Consultations
Completed Consultations & Final Decisions
Consultations from 2006
Review of Sustainability Measures
Review of Sustainability Measures and Other Management Controls for the 2006-07 (1 October) Fishing Year - Submissions
Orange Roughy (ORH 3B)
Appendix 1 - Techincal Information
Appendix 2 - Seamount Closures and Benthic Protected Area Proposals in ORH 3B

APPENDIX 1 - TECHNICAL INFORMATION

Fishery assessment

  1. Genetically two main stocks are recognised within ORH 3B (Chatham Rise and Puysegur; Smith and Benson 1997) and these are considered to be distinct from stocks in adjacent areas (Cook Canyon and Wairarapa). However, it is likely, because of their geographical separation and discontinuities in the distribution of orange roughy, that recently discovered concentrations of spawning fish on the Arrow Plateau, near the Auckland Islands, and west of the Antipodes Islands form separate stocks.
  2. For this assessment, the Chatham Rise has been divided into three areas: the Northwest, the East and the South. There is some evidence of stock separation between the Northwest and the East. Spawning occurs simultaneously in the two areas and the post-spawning migration out of the Spawning Box (in the East) is away from the boundary between the two areas. However, there may be more than one spawning stock in the East and there is no clear boundary between this area and the South Chatham Rise, which may or may not contain a separate stock.
  3. New stock assessments conducted in 2006 are available for two sub-areas within ORH 3B: the Northwest Rise and the East Rise. The stock assessment for the East Rise was further divided into three sub-areas to incorporate CPUE indices from the 25-year chronological development of fishing within the East Rise although there is no direct evidence that these three sub-areas are independent. The sub-area within the East Rise with the largest cumulative catch encompasses the Spawning Box and the flat area to the east known as the Northeast Flats. The Northeast Flats fishery concentrates on the post spawning migration of fish from the Spawning Box. The second sub area is the Northeast Hills. This is presumed to target a population separate from that in the Spawning Box because spawning occurs simultaneously in the two sub-areas and historical trends in CPUE differ. The Northeast Hills sub-area is composed of eight hill regions, each with a three nautical mile radius centred on the top of known undersea features. The third sub-area is about the Andes complex of hills where no major spawning aggregations have been found. It is treated as a separate sub area in the stock assessment because the biomass trend in this sub-area is different from that in the other two sub-areas.

Northwest Chatham Rise

  1. For the Alldata run, B0 was estimated to be 55000 t (95% confidence interval 51 400-59 500 t; Table A1.1), the current biomass was 6000 t (4 200-9 300 t), or 11% (8 -16%) B0;. The Nobiomass run produced slightly lower estimates of all biomass metrics. The NoCPUE run produced higher estimates of B0 (79 800 t; 59 600-128 600 t) and Bcurrent (30 900; 12 400-77 500 t) with the median estimate for the ratio of the two being 39% (21-61%) B0.
  2. Neither of the runs that included the survey estimates fit all four biomass indices well (Figure A1.1). For the Alldata run, the estimated biomass trajectories provided a reasonable fit to the acoustic biomass indices, but not the egg survey. The NoCPUE run provided a reasonable fit to the egg survey and the first two acoustic biomass indices, but was above the upper confidence interval of the most recent (2005) biomass index.

Table A1.1: Biomass estimates (medians, with 95% confidence intervals in parentheses) for three runs. Bcurrent is the mid-year biomass in 2006.

RunB0(t)Bcurrent (t)Bcurrent (% B0)
Alldata 55000 (51 400-59 500) 6000 (4 2009 300) 11 (8-16)
Nobiomass 52500 (48 300-56 400) 4 400(3 200-6 900) 9 (6-13)
NoCPUE 79800 (59 600-128 600) 30 900(12 400 77,500) 39 (21-61)
  1. The large discrepancy between the NoCPUE run and the other two runs reflects the relative influence of biomass vs. CPUE indices. When CPUE data are included, they dominate the result (as in the Alldata and Nobiomass runs) because there are a large number of CPUE observations and they cover a period in the fishery when the biomass changed a lot (Figure A1.1).In contrast, there are only four fishery-independent indices of biomass and they occur in recent years when the biomass is not likely to have changed much. In addition, two of these indices have extremely high CVs.
  2. The Plenary noted that the three runs presented should not be given equal weight. The NoCPUE run was not considered to give a reliable assessment of stock status because it relies on survey estimates that are few in number, have high c.v.s, and are restricted to the end of the time series when there is relatively little contrast in stock size. However, it should also be noted that there is uncertainty in the other two runs that include CPUE because the extent to which the CPUE (which is based only on flat tows) indexes the entire stock is unknown.

