Reducing surface water pollution through the assessment of the cost-effectiveness of BMPs at different spatial scales.

Two kinds of agricultural Best Management Practices (BMPs) were examined with respect to cost-effectiveness (CE) in reducing sediment, nitrates-nitrogen (NO(3)-N) and total phosphorus (TP) losses to surface waters of the Arachtos catchment in Western Greece. The establishment of filter strips at the edge of fields and a non-structural measure, namely fertilization reduction in alfalfa, combined with contour farming and zero-tillage in corn and reduction of animal numbers in pastureland, were evaluated. The Soil and Water Assessment Tool (SWAT) model was used as the non-point-source (NPS) estimator, while a simple economic component was developed estimating BMP implementation cost as the mean annual expenses needed to undertake and operate the practice for a 5-year period. After each BMP implementation, the ratio of their CE in reducing pollution was calculated for each Hydrologic Response Unit (HRU) separately, for each agricultural land use type entirely and for the whole catchment. The results at the HRU scale are presented comprehensively on a map, demonstrating the spatial differentiation of CE ratios across the catchment that enhances the identification of locations where each BMP is most advisable for implementation. Based on the analysis, a catchment management solution of affordable total cost would include the expensive measure of filter strips in corn and only in a small number of pastureland fields, in combination with the profitable measure of reducing fertilization to alfalfa fields. When examined for its impact on river loads at the outlet, the latter measure led to a 20 tn or 8% annual decrease of TP from the baseline with savings of 15€/kg of pollutant reduction. Filter strips in corn fields reduced annual sediments by 66 Ktn or 5%, NO(3)-N by 71 tn or 9.5% and TP by 27 tn or 10%, with an additional cost of 3.1 €/tn, 3.3 €/kg and 8.1 €/kg of each pollutant respectively. The study concludes that considerable reductions of several pollutant types at the same time can be achieved, even at low total cost, by combining targeted BMP implementation strategies only in small parts of the catchment, also enabling policy makers to take local socio-economic constraints into consideration. The methodology and the results presented aim to facilitate decision making for a cost-effective management of diffuse pollution by enabling modelers and researchers to make rapid and reliable BMP cost estimations and thus being able to calculate their CE at the local level in order to identify the most suitable areas for their implementation.