Estimating natural mortality for Atlantic Sea scallops (Placopecten magellenicus) using a size-based stock assessment model Fisheries Research

We estimate temporally varying natural mortality within a size-based stock assessment model for U.S. sea scallops in three different regions, and compare the results to corresponding models where natural mortality is not estimated, and (in one case) where only the mean natural mortality is estimated. In one model, for the Georges Bank Closed Areas, natural mortality was estimated by year for all sizes, whereas in the other two models (for the Georges Bank Open Areas and the Mid-Atlantic Bight), annual variation in natural mortality was estimated for juveniles only. Estimating natural mortality by year improved the performance over fixed natural mortality in all three models, as measured by Akaike’s information criterion (AIC) and reduced retrospective patterns. There was evidence that natural mortality is density-dependent. Large recruitment events were associated with substantial increases in juvenile natural mortality, and natural mortality in the Georges Bank Closed Areas tended to increase at high biomass. Estimation of natural mortality was facilitated by survey data that gave estimates of absolute abundance together with low fishing mortality that minimized the confounding between natural and fishing mortality. Natural mortality is an important process that should be estimated when there is sufficient information to do so.