4.1 Diversity and spatial variations of rodents and shrews
The overall rodent and shrew species composition inhabiting the forest ecosystems of Mabira Central Forest Reserve is relatively diverse as reflected in the high number of determined species (20 rodents and 5 shrews). Previous studies have also recorded high species numbers. For example, Ssuuna et al. (2020) reported 14 rodent species, Kasozi (2017) reported 11 rodents and five shrews, Waswa et al. (2016) reported 13 rodents and seven shrews, while Dickson & Kityo (1996) reported 13 rodents and 2 shrews. This could be attributed to the fact that MCFR represents the only contiguous large mass of forested habitat acting as refugia for many forest dependent species (Kityo 2008, MWE 2017). The rarefaction curves for the four sampling areas in MCFR (Fig. 2) suggest that the techniques used were effective for sampling the rodent and shrew communities but more sampling effort is still required in forested areas because of the fewer shrew species collected compared to those in Waswa et al. (2016) who additionally reported aCrocidura novice. This could be because we did not employ pitfall traps, which are a more effective method of trapping shrews (Umetsu et al. 2006, Stromgren & Sullivan 2014).
The primary forest interior has some habitat specialists such asMalacomys longipes which was collected mostly along the river (Kerbis & Patterson 1995) and Deomys ferrugineus, while along the degraded forest edge only adaptive habitat generalist species such as P. jacksoni and L. stanleyi were recorded. L. stanleyi , although considered an Albertine rift endemic (Monadjem et al. 2015) it is believed to be widely occurring over arrange of habitats (Monadjem et al. 2015). Pardini et al. (2005), Lucie & Séverine (2016) reported that specialized species of rodents and shrews disappear following primary forest disturbance. The fact that more species have been recorded in adjacent fallows and along the degraded forest edge suggests that, many species are adapting or prefer to living in disturbed habitats Lucie & Séverine (2016). Any form of disturbance in a forest opens up space giving rise to various plants that were previously surpassed by low light (Catford et al. 2012) which in turn provide more food alternatives. This abundance of herbaceous plants provides cover and food for various species hence increase in diversity (Brockerhoff et al. 2017). However, the plight of specialized species is in balance with chances of being out-competed by the more aggressive species and ultimately eliminated from their habitat (Thomas et al. 2013) . The continuing illegal activities of resource extraction in MCFR (MWE 2016) continue to reduce the habitat quality for rodents and shrews in the reserve. The recording of new records for the forest is not in its self a testament of the quality of habitat but rather a fact that the forest is understudied. The continued decline in habitat quality in Mabira CFR and its surrounding, is reflected in the lack of records of grassland specialists like Mylomys dybowskyi and Lemniscomys zebra (as barbarus ) recorded in Davenport et al. (1996) but have not been reported in the preceding surveys. Habitat quality is compromised through loss of canopy cover, reduction in decaying logs since many are extracted as firewood, and reduction in plant species richness through logging which gives chance to the evasive paper mulberry to proliferate (Mulugo et al. 2019), this in time will lead to loss of rodent and shrew species forest specialists (Lucie & Séverine 2016, Pardini et al. 2005).