Discards of cod (Gadus morhua) in the Norwegian coastal fisheries: improving past and future estimates

Discarding can be an unknown source of biases and uncertainties in stock assessments. Discarding patterns and quantities vary so a routine methodology for estimating discards is important to give a better picture of total catches, and potentially mortality, in fisheries. Using data from the Norwegia...

Full description

Bibliographic Details
Published in:ICES Journal of Marine Science
Main Authors: Berg, Hilde Sofie Fantoft, Clegg, Thomas L, Blom, Geir, Kolding, Jeppe, Ono, Kotaro, Nedreaas, Kjell Harald
Format: Article in Journal/Newspaper
Language:English
Published: Oxford University Press 2022
Subjects:
Online Access:https://hdl.handle.net/11250/2999554
https://doi.org/10.1093/icesjms/fsac081
Description
Summary:Discarding can be an unknown source of biases and uncertainties in stock assessments. Discarding patterns and quantities vary so a routine methodology for estimating discards is important to give a better picture of total catches, and potentially mortality, in fisheries. Using data from the Norwegian Reference Fleet between 2012 and 2018, this study presents a revised methodology for estimating discards of cod (Gadus morhua) in the Norwegian coastal gillnet fisheries, which accounts for variations in discarding between vessels and uncertainties in the conversion of numbers to weight discarded. The estimated average discard rate of cod (weight of cod discarded as percentage of total weight caught) is 0.55% (95% confidence interval: 0.45–0.70%), although discard rates in southern areas were an order of magnitude higher than in northern areas. We also present an exploratory analysis of the drivers behind discarding using a random forest regression model. Spatial variations and fishing intensity were identified as the most important drivers of discarding. Results from this study suggest ways in which self-sampled data can be used to estimate discards in Norwegian coastal fisheries, and where the accuracy of future estimates can be improved when a higher resolution data collection programme is established. publishedVersion