Scottish Marine and Freshwater Science Volume 5 Number 5: Identfying a Panel of Single Nucleotide Genetic Markers for River Level Assignments of Salmon in Scotland

Discussion

The analysis presented here has shown that, depending on baseline coverage, it is possible, to be able to assign fish to both region, and where baseline coverage is sufficient, river with high accuracy in most situations. The problem of being able to screen out fish from reporting regions not represented in the baseline has also been successfully addressed using hierarchical reporting regions and exclusion techniques.

The identification of a set of SNP markers able to discriminate between fish from the different regional assignment units even if they are from rivers not represented in the baselines of these assignment units means that it would be expected that new rivers can be added to the baseline in these regions without the necessity of rescreening using the full V2 panel and having to choose new regional SNP markers each time. Regional baseline coverage can now be enhanced using just the regional set of markers already identified which will achieve a significant time and cost saving. It should be remembered, however, that the full V2 panel will be needed if river level assignments are required rather than regional level.

Accuracy of assignment to the regional units was very good, and as expected the assignments become more uncertain at more detailed levels of the hierarchy. However, even when examined at the river level (taking into consideration the requirement of combining some rivers), assignments were found to be very good in most cases. Assignments to river, unlike those to region, will still however, require screening of new rivers using the V2 SNP panel when these new rivers are added to the baseline.

Differentiation between the Spey and the Dee was difficult, even with large numbers of markers. In general when looking at the genetic character of populations it is the smaller populations that tend to be most differentiated. This is a reflection of the various mechanisms responsible for influencing the genetic character of a population. Principal among these are included genetic drift, founder effects and adaptive evolution. All of these influences happen with greater speed and/or have potentially greater effects in smaller populations. The Spey and Dee represent some of the largest meta-populations of salmon in Scotland and it may be the case that genetically they are still very closely related using the 'neutral' SNP markers employed here. It will be interesting to look at the Deveron which lies between them to see if it is also similar to these two larger systems or if it has differentiated through the processes outlined above. It will also be of interest to examine in more detail the Tay as another large geographically close system to determine its levels of differentiation. Further work could be performed on identifying a set of markers which differentiate between the Spey and Dee but this would perhaps require techniques such as RAD (Restriction site Associated DNA) Sequencing using next generation DNA sequencing. Using this approach, DNA from groups of individuals from the different rivers would be pooled and large number of SNPs identified that are associated with differences between the groups (for examples in salmonids see Hohenlohe et al., 2011; Houston et al., 2012).

The results of the analysis presented here have developed techniques that allow:

• Regional assignment of fish to different hierarchical levels
• Screening out of fish from rivers not present in the baseline within regions
• Assignment of fish to most rivers represented in the baseline

Future developments of the procedures will aim to:

• Increase baseline coverage using the regional SNP markers identified here
• Increase river level coverage using the V2 SNP panel
• Identify a set of SNP markers for within region regional level assignments to river
• Further investigate techniques of examining situations where rivers cannot be differentiated