RESUMEN
The savanna ecosystem is dominated by grasses, which are a key food source for many species of grazing animals. This relationship creates a diverse mosaic of habitats and contributes to the high grass species richness of savannas. However, how grazing interacts with environmental conditions in determining grass species richness and abundance in savannas is still insufficiently understood. In the Kruger National Park, South Africa, we recorded grass species and estimated their covers in 60 plots 50 × 50 m in size, accounting for varying proximity to water and different bedrocks. To achieve this, we located plots (i) near perennial rivers, near seasonal rivers, and on crests that are distant from all water sources and (ii) on nutrient-rich basaltic and nutrient-poor granitic bedrock. The presence and abundance of large herbivores were recorded by 60 camera traps located in the same plots. Grass cover was higher at crests and seasonal rivers than at perennial rivers and on basalts than on granites. The relationship between grass species richness and herbivore abundance or species richness was positive at crests, while that between grass species richness and herbivore species richness was negative at seasonal rivers. We found no support for controlling the dominance of grasses by herbivores in crests, but herbivore-induced microsite heterogeneity may account for high grass species richness there. In contrast, the decrease in grass species richness with herbivore species richness at seasonal rivers indicates that the strong grazing pressure over-rides the resistance of some species to grazing and trampling. We suggest that the relationships between grasses and herbivores may work in both directions, but the relationship is habitat-dependent, so that in less productive environments, the effect of herbivores on vegetation prevails, while in more productive environments along rivers the effect of vegetation and water supply on herbivores is more important.
RESUMEN
To identify factors that drive plant species richness in South-African savanna and explore their relative importance, we sampled plant communities across habitats differing in water availability, disturbance, and bedrock, using the Kruger National Park as a model system. We made plant inventories in 60 plots of 50 × 50 m, located in three distinct habitats: (i) at perennial rivers, (ii) at seasonal rivers with water available only during the rainy season, and (iii) on crests, at least ~ 5 km away from any water source. We predicted that large herbivores would utilise seasonal rivers' habitats less intensely than those along perennial rivers where water is available throughout the year, including dry periods. Plots on granite harboured more herbaceous and shrub species than plots on basalt. The dry crests were poorer in herb species than both seasonal and perennial rivers. Seasonal rivers harboured the highest numbers of shrub species, in accordance with the prediction of the highest species richness at relatively low levels of disturbance and low stress from the lack of water. The crests, exposed to relatively low pressure from grazing but stressed by the lack of water, are important from the conservation perspective because they harbour typical, sometimes rare savanna species, and so are seasonal rivers whose shrub richness is stimulated and maintained by the combination of moderate disturbance imposed by herbivores and position in the middle of the water availability gradient. To capture the complexity of determinants of species richness in KNP, we complemented the analysis of the above local factors by exploring large-scale factors related to climate, vegetation productivity, the character of dominant vegetation, and landscape features. The strongest factor was temperature; areas with the highest temperatures reveal lower species richness. Our results also suggest that Colophospermum mopane, a dominant woody species in the north of KNP is not the ultimate cause of the lower plant diversity in this part of the park.