The profiling of 92 ascidian samples from 31 individuals revealed a high diversity (19,410 ASVs) of the microbiome associated with Styela plicata in a geographically restricted range. Albeit previous studies were made with 97% OTUs, and thus are not directly comparable to ours, Dror et al (2019) detected ca. 12,000 OTUs in 24 specimens of S. plicata from Israel and North Carolina, two highly distant populations, and a similar number of OTUs was found in the congeneric S. clava in 10 individuals from New Zealand and Ireland [65]. These results, as well as those in the present work, confirm the notion that ascidians are host to a rich microbiome [15–17]. However, the multiple factors determining this diversity have been poorly studied and the high values observed here are likely the result of the inclusion of different compartments as well as ontogenetic stages.
We detected significant differences in microbiome structure between body compartments (tunic, gill, gut), as well as between them and water. Overall, α-diversity estimates followed the order water > tunic > gill > gut. For the Shannon index, water had significantly higher diversity than all the ascidian compartments, and ASV richness was also highest for water, but not significantly different from the ascidian compartments. Juveniles had higher diversity values than adults, but the differences varied across populations (for richness) and compartments (for Shannon index). Among localities, the trend was of higher diversity in Blanes (the smaller harbour) and lower in Vilanova, with all pairwise comparisons significant for the Shannon index.
More pronounced were the differences in terms of bacterial taxonomic composition across sample types. Of the most abundant groups, Gammaproteobacteria were more evenly distributed, but Cyanobacteria were more prominent in the gut, Alphaproteobacteria in the tunic, and the unidentified category in the gills. Compositional differences across our categories were also uncovered in terms of β-diversity. The PERMANOVA results showed, first, significant differences of water with respect to the ascidian samples. Among ascidian compartments, the highest amount of variability was explained by compartment, followed by population and, lastly, stage. All these components of variation were significant, as were all two-way interactions. All pairwise compartment comparisons were significantly different across all harbours and stages, except the comparison between tunic and gill in Barcelona. Juveniles and adults were differentiated as per the tunic and gill composition, but not for the digestive contents. These compositional differences point to different roles of the symbiont community in each ascidian compartment, as confirmed also by the functional analyses. Tissue-specific microbiota have been also detected in other filter-feeding invertebrates, consistent with putatively differentiated roles [66, 67].
Our functional analyses should be taken with caution, as the number of ASVs and reads that could be used were in general low, particularly for the gills since a relevant fraction of the associated ASVs remained unidentified. This fact highlights the need to complete gaps in the reference databases. Nevertheless, with the predicted pathways different functional profiles could be demonstrated, showing that compartment-specific associated bacterial communities fulfill different roles as part of the holobiont functioning. In this scenario, the compartment explained the highest amount of variance in the functional analyses, followed by stage and population, being all main factors and the two-way interactions significant. Pairwise tests revealed significant functional differentiation between juveniles and adults for tunic and gill predicted functions, but not for those of the gut contents, suggesting that the microbiome communities of the former tissues change through ontogenetic development. Metagenomic and metatranscriptomic studies on the symbiont communities of S. plicata are necessary for an accurate assessment of the different functions at play in each compartment and stage.
Trace elements are recognized as serious pollutants due to their toxicity, persistence, and ability to accumulate in marine organisms [68–70]. In ascidians, accumulation of toxic elements inside harbours was higher than in outside populations [71], and S. plicata has been shown to be a good bioindicator of heavy metal pollution due to its bioaccumulation potential [72–74]. Our results on 9 trace elements (mostly heavy metals) showed significantly higher concentration in the ascidian compartments than in water in Cu, Al, and Fe, while for B and Se the concentration in water was significantly higher than in the ascidian. Concentrations in adult tunic were always higher than in juveniles, and this trend was significant in all cases except Zn and B, pointing to a bioaccumulation over time in the ascidian tissues. The bioaccumulation trace elements, many of which are toxic, might have beneficial effects on S. plicata, since they could play a defensive role against predators and enhance the individual’s fitness. In fact, it has been reported that high levels of V accumulated in ascidian species tissues turn them unpalatable to their predators [75]. Interestingly, there is a close relationship between these pollutants and the microbiome structure of S. plicata, as the former explains 68.29% and 54.95% of the variance in the ontogenetic and adult tissues redundancy analyses, respectively. Different tissues are correlated with concentrations of different trace elements, in particular gill with Zn, and tunic with Fe, Al, As, Cu and V. Water samples were more related to B and Se concentrations. Of note here is that juveniles appeared in the RDA ordinations separated from adults, and closer to water samples, again highlighting the potential for differential accumulation of trace elements and microbial community components over the ascidian lifespan.
