Morphological and ultrastructural variations in Scliizjaea pectinata (Schizaeaceae: Pteridophyta)

Schizaea pectinata (L.) Sw. was collected from the extreme ends o f its geographical range in South Africa for a study of sporangial development, sporogenesis and tapetal organisation. Differences were noted in the gross morphology , in sporangium size, spore size and in the patterning of the outer exospore from the two sites. Coiled structures were associated with the development of the inner perispore in spores collected from the Transvaal, whereas dense, heterogeneous bodies were associated with the formation of this layer in spores from the Cape. Differences were also noted in the organisation of the tapetum. A cellular, parietal tapetum and a plasmodial tapetum were present in the Cape material when the spores had developed the sculptured outer exospore. In sporangia from the Transvaal, however, only a plasmodial tapetum was present at the same stage of sporoderm development. A detailed study of S. pectinata throughout its distribution is required to determine the taxonomic importance of these findings.


INTRODUCTION
An investigation of sporangial development, sporo genesis and tapetal organisation in Schizaea pectinata (L.) Sw. was carried out on material collected from Ysterkroon in the Northern Transvaal (2429BB Zebediela) and from the Cape of Good Hope Nature Reserve (341 SAD Simonstown).The Transvaal material was collected at an altitude of 2 046 m in an area of vegetation described as Northeastern Mountain Sourveld (Acocks 1988).The plants were small, approximately 80 mm high, and were found in well-protected sites at the bases of outcrops of Black Reef quartzite.The plants were obscured by large tussocks of Themeda triandra Forssk.and other coarse grasses.Plants from the Western Cape were collected from several sites, at altitudes below 150 m. all of which had well-drained, highly leached, shallow sandy lithosols in areas of Mesic Mountain Fynbos (Macdonald. Clark & Taylor 1989).These plants regenerated quickly after fire, becoming fertile after about four months of growth in open situations.They gradually became overgrown and shaded as other fynbos plants regenerated.In the second and subsequent years following fire they produced few new fronds.Voucher specimens (Cape J067273, Trans vaal J067274) are housed in the C.F. Moss Herbarium (J).In studying the ontogeny of the sporoderm.several features were noted w hich differed in the specimens from the two sites.These included variations in the sculpturing of the outer exospore and in structures involved with the development of the inner perispore.This led to a closer study of the material to ascertain whether the differences were constant features.

M ATERI ALS AND METHODS
Material was fixed in the field in 2% paraformaldehyde and 3% glutaraldehvde in 0.08 M Pipes buffer at pH 8.0 (Colhoun & Steer 1981).washed in ice-cold buffer and post-fixed on ice using 2# CKO4 followed by washing and dehydration through a graded acetone series.Specimens were transferred into propylene oxide, in filtrated w ith resin using mixtures of propylene oxide and resin and finally embedded in Fpon (Luft 1961).
Sections 1 |im thick were stained in toluidine blue for light microscopy and 60 nm ultra-thin sections were stained in uranyl acetate and lead citrate for viewing in a Jeol 100S transmission electron microscope at 80 Kv.Specim ens for scanning electron microscopy were selected from material in primary fixative but were not post-fixed.They were washed in buffer, passed through a graded ethanol series, transferred to propylene oxide and dned by the critical point method in a Bal/ers Union CPD model 020.The specimens were mounted onto aluminium stubs coated with pressure sensitive adhesive and coated with gold-palladium prior to viewing in a Jeol JSM-840 microscope at 15 Kv.Sporal features were observed on fixed and processed material from the two sites and a quantitative survey was also carried out on fixed material.
Measurements of the sporangia and the spores from each of the collection sites was based on a sample size of 50 specimens.Sporangium length was measured from the base of the capsule to the top of the annulus and the width was measured at the widest point.The length and width of spores complete with perispore were recorded.

