Title: Modelling environmental effects on growth of cod: fitting to growth increment data versus fitting to size-at-age data
Abstract: If size-at-age data are collected from annual surveys over a number of consecutive years then variability in growth can be investigated for association with environmental influences by formulating annual growth increments as a function of the environmental conditions prevailing at the time. Because the growth increments are not observed directly, but are calculated as the difference of size-at-age measurements, successive growth increments are statistically correlated. Methods of analysis of such data differ in that some studies ignore the correlation while others accommodate it. Here the performance of three different analytical approaches are compared. Their relative performance is found to depend on the relative amounts of random error and process error.