Review of Northeast Fishery Stock Assessments (1998)

• Improve biological sampling of U.S. commercial landings and discards.

• Examine the effects of large tows on overall and age-specific abundance indices, specifically with reference to closed areas.

• Examine effects of abrupt changes in mean weights during the 1990s, specifically with respect to the 1989-1991 year classes in the eastern part of Georges Bank.

• Investigate factors associated with apparent recent improvements in survival ratios (R/SSB).

• Data Used:

  • ANALYTICAL, Age structured assessment

  • ADAPT VPA Calibration of age 1-10+catch at age, 1978-1996

  • NEFSC spring and autumn RV indices (1978-1996)

  • Canadian DFO spring RV indices (1986-1997)

  • CAA estimated for recreational landings 1981-1996

  • Discard ratios estimated for 1989-1996

• CAA:

  • Commercial: sufficient age/length data not available prior to 1978

  • 1994-1996 sampling poorest since 1982, notably in ''large" market category.

  • 1996 "scrod" and "market" samples greater than 1995 samples

  • Recreational: derived from very few length samples, 1981-1996

• Sensitivity Runs:

  • Recreational (1981-1996): slightly higher stock size, with similar [Text incomplete in source] and SSB as base run

  • LPUE (1978-1996): lower stock size and higher F than final run

• Other issues:

  • Transboundary stock: use Canadian CAA and RV survey  

  • Canadian RV survey indices for 1993 and 1994 not used (lack of coverage)

  • LPUE not used as in previous assessments (1994-1996 data unreliable)

  • Discards not incorporated: lack of samples, 1978-1988

  • Recreational not used: poor sampling, lack of samples, 1978-198[Text incomplete in source]

• Current landings (1996) = 8,900; up from 7,900 mt in 1995

• Survey indices fluctuate near record-low values

• LPUE declined by 1993 to the lowest LPUE in time series

• VPA results  

• Current F(1996) = 0.17

• Current SSB(1997) = 46,400 t

• Shift to older ages in PR

• Mean F (4-8, u) declined from record-high in 1994 to record-low in 1996

• Record-low recruitment of the 1994-1996 year classes

• Yield and SSB per Recruit Model  

• F_0.1= 0.17  

• F_max= 0.32  

• F_20%= 0.43

• Projections  

• Short term: At status quo F, landings in 1998 (8,400 t) remain stable, SSB increases 35% during 1996 (41,200 t) to 1999 (55,000 t) Higher F scenarios initially increase landings and SSB, with subsequent declines  

• Medium term: 1998 to 2006 Landings increase from 8,200 mt to 29,400 mt SSB improves from 53,700 mt to 199,900 mt Median recruitment increases from 14 million to 34.4 million

• CONCLUSIONS  

  • Stock remains in overexploited state with low biomass  

  • Recruitment of three most recent year classes at record low  

  • SSB remains near record low value  

  • Current F(96) nearly equal to F0.1

•  Further investigate the effect of closed areas on the use of LPUE as an index of abundance, specifically examine the impact of change in fleet distribution as a result of progressive exclusion from the Canadian zone and then from Closed Area II.

• Further investigate the basis for deriving the recreational component of the cod catch. Specifically, the effect of sampling levels in the party and charter categories at age can be used to augment the commercial landings at age in the VPA.

• Further examine discard rates in years prior to 1989 before discard data can be incorporated into the catch at age.

• CURRENT ASSESSMENT

• ANALYTICAL (Age Structured)

• ADAPT VPA Calibration of Age 2-7+ CAA, 1982-1996  

• ANALYTICAL (Age Structured)  

• ADAPT VPA Calibration of Age 2-7+ CAA, 1982-1996

• 1996 landings were 7,200 t; up from 6,800 t in 1995

• Sample summary  

• 1994-1995 sampling improved over 1993, but low re: landings  

• 1996 much improved

• Catch at Age  

• Reduction in proportion of older 9>7 yr) ages and few fish >9 yr after 1990  

• 1992 year class 62% by number, 57% by weight in 1996

• Commercial LPUE  

• LPUE declined in 1993 to pre-1989 level; not used in 1994-1996

• Survey Indices  

• Low number and weight per tow persist through autumn 1996  

• Low recruitment indices in NEFSC and Massachusetts DMF surveys in 1995 and 1996  

