IPHC_bluebook_2013.pdf

12013 IPHC ANNUAL MEETING HANDOUT

International Pacifi c Halibut CommissionEighty-ninth Annual Meeting

Meeting Information

Navigating the 2013 IPHC Annual Meeting ..................................................................................2Detailed Meeting Agenda ..............................................................................................................3Halibut Terminology – What You May Hear .................................................................................6

Staff presentations

The Pacifi c halibut fi shery, 2012 ....................................................................................................9Heather L. Gilroy

IPHC Management Strategy Evaluation framework ...................................................................28Steven Martell

Assessment of the Pacifi c halibut stock at the end of 2012 .........................................................32Ian J. Stewart, Bruce M. Leaman, Steven Martell, and Raymond A. Webster

IPHC Staff harvest advice and regulatory proposals: 2013 ....................................................... 114Bruce M. Leaman, Ian J. Stewart, Raymond A. Webster, and Heather L. Gilroy

Optimal harvest rates for Pacifi c halibut ....................................................................................131Steven Martell, Ian J. Stewart, and Bruce M. Leaman

Coastwide comparison of alternative setline survey baits .........................................................154Raymond A. Webster, Steven M. Kaimmer, Claude L. Dykstra and Bruce M. Leaman

Halibut size at age research ........................................................................................................172Bruce M. Leaman, Steven Martell, Timothy Loher, Kirstin K. Holsman, Bruce S. Miller, Kerim Y. Aydin, and Gordon H. Kruse

Appendices

Appendix I. IPHC 2013 Annual Research Plan (Preliminary) ..................................................186IPHC Staff

Appendix II. IPHC regulation proposals submitted for the 2013 Annual Meeting ...................205IPHC Staff

Appendix III. Catch limit, research, stock assessment and apportionment comments submitted for the 2013 Annual Meeting .....................................................................................................220

Cover artwork: Joan Forsberg


Nav

igat

ing

the

2013

IPH

C A

nnua

l Mee

ting

Day

Wha

t Tim

e? R

oom

?Gr

oup

Mee

ng/

Sess

ion

Who

is M

ee n

g?W

ho C

an A

end

?W

hat’s

Bei

ng D

iscu

ssed

Mon

day

-Ja

n. 2

1

11:0

0–N

oon

– Bu

ckin

gham

IPHC

Exe

c Se

ssio

nCo

mm

issio

nCl

osed

Mee

ng

issue

s1:

00–5

pm

– C

ryst

al B

allro

omPu

blic

Ses

sion

ICo

mm

issio

n an

d pu

blic

Ope

nFi

sher

y re

sults

and

ass

essm

ent

6:30

-8:3

0 pm

– B

ucki

ngha

mU

S De

lega

on

mee

ng

US

com

miss

ione

rs w

ith U

S a

end

ees

Any

U.S.

a e

ndee

sIs

sues

of i

nter

est t

o th

e U

S in

dust

ry

6:30

-8:3

0 pm

– B

alm

oral

Cana

dian

Del

ega

on

mee

ng

Cana

dian

com

miss

ione

rs w

ith C

anad

ian

a e

ndee

sAn

y Ca

nadi

an a

end

ees

Issu

es o

f int

eres

t to

the

Cana

dian

in

dust

ry

Tues

day

-Ja

n. 2

2

8-11

am

– C

ryst

al B

allro

omPu

blic

Ses

sion

IICo

mm

issio

n an

d pu

blic

Ope

nRe

sear

ch &

Per

form

ance

Rev

iew

1-5

pm –

Buc

king

ham

IPHC

Adm

in I

Com

miss

ion

Ope

nAr

eas 2

A an

d 2B

age

ncy

repo

rts

1-5

pm –

Cry

stal

Bal

lroom

Conf

eren

ce B

oard

IHa

rves

ters

Ope

nCa

tch

limits

, reg

pro

posa

ls1-

5 pm

– B

alm

oral

Proc

esso

r Adv

isory

Gro

up I

Halib

ut p

roce

ssor

sO

pen

Catc

h lim

its, r

eg p

ropo

sals

Wed

nesd

ay -

Jan.

23

8-11

:30

am –

Buc

king

ham

IPHC

Adm

in II

Com

miss

ion

Ope

nAr

eas 2

C/3/

4 ag

ency

repo

rts

1-2:

30 p

m –

Buc

king

ham

IPHC

Adm

in II

ICo

mm

issio

nO

pen

Rese

arch

, Byc

atch

Wor

k Gr

oup

2:30

-5 p

m –

Buc

king

ham

IPHC

Fin

ance

& A

dmin

Com

miss

ion

Clos

edFi

nanc

e an

d Ad

min

8-5

pm –

Cry

stal

Bal

lroom

Conf

eren

ce B

oard

II &

III

Harv

este

rsO

pen

Catc

h lim

its, r

eg p

ropo

sals

8-5

pm –

Bal

mor

alPr

oc A

dviso

ry G

roup

II &

III

Halib

ut p

roce

ssor

sO

pen

Catc

h lim

its, r

eg p

ropo

sals

Thur

sday

-Ja

n. 2

4

8-10

:30

am –

Cry

stal

Bal

lroom

IPHC

, CB,

PAG

join

tIP

HC, C

B, &

PAG

Ope

nCB

and

PAG

reco

mm

enda

ons

10:3

0-N

oon

– Bu

ckin

gham

IPHC

Adm

in IV

Com

miss

ion

Ope

nCB

/PAG

rec’s

, pro

posa

ls10

:30-

5 pm

– C

ryst

al B

allro

omCo

nfer

ence

Boa

rd IV

& V

Harv

este

rsO

pen

ROP;

oth

er m

a e

rs10

:30-

5 pm

– B

alm

oral

Proc

Adv

isory

Gro

up IV

& V

Halib

ut p

roce

ssor

sO

pen

ROP;

oth

er m

a e

rs1:

30-5

pm

– B

ucki

ngha

mIP

HC A

dmin

VCo

mm

issio

nO

pen

Unfi

nish

ed b

usin

ess

Frid

ay -

Jan.

25

7:30

-8:3

0 am

– C

ryst

al B

allro

omIP

HC E

xec

sess

ion

Com

miss

ion

Clos

edM

ee n

g iss

ues

9-11

am

– C

ryst

al B

allro

omPu

blic

Ses

sion

IIICo

mm

issio

nO

pen

FIN

AL D

ECIS

ION

S on

regu

la o

ns

and

catc

h lim

its

12:3

0–2

pm –

Cry

stal

Bal

lroom

IP

HC A

dmin

VI

Com

miss

ion

Ope

nO

ther

bus

ines

s (if

need

ed)


International Pacifi c Halibut CommissionEighty-ninth Annual Meeting

The Fairmont Empress Victoria, BC

January 21-25, 2013

DETAILED MEETING AGENDA

Notes: 1. All sessions are open to the public unless noted as ‘closed.’ All Commission Public and

Administrative Sessions will be webcast and may take audience/webcast comments & questions (CB and PAG sessions will not be broadcast).

2. Breaks and comments/discussion as determined by Commissioners.3. Briefi ng materials will be posted at the IPHC Annual Meeting webpage ((http://iphc.int/http://iphc.int/

meetings-and-events/annual-meeting.htmlmeetings-and-events/annual-meeting.html) as they become available.) as they become available.4. Presentations will be posted at the Annual Meeting webpage following each day’s ses-

sions.5. Breakout rooms available in Kensington & Library (contact IPHC staff for availability).6. IPHC offi ce is located in St. James.

Monday - January 21 Room Location

a.m. 11:00 – 12:00 Executive Session I (closed) Buckingham

p.m. 1:00 – 5:00

Public Session I Opening remarks Staff reports:

o State of the fi shery in 2012o Management Strategy Evaluation

(MSE) frameworko 2012 stock assessmento Harvest decision table and regulatory

proposals Public comments Discussion

Crystal Ballroom

6:30 – 8:30 United States Delegation Meeting Buckingham

6:30 – 8:30 Canadian Delegation Meeting Balmoral


Tuesday - January 22 Room Location

a.m. 8:00 – 11:00

Public Session II Staff reports:

o Optimal harvest rateso 2012 bait experimento Size-at-age research

Public comments Discussion Progress on Performance Review Recommen-

dationso Advisory Bodies Rules of Procedureo Research Plano Management Strategy Advisory Board

(MSAB)o Scientifi c Review Board (SRB)

Public comments Discussion

Crystal Ballroom

p.m. 1:00 – 5:00

IPHC Administrative Session I Agency reports for Area 2A Public comments Discussion Agency reports for Area 2B Public comments Discussion

Buckingham

1:00 – 5:00 Conference Board (CB) Session I Crystal Ballroom

1:00 – 5:00 Processor Advisory Group (PAG) Session I Balmoral

7 :00 – 9:30 IPHC Reception Palm Court

Wednesday - January 23

a.m. 8:00 – 11:30

IPHC Administrative Session II Agency reports for Alaska (Areas 2C-3-4) Public comments Discussion

Buckingham

8:00 – 12:00 Conference Board (CB) Session II Crystal Ballroom

8:00 – 12:00 Processor Advisory Group (PAG) Session II Balmoral

p.m. 1:00 – 2:30

Administrative Session III IPHC research program Halibut bycatch working group (HBWG II) Public comments Discussion

Buckingham


Wednesday - January 23 (continued) Room Location

1:00 – 5:00 Conference Board (CB) Session III Crystal Ballroom

1:00 – 5:00 Processor Advisory Group (PAG) Session III Balmoral

2:30 – 5:00 Finance & Administration Committee (closed) Buckingham

6:30 – 8:30 HANA Reception Sticky Wicket Pub,919 Douglas Street

Thursday - January 24

a.m. 8:00 – 10:30

IPHC, CB, and PAG Joint Session CB and PAG recommendations Public comments Discussion

Crystal Ballroom

10:30 – 12:00

IPHC Administrative Session IV Harvest decision table and regulatory propos-

als Agency and industry proposals Public comments Discussion

Buckingham

10:30 – 12:00 Conference Board (CB) Session IV Crystal Ballroom

10:30 – 12:00 Processor Advisory Group (PAG) Session IV Balmoral

p.m. 1:30 – 5:00

IPHC Administrative Session V Unfi nished business Public comments Discussion

Buckingham

1:30 – 5:00 Conference Board (CB) Session V Crystal Ballroom

1:30 – 5:00 Processor Advisory Group (PAG) Session V Balmoral

Friday - January 25

a.m. 7:30 – 8:30 Executive Session II (closed) Crystal Ballroom

9:00 – 11:00

Public Session III CB and PAG input and recommendations re-

garding advisory bodies Public comments Discussion Adoption of regulations and catch limits Public comments Discussion Other actions as needed

Crystal Ballroom

p.m. 12:30 – 2:00 Administrative Session VI (if needed) Crystal Ballroom


Halibut Terminology – What You May HearIPHC uses many scientifi c terms, acronyms, and plain ol’ jargon in managing the halibut

resource and fi shery. Some are unique to halibut management, whereas others are more common in the fi sheries management business. Here’s a list of some of the most common terms you may hear.

Archival tag ......A type of tag which measures and stores environmental data while attached to a fi sh.

Availability .......The relative proportion of the stock, by age or size, that is present in a given area.

BAWM .............Bycatch and wastage mortality.

Biomass ............Weight in net pounds.

Blue Line ..........A row in the Decision Table, highlighted in blue, which provides the harvest available when the Commission’s target harvest rates are applied to the current estimate of Ebio.

Bycatch ............Halibut caught by fi sheries targeting other species. Often ‘bycatch’ is used interchange-ably with the term ‘bycatch mortality’ but in either case, this represents the amount of halibut killed, generally expressed in pounds, net weight.

Catchability ......Fraction of total available fi sh caught per unit of effort.

CEY ..................Constant Exploitation Yield. IPHC measures CEY in terms of Total CEY, which is the total amount of yield available for harvest in an area, and Fishery CEY, which is the amount of yield available for the directed halibut fi sheries.

CB ....................Conference Board, an advisory body to IPHC, composed of representatives of the fi shing industry.

CSP ...................Catch Sharing Plan. A management program which allocates available yield among spe-cifi c user groups or sectors. Allocations are often described as percentages of the whole.

Decision Table ..A table developed by the IPHC staff, in collaboration with external scientifi c review, in which the probabilities for specifi c metrics of stock and fi shery performance are estimat-ed for various levels of harvest.

Ebio ..................An abbreviation for Exploitable biomass, which is the fraction of the Total biomass ex-ploited by the directed fi sheries and for which the harvest policy is defi ned.

GHL .................Guideline Harvest Level; part of a management program design by the North Pacifi c Fishery Management Council in 2003 to manage the harvest by the sport charter halibut fi shery in Areas 2C and 3A.

MSAB ..............Management Strategy Advisory Board. A group of 15-20 individuals appointed by the Commission, comprised of harvesters, managers, processors, academia, IPHC staff, and the IPHC science advisors. The role of the MSAB is to defi ne clear, measurable fi shery objectives and to provide technical input on the development of an operating halibut fi shery management model that will permit evaluation of various strategies to achieve management objectives.

MSE .................Management Strategy Evaluation. A long-term process in which alternative management strategies for the halibut fi shery are developed and evaluated by an advisory board.


NPFMC ............North Pacifi c Fishery Management Council, headquartered in Anchorage, Alaska. A U.S. federal board which oversees management of marine fi sheries in the federal waters off Alaska. With halibut, the NPFMC’s role is to decide on allocations between user groups and development of programs to manage and reduce halibut bycatch.

NPUE ...............Number of fi sh per unit effort. Effort is measured in terms of a standard skate of fi shing gear, which is defi ned by IPHC as an 1,800-foot skate of gear, with 100 hooks at an 18-foot spacing.

PAG ..................Processor Advisory Group, an advisory body to IPHC composed of representatives from halibut processors.

PFMC ...............Pacifi c Fishery Management Council, headquartered in Portland, Oregon. A U.S. federal board which oversees management of marine fi sheries in federal waters off Washington, Oregon and California. With halibut, the PFMC’s role is to decide on allocations between user groups and development of programs to manage and reduce halibut bycatch.

PHI ...................Prior hooking injuries. A type of injury incurred by a halibut upon being released and returned to the sea during a previous capture.

PIT tag ..............Passive Integrated Transponder tag, used by IPHC in a coastwide tagging study in 2003-2004.

Profi ler..............Oceanographic equipment used to collect water column environmental data aboard IPHC chartered research vessels.

RAB .................Research Advisory Board, an advisory body to the Commission which provides input on research topics and direction. The RAB is composed of representatives of the halibut fi sh-ing and processing industry.

Sbio ..................An abbreviation for female spawning biomass, measured in weight, which is comprised only of sexually mature female halibut. Sexual maturity begins at age 8, reaches 50% at age 13 and 100% at age 20.

Selectivity ........Relative probability of a fi sh being retained by the gear as a function of its size (or age).

SRB ..................Scientifi c Review Board. A group of 4-6 scientists appointed by the Commission to pro-vide independent scientifi c review of the halibut assessment and research.

SSA survey .......Standardized stock assessment survey, conducted annually by IPHC since 1996.

SWHS ..............Statewide Harvest Survey, used by the Alaska Department of Fish and Game to estimate sport harvests.

Tbio ..................An abbreviation for Total biomass, measured in weight, which is the biomass of all hali-but, generally for ages 8 and older.

Wastage ............The mortality (net pounds) incurred during mandatory release of undersize halibut by the commercial halibut fi shery, or from lost and abandoned fi shing gear.

