中文摘要：

近年来有关节肢动物群落的研究报道较多，而传统的、简单的多样性指数分析远远不能满足对高通量群落数据的挖掘。采用多元统计方法可以较好地分析群落数据。其中，在研究某一处理的长期效应时，主响应曲线是一种经典的多元分析方法；而非度量多维尺度分析具有能处理较大量的数据且常能以低维排序图反映结果的优点，是一种具有广泛应用前景的排序技术。本文以Bt棉与非Bt棉不同比例混作棉田的节肢动物群落为例，详尽地介绍了主响应曲线和非度量多维尺度在节肢动物群落分析中的应用。

英文摘要：

Increasing numbers of studies of arthropod communities have been made in recent years, but comprehensive information derived from community analyses is limited by the use of traditional and simple diversity indices. Multivariate analysis methodology is particularly well suited for meeting specialized needs for high-throughput community data mining. The principal response curve is a classic multivariate analysis method which is widely used in environmental monitoring and ecological restoration to explore the long-term effects of certain human activity. Non-metric multidimensional scaling analysis can deal with large amounts of data and show the resuhs in a low-dimensional ordination, making it a promising method for analyzing community ecology. However, these methods have rarely been used in arthropod community analysis in China to date. Here, the uses of principal response curve and non-metric multidimensional scaling are introduced in a study of arthropod communities from Bt and non-Bt mixture cotton fields with different acreage ratios.