The genus Schoenoxiphium ( Cyperaceae ) . A preliminary account

The genus Schoenoxiphium of the tribe Cariceae of Cyperaceae is conservatively accepted as being restricted to the African continent and Madagascar. The special features of the inflorescence structure are described. The following species are provisionally recognized: S. basutorum Turrill, S. distinctum Kukkonen, S. ecklonii Nees, S.  filiforme Kukenthal, S. gracile Chermezon, S. lanceum (Thunberg) Kukenthal, S. lehmannii (Nees) Steudel, S.  madagascariense Chermezon, S. perdensum Kukkonen, S. rufum Nees, S. schweickerdtii Merxmiiller & Podlech, and  S. sparteum (Wahlenberg) Kukenthal. A key to the species is provided and their distribution is roughly outlined. The morphological variation within the species suggests separation of taxa below specific level, or perhaps even at species level, but this will require more detailed information about the ecology, distribution and the cytology.

The generic limits within the tribe Cariceae of Cyperaceae are obscure.It seems clear, however, that the whole tribe comprises a well-defined and coherent group.Kukenthal (1909) in his monograph understood it to represent a sub-family and distinguished four genera: Schoenoxiphium, Kobresia, Uncinia and Carex.
It seems to be accepted that the different inflorescence structures exhibit stages in a process of inflorescence reduction, early outlined by Pax (1886, also in Kukenthal, 1909) and discussed by several later authors.A detailed description of this process based on anatomical evidence, however, is still to be given.
It is generally agreed that the reduction proceeded from a large paniculate inflorescence towards, ultimately, a spike.Simultaneously with the reduction in size, specialization took place within the inflorescence.The unisexual flowers became grouped together in different parts of the inflores cence, finally forming separate unisexual spikes (e. g. as in Carex).
Kukenthal (1909) suggested that the genera represent, roughly, successive stages in the general reduction, the genus Schoenoxiphium having the least advanced inflorescences, and Uncinia and Carex with their closed utricles the most advanced types.This might have been plausible at the beginning of the century, but since then much more material has been accumulated and it is now evident that the genera cannot be arranged in a simple linear sequence.It is also evident that the genera Kobresia and Schoenoxiphium are badly in need of a taxonomic revision.Ivanova (1940) produced a revision of the genus Kobresia and made a number of new combinations.For example, she transferred several Carex species to Kobresia and some Kobresia species with panicled inflorescences to Schoenoxiphium.The bulk of the present Schoenoxiphium species she assigned to a new genus, to which she applied the name Archaeocarex.Her suggestions, although they have never been definitely rejected, have not been followed.
The inflorescence of both Kobresia and Schoenoxiphium are rather variable in structure and size.They both show relationships to certain species of Carex.To merge Schoenoxiphium with Kobresia does not, of course, produce any additional evidence of the relationship of these genera (Koyama, 1961).-Kukenthal (1909) gave the restriction of Schoenoxiphium to Africa, and the absence of Kobresia there, as one reason for keeping these genera separate.Later, however, he described Schoenoxiphium kobresioideum from Sumatra (Kukenthal, 1940).This species was subsequently transferred to Kobresia by Kern (1958).
Therefore, the genus Schoenoxiphium as under stood in this paper, occurs only on the African continent and Madagascar (Kukkonen, 1978).The aim of this study is to give a preliminary overview of the genus in this context.

On the inflorescence structure o f Schoenoxiphium
An essential prerequisite for defining the genus Schoenoxiphium and its species is a clear idea of the inflorescence structure.In addition, for accurate identification of the species, mature utricles are required.
The inflorescence vary from a large panicle to, occasionally, a single spike.Mostly, they are branched structures, and it may be noted that a utricle is formed only when the inflorescence branches.The utricle is a prophyll, i.e., the first leaf, or rather the modified sheath of the first leaf on the branch.In the lower parts of the inflorescence, at the base of the large first order branches, it may be sterile.If, in addition, it narrowly sheaths the branch, it can be called a cladoprophyll (Kukenthal, 1909).The branch thus grows through the utricle.
A fertile utricle encloses a single pistillate flower.There are no other pistillate flowers on the same branch.Only staminate flowers with their glumes occur on the branch above the pistillate flower.The branch may, however, rebranch and at the base of the new branches utricles again occur, but these are then each on a branch of the next higher order.
Towards the periphery of the inflorescence the branches protruding from the utricles are smaller, Each branch is produced in the axil of its bract.The bracts of the lowest branches are usually large and leaf-like.A male glume supports the staminate flower.The female glume is homologous with the bracts but as applied here the term refers to a bract supporting the utricle with the female flower and with, as a rule, only a sterile rachilla branch.
The utricle is rather variable, even within the same inflorescence; at the base of the first-order branch, if this is fertile, it is frequently split wide open.In this case a beak can hardly be disting- uished.At the base of the ultimate branches, which frequently remain sterile, the utricles are almost closed and flask-shaped, and, when present, the beak is well defined.Some of the morphological characters are summarized in Table I.The data were mainly obtained from the type specimens, which were studied whenever possible.The inflorescence measurements relate to a single apical inflores cence or to a compact terminal group of lateral inflorescences.Unless otherwise mentioned, the measurements of both the utricles and glumes relate to structures on the ultimate branches of the inflorescence.The length of a utricle includes the beak and, when present, the stipe.
Basically, the word 'rachilla' is used here to denote such a sterile, rudimentary branch, either protruding from the utricle or remaining completely enclosed in it, but this term is sometimes modified with the words fertile or sterile(Figs 1 & 2).

TABLE 1 .
-Schoenoxiphium species: summary of some of the morphological characters C<j C/j Co Co' Co Co Co Co Co Co