 1984), an extinct order of sphenophytes, the SphenophyHales (Taylor, 1986), lycopods from Pennsylvanian sediments with microspores referrable to the genus Endosporites (Brack & Taylor, 1972, Taylor, 1973,

Brack-Hanes & Vaughn, 1978), and others with megaspores

Valvisporltes auritus (Gastaldo, 1981), arborescent

Pennsylvanian lycopods Achlamvdocarpon belgicum and

Lepidocarpon takhtaianii (Taylor, 1974; Taylor & Brack-

Hanes, 1976), two extinct divisions of Devonian plants, the

Trimerophytophyta (Psllophvton forbesii. Gensel & White,

1983), and the Progymnospermophyta fArchaeopteris cf. iacksoniif Pettitt, 1966a), and two extinct fern genera

(Botrvopteris, Millay & Taylor, 1982; Scolecopteris. Millay

& Taylor, 1984). Several investigations of in situ spores have been carried out on specimens from Devonian sporangia of uncertain affinity including, Barinophvton cltrullifonue

(Taylor & Brauer, 1983), and Protobarinoohvton pennsvlvanicum (Taylor, et al., 1984). CHAPTER II MATERIALS AND METHODS

Most of the results contained herein were obtained utilizing techniques which were continually modified throughout the course of the Investigation. Since success in studies which involve electron microscopy depends upon these alterations, a detailed account of attempted solutions, successful and unsuccessful, represents a valuable reference for future workers.

Materials

Fossil specimens were contained in calcium carbonate petrifactions (coal balls) from mid-continent Pennsylvanian sediments. Specimens investigated here include megaspores removed from the cones Mazocaroon (O.S.U. Coal Ball 249 D top), and Lepidocaroon (o.S.U. C.B. 767 G top), as well as an undescribed lycopod cone (O.S.U. C.B. 6089 F2 top) which contained megaspores referred to as Laaenoisporites sp..

In addition, megaspores from isolated sporangial remains circumscribed by the dispersed spore genus Valvisporites (O.S.U. C.B. 7021 G top) were also examined.

Extant material of Selaglnella. from the Royal Botanic

16 Gardens, Kew, cane fron living collections and herbarium sheets. Three species (S^ sulcata, s. aaleottii. and S. brevlpes) were successfully established fron cuttings in the greenhouses of The Ohio State University, and produced cones with negasporangia.

Methads Megaspores were removed from cones contained in coal balls by constructing a paraffin well around the desired area, and dissolving away the calcium carbonate matrix with hydrochloric acid (approximately 1%). Some specimens were dissolved In toto by suspending a fragment of coal ball containing the fertile area in part of a plastic funnel which had been cut and fitted with nylon sieve cloth (this was done by softening the funnel base on a hot plate, pressing it quickly onto the nylon cloth, and then sealing any gaps around the edges with epoxy). This microsieve could then be placed in a dish of fresh HC1 periodically, without losing any of the plant matter. In certain cases, dissolution of the calcareous matrix in clean (mud free) coal balls revealed exquisitely preserved cone parts with all sporangia and spores intact.

Preparation of specimens for transmission electron microscopy involved fixation of living material in 3-6% glutaraldehyde in cacodylate buffer (2% sucrose added; pH

6.8) for 2 hours, under vacuum (-6.9 bar), followed by 1% 18 osmium tetroxide for 1 hour (the glutaraldehyde step was omitted in the preparation of fossil material) . After several washes in fresh buffer, specimens were dehydrated in a graded ethanol series (25%, 40%, 50%, 60%, 75%, 85%,

95%, 100%; 20 min. between changes).

During handling (through dehydration and embedment), specimens were contained in 2 dram snap top vials, either as loose specimens or as agar discs prepared in the following manner. Suction, provided by a faucet aspirator, was used to draw a specimen, in water or alcohol, through an acetone soluble 1 cm millipore filter (statistical pore size 1.2 urn) which was situated atop a plastic holder. The specimen was then removed, with sample attached, and coated on all sides with melted agar. Embedment involved transfer to 100% acetone (2 changes with 20 min. between) - this serves to dissolve the filter as well as provide a miscible intermediate for the epoxy embedment - and gradual infiltration with Spurr low viscosity epoxy resin

(decreasing ratios of acetone to Spurr resin; 3:1, 1:1,

1:3, 2 changes in pure resin, 8-12 hours between changes).

Immediately prior to polymerization, the agar discs can either be chopped up and evenly distributed throughout one to several aluminum embedment dishes (pan embedded), or placed between plastic coated (Vue-all plastic transparancy holders work well) microscope slides (flat embedded). Some loose specimens were polymerized in latex flat block molds. 19 Polymerization was carried out in a 70C oven, under vacuum

(-10 bar), overnight.

Thin embedded and some pan embedded specimens were cut from the bulk specimens and glued to flat blocks of Spurr with the desired orientation with cyanoacrylate based adhesives ("crazy glue"). When possible, pan embedded specimens were cut (using a jeweler's saw) to fit directly into the ultramicrotome chuck. All specimens were trimmed

(by hand or using a Reichert TM 60 Automatic Block

Trimmer), and sectioned (AO Ultracut) with a diamond knife.

A recurring problem in the early phases of the research was wrinkling of sections resulting from section sizes commonly in excess of 500 urn. The method of section collection utilized initially consisted of lowering formvar coated 1X2 mm copper slot grids down onto the boat containing the sections. This resulted in excessive section wrinkling, sometimes even obscuring critical structures. This condition was eventually ameliorated by employing a new method of section collection which involved the use of an apparatus referred to aB a rack. The rack consisted of l mm thick aluminum with a series of 5 mm perforations. After being floated off of a glass slide onto a water surface, a formvar film was used to coat the holes in the rack. Sections were then collected on uncoated grids (floating on a drop of water adhered to the grid with the sections [hopefully] showing through the 20

slot), which were then placed onto the formvar film

opposite one of the holes in the rack. Upon drying down,

the sections show few wrinkles. Grids were removed from

the rack by situating them directly over a supported,

upright 3 mm post (in this case a bolt) which was then

pushed through the hole in the aluminum rack in which the

grid was located. The grid was supported by the post and

the formvar broke away at the edge of the grid.

Another difficulty which hampered significant progress was the pace at which sectioning proceeded when serially

sectioning specimens with large block faces (up to 1 mm

square). When as few as 2 sections will fit on a single grid, saving every section is obviously difficult,

especially when a complete series might include 3000 sections! A resolution of this lay in periodic but regular collection of sections. In this case, with sections approximately 90 nm in thickness, it was decided to collect every eleventh section, an approximate spacing of 1 urn. This interval provided adequate coverage of the coarse structural features, while reducing the large number of specimens saved, as well as easing the excessive time commitment.

In the early phases of the investigation, sections were stained by floating grids on a drop of stain

(potassium permanganate, uranyl acetate, lead citrate). A more rapid and successful method involves the use of a 21

Hiraoka staining kit (Polysciences, Warrington, Pa.) in which 1-40 grids can be stained simultaneously.

Precipitation problems encountered early on were greatly reduced or avoided by minimizing the use of KMn0 4 . A variety of staining schedules were employed, and will be indicated in the caption of each figure with a series of 3 numbers separated by hyphens. The first number indicates the time, in minutes of staining with KMnC>4 (1%), the second is uranyl acetate (1%), and the third is lead citrate (preparation after Venable & Coggeshall, 1965).

Technical difficulties have historically hampered developmental investigations of pteridophyte spores. Large block faces and resultant section size have already been discussed. An additional factor which threatened the results of this study was the persistence of air pockets in and among the various organs in whole cone preparations, and within megasporangia. This caused small specimens to float on top of preparatory solutions, and resulted in specimens so poorly infiltrated that hollow megaspores fell from block faces of megasporangia during specimen trimming.

The best results were achieved when whole megasporangia were fixed and embedded after being removed from cones.

Despite the success with this approach, megasporangia of s. caleottii 200-288 urn in diameter consistently contained a central unembedded lumen. This phenomenon limited information on the initiation of various wall layers. 22

Terminology

Spore wall terminology can be evaluated In two ways;

1) from the point of view of the mature form i.e. positionally, or 2) from a developmental perspective. The most meaningful approach integrates both concepts, one set of terminology which is based in part on consideration of both aspects was established by Lugardon (1978a), but applies to isospores and microspores of pteridophytes

(endospore, exospore, perispore, etc.). A similar set of terms was utilized by Kempf (1973) for megaspores, but was based strongly on position and role in formation of the ornamentation. According to Kempf, the perispore is the only wall layer which participates in spore sculpturing.

Based on relative position, therefore, the thin inner separable layer present in Selaainella represents the exospore. This view is not consistent with the original concept of the perispore. As originally defined, the perispore results from the activity of a periplasmodium

(Bower, 1935), and can, therefore, be present only in taxa which possess a plasmodial tapetum (e.g. ferns). In addition, Lugardon (1978a) established, based on his ontogenetic investigation of ferns, that the perispore does not participate in formation of the germinal suture. This is significant since the suture represents a universal topographic reference point in pteridophyte spores. A 23 terminology which incorporates developmental and developmentally based topographic information is preferable

to one which relies solely on topography. In this regard, a modified terminology after that suggested by Lugardon will be employed in this investigation. This includes; 1) exospore, for the layer which contains the pre-formed site of germinal exit, or suture, 2) mesospore, for a layer enclosed by the exospore, the definition of which will be addressed in the discussion section of Chapter V, and 3) intine, for the pecto-cellulosic wall of the spore protoplast. Other terms will be introduced when necessary, and all terms are defined in the Glossary (Appendix A). 24

CHAPTER III

WALL DEVELOPMENT IN FOSSIL SYSTEMS

Introduction

The investigation of in situ spores from structurally preserved fossil plants introduces a new level of biological information beyond the analysis of dispersed grains (Schopf, 1941b). Studies of in situ spores can contribute to a clearer understanding of the affinities of dispersed spores, and thus to the stratigraphic, evolutionary, and phytogeographic utility of dispersed spore assemblages. A number of pteridophytic groups have received attention of this type. Among the Sphenophyta,

Good (1975) was able to demonstrate that all spores produced by the arborescent Pennsylvanian sphenophytes actually possessed spores with elaters. This discovery resulted in a note of caution on the distinctness of certain sporae disoersae taxa (Good, 1977) and a greater appreciation of biological variability. Additional information was provided at the ultrastructural level utilizing the most common elater-bearing spore type removed from arborescent sphenophyte cones (Elaterltes triferens), which was compared to modern elater-bearing spores produced 24 25 by Ecrulsetum (Kunnann & Taylor, 1984).

An extinct group of Pennsylvanian sphenophytes - the

Sphenophyllales - possess what appears to be an overly broad range of spore types. Investigation of this variability reveals a set of unique characteristics which are combined in various ways in the several spore types, and serve both to delimit each type, and to distinguish the spores assignable to that group from spores of all other plant groups (Taylor, 1986). In this context, therefore, the information gathered from in situ spores can conceivably be used to infer affinities of spores known only from the dispersed record (Potonie, 1962; Courviosier

& Phillips, 1975). Similar information on affinity can be acquired by studying compression fossils, but cytological information is not obtainable. The additional information made available by the cellular detail of permineralized fossils has been tapped to some degree in establishing certain developmental sequences (e.g., Pigg & Rothwell,

1983).

Older investigations at the level of the light microscope lacked the resolution necessary to demonstrate changes in sporoderm development, but provide significant information on other pertinent aspects of reproductive biology. An example of this involves megasporophyll development in Lepidocaroon (Balbach, 1962). Developmental information of fossil spores can be gathered if either of 26 the following conditions apply; 1) aborted tetrad members can be recovered which may record an ontogenetic stage through which the functional members passed, or 2) apical cone regions can be defined which possess acropetally developing sporangia. Perhaps the first investigation to capitalize on information provided by aborted tetrad members involved monosaccate microspores assignable to

Endosporltes. which were described in various stages of saccus inflation (Brack & Taylor, 1972; Taylor, 1973), and with what appear to be dividing chromosomes (Brack-Hanes &

Vaughn, 1978).

Studies documenting the changes in the megaspore walls of heterosporous lycopods were made possible with the application of the electron microscope (Taylor, 1974;

Taylor & Brack-Hanes, 1976). The first ultrastructural investigation which concentrated on lycopod megaspores capitalized on both of the above conditions (Taylor, 1974) and concentrated at the level of the SEM, but incorporated some TEH images as well. Both this investigation and the earliest ultrastructural study of fossil megaspores

(Pettitt, 1966a), had as their underlying objective the structural and functional comparison between early seeds and seed-like reproductive propagules of arborescent heterosporous pteridophytes, principally the lycopods.

Significant changes were demonstrated to take place in the spore wall of these taxa in response to megaspore 27 enlargement (they enlarge an extraordinary degree compared to other lycopods). These specialized systems are atypical, and similar investigations on other genera of heterosporous lycopods (principally smaller arborescent and herbaceous types) have not been completed. Taxa which produce multiple megaspores of smaller size would undoubtedly provide a more meaningful opportunity for comparison to other fossil and extant heterosporous lycopods. One purpose of the present investigation is to provide developmental information on additional Pennsylvanian lycopod systems. How do the developmental sequences of these other lycopod megaspores compare with the lepidodendralean types, as well as with modern lycopod megaspores? In an attempt to address these questions, two additional taxa of fossil heterosporous lycopods have been considered from a developmental perspective, including a specimen of Hazocarpon in which two of the tetrad members have aborted, and an additional undescribed Pennsylvanian lycopod cone which also exhibits some spore abortion, as well as immature spores in the apical region of the cone.

This latter specimen contains megaspores assignable to

Lagenoisporites on which no developmental information is currently available. In addition, new specimens of

Lepidocarpon and mature megaspores assignable to

Valvisisporltes have been included to broaden the scope of the investigation in terms of specimen numbers and 28 reproductive types.

Results

Mazocaroon

The genus Mazocarpon was defined by Benson (1918) for permineralized cones thought to belong to the

Sigillariaceae. She described the megaspores of M. shorensef the type species, as being 1.2-2.0 nun in diameter, plano-convex or concavo-convex, and arranged around a subarchesporial pad with the trilete sutures centroscopically oriented (toward the sporangium center).

This is presumably due to the nutritive role of the subarchesporial pad in gametophyte development. Also figured were two abortive sporangia with subarchesporial pads covered with palisade tissue reminiscent of a tapetum.

The maximum number of spores per sporangium is eight in M. shorense. but at least one other species may possess up to twelve megaspores per sporangium. Specimens examined for this investigation compare most favorably with M. oedipternum (Schopf, 1941b) which possesses eight spores per sporangium. 29

Mazocaroon - functional megaspores

Functional megaspores of Mazocaroon are approximately

2.0 mm in diameter and are ornamented by a patchy surface reticulum (Plate II, fig. 4). A trilete suture with laesurae up to 300 urn in length is situated on the megaspore half which is directed toward the sporangium center. This surface is concave due to growth in contact with the mass of sterile intrasporangial tissue

(^subarchesporial pad). The opposite surface is convex and possesses no surface reticulum or other ornamentation. At the ultrastructural level, the walls are 33-36 urn thick in the proximal region, and thin to about 26 urn in distal areas. The inner 3/4 of the exospore in the proximal region and the entire exospore in the distal region is composed of flattened tubular elements (Plate I, fig. 1) which appear in cross-section as irregularly shaped vesicles of varying sizes (Plate II, figs. 5, 6). The tubules are loosely packed, but connected with an adjacent tubule on at least one side. Disposition of the tubeules is clearly parallel to the megaspore surface, but interconnections perpendicular to the surface are also infrequently present. In the region of the suture, the outer exospore becomes vesiculate (Plate II, fig. 7), and results in exospore thickening primarily due to the formation of inter-vesiculate material (i.e., the spaces between adjacent vesicles thicken with wall material). 30

Some thickening of the inter-vesiculate material is also apparent in the inner exospore. On average, the vesicles are larger in diameter than the more inwardly located tubular elements. No obvious compressional features are present (e.g., collapsed vesicles or tubules). A thin (50 nm) nearly continuous layer is present at the innermost extent of the exospore (Plate II, fig. 9).

The aforementioned surface reticulum bears a superficial resemblance to adhering cell walls, e.g., from contact with intrasporangial tissue, however, at the ultrastructural level, these projections do not possess a fibrillar cell wall-like structure, but rather an irregularly vesiculate organization *(Plate II, fig. 8).

One of the interesting features of the spore wall near the suture is the presence of cell wall material inside the megaspore (Plate II, fig. 5). This tissue represents the remains of megagametophyte tissue. The cells are variably sized and shaped, and some possess detritus which could be cytoplasmic in origin. No unequivocal subcellular components or sex organs could be identified.

Mazocaroon aborted megaspores

One tetrad was recovered which possessed two smaller adhering spores which are regarded as aborted (Plate III, fig. 10). They measure approximately 1 mm in equatorial diameter, and display a thickened trilete suture. No 31 reticulum or ornamentation is present on these spores. The walls are approximately 65 urn in thickness and loosely constructed throughout (Plate III, fig. 11). The outer 25 um is composed of variably sized and shaped thin walled chambers which are highly interconnected; chamber orientation is random. When viewed with the three dimensional perspective of stereo micrographs (Plate I, figs. 3, 4) the vesiculate nature of the chambers in the outer portion of the exospore is evident. The remainder of the wall consists of a series of anastomosing, partially interconnected tubular chambers (Plate I, fig. 3) that demonstrate a crude radial orientation in the center of the exospore; toward the wall interior, and roughly paralleling the circumference of the spore near the inner surface, the units become more random (Plate III, fig. 11). 32

Plate I. Stereo pairs Mazocaroon functional megaspores, fig. l Mazocaroon aborted megaspores, figs. 2, 3. 0 = outside (the megaspore); L «= lumen (of the megaspore) ,* E « exospore

figure 1. Interconnected tubules which make up the exospore. The megaspore surface is to the upper right, size bar « 50 um.

figure 2. Vesiculate outer exospore. The megaspore surface is to the upper right, size bar » 50 um.

figure 3. Interconnected tubules of the inner exospore. The megaspore surface is to the upper right, size bar = 50 um. Plat# I 33 34

Plate II. Mazocarpon functional megaspore O *= outside (the megaspore); L = lumen (of the megaspore); E « exospore

figure Proximal view showing trilete suture (T) and surface ramentum. Size bar =» 500 um.

figure Cross-section of wall showing the tubular wall elements with their smaller average size nearest the spore lumen, and cells of the megagametophyte. Size bar « 5um.

figure Cross-section of several tubular wall elements Size bar ■ 1 um.

figure Section through sporoderm in suture (S) region. The exospore on only one side of the separated suture is visible. Note intervesiculate thickenings in the center of the exospore. Size bar = 5 um.

figure Elongated process of the surface reticulum (•'ramentum"). Note the vesiculate organization. Size bar = 5 um.

figure Innermost exospore (E) and adjacent thin continuous layer (arrows) lining spore locule (L). Size bar « 1 um. Plat* II 36 36

Lepidocarpon functional megaspores

Mature megaspores of Lepidocarpon were retained within the confines of extra-sporangial leaf tissue. Presumably because of this structural organization, the megaspore wall

(« megaspore membrane) was not appreciably involved in support, and the megaspores grew as large as 2.0 cm. The organization of the wall is striking in surface view (Plate

III, fig. 16) and consists of appressed layers of plate­ like units interconnected within layers by narrow strap­ shaped extensions. In a cross-section of the wall (Plate

III, fig. 15), units of several layers can be seen. Their apparent thickness depends on the plane of section in which they are viewed and the degree of compression due to spore expansion. The overall thickness of this zone of interconnected units is probably no thicker than 15 um, but is difficult to determine with certainty due to the weak construction of the wall which may become highly folded during embedment, making clean cross-sections difficult to recognize. An irregularly thickened inner separable layer

1-3 um in thickness is also present (Plate III, fig. 14).

Lepidocarpon aborted megaspores

The wall of an aborted meiotic product in a megasporangium of Lepidocarpon consists of an interconnected network of spherical-irregularly shaped (in 37 cross-section) sporopollenin bars which show a preferred radial orientation. Cross-sectional dimensions of discrete units (i.e. those not cut at junction points with other units) range from 0.3-2.1 um in diameter. Units of all sizes, including irregularly shaped ones (i.e. junction points), seem to be randomly interspersed throughout the thickness of the wall.

There is some variability with relation to the density of the wall units. In one region (Plate III, fig. 12), there is perhaps 30-40 percent space and 60-70 percent material, consisting of highly interconnected and strongly radially aligned units. The more common situation has nearly equal proportions of space and material (Plate III, fig. 13).

The overall wall thickness of the stage examined ranges from 35-50 um, and is differentiated into two consistently recognizable layers. The inner layer (Plate

III, fig. 14) is thin (0.1-0.8 um) and may separate from the remainder of the exospore. The innermost extent of the exospore forms a continuous coating over the thin separable layer, and is partially composed of units with a comparable size and shape to those of the remainder of the exospore (Plate III, fig. 14). 38

Plate III. Mazocarpon aborted megaspore, figs. 10, 11 Lepidocarpon aborted, figs. 12-14 Lepidocarpon functional, figs. 15, 16 O « outside (the megaspore); L *= lumen (of the megaspore); E « exospore

figure 10, Low mag view of aborted megaspore (top) attached to reticulum covered functional megaspore (bottom). Pointers indicate laesurae of the trilete suture (T). Size bar » 500 um.

figure 11. Cross-section through exospore. Size bar = 5 um.

figure 12. Exospore cross-section with strong radial lineation. Size bar = 5 um.

figure 13, Exospore cross-section with less obvious radial lineation. Size bar = 5 um.

figure 14. Innermost exospore and inner separable layer (arrow). Pointers indicate variably shaped exospore units which fuse to form a continuous layer. Size bar = l um.

figure 15. Cross-section of megaspore ••membrane". Arrows indicate inner separable layer. Size bar » 5 um.

figure 16. Surface of the megaspore "membrane", size bar » 5 um. Plat# III 39

s-'iPS.^* CJ'M^ mfM-frMtfg 40

Laaenolsporltes nudus (Nowak & Zerndt) Potonie & Kremp

exospore development

Laqenoisporites was established by Potonie and Kremp

(1955) for megaspores with prominently thickened and elevated laesurae, a feature which they share with some

species of Lagenicula and Setosisoorites. but which is distinguished from those genera by a relative lack of surface ornamentation. Specimens like those examined here were reported as Triletes (?) nudus from the Herrin (No. 6) coal bed in Illinois (Schopf, 1938).

The smallest megaspore recovered from this cone (300 um) appears shriveled, and was probably aborted. The wall of this specimens ranges from 42-49 um in thickness and has several clearly differentiated internal regions (Plate IV, fig. 17). The outer 20 um is densely packed at the surface and becomes gradually less so toward the inside. Wall units are highly interconnected and strongly disposed perpendicular to the surface. The next 20 um is loosely packed and displays a low degree of interconnection with units disposed randomly or nearly parallel to the surface.

The inner 5 um shows a return to perpendicular disposition and a higher degree of interconnection. A thin continuous layer (0.2-0.3 um) lines the inner surface (arrow).