Gross morphology
Plants from the Transvaal were small, approximately 100 mm high with only a few fronds which had regenerated after fire (Figure 1A).Between seven and nine pairs of modified pinnae were supported on a straight rachis held at an angle of about 100° from the photosyn thetic stipe.
Plants collected from the Cape (Figure IB) were robust, about 200 mm high and formed extensive clumps with large num bers of fronds (± 50) w hich had regenerated after fire.The pinnae pairs numbered between 14 and 18.The rachis passed through a straight phase during the unfolding of the crozier but was distinctly curved at maturity.Variations in the gross morphology of the plants from the two sites which are approximately 2 000 km apart, were initially thought to indicate ecotypes of S. pectinata.Variation in size of specimens collected at different al titudes had been commented on by Roux (1979).There are known cases of exceptional morphological variability within populations in other ferns from disjunct parts of their range, as in Thelypteris palustris Schott (Tryon & Lugardon 1991).Tryon & Tryon (1982) stated that the systematics of the tropical American Schizaea Sm. are not adequately known, largely due to a lack of field studies necessary to establish the extent of variation within the species and that there may be greater morphological diver sity than is currently recognised.The latter comment probably also applies to the southern African members of the genus, as confirmed by the anomalous results obtained in this material.

Spore and sporangium size
Minimum and maximum dimensions of spores and sporangia are presented in Table 1 and sample mean values and standard deviations are presented in Table 2.All measurements were statistically analysed by the Student's two sample t test (Parker 1979) and the differen ces were found to be statistically significant.while 80% of the Transvaal spores lay in a range between 105.5 pm and 115 pm.
A narrow range of variation in width was also found in the Cape spores where 72% measured between 55.2 pm and 60 pm and 66% of the spores from the Transvaal population measured between 60 pm and 67.2 pm.When individual measurements of spore length and width were plotted, the scattergram (Figure 2A) showed that there was no overlap between the two populations with respect to spore size.
A similar trend was seen in dimensions of sporangia from the Cape sample where 84% of the sporangia lay within a relatively narrow range between 580 pm and 650 pm in length, and in the Transvaal sample 76% of the sporangia were found in the range between 698 pm and 744 pm.
Sporangial width showed a greater degree of variation than sporangial length with 60% of the sporangia from the Cape measuring between 372 pm and 418 pm and 78% of the sporangia from the Transvaal sample ranged from 442 pm to 512 pm wide.
When length and width of individual sporangia were plotted (Figure 2B) the scattergram revealed only a min imal amount of overlap between the two populations.
The systematic potential of spores with respect to size and spore morphology has long been recognised (Wood 1973).Differences in spore size in the two Schizaea samples could indicate a difference in ploidy between the two populations.Tryon & Lugardon (1991) found spores of diploid Polystichum Roth species to have smaller spores than tetraploids but they also found that spores from disjunct populations of Thelypteris palustris showed significant variation in size.It will be necessary to inves tigate the ploidy of the two populations and also to sample populations from the entire distribution range before the full significance of the present findings may be assessed.

Outer exospore sculpturing
The exospore in S. pectinata consists of two layers (Parkinson 1991).The inner exospore is a smooth layer approximately 85 nm wide.The outer exospore which forms the bulk of the exospore is sculptured and reaches a maximum thickness of between 4 pm and 5 pm.The sculpturing differs in the Transvaal and Cape material (Figures 3 & 4).Spores from the Transvaal site (Figures 3A, C; 4A) have indentations (punctae) which are ap proximately 1 pm in diameter with ridges of the same dimension separating them.The ridges have an uneven but continuous surface and the inner surface of the punctae is smooth.Spores from the Cape site (Figures 3B.D; 4 B) are also punctate.The punctae however, are 1.5 pm to 2.0 pm in diameter with uneven, 1 pm to 2 pm wide ridges separating them.The whole surface is uneven, con sisting of irregular, juxtaposed granules.
The differences in exospore structure could be the result of the differences in the habitats of the two popula tions sampled.It has been suggested that the surface com plexity of spores of Pyrmsia Mirb.and Asplenium L. may be related to ecological specialisations, i.e. whether the plants are epiphytic, lithophytic etc. (Tryon 1990).Plants of S. pectinata at both sites were terrestrial.Correlations between spore morphology and overall morphology in the T helypteridaceae have been shown to follow the taxonomic groupings of other authors based on purely morphological grounds (Wood 1973).However, since there have been few studies on the ultrastructure of the exospore in this genus or in the ferns as a whole, it is difficult to assess the taxonomic significance of the varia tion in exospore structure in S. pectinata without further study.

