Effectiveness of unfertilized buffer strips for reducing nitrogen loads from agricultural lowland to surface waters

Unfertilized buffer strips (BS) are widely accepted to reduce nitrogen (N) loads from agricultural land to surface water. However, the relative reduction of N load or concentration (BS effectiveness, BSE), varies with management and local conditions, especially hydrogeology. We present novel experimental evidence on BSE for 5-m-wide grass BS on intensively drained and managed plain agricultural lowland with varying hydrogeology. We selected characteristic sites for five major hydrogeological classes of the Netherlands and installed paired 5-m-wide unfertilized grass (BS) and reference (REF) treatments along the ditch. The REF was managed like the adjacent field, and BS was only harvested. Treatments were equipped with reservoirs in the ditch to collect and measure discharge and flow proportional N concentration for 3 or 4 yr. In addition, N concentration in upper groundwater was measured. We found a statistically significant BSE of 10% on the peat site. At the other sites, BSE for N was low and statistically insignificant. Low BSE was explained by denitrification between adjacent field and ditch, as well as by the site-specific hydrologic factors including low proportion of shallow groundwater flow, downward seepage, low residence time in the BS, and surface runoff away from the ditch. We emphasize that a REF treatment is needed to evaluate BSE in agriculture and recommend reservoirs if drainage patterns are unknown. Introduction of a 5-m-wide BS is ineffective for mitigating N loads from lowland agriculture to surface waters. We expect more from BS specifically designed to abate surface runoff.