Bacterial Communities Show Algal Host (Fucus spp.)/Zone Differentiation Across the Stress Gradient of the Intertidal Zone

Abstract

The intertidal zone often has varying levels of environmental stresses (desiccation, temperature, light) that result in highly stress-tolerant macrobiota occupying the upper zone while less tolerant species occupy the lower zone, but little comparative information is available for intertidal bacteria. Here we describe natural (unmanipulated) bacterial communities of three Fucus congeners (F. spiralis, high zone; F. vesiculosus, mid zone; F. distichus, low zone) as well as those of F. vesiculosus transplanted to the high zone (Dry and Watered treatments) and to the mid zone (Procedural Control) during summer in Maine (United States). We predicted that bacterial communities would be different among the differently zoned natural congeners, and that higher levels of desiccation stress in the high zone would cause bacterial communities of Dry transplants to become similar to F. spiralis, whereas relieving desiccation stress on Watered transplants would maintain the mid-zone F. vesiculosus bacterial community. Bacteria were identified as amplicon sequence variants (ASVs) after sequencing the V4 hypervariable region of the 16S rRNA gene. Microbiome composition and structure were significantly different between the differently zoned congeners at each tissue type (holdfasts, receptacles, vegetative tips). ASVs significantly associated with the mid-zone congener were frequently also present on the high-zone or low-zone congener, whereas overlap in ASVs between the high-zone and low-zone congeners was rare. Only 7 of 6,320 total ASVs were shared among tissues over all congeners and transplant treatments. Holdfast bacterial community composition of Dry transplants was not significantly different from that of F. spiralis, but Watered holdfast communities were significantly different from those of F. spiralis and not significantly different from those of procedural controls. Additional stressor(s) appeared important, because bacterial communities of Dry and Watered transplants were only marginally different from each other (p = 0.059). The relative abundance of Rhodobacteraceae associated with holdfasts generally correlated with environmental stress with highest abundance associated with F. spiralis and the two high-zone transplant treatments. These findings suggest that the abiotic stressors that shape distributional patterns of host species also affect their bacterial communities.

Keywords: Granulosicoccus; Octadecabacter; Rhodobacteraceae; desiccation; microbiome; peptide nucleic acid clamps (PNAs); transplant; zonation.

FIGURE 1

Sampling scheme for both the natural survey and transplant experiment in one of the two study areas at Schoodic Point, Maine (United States). Dry (D), procedural control (PC), and watered (W) refer to our transplant treatments for Fucus vesiculosus.

FIGURE 2

Fucus congeners, and transplanted F. vesiculosus from Schoodic Point during summer 2016. Natural communities of (A) F. spiralis, (B) F. vesiculosus, and (C) F. distichus; (D) an example of a clump of F. vesiculosus that was transplanted from the mid zone to the high zone; (E) a clump of F. vesiculosus (near pink tag) transplanted into the mid zone as a procedural control.

FIGURE 3

Non-metric multidimensional scaling (NMDS) using Bray–Curtis dissimilarity for all samples of natural congeners (n = 105). Fs, Fucus spiralis; Fv, Fucus vesiculosus; Fd, Fucus distichus; H, holdfast; V, vegetative; and R, receptacles. All samples within the encircled area are from study Area A. Stress = 0.106.

FIGURE 4

Bar plot of relative proportion (± SE) of major classes of bacteria across replicates within: Fs, Fucus spiralis; D, dry transplant of Fucus vesiculosus; W, watered transplant of Fucus vesiculosus; PC, procedural control of Fucus vesiculosus; Fv, natural Fucus vesiculosus; and Fd, Fucus distichus. Class abbreviations: Alphaproteobac, Alphaproteobacteria and Gammaproteobac, Gammaproteobacteria.

FIGURE 5

The number of significantly different ASVs (p < 0.05 in mvabund analysis) identified as being indicative of Fucus spiralis, Fucus vesiculosus, or Fucus distichus.

FIGURE 6

Non-metric multidimensional scaling (NMDS) using Bray–Curtis dissimilarity for all samples (n = 226). D, dry transplant; W, watered transplant; Fs, Fucus spiralis; PC, Procedural control; Fv, Fucus vesiculosus; Fd, Fucus distichus; WC, water column; H, holdfast; R, receptacles; and V, vegetative. Stress = 0.142.

FIGURE 7

Non-metric multidimensional scaling (NMDS) using Bray–Curtis dissimilarity for vegetative tissue from transplant treatments (n = 68). D, dry transplant; W, watered transplant; Fs, Fucus spiralis; PC, Procedural control; Fv, Fucus vesiculosus; Fd, Fucus distichus; H, holdfast; R, receptacles; and V, vegetative. Stress = 0.141. The encircled samples from July 20 are driving spread in the plot and are from different sampling areas, but some samples from both areas (not circled) overlapped.

FIGURE 8

Adjusted p-values of pairwise comparisons of microbial composition and structure for the significant interaction term (Tissue:Treatment) over the three transplant treatments (Dry, Watered, Procedural control), unmanipulated Fucus vesiculosus and unmanipulated Fucus spiralis. Shaded values are not significant (p > 0.05).

FIGURE 9

Non-metric multidimensional scaling (NMDS) using Bray–Curtis dissimilarity for transplant treatments, Fucus vesiculosus, and Fucus spiralis on 20 July 2016 (n = 101). Figure legend as in Figure 6. Stress = 0.150.

FIGURE 10

Scatter plot of mean reads of selected major families of bacteria across replicates. D, dry transplant; W, watered transplant; Fs, Fucus spiralis; D, dry transplant; W, watered transplant; PC, procedural control, Fv, Fucus vesiculosus; Fd, Fucus distichus; H, holdfast; R, receptacles; V, vegetative; B, Burkholderaceae; R, Rhodobacteraceae; and T, Thiohalorhabdaceae.