THE SERRANOCHROMINE GENERA REVIEWED 



41 



phylogenetic context how are these similarities and differ- 

 ences, to be evaluated? 



Again, a pattern of reduced ctenoidy like that in the 

 serranochromines occurs in some but not all species of 

 Thoracochromis (Greenwood, 1979; 291; 1984:192 & 200), a 

 genus in which, apparently, there are both true ocellar and 

 non-ocellate types of anal fin markings (see above), but with 

 both kinds occurring only in males. Possibly the 'genus' 

 Thoracochromis is polyphyletic and that some of its species 

 should be included in the serranochromine assemblage. 

 Here, as is so often the case, one is hampered both by a 

 paucity of detailed information on live coloration and the 

 relatively few specimens available for anatomical and mor- 

 phological studies. 



Another difficulty, linked with lack of information, lies in 

 the possibility that further research could establish that there 

 really is a close phylogenetic relationship between the serra- 

 nochromines and certain haplochromines of Lake Malawi. In 

 that eventuality, it is possible that the nearest relative of one 

 or more of the fluviatile serranochromine taxa is to be found 

 in the lake's fauna, thus rendering the serranochromines, as 

 currently conceived, either a para- or a polyphyletic group. 



Although at present the monophyly of the serranochrom- 

 ines cannot be established or refuted, it is possible to con- 

 struct, on the basis of shared derived features, a tentative 

 intragroup taxonomy. 



As compared with Pharyngochromis , the genera Chetia, 

 Serranochromis and Sargochromis all share two derived fea- 

 tures (see Liem, 1991) associated with the upper jaw skele- 

 ton, viz an increase in the shank length of the maxilla relative 

 to its pre-shank length, and an increase in the length of the 

 alveolar process of the premaxilla relative to the length of the 

 entire ascending process of that bone (see Methods). In 

 Pharyngochromis the preshank portion of the maxilla is from 

 1.2-1.3 times longer than its shank length, whereas in Chetia 

 and Sargochromis the two parts are of equal length (with, in 

 some Sargochromis species the preshank portion slightly 

 shorter) and in Serranochromis the shank is noticeably longer 

 (as much as 1.3 times so). In Pharyngochromis the length of 

 the alveolar process of the premaxilla's ascending process is 

 60-66% of the length of the entire ascending process; in 

 Sargochromis it is from 69-76%, in Chetia 13-11% and in 

 Serranochromis 73-83%. 



Neither of these ratios, either inter- or intragenerically, 

 appears to be influenced by the size of the 17 specimens 

 examined, all in the size range 79-410 mm S.L. and repre- 

 senting ten species. 



On the basis of those two characters, a Pharyngochromis 

 and a Chetia - Serranochromis - Sargochromis subgroup can 

 be recognised within the serranochromines. Both these 

 derived features are the only ones shared by the three latter 

 genera (the increased number of abdominal vertebrae used 

 previously to unite Sargochromis and Serranochromis in a 

 single genus [Greenwood, 1979] is now thought to be a 

 homoplasy; see p. 37). 



Chetia and Serranochromis both share three derived fea- 

 tures not found in Sargochromis or in Pharyngochromis viz. 

 (i) an increased modal number of caudal vertebrae (range 

 15-17, modes 16-17 in Chetia, and 15-18, modes 16 and 17 in 

 Serranochromis) compared with Sargochromis (range 12-16, 

 modes 14 and 15) and Pharyngochromis (range 14-16, mode 

 15). (ii) Two or more vertical rows of postorbital scales (c/a 

 single row in Sargochromis and Pharyngochromis , see p. 36). 

 (hi) An increased number of scale rows around the caudal 



peduncle, i.e. 18-20, rarely 16, compared with 16-18 rarely 

 15, (mode 16), in Sargochromis, and 15 or 16 (mode 15) in 

 Pharyngochromis . 



On the basis of those three derived characters, and using 

 the term 'sister taxon' without any phylogenetic implications, 

 then, within the serranochromines, Pharyngochromis is the 

 sister taxon to the group Sargochromis, Chetia and Serrano- 

 chromis, and within that latter group, Sargochromis is the 

 sister taxon of Serranochromis and Chetia combined. 



That scheme bears, in broad outline, a close resemblance 

 to Trewavas's (1964) diagram suggesting the interrelation- 

 ships of Serranochromis, a scheme based essentially on 

 lateral-line and dorsal fin ray counts (which may, of course, 

 be correlated, in part, with the vertebral counts used here) 

 and certain characteristics of the pharyngeal jaws. 



When comparing the two schemes, allowances must be 

 made for the fact that two of Trewavas' Haplochromis species 

 (lucullae and darlingi) are now treated as synonyms of 

 Pharyngochromis acuticeps (Greenwood, 1992), that the sta- 

 tus of H. angolensis, H. humilis and H. toddi is uncertain or 

 unknown (Bell-Cross, 1975, Greenwood, 1979), and that H. 

 welwitschi is now considered to be a species of Chetia (see 

 p. 39). Also, the two Haplochromis species, mellandi and 

 frederici, placed in limbo between the genera Haplochromis 

 and Sargochromis in Trewavas' diagram, are now included in 

 Sargochromis (Bell-Cross, 1975; Greenwood, 1979 and above 

 p. 37), as is Haplochromis carlottae. 



Trewavas (op. cit. 9-10) superimposed on her diagram a tree 

 indicating the possible phylogenetic relationships of the taxa, 

 both inter- and intragenerically. It is here that our views would 

 not coincide, mainly because I do not think there is the evidence, 

 based on cladistic methodology, to justify the relationships 

 proposed, even at an intergeneric level (see p. 40). Certainly 

 there are no features justifying Trewavas' (op. cit. 10) suggestion 

 that Serranochromis is a diphyletic and gradal taxon or that a 

 cladal grouping would recognise Chetia, Serranochromis robus- 

 tiis and S. thumbergi on the one hand, and Chetia welwitschi 

 (Trewavas' Haplochromis welwitschi), and the remaining Serra- 

 nochromis species on the other. Nor can I accept Trewavas' 

 uncertainty about the separation of Sargochromis codringtoni, 

 mellandi, carlottae and greenwoodi (as Haplochromis frederici in 

 her scheme; see Bell-Cross, 1975) from the 'Haplochromis' (i.e. 

 Pharyngochromis,) root of her tree (see Greenwood, 1979, 1992 

 and above). A truly phylogenetic assessment of the groups' 

 relationships must, as discussed on p. 36, await the results from 

 further and preferably multidisciplinary research into the system- 

 atics of all the serranochromine species, and those many Malawi 

 species that Eccles & Trewavas (1989: 21) have placed in their 

 Pharyngochromis - Chetia - Serranochromis group. 



Generic key and diagnoses 



A single row of scales between the posterior orbital margin 



and the vertical limb of the preoperculum A 



Two or more scale rows between the posterior orbital margin 

 and the vertical limb of the preoperculum B 



A. Last 2 or 3 pored scales in the upper lateral-line series 

 separated from the dorsal fin base by not less than two scales 



of approximately equal size A(i) Sargochromis 



Last 5 to 7 (rarely 4 or 8) pored scales in the upper lateral line 

 separated from the dorsal fin base by one large and one small 

 scale A(ii) Pharyngochromis 



