254
M.L. Cancela et al.
leading the way due to the high value of commercial and recreational fisheries but
probably most importantly due to their additional importance for aquaculture breeding and associated development of genomic resources (see Section 7.2 on Genomic
tools and resources). For example, Hayes et al. (2007) identified 2,507 SNPs from
the alignment of Atlantic salmon ESTs, which can be readily used for population
genomic studies.
An example of improving the performance of the number of genomic markers in salmonids can be found in Smith et al. (2007). They compared estimates of population differentiation among 16 collections of Chinook salmon
(Oncorhynchus tshawytscha) using traditional allozyme loci (22 loci), small tandem
repeats/microsatellites (nine loci) and 39 newly developed nuclear SNP loci. The
larger number of markers allowed testing for “outlier loci” (Beaumont and Nichols
1996) identifying five SNPs likely to be influenced by selection. Exclusion of these
loci, and three other, which had previously been suspected to be under selection,
provided a reduction in the estimates of differentiation and an associated increase in
the estimated migration among putative populations. Similarly, slight differences in
relationships among populations were apparent when comparing results from different marker classes, where SNPs generally, even after exclusion of outliers, displayed
higher levels of genetic differentiation. In conclusion, better estimates of population
structure could be achieved for this species by increasing the number of markers to,
in genomic terms, a relatively modest level.
Also for “classical” marine fish, i.e. widespread species with large population
sizes, high fecundity and pelagic eggs and larvae, examples of population genomics
has started to emerge. Moen et al. (2008b) identified and characterised 318 SNPs
in Atlantic cod from alignment of EST sequences. Overall their results demonstrate
and substantiate that Norwegian Coastal Cod and North-East Arctic cod represent
two highly isolated populations, which differ not only in neutral genes due to the
lack of migration, but have also diverged on a genomic level by adaptation to the
differences in local environment that they experience.
As can be seen from the examples above, SNPs are becoming the marker class
of choice for population genomic studies of marine fish. Although large numbers
of microsatellites are isolated (e.g. Karlsson et al. 2008) the attention of genomic
resource development for population genetics is turning more towards SNPs. They
possess several attractive features for ecology, evolution and conservation in general (see Morin et al. 2003) but for fisheries management in particular. First of
all management of most marine fish species are transnational and management
advice is delivered by international bodies such as ICES (International Council for
Exploration of the Sea). Accordingly, there is a need for easy calibration of markers among national fisheries labs to provide concerted data sets and help with the
validation. Since SNP genotyping is qualitative (i.e. different bases represent different alleles) compared to microsatellite scoring, which is quantitative (i.e. fragment
length is translated into alleles with different numbers of base-pairs) calibration
is much facilitated. On the down side, SNPs are less variable than microsatellites
(fewer alleles, commonly two) requiring 4–12 times more loci than microsatellites to attain the same statistical power (Liu et al. 2005a, Smith and Seeb 2008).
M.L. Cancela et al.
leading the way due to the high value of commercial and recreational fisheries but
probably most importantly due to their additional importance for aquaculture breeding and associated development of genomic resources (see Section 7.2 on Genomic
tools and resources). For example, Hayes et al. (2007) identified 2,507 SNPs from
the alignment of Atlantic salmon ESTs, which can be readily used for population
genomic studies.
An example of improving the performance of the number of genomic markers in salmonids can be found in Smith et al. (2007). They compared estimates of population differentiation among 16 collections of Chinook salmon
(Oncorhynchus tshawytscha) using traditional allozyme loci (22 loci), small tandem
repeats/microsatellites (nine loci) and 39 newly developed nuclear SNP loci. The
larger number of markers allowed testing for “outlier loci” (Beaumont and Nichols
1996) identifying five SNPs likely to be influenced by selection. Exclusion of these
loci, and three other, which had previously been suspected to be under selection,
provided a reduction in the estimates of differentiation and an associated increase in
the estimated migration among putative populations. Similarly, slight differences in
relationships among populations were apparent when comparing results from different marker classes, where SNPs generally, even after exclusion of outliers, displayed
higher levels of genetic differentiation. In conclusion, better estimates of population
structure could be achieved for this species by increasing the number of markers to,
in genomic terms, a relatively modest level.
Also for “classical” marine fish, i.e. widespread species with large population
sizes, high fecundity and pelagic eggs and larvae, examples of population genomics
has started to emerge. Moen et al. (2008b) identified and characterised 318 SNPs
in Atlantic cod from alignment of EST sequences. Overall their results demonstrate
and substantiate that Norwegian Coastal Cod and North-East Arctic cod represent
two highly isolated populations, which differ not only in neutral genes due to the
lack of migration, but have also diverged on a genomic level by adaptation to the
differences in local environment that they experience.
As can be seen from the examples above, SNPs are becoming the marker class
of choice for population genomic studies of marine fish. Although large numbers
of microsatellites are isolated (e.g. Karlsson et al. 2008) the attention of genomic
resource development for population genetics is turning more towards SNPs. They
possess several attractive features for ecology, evolution and conservation in general (see Morin et al. 2003) but for fisheries management in particular. First of
all management of most marine fish species are transnational and management
advice is delivered by international bodies such as ICES (International Council for
Exploration of the Sea). Accordingly, there is a need for easy calibration of markers among national fisheries labs to provide concerted data sets and help with the
validation. Since SNP genotyping is qualitative (i.e. different bases represent different alleles) compared to microsatellite scoring, which is quantitative (i.e. fragment
length is translated into alleles with different numbers of base-pairs) calibration
is much facilitated. On the down side, SNPs are less variable than microsatellites
(fewer alleles, commonly two) requiring 4–12 times more loci than microsatellites to attain the same statistical power (Liu et al. 2005a, Smith and Seeb 2008).