| The phosphorus fertilization is essential for obtaining high potato yield but is also a source of non-point pollution. We propose an improved P recommendation model for Quebec, using results of soil tests and fertilizer trials conducted on coarse-textured soils.; The P recommendation model comprised an environmental risk assessment and P response curves calibrated on the 100˙PMehlich-3Al Mehlich-3 ratio. The environmental model showed an upper critical value of 15%. The lower agronomic critical value was 8.6. Below 15% as 100(P/Al)M-III , a liberal recommendation model covering 80% of the cases would still be environmentally safe, but soil test P will build up more rapidly compared to a more conservative model covering 50% of the cases. Thus, the potato crop was at risk when the P saturation exceeded 15%.; The systemic approach to fertilizer trials in view of improving P fertilization requires a methodology to verify the assumption that one nutrient is being varied while other nutrients are in sufficient, non excessive, amounts. We derived DRIS and CND leaf nutrient norms from a survey of 663 observations on Quebec farms. The norms were validated for nutrient imbalance and P using 113 fertilizer trials conducted during three years. The datasets comprised tuber yield and N, P, K, Ca and Mg composition of the first mature leaf at the beginning of blooming. Variance functions were used to separate high-yield from low-yield subpopulations. The cutting yield level was 35 Mg ha --1. We applied the following diagnostic approaches to our validation dataset: the Critical Value Approach (CVA), DRIS (Diagnosis and Recommendation Integrated System) and CND (Compositional Nutrient Diagnosis). In the validation dataset (P fertilizer trials), N and Ca were not deficient. CVA diagnosed Mg, while DRIS and CND detected Mg deficiency and K excess. The K and Mg indexes were negatively correlated at each experimental site. Critical nutrient imbalance indexes were 44.1 for DRIS and 6.7 for CND. Radiating results from the Mg-deficient site that was not conformed to the nutrient sufficiency assumption. Critical P indexes were --16 for DRIS and --1.9 for CND, were symmetrical and significantly correlated with nutrient imbalance indexes as would be expected from single nutrient experiments.; Our results could be used for extension and the theory verified with other nutrients and crops under various pedoclimatic conditions. (Abstract shortened by UMI.)... |
