René Holst

Selectivity experiments with escape windows in the North Sea Nephrops (Nephrops norvegicus) trawl fishery

Abstract Selectivity experiments were carried out in the northern North Sea at the Fladen Ground and the East Ground from a Danish commercial vessel. A conventional 70xa0mm Nephrops codend and a similar codend with a 90xa0mm square mesh panel window inserted in the top panel were tested simultaneously in a twin trawl rig. Selectivity parameters and differences in catches of the target species Nephrops and the by-catch species whiting, haddock and cod were estimated. The logistic function was found to give the best description of fish selection ogives, whereas the asymmetric complementary loglog function described Nephrops selection best. The Fryer model was used to describe and compare the selection ogives for haddock, whiting, cod and Nephrops in the two codends. Due to low catches of cod and Nephrops only four out of 11 hauls could be used for modelling the selection of these two species. Selection factor and selection range were larger for the window codend for all fish species. The codend catch size had a significant negatively correlated influence on l 50 and selection range for whiting in the window codend. There were large significant reductions in numbers of cod and haddock under the Danish MLS with the window codend. Catches of small cod, however, were low. Very few whiting under MLS were retained in either codend but there was a large significant reduction in individuals over MLS with the window codend. A large proportion of the Nephrops in the codends were under MLS. A reduction in catch weight of Nephrops over MLS was found for the window codend, but catches of Nephrops were low for several hauls.

Size selectivity of sole gill nets fished in the North Sea

The size selectivity of sole nets was measured indirectly on a Danish commercial gill net vessel in the North Sea by using nets made in seven different mesh sizes fished simultaneously. Maximum likelihood procedure was used to fit different forms for the selection curve to the catch data for sole, plaice and cod for each setting of the gear. It was found that a bi-normal form for the selection curve gave the best fits. Mean selection curves were then estimated by combining sets using a model of between-set variation. The ratio between length of maximum retention and mesh size was estimated to be 3.28 for sole, 2.60 for plaice and 4.56 for cod. Selection curves were also fitted to the catch data pooled over all sets. The model deviance for the sole and plaice data indicated lack of fit when pooling the catch data.

The effect of twine thickness on the size selectivity and fishing power of Baltic cod gill nets

Abstract Sea trials were carried out on a Danish commercial vessel measuring the size selectivity and fishing power of gill nets used to catch Baltic cod ( Gadus morhua ). A comparison was made of two different twine thicknesses at two different times of the year. Nominal mesh sizes of 70–130xa0mm were used. Method of capture, condition factor and girths were measured for sub-samples of the cod caught. A model of the size selectivity of the gill nets was adapted to the experimental conditions where two gears were fished on the same population. This model was fitted to the catch data for each set. Subsequently a model was fitted for the mean selectivity taking between-set variation into account. The selectivity curve that fitted the data best was given by the sum of two normal distributions. It was found that twine thickness and trials period had relatively little effect upon the shape of the selectivity curve. Twine thickness had a substantial effect upon the fishing power of the nets.