Selection and placement of best management practices used to reduce water quality degradation in Lincoln Lake watershed

Abstract

An increased loss of agricultural nutrients is a growing concern for water quality in Arkansas. Several studies have shown that best management practices (BMPs) are effective in controlling water pollution. However, those affected with water quality issues need water management plans that take into consideration BMPs selection, placement, and affordability. This study used a nondominated sorting genetic algorithm (NSGA‐II). This multiobjective algorithm selects and locates BMPs that minimize nutrients pollution cost‐effectively by providing trade‐off curves (optimal fronts) between pollutant reduction and total net cost increase. The usefulness of this optimization framework was evaluated in the Lincoln Lake watershed. The final NSGA‐II optimization model generated a number of near‐optimal solutions by selecting from 35 BMPs (combinations of pasture management, buffer zones, and poultry litter application practices). Selection and placement of BMPs were analyzed under various cost solutions. The NSGA‐II provides multiple solutions that could fit the water management plan for the watershed. For instance, by implementing all the BMP combinations recommended in the lowest‐cost solution, total phosphorous (TP) could be reduced by at least 76% while increasing cost by less than 2% in the entire watershed. This value represents an increase in cost of $5.49 ha−1 when compared to the baseline. Implementing all the BMP combinations proposed with the medium‐ and the highest‐cost solutions could decrease TP drastically but will increase cost by $24,282 (7%) and $82,306 (25%), respectively.