Figure A1.1.
Figure A1.1: Estimated biomass trajectories (lines) and fitted data (points) from all model runs. The data are identified by the plotting symbol (‘c’ = CPUE, ‘6’ = 1996, ‘9’ = 1999, ‘2’ = 2002, ‘5’ = 2005). CPUE data are scaled up to the biomass. Vertical bars (for biomass indices only) show 95% confidence intervals. Plots are from the medians of the posterior distribution.

  1. For the Alldata and Nobiomass runs, exploitation rates appear to have been higher than the exploitation rate associated with a CAY strategy, E CAY (0.064) for most of the history of the fishery (Figure A1.2). This is to be expected since the fishery was purposely managed to have a fishing down phase. Estimated exploitation rates for2004-05 were 0.26 and 0.34 for the Alldata and Nobiomass runs respectively, both of which were considerably higher than the estimate for the NoCPUE run (0.053).

Figure A1.2: Estimated exploitation rates (solid line) with 95% CI (dashed line) for all model runs. The

horizontal dotted line shows the exploitation rate under a CAY policy, E CAY (0.064).

  1. Yield estimates – Northwest Chatham Rise. The results for the Alldata and No biomass runs suggest that the current catch limit of 1500 t is 3.7-4.8 times than the estimated CAY, and 1.0-1.5 times the associated long-term average yield (MAY) (Table A1.2). In contrast, for the NoCPUE run the current catch limit is similar to the long-term average yield and about three-quarters of the estimated CAY.

Table A1.2.
Table A1.2: Estimated yields: CAY for 2007 and long-term yield under a CAY policy (MAY). The median is shown with the 95% confidence interval in parentheses. All yields were adjusted to allow for an assumed over-run of 5% in future catches.

RunCAY2007 (t)MAY (t)
Alldata 410 (300-610) 1040 (970-1130)
Nobiomass 310 (230-470) 990 (910-1070)
NoCPUE 1950 (810-4790) 1510 (1130-2440)

East Chatham Rise - spawning box and Northeast Flats, Northeast Hills, Andes

  1. The stock assessment results for the East Chatham Rise indicate that there has been some moderate recovery of the aggregation known as the ‘Spawning Box’. The Spawning Box was closed to fishing from 1992 -93 to 1994 -95 in response to declining catch rates. The extent of the rebuild, is however, subject to assumptions regarding recruitment and productivity of the stock. The assessment results for the Spawning Box and Northeast Flats estimated the stock to be 51-61% B 0. Recent exploitation rates and catch are below the rates and long-term yield under a CAY policy. Estimates for the Northeast Hills indicate the sub area has been fished down to a median of 14% B0 (range 7-32% B0). Estimates for the Andes Hills indicate the sub area has been fished down and the biomass median estimate is 29% B0 (range 14-74% B0). Exploitation rates and catch from the Northeast Hills and Andes sub-areas are above the long term CAY policy and five year projections indicate that current catch rates would reduce the sub-areas. Overall catch from the East Chatham Rise has exceeded the catch limits (Table A1.3).

Table A1.3: Catches by subarea in the East Chatham Rise rounded to the nearest 10 t.

Spawning Box and Northeast Flats
Year Northeast Flats Spawning Box Total Northeast Hills Andes
2000-01 240 1 260 1 500 880 1 300
2001-02 200 3 260 3 460 1 040 2 540
2002-03 360 3 360 3 720 870 2 870
2003-04 550 4 510 5 060 640 1 540
2004-05 1 190 4 300 5 490 540 1410

  1. The stock assessment for the East Chatham Rise was divided into three sub-areas in order that all CPUE indices could be used and information is presented separately (Figure A1.3). It was not possible to carry out an overall assessment for the whole East Chatham Rise area due to pronounced differences in the CPUE trends for different sub-areas.

Figure A1.3.
Figure A1.3: The three subareas of the East Chatham Rise for which assessments were done in 2006: Spawning Box and Northeast Flats, Northeast Hills, and Andes.


