Arvicanthis niloticus (E. Geoffroy 1803)

Wilson, Don E. & Reeder, DeeAnn, 2005, Order Rodentia - Family Muridae, Mammal Species of the World: a Taxonomic and Geographic Reference (3 rd Edition), Volume 2, Baltimore: The Johns Hopkins University Press, pp. 1189-1531 : 1289

Arvicanthis niloticus (E. Geoffroy 1803)

[Lemmus] niloticus É. Geoffroy Saint-Hilaire 1803 , Cat. Mamm. Mus . Natl. Hist. Nat., Paris: 186.

Type Locality: Egypt.

Vernacular Names: African Arvicanthis.

Synonyms: Arvicanthis centralis Dollman 1911 ; Arvicanthis centrosus Hollister 1916 ; Arvicanthis dembeensis (Rüppell 1842) ; Arvicanthis discolor (Wagner 1842) ; Arvicanthis jebelae Heller 1911 ; Arvicanthis kordofanensis Wettstein 1916 ; Arvicanthis luctuosus Dollman 1911 ; Arvicanthis major (Sundevall 1843) ; Arvicanthis mearnsi Frick 1914 ; Arvicanthis minor (Sundevall 1843) ; Arvicanthis muansae Matschie 1911 ; Arvicanthis naso Pocock 1934 ; Arvicanthis nubilans Wroughton 1909 ; Arvicanthis ochropus (Heuglin 1877) ; Arvicanthis pelliceus Thomas 1928 ; Arvicanthis raffertyi Frick 1914 ; Arvicanthis reichardi ( Noack 1887) ; Arvicanthis rhodesiae St. Leger 1932 ; Arvicanthis rossii de Beaux 1925 ; Arvicanthis rubescens Wroughton 1909 ; Arvicanthis solatus Thomas 1925 ; Arvicanthis tenebrosus Kershaw 1923 ; Arvicanthis testicularis (Sundevall 1843) ; Arvicanthis variegatus (Lichtenstein 1823) ; Arvicanthis zaphiri Dollman 1911 .

Distribution: Grassland and bush of Sahelian and northern Sudanian savannas, steppe, and semidesert in subsaharan N Africa, primarily in anthropogenic habitats, from Senegal (north of Gambia River), S Mauritania eastward through Mali, Burkina Faso, C and S Niger ( Dobigny et al., 2002 a), and C and S Chad to Sudan and western half and EC portion of Ethiopia (Bâ et al., 2001; Corti et al., 1996 b, as dembeensis ; Dobigny et al., 2001 a, b; Granjon et al., 1992, 2002 b; Osborn and Hemy, 1980;); north along the Nile Valley through Egypt; south through NE Dem. Rep. Congo, Uganda, S Burundi, and W Tanzania into E Zamibia, where the population is isolated from the nearest one in SW Tanzania ( Ansell, 1978). There is a record from N Malawi (1 skull only, living animals never found; Ansell and Dowsett, 1988), and the species also occurs in SW Arabian Peninsula ( SW Yemen and W Oman; Al-Jumaily, 1998; Snowden et al., 2000; and references cited therein). The only truly Saharan record is from SE Algeria (Hoggar), and Dobigny et al. (2001 a:216) speculated that A. niloticus "may persist across the whole Sahara, but only in some massifs or wadies where ecological conditions are more of the subdesert and/or steppe types than purely arid. Gardens also seem to be a good refuge for this species, and its commensalism may also explain its maintaining across the Sahara, via relictual populations." In Niger, the species was frequently encountered "in most wild bushy areas, gardens, villages and towns" ( Dobigny et al., 2002 b:498).

Conservation: IUCN – Lower Risk (lc).

Discussion: 2n = 62, FNa = 62 or 64 ( Corti et al., 1996 b, as dembeensis ; Volobouev et al., 2002 a). This is the species identified by the chromosomal cytotype ANI-1 ( Volobouev et al., 1988 a; see review in Ducroz et al., 1997) with two different forms identifying a different FN: ANI-1a (FNa = 62) in samples east of Mali and Burkina Faso to Ethiopia and Egypt, including the type locality; and ANI-1b (FNa = 64), found west of there ( Volobouev et al., 2002 a). Specimens with FNa = 66 were identified as dembeensis by Orlov et al. (1992 a) and others (see references in Baskevich and Lavrenchenko, 2000). Closest phylogenetic relationships of A. niloticus appears to be with the Ethiopian A. abyssinicus , judged by analysis of chromosomal, allozymic, and morphometric data along with mtDNA cytochrome b sequences ( Capanna et al., 1996 b; Capula et al., 1997; Corti et al., 1996 b; Ducroz et al., 1998; Fadda and Corti, 2001; niloticus was reported as dembeensis in most of these reports).

In 1993, Musser and Carleton reported dissatisfaction with their definition of A. niloticus (which included A. ansorgei and A. rufinus ), noting appreciable intra- and interpopulational variation in body size, pelage coloration, and cytological and biochemical traits, and speculated that more than one species may be present in the complex, as also suggested by the morphometric ( Rousseau, 1983), chromosomal (Viegas-Péquignot et al., 1983; Volobouev et al., 1987) and electrophoretic ( Kaminski et al., 1984) analyses available at the time. Subsequent analyses of a combination of data sets has been used to separate A. ansorgei and A. rufinus from A. niloticus (see those accounts) and identify a third yet unnamed species from the Central African Republic (Ducroz, 1998; Ducroz et al., 1997, 1998; Fadda and Corti, 2001; Volobouev et al., 2002 a). Furthermore, the geometric morphometric analysis by Fadda and Corti (2001) has questioned the identity of samples from Uganda and E Tanzania that we consider to be A. niloticus so our definition of A. niloticus may still be paraphyletic; additional study of samples from Uganda and E Tanzania (which should include the NE Zambian population) employing data from morphology, chromosomes, and gene sequences is certainly warranted.

