Persistent tissue-specific resident microbiota in oysters across a broad geographical range

Abstract

Marine animals often harbour complex microbial communities that influence their physiology. However, strong evidence for resident microbiomes in marine bivalves is lacking, despite their contribution to estuarine habitats and coastal economies. We investigated whether marine bivalves harbour stable, resident microorganisms in specific tissues or if their microbiomes primarily consist of transient members reflecting the environmental microbial pool. Conducting a latitudinal study of wild eastern oysters (Crassostrea virginica) along the East Coast of the United States, we aimed to identify resident microorganisms that persist across a wide geographical range. Our results revealed that microbial communities in seawater and sediment samples followed latitudinal diversity patterns driven by geographic location. In contrast, oyster-associated microbiomes were distinct from their surrounding environments and exhibited tissue-specific compositions. Notably, oyster microbiomes showed greater similarity within the same tissue type across different geographic locations than among different tissue types within the same location. This indicates the presence of tissue-specific resident microbes that persist across large geographical ranges. We identified a persistent set of resident microbiome members for each tissue type, with key microbial members consistent across all locations. These findings underscore the oyster host's role in selecting its microbiome and highlight the importance of tissue-specific microbial communities in understanding bivalve-associated microbiomes.

FIGURE 1

Map of sampling sites and principal component analysis (PCoA) plots of Bray–Curtis dissimilarities of environmental samples. (A) Map of geographic locations of the six oyster reefs sampled on the East Coast of the United States. From North to South: Damariscotta River (44°01′38.1″ N 69°32′35.7″ W) in Maine (ME), Barnstable (41°42′37.6″ N 70°18′18.5″ W) in Massachusetts (MA), Green Hill Pond (41°22′16.1″ N 71°37′13.4″ W) in Rhode Island (RI), Horse Island (37°17′15.5″ N 75°55′02.0″ W) in Virginia (VA), Atlantic Beach (34°42′24.9″ N 76°45′05.7″ W) in North Carolina (NC), and St. Augustine (29°40′17.7″ N 81°12′53.5″ W) in Florida (FL). PCoA of Bray–Curtis Dissimilarity distances of microbial communities in (B) seawater and (C) sediment samples. Sampling geographic locations are indicated across all plots with dots coloured coded in a rainbow gradient based on latitude.

FIGURE 2

Scatter plot showing the relationship between Bray–Curtis dissimilarity distances along the first principal coordinates axis and latitude of (A) seawater and (B) sediment microbial communities in the sampled geographic locations.

FIGURE 3

Bray–Curtis dissimilarities of microbial communities associated with oyster tissues and their surrounding seawater and sediment, visualized with principal coordinate analysis (PCoA) (A) across all sample types, and specific to oyster tissues (B) coloured by geographic location or (D) tissue type. As well as a (C) scatterplot showing no relationship between Bray–Curtis dissimilarity distances along the first principal coordinates axis and latitude of tissue‐associated microbial communities. Finally, (E) violin plot showing oyster‐associated microbial communities compared with the same tissue type across all geographic locations (e.g., gill from ME compared with gill from FL) and compared with different tissue types within the same geographic location (e.g., gill form ME compared with stomach from ME). Large dots inside the violin plot indicate group means.

FIGURE 4

Heat map of the 30 most important amplicon sequence variants (ASVs) in contributing to a Random Forest classification model trained to predict tissue type from microbial community composition. The heat map shows the relative abundances of ASVs in samples of different tissue types, samples are clustered using Bray–Curtis dissimilarity distances. More detailed taxonomic information can be found in Table S5.

FIGURE 5

Stacked bar plot of the relative abundance of amplicon sequence variants (ASVs) identified as core members, present at >1% abundance in more than 50% of samples, in the gill (shades of purple n = 5), mantle (yellow n = 1), and stomach (shades of green n = 5) microbiomes. More detailed taxonomic information can be found in Table S6.