Skip booklet index and go to page content

COSEWIC assessment and status report on the Bocaccio in Canada

Population Sizes and Trends

Commercial Fishing Trawl Landings and Abundance Indices

The initial part of the discussion, regarding trends in population abundance, focuses on trends in landings or catch, or catch-per-unit-effort (CPUE) in the commercial fishery. We emphasize that the catch and CPUE time series have questionable utility as abundance indices for bocaccio. Catch statistics, since the introduction of 100 % dockside and observer coverage by 1996, are considered reliable. Prior to this period, and particularly in the 1985-1995 period, estimates of catches are unreliable (see Stanley and Kronlund 2001).

Readers should note that a management plan based on Individual Vessel Quotas (IVQ) was introduced for the B.C. trawl fishery in 1997. Thus, harvests for all species managed through quotas prior to 1997 are now controlled by assigning area-specific annual catch (retained and discarded) limits to each vessel. Bocaccio are not limited through IVQ’s but are constrained by a 15 000 lb trip limit for all non-quota rockfish combined. The frequency distribution of catch weights of bocaccio in bottom and midwater tows is shown in Fig. 13. Current total coastwide catches of 295 t (Table 1) correspond to about 74 000 individuals, assuming a mean weight of 4 kg (PacHarvHL database, see Appendix Table 2).


Figure 13: Frequency Distribution of Catch Weights of Bocaccio in Bottom and Midwater Tows

Figure 13: Frequency distribution of catch weights of bocaccio in bottom and midwater tows.

Recorded coastwide catches of bocaccio have varied from 90 to 1322 t with a mean of 418 t since 1967 (Table 1, Appendix Tables 1 and 2). Coastwide commercial trawl catches have ranged between 200 and 300 t since the introduction of 100% observer coverage in the trawl fleet in 1996. Not included in commercial statistics are the discards in the commercial hook-and-line fisheries and recreational fisheries. While bocaccio are reported as a nuisance during salmon troll fishing, particularly in Area 5E (F. Crabbe and A. Amos, pers. comm. Appendix 1), actual catches are probably low relative to reported commercial trawl catches.

The trawl landings from the southwest coast of Vancouver Island (3C) and Queen Charlotte Sound (5A and 5B) show no consistent trend in recent years (1996-2001) (Table 1). While catches of bocaccio are widespread in northern waters, they have always been much lower than in the central and southern areas (Table 1). They currently average less than two-thirds their long-term mean. Landings are constrained by the 15 000-lb trip limit for “non-quota” rockfish. Trawl fishers report that landings could return to the long-term average of 400 t if the restrictions were relaxed (R. Gorman, pers. comm. Appendix 1). Thus, bocaccio may be more numerous in northern waters than what can be inferred from current trawl catches.

We were also informed by trawl fishers that bocaccio are often caught when targeting on canary rockfish (S. pinniger) (B. Dickens, pers. comm. Appendix 1). The current IVQ’s for canary rockfish are so low that fishers rarely target on them. They catch their canary rockfish IVQ’s as incidental to other targeting. Thus, the low IVQ’s of canary rockfish and the 15 000 rockfish trip limit act to constrain the trawl catches of bocaccio.

We present median CPUE by region for the 1996 to 2001 time period (Fig. 14). Earlier catch rate data cannot be used to infer abundance trends owing to the large variation in management actions that acted to change fishing strategies over time, the variable amounts of mis-reporting of catches over time, and inadequate data on species composition and effort from earlier years (see Stanley and Kronlund 2001). The CPUE indices were derived from bottom trawl tows for which the midpoint of bottom depth was between 77 and 309 m. The commercial CPUE indices show little change since 1996 for all four areas.

There have been no reported trawl landings of bocaccio from the Strait of Georgia (Minor Areas 13-19, 28 and 29) since 1983 (Table 3). However, in recent years commercial trawlers have been prohibited from retaining rockfish from all of Major Area 4B and no observers have been placed on these vessels. Landings in the hook-and-line fishery are too small to provide abundance indices for more recent years (see Appendix Table 1 in “unknown” category). Salmon trollers have commented that there can be a significant by-catch of bocaccio discarded in the outer coast salmon troll fishery (A. Amos, R.A. Best, R.N. Best, I. Bryce, F. Crabbe, pers. comm. Appendix 1), however this fishery is much smaller than in previous years.


