干旱区生态环境知识资源中心

This study employs scale-dependence as an analytical approach to understanding effects of livestock grazing on rangeland degradation in northern Kenya. It used extensive datasets previously collected from 13200 km(2) rangelands where grazing pressure gradients of livestock (varied from none, light, moderate, heavy and very heavy grazing) in conjunction with seasonality across different ecological scales influenced plant responses and probably contributed to land degradation. The data representing spatial and temporal scales were used to test the equilibrium and non-equilibrium-grazing models and to verify scales at which the models appropriately described range degradation. The equilibrium-grazing models operated at the coarse scales (e.g. range units, km(2)) and non-equilibrium-grazing models at multiple scales (e.g. spatial, temporal and fine scales - plots, landscape patches). The study showed that the equilibrium-grazing hypothesis, which stated that responses of plant species richness, cover and biomass varied along grazing pressure gradients at the coarse scale, was rejected, while the nonequilibrium-grazing hypothesis, which stated that the factors responded to temporal and spatial scales combined with grazing pressure gradients at the fine scale, was accepted. This study emphasized that in future discussions on shifts in the thinking of range science from equilibrium- to non-equilibrium-grazing models should clarify scales at which land degradation is assessed. In conclusion, the paper suggests that understanding plant species responses to grazing pressure and seasonality needs to consider multiple scale effects and that the dogmatic notions about degradation of the and zone rangelands at the coarse scales should be reconsidered. Land degradation assessments in the and zones should focus at the fine scale. Copyright (C) 2002 John Wiley Sons, Ltd.