Population structure of Atlantic herring, Clupea harengus L., in the northeast Atlantic using phenotypic and molecular approaches
Most genetic attempts to study the population structure of herring (Clupea harengus L.) have been limited by the low levels of genetic differentiation observed among discrete spawning aggregations over large geographic scales. Thus, the population genetic structure of Atlantic herring remains undefined. Three sets of phenotypic (meristics, morphometries, otolith shape) and molecular (allozymes, mitochondrial DNA RFLPs, micro satellites) markers were simultaneously used to investigate the morphological and genetic structure of herring populations in the Northeast Atlantic and also to assess the relative usefulness of phenotypic and genetic markers in population identification. Samples were collected from the Celtic Sea, North Sea and fjords, Baltic Sea, Norwegian Sea, Barents Sea, and Pacific Ocean (pacific herring, Clupea pallasi L.). The Truss morphometric method was very sensitive in detecting morphological differences, revealing significant differences among all discrete spawning aggregations. Otolith analysis showed a lower discriminatory ability than the morphometries, differentiation of more widely separated populations, revealing a clear discreteness in the Icelandic, Baltic and Trondheimsfjord herring populations. Meristic analysis was more effective for the identification of different species rather than conspecific populations, exhibiting a clear divergence of the Trondheimsfjord and Icelandic herring samples. With allozyme electrophoresis, twenty-eight putative enzyme-coding loci were examined, and the result was in accordance with previous allozyme studies, showing genetic homogeneity among widely separated populations and localised heterogeneity in the Norwegian fjords (Trondheimsfjord). Also, an allozymically unique Norwegian spring-spawning (NWl) population was detected off the northern Norwegian coast. The number of low-frequency alleles apparently was a major problem with allozymes, limiting the overall ability to detect weak differences in allele frequencies between populations. PCR-based RFLP analysis of ND 3/4 and ND 5/6 regions of mtDNA with six restriction enzymes revealed significant genetic discreteness of the Baltic, Icelandic and Norwegian spring-spawner (NWl) herring. The results also showed a high level of haplotype diversity at the ND genes which contrasts with low levels of genetic divergence. This is apparently due mainly to the high number of unique haplotypes, and low number of common haplotypes detected, which reduced the power of the statistical test. Microsatellites were the most effective molecular marker, revealing genetically distinct Icelandic, Trondheimsfjord, Balsfjord and Norwegian spring-spawner (NWl) herring populations. The detected a high number of alleles and heterozygosity at microsatellite loci provide a new perspective on past estimates of detectable low levels of genetic differentiation among Atlantic herring populations. In identification of Atlantic and Pacific herring, meristic characters is most effective among phenotypic markers, allozymes and mtDNA are good choices among molecular markers. The significance of these findings m terms of stock separation and management are discussed.