Spawning Box and Northeast Flats - Estimates of B0 ranged from 292 800 to 366 200 and the current biomass ranged from 33%B0 to 61% B0 (Table A1.4). Awatea runs tended to estimate smaller biomasses and lower Bcurrent (%B0). In all runs, biomass was estimated to have increased since either the 1990-91 or 1991–92 fishing year when it was at a minimum ( Bmin), just prior to the period when the Spawning Box was closed to fishing. Estimates of Bmin ranged from 25-44% B0. Halving natural mortality (HalfM run) resulted in the smallest increase of the biomass in this subarea and the lowest %B0. Recent years have seen a reduction in the biomass increase due to larger catches but the exploitation rate has remained below ECAY since the closure of the Spawning Box in 1991–92 for the runs with natural mortality equal to 0.045.

- Estimates of B0 ranged from 292 800 to 366 200 and the current biomass ranged from 33%B0 to 61% B0 (Table A1.4). Awatea runs tended to estimate smaller bio masses and lower Bcurrent (%B0). In all runs, biomass was estimated to have increased since either the 1990-91 or 1991–92 fishing year when it was at a minimum ( Bmin), just prior to the period when the Spawning Box was closed to fishing. Estimates of Bmin ranged from 25-44% B0. Halving natural mortality (HalfM run) resulted in the smallest increase of the biomass in this subarea and the lowest %B0. Recent years have seen a reduction in the biomass increase due to larger catches but the exploitation rate has remained below ECAY since the closure of the Spawning Box in 1991–92 for the runs with natural mortality equal to 0.045.

Table A1.4: Biomass estimates (medians from the posterior distribution, with 95% CI in parentheses) for the Spawning Box and Northeast Flats assessments. Bmin is the minimum estimated biomass in the time series (1990-91 or 1991-92).

CASAL
Run B0(t) Bcurrent (t) Bcurrent (%B0) Bmin (%B0)
All W 323 800 (268 200–439 700) 181000 (127 300-297 300) 56 (47-68) 37
No.cpue 366 200 (267 900-596000) 223 100 (128 900-449 500) 61 (48-76) 44
Half.M 342 700 (304 300-418 600) 129 200 (93000-203 700) 38 (30-48) 30
Awatea
Run B0(t) Bcurrent (t) Bcurrent (%B0) Bmin (%B0)
All.W 292 800 (250 800–378 800) 148 500 (107 900-232 500) 51 (43-61) 30
No.cpue 317 500 (248 900–466 500) 172000 (105 800–318 300) 54 (42–69) 35
Half.M 314 500 (281 300–384 600) 103 100 (72 100–170 500) 33 (26–44) 25


Figure A1.4.
Figure A1.4: Estimated spawning biomass and exploitation rates for the Spawning Box and Northeast Flats subarea for the CASAL runs only. Patterns for the Awatea runs were qualitatively similar. The observed CPUE series are plotted without confidence bounds (+, Spawning Box pre-closure; Ï, Spawning Box post-closure; p, Northeast Flats) and the wide-area acoustic estimates are plotted with 95% confidence bounds. The 95% confidence interval from the estimated posterior distribution of the exploitation rates is also shown. The horizontal line on the right panel indicates ECAY ( 0.064 for M=0.045, and 0.032 for HalfM).

  1. All runs indicated that biomass decreased until1990–91 or 1991–92 followed by an increase, except for the CASAL All. R run, which indicated a biomass decrease until 1997–98 followed by an increase. The main differences between the runs were the starting point of the biomass trajectory and the extent to which the stock increased relative to its minimum, resulting in different current estimates of biomass (Figure A1.4). The three runs that estimated substantially different patterns from the rest are the NoCPUE, HalfM, and All.R runs. The NoCPUE runs gave the highest %B 0 because biomass did not decline as steeply in the pre-closure period. The HalfM runs resulted in relatively small increases in biomass since the year it was at a minimum, and gave the lowest estimates of Bcurrent. Estimating initial recruitment deviates (All.R run) produced a pattern of recruitment where a large pulse of recruitment coincides with the start of the fishery, followed by a large reduction in recruitment during the period of the fishery. This causes a starting biomass larger than average virgin biomass and little increase in biomass when catches are reduced. The NoPostClose run shows that the post-closure biomass increase indicated in the All.W run is not driven by the post-closure data. This suggests that this increase in biomass is simply a product of the pre-closure data and the productivity assumptions built into the model. The HalfM and All.R runs show that when these productivity assumptions are altered then so is the strength of the increase.
  2. Yield estimates for the Spawning Box and Northeast flats. The All.W and NoCPUE runs generate a 2007 CAY and MAY that are considerably greater than the 2004–05 catch of 5 490 t. However, these yields are less than the 2004–05 catch for the HalfM run sensitivity analysis due to the lower implied productivity (Table A1.5).

