Studies in the genus Riccia (Marchantiales) from southern Africa. 23. R. bullosa: typification and a full description

In the protologue of Riccia bullosa, Lindenberg (1829) based his vague and incomplete description of the taxon on two heterogeneous elements: a specimen from Portugal and one from the Cape. Stephani (1898) retained the name Riccia bullosa for the Cape element and referred the Portuguese plant to Exormotheca welwitschii (Stephani 1899). Although generally accepted until 1940. K. Muller (1940) rejected Stephani's segregation and referred the Portuguese element to his new combination. Exormotheca bullosa (Link) K. Muller, based on the false assumptions that Stephani had failed to recognize that the Portuguese plant was an Exormotheca. and that the Cape plant was insufficiently known and had not been found again. Counter arguments are set out here, urging a return to Stephani’s earlier segregation. A lectotvpe for Riccia bullosa is selected, which results in the repudiation of Muller's (and Grolle's) lectotypification of Exormotheca bullosa (=Riccia bullosa).

In the protologue of Riccia bullosa, Lindenberg (1829) based his vague and incomplete description of the taxon on two heterogeneous elements: a specimen from Portugal and one from the Cape.Stephani (1898)  Stephani's interpretation of R. bullosa was generally accepted by South African authors, namely Sim (1926).Duthie & Garside (1937).as well as by the Swedish hepaticologist, Arnell (1963).who even cited Stephani as the author of this species, and also by myself (Perold 1989).Furthermore.Goebel (1905) in a footnote, had accepted Exormotheca welwitschii Steph.(Riccia bullosa Link) and so had K. Muller (1906Muller ( -1911)).although the latter admitted that he had not seen the original material at that time.Subsequently, however.K. Muller (1940) 1829) had clearly indicated that the Portuguese element represented the only authentic and original material, because Lindenberg had cited Link (who had brought this specimen from Portugal) as the author and had also adopted Link's epithet, namely Riccia bullosa.In addition, Muller argued that Link had not even known about the Cape plant.In Muller's view, he was acting in accordance with the author's (i.e.Lindenberg's) intent.Furthermore, in his opinion Lindenberg's description of the Link species was wrongly expanded by his inclusion of the Cape species and, of even greater importance to Muller, he thought that the drawing of the thallus in cross section, on Table 23, Figure  In considering the various arguments put forward by Muller (1940Muller ( , 1941Muller ( , 1947Muller ( , 1951Muller ( -1958) ) in support of his conclusions, Lindenberg's (1829) protologue must be examined first.The fact that Lindenberg cited Link as the author and adopted Link's name for the species are not significant nomenclatural issues, since Link had not validly published the name or a description of the plant: it was merely a manuscript name, as was Nees's name of R. crassa for the Cape plant that Lindenberg also referred to in the same description.The two references in this protologue (at the beginning and again at the end) to a groove, in my opinion clearly point to the Cape Riccia which, especially in dried material, is distinctly grooved medianly.Stephani (1898) described it as 'profunde sulcata', whereas Exormotheca is less obviously so, and except for E. tuberifera Kash.(Schiffner 1942), members of this genus are not reported elsewhere in the literature as being grooved.
For the following reasons I also do not agree with Muller that Lindenberg's (1836) drawings are those of an Exormo theca: a narrow, deep dorsal groove is clearly shown from above as well as in the transverse section of the thallus; there are no assimilation filaments arising from the base of the air chambers: no stomata are drawn in Lindenberg's figures (Figure 1.5, 1.7 & 1.8) and in fact, in his more detailed description, Lindenberg ( 1836) stated that open ings or pores are absent.Schiffner (1942) also remarked that stomata are absent, which is not strictly correct, but such an observation could never apply to Exormotheca, a genus in which stomata are very obvious.Ventral scales, another striking character of Exormotheca, are also not illustrated by Lindenberg (1836) and finally, about twice as many air chambers are shown across the width of the thallus in the cross section, as the six (or so) that Muller described for Exormotheca.Muller (1947), however, was so confident that Lindenberg's cross section illustrated an Exormotheca that he accused Schiffner of remaining silent about it, when Schiffner (1942)  To summarise: judging from Lindenberg's (1829) refer ence to a 'grooved' plant, as well as his drawings and expanded description (Lindenberg 1836), he was referring to the Cape plant, even though he had credited Link as the author and had adopted Link's unpublished name for the species.These are, however, without nomenclatural significance, as remarked upon before.
