Pohnpei, Federated States of Micronesia STOCK ASSESSMENT OF SWORDFISH (Xiphias gladius) IN THE SOUTHWEST PACIFIC OCEAN

2 Executive summary This paper presents the 2013 assessment of swordfish in the southwest Pacific Ocean, updating the previous assessment done in 2008. This assessment is supported by several other analyses which are documented separately, but should be considered when reviewing this assessment as they underpin many of the fundamental inputs to the models. These include standardised CPUE analyses of aggregate Japanese and Chinese Taipei longline catch and effort data (Hoyle et al. 2013); standardised CPUE analyses of operational catch and effort data for Australian longline fishery (Campbell 2012); and New Zealand (OFP, 2013), and for the Spanish (i.e. European Union) (OFP, 2012). The assessment includes a new “reference case 2 ” model (Ref.case), and then a series of “one-off ” sensitivity models that represent a single change from the Ref.case model run. The key model runs were taken as representing a set of plausible model runs, and these were included in a structural uncertainty analysis (grid) for consideration in developing management advice. The main developments to model structural assumptions were to: assume two model regions, biologically connected, based on the results of recent electronic tagging programmes, and relaxing assumptions such as the relative recruitment to each region; fixing steepness at 0.8; estimating spline and non-decreasing selectivities for the main longline fisheries. A new statistical assumption was to include time-variant precision in fitting the model to standardized CPUE indices. A summary of these and the alternative assumptions for Ref.case and the other key model runs as agreed at a pre-assessmen