Homework done, budgets in, top brass is happy: a Selective Breeding Programme is underway. There are 400 families each in it’s own tank with 200 fry. All the expense is justified by improvements to productivity.

Some of the current expense is in Passive Integrated Transponder tags or PIT tags. These little radio chips have a 13-digit code that is read externally after the tag has been placed in the fish. When it comes time to combine family groups, the PIT tags will keep the families identifiable. The downside comes when you lose a tagged fish to disease or whatever, or feel a little crunch in the middle of an otherwise healthy, wholesome and nutritious meal.

If only there were some way to identify fish without a PIT or other tag. This way, families could be combined to live in real tanks, have the communal reality of life on the farm and save money. This would require knowing the pedigree of individuals without some sort of tag.

Think about why tags are used. The fish are measured and weighed, FCR, SGR, and mort rate are calculated, heritable traits are catalogued and determined and a host of other desirable things are collected in order to identify families with superior traits. The tags are necessary to identify those individuals and families that carry those traits.

Now think about what drives the heritable traits to make them desirable: the fishes’ genes. If it were possible to routinely and retrogressively (after the fact) track a family through its life by its genes, environmental effects among groups would be minimized and breeding potential optimized. As well, a big concern in programmes is inbreeding: selecting for a positive characteristic but all the while eliminating diversity of the overall gene pool. Tracking performance based on genes (genetic profiling) obviates the need to keep fish separate, removes bias form sampling and minimizes inbreeding.

This may be the stuff of dreams. It may take all the resources of the Bat Cave to figure out genetic profiling. Yet again, it may not.

Micro satellites (MS) are little non-coding (they don’t result in a protein) tandem repeats of DNA. They are often associated with a coding region (gene) of DNA and therefore are of use in gene mapping. They can also be used to as identifiers of family and heritable traits. The trouble is identifying the MS and associating it with a family or a trait.

There are several published MS markers for Atlantic salmon. They permit the retrogressive tracking of family groups with a very high probability. In themselves, they achieve the objectives of identification by parentage and protection against inbreeding. Again, it’s the association of the MS marker with a heritable trait that is the sticky point. And the fish still have to be sampled as before.

To get a grip on why MS markers aren’t the ‘be all to end all’ in genetic profiling, the concept of polygenic traits for quantitative inheritance needs explanation. Simply put, there are more than one or two genes that contribute to growth. There are probably dozens. A MS marker could be association with one or more of these genes and have a high impact on how big or fast the fish grows. But it might only be associated with one of the contributed genes, not all of them.

Perhaps amount of growth hormone expression is one of these traits. But, even GH level itself relies on several other genetic factors. So the trait (growth) relies on the relationship of several contributing genes to achieve the desired effect. In different families, theses genes may have different expression levels or even be variants of that gene. Thus the quantitative trait is a function of what type and how many of the contributing genes are present.

If there were bits of known DNA sequence that were associated with each of these contributing genes, and they were consistently found near those genes, they could be called Quantitative Trait Loci or QTL. In other words, they were genetic markers that identified the presence of one gene in a polygenic trait. Get several of these QTLs identified for the trait of interest and the genetic secret of optimal growth is nearer. Get more of the important genes together by QTL and we get superior fish. Sounds so easy.

QTLs are known for may polygenic human diseases and some other stuff like flower colour, but are not widely known for useful things like fish genetics. The state of gene identification or the recognition of QTLs associated with them is growing. Once there, its possible to identify the gene associated with the QTL, and figure out what the function of the gene really is. Then a gene chip can be built.

Another way to get genes for your chip is to take salmon tissue and copy all the genes in the sample. The genes are sequenced and the codes matched to know similar genes in other organisms. The genes can then be placed on the chip.

Where gene chips help geneticists is the positive identification of genes that are turned on (or off) in specific traits of interest. It is like the QTL process, but a little different because the actual genes involved are identified. Gene chips however can be too sensitive and sometimes can identify genes that respond to unrelated things like what kind of dinner was eaten.

Before outfitting a complete genetics lab and abandoning the Selective Breeding Programme already up and running, these tools are costly and generally in the hands of researchers. There are gene chips for Atlantic salmon that currently have 8,000 genes on them, but there is still the need for lab coats to process the samples and interpret the results. The point here is that there are good things on the horizon for salmon breeding programmes. More tools are becoming available and the science behind them is accumulating at Bat-speed.