Pacific salmon gene expression


Salmonid fishes have been the focus of captive breeding programs worldwide because of declining wild populations. Captive breeding programs also augment fisheries. Human activities, such as logging and agriculture, have degraded the freshwater habitat and are associated with the population declines. Increased environmental protection and stewardship activities are helping address these freshwater habitat issues. However, there are also changes in the marine habitat, such as changes in temperature and prey availability, which are also associated with the population declines. Due to the large spatial extent, these marine habitat issues are very difficult to address, and populations continue to decline. In southern British Columbia (BC) for the past 15-25 years, Sockeye salmon (Oncorhynchus nerka), Chinook salmon (O. tshawytscha), and Coho salmon (O. kisutch) populations have declined, with captive breeding programs being used to enhance populations and augment fisheries. Low adult returns of Chinook salmon and Coho Salmon triggered reduced commercial and recreation fisheries for a few years in the early 2000s. As recently as 2015, recreational fishing of Sockeye salmon on the Fraser River, BC was banned because of low adult returns. Therefore, within Canada beyond the need to improve the habitat of salmonid fishes, there is a timely need to improve the quality of the captive fish that are released into the natural environment.

Coho salmon
Coho salmon smolt

Salmonid fishes may be especially sensitive to individual fitness declines from rearing in the captive (hatchery) environment. The typical hatchery environment differs from the natural environment in many respects, including a higher rearing density and food availability and exposure to fewer predators. Such differences can produce juvenile fish with suboptimal behaviours and lower physiological condition. Also, release of hatchery fish into the natural environment at a time when they are not physiologically prepared for the transition from freshwater to seawater usually results in rapid mortality from osmotic stress. The willingness to change hatchery breeding practices to improve the quality of hatchery fish for the natural environment already exists. However, a major question for these programs is to identify which critical rearing factors require improvement. This problem is exacerbated by the difficulty of identifying the locations and causes of mortality once a fish is released into the natural environment. All the same, new technologies, namely fish acoustic tagging (biotelemetry), has been successfully applied to this issue. Moreover, recent advances in physiological (gene expression) profiling of biopsied tissues from tagged fish has greatly increased our understanding of the potential associations between survival and pre-existing physiological condition in the natural environment.

Suggested references:

Fraser D.J. 2008. How well can captive breeding programs conserve biodiversity? A review of salmonids. Evolutionary Applications 1: 535-586. doi: 10.1111/j.1752-4571.2008.00036.x.

Wikelski M., Cooke S.J. 2006. Conservation physiology. Trends in Ecology and Evolution 21: 38-46. doi: 10.1016/j.tree.2005.10.018.

Gene expression from microarray studies

RNA tubes
Gene expression (RNA) collection
tubes for seven tissues

Hatchery and wild Sockeye salmon, Chinook salmon, and Coho salmon were collected along their migration route over multiple years by Fisheries and Oceans (DFO). Specifically, in the spring, fish (smolts) were collected in the freshwater natal rearing environments just prior to downstream migration, and again in the lower river environments prior to ocean entry. In the spring, summer, and fall, fish (post-smolts) were collected in the early marine environments prior to their long distance migration. Collected fish were euthanized, measured for body length and mass, and sampled for several tissues, including gill, brain, liver, muscle, kidney, heart, and spleen.

The genome-wide physiological (gene expression) profile of four tissues (gill, brain, liver, and muscle) was quantified using a 44K cDNA microarray at the Molecular Genetics Laboratory within the DFO Pacific Biological Station, Nanaimo, BC. The Coho salmon and Chinook salmon gene expression studies are each comprised of 150-250 fish sampled across two (Coho) to four (Chinook) years, with the same fish analysed across the four tissues. Each study incorporates hatchery and wild fish sampled across multiple environments and seasons. I will use supervised multivariate (MANOVA) and univariate (paired t-tests) analyses to determine the transcriptional differences between hatchery and wild fish. I will also use unsupervised (PCA) analyses to identify dominant physiological patterns.

qPCR biomarkers (Salmon 'Fit chip')

Genes associated with the strongest signature differences between hatchery and wild fish will be developed into qPCR biomarkers and validated across a broader array of samples on the Fluidigm BioMark™ microfluidics platform which can assess 96 assays across 96 samples at once. I am working with DFO Salmon Enhancement Program hatcheries to collect samples from hatchery and wild fish for validating smoltification (seawater preparedness) and domestication (hatchery vs. wild) qPCR biomarkers. I am also collecting blood and gill tissues to incorporate additional measures of osmoregulatory performance (osmolality, Na-K ATPase, ion balance) and stress (cortisol, lactate). Additionally, I am collecting samples from four different hatchery rearing groups at Nitinat Hatchery, i.e. traditional large body size, traditional small body size, enriched (semi-natural) large body size, enriched small body size, to associate differences in gene expression among groups with survival in the marine environment. The results will offer an important validation of biomarkers as well as provide targets for additional improvements to hatchery rearing regimes for increasing the survival of fish released into the natural environment.

Fluidigm Biomark plate
Fluidigm BioMark™ dynamic
array plate for 96 assays
across 96 samples

The Miller Laboratory is also mining controlled laboratory studies to elucidate biomarkers associated with specific stressors, with the intent of developing a ‘Fit Chip’ (based on BioMark™ technology) that can rapidly assess the physiological condition of 100s to 1000s of fish. In particular, this chip will measure gene expression as well as the loads of several pathogenic microbes, providing a unique opportunity to compare these two variables with disease status. I am working with DFO Salmon Enhancement Program hatcheries to validate the full suite of qPCR biomarkers and test their utility for measuring the variance in physiological condition (e.g. seawater preparedness, feeding activity, disease exposure and resistance, and other potential stressors).


The identification of physiological mechanisms contributing to differences between hatchery and wild fish can serve as targets for much needed improvements to hatchery programs. For example, a lower feeding behaviour profile (e.g. metabolism and body growth) for hatchery than wild fish may indicate that more live feed or experienced demonstrators are needed in the hatchery. A lower seawater preparedness profile (e.g. osmoregulation) for hatchery than wild fish may indicate that changes in photoperiod and temperature and improved monitoring of seawater preparedness are needed in the hatchery. A lower disease resistance profile (e.g. immunity stimulation) for hatchery than wild fish may indicate that broader pathogen and disease screening and/or enhanced mitigation measures such as vaccination of hatchery fish is needed. Overall, for several salmonid fishes within Canada, improvements in hatchery programs can increase the survival of hatchery fish upon release into the natural environment.



Research Projects

Pacific salmon
gene expression

Lake Ontario
Atlantic salmon



Last updated January 2018
© Aimee Lee Houde