Farm-level optimization of BMP placement for cost-effective pollution reduction

With best management practices (BMPs) being used increasingly to control agricultural pollutant losses to surface waters, establishing the environmental effectiveness of these practices has become important. Additionally, cost implications of establishing and maintaining environmentally effective BMPs are often a crucial factor in selecting and adopting BMPs. This article considers both water quality and economic concerns and presents a methodology developed for determining cost-effective farm- or watershed-level scenarios through optimization. This optimization technique uniquely incorporates three existing tools: a genetic algorithm (GA), a watershed-level nonpoint-source model (Soil and Water Assessment Tool, SWAT), and a BMP tool. The GA combines initial pollutant loadings from SWAT with literature-based pollution reduction efficiencies from the BMP tool and with BMP costs to determine cost-effective watershed scenarios. The methodology was successfully applied to a 300 ha farm within the Cannonsville Reservoir watershed, a phosphorus (P) restricted reservoir within New York City's water supply system. An average reduction in dissolved P of 60% over the lifetime of the BMPs was set as the pollutant target. A baseline scenario was established to represent practices on the farm before BMP implementation. The most cost-effective scenario for the farm, under the presented methodology, achieved a cost-effectiveness of 0.6 kg dissolved P reduction per dollar spent per year. Additionally, the methodology determined alternative scenarios for the farm, which met the pollution reduction criterion cost-effectively. The methodology, as developed, is extendable to multi-farm or watershed-level evaluations.