Over the last several years IPHC staff have noted and discussed changes taking place in the halibut fishery that would likely lead to changes in interpretations and assumptions associated with the Pacific halibut stock assessment. In particular, we indicated that Pacific halibut have undergone a rapid reduction in individual growth in recent years, with average length at age now being 20-25% lower than what it was 10-15 years ago. We also noted that changes in the fishery, prompted by initiation of individual-quota programs, would likely have an effect as well. Last year we proposed a new approach that accounts for changes in individual size at age and its likely effect on the catchability of halibut. The approach, presented in preliminary form at last year's IPHC Annual Meeting, indicated that both stock biomass and recruitment might be higher than that estimated under previous stock assessment procedures. This year was spent confirming these preliminary results, while continuing to incorporate other important sources of information into the assessment. The new assessment procedure not only takes into account commercial age-composition, catch, and CPUE as it has in the past, it also includes size at age of the commercial catch, and catch, CPUE, age-composition, and size at age of IPHC standardized setline surveys. In addition, it now accounts for the mortality of legal-sized halibut associated with bycatch in non-directed fisheries (Figure 1). These features of the new assessment procedure aid in adjusting for changes in growth, in accounting for changes in the fishery, and in better tracking the influence of bycatch mortality on the stock.

Exploitable biomass estimates have increased under the new stock assessment. The increase in the estimates can be broken down into three major components. (1) Halibut size at age is now better represented in the assessment model. We recognize that halibut size at age has been decreasing in recent years as a result of slower growth. This reduction in size has reduced the catchability of younger age groups by setline gear through fish behavior and thresholds imposed under the legal size limit. The "poor recruitment" of age 8 halibut into the fishery was interpreted as low abundance in earlier assessments rather than as poor catchability due to smaller size. This lower catchability can now be estimated, and the estimated abundance of both younger and older age groups has increased accordingly. (2) Bycatch mortality of legal-sized halibut is now included as removals directly in the assessment along with other removals (commercial and sport catches, wastage, and personal use). The estimated biomass must increase to account for the increased level of removals. The magnitude of the increase depends on the amount of legal-sized bycatch mortality relative to total stock biomass in each area. (3) Information from IPHC setline surveys can now be explicitly incorporated. Survey CPUE trends support trends seen in commercial fishery CPUE, lending greater weight to the belief that abundance has increased since the 1980s, while helping to point out changes that have taken place in halibut catch statistics under the recently implemented individual-quota programs.

The Pacific halibut stock assessment continues to show a slight downward trend in coastwide stock biomass over the last five years (Figure 2). This trend, however, is not as severe as that reported under previous assessments. In contrast, some IPHC regulatory areas show a leveling off (Areas 2A, 2B), or an increase (Areas 3B, 4), after accounting for the effects of slower growth and bycatch mortality (Figures 3-8). IPHC systematic survey catch per unit effort (CPUE), now incorporated in the assessment, can be compared with commercial setline CPUE (in number of halibut per skate) in Areas 2B, 2C, and 3A (Figure 9). Survey CPUE, while generally lower than commercial CPUE in Area 2B, shows a greater relative increase between observations taken in the 1990s and those taken in the 1970s to 1980s. Area 2C surveys also show an increase in contrast to the decline shown by commercial CPUE. Area 3A survey and commercial CPUEs both are quite consistent in indicating an increase since the 1980s, with similarly high levels occurring currently. The sublegal-sized component of the fishery is making up a greater proportion of the survey catch in recent years (Figure 9) again indicating the influence of smaller individual size on observed measures of abundance. The assessment now follows changing trends in growth, and takes account of changes in gear selectivity which are likely to occur simultaneously. In areas where this change in growth is great (e.g. Areas 3A and 3B), the result is generally a greater increase in the estimated level of abundance. Apparent poor recruitment to the fishery by more recent cohorts shown in earlier assessments actually resulted from a reduced vulnerability to the fishery, rather than a diminished abundance.

Commercial CPUE (in pounds per skate) is stable or on the upturn this year, with a coastwide increase of 10% from 283 pounds per skate (lbs/sk) in 1995 to 311 lbs/ks in 1996 (Figure 2). CPUE on an area-by-area basis increased 74% to 155 lbs/sk in 2A and 8% to 221 lbs/sk in 2B, decreased 5% to 221 lbs/sk in 2C, increased 13% to 442 lbs/sk in 3A, decreased 3% to 462 lbs/sk in 3B, and increased 25% lbs/sk in Area 4 (Figures 3-8).

