Rougheye rockfish species types I and II COSEWIC assessment and status report: chapter 9

• Previous page
• Table of contents
• Next page

Population Sizes and Trends

• Fishery History
• Abundance
• Fluctuations and Trends
• Changes in Age Composition and Mortality
• Rescue Effect

Fishery History

The frequent occurrence of rougheye rockfishes in the Canadian bottom trawl fishery (Figure 5) indicates a widely distributed species pair (and possibly relatively large biomass) compared to other rockfish. Table 2 gives the annual catch and quota history of this species pair along the BC coast. In 1989, DFO separated rougheye rockfish, Sebastes aleutianus (Jordan and Evermann, 1898) out from a managed aggregate (Pacific ocean perch, yellowmouth rockfish, rougheye rockfish) and set specific quotas for the Western Queen Charlotte Islands (PMFC 5E) where the species pair is primarily targetted. Quota levels reached a peak in 1996, the year that 100% observer coverage was implemented for the trawl fleet. In 1997, an individual-vessel quota (IVQ) system was started, and since 1998 the quotas have remained constant (coastwide = 950 t).

Abundance

No estimates of absolute abundance exist. From 1971 to 2005, the combined trawl and longline fleets removed 25 590 t of biomass of the rougheye rockfish species pair from BC coastal waters. This is equivalent to 16.2 million fish, assuming a conversion rate of 1.585 kg/fish (Haigh et al. 2005).

Fluctuations and Trends

Research surveys cover large areas of the BC continental shelf and slope, but do not currently provide comprehensive spatial sampling (Figure 13). Commercial fisheries catch-per-unit-effort (CPUE) data are also available to support status assessment.

Hecate Strait Assemblage Survey

The Hecate Strait (HS) bottom trawl survey, initiated in 1984, was designed to collect data on species interactions (Fargo and Tyler 1991), and continued on a roughly biennial basis until 2003. Prior to 2003, the survey focused on determining abundance indices for flatfish with the consequence that the habitats surveyed do not appear to be appropriate habitat for rougheye rockfishes; 3 out of 1048 trawl collections provided rockfish specimens. An extended survey area, ranging from Dixon Entrance in the north to Queen Charlotte Sound in the south, initiated in 2005 should improve the utility of the HS survey for many species. However, the depth ranges of the revised survey do not yield large areas of depth preferred by rougheye rockfishes (Greg Workman¹, pers. comm.), and consequently, this survey will not yield useful results for this species.

^AJan-Mar for Trawl;  ^BJan 97 – Mar 98 for HL; ^CApr-Dec (incomplete catch records)
agg = aggregate trawl quota for Pacific ocean perch, yellowmouth rockfish, and rougheye rockfishes
^5EPMFC area 5E only; ^RSRougheye rockfishes and shortraker rockfish combined

^* Catches are rounded to the nearest tonne; entries marked ‘---’ indicate no recorded catch or quota. Trawl data from 1971 to 1995 are stored in the GFCatch database; data from 1996 on reside in PacHarvTrawl. Hook and line data from the Zn, Schedule II, and halibut fisheries reside in PacHarvHL. From 1971 to 1996, fishing year = calendar year; thereafter, fishing year spans Apr to Mar. Source: Haigh et al. (2005). Total allowable catches (TACs) can be found in various fisheries management plans at: http://www-ops2.pac.dfo-mpo.gc.ca/xnet/content/MPLANS/MPlans.htm?lang=en

Figure 13: Overview of Surveys Used as Abundance Indices for Rougheye Rockfishes

Depth strata (specific to surveys) are represented by various shades. Source: Haigh et al. (2005).

Queen Charlotte Sound Synoptic Survey

The Queen Charlotte Sound (QCS) bottom trawl survey was initiated in 2003 to meet multispecies survey requirements for ecosystem-based management (Stanley et al. 2004). The survey area covers the region north of Vancouver Island to southern Hecate Strait and depths of 50 to 500 m. The survey is designed to capture all groundfish species using a tow allocation budget that minimizes the CPUE-estimate CVs for stocks representing a variety of concerns and interests. Rougheye rockfishes were included, but as a “species” whose preferred habitat is in depths too great to be well sampled by the survey. At present, the time series spans only three years (Figure 14), a period too short to detect abundance changes for any species.

