Heino Jani, Alahuhta Janne
Finnish Environment Institute, Natural Environment Centre, Biodiversity, PO Box 413, FI-90014, Oulu, Finland.
Department of Geography, University of Oulu, PO Box 3000, FI-90014, Oulu, Finland.
J Anim Ecol. 2015 Mar;84(2):427-41. doi: 10.1111/1365-2656.12287. Epub 2014 Oct 20.
Regional faunas are structured by historical, spatial and environmental factors. We studied large-scale variation in four ecologically different beetle groups (Coleoptera: Dytiscidae, Carabidae, Hydrophiloidea, Cerambycidae) along climate, land cover and geographical gradients, examined faunal breakpoints in relation to environmental variables, and investigated the best fit pattern of assemblage variation (i.e. randomness, checkerboards, nestedness, evenly spaced, Gleasonian, Clementsian). We applied statistical methods typically used in the analysis of local ecological communities to provide novel insights into faunal compositional patterns at large spatial grain and geographical extent. We found that spatially structured variation in climate and land cover accounted for most variation in each beetle group in partial redundancy analyses, whereas the individual effect of each explanatory variable group was generally much less important in accounting for variation in provincial species composition. We also found that climate variables were most strongly associated with faunal breakpoints, with temperature-related variables alone accounting for about 20% of variation at the first node of multivariate regression tree for each beetle group. The existence of faunal breakpoints was also shown by the 'elements of faunal structure' analyses, which suggested Clementsian gradients across the provinces, that is, that there were two or more clear groups of species responding similarly to the underlying ecological gradients. The four beetle groups showed highly similar biogeographical patterns across our study area. The fact that temperature was related to faunal breakpoints in the species composition of each beetle group suggests that climate sets a strong filter to the distributions of species at this combination of spatial grain and spatial extent. This finding held true despite the ecological differences among the four beetle groups, ranging from fully aquatic to fully terrestrial and from herbivorous to predaceous species. The existence of Clementsian gradients may be a common phenomenon at large scales, and it is likely to be caused by crossing multiple species pools determined by climatic and historical factors on the distributions of species.
区域动物区系由历史、空间和环境因素构成。我们研究了四个生态不同的甲虫类群（鞘翅目：龙虱科、步甲科、水龟虫总科、天牛科）在气候、土地覆盖和地理梯度上的大规模变化，研究了与环境变量相关的动物区系断点，并调查了群落变化的最佳拟合模式（即随机性、棋盘状、嵌套性、均匀间隔、格里森式、克莱门茨式）。我们应用了通常用于分析局部生态群落的统计方法，以提供关于大空间粒度和地理范围内动物区系组成模式的新见解。在偏冗余分析中，我们发现气候和土地覆盖的空间结构变化占每个甲虫类群变异的大部分，而每个解释变量组对省级物种组成变异的个体影响通常要小得多。我们还发现气候变量与动物区系断点的关联最为强烈，仅温度相关变量就占每个甲虫类群多元回归树第一个节点变异的约20%。“动物区系结构要素”分析也表明了动物区系断点的存在，这表明各省存在克莱门茨式梯度，即有两个或更多明显的物种组对潜在生态梯度有相似反应。在我们的研究区域内，这四个甲虫类群呈现出高度相似的生物地理模式。温度与每个甲虫类群物种组成中的动物区系断点相关这一事实表明，在这种空间粒度和空间范围的组合下，气候对物种分布设置了强大的筛选作用。尽管这四个甲虫类群在生态上存在差异，从完全水生到完全陆生，从食草到食肉物种，但这一发现仍然成立。克莱门茨式梯度的存在可能是大尺度上的普遍现象，很可能是由气候和历史因素决定的多个物种库交叉对物种分布造成的。
