Long terme responses of crop yield, soils and nitrate losses to best agricultural practices at the catchment scale | Réponses à long terme des bonnes pratiques agricoles sur le rendement des récoltes, le sol et les pertes de nitrate à l'échelle du bassin versant

Facing the nitrate pollution problem, the European Union has encouraged a code of "Good Agricultural Practices" (GAPs) in order to recover a good biochemical and ecological status of waterbodies. However, the efficiency of GAPs application remains greatly unknown on the long term and at a large spatial scale. In this work, we studied the water and nitrogen fluxes in a small (187 ha) agricultural catchment during 22 years after GAPs implementation (Beaudoin et al., submitted, J. of Environ. management). This prevention action relied on the hypothesis that a systematical GAPS application would allow nitrate at the outlet to meet the US standard (50 mg NO3 L-1). GAPs consisted in adjusting N fertilization rates, establishing catch crops and converting two fields into grassland. The aquifer outlet is a set of water springs which were monitored for water flow and nitrate concentration. The water (RN) and solute (RS) renewal times in the catchment were assessed by three methods. RS was estimated at 13-25 years and RN was 14-74 years, all indicating a strong inertia of the hydrological system.Soil water and mineral N were measured two or three times per year on 36 permanent sites representative of soil and crop variability. These data were used to initialize the STICS model which predicted water drainage and N leaching. At field scale, without integrating set-aside fields, drained water below rooting depth was on average 179 mm yr-1, N leached 19 kg N ha-1 yr-1 and nitrate concentration 44 mg NO3 L-1. Nitrate concentration decreased with soil water retention capacity. Nitrate concentration in the main water spring began to decline 11 years after GAP implementation and levelled off to 49 mg NO3 L-1. The STICS-MODCOU agro-hydrogeological modelling chain satisfactorily predicted water flow and nitrate concentration of the springs but overestimated the response time. Observations and modelling proved the positive impact of GAPs which appear economically competitive, as compared to the curative water treatment, but require long term management. The role of set-aside was here marginal and with higher costly/efficiency ratio than GAPs. A more flexible land occupation could combine land sparing and land sharing. GAPs are expected to be components of the transition towards agro-ecological systems within a loop of progress.