Studies in the genus Riccia ( Marchantiales ) from southern Africa . 27 . Riccia lanceolata and R . radicosa now also locally reported

Riccia lanceolata and R. radicosa are newly reported from southern Africa. Riccia lanceolata is quite widespread in sub-Saharan Africa, whereas R. radicosa has only been collected once before, namely in the Democratic Republic of the Congo (Zaire). Both species are assigned to the group ‘Squamatae’, section Riccia, subgenus Riccia. but are easily distin­ guished from one another. Riccia lanceolata Steph. in Hedwigia 27: 110 (1888).


INTRODUCTION
Riccia lanceolata was originally described by Stephani (1888) from a specimen collected by Newton (G) in Canné.Dahomey (Benin).It was subsequently shown to be widespread in sub-Saharan Africa, Jones (1957) reporting it from Nigeria and Uganda, while Jones & Harrington (1983) mention its occurrence in Ghana as well.Perold (1993) described and illustrated a specimen collected by Frahm in Rwanda, whereas Fischer (1995), besides illustrating it, also provided notes on its commu nities and distribution patterns in Rwanda.Perold (1995) extended the list to include Ivory Coast, Malawi, Mozambique, Niger, Tanzania and Zambia.The present report is, however, the first record for southern Africa, where it was collected in South Africa, near the Gauteng-Mpumalanga border and in Limpopo (Northern Province), at Penge (Figure 1).

DIAGNOSTIC CHARACTERS
Although the thalli of R. lanceolata are more robust than those of R. atropurpurea Sim, there is a close resemblance between the two species: both are dorsally glaucous green and have wavy, hyaline thallus margins.
In R lanceolata however, the finely reticulate dorsal surface lacks any deposit of salts, as is usually found in R. atropurpurea and its purple flanks often appear to be vertically 'striped', due to the 'leading' hyaline margins of the overlapping scales.The spores are quite large, sometimes up to 12 0 //m in diameter, triangular-globular, polar and reddish brown, with the wing distinct, ± 5 //m wide and faintly granular with the marginal angles unperforated.The surface is often incompletely reticu late on both spore faces: the distal face is convex, the are olae.(5)6-8 across, variously incomplete and not extending to the margin.( 10-) 12-22 //m wide, the cen tral walls up to 5 //m high and raised at the nodes, the outer walls low and often absent.The proximal face has It should be stressed, however, that the ornamentation of the spores of even the type specimen (Jones 1957: 219) of this species, Newton 6 (G!), shows some variation (Figure 2A-C).as do others as well, e.g.L.A. Lye B320 (E.ex Herb.Jones!) (Figure 2D-F).Jones & Harrington (1983). in an extreme example, described 'the outer face as areolate over the whole of its surface' in Woodell's col lection from Ghana.Micrographs of one of the South African specimens are included (Figure 3A-C).Previous publications including spore micrographs of R. lanceola ta are Perold (1993. 1996) and Fischer (1995).

South African specimens examined
MPUMALANGA -2529 (Witbank): near Gauteng/Mpumalanga border.20 km N of Witbank Railway Station, on dirt road to Zaaihoek, turnoff to Farm Kwarsspruit and ± 5 km along rough track to Olifants River.(-CA ). on soil in vlei area, between grass. 20-02-1994. Perold 3189.3194 (PRE).Jones (1957) studied an isotype specimen held at BR, adding some details to the original description and illus trating it with, inter alia, good drawings of the spores.In 1995 Perold briefly referred to this species (BR speci men) and later, (Perold 1996) published two SEM micro graphs of the spores.

LIMPOPO (Northern Province
Until recently, the Vanderyst s.n.specimens (MANCH, holo.!; BR, iso.!) were the only known examples of this species.Fortunately, in May 1995 Mrs Colleen Mannheimer of the Windhoek Herbarium, collected bryophytes in the Caprivi Strip, Namibia, at a branch of the Zambezi River near Schuckmannsburg (1724 DB).Mannheimer 134 proved to be a mixed gathering of R. crinita Taylor and, on closer inspection, of R. radicosa.
The collection site was in a floodplain that is seasonally waterlogged and the substrate was wet, loamy clay, where grasses and sedges, as well as algae also grew.The vegetation type at this moist, tropical, northeast comer of the 'Strip' is forest savanna and woodland, the altitude ± 900 m above sea level and the annual rainfall 700-750 mm (Barnard 1998).This locality is quite a considerable distance from where the type specimen was found, although this does not rule out its possible occurrence in the intervening countries of Angola and Zambia, where bryophytes are undercollected.