Small specimens removed from near the cone tip have a thinner wall, approximately 27 um in thickness, and less 41

internal differentiation (Plate IV, fig. 18). A nearly continuous irregular surface layer overlies a region which occupies about half of the wall and is composed of highly

interconnected, loosely packed units disposed roughly perpendicular to the surface. The inner half of the wall has units which are oriented parallel to the surface, and are not highly interconnected except near the inner surface. A nearly continuous, irregularly thickened layer marks the inner extent of the exospore (Plate IV, fig. 18), and contacts a continuous, Bomewhat irregular (50-150 nm), more lightly staining layer which lines the spore locule.

Larger specimens from the near the cone base have a wall which is thicker, but possesses similar structural characteristics (Plate IV, fig. 19). There is also a more extensive network of spore contents. It is probably significant that the aborted megaspore possesses no such contents. 42

Plate IV. Laqenolsporites nudus. figs. 17-19, 23 Valvlsisporltes sp., figs. 20-22, 24 O “ outside (the megaspore)f L = lumen (of the mega spore); E <* exospore

figure 17. Wall of an aborted megaspore. Arrow indicates innermost continuous layer, Size bar « 5 um.

figure 18. Wall of small megaspore from the cone tip. Arrow indicates innermost continuous layer. Size bar “ 5 um.

figure 19. Wall of a fully mature megaspore. Arrow indicates innermost continuous layer. Size bar « 5 um.

figure 20. Cross-section of exospore. Note extreme degree of compression. Size bar ** 5 um.

figure 21. Detailed cross-section of surface layer. Note the irregularly stained appearance of the thicker portions, and the electron dense flecks which coat the surfaces. This layer is visible at the top of figure 20. Size bar “ 1 um.

figure 22, Detailed cross-section of inner separable layer. A lateral equivalent of this layer is present at the bottom of figure 20. Size bar « 1 um. Plat, iv 43 44

yalvlsisporites auritus exospore ultrastructure

Valvlslsporites auritus (Zerndt) Bhardwaj was produced by isoetalean lycopods which inhabited Pennsylvanian coal swamps of the mid-continent of North America. The species was established to contain triangular megaspores with three bulbous projections at the extremities of their laesurae.

Megaspores range in size from 500-1500 um, and when found ill situ in structurally preserved specimens, large numbers are produced in each sporangium (100+).

Specimens examined for this investigation average 700 um in equatorial diameter. The exospore of this spore type ranges from 18-21 um in thickness, and is constructed of plate-like lamellae 80-240 nm in thickness (Plate IV, fig.

20). The lamellae are oriented parallel to the megaspore surface, and are often appressed to one another. The high degree of compression renders structural interpretation difficult, especially at the innermost extent of the sporoderm where an inner separable layer, 1.2-1.8 um in thickness is present (Plate IV, fig. 22). The pre­ compression organization did, however, undoubtedly possess a significant amount of open space within the wall. At the surface, there is a region approximately 3 um thick which is constructed of more irregularly thickened, interconnected, and randomly oriented units (Plate IV, fig.

21). These units often show a finely mottled staining 45 reaction in the center where they are the thickest. In addition, nearly all surfaces of the units in this outer region are sparsely covered with tiny electron dense flecks which could represent remnants of a siliceous surface coating or some diagenetically induced anomaly. 46

Discussion Maapcarpon Megaspore characteristics similar to those documented for Hi. shorense were ascribed to H±. oedipternum. with the exception of a lack of ornamentation, and presence of adhering subarchesporial pad tissue ("raraentum"; Schopf,

1941b). Long (1968) described additional specimens of M. pettvcurense (initially established by Benson [1918]), one of which possesses up to twelve megaspores per sporangium.

Two additional species were described by Pigg (1983; M. villosum. Ht. bensonlil. One of the described specimens of

Hi. villosum possesses apical sporangia containing a number of immature tetrads 42-64 um in diameter with a sporoderm approximately 0.8 um in thickness. Megaspores of M. bensonii range from 1.5-2.5 mm in diameter with a sporoderm

18-34 um thick.

Chaloner (1953) considered the intrasporangial pad, and the resulting saucer-shaped megaspores which were produced therein, to be the main distinction between

Mazocaroon and Leoidostrobus. He proposed using compressed megaspore shape to determine the original form, and thus imply the taxonomic affinities of datached cones and dispersed spores.

The mature exospore ultrastructure of Mazocaroon oedloternum was investigated by Zimmerman and Taylor

(1970). The structure of the wall which was figured by 47 those authors is consistent with that presented here. With regard to the surface reticulum, the evidence presented here would seem to suggest that the reticulum may be the result of tapetal deposition, i.e. a perispore, but for the rather convincing demonstration by Schopf (1941b) of the continuity between the reticulum ("ramentum") and the intrasporangial tissue. At the fine structural level, the spongy organization of the surface reticulum supports the interpretation of this layer as a perispore.- An alternative explanation is that the original fibrillar organization of the walls have been lost due to diagenetic alteration. This seems unlikely since megagametophyte cell walls in the same megaspore have retained this structure.

An additional possibility is that both interpretations could be correct. In modern plants, sporoderm deposition is a secretion phenomenon which is followed by protoplast disorganization (Echlin, 1971). If the tapetal cells were to release their protoplasts into a sporangial locule whose megaspores would come to fill locule as completely as in

Mazocarpon. it is conceivable that the perispore which resulted could adhere to both the megaspore surface and the old tapetal cell walls, an arrangement like that observed in Mazocarpon. Clearly, something causes adhesion between the pad tissue and the megaspore surface after the megaspores have enlarged to some degree, and there are few other possibilities besides the tapetum. However, no 48 lycopods have yet been documented to possess a plasmodial tapetum, and in the strictest sense, only those sporangia which possess a plasmodial tapetum can produce a perispore since this layer is represented by the remnants of the periplasmodium (Bower, 1935; Pettitt, 1966b). The only logical possibility which remains is that the ramentum is exosporial in origin, a hypothesis which is easily tested by performing acetolysis and observing whether a portion of the ramentum survives.

Valvisisporites

Upon initial examination, the megaspores assignable to

Valvisisporites appear to possess many of the characteristics of a modern Selaoinella megaspore. The organization of the bulk of the sporoderm is uniform. An inner highly compressed layer is situated next to the spore lumen, and the external surface is differentiated, and may have some siliceous deposits. The basic construction differs in being highly compressed with wall units that appear plate-like. This may well be due to the larger size of Valvisisporites megaspores. However, Gastaldo (1981) examined slightly larger megaspores assigned to

Valvisisporites auritus from southern Illinois, which possess an organization which is more similar to

Selacrinella primarily due to a lower degree of compression.

Wall thickness in the Illinois specimens is about three 49 tines higher and unit thickness is approxinately twice that of the negaspores exanined here. The two sets of specinens are consistent as far as basic structure is concerned; conpressing the wall seen in the Illinois naterial could produce the organization observed in the present naterial.

However, larger negaspore size, coupled with thicker walls in the Illinois specinens suggests that they were larger overall. If sinilar developnental processes were operative, it stands to reason that the Illinois negaspores could have withstood additional expansion. As for the negaspore surface, no continuous layer, either organic or siliceous, seens to be present, as in Selaainella. There is, however, sone differentiation of the surface layer.

Despite their larger size and possible lack of a continuous surface layer, the basic construction of these

Valvisisporites negaspores is, in fact, quite sinilar to a nodern Selaoinella negaspore.

It now seens clear that nost, if not all, of the lycopods described as possessing Valvisisporites negaspores are probably nembers of the Chaloneriaceae, and therefore belong to the Isoetales. Unfortunately, sporodem developnent has yet to be investigated in living nembers of this order.

Lepidocarpon

The genus Lepidocarpon was instituted by Scott (1900), 50 and later revised by Schopf (1941a). This revision included a recognition of the aborted megaspores, which were first figured (line drawings) by Chaloner (1952) from

Lepidocarpon waltonl. The mature megaspores were first described from a dispersed assemblage as Triletes alaanteus

(Zerndt, 1930) until the genus Cvstosporltes was proposed to encompass them (Schopf, 1938). Balbach (1962) provided developmental information on the megasporophylls, and

Darrah (1968) reported the structure of the megaspore membrane, but both of these light microscope investigations lacked the resolution necessary to critically examine the megaspore wall. This became possible only after the application of the electon microscope. Pettitt (1966a) included a transmission electron micrograph of the mature megaspore wall of Cvstosporltes aiaanticus. and developmental information was provided by Taylor (1974;

Leplflpparpontakhtaianii and Achlamvdocaroon beloicum) and Taylor & Brack-Hanes (1976; Achlamvdocarpon) in the form of both aborted products, and immature specimens from a developing cone tip. Wall development in Lepidocarpon conducted for this investigation shows results which are generally consistent with these authors. The dramatic alteration of unit shape and nearly complete obliteration of space in the wall are obviously due to the drastic expansion of the megaspore. The wall development pattern proposed for Lepidocarpon poses some interesting questions, 51 and raises some concerns as to the reliability of developmental analyses of this type. It is difficult to envision the striking structural differences between the wall structure of mature and aborted megaspores arising as a result of the developmental mechanisms which have been elucidated for heterosporous lycopods. Two factors could account for these discrepancies. First, the developmental changes which these aborted megaspores record may have been subject to deposition following protoplast abortion, much like the activity which contributes to mesospore thickening in aborted Selaoinella pulcherrima megaspores (Pettitt,

1966a). Secondly, the high degree of expansion which occurred in Lepidocarpon may have dictated that a different set of developmental controls operate, especially in light of the additional support offered by the sporangial wall or enveloping megasporophyll.

&aqgn

A number of cones have been described which contain megaspores assignable to the the genus Laaenolsporites. but few have been adequately illustrated. All of these cones were initially assigned to Lepidostrobus. but some were transferred to Flemingites, a genus of monosporangiate cones, by Brack-Hanes and Thomas (1983). The only species of Lagenoisporites which has been reported from Fleminqites cones is rugosus Loose, which is not the same as that 52

investigated here. This cone is probably, therefore, a new species of Flemingites.

Morphogenetic changes through wall development

In all cases, the dominant morphogenetic event is one of increase in megaspore size creating compressional forces in the exospore which must be accomodated by some change in wall organization. In Laaenoisporites. all strata were affected either by changes in unit disposition, or by volume changes of the space present in the wall. Changes in unit disposition take place in the innermost five micrometers and in the central interconnected region.

Apparent space modifications are seen in both the central loosely organized zone, in which a significant amount of free space is available, and in the dense region just beneath the outer surface. In the latter case, chambers become flattened tangentially, but no interruptions form, despite the stresses.

Similar structural modifications are observed in

Mazocarpon in response to megaspore expansion - i.e., wall thinning and reorientation of spaces (random to tangential). The spaces within the tubular units, however, do not appear to have been compressed. As in

Lagenoisoorites. just beneath the surface is a more highly interconnected zone, but it is less dense. Continuity of the surface in Mazocarpon is maintained by the perispore- 53 like layer and not a dense surface layer.

In general, a megaspore wall can accomodate a limited amount of expansion by changes in unit disposition and space modification. Beyond a certain threshold, the wall of a megaspore must compress; that threshold is a function of the type of construction.

Within the lycopod dominated Pennsylvanian coal swamps of mid-continent North America, a number of different wall types evolved to accomodate megaspore expansion.

Sequential changes which take place within the wall to accomodate increasing megaspore size are: 1) changes in disposition of wall units (e.g., Laaenolsporltes). followed by, 2) loss of excess space (i.e., between unconnected units in loosely constructed regions; e.g.,

Lagenoisporltes), and then, 3) loss of all or nearly all space centrifugally (e.g., Valvisisporites}. The megaspore sizes at which these changes occur in each taxon are strongly affected by the basic construction type of the wall.

Correlation of ultrastructure with dispersive strategy

Of the four types of reproduction suggested by

Phillips (1979) for arborescent Carboniferous lycopods, three are represented here, and some developmental information is provided on two. The remaining two types were examined by Taylor (1974) and Taylor and Brack-Hanes 54

(1976). The first type, the free-sporing Leoidostrobus type, is represented here by both the Valvisisporites and the Laaenoisporites bearing cones. These contain relatively large numbers of megaspores, which are dispersed separately; i.e. the dispersal unit is the individual megaspore. Structurally, the wall is spongy, not unlike a modern Selaqinella megaspore, which is not surprising since they share the same reproductive strategy. In addition, all have some type of outer bounding layer, either organic

(sporopollenin) or siliceous, which probably affords some degree of protection, beyond the rather porous spongy wall, to the enclosed megagametophyte.

The second type, monosporangiate cones with intrasporangial modification, is representad by Mazocarpon.

The individual megaspores come to be separated from one another as a result of the post meiotic proliferation of the intrasporangial tissue, into which the proximal suture is directed. Here, the dispersal unit is hypothesized to be the entire megasporophyll with its attached sporangium

(Benson, 1918; Schopf, 1941b). Prior to fertilization, the sporangia fragment and each megaspore remains attached to a piece of sporangial wall to the outside, and a layer of intrasporangial pad tissue to the inside. This has been hypothesized to retard desiccation (Phillips, 1979). This unique reproductive strategy is coupled with a unique ultrastructural organization consisting of tubular wall 55 units. In the absence of an outer bounding layer, protection was probably provided by the adhering sporangial material. CHAPTER IV

WALL ULTRASTRUCTURE IN SELAGINELLA

Introduction

The earliest work to focus on negaspores of heterosporous pteridophytes was conducted on North American representatives of the genus Selaainella. Megaspores of four species were illustrated by Reeve (1935), and an additional thirty-two species by Tryon (1949). A lack of resolution and depth of field hampered these investigations such that the full taxonomic potential of megaspore surface morphology was not realized until the scanning electron microscope became commercially available in the 1960s.

Even so, the non-fern pteridophytes received only limited attention. In fact, no broad monographic treatment of the genus Selaainella using the SEM has yet appeared, although one is currently in preparation (Tryon, personal communication), but is mainly focused on the microspores.

The majority of the research, to date, which focuses on the megaspores of Selaainella is in the form of regional treatments including, Hong Kong (Dahlen, 1982), and

Argentina (Morbelli, 1977). The most recent and comprehensive study incorporates scanning electron

56 57 microscopy of surface morphology and wall sections of thirty-eight species from five continents in an attempt to classify the various megaspore wall types (Minaki, 1984).

Additional lines of investigation involving megaspores of

Selaainella have revolved around the variety of wall structural types which can be visualized with the transmission electron microscope.

Megaspores of Selaainella selaainoldes were among the first biological specimens to be examined utilizing the transmission electron microscope (Afzelius, et al., 1954).

Since that time, several workers have undertaken more advanced investigation involving selected members of this genus with the following emphases: 1) elucidating unusual structural characteristics, particularly with regard to a certain wall structural type with a highly ordered appearance (e.g., Martens, 1960a, 1960b; Stanier, 1966;

Kempf, 1970; Tryon & Lugardon, 1978). This structural type is reviewed and analyzed by Taylor & Taylor (1988), and has been reported in five species to date, with two additional taxa added in this presentation. 2) other wall structural types (Kempf, 1970; Tryon & Lugardon, 1978; Pettitt,

1966a), 3) various phases of megasporogenesis (e.g.,

Pettitt, 1971a, 1979; Sievers & Buchen, 1970), and 4) potential controlling mechanisms of sporangial differentiation (e.g., Pettitt, 1974, 1976a, 1976b, 1977).

These investigations augment information provided at lower 58 levels of resolution in these various areas by other workers (e.g., Fitting, 1900? Lyon, 1901, 1905).

Pettitt (1966a) was the first to compare fossil and extant spore walls and specifically to use megaspores of

Selaainella to address the significance of a particular wall stratum found in both (mesospore). More information can be gathered with this type of approach, and this constitutes a major premise of this investigation. A unified appraisal of fossil and living systems with regard to structure, development, and variations between groups over time may provide key insight to the potential significance of the prevalent structural organization in megaspores, commonly referred to by the rather imprecise term "spongy11. Pursual of this goal requires a careful consideration of the various types of spongy morphology.

The major aim of this presentation, therefore, is to gather ultrastructural information on a number of different wall construction types within the genus Selaainella. This information, in combination with other published descriptions, will aid greatly in elucidating the parameters which interact to create the variety of structural types present within megaspores of heterosporous pteridophytes. A second aim of this study is to consider the potential significance of consistent ultrastructural features, such as the siliceous surface coat. For the sake of comparison, an approximate equatorial diameter, and, 59 where possible, a proximal view of the spore will be included in each description.

The sporoderm terminology originally used by Fitting

(1900) for Selaainella will be employed in this investigation (exospore, mesospore), with the addition of one general term, the inner separable layer (isl). This layer is consistently present and is composed of a darkly staining, often lamellated region approximately 500 um thick adjoined to a thin portion of the innermost exospore, whose adjacent surface often stains with a similar intensity. The inner separable layer may be bounded on the protoplast side by an intine.

Terms and concents

Published descriptions of megaspore wall ultrastructure have utilized such terms as labarynthine, or, more commonly, spongy, for designating structural organization. The definition of this term is a critical step in the refinement of this descriptive process, and will be accomplished by highlighting key structural components of "sponginess". This section is specifically designed to define the terms and concepts which will be used throughout the descriptive portions of this contribution.

In all likelihood, all structural components viewed in a particular cross-section are ultimately connected. While 60 this fact is considered, it does not preclude treatment of

the various structural components as somewhat separate

entities in order to gain some appreciation for the subtle

differences which characterize these megaspore walls.

Unit size (or thickness) and shape

In each cross-section of a particular wall region in

Selaginella, the relative proportion of variously shaped

solid wall units imparts a local appearance which may distinguish that region. The number of possible unit

shapes can be described with a relatively small number of types: 1) A particular unit may appear circular in free cross-section, in which case it is probably rod-like in

Bhape (at least locally; Plate V, fig. 24, A), or 2) it may be more sheet-like - i.e., elongated in cross-section - and

form the wall of an open or closed vesicle, or occur as a free lamella of variable thickness and lateral dimension

(Plate V, fig. 23). 3) The fundamental unit may be less uniformly thickened, and appear to be constructed of small spherules fused into larger plate-like elements which are regularly or irregularly ornamented with rounded protrusions. The regularly adorned plate-like element which will be referred to is the ordered element. It consists of a plate-like structure with a regular pattern of protuberances on one side (Taylor & Taylor, 1987).

These plates may occur in stacks (which may themselves 61 possess a high degree of organization) or may only be

suggested in isolated regions. Plate-like elements with

irregularly organized protrusions are far less ordered, and

will be referred to as displaying lateral fusion of wall units. Regardless of whether the wall is composed of units which are laminar, unordered spherules or ordered

spherules, there is a clear maximum and minimum size which

is characteristic of a particular region, and is arrived at based upon a visual survey and measurement of the extremes.

Unit disposition

Elongated units (plates.or rods) are either parallel to, perpendicular to, or randomly disposed with respect to the megaspore surface. This is used in a regional sense to characterize particular strata within the wall.

Unit density

The degree of packing of adjacent solid units can define all or part of a megaspore wall.

Degree of interconnection

Two separate megaspore walls can be composed of similarly sized and shaped units but differ significantly in overall structure if adjacent units in one tend to fuse while in the other they remain free. There are several 62 important structural arrangements which will be referred to in describing megaspore walls. One is the free circular cross-section, i.e. a cross-section of a solid rod-like portion of a wall. It is assumed that these circular views represent either cross-sections of anastomosing loops of wall material where they happen to be separated from the margins of sheet-like units (Plate V, fig. 24, A), or cross-sections of units which are actually rod-like for much of their length (Plate V, fig. 24, B). A second component is the presence of free sheet-like margins, i.e. not connected to form what appear as closed chambers or vesicles in cross-section (Plate V, fig. 23, A). A highly interconnected construction - one with no free margins or circular cross-sections - approaches an organization consisting solely of completely enclosed vesicles.

Overall, wall construction in Selaginella can be broken down into two primary types: 1) those consisting primarily of uniformly thick laminar units - the concepts of degree of interconnection, presence of free sheet-like margins, and free cross-sectional profiles are useful in delimiting subcategories in this case (Plate V, fig, 23) - and 2) those constructed of irregularly thickened plate­ like elements which can be best described as being composed of spherical to irregularly shaped units which display various degrees of lateral fusion (Plate V, fig. 24). 63

Plate V. Diagrammatic representations of basic megaspore wall construction types

figure 23. Wall construction type which is composed of uniformly thick laminar units. Those which do not form the wall of a closed vesicle are termed free sheet-like margins (A).

figure 24. Wall construction type which displays lateral fusion of spherical or rod-like units, some of which are joined to more plate-like units in other planes (A), while others are rod-shaped for most of their length (B). 64 Plate V 65

Species descriptions

Types exhibiting strictly laminar construction

Selaginella flabellata type - Highly interconnected, dense construction, thick surface coat incorporating extensive loose mesh reticulum, complex stratification in inner separable layer - 5^. flabellata. S-i. usta. and Sj. viridanaula

Selaginella flabellata (L.) Spring

These megaspores are approximately 250 urn in equatorial diameter (Plate VI, fig. 25), with wide elevated laesurae which extend the full radius of the megaspore.

Interradial regions are ornamented with low spinae and verrucae, while low ridges arise near the equator and become organized into a coarse reticulum on the distal surface (Plate VI, fig. 26). The sporoderm is 18-22 um in thickness, loosely constructed and highly interconnected, i.e. free circular cross-sections are rare (Plate VI, fig.

27). The wall is constructed primarily of sheetlike units organized into closed loops. Disposition of elongated units is roughly parallel to the spore surface. Some compression is present in the inner 1/3 of the wall. An irregularly thickened, inner separable layer is present and. includes one or two lamellae of the exospore (Plate VI, 66

fig. 29). The thickness of the wall units is variable throughout, ranging from 0.3-0.5 utn with a slight narrowing trend toward the surface (to 0.2 urn; Plate VI, fig. 28). A well developed silica deposit covers the surface and incorporates a loose network of thin bar like sporoderm units (Plate VI, fig. 28). The silica deposit penetrates to a depth of 4.0 um.

Selaginella usta Viell. ex Baker

The sporoderm of this taxon is 20 um in thickness, densely constructed, and highly interconnected (Plate VII, fig. 30). The disposition of wall units is only weakly circumferential in places, especially near the inner surface. The thickness of wall units is somewhat variable

(0.3-0.5 um) throughout, except at the outer surface where a network of wall units with decreasing dimensions (to 80 nm) is embedded in a thick continous siliceous surface coating (Plate VII, fig. 31), which penetrates to a depth of 4.5 um. Some small spines can be seen on this surface coating. Complex layering is associated with the inner separable layer (Plate VII, fig. 32)

Selaginella viridangula Spring

The exospore of this taxon is approximately 18 um in thickness, and is rather dense and highly interconnected

(Plate VII, fig. 33). Chambers enclosed by the wall units 67 are slightly flattened and roughly parallel to the outer surface in the outer exospore, but become highly flattened and elongated parallel to the surface in the inner exospore. Compression is apparently the reason for the flattening of vesicles, but unit thickness is only slightly lower on average in the inner exospore than the outer

(0.15-0.25 um throughout the wall). A siliceous surface coating up to 3 um thick penetrates to a depth of 2 um, but may be present throughout (Plate VII, fig. 34). Complex layering is associated with the inner separable layer

(Plate VII, fig. 35). 68

Plate VI. Selaginella flabellata. Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph in the following transmission electron micrographs, and any lamellations or other linear features (e.g., siliceous surface layers) are parallel to the spore surface. E = exospore, I = intine, isl = inner separable layer.

figure 25.

figure 26.

figure 27. Cross-section of exospore (except inner separable layer). X3300.

figure 28. Detail of reduced surface elements (e.g., arrow) and surrounding siliceous coating. X16,000.

figure 29. Inner separable layer. Spore lumen is to the right side of the micrograph. X40,000. Plate VI 70

Plate VII. Selaginella usta figs. 30-32 g.elaqlnellfl viridangula figs. 33-35 Unless, otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. E = exospore, I = intine, isl = inner separable layer

figure 30 Cross-section through exospore. X3300.

