Where and why do people hunt bats? Predicting the drivers and distribution of a conservation-relevant human behavior using spatially explicit latent variable models

Behavioral and ecological factors drive overexploitation of wildlife, suggesting that it is possible to integrate human behavioral models and ecological models to map and model the drivers of conservation behaviors. Current modeling techniques are siloed within either spatial ecology or social science, meaning conservation biology lacks an approach that has a clear conceptual framework, is broadly applicable, and is spatially predictive. Here, I use bat hunting in Sub-Saharan Africa as a case study to novelly combine structural equation modeling with spatially explicit data, to create spatial structural equation models that can map the relative likelihood of bat hunting. My dissertation unfolds in a stepwise manner, with each chapter building on the next. In Chapter II, I create a spatially explicit dataset of bat hunting, selling, and consumption. Chapter III I use these data to model bat hunting in two African hunting hotspots: West Africa and Madagascar, and in Chapter IV I build on the West Africa model by including human culture and expanding models to include other mammalian taxa. I found that West Africa and Madagascar have different drivers of bat hunting, though climate and human footprint were statistically significant in both regions, and that maps derived from these models could differentiate hunted from non-hunted areas. In West Africa, I found that all wild meat hunting is best predicted by climatic stability, with other variables playing smaller, more taxon-specific roles. I conclude it is possible to combine available spatial data, ecological modeling techniques, and structural equation modeling to model the drivers and map the distribution of a conservation-relevant human behavior.