Habitat preferences and niche breadth/overlap
Habitat represents one of the main niche dimension, which has often been considered in community ecology studies (Morris, 1996). Nevertheless, most studies conducted to date concerned communities in outdoor environments, where habitats may differ according to several factors like vegetation, soil, elevation, among others. Here, room type and rooms characteristics were chosen as easy-to-describe proxies of habitat / microhabitat structure that may be relevant for commensal small mammals. Indeed, the variables recorded here enable us to distinguish between categories of domestic spaces, in terms of “privacy” (from bedrooms to outdoor spaces or shops), type of activities hosted, hiding places and food resources present. Also, the nature of construction materials used for rooms can help to distinguish between traditional (use of clay for floor and walls, and of straw for ceiling) and more modern buildings (use of concrete for floor and walls, and of corrugated iron for ceiling), the latter being expected in villages that are more integrated in commercial networks and directly connected to large cities, thus more prone to the introduction of exotic rodents (Diagne et al. 2016, but see Lucaccioni et al. 2016).
Our analyses present M. musculus as more abundant in some room types (kitchens and stores), especially when built with non-traditional construction materials (cement and iron, particularly), similarly toM. natalensis in its area of occurrence. These habitat types contrast with those where A. niloticus is found more often than expected (granaries and outdoors), which is coherent with the ecology of the latter species, more abundant in grassy habitats and grain fields in outdoor environments (Granjon et al., 2013). Arvicanthis niloticus also shows both the larger niche breadth and the lower mean overlap with other species. These characteristics may represent attributes of the “occasional commensal” category of Hulme-Beaman et al. (2016), of which A. niloticus is probably the most extreme representative. Interestingly, P. daltoni , which is regularly found indoors in West Africa (Bryja et al., 2010), has the smallest niche breadth – being often under-represented in the room types sampled, and nearly only over-represented in bedrooms, mainly in the extreme East of the country. This species may suffer from the arrival of exotic invasive species, and be pushed back into the innermost rooms of the houses until it is excluded. The only non-rodent species, i.e. the shrew C. olivieri appears as very catholic in its habitat preference, being found in all room types in numbers close to those expected from their proportions in the overall sample. This also translates in a relatively high value of niche breadth, and also high niche overlap values with all rodent species. The wide range of habitats occupied and adaptability of this species have already been underlined (Churchfield & Hutterer, 2003). The habitat niche overlap with rodent species here observed probably relates to the fact that this shrew does not belong to the same ecological guild (sensu Simberloff & Dayan, 1991) and as such, is probably not submitted to competitive interactions with them likely to constrain its ecological distribution. Here, another type of interaction may rather be at work between shrews and rodents, namely a predator-prey relation: a preliminary metabarcoding study of the gut and faeces content of C. olivi eri individuals provides support for such a hypothesis, that would imply active predation, possibly mostly on neonates or non-active unweaned juveniles, directed primarily against M. musculus (Galan et al. 2023). Rattus rattus , which was very abundant in stock rooms where it probably causes important damages to food stuff (see Dossou et al., 2020 for an example in Cotonou, Benin), presents an average value of niche breadth compared to other species, and high overlap values with other species. Using telemetry in a rural area around Berega in Tanzania, Monadjem et al. (2011) found that within the houses or buildings they live in, black rats (also called roof rats) were located in the roof (37% of fixes), in the bedroom (35%), kitchen (14%) and in walls and windows (14%). Even if some R. rattus were caught in traps set on top of furniture items or wall tops, we were not able to quantify the three-dimensional activity of the species known to be at home in the upper parts of dwellings (Granjon & Duplantier, 2009; Monadjem et al., 2011). This vertical component of its spatial niche may however participate to the ecological distribution of the species and help its coexistence with the other ones. Its tolerance for quite traditional and rural conditions also makes it a good candidate for long-term persistence in relatively marginal areas, even in the absence of intense and regular road traffic (Lucaccioni et al. 2016)
As precised by Colwell & Futuyama (1971), such raw measures of actual niche breadth and overlap cannot per se give conclusive answers on the potential competition between coexisting species from a community. However, they can help formulate hypotheses to be tested via experimental procedures. In between, co-occurrence analyses may also help go further in the understanding of actual interspecific relationships at various spatial scales.