Tuesday, 21 January 2014

Controlled, safe, clean hatcheries actually enhance salmon disease

 Selective fish hatchery breeding rarely uses quantitative genetic techniques to maintain genetic diversity as they require 10-20 years to affect production and are therefore expensive. Most salmonids are bred by artificial random mating with selection of desirable traits; the resulting production increases associate with genetic diversity decreases. Whilst the genetic changes contrived by aquaculture may benefit artificial rearing of salmonids, they may be maladaptive in natural conditions. Such selection may have counterproductive effects when the salmon are moved to ocean net pens and partially explain mass mortalities from bacterial and viral infections. Some semi-wild breeding practises have been implemented, providing a breeding habitat that exposes salmonids to wild microbe communities and sexual selection. The role of major histocompatibility (MH) genes in vertebrate sexual selection is well known; these genes are central to the adaptive immune system, allowing T cells to produce a range of very polymorphic immune proteins. Heterozygous MH individuals are therefore better able to identify a wider array of pathogenic particles and be favoured in nature; particular alleles conferring resistance to a specific disease may also be favoured and amplified within a population. The diversity of MH genes is likely generated by sexual selection and has important influence on fitness.
This study attempted to quantify the advantages of outdoor spawning of Chinook salmon (Oncorhynchus tshawytscha) for aquaculture by testing captive bred versus semi-wild bred fish immune systems, challenged by Vibrio anguillarum. Mortality, antibody-mediated (humoral) immune response and MH genotypes were assessed in survivors and mortalities.
All salmon were obtained from a farm which had previously lost stock to vibriosis and bacterial kidney disease from Renibacterium salmoninarum and V. anguillarum, so were probably selectively bred for resistance to these diseases. Fish were bred in an exposed, semi-wild channel and left to spawn non-randomly (CH) or in a traditional hatchery with random mating (H); in the CH group the fertilised eggs were left buried until they hatched into fry. After 6 months the H and CH groups were reciprocally transplanted for 5 months, creating four groups. Afterwards, all four groups were exposed to V. anguillarum.
Mortality did not significantly differ between CH/CH and CH/H groups, but H/CH fish had much higher mortality than H/H fish. 12 different MH alleles were identified, a low amount compared to wild populations; H fish more frequently had low allelic diversity than CH fish, however this was not significant and nor was there a significant difference in allelic diversity between the 31 survivors and the 31 mortalities.
Temperature, photoperiods and other abiotic factors are known to affect fish immune systems, which are altered from natural levels in hatcheries. The vibriosis susceptibility of hatchery raised fish transplanted to the channels may be a product of relaxed selection for robust immune systems; genotype is stated to outweigh the effects of the environment, which is why
transplanted CH fish had only moderate mortality. The MH genotypes findings align with the theory that heterozygous MH individuals can cope better with disease. The increased genetic diversity from sexually selective breeding in semi-wild conditions is recommended as a broodstock management technique in aquaculture.
Note that pathogen exposure in semi-wild is not controlled for; meaning that selective pressure or disease from pathogens other than V. anguillarum could have been influencing mortality and genotypes. Another coinciding factor could be that natural microbe levels may stimulate immune systems, affecting development and preparing the fish for later infection, possibly explaining the inferior immune systems of the more microbe-naive H fish. Also, allowing eggs and juveniles to develop in natural conditions may allow colonisation of fish gut or skin with beneficial bacteria able to exclude pathogens, but this study did not control for this. If we could identify advantageous heterozygous MH combinations then quantitative genetic techniques (which are getting cheaper and more efficient), could allow us to combine hatchery breeding with artificial sexual selection, which would probably be more efficient than semi-wild breeding. Hatcheries could also use less sterile conditions to imitate the effects of the environment on wild fish immune systems. This study focused on freshwater rearing, but it relates to the marine environment, because when salmonids are transported to the sea to mature in sea pens, they are exposed to a sudden new range of microbial threats, so a strong immune system is crucial for this stage of salmonid aquaculture. The implications are worse for salmon hatcheries used for replenishing wild populations.

Becker, L. A., Kirkland, M., Heath, J. W., Heath, D. D., & Dixon, B. (2014). Breeding strategy and rearing environment effects on the disease resistance of cultured Chinook salmon (Oncorhynchus tshawytscha). Aquaculture,422, 160-166.


  1. It is a fine line between exposing the young fish to levels of pathogens that will trigger an immunity response, and exposing them to levels that will cause fatalities or possible epizootics. The level of exposure would be difficult to control when exposing them to the environment, this is why vaccines are becoming more popular. Though expensive, they do at least allow some reassurance that you will not accidentally kill the fish whilst trying to trigger immunity in them.
    Do you think genetically profiling individual fish and selectively breeding desired genotypes is feasible for such large quantities of salmon?

    1. At the moment no I do not, currently it would be more feasible to let the salmon sexually select each other. Perhaps vaccine developments instead could reinforce salmon immunity more economically than large genetic techniques. But perhaps novel genetic techniques could be cheap and efficient enough to aid salmon selection in future selection; it has been shown that such methods have been used to effectively select against grilsing by identifying quantitative trait loci.

  2. It makes sense that fish raised in virtually sterile hatcheries fail to adequately develop immunity to pathogens that they will later encounter in sea pens. I wonder whether interfering RNA may be a consideration here? It may be possible to include RNAi into feeds, targeting the knockdown of viral mRNA to reduce the mortality and severity of viral outbreaks. Although, the RNAi may need to be sequence specific which would therefore only target certain viruses. Food for thought though!

    1. It is silly of aquaculture to expect their livestock to be well adjusted in the face of threats which they have previously been completely sheltered from. With time and money, RNAi feeds could certainly reduce viral infections, but I bet it would be more cost effective to adopt semi-wild breeding programs.

  3. I wonder if this could be solved by something as simple as using pumped ocean water in young fish hatcheries strengthened with probiotics? Surely over time, by allowing a certain degree of die-off, you're ensuring strength in future salmon stocks. However, I guess then there is the risk of 100% die-off but perhaps the odd die-off is more cost-effective than bi-monthly vaccines for new generations of fish.