Department of Ecology, Evolution, and Organismal Biology, Iowa State University Ames, IA, USA.
Mathematics and Computer Science, Argonne National Laboratory Argonne, IL, USA ; Microbiology and Microbial Genetics, Michigan State University East Lansing, MI, USA.
Front Microbiol. 2014 Jul 18;5:358. doi: 10.3389/fmicb.2014.00358. eCollection 2014.
Co-occurrence patterns are used in ecology to explore interactions between organisms and environmental effects on coexistence within biological communities. Analysis of co-occurrence patterns among microbial communities has ranged from simple pairwise comparisons between all community members to direct hypothesis testing between focal species. However, co-occurrence patterns are rarely studied across multiple ecosystems or multiple scales of biological organization within the same study. Here we outline an approach to produce co-occurrence analyses that are focused at three different scales: co-occurrence patterns between ecosystems at the community scale, modules of co-occurring microorganisms within communities, and co-occurring pairs within modules that are nested within microbial communities. To demonstrate our co-occurrence analysis approach, we gathered publicly available 16S rRNA amplicon datasets to compare and contrast microbial co-occurrence at different taxonomic levels across different ecosystems. We found differences in community composition and co-occurrence that reflect environmental filtering at the community scale and consistent pairwise occurrences that may be used to infer ecological traits about poorly understood microbial taxa. However, we also found that conclusions derived from applying network statistics to microbial relationships can vary depending on the taxonomic level chosen and criteria used to build co-occurrence networks. We present our statistical analysis and code for public use in analysis of co-occurrence patterns across microbial communities.
共现模式在生态学中被用来探索生物之间的相互作用以及环境对生物群落共存的影响。微生物群落共现模式的分析范围从所有群落成员之间的简单两两比较到焦点物种之间的直接假设检验。然而，在同一研究中，共现模式很少在多个生态系统或多个生物组织尺度上进行研究。在这里，我们概述了一种产生共现分析的方法，该方法侧重于三个不同的尺度：群落尺度上的生态系统之间的共现模式、群落内共现微生物的模块以及嵌套在微生物群落内的模块内的共现对。为了展示我们的共现分析方法，我们收集了公开可用的 16S rRNA 扩增子数据集，以比较和对比不同生态系统中不同分类水平的微生物共现。我们发现了反映社区尺度环境过滤的群落组成和共现差异，以及可能用于推断关于了解甚少的微生物类群生态特征的一致的成对出现。然而，我们还发现，应用网络统计来分析微生物关系的结论可能会因选择的分类水平和构建共现网络所使用的标准而有所不同。我们提出了我们的统计分析和代码，供公众在分析微生物群落中的共现模式时使用。
