Hodgson Jenny A, Moilanen Atte, Thomas Chris D
Department of Biology, University of York, York YO10 5YW, United Kingdom.
Ecology. 2009 Jun;90(6):1608-19. doi: 10.1890/08-1227.1.
Many species have to track changes in the spatial distribution of suitable habitat from generation to generation. Understanding the dynamics of such species will likely require spatially explicit models, and patch-based metapopulation models are potentially appropriate. However, relatively little attention has been paid to developing metapopulation models that include habitat dynamics, and very little to testing the predictions of these models. We tested three predictions from theory about the differences between dynamic habitat metapopulations and their static counterparts using long-term survey data from two metapopulations of the butterfly Plebejus argus. As predicted, we showed first that the metapopulation inhabiting dynamic habitat had a lower level of habitat occupancy, which could not be accounted for by other differences between the metapopulations. Secondly, we found that patch occupancy did not significantly increase with increasing patch connectivity in dynamic habitat, whereas there was a strong positive connectivity-occupancy relationship in static habitat. Thirdly, we found no significant relationship between patch occupancy and patch quality in dynamic habitat, whereas there was a strong, positive quality-occupancy relationship in static habitat. Modeling confirmed that the differences in mean patch occupancy and connectivity-occupancy slope could arise without changing the species' metapopulation parameters-importantly, without changing the dependence of colonization upon connectivity. We found that, for a range of landscape scenarios, successional simulations always produced a lower connectivity-occupancy slope than comparable simulations with static patches, whether compared like-for-like or controlling for mean occupancy. We conclude that landscape-scale studies may often underestimate the importance of connectivity for species occurrence and persistence because habitat turnover can obscure the connectivity-occupancy relationship in commonly available snapshot data.
许多物种必须世代追踪适宜栖息地空间分布的变化。要理解这类物种的动态变化，可能需要空间明确的模型，基于斑块的集合种群模型可能是合适的。然而，在开发包含栖息地动态的集合种群模型方面，相对较少受到关注，而对这些模型预测的检验则更少。我们利用来自两种银斑弄蝶集合种群的长期调查数据，检验了理论上关于动态栖息地集合种群与其静态对应物之间差异的三个预测。正如预测的那样，我们首先表明，栖息在动态栖息地的集合种群的栖息地占有率较低，而这无法用集合种群之间的其他差异来解释。其次，我们发现，在动态栖息地中，斑块占有率并未随着斑块连通性的增加而显著增加，而在静态栖息地中，连通性与占有率之间存在很强的正相关关系。第三，我们发现，在动态栖息地中，斑块占有率与斑块质量之间没有显著关系，而在静态栖息地中，质量与占有率之间存在很强的正相关关系。建模证实，平均斑块占有率和连通性 - 占有率斜率的差异可能会出现，而无需改变物种的集合种群参数——重要的是，无需改变定殖对连通性的依赖性。我们发现，对于一系列景观情景，演替模拟总是比具有静态斑块的可比模拟产生更低的连通性 - 占有率斜率，无论是同类比较还是控制平均占有率。我们得出结论，景观尺度的研究可能常常低估连通性对物种出现和持续存在的重要性，因为栖息地更替可能会掩盖常见快照数据中连通性与占有率之间的关系。