Table A1.5. CAY catch and MAY (medians with 95% CI) for the Spawning Box and Northeast Flats subarea.

CASAL
Run
CAY (t)
MAY (t)
All W
11 200 (7 900-18 300) 6 100 (5 100-8 300)
NoCPUE
 13 800 (8000-27 600) 6 900 (5 100-11 300)
Half.M
 4000 (2 880–6 300) 3 250 (2 900– 4000)
Awatea
Run
CAY (t)
MAY (t)
All W
9300 (6800-14 400) 5500 (4800-7200)
NoCPUE
 10 700 (6700-19 600) 6000 (4700-8800)
Half.M
 3200 (2300-5200) 3000 (2700-3600)

  1. Northeast Hills – B0 was estimated to be 17 800 t (95% confidence interval
    14 700-22 900 t) and the current biomass 14% B0 (95% confidence interval 7–32%).
    Estimated biomass has declined rapidly, but after substantial reductions in catches in the last two years it has increased by approximately 4.5% per annum. Catch-rates over the last three years have also been relatively stable (Table A1.6). Estimated exploitation rates have declined in recent years but are still almost three times E CAY (Figure A1.5). The 2007 median estimate for the CAY policy at 170 tonnes (95% confidence interval 80 to 460 tonnes) is less than recent catch. It was agreed that the only one run reported for the Northeast Hills subarea would be that using a shortened CPUE series (first three years eliminated).

Figure A1.5.
Figure A1.5. Estimated spawning biomass trajectory and exploitation rates for the Northeast Hills sub-area.The observed CPUE with 95% confidence bounds is plotted with the spawning biomass trajectory. The 95% confidence interval from the estimated posterior distribution of the exploitation rates is also shown. The horizontal line on the right panel indicates ECAY (0.064).

Table A1.6. Biomass estimates (medians, with 95% CI) for the Northeast Hills

B0(t)Bcurrent (t)Bcurrent (% B0)
17 800 (14 700–22 900)
2 500 (1 100–7 200)
14 (7–32)

  1. Yield estimates for the Northeast Hills (Table A1.7). The CAY for 2007 (170 t) and the MAY (350 t) are less than recent catches (870, 640, and 540 t for the three most recent years).

Table A1.7. CAY catch and MAY for the Northeast Hills subarea. The median is shown with 95% CI.

CAY Catch (t)MAY (t)
170 (80 – 460)
350 (290 – 460)

  1. Andes Complex – It was agreed that the only run reported for the Andes subarea would be that using a shortened CPUE series (first three years eliminated). B0 was estimated to be 35 700t (95% confidence interval 29 300–97 600t), and the current biomass 29% B0 (95% confidence interval 14–74%). Biomass has continued to decline despite recent reductions in catches (Table A1.8). Exploitation rates have been variable but in recent years they have been above the exploitation rate under a CAY policy, E CAY (0.064) (Figure A1.6).

Table A1.8. Biomass estimates (medians, with 95% CI) for the Andes subarea assessment.

B0(t)Bcurrent (t)Bcurrent (% B0)
35 700 (29 300–97 600)
10 300 (4 400–72 100)
29 (14–74)

Figure A1.6.
Figure A1.6: Estimated spawning biomass trajectory and exploitation rates for the Andes subarea assessment.The observed CPUE with 95% CI is plotted with the spawning biomass trajectory. The 95% CI from the estimated posterior distribution of the exploitation rates is also shown. The horizontal line on the right panel indicates ECAY (0.064)

  1. Yield estimates for the Andes Hills complex. The CAY for 2007 (650 t) and the MAY (710 t) (Table A1.9) are about quarter to half of the catches in recent years (2 870, 1 540, 1 410 t for the three most recent years).

Table A1.9 . CAY catch and MAY for the Andes subarea. The median is shown with 95% CI.

CAY Catch (t)MAY (t)
650 (280 – 4 420)
710 (580 – 1 940)

Previous | Next
Updated : 16 November 2007