Effects of cattle trampling on vegetation, infiltration, and erosion in a tropical rangeland

Cattle trampling without forage consumption at stocking densities of 0.03–1.4 cows ha −1 was simulated on two dry-season rangelands in Kenya. Experiments under artificial rainfall documented the response of plant cover and production, infiltration, and erosion on a Luvisol and a Vertisol. Trampling reduced plant cover, biomass, and, at the highest rate, regeneration in the ensuing wet season. Infiltration was reduced on the Vertisol but not the Luvisol, although increases in runoff due to trampling were slight. Trampling increased soil loss partly by reducing vegetation cover but mainly by disrupting surface layers of sand on the Luvisol and of clay aggregates on the Vertisol. Soil loss normalized by runoff and rainfall energy declined in a sequence of erosive rainstorms as the sandy surface layer became re-established, but before vegetation recovered. Establishment of a sandy armor layer during runoff events and its disruption by dry-season trampling thus strongly affect soil-loss rates. Trampling limits plant recovery in the ensuing wet season only at intensities typical of settlement and watering centers. The experimental results, generalized with a spatial model of stock density, can be used to estimate the contribution of trampling to forage production and erosion as herding patterns change in response to sedenterization and water development. ► Cattle trampling has strong, nonlinear effects on ground cover and soil erosion. ► A mathematical model relates trampling intensity to stocking rate and herding pattern. ► Modeling relates landscape-scale effects to plot-scale measurements. ► Disruption of sandy surface layers affects erosion more than does cover reduction.