论文部分内容阅读
为综合治理谷田杂草,探索谷田阔叶杂草对谷子的竞争规律。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,谷子产量损失率与阔叶杂草密度呈显著正相关,随着阔叶杂草密度的增加呈不对称S型增加的趋势;幂双曲线函数模型拟合优度(99.5%)最大,残差平方和(22.878)最小,是模拟阔叶杂草与谷子竞争关系的最优模型。预测方程式为Y=d1.179/(3.740+0.021d1.179),阔叶杂草单株最大竞争力为0.2674,群落竞争力为0.0056,谷子产量最大损失率为47.62%。研究确定了一种拟合优度高而又实用的预测模型,可为综合治理谷田以马齿苋为主的阔叶杂草提供技术支撑。
In order to comprehensively control the weeds in the valley, we explore the competition law of the broad-leaved weeds in the valley. Field trial and non-linear regression analysis methods were used to simulate and compare the main weed competition experience models. The results showed that the yield loss of millet was significantly and positively correlated with the broadleaf weed density, and the asymmetric sigmoid type increased with the increase of the broadleaf weed density. The hyperbolic function model fitting power (99.5%) was the largest , The residual sum of squares (22.878) is the smallest, is the best model to simulate the competition between broadleaf weeds and millet. The prediction equation was Y = d1.179 / (3.740 + 0.021d1.179). The maximum single herbivores of broadleaf weeds was 0.2674, the community competitiveness was 0.0056, and the maximum loss of millet was 47.62%. The study determined a good and good fitting prediction model, which could provide technical support for the comprehensive management of broadleaf weeds dominated by purslane in Gutian.