7 Genomic Approaches in Aquaculture and Fisheries
257
impact of the Leucine Aminopeptidase (LAP) gene involved in the osmoregulation of bivalves (e.g. Milkman and Koehn 1977, Koehn and Immermann 1981).
However, the number of loci which can be assessed is relatively limited and many
candidates may turn out not to be subject to selection and therefore unsuitable. A
good example of the inherent problems of using a candidate gene approach can be
found in Ryynanen and Primmer (2004), who examined variation in and around the
Growth Hormone (GH) gene in Atlantic salmon. Despite extensive sequence analysis of one of the most prominent candidates for growth differences in fish (De-Santis
and Jerry 2007), they were not able to find any genetic variation in exons, which
could be subject to differential selection among populations. Therefore “genome
scans” (e.g. Storz 2005) of large numbers of gene associated markers such as SNPs
and SSRs has become the preferred genomic tool for identification of adaptive divergence among populations (Beaumont 2005). Loci that display elevated levels of
genetic divergence among populations (high F st ) are most likely subject to selection.
The genes included can be completely random, thus providing evidence of the proportion and types of genes involved in local adaptation. For example, Lemaire et al.
(2000) detected in European sea bass caught in inshore and offshore habitats, six loci
(all allozymes), among 6 microsatellite and 18 allozymes, with above average F st
values. Alternatively, the scan can be “directed” by combining the genome scan with
a candidate gene approach, preferentially including genes with known function and
suspected to be involved in adaptation. This approach allows the information content
of the scan to be maximised and inferences on the genetic architecture of adaptive
traits to be made. Mäkinen et al. (2008) conducted a genome scan among seven
marine and freshwater populations of the three-spined stickleback (Gasterosteus
aculeatus). They found strong signatures of directional selection for two of the EST
derived microsatellites and the Eda associated indels, thus strongly suggesting local
adaptation. Likewise, Vasemägi et al. (2005) conducted a genome scan among populations of adult Atlantic salmon from different habitats (i.e. freshwater, brackish
and marine). They found that nine EST-associated microsatellites displayed highly
divergent patterns of genetic differentiation, and were thus likely to be candidate
genes for local adaptation in Atlantic salmon. Still the genome scans presented here
are based on a relatively small number of gene markers and could be vastly improved
by applying new large-scale sequencing methods such as “sequencing by synthesis”
(e.g. 454 sequencing). Naturally, genome scans are not restricted to spatial analysis, but can also be applied on a temporal scale to study potential evolutionary
change mediated by global change or selective fisheries. No large-scale temporal
genome scans have been published to date. However, Nielsen et al. (2007) investigated potential temporal selection on the Pan I (Panthophysin) gene in Atlantic cod.
By extracting DNA from up to 69 year old otoliths, they were able to compare levels
of genetic differentiation at the Pan I locus with a suite of microsatellites and found
that the temperature change in the investigated time period and area was too small
to have had any profound effect on Pan I allele frequencies. With the increasing
awareness of global change and the need to understand evolution in order to predict
the distribution and abundance of marine fish in a changing world, we expect temporal genome scans to play an increasingly crucial role. We also expect that in the
257
impact of the Leucine Aminopeptidase (LAP) gene involved in the osmoregulation of bivalves (e.g. Milkman and Koehn 1977, Koehn and Immermann 1981).
However, the number of loci which can be assessed is relatively limited and many
candidates may turn out not to be subject to selection and therefore unsuitable. A
good example of the inherent problems of using a candidate gene approach can be
found in Ryynanen and Primmer (2004), who examined variation in and around the
Growth Hormone (GH) gene in Atlantic salmon. Despite extensive sequence analysis of one of the most prominent candidates for growth differences in fish (De-Santis
and Jerry 2007), they were not able to find any genetic variation in exons, which
could be subject to differential selection among populations. Therefore “genome
scans” (e.g. Storz 2005) of large numbers of gene associated markers such as SNPs
and SSRs has become the preferred genomic tool for identification of adaptive divergence among populations (Beaumont 2005). Loci that display elevated levels of
genetic divergence among populations (high F st ) are most likely subject to selection.
The genes included can be completely random, thus providing evidence of the proportion and types of genes involved in local adaptation. For example, Lemaire et al.
(2000) detected in European sea bass caught in inshore and offshore habitats, six loci
(all allozymes), among 6 microsatellite and 18 allozymes, with above average F st
values. Alternatively, the scan can be “directed” by combining the genome scan with
a candidate gene approach, preferentially including genes with known function and
suspected to be involved in adaptation. This approach allows the information content
of the scan to be maximised and inferences on the genetic architecture of adaptive
traits to be made. Mäkinen et al. (2008) conducted a genome scan among seven
marine and freshwater populations of the three-spined stickleback (Gasterosteus
aculeatus). They found strong signatures of directional selection for two of the EST
derived microsatellites and the Eda associated indels, thus strongly suggesting local
adaptation. Likewise, Vasemägi et al. (2005) conducted a genome scan among populations of adult Atlantic salmon from different habitats (i.e. freshwater, brackish
and marine). They found that nine EST-associated microsatellites displayed highly
divergent patterns of genetic differentiation, and were thus likely to be candidate
genes for local adaptation in Atlantic salmon. Still the genome scans presented here
are based on a relatively small number of gene markers and could be vastly improved
by applying new large-scale sequencing methods such as “sequencing by synthesis”
(e.g. 454 sequencing). Naturally, genome scans are not restricted to spatial analysis, but can also be applied on a temporal scale to study potential evolutionary
change mediated by global change or selective fisheries. No large-scale temporal
genome scans have been published to date. However, Nielsen et al. (2007) investigated potential temporal selection on the Pan I (Panthophysin) gene in Atlantic cod.
By extracting DNA from up to 69 year old otoliths, they were able to compare levels
of genetic differentiation at the Pan I locus with a suite of microsatellites and found
that the temperature change in the investigated time period and area was too small
to have had any profound effect on Pan I allele frequencies. With the increasing
awareness of global change and the need to understand evolution in order to predict
the distribution and abundance of marine fish in a changing world, we expect temporal genome scans to play an increasingly crucial role. We also expect that in the