二化螟种群密度的克力格估值及其模拟抽样

为设计可靠合理的二化螟幼虫种群密度抽样方案，从二化螟幼虫空间分布原始总体出发，另构建了一个随机总体和一个顺序总体，采用无放回随机抽样、间隔变程以上无放回随机抽样和基于克力格估值且初始点随机的顺序抽样对3总体进行了模拟抽样比较．结果表明，间隔变程以上随机抽样对原始总体平均数的估计优于随机抽样，且随总体聚集程度增加，间隔变程以上随机抽样愈优；正确识别种群空间格局极为重要，对聚集分布总体采用随机抽样和对随机分布总体采用间隔变程以上随机抽样均将降低抽样估计精度．针对随机抽样在应用上的局限性，提出了一种基于地统计学克力格估值、初始点随机的顺序抽样方案：它以初始点随机保证随机性，以顺序抽样保证可操作性，以二化螟种群空间分布的区域变量属性保证克力格样本较调查样本对局域样本和总体的平均数估计为优；且聚集范围一定时，总体聚集强度愈大，克力格样本局域估计和全局估计愈优于调查样本；取样间隔(以变程为标准)极为重要，样方的空间布局要平衡考虑相互独立的样方对数和变程范围内的样方对数。

Krigle estimation and its simulated sampling of Chilo suppressalis population density

In order to draw up a rational sampling plan for the larvae population of Chilo suppressalis,an original population and its two derivative populations,random population and sequence population,were sampled and compared with random sampling,gap-range-random sampling,and a new systematic sampling integrated Krigle interpolation and random original position.As for the original population whose distribution was up to aggregative and dependence range in line direction was 115 cm (6 9 units),gap-range-random sampling in line direction was more precise than random sampling.Distinguishing the population pattern correctly is the key to get a better precision.Gap-range-random sampling and random sampling are fit for aggregated population and random population,respectively,but both of them are difficult to apply in practice.Therefore,a new systematic sampling named as Krigle sample (n=441) was developed to estimate the density of partial sample (partial estimation,n=441) and population (overall estimation,N=1500).As for original population,the estimated precision of Krigle sample to partial sample and population was better than that of investigation sample.With the increase of the aggregation intensity of population,Krigel sample was more effective than investigation sample in both partial estimation and overall estimation in the appropriate sampling gap according to the dependence range.