一、前 言 长期以来,在农业科研工作中人们习惯于用单因子的研究分析方法来考察某一因子对农牧产品的产量、品质等方面的影响。这种研究方法的优点是因子作用明确,试验技术比较简单,便于分析比较。“一次变动一个因素,全面进行记录”至今仍是一种有用的研究方法。但是,在农业生态系统中,因子数量多,诸因子间叉存在着正或负的相互效应,单因子试验往往不能反映出事物变化的真实面貌。 当农作物受到两种肥料的共同作用时,其产量反应可以用立体的“肥料效应曲面”来表示。所谓“肥料效应曲面”是指,当用三向座标系统来表示产量(y)随肥料(x)和肥料(z)的施用量变化而发生变化时,由三联数(x、y、z)决定的许多点子所构成的回归曲面。代表“肥料效应曲面”的数学模型中应用较广泛的是二元二次回归方程: I. Preface For a long time, people are accustomed to studying the influence of a certain factor on the output and quality of agricultural and animal husbandry products in the agricultural research work. The advantage of this research method is the clear effect of factors, test technology is relatively simple, easy to analyze and compare. “One change at a time, a comprehensive record” is still a useful research method. However, in agro-ecosystems, there are many factors and there are positive or negative mutual effects among factors. Univariate tests often can not reflect the true appearance of things. When the crop is affected by both fertilizers, the yield response can be expressed in terms of the three-dimensional “fertilizer effect surface.” The term “fertilizer effect surface” means that when a three-dimensional coordinate system is used to indicate that the output (y) changes with the change of the application amount of the fertilizer (x) and the fertilizer (z) ) Many of the ideas decided by the regression surface. The most widely used mathematical model representing “fertilizer effect surface” is the binary quadratic regression equation: