286 
Fishery Bulletin 108(3) 
Model input 
Expert judgment on the fraction of rockfish in the 
water column that lie below the height of the 
headrope of the approaching trawl gear 
Expert judgment on the fraction of rockfish above 
the headrope that dive below the height of the 
headrope as the vessel passes 
Expert judgment on the fraction of rockfish outside V 
of the path of the net that are herded by the doors ! 
and warps into the path of the net ■ 
■■■■■■■■■■■■■■■■■a... 
( 
Expert judgment on the fraction of rockfish in front of « 
the path of the net that are captured by the net : 
; Observations of the ratio of catch rates between * 
different survey gears in the same area ■ 
; : •. : : ■. ; ■„ : ; : •. ; : : ; : : : : ■. ; ; : : : : : : : : : : ; ; : : ■. : : ; : ; : : . 
■ Expert judgment and experimental data on the 
l difference in rockfish density between trawlable and 
• untrawiable areas 
; Data on the fraction of trawlable area in each survey 
■ 
I Average swept area estimates of rockfish population in 
: each survey area from groundfish trawl surveys 
Minimum uncertainty factor 
Submodels 
q ne i model for each 
survey gear for each 
trawl expert 
Model to combine q nel priors 
from different experts for each 
survey gear 
Model results 
q ne i distribution for 
each survey gear for 
each trawl expert 
Model for the bias in swept 
area biomass from 
differences in density 
between trawlable and 
untrawiable bottom in each 
survey area 
Mixture distribution for 
q ne t for each survey 
gear 
Estimate of the bias in swept 
I area biomass from 
1 differences in density 
. between trawlable and 
Model for the fraction of the 
population in each survey 
area for each survey gear 
untrawiable bottom in each 
survey area 
Estimate of the fraction of the 
population in each survey 
area for each survey gear 
Qgross model for each survey 
Distribution of q gross for 
each survey including 
the correlation in q gross 
between the different 
surveys 
Figure 2 
Outline of the structure of the trawl survey catchability model, showing the flow of information from model inputs to 
the model components (submodels) where the inputs are processed and integrated, to the final model results. See Table 
1 for definitions of the variables. 
q available ~ the proportion of S y in the survey area; 
and 
q trawlable ~ the average ratio of rockfish density be- 
tween trawlable and untrawiable areas 
adjusted by the fraction of the seabed 
within the surveyed area that is trawlable. 
We present conceptual models and equations for each 
of these components below. 
Quantifying catchability with the trawl survey net ( q net ) 
In most instances, results from experiments designed to 
estimate q net for the survey gears and fish populations of 
interest are unavailable. Since the 1990s, some research- 
ers have developed priors for q net by integrating, within 
a Monte Carlo simulation model, expert judgment on the 
components of q net and, in some instances, auxiliary data 
(Punt et al., 1993; McAllister and Ianelli, 1997; Boyer 
et al., 2001). Here we present a protocol for an approach 
that can be applied to estimates of biomass from trawl 
surveys when several experts provide key information, 
the population is surveyed by one or more types of trawl 
gears and, in one or more areas, records of catch rates 
from two or more types of trawl gears are available. 
We first present a conceptual model for the compo- 
nents of q net . It is assumed that a trawl net captures 
less than 100% of the fish that lie in its path, defined 
over the horizontal as the path between the trawl doors 
and over the vertical as the area from the surface to the 
bottom. Fish can escape for a variety of reasons includ- 
ing, but not limited to (Fig. 3), the following: 
1 they are initially high up in the water column and 
do not “dive” to lie below the oncoming headrope of 
the trawl; 
2 they are near bottom but are driven away horizon- 
tally by the influence of the warps near the doors as 
they spread outwards towards the doors; 
3 they are initially in front of the paths of the sweeps 
and bridles but are not herded into the path of the 
net; 
4 they escape over the headrope or under the foot- 
rope; 
