中文摘要：

本研究提出了用研究复合种群的关系研究群落的方法,将花椒园中昆虫群落按其与花椒的利害关系分成不同的复合种群的功能团,建立多复合种群Latka Volterra模型,并对模型结合试验数据进行求解,其目的是为解决Lotka Volterra生态模型中参数确定困难的问题,并提供一种将模型与调查试验数据结合起来进行模拟分析和检验的解决思路。结果显示,利用差分方程确定模型系数的初值,然后借助计算机软件Matlab的微分方程求解功能和参数拟合功能,再对系数进行优化确定,从而确定花椒园中各复合种群的关系模型,并量化了各复合种群功能团间的关系,所得拟合值与调查值残差平方和较小,拟合效果较好。另外,本研究在群落层面以复合种群研究整个群落,得到植食性、捕食性、寄生性和中性昆虫复合种群功能团间的关系与一般研究种群设定在“理想环境”中研究较为吻合,表明以复合种群研究群落可规避许多干扰因素,其研究结果较为理想。

英文摘要：

This paper presents a method for studying community by using the metapopulation relation- ship, and gives an example for its use in a study of the insects in a Zanthoxylum bungeanum garden. Based on the power of relations with Z. bungeanum, the insect communities in the Z. bungeamtm gar- den were divided into different functional groups within metapopulations. A Latka-Volterra model was es- tablished for the muhi-metapopulations, and the initial coefficients of the model were determined using a differential equation. The purpose of the study is to solve the parameters of the Lotka-Volterra model to determine the ecological problems, as well as for the previous few experimental data model and investiga- tion of combined to carry out simulation analysis and test questions provide a solution. The differential e- quations were analyzed using Matlab software to determine the functions and parameters for further opti- mizing of the coefficient, and eventually quantifying the relationship between the various functional groups within the insect metapopulation in the Z. bungeanum garden. We found that the parasitic insectpopulations are the predominant insect metapopulation in the Z. bungeanum garden. The sum of the squared residuals was small, and the fitting effect was good with given values. The entire community was also studied at the community level with the metapopulation method, and the results were consistent with those values in ideal environment conditions in general population studies. This indicates that community studies using the metapopulation method can avoid many confounding factors with desirable efficiency.