Background/Question/Methods Understanding factors that influence the relationship between ecosystem productivity and biodiversity may provide important insights into the conservation of ecosystems that capture carbon, as well as the preservation of species diversity. Productivity-diversity relationships can be positive, negative, and hump-shaped and depend on competition, resource availability, disturbance, traits of species, and spatial scale of analyses. Human land use and management may also influence these relationships because humans harvest productivity and manipulate species abundance (e.g., through silviculture). Additionally, the shape of productivity-diversity relationships may have conservation implications. For example, if the most potentially productive ecosystems (determined by climatic and edaphic factors) are the least diverse, then harvesting natural resources and converting land to agricultural may not detrimental influence regional species diversity. Alternatively, if productivity-diversity patterns are positive, then protecting highly productive areas may be needed to sustain species diversity. With these concepts in mind, I used publically available data to explore relationships between productivity and tree diversity and how these relationships varied with spatial scale, ecoregions, forest types, land use zones, and along gradients of productivity and diversity. Productivity estimates were derived from Moderate Resolution Imaging Spectroradiometer (MODIS), and tree richness was estimated from over 18,000 Forest Inventory and Analysis plots in a five state region of the southeastern U.S.

Results/Conclusions Across the five-state region, productivity-diversity relationships (PDRs) were either negative or hump-shaped depending on whether richness was measured at individual plots or the average across plots within ecoregions. PDRs within ecoregions were either positive or negative. Specifically, relationships were positive in the Appalachian Mountains but negative in forests of the Atlantic Coastal plain. PDRs varied by forest type and land use zones, but were typically influenced by ecoregions. Interestingly, PDRs tended to be negative in the most productive and least diverse ecoregions and positive in the least productive and most diverse ecoregions. While, rigorously determining mechanisms is difficult given the observational nature of this study, these patterns may be due to competition, water stress at extreme ends of moisture gradients, and possibly land use. For instance, local tree diversity appears to be strongly influenced by regional land use. In sum, these results suggest that adopting one conceptual model for how productivity and diversity are related may not be appropriate. Conservation planning for carbon sequestration and biodiversity should consider the potentially idiosyncratic regional ecology of species and forests and how land use influences both productivity and species diversity.