A new species of Trematocranus, T. pachychilus sp. n., is described from Lake Malawi. So far, it has only been found at Jafua Bay, Mozambique. It can easily be distinguished from T. labifer by its molariform pharyngeal dentition. A morphometric study, including 24 measurements and 15 counts, was done to compare the new species with T. microstoma and T. placodon. Trematocranus pachychilus is characterised by its thick lips. This species further differs from T. microstoma by its bicuspid (vs. unicuspid) outer oral teeth, wide (vs. small) pharyngeal bone, and its head shape. It resembles T. placodon, from which it can be distinguished by its straight to concave head profile (vs. rounded), less-developed pharyngeal bones (vs. hypertrophied), and the presence of small to minute teeth on the lateral parts of the dentigerous area on the lower pharyngeal bone. A key to the species of Trematocranus is provided.

Cichlid, Lake Malawi, new species, thick lips, Trematocranus pachychilus sp. n.

Lake Malawi, like most other African Rift lakes, is home to a large number of endemic cichlids, that in each lake radiated into many genera and species in a very short period. Some sources estimate that there are more than 800 species of cichlids in Lake Malawi alone (e.g., Snoeks 2000). Because of the explosive radiation, many species and genera complexes still need to be taxonomically resolved.

In Lake Malawi, haplochromines constitute the major part of the cichlids. The endemic genus Trematocranus Trewavas, 1935, currently includes three valid species: T. microstoma Trewavas, 1935, T. labifer (Trewavas, 1935), and T. placodon (Regan, 1922) (Eccles and Trewavas 1989). These fish can grow to sizes of 250 mm TL (Maréchal 1991). The genus is characterised by a typical melanin pattern with large supra-pectoral and supra-anal spots situated on the upper lateral line, extending to the base of the dorsal fin and usually overlaid on fainter bars. There is also an opercular spot and one at the end of the caudal peduncle (Eccles and Trewavas 1989). The teeth are placed in 3 - 8 inner teeth series in the lower oral jaw and the outer teeth are long and recurved, either bicuspid or simple (Eccles and Trewavas 1989).

Trematocranus microstoma is the type species of the genus. It has simple, long, and slender teeth in the oral jaws. They are recurved and placed in 6 - 8 inner teeth series anteriorly (Eccles and Trewavas 1989). The pharyngeal teeth are considerably enlarged and molariform (Snoeks and Hanssens 2004). Trematocranus labifer has bicuspid, or simple anterior, oral teeth in three or four series. The teeth on the lower pharyngeal bone are all compressed, sharp, and not enlarged or molariform (Eccles and Trewavas 1989, Snoeks and Hanssens 2004). The third species, T. placodon, has bicuspid, or simple anterior, oral teeth in three or four inner teeth series. The lower pharyngeal bones are massively enlarged with large, flattened, and molariform teeth (Eccles and Trewavas 1989, Snoeks and Hanssens 2004). This adaptation is clearly related to a diet of snails (Mtethiwa et al. 2015).

Although Eccles and Trewavas (1989) and Snoeks and Hanssens (2004) limited the genus to these three large species, six more species have been placed in the genus at some point during their taxonomic history. Trematocranus auditor Trewavas, 1935, T. brevirostris Trewavas, 1935, T. jacobfreibergi Johnson, 1974, and T. trematocephala (Boulenger, 1901) are now assigned to the genus Aulonocara (Eccles and Trewavas 1989, Maréchal 1991). Eccles and Trewavas (1989) placed T. brevirostris in Aulonocara. Snoeks and Hanssens (2004) proposed to assign it to Aulonocara or Otopharynx, since its sensory pores are less developed than in most other Aulonocara species. The species is still mentioned as T. brevirostris in a popular account of Lake Malawi cichlids by Konings (2016). Trematocranus peterdaviesi Burgess and Axelrod, 1973, is now in the genus Alticorpus, and T. intermedius (Trewavas, 1935) in Tramitichromis (Eccles and Trewavas 1989).

Already in 2004, Snoeks and Hanssens mentioned the possibility of a fourth species of Trematocranus, then known as T. sp. ‘thicklip-bicuspid’, based on the oral dentition and the hypertrophied lips. These specimens were all collected at Jafua Bay, north of Cobue, and near the island of Likoma, Mozambique, in 1998. In this bay, T. placodon, which has a wider distribution, is also found (Snoeks and Hanssens 2004). So far, the new species of Trematocranus has not been found in another location.

