The genus Trematocarpus ( Sarcodiaceae , Rhodophyta ) in southern Africa and the exclusion of Sphaerococcus ( Chondrus ) scutellatus

Trematocarpus Kiitzing in southern Africa is found to comprise four taxa in three species: newly described is T. fragilis (Ag.) De Toni var. divaricatus Simons, var. nov. Two of the species, T. flabellatus (J. Ag.) De Toni and T. affinis (J. Ag.) De Toni are restored to independent specific rank after being regarded as synonyms of T. scutellatus (Her.) Searles. It is shown that Sphaerococcus (Chondrus) scutellatus Her., the basionym of T. scutellatus refers to a species of Gigartina, therefore a new combination Gigartina scutellata (Her.) Simons is made for this species which was formerly known as G. fastigiata J. Ag. and G. scabiosa (Kiitz.) Papenf. Trematocarpus elongatus Kiitz. is relegated to synonymy under T. fragilis var. divaricatus.

My studies have, however, convinced me that T. flabellatus (J.Ag.) De Toni and T. affinis (J.Ag.) De Toni do indeed have separate status and that T. scutellatus (Her.)Searles is attributable to another genus.I have found, too, that there are forms assignable to Trematocarpus that do not fit any of the published descriptions except, perhaps, that of T. elongatus Kiitz., which was described (Kutzing, 1868)  elongatus in the Leiden Rijksherbarium (NBV) and these are Pappe specimens collected at the Cape of Good Hope.T. elongatus Kiitz. is quite clearly identifiable with T. fragilis (Ag.)De Toni and more particularly with what I regard as a new variety of this species, and that I here describe as T. fragilis var.divaricatus Simons, var.nov.Trematocarpus, as interpreted by Searles (I.e.), is basically cellular in structure and the fusion-cell of the cystocarps forms a complex network.I have found that the South African representatives assigned to Trematocarpus have these characters and, although the internal cells of terminal portions of both T. flabellatus and T. fragilis var.divaricatus are somewhat loosely associated, they are more compacted than are those of T. dichotomus specimens I have seen.It is possible that a separation of Dicurella and Tremato carpus is justifiable, but I hesitate to do so without better evidence.Figs 1-4 illustrate structures of the respective South African taxa.Cystocarps are hemispherical with an ostiole making a minute apical pore.
The identity of Sphaerococcus (Chondrus) scutel latus Hering seemed to have been resolved by Papenfuss (1940) when he found a specimen bearing this name in Herbarium Hamburgense (HBG) (Fig. 5).Not only did the specimen bear Hering's name, but also a description in Hering's own hand, little different from that published (Hering, 1841;Krauss, 1846); two specimens were apparently present, one fertile the other sterile (Papenfuss, I.e.) as was inscribed by Hering on the sheet.The evidence for supposing that the original Hering specimen of 5. (C.) scutellatus had been traced seemed therefore to be complete and irrefutable.Having established that the relevant plant was, in his opinion, a Dicurella and that it was to be equated with the later names D .flabellata J. Ag. and D. affinis J. Ag., Papenfuss relegated these to synonymy of Dicurella scutellata (Her.)Papenf.My own researches led me to distinguish once again between the two J. Agardh taxa, so that it became necessary for me to decide to which of the two the combination D. scutellata should be applied.This could only be resolved by seeing and examining the type of D. scutellata sensu  Papenfuss.This I managed with the help of Dr. G. F. Papenfuss among others, but was disappointed to find only fragments of the relevant material.Nevertheless, I was able to ascertain that all the fragments were of a kind and probably came from the same plant, a tetrasporophyte, and that this plant is probably the same as that seen by Papenfuss, who described (I.e.) 'seriately divided tetrasporan-gia\ This was significant, because the inscription on the sheet included a reference to two specimens 'exempl.sterile et fructiferum'; and also to 'capsulis sphaericis'.The tetrasporangiferous specimen might easily be thought to be sterile, since there is no obvious evidence of fruiting, whereas a cystocarpiferous plant would not escape such notice, and it could well be carrying 'capsulis sphaericis' in the manner of so many Red Algae.I immediately suspected that the 'fruiting' specimen had been misplaced, but diligent searching on the part of the staff at HBG failed to produce it, and it must be accepted that the present sheet is incomplete.This would have no importance were it not for the fact that Barton (1896) identified