Predicting Nitrate Leaching from Drained Arable Soils Derived from Glacial Till

Abstract Regional‐scale prediction of ground and surface water pollution by nitrate (NO3) requires field‐scale prediction of NO3 leaching. This can be done by a N balance sheet approach, but this is not robust, because no account is taken of organic matter dynamics or denitrification. In this paper, predictions of NO3 leaching made using a balance sheet method and those using a dynamic N cycle model, Agricultural Nitrogen Model (ANIMO), are compared with observations of NO3 leaching from drained plot experiments on two Scottish arable soils, a clay loam (aquic Haplaquept) and a sandy loam (dystric Eutrochrept). Mineral N fertilizer applications in these experiments varied from 0 to 210 kg N ha−1 and the mean observed NO3‐N leaching to drains was 21 kg N ha−1 for the clay loam and 53 kg N ha−1 for the sandy loam. The balance sheet method on average overpredicted NO3 leaching by 120% for the clay loam and underpredicted by 16% for the sandy loam. ANIMO predictions using default parameters were poor but when subsoil respiration was included to a depth of 0.7 m at the clay loam site and 2.0 m at the sandy loam site, the mean predicted NO3‐N leaching was within 2% of observed. The root mean square error was still quite large (76 and 53% for the clay loam and sandy loam, respectively) and the correlation coefficient was poor at the clay loam site (r = 0.09 and 0.91 for the clay loam and sandy loam, respectively). Simulated denitrification losses are strongly dependent on the depth to which soil respiration occurs. Work is needed to obtain more direct evidence of subsoil denitrification. Heterogeneity of topsoil organic matter and respiration may also need to be taken into account.