1. ASSESSMENT OF BIGEYE TUNA (THUNNUS OBESUS) IN THE EASTERN PACIFIC OCEAN January 1975 – December 2006

2. Outline Stock assessment –Overview of assessment model –Fishery data –Assumptions –Results of base case model –Projections Sensitivity analyses Summary and conclusions Discussion

3. Overview of assessment Age-structured, statistical, catch-at-length model (Stock Synthesis II). Same type of model as A-SCALA or MULTIFAN-CL Differences between SS2 and A-SCALA

4. Bigeye fishery definitions Recent FLT (2-5) Discards (10-13) N Longline (8) S Longline (9) Early FLT (1) Early & Recent UNA (6, 7) 1, 6-7 8 9 2, 10 3, 11 5, 13 4, 12 FLT – Floating objects; UNA - Unassociated

5. Data - catch Early FLT N Offshore FLT Equatorial FLTCoastal FLTS Offshore FLT Early UNARecent UNAN LL S LL Disc - S FLT Disc – Eq FLT Disc – coastal FLT Disc -N FLT

6. Data - discards Southern FLT Equatorial FLT Northern FLT Inshore FLT

7. Data - effort Early FLT N Offshore FLT Equatorial FLT Coastal FLT S Offshore FLT Early UNA Recent UNA N LL S LL Disc - S FLT Disc – Eq FLT Disc – coastal FLT Disc -N FLT

8. Data - CPUE Early FLT N Offshore FLT Equatorial FLTCoastal FLTS Offshore FLT Early UNARecent UNAN LL S LL Disc - S FLT Disc – Eq FLT Disc – coastal FLT Disc -N FLT

9. Data - Length frequency data Quarter 2000 1997 2005

10. Assumptions (base case) - movement Tagging records indicate little exchange of bigeye between E and W Pacific Results from conventional and archival tagging indicate regional fidelity for bigeye in EPO Different CPUE trends between EPO and WCP DATA Single stock of bigeye in EPO No net movement of fish between the eastern and western Pacific SA for EPO and Pacific wide are consistent ASSUMPTIONS

11. Assumptions (base case) - growth Von Bertalanffy – fixed parameters

12. Assumptions (base case) – M

13. Assumptions (base case) - cont. Age-specific maturity and fecundity indices No S-R relationship (steepness = 1)

14. Results (base case) Fit to the length frequency Fishing mortality Selectivity Recruitment Biomass

15. Fit to LF data – Pearson residuals

16. Fit to CPUE data – Floating object

17. Fit to CPUE data – Longline

18. Fishing mortality Ages 1-4 Ages 5-8 Ages 13-16 Ages 17-20 Ages 9-12 Ages 21-24 Ages 25-28Ages 29-32 Ages 33-36 Ages 37-40

19. Age-specific fishing mortality 1975-1992 1993-2006

20. Size selectivity Early FLT N Offshore FLT Equatorial FLT Coastal FLT S Offshore FLT Early UNA Recent UNA N LL S LL

21. Recruitment

22. Biomass Biomass of fish 0.75 + years old

23. Spawning biomass Population fecundity

24. No-fishing plot

25. Average weight

26. Retrospective analysis - biomass

27. Retrospective analysis - recruitment

28. Comparisons with A-SCALA

29. Comparisons with A-SCALA assessments

31. Comparisons to reference points Spawning biomass depletion (SBR)

32. Spawning biomass ratio

33. SBR comparison with A-SCALA

34. Time varying indicators

35. AMSY-quantities

36. AMSY-quantities – by fishery

38. Forward simulations Biomass Spawning biomass depletion Surface fishery catch Longline catch

39. Spawning biomass ratio

40. Predicted catches – purse-seine

41. Predicted catches – longline

42. Sensitivity analyses 1.Spawner-recruitment relationship (steepness = 0.75) 2.Growth -Growth estimated -Linf fixed (171.5 and 201.5) 3.Fitting to the initial equilibrium catch 4.Use of iterative reweighting of data 5.Time blocking of selectivity and catchability for southern longline fishery 6.Inclusion of new Japanese longline data

43. Stock-recruitment relationship (h = 0.75)

44. Biomass

45. Recruitment

46. Spawning biomass ratio

47. Spawner-recruitment curve

48. Use of CPUE time series for southern longline fishery only

49. Spawning biomass ratio

50. Model fit to CPUE data

51. Assumed value for the asymptotic length parameter of the VB growth curve

52. Growth curves

53. SBR

54. Fit to initial equilibrium catch

55. Spawning biomass ratio

56. Use of iterative reweighting

57. Iterative reweighting

58. Spawning biomass ratio

59. Fit to CPUE data

60. Residual plot

61. Use two time blocks for selectivity and catchability of the southern longline fishery

62. Fit to LF data – base case

63. Spawning biomass ratio

64. Fit to CPUE data With iterative reweighting Without iterative reweighting

65. Residual plot

66. Inclusion of the new Japanese longline data

67. Biomass

68. Recruitment

69. SBR

70. Comparisons between models

71. Summary: Main results Both total and spawning biomass is estimated to have substantially declined since 2000 Current biomass level is low compared to average unexploited conditions The current effort levels are too high to maintain the population at level that will support AMSY Yileds could be increased if more of the catch was taken in the longline fisheries

72. What is robust Fishing mortality levels are greater than that necessary to achieve the maximum sustainable yield Two exceptions: Lmax fixed and time blocks

73. Plausible Sensitivities and Uncertainties Results are more pessimistic with the inclusion of a stock-recruitment relationship Biomass trends are sensitive to the weighting of different datasets Recent estimates are uncertain and subject to retrospective bias

74. Conclusions Current spawning biomass is unlikely to remain at or above the level required to produce AMSY. In the most recent years the fishing mortality is greater than that required to produce AMSY. Under average recruitment, the stock is predicted to be below the level that would support AMSY unless fishing mortality levels are reduced further than the current restrictions.

75. Comparisons between models