Evaluating the efficacy of targeting options for conservation practice adoption on watershed-scale phosphorus reductions.

Conservation identities of farmers in the Maumee River watershed, derived from farmer surveys, were embedded into a SWAT watershed model. This was done to improve the representation of the heterogeneity among farmers in the decision-making process related to the adoption of conservation practices. Modeled farm operations, created with near field-level Hydrologic Response Units (HRUs) within the SWAT model, were assigned a modeled primary operator. Modeled primary operators held unique conservation identities driven by their spatial location within the watershed. Five pathways of targeting the adoption of subsurface placement of phosphorus and buffer strips to HRUs within the watershed were assessed. Targeting pathways included targeting by HRU-level phosphorus losses, conservation identity of model operators, a hybrid approach combining HRU-level phosphorus losses and conservation identity of the model primary operator managing the HRU, and a proxy measure for random placement throughout the watershed. Targeting the placement of subsurface phosphorus application to all agricultural HRUs resulted in the greatest reduction in total phosphorus losses (32%) versus buffer strips (23%). For both conservation practices, targeting by HRU-level total phosphorus losses resulted in the most efficient rate of phosphorus reduction as measured by the ratio of phosphorus reduction to conservation practice adoption rates. The hybrid targeting approach closely resembled targeting by phosphorus losses, indicating near optimal results can be obtained even when constraining adoption by farmer characteristics. These results indicate that by developing management strategies based on a combination of field-level information and human-operator characteristics, a more efficient use of limited resources can be used while achieving near-maximal environmental benefits as compared to managing environmental outcomes solely based on field-level information.