Modelling length-at-age variability under irreversible growth

Abstract

In this paper, we describe a discrete-time formalism for describing the dynamics of the size-at-age distribution of a cohort of individuals exhibiting irreversible von Bertalanffy growth in a statistically uniform random environment. This formalism yields a highly efficient numerical implementation, which is particularly suited to automatic optimization. In the special case where mortality is sufficiently size-independent not to vary substantially across the bulk of the size distribution at any given age, we can further increase this efficiency by deriving compact update rules for the mean and coefficient of variation of size-at-age. In this case, we also demonstrate that the depensatory effect of random growth variability and the compensatory effect of deterministic von Bertalanffy growth balance to yield an attracting (initial condition independent) trajectory of mean length and length coefficient of variation against age. We demonstrate the applicability and extensibility of this formalism by two exemplary applications - juvenile salmonids and demersal cod.