The piranha gut microbiome provides a selective lens into river water biodiversity

Abstract

Advances in omics technologies have enabled the in-depth study of microbial communities and their metabolic profiles from all environments. Here metagenomes were sampled from piranha (Serrasalmus rhombeus) and from river water from the Rio São Benedito (Amazon Basin). Shotgun metagenome sequencing was used to explore diversity and to test whether fish microbiomes are a good proxy for river microbiome studies. The results showed that the fish microbiomes were not significantly different from the river water microbiomes at higher taxonomic ranks. However, at the genus level, fish microbiome alpha diversity decreased, and beta diversity increased. This result repeated for functional gene abundances associated with specific metabolic categories (SEED level 3). A clear delineation between water and fish was seen for beta diversity. The piranha microbiome provides a good and representative subset of its river water microbiome. Variations seen in beta biodiversity were expected and can be explained by temporal variations in the fish microbiome in response to stronger selective forces on its biodiversity. Metagenome assembled genomes construction was better from the fish samples. This study has revealed that the microbiome of a piranha tells us a lot about its river water microbiome and function.

Keywords: Serrasalmus rhombeus; Amazon river microbiome; Biotechnology; Host-associated microbiome; Metagenomics.

Fig. 1

Photos of fish and river sampling on the Rio São Benedito, located in the Amazon Basin in Pará, Brazil. (a–c) Example specimens of the endemic Redeye Piranha (Serrasalmus rhombeus) sampled in the study. (d) Demonstration of river water sampling using a Sterivex filter unit.

Fig. 2

Taxonomic profile of the Rio São Benedito fish and water metagenomes. The graph shows distribution of the 15 most abundant prokaryotic genera from anal swab and water samples and their respective phyla.

Fig. 3

Diversity measures of Rio São Benedito water and piranha anal swab samples at the genus level for prokaryotic communities. (a) Alpha diversity represented by Shannon Index showing significant differences between anal swab and river samples (one-way ANOVA, P < 0.001). (b) Beta diversity represented by principal component analysis plot showing differences in the structure of prokaryotic communities from anal swab and river water samples. The axes show the Principal Components (PC), with PC1 and PC2 respectively explaining 65.5% and 22.8% of the total variance. Only entries where at least 10 reads were classified within a given sample from either swab or water metagenomes were used for diversity measures. Venn diagram representing unique and common counts of prokaryotic genera between (c) anal swabs, (d) water and (e) both sampling groups overall. For Venn Diagram counts, only entries where at least 10 counts were classified to a genus within an individual swab or river sample were included.

Fig. 4

Differential abundance of metabolic genes from the anal swab and river water metagenomes. Differences in relative abundance of metabolic gene groups categorized at SEED Level 1 were visualized via (a) heatmap of all classified functional genes with (b) the significant differences in differential abundance displayed as a bar plot with error bars (two-sided Welch’s t test, with Benjamin-Hochberg method for false discovery rate, q < 0.05).

Fig. 5

Diversity and abundance measures of Rio São Benedito water and piranha anal swab samples for functional genes at SEED level 3. (a) Alpha diversity represented by Shannon Index between anal swab and river samples. (b) Beta diversity represented by principal component analysis plot showing differences in the community metabolic structure from anal swab and river water samples. The axes show the Principal Components (PC), with PC1 and PC2 respectively explaining 72.5% and 11.9% of the total variance. Only entries where at least 10 reads were classified within a given sample from either swab or water metagenomes were used for diversity measures. Venn diagram representing unique and common counts of functional genes categorized at SEED level 3 between (c) anal swabs, (d) water and (e) both sampling groups overall. For Venn Diagram counts, only entries where at least 10 counts were classified to a category within an individual swab or river sample were included.