Population Ecology Group, Institut Mediterrani d'Estudis Avançats (CSIC-UIB), Miquel Marqués 21, Esporles, 07190, Spain.
J Anim Ecol. 2014 Jan;83(1):276-85. doi: 10.1111/1365-2656.12118. Epub 2013 Aug 19.
Many species are found today in the form of fragmented populations occupying patches of remnant habitat in human-altered landscapes. The persistence of these population networks requires a balance between extinction and colonization events assumed to be primarily related to patch area and isolation, but the contribution of factors such as the characteristics of patch and matrix habitats, the species' traits (habitat specialization and dispersal capabilities) and variation in climatic conditions have seldom been evaluated simultaneously. The identification of environmental variables associated with patch occupancy and turnover may be especially useful to enhance the persistence of multiple species under current global change. However, for robust inference on occupancy and related parameters, we must account for detection errors, a commonly overlooked problem that leads to biased estimates and misleading conclusions about population dynamics. Here, we provide direct empirical evidence of the effects of different environmental variables on the extinction and colonization rates of a rich butterfly community in the western Mediterranean. The analysis was based on a 17-year data set containing detection/nondetection data on 73 butterfly species for 26 sites in north-eastern Spain. Using multiseason occupancy models, which take into account species' detectability, we were able to obtain robust estimates of local extinction and colonization probabilities for each species and test the potential effects of site covariates such as the area of suitable habitat, topographic variability, landscape permeability around the site and climatic variability in aridity conditions. Results revealed a general pattern across species with local habitat composition and landscape features as stronger predictors of occupancy dynamics compared with topography and local aridity. Increasing area of suitable habitat in a site strongly decreased local extinction risks and, for a number of species, both higher amounts of suitable habitat and more permeable landscapes increased colonization rates. Nevertheless, increased topographic variability decreased the extinction risk of bad dispersers, a group of species with significantly lower colonization rates. Our models predicted higher sensitivity of the butterfly assemblages to deterministic changes in habitat features rather than to stochastic weather patterns, with some relationships being clearly dependent on the species' traits.
如今,许多物种都以碎片化的种群形式存在,这些种群分布在人类改造景观中的残余栖息地斑块中。这些种群网络的持续存在需要在灭绝和定居事件之间取得平衡,这些事件主要与斑块面积和隔离有关,但斑块和基质栖息地的特征、物种特征(栖息地特化和扩散能力)以及气候条件的变化等因素的贡献很少同时得到评估。确定与斑块占有和周转率相关的环境变量可能特别有助于在当前全球变化下增强多种物种的持续存在。然而,为了对占有率和相关参数进行稳健推断,我们必须考虑到检测误差,这是一个经常被忽视的问题,会导致对种群动态的有偏差估计和误导性结论。在这里,我们提供了直接的经验证据,证明了不同环境变量对西地中海丰富蝴蝶群落灭绝和定居率的影响。该分析基于包含 73 种蝴蝶物种在西班牙东北部 26 个地点的 17 年检测/未检测数据的数据集。使用多季节占有模型,该模型考虑了物种的可检测性,我们能够为每个物种获得可靠的局部灭绝和定居概率估计,并测试了站点协变量(如适宜栖息地面积、地形变异性、站点周围景观渗透性和干旱条件下气候变异性)的潜在影响。结果显示,物种之间存在一般模式,局部栖息地组成和景观特征是比地形和局部干旱更能预测占有动态的因素。站点内适宜栖息地面积的增加强烈降低了局部灭绝风险,并且对于许多物种来说,更多的适宜栖息地和更具渗透性的景观都增加了定居率。然而,地形变异性的增加降低了扩散能力较差的物种的灭绝风险,这是一组具有显著较低定居率的物种。我们的模型预测蝴蝶组合对栖息地特征的确定性变化更为敏感,而不是对随机天气模式敏感,其中一些关系显然取决于物种的特征。