Comparison of fertilizer-effect models on winter wheat response to nitrogen and phosphorus fertilizers in saline soils in the Yellow River Delta

Salt-affected soils distribute widely across the Yellow River Delta, which inhibit plant growth and crop yield in this area.Although fertilizer application can alleviate the adverse impact of salinization on crop, little remains known about the performance of different fertilizer-effect models on simulation of'3414'fertilizer experiments in winter wheat field in saline soils.Among the essential elements, nitrogen (N) and phosphorus (P) are usually the most important nutrients limiting plant growth in saline or non-saline soils.In this study, four fertilizer-effect models were used to simulate the impact of N and P fertilizers on wheat yield in a field experiment in Lijin County of Shandong Province.The study aimed at selecting suitable fertilizer-effect model for wheat grown in saline soils in the Yellow River Delta.Four levels of N and P fertilizers (N:0 kg·hm^-2, 135 kg·hm^-2, 170 kg·hm^-2, 405 kg·hm^-2; P:0 kg·hm^-2, 53 kg·hm^-2, 105 kg·hm^-2, 157.5 kg·hm^-2) were set in eight treatments based on the'3414'test in order to investigate the fit effect of the four fertilizer-effect models-linear plus plateau, unary quadratic, square root and binary quadratic.To ensure the accuracy of fertilizer-effect models, the least square method was used in the statistical regression analysis.The results suggested that the four fertilizer-effect models had an extremely significant level (P < 0.01) based on the fit test.Unary quadratic model of effects of N and P fertilizers was the best among unary fertilizer-effect models, which suggested that the highest income levels were 7 448.3 ¥·hm^-2 and 7 357.7 ¥·hm^-2 and the recommended N and P rates were 254.4 kg·hm^-2 and 98.6 kg·hm^-2, respectively.Compared with the unary models, binary model was better.In binary model, the best N and P application amounts were respectively 244.1 kg·hm^-2 and 94.2 kg·hm^-2, with N to P ratio of 2.6.In the model, the economic benefit and agronomic efficiency of N and P were 7 432.4 ¥·hm^-2, 6.2 kg (grain)·kg^-1(N) and 13.8 kg (grain)·kg^-1(P₂O₅), respectively.On the basis of the fitted curve, the binary quadratic model had the better economic fertilizer rate, agronomic efficiency, economic benefits than the unary models.Then the binary quadratic fertilization model was the best for wheat cultivation in saline soils in Yellow River Delta.