Data from: Season-specific and guild-specific effects of anthropogenic landscape modification on metacommunity structure of tropical bats

1. Fragmentation per se due to human land conversion is a landscape-scale phenomenon. Accordingly, assessment of distributional patterns across a suite of potentially connected communities (i.e. metacommunity structure) is an appropriate approach for understanding the effects of landscape modification, and complements the plethora of fragmentation studies that have focused on local community structure. To date, metacommunity structure within human-modified landscapes has been assessed with regard to nestedness along species richness gradients. This is problematic because there is little support that species richness gradients are associated with the factors moulding species distributions. More importantly, many alternative patterns are possible, and different patterns may manifest during different seasons and for different guilds because of variation in resource availability and resource requirements of taxa. 2. We determined the best-fit metacommunity structure of a phyllostomid bat assemblage, frugivore ensemble, and gleaning animalivore ensemble within a human-modified landscape in the Caribbean lowlands of Costa Rica during the dry and wet seasons to elucidate important structuring mechanisms. Furthermore, we identified the landscape characteristics associated with the latent gradient underlying metacommunity structure. 3. We discriminated among multiple metacommunity structures by assessing coherence, range turnover, and boundary clumping of an ordinated site-by-species matrix. We identified the landscape characteristics associated with the latent gradient underlying metacommunity structure via hierarchical partitioning. 4. Metacommunity structure was never nested nor structured along a richness gradient. 5. The phyllostomid assemblage and frugivore ensemble exhibited Gleasonian structure (range turnover along a common gradient) during the dry season and Clementsian structure (range turnover and shared boundaries along a common gradient) during the wet season. Distance between forest patches and forest edge density structured the phyllostomid metacommunity during the dry and wet seasons, respectively. Proportion of pasture and forest patch density structured the frugivore metacommunity during the dry season. 6. Gleaning animalivores exhibited chequerboard structure (mutually exclusive species-pairs) during the dry season and random structure during the wet season. 7. Metacommunity structure was likely mediated by differential resource use or interspecific relationships. Furthermore, the interaction between landscape characteristics and seasonal variation in resources resulted in season-specific and guild-specific distributional patterns.1. 人类土地开发转换引发的生境破碎化本身属于景观尺度现象。因此，针对一系列潜在连通群落的分布格局（即集合群落（metacommunity）结构）开展评估，是理解景观改造效应的合理方法，可补充诸多聚焦局域群落结构的破碎化研究。迄今为止，针对人类改造景观内的集合群落结构的评估，多围绕物种丰富度梯度上的嵌套性展开，但该方法存在局限：现有研究几乎无法证实物种丰富度梯度与塑造物种分布的关键驱动因素存在关联。更关键的是，还存在诸多其他潜在格局，且由于不同类群的资源可获得性与资源需求存在差异，不同格局可能在不同季节、针对不同类群呈现。 2. 本研究以哥斯达黎加加勒比低地人类改造景观为研究区，针对旱季与雨季的叶口蝠科蝙蝠类群（phyllostomid bat assemblage）、食果动物类群（frugivore ensemble）以及拾食性食肉动物类群（gleaning animalivore ensemble），分别确定其最优拟合集合群落结构，以阐明群落构建的核心机制；此外，本研究还识别了支撑集合群落结构的潜在梯度所对应的景观特征。 3. 本研究通过对排序后的样地-物种类矩阵的连贯性、范围周转与边界聚集进行量化评估，实现对多种集合群落结构的判别；同时通过分层划分（hierarchical partitioning）方法，识别出与集合群落结构潜在梯度相关的景观特征。 4. 本研究区内的集合群落结构从未呈现嵌套性，也未沿物种丰富度梯度形成结构化格局。 5. 叶口蝠类群与食果动物类群在旱季表现为格莱松式结构（沿共同梯度发生范围周转），在雨季则表现为克莱门茨式结构（沿共同梯度发生范围周转且存在共享边界）。林斑块间距与林缘密度分别为调控旱季、雨季叶口蝠集合群落的关键景观因子。旱季时，牧场占比与林斑块密度为调控食果动物集合群落的关键景观因子。 6. 拾食性食肉动物类群在旱季呈现棋盘式结构（物种对间存在互斥分布），在雨季则呈现随机结构。 7. 集合群落结构大概率由差异化的资源利用策略或种间相互关系介导。此外，景观特征与资源的季节变化之间的交互作用，共同造就了季节特异性与类群特异性的分布格局。