WPUE ..............Pounds of fi sh per unit effort. Effort is measured in terms of a standard skate of fi shing gear, which is defi ned by IPHC as an 1,800-foot skate of gear, with 100 hooks at an 18-foot spacing.


Notes


The Pacifi c halibut fi shery, 2012

Heather L. Gilroy

The International Pacifi c Halibut Commission (IPHC) estimates all halibut removals taken off the Pacifi c coast and uses this information in its harvest policy and the stock assessment. When the Commission was established in 1923, the commercial fi shery, which dates back to the late 1800s, was the only documented source of harvest. Since that time, the commercial catch has continued to be the largest removal coastwide. Through the years, estimates of other removals have been added: bycatch mortality in the 1960s, sport catch in the late 1970s, wastage in the 1980s, and personal use (subsistence) in the 1990s. In 2012, the removals from all sources totaled 51.7 million pounds (Table 1). Total removals have declined from the 90-100 million range which occurred during 1998-2007, and are now at a level similar to removals seen in the early 1980s.

This report reviews the catch by commercial and sport fi sheries, bycatch mortality estimates, and the allocation programs within each country. In addition, catch versus catch limits or harvest limits are also reported.

Personal use removals include the Washington State treaty tribal ceremonial and subsistence (C&S) fi shery, the British Columbia Food, Social, and Ceremonial (FSC) fi shery, and the Alaska subsistence fi shery (Williams 2013a). The C&S fi shery is part of a catch sharing plan and is estimated at 24,500 pounds for 2012 by the treaty tribes. The Canadian Department of Fisheries and Oceans (DFO) provided the estimate of the FSC fi shery catch. For 2012, FSC catch was estimated at 0.405 million pounds, which has been unchanged since 2007. The Commission has not seen documentation on the estimation methodology; it is reported simply as the sum of all allocations within the FSC communal licenses including recent treaty settlements (T. Karim, DFO, pers. comm.). The Alaska subsistence harvest was estimated by Alaska Department of Fish and Game (ADF&G) to be approximately 0.7 million pounds for 2011, the latest year for which information is available. The 2011 harvest is lower than in 2010, and the harvest has continually declined since the 2004-2005 high of 1.2 million pounds. The Alaska program was implemented in 2003.

The incidental mortality of halibut from the commercial halibut fi sheries (1.6 million pounds) is termed wastage (Gilroy and Stewart 2013). The legal minimum size limit for the commercial halibut fi shery is 32 inches in total length, so the wastage estimates include the mortality of halibut 32 inches and over (O32) killed by lost and abandoned longline gear, and the proportion of halibut under 32 inches (U32) that must be released and subsequently die. A breakdown for halibut over 26 inches (O26) and under 26 inches (U26) in total length has been provided in Table 1 for wastage and bycatch mortality as the removals are treated differently within the harvest policy (Stewart et al. 2013).

In addition to data compiled by the IPHC, other sources of halibut data include federal and state agencies. All 2012 data are net weight and considered preliminary at this time. These data were updated in December 2012 after the completion of the IPHC Report of Assessment and Research Activities (RARA) 2012 and the stock assessment tables, so numbers provided here may differ slightly from those reports. Catch limit, as referred to here, is the IPHC catch limit. The adjusted catch limit represents the IPHC catch limit with adjustments by DFO and NMFS from the underage and overage programs from the previous year’s quota share program and other


quota leasing programs determined by DFO in British Columbia. The IPHC regulatory areas are provided in Figure 1.

Allocation programs, updates, and totals

The authority for domestic allocation among the user groups rests with the national governments. Currently, both the United States and Canadian governments have allocation plans for halibut fi sheries within a single regulatory area or within smaller local areas.

Washington, Oregon, and California – Area 2AComprehensive user group allocation occurs in Area 2A. The Commission determines the

Area 2A total catch limit for all user groups, and the Pacifi c Fishery Management Council (PFMC) allocates halibut catch limits among user groups through a Catch Sharing Plan (CSP). In this plan, 44.4% is allocated to the sport fi shery, 35% to the treaty tribes, and 20.6% to the non-treaty commercial fi shery.

The PFMC CSP specifi es that a primary limited-entry longline sablefi sh (Anoplopoma fi mbria) fi shery north of Point Chehalis, WA (4653’18”N) would be allocated part of the Washington sport allocation only if the Area 2A catch limit was over 900,000 pounds and the amount available to the incidental fi shery was over 10,000 pounds. These criteria were met for the fi rst time since 2009, and an incidental halibut fi shery during the limited-entry sablefi sh fi shery occurred in 2012. Therefore, in 2012 the Area 2A catch limit was divided among the two sport fi sheries, a treaty Indian commercial fi shery and a ceremonial and subsistence fi shery, a non-treaty commercial directed fi shery, and two incidental halibut fi sheries (during the salmon troll fi shery and limited-entry sablefi sh fi shery) (Table 2). The total 2012 Area 2A catch (not including IPHC research) of 1.013 million pounds was over the catch limit (0.989 million pounds) by 2.5%.

British Columbia – Area 2BIn 2005, DFO implemented an allocation framework for the commercial and recreational

sectors, with the recreational sector receiving a 12 percent “ceiling” of a combined commercial/recreational catch limit. In 2011, the Minister of Fisheries and Oceans Canada initiated a review of the Pacifi c halibut allocation between users, which resulted in a new allocation ratio of 85% commercial to 15% sport for 2012. When managed to the allocation ceilings, both sectors’ catch will fl uctuate with stock abundance.

The IPHC adopted an Area 2B combined sport and commercial catch limit of 7.038 million pounds that was allocated to the user groups by the DFO. The 7.038 million pounds combined catch limit was arrived at after 0.193 million pounds of commercial wastage for O26 halibut had been deducted from the Area 2B constant exploitation yield. To avoid penalizing the sport sector for commercial wastage, DFO added back the previously deducted 0.193 million pounds of wastage and partitioned this new total by the 85:15 (commercial to sport) ratio. The commercial wastage amount was then deducted solely from the commercial portion. The newly calculated catch limits available to the commercial and sport fl eets were 5.953 million pounds and 1.085 million pounds, respectively. DFO made additional adjustments to the quotas, which included pounds available from the underage/overage plan, quota leased from the commercial fi shery to the sport fi shery, and quota held by DFO for First Nations through a relinquishment process. The total 2012 Area 2B catch of 6.955 million pounds was under the catch limit (7.038 pounds) by 1% (Table 3).


Alaska – Areas 2C, 3A, 3B, and 4 Two allocation plans exist for fi sheries within Alaska. The fi rst applies to Area 4CDE, in which

the Commission adopts a single commercial fi shery catch limit, as the area is considered to be one biological unit. A CSP developed by the North Pacifi c Fishery Management Council (NPFMC) in 1996 specifi es individual catch limit allocations for Areas 4C, 4D, and 4E. This CSP also allows Area 4D Community Development Quota (CDQ) to be harvested in Area 4E, and Area 4C quota shares to be fi shed in Areas 4C or 4D. The total commercial catch of 2.294 million pounds was under the combined Area 4CDE catch limit (2.464 million pounds) by 7%.

In 2003, a Guideline Harvest Level (GHL) program was implemented to manage the harvest by the sport guided (charter) fi sheries in Areas 2C and 3A. The GHL program includes a provision that the GHL adjusts by specifi ed increments to changes in halibut abundance, yet will not increase above a maximum level, nor decrease below a minimum level (Williams 2013a). In 2012, the Areas 2C and 3A sport charter harvests were under their GHLs by 31%, and 23%, respectively (Table 4). The GHL program remains in place until superseded by a commercial-sport charter CSP that was adopted by the NPFMC. The CSP is expected to be implemented by the 2014 fi shery.

Detailed catch data

The commercial fi sheryA detailed summary of fi shing seasons, catch limits, and catch by IPHC regulatory area is

provided in Table 5. Commercial catch occurs in: an open-access fi shery, two incidental catch fi sheries, and a treaty Indian fi shery in Area 2A; the quota share (QS) fi sheries in British Columbia and Alaska; and the Metlakatla fi shery within the Annette Island Reserve in southeast Alaska (Area 2C).

Area 2AIn 2012, IPHC issued 619 Area 2A vessel licenses, including 311 licenses for retaining

incidental halibut during the salmon troll fi shery, 177 licenses for the directed commercial fi shery and incidental halibut during the sablefi sh fi shery, and 131 licenses for sport charter vessels. The largest change in number of licenses issued compared to 2011 was for the directed commercial fi shery/incidental halibut during the sablefi sh fi shery which increased by 30 licenses, in part due to the allowance of an incidental halibut fi shery during the sablefi sh fi shery. The number of licenses issued for the sport charter fi shery and the incidental halibut during the salmon troll fi shery decreased by 10 and fi ve licenses, respectively.

The Area 2A directed commercial fi shery closed after two 10-hour openings with fi shing period limits (Table 6). The fi shing period limits were assigned by vessel class; the H-class vessels received 9,000 pounds for the June 27 opening and 1,500 pounds for the July 11 opening (Table 5). The total directed commercial catch of 179,000 pounds was 3.3% over the catch limit (173,216 pounds).

The management objective for the incidental halibut fi shery during the salmon troll season is to harvest the majority of the halibut catch in May and June. The allowable incidental catch ratio of halibut during the salmon troll fi shery was one halibut per four chinook, plus an “extra” halibut per landing. However, the total number of incidental halibut per vessel per landing could not exceed 20. In 2010 and 2011 the ratio was 1:3 for halibut to chinook, with the total number of halibut per vessel per landing not to exceed 35. The landing restrictions have become increasingly


conservative since 2008 and 2009, when the landing ratio was 1:2. The objective was met, as the season opened on May 1 and closed on July 3, with a total catch of 35,300 pounds, 15% over the catch limit (30,568 pounds). Unfortunately, a perfect storm occurred with the four agencies that work together to track landings and close the fi shery, as one agency had communication infrastructure problems, and there were issues with all four coordinating over a holiday weekend such that the catch limit was exceeded.

Incidental halibut retention during the limited-entry sablefi sh fi shery remained open from May 1 to October 31, closing at the end of the sablefi sh fi shery season. The allowable landing ratio was 50 pounds of halibut to 1,000 pounds (net weight) of sablefi sh, and up to two additional halibut in excess of the ratio limit. The total catch was 4,400 pounds, 79% under the catch limit (21,173 pounds).

In Area 2A-1, the treaty Indian fi sheries management plan allowed for both unrestricted fi sheries with no landing limits and restricted fi sheries with limits, as well as a late season mop-up fi shery that could be set up with or without landing limits. The restricted fi shery opened at noon on March 17 with a 500 pound per vessel per day limit. The restricted fi shery lasted 55 hours and closed on March 19. The unrestricted fi shery opened at noon on March 24 and closed 48 hours later at noon on March 26. A mop-up fi shery, with no landing restrictions, occurred on May 2 and lasted 13 hours. The weather was good during this 13-hour opening and the landings were higher than predicted. The total treaty Indian commercial catch was 355,100 pounds or 10% over the catch limit (321,650 pounds).

Area 2C Metlakatla fi sheryThe Metlakatla Indian Community was authorized by the United States government to

conduct a commercial halibut fi shery within the Annette Islands Reserve. There were eleven 2-day openings between April 20 and September 23 for a total catch of over 49,000 pounds This was almost 13,000 pounds lower than the 2011 catch, and within the historical catch range that has varied over time from a low of 12,000 pounds in 1998 to a high of 126,000 pounds in 1996.

The Quota Share fi sheries – British Columbia and AlaskaIn Area 2B, the halibut IVQ program was implemented in 1991, and 435 vessels received

IVQs. Each initial IVQ was split into two shares called blocks. Numerous changes have been made since then, including allowing temporary block transfers (1993) and then permanent block and IVQ transfers (1999). The Groundfi sh Integrated Fisheries Management Plan (IFMP) was implemented in 2006 to address rockfi sh conservation concerns and improve catch monitoring. The IFMP includes quota shares for all hook and line groundfi sh fi sheries, transferability with limits between license holders, 100% at-sea and dockside monitoring, and vessel accountability for all catch, both landed and discarded. There is 100% monitoring through logbook recordings, video camera coverage, and dockside coverage.

In 2012, the number of vessels landing Area 2B halibut was 234, with 150 vessels with halibut licenses (L licenses and First Nations communal commercial licenses or FL licenses) and 84 vessels licensed for other fi sheries but landing halibut as incidental catch as part of the IFMP. The commercial catch of 5.8 million pounds for Area 2B was 2% under the catch limit.

In Alaska, the Individual Fishing Quota/Community Development Quota (IFQ/CDQ) halibut and sablefi sh fi shery program has been in effect since 1995. NMFS Restricted Access Management (RAM) allocates halibut QS to recipients by IPHC regulatory area. Quota share transfers were


permitted with restrictions on the amount of QS a person could hold and the amount that could be fi shed per vessel. As of November 26, 2012, RAM reported that 2,574 persons held quota shares, down from the initial 4,831 persons at the start of the program (Table 7). The number of vessels catching halibut has decreased by 44% since the implementation of the QS program.

The total catch in 2012 from the IFQ/CDQ halibut fi shery for the waters off Alaska was 24.7 million pounds, which was 3.3% under the catch limit. For comparison, the 2011 commercial catch was 2.5% under the catch limit. For Areas 2C, 3A, 3B, and 4A, the commercial QS catches were under the catch limits by 2%. For Areas 4B and 4CDE, the catches were 10 and 7% under the catch limit respectively. As mentioned previously, the NPFMC CSP allowed Area 4D CDQ to be harvested in Areas 4D or 4E and Area 4C IFQ and CDQ to be fi shed in Areas 4C or 4D. These two regulations were the reason the catches in Areas 4D and 4E exceeded the catch limits.

Landing patternsCommercial trips from Area 2B were delivered into 11 different ports in 2012. The ports of

Port Hardy (including Coal Harbour and Port McNeill), Prince Rupert/Port Edward, and Vancouver were the major landing locations, receiving 93% of the Area 2B commercial catch. Port Hardy received 49% (2.8 million pounds) of the Area 2B commercial catch. Prince Rupert/Port Edward and Vancouver received the second and third largest landing volumes, representing 39% and 5% of the Area 2B commercial catch, respectively. All of the IVQ catch was landed in Area 2B.

Homer and Kodiak received approximately 18% (4.4 million pounds) and 19% (4.8 million pounds) of the commercial Alaskan catch, respectively. Landed pounds for these two ports were closer in 2012 and in 2011 than they had been previously. For example, in 2010 Homer and Kodiak received 25% and 14% of the landed poundage, respectively. Kodiak had been the busiest port in the derby days and is again one of the top ports for pounds landed. Seward received the third largest landing volume at 11% of the Alaskan commercial catch. In SE Alaska, Sitka received 5%, Petersburg 4%, and Juneau 4%. The Alaskan QS catch that was landed outside of Alaska was 2.3%.

The 2012 QS fi shery landings were spread over nine months of the year (Table 8). The proportion of catch landed earlier in the season in British Columbia shifted in the last two years with approximately 38% of the Area 2B catch landed by the end of May in both 2011 and 2012 compared to 49% landed by the end of May in 2010. On a month-to-month comparison, August took the lead as the busiest month for total poundage (19%) from Alaska. In 2011, May was the busiest month for total poundage from Alaska, but ranked third for 2012 landings. August was the busiest month for poundage delivered in British Columbia with approximately 18% of the catch.

The sport fi sheryThe 2012 sport harvest was estimated at 6.9 million pounds, a slight decrease of 2.2% over

last year, and below the 10-11 million pound amounts from 2004-2008 (Table 9). Decreases were seen in all regulatory areas, except for in Areas 2A and 2C.

