摘要： 1987~1988年在山西省屯留试点进行试验，采用二次回归正交旋转组合设计，对影响谷子产量形成的密度、播期、氮、磷和农家肥五项关键农艺措施进行了定量化的综合研究，初步确定了最佳农艺措施组合方案，并利用产量反应模型研究了各项措施对产量形成的单独反应以及各项农艺措施间的互作效应。优化方案经1988和1989年的示范对比和小面积推广，证明是切实可行的。

关键词: 旱地, 谷子, 栽培措施, 数学模型, 产量

Abstract: The experiment was carried out on the experimental farm in Tunliu of Shanxi province during 1987 - 1988.Through quadratic regression orthogonal rotational combinatorial design the quatitative comprehensive study was made about the key agronomic measures of sowing density, sowing time, N, P and manure fertilizer application amount, and the optimum synthetic design of agronomic measures was preliminarily determined.Both single reactions on the millet yield and interactions of all agronomic measures were also studied using yield reaction model.After application on a small scale and comparison, the optimization plant was proved to be feasible and profitable.

Key words: Millet, Arid land, Cultural measures, Mathematical models, Yield