Convergence and contingencies in savanna grasslands

Savanna grassland distribution, structure and function are a product of three interacting drivers - fire, grazing by large herbivores, and extreme climatic fluctuations. Although all savanna grasslands share these drivers, evidence suggests that fire and grazing affect ecosystem structure and function in fundamentally different ways in southern Africa (South Africa (SA), in particular) and North America (NA).

These differences have been attributed to the contingent factors of greater age, longer evolutionary history, lower soil fertility, and greater diversity of plants and large herbivores in SA. An alternative hypothesis is that differences in methods and approaches used to study these systems, and a strong NA bias in the number of studies of savanna grasslands, has led to differing perspectives on the role of these drivers. It is important to differentiate between these alternatives. If the impacts of shared and fundamental ecosystem drivers - fire and grazing regimes, which are being extensively altered by humans worldwide - truly differ with age and evolutionary history, this calls into question our basic understanding of savanna grasslands and our ability to forecast change and maintain ecosystem services. Thus, our goal is to quantify, in directly comparable ways, ecosystem and community responses to fire and grazing in savanna grasslands of SA and NA, and to identify those ecological processes that are similar (convergent), despite potential contingent factors of differing evolutionary history, herbivore diversity and soil fertility. To address this goal we are building on three ongoing long-term (20-50+ yr) manipulations of fire and grazing in both NA and SA. We are testing specific hypotheses related to responses of ecosystem processes (ANPP, N and C cycling) and plant community structure (richness, diversity and dominance) and dynamics (compositional change, species turnover, species associations) to manipulations of 1) fire regime, 2) grazing and fire/grazing interactions, and 3) megaherbivore diversity (1 vs. 14 species).