the fruiting plant as Gigartina fastigiata J. Ag.In itself this is not enough to justify the view that the Hering mount contained two different taxa, but this opinion becomes undeniably weighted when Barton's (I.e.) statements are considered; she states unequivocally that the fruiting specimen = G. fastigiata J. Ag. is 'quite different' from the sterile one which has the 'whole interior filled with roundish, thick-walled cells'.Such an interior characterizes the specimen that Papenfuss saw ('the plant is entirely cellular') and the fragments of the plant now associated with the Hering specimen.Since the Hering description applies more appropriately to the missing fruiting specimen = Gigartina fastigiata J. Ag., especially with respect to the cystocarps with their scutellate ostioles, the epithet cannot justifiably be assigned to what Papenfuss and I recognize as belonging in Dicurella J. Ag.On the contrary, G. fastigiata J. Ag. and G. scutellata (Hering) Simons are to be regarded as synonyms; the Hering name is the earlier and therefore becomes the correct name for the taxon.This combination replaces G. scabiosa (Kiitz.)Papenf., which substituted for G. fastigiata J. Ag., an illegitimate later homonym.
-\== <o; i- ---- (a) var.fragilis This variety (Fig. 6) has long axes, up to 300 mm, that are fairly rigid and branched in the distal half only.Branching is dichotomous-subflabellate with internodes becoming progressively shorter distally.The axils are more or less rounded and the ultimate segments look like the prongs of a hay-fork, that is, widely divergent at first, then becoming more or less parallel.The often entirely terete axes may in some specimens be somewhat compressed, but there is always a terete stipe and the ultimate segments are little, if at all, compressed.Gigartina scutellata (Her.)Simons, comb.nov.Tufts of many erect wiry fronds up to 300 mm high, emerging from a disciform holdfast; fronds scarcely branched in the lower half, thereafter dichotomising subflabellately at progressively short er intervals and attenuated to bluntly rounded or subacute tips.
Two varieties are here distinguished for the first time.The epidermal layer consists almost always of more or less cylindrical cells, with the long axes vertical to the surface, and up to four times longer than wide.Although entirely cellular (in transverse section), there is tissue differentiation into epider mis, outer cortex, inner cortex and medulla.The medulla can, when axes are terete, occupy up to half the sectional area, and consists of cells markedly variable in cross-sectional shape, but they are generally smaller than those surrounding them.The outer cortex contains two or three rows of roughly isodiametric cells, inside which is an inner cortex filled with coffin shaped cells whose long axes are at  Cystocarpiferous plants are often remarkable for their more slender distal segments so that the cystocarps, which in any case are rather prominent, appear disproportionately large.Tetrasporangia occur in pseudonemathecia that more or less cover terminal and subterminal segments.This is mainly sublittoral but, not infrequently, it grows in tidal pools in the lower reaches of the littoral zone.It is characteristic of the Benguela System on the west coast of southern Africa and does not, as far as I know, reach east of Cape Agulhas.There is a record, via Suhr (1834), of Halymenia furcellata var.cartilaginea (= Dicurella fragiliSy fide J. Agardh, 1852) from Algoa Bay (= Port Elizabeth), a locality well east of Cape Agulhas; unlike J. Agardh, I am of the opinion that the plant in question is Trematocarpus flabellatus, because, not only does Port Elizabeth lie beyond the known range of the plant but the illustration by Suhr (I.e.,t. 22,f. 16) shows what looks like a sorus typical of T. flabellatus (<?.v.), a species known to occur as far east as Port Elizabeth.I have, therefore, excluded Halymenia furcellata var.cartilaginea Suhr from the synonymy of T. fragilis and placed it with T. flabellatus.
The most remarkable differences between this variety and the typical variety are the flattened thallus and divergent forking at dichotomies, a characteristic that is not invariable.One or two branchings frequently occur in the proximal half.I have seen and collected many examples of this variety at Bakoven just a kilometre or two north of Hottentotshuisie.It seems always to be submerged at least 2 m below mean sea level.Anatomically there are slight differences between the two varieties such as the rather laxer tissues of var.divaricatus in the ultimate segments, but these are less obvious when comparing the respective female vegetative tissues; the pericarps making up the protective walls of cystocarps, however, seem to be distinguishable.Fig. 2 shows the main feature of the mature tissue-disposition of this variety.