DESCRIPTION
Thalli small, annual, scattered, dorsally light grey, dull, finely reticulate, once or twice to three times sym metrically or asymmetrically furcate; main branches up to 6 mm long, narrow at base, widening to ± 2 mm before bifurcating, diverging by ± 60°, secondary, and if present, tertiary segments by ± 40°.Branches (Figure 4A) 4.5 mm long, including terminal segments, 2 .0-2.5 x ± 1.5 mm, linear-oblong, gradually tapering toward blunt or wedge-shaped apices when remaining unbranched, oth erwise apically widening if bifurcating again; distally with one or two sharp grooves, the latter soon converg ing and proximally single, eventually becoming only shallowly indented medianly; margins thin, hyaline, scal loped; flanks (Figure 4N) covered with overlapping scales, partly hyaline, partly deep purple-red and gener ally shiny; ventral face bearing dense mat of rhizoids; when dry, thalli dorsally white, occasionally with faint green blotches; margins also white, incurved, revealing scales along sides, sometimes their 'leading' hyaline edge forming a white tuft.In cross section toward apex of single branch (Figure 4G), mid-dorsally narrowly Vshaped.only 0.39 mm thick from bottom of groove to ventral face, at slightly sloping sides 0.7 mm thick, width across 1.2 mm, i.e. 1.7 times wider than thick, ventrally rounded; in section of bifurcating branch at ± 1.25 mm from apices (Figure 4H), deep dorsal groove on either side separating off rounded, central hump of tissue.0.43 mm high, laterally wings arched, flanks rising obliquely to steeply, up to 0.75 mm thick, width across 1.14 mm, ventrally rounded; in progressively proximal sections (Figure 41.J) central hump becoming smaller and even tually wedge-shaped, wings not arched, sloping oblique ly; in next section (Figure 4K) groove single, dorsally Vshaped, 0.41 mm thick from bottom of groove to ventral face, flanks steep, up to 0.9 mm thick, width across ± 1.14 mm, nearly 1.3 times wider than thick; in section toward base (Figure 4L) medianly only shallowly indent ed, wings laterally shortly expanded, flanks obliquely rising to slightly recurving, ± 0.54 mm thick, width across 1.2 mm.Dorsal epithelium with cells hyaline, in one layer, rounded above (Figure 4C, E), and often encrusted with some deposit, 25-45 x 27.5-45.0j/m.some collapsing and sunken in the middle, forming 'ringed' cells (Figure 4D), or shallow cups (Figure 4F), air pores between epithelial cells small, mostly triangular; margins (Figure 4M) acute, hyaline, quite delicate, in cross section only one cell projecting above epithelium; assimilation tissue ± '/2 thickness of thallus, to slightly more than up to 10 rectangular cells attached end to end in each row, 45-65 x 37.5-50.0p m , separated by narrow air canals; storage tissue forming the remaining thickness of thallus, cells crowded together, ovoid.± 52.5 x 65 p m; outer cells of flanks and ventral face greenish brown, not clearly thick ened; rhizoids arising from ventral epidermal cells and from lower 'front' or 'leading' edge of scales, 1 0 .0-22.5 pm wide, mostly smooth, but some tuberculate.Scales bicoloured (Figure 4 0 .P), quite fragile; except for loose, 'leading' edge, the rest closely adherent to thallus flanks and difficult to detach intact, overlapping and consisting of the 'leading', hyaline part.450-550 p m high and 400-590 pm wide, at the 'front' margin sloping oblique ly outward and down, with outer row of cells mostly rec tangular, 30.0-1-2.5 x 37.5-50.0/^m.adjacent and near by cells 4-6-sided.62.5-100.0x 25-40 p m .these hya line cells surrounding inner group of ± 35 or more, pur ple-red cells of similar shape and size: at rear of coloured cell group, hyaline cells mostly rectangular, 62.5-87.5 x 37.5-50.0pm .and joined end to end in rows: then fused to flanks of thallus, where cells also joined together at sides for 12 to 14 rows (up to ± 400 pm wide), some times tom assimilation filaments remaining stuck to inner surface; at the top a row of laterally fused margin al cells, short-rectangular.42.5-57.5x 30-40 pm , exceeding thallus margin, sometimes slightly dusted with fine, powdery deposit.
Monoicous.Antheridia inconspicuous, with short necks turned white, emerging from slightly sunken, faint, brownish flecks in a row along middle of branch.Archegonia with thin, dark red necks along midline of same or a different branch, persistent and toward base projecting from bulging sporangia in close proximity.± 500 pm wide.Spores (Figure 5A-F) 90-120 pm diam., triangular-globular, polar, brown to reddish brown, ridges more deeply coloured, wing undulating, 7.5-10.0pm wide, marginal angles not perforated or notched, margin crenulate: distal face convex, with 7 or 8 angular or rounded areolae across diameter.7.5-15.0x 7.5-15.0p m .bordered by ridges, usually raised and thickened, rarely faint, almost always with tubercles at nodes where joining, up to 5 pm high and stout, occasionally wider and truncate, then plate-like: proximal face with distinct triradiate mark, ends of rays extending across wing, each of 3 facets with 11 or 12 areolae, round or angular.7.5-10.0x 7.5-12.5 //m, ridges quite low.but slightly raised at nodes, sometimes vestiges of areolar ridges reaching partly across wing.