Figure 14: Box and Whiskers Plot of Bocaccio CPUE (log kg/hr) in the Commercial Trawl Fishery in Each Coastal Region: (a) 3CD; (b) 5AB; (c) 5CD; (d) 5E


Figure 14. Box and whiskers plot of bocaccio catch per unit effort (log kg/hr) in the commercial trawl fishery in each coastal region: 3CD; 5AB; 5CD; 5E.

For each box, the upper and lower bounds indicate the 75th and 25th percentiles, respectively; the central horizontal line indicates the median; the upper and lower whiskers are positioned at 1.5 times the inter-quartile range; and the open circles indicate values that fall outside the whiskers. However, these data may have questionable utility as abundance estimates.


Survey-based Indices

A number of surveys have been conducted on the B.C. coast. Although the surveys were not designed to focus on bocaccio, we examined them for utility in tracking bocaccio abundance. We summarize the results below.

U.S. Triennial Bottom Trawl Survey (1980-2001)

The strongest data set on population trends is the U.S. triennial bottom trawl survey, which began in 1977 and typically covers northern California to the U.S./Canada border in northern Washington (Shaw et al. 2000). It was extended into southern B.C. waters for 7 sampling years. The first two surveys extended to 49°15' N; the latter four surveys extended further north to 49°40' N (Fig. 15a and d). Biomass estimates are computed for all depths combined with 95% confidence limits. The initial trawl catch rate data are extrapolated to a biomass estimate based on an area-swept logic. While presented as biomass estimates in Fig. 15a, the presumed low catchability in bottom tows for a semi-pelagic species implies that the survey is used best as a relative index. The survey is also imprecise owing to the low number of tows conducted in bocaccio depths near the 100-fathom contour (see Figs. figure16 and figure17).


Figure 15: Abundance Indices from (a) Biomass Estimates of the U.S. Triennial Survey in Area 3C and Part of Area 3D Plotted on a Natural Log Scale, (b) Bocaccio Stratified Mean CPUE (kg/hr; natural log scale) in the Queen Charlotte Sound Pacific Ocean Perch Survey, (c) Bocaccio Stratified Mean CPUE (kg/hr; natural log scale) in the Hecate Strait Multispecies Assemblage Survey

Figure 15: Abundance indices from biomass estimates of the U.S. triennial survey in Area 3C and part of Area 3D plotted on a natural log scale, bocaccio stratified mean CPUE in the Queen Charlotte Sound Pacific ocean perch survey, and bocaccio stratified mean CPUE in the Hecate Strait multispecies assemblage survey.


Figure 16: Trawl Locations for the Northern Portion of the U.S. Triennial Survey

Figure 16: Trawl locations for the northern portion of the U.S. Triennial Survey.

Zero catch represented by “+”’s (figure provided by Mark Wilkins, pers. comm. Appendix 1.


Figure 17: Number of Bocaccio caught in Canadian waters during the U.S. Triennial Surveys of 1980, 1983, 1989, 1992, 1995, and 1998

Figure 17: Number of Bocaccio caught in Canadian waters during the U.S. Triennial Surveys of 1980, 1983, 1989, .Figure 17: Number of Bocaccio caught in Canadian waters during the U.S. Triennial Surveys of 1992, 1995, and 1998.

Figure provided by Mark Wilkins, NMFS.

Over the past 10 years, the decline has been over 90% (1989 = 2348 t (CV = 0.52), 2001 = 157 t (CV = 0.84), 93.3% decline). Over the past 20 years, the decline has been over 95% (1980 = 6541 t (CV=0.94), 2001 = 157 t (CV = 0.84), 97.6%). These results (Fig. 15a) indicate a decline of almost two orders of magnitude over the last two decades.

 

Area 3C and 3D Shrimp Bottom Trawl Survey (1973-2001)

Results of all shrimp trawl surveys conducted by DFO were examined for presence of bocaccio (Fig. 4). Many of these surveys now use fish excluders; however, they did reveal the presence of bocaccio in semi-enclosed waters. 

We also show bocaccio biomass estimates generated from the annual shrimp trawl survey on the west coast of Vancouver Island, when fish excluders were not employed (Fig. 18) (Boutillier et al. 1998 and Appendix 3). The B.C. area covered in this survey is roughly similar to that of the U.S. survey.


Figure 18: Bocaccio Biomass Estimates from Shrimp Trawl Surveys of the West Coast of Vancouver Island


Figure 18: Bocaccio biomass estimates from shrimp trawl surveys of the west coast ofVancouver Island. The shaded region on the inset chart indicates the area that was surveyed.

