A TIME SERIES APPROACH TO STOCK-RECRUITMENT ANALYSIS: TRANSFER FUNCTION NOISE MODELLING

Perceived relationships between stock size and recruitment have long been a corner-stone of fisheries management. These relationships are often used to design harvest strategies for ensuring that sufficient spawning stock exists to generate desired levels of recruitment in subsequent years. However, existing models fail to recognize and exploit the autocorrelation structure of both the recruitment and stock time series. The time series approach to modelling stock and recruitment presented in this paper takes this autocorrelation structure into account. The performance of the time series model is compared to existing stock-recruitment models using North Sea herring and Pacific halibut data.