The role of macrobiota in structuring microbial communities along rocky shores

Abstract

Rocky shore microbial diversity presents an excellent system to test for microbial habitat specificity or generality, enabling us to decipher how common macrobiota shape microbial community structure. At two coastal locations in the northeast Pacific Ocean, we show that microbial composition was significantly different between inert surfaces, the biogenic surfaces that included rocky shore animals and an alga, and the water column plankton. While all sampled entities had a core of common OTUs, rare OTUs drove differences among biotic and abiotic substrates. For the mussel Mytilus californianus, the shell surface harbored greater alpha diversity compared to internal tissues of the gill and siphon. Strikingly, a 7-year experimental removal of this mussel from tidepools did not significantly alter the microbial community structure of microbes associated with inert surfaces when compared with unmanipulated tidepools. However, bacterial taxa associated with nitrate reduction had greater relative abundance with mussels present, suggesting an impact of increased animal-derived nitrogen on a subset of microbial metabolism. Because the presence of mussels did not affect the structure and diversity of the microbial community on adjacent inert substrates, microbes in this rocky shore environment may be predominantly affected through direct physical association with macrobiota.

Keywords: 16S; Ammonium; Animal excretion; Host-microbe; Mytilus californianus; Nitrification; Nitrogen cycling; Rocky intertidal; Tatoosh Island; Tidepool.

Figure 1. The proportional representation of OTUs and the mean observed OTU richness among substrates sampled.

The proportional representation of OTUs among the major microbial groups (colored bars), with the overall mean observed OTU richness (+SE) among all substrate types at (A) Tatoosh Island, and (B) in tidepools where natural rock substrate and coverslips were sampled in the context of an experimental removal of mussels. The substrates in (A) showed significant differences in observed richness (ANOVA, F_7,16 = 4.968, p = 0.004) with rocks (n = 3), crucible lids (n = 9) and filtered plankton (n = 3) showing the greatest richness while the lowest observed richness was associated with mussel gill (n = 3) and siphon (n = 1) tissue. OTU richness of mussel shell (n = 2), anemone (n = 2), and red algae (n = 2) was intermediate to the others. In (B) Tidepools with mussels removed had greater OTU richness than those with mussels (Two-Way ANOVA, F_1,18 = 12.759, p = 0.002) while rock had over twice the OTU richness of coverslips (F_1,18 = 140.59, p < 0.001); there was no interaction between substrate and mussel presence.

Figure 2. A PCoA of the OTU beta diversity of substrates on Tatoosh Island.

A PCoA of the OTU beta diversity of substrates on Tatoosh Island, demonstrating the clustering among the different microbial assemblages associated with each substrate. The weighted UniFrac metric was used to incorporate relative abundance; the first axis explained 40.2% of the variance, while the second explained 14.8%. Differences among substrates were significant (PERMANOVA, F_5,18 = 6.570, p < 0.001), and groupings that included anemone, Prionitis, mussel shell, mussel tissue, and inert substrates were differentiated while plankton were indistinguishable from all.

Figure 3. The relative abundance of the 10 OTUs that differed among the Tatoosh Island substrates.

(Boniferroni-corrected ANOVA, p < 0.05).

Figure 4. Shared OTU diversity among microbes sampled from the substrate groupings at Tatoosh Island and portrayed as a spring-embedded layout.

Shared OTU diversity among microbes sampled from the substrate groupings as in Figs. 1–3 at Tatoosh Island and portrayed as a spring-embedded layout, where OTUs that are in common bring nodes or samples together and OTUs that are distinct repel nodes. In (A) only common OTUs detected more than 5,000 times are included, while (B) shows only rare OTUs that were present 5–10 times across the entire dataset.

Figure 5. A PCoA of the OTU diversity of tidepool rock versus coverslip substrates at Second Beach, WA.

A PCoA of the OTU diversity of tidepool rock (n = 10) versus coverslip (n = 12) substrates at Second Beach, demonstrating strong clustering among the microbial assemblages from the two substrates, while the presence of mussels (filled symbols) versus removal of mussels (open symbols) were not a factor for explaining beta diversity. Using weighted UniFrac, the first axis explained 46.5% of the variance, while the second explained 20.3%.

Figure 6. Shared OTU diversity among microbes sampled from tidepool rock versus coverslip substrates in tidepools at Second Beach, WA.

Shared OTU diversity among (A) microbes sampled from tidepool rock versus coverslip (lighter green) substrates and (B) samples distinguished by whether mussels were present (blue) or absent (red) from tidepools at Second Beach. The spring-embedded layout shows OTUs that are in common bring nodes or samples together and OTUs that are distinct repel nodes. Only common OTUs greater than >5,000 are included. Analyses of relatively rare OTUs did not change the network pattern.