.figure 31 outer exospore elements (note reduced size) and thick surface coating. X12,500.

figure 32 Complex layering associated with innermost exospore.. Spore lumen is toward the left side of the micrograph. X31,500.

figure 33 Cross-section through exospore. X3300.

figure 34 Highly magnified exospore showing possible siliceous coating of elements throughout. X40,000.

figure 35. Innermost exospore and inner separable layer. X12,500. Plato VII 71 72

SfilaglneUa lnaeoualifolia type - Highly interconnected, loose construction, thick surface coat incorporating mainly free loops. Highly compressed and lamellated inner separable layer, fi* Inaeouallfolia. pjUIepcenB* and fij. PHlPhgrcima Selaginella inaeaualifolia (Hook. & Grev.) Spring

Megaspores of this taxon measure 300 um in equatorial diameter and possess a slightly elevated suture whose laesurae extend nearly to the equator. Ornamentation consists of both larger evenly distributed spinae, and smaller less regularly distributed verrucae (Plate VIII, fig. 36). The exospore is 20-25 um in thickness (Plate

VIII, fig. 37). Two intergrading regions are detectable in the body of the exospore, the outermost of which is, approximately 8 um thick, densely constructed, highly interconnected, and uncompressed. The inner stratum is 12-

18 um thick, more loosely constructed, and more compressed; closed loops are consistently flattened parallel to the spore surface in this region. Fewer loops occur in the inner exospore (Plate VIII, fig. 37). Unit thickness is highest in the outer region, averaging 0.2 um, thins slightly to an average of 0.15 um in the exospore center, and thins once again to o.l um near the inner surface. A

0.5 um thick inner separable layer is present, which includes a highly compressed portion of the exospore 73 several lamina in thickness (Plate VIII, fig. 38, isl). An lntlne and a megagametophyte cell wall are also present In the Illustrated specimen. At the megaspore surface (Plate

VIII, fig. 39), a well developed silica deposit penetrates to a depth of 1.5 um and encloses a number of closed loops.

Sglaqlnellfl Rflllgpceng (Presl) Spring These megaspores are approximately 240 um In equatorial diameter (Plate VIII, fig. 40). A trllete suture with slightly sinuous laesurae extends 3/4 of the spore diameter. Interradlal ornamentation consists of tightly packed baculae. Near the spore equator, the baculae fuse to form a relatively coarse reticulum. The sporderm is approximately 17 um in thickness between ornamental elements, is loosely constructed, and highly interconnected (Plate IX, fig. 42). Unit orientation appears random and there is little observable compression.

An inner separable layer is present (0.20-0.25 um thick) which is composed of lamellae approximately 30 nm in thickness (Plate VIII, fig. 41, arrow). Unit thickness is uniform throughout at 0.2-0.3 um. A discontinuous surface layer of wall material is coated with a thick (to 2 um) nearly continuous silica deposit which penetrates to a maximum depth of 4 um (Plate IX, fig. 43). Smaller megaspores of this species possess thinner walls, compression (espcially in ornamental elements), and sparse 74 silica deposits, but wall units of comparable thickness and disposition (Plate IX, fig. 44).

Selaginella pulcherrlma Liebm.

Megaspores of this taxon average 300 um in equatorial diameter (Plate IX, fig. 45), with a strongly elevated suture which extends 3/4-7/8 of the way to the spore equator. Ornamentation consists of spinae or verrucae which are free from one another in the interradial regions, but unite to form a reticulum near the spore equator. The exospore is approximately 25 um thick and is highly interconnected throughout (Plate X, fig. 46), In cross section there are a number of closed loops within the wall, a few free lamellar tips but few free circular cross sections. All closed loops are slightly flattened, and lamellar portions oriented parallel to the megaspore surface, but otherwise compression is minimal throughout most of the exospore. Unit thickness is uniform (0.2 um) in the outer half of the wall, but thins slightly (to 0.1 um) in the inner half of the exospore, and at the surface, where units are partially embedded in a siliceous coating

(Plate X, fig. 47). This siliceous coating is 4.0 um thick and penetrates to a depth of 2.5 um. A highly compressed inner separable layer is also present (isl; Plate X, fig.

48). The portion of this layer which is in immediate contact with circumferentially oriented innermost exospore units probably represents the compressed remains of the mesospore. A slightly lower contrast region which could represent an intine occurs inside of this (Plate X, fig. 48). 76

Plate VIII. Selaginella ineaualifolia figs. 36-39 Selaginella pallescens figs. 40, 41 Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. E « exospore, I = intine, isl ■= inner separable layer

figure 36. View of proximal surface. X175.

figure 37. Cross-section through exospore. X2650.

figure 38. Detail of inner separable layer and adjacent strata. X20,100.

figure 39. Surface elements and siliceous coating. X12,500.

figure 40. Proximal surface. X2S0.

figure 41. Inner separable layer. Note lamellations (arrow). X50,500. LL III A W i d 78

Plate IX. Selaginella pallescens figs. 42-44 Selaginella pulcherrlma fig. 45 Unless otherwise Indicated, the lumen of the megaspore Is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. E = exospore, I « Intlne, lsl = Inner separable layer

figure 42. Cross-section through exospore. X2650.

figure 43. Detail of outer exospore elements in an ornamental process. Note deep penetration of surface coating and its participation in formation of background ornamentation (arrows). X10,100.

figure 44. Cross-section through exospore of underdeveloped tetrad product. Note compression of wall spaces beneath ornamental processes, and lack of surface coating. X2650.

figure 45. Proximal surface with attached microspores. X300. Plate IX 7 9 80

Plate X. Selaginella pulcherrlma figs. 46-48 selaginellaaraentea figs. 49-51 Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. E « exospore, I « intine, isl « inner separable layer

figure 46. Cross-section through exospore. X2100.

figure 47. Detail of outer exospore elements and fragmented surface layer. Note reduced thickness of outermost exospore elements (arrow) as compared to those deeper in the exospore. X12,500.

figure 48. Inner separable layer. X31,500.

figure 49. Proximal surface. X300.

figure 50. Cross-section through exospore. X3300.

figure 51. Exospore surface. Note surface deposit and thin surface layer (arrows). X31,500 Plate X 81 82

Selaginella argentea type - Highly interconnected, extremely densely constructed. Sj. araentea

Selaginella argentea (Hall.) Spring

Megaspores of this species are approximately 200 um in equatorial diameter (Plate X, fig. 49). Laesurae extend to the equator, and the proximal surface is ornamented with closely spaced baculae which are formed from the siliceous surface coating (Plate X, fig. 50). Equatorial and distal ornamentation consists of a low coarse reticulum. The wall of this species is approximately 18 um in thickness (Plate

X, fig. 50). At the fine structural level, the wall is densely packed throughout, with some slight loosening immediately beneath the external surface. The construction is highly interconnected, with no observed free circular cross sections. Hall unit thickness is highest near the cell interior (0.4-0.6 um) and thins slightly (to 0.3-0.4 um) just beneath the surface (Plate X, fig. 51). At the surface is a continuous set of thin (0.1 um) laminae (Plate

X, fig. 51; arrows), which is coated by a discontinuous siliceous deposit. Some silica is present to a depth of

1.4 um. No thin inner separable layer was observed, but the sections examined were slightly oblique. 83

Selaginella ornata type - Thin, highly interconnected exospore;. well developed specialized outermost exospore which forms a nearly continous surface layer; spiny, continuous siliceous surface coating. ornata

Selaginella ornata (Hook. & Grev.) spring

These megaspores are approximately 240 um in equatorial diameter with straight elevated laesurae which extend up to 3/4 of the spore diameter (Plate XI, fig. 52).

The proximal surface is ornamented with verrucae of variable size which are least conspicuous near the proximal pole, and become more so toward the equator. The sporoderm ranges from 10-12 um in thickness, and is highly interconnected especially in the inner half of the exospore

(Plate XI, fig. 53); some sheet-like margins and circular cross-sections are present in the outer half of the exospore. Unit thickness ranges from 0.4-0.5 um throughout, except for the thin (0.2-0.4 um) nearly continuous surface layer (Plate XI, fig. 54; arrow). These thick wall units impart a dense appearance to the wall despite a significant degree of open space between the units. Compression in the bases of some ornamental elements (Plate XI, fig. 53) is probably associated with aborted megaspores. Unit disposition is roughly parallel to the spore surface throughout. A lamellated inner separable layer, 0.2-0.4 um in thickness (individual lamellae 15-45 nra), is present (Plate XI, fig. 56), and

Includes a single sheet of Inner exospore with a densely staining Inner surface and an Irregular thickness. The megaspore surface Is covered with a nearly continuous spiny siliceous coating (Plate XI, fig. 54, 55). 85

Plate XI. Selaqinella ornata figs. 52-56 g.elaqinellaervthropus figs 5 7 , 58 Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. E «= exospore, I = intine, isl ■= inner separable layer

figure 52. View of proximal surface. X250.

figure 53. Cross-section through exospore of slightly underdeveloped tetrad product. Note compression of chambers beneath ornamental element and irregular surface coating. X4150.

figure 54. Detail of outer exospore and surface coating. Note more uniform thickness of coating on this fully developed specimen and the nearly continous thinner surface element which parallels the megaspore surface (arrow). X12,500.

figure 55. Detail of surface coating illustrating its role in formation of background ornamentation (arrows). X12,500.

figure 56. Lamellations in the inner separable layer. X64,000.

figure 57. Detailed view of lamellated intine, and densely stained inner separable layer. X12,500.

figure 58. View of proximal surface. Thin linear streaks represent fungal contaminant. X200.

87

Selaglnella ervthropus type - Outer regions highly interconnected, inner exhibiting numerous circular cross- sections and bulbous sheetlike margins. Surface layer incorporating mainly laminar units and closed loops. S.

SEY-thrppMP, Sj. pllifera Selaqinella ervthropus (Mart.) Spring

Megaspores of this taxon measure approximately 300 urn in diameter in proximal view (Plate XI, fig. 58), and display a slightly elevated trilete suture which extends approximately 1/2 way to spore equator. Ornamentation is spinate to baculate on the proximal surface, except in the interradial regions where it is reduced to form a contact area. The exospore is approximately 20 urn in thickness and is densely constructed (Plate XII, fig. 59). The outer 1/3 of the exospore is highly interconnected, but the inner 2/3 is less so due to the presence of numerous sheet-like margins and circular cross-sections, compression is lowest at the outer surface, increases slightly in the center of the exospore, and becomes prominent near the inner surface.

Unit disposition is random outside, roughly parallel to the megaspore surface in the center, and more prominently parallel near the inner surface. Unit thickness is low at the outer surface (0.1 um), variable in the exospore center

(0.2-0.3 um), and low near the inner surface (0.1 um). An inner separable layer is present (1.5-3.5 um thick), 88 including a portion of the exospore 1-3 um in thickness. A

2-4 um thick intine is present in the figured specimen

(Plate XII, fig. 60). A discontinous siliceous coating occurs at the surface (Plate XII, fig. 59), and penetrates to a depth of 7 um.

Selaqilnenq uil if era a . Br. Megaspores of this taxon are approximately 200 um in equatorial diameter with thin elevated laesurae which extend nearly the entire spore radius (Plate XII, fig. 61).

Ornamentation consists of a fine reticulum. The exospore is approximately 15 um in thickness, and is loosely constructed (Plate XII, fig. 62). Laminar units are prevalent, but are not highly interconnected except in the outer 1/3 of the wall, where most of the closed loops are present. The remainder of the wall contains numerous sheet-like margins many of which are bulbous at their tips.

Some laminar units are quite laterally extensive parallel to the spore surface. There are also a number of large lacunae in the wall. Compression is low near the outer surface, and increases gradually to the inside, reaching a maximum in the continous innermost darkly stained exospore layer which is part of the inner separable layer (Plate

XII, fig. 64). Unit thickness is variable throughout the exospore, ranging from 0.1-0.3 um, with the average thickness slightly higher in the outer 5 um of the wall, 89 and slightly lower in the more compressed inner regions. A nearly continuous siliceous surface coating (Plate XII, fig. 63) approximately 1 um thick penetrates to a depth of

2 um, but may be more extensive if the surface coatings on units which occur throughout the wall (Plate XII, fig. 64y arrow) are identified as silica. 90

Plate XII. g.elaglneUa ervthropus figs. 59, 60 Selaainella oilifera figs. 62-64 gglfl.gineJ.la brevipes figs. 65-68 Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. E = exospore, I « intine, isl = inner separable layer

figure 59 Exospore cross-section excluding inner separable layer. X2650.

figure 60 Inner separable layer and adjacent intine. X12,500.

figure 61 Proximal view. X250.

figure 62 Cross-section through exospore. X3300.

figure 63 Outer exospore and fragmented surface coating. X10,100.

figure 64 Inner exospore and inner separable layer and adjacent intine. Note possible siliceous coating of wall elements (arrow). X16,000.

figure 65 Cross-section through exospore and intine. X3300.

figure 66. Highly compressed inner separable layer appressed to innermost exospore. The former line of separation of these two layers is visible (arrow). X25,100.

figure 67. Reduced outer wall elements (arrow) and surface deposit. X20,100.

figure 68. Folded intine and pseudo-endospore (pe) in suture region. The sutural groove extends out of the micrograph toward the upper right. X1650. Plate XII 91 92

Selaglnella brevipes type - Lateral fusion throughout, thick surface coat incorporating loose elements of reduced size; highly compressed but clearly delimited inner separable layer. £_». brevipes. S. elmeri. and Plana

Selaginella brevipes

The exospore in this species is approximately 20 um in thickness (17-21 um) and is densely constructed, displaying lateral fusion throughout (Plate XII, fig. 65). Unit thickness is consistent (at 0.3-0.4 um) throughout most of the central exospore with a slight thinning trend toward the inner surface. At the outer surface unit thickness decreases to 0.15 um (Plate XII, fig. 67) in association with the surface coating. Unit disposition is roughly parallel to the spore surface throughout the exospore.

Compression is low in the outer exospore but increases steadily to reach a maximum in the inner separable layer where units thin to 40 nm (Plate XII, fig. 66). The exospore and inner separable layer separate during the course of development resulting in different degrees of unit compression. This size difference allows the line of separation to be recognized (Plate XII, fig. 66; arrow). A surface silica deposits is present which penetrates to a depth of approximately 1 um, and incorporates a network of wall elements of reduced size (Plate XII, fig. 67; arrow).

Two additional layers are developed inside of the exospore 93 (Plate XII, fig. 68). One is faintly fibrillar, and approximately 1 um in thickness, except beneath the suture, where it thickens to 5 um. This may correspond to a pseudoendospore (pe) (Lugardon, 1971). Inside of this is a well developed, distinctly fibrillar intine, 3.6 um in thickness, folded beneath the suture, and composed of 8-10 lamellae.

elmeri Hieron.

The sporoderm of this taxon is approximately 13 um in thickness and densely constructed; lateral fusion is prominent (Plate XIII, fig. 69). Compression is evident in the outer portion of the exospore and increases steadily toward the inner surface where a finely lamellated inner separable layer contacts the 1 um thick intine (Plate XIII, fig. 71). Unit thickness ranges from 0.2 um just beneath the megaspore surface (Plate XIII, fig. 70), and thins gradually to 10 nm proir to structural obliteration in the inner separable layer. Unit disposition is parallel throughout. The surface coating is fragmented, but appears to be continuous, penetrating to a depth of 1 um, unless the dark surface coating of units throughout the wall is siliceous.

Selaginella plana (Desv.) Hieron.

Megaspores of this taxon are approximately 270 um in 94 equatorial diameter, with elevated laesurae which extend up to 7/8 of the spore radius (Plate XIII, fig. 72).

Interradial ornamentation is nearly smooth. In the equatorial and distal regions, low verrucae unite to form an imperfect reticulum. The exospore is approximately 17 um thick, and composed of densely packed units which display lateral fusion (Plate XIII, fig. 73). Compression is low near the megaspore surface, but increases steadily toward the inside, and is most obvious in the inner separable layer next to the low contrast intine (2 um thick; Plate XIII, fig. 74). Unit thickness in the inner separable layer is as low as 15 nm. Elsewhere in the exospore, unit size is variable ranging from 0.1-0.2 um.

The inner separable layer is separated from the more external exospore layers at some point in development.

This is evidenced by a distinct suture where the two wall layers have again come in contact at maturity (Plate XIII, fig. 73; arrow). Surface units (Plate XIII, fig. 73) are embedded in a siliceous coating which appears to be present in the interstices throughout the wall, and as a coating on all surfaces. 95

Plate XIII. Selaainella elroerl figs. 69-71 Selaginella plana figs. 72-24 Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. E = exospore, I = intine, isl = inner separable layer

figure 69. Cross-section through exospore and intine. X4150.

figure 70. Outermost exospore element morphology and surface deposit. X16,000.

figure 71. View of intine and highly compressed inner separable layer. X40,000.

figure 72. Proximal surface. X250.

figure 73. Cross-section through exospore. Mote line of separation between exospore and inner separable layer (arrow). X3300.

figure 74. Inner separable layer and lamellated ‘intine. X20,100. 6 6 |||X •leid 97

Selaglnella flssldentoldes type - Lateral fusion throughout, variable unit size, rod-like construction of innermost exospore units. £_l. fissidentoidesf S. vemensis

Sglaglnelia fissidentoides (Hook. & Grev.) Spring

The sporoderm of this species is approximately 27 um in thickness and is densely constructed (Plate XIV, fig.

75). Lateral fusion is high throughout the wall. The inner 1/3 of the wall may be composed of rod shaped elements 0.2-0.3 um in thickness. This inner region is compressed and strongly layered parallel to the megaspore surface. Compression decreases slightly while lateral fusion increases in the outer 1/2 of the exospore. Just beneath the surface units are randomly disposed and thickest (0.4 um). At the surface, units decrease slightly in thickness, and become roughly aligned perpendicular to the surface (Plate XIV, fig. 76). A silica deposit forms a discontinuous coating over the megaspore surface, and penetrates to a depth of 2.0 um. In places, the coating forms peaks which possess an ordered internal structure

(Plate XIV, fig. 77). An inner separable layer is present and displays irregular staining characteristics and an extreme degree of compression (Plate XIV, fig. 78). The exospore elements associated with the inner separable layer are probably rod-shaped and randomly disposed within planes parallel to the megaspore surface. 98

Plate XIV. S.elaqln

figure 75. Cross-section through exospore. X2650.

figure 76. Detail of surface layer. X12,500.

figure 77. Magnified view of siliceous ornamental process showing ordered arrangement of tiny crystallites. X25,100.

figure 78. Inner separable layer and intine. X12,500.

figure 79. Cross-section through exospore. Four distinct layers are indicated by letter designations (A-D). X2100.

figure 80. Outer exospore and surface coating. X10,100.

figure 81. Inner separable layer. Note lamellations. X10,100. Plate XIV 100

Selaainella vemensls (Sw.) Spring ex. Dene.

The main exospore of this taxon Is approximately 36 um

in thickness and possesses at least four distinct wall regions (Plate XIV, fig. 79; A-D). The inner most (D) is approximately 8 um thick, and is composed of loosely organized rod-like elements (0.2-0.3 um in thickness), which are disposed parallel to the spore surface. This layer grades into the second region (C; approximately 5 um thick) with increasing unit size (to 0.4-0.6 um) and increasingly random unit disposition. Units of the third region (B; approximately 15 um thick) increase to 0.4-1.0 um in thickness, become highly interconnected, and disposed largely perpendicular to the spore surface. The final region (A) extends roughly to a depth of 8 um, and is composed of units whose thickness decreases steadily (from the maximum thickness in the underlying layer) to 0.2 um.

A patchy surface deposit of opaline silica penetrates to a depth of 4 um (Plate XIV, fig. 80). In addition, an inner separable layer composed of 6-10 lamellae separates from the main exospore and includes a thin portion of exospore layer D (Plate XIV, fig. 81). 101

Sfilflglnella Intermedia type - Constructed primarily of laterally fused spherules but with some extreme lateral fusion into laminar units. intermedia

Selaginella intermedia (Bl.) Spring

Megaspores of this taxon are approximately 160 um in equatorial diameter, with elevated laesurae which extend to the equator (Plate XV, fig. 82). Proximal ornamentation is variable (echinate - baculate with some complex ridges).

The exospore is approximately 11 um in thickness between ornamental elements, and displays an extreme degree of lateral fusion resulting in a number of closed loops, especially in the inner 1/2 of the exospore (Plate XV, fig.

83). Units are disposed nearly parallel to the surface in the inner 1/2 of the wall, and at the outer surface, but are more random elsewhere. Unit thickness ranges from

0.4-0.8 um except at the outer surface where it thins to

0.3 um. Laminar units form a thin, discontinuous covering of the megaspore immediately beneath an irregular and discontinuous silica coating which penetrates to a depth of approximately 2 um (Plate XV, fig. 84). A 1 um thick, darkly staining inner separable layer is present (Plate XV, fig. 85), and is composed of 15-20 lamellae of irregular and variable thickness (16-200 nm). Also included in this inner separable layer is the innermost portion of the exospore which stains with a similar intensity (Plate XV, fig. 85; arrow). There are no clear zones of compression. 102

Plate XV. g.elaging 11a intermedia figs. 82-85 Selaglnella frondosa fig. 86 Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface.

figure 82. View of proximal surface. X250.

figure 83. Cross-section through exospore (except for inner separable layer). Note discontinuous surface deposit (arrow). X5150.

figure 84. Outermost exospore and laminar surface elements which parallel megaspore surface (arrow). X16,000.

figure 85. Lamellated inner separable layer and similarly stained innermost exospore unit (arrow). X25,100.

figure 86. View of proximal surface. X250. Plate XV 104

Selaginella frondosa type - Lateral fusion apparent; thick, highly compressed and convoluted innermost exospore; spaces in exospore. ^ frondosa

SelaqlneUa frondosa warb.

Megaspores of this taxon possess a broad cingulum

(Plate XV, fig. 86). The diameter of the body of the megaspore is approximately 150 um, and the cingulum, which completely surrounds the spore at the equator, is about 50 um wide. Slightly elevated laesurae extend to the outer edge of the cingulum. Proximal ornamentation is sparsely rugulate. The sporoderm is approximately 10 um in thickness between ornamentation (Plate XVI, fig. 87). Unit thickness is consistent, at 0.1-0.2 um throughout. To a depth of approximately 3 um, units are extremely densely packed, laterally fused, and disposed roughly parallel to the spore surface (Plate XVI, fig. 88). An irregular coating of opaline silica penetrates to a depth of approximately 0.5 um (Plate XVI, fig. 88). The central 6 um thick region varies with regard to unit density (Plate

XVI, fig. 87); some regions possess the density characteristics of the outer exospore, others are far less dense with large lacunae* In general, units are randomly disposed in the outer portion of this region, and become more circumferentially aligned toward the spore interior.

