Title: Comparison of different stock assessment models using data on the Icelandic summer-spawning herring (Clupea harengus) stock

In this study the results obtained by different stock assessment models are compared and their sensitivity to various assumptions is analysed. Extended Survivors Analysis (XSA), Adapt and Statistical Catch-at-Age methods are applied on the Icelandic summer-spawning herring data. The quality of the catch-at-age and acoustic survey data on the Icelandic herring is also analysed using the Shepherd-Nicholson method. The largest differences in results between the methods were obtained from XSA and Statistical Catch-at-Age. The estimate of stock size by XSA was around 30% lower and fishing mortality higher compared to the point estimates from Statistical Catch-at-Age. The difference between short-term predictions of yield was around 40%.

Uncertainty of results from Adapt and Statistical Catch-at-Age was estimated by bootstrapping. Significant differences between the results from these two methods were not found.

Retrospective analysis showed overestimation of spawning stock biomass and underestimation of fishing mortality for all methods. Differences in retrospective patterns between the methods were not observed.