Annual habit and apomixis as drought adaptations in Selaginella tenerrima

Selaginella tenerrima A. Braun ex Kuhn, widely distributed in the savanna-woodland and savanna zones of tropical Africa, is strictly annual. Dormant megaspores survive the yearly drought. Apomixis has been confirmed through laboratory cultures. Microsporangia are usually absent or very few. and the sporophyte chromosome number is triploid: 2n = 30. Selaginella tenerrima A. Braun Kuhn, largement répandue dans la savane boisée et les zones de savanes de I’Afrique tropicale, est strictement annuelle. Des mégaspores dormantes survivent aux sécheresses annuelles. L’apoximie a été confirmee par des cultures en laboratoire. Les microsporanges sont habituellement absents ou três limités et le nombre de chromosomes sporophytes est triploide: 2n = 30.


INTRODUCTION
The annual habit is very common in the angiosperms, especially those living under a season ally dry climate. Such plants survive the rainless part of the year as dormant seeds, emerge during the rains, grow very rapidly, produce large quantities of seeds, and die away at the beginning of the next dry season. Similar life cycles (with the spores as drought surviving organs) are extremely rare in the pteridophytes, from among which only a few dwarfish selaginellas are unquestionably annuals. Until recently, very little was known about their biology.
S. tenerrima A. Braun ex Kuhn from tropical Africa, combines both these remarkable features of annual habit and apomixis. A report on the seasonal pattern and reproductive biology of this species seems, therefore, to be of interest for students of pteridophyte ecology and evolution.  (Alston, 1957(Alston, , 1959Kornas, 1979;Kuhn, 1868;Schelpe, 1970Schelpe, , 1977Schelpe & Diniz, 1979;Tardieu-Blot, 1964a, 1964b. It certainly is much more widespread than indicated on our map ( Fig. 1), but is very easily overlooked, because of its small size and moss-like appearance. It seems to be more common in the seasonally dry areas of the savanna-woodland and savanna zones, although a few stations in the rain forest zone (in S. Nigeria and Gabon) have also been recorded. It usually occurs in bare soil, forming small tufts, often associated with tiny bryophytes (e.g. Fissidens spp.), in rock crevices, under overhangs, around the bases of boulders in holes in the soil, and in other similar microhabitats, where it enjoys an increased humidity and some kind of protection from the competition of larger vascular plants.

PHENOLOGY AND LIFE-FORM
The seasonal pattern of growth and sporulation of S. tenerrima has been followed in two areas, one situated south of the equator, in Zambia, and the other north of the equator, in N.E. Nigeria and N. Cameroon. In both of them, young sporophytes emerged from the megaspores near the end of the rainy season: in April-June in Zambia and in September-November in Nigeria and Cameroon (Fig. 2). Germination of megaspores was apparently stimulated by the increase in soil moisture; laboratory experiments conducted in Poland have demonstrated, that the fresh megaspores from N. E. Nigeria were able to germinate in January and February, therefore possibly lacking an inherent dormancy period. Sporophytes developed very fast under natural conditions and set strobili in a few weeks. In more humid habitats, they attained a height of up to 6 cm, but in the drier ones they remained dwarfish and often unbranched, forming a single strobilus at the apex of a stem of only 0,5 cm tall (Fig. 3).
The life duration of the sporophytes was extremely short. Soon after the end of the rains, they turned yellow and died. Megaspores were shed in such quantities that they were easily detectable on the soil surface near the tufts. 5. tenerrima is strictly annual: no perennating vegetative organs have ever been detected, and most of the mature sporophytes retained the empty wall of the megaspore at their base (Fig. 3).

REPRODUCTIVE BIOLOGY
All available living and herbarium material of S. tenerrima has been checked for the presence of strobili and the two types of sporangia (see Appendix). Practically every individual plant studied, even the smallest one, bore strobili, the number of which ranged from 1 to 40. Most of the sporophytes contained only megasporangia; in 5 samples (including the type collection from Angola -Welwitsch 45, BM, K) very few microsporangia were also seen (usually no more than 1-3 in c. 100 strobili inspected). The normal number of mega spores per sporangium was four, and their output in all populations was enormous -a feature which is similar to the very high production of seeds in annual angiosperms.

CULTURE EXPERIMENTS AND CHROMOSOME NUMBER
The populations of S. tenerrima studied in the field persisted from year to year even in the absence of microspores and microgametophytes. Apomixis is the only explanation of this situation. We were able to confirm this mode of reproduction in laboratory Megaspores are furnished with a trilete scar marking the commissures along which they dehisce. After a few weeks of watering, a germinating megaspore opened along the scar, expressing the prothallium with three tufts of rhizoids, which were twice as long as the diameter of the megaspore (Fig.  4). Such a quick germination is characteristic of tropical selaginellas as opposed to those of the temperate zones in which the dormant condition may last for many months or even for years (Bruchmann, 1912;Hofmeister, 1867). The prothallia of 5. tenerrima never turned green, even on their tops. Several archegonia were formed on the upper surface of the prothallium. The short neck of each archegonium consisted of four rows of cells and was two cells high. Sporelings which grew in a permanent layer of water (1-2 mm) developed rhizoids which were visible up to the stage of two cotyledons, whereas sporelings growing in drier conditions were without rhizoids and had shorter hypocotyls (Figs 4 & 5).  thick were stained with Newton's gentian violet.
Chromosomes are small and have approximately the same length within the somatic plate (c. 0,7 |im). The chromosome number is 2n = 30 (Fig. 6).
The chromosome numbers in the genus Selaginel la have been established for only c. 80 species out of the c. 700 known (Love et al., 1977). The basic numbers are x = 7, 8, 9 and 10. Therefore the investigated species S. tenerrima proved to be triploid. Unfortunately the mode of reproduction of the remaining triploid species ( (Hieronymus, 1902;Velenovsky, 1905). In both species, empty megaspores are usually found attached to the base of mature sporophytes. Possibly a few other dwarfish selagi nellas, closely resembling those listed above, have a similar life cycle, e.g. 5. ciliaris (Retz.) Spring from S.E. Asia, N. Australia and W. Pacific, or S. perpusilla Baker from Africa and Madagascar. In S. ciliaris, apomixis seems to occur, as indicated by the lack of microsporangia (Alston, 1951;Hieronymus, 1902). Little is known about the habitat require ments of these tiny plants, but most probably they are similar to those of S. tenerrima. On the other hand, the annual selaginellas are included in various infrageneric taxa (5. pygmaea and 5. gracillima in the subgenus Homeophyllum, S. tenerrima in the subgenus Heterophyllum -Hieronymus, 1902) and therefore their striking similarities may result rather from a convergent evolution, stimulated by the water stress in seasonally dry climates, than from a real phylogenetic affinity.