生物入侵风险分析的研究进展

Research Progress of Risk Analysis of Biological Invasion

随着全球经济一体化的加快,外来物种入侵对中国生态环境带来非常严重的危害和巨大的经济损失。生态位模型(ecologicalniche models,ENMs)为外来入侵生物的风险分析提供了重要的定量化分析工具,已经广泛应用于入侵生物适生区的预测,但如何评价不同模型之间的预测效果成了当代研究的热点问题。介绍了4种生态位模型对其应用原理和分析方法进行比较。在这4种模型中:MAXENT和GARP是利用物种的已知分布数据和环境数据产生以生态位为基础的物种生态需求,然后研究生物多样性和物种的潜在分布;CLIMEX是通过物种在已知地理分布区域的气候参数来预测物种的潜在地理分布,CLIMEX运行结果的EI值是一个相对值,反映了物种在各地分布的相对丰度,并能以图表的形式表现出来;GIS即可以单独构建对目标物种的适生区预测,也可以与其他模型相结合,主要用到了他的两个功能:插值和叠加。

With the economic globalization speeding up, invasion of the alien species resulted in very serious harm and huge economic losses for ecological environment in China. In the target region, risk evaluation of invasive species is the precondition of reasonable quarantine measures and implementation of scientific planning, while potential geographic distribution of invasive species in the target region is the foundation of scientific risk evaluation of invasive species. Ecological niche models（ENMs）, which are widely applied to predict the potential geographic distribution of species,provide an important tool to quantify the risk evaluation of invasive alien species. One of the frequently discussed issues is how to evaluate the effectiveness among different models. In this article, four common ecological niche models were introduced and their advantages and disadvantages were discussed. Using known distribution and the environmental data, MAXENT and GARP produce ecology demand of species based on ecology position, then study biodiversity and potential distribution of species; CLIMEX is to forecast the potential geographic distribution of species by climate parameter collected from the known geographic distribution region of that species. EI value generated from CLIMEX, which is a relative value, has reflected relative abundances of the species in the region, and can be expressed by the graph. GIS can construct potential geographic distribution of invasive species in the target region and unify with other models. This ability mainly depends on its two functions： Interpolation and superimposition. With all this knowledge of the four models, researchers can wisely and effectively choose the model（s） to meet their requirements.