Many mammals form close associations with conspecifics (Lott, 1991). Groups vary in size (few individuals to large aggregations), reproductive skew (plural vs. single breeders), context (e.g. foraging vs. communally nesting groups) and duration (seasonal vs. permanent groups; Silk, Z-VAD-FMK 2007). Groups can comprise kin (e.g. brushy-tailed wood rats Neotoma cinerea; Moses & Millar, 1992) and non-kin (e.g. Leisler’s bat Nyctalus leisleri; Boston et al., 2012). Because relatives are often

more tolerant of one another than of non-relatives (Charnov & Finerty, 1980), kin typically form more cohesive, less competitive groups (Ebensperger, 2001), but kinship is not necessarily a predictor of reproductive success (Silk, 2007). Rodent social systems result from complex interactions between the species’ life history,

behaviour, phylogeny and ecology (Thorne, 1997). Group living evolves when the net benefits (e.g. reduced predation risk, social thermoregulation and group foraging) outweigh the costs (e.g. PD0325901 resource competition, disease and reproductive suppression; Silk, 2007). The adaptive function of group living in rodents is embodied in six important hypotheses, including (1) predatory risk, (2) social thermoregulation (i.e. huddling under low temperatures; Edelman & Koprowski, 2007) and (3) burrow sharing (i.e. burrows are limited or their construction reduces per capita energy expenditure of group members; Taraborelli, 2009). Three more hypotheses consider the distribution of resources. (4) The resource-defence hypothesis proposes that group living

is related to increased resource abundance (Slobodchikoff, 1984), whereas (5) the food competition hypothesis maintains that competition affects interactions between individuals (Gliwicz, 1981), such that group living is not favoured because of competition of patchily distributed resources (Ranta, Rita & Lindström, 1993). (6) The resource dispersion hypothesis proposes that spatio-temporal heterogeneity of resources favours group living by allowing individuals to utilize the same resources without communal foraging, thereby 上海皓元 minimizing competition and facilitating social tolerance (Carr & MacDonald, 1986). Functional explanations assume, but often do not test, the fitness consequences of group living and moreover do not consider social interactions between individuals (Silk, 2007). Because group living is the outcome of social relationships, testing the various hypotheses must consider spatio-temporal variation in conspecific interactions. Otherwise, functional explanations may become tenuous. Therefore, to test some of these hypotheses of group living, we investigated home-range size and social behaviour of the African ice rat Otomys sloggetti robertsi, a southern African endemic taxon. The ice rat is a diurnal, colonial, murid rodent, occupying the alpine–sub-alpine phytogeographic belts (exceeding 2000-m altitude) in the Drakensberg and Maluti mountains of southern Africa.