Rose et al.: Mortality rates for Chionoecetes opilio, C. bairdi, and Parahthodes camtschaticus after trawls on the seafloor 
45 
mercial trawler equipped with a trawl configured simi- 
larly to the one used by many of the bottom trawlers 
that are used in Bering Sea groundfish fisheries. The 
2-seam trawl net had a 36.0-m headrope and a 54.6- 
m footrope, which was made of 19-mm-long link -t-steel 
chain and equipped with bobbins 46 cm in diameter. 
The ~70-cm sections between bobbins were covered 
with 2 steel-chain toggles, weighing 6.4 kg each, rubber 
disks of 4-20 cm, and one 5-kg circular weight. Wing 
extensions, installed ahead of the forward ends of the 
footrope, were made of 20-em disks strung over 19-mm- 
long link chain. The cables (sweeps) that ran forward 
from the trawl to the doors were made of 48-mm combi- 
Figure 2 
Diagram of the trawl net (not to scale) used in our 
study of unobserved mortality rates for snow crab (Chi- 
onoecetes opilio ), southern Tanner crab (C. bairdi), and 
red king crab (Paralithodes camtschaticus), showing po- 
sitions of recapture nets designed to retain crabs after 
contact with various trawl components. No more than 2 
of these nets were fished during the same tow, and the 
control net always was fished separately. Illustration 
by Kama McKinney. 
nation rope, a product made of both steel and synthetic 
materials and used by most Bering Sea flatfish trawl- 
ers. The red king crab study included tests of sweeps 
equipped with disk clusters spaced at 14-m intervals 
and raised the combination rope 7.5 cm above the sea- 
floor. Rose et al. (2010) found that such raised sweeps 
reduced seafloor contact while still herding groundfish 
effectively. 
Crabs were captured immediately after contact with 
the components of the main trawl by small recapture 
nets fished behind these 3 gear regions: 1) at center of 
the footrope, 2) at the footrope wings (including their 
extensions), and 3) behind the sweeps (Fig. 2). These 
recapture nets were small trawls designed to minimize 
fish capture and maximize crab capture. The recapture 
nets used behind the wings and sweeps had unequal 
bridle lengths, which were adjusted until water passed 
perpendicular to the center of the headrope of each net, 
as observed with an underwater camera. An identical 
recapture net was fished ahead of the trawl as a con- 
trol to assess damage and mortality due to handling. 
A rope between the sweeps ahead of the control net 
was necessary to avoid overspreading. That rope was 
raised 23 cm off the bottom of the seafloor to avoid 
affecting crabs. Only 1 recapture net was used at a 
time during every tow in 2008 to ensure that nets did 
not tangle when launched. Experience allowed us to 
expand to 2 nets (1 sweep and 1 footrope) at a time 
during some tows in 2009; however the control net 
was always fished alone because it would potentially 
have damaged crabs before they reached the footrope. 
The time required to change positions of the recapture 
nets on the trawls precluded alternating them between 
trawl components on a tow-by-tow basis; therefore, all 
tows that addressed each gear component were done 
in 1 or 2 blocks of sequential tows. To maximize hold- 
ing times for crabs affected by the trawl, the control 
tows were done last. The codend of the main trawl was 
not closed because the tows were too short to represent 
typical mortality due to capture by the trawl, and catch 
volume was considered unlikely to significantly affect 
sweep and footrope mortality. 
Towing speeds were 3-3.5 kn. Tow lengths were kept 
short to minimize damage to crabs from the recapture 
process but varied from 7 to 25 min to capture suf- 
ficient numbers of crabs. These speeds reflect industry 
practice, and, although commercial tows last much lon- 
ger, the shorter lengths of the tows in our study did not 
change the relatively brief interactions between indi- 
vidual crabs and the ground contact components of the 
trawl. The main trawl was monitored with trawl sonar, 
which would detect any significant net asymmetry, and 
video observations of ground-gear components were 
used to check for atypical contact with the seafloor. 
Tow sites (Fig. 1) were selected to provide adequate 
numbers of the targeted crab species during relatively 
short tows. Both snow and southern Tanner crabs were 
sufficiently abundant to be studied at a single site in 
2008, but red king crab research in 2009 required an 
