Divergence in savanna grassland community responses to fire and grazing in North America and South Africa

Background/Question/Methods

Although fire and grazing impact savanna grasslands globally, these shared ecosystem drivers appear to influence savanna grassland plant communities and ecosystem structure and function differently in South Africa (SA) vs. North America (NA). In 2005, we initiated a comparative experiment to determine the degree of divergence (or convergence) in herbaceous plant community responses to fire, grazing, and their interaction in savanna grasslands in SA and NA. We erected replicate permanent herbivore exclosures (37 m² each) in grassland sites at the Kruger National Park, South Africa and the Konza Prairie Biological Station, KS that have experienced different fire return intervals (1-yr, 3-4 yr and unburned) for >25 yrs and that have either a single grazer (NA, bison) or diverse suite of herbivores (SA) present. This allowed us to assess in directly comparable ways the interactive effects of fire regime and grazing on herbaceous plant community structure, composition, and temporal dynamics. We hypothesized that although NA and SA savanna grasslands are structurally and functionally similar (i.e., dominated by C₄ grasses), they would diverge in their plant community responses to fire and grazing potentially as a result of differences in evolutionary history, age and diversity of plant species.

Results/Conclusions

We found plant diversity was highest with intermediate fire frequency and lowest with annual burning in NA, whereas plant diversity was highest with annual burning and lowest in unburned treatments in SA. Plant community compositional responses also differed dramatically between the two continents, with SA savanna grasslands experiencing greater turnover of grass species among the different fire regimes both with and without grazers. Additionally, trajectories of community change differed between NA and SA, with NA grasslands experiencing greater changes in plant community structure over time with removal of grazers, particularly in conjunction with frequent fire. Our results suggest divergence in community structure in response to fire and grazing in NA and SA. These results have important implications for our ability to generalize how savanna grasslands may respond to the shared drivers of fire and grazing at a global scale. New phylogenetic and trait-based studies will allow us to examine how biogeographic history and variation in traits may contribute to divergence in the herbaceous community responses observed.