Effects of Seasonal Anoxia on the Microbial Community Structure in Demosponges in a Marine Lake in Lough Hyne, Ireland

Abstract

Climate change is expanding marine oxygen minimum zones (OMZs), while anthropogenic nutrient input depletes oxygen concentrations locally. The effects of deoxygenation on animals are generally detrimental; however, some sponges (Porifera) exhibit hypoxic and anoxic tolerance through currently unknown mechanisms. Sponges harbor highly specific microbiomes, which can include microbes with anaerobic capabilities. Sponge-microbe symbioses must also have persisted through multiple anoxic/hypoxic periods throughout Earth's history. Since sponges lack key components of the hypoxia-inducible factor (HIF) pathway responsible for hypoxic responses in other animals, it was hypothesized that sponge tolerance to deoxygenation may be facilitated by its microbiome. To test this hypothesis, we determined the microbial composition of sponge species tolerating seasonal anoxia and hypoxia in situ in a semienclosed marine lake, using 16S rRNA amplicon sequencing. We discovered a high degree of cryptic diversity among sponge species tolerating seasonal deoxygenation, including at least nine encrusting species of the orders Axinellida and Poecilosclerida. Despite significant changes in microbial community structure in the water, sponge microbiomes were species specific and remarkably stable under varied oxygen conditions, which was further explored for Eurypon spp. 2 and Hymeraphia stellifera However, some symbiont sharing occurred under anoxia. At least three symbiont combinations, all including large populations of Thaumarchaeota, corresponded with deoxygenation tolerance, and some combinations were shared between some distantly related hosts. We propose hypothetical host-symbiont interactions following deoxygenation that could confer deoxygenation tolerance.IMPORTANCE The oceans have an uncertain future due to anthropogenic stressors and an uncertain past that is becoming clearer with advances in biogeochemistry. Both past and future oceans were, or will be, deoxygenated in comparison to present conditions. Studying how sponges and their associated microbes tolerate deoxygenation provides insights into future marine ecosystems. Moreover, sponges form the earliest branch of the animal evolutionary tree, and they likely resemble some of the first animals. We determined the effects of variable environmental oxygen concentrations on the microbial communities of several demosponge species during seasonal anoxia in the field. Our results indicate that anoxic tolerance in some sponges may depend on their symbionts, but anoxic tolerance was not universal in sponges. Therefore, some sponge species could likely outcompete benthic organisms like corals in future, reduced-oxygen ecosystems. Our results support the molecular evidence that sponges and other animals have a Neoproterozoic origin and that animal evolution was not limited by low-oxygen conditions.

Keywords: Demospongiae; Porifera; Thaumarchaeota; anoxia; deoxygenation; host-microbe interactions; microbiome.

FIG 1

Map of Lough Hyne with sampling sites and oxygen and temperature profiles taken from the middle of the Western Trough in Lough Hyne. (A) Map of Lough Hyne showing sampling sites and in situ pictures of encrusting sponges at 27-m depth. Upper left image, the sponge actively pumping as visualized using fluorescein dye under hypoxic conditions; upper right image, a sponge without dye. (B) Dissolved oxygen concentrations versus depth for the three different sampling trips, corresponding to three different oxygen conditions at ∼27 m (solid line): July 2018, anoxic (red), July 2019, hypoxic (green), and April 2019, normoxic (blue). Samples were collected at ∼27 m (solid black line, below the thermocline) and at ∼20 m (dashed black line, above the thermocline) for comparison across time and oxygen condition. (C) Temperature versus depth during the different sampling oxygen conditions.

FIG 2

28S Bayesian inference (BI) phylogeny of sponges from Lough Hyne (indicated by DC numbers after taxon names). For visualization, subtrees were pruned from the complete phylogenetic tree (see Fig. S3 in the supplemental material). Posterior probability values of >0.95 are given above branches. See Fig. S4 for cox1 BI phylogeny.

FIG 3

(A) PCA of the focused subset. (B) CCA of the focused subset. Individual OTUs are identified in gray. (C) Heat map of the OTUs (order level) driving the separation based on anoxia in the CCA, i.e., OTUs contained within the red shaded area in panel B.

FIG 4

Heat map of top seven most abundant OTUs in the focused subset.

FIG 5

Archaeal maximum likelihood phylogeny of partial 16S rRNA gene (292 bp), with bootstrap support values (1,000 replicates; GTRCAT model) given for nodes with 70% or greater. OTUs of interest are marked in red.

FIG 6

Bacterial maximum likelihood phylogeny of partial 16S rRNA gene (292 bp), with bootstrap support values (1,000 replicates; GTRGAMMA model) reported for nodes with 75% or greater. Sponge OTUs including all Delta- and Gammaproteobacteria OTUs from heat maps in Fig. 3C, 4, and 7 from this study are highlighted in red. The OTUs that were absent or negligible in sponge samples and present only in anoxic water are marked with an asterisk.

FIG 7

Top 26 most abundant OTUs present in sponge species and environmental samples from the location of seasonal anoxia (i.e., Labhra Cliff, designated as anoxic, hypoxic, or normoxic) and outgroup samples.

FIG 8

Hypothetical metabolic cycling processes in the sponge holobiont showing the three major symbiont combinations i, ii, and iii. Dashed, red arrows indicate the cycling processes under anoxia, and solid black arrows show processes under normoxic conditions.