The shaded region on the inset chart indicates the area that was surveyed.

While regression analysis of the time series indicates a negative slope, the slope is not significantly different from 0 (p>0.05). The catchability of bocaccio in this survey must be very low, owing to the low towing speed, thus, as with the U.S. survey, the estimates are imprecise and should only be viewed as relative. The variation precludes inferring a decline, although it does not refute the decline indicated by the U.S. survey for the same general area.

 

Areas 5A and 5B Pacific Ocean Perch Bottom Trawl Survey (1966-1995)

Pacific ocean perch (Sebastes alutus) surveys have been conducted intermittently since 1966 (Yamanaka et al. 1996) in Queen Charlotte Sound (Table 5, Fig. 15b, d). Bocaccio have been a minor component of the catch. The depth-stratified catch rate estimates are imprecise and low but do not show any trend, unlike the west coast of Vancouver Island (Table 5).

Table 5: Bocaccio CPUE (log kg/h): Stratified Mean and Coefficient of Variation (%) from Surveys Conducted in Queen Charlotte Sound and Hecate Strait
Survey AreaYearNumber of SetsMean CPUE (kg/h)CV (%)
Queen Charlotte Sound19661316.528.6
19673326.315.0
19693220.521.6
19713923.724.3
19733328.536.7
197636235.577.4
1977497.530.0
19842734.659.5
19843716.731.0
19955421.661.5
19955719.636.8
Hecate Strait1984829.165.2
1987875.850.7
1989900.455.9
1991970.468.9
1993951.153.2
19951020.639.8
1998860.450.5
20001060.764.5


Hecate Strait Bottom Trawl Assemblage Survey (Areas 5C and 5D)

The assemblage surveys are part of a long-term ecosystem study of Hecate Strait (Workman et al. 1997). The focus of the survey has been to classify species assemblages by depth. Bocaccio are a very minor component of the total catch (Table 5, Fig. 15c,d). The first surveys in the late 1980s encountered more bocaccio, but there has been no apparent trend since 1989.

Other Groundfish Related Surveys in B.C. Waters

Excluded from this document are summaries from acoustic and midwater trawl surveys directed at inshore and offshore hake (Merluccius productus) populations. Incidence of bocaccio was limited to only a few fish per survey (M. W. Saunders, pers. comm. Appendix 1). Similarly, extensive midwater and surface trawling has been conducted in the Strait of Georgia as part of an ecosystem study, but they reported no bocaccio in their catches (R. J. Beamish, pers. comm. Appendix 1).

We also did not include results from the International Pacific Halibut Commission (IPHC) standardized hook-and-line survey (Kronlund 2001). Until the early 1990s, rockfishes were most often lumped into a general rockfish category. Attempts to identify rockfish have gradually increased since the early 1990s to the point where most are recorded to species. However, the total catch of fish identified as bocaccio from 1993 to 1999 was 23 animals. This survey could be useful for indicating the presence/absence of bocaccio in the survey area; however, the trawl fishery already confirms this.

Recreational Creel Survey

Fisheries and Oceans, Canada conducts creel surveys of the recreational angling fishery in the Strait of Georgia and elsewhere (K. Hein, pers. comm. Appendix 1). The focus of the monitoring has been chinook (Oncorhynchus tshawytscha) and coho salmon (O. kisutch). There are no records of bocaccio in the database but we suspect that bocaccio are still not explicitly enumerated apart from “other” rockfish.

Abundance in Adjacent Waters

The World Wildlife Fund (1999) lists bocaccio as an endangered species and the International Union for the Conservation of Nature (IUCN) has categorized bocaccio in the Pacific eastern central and Pacific northeast as “critically endangered” (Hilton-Taylor 2000). These conclusions are based on declining trends in the U.S. triennial bottom trawl survey, the commercial trawl CPUE, the recreational CPUE index and a juvenile abundance index (MacCall et al. 1999). Authors of the stock assessment (MacCall et al. 1999) which summarizes this material commented that there has been consistent recruitment failure in California from 1990-1998. An update on the status of this species in U.S. waters will be presented in 2002 (A. MacCall, pers. comm. Appendix 1).

Bocaccio, once reported to have been common in Puget Sound (immediately contiguous to the Strait of Georgia), are now thought to be very rare (Love et al. 2002) and were included in an Endangered Species Act (ESA) petition list for consideration by NMFS.