Washington, Oregon, and CaliforniaThe Area 2A sport harvest of 0.415 million pounds (Williams 2013b) was very close to the

allocation (Table 10). To manage the sport fi shery to the CSP allocation, the area is subdivided into six subareas with their own quotas, with four core areas having in-season catch monitoring. A subarea is closed when the subarea quota is achieved. The in-season management includes fi shery


openings in the spring and fall, changes in the number of fi shing days allowed within a week, and closing fi shing seasons from weekly to every other week (Table 11). Tracking and validation of the landings are done dockside by state creel samplers. The two subareas without in-season monitoring (Inside Washington waters and South OR/California) have daily bag limits and a fi xed season length typically based on fi shing success in the preceding year. In 2011, the Oregon portion of the South OR/California subarea had exceptional fi shing success for the area and it appears the same unprecedented success was not repeated in 2012. New estimates for the California portion of this area for 2004-2011 have been provided to IPHC by California Department of Fish and Wildlife. Agency staffs worked in 2012 to review the catch estimation and monitoring procedures in the South OR/California subarea and are examining methods for longer term monitoring and fi shery management methods.

British ColumbiaThe DFO provides the sport estimates that are from mixed sources including overfl ights,

on-water vessels counts, lodge logbooks, and creel sampling by DFO and the Haida Fisheries Program. Piece counts are available from lodge logbooks and creel samples, but size composition data are limited, and lodge data are mostly self-reported and unverifi ed.

DFO instituted several management restrictions on the sport fi shery to constrain the harvest to stay within the sport allocation and extend the season as long as possible. Two management restrictions were unchanged from 2011, and included a reduction of the daily bag limit from two to one halibut and the prohibition on halibut retention in Area 121 seaward of 12 nmi (SW off Vancouver Island). In 2011, the possession limit changed from three-fi sh to two-fi sh, and additional restrictions were implemented in 2012. Starting April 1, a maximum length restriction was implemented with the two fi sh possession limit, where one halibut had to be less than 83 cm in total length.

The 2012 sport season opened on March 1 and closed on September 9. The 2012 sport catch of 1.14 million pounds exceeded the sport fi shery allocation by 5.5%. The 2012 measures brought the sport harvest closer to the allocation as it had exceeded the allocation by 28% in 2011 and 2010.

For the second season, DFO implemented an experimental leasing program, in which interested recreational fi shers could receive experimental licenses that would allow them to lease halibut quota from commercial fi shery quota holders which would permit continued sport fi shing after the general sport fi sh closure. In 2012, a total of 59 licenses were issued, with 2,010 pounds transferred to the sport fi shery. The preliminary estimate is that 814 pounds were caught under this leasing program. DFO is currently determining whether to make the program permanent.

AlaskaEstimates of the sport fi shery harvest off Alaska are supplied by ADF&G and different methods

were used to project guided (charter) and private (unguided) catch estimates. The number of fi sh harvested by the guided fi shery is estimated from data reported by the ADF&G mandatory charter logbook program. The unguided (private) fi shery harvest, in number of fi sh, is estimated from the Statewide Harvest Survey. Average weight data, from creel sampling in the current year, is then used to estimate the pounds caught in both sectors. The 2012 projected sport harvest for Alaska is 5.4 million pounds, similar to the fi nal 2011 estimate of 5.5 million pounds.

To improve management of the Area 2C sport charter fi shery in 2012, the IPHC adopted a recommendation from the NPFMC for a reverse slot limit, whereby halibut could only be retained


if the total length was ≤ 45 inches (114 cm) or ≥ 68 inches (173 cm). This U45/O68 measure was recommended by the Council following analysis by ADF&G which showed that the resulting harvest would be below the 2012 GHL of 0.931 million pounds.

The Area 2C sport harvest is projected to have increased in 2012, to 1.4 million pounds from the 2011 harvest of 1.0 million pounds (Table 9). Harvests by both the charter and private sectors increased, though more so in the charter sector, likely due to the relaxation of the maximum size limit of 37 inches in 2011 to the reverse slot limit. The projected charter fi shery harvest of 0.645 million pounds was below the GHL of 0.931 million pounds (Table 4). The guided fi shery harvest was above the GHL from 2004 to 2010, often substantially. With the more restrictive length regulations of 2011 and 2012, the harvest has been under the GHL. The new fi shery management process has a learning curve and although the harvest was under the allocation in 2012 by 30%, the harvest was closer than in 2011 (56%).

In Areas 3A, 3B, and 4, the daily bag limit of two halibut for the guided and unguided fi sheries remained unchanged. The estimated harvest in Area 3A by the charter and private sectors decreased from 2011 (Tables 4 and 9). The average weight of halibut caught by both the charter and private fi shery sectors dropped slightly, although the number of fi sh caught decreased only in the charter fi shery. The harvest by the charter sector was well below the GHL of 3.103 million pounds (Table 4).

Estimates of the harvest in Areas 3B and 4 remained relatively low, at 13,000 pounds in Area 3B and 16,000 pounds in Area 4A. Halibut sport fi shing is much less common in the western Gulf of Alaska and Bering Sea due to the relative remoteness of the ports. Since 2005, annual harvests have ranged from 13,000 to 30,000 pounds in Area 3B, and 16,000 to 50,000 pounds in Area 4.

Bycatch mortality IPHC receives bycatch estimates, size, and release condition data from the observer programs

which operate off Canada and the United States. Not all fi sheries are observed therefore bycatch rates and discard mortality rates from similar fi sheries are used to calculate bycatch mortality in unobserved fi sheries.

In 2012, bycatch mortality was estimated at 9.9 million pounds (Table 12), unchanged from 2011, and the lowest since 1986 (Williams 2013c). Historically, bycatch mortality peaked at 20 million pounds in 1992 due to the growth and expansion of the Alaska groundfi sh fi sheries, declined to between 12-14 million pounds since the late 1990s, and has been below 10 million for the last two years.

A notable program implemented in 2011 by the PFMC has had a major impact on lowering halibut bycatch mortality in Area 2A. In 2011, an Individual Quota management program for the shoreside groundfi sh trawl fi shery was started, which provides for an individual bycatch quota (IBQ) system, 100% observer coverage, and real-time reporting. The program is similar to the B.C. trawl program, where there is individual vessel accountability for bycatch mortality. The fi nal 2011 estimates are available, and with the IBQ incentives the overall Area 2A halibut bycatch mortality was lowered by more than half from 0.3 to 0.1 million pounds, from 2010 to 2011.


Update on the workshop and bycatch estimatesAt the 2012 Annual Meeting the staff was directed to work on several projects regarding

learning more about and educating the public on the bycatch mortality of halibut in non-directed fi sheries. The projects included a workshop on bycatch mortality, size, growth, and harvest strategy jointly organized by the IPHC and the NPFMC, a review and update of the bycatch mortality estimates in state managed fi sheries in Alaska, and continued work by the IPHC Halibut Bycatch Project Team and Work Group.

The workshop was held in Seattle on April 24-25 and was webcast for the public. The purpose was to review the methodology and accuracy of the estimation of halibut bycatch off Alaska, the impacts of halibut bycatch on the halibut stocks coastwide and by area, and the current knowledge of halibut ecology and migration. The report of the workshop including a list of potential future research is available on the IPHC website: http://www.iphc.int/documents/2012bycatch/FINAL_Halibut_Bycatch_Workshop_Meeting_Summary_April_24-25_2012.pdf

The IPHC staff has started the review of bycatch estimates for state-managed fi sheries in Alaska and will fi nalize the estimates in 2013. The fi sheries being reviewed include the beam trawl fi shery for shrimp and fl ounders of inside waters in Southeast Alaska, hook-and-line fi sheries in Chatham and Clarence Straits in Area 2C, sablefi sh fi sheries in Prince William Sound in Area 3A, and king/tanner crab and shrimp fi sheries in the Gulf of Alaska and the Bering Sea. The original estimates for these fi sheries were derived from research surveys as these fi sheries are unobserved.

The Bycatch Project Team and Work Group met several times in 2012, primarily in conference calls. Progress has been made on a comprehensive report which reviews available information on bycatch and other removals from the resource, impacts to the resource, and treatment of bycatch mortality within IPHC’s harvest policy. These efforts are expected to be completed in 2013, with the report being available to the public at that time.

Observer coverage of the Alaska halibut fi sheryWith the start of 2013, the commercial halibut fi shery off Alaska is now subject to some

new fi shery monitoring requirements. The redesigned North Pacifi c Observer Program covers all groundfi sh fi sheries, including the halibut fi shery. Vessels which previously had less than 100% at-sea coverage requirements, including the halibut fi shery, will now be subject to either trip selection or vessel selection coverage requirements, under a deployment plan developed by NMFS. The groundfi sh and halibut vessels in these categories will be managed under an ex-vessel fee-based structure. NMFS is currently developing and testing an electronic monitoring program as a monitoring option; an implementation date is unknown. Vessels which have had coverage requirements of greater than 100% coverage will continue with the current program by contracting directly with observer providers. More information is available at the NMFS Alaska Region webpage: http://www.alaskafi sheries.noaa.gov/sustainablefi sheries/observers/

References

Gilory, H. L and Ian J. Stewart. 2013. Incidental mortality of halibut in the commercial halibut fi shery wastage). Int. Pac. Halibut Comm. Report of Assessment and Reseach Activities 2012: 53-60.


Stewart, I.J., Leaman, B.M., Martell S., and Webster, R.A. 2013. Assessment of the Pacifi c haliubt stock at the end of 2012. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2013: 93-186.

Williams, G. H. 2013a. The personal use harvest of Pacifi c halibut through 2012. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2012: 61-66.

Williams, G. H. 2013b. 2012 sport fi shery review. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2012:43-52.

Williams, G. H. 2013c. Incidental catch and mortality of Pacifi c halibut, 1962-2012. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2011:315-336.


Table 1. The 2012 preliminary estimates of total removals (thousands of pounds, net weight), 2012 catch limits and catch of Pacifi c halibut by regulatory area, and 2012 sport guideline harvest level and sport guided harvest for Areas 2C and 3A (Preliminary, November 7, 2012).

Area 2A 2B 2C 3A 3B 4 Total Commercial 574 5,811 2,568 11,649 4,954 5,511 31,067Sport 415 1,144 1,405 3,938 13 16 6,931Bycatch Mortality1: O26 103 175 6 1,259 1,109 3,685 6,337 U26 fi sh 2 14 1 681 470 2,362 3,530Personal Use2 253 405 387 266 22 434 1,148Wastage Mortality: O26 11 165 78 561 467 196 1,478 U26 fi sh 0 6 5 30 57 28 126IPHC Research 18 109 119 297 112 76 731

Total Removals 1,148 7,829 4,569 18,681 7,204 11,917 51,3482012 Catch Limits5 9896 7,0387 2,565 11,918 5,070 5,900 33,480

2012 Catch 1,0136 6,9557 2,568 11,649 4,954 5,511 32,6502012 Sport GHL 931 3,103 NA

2012 guided harvest 645 2,375 NA1 Area 2A bycatch is the 2011 estimate as the 2012 estimate will not be available until 2013.2 Includes 2011 Alaskan subsistence harvest estimates. 3 Treaty Indian ceremonial and subsistence fi sh authorized in the 2012 catch sharing plan.4 Includes 20,000 pounds of sublegal halibut retained in the 2012 Area 4DE Community Development Quota.5 Does not include poundage from the underage/overage programs in Area 2B or Alaska6 Includes commercial, sport, and treaty subsistence catch7 Includes commercial and sport catch


Table 2. The Area 2A 2012 catch limits allocated by the Pacifi c Fishery Management Council Catch Sharing Plan and preliminary catch estimates (net weight).

Area Catch Limit CatchNon-treaty directed commercial 173,216 179,000Non-treaty incidental commercial with salmon troll fi shery 30,568 35,300Non-treaty incidental commercial with sablefi sh fi shery 21,173 4,400 Treaty Indian commercial 321,650 355,100Treaty Indian ceremonial and subsistence 24,500 24,5001

Sport – Washington 214,110 213,296Sport – Oregon/California 203,783 202,087Total allocation 989,000 1,013,383IPHC research catch 18,000Total 989,000 1,031,383

1 Ceremonial and subsistence allocation amount.

Table 3. The 2012 Area 2B catch limits as allocated by the Canadian Department of Fisheries and Oceans and estimated catches (thousands of pounds, net weight).

Fishery Allocation CatchCommercial fi shery 5,953.35 5,811Sport fi shery 1,084.65 1,144Total allocation/catch 7,038.00 6,955IPHC research catch 109Total 7,038.001 7,064

1 Adjustments totaling 5,000 pounds were made to the commercial fi shery catch limit from the underage/overage plan. In addition, there was an opportunity for individual leasing of quota to the sport sector from the commercial quota share holders and 814 pounds were landed under these recreational licenses.


Table 4. Estimated harvest by the private (unguided) and charter (guided) sport halibut fi shery in millions of pounds (net weight) in Areas 2C and 3A, 2000–2012. Also shown is the Guideline Harvest Level (GHL) applicable to the guided fi shery.

Area 2C Area 3AYear Private Charter Total GHL Private Charter Total GHL2000 1.121 1.130 2.251 - 2.165 3.140 5.305 -2001 0.721 1.202 1.923 - 1.543 3.132 4.675 -2002 0.814 1.275 2.090 - 1.478 2.724 4.202 -2003 0.846 1.412 2.258 1.432 2.046 3.382 5.427 3.6502004 1.187 1.750 2.937 1.432 1.937 3.668 5.606 3.6502005 0.845 1.952 2.798 1.432 1.984 3.689 5.672 3.6502006 0.723 1.804 2.526 1.432 1.674 3.664 5.337 3.6502007 1.131 1.918 3.049 1.432 2.281 4.002 6.283 3.6502008 1.265 1.999 3.264 0.931 1.942 3.378 5.320 3.6502009 1.133 1.249 2.383 0.788 2.023 2.734 4.758 3.6502010 0.885 1.086 1.971 0.788 1.587 2.698 4.285 3.6502011 0.685 0.344 1.029 0.788 1.615 2.793 4.408 3.65020121 0.761 0.645 1.405 0.931 1.563 2.375 3.938 3.103

1 Preliminary


Table 5. Commercial fi shing periods, number of fi shing days, catch limit, commercial, research and total catch (thousands of pounds, net weight) by regulatory area for the 2012 Pacifi c halibut commercial fi shery (preliminary, as of November 7, 2012).

Area 2A Fishing PeriodCatchLimit

No. ofDays

CommercialCatch

ResearchCatch

Total Catch

Treaty Indian

Total

Unrestricted:3/24 –26

5/1Restricted:

3/17-19

321.7

48-hours13-hours

55-hours

155.5132.6

67.0355.1

155.5132.6

67.0355.1

Incidental in Salmon Fishery 5/1 – 7/3 30.6 64 days 35.3 35.3

Incidental in Sablefi sh Fishery 5/1 – 10/31 21.1 184 days 4.4 4.4

Directed1

Directed Total

6/277/11

173.2

10-hours10-hours

150.029.0

179.0 179.0

2A Total 546.6 573.8 18 591.8

Area Fishing PeriodCatch Limit

Adjusted Catch Limit2

Commercial Catch

Research Catch

Total Catch3

2B 3/17 – 11/7 5,953 5,958 5,8114 109 5,9202C 3/17 – 11/7 2,624 2,656 2,5685 119 2,6873A 3/17 – 11/7 11,918 12,033 11,649 297 11,9463B 3/17 – 11/7 5,070 5,225 4,954 112 5,0664A 3/17 – 11/7 1,567 1,627 1,538 40 1,5784B 3/17 – 11/7 1,869 1,922 1,679 23 1,7024C 3/17 – 11/7 1,107 1,136 5696 4 5734D 3/17 – 11/7 1,107 1,140 1,410 6,7 9 1,4194E 3/17 – 11/7 250 250 3157 0 315

Alaska Total 25,512 25,989 24,682 604 25,286Grand Total 32,0128 32,4948 31,066.8 731 31,797.8

1 Fishing period limits by vessel class.2 Includes adjustments from the underage/overage programs and in 2B, quota held by DFO for First Nations through relinquishment processes.3 Includes pounds from November 7, 2012 Prior Notice of Landings in Alaska and hail-ins from Fishery Operations System in Canada. 4 Includes the pounds that were landed by Native communal commercial licenses (FL licenses).5 Includes the pounds taken in the Metlakatla fi shery within the Annette Island Reserve.6 Area 4C IFQ and CDQ could be fi shed in Area 4D by NMFS and IPHC regulations.7 Area 4D CDQ could be fi shed in Area 4E by NMFS and IPHC regulations.8 Includes Area 2A catch limit.


Table 6. The fi shing period limit (pounds, net weight) by vessel class used in the 2012 directed commercial fi shery in Area 2A.

Vessel Class Fishing Period & Limits Letter Feet June 27 July 11

A 0-25 755 200B 26-30 945 200C 31-35 1,510 250D 36-40 4,165 695E 42-45 4,480 745F 46-50 5,365 895G 51-55 5,985 1,000H 56+ 9,000 1,500

Table 7. The number of vessels catching halibut since the implementation of the Area 2B IVQ fi shery and number of vessels and permit holders catching halibut since the implementation of the Alaska IFQ fi shery.

YearNumber of vessels fi shing in Area 2B

Number of vessels fi shing in Alaska

Number of quota share holders in Alaska

1991 435 no IFQ fi shery no IFQ fi shery1992 433 no IFQ fi shery no IFQ fi shery1993 355 no IFQ fi shery no IFQ fi shery1994 318 no IFQ fi shery no IFQ fi shery1995 295 2,206 4,8311996 278 2,130 NA1997 284 2,163 NA1998 287 1,802 NA1999 268 1,847 NA2000 238 1,841 3,5412001 239 1,728 3,5072002 216 1,643 3,5002003 225 1,592 3,4352004 219 1,513 3,3152005 222 1,495 3,2392006 214 1,476 3,2102007 210 1,499 3,0762008 200 1,417 2,9072009 187 1,326 2,8512010 236 1,316 2,7802011 217 1,286 2,7402012 234 1,227 2,574


Table 8. The total pounds (thousands, net weight, preliminary) of 2012 commercial landings (not including IPHC research catch) of Pacifi c halibut for Alaska and British Columbia by regulatory area and month.

Regulatory Area March April May June July Aug. Sept. Oct. Nov. Total2B1 628 838 714 568 668 1,026 514 759 96 5,811 2C 336 378 526 391 176 308 275 165 13 2,5683A2 670 1,747 1,731 1,833 1,085 1,570 1,204 1,724 85 11,6493B2 27 220 777 771 709 1,298 515 557 80 4,9544A2 2453 294 236 459 3044 1,5384B2 2643 242 283 479 281 123 7 1,679

4CDE2 72 321 740 602 422 129 8 2,294Alaska Total 1,033 2,345 3,615 3,852 3,229 4,716 2,913 2,786 193 24,682

Total 1,661 3,183 4,329 4,420 3,897 5,742 3,427 3,545 289 30,4931 Based on landings from DFO Fishery Operations System (FOS).2 Based on landings from NMFS Restricted Access Management (RAM) Division3 Weight combined with the previous month(s) for confi dentiality purposes.4 Weight combined for the months of Sept, Oct, and Nov for confi dentiality purposes.


Table 9. Harvest of halibut by sport fi shers (millions of pounds, net weight) by IPHC regulatory area, 1977-2012.

Year Area 2A Area 2B Area 2C Area 3A Area 3B Area 4 Total1977 0.013 0.008 0.072 0.196 - - 0.2891978 0.010 0.004 0.082 0.282 - - 0.3781979 0.015 0.009 0.174 0.365 - - 0.5631980 0.019 0.006 0.332 0.488 - - 0.8451981 0.019 0.012 0.318 0.751 - 0.012 1.1121982 0.050 0.033 0.489 0.716 - 0.011 1.2991983 0.063 0.052 0.553 0.945 - 0.003 1.6161984 0.118 0.062 0.621 1.026 - 0.013 1.8401985 0.193 0.262 0.682 1.210 - 0.008 2.3551986 0.333 0.186 0.730 1.908 - 0.020 3.1771987 0.446 0.264 0.780 1.989 - 0.030 3.5091988 0.249 0.252 1.076 3.264 - 0.036 4.8771989 0.327 0.318 1.559 3.005 - 0.024 5.2331990 0.197 0.381 1.330 3.638 - 0.040 5.5861991 0.158 0.292 1.654 4.264 0.014 0.127 6.5091992 0.250 0.290 1.668 3.899 0.029 0.043 6.1791993 0.246 0.328 1.811 5.265 0.018 0.057 7.7251994 0.186 0.328 2.001 4.487 0.021 0.042 7.0651995 0.236 0.887 1.751 4.511 0.022 0.055 7.4621996 0.229 0.887 2.129 4.740 0.021 0.077 8.0831997 0.355 0.887 2.172 5.514 0.028 0.069 9.0251998 0.383 0.887 2.501 4.702 0.017 0.096 8.5861999 0.338 0.859 1.843 4.228 0.017 0.094 7.3792000 0.344 1.021 2.251 5.305 0.015 0.073 9.0092001 0.446 1.015 1.923 4.675 0.016 0.029 8.1042002 0.399 1.260 2.090 4.202 0.013 0.048 8.0122003 0.404 1.218 2.258 5.427 0.009 0.031 9.3472004 0.476 1.613 2.937 5.606 0.007 0.053 10.6922005 0.477 1.841 2.798 5.672 0.014 0.050 10.8522006 0.511 1.773 2.526 5.337 0.014 0.046 10.2072007 0.504 1.556 3.049 6.283 0.025 0.044 11.4612008 0.487 1.536 3.264 5.320 0.026 0.040 10.6732009 0.490 1.098 2.383 4.758 0.030 0.024 8.7832010 0.393 1.156 1.971 4.285 0.024 0.016 7.8452011 0.401 1.220 1.029 4.408 0.014 0.017 7.08920121 0.415 1.144 1.405 3.938 0.013 0.016 6.931

2011-2012 changePounds 0.014 -0.076 0.376 -0.470 -0.001 -0.001 -0.158Percent 3.5 -6.2 36.5 -10.7 -7.10 -5.9 -2.2

1 Preliminary


Table 10. 2012 Area 2A sport harvest allocations and preliminary harvest estimates (pounds, net weight) by subarea.

Harvest Pounds PercentSubarea Allocation Estimate Over/(Under) TakenWA Inside Waters1 57,393 57,393 0 100.0WA North Coast 108,030 105,478 (2,552) 97.6WA South Coast 42,739 42,467 (272) 99.4Columbia River 11,895 7,958 (3,937) 66.9OR Central Coast 191,780 196,031 4,251 102.2South OR/California1 6,056 6,056 0 100.0Total 417,893 415,383 (2,510) 99.41A harvest estimate is not yet available so the allocation is shown as the preliminary harvest.


Table 11. Summary of the 2012 Pacifi c halibut sport fi shery seasons. No size limits were in effect unless otherwise noted.

Regulatory Area & Region Fishing DatesFishing Days

per week

No. of Fishing

Days

Daily Bag

LimitArea 2A - Washington, Oregon & California

WA Inside Waters East of Low Point May 3 – 19 3 (Thur – Sat) 9 1

May 24 – 27May 28, 31

Jun 1-2

4 (Thur – Sun)2 (Mon, Thur)

2 (Fri, Sat)

422

111

Low Point to Sekiu River May 24 – 27 4 (Thur – Sun) 4 1May 28, 31 2 (Mon, Thur) 2 1

Jun 1-2 2 (Fri, Sat) 2 1Jun 7 – Jun 23 3 (Thur – Sat) 9 1

WA North Coast (Sekiu Rvr to Queets Rvr) May 10 – 19 2 (Thur, Sat) 4 1May 31, Jun 2

Jun 142 (Thur, Sat)

1 (Thur)21

11

WA South Coast (Queets Rvr to Leadbetter Pt.) All depths May 6 – 20 2 (Sun, Tues) 5 1 Northern nearshore May 7 – Jun 8 7 (Mon – Sun) 33 1Columbia River (Leadbetter Pt. to Cape Falcon) May 3 – Jul 14 3 (Thur – Sat) 33 1

Aug 3 – Sep 29 3 (Fri – Sun) 27 1OR Central Coast (Cape Falcon - Humbug Mtn.) All depths May 10 – Jun 30 3 (Thur – Sat)1 17 1

Aug 3 – 18 2 (Fri – Sat)2 4 1 Less than 40 fathoms May 1 – Jul 22 7 (Sun – Sat) 83 1

Sep 24 – Oct 31 7 (Sun – Sat) 38 1OR/CA (South of Humbug Mtn.) May 1 – Oct 31 7 (Sun – Sat) 184 1

Area 2B - British Columbia Mar 1 – Sep 9 7 (Sun – Sat) 192 1Area 2C - Alaska

Guided anglers Feb 1 – Dec 31 7 (Sun – Sat) 334 13

Unguided anglers Feb 1 – Dec 31 7 (Sun – Sat) 334 2Areas 3 and 4 - Alaska Feb 1 – Dec 31 7 (Sun – Sat) 334 2

1 Fishing was prohibited during June 7-9 and June 21-23.2 Fishing was prohibited during August 10-11.3 A reserve slot limit defi ning retained halibut as ≤45 inches or ≥68 inches in total length was in effect in 2012.


Table 12. Estimates (thousands of pounds, net weight) of bycatch mortality of halibut by year and area, 2003 - 2012. Area 2003 2004 2005 2006 2007 2008 2009 2010 2011 20121

Area 2A 541 351 504 548 322 383 509 346 106 1062

Area 2B 244 251 346 294 320 143 213 181 232 189Area 2C 341 362 340 341 342 346 344 343 342 7Area 3A 3,180 3,671 3,220 2,975 2,843 3,066 2,722 2,532 2,726 1,940Area 3B 1,734 1,274 1,126 1,400 1,115 1,353 1,294 1,147 1,165 1,579Area 4 6,822 6,735 7,692 7,491 7,262 6,555 6,297 6,082 5,382 6,048TOTAL 12,862 12,644 13,228 13,049 12,204 11,846 11,378 10,631 9,953 9,869

1 Preliminary (Alaska state-managed fi sheries not included).2 2011 estimate carried forward.

Figure 1. IPHC regulatory areas for the 2012 fi shery.


IPHC Management Strategy Evaluation framework

Steven Martell

Overarching objective

To develop a formal process in which to evaluate the performance of alternative management procedures for the Pacifi c halibut stock against a range of scenarios that encompass observation and process uncertainty in stock ass essments, alternative hypotheses about stock dynamics and structural assumptions. This process is commonly referred to as Management Strategy Evaluation (MSE) in fi sheries science.

Management Strategy Advisory Body

The MSE process will be overseen by a Management Strategy Advisory Body (MSAB) that is comprised of harvesters (commercial, sport, and subsistence), fi sheries managers (DFO and NMFS), processors, IPHC staff, commissioners and academia. The advisory body will be broadly based, both geographically and by harvesting sector. The advisory body will be nominated from existing Commission advisory bodies, nominations from partner agencies, and direct application from the public. The MSAB will consist of approximately 15-20 individuals contemplated by the Commission and a key consideration for members is that they be available to participate in the process from initial development of fi sheries objectives, defi ning performance measures and iteratively refi ning management procedures. Continuity on the MSAB is key to the success of this process.

Role of the MSABThe MSAB will work interactively with analysts on the Commission staff and Research

Advisory Board to initially defi ne clear, measurable objectives for this fi shery, defi ne candidate management procedures (MP) for testing within the MSE framework, and defi ne the performance measures to evaluate alternative MPs. A management procedure constitutes the entire decision- making process starting with what data to be used in stock assessment, a stock assessment method to interpret the data, a harvest control rule in which to compute yield options and a projection model in which to evaluate impacts of alternative yield options on the stock. A series of quantitative measures must be defi ned in which to evaluate how well each MP performs relative to perfect information. The central role of the MSAB is to: defi ne fi shery objectives, articulate management procedures, and defi ne performance measures in which to evaluate MPs.

MSAB membershipThe following table identifi es the broad membership of the MSAB by group. The science

advisors group would provide input on the technical aspects of developing the operating model(s) to be used in the MSE framework. The processors group would provide inputs on fi sheries objectives and the development of performance measures for evaluating candidate management procedures. The harvesters group would also provide inputs on fi sheries objectives, the development of management procedures and performance measures. Fisheries managers are key for defi ning the overall fi sheries objectives, as well as, providing input on designing candidate management


procedures and ensuring that the performance measures encompass fi sheries legislation in respective countries. Membership should also include at least one Commissioner from each of the respective member countries to ensure the objectives and performance measures encompass commercial, sport, and aboriginal sectors. The IPHC staff will work in collaboration with all members of the MSAB and implement the technical details of the MSE framework; Specifi cally develop and modify existing software for testing alternative MPs in the MSE framework.

Group Number Potential CandidatesScience Advisors 2-4Processors 2+Harvesters 5Managers 2-3Commissioners 2IPHC Staff 4+ Martell, Stewart, Webster, Leaman, ...

Members of the MSAB should be able to make a long-term commitment in order to ensure that their respective ideas are fully vetted in the MSE process. Other criteria that the Commission will take into consideration in selecting advisors is the breadth of experience in the halibut fi shery, experience and expertise in the appropriate group, and the ability to provide objective input into the process and disseminate information. It is also important to ensure a balanced geographic representation from each of the Regulatory Areas.

Timeline and project milestonesThe MSE framework was introduced at the Commission’s interim meeting in November 2012,

and the process of forming the MSAB will be initiated at the Annual meeting in January 2013. For the initial formation of the MSAB, the IPHC would like to solicit nominations from existing Commission advisory bodies (RAB, CB, PAG) and direct application from the public. Following, the list of nominations will be categorized (processors, harvesters, managers, etc.) and ranked by the IPHC staff, and then submitted for fi nal selection by the IPHC Commissioners.

A preliminary meeting for the MSAB will be held at the earliest possible date after the formation of the MSAB. The purpose of the preliminary meeting is to give members of the MSAB a broad overview of the objectives of MSE. We also hope to include presentations from one or more MSE-based projects that have had practical application (e.g., the Canadian Sablefi sh Association) with MSE as a general introduction. This initial meeting will also serve the purpose of defi ning objectives for the halibut fi shery, scoping out performance measures in which to compare alternative harvest policies, and fl ush out key operating model components that will be required to address alternative MPs. Supplemental follow-up meetings via teleconference or webinar will be conducted as required to present preliminary results and reiterate ideas on appropriate performance metrics and fi sheries objectives.

