Diets of savanna ungulates from stable carbon isotope composition of faeces

Hypotheses to explain diversity among African ungulates focus largely on niche separation along a browser/grazer continuum. However, a number of studies advocate that the browser/grazer distinction insufficiently describes the full extent of dietary variation that occurs within and between taxa. Disparate classification schemes exist because of a lack of uniform and reliable data for many taxa, and failure to incorporate spatio-temporal variations into broader assessments of diet. In this study, we tested predictions for diet and dietary niche separation of African savanna ungulates using stable carbon isotope evidence from faeces for proportions of C3 (browse) to C4 (grass) intake among 19 species from the Kruger National Park, South Africa. Dietary predictions from the literature are confirmed in the case of browsers (black rhinoceros Diceros bicornis, giraffe Giraffa camelopardalis, bushbuck Tragelaphus scriptus, kudu Tragelaphus strepsiceros), mixed-feeders (impala Aepyceros melampus, nyala Tragelaphus angasii), and most grazers (white rhinoceros Ceratotherium simum, Burchell’s zebra Equus burchellii, warthog Phacochoerus africanus, hippopotamus Hippopotamus amphibius, blue wildebeest Connochaetes taurinus, tsessebe Damaliscus lunatus, waterbuck Kobus ellipsiprymnus). In contrast, several species showed results differing from most expectations derived from the available literature, including eland Taurotragus oryx, steenbok Raphicerus campestris, grey duiker Sylvicapra grimmia, buffalo Syncerus caffer, roan antelope Hippotragus equinus and sable antelope Hippotragus niger. Many of these discrepancies can be accounted for by seasonal and/or regional dietary differences. Cluster analysis based on a data matrix that incorporates the extent of spatio-temporal dietary variation among Kruger Park ungulates reveals several distinct categories of feeding preferences that extend beyond a two-edged browser/grazer dichotomy, such as mixed-feeders with a preference for either forage class, and spatial/seasonal shifts between uniform and mixed-feeding styles among variable browsers (e.g. grey duiker) and variable grazers (e.g. buffalo). These results highlight the need for approaches that are sensitive to spatio-temporal variations and the continuity of diet.