In many landscapes, habitat edges have critical impacts on biodiversity conservation and the provision of ecosystem services. Edge effects may be altered by the presence or proximity of other nearby edges. This phenomenon – edge interaction –is increasingly recognized as a driver of landscape pattern and process. As landscapes become more fragmented or patchy, edges become more dense, and edge interactions become more important. I investigated the prevalence and development of edge interactions in an East African savanna. For centuries in this landscape, abandoned cattle corrals have developed into treeless, nutrient-rich, preferentially grazed “glades” that persist for many decades. Glades represent a structurally dramatic and functionally critical source of heterogeneity with major impacts on plant and animal distributions. Using existing variation in inter-glade distance, I compared edge effects of less isolated glades (<150m from a second glade) to edge effects of more isolated glades (>250m from another glade). In January 2009, I began manipulating cattle corral density and spacing. Each of twelve 400x400m sites contained either one corral, two corrals that were 100m apart, or two corrals that were 200m apart. All corrals were abandoned after one month of intensive use.

Results/Conclusions

Edge effects around existing glades differed dramatically when a second glade was present within 150m. Compared to areas outside of isolated glades, areas between nearby glades had almost twice as many trees, half as much wildlife dung, and substantially different community composition patterns for non-tree plant species and Acacia-ant species. In many cases, edge interaction patterns did not indicate a convergence or merging of glade environments. From the corral manipulation experiment, initial surveys demonstrated that during active corral use, wildlife avoided sites with two corrals separated by 100m more than sites with two corrals separated by 200m. One year after corral abandonment, corral proximity continued to have broad-scale impacts on plant and wildlife communities. These findings suggest that in complex landscapes, edge effects may be very sensitive to patch configuration. As fragmentation and landscape complexity increase, an understanding of edge interaction will be critical for effective conservation and sustainable use.