One of the central objectives of microbial ecology is to study the distribution of microbial communities and their association with their environments. Biogeographical studies have partitioned the oceans into provinces and regions, but the identification of their boundaries remains challenging, hindering our ability to study transition zones (i.e. ecotones) and microbial ecosystem heterogeneity. Fuzzy clustering is a promising method to do so, as it creates overlapping sets of clusters. The outputs of these analyses thus appear both structured (into clusters) and gradual (due to the overlaps), which aligns with the inherent continuity of the pelagic environment, and solves the issue of defining ecosystem boundaries. We show the suitability of applying fuzzy clustering to address the patchiness of microbial ecosystems, integrating environmental (Sea Surface Temperature, Salinity) and bacterioplankton data (Operational Taxonomic Units (OTUs) based on 16S rRNA gene) collected during six cruises over 1.5 years from the subtropical frontal zone off New Zealand. The technique was able to precisely identify ecological heterogeneity, distinguishing both the patches and the transitions between them. In particular we show that the subtropical front is a distinct, albeit transient, microbial ecosystem. Each water mass harboured a specific microbial community, and the characteristics of their ecotones matched the characteristics of the environmental transitions, highlighting that environmental mixing lead to community mixing. Further explorations into the OTU community compositions revealed that, although only a small proportion of the OTUs explained community variance, their associations with given water mass were consistent through time. We demonstrate recurrent associations between microbial communities and dynamic oceanic features. Fuzzy clusters can be applied to any ecosystem (terrestrial, human, marine, etc) to solve uncertainties regarding the position of microbial ecological boundaries and to refine the relation between the distribution of microorganisms and their environment.

中文翻译：

---

减少任意性：微生物生态系统和生态系统的模糊检测–集中在远洋环境

微生物生态学的主要目标之一是研究微生物群落的分布及其与环境的关系。生物地理学研究将海洋划分为多个省和地区，但是确定其边界仍然具有挑战性，这阻碍了我们研究过渡区（即过渡带）和微生物生态系统异质性的能力。模糊聚类是一种有前途的方法，因为它会创建重叠的聚类集。因此，这些分析的结果既有条理（成簇）又是渐进的（由于重叠），与远洋环境的固有连续性保持一致，解决了界定生态系统边界的问题。我们展示了应用模糊聚类解决微生物生态系统的零散性的适用性，整合了在1.5年的时间里从新西兰亚热带副热带地区进行的六次航行中收集的环境（海表温度，盐度）和浮游细菌数据（基于16S rRNA基因的操作分类单位（OTU））的数据。该技术能够精确识别生态异质性，区分斑块和它们之间的过渡。特别是，我们显示了亚热带锋面是一个独特的，尽管短暂的微生物生态系统。每个水团都有一个特定的微生物群落，其生态过渡带的特征与环境转变的特征相匹配，突显了环境混合导致群落混合。对OTU社区组成的进一步研究表明，尽管只有一小部分OTU解释了社区差异，它们与给定水量的关联在时间上是一致的。我们证明了微生物群落和动态海洋特征之间的经常性联系。模糊聚类可应用于任何生态系统（陆地，人类，海洋等），以解决有关微生物生态边界位置的不确定性，并改善微生物分布与其环境之间的关系。