Lemniscomys rosalia Thomas, 1904

Denys, Christiane, Lalis, Aude, Lecompte, Émilie, Cornette, Raphaël, Moulin, Sibyle, Makundi, Rhodes H., Machang, Robert S., Volobouev, Vitaly & Aniskine, Vladimir M., 2011, A faunal survey in Kingu Pira (south Tanzania), with new karyotypes of several small mammals and the description of a new Murid species (Mammalia, Rodentia), Zoosystema 33 (1), pp. 5-47 : 16-19

Lemniscomys rosalia Thomas, 1904

Lemniscomys griselda rosalia Thomas, 1904: 414 .

The single-striped mouse of the genus Lemniscomys found in fallow bush and degraded woodlands in KP displays only one dark median band on a yellow bright pelage which identifies it as a member of the species complex including L. zebra Heuglin, 1864 , L. griselda Thomas, 1904 and L. rosalia Thomas, 1904 . Van der Straeten (1980) has separated several species within this complex on a morphometric basis, and L. griselda is only known from Angola, Zambia, and DRC ( Wilson & Reeder 2005). Cytogenetic studies have shown that three species coexist in Tanzania: L. zebra , L. rosalia and L. striatus Linnaeus, 1758 ( Fadda et al. 2001; Castiglia et al. 2002a; Corti et al. 2005). On a cytogenetic basis, Castiglia et al. (2002a) proposed without justification to rename the South African representatives of L. rosalia into L. calidior Thomas & Wroughton, 1908 , a decision that was not validated by Wilson & Reeder (2005).

By their general size and morphology the KP specimens are closer to the other Tanzanian specimens of L. rosalia from Berega ( Tanzania) than to the South African specimens of Tembe Elephant Park, especially in the ear size and both populations may belong to distinct taxonomic units ( Table 8).

Karyological data provide a 2n = 54, FNa = 62 for the two specimens of KP ( Fig. 7 ) and these values do not correspond to L. rosalia from Tembe ( South Africa) (2n = 48, FNa = 62) described by Ducroz et al. (1999) or L. zebra ( Fadda et al. 2001; Castiglia et al. 2002a; Corti et al. 2005) from Tanzania (2n = 54, FNa=58). But this karyotype corresponds to L. rosalia from Tanzania described by Castiglia et al. (2002a), Corti et al. (2005) and by Fadda et al. (2001) in northern Tanzania ( Table 9).

Because the type locality of L. rosalia is Nguru Mts, Monda ( Tanzania), we agree with Castiglia et al. (2002a) that the South African specimens called L. rosalia may represent a different species,

19 20

XX

24 25 26

XX

but without a careful examination of the type specimens of the various synonyms of L. griselda or of L. roseveari Van der Straeten, 1980 from Zambia, we cannot here name the Kwazulu Natal karyotypic form described by Ducroz et al. (1999). We cannot follow the nomenclatural proposal of Castiglia et al. (2002a) and Corti et al. (2005) because the type specimen of L. calidior (Thomas & Wroughton, 1908) came from Mozambique and not from South Africa. Similarly, these authors propose to give the name L. zebra for a given Tanzanian karyotype without any morphological or geographical justification (the oldest described?). Lemniscomys zebra was described from the Bahr el Ghazal ( Sudan), which corresponds to a very different region and a distant one from Tanzania. Moreover, there are other available names from old descriptions of the L. griselda complex and pending their revision we will not propose synonymies for the South African and north Tanzanian forms.

Consequently, here we attribute the Selous specimen to L. rosalia s.s. due to its geographical proximity with the holotype, pending a further revision of this taxon.

Figures

FIG. 7. — Standard karyotype of KP Lemniscomys rosalia Thomas, 1904 specimen (MNHN-CG2007-1206).