辽宁省花生“3414”肥料试验施肥模型探讨

以农业部测土配方施肥项目2006～2007年在辽宁省进行的34个花生试验数据为依据,比较向后回归法建立肥料效应回归模型与强迫法建立肥料效应回归模型的优劣,探讨对关键处理设置重复在建立数学模型中起到的作用。结果表明,采用强迫法进行三元二次方程拟合的成功率只有12%,而用向后回归法拟合的成功率有55%,2007年对处理1、处理2、处理4、处理6、处理8设置重复后,三元二次方程模型和一元二次方程模型方程的拟合的成功率分别占当年试验总数的84%、16%。2006年不设置重复,有20%不能够拟合。因此,"3414"肥料试验关键处理设置重复,并采用向后回归法,对"3414"肥料试验得到合理施肥模型具有重要作用。

A Study on Peanut Simulation of "3414"Fertilizer Experiments in Liaoning

This paper studied the effect of multiplicating key treatments by mathematical models, based on 34 peanut experimental data which were conducted from 2006 to 2007 in Liaoning province. The results showed that the success ratio of simulation by three-factor-two times formulation of enter method and by back regressed method was 12% and 55% respectively. the success ratio of simulation by three-factor-two-times formulation model and by one-factor-two-times formulation model was 84% and 16% respectively in the experiment in 2007 with multiplicating treatment 1, treatment 2,reatment 4,treatment 6 and treatment 8 in 2007, while that simulation unsuccessfully was 20% without multiplicating treatment in 2006. So in'3414'fertilizer experiment,it played important role in the user of utilized fertilizer model that multiplicating key treatments and back regress method.