The Oak Openings Region of northwest Ohio is a unique and complex collection of ecosystems. The heterogeneity of the area is extreme, and permeates all scales. Oak savanna, one of the habitats within the mosaic, is vital for the federally endangered Karner Blue Butterfly (Lycaeides melissa samuelis).
We sought to shed light on the interactions between the savanna habitat, wild blue lupine (Lupinus perennis) and white-tailed deer (Odocoileus virginianus). To this end we measured browsing on oak seedlings and lupine leaves within 1m2 quadrats at six sites within three different natural areas. We compared trends in lupine and oak abundance and browsing pressure over space, time and at the site scale as well as the quadrat scale using logistic regression and a spatially explicit test for clustering. The results indicated that there were complex interactions at both scales that were not adequately captured by the variables measured. In the logistic regressions of quadrat scale browsing on both lupine and oaks, browsing on one predicted variation in browsing on the other. Interesting correlations were found at the site scale between management actions such as burning and mowing and proportion of the site classified as closed canopy. Examination at a single scale would have excluded part of the picture present in these natural areas. Results from the spatial clustering test suggest different spatial trends in distribution of lupine. Spatial statistics added much to our understanding of trends in the data that were invisible to other measures.
We also performed weekly road-based surveys within the largest natural area, Oak Openings Preserve Metropark, to estimate relative deer abundance. Deer abundance observations were analyzed using forward stepwise logistic regression. The results of our forward stepwise logistic regression model indicated that length of streams within a zone was an important variable for predicting overall deer abundance in a zone. This suggests that the deer may be moving throughout the park exploiting patchy resources.
Scale effects were evident in trends for browsing and distributions of lupine and oaks, and both were affected by the heterogeneous nature of the area. Both spatial and temporal effects were evident in deer utilization of the park. This study deepened our understanding of complex interactions between deer, oak, and lupine within the globally rare, early successional oak savanna. Our methods to study these interactions were easy to replicate, initiate, extend and monitor.