Checklist of small mammals of Mpigi Forest Reserve, Uganda

Study Extent

These mobilised data on small mammals was recorded from Mpigi forest reserve

Sampling

A combination of trap types was used, including Sherman, Longworth, pitfall and break back traps. The latter included locally made rat traps (18 x 8 cm), Museum Specials (14 x 7 cm), and large and small commercially available break back traps from the USA (18 x 8 cm and 10 x 4 cm respectively). Pitfall traps were used with the particular intention of collecting shrews. These traps consisted of a drift fence constructed from metal wire mesh (mesh size 5 mm x 5 mm) approximately 30 cm in height and 20 to 25 metres in length. The fence acted as a barrier to foraging animals, guiding them into sunken 5 litre plastic buckets (20 cm diameter x 25 cm depth) situated every five metres. The pitfall fences were situated in a variety of forest habitats concentrating on undisturbed forest and sites in close proximity to streams. In order to establish a degree of uniformity and enable comparison of data between sites, a standardised trapping regime was adopted at each site. A variety of bait types was used in order to catch the full range of rodent and shrew species, and traps were set in a range of different habitat types and locations, ten metres apart according to the following guidelines: • Fifty percent of traps were set at ground level in a variety of habitats, including disturbed and open areas. Two-thirds of these were baited with peanut butter and oatmeal or termites and one-third with rehydrated fish. • Forty percent of traps were set near streams or close to the water’s edge, where possible. Of these traps, half were baited with rehydrated fish and half with peanut butter and oatmeal, or peanut butter and termites. If there were no streams in the vicinity of the trapping area the traps were set as detailed above, in a variety of different habitat types. • Ten percent of traps were situated 1-6 metres above ground level in trees. Half of these traps were baited with mashed banana and the other half with peanut butter and oatmeal

Quality Control

In order to establish a degree of uniformity and enable comparison of data between sites, a standardised trapping regime was adopted at each site. A variety of bait types was used in order to catch the full range of rodent and shrew species, and traps were set in a range of different habitat types and locations, ten metres apart according to the following guidelines: • Fifty percent of traps were set at ground level in a variety of habitats, including disturbed and open areas. Two-thirds of these were baited with peanut butter and oatmeal or termites and one-third with rehydrated fish. • Forty percent of traps were set near streams or close to the water’s edge, where possible. Of these traps, half were baited with rehydrated fish and half with peanut butter and oatmeal, or peanut butter and termites. If there were no streams in the vicinity of the trapping area the traps were set as detailed above, in a variety of different habitat types. • Ten percent of traps were situated 1-6 metres above ground level in trees. Half of these traps were baited with mashed banana and the other half with peanut butter and oatmeal.

Method steps

1. The national survey of Uganda forests collected data in 65 forests. The original field survey results including all data were published in a series of reports by the Forest Department of Uganda (Howard et al. 1996). Compilation of species lists enabled comparison with other forests and provided a basis for highlighting species that are of particular biogeographical or conservation significance. All the small mammal species currently known from Uganda were classified according to their habitat requirements using Rosevear (1969), Kingdon (1971-74), Delany (1975), Hutterer (1987) and Skinner and Smithers (1990). This form of analysis assists in classifying forests and also in determining the importance of an individual forest to a defined group of species with a known habitat requirement. As this study is concerned mainly with forest habitats, the emphasis was placed particularly on forest habitat-type divisions, whilst non-forest habitats tended to be grouped under one classification as open habitat. The following ecological type categories are distinguished: • Forest-dependent species (F-species) are largely confined to closed-canopy forest and would be unlikely to tolerate any form of major large-scale habitat modification, although they may persist in secondary forest and isolated forest fragments. • Forest non-dependent species (f-species) are not restricted to closed-canopy forest and may occur in forest edge, gallery forest and dense savannah woodland.
   • Non-forest (open habitat) species are characteristic of open grasslands and semi-arid environments (O), aquatic or swamp habitats (A) or occur in a wide range of habitats (W). Species are further classified according to whether they occur only in highland areas above 1500 m (H) or lowland areas below 1500 m (L). The lack of knowledge concerning the ecological requirements of many of the small mammal species, particularly the shrews, makes it impossible to classify all species and nine shrews and four rodents remain of unknown ecological type (U). Using the daily trapping records, species accumulation rates were examined in order to estimate the extent to which further trapping would add to the species list, and facilitate comparisons of species richness between forests for any given level of sampling effort. The current dataset comprises records of small mammals recorded from Mpigi forest reserve during this survey. Electronic copies of the written reports and a separate electronic database of records were made available for this purpose by the original authors of the reports (Peter Howard).We extracted the mammal records from the electronic database and supplemented these with location data provided in the narrative reports. The location data in the original reports were provided in Military Grid Reference System. These were first converted to standard UTM coordinates, noting that the MGRS was using an old datum (MGRS-AL scheme also called "MGRS old"). UTM coordinates were then converted to decimal degrees using an online conversion spreadsheet (https://giscrack.com/download-excel-template-convert-geographic-coordinates-utm/). The converted locations were then checked against Google Maps imagery to confirm that a satisfactory conversion had been made. Species records from each Forest Reserve were checked against the distribution maps in Carswell et al. (2005) and through expert assessment to ensure that no unsubstantiated or extralimital records were included.

A total of twelve small mammals were identified to species level.

1. Crocidura fuscomurina (Heuglin, 1865)
   common name: Tiny Musk Shrew

The data were collected from Mpigi Forest Reserve