A full-scale fishing experiment on north-east Arctic cod (Gadus morhua) and
haddock (Melanogrammus aeglefinus), involving a trawler, a longliner and a
gillnetter was conducted in order to determine how the length distribution
s and life history parameters of the catches were affected by gear type. Re
lative selectivity was analysed for catches taken when the boats fished sim
ultaneously in the same area, and when the skippers were allowed to fish un
der conditions as close as possible to commercial operation. Trawl and long
line-caught cod had similar length distributions, but cod caught by trawl w
ere on average 2.3 cm smaller than longline-caught cod. Cod caught by longl
ine had a lower condition factor and length at age. The longliner caught sm
aller haddock than the trawler when the boats fished in the same area. When
the skippers were allowed to change fishing ground, the mean length of cod
increased in the trawl catches and became larger than the mean length in t
he longline catches. The mean length of haddock in the longline catches inc
reased by 2.8 cm and became larger than that of the trawl catches. Gillnet
catches consisted almost solely of large cod. Mean length of gillnet caught
fish in 186 mm nets decreased from 86.2 cm to 82.3 cm even if the boat did
not change area. Separate selectivity experiments for trawl and gillnet we
re conducted, and the parameters for the mesh selection are presented. To f
ind the functional form of the longline selectivity we used relative catch
proportion in each 5 cm length-group for longline compared to trawl and gil
lnet, whose selectivities were assumed to be known. When compared to that o
f the gillnet, the plots indicate that the longline selectivity curve takes
the form of a unimodal distribution. When longline catches was compared to
trawl catches none of the suggested models seemed to fit, and to conclude
anything about the form of the longline selection curve is not possible. (C
) 2000 International Council for the Exploration of the Sea.