Figure 14: Abundance Index from Trawl Survey for Rougheye Rockfish Species Pair in Queen Charlotte Sound

Index is standardized to 2003. Vertical bars indicate 90% confidence intervals from 1000 simulated index estimates. Source: Haigh et al. (2005).

Shrimp Trawl Surveys

Sinclair et al (2001) describe, in detail, the shrimp trawl survey off the west coast of Vancouver Island (WCVI). For some species of rockfish, relative abundance indices from the shrimp trawl survey show coherence with those from surveys targetting groundfish. In the case of the rougheye rockfish species pair few tows caught this species, which renders this index series useless (Figure 15, top panel). As seen above for the HS assemblage survey, the WCVI shrimp survey consistently covers depths too shallow (80 m to 175 m) to provide a reliable index for this species.

Boutillier and Olsen (2000) describe the Queen Charlotte Sound (QCS) shrimp trawl survey. Unlike the WCVI shrimp survey, 50-60% of the tows capture rougheye rockfishes. The trend analysis excludes the years 1998, 1999, and 2005 as the tows cover the survey region inconsistently. Overall, the index shows no trend from 2000 to 2004 (Figure 15, bottom panel). However, the 2004 index confidence interval does not overlap that for 2000. This survey, if continued, might prove useful for tracking populations of rougheye rockfishes in the region of QCS that it covers.

Figure 15: Relative Abundance Index for Rougheye Rockfish Species Pair from Shrimp Trawl Surveys in Waters off the WCVI (top) and Queen Charlotte Sound (bottom)

Vertical bars indicate 90% confidence intervals from 1000 simulated estimates. Source: Haigh et al. (2005).

NMFS Triennial Trawl Survey

The US National Marine Fisheries Service (NMFS) conducted triennial groundfish trawl surveys at depths 55-500 m along the US Pacific coast and the WCVI from 1977-2001 (Weinberg et al. 2002), using statistical areas set up by the International North Pacific Fisheries Commission (INPFC). Mark Wilkins² provided tow data from a trans-boundary INPFC region named “Vancouver” for the seven years that surveyed Canadian waters. NMFS assigned these tows to depth strata, but the size and definition of these strata changed over the life of the survey. In particular, NMFS added deep strata (367-500 m) in the final years.

Haigh et al. (2005) assessed and analyzed these data (excluding deep strata) for rougheye rockfishes for the entire Vancouver region and for the Canadian and US sub-regions (Figure 16). The relative biomass estimates show no significant trend for the Vancouver region, nor for the two sub-regions. The trend for the total Vancouver region reflects that in the Canadian sub-region. The relative biomass estimates are more precise in the US sub-region than in the Canadian sub-region due to the reasonably consistent catches of rougheye rockfishes in US waters. In Canadian waters, the surveys captured almost no rougheye rockfishes in 1980, 1983 and 1989; whereas the 1995 survey had one tow with a large catch of this species pair.

Observed Commercial Trawl CPUE

Haigh et al. (2005) analyze commercial catch data for rougheye rockfishes from the DFO PacHarvTrawl database using two general linear regression models (GLM): one assuming a log-normal distribution based on non-zero catches of rougheye rockfishes, and the other assuming a binomial distribution based on the presence/absence of this species pair in the catch. The CPUE analysis uses only data starting April 1, 1996 when 100% observer coverage on the trawl fleet began. The analysis also restricts tows to depths where rougheye rockfishes were captured and to vessels that participated in the fishery for at least three years with at least five trips per year. The analysis considers three fisheries for rougheye rockfishes: the west coast of Vancouver Island (WCVI: PMFC areas 3C and 3D), the combined Queen Charlotte Sound and Hecate Strait (QCS: PMFC areas 5A, 5B, 5C and 5D), and the west coast of the Queen Charlotte Islands (WQCI: PMFC area 5E).

Figure 16: Three Relative Biomass Indices for Rougheye Rockfish Species Pair from US National Marine Fisheries Service (NMFS) Triennial Groundfish Survey off the West Coast of Vancouver Island (total region, Canadian waters only and US waters only) with 95% Bias Corrected Error Bars Estimated from 5000 Bootstraps

Source: Haigh et al. (2005).

