Dissertation Research: Testing the effects of large mammalian herbivores on savanna dynamics and community structure with regional- and continent-scale natural experiments

Princeton University, Princeton NJ

Investigators

Abstract

The goal of this project is to determine how population declines of large mammalian herbivores (e.g., elephants, wildebeest, buffalo, antelopes, and zebra) affect African savanna habitats. Small-scale experimental studies have shown that because these large herbivores consume huge quantities of plant material, losing these species is likely to profoundly affect vegetation. The researchers will test how population declines and consequent changes in the frequency of two types of large herbivore-plant interactions alter savanna vegetation. In order to determine whether tree species reliant on large-mammal seed dispersal have declined in abundance, the research team will conduct fieldwork in Gorongosa National Park, Mozambique, where poaching during the 1977-1992 civil war nearly eliminated all large mammals. Second, by analyzing historical and modern satellite imagery, the researchers will test whether recent, human-driven declines in elephant populations across African protected areas lead to increased tree cover and possibly consequent threats to species adapted to historically sparse tree cover in African savannas. Elephants are particularly important savanna herbivores because they topple trees, potentially maintaining open savanna habitats. Many U.S. ecosystems have also experienced recent, dramatic changes in mammalian herbivore pressure as deer populations have exploded, and this research will advance our understanding of ecosystem responses to changed herbivory regimes. Project results will be incorporated into an undergraduate course on ecology and conservation. Results will be disseminated through scientific articles, blogs on park and affiliated websites, and online videos produced by the E.O. Wilson Biodiversity Foundation. Vegetation structure and composition is shifting in savannas across Africa due to altered fire regimes and climate change, but the extent to which declining herbivore populations are responsible for changing vegetation is unclear. Fieldwork in Gorongosa will include comparisons of vegetation in woodland patches that recruited prior to and since large mammal declines to test whether new woodland communities are relatively devoid of mammal-dispersed plants. This pattern is known from tropical forests where seed-dispersing mammals have declined, but the extent to which savanna trees depend on large mammals for dispersal is unknown. To test support for the hypothesis that recent declines in elephant populations lead to increased savanna tree cover, the researchers will build statistical models predicting changes in remotely sensed tree cover for African parks from published estimates of elephant density along with remotely sensed fire and rainfall data.