The development of an operating model is currently underway by IPHC staff, and this work will evolve continuously with the development and revisions of the MSE framework. Input from the MSAB, as well as the available historical data, will help shape the structure of the reference and observation models to be used in the MSE efforts. In addition to the current coast-wide assessment model, alternative assessment models will also undergo simulation testing using the


MSE framework. The reference and observation model platforms provide “known” state variables in which to evaluate alternative assessment models, or changes to the current assessment model. Prior to the second MSAB meeting, it is anticipated that the “alpha” version of the MSE software will have the capability of exploring alternative estimators (or structural assumptions), alternative harvest control rules, and establish base-line metrics based on “perfect information” over a range of alternative hypotheses about stock structure.

The second annual meeting of the MSAB will receive presentations on progress of the development of the MSE software and some preliminary results with respect to initial ideas from the fi rst annual MSAB meeting (and interim meetings if necessary). This will be the fi rst real opportunity for the MSAB to provide feedback on the MSE software and help refi ne and devise alternative management procedures for testing in the next iteration.

We anticipate that many of the issues surrounding the assessment, monitoring and management of the halibut resource will be of signifi cant interest to members of the Research Advisory Board (RAB) the Conference Board (CB) and the Processors Advisory Group (PAG). Given the overlap of interests, MSAB annual meetings are likely to immediately follow the annual RAB meeting, just prior to the interim meeting. Additional presentations will also be given at the Interim Meeting, the Annual Meeting in January each year, and summarized in the Report of Assessment and Research Activities each year.

The following bullet points identify the major milestones and summarizes the work to be performed in more-or-less chronological order.• Assemble MSAB

• Send out a call for nominations for the MSAB (Public, RAB, CB, and PAG)• Commissioners approve choices for the MSAB

• MSAB (I) preliminary meeting• overview of project & presentation from MSE experts (Cox/Kronlund/A’Mar)• defi ne tasks for MSAB• develop fi sheries objectives• initial scoping of performance measures• operating model skeleton

• Interim MSAB (webinar)• revise performance metrics and fi sheries objectives

• Develop MSE software (alpha version)• Operating model:

• reference model (true state of nature)• observation models (simulating data programs)

• estimator(s) (stock assessment models)• harvest control rules• projection scenarios (alternative hypotheses)• establish metrics and a status quo management procedure

• MSAB (II)• presentation of alpha model & preliminary results• refi ne, add, drop additional MP and performance metrics• feedback from MSAB, re-iterate and revise as necessary

• MSE presentation at annual meeting and in RARA (2014-2016)• outreach• continue to work with MSAB to refi ne and test current and alternative MP


Figu

re 1

. An

illus

trat

ion

of th

e fou

r maj

or co

mpo

nent

pro

cess

es u

sed

in M

anag

emen

t Str

ateg

y Ev

alua

tion.

Fir

st, fi

sher

ies o

bjec

tives

m

ust b

e defi n

ed th

at u

ltim

atel

y defi n

e pe

rfor

man

ce m

easu

res.

A su

ite o

f alte

rnat

ive

man

agem

ent p

roce

dure

s are

dev

elop

ed a

nd

used

to fi

ll ea

ch r

ow o

f the

per

form

ance

mea

sure

s ta

ble.

The

eva

luat

ion

of e

ach

man

agem

ent p

roce

dure

is d

one

over

a r

ange

of

sce

nari

os th

at s

pan

the

rang

e of

unc

erta

inty

in o

ur c

urre

nt u

nder

stan

ding

of h

alib

ut d

ynam

ics.

Las

tly, c

omm

unic

atio

n w

ith

stak

ehol

ders

is a

key

step

in refi n

ing fi s

heri

es o

bjec

tives

and

des

igni

ng n

ew m

anag

emen

t pro

cedu

res t

hat a

re ro

bust

to u

ncer

tain

ty.

Fish

ery

Ob

ject

ives

man

ager

s,st

akeh

olde

rs

Man

agem

ent

pro

ced

ures

data

, as

sess

men

t, de

cisi

on r

ules

Man

agem

ent

pro

ced

ures

data

, as

sess

men

t, de

cisi

on r

ules

MMMMMMaaann

agem

ent

pro

ced

ures

data

,as

sess

men

t,

gp

data

Man

agem

ent

pro

ced

ures

MMaan

agem

ent

pro

ced

ures

Man

agem

ent

pro

ced

ure

Man

agem

ent

pro

ced

ure

dt

dt data

,at

a,at

a,ss

tss

essm

ent,

smen

tat

a,sm

esm

esm

ent

Man

agem

ent

pro

ced

ures

data

, as

sess

men

t, de

cisi

on r

ules

Eva

luat

ion

scen

ario

s,si

mul

atio

ns,

perfo

rman

ce m

easu

res

Co

mm

unic

atio

nun

cert

aint

y,tr

ade-

offs

,ris

kP

erfo

rman

ce m

easu

res

MP

Catch

AAV

Depletion

Bycatch

Profit

as decas

sess

men

t,ss

essm

ent,,

ec deci

sion

rule

sece

smen

tnnt

,ci

sion

rule

sci

sion

rule

sss

essm

ent,,

cisi

onle

si

il

cisi

onle

sec

isio

nle

sci

sion

rule

ssi

onru

leci

sion

rule

ssi

onru

lede

cisi

onru

les

ecis

ion

rule

sec

isio

nru

le

Eva

luat

ion

scen

ario

s,si

mul

atio

ns,

perfo

rman

cem

easu

res

trad

e-of

fs,

risk

Per

form

ance

mea

sure

s

MP

Catch

AAV

Depletion

Bycatch

Profit


Assess ment of the Pacifi c halibut stock at the end of 2012

Ian J. Stewart, Bruce M. Leaman, Steven Martell, and Raymond A. Webster

Abstract

This stock assessment reports the status of the Pacifi c halibut resource in the northeastern Pacifi c, including the territorial waters of the United States and Canada. Annual removals were above the 100 year average of 64 million pounds from 1985 through 2010, dropping to 60 million pounds in 2011, and then 51 million pounds in 2012, in response to management measures intended to stabilize declining trends in the IPHC setline survey and stock assessment estimates. All data sources were updated through 2012. The 2012 setline survey showed an increase in WPUE (+12% relative to 2011), the fi rst since the current geographically comprehensive survey began in 1997. Age distributions in 2012 from both the survey and fi shery were very similar to those observed in 2011, indicating a relatively stable stock, but not showing any evidence of strong recruitments in recent years. Individual size-at-age continues to be low relative to the rest of the time-series, although there were signs of fl attening in the declining trend for some ages for both males and females.

For 2012, there was a full review of the data, specifi c model equations and general approach used to assess the stock in recent years. This effort consisted of three parts: 1) investigate and address the cause of the retrospective pattern observed in recent assessments, 2) improve the way uncertainty is propagated through data processing, model estimation and into the results used for management, and 3) identify additional work needed to create a more stable and easily reviewed stock assessment for the future. This work culminated in a successful Scientifi c Review Meeting, 24-26 October, 2012. Allowing for time-varying availability in the assessment model removed the retrospective bias in recent status estimates and is consistent with observed geographic and demographic trends. This change to the assessment model resulted in a much more pronounced decline in the estimated stock trend in recent years, a large reduction in the scale of current population estimates, and also a decrease in the estimated average level of productivity.

The 2012 assessment indicated that the Pacifi c halibut stock has been declining continuously over much of the last decade as a result of decreasing size-at-age, as well as poor recruitment strengths. The population decline is estimated to have slowed and the stock trajectory is now relatively fl at at 35% of the reference level, just above the harvest policy threshold (30%). Despite reductions in harvest levels in 2011 and 2012, the assessment estimates that, in retrospect, harvest rates have been well above the coastwide targets implied by the current harvest policy. The 2013 estimate of exploitable biomass is 186.49 million pounds, signifi cantly smaller than the 2011 estimate of 260 million pounds. A bridge with the 2011 stock assessment results is provided, along with revised results from that model using the fi nal data sets available through 2011 and through 2012 for direct comparison to the 2012 assessment results. Forecast projections were conducted for a range of alternative management actions, and probabilities of various risk metrics are reported in a decision-making table framework. The application of the current harvest policy, consistent with the approach used in recent stock assessments, results in the Blue line of the decision-making table indicating a coastwide TCEY of 36.63 million pounds and FCEY of 22.7 million pounds.


Introduction

Stock, management, and removalsThis stock assessment reports the status of the Pacifi c halibut resource in the northeastern

Pacifi c Ocean, including the territorial waters of the United States and Canada. As in recent assessments (Hare 2011), the resource is modeled as a single stock extending from northern California to the Aleutian Islands and Bering Sea, including all inside waters, the Strait of Georgia and Puget Sound. Potential connectivity with the western Pacifi c resource is considered slight and is not explicitly accounted for.

The halibut fi shery has been managed for nearly 100 years, and much is known about the history of fi shery removals, population trends and life-history characteristics. Total annual halibut removals (including all sources of mortality: target fi shery landings and discards, bycatch in non-target fi sheries, research, sport and personal use) have ranged from 34 to 100 million pounds over the last 100 years (Fig. 1; all weights in this document are reported as ‘net’ weights, head and guts removed; this is approximately 75% of the round weight). The average annual removal over this period has been 64 million pounds. Annual removals were above the 100 year average from 1985 through 2010, dropping to 60 million pounds in 2011 and then 51 million pounds in 2012 (Table 1).

After a peak in 2004, annual removals have decreased each subsequent year in response to management measures intended to stabilize declining trends in IPHC setline survey WPUE (Weight-Per-Unit-Effort, net lb/skate) and stock assessment estimates. These reductions have been proportionally greatest for the directed longline fi shery (Fig. 2), although all components have seen reductions during this period. Management of halibut includes a complex set of regulations that vary among fi shery sectors, regulatory areas and management bodies. The management of annual removals begins with the Total Constant Exploitation Yield (TCEY), which is defi ned as the target level of harvest of halibut exceeding 26 in. (66 cm) in fork length (O26). This value is obtained by applying regulatory area-specifi c harvest rates (based on current harvest policy) to the coastwide estimate of Exploitable biomass (frequently referred to as Ebio in previous assessments) after it has been apportioned among areas. Apportionment calculations consist of adjusting raw survey WPUE in each regulatory area to account for the fraction of the annual fi shing mortality that occurred prior to the survey, competition for baits as measured by the proportion of hooks returning with baits, and the geographic extent of the area. These calculations and results are described in Webster and Stewart (2013). Over the last decade, management has followed this prescription quite closely, with the TCEY tracking the declines in each subsequent estimate of exploitable biomass (from the annual stock assessment results), available at the time of each decision (Fig. 3 ). Both the setline survey and commercial fi shery logbook WPUE trends have shown a very similar trend when compared with assessment results (Fig. 3). The Fishery Constant Exploitation Yield (FCEY) is calculated by subtracting from the TCEY the removals associated with directed fi shery wastage (sublegal fi sh discarded that are then assumed to die due to hooking mortality as well as mortality of all sizes from lost fi shing gear), bycatch, sport and personal use fi sheries (with the exceptions of Area 2A, where personal use and sport removals are included in the FCEY, and Area 2B where sport removals are included in the FCEY). Since 2004, reductions in the TCEY have translated into reductions in the total removals from all sources, as well as the FCEY. During this period, there has been an increase in the proportional contribution of other removals (Fig. 4).


Data sources

The data included in this assessment are comprised of fi shery-dependent sources, fi shery-independent sources, and auxiliary biological information (Table 2). These data represent the survey trends, biological characteristics associated with these trends (proportions-at-age by sex, length-at-age, and weight-at-age), removals from the stock from each source, as well as the biological characteristics of those removals. All raw observations undergo various processing steps to account for sampling methods, and then a process of aggregation to the coastwide level. Data are fi rst summarized by regulatory area, and then weighted appropriately, such that inputs to the assessment represent total coastwide mortality, or the geographically-weighted average values for indices of abundance and biological summaries (length-, weight-, and age-composition data). Further, where different methods have been used in different years (e.g., ageing methods before and after 2002) these changes are also accounted for. The primary sources of information for this assessment and the methods used to process them remain unchanged from recent stock assessments, although guidance on potential improvements for the 2013 stock assessment is outlined in the report of the recent Scientifi c Review Meeting (Stewart et al. 2012).

Data through 2011 were updated to include additional observations either not available for the 2011 stock assessment (e.g., fi nalized 2011 catches, additional logbook information), or improved estimates over the time-series for all sources. All data sources were fi nalized on 7 November, 2012 in order to provide adequate time for analysis and modeling prior to the IPHC’s interim meeting.

Fishery-independent dataThe annual IPHC setline survey provides the primary source of fi shery-independent data

available for the stock assessment. Trend information included in the assessment model includes both WPUE and Numbers-Per-Unit-Effort (NPUE) for halibut over 32 in. (81.3 cm) in fork length (O32). Trends in these indices of abundance had been steadily declining through 2011 since 1997 when the current geographically comprehensive survey design was initiated (Table 3, Fig. 5). However, the 2012 setline survey indices of WPUE and NPUE were 12% and 9% higher than those from 2011. Despite the increase in the coastwide aggregate, this trend was not ubiquitous across all regulatory areas (Figs. 5 and 6). The largest WPUE increase was observed in area 2B (30%), with area 4B showing a large decline (29%).

Coastwide age distributions observed by the survey in 2012 were very similar to those observed in 2011, indicating a relatively stable stock, but not showing any clear evidence of strong recent recruitments (Fig. 8). The majority of the halibut encountered by the survey has consistently remained in the 7-15 year-old range for the recent decade (Fig. 9). Prior to that time, stronger recruitment around the 1987 cohort was quite dominant in the proportions-at-age. The average age observed in the survey increased until 2002 (following the aging of the 1987 cohort) and has since declined to 12.2 years in 2012 (Fig. 10). This is slightly older than the age at 50% maturity. Average individual weight has declined signifi cantly from 21.2 lb in 1997 to 13.7 lb in 2012. This refl ects not only the decline in the average age, but also a strong decline in both length- and weight-at-age (estimated from the weight-length relationship) over all ages observed and for both female and male halibut (Figs. 11 and 12). The apparent ‘kink’ in these size-at-age patterns is likely due to a change in ageing methods (described below), rather than an actual biological phenomenon. Some fl attening of recent trends is evident in the 2012 length- and weight-at-age data, particularly


for female halibut between ages 10-20, which are those ages making the greatest contribution to the spawning biomass.

NOAA trawl surveys provide direct calibration information used to extrapolate the setline survey estimates in the Bering Sea, and also indirect comparisons of stock trend for the assessment results. The trawl surveys capture a relatively high proportion of age-2 to age-5 halibut, much younger than frequently observed in the commercial fi shery or setline survey, and therefore also potentially provide an index for recruitment.

The NOAA Bering Sea trawl survey conducted in 2012 observed a slight (1%) decrease in total biomass (estimated via density multiplied by area-swept), and a modest increase (17%) in the biomass thought to be available to the setline survey (Fig. 13). These differing trends appear to be caused by a decline in the number of small halibut, but an increase in the number of fi sh available to the survey (Fig. 14). No clear evidence for strong incoming recruitments can be seen in the annual size distributions from the Bering Sea, since the relatively strong 2005 cohort was fi rst observed (Fig. 15).

Although there was a NOAA Aleutian Islands trawl survey conducted in 2012, there was insuffi cient time to process and analyze these data for this assessment document. Using data through 2012, different trends have been observed in Area 4B and 4A. Specifi cally, both total and exploitable numbers have declined in 4B, while total numbers have increased in 4A (Fig. 16 Since the exploitable numbers have also declined in 4A, there have apparently been some increases in smaller fi sh in that area. There was no NOAA trawl survey conducted in the Gulf of Alaska during 2012. Due to the spacing in the timing of the surveys, it is diffi cult to draw strong conclusions; however, the numbers of exploitable fi sh appear to have declined over the last decade (Fig. 17) consistent with setline survey and fi shery observations.

