Studies in the genus Riccia ( Marchantiales ) from southern Africa . 16

A description of Riccia albomarginata Bisch. ex Krauss, augmented with the aid of new collections which most closely match the relevant fragments of the type material, is presented. The type collection comprises two species. The fragments that I consider to be in closest agreement with Krauss’s protologue are selected as lectotype. As far as could be established the rest of the type material is probably referable to R. concava Bisch. As previously mentioned by Perold (1989b), Sim (1926) and subsequent authors had applied the name *R. albomarginata' to a different taxon. This taxon, R. albomarginata auct. non Bisch., is now described and illustrated as R. simii, sp. nov. Its distribution and ecology are also noted.


INTRODUCTION
The collections discussed below are in a poor to very poor condition.The relevant descriptions by early authors are very incomplete and lack reference to important diagnostic features such as the free-standing dorsal cell pillars.Furthermore, no illustrations were provided by these authors.In addition, species of section Pilifer, to which the species under discussion belong, are often very difficult to distinguish, particularly in a dry state, as the dorsal pillars remain collapsed and cannot be reconstituted to examine their shape and size.Nevertheless, after a thorough investigation of all available evidence and on the basis of expertise gained through the study of numerous specimens of species belonging to the section Pilifer, I have arrived at the following conclusions: the collection of R. albomarginata, annotated by Bischoff and held at BM (presumably Krauss's collection) is decidedly a mixed collection; the parts alpha and beta referred to by Bischoff on the specimen label (or varieties alpha and beta according to Gottsche et al. 1846)  Thallus monoicous (?), perennial, in crowded gregari ous patches, in partial rosettes or scattered, olivaceous green to green when actively growing, somewhat velvety, with hyaline scales extending above thallus margins (Figures 1A; 2D); rather small (Figure 2A), branches once to several times symmetrically or asymmetrically furcate, medium divergent, 5-7 mm long, terminal segments 1-3 mm long (Figure 2C), 0,7-1,8 mm wide and 0,6-1,1 mm thick, i.e. scarcely wider to nearly twice wider than thick in cross section (  As can be seen from Figure 3B, C, E, the spore ornamentation in R. albomarginata is quite variable, ranging from smallish areolae and a few thick radiating ridges, to mostly very thick, prominent ridges.On the whole, the ornamentation is not markedly different from that of R. concava (Perold 1989) or from the Zeyher collection of R. albomarginata (see below) (the part provisionally referred by me to var.beta), and the spores of these taxa would only be distinguished with difficulty by conventional light microscopical examination.

R. albomarginata has no outstanding vegetative
characters by which it can be readily recognized; however, it is generally rather smaller than most other species in section Pilifer.In the dry state, it can frequently be distinguished by the light brown colouring of the dorsal face and the incurved flanks, fringed with hyaline or crisp, white scales that contrast strongly with the brown flanks, to which, it is thought, the specific epithet refers.Stephani (1898) described the scales as being prominent.Later authors, such as Sim, apparently assumed that the word 'prominent' also meant 'large' and therefore erroneously concluded that R. albomarginata is a species with large scales.In some cultured specimens, the brown colouration of the thalli and pale yellowish stain of the scale cell walls are persistent, even after a year.Sometimes tarry smudges are found at the flanks, as was also noticed in the var.alpha part of the Krauss specimen.It is doubtful whether this is the result of algal or fungal infestation.In their description of R. concava, which follows immediately on that of R. albomarginata (but which they placed in a different section, 'Subtus Squamatae' and not in the Ciliatae, Gottsche et al. noted that R. concava was similar to R. albomarginata var.beta maior (and to R. lamellosa).This apparent similarity between a part of both collections and R. concava, was also found in the present investigation as noted above.