Change continues to be observed in the average weight at age of individual halibut. Figure 10 shows the trend in the weight of age-12 halibut for each regulatory area. Dramatic decreases can be seen in the average weight of fish landed in the central regulatory areas Area 3A and Area 3B. Decreasing, though less dramatic, trends can be seen in Area 2AB and Area 4, while some increase can now be seen in the weight of halibut caught in Area 2C. Halibut younger than age 12 (not shown) have begun to exhibit an upturn in weight for all areas except Area 3A. The implications of these continually changing weights for determining the status and production levels of future stock biomass is complex and will continue to be monitored.

The incorporation of growth into the assessment has had a major effect on our estimates of year-class strength and trends in recruitment. The stock assessment figures show total biomass of 8-year-old halibut labeled as recruitment (Figures 2-8). This statistic represents the relative year-class strength in biomass of potential recruits rather than a reflection of their level of entry into the fishable portion of the stock. Under previous assessment methods the trends in these recruitment estimates were in severe decline. Some decline can still be seen on average coastwide and in most areas. However, the decline is not severe and the strength of more recent cohorts is better represented. The 1987 year class in particular, indicated as being strong in abundance in National Marine Fisheries Service trawl surveys (Clark and Walters, 1995), continues to show its strength as it enters into the fishery. These recruiting halibut (shown as a peak in eight-year-old recruitment biomass in 1995 in Figures 2-8) will be ten years of age during the 1997 season. Of these fish, approximately one third are estimated to be available to the fishery. The presence of this year-class appears to be greatest in Area 4; however, great uncertainty is associated with the Area 4 estimate. Recruitment biomass estimates in this and other areas are highly imprecise in the most recent years, when cohorts have been observed only once or twice in the fishery. Furthermore, given the generally smaller size of these fish, the percentage available for harvest is estimated to be very low, which in turn implies that the estimates themselves may be quite unreliable as only a very small fraction is observed in the catch. An additional consequence of the reduced size-at-age is that the overall contribution to exploitable biomass of these year classes is likely to be smaller in the long term than the strong year classes of larger individuals observed in the mid-1980s. The strength of the 1987 year class, never-the-less, is a positive sign for the fishery.

As can be noted in the accompanying figures, each area's assessment demonstrates its own unique representation of stock trends and recruitment levels. The total quantity of information available for each area's assessment is not the same however. Areas 2A-2B, 2C, and 3A, for example, all have long term IPHC setline survey data that provide information on trends in total abundance and year-class strength. Area 3B and Area 4 lack such systematic and longer term survey information. The resulting estimates are considerably less precise with one half to one third the level of confidence of the estimates given in the other areas.

In Area 3B, inconsistencies can be noted in relative abundance as estimated in independent assessments conducted on Areas 3A and 3B. The independent estimates, shown in this document, indicate that Area 3B exploitable biomass is roughly 30% of that estimated for Area 3A. The 1996 IPHC setline survey and NMFS trawl survey averages conducted over the two areas, on the other hand, indicate that Area 3B exploitable biomass should be roughly 60% of that shown for Area 3A (Clark 1996). No merging of these data has yet brought about an estimate that is consistent with all available information. Unfortunately, long-term setline survey information is lacking in Area 3B. Such information would be invaluable in addressing observed differences in estimates of relative abundance. Commission staff will continue to follow closely trends and statistics collected in Area 3B relative to the neighboring Area 3A.

In Area 4, low harvests in the 1970s have reduced the level of information available from the commercial catch for this area. Furthermore, there is sparse commercial coverage of all grounds known to contain halibut in Area 4. The lack of complete data coverage over time and area is a serious concern in the assessment of the Area 4 stock. As noted by the measures of relative uncertainty shown in the stock biomass and recruitment figures, greater risk is associated with managing the stock in these areas under the current management protocols. Commission staff will consider alternative assessment and management strategies for Area 4

Given the changes occurring in the biology of the stock, and the associated change in the assessment, exploitation rates used in calculating the constant exploitation yield (CEY) must be reevaluated. How different exploitation rates perform hinges on the relationship between adult biomass levels and future levels of recruitment, as well as the average reproductive contribution of recruits. In conformance with a change in method of bycatch accounting, the choice of harvest rate now reflects the loss due to pre-recruitment bycatch mortality. The analysis of alternative harvest rates conducted using a definition of exploitable biomass that is consistent with current estimates of selectivity indicates that harvest rates in the range 0.20-0.25 may achieve close to maximum yields under a variety of possible future recruitment scenarios with a high likelihood that the stock is maintained within the range of historically observed levels. These and other issues are discussed more fully in a separate document. Setline CEYs computed using a harvest rate of 0.20 are shown in Table 1.

