34 



PH. GREENWOOD 



designate 'The Pharyngochromis - Chetia - Serranochromis 

 group'. 



The term 'serranochromine' will be used in this paper as a 

 group name for the genera Chetia, Serranochromis, Sar- 

 gochromis and Pharyngochromis. Its use should not be 

 construed as an indication or even a presumption of the 

 group's ultimate and formal recognition as a Tribe. In the 

 sense employed here it is comparable with my earlier use of 

 the informal categories 'haplochromines' and 'pelmatochrom- 

 ines' (Greenwood, 1979 & 1987). Such continued use, and 

 introduction, of informal groupings clashes with the tribal 

 status given by Poll (1986) to several cichlid assemblages in 

 Lake Tanganyika, and with the geographically and taxonom- 

 ically more extensive tribe Haplochromini defined by Eccles 

 & Trewavas (1989), a tribe which also includes the serrano- 

 chromines. In my view, these authors' actions are premature. 

 Too few critical higher-level taxonomic studies have yet been 

 made on the Cichlidae to support the phylogenetic relation- 

 ships that are (or should be) implicit in the award of formal 

 tribal status. For instance, Eccles & Trewavas (1989: 21) 

 define the Haplochromini as: "Maternal mouth-brooding 

 cichlid fishes of Africa and the Jordan Valley in which the 

 basioccipital bone participates with the parasphenoid to form 

 the apophysis for the upper pharyngeal bones". The value of 

 the apophyseal character has been questioned by several 

 authors (see review in Greenwood, 1978, also Greenwood 

 1986) and it may have evolved more than once among 

 African taxa (Greenwood, 1987); maternal mouthbrooding 

 has apparently evolved independently in both the Tilapiini 

 and Haplochromini (the tribes, respectively, sensu Trewavas, 

 1983, and Eccles & Trewavas, 1989), and in one species of the 

 genus Chromidotilapia of the pelmatochromines (sensu 

 Greenwood, 1987: 169) in which paper it is also argued (op 

 tit: 194-199) that this group should not be included, as it was 

 by Trewavas (1983), in the tribe Tilapiini. 



Thus, the purpose of the present paper is simply to clarify 

 the taxonomic status of the 'serranochromine' genera, and to 

 establish a basis for further and phylogenetic studies of those 

 taxa and those of Lake Malawi. 



METHODS AND MATERIALS 



Anal fin spots and true ocelli. One of the features used 

 to define the serranochromines is the presence of maculae on 

 the anal fin, usually in both sexes (Greenwood, 1979). A 

 distinction was made there between true ocelli (such as occur, 

 but almost exclusively in males, in a large number of hap- 

 lochromine species, e.g. those in Lakes Victoria, Edward and 

 Kivu), and the spots or maculae found in the serranochrom- 

 ines and the haplochromines of Lake Malawi (see figure in 

 Eccles & Trewavas, 1991). Judging by a recent description of 

 a new Serranochromis species (Winemiller & Kelso- 

 Winemiller, 1991) it is clear that some confusion still exists 

 when discriminating between these two kinds of anal fin 

 markings. Granted, it is often difficult to do so when only 

 preserved material is available, but in life the difference is 

 obvious, as colour photographs in aquarium books will show 

 (e.g. Konings, 1991). 



The densely pigmented ovoid or near circular centre of the 

 true ocellus, usually circumscribed by a narrow, darkly pig- 

 mented ring, is surrounded, or almost surrounded, by a zone 

 of virtually transparent, or at least freely translucent, and 



apparently unpigmented fin membrane. This clear zone is of 

 variable width and outline, but is often concentric with that of 

 the pigmented centre. In life, the clear zone seems to 

 emphasise the coloured central area, thereby making it stand 

 out from the rest of the fin membrane, be that membrane 

 pigmented or hyaline. In colour photographs of live or freshly 

 dead specimens, the clear zone often appears to be dark or 

 even black, a result either of the dark background against 

 which the fish was posed, or the shadow cast by the fish's 

 body and the anal fin itself. To the best of my knowledge, 

 true ocelli do not occur, at least in nature, on any of the other 

 paired fins, although these fins are often maculate. Anal 

 ocelli are also of rare occurrence in females, but large 

 non-ocellate spots are sometimes present on that fin in the 

 females of species whose males have true ocelli. The spots in 

 such females occupy the same position as the ocelli in males, 

 and are often of the same size. As compared with maculae 

 (i.e. non-ocellate spots) on the anal fin, true ocelli are 

 generally larger, and are always readily distinguishable from 

 the maculae on the other unpaired fins. 



In contrast, non-ocellate anal spots, besides usually being 

 smaller than ocelli, are, with few exceptions (see Greenwood, 

 1992) more numerous and differ little in their overall appear- 

 ance from those on the other fins, although the central 

 pigmented portion may differ quite markedly in colour. The 

 essential difference between maculae and ocelli, however, 

 lies in the absence of a transparent or freely translucent area 

 surrounding the pigmented centre of a macula, which, like 

 that of an ocellus, is bounded by a very thin ring of dark 

 pigment. Instead, the macula's pigmented centre is circum- 

 scribed by a ring, usually narrow, of lighter pigment which 

 separates it from whatever ground colour the fin membrane 

 may have. 



This outer, lightly pigmented ring is not visible in some of 

 the preserved serranochromine specimens I have examined, 

 and the central spot is bounded only by the very narrow ring 

 of dark pigment separating it from the chromatophores in the 

 fin membrane. Whether or not this situation is a preservation 

 artefact cannot be determined at present. 



In their account of the anal fin markings in the newly 

 described species Serranochromis altus, Winemiller & Kelso- 

 Winemiller (1991: 679) describe, unfortunately without an 

 illustration, the fin in both sexes as having ". . .30-40 large 

 pink or pink-orange ovate spots, each ringed with a transpar- 

 ent, white ocellus. . .". I would suggest that the use of the 

 words 'transparent and white' in apposition is somewhat 

 contradictory, and that 'translucent white' would describe the 

 condition more accurately, especially since it is one I have 

 seen in fresh specimens of Serranochromis (and Sar- 

 gochromis) species from the Okavango river and swamp 

 system in Botswana. 



PRESHANK LENGTH OF THE MAXILLA; LENGTH OF THE PRE- 

 MAXILLARY ASCENDING AND ALVEOLAR PROCESSES. The 



preshank length of the maxilla, relative to its shank length, 

 and the length of the alveolar process of the premaxilla 

 relative to the length of the entire ascending premaxillary 

 process, are two morphometric characters not used in earlier 

 papers (Greenwood, 1979, 1989, 1981). 



Preshank length of the maxilla is measured, on the bone's 

 medial aspect, from the anterior tip of the medial arm of the 

 maxilla's saddle process, to the mid-point of the anterior 

 vertical projection from the upper margin of the bone's shank 

 (see Fig. 1). Shank length is measured, also directly and on 