Article 8.1 of the I.C.B.N. states that: the author who first designates a lectotype must be followed (in this case Muller), but his choice is superseded if various conditions are not met: (b) if it can be shown that it is in serious conflict with the protologue and another element is available which is not in conflict with the protologue (the Cape element is 'grooved', whereas Link's is not, but otherwise the descrip tion is vague, and one may hesitate to invoke the term 'in serious conflict w ith....'); (c) if it was based on a largely mechanical method of selec tion (Muller could not locate the Cape element and only the Link element was available to him, which can be interpreted as a 'largely mechanical method of selection'); (d) if it is contrary to Article 9.2, which states 'if it is proved that such a type herbarium sheet or preparation contains parts belonging to more than one taxon, the name must remain attached to the part (lectotype) which corresponds most nearly with the original description' (in this case 'grooved').Thallus dioicous (Figure 2A,B), ?perennial, in crowded, gregarious patches or scattered (Figure 3A); apically light green, soon turning straw-coloured, deeply grooved, laterally swollen to bloated, with small, polygonal, domed areas, becoming proximally pitted and spongy; when dry, rather deflated and with folds across, sides not inflexed, margins scalloped; branches once, occasionally twice dichotomously furcate, sometimes simple, shortly to deeply divided, moderately to widely divergent; broadly ovate to oblong, 5,0-15,0 x 3 ,5 -5 ,5 mm, 1 ,5 -2 ,5 mm thick, in section 2 ,0 -2 ,5 times wider than thick; apex obtuse to truncate, emarginate.Groove deep and narrow, sometimes split into two by raised wedge of tissue (Figure 3B), disappearing toward base or at sporangia.Thallus margins obtuse, rounded, often overhanging.Flanks sloping obliquely, ventral face rounded to keeled, green.Scales hyaline, vestigial (Figure 3C  Anatomy o f thallus: dorsal epidermis forming a domed roof over each air chamber (Figure 2C).cells 4-6-sid ed , 6 2 -7 5 X 3 5 -4 0 /xm, air pores ringed by 6 or 7 wedgeshaped cells, smaller (Figures 2D, E; 3B) and often Riccia bullosa is endemic to southern Africa and is found at seepages or on damp sandy soil underneath brush, or at granitic, basaltic or sandstone outcrops in the western and southern Cape and in the subalpine belt of the Drakensberg Mountain range of Natal and Lesotho in grasslands or bogs (Figure 4).It is placed in the subgenus Ricciella and can be distinguished from related species, R. garsidei and R. volkii, by its straw-coloured or yellowish green, rather swollen thalli.R. garsidei Sim is often larg er, almost white when dry, with many exposed air cham bers; its spores have fewer and larger areolae.R. volkii is less robust and not so bloated, with narrowly winged, smaller spores and its distribution is restricted to the sum mer rainfall areas.Arnell (1963) regarded R. bullosa as intermediate between (Eu-) Riccia and Ricciella with its narrow air spaces in the median part of the thallus and wide air chambers in the lateral parts.This is rather similar to Stephani's (1898) interpretation of R. vesiculosa Carr.& P. from Australia, which prevented him from dividing the genus Riccia into two subgenera, but as Na-Thalang (1980) and Volk (1983) have shown, it is not correct to separate these two subgenera on the size of the air chambers; the anatomy of the air chambers, the pores and the arrange ment of the epidermal cells should be considered to be of greater importance.Arnell (1952) described a new species, R. montaguensis, admitting that it was very similar to R. bullosa in habit and colour, although somewhat smaller.The supposedly smaller spores, which he reported to be 80 /xm in diameter, were found to be rather larger at 100-130 /xm (refer also Specimens of R. montaguensis were closely examined and found to be indistinguishable from R. bullosa, which can vary considerably in size from rather small to large.Spores from all the sporulating material