【摘要】 以贵州省的71个3414水稻施肥试验数据为基础,研究其三元二次施肥模型、一元施肥模型、线性+平台型对3414肥料试验结果的拟合情况。结果表明,能用三元二次拟合的肥效试验仅占11.27%,能用一元肥料效应模型拟合的氮、磷、钾肥试验分别占56.34%、33.80%和43.66%;能用线性+平台型拟合的氮、磷、钾肥试验分别占8.45%、11.27%和5.63%;表现为增产趋势的分别占22.54%、22.54%和11.27%;表现为非增产趋势的分别占2.82%、1.14%和2.82%;而由于非肥力因素影响的氮、磷、钾肥试验分别占9.86%、30.99%和36.62%。在同一试验中,分别采用一元施肥模型、三元二次施肥模型计算推荐施肥量,结果三元二次模型拟合N肥平均推荐用量较一元肥料效应模型高40.9%。P肥平均推荐用量较一元肥料效应模型高73.6%。K肥平均推荐用量较一元肥料效应模型高54.5%。因此,在贵州省的水稻3414施肥试验数据的计算过程中,一元肥料效应模型和线性+平台型的拟合是对三元二次肥料模型拟合的补充和优化。更多还原

【Abstract】 This paper discussed the fit of rice fertilization models such as ternary quadratic model,single variable model,and linear + platform models based on the data from 71 experiments of 3414 formulation.The results showed that the model of experiments can fit with ternary quadratic was only 11.27%,the one can fit with single variable were 56.34%,33.80%,and 43.66%,and the one fit linear + platform were 8.45%,11.27%,and 5.63% respectively.The experiments had yield increase potential were 22.54%,22.54%,and 11.27%,oppositely the experiments had no yield increase potential were 2.82%,1.14%,and 2.82%.And the experiments which affect by non-fertilizer factors were 9.86%,30.99%,and 36.62%.In one experiment,the fertilizer amount calculated by single variable,ternary quadratic model,results showed that the amount of N fertilizer recommendation calculate by ternary quadratic was 40.9% higher than single variable model,73.6% higher in P fertilizer,and 54.5% higher in K fertilizer.So the fit of single variable model and linear + platform model was replenish and optimization for the fit of ternary quadratic in calculation of 3414 formulation for rice plating in Guizhou.更多还原