Our study is the first to demonstrate an unexpectedly diverse symbiont community in the gills of ascidians. This community is distinct both in composition and in function from the other ascidian compartments, and the relatively high abundance of unidentified ASVs in it (27.78% of reads) points to a largely unexplored and specific microbiome. While the tunic of ascidians can act as a relatively inert outer covering, the branchial sac is highly vascularized and in intimate contact with the circulating haemolymph. The gill thus provides a direct contact point between the ascidian tissues and the environment, as shown by the high percentage of bacterial reads (58.3%) that this compartment has in common with the water core. This is coupled with high water pumping rates [76] making branchial tissues particularly prone to incorporate water-borne chemicals and organisms. Indeed, the branchial tissue of solitary ascidians accumulates several heavy metals at higher concentrations than other tissues [73, 77] although in our study the same individual has generally higher concentration of trace elements in the tunic with the exception of Zn. The branchial sac is also likely to be the entry point of potential pathogens and toxic algae inhabiting anthropized habitats where S. plicata thrives [78, 79]. The gill microbiome can therefore be a crucial factor to aid the ascidian to cope with environmental conditions and to mediate immune responses, akin to the role assumed for the skin of other groups [80, 81]. It has been shown that microbes can play a role in detoxifying potential toxic compounds in the diet of metazoans (“gut microbial facilitation hypothesis” [82]). In the case of ascidians, it seems reasonable to assume that this role is mainly played by the gill microbiome, at the main point of contact with environmental biotic and abiotic influences. Indeed, the redundancy analysis indicated a different behaviour of the gill compartment as explained by environmental variables, with a high correlation with Zn values, an element that is concentrated in the branchial basket of S. plicata [73]. In any case, gill-associated bacteria should be taken in consideration as future candidates for research on biotechnological applications such as bioremediation or bioindicators. Furthermore, anti-pathogenic functions can be expected since gill bacteria can act as a kind of symbiotic immune system in S. plicata.
Previous studies of ascidian gut-related microbiomes explored different sources of the symbiont community, making comparisons difficult. For instance, Dishaw et al. (2014) analysed stomach and gut tissues together with their digestive contents. Utermann et al (2020a) flushed the contents and retained only the tissues. All these authors performed these studies on Ciona intestinalis, a tunicate model species. On non-model species, Wei et al. (2020) analysed gut contents separated from the ascidian tissue over a seasonal cycle of Halocynthia papillosa, finding changes related to season and starvation stress. It is noteworthy that in our study the gut microbiome did not change substantially, both in ASV composition and in predicted functions, between juveniles and adults, contrary to what happens in the other ascidian compartments investigated. Very likely the microbiota associated with the digestive tissues should be fully functional from the very beginning, experiencing little change over time since it plays an important role in the immune interactions of ascidian hosts [28, 83], and possibly on feeding and metabolic processes [31]. Thus, our results point to a microbiota specific of the gut contents, being the higher abundance of Cyanobacteria with respect to other compartments likely a result of the capture of food items by the ascidian. As cyanobacteria are related to eutrophication [84], this indicates a potential role of the ascidian in remediation of pollution, in agreement with reports that S. plicata can effectively remove bacteria and bloom-forming microalgae from the environment [85, 86].
