Evaluating elephant, Loxodonta africana, space-use and elephant-linked vegetation change in Liwonde National Park, Malawi

Heterogeneity, the spatio-temporal variation of abiotic and biotic factors, is a key concept that underpins many ecological phenomena and promotes biodiversity. Ecosystem engineers, such as African savanna elephants (hereafter elephant), Loxodonta africana, are organisms capable of affecting heterogeneity through the creation or modification of habitats. Thus, their impacts can have important consequences for ecosystem biodiversity, both positive and negative. Caughley’s “elephant problem” cautions that confined or compressed, growing elephant populations will inevitably lead to a loss of biodiversity. However, a shift in our understanding of elephants suggests that not all elephant impacts lead to negative biodiversity consequences, as long as there is a heterogeneous spread of elephant impacts that allows for spatio-temporal refuges promoting the persistence of both impact-tolerant and impact-intolerant species. To date, little empirical evidence is available in support of managing elephants under this paradigm and few studies are available that infer the consequences of the distribution of elephant impacts on biodiversity. In addition, most studies use parametric statistics that do not account for scale, spatial autocorrelation, or non-stationarity, leading to a misrepresentation of the underlying processes and patterns of drivers of elephant space-use and the consequences of their impacts on biodiversity. Here, I evaluate spatio-temporal patterns and drivers of elephant space-use, and how the distribution of their impacts affects biodiversity through vegetation changes, using a multi-scaled spatial approach, in Liwonde National Park, Malawi. My study demonstrates that elephant space-use in Liwonde is heterogeneous, leading to spatio-temporal variation in the distribution of their impacts, even in a small, fenced reserve. The importance of the drivers of this heterogeneous space-use varied based on the scale of analysis, water was generally important at larger scales while vegetation quality (indexed by NDVI) ...