Towards a new recruitment paradigm for fish stocks

Abstract

The stock recruitment paradigm involves the hypothesis that recruitment (R) to a fish stock is positively related to the spawning stock biomass (SSB) of the stock, at low SSB. I propose a ``recruitment states'' hypothesis wherein R is independent of SSB but has different mean values during successive periods. Meta-analysis was used to test the null hypothesis that recruitment is a series of random, independent events, against these two alternative hypotheses, for 153 marine spawning bony fish stocks and 31 salmonid stocks. A test statistic for the stock recruitment paradigm, based on estimating derivatives from the first differences of the time series, was not significant for the marine stocks. The null hypothesis was rejected for the salmonid stocks. Recruitment states models significantly fitted time series for the marine stocks. Ricker models also significantly fitted these data, conflicting with the derivatives test result. However, because SSB is dependent on R, lagged by the age at maturity, a period in a low recruitment state would tend to lead to a period of low SSB. Therefore, the significance of the fit to the Ricker model may have been spurious. The recruitment states model best explained the meta-dataset for the marine stocks.