The inner separable layer (Plate XVI, fig. 89; isl) is approximately 1 um thick and composed of thinner laminar 105 units (to 50 nm), which are compressed to the point where structural boundaries are indistinct. 106

Plate XVI. Selaqlnella frondosa figs. 87-89 SfilaqlneUa lvaiiii figs. 9 0 - 9 2 Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. E = exospore, isl = inner separable layer

figure 87 Cross-section through exospore. X5150.

figure 88 Outer exospore and surface layer. X16,000.

figure 89 Inner separable layer. X20,100.

figure 90 Proximal view of collapsed megaspore showing coarse reticulate ornamentation and large size. X100.

figure 91 Outer exospore and surface elements. X20,100.

figure 92 Detail of inner separable layer. X16,000. Plate XVI 107 108

Selaglnella lvallil type - Obvious lateral fusion into ordered elements; thick exospore, gj. lvallii. 5. ttillflenpv.U Selaglnella lvallii (Hook. & Grev.) spring

The figured specimen (Plate XVI, fig. 90) is aborted, but the large megaspore size (700 um) characteristic of species whose megaspores possess this wall type is apparent. Ornamentation consists of a coarse reticulum with an intramural background pattern of small units whose precise morphology is unresolved. The sporoderm is approximately 75 um in thickness and is composed of four clearly differentiated regions (Plate XVII, fig. 93; A-D, D not shown). The innermost is an inner separable layer (D) 1 um in thickness (Plate XVI, fig. 92). This grades outward into a loosely organized region (Plate XVII, fig. 93; C) composed of rod-like strands of sporopollenin, 0.2-0.3 um in diameter, strongly layered parallel to the spore surface, but randomly disposed within layers. The next region (B) forms the majority of the sporoderm

(approximately 55 um) and has a patchwork appearance in section view. The interference patterns apparent in some patches are the result of a repeated subunit organization.

Each subunit, or ordered element, is about 0.3 um in thickness and possesses an orderly arrangement of protuberances in rows and columns on one side attached to a 109 plate-like backbone. These ordered elements occur in

stacks. Each patch seen in cross-section represents one

stack of repeated plate-like subunits. The overall patchwork organization results from the differing spatial orientations of adjacent stacks with respect to the spore surface. The outermost stratum (A) is composed of units of comparable size (0.2-0.3 um) to those of the underlying region, but with more random organization (Plate XVII, fig.

93); considerable lateral fusion is apparent. On the surface is a nearly continuous coating of silica which penetrates to a depth of 2 um (Plate XVI, fig. 91).

gQIflqinellfl willdenovii Megaspores of this taxon average 400 um in equatorial diameter (Plate XVII, fig. 94). Laesurae are elevated and extend 1/2-2/3 of the spore radius. Interradial ornamentation consists of low verrucae. Beyond the laesurae, ornamental elements are larger and more complex, in places assuming the form of irregular ridges. The exospore is approximately 40 um in thickness, and is densely constructed, and uncompressed throughout (Plate

XVII, fig. 95). The inner 12 um of the wall is composed of spherical - irregularly shaped interconnected subunits of variable dimension (0.3-0.6 nm; Plate XVIII, fig. 100).

The remainder of the wall is composed of smaller units

(0.2-0.3 um) which display considerable lateral fusion and 110

localized, but apparent organization into ordered elements which are occasionally organized into stacks (Plate XVIII, fig. 99). Wall units'decrease in size (to o.l um) and become embedded in a nearly continous surface coat of silica which penetrates to a depth of 4.5 um, and forms ornamental spines on the megaspore surface (Plate XVII, fig. 95; Plate XVIII, fig. 98). Some of these siliceous spines possess a laminated organization oriented parallel to their long dimension (Plate XVIII, fig. 98). An inner separable layer averaging 1.5 um in thickness includes a continous lamella of inner exospore on the outside, and is bounded by an intine toward the spore interior (Plate XVII, fig. 97). Ill

Plate XVII. Selaqlnella lvaiill fig. 93 Selaglnella wllldenovll figs. 94-97 Unless otherwise Indicated, the lumen of the megaspore Is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface. I >= intine

figure 93. Cross-section of exospore excluding inner separable layer. X2100.

figure 94. Proximal surface. X150.

figure 95. Cross-section of exospore excluding inner separable layer. X1650.

figure 96. Highly magnified view of background ornamental process showing ordered organization of component spherules. X130,000.

figure 97. Inner exospore, inner separable layer, and adjacent intine. X8050.

113

Plate XVIII. Sslaqinella wiiidenovii Unless otherwise indicated, the lumen of the megaspore is toward the bottom of the micrograph, and any lamellations or other linear features are parallel to the spore surface.

figure 98. Outer exospore and surface coating. Note participation of surface coating in background ornamentation. X10,100.

figure 99. Detail of central ordered region of exospore. X12,500.

figure 100. Unordered inner exospore elements. X12,500. Plate XVIII 114 115

RlsfiHgBipn A summary of sizes, thicknesses, and construction types of all of the species examined appears in table 1.

Minaki (1984) produced the most comprehensive analysis to date of the fine structure of megaspore wall in

Selaoinella. He limited his investigation to the level of the SEM. His classification scheme includes four groups of megaspore types and nineteen types in thirty-eight species, most of which differ in subtle ways. He relied, in some cases, on the presence of specific wall layers to define various types. A detailed critique of Minaki's scheme is not within the scope of this presentation, nor is it the aim of this author to attempt to place the taxa examined here into Minaki's groups. However, every effort has been made to group types together which have obvious structural similarities. Some additional observations of basic megaspore types are also in order. 116

Table 1. Megaspore size, vail and unit thickness, and construction type of nineteen species of Selaainella species of aegaspore thickness (us) unit construction Selaainella size vail unit laalnar spherules species exibiting strictly lasinar unit construction flabellata 250 um 18-22 0.3-0.5 + o usta NA 20 0.3-0.5 + 0 xlEldangula NA 18 0.15-0.25 + o lnaegualiftlla 300 ua 20-25 0.1-0.2 o pallucens 240 un 17 0.2-0.3 + o pulcherrlna 300 ua 25 0.1-0.2 + o acgsntga 200 US 18 0.35-0.6 + o omasa 240 US 10-12 0.4-0.5 •fr o aixthrapua 300 ua 20 0.1-0.3 + (outer) o pilifsra 200 ua 15 0.1-0.3 4- (outer) o Species exhibiting lateral fusion

PEfiYlPftl NA 20 0.3-0.4 o + elnerl HA 13 0.1-0.2 o + plana 270 ua 17 0.1-0.2 o +

Xlsffidentoldoa NA 27 0.2-0.4 0 + ysnfinola NA 36 0.2-1.0 o + intermedia 160 ua 11 0.3-0.B + + frondosa 250 ua 10 0.1-0.2 o + lYflllU 700 ua 75 0.2-0.3 o + villdenovii 400 ua 40 0.2-0.6 o + 117

Observations on basic megaspore types

one obvious wall type stands apart from the majority within the genus. This type, the gridwork type, is reported here in lvallii and fij. willdenovii. and is also present in g*. mvosurus (Martens, 1960a, 1960b; Stanier,

1966), Sj. qaleottli (Kempf, 1970; Tryon & Lugardon, 1978),

Sjl remotlfolla (Minaki, 1984), maroinata (Morbelli,

1977), and has undergone both structural (Taylor & Taylor,

1988) and developmental analysis (see Chapter V). It has also been discovered in megaspores from Cretaceous sediments (Bergad, 1978; Hueber, 1982; Huckriede, 1982;

Taylor & Taylor, 1988). In some other taxa (e.g., 5. brevlpes. vemensis), individual wall units approaching or possessing the organization of the "ordered elements" which make up this wall type can be detected. The resemblance in some may be coincidental, but in others it is unmistakable.

One obvious structural correlation with this distinct type is the fact that the thickest sporoderms in the genus

- up to 130 um - are built on this ground plan. An additional correlate involves the innermost stratum of the exospore. In all but one of the taxa so far examined, the inner exospore is constructed of apposed sporopollenin strands randomly disposed in a plane parallel to the spore surface. At a higher level of organization, all of the 118 taxa with the ordered walls produce one basal megasporangium per cone.

Undoubtedly the most interesting taxon with this

feature is wllldenovil. The megaspores of this taxon appear to display this pattern of organization only in part, with certain regions of the central exospore possessing a more random pattern of construction. In addition, the megaspore diameter and exospore thickness are intermediate between those taxa with strictly this organization and those without. Thus, as with many other seemingly unique characters, there may be a continuous range of variation with regard to this feature.

The only other well delineated structural type is circumscribed by Minaki's term labyrinthine (Minaki, 1984), and represents an extreme of unit interconnection, with all chambers within the exospore rounded and enclosed toy solid laminar walls. This is best exemplified by the s. flabellata type. Departures from this extreme in s. flabellata, and more prominently in other taxa, consist of both free sheet-like and/or rod-like units which occur inside otherwise enclosed chambers. The majority of species of Selaginella possess even larger numbers of free circular cross-sections and sheet-like margins (than S. flabellata), which define decreasingly regularly shaped enclosures within the wall, other variables which further complicate a comparison of wall ultrastructure are varying 119 degrees of unit density and unit disposition between taxa, as well as combinations of these parameters within different layers of the same wall.

Internal wall layers

Hearly all of the species investigated possess some type of distinct layering immediately internal to the exospore. In some cases (e.g., Sa. fissldentoides. S. brevloes^ the innermost extent of the exospore is compressed to such a degree that it is scarcely recognizable as a continuation of that stratum.

Alternately, there may be little or no compression or unit size decrease of that zone. In either case a thin (0.2-1.0 um) lamellated (lamellae 15-45 nm), darkly stained zone is consistently present in all investigated species. This zone does not appear to be continuous with the innermost exospore to which it is apposed, but the inner exospore may stain with a similar intensity (e.g., Intermedia. S.

RaHescens, ornata).

Additional layers which may be present are the intine

(endospore sensu Lugardon), which represents the pecto- cellulosic wall of the haploid cell, the pseudo-endospore, which is purported to form immediately prior to germination

(Lugardon, 1971b), and the mesospore, which will be discussed in Chapter V of this investigation. Several 120 species possess complex layering In the Inner exospore region. These layers are comprised of units from the Inner exospore, and could also represent remnants of either the

Intine or the pseudo-endospore. Several of these layers commonly separate as a unit from the exospore - a phenomenon to which the term Inner separable layer (Isl) has been applied. This Is possibly a manifestation of the function of these separable layers In maintaining a physiologically acceptable water balance with the ambient atmosphere by altering the volume of the protoplast locule.

Surface layers

A variety of structural arrangements occur at the outer megaspore surface as well. One component which is consistently present is some type of surface coating. It may be discontinuous (e.g., frondosa. S. ornata. S.

Intermedia. argentea) or continous (e.g., £L. pallfisesns, £*. flabellata. brevipes), and is probably involved in the formation of much if not all of the fine surface ornamentation of megaspores (Britton & Jenny,

1974). Silica in the spore coat of gj. aaleottii has been reported (Kempf, 1970) and documented (Tryon & Lugardon,

1978). Of the taxa included here, the surface coating of jL. brevipes was analyzed using energy dispersive analysis with x-rays (EDAX), and found to contain significant amounts of silicon. This is presumed to occur in the form 121

of amorphous silicon dioxide (opal). The crystalline state of this deposit was determined using x-ray diffraction to be nearly amorphous. High resolution photomicrographs of

&i. brevjpes show a highly fragmented, electron dense surface layer with little detectable organization. This is true of all taxa for which fresh material was utilized. An alternate appearance is apparent in high resolution photomicrographs of herbarium specimens which show tiny spherical crystals, like those illustrated for non-biogenic opal (Sanders & Darragh, 1971). In some instances (e.g.,

willdenovii. and fij. fissidentoides). a distinctly ordered substructure like that reported for precious opal is readily detectable. This clear dichotomy of surface deposit appearances serves as a reminder that caution must be exercised when evaluating ultrastructural differences for which there may be a multitude of causative factors.

In this instance, the differences are likely to represent a recrystallization phenomenon, especially in light of the fact that only those spores removed from older herbarium sheets possessed a surface coating in the form of tiny spherical crystals.

Various relationships exist between the siliceous coatings and the sporoderm surface elements. Some surface elements unite to form a smooth covering over all or part of the megaspore - often constructed of units which are slightly thinner than those in the immediate vicinity 122

(e.g., Sj. arcrentea , £*. ornata). or sometimes of units of similar thickness (e.g., Sj. intermedia. S. frondosa. S. pallescens). Most of the remaining species examined here show a marked decrease in surface unit size with silica deposits external to, and in the interstices of the wall elements to varying depths (e.g., g*. flabellata.

Xfimensis, §_*_ frrevlpes, g. flssldentoldes. g*. lvallill. In extreme cases, an extensive network of thin wall elements are embedded in a thick coating of silica (e.g., s. flabellata). It is interesting to note that in these cases, where the outermost exospore units are thin and sparse, and thus the surfaces the most porous, the silica deposit forms a nearly continuous coating. This evidence could be used to suggest that the siliceous coating acts as a sealing agent in the absence of a more continuous surface of sporopollenin. CHAPTER V

WALL DEVELOPMENT IN SELAGINELLA

Introduction

A considerable degree of controversy has arisen over the development of the various wall layers present in megaspores of Selaqinella. The earliest account (Fitting,

1900) suggested that the formation of the main sporoderm wall (exospore) in Selaqinella took place in the absence of protoplasmic contact until late in development. Soon thereafter, Lyon (1901) published an initial account of megaspore wall development in which she agreed with

Fitting's conclusions. Later she reevaluated her findings and reversed her conclusions (Lyon, 1905), suggesting that the separation between the two principal layers of the sporoderm (exospore and mesospore) was an artefact of fixation.

Pieniazek (1938) was able to cause a separation of the inner body (Fitting's mesospore) from the outer by immersion in a plasmolyzing solution. This was used by

Sievers and Buchen (1970) as evidence for the artefactual nature of the separation of the two portions of the megaspore wall.

123 124

Pettitt (1966a) examined the mesospore histochemically. He demonstrated that 1) megaspores which contain thick mesospores lack a protoplast at maturity and

2) mesospores in this case stain positively for proteins and carbohydrates, in addition to the lipids of the sporopollenin. This suggests that the mesopore sensu

Fitting is at least in part represented by the degraded remains of the aborted protoplast.

All of these results were summarized by Ruf (1975).

The results of the present investigation, in combination with Ruf's conclusions, shed considerable light on the true nature of the mesospore in Selaqinella (see p. 166).

Mature megaspore walls of Selaqinella were among the first biological entities to be examined with the transmission electron microscope (Afzelius, et al., 1954).

However, megaspores of Selaqinella have received limited attention with regard to exospore development. Pettitt

(1971a) first reported aspects of wall development in the context of a composite overview of sporoderm development in the pteridophytes as a whole, while Buchen and Sievers

(1978b) addressed formation of the wall, but only detailed early stages of development. Until now, there have been no studies which have specifically addresses the stages between wall initiation and attainment of mature form.

For this investigation, wall development in

Selaqinella galeottli Spring will be traced. Megaspores of 125 this species were examined with transmission electron microscopy by Kempf (1970) and found to possess the same unusual sporoderm structure first observed in JL*. mvosurus

(Stanier, 1966). Further interest was generated when several fossil megaspores of Cretaceous age (65-120 million years old) were found to possess an identical organization

(Bergad, 1978; Hueber, 1982; Huckriede, 1982; Taylor &

Taylor, 1988) thus providing the opportunity to consider this unusual structural type from an evolutionary perspective.

At the inception of the developmental portion of this investigation, a different species of Selaqinella appeared to offer the greatest promise as a model system. Specimens figured as sulcata in the literature (Morbelli, 1977) possess the ordered wall organization which was to be studied from a developmental perspective. In addition, contrary to literature reports which suggest that species with this wall type possess only one basal megasporangiuro per cone, the available specimens attributed to sulcata produce long cones with interspersed patches of micro- and megasporangia. Within sporangial patches, close to the fertile shoot tips, successive megasporangia gradually decrease in size and stage of development. All required stages appeared to be present in such a specimen. Upon examination, the available specimens of s.*. sulcata did not possess the ordered wall construction type. The 126 information on this species, however, provides independent substantiation of the validity of the processes established based on the study of gj. caleottii. as well as strengthening the case for a more widespread developmental mechanism.

In addition to establishing the developmental events which transpire during megaspore wall formation in

Selaqinella. the following questions will also be considered: 1) How does the tapeturn change through the course of wall development? 2) How does the nature of the secreted lipids change? Specifically, do they appear to possess a preformed shape prior to apposition to the accreting wall? 3) How are the various internal bodies referred to as mesospores related to one another? 4) Does control of exospore patterning lie with the gametophyte, the sporophyte, or both? 5) Do the developmental pathways of wall growth differ depending on the mature form of the wall, i.e. are there basic developmental processes, and are they consistent throughout the genus? 6) Are there correlations with wall construction, e.g. with megaspore size, and what might this suggest as to the functional significance of the various wall types?

In his description of megasporogenesis in Selaqinella.

Pettitt (1971b) introduced several terms which will be employed in this investigation. Prior to meiosis, two distinct types of cells exist in the locule of the 127 developing megasporangium. Pettitt used the terms viable and non-viable megasporocytes for these cells. The former are slightly larger with a fairiy well developed plastid associated with the nucleus, centrally located mitochondrial profiles, large amorphous inclusions and densely packed free ribosomes. Cells of the latter type are slightly smaller, have reduced plastids, dispersed mitochondria, fewer ribosomes, and evidence of secretory activity in the form of dilated smooth endoplasmic reticulum and active Golgi bodies. 128

Description

Selaqinella sulcata (Poiret) Martius sens. lat.

The most informative developmental sequence of this taxon supplies information on four stages of megaspore wall morphogenesis. The following descriptions will proceed from the earliest to the latest stage of development, as determined by megaspore size.

S.elaqinellfl sulcata stage 1 The wall at this stage ranges from 4-6 um in thickness

(Plate XIX, fig. 101), and is uniformly granular throughout the exospore; individual wall units range from 0.01-0.03 um in size (Plate XIX, fig. 104). The units are densely packed and highly interconnected. Considerable quantities of weft-like wall material are provided by the persistent megasporocytes (Plate XIX, fig. 102), and similar material throughout the lumen of the sporangium can be seen accreting onto the megaspores. At the accreting surface, lightly staining patterns are apparent as if some informational framework is becoming more visible due to the accumulation of stainable material (Plate XIX, fig. 104).

A mesospore is also present in this specimen, the wall of which ranges from 4-9 um in thickness, and is distinctly two parted (Plate XIX, fig. 103).. The outer 2 um is indistinguishable from the exospore, while the units of the 129 remainder of the mesospore wall gradually elongate into fibrous strands that run parallel to the megaspore surface.

Hew material seems to be accreting onto the inner surface of the mesospore, and the cell lumen contains thin irregularly shaped wefts similar to those present throughout the sporangium. No viable cytoplasm can be detected in any mesospore bearing spore, only an occasional patch of residue near the proximal pole.

SfilflglneUfl sulcata stage 2 The exospore thickness of So. sulcata megaspores at this stage of development ranges from 4-14 um between ornamental elements, with an average thickness of approximately 8 um (Plate XIX, fig. 107). The wall is composed of densely packed and highly interconnected units

0.03-0.06 um in size (Plate XIX, fig. 105). The central region of the wall has slightly larger units (0.04-0.06 um), while the interior and exterior surface units are, on average, smaller (0.03-0.05 um). The apposition of weft-like material to both the inner and outer surfaces of the exospore (Plate XIX, fig. 105, 106) suggests that unit formation is taking place at these locations. A mesospore is present in this specimen with a wall 7-13 um thick

(Plate XIX, fig. 107). The wall is similar to the main sporoderm in its outer portion but becomes fibrillar toward the spore center. No cytoplasm is evident, but apposition 130 of wall material appears to be taking place. Tapetal cells and persistent megasporocytes show extensive cytoplasmic disorganization, but are actively secreting weft-like material nonetheless.

Selaqinella BUlcata stage 3

The megaspore wall at this stage of development is approximately 9 um thick, densely constructed, and highly interconnected (Plate XIX, fig. 108). Hall unit thickness ranges from 0.1-0.2 um throughout, except for a slight thinning toward the external surface. Some weft-like material is present on or near both the external and internal surfaces, but accretion of new wall material does not appear to be taking place at either position. An 3 um thick inner separable layer is present, highly compressed, and tightly adhered to the inner surface of the mesospore

(Plate XIX, fig. 112). An external portion, composed of units much like the exospore, grades inward to a region composed of lamellae, 0.02-0.04 um in thickness.

Immediately internal to this inner separable layer is the plasmalemma which lines the cell lumen and bounds a thin layer of cytoplasm in which various subcellular components are embedded. A distinct fibrillar deposit is present on the megaspore surface which is probably the result of early silica deposition. 131

Selaqinella sulcata stage 4

At this stage of development, the megaspore wall ranges from 14-18 um in thickness and is composed of densely packed interconnected units which vary somewhat in size (Plate XIX, fig. 109). Units are thickest (up to 0.25 um) approximately 3 um beneath the outer megaspore surface.

Unit size decreases abruptly to average 0.1 um at the megaspore surface, and more gradually, through 0.2 um, to

0.1 um at the inner megaspore surface. Unit disposition is roughly parallel to the circumference throughout the wall, and a low degree of compression may have affected the inner

5 um of the exospore. An inner separable layer 3 um in thickness is present, and is tightly appressed to the inside of the exospore (Plate XIX, fig. 109). The outer half of this layer is similar to the main sporoderm but highly compressed, while the inner half appears more fibrillar. There does not appear to be any surface residue present on megaspores of this developmental stage.

