Abstract
The historical development of the idea of isolated stocks of fish that can be managed as separate management units has been as strongly tied to the intuitive idea of separate races as it has been to the practical necessities of jurisprudence and the estimation of both biomass and sustainable yield by fisheries managers. Demonstrating the existence of isolated fish stocks and delineating their boundaries has generally proved unsuccessful. Various techniques ranging from meristic count differences to polymorphic allelism have usually failed. However, in the pursuit of isolated stocks, biochemists have uncovered a great deal of information about the variation of polymorphic allele frequencies over time and space. Following the shift in opinion away from stochastic to natural selection mechanisms in allele frequency variation, it is evident that the observed variation in allele frequencies allows more insight into the biology of fishes than into the breeding structure of populations. These insights argue against the idea of isolated stocks of fish with homogeneous growth rates that are the basis of the sustainable yield models in favour of migration-linked stocks with heterogeneous growth rates.