利用反应-扩散模型预测生物入侵  

The application of biological invasion prediction-illustrated Reaction-Diffusion model

作　　者：张林艳[1,2] 叶万辉[2] 江洪[3] 

机构地区：[1]云南大学资源环境与地球科学学院,昆明650091 [2]中国科学院华南植物园,广州510650 [3]南京大学国际地球系统科学研究所,南京210093

出　　处：《生态环境学报》2009年第4期1565-1572,共8页Ecology and Environmental Sciences

基　　金：国家自然科学基金重点项目(30530160);广东省自然科学基金重点项目(05200701)

摘　　要：生物入侵严重威胁着各国的生态环境,各国在生物入侵的预警、预防、预测方面给予了高度关注。文章引介了当前入侵预测中主要采用的3种方法(即统计方法、生理方法和分析方法),对入侵模拟时所需考虑的影响因素及模型的输入和输出变量进行了描述分析。以分析方法中的反应–扩散模型(简称R–D模型)为例探讨建模方法对生物入侵的预测,并基于一维反应–扩散模型的推导,阐述了R–D模型的模型特点及应用时的限制条件。R–D模型可将空间和种群过程融合进入侵速率的预测之中,且模型中连续参数的使用不受尺度限制,适用空间尺度较广。该模型以其简洁易表达的形式在生物入侵的研究中曾受到广泛应用,范围广至孢子、病毒,无脊椎农业害虫及麝鼠Ondatra zibethicus、水獭Enhydra lutris等脊椎生物。然而,随着入侵理论的不断完善,R–D模型在实际应用中的缺陷日益突出,它对空间现象模拟预测的无能为力、对生态过程相互作用模拟的失败,致使该模型应用受限、预测能力降低。目前,入侵预测倡导基于个体的空间模拟模型,并开始关注偶然发生却起着关键作用的长距离扩散的预测。从某种程度上来说这些模型比R–D模型更接近真实,但R–D模型有助于理解现有的入侵格局,对生态入侵研究提供最基本的估测,可以加深对入侵过程的理解。过去,R–D模型对入侵生态理论的发展做出过的重要贡献,今后该模型仍为我们初步预测生物入侵提供一个简单便捷的工具。Presently, the biological invasions emerge as the major threat factor to the ecological environment, as well as the trade globalization made such invasions even more inevitable. Consequently, the alteration, prediction and prevent the biological invasions had been got more attention in most countries. The statistical method, physiological method, and analysis method for biological invasion prediction were introduced in this paper. We also evaluated these variables, which driving the invasion model, and discussed the requirement of input and output variables. The Reaction-Diffusion model （a.k.a. R-D model） was an example to show how the modeling methods can be useful in the biological invasion prediction. The characteristics and limitation of the R-D model was analyzed through the construction of one-dimensional R-D model. The R-D model can combines the spatial and population process within the prediction of invasion rate, also the use of continuous parameters in the model made it free to broad spatial scale. The R-D model was once widely applied because of its function is simple and easy to calibration. It had been applied from spore, virus, invertebrate agricultural pest, and vertebrate such as Ondatra zibethicus, Enhydra lutris. Biological invasion prediction is apt to individual-based spatial simulation and begins to focus on the long distance diffusion prediction in current time. The R-D model is unable to simulate the spatial dispersal and it fails to address the interaction of ecological processes in application. However, R-D model is helpful in understanding the temporal invasion pattern and it provides the basic evaluation for biological invasion research. R-D model had made many contributions to the invasion theory in the past and it continues to be a simple tool of predicting biological invasion in the future.

关 键 词：反应-扩散模型 入侵预测 生物入侵 入侵管理 生态环境 

分 类 号：X17[环境科学与工程—环境科学]