DISCUSSION
It appears that Riccia radicosa is a very rare species, but presumably, it could also be confined to undercol lected areas.It can be recognized by its rather small size, by the deeply grooved apical branches of the thalli, by the closely adherent, overlapping, hyaline and purple-red scales that scarcely exceed the scalloped thallus margins, by the dense mass of ventral rhizoids (for which it had been aptly named), as well as by the distinct ornamenta tion of the winged spores.
The Mannheimer collection of R. radicosa is a small one.consisting of only about 14 thalli at various stages of IE 4 -Riccia radicosa, C Mannheimer 134.A, repeatedly bifurcating thallus branches; B, copy of Pearson's 'radiate' thalli in his plate XIII, fig. 1, C, intact dorsal epithelial cells of thallus and air pores from above; D, collapsed, 'ringed' dorsal epithelial cells and air pores from above, E, c/s intact dorsal epithelial cells and upper cells of assimilation tissue (at this magnification, air canals not visible); F, c/s collapsed dorsal epithelial cells and some assimilation tissue.G-L ,c/s thallus branch: G, apical part with deep, narrow, single groove; H-J doubly grooved part; K, wider, shallowly grooved part; L, basal part.M, c/s thallus at margin, with overlapping scales on outside (left) and dorsal cells and cell columns on inside (right); N, flank of thallus with scales seen from side, broken line indicating cut edge; O, P, bicoloured scales, proximally fused to cells covering flanks.and represents a 'main' branch.This does not exceed by much the longest branches in Mannheimer 134, at 6 mm.The absence of rosettes in the latter specimen could per haps be ascribed to a shortened growth period on an unsta ble alluvial substrate.In addition, some of the smaller thalli were partly overgrown by algae, which may have had a detrimental effect.At the time of study, nine years had elapsed since this specimen was collected.On wetting a small portion, the thalli did not fully recover their form, the flanks remaining somewhat inflexed.
Jones (1957) refers to the MANCH specimen (which he did not see) as the type, since Pearson was resident in Manchester.Pearson clearly kept the portion of Vanderyst's collection, which was sent to him by Mr Naveau, whom he thanks for the opportunity of examin ing the collection (the rest of the specimen is held at BR).In the holotype specimen, there is very little soil mixed with the abundant rhizoids underneath the thalli, which appear to have been pressed, as the flanks have been flat tened.The scales are nearly black, with the hyaline 'lead ing' parts barely visible.In Pearson's illustration (plate XIII, fig.2), the scales, when seen from above, are drawn as cross-hatched squares separated by white rectangles, and he describes them as 'subquadrate, distant, dark pur ple'.He also notes that they 'are not imbricate as in R. nigrella DC., but are remote'.Jones (1957) on the other hand, depicts the scales along one of the flanks, as seen from the side, in his fig.5b; they are clearly overlapping, with a black patch at the lower, rear comer, the rest remaining ± white.He also remarks that the 'ventral scales are imbricate or locally distant, deep violet with broad hyaline margin'.The scales are undoubtedly over lapping, but because they are so difficult to detach with out tearing, neither of the previous two authors attempt ed to illustrate them individually.It would appear that in the type specimens, the walls of the flanks underneath the scales are also purple.
Pearson's 'dimensions' for the spores are incorrect, but this is probably due to a typographical error, as the decimal point is in the wrong position, i.e. not ' .9mm..1.mm.x .85mm.', but are meant to be 90-100 x 85 pm , with ' 6 to 7 areolae across surface'.Jones's measure ments are '80-95 p m diam.with 8 or 9 areolae across the face', but he concedes that 'the spores are so opaque that it is by no means easy to ascertain the exact number'.My measurements of the BR spores are 90-115 p m and are illustrated in Figure 6 A-F.
The above two species both belong to subgenus Riccia, section Riccia and to group 'Squamatae'.How ever, R. lanceolata is more robust than R. radicosa, its scales are also difficult to detach from the flanks, and are deep violet to mauve, with a hyaline margin, but with some gradual shading, not clearly delimited to a purple blotch as in R. radicosa.Its winged spores are variably incompletely reticulate on both faces, as opposed to completely reticulate in R. radicosa.