, 2010; Pitman, Swanepoel & Ramsay, 2012). While both techniques are effective, they are not exempt from bias. Faecal analysis has been found to overestimate prey biomass and underestimate the consumption of small species (Mills, 1992; Marker et al., 2003). Faecal analysis can also be influenced by collection and identification procedures that could result in inaccurate dietary estimates (Klare et al., 2011). Reconstructing carnivoran diets from kills found through GPS cluster investigations overestimates the consumption of large prey. This is because Venetoclax chemical structure large carnivores exhibit longer handling times at larger kills and researchers

often find it easier to locate larger kill sites in the field (Anderson & Lindzey, 2003; Martins et al., 2011; Tambling et al., 2012). The GPS cluster method can also fail to detect kills (often small prey) made by predators, especially if prey species are consumed quickly. For the GPS cluster method to be widely adopted, calibration against more traditionally accepted diet determination techniques like faecal

analysis is required. Additionally, the potential exists to enhance the GPS cluster method with faecal samples located at cluster sites. Combining both faecal and GPS-located kill datasets may offer a way of U0126 reconstructing carnivoran diets at very high resolution by incorporating undetected kills from either technique (e.g. as with lions Panthera leo in Kruger National Park, South Africa; Tambling et al., 2012). While GPS cluster

investigations have been used to locate leopard kills (Martins et al., 2011; Pitman et al., 2012), combining GPS-located leopard faecal data with leopard kill site data has not yet been attempted. The aim of this study was to compare estimates of leopard prey composition and biomass intake using (1) ‘GPS cluster analysis’ and ‘faecal analysis’ thereby assessing whether the GPS cluster method yields comparatively similar dietary estimates to 上海皓元医药股份有限公司 that of faecal analysis; (2) ‘GPS cluster analysis’ and ‘GPS cluster analysis supplemented with faecal samples’ found at cluster sites to evaluate whether dietary estimates generated by the GPS cluster method could be improved by the addition of GPS-located faecal samples (e.g. by incorporating undetected kills). Welgevonden Private Game Reserve (24°10′–24°25′S and 27°45′–27°56′E, hereafter ‘Welgevonden’) is situated on the Waterberg Plateau in Limpopo province, South Africa. The topography is characterized by undulating mountains and flat hilltop plateaux (altitude 1080–1672 m), dissected by deep valleys and ravines (Parker, 2004). The vegetation is classified as Waterberg Mountain Bushveld.