Genus Dephomys Thomas, 1926

(Figs 7; 8; Tables 7; 8)

REMARKS

The genus Dephomys was created by Thomas (1926) for Mus defua Miller, 1900 from a specimen collected from Mount Coffee, Liberia. A subspecies named D. d. eburneae Heim de Balsac & Bellier, 1967 from Lamto, Côte d’Ivoire provided different chromosomal diploid numbers between northern and southern Côte d’Ivoire (Tranier & Dosso 1979). Then D. d. eburneae was further elevated to species rank by Van der Straeten (1984) based upon morphometric and chromosomal characters; but no diagnosis was ever published. Based on these data, it would appear that D. defua occurs in the Guineo-Liberian forest block, while D. eburneae occurs in the forests of Côte d’Ivoire and Ghana to the east (Heim de Balsac & Aellen 1965; Rosevear 1969; Musser & Carleton 2005). Dephomys under the name D. defua was recorded by Heim de Balsac & Lamotte (1958), Gautun et al. (1986), Coe (1975) and Misonne & Verschuren (1976) on both Guinean and Liberian sides of Mount Nimba.

We found several karyotyped specimens in the MNHN collections displaying two chromosomal formulae (2N> 50 attributed to D. defua and 2N <50 attributed to D. eburneae according to Tranier & Dosso (1979) and originating both from Grabo, Adiopodoumé in Côte d’Ivoire. This could indicate that they may be sympatric in part of their distributional range and therefore both may be present at Mount Nimba. Tranier & Dosso (1979, unpublished data) provided chromosomal data and showed a great pelage color variability with age between the two taxa. Also, the dental character of the absence or presence of a cusp t3 on the upper M1 was variable, depending on the wear stage. The latter character was not so discriminant than previously reported (Tranier pers. comm.). Tranier & Dosso (1979) report a relative absence of morphological characters to distinguish between the two chromosomal species but found the width of the upper M1/, the height of the rostrum and the chromosome formula to be effective distinguishing criteria.

We performed molecular and morphometric analyses to verify the possible coexistence of two Dephomys species on Mount Nimba and in West Africa. The Neighbour Joining molecular tree indicates the existence of different lineages in West Africa (Fig. 7). The newly collected specimens 2021-1694 (SER167) +NIMII24 are in the same lineage (named Clade A below) and very distinct from the new specimen DM13172 of the Liberian side of the Mount (Clade B). Moreover, the SER 167 specimen harbours a 2N = 42 karyotype (Appendix 4). The K2P genetic distance between clades A and B is of 12% and they are sympatric in Côte d’Ivoire and Guinea suggesting the presence of two distinct species. We could not recover the holotype and paratype of D. eburneae, which was described from Lamto and Adiopodoume and should be deposited in the MNHN.

The DFA performed on skull measurements of genetically sequenced specimens from Côte d’Ivoire, Guinea and Liberia provides some discrimination between the two clades (Fig. 7). The karyotyped specimens are separated by a size axis (nearly all variables are positively correlated with axis 1), the 2N<50 specimens having on average larger skull measurements. Karyotyped specimens (1979-144 & 1979- 145 specimens) from Gabro (Côte d’Ivoire) with 2N= 52-53 are well classified in clade B close to molecularly typed specimens of Ziama (Guinea), Taï (RCI), Zogota (Liberia) (MNHN2011-346, P2357 and MNHN2015-1257 R24148). Specimens with 2N= 42-44 from Adiopodoume (Côte d’Ivoire, 1967-1560 and 1982-565) are close to molecularly typed specimens of clade A from Ziama (MNHN2011-344, P1539) and to the new specimen SER 167 (MNHN-ZM-2021-1694). All specimens used in the FFA were correctly classified (100%) and the axis 1 represents a size axis with specimens of Clade A slightly larger on average (Fig. 8).

The canonical analysis performed on genetically typed specimen on four external measurements provided a lower level of discrimination but the average values of HF length was the most discriminant variable (not figured). Clade A on average had a HF length of 27 mm against clade B with 25 mm. Clade A specimens are larger.

Consequently, here we can be confident that two taxa of Dephomys coexist on Mount Nimba as well as in some other West African localities, but naming the non-genotyped specimens remains difficult pending further studies and a larger sample of genetically typed specimens. Following Van der Straeten (1984) the Taï (Côte d’Ivoire) specimens belong clearly to D. defua, specimen from Taï in our molecular tree are found in clade B, while specimens from Lamto and Adiopodoumé (Côte d’Ivoire) belong to D. eburneae (Clade A). We provisionally attribute clade A to Dephomys cf. eburneae and clade B to Dephomys cf. defua pending on a further revision of the whole genus.