Abstract
Nitrate leaching was studied in an Andisol upland field in Abashiri area in Hokkaido treated with six level of nitrogen input (successive applications of cow compost (0, 20, 40 Mg ha^<-1> year^<-1>) combined with treatments of N fertilization) for the last 3 years in 5 years' treatment (1998-2002). Following the standard crop rotation system in Abashiri area, wheat (Triticum aestivum L.), sugar beet (Beta vulgaris L.) and potato (Solanum tuberosum L.) were grown. The N0_<3^->N concentrations in percolate collected by pan lysimeter at a 90 cm depth decreased remarkably in the growing period of wheat and sugar beet, which have deep root systems, while the N0_<3^->N concentrations gradually increased in the growing period of potato. Mean annual subsurface percolate flow estimated by crop model during 3 years was 335 mm. The estimated NO_<3^->N loss in percolate averaged 22 kg ha^<-1> year^<-1> with the N input (sum of fertilizer N and mineralized N from compost) 117 kg ha^<-1> year^<-1>, and 62 kg ha^<-1> year^<-1> with the N input 210 kg ha^<-1> year^<-1>. Flow weighted average N0_<3^->N concentrations were related (R^2=0.96) to the N input, and estimated to exceed 10 mg L_<-1> when N input exceeds 150 kg ha^<-1> year^<-1>. According to the fertilizer recommendations given by Hokkaido Prefecture, the average of optimum N application for wheat, sugar beet and potato is around 120 kg ha^<-1> year^<-1>. Therefore, we conclude that N0_<3^->N concentrations in percolate can be maintained at less than 10 mg L^<-1> for the standard crop rotation system in Abashiri area, when N fertilization is based on the recommendations.