Selaqinella sulcata mature megaspores

Mature megaspores are approximately 200 um in equatorial diameter (Plate XIX, fig. Ill). Proximal ornamentation consists of small verrucae which are present on the thick elevated laesurae as well as the interradial areas. In the equatorial and distal regions the verrucae become much coarser. The mature exospore of S_s. sulcata is 132 approximately IS um in thickness, and displays considerable lateral fusion (Plate XIX, fig. 110). Gradually increasing compressional forces in the inner exospore obscure unit boundaries and create a lamellated organization at the inner surface (Plate XIX, fig. 110); units as thin as 40 nm are measurable. Unit size in the uncompressed outer region ranges from 0.2-0.3 um, except at the surface, where units as small as 75 nm are detectable (Plate XIX, fig. 113), embedded in a silica surface coat which penetrates to a depth of 4 um. A well developed 5 um thick intine is formed by this stage (Plate XIX, fig. 110), around a protoplast which contains few recognizable organelles. 133

Plate XXX. Selaqinella sulcata stage l, figs. 101-104; stage 2, figs. 105-107; stage 3, figs. 108, 112; stage 4, fig. 109; nature megaspore, figs. 110, 111, 113 The following key applies to all of the figures in the developnental portions of this contribution: E « exospore; H «= mesospore; pn « persistent negasporocyte; pmw « persistent negasporocyte wall; 0 « outside of respective wall layer (toward the sporangium locule); si « sporangium locule; T ■= tapetal cell(s); L = lumen of megaspore. Other letter designations will be included in the figure captions. The section staining schedule is indicated at the end of each figure caption (time in minutes of staining in potassium permanganate-time in uranyl acetate-time in lead citrate).

figure 10l Cross-section of immature megaspore inside of sporangium. Note degraded tapeturn, persistent megasporocytes, and well developed mesospore. Size bar *» Sum. 10-10-2 .

figure 102 Weft-like material accreting to form the exospore. Individual wall units are 0.01 -0.03 um in thickness at this stage. Size bar « 0.2um. 10-10-2.

figure 103 Cross-section of accreting exospore and mesospore. There are four accretion surfaces visible (arrows). Size bar = Sum. 10-10-2 .

figure 104. Weft-like and granular material between persistent megasporocyte and accreting exospore, size bar = 0.2um. 10-10-2.

figure 105. Cross-section through accreting exospore. Two accretion surfaces are visible (arrows). size bar ** 2um. 10-20-4.

figure 106. Weft-like wall material accreting to form exospore. Individual wall units are 0.03 -0.06 um in size at this stage, size bar = 0.2um. 10-20-4.

figure 107. Cross-section of immature megaspore. Size bar ® Sum. 10-20-4. 134

figure 108. Cross-section of exospore. Wall units are 0.1-0.2 tun at this stage. Size bar ** 2um. 0-15-7.

figure 109. Exospore of viable megaspore with highly compressed mesospore. Wall units average 0.2 um at this stage, size bar » Sum. 10-10-2 .

figure 110. Cross-section through exospore. The zone of separation between the compressed mesospore and the exospore is no longer discernible. Note well developed intine. Size bar *> 5um. 0-5-2. figure 111. Proximal surface of mature megaspore. Size bar “ 50um. figure 112. Cross-section of compressed mesospore in a viable megaspore. Note plasma membrane (arrow) and adjacent parietal cytoplasm. Size bar *» lum. 0-15-7. figure 113. Outer exospore and surface deposit, size bar - lum. 0-5-2.

136

Selflginella qaieottii Unlike the situation in gj. sulcata. where a single cone nay contain a sequence of several negasporangia at different developnental stages, cones of fij. aaleottii possess only a single basal negasporangium. Relative developmental stages must, therefore, be based on some common measurement which in this case is megasporangium size. The following descriptions, procede from smallest to largest megasporangium, or from earliest to latest developmental stage.

Selaginella qaieottil stage 1; 135 um

Sporangia 135 um in diameter are clearly differentiated into three regions. The innermost is a central zone consisting of isodiametric cells which display early stages of autolysis including several small vacuoles per cell and sequestered portions of cytoplasm (Plate XX, fig. 114) ~ this represents the sporogenous tissue. The middle region is 14-16 um in thickness, parallels the sporangium surface and consists of radially aligned cells with dense unvacuolated cytoplasm - this region will differentiate into the tapetum. Each cell has a well developed nucleus with a densely staining nucleolus and a closely associated plastid as well as the usual complement of remaining organelles. The outermost region is a 137 multilayered sporangium wall which is characterized by the presence of prominent intercellular spaces.

Selaainella qaieottii stage 2; 200 um

By the time megasporangia reach 200 um in size, the irregularly Bhaped cells of the sporogenous tissue have almost completely separated. Many cells are at intermediate stages of autolysis with numerous membranous vacuolar inclusions (Plate XX, fig. 117).

Each sporangium at this stage of development contains one cell which differs from all of the remaining cells in the sporangium. This cell represents the functional megasporocyte, or megaspore mother cell, which will divide meiotically to give rise to a tetrad of megaspores. It is

17 um in diameter, appears less vacuolate and possesses denser cytoplasm than adjacent cells (Plate XX, fig. 118).

In addition, the cell has become nearly spherical and enclosed itself in a special cell wall 0.2-0.7 um in thickness inside the old fibrillar wall (Plate XX, fig.

119).

The locule of the sporangium is filled with granular material which appears weft-like when in contact with the surface of a megasporocyte (Plate XX, fig. lie). The source of the granular material is not known but is presumed to be the result of break down of cell wall material between the megasporocytes. Darkly stained 138 tapetal cells average 15 um in radial dimension but do not appear to be actively secreting material into the sporangial chamber (Plate XX, fig. 115).

Selaolnella ctaleottii stage 3; 238 um

Tapetal cells at this stage of development (Plate XX, fig. 120) are approximately 24 um in length parallel to the radius of the sporangium, and possess dense ribosome rich cytoplasm (Plate XX, fig. 121). They appear by this stage to have fibrillar material and wefts moving away from the surface of the cell wall (Plate XX, fig. 121). No information is available on cells inside of the tapetum. 139

Plate XX. Selacrinella qaieottil stage 1, fig. 114; stage 2, figs. 115-119; stage 3, figs. 120, 121; stage 4, figs. 122, 123 E = exospore; H = mesospore; si = sporangium locule; T « tapetal cell(s); pm = persistent megasporocytes. Key to staining time designations; KMn04-UA-Pb Cit

figure 114. Early differentiation of a megasporangium. Relatively vacuolated cells to the lower left represent the sporogenous cells; elongated cells of nascent tapetum are radially oriented; cells with prominent intercellular space (upper right) are of the sporangium wall. Size bar “ 5um. 2-20-10.

figure 115. Inactive tapetal cells. The center of the sporangium is toward the upper right. Size bar = Sum. 2-5-2.

figure 116. Megasporocyte wall showing the weft-like nature of the wall material (arrow). Size bar « 0.2um. 2-5-2.

figure 117. Separating sporogenous cells. The degraded condition of these cells suggests that they would have aborted. Size bar = Sum. 0-20-10.

figure 118. Megaspore mother cell (MMC). Size bar = Sum. 5-20-10.

figure 119. Megaspore mother cell wall. The outer fibrillar layer is probably the primary cell wall, while the inner was undoubtedly deposited by the MMC following separation from neighboring cells. Size bar « 0.2um. 5-20-10.

figure 120. Tapetal cells. Size bar “ 5um. 2-20-10.

figure 121. Possible secretion at the inner surface of the tapetum. Size bar = 0.5um. 2-20-10.

figure 122. Oblique cross-section of developing megaspore in glancing section. The exospore is differentially thickened on the proximal and distal surfaces. A rudimentary suture is apparent on the proximal surface (arrow). The darker outer exospore (El) appears at high-mag in figure 123. Size bar = lOum. 2-5-2. figure 123. Outer exospore (El) surface with little detectable ordered structure. Size bar 0.2um. 2-5-2.

142

SfilagjtngliU qaieottii stage 4; 288 um At this stage of development, the four members of the functional tetrad have expanded to a size of 31-42 um in maximum diameter (along the polar axis). The exospore is two parted (El and E2) and is differentially thickened on the proximal and distal surfaces (Plate XX, fig. 122). The difference in the inner zone (E2), which is composed of randomly oriented strands of material 60-80 nm in diameter, is slight, and is obscured in the proximal region due to continuity with the adjacent mesospore. The difference is more pronounced in the external densely stained layer (El) which is as thick as 1.9 um near the distal pole, 1.2 um in the lateral region and is just condensing on the proximal surface. There is little detectable structure in this outer layer aside from a slight hint of coarse lamellations that parallel the megaspore surface (Plate XX, fig. 123).

The megaspores contain bi-layered mesospores (M), 4-7 um thick, which surround cytoplasmic remnants (Plate XXI, fig.

124). The outer half of the mesospore (Plate XXI, fig.

125) is similar in construction to the inner exospore and is continous with it on the proximal surface. The inner half of the mesospore wall is darkly stained and composed of a mixture of short, thick fragments and fibers of wall material (Plate XXI, fig. 125).

During this stage of development, active secretion of 143 wall material by both tapetal cells (Plate XXI, fig. 126-

129) and persistent megasporocytes (Plate XXI, fig. 129,

131) takes place, as well as accretion of this material onto the developing megaspores (Plate XXI, fig. 130).

Surfaces of active secretion are bounded by a lightly staining zone immediately outside of the plasmalemma (Plate

XXI, fig. 127, 128, 131). The innermost extent of this coating is marked by a continuous surface where the secreted material begins to take on stain and forms beaded wefts and individual droplets of wall material (Plate XXI, figs. 127, 128). Wefts from persistent megasporocytes

(Plate XXI, fig. 128) range from 32-60 nm in thickness between the beads which themselves are 90-130 nm in size.

Wefts from tapetal cells (Plate XXI, fig. 127) appear to be slightly thicker than those from megasporocytes, ranging from 50-90 nm thick between beads, which themselves are 75-

120 nm in thickness. Droplets not attached to wefts are variable in size ranging up to 270 nm. Little or no fibrillar construction suggestive of a cellulosic cell wall can be seen on either tapetal cells or persistent megasporocytes where they interface the locule of the sporangium. Near the center of the tetrad of developing megaspores three intermingled sets of beaded wefts are visible (Plate XXI, fig. 132), one set roughly parallel to each of the accreting surfaces. Individual droplets are often seen impinging on the surfaces of the megaspores 144

(Plate XXI, fig. 130, 133) but never wefts. It is possible that the wefts disaggregate into droplets prior to depostion. 145

Plate XXI. Selaainella qaieottii stage 4 E = exospore; H - mesospore; pm = persistent megasporocyte; si = sporangium locule; T = tapetal cell(s); L ■ lumen of the spore. Other letter designations will be included in the figure captions. Staining schedule key; KMn04-UA-Pb Cit

figure 124 Equatorial cross-section of a developing megaspore. The mesospore is apparent in this specimen, and encloses cytoplasmic remnants (arrow). Size bar = lOum. 2-5-2.

figure 125 Bi-layered mesospore wall, size bar = 2um. 2-5-2.

figure 126 Actively secreting tapetal cells. Note continuous secretion surface extending into embayments between separated tapetal cells (arrow), size bar - 2um. 2-5-2.

figure 127 Detail of secretion at the inner surface of a tapetal cell. The continuous surface from which material is blebbing off into the sporangium locule (arrow) is separated from the tapetal cell by a low contrast area in the position of the cell wall. Size bar «=* lum. 2-5-2.

figure 128. Tapetal cell surface. Note well developed beaded wefts and continuous secretion surface (arrow). Size bar « 0.5um. 2-5-2.

figure 129. Oblique section through tapetum. All of the cells viewed (except the persistent megasporocyte in the upper left) are tapetal cells. Due to the oblique plane of section, moving from upper left to lower right each row of cells is viewed at a plane of section closer to the sporangial wall. This view shows the size and depth of some of the embayments in the tapetum (em), and the active secretion taking place at all surfaces. Size bar = Sum. 2-5-2. 146

figure 130. Droplets accreting onto the developing megaspore surface. Size bar ** 0.5um. 2-5-2.

figure 131. Persistent megasporocyte. A low contrast zone and continuous secretion surface (arrow) similar to that which is present in tapetal cells also surrounds these active cells. Size bar *» 2um. 2-5-2.

figure 132. Inside sporangium locule between three tetrad members. Many beaded wefts are present but only droplets can be seen in contact with accreting surfaces (arrows). Size bar «= 2um. 2-5-2. figure 133. Beaded wefts near the accreting surface of a megaspore. Size bar « 0.5um. 2-5-2.

148

Selacrlnella qaieottil stage 5; 340 um

Negasporangia this size contain developing megaspores approximately 80 um in maximum diameter. The exospore is divided into two distinct layers (Plate XXII, fig. 134), the inner of which (E2) is 3.4-5.4 um thick, and constructed of randomly oriented fibers (Plate XXII, fig.

135). The outer region (El) is 2.0-4.8 um in thickness and has a patchwork organization which reflects a three dimensional mosaic construction. Units of this region are densely packed, highly interconnected, and 22-30 nm in size. Material is accreting onto the megaspore surface

(Plate XXII, fig. 136) with the same geometric organization visible in the adjacent wall region. Larger and more lightly staining structureless masses are occasionally present at the accreting surface (Plate XXII, fig. 137).

The material which impinges on the accreting wall sometimes displays a somewhat beaded appearance (Plate XXII, fig.

137), but more often is fibrous or irregularly shaped.

Surface ornamentation is in the process of forming at this stage, and the units which comprise the hollow surface projections do not possess the same orderly arrangement as the adjacent sporoderm (Plate XXII, fig. 138).

A mesospore is present in the specimens studied here

(Plate XXII, fig. 134). The mesospore ranges from 24-28 um in maximum diameter and has a two parted wall. The outer

1-3 um is indistinguishable from the inner exospore (E2) 149 and Is continuous with it at the proximal pole. The inner

3-4 um is densely packed, darkly stained, and composed of short angular and fibrous wall material (Plate XXII, fig.

139). The mesospore encloses shrunken cytoplasmic remnants, some with recognizable subcellular components

(Plate XXII, fig. 139). Accretion of wall material appears to be taking place on both the inner and outer surfaces of the exospore. A rudimentary germinal suture is obvious at this stage of development (Plate XXII, fig. 134, arrow).

Both tapetal cells (Plate XXII, fig. 140, 141) and persistent megasporocytes (Plate XXIII, fig. 143) are actively secreting material. Cells of the tapetum are approximately 20 um in length parallel to the radius of the sporangium. Embayments commonly form between adjacent tapetal cells by separation along their common cell wall

(Plate XXII, fig. 140). Secretion takes place all along the boundaries of these embayments. A nearly continuous sheet of overlapping beaded wefts covers all the free tapetal surfaces (Plate XXII, fig. 140). As the material moves away from the tapetal region, it separates into small fragments of irregular shape (Plate XXII, fig. 141).

Secretion from persistent megasporocytes appears to conform to the pattern just described fortapetal cells, except that material is liberated on all sides into the locule of the sporangium (Plate XXII, fig. 142? Plate XXIII, fig. 143). 150

Plate XXII. Selaainella galeottil stage 5 E ■ exospore; M *= mesospore; si = sporangium locule; T « tapetal cell(s). Other letter 'designations will be included in the figure captions. Section staining schedule; KMn0 4 ~ UA-Pb Cit

figure 134. Immature megaspore with thick-walled mesospore. The outer exospore (El) has a slightly lower affinity for stain than in the previous stage. Fibrillar material is present throughout the sporangium locule. The arrow marks the position of the developing germinal suture. Size bar « lOum. 2-20-10.

figure 135. Bi-layered exospore showing fibrous structure of inner exospore (E2). Size bar *» 2urn. 2-20-10.

figure 136. Highly magnified view of accretion at the surface of the ordered wall region, size bar = O.lura. 2-5-2.

figure 137. View of accretion surface showing ordered structure of outer exospore, fibrillar -irregular material approaching wall, and globules at the accreting surface (arrow). Size bar = 0.2um. 2-5-2.

figure 138. Hollow ornamental element with diminished ordered organization from that of the outer exospore. Size bar = lum. 2-5-2.

figure 139. Cytoplasmic remnants inside of well developed mesospore. Size bar = 5um. 2-20-10.

figure 140. Vacuolated tapetal cells with darkly stained membranous organelles and secretion at all surfaces adjacent to sporangium locule. Size bar = 5um. 2-5-2.

figure 141. Two adjacent tapetal cells actively secreting wall material. Note concentration of smooth ER (arrow) adjacent to the secretion surface. Size bar ** lum. 2-5-2.

figure 142. View of secretion surface showing low contrast region immediately adjacent to the cytoplasm (arrow). Also note the weft-like material near the secretion surface, but predominance of irregular material in sporangium locule. Size bar = 0.2um. 2-5-2. mm

IIXX •tBId 153 g.elflqineUaqaieottii stage 6; 500-544 um Sporangia at this stage of development contain megaspores with walls 16-20 um in thickness (Plate XXIII, fig. 144). Units are densely packed (nearly solid in places), ranging from 90-120 nm in size in the outer exospore (El), which is 4-8 um thick, and 60-90 nm in the inner exospore (E2; overall exospore thickness approximately 10 um). The formation of new ordered elements at the megaspore surface appears to involve the coagulation of coarser material than has been seen to this point (Plate XXIII, fig. 145). Interconnected lamellae 6-

18 nm in thickness are abundant in the region immediately surrounding the megaspores, and can be seen in the interstices of recently formed units in the densely packed outer wall (Plate XXIII, fig. 146). Also forming at the surface are ornamental ridges (Plate XXIII, fig. 144, 147).

At this stage, the ridges are constructed of loosely packed, unordered units intermixed with a considerable number of coarse lamellae (Plate XXIII, fig. 147). The mesospore has retained the bi-parted organization expressed in earlier stages. It has, however, thickened to approximately 15 um, particularly in the inner region

(Plate XXIII, fig. 148).

Material that will form the wall is secreted through the activity of tapetal cells (Plate XXIII, fig. 149) and 154 persistent megasporocytes. Tapetal cells are approximately

14 um in longest dimension (parallel to the radius of the sporangium), and have a substantial number of vacuoles in and among the usual complement of organells, including some darkly staining mitochondrial profiles. Deep embayments are common between adjacent tapetal cells. Secretion of the lamellae present throughout the sporangium, is evident near the inner face of the tapetal cells (Plate XXIII, fig.

150). A lightly staining zone occupies a position between the membrane of tapetal cells and the forming lamellae

(Plate XXIII, fig. 150). Stacked ER is evident just beneath the plasma membrane. 155

Plate XXIII. Selaqinella qaieottil stage 5, fig. 143; stage 6, figs. 144-150. E ■ exospore; si ** sporangium locule; T ■= tapetal cell(s). Other letter designations will be included in the figure captions. Section staining schedule; KMn0 4 -UA-Pb Cit

figure 143. Actively secreting (arrow a secretion surface), highly vacuolated persistent megasporocyte. Size bar = 2um. 2-5-2.

figure 144. Bi-layered exospore and ornamental element. Size bar «• 5um. 2-5-2.

figure 145. Fibrous wall material at outer surface of ordered unit formation. Size bar = 0.2um. 2-5-2.

figure 146. Internal region of ordered wall. Fibrous wall material can be seen in the interstices of these wall units. The separate nature of the units is a function of the plane of section* Size bar ** 0.2um. 2-5-2.

figure 147. Loosely constructed ornamental element. Size bar = 2um. 2-5-5.

figure 148. Detail of inner mesospore wall. Size bar « 5um. 2-5-5.

figure 149. Tapetal cells. Size bar = 2urn. 2-5-5.

figure 150. Tapetal cell surface showing low-contrast region immediately external to cytoplasm, and continuous secretion surface with blebbing off of fibrous material. Size bar “ o.5um. 2-5-5.

157

Discussion

S. sulcata - wall development

In the course of development, both megaspore diameter and exospore thickness increase. The structural basis for the increase in exospore thickness lies at the level of the units which compose the wall. This phenomenon of growth is summarized diagramatically for sulcata (Plate 24, fig.

1-4). Megaspores with an average wall thickness of 5 urn

(stage 1), 8 urn (stage 2), 9 urn (stage 3), and l€ urn (stage

4), have wall units 20 nm, 45 nm, 100 nm, and 200 nm respectively. Accretion of new wall material also takes place in the earliest stages on both the external and internal sporoderm surfaces, as well as between the mesospore and exospore. The fibrillar material has a consistent appearance in the early stages examined here

(stage 1 and 2), is present throughout the sporangia1 locule, and is secreted by tapetal cells and persistent megasporocytes, both of which are at advanced stages of degeneration.

When the sporoderm reaches approximately 9 um in thickness (stage 3) synthesis of wall material seems to be largely completed. With no new influx of wall material, the sole mode of wall thickness increase from this point on must be enlargement of component units. A continuous membrane is present in all subsequent stages as well as an 158 inner separable layer with an ultrastructure consistent with that which would result from the compression of the mesospore of the earliest developmental stages. Wall units of mature megaspores reach 300 nm in thickness, but the megaspores display no observable increase in wall thickness over the immediately preceding stages. This increase in component unit size without a compensatory increase in wall thickness results in a high degree of compression which is particularly evident in the mesospore.

structural basis for expansion

Pettitt (1971, 1977) utilized £L±. sulcata as a model system for the analysis of developmental mechanisms in heterospory. He noted examples of post meiotic regression in this species in the form of sporangia possessing smaller tetrads in and among the mature megasporangia, as well as a regular occurrence, in each mature sporangium, of two smaller aborted megaspores with two larger functional ones.

Several megasporangial rows with distally decreasing megaspore size (aborted by post-meiotic processes) were utilized in this investigation as stages in sporoderm development. This is justified since the basic informational pattern for the sporoderm is established soon after meiosis, and possibly even prior to breakup of the tetrad (Pettitt, 1971a, 1977). Expansion of the spore and its enveloping sporoderm proceeds in the presence or 1S9 absence of a viable megaspore protoplast. What was unclear, until now, was the precise structural basis for this expansion, and the changes which take place in the sporoderm construction. It is now clear that the informational framework is present early on, and that expansion involves a massive synthesis and accretion of sporopollenin at the surface, concomitant with a uniform thickening of the wall units already present, with or without addition of new material. In this context it is important to differentiate between wall development - the establishment of an informational framework - and wall expansion, or growth - addition of new material to a pre­ existing informational framework.

Structural basis for unit enlargement

Size increase of an individual wall unit at a particular time must result from addition of new material

{« accretion) or enlargement of the unit without the addition of new material, or both processes simultaneously.

Visual evidence in support of surface accretion is obtainable, but demonstrating structural change at the macromolecular level without addition of new material - i.e., swelling of units which have already formed - is difficult. Change in staining intensity is one possible criterion for evaluating macromolecular changes, but observed differences in staining reactions could result 160 from structural changes, addition of new material, or both.

It should be noted, however, that the uniform size increase of the wall units, regardless of their position in the wall, intuitively is more easily explained by a structurally based expansion in which the process of unit expansion can take place in all units simultaneously throughout the wall, independent of the availability of new wall material. Therefore, both processes appear to be active until the wall is approximately 9 urn in thickness - i.e., until visual evidence of surface accretion ceases - at which point unit enlargement becomes the primary mode of size increase until maturity is reached.

Sj. aaleottii - wall development

Collection of the single megasporangium from each cone requires significantly more time and effort than the self contained system in sulcata. but results in a much more complete developmental analysis, including pre-meiotic stages.

Early in sporangial development (stage 1), tapetal and sporogenous tissues are differentiated based on cell shape and degree of vacuolation. Cells of the tapetum elongate parallel to the radius of the sporangium and contain dense cytoplasm. Isodiametric cells of the sporogenous tissue occupy the center of the sporangium. As development proceeds, the sporogenous cells show increased levels of 161 vacuolation and cytoplasm sequestered for digestion by autophagic vesicles. Additional evidence for autophagy

* includes vesicle in vesicle structures which are typical of lysosomal residual bodies (de Duve & Wattiaux, 1966;

Pettitt, 1971b).

