中国40，000种昆虫蜘蛛区系的多元相似性聚类分析

The multivariate similarity clustering analysis for 40,000 species of insect and spider fauna in China

使用多元相似聚类分析（Multivariate similarity clustering analysis，MSCA）方法，封隶属于22目339科7953属的40，348种昆虫和蜘蛛在中国28省区的分布进行相似性分析。结果显示，28省区的总相似性系数为0．079，在0．22的相似性水平上，全国分为9个昆虫地理分布区。这一结果肯定和支持了中国现有动物地理和昆虫地理研究的基本成果，同时也提出了需要进一步改进与完善的建议，如内蒙古和东北3省关系密切；宁夏和华北3省距离较近；台湾和华南3省比较疏远；河南和陕西、甘肃关系密切。和传统的相似性聚类分析法（SCA）相比，直接计算多省区的相似性系数和不再进行逐省合并，是MSCA法的两个特点，快捷简便和准确合理是MSCA法的两个优点，从而能够为生物地理学研究和地理区划提供数理统计学的支撑、

The authors analyzed the similarity for 40,348 insect and spider species belonging to 22 orders, 339 families and 7,953 genus distributed in 28 provinces in China by using Multivariate Similarity Clustering Analysis （MSCA）. The results showed that the general similarity coefficient of 28 Provinces were 0.079. Whole country were divided into 9 geographical distribution regions at 0.22 level. This result not only affirmed and sustained the product of Chinese Animal Geography and Insect Geography at present, but also put forward some suggestions need more improved and perfect, such as the insects in Inner Mongolia were closely related to those in 3 northeastern Provinces, insects in Ningxia Province were relatively close to those in 3 North China Provinces; insects in Taiwan were relatively distanced from those in 3 South China Provinces, insects in Henan Province were closely related to those in Shaanxi and Gansu Provinces. Compared with the traditional Similarity Clustering Analysis, MSCA directly calculate the similarity index for multiple Provinces, which do not require consolidating province by province with largest similarity index. MSCA is rapid, simple and precise, providing a statistical support for the studies on the geographical division of organisms.