9% of the overall variation (P < 0.001 based on 1000 permutations). In concordance with the expectation of random sampling before treatment assignment we found no significant difference between “ambient” and “disturbed” oysters in terms
of their genetic variation (R2 = 0.031, P = 0.159 based on 1000 permutations) and no significant interaction effect (R2 = 0.053, P = 0.257 based on 1000 permutations). Due to high within locus polymorphism the majority of variation was found among individuals (R2 = 0.797). Microbial AZD4547 supplier communities of oysters before and after disturbance Out of the 52,092 reads that could successfully be assigned to an amplicon library for each individual, 38,029 reads passed our quality selection and de-noising criteria for further analysis. The resulting average library size per individual
was 825 ± 80. With a total number of 4,464 unique operational taxonomic units (OTUs) 4SC-202 manufacturer distributed over 213 genera, microbial species richness was very high. However, only few OTUs occurred frequently and most OTUs occurred rarely (<1% within whole data set). After rigorous de-noising of our sequencing data we potentially underestimated species richness of the respective communities, but we could reliably calculate diversity 3Methyladenine (Shannon’s H’) for most experimental groups (Figure 2A). Microbial diversity was significantly lower in oysters originating from DB (GLM, F2,36 = 3.55, P = 0.039) especially under ambient conditions (Figure 2A,B). The disturbance treatment led to a significant decrease of bacterial diversity in oysters from all beds (Figure 2B, disturbance: GLM F1,36 = 7.52, P = 0.009, disturbance × oyster bed interaction: F2,36 = 0.80, P = 0.456). Figure 2 Bacterial diversity (Shannon’s H’) of oyster gill microbiota stemming from different oyster beds. A) Rarefaction curves of Shannon’s H’ in different oyster beds under ambient field conditions and after disturbance. Shown are rarefied means for treatment and origin groups from 10 resamples with a maximum number corresponding to the lowest coverage of a single microbiome in each group.
Solid lines represent ambient conditions and dashed lines disturbed microbial communities. Amino acid B) Observed values of Shannon’s H’ for individual oysters stemming from different oyster beds (mean ± se) showing significant differences between oyster beds (F2,36 = 3.55, P = 0.039) and a significant decrease of diversity after disturbance (F1,36 = 7.52, P = 0.009). Non-metric multidimensional scaling of the full bacterial communities from individual oysters suggested that communities were differentiated by treatment along both axes (Figure 3), which was confirmed by Permanova (effect of disturbance: R2 = 0.077, P = 0.006). Clustering of ambient group centroids in the ordination suggests that initially there was no significant difference between beds and large variation within beds under ambient conditions (Figure 3, Permanova, effect of oyster bed: R2 = 0.058, P = 0.211).