The two taxa that Papenfuss (1940) sank under Dicurella scutellata (= Trematocarpus scutellatus sensu Searles), namely D. flabellata and D. affinis, are in fact distinguished from each other in the features detailed by J. Agardh (1852 and elsewhere).The differences are more subtle than they are obvious: branching in both is dichotomous in one plane and the plants are much the same height.T. flabellatus, though, is dorsiventral (Fig. 9) against the bilateral symmetry of T. affinis (Fig. 10): not only is this true of the vegetative axes, but it applies also to the distribution of the sori, both male and tetrasporangial.In T. flabellatus the segments are plano-convex in section and the sori confined to the plane surface, whereas in T. affinis sori discontinu- ously coat the surface of fertile segments usually causing some distortion of the affected parts.
Epidermal cells are usually cylindrical, being anticlinally elongated.The central tissue is com posed of more or less large, and angled cells in which there may be a core of distinctly smaller cells buried in a thick matrix (Fig. 3).This species commonly occurs in tidal pools forming low-growing (up to 80 mm, rarely more) magenta-coloured compound fan-like tufts that are distinctive and very pretty.The fans are slightly retrorse.Although more typically a Benguela System (i.e.west coast) species, some collections have come from as far east as Port Elizabeth.
Ultimate segments of the axes are subspatulate (Fig. 12) with a convex surface often on the dorsal side, i.e. back of the upward-facing surfaces of the slightly arched lobes of the fan.The other surface is either plane (Fig. 9) or very little concave.The linear sori make a raised pad of tissue on the I have found anomalous forms, some with sori containing only undivided sporangia (?monosporangia) that suggest immature tetrasporangia and others with parts having morphological and anatomical features similar to those of T. affinis.Monospores are not unusual features of many Red Algae and these are variously interpreted.In this instance, their production is sometimes associated with the vegetative anomalies mentioned above.I consider these forms to be hybrids involving T. flabellatus and T. affinis as parents.
Trematocarpus affinis (J.Ag.) De Toni Syll.Alg. 4: 419 (1900) This species is quite a frequent constituent in samples of seaweeds dived out in Cape west coast waters as much as 15 m deep.It does, however, approach the intertidal zone, where, I think, a certain amount of natural crossing with T. flabellatus takes place.Its general appearance is very similar to that of T. flabellatus.Its naviculoid cross-sections (Fig. 10) and its tissue-construction and distribution, nevertheless, distinguish it from that species: epidermal cells are often cuboid (Fig. 13A), whereas the medulla occurs as scarcely more than a median plate of flattened cells (Fig. 13B); cortex consists of more or less isodiametric round-angled cells that may become slightly flattened parallel with the surface; sori more or less surround terminal and subterminal segments and their uneven distribution leads to some superficial distortion.
as having come from the Pacific Ocean.I discovered the representative collections of T. * Seaweed Research Unit, c/o Department of Botany, University of Cape Town, Private Bag Rondebosch 7700, South Africa.

Fig. 5 .
Fig. 5. -Dicurella scutellata.'Type' of Sphaerococcus (Chondrus) scuíellatus Hering in Herb.Hamburg: copy of a photograph of the sterile plant fragments associated with the Hering MS description of S. (C.) scutellatus as seen by Dr G. F. Papenfuss and referred to in his paper (Papenfuss, 1940).

F
i g .8. -Trematocarpus fragilis var.divaricatus var.nov.Portion of a tuft sharing the same discoid holdfast.right-angles to the surface.Fig. 7A & B show the various tissues.

F
i g .12. -Trematocarpus flabellatus.Agardh.Herb.No. 28690 in Lund.Herb.(LD) -part of type material.L. S. through a terminal segment that is tetrasporiferous.Note subspatulate character, and sorus on one surface only.
This species is a common constituent of the intertidal flora of the whole Benguela System that washes the west coast of the Cape Province of South Africa.Its dichotomously branched axes present a similar appearance to those of Trematocarpus, but they are generally more fleshy in the fresh state, and either a deeper purple or a brownish-yellow.