Data from: Community trees: identifying codiversification in the páramo dipteran community

Groups of codistributed species that responded in a concerted manner to environmental events are expected to share patterns of evolutionary diversification. However, the identification of such groups has largely been based on qualitative, post hoc analyses. We develop here two methods (PPS, K-F ANOVA) for the analysis of codistributed species that, given a group of species with a shared pattern of diversification, allow empiricists to identify those taxa that do not codiversify (i.e., "outlier" species). The identification of outlier species makes it possible to jointly estimate the evolutionary history of co-diversifying taxa. To evaluate the approaches presented here, we collected data from Páramo dipterans, identified outlier species, and estimated a "community tree" from species that are identified as having co-diversified. Our results demonstrate that dipteran communities from different Páramo habitats in the same mountain range are more closely related than communities in other ranges. We also conduct simulation testing to evaluate this approach. Results suggest that our approach provides a useful addition to comparative phylogeographic methods, while identifying aspects of the analysis that require careful interpretation. In particular, both the PPS and K-F ANOVA perform acceptably when there are one or two outlier species, but less so as the number of outliers increase. This is likely a function of the corresponding degradation of the signal of community divergence; without a strong signal from a co-diversifying community, there is no dominant pattern from which to detect and outlier species. For this reason, both the magnitude of K-F distance distribution and outside knowledge about the phylogeographic history of each putative member of the community should be considered when interpreting results.对环境事件做出协同响应的共分布物种（codistributed species）类群，其进化分化（evolutionary diversification）模式理应趋于一致。然而，此类类群的识别迄今大多基于定性的事后分析（post hoc analyses）。本文开发了两种用于共分布物种类群分析的方法（PPS、K-F ANOVA）：当给定一组具有共同分化模式的物种类群时，该方法可帮助实证研究者识别未发生共分化的类群，即“异常物种（outlier species）”。通过识别异常物种，研究者得以联合估算共分化类群（co-diversifying taxa）的演化历史。为评估本文提出的方法，我们收集了帕拉莫（Páramo）双翅目昆虫（dipterans）的相关数据，识别出其中的异常物种，并基于被判定为共分化的类群构建了“群落演化树（community tree）”。研究结果显示，同一山脉不同帕拉莫生境中的双翅目群落，相较于其他山脉的群落，亲缘关系更为紧密。此外，我们还通过模拟测试对该方法进行了验证：结果表明，本方法可为比较系统地理学方法（comparative phylogeographic methods）提供有益补充，同时也明确了分析中需谨慎解读的环节。具体而言，当异常物种数量为1或2时，PPS与K-F ANOVA均表现良好，但随着异常物种数量增加，其性能会有所下降。这一现象可能源于群落分化信号的相应衰减：若共分化群落缺乏强信号，则无法形成可供识别异常物种的主导模式。因此，在解读结果时，需同时考量K-F距离分布（K-F distance distribution）的幅度，以及关于群落各潜在成员系统地理学历史（phylogeographic history）的外部先验知识。