Introducing inter-individual growth variability in the assessment of a cephalopod population: application to the English Channel squid Loligo forbesi

A new approach is presented here to better take into account inter-individual growth variability in age-structured models used for stock assessment. Cohort analysis requires knowledge of the age structure of the catch, generally derived from an age–length key and length-structure information. Age distribution at length is estimated by applying conditional quantile regression to a data set of lengths and ages estimated from calcareous parts. A “stochastic” age–length key that describes the probability of age-at-length is applied to the English Channel squid Loligo forbesi. Age distribution at length from quantile regression proved to be considerably less biased than that resulting from both polymodal decomposition (PD) and two separate slicing methods. Both catch-at-age and stock size were underestimated using classical methods. Estimations of fishing mortalities from classical methods were higher causing underestimation in yield simulations. Quantile regression offers a more complete statistical analysis of the stochastic relationships among random variables than mean regression and PD.