以植物丰度为辅助变量的钉螺空间分层采样研究

Study of spatial stratified sampling strategy of Oncomelania hupensis snail survey based on plant abundance

目的优化湖沼型血吸虫病流行区钉螺调查方法,为提高查螺效率提供依据。方法基于一处50 m×50 m的草洲范围内推扫式的钉螺调查数据,提出一种基于植物丰度分布特征为辅助变量的钉螺空间采样策略;根据植物丰度数据进行空间分层,通过Hammond McCullagh方程计算各子区域内的最优采样点数量;采用多向插值离散度的空间采样布局方案(MDI),确定采样点部署策略;并对比空间随机采样、系统采样、空间分层采样和三明治采样。结果 5种采样策略中,本研究提出的采样策略所得到的抽样绝对误差最小,为0.213 8,系统抽样误差最大为0.924 4。结论本研究提出的钉螺采样策略能够获得比空间随机采样策略、系统采样策略、空间分层采样策略和三明治采样策略更高的估算精确度。

Objective To optimize and simplify the survey method of Oncomelania hupensis snails in marshland endemic re-gions of schistosomiasis,so as to improve the precision,efficiency and economy of the snail survey. Methods A snail sam-pling strategy（Spatial Sampling Scenario of Oncomelania based on Plant Abundance,SOPA）which took the plant abundance as auxiliary variable was explored and an experimental study in a 50 m＆#215;50 m plot in a marshland in the Poyang Lake region was performed. Firstly,the push broom surveyed data was stratified into 5 layers by the plant abundance data;then,the required numbers of optimal sampling points of each layer through Hammond McCullagh equation were calculated;thirdly,every sample point in the line with the Multiple Directional Interpolation（MDI）placement scheme was pinpointed;and finally,the compari-son study among the outcomes of the spatial random sampling strategy,the traditional systematic sampling method,the spatial stratified sampling method,Sandwich spatial sampling and inference and SOPA was performed. Results The method（SOPA） proposed in this study had the minimal absolute error of 0.2138;and the traditional systematic sampling method had the largest estimate,and the absolute error was 0.9244. Conclusion The snail sampling strategy（SOPA）proposed in this study obtains the higher estimation accuracy than the other four methods.