Midwest oak savannas are among the most imperiled North American plant communities, having declined >99.9% since European settlement.  The 480-km² Oak Openings region of northwestern Ohio features one of only a few landscape-scale oak savanna systems remaining in the Midwest.  Approximately 80% of the region’s natural land cover has been lost to urban and agricultural development while many of the remaining natural plant communities have been substantially modified by anthropogenic disturbance.  Despite these changes, the region’s remnant plant communities are highly diverse and continue to be influenced by natural disturbances such as seasonal variability in the water table, and prescribed fires set by land managers.  However, these plant communities have yet to be characterized in relation to the surrounding mixed-disturbance landscape.  We examined relationships between plant diversity (total, native, non-native), environmental variables (e.g., soil, topography) and landscape characteristics (e.g., patch metrics, management) within 15 randomly selected remnant Oak Openings prairies using 1000-m² modified-Whittaker, multi-scale plots.  Using readily available GIS / remote sensing data, we evaluated the influence of landscape heterogeneity, edaphic characteristics, topographic variables and site management history for each site at four nested spatial extents.  

Results/Conclusions

Collectively, we recorded 254 vascular plant species at these 15 sites, capturing 20% of the region’s known plant species.  The average number of species per 1 m² subplot ranged from 6.7 to 15.3, of which approximately 13.4%, on average, comprised non-native species.  Species richness increased from barrens (average 8.7 species/subplot) to dry prairies (10.9 species/subplot) to mesic prairies (13.8 species/subplot).  The relationship between plant diversity and characteristics of the surrounding landscape varied depending on the spatial extent being considered.  Site management history (i.e. the application of restoration treatments) was positively related to mean plant species richness at the sub-plot level, but was less predictive of species diversity at larger spatial extents.  The Ohio floristic quality assessment index (FQAI) was a better predictor of site management history at larger spatial extents.  By evaluating the associated environmental and landscape factors, we are able to highlight some of the important factors influencing species diversity in these ecosystems.  In addition, our study highlights the importance of considering scale and landscape context for conservation planning and restoration.