A comparison of the three areas for each type of GLM analysis shows similarities between series across areas (Figure 17). Although regression analysis of the six series reveals increasing trends ranging from 2% to 11% per year (depending on the area and regression model used), the three binomial series exhibit little long-term trend although the most recent one or two fishing years show higher values than earlier. The three lognormal series overlay each other with no strong trend, even though maxima and minima are not synchronized. The WQCI lognormal series peaks in both 1998/99 and 2003/04, inconsistent with the other two areas. This could be due to a strong directed fishery for rougheye rockfishes in WQCI while the other two areas only catch this species incidentally. These relative indices should be interpreted with caution, however, as they come from fishery-dependent data that reflect between-year effects originating from sources other than fish abundance (primarily fishermen behaviour – compliance with quotas and regulations, bycatch avoidance, minimization of fuel costs, etc.).

Figure 17: Comparison of Two Sets of CPUE Indices Each Based on Different Regression Model Assumptions for Each of Three Areas

Each series has been standardized relative to the geometric mean of the period 1996/97 to 2004/05. The error bars show ± 95% confidence bounds. Source: Haigh et al. (2005).

Trend Summary

The available survey index trends are generally flat to increasing, but available surveys either do not cover the distribution of rougheye rockfishes well or are of short duration. The two shrimp trawl surveys – one off WCVI and the other in QCS (Figure15) – suffer from restricted spatial (depth and area) ranges. The US NMFS triennial survey (Figure 16) shows no significant trend, although there is some tendency for increase in the latter part of the time series. However, this survey covers depth ranges too shallow for rougheye rockfishes. While surveys usually give the most reliable index for monitoring demersal marine species, the large error bars from these surveys mean that any estimated trends would be highly uncertain. Indices from the more appropriate groundfish synoptic surveys initiated in 2003 should describe abundance trends more reliably once the survey series cover longer time periods.

All six sets of CPUE abundance series (two models: lognormal and binomial for each of three areas outlined above) show little overall trend, although higher values in the last 1-2 years are seen in some (Figure 17). These indices may be influenced by factors other than abundance of rougheye rockfishes.

Changes in Age Composition and Mortality

Age-proportion data from 1996 to 2004 off the west coast of the Queen Charlotte Islands (WQCI, PMFC³ area 5E; Figure 10) suggest a decline in older ages (plus class), and a shift to younger fish (Figure 11). Catch-curve analysis using the method of Schnute and Haigh (2006) suggest that total mortality (natural + fishing) has doubled from 1996 to 2003 (Figure11). The model comprises three components – survival, selectivity, recruitment anomalies – and assumes full selectivity by age 40. Posterior model estimates of total mortality rate Z for the survey year 1997 have mean 0.048 with 95% limits of (0.039, 0.058). Commercial age proportions in 1996 yield essentially the same estimate of Z with mean 0.045 and 95% limits (0.038, 0.054). In 2003, the shift in age classes yields a posterior distribution of Z with mean 0.091 and 95% limits (0.072, 0.107). While the increase suggests that fishing mortality has escalated, non-representative catch sampling may account for some of the perceived difference in proportion-at-age data. A survey in PMFC 5E would be advisable to clarify the situation.

Rescue Effect

Bordering populations in Washington and Alaska could act as population sources though there are no data suitable to test this hypothesis. Population trends in the Gulf of Alaska, from both bottom trawl indices and sablefish longline survey indices, are stable (Shotwell et al. 2005). In Washington, the INPFC does not assess rougheye rockfishes, but most Sebastes stocks have been in decline. The US Vancouver region NMFS survey index shows no trend (Fig 16).

¹ Marine Ecosystems and Aquaculture Division, Fisheries and Oceans Canada: Pacific Biological Station, 3190 Hammond Bay Road, Nanaimo, BC, V9T 6N7
² Alaska Fisheries Science Center, Department of Commerce, National Oceanic and Atmospheric Administration: National Marine Fisheries Service, 7600 Sand Point Way N.E., Bin C15700, Building 4, Seattle, Washington 98115-0070
³ Pacific Marine Fisheries Commission, Portland OR.