of this species at BOL and the many recent collections at PRE have been measured and photographed and they exhibit a continual gradation in size and also in the completeness or incompleteness of the ornamentation, so that a broader circumscription of R. bullosa, which includes these variations, is necessary.R. montaguensis S. Arnell is there fore included in the synonymy of R .bullosa.Close examination of Nees's material, now held at STR, shows that the contents of the two packets labelled Riccia crassa are indeed R. bullosa.One of these two specimens is a single thallus containing a sporangium with ripe spores (Figure 3E, F).The other specimen is here selected as lectotype because it agrees well with the protologue, its locality is clearly stated and it is reasonably good material; the specimen is identified on the label and in the proto logue as R. crassa, but no collector's name is given.It must have been Ecklon, however, (Lehmann 1829), as also cited by Stephani (1898).The latter's three herbarium specimens at G were seen, and these are most probably duplicates.They had been identified as R. bullosa and no mention is made of R. crassa.For this reason none of these specimens were selected as lectotype.The contents of a third packet, labelled R. bullosa, cannot be identified.The packet labelled R. ecklonii c. tr.R. bullosa, appears to contain a male plant of R. purpurascens Lehm.This is also inferred from Nees's (1838) description, where it is added in a footnote that the male plant he described was from kVorgebirge der guten Hoffnung', meaning the Cape of Good Hope, but the word, Cape, was left out [as Lindenberg (1836) also did].Nees (1838) also described the antheridia causing bulging of the frond above and be low.The bulges below were, according to him, richly sup plied with rhizoids and therefore resembled sporangia of R. natans (= Ricciocarpos natans (see Perold 1990, Figure 6F).Such dorsiventral swellings are quite conceivable in the thin thallus of R. purpurascens but hardly in a thallus as thick as that of R. bullosa.

Typification of
retained the name Riccia bullosa for the Cape element and referred the Portuguese plant to Exormotheca welwitschii (Stephani 1899).Although generally accepted until 1940.K. Muller (1940) rejected Stephani's segregation and referred the Portuguese element to his new combination.Exormotheca bullosa (Link) K. Muller, based on the false assumptions that Stephani had failed to recognize that the Portuguese plant was an Exormotheca.and that the Cape plant was insufficiently known and had not been found again.Counter arguments are set out here, urging a return to Stephani's earlier segregation.A lectotvpe for Riccia bullosa is selected, which results in the repudiation of Muller's (and Grolle's) lectotypification of Exormotheca bullosa (=Riccia bullosa).UITTREKSEL In die protoloog het Lindenberg (1829) sy vae en onvolledige beskrywing van Riccia bullosa op twee heterogene elemente gebaseer: 'n eksemplaar uit Portugal en een uit die Kaap.Stephani (1898) het die naam Riccia bullosa vir die Kaapse element behou en die Portugese plant na Exormotheca welwitschii (Stephani 1899) verwys.Alhoewel algemeen aanvaar tot 1940.het K. Muller (1940) Stephani se segregasie verwerp en die Portugese element verwys na sy nuwe kombinasie.Exormotheca bullosa (Link) K. Muller, op grond van die verkeerde aannames dat Stephani nie besef het nie dat die Portugese plant "n Exormotheca is en dat daar te min oor die Kaapse plant bekend was en dit nie weer gevind is nie.Teenargumente ten gunste van die terugkeer na Stephani se vroeere segregasie, word hier aangevoer.n Lektotipe vir Riccia bullosa word aangewvs.met die gevolg dat Muller (en Grolle) se lektotipifisering van Exormotheca bullosa (=Riccia bullosa) verwerp word.HISTORY AND TYPIFICATION OF RICCIA BULLOSA Recently Dr R. Grolle (in litt.)drew my attention to nomenclatural problems concerning Riccia bullosa Link ex Lindenb., that have apparently not yet been satis factorily resolved.This is due to Lindenberg (1829) having based his original