Assessing the strength and sensitivity of the core microbiota approach on a highly diverse sponge reef

Marine sponge reefs usually comprise a complex array of taxonomically different sponge species, many of these hosting highly diverse microbial communities. The number of microbial species known to occupy a given sponge ranges from tens to thousands, bringing numerous challenges to their analysis. One way to deal with such complexity is to use a core microbiota approach, in which only prevalent and abundant microbes are considered. Here we aimed to test the strength and sensitivity of the core microbiota approach by applying different core definitions to 20 host sponge species. Application of increasingly stringent relative abundance and/or percentage occurrence thresholds to qualify as part of the core microbiota decreased the number of 'core' OTUs and phyla and, consequently, changed both alpha- and beta-diversity patterns. Moreover, microbial co-occurrence patterns explored using correlation networks were also affected by the core microbiota definition. The application of stricter thresholds resulted in smaller and less compartmentalized networks, with different keystone species. These results highlight that the application of different core definitions to phylogenetically disparate host species can result in the drawing of markedly different conclusions. Consequently, we recommend to assess the effects of different core community definitions on the specific system of study before considering its application.