Below, the species new to science as Trematocranus pachychilus sp. n. will be proposed.

In total, 30 specimens of Trematocranus of the collection at the RMCA (Royal Museum of Central Africa, Tervuren) were examined: seven T. pachychilus sp. n., 13 T. placodon, and ten T. microstoma. All specimens used in this study were collected by the taxonomy team of the SADC/GEF Lake Malawi/Nyasa/Niassa project during several expeditions on Lake Malawi in 1997 and 1998. They were all caught by bottom trawling using the RV Usipa.

The measurements and counts performed here follow Snoeks (2004). On the lower pharyngeal bone, one additional measurement was taken. The pharyngeal depth was measured from the anteriormost part of the dentigerous area to the indentation between the central area and the rostral keel on the ventral side of the bone. In total, 24 measurements, 16 counts (including the number of vertebrae via X-rays), and some observations on the body, head shape, and the colour pattern in alcohol were made. The abbreviations of these measurements and counts are given in Table 1.

Principal component analysis (PCA) was performed in R to explore the multivariate data set. Measurements were log-transformed and the correlation matrix used. When using log-transformed measurements, the individual loads of all variables on the first principal component (PC 1) are of the same magnitude and sign and PC 1 can therefore be regarded as a proxy for multivariate size (Jolicoeur 1963, Snoeks 2004, Van Steenberge et al. 2015). The correlation matrix was used for the raw meristic data. The number of vertebrae and anal and pelvic fin spines and rays were all without variation and thus were excluded from the analysis.

According to the revised classification by Eccles and Trewavas (1989), there were three valid species in the genus Trematocranus. Fish caught at Jafua Bay, Lake Malawi, by Snoeks and Hanssens in 1998 are found here to be a new species, Trematocranus pachychilus sp. n.

Trematocranus pachychilus is distinguished from all other congeners by its thickened and fleshy lips. Trematocranus labifer is known to have slender teeth on the pharyngeal jaws, while all other Trematocranus species have a molariform dentition (Eccles and Trewavas 1989, Snoeks and Hanssens 2004). Therefore, it was not included in the detailed morphometric analyses. In these analyses, it was shown that the new species differs in several characters from T. microstoma and T. placodon. T. pachychilus has thick lips and bicuspid oral teeth on the external row, while in T. microstoma lips are generally less thick and the outer oral jaw teeth are unicuspid. In addition, there are some differences in measurements, but there appears to be no clear separation between T. pachychilus and T. microstoma based on meristic data (Table 1, Fig. 2). They do differ in head shape. Trematocranus pachychilus can be distinguished from T. placodon by its thick lips and less hypertrophied pharyngeal bone, oral and pharyngeal dentition, and in a few measurements and meristics (Table 1, Figs 3–4).

While all specimens of T. pachychilus have hypertrophied lips, some variability in this characters in the two other species was observed. Two specimens of T. placodon were a bit aberrant. They clearly have the massive pharyngeal bones typical for T. placodon (Fig. 6C) and, in the analyses of the measurements and meristics, are associated with the other members of this species (Figs 3–4). However, one specimen (MRAC 99-041-P-4779) has both the upper and lower lips thickened, while the other specimen (MRAC 99-041-P-4780) only has a slightly thicker upper lip. Three T. microstoma specimens also show enlarged lips (MRAC 99-041-P-4789-4791) that are thicker than in their conspecifics. At present, we have difficulties in interpreting the variability in lip morphology in both species. Genetically determined intraspecific variability may be large in these species, but on the other hand, lips are also known to be morphologically plastic in cichlids. In an experimental study on Neotropical cichlids, both a genetic and a phenotypic plastic factor was found to be involved in the development of thick lips. (Machado-Schiaffino et al. 2014). Whether or not this is also the case in Lake Malawi and therefore explains the case of the two specimens of T. placodon and the three of T. microstoma is unclear. Alternatively, the hypothesis of hybridization cannot be excluded either. However, it should be stressed that except for lip thickness, the two specimens of T. placodon and the three of T. microstoma display the diagnostic characters typical for their respective species.