It has been shown that introduced ascidians have a core microbiome, generally of low diversity but high abundance, representing species-specific symbionts, and then a dynamic component of high diversity but low abundance, likely representing locally acquired symbionts that may confer resilience and adaptive value to the populations [17, 25, 42]. Our multidimensional study showed that the concept of core and dynamic components in ascidians should be qualified with respect to which compartment is considered. Only one ASV was present in all ascidian samples (and in all water samples as well), and the core and variable communities are best defined as per compartment. The tunic had the least diverse core community, with high stage specificity (4 ASVs in juveniles, 6 in adults, 1 shared). These ASVs constituted a variable proportion of the tunic reads, ranging from values below 10% in juveniles to values above 40% in adults. The gills and gut featured more diverse core communities, and they represented in all samples more than 20% and 40% of the reads, respectively. The multiple interplay between compartment- and stage-dependent core and variable components of the microbiome of S. plicata implies a wide scope for adaptation to different environments and stresses.
We also detected a significant ontogenetic component in the structure of symbiont communities. Some vertical transmission mechanisms have been demonstrated for colonial ascidians with obligate symbionts [87], and there is also evidence of the presence of prokaryotes in the embryos and larvae of colonial ascidians [88–90]. However, to date this vertical transmission has not been demonstrated in solitary ascidians. Moreover, no bacteria have been found in the gonad tissue of S. plicata [41, 42], making transmission to embryos unlikely. In any case, even if some symbionts are passed vertically, the ascidians should acquire their complete microbe complement horizontally from the environment, as described in other marine invertebrates [91, 92]. In the present study, most of the core bacteria found in the different tissues at different ontogenetic stages are also found in the surrounding water, although at low concentrations. This reinforces the idea of a horizontally-transmitted microbiome in S. plicata, which can capture bacteria already adapted to local conditions to enhance the ascidian introduction success. Overall, ASV richness was not significantly different between adults and juveniles, but diversity values were higher in juveniles (except for gut). This pattern may be due to selection acting on the established microbiome, promoting the proliferation of few bacteria which can overcome trace element bioaccumulation. In general, ascidian compartments are more differentiated for adults, as is also found for their functional profiles, likely indicating a specialisation of the different microbiomes. The exception to this trend is the gut component, with little changes. Juveniles are also closer in terms of functionality to water. All evidence points to a changing and specialising microbiome as ascidians grow, likely promoted by a passive acquisition of the microbiome from the environmental water by filtering. Those bacteria that find in S. plicata tissues a suitable habitat from which they benefit while enhancing S. plicata fitness could be selectively maintained and amplified through adulthood. Furthermore, the microbiome specialisation of S. plicata is likely linked to the progressive bioaccumulation of pollutants in S. plicata tissues, since mature microbiomes need to face and/or benefit from high concentrations of many trace elements by metabolically getting rid of them or by taking advantage of them during bacterial metabolite synthesis [93–95].
Finally, even in a restricted geographic range, we found evidence for a significant locality effect in some variables. Intra-specific differences in the microbiome of several ascidian solitary species have been reported to be more marked between locations than within them [65]. That work, however, encompassed widely separated geographic locations. Casso et al., (2020) uncovered for the worldwide distributed colonial ascidian Didemnum vexillum that the main factor explaining microbiome differentiation was the geographic component, but a significant effect of the genetic relatedness of ascidian colonies was also detected. Styela plicata has been shown to have a strong spatial and temporal genetic structure [32, 40]. In our study we fixed the temporal dimension to avoid too many variables, but the populations of two of the harbours studied here (Blanes and Vilanova) were known to be genetically diverse and differentiated among them [96]. Likewise, there are differences among the studied harbours’ features (pollution, trace elements, dimensions, activities). We cannot at present completely disentangle environmental factors from genetic influences but, in any case, locality differences were detected in α-diversity, β-diversity and functional analyses, being this fine-scale geographic variability indicative of potential adaptive mechanisms in the microbiomes of these populations to face differential environmental conditions.