Abstract

Body mass affects many components of demography and therefore has consequences for management and conservation. Growth is typically affected by environmental variables such as population density and weather. Terrestrial animal demographers often model growth in body size using time-series of body masses, whereas fish demographers more frequently use length-at-age models such as the Von Bertalanffy or Gompertz growth models. However, typical length-at-age models are less flexible for quantifying transient variability in growth. To use length-at-age models more flexibly, we converted masses to lengths and modeled growth increments from Von Bertalanffy and Gompertz growth models. We compared these methods with regression models of weight-at-age time series. Regression methods we considered included glmmTMB, glmmLasso, and a state-space model that controls for observation error. We then evaluate the ability of these models to forecast with the goal of projecting future fish body weights under various climate change scenarios.