Pharyngeal jaw morphology has also been reported as phenotypically plastic. For instance, by feeding on hard food items, the pharyngeal jaws of cichlids have been reported to become larger with a more developed horn and keel and their teeth more numerous and stouter during ontogeny (Greenwood 1965, Hoogerhoud 1986, Gunther et al. 2013). However, in the case of Trematocranus, we see a clear distinction in pharyngeal jaw morphology and its dentition between the species, supported by other morphological features and therefore have no indication of such plasticity in the populations examined. In addition, significant differences in lower pharyngeal jaw morphology due to phenotypic plasticity have only been obtained by a highly differential feeding regime (soft versus hard items) in an experimental setup (Hoogerhoud 1986, Gunther et al. 2013), while intraspecific differences in nature have been found only in one African cichlid Astatoreochromis alluaudi in allopatric populations living in different environments and probably linked to the presence/absence of snails (Greenwood 1965). Such diverging (experimental and natural) conditions do not seem to correspond to the conditions in which T. pachychilus and T. placodon were found, both in trawls over soft bottoms far remote from rocks in order not to damage the net, and in the presence of snails (see below).

Trematocranus pachychilus and T. placodon are both found at Jafua Bay. Because of the obvious differences in morphology we expect their trophic niche to be different. Nevertheless, both appear to feed on molluscs. Trematocranus placodon has been reported to eat molluscs (Mtethiwa et al. 2015). Indeed, some fragments of bivalve shells and snails were found in the pharyngeal area of specimens of this species, while dissecting the pharyngeal bones. Shell fragments of snails were observed in the stomachs of some specimens of both T. placodon and T. pachychilus on X-rays. It would be very interesting to see if T. pachychilus also lives in other bays across the lake, and whether they live sympatrically in other areas with T. placodon. If more specimens and data become available, a more in-depth study could relate the morphological differences found with their trophic ecology.

Morphometric analyses confirm that T. pachychilus is a new species within the genus Trematocranus, next to the three valid species T. microstoma, T. labifer and T. placodon. Found at Jafua Bay of Lake Malawi, it co-occurs with the latter. It was compared with T. microstoma and T. placodon using a principal component analysis on measurements and counts. The most striking characteristics to separate the four species are the dimensions of the pharyngeal bone, the number of pharyngeal teeth rows, pharyngeal dentition, oral dentition, and lip thickness. More data on the distribution and morphological variability are needed to resolve some questions regarding the sympatry of T. placodon and T. pachychilus sp. n.

Trematocranus placodon: MRAC 99-041-P-4776-4778 (3), 125.1–153.8 mm SL, Jafua Bay, Lake Malawi, Mozambique, 12°07.13'S, 34°45.89'E, 4 April 1998; MRAC 99-041-P-4779-4780 (2), 161.1–163.1 mm SL, Jafua Bay, Lake Malawi, Mozambique, 12°07.13'S, 34°45.89'E, 4 April 1998; MRAC 99-041-P-4798-4806 (6 of 9), 116.0–149.3 mm SL, Senga Bay, Lake Malawi, Malawi, 13°46.73'S, 34°37.91'E, 21 September 1997; MRAC 6539-6540 (2), 138.1–162.0 mm SL, Senga Bay, Malawi, Lake Malawi, 13°46.73'S, 34°37.91'E, 21 September 1997.

Trematocranus microstoma: MRAC 99-041-P-4785-4786 (2), 166.7–182.7 mm SL, Kanda Bay, Lake Malawi, Malawi, 11°56.84'S, 34°08.28'E, 3 June 1997; MRAC 99-041-P-4787-4788 (2), 143.7–154.9 mm SL, Mazinzi Bay, Lake Malawi, Malawi, 14°09.60'S, 34°58.72'E, 11 October 1997; MRAC 99-041-P-4789-4791 (3), 135.4–158.2 mm SL, Mazinzi Bay, Lake Malawi, Malawi, 14°05.33'S, 35°01.94'E, 11 October 1997; MRAC 99-041-P-4792-4797 (3 of 6), 126.7–144.9 mm SL, Chembe Bay, Lake Malawi, Malawi, 14°01.57'S, 34°49.88'E, 27 May 1997.

We acknowledge the project members of the SADC/GEF taxonomy team and the crew of the RV Usipa for their logistical support; Jonathan Brecko and Nathan Vranken (RMCA) assisted with taking pictures. We would also like to thank Vincent Ceyssens for proofreading our article.