Determining termite diversity in arid Namibian rangelands - A comparison of sampling methods

Three methods of sampling termite diversity in arid rangelands were tested in Namibia during the wet (March) and dry (October) seasons of 1998. Six sites were chosen: one pair on each of three farms representing a gradient of land-use intensity. At each site, two adjacent plots of 1 ha each were sampled: one plot by a modified standard transect system, with superficial ground-breaking, and the second plot by visual searching and/or baiting methods. All six sites were similar in climate, soil conditions and topography. A total of 11 termite genera was found, including at least 19 species. No sampling method recorded all taxa. The baiting method detected 69% of the taxa, and the soil-excavating transect and the visual search method 63% each. Some taxa were detected with one method only, and could be absent seasonally. All methods, therefore, were required to maximize a species inventory. More termite taxa were found on the commercially-managed than on the two communally-managed farms. More taxa were found at the perceived high-intensity land-use site than at the low-intensity land-use site on the commercially managed farm, the reverse being true on the two communally-managed farms.