Implications of assumption violation in density estimates of antelope from dung-heap counts:: a case study on grey duiker (Sylvicapra grimmia) in Zimbabwe

Dung-heap counts were used to estimate density of grey duiker (Sylvicapra grimmia Linnaeus 1758) in the Matobo National Park, Zimbabwe. To test assumptions of this method, defecation rate and defecation site selection were investigated under captive and field conditions, and densities were compared with independent estimates derived from territory mapping. Many assumptions were violated: males defecated more frequently than females with mean dry mass per deposit greater in females, but total daily faecal production was similar between sexes. Spatial distribution of faeces was clumped, and 52.8% of locations contained multiple deposits. Duikers exhibited habitat type preferences (i.e. low- to medium-density woodland) with herbaceous layer heights 40-100 cm and visibility > 20 m. Calculated grey duiker density from dung-heap counts in cleared plots was 9.7 +/- 1.3 animals km(-2), approximately double the territory-mapping estimate based on Minimum Convex Polygons (5.13 animals km(-2)) but similar to the 75% Fixed Kernel estimate (10.95 animals km(-2)). Provided that sex ratios approach parity and sampled area is representative of all utilized habitats, violation of basic assumptions of the dung-heap count method has a minor effect on density estimate accuracy.