Abstract
This paper describes a method of evaluating virtual codend selectivity with covernet selectivity when the mesh size of the covernet is not small enough to retain all the fish that enter the covernet. If the covernet mesh size is not small enough, small fish can escape through the covernet mesh after passing through the codend, and then their body length proportion retained in the codend will be overestimated. We developed a model in which covernet selectivity can be estimated at the same time as accurate codend selectivity. The validity of the method was examined with data from a covered-codend experiment on the Patagonian grenadier Macruronus magellanicus collected in the southwest Atlantic. The model fits the data well even though the plots of proportion retained were U-shaped. For both the codend and covernet selectivity estimated in the model, fish of length close to 100% selectivity had a girth close to mesh perimeter. This supported that both the codend and covernet selectivity estimated in the model were reasonable. The analysis proposed in the present study would be useful to obtain accurate estimates of codend selectivity by evaluating the effects of covernet selectivity on the data.
