Effect of Intensive Agriculture on Small Mammal Communities in and Adjacent to Conservation Areas in Swaziland
Hurst, Zachary Matthew
MetadataShow full item record
I examined the effect of sugarcane plantations on small mammal communities at 3 sites in the Lowveld of Swaziland during the dry and wet seasons of 2008. I evaluated changes in species abundance and community parameters in relation to distance to the interface, as well as the relationship between small mammal communities and environmental variables. I used pitfall arrays and Sherman live traps to sample small mammals along 9 traplines at the land-use interface and on a gradient extending 375 m into each land-use. I used point-centered-quarter, range pole, and line-transect sampling to characterize plant community structure. Two generalist small mammal species had increased abundance as distance into the sugarcane increased. Two species with wide geographic ranges appeared to select areas within 75 m of the interface. Four species with restricted habitat tolerances or diets were negatively affected by sugarcane, as was 1 species that selects for low ground cover. Two species may have avoided the interface. For the majority of species in the Lowveld, sugarcane does not provide habitat. Sugarcane monocultures > 375 m in width may form a barrier to movement of small mammal species. Species richness and diversity significantly decreased at the interface of 2 sites, however, 1 site had increased diversity associated with the interface. My analysis indicated a difference in community composition between the 2 land-uses and differences between the farthest interior conservation area (375 m)-interface (0 m) and the farthest interior sugarcane (375 m). There was no difference in community composition between seasons or distances within the conservation area. The farthest interior sugarcane trapline had distinctness from other traplines within the sugarcane, and may be of importance for minimizing the effects of habitat fragmentation in lowveld savanna. The effects of sugarcane did not extend into adjoining natural vegetation. My results indicated grass biomass, litter depth and shrub density played important roles in structuring the communities. Between sites, variation in community structure attributable to the sugarcane interface varied. The site with poorest vegetative cover had the highest relative importance of distance to the interface. One species (Steatomys pratensis) was negatively affected by distance to the interface.