Microbiomes from feces vs. gut in aquatic vertebrates: distinct community compositions between substrates and preservation methods

Abstract

Sample type and preservation methods are likely to influence microbiome analysis results. Relatively few studies have explored the differences between feces and gut as well as ethanol-stored and frozen samples. Here, we sampled the same individuals of three aquatic vertebrates from the Qinghai-Tibetan Plateau non-invasively for feces, and subsequently for hindgut through dissection. Our study species, two fishes (Gymnocypris cf. namensis and Triplophysa sp.) and one amphibian (tadpoles of Nanorana parkeri), were all collected at the same time and site. Gut and fecal samples were stored in ethanol, and additionally, part of the gut samples were frozen, but temporarily thawed during transport as it often happens under difficult field conditions. Our results showed that both substrate (gut content vs. feces) as well as preservation method can influence the analysis of intestinal microbiomes. Frozen gut samples strongly differed from ethanol-stored samples, and especially in Nanorana most frozen samples were dominated (in relative abundance) by a set of Proteobacteria OTUs that were completely absent from the ethanol-stored samples. This blooming of contaminant bacteria occurred after less than 12 h of thawing, thus caution should be taken when constancy of cold temperatures cannot be maintained in the field for sample preservation purposes. Among ethanol-stored samples, bacterial communities from feces differed from those recovered from guts, but in part recovered similar patterns, such as a higher bacterial richness in the more herbivorous Nanorana tadpoles. Although our results argue against combining gut and fecal samples in analyses of host-specific microbiome differences, they also confirm that non-invasive sampling of feces can provide useful information of gut microbiomes in aquatic vertebrates, which may be important especially when working with endangered species.