Partitioning tree diversity patterns to prioritize conservation investments

The available tools and approaches to inform conservation decisions commonly assume detailed distribution data. We examine how well-established ecological concepts about patterns in local richness and community turnover can help overcome data limitations when planning future protected areas. To inform our analyses, we surveyed tree species in protected areas in the southern Appalachian Mountains in the eastern USA. We used the survey data to construct predictive models for alpha and beta diversity based on readily observed biophysical variables and combined them to create a heuristic that could predict among-site richness in trees (gamma diversity). The predictive models suggest that site elevation and latitude in this montane system explain much of the variation in alpha and beta diversity in tree species. We tested how well resulting protected areas would represent species if a conservation planner lacking detailed species inventories for candidate sites were to rely only on our alpha, beta and gamma diversity predictions. Our approach selected sites that, when aggregated, covered a large proportion of the overall species pool. The combined gamma diversity models performed even better when we also accounted for the cost of protecting sites. Our results demonstrate that classic community biogeography concepts remain highly relevant to conservation practice today.