Fishery-dependent dataUp-to-date estimates of commercial fi shery landings from the stock are compiled from a

variety of sources (Table 1). These commercial landings are infl ated to account for lost gear, as well as discarded sublegal halibut and also include annual research catches. Because there are no direct observations of discarding for most of the directed fi shery, a proxy method is required to infer the number and size of discarded fi sh. For all recent assessments, these quantities have been estimated via the ratio of legal to sublegal halibut captured by the setline survey for each year and area combination. Only the survey stations with the highest 33% of all catch-rates are included in this analysis, as the catch rates observed for these stations have been found to track the observed commercial rates quite well (Gilroy and Stewart, 2013). This approach makes the implicit assumption that temporal and spatial difference between survey and fi shery catches do not result in signifi cant demographic differences in availability and should be a point of future investigation.

Logbooks collected from the commercial fi shery generate indices of both WPUE and NPUE. These indices indicate very similar trends to those observed in the setline survey (Table 3). Many of the general patterns observed in the logbooks are also similar to those from the setline survey, particularly the observed recent increasing trends throughout Area 2. However, unlike the survey WPUE, the coastwide commercial fi shery WPUE was almost unchanged from 2011 to 2012, and there were somewhat more pronounced declines in Area 3 (Figs. 18 and 19).

The length- and age-frequency distributions of commercially landed halibut are sampled by IPHC port samplers. Because these fi sh have been gutted at sea the sex cannot be determined at


the time of sampling. Sex-ratios observed in the setline survey generally show a tight relationship with size within a given age, due to the pronounced sexually dimorphic growth pattern of females attaining much larger sizes than males. Because of this consistency, the relationship between sex-ratio and size by age has historically been estimated from the survey and then applied to the fi shery biological samples in order to infer the ages and lengths-at-age by sex. Although representing a very reasonable approach, this processing step has implications for calculation of uncertainty and was recommended for revisiting in the future by the Scientifi c Review Meeting (Stewart et al. 2013).

Age distributions for most years observed in the commercial fi shery are very similar to those observed in the setline survey, but generally show fewer fi sh less than age-10 (due to a high proportion of these fi sh being sublegal). This was again the case in 2012 (Fig. 9). Also discernible in the commercial age-frequency distribution for 2012 are relatively fewer males (again, a greater proportion are sublegal) as well as slightly more of the oldest fi sh at age-25 or greater. These old fi sh were observed primarily in Areas 4A and 4B, with fi sh greater than age-20 almost entirely absent from Area 2 (Fig. 20). As in the survey data, recent age distributions have been relatively stable, with most of the commercial catch ranging from 8-15 years old (Fig. 21). The coastwide average weight and age of commercially caught fi sh are consistently greater than those observed in the survey (Fig. 10), but area-specifi c patterns are very pronounced. The average commercial landed fi sh weight has declined over the recent time-series; however, this trend has not been consistent across all regulatory areas. Specifi cally, average weight in Areas 2B, 2C and 4A has been increasing for the last several years, while in Areas 3, 4B and 4D there have been very strong and consistent declines (Fig. 22). The heaviest average fi sh observed in 2012 were found in Area 2C. Declines in length- and weight-at-age for both females and males appear to have been more pronounced in commercially caught fi sh than was observed in the survey data (Figs. 23 and 24). Note that for some ages there are very few observations and therefore historically fi xed values are assigned; however these values are applied to so few fi sh they are virtually irrelevant in subsequent calculations. In general, there also seems to have been less fl attening of the declines in size-at-age in the commercial data between 2011 and 2012 than in the survey observations.

Auxiliary informationThere is a variety of auxiliary information that is analyzed external to the assessment model

and contributes to the analysis either as fi xed parameter values or as structural assumptions built into the model framework. This includes both biological relationships, as well as information about the methods used to collect the raw data.

Although the stock assessment compares predictions of age frequency to the sampled proportions-at-age, age itself is observed imprecisely and sometimes with some degree of bias. Current ageing is conducted using the Break-and-Bake (BB) method, which has been shown to be unbiased (Piner and Wischnioski 2004). Even unbiased ageing methods are still subject to observation error (imprecision), and this has been thoroughly quantifi ed for the BB method (Clark 2004, Clark and Hare 2006a). The degree of BB ageing imprecision previously estimated and used in all recent stock assessments has not been altered for this analysis. Prior to 2002, ages were read using surface reading, a method that is known to be biased for older fi sh across many species. Further, the method is much less precise than BB ageing, resulting in lower quality information available to the stock assessment. Although the bias and imprecision have both been quantifi ed for surface-read halibut ages (Clark and Hare 2006a), there has been some concern regarding the


quality of these estimates (J. Valero, unpublished analyses) and an effort to re-analyze historical ages is currently underway at the IPHC. Observations-at-age prior to 2002 are corrected to account for the perceived degree of bias and imprecision; however discontinuities are still discernible in the time-series’ of length- and weight-at-age for both the survey and fi shery (Figs. 11, 12, 23 and 24). This issue was identifi ed during the Scientifi c Review Meeting (Stewart et al. 2013), and will be the subject of further analysis as re-ageing data become available.

The maturity schedule for Pacifi c halibut has also been investigated historically and the relative maturity-at-age found to be very stable despite long-term changes in size-at-age (Clark and Hare 2006). Estimates of the age at which 50% of female halibut are sexually mature average 11.6 years, with very few fi sh mature at ages less than fi ve and nearly all fi sh mature by about age-17 (Fig. 25).

Natural morality is a notoriously diffi cult quantity to collect to information on for any fi sh species, and halibut is no exception to this. Tagging studies can, when correctly accounting for tag loss, tagging morality, reporting rates, and fi shing intensity, provide some information on the magnitude of natural mortality. Estimates produced as priors for the IPHC’s PIT tagging analyses included a range of values from roughly 0.1 to 0.2. Although there was little statistical difference between an estimated value vs. a fi xed value of 0.15, the point estimate produced from the PIT tag analysis was 0.124 (Webster 2010). The asymmetric implications of over- vs. underestimating the true value for natural mortality in a stock assessment (Clark 1999) and a reevaluation of life-history information led to the adoption of the current value of 0.15/year for female halibut in the 1998 stock assessment (Clark and Parma 1998), revised downward from the 1997 assessment, which used a value of 0.2/year (Sullivan and Parma 1997). Uncertainty in this value is discussed in more detail below, and directly included in the results of this assessment via the decision-making table.

Assessment

The evolution of the stock assessment for Pacifi c halibut has closely tracked that of fi sheries science in general (Clark 2003), moving from simple equilibrium-based models to current age-structured approaches (Table 4). Key transitions in this evolution relevant to the changes made for 2012 occurred with the change in natural mortality in 1998 and with the shift to a coastwide stock assessment in 2006. Both of these changes were logical and substantially improved the performance of the models at the time.

Changes to the 2011 assessmentIn 2012, the model input data, specifi c model equations, and the general approach used to

assess the halibut stock in recent years was fully reviewed by IPHC staff, and an external review panel. The primary focus of this effort consisted of three parts: 1) investigate and address the cause of the retrospective pattern observed in recent assessments, 2) improve the way uncertainty is propagated through data processing, model estimation and into the results used for management, and 3) identify additional work needed to create a more stable and easily reviewed stock assessment for in the future. This work culminated in a successful Scientifi c Review Meeting (24-26 October, 2012), from which a detailed summary report is available (Stewart et al. 2013).

The most pressing issue to resolve for 2012 was the pronounced retrospective pattern (Fig. 26) observed among recent Pacifi c halibut stock assessments (Clark and Hare 2006b, 2007, Hare and Clark 2008, Hare 2009, 2010, 2011). This retrospective pattern resulted in each stock


assessment estimating a lower absolute stock level than the previous assessment, which can have strong potential implications for harvest policy (Valero 2011). However, it is diffi cult to correct adequately for such a bias when the cause is unknown.

The retrospective pattern was clearly evident within the 2011 model (the wobblesq confi guration, on which most of the 2011 results were focused) when evaluated by sequentially removing the terminal year of data and re-estimating the time-series of spawning biomass (Fig. 27). As was documented in the 2011 assessment, the retrospective bias in stock size was a direct result of transient overestimation of incoming year-class strengths during the period for which they were relatively poorly informed by the data but contributing signifi cantly to the spawning biomass (i.e., the 1998-2000 cohorts in 2011; Fig. 28).

In order to resolve the retrospective pattern, a detailed investigation of the stock assessment model code and structural assumptions was performed during August and September, 2012. No signifi cant coding errors or inconsistencies in data preparation that appeared to be contributing to the retrospective bias were discovered. Treatment of bycatch mortality (selectivity and magnitude), commercial and survey catchability (identifi ed as a potential factor during the 2011 process; Valero, unpublished analyses), the translation of length- to age-based selectivity, smoothing of recruitment estimates as well as many other potential mechanisms were tested, but none showed any strong correlation with retrospective performance. The most informative tests conducted consisted of: 1) directly penalizing large recruitments, 2) substantially increasing the relative weight placed on the survey trend during model fi tting, and 3) evaluating the potential for time-varying availability (Fig. 29). The fi rst of these tests merely provided a means to determine how the fi t to all data sources in the model changed as ‘brute force’ was applied to directly remove the retrospective pattern, without any understanding of its underlying cause. This analysis indicated that the age data were clearly linked to the retrospective behavior, as the fi t to these data was consistently degraded as the retrospective bias was removed. The second test provided insight into the relative weighting of the various data sets, particularly the consistency of the survey trend with retrospective patterns. It was discovered that the retrospective bias was removed if the survey WPUE trend was substantially up-weighted (thereby decreasing the relative weight on the age data). Although informative, neither of these tests provided an explanation for the retrospective patterns, only a highlighting of which data (the age information) were most closely implicated.

Availability (also called ‘selectivity’) provides the link between the underlying estimated population age-structure and the observed age data in a stock assessment model. When modeled at a small spatial scale the dominant component of this process is represented by vulnerability: which demographic components (i.e., small vs. large fi sh, old vs. young fi sh) are most likely to be captured when the gear is deployed. At a coastwide scale, availability includes not only the capture effi ciency of the fi shing gear, but also the interaction between the spatial distribution of the stock and the differences in population characteristics (i.e., age, length, weight- or length-at-age ) among areas. Historical closed area assessment models had maintained a rigid assumption that availability could not vary over time, and this assumption had been carried forward to the current coastwide model, despite the difference in effective application of the relationship at over a much broader spatial scale. Further, the large amount of weight carried by the age-composition data in the assessment model relative to the survey index of abundance was both found to have contributed the observed retrospective bias and the diffi culty in identifying it. The age data were largely responsible for stock estimates. As has been the case twice before in the history of the halibut stock assessment (1994 and 2002), a change in the parameterization of availability was


required to improve model performance, and remove the retrospective bias. Several different approaches to implementation of time-varying availability were investigated, and all produced results that much more closely matched the observed time-series of survey catch rates than did the 2011 model.

The approach that was selected utilized the same smoother implemented to create continuity between availability of adjacent size bins (Clark and Hare 2006; page 29). This is merely a second differencing equation with a standard error input via the data fi le. In recent assessments, the standard deviation for the smoothing function over size has been set to 0.05. Values ranging from 0.001 to 0.1 for the standard deviation for the smoother applied to inter-annual changes in availability for each size bin were explored. Values approaching zero produced identical results to a model with no variability over time permitted and larger values quickly converged to relatively stable model estimates of stock size. A working value of 0.025, implying somewhat less change over time than among sizes, produced stable availability estimates and model behavior. This approach also successfully removed the retrospective bias (Fig. 30) in recent status estimates and is consistent with observed geographic and demographic trends.

This change to the assessment model resulted in a much more pronounced decline in the estimated stock trend in recent years (Fig. 31), as a result of much lower estimates of recent recruitments (Fig. 32). These revised estimates also correspond to a large reduction in the average level of productivity, and therefore the absolute value of the spawning biomass reference points (Fig. 33). In tandem, these results suggest only a modest decrease in stock status relative to the harvest policy target. Using only data updated through 2011, the 2011 model estimate of 2012 spawning biomass was 40.9% of the reference level, which was reduced to 31.8% in the revised model, despite a 40% reduction in the absolute estimates. The largest change can be observed in the estimated time-series of age-8+ biomass, which no longer shows a rapid increase in the most recent years (Fig. 34).

Summary of the 2012 model

Little change was made to the 2011 model framework other than the addition of time varying availability, and making aggregate catchability a single estimated quantity. The annual curves remained quite similar for the commercial fi shery over time; however the setline survey is estimated to have experienced increasing availability of smaller halibut and decreasing availability of larger fi sh (Fig. 35). These estimates are consistent with observed increases in abundance in Area 2 and decreases in Areas 3-4, and the biological characteristics generally observed in those areas.

As has been the case in recent assessments, the sport and personal use/subsistence fl eets are assigned the same selectivity pattern as the survey. Bycatch morality is assumed to follow a fi xed selectivity pattern with a dome-shape, selecting more 40-50 cm halibut than 60+ cm (Fig. 36). Commercial fi shery catchability is allowed to vary over time, with the estimated trend for both males and females similar to that seen in previous assessments (Fig. 37). Natural mortality is fi xed at a value of 0.15/year for females and estimated to be 0.143 for males. Likelihood equations and model dynamics were also unchanged from previously documented equations (Clark and Hare 2006a).

An effort to fully document all steps involved with data preparation, and the stock assessment model itself, was begun during 2012; however, it was deemed ineffi cient to proceed with this process until clear guidance on future improvements could be identifi ed. This was achieved during


the Scientifi c Review Meeting (Stewart et al. 2013), and will be completed for the 2013 process. This effort will also focus on developing models with implicit treatment of spatial patterns, better use of data collected prior to 1996, and other improvements identifi ed during the Scientifi c Review Meeting.

Goodness of fi t

The 2012 stock assessment model is able to fi t the primary indices of abundance (WPUE and NPUE) from both the setline survey and commercial fi shery reasonably well, similar to those fi ts reported for previous models (Fig. 38). Because the age data are included in the stock assessment in several different forms (e.g., setline survey proportions-at-age and NPUE-at-age) there is a great deal of redundancy in the fi tting. For this reason, only one set of fi ts to age data from the setline survey and commercial fi shery are presented graphically. Redundant sources produced very similar patterns. The fi t to the total setline survey proportions-at-age captures the general modal structure of the observed data (Fig. 39), and is similar for both females (Fig. 40) and males (Fig. 41). The fi t to the commercial fi shery total catch-at-age (Fig. 42), as well as the females (Fig. 43) and males (Fig. 44) separately is similar to that of the setline survey.

As is often the case patterns in goodness-of-fi t can be much more readily identifi ed through examination of residual plots than through direct examination of fi ts to the data. Residual patterns for survey proportions-at-age indicate some lack-of-fi t associated with the above average 1987 year-class (Fig. 45). This lack of fi t shows a transition from positive to negative values occurring at age-16 in 2003, right after the change from surface to break-and-bake ageing methods. A similar pattern can be seen in the female residuals (Fig. 46), which also show positive values for the oldest fi sh. Male survey residuals tend to show more negative residuals across the youngest and oldest ages (Fig. 47), but a similar transition after 2002. Fishery residuals also display these patterns for fi t to total numbers-at-age (Fig. 48), as well as females (Fig. 49) and males (Fig. 50) separately.

