Spatiotemporal modeling of mature‐at‐length data using a sliding window approach

• Published in: Environmetrics
• Main Authors: Yan, Yuan, Cantoni, Eva, Field, Chris, Treble, Margaret, Mills Flemming, Joanna
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2022
• Subjects: info:eu-repo/classification/ddc/310; Arctic; Greenland; Somniosus microcephalus
• Online Access: https://archive-ouverte.unige.ch/unige:163573

Excess bycatch of marine species during commercial fishing trips is a challenging problem in fishery management worldwide. The aims of this paper are twofold: to introduce methods and provide a practical guide for spatiotemporal modelling of bycatch data, as well as to apply these methods and present a thorough examination of Greenland shark (Somniosus microcephalus) bycatch weight in a Canadian Arctic fishery. We introduce the spatially explicit two-part model and offer a step by step guide for applying the model to any form of bycatch data, from data cleaning, exploratory data analysis, variable and model selection, model checking, to results interpretation. We address various problems encountered in decision making and suggest that researchers proceed cautiously and always keep in mind the aims of the analysis when fitting a spatiotemporal model. Results identified spatiotemporal hotspots and indicated month and gear type were key drivers of high bycatch. The importance of onboard observers in providing robust bycatch data was also evident. These findings will help to inform conser