For 430 dph fish transferred from the ocean net pen back to the land based facility, both air diffusers and the water column were the largest microbial sources to the gill and skin microbiomes . For the 430 dph net pen reared fish, gill, and skin samples were primarily colonized by microbes from unknown sources followed by small proportions from air diffusers, airstones, and water from pre-transfer. Common planktonic marine microbes from sea water and netpen biofouling were not collected in this study and thus is likely a meaningful “source” which would fall into “unknown sources” in this study. Interestingly, digesta samples for both 430 dph seapen and 430 dph indoor fish were primarily colonized from water samples from the 137 dph . This would suggest that the water community which fish are exposed to prior to transfer to ocean net pen is very important to the gut microbiome colonization and that these microbes remain in the gut even after long term growout in seapens. The finding that the microbiome of the fish digesta originates primarily from water sources rather than feed sources is intriguing. It is important to note, however, that the feeds used in this study were normal extruded pellet feeds with no added probiotics. Results from the Sourcetracker2 analysis reinforce and support the observations from the beta diversity comparisons,growers equipment that fish mucosal sites are influenced uniquely by the BE which also show succession patterns as a function of age.
One of the primary questions in this dataset is understanding how the surrounding environment influences mucosal microbiomes. Specifically, we were interested in understanding the specificity and stability of these microbial communities as a function of ontongeny. To compare fish of the same age and genetic cohort, we sampled fish which were being reared in ocean net pens along with fish which had been in seapens but were brought back to the indoor facility. Digesta samples were previously shown to have large decreases in alpha diversity at 430 dph particularly when comparing the fish in the seapen vs the indoor fish. Interestingly, much of this microbial diversity loss can be attributed to a single uncultured representative ASV, from the family Mycoplasmataceae , which becomes more dominant in the fish gut with age especially in the outdoor seapen. This ASV was observed in 100% of the 430 dph fish yet was found in only 75% of the 43 dph and 137 dph fish, while less frequently observed in the BE . At 430 dph this ASV made up a large proportion of total reads in the seapen and FT indoor tank fish but significantly less abundant in younger fish at 137 dph and 43 dph . Thus, although the Mycoplasmataceae is present in younger fish, the proportion of reads is much smaller. Since these are proportions, it’s important to realize that this does not implicate a biomass change, but only representation in comparison to total microbial diversity. Seafood is an important source of protein globally which has led to the steady positive growth in aquaculture over the past 30 years. Marine finfish production has tremendous opportunity for growth yet challenges and concerns have arisen over the sustainability of such practices . One of the primary concerns is animal welfare and preventing disease transmission from farmed fish to wild stocks .
The mucosal microbiome is an important component of fish health as microbes colonizing the gill, skin, and gastrointestinal tract can either be a source of infection or inversely, protect the animal from infection by inhibiting the colonization of pathogens, producing antimicrobial compounds, or eliciting an immune response . Our research sought to evaluate how the mucosal microbiome develops and to estimate its stability in different body sites over time in the economically important cultured marine fish S. lalandi. We describe the potential sources of microbes from the “BE” that drive these changes across three unique body sites including the gill, skin, and digesta communities. Previous fish microbiome studies have focused primarily on one body site at a time, particularly the gut, while our approach aimed to more fully describe diversity dynamics across multiple mucosal body sites. Gill microbiomes were the most sensitive to changes in the indoor and outdoor culture environment followed by skin with digesta demonstrating a more deterministic or enriched microbiome with ontogenic development. Specifically, while both gill and skin microbial communities increased in diversity with age, the digesta decreased. The progression of decreasing microbial diversity in the fish gut samples suggest that the gut environment is more deterministic rather than stochastic in microbial community composition. Conversely, the gill and skin generally increase in diversity with age which could be due to additive exposure and increased surface area over time.
In addition to variable exposure to the external environment, individual body sites maintain unique physical and chemical properties that confer selection for specific microbial groups. Neutral theory ascribes that biodiversity formation and change over time occurs from random dispersal and exposure events and while it is largely conceptualized in macrofauna and flora , it can also be applied to microbial communities . In contrast, a niche-based model describes how select species evolve and adapt to certain conditions as the result of interspecies interactions and niche differentiation. In this study, we demonstrate that while the gill and skin do have unique microbial communities, the processes for microbial colonization are largely stochastic whereas the gut environment demonstrates a more deterministic process for microbial colonization. In adult Atlantic salmon sampled from marine net pens, gut microbial diversity decreased as the age of fish was increased while the presence of most individual gut microbes were random and only a few deterministic, which was primarily driven by Mycoplasma . In zebrafish and sturgeon , both freshwater fish, higher proportions of gut microbes were non-neutral or deterministic as fish matured . In catfish skin microbiomes, geographic location drove community composition with most microbes being neutral . The implications of different body sites demonstrating a more neutral or deterministic microbiome is important for understanding both the impact of environmental change on wild fish stocks as well as improving aquaculture production. Negative anthropogenic impacts to the marine environment include contaminant and nutrient pollution which can causedisturbances of primary productivity. In a wild marine fish,plant benches the Pacific chub mackerel, the composition of external mucosal microbiomes of gill and skin were most influenced by temporal changes, coinciding with temperature, along with gill alpha diversity positively correlated with age . The gill is an important organ for excretion of nitrogenous waste and gas exchange which is critical for highly active swimming fish like Seriola spp. . In aquaculture settings, microbes which produce compounds causing off-flavor in flesh have been found to be enriched and primarily taken up through the gills of fish . Since the gill is a critical component of maintaining homeostasis, and in this study appears most susceptible to changing environmental conditions, further research is needed to understand how changes in the microbiome may negatively or positively impact fish physiology. Additionally, skin is an important physical barrier for disease prevention. The skin microbiomes of two coastal pelagic marine fish, Scomber japonicus and S. lalandi, were strongly influenced by increased temperature that coincided with increased proportions of a potential marine pathogen, Photobacterium spp. . Body site microbiomes of S. lalandi were most similar to the BE surfaces at the earliest age .