Biomass, recruitment and reference point results

The results of the 2012 stock assessment indicate that the Pacifi c halibut stock has been declining continuously over much of the last decade (Fig. 51, Table 5). This decline has been a result of decreasing size-at-age, as well as relatively poor recruitment strengths (Fig. 52, Table 5). In the last few years, both the exploitable and spawning stock biomass appear to have stabilized, and the predicted numbers-at-age have remained relatively consistent for older ages (e.g., 15-25 years; Table 6). Based on the reductions in recent harvest levels and evidence from the survey index of abundance as well as the age-composition data, the stock assessment estimates that the current stock trajectory is relatively fl at. Spawning biomass is estimated to have increased from 197 to 201 million lb from 2012 to 2013 and exploitable biomass from 179 to 186 million lb over the same period.

The current harvest policy for Pacifi c halibut utilizes a ramp from target harvest rates to no fi shing between 30% relative spawning biomass and 20% relative spawning biomass (Fig. 53). At the beginning of 2013, the stock is estimated to be at 35% of the reference level, just above the harvest policy threshold (Fig. 54, Table 5). The details of the calculation of relative spawning biomass have not changed from recent assessments. Briefl y, this calculation relies on a historical estimate of spawning-biomass-per-recruit (118.5 lb/age-6 recruit), using size-at-age from the 1960s to 1970s (Hare 2012). Average estimated age-6 recruitment is calculated from the assessment,


corrected for environmental regime (Clark and Hare 2006), and then multiplied by the historical spawning-biomass-per-recruit to produce and estimate of the average spawning biomass in the absence of fi shery removals.

The current harvest policy assigns a harvest rate of 21.5% to Areas 2A, 2B, 2C and 3A, and a harvest rate of 16.125% to Areas 3B, 4A, 4B, and 4CDE. Because the harvest policy is defi ned at the Area-specifi c level, the results of apportionment calculations must be used (Webster and Stewart 2013), to evaluate the relative fi shing intensity, even though the assessment is conducted at a coastwide scale. Specifi cally, in order to compare the effective coastwide harvest rate (ECHR) estimated in the stock assessment to a target level, exploitable biomass must be apportioned to area, with area-specifi c catch limits aggregated back to the coastwide level (Fig. 55). Using this method, harvest rates are estimated to have been well above targets for the last decade (Fig. 56). This calculation is made in hindsight, and does not correspond to the estimates and targets as historical management decisions were being made, but to the realized harvest rates now estimated in the 2012 stock assessment. Reductions in harvest levels in 2011 and 2012 have brought realized harvest rates much of the way back toward the coastwide target, and declines in spawning biomass appears to have moderated and reversed slightly in 2012 (Fig. 57).

To provide a direct link, or bridge, with the 2011 stock assessment results, those values are reported along with revised results from that model using the fi nal data sets available through 2011 and through 2012 (Table 7). The estimated trend and absolute level of spawning biomass are both very similar between the two models, with the primary divergence visible in the most recent fi ve years (Fig. 58). The very sharp increase estimated by the 2011 model (and previous models) was an artifact of the retrospective pattern, and such increases had been predicted for several sequential assessments, but had never been subsequently observed in the data. The trend in the estimate of spawning biomass for the 2011 stock assessment model updated through 2012 indicated a continued retrospective pattern, further reinforcing the improvements made in the 2012 stock assessment model.

Major sources of uncertainty

Estimation uncertainty, or the portion of uncertainty associated with estimating the most likely values for the parameters of the stock assessment from the available data, is relatively small. Although this is common for many fi sheries stock assessments, the degree of pre-model processing and redundancy in the halibut data sets likely result in a substantial underestimation of this source of uncertainty. Nonetheless, it is included in the decision-making framework described below. Additional sources of uncertainty include choices made in structuring the assessment model (e.g., explicit inclusion or exclusion of spatial processes), steps taken during data processing, and many other sources that are not included in the results. The Scientifi c Review Meeting identifi ed a number of data and model related aspects of uncertainty that could be included in future stock assessments, but for which there was insuffi cient time during the 2012 process to adequately pursue.

During the 2012 stock assessment process there was substantial discussion regarding estimates of total removals used in the halibut stock assessment. Some of these removals are observed directly through landings, but many others, such as discard mortality and some sources of bycatch in non-target fi sheries are inferred from sparse or incomplete data. Using methods consistent with previous years’ analyses, this stock assessment includes estimates of removals including all sizes of halibut from all sources for which an estimate is available. To the extent that these estimates


are incorrect or incomplete, the results of the assessment will be biased. It is diffi cult to predict how changes in estimated removals might infl uence model results, and potential effects are likely to depend on the trends and absolute scale of such changes. This is the case for nearly all stock assessment analyses, and is an important source of uncertainty if the differences among current estimates and actual removals are large. If improved estimates are made available, these can be directly incorporated into the 2013 stock assessment. If uncertainty estimates can be generated for currently used values, or even if plausible ranges removals can be identifi ed, this is a source of uncertainty that could be directly incorporated into the decision-making framework outlined below.

Recent trends of below average recruitment and decreasing size-at-age have been important contributing factors in the overall stock decline. Unfortunately, although the stock assessment can track these trends quite precisely, it does not provide information on the mechanisms causing these trends. The effects of recent poor recruitment are likely to infl uence spawning biomass trends in the near-term, as these weak cohorts mature. Regardless of harvest levels, potential increases in stock biomass will also be very sensitive to future trends in size-at-age and recruitment. Until these processes are better understood they represent a substantial source of uncertainty that is diffi cult to include in the forecast projections. Extending the time-series of data included in the stock assessment may help to better identify covariates (e.g., Clark and Hare 2002) which will improve understanding of these population mechanisms. Extending the time-series may also help to reduce the effects of the current ‘one-way-trip’ of decreasing indices of abundance. Such trends are known to create problems for stock assessment models in delineating between productivity and absolute population size (e.g., Hilborn and Walters 1992).

Sensitivity analyses using the 2011 modelBecause survey catchability was a major source of discussion during the 2011 stock assessment

process, and was suggested to be a potential factor contributing to the retrospective bias (Valero, unpublished analyses), a sensitivity to the treatment of this parameter is presented for the wobblesq model. Using data updated through 2012, a single value for catchability was (constant over time), was estimated and compared with the time-varying implementation in the wobblesq model. This analysis revealed a similar pattern as in 2011: the absolute stock estimate decreased slightly, but over a relatively small range compared to the full application of time-varying availability (Fig. 59). This is likely due to time-varying catchability capturing a small amount of the demographic shift in availability, but not enough to remove the retrospective pattern.

Sensitivity analysesDuring preliminary model investigation conducted during 2012, a wide range of sensitivity

analyses were conducted in order to better understand the general modeling approach, identify important aspects of the data and weighting of these data during model fi tting, and determine which components in the entire analysis had the most direct effect on absolute estimates of stock size. A discussion of a number of these analyses is included in the Scientifi c Review Meeting document (Stewart et al. 2013), and several were identifi ed as high-priority for a full investigation during the 2013 stock assessment process. For 2012, natural mortality was identifi ed as the most infl uential fi xed parameter or assumption in the Pacifi c halibut stock assessment.

Natural mortality is a dominant source of uncertainty in many fi sheries stock assessments. A value of 0.15/year for female halibut has been used for all recent halibut assessments (the value


for males has been estimated, the relative difference between the sexes being well-informed by the observed sex-ratios in the age data), based on a downward revision made for the 1998 assessment. That well-justifi ed revision resulted in a signifi cant change to the estimates of stock size at the time. In order to avoid abrupt changes in future estimates uncertainty in the natural mortality rate is explicitly included in this assessment. The approach is based on selecting two alternate values of natural mortality, each approximately half as likely as the current best estimate (0.15). These values were 0.10 and 0.2, which produced estimates of spawning biomass that differed by more than 80 million lb (Fig. 60). The method of including and reporting this uncertainty in the decision-making table is described below, as part of the forecasting approach.

An extensive bait-comparison experiment was conducted during the 2012 setline survey (Webster et al. 2013), which included setting somewhat fewer skates baited with the standard bait of chum salmon. Although the number of skates deployed was still within historical survey protocols, there is some information to suggest slightly higher catch-rates for skates at the ends of the gear (Webster and Hare 2011), and this was factored in to the bait experiment analysis. Since number of skates deployed is not a direct component in the standard survey analysis, a stock assessment sensitivity run was conducted by decreasing the 2012 survey catch rate to by 3%. This value represents the maximum difference estimated for any area-specifi c comparison of end- vs. middle-skates that had a substantial quantity of data available (the values actually range below zero as well). This analysis confi rmed that current assessment results were not sensitive to small changes in survey catch rates (Fig. 61).

Retrospective analysesA retrospective analysis of the 2012 model revealed little pattern in recent spawning biomass

estimates as data are sequentially removed from the model (Fig. 62). This is important information for the decision-making process, as previous estimates have been known to include bias, which indicated that reduced catches might be warranted (Valero 2012). Although the improvement in model performance in 2012 is no guarantee that future retrospective patterns may not arise from other aspects of the model (or data sources), the lack of a retrospective pattern means that the results of the 2012 stock assessment are likely to be more reliable than those reported in recent years.

Forecasts and decision-making table

For the 2012 assessment, signifi cant improvements to the methods used to forecast future stock size and to calculate the uncertainty associated with these predictions were made. Changes from previous assessments included integrating the forecasting step into the stock assessment model (rather than treating it as a subsequent independent calculation), which enabled direct inclusion of estimation uncertainty. In addition, given the pronounced declining trends in recent size-at-age, alternative projections were run using observed size-at-age from 2012, as well as fi tting a linear trend to the most recent three years of data.

Stock projections rely on the results from the stock assessment, summaries of the removals in 2012, as well as the results of apportionment calculations. The steps included: 1) analyzing the survey catch rates consistent with recent approaches (Webster and Stewart 2013), 2) using the estimated survey biomass distribution to apportion the coastwide exploitable biomass estimates from the stock assessment, 3) applying the area-specifi c harvest rates to generate the TCEY for


each area, 4) subtracting the other removals (O26), assuming that the values for 2013 remain constant at 2012 levels (with the exception of commercial wastage, and sport and personal use in Area 2A and sport in 2B, which are scaled proportional to the TCEY), and 5) calculating the total coastwide mortality (including U26). This calculation results in the application of the current harvest policy, completely consistent with the approach used in recent stock assessments.

The projected removals consistent with the current harvest policy are identifi ed as the Blue Line in the decision-making table and forecast results. For alternative levels of coastwide harvest, the TCEY values were scaled according to the calculations above, and a range of levels from no removals (useful as a comparison to potential management actions; this represents the predicted stock dynamics in the absence of future harvest) to a total mortality of 60 million lb was considered. Increments among alternatives were initially set to roughly 10 million lb of total mortality, with additional rows for a zero FCEY, as well as a row corresponding to the current harvest policy applied to the updated results of the 2011 stock assessment (for bridging comparison). At the request of the Commissioners during the Interim Meeting, two additional rows adjacent to the Blue Line, but differing by only 5 million lb total mortality were added to the decision making table.

Models using all three values of natural mortality were projected three years into the future assuming constant catches at the value identifi ed for each row of the table. Each of these results includes a distribution for forecast quantities, representing estimation uncertainty. These distributions were then weighted and combined, assigning 50% of the probability to the best estimate of natural mortality (0.15) and 25% to the higher and lower alternatives, in order to generate a single probability distribution for a suite of risk metrics. Risk metrics included harvest intensity, stock status relative to harvest policy reference points, stock trend relative to 2013 estimates, as well as catch trend relative to the catch associated with each row.

Initial projections were found to be sensitive to the treatment of future size-at-age. Therefore, based on the advice of the Scientifi c Review Meeting, forecasts were conducted using recent trend, which indicated a reduced level of future biomass relative to simply assuming the 2012 size-at-age would persist (Fig. 63).

Although the stock is projected to remain stable or increase slightly in the absence of mortality during 2013, all levels of harvest evaluated resulted in declines in the current stock size by 2014 (Fig. 64). There is a 25% probability that the stock will be below the harvest policy threshold of 30% of the reference level of spawning biomass, regardless of the removals in 2013 (column b, Table 8); however there is less than a 1% probability that the stock is below the harvest policy limit of 20% relative spawning biomass (column c). There is a 23% probability that the stock will be smaller in 2014 than it is estimated to be in 2013 in the absence of any removals. Because the stock trajectory is estimated to be very fl at, any removals in 2013 yield a much larger probability of a smaller stock in 2014 than 2013, ranging from 76% to 86% over the range of alternatives evaluated (column d). Despite the high probability of a one-year decline in the stock abundance, there is a very low probability that this decline will be large. Probabilities of a greater than 5% stock decline are all 4% or less (column e).

Given recent poor recruitment, declines in spawning biomass are projected to be very likely over a three-year projection, with a probability of 41% in the absence of harvest, increasing rapidly to 95-99% over the alternatives considered (Fig. 65; Table 8, column f). However, the probabilities of dropping below management reference points by 2016 do not change appreciably from those for 2014 (Table 9). Given the current harvest policy, if a fi shery CEY of 22.7 million pounds (the Blue line) is removed in 2013, there is an almost even chance (48%) that the exploitable biomass in


2014 could produce a catch at least as large (column g). For smaller removals, there is a very high probability that the harvest policy catch would be larger in 2014, but there is a very low probability of the same or larger FCEY for catches above 22.7 million pounds. Similarly, all harvests smaller than 17.7 million pounds result in a very low probability (1% or less) of exceeding the coastwide target, but those above 22.7 have very high probabilities (75% to > 99%; Table 8, column a).

Future research

Historically, there has been signifi cant investigation into the performance of both area-specifi c and coastwide models for conducting the halibut stock assessment. Recently, new fi sheries approaches have been developed and tested for dealing with spatial processes. These improvements represent an opportunity to revisit the problem in the near-term. Building upon the work completed for 2012, and following the guidance of the Scientifi c Review Meeting (Stewart et al. 2013), future efforts will focus on several key aspects of the stock assessment:

1) Improved accounting for additional sources of uncertainty through reduced data processing, use of more fl exible model structures capable of directly including alternate structural hypotheses, Bayesian methods for fully integrating parameter uncertainty and model averaging.

2) Development of implicitly and explicitly spatial models to better incorporate the spatial variability observed for halibut.

3) Further investigation of the factors contributing to recruitment strength and observed size-at-age in order to better forecast trends in these quantities.

4) Simulation testing the stock assessment model based on data generated from a research model.Additional work during 2013 will address the specifi c items relating to data processing

and model details listed in the report from the Scientifi c Review (Stewart et al. 2013).

Acknowledgements

We thank all of the IPHC staff for their contributions to data collection, analysis and preparation for the stock assessment. Ian Stewart particularly wishes to thank everyone involved in the assessment process for helping me to better understand 100 years of scientifi c research in less than three months. The improvements made for 2012 were made possible by extensive analysis and investigation by previous assessment authors and IPHC staff including: Steven Hare, Bill Clark, Juan Valero, as well as many others. The Scientifi c Review Meeting held 24-26 October, 2012 provided a great deal of insight and creative ideas for best presenting the 2012 assessment results, particularly the format and contents of the decision-making table. Contributions from Jim Ianelli and Robyn Forrest substantially improved the content and clarity of this analysis.


References

Clark, W. G. 1999. Effects of an erroneous natural mortality rate on a simple age-structured stock assessment. Can. J. Fish. Aquat. Sci. 56:1721-1731.

