基于元胞自动机的多尺度草地蝗虫复合种群动态模拟

With the influence of climate warming and grassland degradation, there has been grasshopper plague in Xilinguole League, Inner Mongolia in the past ten years. The plague has seriously affected grassland ecosystems, agricultural and grazing production and local people's lives. Therefore, it's very necessary and important to precisely monitor and predict the spatial distribution and effects of potential occurrence and dispersal of grasshoppers to put forward effective preventive and control measures. After evaluating the climatic and habitat suitability for potential occurrence of grasshoppers in Xianghuangqi county, Xilinguole League and obtaining the final potential occurrence ranks, we will try to build a simulation system based on cellular automata to model multi-scale dynamics of grasshopper metapopulation. Based on biological characteristics and behaviors of grasshoppers in inhabiting, diet and movement in a heterogeneous landscape, the system is to be used to simulate the dispersal processes of grasshoppers affected by spatial patterns among different cells, among different lower-level sub-patches, and among different sub-patches within a grasshopper metapopulation at four spatial scales, respectively, in each growing period from third-instar larvae to adults. The possibility, intensity, and spatial ranges of occurrence and dispersal of grasshoppers and thus the losses to plants can be derived from the simulations. The complicated dynamics of grasshopper metapopulation will be expected to be clarified by studying the interactions between spatial pattern and dispersal process at multiple scales. The study may provide a new approach to predict occurrence and dispersal of grasshoppers, and suggest a new perspective from landscape ecology to prevent and control grasshopper plague by altering spatial patterns of landscape.

内蒙古锡林郭勒盟草地蝗虫已连续十年暴发成灾，对草地生态系统、牧区人民的生产和生活构成严重威胁。准确地监测和预警蝗虫的发生和扩散有利于组织防控，并提高防控效果。本研究在评价锡林郭勒盟镶黄旗草地蝗虫发生的气象和生境适宜性，获取蝗虫潜在发生源地的基础上，拟创新性地将元胞自动机融入到蝗虫复合种群动态模型的构建，对蝗虫的动态进行自下而上的多尺度空间显式模拟。该模型利用蝗虫在异质景观中的栖息、觅食和迁移特点，从元胞、次级亚种群斑块、亚种群斑块、复合种群景观这四个尺度上，模拟蝗虫在从三龄蝗蝻到成虫的每个时段，在元胞之间、次级亚种群斑块之间、亚种群斑块之间扩散迁移的动态变化，以预测不同时期蝗虫发生和扩散的可能性、强度、空间范围和危害损失。拟通过多尺度的格局-过程模拟揭示草地蝗虫复合种群的复杂动态，为蝗虫发生和扩散的预测提供一种新途径，并从改变空间格局的景观生态学角度为蝗灾的防控提供一种新思路。

内蒙古锡林郭勒盟草地蝗虫连续十多年暴发成灾，对草地生态系统、牧区人民的生产和生活构成了严重威胁。准确监测和预警蝗虫的发生和扩散有利于提高防控效果。本研究进行了连续4年的野外调查。在此基础上，逐步完善草地蝗虫生境适宜性模糊评价模型系统；并融合气象适宜性评价结果，探讨了亚洲小车蝗潜在发生的时空分布及其与生境因子和气象因子的关系。同时，探讨了景观镶嵌体的多尺度空间格局、亚洲小车蝗的多尺度分布格局，以及蝗虫密度与生境因子的局域空间关系。结果表明，所构建的生境适宜性评价模型具有可靠性，获得的生境适宜性评分值可用来指示蝗虫的潜在发生。亚洲小车蝗潜在发生源地的空间异质性受生境因子（地形、土壤、植被）的影响很大，且不同生境因子的两两交互作用对蝗虫密度的影响均比单个生境因子的影响大。除全局影响之外，生境因子（特别是土壤含水量和坡度）的局域变化也会对蝗虫密度产生很大影响。亚洲小车蝗在<0.3 km、<0.7–1.0 km和<1.9 km这3个尺度域上出现斑块化分布，这是不同尺度上相关生境因子共同作用的结果。蝗虫潜在发生的年际变化与植被盖度和和气象因子的年际变化相关。在获取蝗虫潜在发生源地的基础上，融合亚洲小车蝗的生物学习性、植物-蝗虫的相互作用机理，先后构建了元胞自动机模型、基于元胞自动机的多尺度复合种群模型、基于元胞自动机和最优路径的复合种群网络模型，逐步完善种群动态模型，以模拟草地蝗虫从蝗蝻孵化到成虫死亡的整个发育期，在元胞内部，以及在元胞之间和生境斑块之间扩散迁移的动态变化。模型适用于典型半干旱草原蝗虫密度为3–12 ind·m-2的情形。已被应用于生境斑块、景观和整个镶黄旗，蝗虫密度时空动态的模拟结果已获得验证。本研究将复合种群途径引入草地蝗虫研究，将元胞自动机的思想和方法融入到复合种群动态模型的构建中，实现元胞-斑块-景观-区域的多尺度格局-过程关系模拟，这在思想理论上是一次创新性尝试；在实践中，可预测蝗虫发生和扩散的时空动态，并为蝗灾的防控提供重要依据。