In computing the setline CEY from the total CEY under a 0.20 harvest rate, a new method of accounting for bycatch has been implemented. In past reports, total bycatch mortality was reported in Table 1 and a pound-for-pound adult reproductive compensation poundage was computed as a reduction in each IPHC area in proportion to the biomass in that area. This year, we instead incorporated legal-sized bycatch mortality into the calculation of stock abundance as removal. This contributed to raising the estimated stock levels. Legal-sized bycatch mortality is now the only component removed in the CEY calculation, with the sublegal-sized bycatch component resulting in a reduction of the recommended range of harvest rates. The legal-sized bycatch mortality reduction represents the current year's losses, and the CEY is reduced in each area by the level of legal-sized bycatch mortality that has taken place in that area.

We believe that these estimates better reflect the stock biomass and harvest levels on which management should be based, especially in areas where the assessment is supported by fishery-independent data. However, we recognize that the assessment method is new and will continue to evolve as we incorporate new data and further evaluate sensitivity to differences in model assumptions. The uncertainty demonstrated for the estimates given in Area 3B and especially Area 4, areas with little or no fishery-independent data, must be considered in setting catch limits for the upcoming season.

Clark, W. G. 1996. Survey information on distribution and trends in abundance. IPHC Report of Assessment and Research Activities.

Clark, W. G., and G. E. Walters. 1995. Results of the 1994 NMFS Bering Sea trawl survey. IPHC Report of Assessment and Research Activities. Pages 271-276.

Figure 2. Coastwide size-age based estimates of exploitable stock biomass, commercial CPUE, and total biomass of eight-year-old halibut as an indicator of recruitment. Vertical lines represent confidence measures (plus or minus two standard deviations) on the biomass estimates.

Figure 3. Area 2A size-age based estimates of exploitable stock biomass, commercial CPUE, and total biomass of eight-year-old halibut as an indicator of recruitment. Vertical lines represent confidence measures (plus or minus two standard deviations) on the biomass estimates. Area 2A biomass estimates represent 7% of the Area 2A-2B combined estimate.

Figure 4. Area 2B size-age based estimates of exploitable stock biomass, commercial CPUE, and total biomass of eight-year-old halibut as an indicator of recruitment. Vertical lies represent confidence measures (plus or minus two standard deviations) on the biomass estimates. Area 2B biomass estimates represent 93% of the Area 2A-2B combined estimate.

Figure 5. Area 2C size-age based estimates of exploitable stock biomass, commercial CPUE, and total biomass of eight-year-old halibut as an indicator of recruitment. Vertical lines represent confidence measures (plus or minus two standard deviations) on the biomass estimates.

Figure 6. Area 3A size-age based estimates of exploitable stock biomass, commercial CPUE, and total biomass of eight-year-old halibut as an indicator of recruitment. Vertical lines represent confidence measures (plus or minus two standard deviations) on the biomass estimates.

Figure 7. Area 3B size-age based estimates of exploitable stock biomass, commercial CPUE, and total biomass of eight-year-old halibut as an indicator of recruitment. Vertical lines represent confidence measures (plus or minus two standard deviations) on the biomass estimates.

Figure 8. Area 4 size-age based estimates of exploitable stock biomass, commercial CPUE, and total biomass of eight-year-old halibut as an indicator of recruitment. Vertical lines represent confidence measures (plus or minus two standard deviations) on the biomass estimates.

Figure 9. Commercial setline catch per unit effort (CPUE) in number of halibut per skate contrasted with IPHC setline survey CPUE in total number of halibut per skate and number of legal-sized halibut per skate.

Figure 10. Trends in average individual weight at age for age 12 halibut in each IPHC regulatory area. Weight given in pounds net weight.