species description on two hetero geneous elements: a specimen from Portugal, now referred to Exormotheca and a Riccia specimen from the Cape.Lindenberg stated that he had received the Portuguese specimen under the manuscript name.Riccia bullosa Link, and the Cape specimen from Nees under the name R. crassa.The protologue (Lindenberg 1829) was incomplete and rather vague, and for the most part, it could apply to either plant.In his Monographic der Riccien.Lindenberg (1836) gave an expanded description and drawings (Figure 1) of the entity which he referred to as Riccia bullosa Link.He still considered the Portuguese and the Cape specimens as both belonging to the same entity, even though he observed clear differences between them.He described the Portuguese plant as being less yellow and more lax and the layer of 'horizontal' cell tissue as being much thinner.He also mentioned that the Cape specimen had been collected b\ Ecklon.The illustrations of R. bullosa in Lindenberg (1836) depicts the Cape plant, as will be shown below.Stephani (1898) Ux>k up the name Riccia bullosa for the Cape element and excluded the element from Portugal which in his opinion was clearly not a Riccia.At the time he was uncertain to which genus it belonged.He stated, however, that it matched a specimen collected in 1847 by Welw itsch ( Welwitsch 33.see Mitten 1853) at Vendas on the banks of the Tejo River.In the follow ing year Stepha ni (1899) referred the Portuguese plant to the genus Exor motheca and described it as E. welwitschii Steph.. placing Riccia bullosa Link in Lindenb.ex parte in synonymy un der it.
1.6 (Lindenberg 1836) reproduced here as Figure 1, left no doubt at all that Riccia bullosa was identical with Exormotheca welwitschii.By his use of the words 'authentic' and 'original' for Link's Portuguese element, Muller (1940) effectively lectotypified Riccia bullosa (see Recommendation 7B.4 of the I.C.B.N.).This was accepted and followed by Grolle (1976), who specifically cited the Lindenberg Herbarium No. 9037 as the lectotype.
had clearly stated that the description and illustration referred to the Cape Riccia only.Schiffner's (1942) own drawings of Link's specimen from Portugal (Herb.Lindenberg No. 9037) (Figure 5a, b, c) are obviously those of an Exormotheca with triangular scales and with tall air chambers which contain assimila tion filaments at the base.Muller (1947) admitted to not having seen the Cape plant, because he could not locate the specimens in Lindenberg's herbarium in Vienna, or at the herbarium of the Botanical Museum in Berlin.He concluded that they must have been with Schiffner in Vienna.In Muller's (1947) opinion, the 'ziemlich ungeklarte Riccia vom Kap' was not found again, and because he regarded it as a dubious species, it should get another name!None of these arguments, forwarded by Muller to justify his actions, are correct.Schiffner had by 1942 decided that Exormotheca bullosa (Link) K. Miill.nov.comb, should be withdrawn as a species, 'als Art einzuziehen', and he referred to the Portuguese plant as Exormotheca welwitschii.Muller, however, (1947, 1951-1958) continued to defend his point of view.In passing, it could perhaps be mentioned that both Schiffner (1862-1944) and Muller (1881-1955) were engaged in studying and describing Exormotheca species at about the same time and there may have been a measure of competition between them.

FIGURE 1 .
FIGURE 1. Reproduction of TAB.XXIII.Fig. 1.Vol.XVIII PI Monographie der Riccien by J.B.G.Lindenberg.p. 443.Captions translated into English: 1.1, Riccia bullosa, natural size; 1.2, enlarged.The walls o f the air chambers appear as veins on the surface; 1.3, a piece of the thallus, enlarged, from below; 1.4.a piece of the thallus, much enlarged, so that the texture together with the transparent air bladders are visible; 1.5, a piece of the thallus, even more enlarged; 1.6, transverse section: a.a. the air canals; b.b. the same with side walls; 1.7, a piece of the epidermis, much enlarged: 1.8, some air canals, of which some are divided by a cross wall.