Publication Abstract

A positive relationship between fish abundance and spatial distribution suggests that indicators of spatial distribution could be used as proxies of abundance and stock status. This study evaluated the ability of 10 spatial indicators to classify the health status of 25 data-rich stocks, identifying the best performing indicators and the conditions that influence their performance using a combination of machine learning techniques. Spatial indicators were able to detect health status better than a random classifier for the majority of stocks studied. The occupancy and aggregation indicators showed the best overall performance, suggesting that years of healthy stock status were often associated with presence in more survey hauls or more evenly distributed fish density. However, indicator performance depended on the source of survey data used and varied stock-by-stock. Explanations of these findings are provided and future research is suggested to further explore the impact of survey design and environmental variables. Our findings highlight the potential for spatial indicators to improve the assessment of category 5 and 6 data-limited stocks managed by the International Council for the Exploration of the Seas (ICES).

Peter J. Kidd*, Laurence T. Kell, José A. A. De Oliveira*, Alexandra M. Collins

https://data.mendeley.com/datasets/3b3k3vrvpt/1