McAllister et al.: Using experiments and expert |udgment to model catchability of Pacific rockfishes 
287 
A 
Wingtips 
Footrope 
B 
Between doors 
Direction of 
movement 
Between wingtips C 
Dl Live zone (to be herded) . 
rope 
^D2 Dead zone.,--''" i 
± Heac 
W; 
irp D 
1 
3or 
Foot 
rope £ 
Figure 3 
Diagram of fishing zones in the path of the bottom trawl. (A) Area A is the area 
below the boat between the doors and beneath the headrope. Area B is the area 
to the surface above area A. Area C is the area beneath the headrope to the 
footrope and between the net wingtips. Area D is the area beneath the headrope 
to the footrope and between the doors and the net wingtips. Area E is the area 
between the wingtips and beneath the headrope immediately in front of the 
trawl, i.e., the so called “kill zone,” or capture zone. (B) Expanded top view and 
(C) side view showing the trawl warps, sweeps, and net configuration. Dl, the 
“live zone” of the gear is the segment immediately outside of the wingtips in 
which rockfish are herded into the net path. D2, the “dead zone” of the gear, is 
the region just inside the trawl doors in which rockfish present escape from the 
pathway of the net. This conceptual scheme was used in interviews to enable 
trawl captains to express their views about the faction of rockfish in the water 
column that ended up in the trawl net. 
5 they are captured in the last 
few minutes of the tows and 
escape during retrieval (note 
that the DFO groundfish 
survey tows along the bottom 
last usually 19 minutes and 
in our application none of 
these fish were assumed to 
have escaped). 
All of these potential sources 
of escape are factored into our 
catchability model. 
We assumed that 
1 q net is constant among areas 
for the same type of trawl 
net; 
2 q net pertains to fishing during 
a bottom trawl survey, as 
opposed to commercial fish- 
ing. This assumption was 
emphasized to the trawl 
captains so that they would 
provide specifications based 
on standard trawl survey 
operations as opposed to com- 
mercial operating conditions; 
and 
3 q net does not vary with abun- 
dance. 
t ? fral v/ 0 6/e- differe,1CeS « n fisl1 
density between untrawlable 
and trawlable areas 
Trawl captains and groundfish 
researchers believe that the den- 
sities of rockfishes are higher 
over untrawlable bottom than 
over trawlable bottom. Note 
“untrawlable” is an operational 
distinction that reflects any type 
of bottom relief that trawl cap- 
tains (research and commercial) 
judge as presenting too much 
risk for damage to the trawl 
gear. These opinions are based 
on the tendency for catch rates 
for virtually all rockfish spe- 
cies to be higher on, or nearer, 
rougher bottom, as well as the 
tendency for untrawlable bottom 
to be associated with a much 
stronger acoustic signal for rockfish, and on the basis 
of submersible studies, which indicate the tendency of 
rockfishes to be associated with rugged habitat (Krieger, 
1993). 
Estimates of biomass over swept areas are usually 
computed by assuming that the average catch rate of 
survey hauls in a given area (stratum) is a random 
sample of the entire survey area and, when multiplied 
by the total survey area, will provide the biomass index 
for the survey area. In this section, we present a bias 
correction factor to account for the average relative dif- 
ference in fish density between trawlable and untraw- 