Formation o f the inner perispore
In the spores from both the Cape and the Transvaal, the inner perispore consisted of a darkly staining, discon tinuous and heterogeneous layer which was laid down on the ornamented surface of the outer exospore.In the material from the Cape (Figures 4B; 5A) the deposition of the inner perispore was associated with dense, heterogeneous spherical bodies.In the spores from the Transvaal at the same stage of development (Figure 4A) coiled structures were found in close proximity to the developing inner perispore.There is some evidence (Parkinson 1991) that these bodies are derived, at least in part, from the plasmodial tapetum which disappears at about this time.They are thought to contain some sporopollenin as the perispore has been shown to be acetolysis resistant (Parkinson 1991).These structures also differ structurally and functionally from the spheroids in Psilotum nudum (L.) P. Beauv.(Parkinson 1988) and spherical structures in S. pectinata (Figure 5B) termed composite bodies (Parkinson 1990) which are present during the formation of the outer perispore.I suggest that the coiled and dense bodies are vehicles for the transport of materials, particularly sporopollenin or its precursors, from the plasmodial tapetum to the developing inner perispore.

Outer perispore
The junction between the inner and outer perispore layers was clearly demarcated in section, by an electron lucent area.The inner perispore was a narrow, darkly staining, heterogeneous layer, whereas the outer perispore was not so darkly stained and often had a striated ap pearance in section (Parkinson 1991).Composite bodies were associated with its development in both the Cape and the Transvaal material (Figure 5B).The outer perispore and the composite bodies were shown to contain silicon (Parkinson 1990) and phenolics (Parkinson 1992).
The outer perispore was frequently displaced from the un derlying layers during handling of the material.There were no observable differences between the outer perispore of spores from the Transvaal and the Cape.
The structure of the perispore in ferns and fern allies has been discussed by Hennipman (1970), Lugardon (1974), Schraudolf (1984) and Uehara & Kurita (1989a & b) but the ontogeny of the perispore layer is poorly known.The dense bodies and the coiled structures reported here are not the globules scattered on the spore surface in species with a thin perispore (Lugardon 1981) and which were regarded as the counterpart of orbicules or Ubisch bodies in pollen.Fern spores used to be described as perinous or non-perinous but work at the electron microscope level (Lugardon 1971(Lugardon , 1974) ) has shown that whereas the perine/perispore may be difficult to detect with light microscopy, genera which were originally described as being without a perispore have now been shown to possess this layer.Tryon & Tryon (1982) called for a re-examination of distinctions between genera based on the presence or absence of a perispore and characteristics of its formation and final structure, as it may be taxonomically important.

