Reestablishing native plant communities is an important focus of ecosystem restoration. In complex landscapes containing a diversity of ecosystem types, restoration requires a set of reference vegetation conditions for the ecosystems of concern, and a predictive model to relate plant community composition to physical variables. Restoration also requires an approach for prioritizing efforts, to facilitate allocation of limited institutional resources. Hierarchy theory provides a conceptual approach for predicting plant communities of disturbed ecosystems and, ultimately, for prioritizing restoration efforts. We demonstrate this approach using a landscape in southwestern Georgia, USA. Specifically, we used an existing hierarchical ecosystem classification, based on geomorphology, soil, and vegetation, to identify reference plant communities for each type of ecosystem in the landscape. We demonstrate that ecosystem identity is highly predictable using: only geomorphic and soil variables, because these upper hierarchical levels control the development of vegetation, a lower hierarchical level. We mapped the potential distribution of reference ecosystems in the landscape and used GIS (geographic information systems) to determine relative abundance of each ecosystem, as a measure of its historical rarity. We joined the reference ecosystem map with a current cover map to determine current abundance of each reference ecosystem, and percentage conversion to different disturbance classes. We show that over half of the landscape supports something other than reference plant communities, but degree of rarity varies widely among ecosystems. Finally, we present an index that integrates information on historical and current rarity of ecosystems, and disturbance levels of individual polygons, to prioritize restoration efforts. The premise of the index is that highest priority be given to restoring (1) currently rare ecosystems that were also historically rare and (2) the least disturbed examples of these ecosystems, as these will require the least effort to restore. We found that 80% of high-priority sites occur within just three (of 21) ecosystems. Moreover, the high-priority ecosystems all occur within stream valleys. Our approach provides managers with a straightforward methodology for determining potential distribution of reference ecosystems and for allocating efforts and resources for restoration in complex landscapes. Development of a priority index for a specific landscape requires an understanding of the hierarchical relationships among geomorphology, soil characteristics, and plant communities, in addition to well-defined restoration objectives.
Read full abstract