Autodigesiton results in highly degraded cytoplasm within all of the megasporocytes except one. This cell, the megaspore mother cell (stage 2), becomes spherical and secretes a structureless granular wall inside of the remains of the primary cell wall. Pettitt (1971b) described and figured megasporocytes from £j. sulcata stating that "Viable and non-viable appear to be represented in equal numbers..." In the case of S. craleottiif the megaspore mother cell is the only cell which fits the description of a viable megasporocyte sensu

Pettitt. The remaining megasporocytes are irregularly shaped (apparently still reflecting their shape as influenced by contact with adjacent sporogenous cells prior to separation), and appear to be at an advanced stage of degeneration, unlike the smooth ovoid non-viable megasporytes described by Pettitt. The viable megasporocytes which do not undergo meiosis, Pettitt refers to as persistent megasporocytes. In £L*. aaleottii these cells resemble the degenerative non-viable cells, but they persist nonetheless, and supply a significant amount of wall material to the developing sporoderm even at seemingly 162 advanced stages of autophagy. These cytological differences suggest that developmental variability of megasporogenesis may exist within the genus. It will be interesting to see if future research reveals correlation of this pattern with the unusual wall structure type present in Sj. qaleottii. or if the differences are more widespread.

Abundant fibrillar material is evident in the sporangium locule at all stages prior to sporopollenin deposition. The primary cell wall breakdown suggests that the fibrillae result from cell wall disaggregation. This event takes place prior to the onset of secretory activity by the tapetum. Callose, or some similar sealing agent reaches its peak of production immediately prior to meiosis

(Pettitt, 1971a), when Bporangia range from 170-250 um in diameter (stage 3). At this size, coincidentally or not, penetration is a chronic problem.

After meiosis and separation of the individual megaspores, sporoderm deposition begins (stage 4). With the onset of secretory activity, tapetal cell walls undergo a chemical change, or are replaced altogether, and now appear electron-translucent. This phenomenon has been reported previously in angiosperms (Echlin, 1971). Similar changes take place in the persistent megasporocytes. The secreted material consists of wefts and globules which could represent different cross-sectional profiles of the 163 sane shape. Material of both shapes is present near the tapetum surface, but only globular material fuses with the forming megaspore walls. Secretion in persistent megasporocytes, which float freely in the sporangium locule, takes place on all sides, while embayments form between adjacent tapetal cells. This maximizes the area of interface along which the abundant wall material which is required can be secreted. The pattern in the ordered wall is detectable around the time the individual units reach 22-30 nm in size (stage

5). Tapetal cells and persistent megasporocytes still actively secrete large amounts of material, but of a different morphology than in previous stages. Beaded wefts occur near the secretion surfaces, but are smaller and tend to disaggregate as they move away from the tapetal cells.

Irregularly shaped fibrillar material is now deposited at the megaspore surface and gives rise to the units of the ordered region. Ornamental ridges expand, at least in part, at the expense of their interiors which become hollow.

At later stages (stage 6), the ordered portion of the exospore continues to thicken (to 4-8 um) due to an increase in constituent unit size (to 90-120).

Interestingly, an approximate doubling of the wall thickness (2.8-4.8 to 4.0-8.0) corresponds to an approximate quadrupling of the unit size (22-30 to 90-120 nm). An increase in volume of the entire megaspore is partly responsible for the discrepancy. The outer wall must thin to maintain continuity over the increased surface area resulting from increased spore volume, ornamental ridges become even more sparsely constructed. The nature of the secreted material changes significantly once again.

At the tapetal cell surface, an electron lucent bounding layer is still present in contact with the plasmalemma, but the material entering the sporangium locule is no longer beaded. Furthermore, in the accretion zone this material is thicker, highly interwoven, and appears to be tangled together around the free surfaces of many of the wall units, possibly giving rise to the outermost units.

Published descriptions of mature megaspores of gj. aaleottii

Mature megaspores of g^ aaleottii were first examined and described by Kempf (1970). Relative to the latest stage described here, mature megaspores increase in wall unit size in the ordered outer exospore (to 190-230 nm), and accumulate a siliceous surface coating. The "ribbon" morphology of the presumed siliceous deposits reported by

Kempf (1970) is unlike that seen by other investigators in gj. aaleottii (Tryon & Lugardon, 1978), or other species within the genus. Furthermore, the tubular ribbons in

Kempf's specimens occur throughout the sporoderm, but do not possess the same morphology as the siliceous surface 165 deposits. Demonstration of the chemical composition of these ribbons would be of interest, other changes which take place are the further expansion of the ornamental ridges at the megaspore surface, and the formation of a thick intine.

Sporoderm morphogenesis within Selaoinella

As was mentioned previously, some variation exists between species of Selaoinella with regard to cytological aspects of megasporogenesis. On the other hand, there are certain similarities which tend to unite seemingly divergent developmental types. Most notably, early stages of wall formation, when the units are small, are quite similar except for the rows and columns evident in s. oaleottii (stage 5, compare to Sa sulcata stage 1). A comparable organization consisting of small spherical units fusing at the surface of an accreting wall was figured by

Pettitt (1971a; fig. 8a) in broadwayii. This is especially significant considering the fact that s. broadwavii possesses the third major type of wall organization present within the genus, which consists of highly interconnected laminar units. One additional species of Selaginella. which has also been examined at the fine structural level (£*. caulescens. Buchen & Sievers,

1978b), also possesses a laminar construction type. In this species, wall units display similar ontogenetic 166 patterns of Initiation and enlargement, but their mature morphology seems to be established at an early stage.

Comparison with previous literature reports is difficult given the lack of control over the precise developmental stage examined. It does seem likely, however, that the initiation of unit morphology at an early stage, and the subsequent gradual enlargement throughout the course of development are probably consistent features of the genus as a whole.

In this context, it is worth noting the limited state of knowledge regarding not only variation of morphogenetic processes, but numerous aspects of Selaainella. including matters as fundamental as the type of tapetum present in the genus. A plasmodial tapetum was reported for s. broadwayii (Pettitt, 1971a), while the results of this investigation confirm a secretory tapetum in both S. sulcata and Qj. aaleottii. which Foster and Gifford (1974) hold to be true for the entire genus.

The role of the tapetum

The increase in wall thickness throughout development corresponds to an increase in wall unit size, which is accompanied by a change in the nature of the source material for the accreting units. There are two possible explanations for these changes: 1) the tapetal cells respond to some developmental signal and provide the 167 appropriately modified precursors or, 2) the tapetal cells merely produce small monomeric units, and the chemical nature of the sporangial locule determines the nature of the polymerized material. Since the physiochemical environment of the sporangium locule is under the influence of the tapetum, the question becomes, at what level does the tapetum exercise its control. No pre-formed sporopolleninous bodies were observed in tapetal cells.

This precludes only an exclusive role of the tapetum in direct modification within the limits of resolution of the transmission electron microscope. The tapetum could modify the sporopollenin precursors at the macromolecular level.

The second possibility, that monomeric units are provided by the tapetum and modified by the environment of the sporangium, is supported by evidence from a later developmental stage; the morphology of siliceous surface deposits in Selaainella. These deposits vary throughout the genus, from spines to small rounded structures, but undoubtedly are constructed from essentially identical precursors. If the tapetum is capable of modifying the sporangial environment enough to create a number of types of siliceous coatings, a similar situation could entail with the exospore. In the case of sporopollenin, the potential for chemical variation of the precursors is obviously much greater. 168

The mesospore

Ah least three types of bodies occur inside megaspores of Selaainella. all of which have been referred to as mesospores. Positionally, they are all the sane, occurring immediately outside of the protoplast, and connecting with the exospore only in the proximal area. Structurally, they are quite different; one is thick and densely constructed

(e.g., Qj. aaleottii). one is thick and loosely constructed of fine material (e.g., SL. sulcataf stage 1) and the third is thin and highly compressed (e.g., sulcata, stage 2).

The functional difference which separates these three types is believed to be their potential for expandability. The thick dense variety possesses no viable protoplast, and appears to be too densely organized to allow any appreciable degree of volume change of the inner body. The other two possess a construction type which is capable of increase in internal body volume in a relatively short period of time (Ruf, 197S). This flexibility was experimentally demonstrated and examined ultrastructurally by varying the concentration of a surrounding sugar solution and observing the resulting change in internal body diameter (Pieniazek, 1938; Sievers & Buchen, 1971).

The thicker body will never realize this potential for expansion since it lacks the viable membrane which is responsible for the osmotically induced volume changes. 169

However, these two expandable mesospores in gj. sulcata are developmentally related. The thicker variety expands to become the thinner provided protoplast abortion does not take place. Evidence in support of this conclusion is provided by the developmental analysis of wall morphogenesis in fLi. sulcata. Detailed examination of the thin internal layer present in megaspores which possess a viable protoplast reveals an identical, albeit highly compressed organization compared to that present in non­ protoplast bearing megaBpores. Even if such means were available, it would be physically impossible to compress the mesospore of a fully aborted megaspore to the thickness of that of a mature functional megaspore since the aborted protoplast undoubtedly adds new lipoidal material to the mesospore upon degeneration. Pettitt (1966a) also demonstrated the presence of carbohydrates and protein in thick mesospores of pulcherrima.

The unexpandable mesospore in £j. aaleottii probably also represents a late abortion product. It is interesting that the resultant form of the internal body is so different in the two taxa. This is no doubt reflective of the prevailing physiochemical milieu which is also directly responsible for the widely differing exospore organization between the two taxa. 170

Control of sporoderm patterning - sporophytic or gametophytic

The controlling agents for determining the ultrastructural organization of spore walls in Selaainella may be sporophytic, gametophytic, or a combination of both.

The accumulation of wall material after the complete degeneration of the spore protoplast suggests that strictly gametophytic control is unlikely, unless some undetected framework persists after abortion. Pettitt (1979) was able to visualize what he interpreted as such a framework by staining for acidic polysaccharides in the developing microspores of gj. plana. Sporophyte control seems more plausible possibly through the activity of the tapetum in creating a physiochemical environment in the sporangium conducive to polymerization of a wall with certain structural characteristics. An alternative possibility which involves sporophytic control is that the patterning of sporoderm form begins prior to meiosis of the mother cell. This situation entails in many bryophytic systems in which structural precursors to exospore ornamental elements are detectable prior to meiosis (Brown, et al., 1986).

Concordance with prevalent models of sporoderm morphogenesis

With regard to specific models of sporoderm patterning 171 which have been proposed, perhaps the most comprehensive is that of Rowley and his colleagues (Rowley, Dahl & Rowley,

1981, Rowley, Dahl, Sengupta & Rowley, 1981; Abadie, et al., 1986). It is difficult to envision the organization present in the exospore of g* aaleottii (or any other species of Selaainella) being constructed in the manner proposed by these workers based on an analysis of dissected exines. This is perhaps not surprising given the evolutionary distance which separates the lycopod and angiosperm lineages. However, the most fundamental and resistant framework sporoderm unit discovered to date was identified in degradation experiments using Lvcopodlum clavaturn spores (Sengupta & Rowley, 1974). These "tapes' of unknown chemical composition withstand hot 2- aminoethanol treatment and elevated pressures and temperatures which are destructive to sporopollenin, and are, therefore, assumed to be even more chemically and physically resistant. It is conceivable that the early stages of exospore development in Selaainella. which are initiated at the macromolecular level, could consist of structural units like those in Lvcopodlum clavatum. In this context, Buchen and sievers' (1978b) analysis of early sporoderm morphogenesis in Selaainella is worthy of discussion. They observed tripartite white lines (=tapes) being secreted by the megaspore protoplast early in wall formation and in and among the compressed chambers in the 172 mature mesospore of Selaainella caulescens. These white lines are felt by some to represent the fundamental structural unit of sporopollenin deposition (Rowley &

Southworth, 1967; Dickinson & Heslop-Harrison, 1968).

Rowley's modified interpretation is indicated above.

However, the developmental processes which act to produce the organization present in £j. aaleottii do not appear to conform well to Rowley's model in several respects. The information for sporoderm patterning in angiosperms is based on the glycocalyx model (Bennett,

1969), which involves the network of macromolecules which protrude above the plasma membrane and presumably confers spacial differentiation to the cell surface. Accretion of sporoderm subunits in Selaainella takes place remote from the plasma membrane relative to pollen grains, certainly beyond the range where a macromolecular ramentum could be expected to exert control. Furthermore, new wall units form after the protoplast aborts suggesting a lack of protoplasmic intervention. Little more definite can be said, based on the results of this investigation, about the earliest stages of sporoderm initiation, which are admittedly the most crucial in transferral of information for sporoderm patterning. It is not difficult to visualize the formation of the laminar sporoderm of caulescens

(Buchen and Sievers, 1978b) within the conceptual framework of the glycocalyx model. However, the subunit shape 173 encountered in Sj. aaleottii and sulcata are more problematical.

One specific area of conceptual difficulty involves the extreme degree of expansion encountered in all species of Selaainella. While proposed substructural sporoderm units (Rowley, Dahl, Sengupta & Rowley, 1981) do show some elasticity after removal of the sporopollenin matrix, there is no indication that a size modification takes place in the normal course of development. The model also does not account for the magnitude of size increase demonstrated here.

Correlations with the ordered wall construction type

The most prominent correlate with possession of the ordered wall organization like that present in aaleottii is with megaspore size. The average megaspore diameter within Selaainella exclusive of the taxa with the ordered organization (computed from all taxa reported in Minaki,

1984? and Tryon, 1949) is approximately 400 um. Average diameter of the ordered taxa is approximately 700 um, including the largest megaspores in the genus, in excess of

1.2 mm. There is approximately a three-fold difference in surface area between a 400 um megaspore and a 700 um megaspore. For a sporoderm to maintain continuity over the surface of an enlarging megaspore, it must employ one of the following strategies: l) enlarging the size of the 174

individual component units or 2) compression of one or more of the sporoderm strata. Both of the species examined for this investigation (and perhaps all species within this genus) employ the former strategy. Many of the unordered varieties also exhibit compression of inner sporoderm layers primarily late in development. It is hypothesized, however, that the unordered wall types are limited as to their maximum attainable megaspore diameter. Above a certain threshold, the ordered organization compensates more easily for extreme size increases, perhaps due to a capacity for shear to occur between the modular component subunits of the exospore.

Implications on the nature of sporopollenin

cells in a developing megasporangium are either the site of active secretion of wall material, the site of deposition, or both. The tapetum and persistent megasporocytes degenerate before the megaspores are mature, but continue to secrete after considerable structural disorganization. This fact, in combination with the observed thickening of the mesospore after protoplast abortion, suggests that sporopollenin may consist largely of waste products from degraded sporangial protoplasts.

Considering the resistance of sporopollenin, it seems logical that it would represent those products of cytoplasmic degeneration which the plant could not digest. 175

In the controlled environment of a sporangium locule, these waste products polymerize into a highly resistant substance well suited to providing protection for the propagules.

This hypothesis suggests a potential origin of sporopollenin which is attractive in its simplicity. CONCLUSIONS

Utilizing a newly developed set of terns and concepts for megaspore walls, ultrastructural analysis of both fossil and living lycopod negaspores was undertaken. In the case of Pennsylvanian megaspores, wall ultrastructure type can be correlated with dispersal strategy.

Megaspores of Mazocaroon possess a unique wall structure comprised of interconnected tubules, which correlates with a unique dispersal strategy. Valvislsporltes and

Laaenoisporites probably dispersed like modern Selaainella megaspores and, consequently, display a comparable wall organization. It is possible to elaborate a general sequence of developmental stages which occur within the walls of these Pennsylvanian megaspores in response to spore enlargement. These stages are: 1) changes in disposition of wall units, 2) loss of excess space between units, and 3) loss of all or nearly all space centrifugally.

Three basic construction types are present in megaspore walls of the genus Selaainella - those constructed of laminar units, those constructed of laterally fused spherules, and those constructed of

176 177 ordered units. All megaspores within the genus possess an inner separable layer which nay affect changes in the volune of the spore locule in response to water balance of the spore protoplast. Water balance nay also be regulated by sone type of protective layer at the negaspore surface, either a nodified set of outemost wall units, or a coating of silica.

The walls of two species of Selaainella arise as a result of sinilar developnental processes, and the pattern for the form of the wall is established at an early stage of developnent - probably at the nacronolecular level.

One unusual highly ordered type of construction is hypothesized to be adapted for acconodating large increases in negaspore volune during the course of developnent. Wall naterial for the developing spores is provided by cells of the tapetun - and by a number of cells which remain in the sporangium throughout the developnental process. The form of the naterial changes with advancing stages of developnent, but continues to be secreted by the source cells even when these cells are at advanced stages of degeneration. This is cited in support of an evolutionary hypothesis to explain the wide occurrence of sporopollenin, which makes up the spore wall. It is hypothesized that sporopollenin represents, in an evolutionary sense, the most resistant breakdown products which result from cell autolysis. 178

The salient points of this dissertation are summarized below:

1) The fine structural organization of the megaspore wall of Mazocarpon is unique, which reflects its unique dispersive strategy.

2) Pennsylvanian lycopods which produce

Valvisisporites aurltus and Lagenoisporites nudus. which are similar to modern Selaainella in their dispersive strategy, also share a similar sporoderm ultrastructure.

3) Morphogenetic changes in fully formed megaspore walls of Pennsylvanian lycopods, in response to spore enlargement, are probably dependant upon the construction type of the wall, but may display a consistent sequence of structural changes regardless of the wall type. This sequence is hypothesized to include bhe following steps:

1) changes in disposition of wall units (e.g.,

Lagenoisporites), followed by, 2) loss of excess space

(i.e., between unconnected units in loosely constructed regions; e.g., Lagenoisporites). and then, 3) loss of all or nearly all space centrifugally (e.g., Valvisisporites).

A) Use of a newly formed set of concepts which 179 includes unit shape, unit size, unit density, and degree of unit interconnection, has allowed a number of megaspore wall construction types to be recognized within the genus

Selaainella. The units which compose these walls fall into three basic categories; laminar units, laterally fused spherules, and ordered elements. With increased lateral fusion, laterally fused units approach laminar unit.

5) All species of Selaainella appear to possess an inner separable layer at the innermost extent of the exospore. This layer may affect changes in the volume of the spore locule, and thus help maintain a physiologically acceptable water balance between the spore protoplast and the ambient atmosphere.

6) All of the lycopod megaspores examined possess some form of modified surface which partially or completely seals the protoplast from the environment.

This surface takes a number of forms which include; 1) a siliceous surface coating which penetrates otherwise porous outer wall elements (several species of

Selaainella^, 2) a series of modified surface elements

(usually thinner than those in the underlying exospore) which run parallel to the surface, and are penetrated by a discontinuous siliceous coating (several species of 180

Selaainella^, 3) a denser network of exospore elements without an obvious siliceous coating fValvisisporites. and

Lagenoisporites. and 4) an additional wall layer of uncertain origin which is continuous with the subarchesporial pad and is, therefore, only present on the proximal surface (Mazocaroonl.

7) Megaspore walls of Selaainella appear to arise as a result of similar developmental processes regardless of which wall unit construction type they possess. The form of the units arises at the macromolecular level, and subsequently enlarges, first by both swelling and addition of new material, then later by swelling only.

8) Presence of the ordered wall construction type in the largest megaspores in the genus selaainella suggests a functional correlation. Megaspores which possess this organization may be better adapted than megaspores with other construction types to accomodate changes in surface coverage which result from volume increases. The adjustments result from shearing action which can take place between the elements of the modular wall.

9) Wall material for the developing exospore in

Selaainella aaleottii is provided by the tapetum and a number of megasporocytes which persist in the sporangium 181 locule. The structural nature of the secreted material changes with advances in stage of development. Secretion by these cells proceeds until the cells are at an advanced dergee of degeneration. This information is cited in support of a hypothesis which suggests that sporopollenin represents, in an evolutionary sense, waste products from a degenerating protoplast.

10) Three types of structures which could be referred to as mesospores occur in megaspores of

Selaainella. Two of these occur in spores which have aborted, and are incapable of volume increase. This is due to the fact that the abortion of the protoplast has caused the mesospore to thicken through the deposition of waste products which result from cytoplasmic degeneration.

In addition, there is no viable membrane to provide the force necessary for mesospore compression. The third structure appears in early developmental stages of viable megaspores, and will be compressed against the inside of the exospore at maturity. APPENDIX A - GLOSSARY OF TERMS acropetally proceeding in development from the base of the organ to the apex anastomosing intertwined or interwoven in an irregular fashion, as in a mat of hair baculae ornamental processes found on a spore surface which are taller than wide, and blunt tipped cingulum a usually broad flap of sporopollenin which encircles a spore at the equator contact area region on the proximal surface, made up of three contact faces in trilete spores, which corresponds to the position of contact with the three sister spores of the tetrad contact face each of three regions which make up a contact area whose boundaries are two arms of the trilete suture and a curvaturae if present curvaturae most distal boundary of a contact face; often a raised rim of sporopollenin distal away from the center of the tetrad element higher level sporopollenin unit of a sporoderm which is composed of units equatorial located at the imaginary plane which is equidistant between the proximal and distal poles exospore main sporopollenin layer in a pteridophyte sporoderm; forms the suture inner stratum which is consistently present in separable lycopod megaspores which is composed of a layer thin piece of the innermost portion of the main exospore, and all of a thin lamellated innermost exospore layer interradial triangular region whose boundaries consist region of two arms of the trilete suture and a curvaturae if present; three per trilete spore

182 183 inter- space not contained within any vesicle of a vesiculate vesiculate wall region. The boundaries of an inter-vesiculate space consist of the outside edges of a group of adjacent vesicles

intine cellulosic wall of a spore protoplast; situated immediately inside of the exospore

intramural on the surface of a spore in a region surrounded by the ridges (muri) of a reticulum

intra- within a sporangium sporangial

in £ltU refers to spores contained within their parent sporangium (especially to fossil spores recovered from their sporangium as opposed to dispersed spores recovered from sediments = sporae dispersae!

laesura (-ae, plural) the raised ridges which form the arms of a suture; three of these make up a trilete suture

mesospore structure, made of sporopollenin, which is contained within the exospore and is usually connected with it at the proximal pole

perispore wall layer which is situated outside the exospore and is deposited by a plasmodial tapetum; covers the suture

proximal directed away from the center of the tetrad

reticulum ornamentation consisting of a series of raised ridges which form a net-like appearance in surface view; in a perfect reticulum the ridges define only completely closed loops

rugulate ornamentation consisting of irregular elongate ridges

spina (-ae, plural) ornamental element which is tapered and pointed

soorae fossil spores (and pollen) which are dispersae recovered from sediments - i.e., not from within their parent sporangia 184 sporoderm general term encompassing all layers of a spore wall sporopollenin highly resistant oxidative polymer of carotenoid and carotenoid esters which forms the outer covering of spores and pollen grains suture the structurally modified weak point in a spore wall through which the gametophyte will emerge upon germination unit the basic structural sporopolleninous bodies which-make up a spore wall; several form a wall element verruca (-ae, plural) wart-like sculptural projections vesicles generally spherical completely enclosed structures which, in the case of spore walls, probably enclose space APPENDIX B - LIST OP SPECIMENS

Kew LC# = Living collection specimen number, Royal Botanic Gardens, Kew

Kew HB# « Collector and. herbarium sheet number, Herbarium, Royal Botanic Gardens, Kew

? «= Herbarium sheet number unrecorded

S.elfl.qingllfl flabellata Kew LC# 794-58-79403 Selaqinellausta Kew LC# 599-63-59908 SelaqinellaViridanqula Kew LC# 319-71-02760 SelaqinellaInaeaualifolia - Kew LC# 355-69-03071 Selaqinellaoaiiescens Kew HB# ? Selaqinellanulcherrima Kew LC# 000-73-13798 Selaainella arqentea Kew HB# Molesworth-Allen 4502 Selaqinella ornata Kew HB# Reid 16195 Selaqinella ervthroous Kew LC# 262-66-26201 Kew HB# ? Selaqinella brevipes Kew LC# 355-69-03070 Selaqinella elmerl Kew LC# 207-71-02023 Selaqinella plana Kew LC# 407-74-03162 Selaqinella fissidentoldes - Kew HB# ? Selaqinellavemensis Kew HB# ? Selaqinella Intermedia Kew HB# Mohd. Shah 1412 Selaqinella frondosa Kew HB# Molesworth-Allen 2657 Selaqinellalvaiiii Kew HB# ? Selaqinellawiiidenovii Kew HB# Turnau 897 Selaqinellasulcata Kew LC# 576-65-57609 Selaqinellaqaieottii Kew LC# 234-66-23403

185 18 6

LIST OF REFERENCES

Abadie, M., M. Hideux and J.R. Rowley. 1986. Ultrastructural cytology of the anther. II. Proposal for a model of exlne considering a dynamic connection between cytoskeleton, glycolemma and sporopollenin- synthesis. Ann. Sci. Nat., Bot., Paris, ser. 13, 8; 1- 16.

