Workshop 3 on Stock Identification and Allocation of Catches of Herring to Stocks (WKSIDAC3)

Most herring populations are migratory with variable overlap such that catches and surveys often cover a mixture of stocks. To monitor specific stocks, stock composition data based on ver-tebrae, otoliths or genetics have routinely been collected; however, the design of sampling strate-gies has received limited attention for this specific purpose. WKSIDAC3 reviewed current sam-pling strategies and designs. Further, simulation and case studies evaluated different sampling strategies and the effect of different sample sizes.

Most institutes are collecting genetic samples from herring in surveys and commercial catches. However, some gaps are still present, particularly in the southwestern North Sea, eastern Baltic Sea, and Icelandic Sea. The number of individual genotypes needed for inference on stock com-positions depends on the genetic difference between stocks, between haul variation, and the sub-sequent analysis. Simulation studies indicated that random sampling strategies should geno-type at least 30–50 individuals per haul and 300 per stratum. For length stratified sampling the minimum number per haul should be 50–70. When possible, increasing the number of hauls is more effective at increasing precision than increasing the number of genotyped individuals. The two case studies have also demonstrated this. However, further studies based on the specific sampling programs are needed. In some cases, fewer samples may be necessary.

Further, advancing stock identification and allocation of catches of herring to stocks depends on developing common, standardised methods for analysing genetic samples including across all herring populations as well as guidelines for the storage and accessibility of genotype data.