夏谷田杂草为害损失预测模型的研究

为了综合治理谷田杂草，探索谷田杂草对谷子的危害及竞争规律。采用田间小区试验和非线性回归分析的方法，对主要的杂草竞争经验模型进行模拟和比较。结果表明，谷子产量损失率与杂草密度呈显著正相关，且随着杂草密度的增加呈减速增加的趋势。双曲线模型的决定系数R2为0.99712，最小残差平方和为16.174，是模拟谷田杂草与谷子竞争关系的优化模型。预测方程式为Y=d/(1.733+0.018d)，杂草的种间竞争力为0.5770，种内竞争力为0.0103，谷子产量最大损失率为55.56%。本研究确定了谷田杂草对谷子为害的预测模型，将为谷田杂草综合治理提供有益帮助。

Study on Prediction Model of Millet Yield Lossing Caused by Weeds in Summer Season Millet Field

The study focused on weed management in summer season foxtail millet field by evaluating the weed damage and exploring competition between weeds and foxtail millet, a few fitting models were simulated and compared by employing field plot experiment and nonlinear regression analysis. The results showed that, the millet yield losses and weed density were highly significantly correlated. Among the tested models, the determination coefficient (R2) of hyperbolic model was 0.99712, and minimum sum of squared residuals was 16.174, which was considered the optimal model to simulate the competition relation between weeds and millet. The Predicted equation was Y=d/(1.733+0.018d). The competitiveness between weed species was 0.5770, within the species was 0.0103. The maximum millet yield loss was 55.56%. This study has established a prediction model with high goodness-of-fit and practical prediction which could help weed management in summer season millet field.