Estimation of grazing intensity along grazing gradients - the bias of nonlinearity

Distance to grazing hotspots like watering points or farms is regularly used as proxy for grazing intensity in extended rangelands. In many studies the resulting patterns between distance and grazing dependent variables are nonlinear with strong changes in the vicinity of the centre, a distinct transition zone, and hardly any changes on remote sites. Here we propose that this typical pattern just reflects the nonlinear relationship between grazing intensity and distance in circular grazing hotspots. Theoretical considerations as well as other proxies of grazing intensity like dung density or salinity suggest that reciprocal distance better represents grazing intensity. It is a more realistic description of the temporal distribution of livestock and often reveals linear relationships to grazing dependent variables preventing misinterpretations of thresholds that might be important for rangeland management.