中国昆虫区系的多元相似性聚类分析和地理区划  

The multivariate similarity clustering analysis and geographical division of insect fauna in China

作　　者：申效诚[1,2] 刘新涛[1] 任应党[1] 申琪[3] 刘晓光[1] 张书杰[2] 

机构地区：[1]河南省农业科学院植物保护研究所,郑州450002 [2]郑州大学生物工程系,郑州450001 [3]河南中医学院第一临床医学院,郑州450000

出　　处：《昆虫学报》2013年第8期896-906,共11页Acta Entomologica Sinica

基　　金：河南省重点实验室建设专项经费(112300413221)

摘　　要：为了总结分析中国昆虫的分布规律,为昆虫地理区划提出依据,作者2008年创建了新的多元相似性聚类分析方法,它与传统的聚类分析方法的主要不同是直接计算任意多个地区之间的相似性系数,从而摈弃原来的合并环节。我们应用这种方法对中国823科17 018属93 662种昆虫的分布进行了定量分析。当种级水平相似性系数分别为0.19和0.14时,全国64个基础地理单元聚类为20个小单元群和9个大单元群。根据聚类分析结果,提出我国第一个由定量分析产生的9个昆虫区20个昆虫亚区的中国昆虫地理区划的初步方案。西北昆虫区、东北昆虫区、华北昆虫区、青藏昆虫区稳定地组成中国北方昆虫大区,江淮昆虫区、华中昆虫区、西南昆虫区、华东昆虫区、华南昆虫区稳定地组成中国南方昆虫大区。江淮昆虫区的设立是因为这里多平原和丘陵,为重要农业区,昆虫区系简单,种类显著少于华东、华中、华南昆虫区,而和华北昆虫区相当;包括台湾在内的华东昆虫区是昆虫多样性最丰富地区之一,显著高于邻近的华北、江淮、华南昆虫区。最后,对昆虫地理定量分析中的重要环节进行讨论,还对今后昆虫区系调查的薄弱地区提出建议。这项研究证明多元相似性聚类分析方法是简便、快捷、实用的,能够在更广泛的范围应用,这预示着生物地理学研究将会由定性研究进入到定量研究阶段。In order to analyze the insect distributional law in China and provide scientific basis of insect geographical division, in 2008 we created a new method, i. e., the multivariate similarity clustering analysis （MSCA）. The key difference of this method from traditional similarity clustering analysis methods is directly computing the similarity coefficient of multi-region and throwing away the old merged steps. By using the MSCA, distribution information of total 93 662 insect species belonging to 17 018 genera of 823 families in 64 basic geographical units （BGUs） of China was analyzed. The results showed that 9 large unit groups and 20 small unit groups are clustered by 64 BGUs when similarity levels are 0.14 and 0.19, respectively. First insect fauna geographical division plan of China was putted forward by quantitative analysis, in which the insect fauna of China is divided into 9 regions and 20 sub-regions. The 9 regions were named as Northwestern China, Northeastern China, Northern China, Qinghai-Xizang, Changjiang-Huaihe, Central China, Southwestern China, Eastern China and Southern China insect region. A northern large region was clustered by the preceding 4 regions and a southern large region by the others stably. Taiwan was clustered to East China, but not to South China, because its similarity degree with Fujian is closer than that with Guangdong, Guangxi or Hainan. Several important aspects in quantitative analysis on insect biogeography are discussed, and some suggestions about insect fauna investigation are given. This study proves that the MSCA method is very applicable, simple and quick, and can be used in a wider range. It is presaged that bio-geographical study will get from qualitative analysis into a quantitative stage.

关 键 词：昆虫 区系 分布 动物地理学 地理区划 多元相似性聚类分析 中国 

分 类 号：Q968[生物学—昆虫学]