Analytical challenges and characterization of skin and gut microbiota of Atlantic salmon fry in a commercial smolt production facility.
MetadataVis full innførsel
Although the Atlantic salmon is an economically important specie, little information exist on the composition of its gut and skin microbiota during early lifestages. Fish microbiota play an essential role in health of fish larvae. Initially, this study aimed at the characterization of the gut and water microbiota of salmon fries from a first feeding experiment in a research facility at Frøya. Despite our efforts to optimize the PCR protocol by applying different PCR facilitators, annealing temperatures, PCR primers, PCR cycles and DNA concentrations, we were not able to amplify 16S rDNA from this samples. This caused the study from the Frøya experiment to be discontinued. Nevertheless, processing and analysis of an Illumina sequencing data set for V4 16S rRNA gene amplicons was used to characterize the gut and skin microbiota of salmon fries of different age in three production batches sampled at three-time points (t1, t2, t3) from a commercial production farm at Follafoss (Salmar Settefisk AS). Each production batch represented fries of a distinct age group reared in different rearing systems: Production batch 1R was reared in RAS and represented the youngest fry (50 dph at the first sampling), while production batch 1A (97 dph at the first sampling time) and 5S (287 dph at the first sampling time) were reared in FTS and RAS, respectively. Sampling time represented the increase in age of fries in each production batch (first sampling time (t1): 22.02.2017, t2: 36 days after t1, t3: 63 days after t1). At the class level Betaproteobacteria, Actinobacteria and Gammaproteobacteria dominated the gut and skin microbiota. OTUs that contributed most dissimilarity between gut and skin microbiota included OTU-1(Propionibacterium), OTU-9 (Zoogloea), OUT-8 (Delftia), OTU-12 (Vibrio) and OTU-4 (Brevinema). Propionibacterium, Zoogloea and Delftia were more abundant in the skin microbiota while Vibrio and Brevinema were more abundant in the gut microbiota. The gut microbiota had a higher alpha diversity than the skin microbiota while the skin microbiota had a higher beta diversity than the gut microbiota. In the gut microbiota, older fries in the production batch 1A and 5S had a higher alpha diversity than youngest fries in production batch 1R. The skin microbiota seemed to be more influenced by the sampling time and production batch than the gut microbiota. Next, we investigated how the gut and skin microbiota varied between production batches and sampling times. The production batches seemed to influence the gut and skin microbiota more than the sampling time. The difference in skin and gut microbiota between production batches at each sampling time and between sampling time in each production batch was also studied. There was no significant difference in gut microbiota between production batches or their sampling time, although the PCoA plot suggested differences between sample groups. For the gut microbiota at t2, OTU-4 (Brevinema), OTU-12 (Vibrio), OTU-20 (Moritella) and OTU-1312 (Eubacterium) was far more abundant for the oldest fry in production batch 5S than in the younger fries within production batch 1R and 1A. Moreover, at t1 and t2, OTU-8 (Delftia), OTU-9 (Zoogloea), OTU-13 (Acinetobacter) and OTU-23 (Comamonas) were more abundant in the younger fires in production batch 1A and 1R than older fires in production batch 5S. The most abundant genera in the salmon skin microbiota were Propionibacterium, Zoogloea and Bacillus. There was no significant difference in skin microbiota between production batches or sampling time. This is the first study reporting a relatively high abundance of genus Zoogloea on the skin microbiota of a teleost. The result from our study indicates that age probably had more influence on the microbial composition of the gut and skin than the rearing systems used for the production batches. Furthermore, the gut and skin microbiota was more similar for the younger fries in production batch 1A and 1R than older fries in production batch 5S. Gut microbiota of fries in production batch 1A and 5S became more similar with increasing sampling time (t1-t2). The older fries in production batch 5S had the most variable skin microbiota among individuals at second sampling time (t2). Surprisingly, PcoA analysis indicated that both the gut and skin microbiota clustered together at t3, independent of the production batch. It was found that DNA from the samples at the third sampling time had been extracted by a separate DNA extraction kit and probably were influenced by contaminating DNA. OTU representing Propionibacterium was highly abundant in these samples. This reason lead to the exclusion of results from the third sampling time when the investigated factors were discussed. This study indicates that age is a major determinant in the architecture of the gut and skin microbiota of developing salmon larvae.