微环境对山区钉螺分布的影响

目的探索山区钉螺分布与微环境因素之间的关系,为生态灭螺、控制血吸虫病流行提供依据。方法 2008年2月-2009年2月选择普格县城附近的山坡草滩为现场,随机抽样查螺测量部分植被盖度、土表温度,并采集土样。实验室压螺鉴定死活并分框记数,测量土壤含水量、pH。首先计算中位数、方差、最小值、最大值等指标描述钉螺及其影响因素,然后绘制钉螺与影响因素间的散点图并做Spearman秩相关分析,最后拟合钉螺与因素间的单因素曲线模型、多因素广义相加模型。结果共查螺75框;土壤pH为4.88～7.77,温度为19.3～27.1℃,含水量为0.14～0.85。植被盖度1%～100%。Spearman秩相关分析显示仅pH与钉螺数有明显相关性(P0.01)。拟合钉螺密度与各因素间曲线模型,校正r2均0.8。多因素广义相加模型拟合各因素均有统计学意义,在扣除线性及其他因素的影响后,各微环境因素与钉螺间曲线关系表现与单因素拟合不同。结论山区钉螺分布与微环境因素间为平滑函数关系,不能简单采用负二项回归或其他线性回归模型拟合。

Impact of micro-environment on Oncomelania hupensis distribution in mountains areas

Objective To explore the relationship between Oncomelania hupensis snail distribution and micro-environment factors in mountain areas so as to supply the information for eliminating the snails with ecology methods and controlling schistosomiasis.Methods A hillside marsh in Puge County,Sichuan Province was selected as the study field from February 2008 to February 2009.The snails were investigated by the random sampling method.The soil surface temperature and vegetation cover were measured partly when the snails were investigated,and the soil samples were collected.The number of living snails was counted by the method of dissection at the laboratory and the soil pH and moisture were measured,too.Statistical indices of snails such as median,variance,minimum value,maximum value etc.were first calculated to conduct a preliminary analysis,then the scatter diagram and spearman analysis between the factors and snails were done,and finally the general additive model for all factors and the curve model for each factor were fitted.Results Seventy-five frames were surveyed;the soil pH values were from 4.88 to 7.77,the coverages of the grass were from 1 percent to 100 percent;the soil temperatures were from 19.3 ℃ to 27.1 ℃ ;the water contents of soil were from 0.14 to 0.85;and the spearman analysis showed that pH was significant statistic.The curve models were fitted effectively,and adjusted r2 were all above 0.8.The general additive model for all the factors was fitted and the factors were all significant.After the deduction of linear and other factors,the curve relationship between the micro-environmental factors and snails was different from that single-factor fitted.Conclusion There is a smoothness function relationship between the snails and micro-environment factors,so the negative binominal regression or other linear regress models are not fitted simply.