Enhancing the resolution of fishing dynamics in Taiwan's mixed-trawl fisheries: Investigating gear type operation patterns through vessel tracks

Abstract

Vessel tracking systems, such as voyage data recorders (VDRs), facilitate the spatial and temporal analysis of fishing behaviors. In multi-gear fisheries, diverse operational characteristics influence the accuracy of speedbased classification of fishing activities. Our approach adjusts fishing speed criteria according to vessel tonnage and gear type, addressing the operational differences among bottom and mid-water trawls. To enhance information on fishing effort and investigate the distribution of fishery resources in Taiwan's mixed-trawl fisheries, we conducted stratified random sampling from 2010 to 2020, covering various vessel sizes and gear types. Manually labeled VDR data were used to establish speed criteria tailored to these factors, achieving overall accuracy and true skill statistic values of 0.91 and 0.80, respectively. This approach notably reduces misclassification of activities for mid-water trawls. Additionally, Pearson correlation analysis demonstrates a strong positive relationship between fishing effort and catch, supporting more precise spatial allocation of landings. These findings offer valuable insights for improving fishery management, providing a robust framework for resource allocation and enhancing our understanding of the operational differences in multi-gear fisheries. Finally, this research lays a foundation for future studies aimed at refining classification methods for fishing activities.