• Survey Zx 1.1 in 1991-1993 and 1994-1996

• VPA Calibration  

• Unusual F pattern in 1994 for fully recruited ages 4 and 5  

• Record low recruitment of 1994 and 1995 year classes <1 million fish

Commercial CAA

• Catch at Age

  • Majority of catch at ages 3 and 4 (by number) and ages 4 and 5 (by weight)

  • Reduction in proportion of older (>7 yr) ages and few fish > [Text incomplete in source] yr after 1990

  • 1992 year class 62% by number, 57% by weight in 1996

  • High mean wts. at age in 1995 at ages 5 and 6 ==> Low Numbers may have influenced high F at age 4 and 5 in 1994

• Sampling Intensity

• Poor sampling in 1993

• 1994-1995 sampling improved, but low for older ages (Large market category)

• Overall improvement in 1996

Recreational CAA

• Catch at Age

• Length composition indicates higher proportion of smaller fish compared to commercial catch

• Majority of catch at ages 2 and 3 (by number) and ages 3 and 4 (by weight)

• Note: Commercial age/length key used to distribute number at length to numbers at age

• Sampling Intensity

• Generally fewer than 1,000 (often <500) fish measured per year

• Length data (and age) samples pooled on annual basis

• Six alternative assessment models using Stock Synthesis Approach (Fournier and Archibald, 1982)

• Employed same survey indices as ADAPT  

• Included commercial CAA alone and commercial and recreational CAA  

• Lognormal, multinomial, and robust error structure assumptions

• Lognormal model results

• Log-normal error structure most closely approximated ADAPT/VPA

• 1996 and 1997 Fs for fully recruited ages were estimated to be approximately 1.0

• Recent recruitment (1994 and 1995 year classes) poorest ever in both analyses

• SSB declining sharply since 1989, remaining lowest ever through 1997

• Multinomial model results

• Better fit to older ages and slightly lower Fs

• Lower Fs at older ages in earliest years

• Slightly lower Fs in recent years at all ages

• Slightly better recruitment of 1994 and 1995 year classes

• Marked increase in catchability by the fishery over time

• Robust likelihood results similar to multinomial, but with very high Fs at older ages since 1988

• Conclusions

• All model formulations produce results similar to ADAPT/VPA:

  1. Sharp decrease in SSB since 1989

  2. Poorest recruitment on record for 1994 and 1995 year classes

  3. F on fully recruited ages has remained at or above 1.0 since early 1980s

• Four Sensitivity Runs (including final)

• Two with commercial CAA alone

• Two including recreational CAA

• VPA results

• 1996 terminal F for fully recruited ages ranged from 0.96 to 1.04 (essentially 1.0)

• Recruitment estimates slightly higher when recreational CAA included

• Recent recruitment (1994 and 1995 year classes) poorest ever in all runs

• Retrospective patterns in recruitment not persistent

• SSB remains low in all runs

• Final VPA calibration used commercial CAA alone due to

• poor sampling of recreational CAA

• uncertain allocation of recreational cod catch between Gulf of Maine and Georges Bank

•  Current landings (1996) = 7,200 t

• LPUE declined sharply in 1992, remained low in 1993-1996

• Survey Indices remain close to record-low levels in 1995-1996

• VPA results

• current F(1996) = 1.04

• current SSB(1996) = 9,200 t

• Yield and SSB per Recruit Model

• F0.1 = 0.16

• Fmax = 0.29

• F20% = 0.37

• Projections

• At status quo F, SSB declines to unprecedented record low

• Landings decline in 1998 under all F scenarios

• SSB increases over the medium term at Fmax (0.29) to 13,000 t

• CONCLUSIONS

• The stock is overexploited, current F(1996) is well above Fmax and about 3X F20%

• Spawning stock biomass is at a record low level and is projected to decline further in 1998 and 1999

• Further investigation of the changes in effort and LPUE in the VTR data set is required before LPUE can be used to calibrate VPA.

• Further investigation of the basis for deriving the recreational component of the cod catch, specifically the effect of sampling levels in the party and charter categories, is required before the recreational landings at age can be used to augment the commercial landings at age in the VPA.

• Further examination of discard rates in years prior to 1989 is required before discard data can be incorporated into the catch at age.

