Pushing the limits of a data challenged stock: A size- and age-structured assessment of ling (Molva molva) in Icelandic waters using Gadget

In recent years a greater emphasis has been placed on developing management strategies that prevent over-exploitation. Harvest control rules (HCRs) have therefore, in many places, been developed and implemented. Commonly these HCRs are developed for stocks that are assessed using age-structured models, and various platforms exist to evaluate their performance and analyze various sources of bias for that particular class of models. Many stocks, however, cannot be assessed reliably using classical age-structured methods due to data limitations (gaps in data series, unreliable age readings, etc.). One such stock is the common ling (Molva molva) in Icelandic waters. Availability of data on the stock dynamics, in particular age data for both survey and commercial samples, has been a limiting factor when assessing the stock. When modeling stocks such as this, data limitations need to be considered, and how associated uncertainty is propagated both through the assessment and into the advice. In this study, ling was assessed using the size- and age-structured model Gadget after synthesizing all available data. Having limited age data available causes high uncertainty in the model fitting process, especially in estimating growth. However, including this key uncertainty in the assessment allowed the subsequent management strategy evaluation to take it into account directly while deriving common management reference points and estimating uncertainties in stock status and other derived quantities. Uncertainty was estimated using a specialized bootstrap for disparate data sets that mimics the sampling process. The process of assimilating data for the assessment model and the bootstrap procedure was performed using a specialized database program, MFDB, ensuring that the whole process is reproducible.