Background/Question/Methods:   The integration of agricultural production systems with ecological systems is increasingly viewed as an essential step toward achieving global and regional conservation goals.  The Wisconsin Healthy Grown Potato® eco-label provides an economic incentive for growers to implement conservation and restoration plans on their non-crop lands. The goal of this research was to 1) assess the contribution of a range of upland plant communities for supporting regional biodiversity and conservation goals on participating farms, 2) to model how plant species composition, soil properties and light conditions change across the boundaries between potato fields and adjacent non-crop lands and 3) to provide practical land management recommendations to participating growers.  We sampled 38 habitat patches across six participating Healthy Grown® potato farms.   Sites were classified as afforested woodland, pine plantation or grassy field, the dominant vegetation classes on the farms.  At each site vegetation, soil and light variables were measured along transects spanning from 20 m into the potato field to 60 m into the adjacent habitat.   

Results/Conclusions:   Afforested woodlands, pine plantations and grassy fields differed significantly in their average species richness, native species richness, prairie-savanna indicator species richness and indices of floristic quality.  As indicated by all four variables, afforested woodlands had the highest conservation value and grassy fields the lowest.  The results of multi-response permutation procedures indicated that edge-effects had a detectable influence on herbaceous plant community composition to a distance of less than 10 m from the habitat edge in afforested woodlands and pine plantations.  There was no significant change in herbaceous composition from habitat edge to interior for weedy fields.  Relative dominance of native species and indices of floristic quality significantly increased from the habitat edge towards the interior for both afforested woodlands and pine plantations.  But the frequency of prairie-savanna indicator species, which were historically dominant in the landscape, was greater at the edges of woodlands than the interior.  Canopy cover and soil variables were significantly correlated with patterns of herbaceous species composition both between and within sites.  Our results suggest that plant conservation efforts on participating farms should prioritize afforested woodlands.  We suggest that gradually opening the canopy toward the interior of afforested woodlands might promote the inward recolonization of prairie and savanna species that are currently more frequent at habitat edges.  Any restoration plan should be adaptive, however, as there is a significant risk of promoting the spread of invasive species.