Mapping hydrological pathways of phosphorus transfer in apparently homogeneous landscapes using a high-resolution DEM

Agriculture is a significant contributor to the diffuse loading of phophorus (P) in fresh water systems. At the landscape level, source areas and targeted surface waters are connected through different hydrological pathways. One of these pathways, P removal through surface runoff, may increase in relative importance when storm events occur and in wet seasons. With Digital Elevation Models (DEMs) it is possible to identify dominant surface pathways in the landscape although resolution of available DEMs is often not detailed enough for areas with low relief or for more detailed landscape-field studies. New high-resolution DEMs that do combine a high resolution with substantial spatial coverage provide the opportunity to develop site-specific management alternatives for sensitive or critical source areas at the landscape level. The objective of this paper is to map hydrological pathways of P for a relatively flat glacial till landscape in the Northern part of the Netherlands and to compare two strategies (A and B) in this landscape for minimizing P loading of surface waters following different scenarios. Strategy A assumes equal lowering of phosphate input rates for all fields and strategy B takes into account the relative sensitivity of fields for removal through overland flow. Available soil data, which did not allow to geo-reference individual fields, showed high soil P levels in about 5% of all fields. Data on field inputs of total phosphate (P2O5) were available for 29 farms and showed that fertilized fields received on average 124 kg ha− 1 in the year 2000. Transfer potential of each individual field was evaluated on the basis of two characteristics in a 5 × 5 m resolution DEM: 1) the degree of internal drainage through re-distribution and 2) the degree of external drainage, which indicates the overland flow potential. Almost every field was characterised by sinks (98%) and external drainage into along-field ditches (99%). In addition, almost half of the fields showed drainage directly into surface waters (47%). Applying a classification scheme to rank the sensitivity for losses through overland flow provided a framework to distribute catchment inputs of phosphate at field level (strategy B). Comparing the strategy of distributed phosphate inputs (B) with generic phosphate inputs (A) showed that catchment input-to-output ratios are higher for strategy B for three out of four of the applied scenarios.