Tapetal organisation
Tapetal organisation has been well studied in the angiosperms and Goebel (1905) recognised two types of tapeta, secretory and plasmodial, between which transition forms occur (Foster & Gifford 1959).The range of tapetal types and structural variations in the Embryophyta is large (Pacini, Franchi & Hesse 1985) but there have been few studies dealing with tapetal ontogeny and structure in the lower vascular plants.
A detailed, ontogenetic study of the tapetum has been carried out on S. pectinata from the Cape (Parkinson 1991) and this work will form the basis of a separate publication (in preparation).Essentially, what has been determined is that a central, tetrahedral, archesporial initial cell is responsible for the formation of a tapetal initial layer.These cells then divide periclinally initiating the for mation of an inner and an outer tapetal layer.The inner layer differentiates into a periplasmodial tapetum as sociated with the developing archesporial tissue and the outer layer differentiates into a cellular, parietal tapetal layer (Parkinson 1991).A complete ontogenetic study of the tapetum from material from the Transvaal has not been completed.In sections through sporangia from the Transvaal material, at the stage when the outer exospore was fully developed but perispore development was not yet complete (Figure 6A), only a periplasmodial tapetum was present.In sections of the Cape material at the same stage of development (Figure 6B) both a cellular, parietal tapetal layer and a periplasmodial tapetum were present.
The presence of a cellular, parietal tapetum and a plasmodial tapetum existing concurrently in S. pectinata from the Cape is similar in some respects to the condition pre viously reported in Psilotum nudum (Parkinson 1987).The condition also exists in other members of the Schizaeaceae, namely Anemia phyllitidis (L.) Sw. (Schraudolf 1984) and Lygodium Sw. (Binford 1907), as I have de duced from illustrations in these papers.It almost certainly exists in Schizaea tenella Kaulf.(Parkinson 1991).
There are examples in the literature where a parietal tapetum breaks down and becomes invasive during later stages of pollen development as in Pinus banksiana Lamb.(Dickinson & Bell 1972, 1976) and Beta vulgaris L. (Hoefert 1971).There are also descriptions of an invasive but non-syncytial type of plasmodium in Canna L. (Tiwari & Gunning 1986a, b) which is regarded by these authors as being intermediate between the secretory and plasmodial forms.The tapetal condition in 5. pectinata from the Cape is interpreted as being dimorphic (Parkinson 1991) and the present author considers it to be a more truly transitional form between the secretory and plasmodial condition than the one described in Canna.
The significance of the differences noted in the condi tion existing in the material from the Cape and from the Transvaal, at the stage when outer exospore development has been completed, lies in the fact that in studies carried out on the angiosperms, the timing of events in the tapetum at specific stages of microspore development is noted as reflecting family-or genus-specific differences.The change from one type to another within families or genera was interpreted as being a significant evolutionary trend (Pacini, Franchi & Hesse 1985).

CONCLUSIONS
The statistically significant differences in sporangium and spore size, the differences in the outer exosporal or namentation, the differences in the formation of the inner perispore and the variation in tapetal organisation indicate that what has always been considered to be a single species may constitute more than one element.S. pec tinata and S. tenella are the only species currently recog nised in southern Africa (Sim 1915;Schelpe 1970;Roux 1979: Jacobsen 1983: Schelpe & Anthony 1986;Burrows 1990).
Tapetal origin and form was a basis for suggesting phylogenetic relationships within members of the Embryophyta which have been studied more completely, particularly the angiosperms and gymnospcrms (Pacini. Franchi & Hesse 1985).Tapetal organisation in the ferns and fern allies may supply information which could be applied to phylogenetic problems within these groups.It will be interesting to see whether work by Roux (1992) on certain Schizaeaceae.particularly Mohria, will support the present findings based on ultrastructural details.
FIGURE 1.-Gross morphology of plants of Schizaea pectinata.A, J067274 from Ysterkroon in the Wolkberg, Transvaal; B, J067273 from the Cape of Good Hope Nature Reserve.

FIGURE 4 .
FIGURE 4.-Structures associated with the formation o f inner perispore in S. pectinata.A, section through part o f spore wall o f Transvaal material.Sculptured outer exospore (E) with deep indentations, covered by narrow, heterogeneous layer o f greater electron density, the inner perispore.Large bodies (arrowed) have convoluted or coiled appearance.Smaller globules (arrow-heads) found attached to or close to developing spore wall.B, section through part o f spore wall o f Cape material.Sculpturing o f outer exospore (E) with shallowly rounded projections covered by narrow, heterogeneous layer o f greater electron density, the developing inner perispore.Spherical bodies (arrowed) de#se, but show some substructure or heterogeneity and are attached to developing spore wall.Scale bars: A. B, 0.5 ^tm.

FIGURE 6 .
FIGURE 6.-Sections through sporangia: A, showing sculptured nature o f outer exospore in S. pectinata from Transvaal material.Spores associated with last remnants o f periplasmodial tapetum (arrow-heads).N o evidence o f cellular, parietal tapetum which would be situated at position marked by asterisks (*) adjacent to cells o f sporangium wall.B, from Cape material.Sculptured outer exospore has developed and spores associated with periplasmodial tapetum (arrow-heads).Attenuated, cellular, parietal, tapetal layer (*) adjacent to cells o f sporangium wall.Scale bars: A, B, 10 |im.