Atlantic salmon microbiome : impacts of environment and diet

It is well known that gut bacteria can play key roles in host health and changes in microbiome diversity and compositions have been associated with diet, disease development and the environment the host encounters. The gastrointestinal microbiota of mammals has been studied extensively and significant progress has been made in the same area for aquacultured finfish. However, there are still gaps in knowledge around how the finfish microbiome is associated with health and productivity, particularly in a commercial aquaculture setting. The aims of this Thesis were to characterise the bacterial constituents of Tasmanian Atlantic salmon (Salmo salar) microbiomes to 1) assess the impact of time from feeding on measurement and quantification of gut microbiota composition and determine the effectiveness of the faecal scoring method as a rapid way to assess digestive status; 2) examine the effect of the early life production cycle and rearing environment on the composition of host-associated microbiomes; and 3) investigate the efficacy of prebiotic feed supplementation in amelioration of pathomorphological changes caused by soybean meal on the gastrointestinal tract, and the impact on digesta-associated microbiota and growth performance. A faecal score system was originally developed as an indicator of summer stress by using digestive status as a measure of ‚ÄövÑv¿gut health‚ÄövÑvp and is currently used on salmon farms in Tasmania, Australia. Although faecal scoring is widely employed as a means to assess digestive status in different animal species as well as humans the specific relationship between faecal score and gut microbiome is still poorly understood in Atlantic salmon, as are the potential confounding effects of inconsistent timing of scoring post feed. We assessed the impacts of time of sampling in relation to last feeding on both faecal scores and gut microbiota composition in adult Atlantic salmon. Two marine cohorts in different seasons (summer and winter) were sampled over a 24 hour period and comparisons were made between hindgut faecal squeeze samples, hindgut digesta and mucosa from dissected fish were made. We found that faecal score varied with time post feed in the summer group and there were significant differences in bacterial diversity and compositions between low and high faecal score groups, with higher relative abundance of beneficial bacterial types such as lactic acid bacteria in low faecal scores. Conversely, higher abundances of Vibrionales were observed in high faecal scores, which include important pathogenic, commensal and mutualistic bacterial species. Whilst further research is required to establish whether faecal score response is associated with high or low performing individuals, our results did indicate that in cooler months consistently low faecal scores result in similar diversity and compositional profiles across a sampling day, whereas, in summer the highest proportion of low faecal scores were recorded at 21 hours post feed. However, low faecal scores were recorded in fish at all time points in summer (varying from 20-60 % of fish sampled), therefore we suggest that (i) time from feeding prior to sampling should be recorded and (ii) sampling within a set time period post-feed be considered, particularly in warmer months. We recommend that studies should be designed to capture both temporal and seasonal effects that are associated with the changes in digestive status and gut microbiome that we observed. We also found that bacterial diversity was significantly lower in hindgut mucus samples compared to hindgut digesta samples. Mucosa samples also host significantly distinct bacterial communities compared to digesta samples and were dominated by the orders Vibrionales, Oceanospirillales and Pseudomonadales. Our results indicated that faecal squeezes samples were a good proxy for capturing transient digesta microbiota but may not capture all resident gut microbiota. Therefore, sampling of both hindgut mucosa and digesta samples is recommended to capture both transient and resident bacterial populations. The gut and skin microbiomes of Atlantic salmon (Salmo salar L.) were characterised in three commercial hatcheries from egg to sea water transfer to evaluate life stage and environment specific microbial signatures. The bacterial communities of whole organisms (eyed eggs and hatched larvae), whole intestine (fry), distal intestine digesta (parr, freshwater smolt and seawater smolt), skin, feed and water were analysed over the first 16 months of commercial production farm cycle using high-throughput sequencing of 16S rRNA gene V1-V3 amplicons. Community composition of the gut and skin microbiomes were highly dynamic over time and between farm environments. There was also evidence that the hatchery microbiome is retained in the gastrointestinal tract, with significant delay in community changes following seawater transfer, whereas skin bacterial communities were reflective of their immediate environment. This study provides evidence that significant transitions in gut and skin microbiomes occur throughout the Atlantic salmon life cycle, driven by both rearing environment and fish development stage. The impact of hatchery rearing and retention of freshwater microbiome in the gastrointestinal tract suggests there could be opportunities to positively influence smolt microbiome and performance over the weeks following marine input. Increased global demand for marine resources has resulted in increased cost of fishmeal as an aquaculture feed ingredient. Therefore, feed manufacturers have explored a number of plant and animal protein sources as fishmeal replacements. The desire to reduce the cost of production (both in monetary and environmental terms) has resulted in the use of alternative dietary protein sources, such as soybean meal, to supplement diets of carnivorous aquaculture species. Whilst soybean meal is a common alternative due to its low cost and high protein content, it has been shown to interfere with digestive function, often causing enteritis in the distal intestine of Atlantic salmon. In other finfish species prebiotic diet supplements have been shown to help ameliorate these effects by modulating immune responses and thus preventing intestinal damage. This study aimed to investigate the effects of dietary supplementation with the prebiotics sodium butyrate and inulin on growth performance, intestinal health and gut microbiome composition in juvenile Atlantic salmon fed high inclusion soybean meal diets. A total of 4 diets were compared, including fish meal (FM), soybean meal at 30% w/v (SBM), soybean meal with 1% w/v inulin (SBMI) and soybean meal with 1% w/v sodium butyrate (SBMB). This diet trial ran for 48 days in a freshwater experimental recirculating aquaculture system with samples collected at D0, D28 and D56. Digesta was collected from the distal intestine for microbiome analysis and histology of distal intestinal tissue was used to score the degree of intestinal damage. There were significant differences in average weight between diets at the end of experiment, with the SBM diet group having the highest average weight and FM having the lowest. Inulin and butyrate did not reduce intestinal damage, with SBM, SBMI and SBMB diet groups all showing some degree of damage. There were no significant differences in gut microbiome diversity or composition between diet groups after 56 days. The results suggest that dietary supplementation with inulin and butyrate had no effect on the degree of intestinal damage or microbiota composition. The overall aim of this thesis was to provide a comprehensive characterisation of the microbiota of intensively farmed Tasmanian Atlantic salmon. We found that the gut and skin microbiomes are significantly affected by fish, life stage and that diet and rearing environment are associated with these changes. Additionally, whilst the characterisation of the gut microbiome can capture relative changes in gut microbiota profiles there can be significant bias in results due to time of sampling relative to feeding, sample type collected and the type of analysis conducted. Significant differences in bacterial composition exist between the mucosa-associated and digesta-associated intestinal communities, however, non-destructive hindgut faecal squeeze sampling methods provide adequate representation of the hindgut transient community. As significant differences likely exist between experimental and commercial systems, future research could involve comparisons on the same cohort to determine the size of these differences. This is important in aquaculture species where results are used to directly inform industry and could be impacted by experimental design.