饱和D-最优设计在高蛋白大豆施肥优化中的应用

Application of saturated D-optimal design for optimum fertilization in the production of high protein soybean

【目的】高蛋白大豆生产是我国农业供给侧改革发展的一个重点产业,合理施肥是实现高蛋白大豆生产的一个必要环节。利用饱和D-最优设计对黄淮海地区高蛋白大豆氮、磷、钾肥施用量进行研究,旨在为高蛋白大豆生产中的肥料配施提供理论依据和参考。【方法】以高蛋白大豆冀豆21为试验材料,采用氮、磷、钾三因素二次饱和D-最优设计,每个因素分别设10个处理,每个处理3次重复,共计30个小区,随机区组排列。各处理肥料均在大豆开花期分批次施用,成熟后测定大豆产量及籽粒蛋白质含量。【结果】建立了以氮、磷、钾施用量编码值为变量因子,高蛋白大豆产量和蛋白质含量为目标函数的三元二次多项式数学模式。通过对模型解析寻优表明,氮、磷、钾肥对高蛋白大豆和蛋白质含量均有显著影响,且氮肥>钾肥>磷肥;当氮、磷、钾肥用量分别为95.46、183.8和128.7 kg/hm^2时,边际产量效应值为0,当氮、磷、钾肥分别为120.8、178.4和141.3 kg/hm^2时,边际蛋白质含量效应值为0。氮磷、氮钾、磷钾对高蛋白大豆产量和蛋白质含量存在明显的交互作用。本试验条件下,大豆产量超过3104 kg/hm^2,籽粒蛋白质含量46.04%以上的施肥方案为施氮量76.13～101.1 kg/hm^2,施磷量131.1～168.5 kg/hm^2,施钾量104.9～134.8 kg/hm^2。【结论】利用饱和D建立的肥料函数模型可以很好的说明氮、磷、钾施肥与大豆产量、蛋白质含量的连续变化关系和氮、磷、钾间的互作效应,适量的氮肥、磷肥、钾肥可有效提高高蛋白大豆产量和蛋白质含量。供试条件下,最佳肥料配比是N76.1～101.1 kg/hm^2、P 131.1～168.5 kg/hm^2、K 104.9～134.8 kg/hm^2。

【Objectives】The production of high protein soybean is an important industry in the supply-side reform in modern agriculture system. Optimal fertilization plays a vital role in the efficient production of high protein soybean. The saturated D-optimum design method was attempted in the fertilization management in highprotein soybean production in Huang-Huai-Hai region.【Methods】A field experiment was conducted using a high-protein soybean cultivar of Jidou21 as the tested crop. Fertilizers of nitrogen, phosphate and potassium were designed using the two factors' saturating D-optimal design. Each factor included 10 levels and with three replicates. The total of 30 experimental plots were randomly arranged. All fertilizers were applied in batches at the flowering stage of soybean. Both the soybean yield and protein contents were determined at the maturity stage.【Results】Using the obtained data, a ternary quadratic polynomial mathematical model was established,with the applied amounts of N, P and K as independent variables, and the soybean yield and protein content as the target functions. The analysis of model showed that application of N, P, and K significantly influenced soybean yield and protein content, and N exhibited better effect than K and P did. When the levels of N, P and K reached 95.46, 183.8 and 128.7 kg/hm^2, the marginal yield effect became zero;When the levels of N, P and K reached 120.8, 178.4 and 141.3 kg/hm^2, the marginal protein effect became zero. There were interactions among the N, P and K rates and the yield and protein content of soybean. Under the experimental conditions, to obtain above 3104 kg/hm^2 of soybean yield with more than 46%of protein content, the optimum fertilization rate was N 76.1-101.1 kg/hm^2, P 131.1-168.5 kg/hm^2 and K 104.9-134.8 kg/hm^2.【Conclusions】The fertilizer function model established using the saturated-D method is satisfactory in clearly expressing the continue change of the relationship between the NPK fertilization and the yields and protein contents of soybean, and the interaction of the three fertilizers. Suitable rates of NPK fertilizers are effectively in improving the yield and protein content of the tested soybean cultivar. The optimal rate of fertilizers is N 76.1-101.1 kg/hm^2, P 131.1-168.5 kg/hm^2 and K 104.9-134.8 kg/hm^2 for the soybean cultivar in the experimental area.