The goal of this project is to address the challenges of creating cost-effective and sustainable agricultural best management practice (BMP) implementation strategies by developing an optimization-based decision support system (SWAT-BMP-OPT) using artificial intelligence techniques. The SWAT-BMP-OPT will be able to comprehensively evaluate the effectiveness, cost, and cost-effectiveness of commonly used agricultural BMPs in reducing nonpoint source (NPS) nutrient loadings; and reliably and efficiently develop optimal BMP implementation strategies (optimal types, quantities, and spatial locations of BMPs) to minimize nutrient loadings at minimum cost.