7 Genomic Approaches in Aquaculture and Fisheries
259
“Barcoding Of Life” initiative (www.fishbol.org; see Chapter 3). However, inferences on the population of origin of a sample of fish and/or individual fish rarely rely
on fixed genetic difference among populations, since they are connected by gene
flow. Instead they have to use statistical methods for calculation of the most likely
population of origin. Most commonly the genome is used, but also the transcriptome may identify populations (Larsen et al. 2007) or separate wild from culture
fish (Roberge et al. 2006). The two most widely used methods for GSI (Genetic
Stock Identification) are “Individual Assignment” (IA; see reviews by Hansen et al.
2001 and Manel et al. 2005) and Mixed Stock Analysis (MSA; Pella and Masuda
2001).
The principle of IA is that an individual is assigned to the population, out of a
number of potential populations of origin, where its multi-locus genotype has the
highest likelihood of occurring. For example, Nielsen et al. (2001) were able to
almost unambiguously assign individual cod form the North Sea, Baltic Sea and
Barents Sea to their population of origin. This ability has subsequently been used
in a court case, where a Danish fisherman was accused of fishing illegally in the
North Sea, while he claimed that the fish were legally caught in the Baltic Sea.
DNA analysis of five cod from the catch and subsequent assignment tests using 10
microsatellites revealed that all fish were assigned to the North Sea. The fisherman
was subsequently convicted (E.E. Nielsen, personal communication). Likewise,
Primmer et al. (2000) used assignment tests to exclude an unusually large Atlantic
salmon presented at a fishing contest in Lake Saimaa (Finland) as a local fish. Its
genotype was 600 times more likely to correspond to populations commonly sold
on fish markets in Finland. When confronted with the evidence the angler confessed
to the fraud.
In contrast to IA, MSA estimates the most likely proportions of individuals, in a
potentially mixed sample, originating from each of a number of sampled base-line
populations. Ruzzante et al. (2006) used MSA to estimate the most likely origin of
mixed population aggregations of herring in Skagerrak. They found that the proportions of fish from the North Sea, Skagerrak and Kattegat/Western Baltic varied
among juveniles and adults and summer/winter, while the results between years
were remarkably stable. It is now possible to manage fisheries in the area, to avoid
overfishing of vulnerable populations by targeting specific areas and time periods
where thriving populations are abundant. MSA has also been employed in forensic
cases, in particular for Pacific salmonids. Withler et al. (2004) reported 17 forensic cases of illegal fishing or selling of Sockeye (Oncorhynchus nerka) and Chinook
salmon from the Fraser River, Canada. Microsatellite analysis and subsequent use of
both MSA and IA for fish sampled at the harvester as well as in restaurants revealed
fraud most of the time. An additional application of genetic methods for traceability
and forensics is for the identification of individuals, parentage and families as used
in livestock (Dalvit et al. 2007). Identification of individual fish in wild populations
of marine fish is not common, but could be highly relevant for large and/or particularly valuable specimens such as tuna or whales. For example, Baker et al. (2007)
assessed unique genotypes of North Pacific minke whales (Balaenoptera acutorostrata spp.) sampled at 12 surveys of fish markets in the Republic of South Korea
259
“Barcoding Of Life” initiative (www.fishbol.org; see Chapter 3). However, inferences on the population of origin of a sample of fish and/or individual fish rarely rely
on fixed genetic difference among populations, since they are connected by gene
flow. Instead they have to use statistical methods for calculation of the most likely
population of origin. Most commonly the genome is used, but also the transcriptome may identify populations (Larsen et al. 2007) or separate wild from culture
fish (Roberge et al. 2006). The two most widely used methods for GSI (Genetic
Stock Identification) are “Individual Assignment” (IA; see reviews by Hansen et al.
2001 and Manel et al. 2005) and Mixed Stock Analysis (MSA; Pella and Masuda
2001).
The principle of IA is that an individual is assigned to the population, out of a
number of potential populations of origin, where its multi-locus genotype has the
highest likelihood of occurring. For example, Nielsen et al. (2001) were able to
almost unambiguously assign individual cod form the North Sea, Baltic Sea and
Barents Sea to their population of origin. This ability has subsequently been used
in a court case, where a Danish fisherman was accused of fishing illegally in the
North Sea, while he claimed that the fish were legally caught in the Baltic Sea.
DNA analysis of five cod from the catch and subsequent assignment tests using 10
microsatellites revealed that all fish were assigned to the North Sea. The fisherman
was subsequently convicted (E.E. Nielsen, personal communication). Likewise,
Primmer et al. (2000) used assignment tests to exclude an unusually large Atlantic
salmon presented at a fishing contest in Lake Saimaa (Finland) as a local fish. Its
genotype was 600 times more likely to correspond to populations commonly sold
on fish markets in Finland. When confronted with the evidence the angler confessed
to the fraud.
In contrast to IA, MSA estimates the most likely proportions of individuals, in a
potentially mixed sample, originating from each of a number of sampled base-line
populations. Ruzzante et al. (2006) used MSA to estimate the most likely origin of
mixed population aggregations of herring in Skagerrak. They found that the proportions of fish from the North Sea, Skagerrak and Kattegat/Western Baltic varied
among juveniles and adults and summer/winter, while the results between years
were remarkably stable. It is now possible to manage fisheries in the area, to avoid
overfishing of vulnerable populations by targeting specific areas and time periods
where thriving populations are abundant. MSA has also been employed in forensic
cases, in particular for Pacific salmonids. Withler et al. (2004) reported 17 forensic cases of illegal fishing or selling of Sockeye (Oncorhynchus nerka) and Chinook
salmon from the Fraser River, Canada. Microsatellite analysis and subsequent use of
both MSA and IA for fish sampled at the harvester as well as in restaurants revealed
fraud most of the time. An additional application of genetic methods for traceability
and forensics is for the identification of individuals, parentage and families as used
in livestock (Dalvit et al. 2007). Identification of individual fish in wild populations
of marine fish is not common, but could be highly relevant for large and/or particularly valuable specimens such as tuna or whales. For example, Baker et al. (2007)
assessed unique genotypes of North Pacific minke whales (Balaenoptera acutorostrata spp.) sampled at 12 surveys of fish markets in the Republic of South Korea