Afzelius, B.M., 6. Erdtman and F.S. Sjostrand. 1954. On the fine structure of the outer part of the spore wall of Lvcopodium clavatum as revealed by the electron microscope. Svensk bot. Tidskr. 48; 151-161.

Anonymous. 1976. Sporae Pteridophytum sinicorum. Acad. Sinica, Peking. Science Press, Peking. 451 pp.

Balbach, M.K. 1962. Observations on the ontogeny of Lepidocarpon. Amer. J. Bot. 49; 984-989.

Baldoni, A.M. and T.N. Taylor. 1985. Megasporas Cretacicas de la Formaclon Springhill en el subsuelo de Argentina y Chile austral. Ameghiniana 21; 151-167.

Batten, D.J. and D.J.P. Ferguson. 1987. Cabochonicus. a new genus for species of gemmate megaspores previously referred to Verrutrlletes. J. Micropalaeontol. 6; 65- 75.

Beer, R. 1909. The development of the spore of Eouisetum. New Phytol. 8; 261-266.

Beisvag, T. 1970. An electronmicroscopic investigation of the young gametophyte of the fern Blechnum spicant (L.) Roth. Grana 10; 121-135.

Bell, P.R. 1979. The contribution of the ferns to an understanding of the life cycles of vascular plants. In A,F. Dyer (ed.), The experimental biology of ferns. Academic Press, New York. pp. 58-85.

Bennett, H.S. 1969. Cell surface: components and configurations, and movements and recombinations. In A. Lima-de-Faria (ed.), Handbook of molecular 187

cytology. North Holland Publ. Co., Amsterdam, pp. 1261-1306.

Benson, H.J. 1918. Mazocaroon or the structural Slgillarlostrobus. Ann. Bot. 32; 569-589.

Bergad, R.D. 1978. Ultrastructural studies of selected North American Cretaceous megaspores of Minecippprltep, Erlansonlsporltes. Horstlsporltes. and Rlclnospora. n. gen. Palynology 2; 39-51.

Berthet, P. and H. Lecocq. 1977. Morphologic sporule des especes francalses du genre Isoetes L. Pollen et Spores 19; 329-349.

Bir, S.S. 1967. Correlation between spore size and polyploid level In the Himalayan asplenloid and athyrioid ferns. J. Palynology 2; 41-48.

Blr, S.S. 1977. Contributions of spore morphology In the taxonomy of some taxa of ferns. In P.K.K. Nalr (ed.), Advances In pollen-spore research. Volume II. To-day & To-morrow's Printers & Publishers, New Delhi, pp. 92- 119.

Bonnet, A. 1955. Contrubutions a 1'Etude des Hydropterldees I. Recherches sur Pllularla globulifera LetL. mlDMfcfl Dur. Cellule 57; 131-239.

Boterburg, A. 1956. Etude sur les Hydropterldales IV. Genese st dlfferenclatlon des porols sporales chez Marsllea diffusa Lepr. Cellule 58; 81-106.

Bower, F.O. 1935. Primitive land plants. MacMillan, London. 658 pp.

Brack, S.D. and T.N. Taylor. 1972. The ultrastructure and organization of Endosporites. Micropaleontology 18; 101-109.

Brack-Hanes, S.D. and B.H. Thomas. 1983. A re-examination of Lepldostrobus Brongniart. Bot. J. Linn. Soc. 86; 125-133.

Brack-Hanes, S.D. and J.c. Vaughn. 1978. Evidence of Paleozoic chromosomes from lycopod microgametophytes. Science 200; 1383-1385.

Breckon, G.J. and R.H. Falk. 1974. External spore morphology and taxonomic affinities of Phvlloglossum drummondii Kunze (Lycopodiaceae). Amer. J. Bot. 61; 481-485. 188

Britton, D.M. 1968. The spores of four species of spinulose wood ferns (Drvopterisl in eastern North America. Rhodora 70; 340-347.

Britton, D.M. and A.C. Jenny. 1974. The spores of Drvopterls felix-mas and related taxa in North America. Can. J. Bot. 52; 1923-1926.

Brooks, J. and G. Shaw. 1968a. Chemical structure of the exine of pollen walls and a new function for carotenoids in nature. Nature 219; 532-533.

Brooks, J. and G. Shaw. 1968b. The post-tetrad ontogeny of the pollen wall and the chemical structure of the sporopollenin of Lilium henrvi. Grana 8; 227-234.

Brooks, K.E. 1973. Reproductive biology of Selaqinella; I. Determination of megasporangia by 2- chloroethylphosphonic acid, an ethylene releasing compound. Plant Physiol. 51; 718-722.

Brown, R.C., B.E. Lemmon and K.S. Renzaglia. 1986. Sporocytic control of spore wall pattern in liverworts. Amer. J. Bot. 73; 593-596.

Buchen, B. and A. Sievers. 1978a. Megasporogenese von Selaqinella. I. Ultrastrukturelle und cytochemische Untersuchungen zur Sekretion von Polysacchariden. Protoplasma 96; 293-317.

Buchen, B. and A. Sievers. 1978b. Megasporogenese von Selaqinella. II. Ultrastrukturelle und cytochemische Untersuchunges zur Sekretion von Lipiden. Protoplasma 96; 319-328.

Buchen, B. and A. Sievers. 1984. Sporogenesis and pollen grain formation. In 0. Kiermayer (ed.) Cytomorphogenesis in plants. Springer-Verlag, New York. 349-376.

Chaloner, W.G. 1952. On Lepidocaroon waltoni sp. n., from the lower Carboniferous of Scotland. Ann. Mag. Nat. Hist. Ser. 5; 572-582

Chaloner, W.G, 1953. On the megaspores of Sigillaria. Ann. Mag. Nat. Hist. 12; 881-897.

Chen, S. and T. Huang. 1974. Spore morphology of Formosan Pteridaceae. Taiwania 19; 179-229

Chrysler, M.A. and D.S. Johnson. 1939. Spore production in Reqnellidium. Bull. Torrey Bot. Club 66; 263-279. 189

Cichan, M.A. 1985a. Vascular cambium and wood development in Carboniferous plants. I. Lepidodendrales. Amer. J. Bot. 72? 1163-1176.

Cichan, M.A. 1985b. Vascular cambium and wood development in Carboniferous plants II. sphenophvllum plurifoliatum Williamson and Scott (Sphenophyllales). Bot. Gaz. 146; 395-403.

Cichan, M.A., T.N. Taylor and D.F. Brauer. 1984. Ultrastructural studies of in situ Devonian spores: Pr.PtafrarlnQBtoton pennsvlvanicum Brauer. Rev. Palaeobot. Palynol. 41? 167-175.

Courviosier, J.M. and T.L. Phillips. 1975. Correlation of spores from Pennsylvanian coal-ball fructifications with dispersed spores. Micropaleontology 21? 45-59.

Crepet, W.L. 1979. Insect pollination: A paleontological perspective. Bioscience 29? 102-108.

Dahlen, M.A. 1982. A taxonomic reassessment of the Hong Kong fern allies with special reference to Selaqinella. Thesis, University of Hong Kong. 209 pp.

Darrah, E.L. 1968. The microstructure of some Pennsylvanian seeds and megaspores studied by maceration. Micropaleontology 14; 97-104. de Duve, C. and R. Wattiaux. 1966. Functions of lysosomes. A. Rev. Physiol. 28; 435-492.

DeMaggio A.E. and D.A. Stetler. 1980. Storage products of Onoclea sensibilis L. Amer. J. Bot. 67; 452-455.

Denke, P. 1902. Sporenentwicklung bei Selaqinella. Bot. Centralbl. 12? 182-199.

Devi, S. 1977. Scanning electron microscopy of some spores of Polvstichium Rev. Palaeobot. Palynol. 23; 139-144.

Dickinson, H.G. and J. Heslop-Harrison. 1968. Common mode of deposition for the sporopollenin of sexine and nexine. Nature 220; 926-927.

DiMichele, W.A. and T.L. Phillips. 1985. Arborescent lycopod reproduction and paleoecology in a coal-swamp environment of late Middle Pennsylvanian age (Herrin Coal, Illinois, U.S.A.). Rev. Palaeobot. Palynol. 44; 1-26. 190

Duckett, J.G. 1970. Spore size in the genus Equisetum. New Phytol. 69; 333-346.

Duerden, H. 1929. Variations in megaspore number in Selaqinella. Ann. Bot. 43; 451-457.

Duthie, A.V. 1923. Studies in the morphology of Selaqinella pumllla. Part 2. The cones, spores, and gametophytes. Trans Roy. Soc. South Africa 2; 131-144.

Echlin, P. 1971. Production of sporopollenin by the tapetum. In J. Brooks, et al. (eds.), Sporopollenin. Academic Press, New York. 220-247.

Erdtman, G. 1957. Pollen and spore morphology/plant taxonomy, Gymnospermae, Pteridophyta, Bryophyta (illustrations). Almqvist and Wiksell, Stockholm, 151pp.

Erdtman, G. and P. Sorsa. .1971. Pollen and spore morphology/Plant taxonomy: Pteridophyta (text and additional illustrations). Almqvist and Wiksell, Stockholm. 302pp.

Esteves, L.M., G.H. Felippe and T.S. Melhem. 1985. Germination and morphology of spores of Trichioteris corcovadensis. Amer. Fern J. 75; 92-102.

Feller, M.J. 1953. Etude sur les Hydropteridales IX. Sporocarpe et sporogenese chez Marsilea hirsuta R. Br. Cellule 55; 307-377.

Fitting, H. 1900. Bau und Entwicklungsgeschichte der Makrosporen von Isoetes und Selaqinella und ihre Bedeuting fur die Kenntniss des Wachstums pflanzlicher Zellmembranen. Bot. Zeit. 58; 107-156.

Foster, A.S. and E.M. Gifford. 1974. comparative morphology of vascular plants. W.H. Freeman and Co., San Francisco. 751 pp.

Fowler, K. 1975. Megaspores and massulae of Azolla prisca form the Oligocene of the Isle of Wight. Palaeontology 18; 483-507.

Fowler, K. and J. Stennett-Willson. 1978. Sporoderm architecture in modern Azolla. Fern Gaz. 11; 405-412.

Fraser, T.W. and D.L. Smith. 1974. Young gametophytes of the fern Polypodium vuloare L. An ultrastructural study. Protoplasma 82; 19-32. 191 Fuchs-Eckert, H.P. 1981. Isoetes Palmer! H.P. Fuchs, aine neue Isoetes-Art des Paramo. Proc. Kon. Ned. Akad. Wetensch. C84? 165-174.

Gantt, E. and H.J. Amott. 1965. Spore germination and development of the young gametophyte of the ostrich fern (Matteucccla struthiopterlB). Amer. J. Bot. 52; 82-94.

Gastaldo, R.A. 1981. An ultrastructural and taxonomic study of Valvisporites aurltus (Zerndt) Bhardwaj, a lycopsid megaspore from the Middle Pennsylvanian of Southern Illinois. Micropaleontology 27; 84-93.

Gastony, G.J. 1974. Spore morphology in the Cyatheaceae. I. The perine and sporangial capacity. General observations. Amer. J. Bot. 61; 672-680.

Gastony, G.J. 1979. Spore morphology in the Cyatheaceae. III. The genus Trichioteris. Amer. J. Bot. 66; 1238- 1260.

Gastony, G.J. 1981. Spore morphology in the Dicksoniaceae. I. The genera Cvstodium. Thvrsooteris. and Culcita. Amer. J. Bot. 68; 808-819.

Gastony, G.J. 1982. Spore morphology in the Dicksoniaceae. II. The genus Cibotium. Can. J. Bot. 60; 955-972.

Gastony, G.J. and R.M. Tryon. 1976. Spore morphology in the Cyatheaceae. II. The genera Loohosoria. Metaxvla. Alsophila. and Nephelea. Amer. J. Bot. 63; 738-758.

Gensel, P.G. and A.R. White. 1983. The morphology and ultrastructure of spores of the early Devonian trimerophyte Psilophvton (Dawson) Hueber fi Banks. Palynology 7; 221-233.

Good, C.W, 1975. Pennsylvanian-age calamitean cones, elater-bearing spores, and associated vegetative organs. Palaeontographica 153B; 28-99.

Good, C.W. 1977. Taxonomic and stratigraphic significance of the dispersed spore, in R.C. Romans (ed.) Geobotany. Plenum Publ. Corp., New York. 43-64.

Graustein, J.F. 1930. Evidences of hybridism in Selaqinella. Bot. Gaz, 90; 46-74.

Greguss, P. 1941. Die Sporen der mittereuropaischen Pteridophyten. Mat. Term. Kozl. Magyar Tud. Akad. 39; 1-36. 192

Guervin, C., J. Laroche, C. LeCoq and J.P. Bossey. 1972. La spore d'Eouisetum arvense L. stereostructure de la perispore et node d'insertion des elateres. C.R. Acad. Sc. Paris (Series D) 274; 3385-3386.

Gullvag, B.M. 1968a. On the fine structure of the spores of Egttlsatmn fluviatile var. verticillatum studied in the quiescent, geminated, and non-viable state. Grana 8; 23-69.

Gullvag, B.M. 1968b. Fine structure of the plastids and possible ways of distribution of the chloroplast products in some spores of Archigoniatae. Phytomorph. 18? 520-535.

Gullvag, B.M. 1969. Primary storage products of some pteridophyte spores, a fine structural study. Phytomorphology 19? 82-92.

Gullvag, B.M. 1971. Microbodies in the spore of Eouisetum: a fixation study. Grana 11; 36-40.

.Hagenau, D.J. 1961. Spore studies- in the genus Cvstopteris. I. The distribution of Cvstopteris with non-spiny spores in North America. Rhodora 63; 181-193.

Hall, J.W. 1974. Cretaceous Salviniaceae. Ann. Miss. Bot. Gard. 61? 354-367.

Hall, J.W. 1975. Ariadnaesporites and Glomerisporltes in the Late Cretaceous: ancestral Salviniaceae. Amer. J. Bot. 62? 359-369.

Hanning, E. 1911. Uber die Bedeutung der Periplasmodien: II. Die Bildung der Massulae von Azolla. Flora, N.F. 2; 243-278.

Harris, W.F. 1955. A manual of spores of New Zealand. Pteridophyta. Bull. N.Z. Dep. Sclent, ind. Res. 116; 1-186.

Haufler, C.H. and G.J. Gastony. 1978. Systematic implications of spore morphology in Bomroeria and related fern genera. Syst. Bot. 3; 241-256.

Hennipman, E. 1970. Electron and light microscopical onservations on the perine of the spores of some Bolbitis species (Filices). Acta. Bot. Neerl. 19; 671- 680. 193

Hennipman, E. 1977. A monograph of the fern genus Bolbltls (Lomarlopsldaceae). Leiden Bot. Ser. 2; 1-331.

Hickey, R.J. 1986. Isoetes megaspore surface morphology: nomenclature, variation, and systematic importance. Amer. Fern J. 76; 1-16.

Hill, S.R. 1977. Spore morphology of Anemia subgenus Coptophvllura. Amer. Fern J. 67; 11-17.

Hill, S.R. 1979. Spore morphology of Anemia subgenus Anemia. Amer. Fern J. 69; 71-80. Horner, H.T., Jr. 1966. Developmental aspects of heterospory in the genus Selaqinella. Amer. J. Bot. 53; 610.

Horner, H.T., Jr. and H.J. Arnott. 1963. Sporangial arrangement in North American species of Selaqinella. Bot. Gaz. 124; 371-383.

Horner, H.T., Jr. and C.K. Beltz. 1970. Cellular differentiation of heterospory in Selaqinella. Protoplasma 71; 335-341.

Huang, T. 1981. Spore flora of Taiwan (Pteridophyta). Tah- jihn Press Co., Taiwan. 111pp.

Huckaby, C.S., R. Nagmani and V. Raghavan. 1981. Spore germination patterns in Anoqraroma. Bommeria. SymnSE&sris, Hemionitls and Pitvogramma. Amer. Fern J. 71; 109-119.

Huckriede, R. 1982. Die unterkretazische Karsthohlen Fullung von Nehden im Sauerland. 1. Geologische, palaozoologische und palaobotanische Befunde und Datierung. Geologica et Palaeontologies 16; 183-242.

Hueber, F.M. 1982. Megaspores and a palynomorph from the Lower Potomac Group in Virginia. Smithsonian Contrib. to Paleobiology 49. Smithsonian Institution Press, Washington, D.c. 69pp.

Jux, U. and E.K. Kempf. 1971. Microstructure of the mesozoic megaspore Tasmanitriletes n.g. Grana 11; 95 -100.

Kanamori, K. 1969. Studies on the sterility and size variation of spores in some species of Japanese Drvopteris. J. Jap. Bot. 44; 207-217. 194

Kawasaki, T. 1970. Of the external figures of spores in the Polypodiaceae and their phylogenetic significance. Bull. Tokio Gakugei Univ. 22; 105-121.

Kempf, E.K. 1969a. Elektronenmikroskopie der Sporodermis von kanizoischen Hegasporen der Wasserfarn-Gattung Azolla. Palaont. Z. 43; 95-108.

Kempf, E.K. 1969b. Elektronenmikroskopie der Megasporen von Azolla teqellensis aus dam Altpleistozan der Niederlande. Palaeontographica 128B; 167-179.

Kempf, E.K. 1970. Elektronenmikroskopie der Sporodermis von Hegaspores der Gattung Selaqinella (Pteridophyta). Rev. Palaeobot. Palynol. 10; 99-116.

Kempf, E.K. 1971a. Electron microscopy of the megaspore Horstisporites semireticulatus from Liassic strata of Germany. Grana 11; 18-22.

Kempf, E.K. 1971b. Elektronenmikroskopie der Sporodermis von Mega- und Mikrosporen der Pteridophyten-Gattung aus dem Tertiar und Quatar Deutschlands. Palaeontographica 136B; 47-70.

Kempf, E.K. 1972. Electron microscopy of Mesozoic megaspores from Denmark. Grana 11; 151-163.

Kempf, E.K. 1973. Transmission electron microscopy of fossil spores. Palaeontology 16; 787-797.

Knox, E.M. 1938. The spores of the Pteridophyta, with observations on the microspores in coals of carboniferous age. Trans. Bot. Soc. Edin. 32; 438- 466.

Knox, E.M. 1950. The spores of Lvcopodium. Phvlloglossum. Selaqinella. and Isoetes and their value in the study of microfossils of Paleozoic age. Trans. Bot. Soc. Edin. 35; 211-357.

Knox, E.M. 1951. Spore morphology in British ferns. Trans. Bot. Soc. Edin. 35; 437-449.

Kott, L.S. and D.M. Britton. 1983. Spore morphology and taxonomy of Isoetes in northeastern North America. Can. J. Bot. 61; 3140-3163.

Kovach, W.L. and D.L. Dilcher. 1985. Morphology, ultrastructure, and paleoecology of Pax illitri11ites vittatus sp. nov. from the mid-Cretaceous (Cenomanian) of Kansas. Palynology 9; 85-94. 195

Kremp, G.O.W. and T. Kawasaki. 1972. The spores of the pteridophytes. Hirokawa Publ. Co., Tokyo, Japan. 398 pp.

Kurmann, M.H. and T.N. Taylor. 1984. Comparative ultrastructure of the sphenophyte spores Elaterites and Eaulsetura. Grana 23; 109-116.

Kurmann, H.H. and T.N. Taylor. 1987. Sporoderm ultrastructure of LophPSPCla and Cvatheacidites (Filicopsida): systematic and evolutionary implications. PI. Syst. Evol. 157; 85-94.

Lecoq, C.C. Guervin and J. LaRoche. 1973. Morphologie de quelches spores de Polypodiacees sensu lato en microscopie electronique a balayage. Pollen et Spores 15; 109-115.

Li Wen-ben and D.J. Batten. 1986. The Early Cretaceous megaspore Arcellites and closely associated Crvbelosporites microspores from northeast Inner Mongolia, P.R. China. Rev. Palaeobot. Palynol. 46; 189-208.

Li Wen-ben, D.J. Batten, Zhang Dahua and Zhang Liangde. 1987. Early Cretaceous megaspores from the Jalainar Group of northeastern Inner Mongolia, P.R. China. Palaeontographica 206B; 117-135.

Liew, F.s. 1977. Scanning electron microscopical studies on the spores of pteridophytes. XI. The family Oleandraceae tOleandra. Nephrolepis. and Arthropterlsl. Gard Bull. Singapore 30; 101-110.

Lloyd, R.M. 1976. Spore morphology of the Hawaiian genus Sadlerla (Blechnaceae). Amer. Fern J. 66; 1-7.

Long, A.G. 1968. Some specimens of Mazocarpon. Achlvmadocarpon. and Cvstosoorites from the Lower Carboniferous rocks of Berwickshire. Trans. Roy. Soc. Edinburgh 67; 359-372.

Lugardon, B. 1963. les spores des Pteridaees de France. Pollen et Spores 5; 325-336.

Lugardon, B. 1965. Structure des parois de la spore de Blechnum soicant (L) Roth. Pollen et Spores 7; 409- 428. 196

Lugardon, B. 1966. Formation de 1'exospore chez Blechnum solcant (L) Roth. C.R. Acad. Sci. Paris 262; 2029- 2031.

Lugardon, B. 1969a. Sur la formation de 1'exospore chez Osmunda regalis (L). C.R. Acad. Sci. Paris 268; 2879 -2882.

Lugardon, B. 1969b. Sur la structure fine des parois sporales d'Eauisetum maximum Lamk. Pollen et Spores 11; 449-474.