R. albomarginata generally grows on rather coarse
Because of the fragmentary nature and generally poor condition of the type material and the adverse effects of cleaning and pressing the Zeyher collection, only the habit, scales and to some extent, the colour of the thalli, could be checked for each species and were found to be correct.Other characters, such as the vitally important shape and size of the cells in the dorsal pillars and the shape of the thalli in cross section, could not be checked at all, unfortunately.Marginal cilia are definitely absent.
Later authors, such as Stephani (1898), who incidentally had only seen Zeyher's collection, again ignored the fact that the collections were mixed, and treated them as a single species with prominent scales, disregarding the reference to small membranes, proximally not divided into separate lamellae (fide Gottsche et al.).Sim (1926) applied the name to a different taxon (see below), Arnell (1957Arnell ( , 1963) to yet another, and Volk (1981Volk ( , 1983) )  Finally, the scales oiR.albomarginata auct.non Bisch.are very prominent, wavy, white, and apparently with some striations on the cell surfaces (Figure 6F), unlike the smaller (± 600 x 325 -450 nm), clear, hyaline or white membranous scales (Figure 2F) of the type.Sim may also have been misled by Stephani's comments that the scales of Zeyher's plants were prominent, extending much above the thallus margins (but only up to ± 150 and 175 /xm in the two varieties alpha and beta respectively, as measured by me), and that the species had been named for this reason (see under R. albomarginata Bisch.).Sim did not describe spores for his species.Regardless of what influenced Sim, it is now indis putable that he misapplied the specific name albo marginata to a species which is here named and described as R. simii, sp.nov.The specific epithet, simii, has been chosen in deference to Sim's work, as he was the first to describe and illustrate free-standing dorsal epithelial cell pillars in a Riccia species.
Bisch.belongs to section Pilifer Volk, characterized by free-standing, multicellular dorsal epithelial pillars and only lately recognized as a distinct group byVolk (1983) and also bySchuster (1984Schuster ( , 1985) ) as his subgenus Pteroriccia.Gottsche  et al. (1846) had incorrectly classified R. albomarginata under the Ciliatae, which have unicellular marginal hairs, although R. concava Bisch., which they regarded as similar to R. albomarginata (var.beta, maior), was classified under a different group, 'Subtus Squamatae'.However, it appears from Bischoff s observations when referring to R. concava (quoted by G.L. & N.), that the 'small scales' in the dry plants could be taken for cilia!.Stephani (1898) placed R. albomarginata in his Inermes (without cilia) and also failed to report on the dorsal epithelial cell pillars.Admittedly, in the pressed, dried plants of the original material, they are very difficult to recognize.In his key to the Riccia species, Arnell (1963) grouped R. albomarginata and R. concava together with a ciliated species, R. natalensis, partly following Sim (1926) who had classified R. albomarginata, R. natalensis and his un traced species, R. coronata, together under the heading 'epidermal cells, or some of them, elongated or mammillate and free'.It is therefore obvious that earlier authors did not have a clear understanding of the difference between unicellular marginal cilia and multicellular hairs covering the entire dorsal face of the thallus.MISAPPLICATION OF R. ALBOMARGINATA From his own and Potts's collections, Sim was familiar with a Riccia species with large white scales to which he applied the specific name 'albomarginata'.This species is further characterized by 'upper pillars quite free from one another...' which he (Sim 1926) proceeded to describe and illustrate, but without citing specimens.Sim could of course examine fresh material (with the dorsal pillars intact and not collapsed, as they are in long-dried specimens) and this admittedly gave him an advantage denied the earlier workers, who in any case, were totally unfamiliar with such cell pillars.Whether the specimens of R. albomarginata Bisch.cited by Krauss and by G.L. & N. were less fragmentary and in a better state of preservation in Sim's time, is not known.However, Sim evidently noticed signs of loose dorsal pillars in the type specimens [although he overlooked them in R. concava (Perold 1989)] and assumed that his and Potts's collections, presumably the only ones to his knowledge with similar pillars, belonged to R. albo marginata.Volk (1983) finds it 'erstaunlich, dass Bischoff die Haare des Epithels nicht erwahnt', but I venture to state, that, if Bischoff had indeed been dealing with R. albomarginata auct.non Bisch., he, Krauss and Gottsche et al. could not have failed to notice the thick velvety dorsal covering of the thallus, which is still very evident in Sim 338, a hundred years after collection and also after experimental pressing by me.Earlier, Stephani had remarked on the 'ganz diinnes Laub' of R. albomarginata Bisch., which could have referred to either var.alpha or var.beta, but certainly not to the plants described by Sim as 2 mm thick in cross ection.Even if subjected to prolonged pressing, such thick lobes could not have been flattened to the almost paper like thinness of some of the type material.