Yield response to N fertilizer and optimum N rate of winter oilseed rape under different soil indigenous N supplies

• Four levels of soil indigenous N supply were graded using 1457 site-year experiments.
• Economic optimum N rates (EONR) were evaluated through the yield-response model.
• Nearly 45% of all sites applied N fertilizers suitably in Yangtze River Basin.
• N fertilizers were overused and/or lacking over 25% of the sites.
• EONR could increase yield with efficient N fertilizers input in different INS levels.

Assessing the optimum N fertilizer rate is a crucial step in nitrogen (N) management of winter oilseed rape (Brassica napus L.). We conducted 1457 site-year experiments with four N application rates, no-N (i.e., no N application), LN (i.e., half the N rate of the MN treatment), MN (i.e., current recommended N rate from local technicians), and HN (i.e., 1.5 times the N rate of the MN treatment) at the regional scale in China, aiming to evaluate the economic optimum N rate (EONR) for winter oilseed rape under different levels of soil indigenous N supply (INS). We used yield under no-N treatment to represent INS. Four INS levels (i.e., L1, L2, L3, and L4) were graded based on the relationship between the yield under no-N treatment and the relative yield (i.e., yield of no-N treatment divided by the yield of the MN treatment). The results indicated that the rapeseed yield was higher under the MN treatment than under the LN and HN treatments across different INS levels, with an average of 2267–3185 kg ha−1. The values of EONR were determined based on the relationship between rapeseed yield and N application rates across sites, with an average of 200, 182, 173, and 162 kg N ha−1 in the levels of L1, L2, L3, L4, respectively. Currently, nearly 45% of the sites apply suitable rate of N fertilizer (i.e., nearly 180 kg N ha−1) in the Yangtze River Basin. However, current N rates were insufficient (Δ N rate, i.e., N difference between the EONR and MN, higher than 10 kg N ha−1) in 37% and 32% of the experiments for L1 and L2 INS levels, respectively, and excessive (Δ N rate ≤ −10 kg N ha−1) in 30% and 36% of the experiments for L3 and L4 INS levels, respectively. The EONR optimized the current N rate with a yield-response model, which could increase rapeseed yield with efficient N fertilizers input under different INS levels. Our study provides data to support regional N fertilizer management systems for winter oilseed rape, especially in those countries that also plant multiple cropping systems and/or plant small-scale sites across the world.