Spatial and Temporal Dynamics of Pacific Oyster Hemolymph Microbiota across Multiple Scales

Abstract

Unveiling the factors and processes that shape the dynamics of host associated microbial communities (microbiota) under natural conditions is an important part of understanding and predicting an organism's response to a changing environment. The microbiota is shaped by host (i.e., genetic) factors as well as by the biotic and abiotic environment. Studying natural variation of microbial community composition in multiple host genetic backgrounds across spatial as well as temporal scales represents a means to untangle this complex interplay. Here, we combined a spatially-stratified with a longitudinal sampling scheme within differentiated host genetic backgrounds by reciprocally transplanting Pacific oysters between two sites in the Wadden Sea (Sylt and Texel). To further differentiate contingent site from host genetic effects, we repeatedly sampled the same individuals over a summer season to examine structure, diversity and dynamics of individual hemolymph microbiota following experimental removal of resident microbiota by antibiotic treatment. While a large proportion of microbiome variation could be attributed to immediate environmental conditions, we observed persistent effects of antibiotic treatment and translocation suggesting that hemolymph microbial community dynamics is subject to within-microbiome interactions and host population specific factors. In addition, the analysis of spatial variation revealed that the within-site microenvironmental heterogeneity resulted in high small-scale variability, as opposed to large-scale (between-site) stability. Similarly, considerable within-individual temporal variability was in contrast with the overall temporal stability at the site level. Overall, our longitudinal, spatially-stratified sampling design revealed that variation in hemolymph microbiota is strongly influenced by site and immediate environmental conditions, whereas internal microbiome dynamics and oyster-related factors add to their long-term stability. The combination of small and large scale resolution of spatial and temporal observations therefore represents a crucial but underused tool to study host-associated microbiome dynamics.

Keywords: Crassostrea gigas; amplicon analysis; distance-decay relationship; host-associated communities; marine invertebrate microbiota; spatiotemporal dynamics; spatiotemporal patterns.

Figure 1

Experimental design and timeline. The spatially stratified sampling design is described on the map. The experimental sites (orange = Sylt, blue = Texel) are marked by an asterisk (^*). The arrows on the map link the oyster collection sites with the deployment sites (i.e., half of the oysters were returned to their collection site while the other half was transplanted). The inserted plots show within-site deployment scheme with all sampling spots. Within each sampling spot there are two (Sylt) or three (Texel) bags with four oysters each. Each oyster in the bag belongs to one of the four treatment groups (i.e., antibiotic treatment/control × Sylt/Texel oyster origin). Note that the colors (orange = Sylt, blue = Texel) can denote site or the oyster origin, depending on the context. Tables show the total and the per treatment sample sizes for each site. The box below the map shows the sampling timeline for both sites.

Figure 2

α-diversity of hemolymph and seawater (+) microbiota on Texel (left, blue frame), and Sylt (right, orange frame) throughout the sampling period. The dotted line represents mean daily temperature at the sites. Legend explanation: Sy and Tx refer to the oyster origin, A and C to antibiotic treatment and control, respectively.

Figure 3

NMDS plot (Bray-Curtis) depicting variability within the complete dataset (all seawater and hemolymph samples). Hulls enclose laboratory and field samples, respectively.

Figure 4

Constrained analysis of principal coordinates (CAP) of β-diversity of hemolymph communities on Sylt and Texel showing the effects of oyster origin and antibiotic treatment after partialling out the effect of distance (variation explained by distance is given in the plots, “CondV”). (A) Pre-deployment, (B) June, (C) July Sylt only, (D) July, (E) August Sylt only, (F) August. The percentages in parentheses represent the variability explained by significant axes. ^***p < 0.001, ^**p < 0.01, ^*p < 0.05, 'p < 0.1. Hulls enclose all samples from the same site (Sylt = orange, Texel = blue).

Figure 5

Distance-decay relationship including (A) all OTUs and (B) resident (excluding seawater) OTUs. The lines represent linear models fitted for the given spatial scale.

Figure 6

Effect of transient OTUs on slope of distance-decay relationship on small, medium and overall spatial scale. ○ = all OTUs, △ = without transient.

Figure 7

Bacterial community turnover, i.e., the percentage of OTUs shared between the samples from the same individual at the initial and the subsequent sampling points (○ = all samples, △ = laboratory samples excluded).