Lugardon, B. 1971a. Contribution a la connaissance de la morphogenese et de la structure des parois sporelas chez les Filicinees isosporees. These, Universite Paul Sabatier. Toulouse. 257 pp.

Lugardon, B. 1971b. L'endospore et la "pseudo-endospore" des spores des Filicinees isosporees. C.R. Acad. Sci. Paris 273; 675-678.

Lugardon, B. 1972a. Sur la structure fine st la nomenclature des parois microsporales chez Selaqinella denticulata (L) Link, et fi*. selaginoides (L) Link. C.R. Acad. Sci. Paris 274; 1256-1259.

Lugardon, B. 1972b. La structure fine de 1'exospore et la perispore des Filicinees isosporees. I, Generalites. Eusporangiees et Osmundales. Pollen et Spores 14; 227- 261.

Lugardon, B. 1973a. Sur les parois sporaleB de Psilotum triouetrum Sw. et leur structure fine. C.R. Acad. Sci. Paris 276; 1277-1280.

Lugardon, B. 1973b. Nomenclature et structure fine des parois acetoresistantes des microspores d'Isoetes. C.R. Acad. Sci. Paris 276; 3017-3020.

Lugardon, B. 1974. La Structure fine de 1 'exospore et de la perispore des Filicinees isosporees II. Filicales. Commentaries. Pollen et Spores 16; 161-226.

Lugardon, B. 1975. Sur le sporoderme des isospores st microspores des Pteridophytes et sur la terminologie appliquee a ses parois. Bull. Soc. bot. Fr. 122; 155- 167.

Lugardon, B. 1976. Sur la structure fine de 1'exospore dans les divers groupes de Pteridophytes actuelles (microspores et isospores), in I*K. Ferguson and J. Muller (eds.), The evolutionary significance of the 197

exine. Academic Press, New York. 231-250.

Lugardon, B. 1978. Comparison between pollen and pteridophyte spore walls. IV Int Palynol. Conf., Lucknow 1; 199-206.

Lugardon, B. 1978. Isospore and microspore walls of living pteridophytes; identification possibilities with different observation instruments. IV Int. Palynol. Conf., Lucknow 1; 152-163

Lugardon, B. 1979. Sur la formation du sporoderme chez Psliotum triouetrum Sw. Grana 18; 145-165.

Lugardon, B. 1986. Donnees ultrastructurales sur la fonction de 1'exospore chez les Pteridophytes. In S. Blackmore and I.K. Ferguson (eds.), Pollen and spores: form and function. Academic Press, New York. 251-264.

Lugardon, B. and P. Husson. 1982. Ultrastructure exosporales et caracteres generaux du sporoderme dans les microspores et les megaspore des Hydropteridees. C.R. Acad. Sci. Paris 294; 789-794.

Lyon, F. 1901. A study of the sporangia and gametophytes of Selaqinella aous and rupestris. Bot. Gaz. 32; 124 -141,170-194.

Lyon, F. 1905. The spore coats of Selaqinella. Bot. Gaz. 40; 285-295.

Martens, P. 1960a. Sur une structure microscopique orientee dans la paroi megasporale d'une Selaginelle. C.R. Acad. Sc. Paris 250; 1599-1602.

Martens, P. 1960b. Nouvelles observations sur la structure des parois megasporales de Selaqinella mvosurus (Sow) Alston. C.R. Acad. Sci. Paris 250; 1774-1775.

Mickel, J.T. and L. Atehortua G. 1980. Subdivision of the genus Elaphoglossum. Amer. Fern J. 70; 47-68.

Millay, M.A. and T.N. Taylor. 1982. The ultrastructure of Paleozoic fern spores: I. Botrvopteris. Amer. J. Bot. 69; 1148-1155.

Millay, M.A. and T.N. Taylor. 1984. The ultrastructure of Paleozoiic fern spores; II. Scolecopteris (Marattiales). Palaeontographica 194B; 1-13. 198

Miller, J.H., T.C. Vogelmann, T.C. and A.R. Bassel. 1983. Promotion of rhizoid elongation by metal ions and the function of the spore coat as an ion reservoir. PI. Physiol, 71; 828-834.

Minaki, M. 1984. Macrospore morphology and taxonomy of Selaqinella (Selaginellaceae). Pollen et Spores 26; 421-480.

Mitui, K. 1977. Spore wall structure of some Japanese species in Polypodiaceae. s. st. Bull. Nippon Dental Univ. 6; 119-129.

Mutui, K. 1982. Spore morphology of Japanese ferns. 206 pp.

Morbelli, M.A. 1974. Analisis palinologico en hibridos interspecificos del genero Blechnum. L. Bol. Soc. Argent. Bot. 15; 446-466.

Morbelli, M.A. 1977. Esporasde las especies argentinas de Selaqinella (Selaginellaceae - Pteridophyta). Obra del Centenario del Museo de la Plata, Secc. Botanica, 36 PP. Moy, C.J. 1988. Variations of fern spore ultrastructure as reflections of their evolution. Grana 27; 39-51.

Murillo, M.T. and M.J. M. Bless. 1974. Spores of recent Colombian Pteridophyta. I. Trilete spores. Rev. Palaeobot. Palynol. 18; 223-269.

Murillo, M.T. and M.J.M. Bless. 1978. Spores of recent Colombian Pteridophyta.il. Monolete spores. Rev. Palaeobot. Palynol. 25; 319-365.

Nayar, B.K. 1964. Palynology of modern pteridophytes. in Advances in Palynology, P.K.K. Nair (ed.), Natl. Botanic Garden, Lucknow. 101-141.

Nayar, B.K. 1970. A phylogenetic classification of the homosporous ferns. Taxon 19; 229-236.

Nayar, B.K. and S. Devi. 1963. Spore morphology of some Japanese Aspidiaceae. Pollen et Spores 5; 355-372.

Nayar, B.K. and S. Devi. 1964a. Spore morphology of Indian ferns I. Aspidiaceae. Grana 5; 80-120

Nayar, B.K. and S. Devi. 1964b. Spore morphology of Indian ferns II. Aspleniaceae and Blechnaceae. Grana 5; 222- 246. 199

Nayar, B.K. and S. Devi. 1964c. Spore morphology of Indian ferns III. Polypodiaceae. Grana 5; 342-396.

Nayar, B.K. and S. Devi. 1965. Spore morphology of Indian ferns IV. Grammitidaceae. Grana 6; 121-127.

Nayar, B.K. and S. Devi. 1966. Spore morphology of the Pteridaceae I. The Pteridoid ferns. Grana 6; 476-502.

Nayar, B.K. and S. Devi. 1967. Spore morphology of the Pteridaceae II. The gymnogrammoid ferns, Grana 7; 568- 600.

Nayar, B.K. and P. Lata. 1965. Spore morphology or some african fern-allies. Pollen et Spores 7; 313-318.

Nayar, B.K., P. Lata, and L.P. Tiwari. 1964. Spore morphology of the ferns of west tropical Africa. Pollen et Spores 6; 545-582.

Oliver, F.W. and D.H. Scott. 1904. On the structure of the Palaeozoic seed Laqenostoma lomaxi. with a statement of the evidence upon which it is referred to Lvginodendron. Phil. Trans. Roy. Soc. Lond. B 197; 193-247.

Peabody, D.M. 1964. Morphologic variation of osmundaceae spores. Thesis, Univ. Arizona, Tucson, Ariz. (unpubl.).

Pearman, R.W. 1976. A scanning electron microscope investigation of the spores of the fern genus Cvstopteris. Brit Fern Gaz. 11; 221-230.

Pettitt, J.M. 1966a. Exine structure in some fossil and recent spores and pollen as revealed by light and electron microscopy. Bull. Br. Mus. (Nat. Hist.) geol. 13; 223-257.

Pettitt, J.M. 1966b. A new interpretation of the structure of the megaspore membrane in some gymnospermous ovules. J. Linn. Soc. (Bot.) 59; 253-263.

Pettitt, J.M, 1971a. Some ultrastructural aspects of sporoderm formation in pteridophytes. in G. Erdtman, and P. Sorsa (eds.), Pollen and spore morphology/Plant taxonomy. Pteridophyta. Almqvist & Wiksells, Stockholm. 227-251.

Pettitt, J.M. 1971b. Developmental mechanisms in heterospory I. Megasporocyte degeneration in 200

Selaqinella. Bot. J. Linn. Soc. 64; 237-246.

Pettitt, J.M. 1974. Developmental mechanisms in heterospory. II. Evidence for pinocytosis in the microspores of Selaqinella. Bot. J. Linn. Soc. 69; 79- 87.

Pettitt, J.M. 1976a. A route for the passage of substances through the developing pteridophyte exine. Protoplasma 88; 117-131.

Pettitt, J.M. 1976b. Developmental mechanisms in heterospory. III.. The plastid cucle during megasporogenesis in Isoetes. J. Cell Sci. 20; 671- 685.

Pettitt, J.M. 1977. Developmental mechanisms in heterospory: Features of post meiotic regression in Selaqinella. Ann. Bot. 41; 117-125.

Pettitt, J.M. 1979. Ultrastructure and cytochemistry of spore wall morphogenesis. In A.F. Dyer (ed.), The experimental biology of ferns. Academic Press, New York. 213-252.

Pettitt, J.M. and A.C. Jermy. 1974. The surface coats on spores. Biol. J. Linn. Soc. 6; 245-257.

Pfeiffer, N.E. 1922. Monograph of the Isoetaceae. Ann. Miss. Bot. Gard. 9; 79-232.

Phillips, T.L. 1979. Reproduction of heterosporous arborescent lycopods in the Mississippian- Pennsylvanian of Euramerica. Rev. Palaeobot. Palynol. 27; 239-289.

Pieniazek, S.A1. 1938. Uber die Entwicklung und das Wachstums der Macrosporenmembranen bei Selaqinella. Sprowozdania Towazystwa Nauk. Warsawskiego Wyzdiat 4 31; 211-230

Pigg, K.B. The morphology and reproductive biology of the Sigillarian cone Mazocarpon. Bot. Gaz. 144; 600-613.

Pigg, K.B. and G.W. Rothwell, 1983. Megagametophyte development in the Chaloneriaceae fam. nov., permineralized Isoetales (Lycopsida). Bot. Gaz. 144; 295-302.

Potonie, R. 1962* Synopsis der Sporae in situ. Beihefte zum Geologischen Jahrbuch 52; 1-204. 201

Potonie, R. and 6. Kremp. 1955. Die sporae dlscersae des Ruhrkarbons ihre Morphographie und Stratigraphie ait Ausblicken auf Arten andere Gebiete und Zeltabschnitte. Palaeontographica 98B; 1-136.

Pringsheim, N. 1853. Notiz uber die Schleudered von Eauisetum. Bot. Zeitung 11; 241-244.

Raghavan, V. 1980. Cytology, physiology and biochemistry of germination of fern spores. Int. Rev. Cytol. 62; 69 -118.

Reeve, R.M. 1934. The spores of the genus Selaqinella in north central and northeastern United States. Rhodora 37; 342-345.

Robert, D. 1970. Participation des cellules del'assise tapetale du sporange a 1 'edification des parois microsporales chez une selaginelle, Selaqinella kraussiana A. Br. C.R. Acad. Sci. Paris 271; 643-651.

Robert, D. 1971a. Nouvelle contribution a 1'etude des parois microsporales chez le Selaqinella kraussiana A.Br. C.R. Acad. Sci. Paris 272; 385-388.

Robert, D. 1971b. Etude en microscopie electronique, des modalites d #edification des parois microsporales chez le Selaqinella selaqinoides (L.), mise en place du feullet interne. C.R. Acad. Sci. Paris 273; 1933- 1936.

Robert, D., F. Roland-Heydacker, J. Denizot, J. Laroche, P. Fougerouz, and L. Davignon. 1973. La parois megasporale de 1'Isoetes setacea Bocs ex Delile. Adansonia, ser. 2, 13; 313-332.

Rowley, J.R., A.o. Dahl and J.S. Rowley. 1981. Substructure in exines of Artemisia vulgaris (Asteraceae). Rev. Palaeobot. Palynol. 35; 1-38.

Rowley, J.R., A.O. Dahl, S. Sengupta and J.S. Rowley. 1981. A model of exine substructure based on dissection of pollen and spore exines. Palynology 5; 107-152.

Rowley, J.R. and D. Southworth. 1967. Deposition of sporopollenin on lamellae of unit membrane dimensions. Nature 213; 703-704.

Ruf, E.W., Jr. 1975. The morphology and reproduction of certain Selaginellas. Thesis, The University of Texas at Austin. 202

Sanders, J.V. and P.J. Darragh. 1971. The microstructure of precious opal. Mineralogical Record 2; 261-268.

Sanlo, C. 1856. Beitrag zur Kenntnlss der Entwicklung von Eoulsetum arvense. Bot. Zeltung 14; 177-185, 193-196.

Sanlo, c. 1857. Elnlge weltere Bemerkungen uber die Sporenentwlcklung bel den Egulseten. Bot. Zeltung 15; 657-664.

Saxena, D.K. 1980. Ultrastructure of spores of Ecruisetuin ramoslsslmum and Rx. diffusum. 5 Int. Palynol. conf, Absts. 355.

Schopf, J.M. 1938. Spores from the Herrin (No. 6) coal bed In Illinois. Illinois Geol. Surv. Rept. Inv. No. 50; 73pp.

Schopf, J.M. 1941a. Contribution to Pennsylvanian paleobotany. Notes on the Lepldocarpaceae. Am. Midi. Natl. 25; 548-563.

Schopf, J.M. 1941b. Contribution to Pennsylvanian paleobotany; Mazocaroon oedinternum sp. nov., and sigillarian relationships. Illinois Geol. Surv. Rept. Inv. No. 75; 53pp.

Scott, A.C. and T.N. Taylor. 1983. Plant/animal interactions during the Upper Carboniferous. Bot. Rev. 49; 259-307.

Scott, D.H. 1900. Note on the occurrence of a seed-like fructification in certain Paleozoic lycopods. Proc. Roy Soc. London 67; 306-309.

Sengupta, S. and J. Rowley. 1974. Re-exposure of tapes at high temperature and pressure in the Lvcopodium clavatum spore exine. Grana 14; 143-151.

Shattuck, C.H. 1910. The origin of heterospory in Marsilea. Bot. Gaz. 49; 10-40.

Sheffield, E. and P.R. Bell. 1979. Ultrastructural aspects of sporogenesis in a fern Pteridium aquilinum (L.) Kuhn. Ann. Bot. 44; 393-405.

Sheffield, E., S. Laird and P.R. Bell. 1983. Ultrastructural aspects of sporogenesis in the apogaraous fern Drvopteris borrerl. J.Cell Sci. 63; 125-134. 203

Sievers, A. and B. Buchen. 1970. Uber den Feinbau der wachsenden Megaspore von Selaqinella. Protoplasma 71; 267-279.

Sievers, A. and B. Buchen. 1971. Contact between the spore cytoplasm and the growing sporoderm of the Selaqinella megaspore. In J. Brooks, et al., (eds.) Sporopollenin. Academic Press, New York. 654-658.

Sitte, P. 1963. Bau und Bewegung der Sporen-haupteren bei Equlsetum arvense L. Ber. Nat. Med. Vereins Xnnsbr. 53; 193-207.

Slagg, R.A. 1932. The gametophytes of Selaqinella kraussiana I. The microgametophyte. Amer. J. Bot. 19; 106-127.

Smith, A.R. 1986. Revision of the neotropical fern genus Cvclodlum. Amer. J. Bot. 76; 56-98.

Smoot, E.L. 1984a. Secondary phloem anatomy in Callistophvton bovssetii (Renault) Rothwell and histological changes in the outer phloem. Bot. Gaz. 145; 395-406.

Smoot, E.L. 1984b. Phloem structure in the Carboniferous fern Psaronius (Marattiales). Amer. J. Bot. 71; 1104- 1113.

Somers, p. and W.R. Buck. 1975. Selaqinella ludoviciana. apoda and their hybrids in the southeastern United States. Amer. Fern J. 65; 76-82.

Sorsa, P. 1964. Studies on the spore morphology of fennoscandian fern species. Ann. Bot. Fennici 1; 179- 201.

Sorsa, P. 1980. Spore morphology of the fern genus Gvronocarpium and its relations to the taxonomy. Ann. Bot. Fennici 17; 86-90.

Sosa, A. and C. Sosa-Bourdouil. 1940. Sur la composition des macrospores et des microspores de Selaqinella. C.R. Acad. Sc. Paris 210; 59-61.

Stainer, F. 1966. Morphological study of the walls of the mega-and microspores of Selaqinella mvosurus (Sw.) Alston. Rev. Palaeobot. Palynol. 3; 47-50. 204

Sussex, I.M. 1966. The origin and development of heterospory in vascular plants, in E.G. Cutter (ed.), Trnds in plant morphogenesis. Longmans, Green and Co., London. 140-152.

Tardieu-Blot, H.L. 1963a. Sur les spores de Lindsceaceae et de Dennstaedtiaceae de Madagascar et des Mascareignes. Etude de palynologie appliquee a la Systematique. Pollen et Spores 5; 69-86.

Tardieu-Blot, M.L. 1963b. Sur les spores de Pterideae Malagaches. Pollen et Spores 5; 337-353.

Tardieu-Blot, M.L. 1964. Sur les spores de Davalliaceae et Vittariaceae Malagaches. Pollen et Spores 5; 537-544.

Tardieu-Blot, M.L. 1965. Sur les spores d'Adiantaceae, Aspleniaceae, Thelypteridaceae, et Athyriaceae de Madagascar. Pollen et Spores 7; 329-338.

Tardieu-Blot, M.L. 1966. Sur les spores de Fougeres Malagaches: Filicales (fin), Marattiales, Ophioglossales. Pollen et Spores 8; 75-122.

Taylor, T.N. 1973. Combined transmission and scanning electron microscopy of in situ Palaeozoic spores. Palaeontology 16: 765-776.

Taylor, T.N. 1974. Scanning electron microscopy of fossil megaspores: wall development. Scanning Electron Microscopy/1974, Part II; 359-366.

Taylor, T.N. and S.D. Brack-Hanes. 1976. The structure and reproductive significance of the exine in fossil lycopod megaspores. Scanning Electron Microscopy/1976, Part VII; 513-518.

Taylor, T.N. and D.F. Brauer. 1983. Ultrastructural studies of in situ Devonian spores: Barinonhvton citrulliforme. Amer. J. Bot. 70; 106-112.

Taylor, T.N., N.J. Maihle, and L.V. Hills. 1980. morphological and ultrastructural features of Nikltlnsporites canadensis Chaloner, a Devonian megaspore from the Frasnian of Canada. Rev. Palaeobot. Palynol. 30; 89-99.

Taylor, W.A. 1986. Ultrastructure of sphenophyllalean spores. Rev. Palaeobot. Palynol. 47; 105-128.

Taylor, W.A. and T.N. Taylor. 1987. Subunit Construction of the spore wall in fossil and living lycopods. Pollen 205

et Spores 29; 241-248.

Taylor, W.A. and T.N. Taylor. 1988. Ultrastructural analysis of selected megaspores from Argentina. J. Micropalaeont. 7; 73-87.

Taylor, W.C., R.H. Mohlenbrock, and J.A. Murphy. 1975. The spores and taxonomy of Isoetes butlerl and I. melanopoda. Amer. Fern J. 65; 33-38.

Tralau, H. 1969. The morphology of microspores of Pllularla Grana 9; 118-123.

Tryon, A.F. 1949. Spores of the genus Selaqinella in North America north of Mexico. Ann. Miss. Bot. Gard. 36; 413-431.

Tryon, A.F. 1971. Structure and variation in spores of Thelvpteris palustris. Rhodora 73; 444-460.

Tryon, A.F. 1972. spores, chromosomes, and relations of the fern Pellaea atroouipurea. Rhodora 74; 220-241.

Tryon, A.F. 1985. Spores of myrmecophytic ferns. Proc. Roy. Soc. Edinb. 86B; 105-110.

Tryon, A.F. 1986. Stasis, diversity, and function in spores based on an electron microscope survey of the Pteridophyta. In S. Blackmore, and I.K. Ferguson (eds.), Pollen and spores: form and function. Academic Press, New York. 233-249.

Tryon, A.F. and B. Lugardon. 1978. Wall structure and mineral content in Selaqinella spores. Pollen et Spores 20; 315-340.

Tryon, A.F., R. Tryon, and F. Badre. 1980. Classification, spores, and nomenclature of the marsh fern. Rhodora 82; 461-474.

Tryon, R.M. and A.F. Tryon. 1973. Geography, spores, and evolutionary relationships in the cheilanthoid ferns, in A.C. Jermy, J.A. crabbe, and B.A. Thomas (eds.), The phylogeny and classification of the ferns. Academic Press, New York. 145-153.

Tryon, R.M. and A.F. Tryon. 1982. Ferns and allied plants. Springer-Verlag, New York. 857 pp.

Uffelen, G.A. van. 1985. Synaptospory in the fern genus Pvrrosia (Polypodiaceae). Blumea 31; 57-64. 206

Uffelen, 6.A. van and E. Hennipman. 1985. The spores in pyrrosia Hirbel (Polypodiaceae), a scanning electron microscope study. Pollen et Spores 27; 155-19.8.

Vide, G. 1974. Genome analysis of the European Cvstopteris fraollls complex. I. Tetraploid taxa. Acta. Bot. Hung. 20; 181-192.

Wagner, W.H., Jr. 1952. The fern genus Dlellla. University of California Publicaitons in Botany 26. 1-212.

Wagner, W.H. 1974. Structure of spores in relation to fern phylogeny. Ann. Mo. Bot. Gard. 61; 332-353.

Wagner, W.H. 1985. Bilateral spores in New World grammitid ferns. Amer. Fern J. 75; 6-11.

Wayne, R. and P.K. Hepler. 1984. The role of calcium ions in phytochrome mediated germination of spores of Onoclea sensibilis L. Planta 160; 12-20.

Wayne, R. and P.K. Hepler. 1985. The atomic composition of Onoclea sensibilis spores. Amer. Fern J. 75; 12-18.

Wilce, J.H. 1972. Lycopod spores X. General spore patterns and the generic segregates of Lvcopodium. Amer. Fern J. 62; 65-79.

Wilson, L.R. 1936. The spores of the genus Lvcopodium in the U.S. and Canada. Rhodora 36; 13-19.

Wollenweber, E., C. Scheele and A.F. Tryon. 1987. Flavenoids and spores of Platvzoroa microphvllum an endemic fern of Australia. Amer. Fern J. 77; 28-32.

Wood, C.C. 1973. Spore variation in the Thelypteridaceae. Ill A.c. Jermy, J.A. Crabbe and B.A. Thomas (eds.). The phylogeny and classification of the ferns. Academic Press, New York. 191-202.

Zerndt, J. 1930. Triletes qjqanteus n. sp., eine riesige Megaspore aus dem Karbon. Bull. Internat. Acad. Polonaise Sci. et Lett., Ser. B, Annee 1930; 71-79.

Zimmerman, R.P. and T.N. Taylor. 1970. The ultrastructure of Paleozoic megaspore membranes. Pollen et Spores 12; 451-468.