Our inclusion of two synonyms require amplification. One is dembeensis (type locality, Deraske, Lake Tana, N Ethiopia), which has been recognized as a distinct species occurring in Ethiopia and nearby regions ( Demeter, 1983; Yalden et al., 1976, 1996;) and treated as a distinct species in recent results of chromosomal, molecular, and morphometric analyses (Baskevich and Lavrenchenko, 1994, 2000; Bekele et al., 1993; Bulatova et al., 2002; Capanna et al., 1996 b; Capula et al., 1997; Corti and Fadda, 1996; Corti et al., 1996 b; Orlov et al., 1992 a). Thomas (1910 b) had once associated dembeensis with Desmomys , noting that it had alredy been referred to Mus , Arvicanthis , Golunda , Pelomys , and Oenomys . Later, however, he ( Thomas, 1928 a) referred dembeensis to Arvicanthis , a view confirmed by Osgood (1936) and reinforced by Dieterlen (1974). The name lacernatus was once used in place of dembeensis for the Ethiopian samples of Arvicanthis (G. M. Allen, 1939; Corbet and Yalden, 1972; Osgood, 1936) but cranium of the holotype is an example of Meriones ( Yalden et al., 1976) . Arvicanthis dembeensis was thought to be a separate species because Ethiopian samples were originally contrasted by Corbet and Yalden (1972, under the name lacernatus) with abyssinicus , which they treated as a subspecies of A. niloticus ; however, abyssinicus is a distinctive species (see that account). From our study of specimens, we observed that the diagnostic features attributed to dembeensis also described the morphology in samples of A. niloticus from Egypt and westward throughout the steppes, semidesert, and N Sudanian belt as outlined above in the distribution. For that reason, Musser and Carleton (1993) included dembeensis within the synonymy of A. niloticus for which we were accused of confusing its specific rank ( Baskevich and Lavrenchenko, 2000). We are not confused nor are Ducroz et al. (1999:114), who noted the small genetic divergence between their samples of dembeensis and Egyptian A. niloticus and the lack of chromosomal differentiation between the two ( Capanna et al., 1996 b; Corti et al., 1996 b; Volobouev et al., 1988 a), and thus declared "there is no convincing evidence for the recognition of A. dembeensis as a distinct species and suggest referring it to as A. niloticus until conflicting evidence is produced." J.-F. Ducroz (in litt., 2001) now has additional information from Sudanese samples "clearly showing that there is a continuity from the genetic or karyologic point of view from Lower Egypt to Ethiopia, thus strengthening the point that there is no reason to distinguish dembeensis ."

The other name is testicularis (type locality, Bahr-el-Abiad, N Sudan). Samples identified as A. testicularis were thought to occur together with samples of A. niloticus at several places in Uganda ( Delany, 1975; Dollman, 1911; Hollister, 1919) and at Khartoum in Sudan ( Fadda and Corti, 1998), but after study of large series from Uganda and Sudan we could not distinguish two kinds at any of the localities ( Musser and Carleton, 1993) and found no morphological or distributional reasons why testicularis should not be included in A. niloticus . In a study of geographic variation among samples of Arvicanthis along the Nile Valley from Cairo to S Sudan using geometric morphometric techniques, Fadda and Corti (1998) recognized A. niloticus in Egypt and in Sudan at the Blue Nile and Khartoum, and A. testicularis at Khartoum and elsewhere in Sudan, but "museum labels were used for identification" (p. 105). They identified a "general trend in size decrease from North to South" (p. 108), and the suggestion of "an opposite clinal trend in shape" (p. 109), and claimed their scatterplot of first two principal compoments to demonstrate two clear clusters, which we could not appreciate and are only evident if the specimens were preidentified from their skin labels as to either niloticus or testicularis . If there are compelling data indicating specific integrity for testicularis they have not been published. J.-F. Ducroz (in litt., 2001) informed us of his "additional data from Sudan clearly showing... a continuity from the genetic or karylogic point of view from Lower Egypt to Ethiopia, thus strengthening the point that there is no reason to distinguish... testicularis as a distinct species." Philippi (1994) has demonstrated normal fertility of F1 and "backcross" individuals between samples of A. niloticus from Egypt and Sudan.

Corbet (1984) questioned including the S Arabian naso with A. niloticus , but current morphometric analysis supports that identity (Fadda and Corti, 2001). Verifiable sympatry with other species of Arvicanthis has been recorded: with A. ansorgei along the southern course and in most of the Inner delta of Niger River in Mali; and with A. neumanni in the Rift Valley of Ethiopia ( niloticus was recorded as either dembeensis or lacernatus; Corbet and Yalden, 1972; Demeter, 1983) and the Mawele region (Mwanasomano's, 30 mi [48 km] S Tabora) of C Tanzania (specimens in MCZ). Allopatry or parapatry, however, is usual. In Ethiopia, A. niloticus occurs up to about 2000 ft (610 m) and is replaced at higher elevations by A. abyssinicus and A. blicki ( Rupp, 1980; Yalden et al., 1976); in Kenya and Tanzania, A. niloticus ranges west of the E Rift Valley and is allopatric with A. nairobae , found in and east of the Rift as we have defined the species .