Spatiotemporal modeling of bycatch data: methods and a practical guide through a case study in a canadian arctic fishery

• Published in: Canadian Journal of Fisheries and Aquatic Sciences
• Main Authors: Yan, Yuan, Cantoni, Eva, Field, Chris, Treble, Margaret A, Mills Flemming, Joanna
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2022
• Subjects: info:eu-repo/classification/ddc/310; Bycatch; Gaussian random field; Geostatistics; Semi-continuous data; Two-part20model; TMB; Arctic; Greenland
• Online Access: https://archive-ouverte.unige.ch/unige:152640

Excess bycatch of marine species during commercial fishing trips is a challenging problem in7fishery management worldwide. The aims of this paper are twofold: to introduce methods and8provide a practical guide for spatio-temporal modelling of bycatch data, as well as to apply9these methods and present a thorough examination of Greenland shark bycatch weight in a10Canadian Arctic fishery. We introduce the spatially explicit two-part model and offer a step by11step guide for applying the model to any form of bycatch data, from data cleaning, exploratory12data analysis, variable and model selection, model checking, to results interpretation. We address13various problems encountered in decision making and suggest that researchers proceed cautiously14and always keep in mind the aims of the analysis when fitting a spatio-temporal model. Results15identified spatio-temporal hotspots and indicated month and gear type were key drivers of high16bycatch. The importance of onboard observers in providing robust bycatch data was also evident.17These findings will help to inform conservation strategies and management decisions, such as18limiting access to spatial hotspots, seasonal closures and gear restrictions.