• Commercial Landings 1973-1996

• Collect data for quarterly estimation

• Per capita sampling adequate 1973-1996

• Commercial Discards 1973-1996

• Need adequate sea sampling

• VTR discard ratios

• Survey Indices 1973-1996: Spring, Autumn, Scallop

• Fishing Mortality: F0.1=0.27, F96=0.12, below reference point in 1996

• Spawning Stock Biomass: SSB-threshold=10,000 mt, SSB97=5,000 mt, below threshold, but increasing

• Short-term: stock improving, could reach 8,000 mt SSB by 1999 at F=0.12

• Medium-term: stock should rebuild if fishing rate remains low

• Improve sea sampling coverage on trawl and scallop vessels.

• Increase sampling frequency on research surveys.

• Improve commercial length and age samples.

• Examine VTR data for mesh-specific discard ratios.

• Evaluate changes in maturity at age in recent years.

• Evaluate performance of the scallop survey as a tuning index.

• Evaluate winter survey as a tuning index.

• Data in the assessment: 1973-1996 U.S. and Canadian commercial landings and discards (recreational catch is negligible), U.S. spring and fall trawl survey indices, U.S. scallop survey indices, and Canadian trawl survey indices.

• Quality of Catch at Age:

• Age composition by quarter and statistical area not available 1994-1996. Catches characterized semiannually for the entire stock area.

• Sexually dimorphic growth not considered in estimates of catch at age.

• Sensitivity Runs: Log-transformed survey indices and removal of uncertain age-1 discard estimates. Results and diagnostics similar among all sensitivity runs.

• Alternative Methods: Biomass dynamics model (ASPIC) using total catch (1963-1996), U.S. spring and fall survey indices, and the Canadian survey index of biomass. Magnitude and temporal pattern of mean biomass and F similar to those from VPA.

• F and SSB Trends:

• F averaged 1.2 during 1973-1994 and decreased to 0.1 in 1996

• SSB was 21,000 mt in 1973, declined to less than 4,000 mt during 1984-1988, fluctuated below 6,000 mt from 1989 to 1994, and increased to 11,700 mt in 1996

• Probability Distributions for F and SSB in 1996:

• 80% chance that F was between 0.08 and 0.14, and nearly 0% probability that F exceeded F0.1 (0.25)

• 80% chance that SSB was between 9,800 and 14,600 mt, and 12% probability that SSB was below the rebuilding threshold of 10,000 mt

• Projections:

• At F0.1, landings and SSB will continue to increase in the next three years. At F96, landings will decrease slightly in 1997, then increase in 1998 and 1999, and SSB will continue to increase in the next three years.

• At F0.1, landings will increase to 8,400 mt in 2006, and SSB will increase to 46,200 mt in 2006. At F96, landings will increase to 5,500 mt in 2006, and SSB will increase to 71,600 mt in 2006.

• Extending the VPA back to the 1960s should be explored.

• Changes in maturity should be closely monitored.

• The NEFSC winter survey should be modified to ensure coverage of Georges Bank.

• Evaluate the feasibility of sex identification in all field sampling to estimate catch at age and survey indices by sex.

• The number of ages in the VPA should be expanded.

•   The Situation for Georges Bay Cod, Haddock & Yellowtail, and SNE Yellowtail Has Improved; Gulf of Maine Cod is Unchanged

•   Except for GM Cod, F Has Been Reduced BELOW Overfishing Reference Points & is NEAR Rebuilding Fs

•   Except for GM Cod, Some Rebuilding of SSB has Occurred. GB Yellowtail is near SSB Threshold, but SSB for all Stocks is LOW relative to Historical Levels & BMSY

•   Recruitment in Recent Years Remains LOW Relative to Historic Levels

•   Short-Term Projections Indicate Maintenance or Modest Increases in SSB at Current Fs, Except for GM Cod

•   Strong Management Measures Are Warranted for GM Cod to Reduce the Risk of Stock Collapse

•   Continued Rebuilding of Four Stocks Will Be Jeopardized if Fishing Mortality Rates Are Allowed to Increase

•   Efforts to Reduce F for GM Cod Should NOT Come at the Expense of Other Heavily Exploited Stocks in the Gulf of Maine or Elsewhere