Heery and Cope: Co-occurrence of bycatch and target species in the groundfish trawl fishery 
37 
Figure 1 
Map of the study area from northern Washing- 
ton to southern California where catch data 
were collected by observers from 2002 to 2009. 
In cluster analyses, clusters containing rebuild- 
ing species were most evident when the data for 
each port were evaluated separately. The map 
highlights examples of port groups in which 
relationships between rare rebuilding rockfish 
and target species were identified in the com- 
mercial catch. Courtesy of M. Bellman 
role of observer data for management is for the calcula- 
tion of bycatch ratios. Bycatch ratios reflect the amount 
of catch of incidental (nontarget) species that occurs 
in relation to the amount of retained catch of species 
that are targeted by the fishery. Managers produce 
projected estimates of bycatch for nontarget species on 
the basis of such ratios in conjunction with anticipated 
landings (Bellman and Heery, 2013). This approach 
assumes that there is a proportional relationship be- 
tween bycatch and landings of target species (Rochet 
and Trenkel, 2005). 
Several previous studies have examined assem- 
blages among groundfish species through the use of 
data from fishery-independent surveys (Gabriel and 
Tyler, 1980; Weinberg, 1994; Jay, 1996; Williams and 
Ralston, 2002; Tolimieri and Levin, 2006; Zimmerman, 
2006; Cope and Haltuch, 2012). There have been fewer 
studies of species associations with the use of fishery- 
dependent data. Lee and Sampson (2000) used logbook 
data to evaluate species composition in the bottom 
trawl fishery. Trawl logbooks are maintained by vessel 
captains and include only species that are retained and 
landed. Their study, therefore, did not include species 
that were also caught by trawl nets but that were dis- 
carded because of economic or regulatory constraints. 
Rogers and Pikitch (1992) used observer data to iden- 
tify species assemblages, but participation in the ob- 
server program that produced their data was volun- 
tary. The data available for that study were collected 
between 1985 and 1987, a period in which fishery prac- 
tices and regulations differed considerably from those 
used more recently. 
This study presents a current view of species co- 
occurrence onboard commercial vessels in the bottom 
trawl fishery, and with the use of data from a man- 
datory at-sea observer program conducted yearly from 
2002 to 2009, is more comprehensive than that of pre- 
vious studies. Three major questions were explored: 
1) Are there identifiable associations between species 
caught in the bottom trawl fishery? 2) Do species that 
are undergoing population rebuilding toward target 
biomass levels (“rebuilding species”) cluster with tar- 
geted species in a consistent way? 3) Are the relation- 
ships between rebuilding species and target species 
more resolved at particular spatial scales or are re- 
lationships spatially consistent across the whole data 
set? All groundfish species that were managed under 
federal rebuilding plans during the study period were 
considered as rebuilding species. The results from our 
analysis relate indirectly to species assemblages in the 
marine environment. However, the study is primarily 
relevant in the context of fisheries management be- 
cause it provides insight into the relationship between 
bycatch of nontarget species and catch of targeted spe- 
cies in the commercial catch of the demersal trawl fish- 
ery on the west coast of the United States. 
Materials and methods 
At-sea observer data 
Observer data were obtained from the West Coast 
Groundfish Observer Program (WCGOP), part of the 
National Marine Fisheries Service (NMFS). The WC- 
GOP employs a stratified multistage random sampling 
design in which permits for the coastwide limited entry 
trawl fishery are selected for 2-month periods without 
replacement until all permits in the fleet are observed. 
The vast majority of permits were linked to individual 
