Nitrate leaching in intensive agriculture in Northern France: Effect of farming practices, soils and crop rotations

The efficacy of ‘Good Agricultural Practices’ (GAP) for reducing nitrate pollution is tested on the scale of a small catchment area (187 ha) which is almost entirely under arable agriculture. GAP have been introduced on all fields since 1990. They consist in applying carefully planned N fertiliser recommendations, establishing catch crops (CC) before spring crops and recycling all crop residues. Soil water and mineral nitrogen (SMN) were measured three times each year on 36 sites representative of crops (wheat, sugarbeet, pea, barley, oilseed rape) and soil materials (loam, loamy clay and rocks, sand loam and limestone, sand) during 8 years (1991‐1999). These measurements (about 3600 soil samples) were used in LIXIM model to calculate water and nitrogen fluxes below the rooting zone. The model could reproduce satisfactorily the water and SMN contents measured at the end of winter. It simulated reasonably well the nitrate concentration measured in the subsoil (3‐10 m deep) of nine fields. The mean calculated amounts of drained water and leached nitrogen below the rooting depth were 231 mm year � 1 and 27 kg N ha � 1 year � 1 , corresponding to a nitrate concentration of 49 mg L � 1 . Leached N varied by a five-, four- and threefold factor according to the year, crop and soil type, respectively. Nitrate concentration was primarily affected by soil type: it varied from 31 mg L � 1 in deep loamy soils to 92 mg L � 1 in shallow sandy soils, and was linked to the soil water holding capacity. The sugarbeet‐wheat rotation gave the lowest concentration (38 mg L � 1 ) and the pea‐wheat rotation the highest one (66 mg L � 1 ). In spite of their moderate growth (mean biomass = 0.8 Mg ha � 1 ), the catch crops allowed to reduce the mean concentration by 50% at the annual scale and 23% at the rotation scale. Straw incorporation was also beneficial since net mineralisation between harvestand late autumn was reduced by 24 kg N ha � 1 when strawresidues were incorporated. Reducing fertilisation below the recommended rate did not significantly reduce further nitrate leaching. Although GAP were not all optimal and therefore less efficient than in well controlled experiments, they appear essential in intensive agriculture in order to comply with the EU standard for nitrate concentration. # 2005 Elsevier B.V. All rights reserved.