A revision of the genus Scleria Bergius (Cyperaceae) in Southern Africa.

Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 1983.

The genus Scleria Bergius (Cyperaceae) in Southern Africa is critically examined, and the generic limits reviewed The taxonomic position of the genus in the family is examined. Diagnosis of the tribe Sclerieae is altered to circumscribe Scleria as the only genus, and diagnoses of the tribes Bisboeckelereae and Sclerieae are made. Infrageneric limits are re-assessed and two subgenera, Scleria and Hypoporum recognised, the relationship of which is postulated as co-lateral, not filial. Evidence is presented that ecological specialisation in subgenus Hypoporum has resulted in taxa which are adapted to open, seasonally dry, temperate habitats, whereas ecological specialisation in subgenus Scleria has given rise to taxa which are adapted to shady, wet, tropical and subtropical habitats. Taxa in subgenus Hypoporum are slender, usually narrow-leaved annuals, or perennials with annual aerial parts, that is, they have evolved drought/cold escape mechanisms, the annuals by completion of the life cycle in a season, the perennials by withdrawal of food reserves into a protected, subterranean perennating organ and sometimes also into enlarged culm-bases. Taxa in subgenus Scleria are more-or-less robust, usually broad-leaved perennials, or, less often, annuals. With few exceptions the plants are evergreen and do not manifest drought/cold escape mechanisms. The annual species occupy tropical habitats in areas where seasonal drought may be experienced and it is suggested that they have acquired the annual habit as a drought-escape mechanism. The only perennial species in subgenus Scleria in Southern Africa which has annual aerial parts, has evolved additional storage regioni in the swollen culm-bases. This species, S. transvaalensis occurs at higher, more temperate altitudes than other species in the subgenus. The fundamental branching pattern of the inflorescence of all species examined has been shown to be the same. It is postulated that the pattern is modified in two ways, namely, by progressive contraction of all or most ramuli leading to the "glomerate-spicate" type of inflorescence characteristic of subgenus Hypoporum, in which the bracts are reduced, glumiform structures, and, by progressive contraction of some ramuli and progressive elongation of others leading to the "interrupted-paniculate" type of inflorescence characteristic of subgenus Scleria, in which the bracts are foliaceous. It is suggested that the branched glomerate-spicate type of inflorescence is less specialised than the simply glomerate-spicate type, and that in the line with interrupted-paniculate inflorescences, the greater the degree of elongation and the greater the number of elongated ramuli, the more highly specialised the inflorescence. Evidence is put forward that the spikelet of Scleria is a monopodial, that is, racemosely-branched structure, and suggestions that it may be sympodial, refuted. It is postulated that unisexual spikelets in Scleria have been derived by reduction from bisexual (androgynaeceous) spikelets. Unisexual female spikelets are unknown in subgenus Hypoporum which has androgynaeceous spikelets and unisexual male spikelets. It is suggested that the higher the ratio of androgynaeceous to male spikelets in the inflorescence, the less specialised the inflorescence. Unisexual male and functionally unisexual female spikelets occur in subgenus Scleria and, rarely, androgynaeceous spikelets. It is suggested that species which consistently produce some androgynaeceous spikelets are more primitive than those which consistently lack them, and that species whose functionally female spikelets consistently lack any vestigial male parts are more advanced than those which have male rudiments. The hypogynium or "disc" which is present on some achenes is considered to be a new modification of the stipe of the achene and not a vestigial structure, therefore it is postulated that the type of achene found in subgenus Hypoporum which has a trigonous stipe lacking any distal elaboration as an hypogynium is primitive, and that the type of achene found in subgenus Scleria which has an obpyramidal stipe elaborated distally as an hypogynium is derivative. It is suggested that development of the hypogynium has attained its most specialised level in one group of species in which this structure serves as a flotation device. Diagnoses of sections in subgenus Scleria are based partly on differences in morphology of the hypogynium. It has been demonstrated that the cell-walls of the pericarp are silicified, that the process of silification is progressive proceeding from the apex towards the base of the fruit, and that abscission of the fruit takes place when silicification is complete and the vascular supply is severed. The achenes of subgenus Hypoporum have all cells silicified; those of subgenus Scleria have all cells except those of the free flange(s) of the hypogynium silicified. Scanning electron microscopy has revealed details of surface ornamentation of the achenes not previously known, which provide additional diagnostic characters at species level. Attemptsto germinate achenes of Scleria have been unsuccessful : the conditions required, physical and physiological, are not understood. Analysis of anatomical evidence, in particular those characters seen in transverse sections of laminas and culms, confirms that there has been specialisation along two divergent pathways; one which has led to successful occupation of relatively dry, temperate habitats (subgenus Hypoporum), and the other to successful occupation of shaded damp, and open aquatic, subtropical and tropical habitats (subgenus Scleria). One section, Hypoporum, is recognised in subgenus Hypoporum, pending survey of the subgenus on a world basis. Four sections are recognised in subgenus Scleria, namely, Scleria, Acriulus, Schizolepis and Ophryoscleria. Section Scleria may comprise several natural groups, the delimitations of which can not be attempted until a world survey has been made. A map showing world distribution of the genus, and regional distribution maps of Southern African species are provided, also a Table showing the total distribution range of species recorded from Southern Africa. Generic, subgeneric, sectional and species descriptions are provided. Two new species are described. Keys to the Southern African species for use in the herbarium, in the field, and one based on anatomical characters of the laminas are presented. Original descriptions and photographs of type specimens of taxa represented in Southern Africa are included as appendices.