As fish aged, digesta samples diverged from all BE surfaces, while gill and skin were differentially influenced by specific BE surfaces. In Atlantic salmon reared in freshwater indoor hatcheries, microbial diversity from both the tank side and water column were highly correlated with the fish skin and gut, but not other BE surfaces . Understanding which surfaces likely contributed to the various body sites over time was calculated using SourceTracker2 analysis. At 43 dph, the biofilm from the tank side along with the aeration equipment were the largest contributors to the gill and skin communities whereas much of the digesta microbes were from unknown sources. Aeration equipment in tilapia culture has been implicated as a source of Acinetobacter in culture systems . While feed had a marginal impact on the microbial community of the various fish body sites, it was not consistent and was generally lower than the surrounding BE surfaces. Although diet has been shown to have a strong influence on gut microbiome development , the importance of live feeds as contributors to the gut microbiome is debated . One explanation is that the microbes colonizing the live feeds have low specificity for successful colonization of the fish gut. Likewise, since the overall exposure to and density of BE surfaces and associated microbes, including the water, is much greater than that of the live feeds , feed-associated taxa may be outcompeted in the gut environment. A unique opportunity of this study was to compare mature fish from an ocean net pen to fish that had been in the ocean but were transported back to an indoor system to be used as broodstock. We are not aware of any other study which has looked at the microbiome transition from ocean to indoor in a marine fish. Selective breeding programs rely on the ability to develop broodstock which are used to maintain genetic lines from previous grow out populations . Ocean net pen fish generally had lower microbial diversity than indoor reared fish for all fish body sites, but was most pronounced in the gill. This further suggests that the mucosal, even in adult fish that are least susceptible to BE impacts, has a high capacity to change which is critical when considering time scales for probiotic effects . Probiotic treatments in fish are common but little is known about dosage for a given treatment along with frequency of administration for having a lasting effect. If the normal microbial community of a fish gill or skin can change rapidly, this would suggest that a sustained administration rather than a “one-time treatment” would be required for maintaining mucosal health in fish. For gill and skin communities, the water column and aeration surfaces contributed the most for indoor reared fish while fish reared in the net pens had many bacteria of unknown sources, presumably from the ocean, e.g., seawater. Digesta samples, however, were primarily colonized by hatchery water associated microbiota and to a lesser extent feeds. The opposite explanation is also true that in land-based systems, fish feces could be contributing more to the water column microbiome as compared to the oceanic conditions where feces is more quickly exported out of the system. These vast differences and the speed at which microbiomes develop and change is a plausible explanation for differences between wild and farmed Seriola . The most abundant microbe in the Seriola digesta was an unresolved Mycoplasmataceae which was strongly associated with transfer of fish from indoor rearing systems to the ocean net pen. Mycoplasma are important gut microbes which can colonize the gut very early in development. Several plausible explanations exist for this observation. First, it is possible that in land-based systems, fish are simply not as heavily exposed to Mycoplasma. Second, it is possible that Mycoplasma microbial density or diversity is higher in ocean net pen systems compared to the indoor system thus allowing the Mycoplasma to dominate the gut microbiome. Lastly, an alternative explanation is that Mycoplasma outcompetes other microbes in the fish gut especially as the fish increase in age. However, since the data are compositional, it is not possible to determine absolute microbial densities thus requiring additional experimentation to resolve. All mucosal environments were influenced by the BE over time with the strongest effects at early fish development. Digesta samples in particular became less influenced by the BE over time and demonstrated a strong selective or deterministic pressure on microbiome development with increasing age. This progression of decreasing microbial diversity in the fish gut suggests that the gut environment is more deterministic rather than stochastic in microbial community composition whereas the gill and skin generally increase in diversity with age which could be due to additive exposure. One of the limitations of this study is that we did not perform quantitative measures of the microbial communities. Part of the reason for this is that these methods can often involve invasive or destructive sampling of tissues.