Clark, W. G. 2003. A model for the world: 80 years of model development and application at the international Pacifi c halibut commission. Nat. Res. Mod. 16:491-503.

Clark, W. G. 2004. Nonparametric estimates of age misclassifi cation from paired readings. Can. J. Fish. Aquat. Sci. 61:1881-1889.

Clark, W. G. and Hare, S. R. 2002. Effects of Climate and Stock Size on Recruitment and Growth of Pacifi c Halibut. N. Am. J. Fish. Man. 22:852-862.

Clark, W. G. and Hare, S. R. 2007a. Assessment and management of Pacifi c halibut: data, methods, and policy. Int. Pac. Halibut Comm. Sci. Rep. No. 83.

Clark, W. G. and Hare, S. R. 2007b. Assessment of the Pacifi c halibut stock at the end of 2006. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2006: 97-128.

Clark, W. G. and Hare, S. R. 2008. Assessment of the Pacifi c halibut stock at the end of 2007. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2007: 177-204.

Clark, W. G. and Parma, A. M. 1999. Assessment of the Pacifi c halibut stock in 1998. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 1998: 89-112.

Gilroy, H. L. and Stewart, I. J. 2013. Incidental mortality of halibut in the commercial halibut fi shery (Wastage). IPHC Report of Assessment and Research Activities 2012: 53-60.

Hare, S. R. 2010. Assessment of the Pacifi c halibut stock at the end of 2009. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2009: 91-170.



Hare, S. R. and Clark, W. G. 2009. Assessment of the Pacifi c halibut stock at the end of 2008. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2008: 137-202.

Hilborn, R. and Walters, C. J. 1992. Quantitative fi sheries stock assessment: choice, dynamics and uncertainty. London, Chapman and Hall.

Piner, K. R. and Wischnioski, S. G. 2004. Pacifi c halibut chronology of bomb radiocarbon in otoliths from 1944 to 1981 and a validation of ageing methods. J. Fish Bio. 64:1060-1071.

Stewart, I. J., Martell, S., Webster, R. A., Forrest, R., Ianelli, J., and Leaman, B. M. 2013. Assessment review team meeting, October 24-26, 2012. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2012: 239-266.

Sullivan, P. J. and Parma, A. M. 1998. Population assessment, 1997. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 1997: 83-210.


Valero, J. L. 2012. Harvest policy considerations on retrospective bias and biomass projections. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2011: 311-329.

Webster, R. A. 2010. Analysis of PIT tag recoveries through 2009. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2009: 177-186.

Webster, R.A. and Hare, S. R. 2011 Adjusting IPHC setline survey WPUE for survey timing and hook competition. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2010: 251-259.

Webster, R. A., Kaimmer, S. M., Dykstra, C., and Leaman, B. M. 2013. Coastwide comparison of alternative setline survey baits. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2012: 569-586.

Webster, R. A. and Stewart, I. J. 2013. Apportionment and regulatory area harvest calculations. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2012: 187-206.


Table 1. Time-series of removals (by source; million lb, net wt.) used in the stock assessment.

YearCommercial fi shery

Commercial wastage Bycatch Sport

Personal use Total

1996 47.69 0.73 14.46 8.08 0.54 71.511997 65.49 1.05 13.51 9.03 0.54 89.611998 70.12 1.20 13.43 8.59 0.74 94.071999 74.70 1.34 13.84 7.38 0.75 98.002000 68.55 1.29 13.29 9.01 0.76 92.892001 70.97 1.44 13.16 8.10 0.77 94.452002 74.95 1.66 12.61 8.01 0.76 97.992003 73.36 1.77 12.58 9.35 1.38 98.442004 73.31 1.93 12.58 10.70 1.53 100.052005 72.11 2.03 13.26 10.86 1.54 99.802006 68.12 2.05 13.08 10.19 1.48 94.922007 63.03 2.29 12.27 11.46 1.49 90.532008 58.70 2.34 11.89 10.67 1.34 84.932009 52.18 2.62 11.38 8.75 1.31 76.242010 49.83 3.04 10.63 7.80 1.24 72.542011 39.61 2.21 9.90 7.08 1.24 60.042012 31.87 1.54 9.87 6.85 1.24 51.36


Table 2. List of data sources included in the assessment.

Years Range Resolution DataSetline survey data1997-2001,2002-2011

Ages: 6-20+, 6-25+

Males, Females, Total

Proportions-at-age, Standard Error (SE) propor-tions-at-age, Numbers-per-unit-effort (NPUE)-at-age, SE of NPUE-at-age, Mean length-at-age, Ageing bias-corrected mean length-at-age, Mean weight-at-age, Ageing bias-corrected mean weight-at-age, Proportion legal (over 32”), Legal weight-at-age

1996-2011 Aggregated Aggregated NPUE, SE of NPUE, Weight-per-unit-effort (WPUE), SE of WPUE

Commercial fi shery data1996-2001,2002-2011

Ages: 6-20+, 6-25+

Males, Females, Total

Numbers-at-age, SE of Numbers-at-age, NPUE-at-age, SE of NPUE-at-age, Mean weight-at-age, Ageing bias-corrected mean weight-at-age

1996-2011 Aggregated Aggregated NPUE, SE of NPUE, WPUE, SE of WPUEBycatch data1996-2011 Lengths:

0-120 cm, 10-cm bins

Aggregated Numbers, SE of numbers

1996-2011 Ages: 6-30 Males, Females, Total

Ageing bias-corrected mean length-at-age, SE of ageing bias-corrected mean length-at-age, Ageing bias-corrected mean weight-at-age

Removals data1996-2011 Aggregated Aggregated Total weight of removals for: commercial, discard,

bycatch, sport, personal useAgeing imprecision dataAggregated Ages:

1-20, 1-30

Aggregated Transition matrix from observed: surface to canon-ical age, break-and-bake to canonical age

Maturity dataAggregated Ages:6-30+ Females Maturity-at-age


Table 3. Indices of O32 abundance used in the stock assessment (WPUE in lb/skate, NPUE in number/skate).

Year

Setline survey WPUE

Setline survey

WPUE SE

Setline survey NPUE

Setline survey NPUE

SEFishery WPUE

Fishery WPUE

SEFishery NPUE

Fishery NPUE

SE1996 NA NA NA NA 415 9 14.4 0.321997 138.2 4.0 8.0 0.2 423 9 14.4 0.311998 133.9 3.7 7.6 0.2 429 9 15.3 0.331999 126.1 3.7 6.9 0.2 398 9 15.1 0.342000 120.6 3.3 6.8 0.2 417 9 15.2 0.342001 112.3 3.3 6.6 0.2 382 9 14.0 0.322002 108.8 3.2 6.6 0.2 379 9 13.8 0.332003 91.6 2.7 6.0 0.2 346 8 12.8 0.312004 88.4 2.6 6.6 0.2 338 8 13.1 0.312005 82.1 2.4 6.1 0.2 314 7 12.5 0.302006 71.1 2.2 5.6 0.1 283 7 11.5 0.282007 65.8 1.9 5.8 0.1 268 6 11.3 0.282008 60.2 1.7 5.7 0.1 249 6 10.6 0.262009 55.4 1.6 5.5 0.1 236 5 10.3 0.242010 47.0 1.5 5.2 0.1 210 5 9.5 0.232011 44.7 1.3 5.1 0.1 209 5 9.6 0.242012 49.9 1.5 5.5 0.1 209 5 9.5 0.23


Table 4. Summary of historical stock assessment models.

Years Model IssuesPre-1977 Yield, yield-per-recruit, simple stock-

production modelsNo growth or recruitment variability

1978-1981 Cohort analysis, coastwide, natural mor-tality (M)=0.2

Unstable estimates

1982-1983 Catch-AGE-ANalysis (CAGEAN; age-based availability), coastwide, M=0.2

Migratory dynamics not accounted for

1984-1988 CAGEAN, area-specifi c, migratory and coastwide, M=0.2

Trends differ by area

1989-1994 CAGEAN, area-specifi c, M=0.2, age-based selectivity

Retrospective pattern

1995-1997 Statistical Catch-Age (SCA), area-specif-ic, length-based selectivity, M=0.2

M estimate imprecise

1998-1999 SCA, area-specifi c, length-based selec-tivity, M=0.15

Poor fi t to data

2000-2002 New SCA, area-specifi c, constant age-based selectivity, M=0.15


2003-2006 SCA, area-specifi c, constant length-based selectivity, M=0.15

Migratory dynamics created bias

2006-2011 SCA, coastwide, constant length-based availability, M=0.15



Table 5. Time-series of population estimates (million lb, numbers in millions). Age-6 recruits in 2013 refl ect the mean, rather than an estimate updated by the data.

Year Total biomassExploitable

biomassSpawning biomass

Relative spawning biomass

Age-6 recruits

1996 1,225.51 518.76 351.35 61% 12.961997 1,270.41 569.70 383.30 67% 11.471998 1,243.76 575.37 401.01 70% 9.911999 1,174.19 555.38 397.35 69% 9.012000 1,062.68 503.88 368.05 64% 17.612001 941.56 445.44 327.82 57% 18.882002 927.20 418.53 318.63 56% 13.562003 893.57 380.42 288.05 50% 12.592004 837.31 339.08 259.68 45% 20.132005 781.80 300.62 233.87 41% 27.202006 772.50 268.46 214.80 37% 21.042007 795.46 236.33 201.97 35% 14.862008 788.50 210.10 192.90 34% 14.492009 750.72 191.32 186.97 33% 9.322010 716.08 180.56 187.94 33% 7.832011 667.25 173.91 190.11 33% 7.202012 632.77 178.84 196.91 34% 3.782013 598.03 186.49 200.68 35% 14.13


Tabl

e 6.

Est

imat

ed n

umbe

rs a

t age

(mill

ions

). A

ge-6

in 2

013

refl e

cts t

he m

ean,

rat

her

than

an

estim

ate

upda

ted

by th

e da

ta.

Age

(yr)

Year

67

89

1011

1213

1415

1617

1819

20+

1996

12.9

613

.32

19.7

625

.77

10.1

25.

745.

755.

923.

121.

752.

071.

780.

941.

920.

6219

9711

.47

11.0

111

.29

16.6

821

.61

8.43

4.74

4.70

4.81

2.52

1.40

1.65

1.41

0.74

2.00

1998

9.91

9.75

9.33

9.53

13.9

517

.90

6.90

3.84

3.77

3.84

2.00

1.10

1.29

1.10

2.11

1999

9.01

8.42

8.27

7.87

7.96

11.5

414

.61

5.57

3.07

2.99

3.03

1.56

0.86

1.00

2.45

2000

17.6

17.

667.

146.

966.

566.

579.

3811

.71

4.42

2.42

2.34

2.35

1.21

0.66

2.61

2001

18.8

814

.97

6.49

6.01

5.81

5.43

5.36

7.55

9.33

3.50

1.90

1.83

1.83

0.93

2.49

2002

13.5

616

.04

12.6

75.

475.

014.

794.

414.

305.

987.

322.

721.

471.

411.

402.

5720

0312

.59

11.5

313

.56

10.6

14.

534.

113.

873.

523.

394.

675.

672.

101.

131.

082.

9620

0420

.13

10.6

99.

7211

.33

8.75

3.69

3.30

3.07

2.76

2.64

3.60

4.35

1.61

0.86

3.00

2005

27.2

017

.08

9.01

8.11

9.31

7.08

2.94

2.59

2.38

2.11

2.01

2.72

3.28

1.21

2.84

2006

21.0

423

.06

14.4

07.

506.

647.

505.

592.

291.

991.

801.

591.

502.

042.

452.

9920

0714

.86

17.8

419

.47

12.0

26.

165.

365.

944.

341.

751.

501.

351.

191.

121.

524.

0020

0814

.49

12.5

915

.09

16.2

99.

904.

984.

254.

603.

301.

311.

121.

000.

880.

834.

0220

099.

3212

.26

10.6

712

.65

13.4

78.

023.

953.

303.

512.

480.

980.

830.

740.

653.

5320

107.

837.

8710

.36

8.92

10.4

410

.94

6.41

3.11

2.56

2.69

1.88

0.74

0.63

0.56

3.11

2011

7.20

6.61

6.65

8.67

7.35

8.47

8.74

5.05

2.41

1.96

2.06

1.42

0.56

0.47

2.74

2012

3.78

6.08

5.58

5.57

7.16

6.00

6.83

6.97

3.98

1.89

1.53

1.60

1.10

0.43

2.47

2013

14.1

33.

195.

124.

674.

605.

874.

885.

515.

583.

171.

501.

211.

270.

872.

27


Table 7. Results of the bridging analysis, comparing the 2011 (wobblesq) model with data through 2011, data through 2011 but updated in 2012, data through 2012 and the current assessment model results (right-hand column).

Model 2011 (wobblesq) 2012End year 2011 2011 2012 2012

Data fi nalized in: November 2011

November 2012

November 2012

November 2012

Quantity2012 Spawning biomass 319 309 272 1972012 Relative spawning biomass 42% 41% 38% 34%2013 Spawning biomass -- -- 324 2012013 Relative spawning biomass -- -- 46% 35%2012 Exploitable biomass 260 252 219 1792013 Exploitable biomass -- -- 258 1862012 Coastwide harvest rate 19.4% 18.9% 21.8% 26.7%


Table 8. Decision-making table. Values indicate the probability of the outcome in each column given the level of removals for that row.

Fishing intensity

Catch trend

Effective coastwide HR

Fishery CEY

2013 2016 2014

is greater than

target

is less than

30%

is less than

20%

is less than

2013

is 5% less than

2013

is less than

2013

is less than

20130.0 (0.0) 0% 25% <1% 23% <1% 41% 0%

0.0 (16.5) <1% 25% <1% 76% 2% 95% 0%3.4 (20.0) <1% 25% <1% 77% 2% 96% <1%

12.9 (30.0) 1% 25% <1% 79% 2% 97% 1%17.7 (35.0) 23% 25% <1% 80% 2% 97% 19%

22.7 (40.2) 50% 25% <1% 82% 3% 97% 48%27.3 (45.0) 75% 25% <1% 83% 3% 98% 75%32.1 (50.0) 84% 25% <1% 84% 3% 98% 85%36.2 (54.3) 97% 25% <1% 85% 4% 98% 97%41.6 (60.0) >99% 25% <1% 86% 4% 99% >99%

a b c d e f g

Coastwide Fishery CEY

(total removals) millions lb

Stock status Stock trend

Spawning biomass2014


T ab le 9. Extended decision-making table columns for 3-year projections.

2016 2016 2016

is less than

2013

is less than

30%

is less than

20%0.0 (0.0) 41% 24% <1%

0.0 (16.5) 95% 27% <1%3.4 (20.0) 96% 27% <1%

12.9 (30.0) 97% 27% <1%17.7 (35.0) 97% 27% <1%

22.7 (40.2) 97% 28% <1%27.3 (45.0) 98% 28% <1%32.1 (50.0) 98% 28% <1%36.2 (54.3) 98% 28% <1%41.6 (60.0) 99% 29% <1%

Stock trend

Spawning biomassCoastwide Fishery CEY



Fi gure 1. Time series of total removals (million lb) from all sources, 1888-2012. Horizontal line indicates the most recent 100-year average.


Fi gure 2. Recent removals by regulatory area and source, 2005-2012. Values below the year labels indicate the total removals from all sources, the percent change in the total from the previous year and the percent change from the directed fi shery removals in the previous year.


Fi gure 3. Time series management decisions (Total CEY), and available information at the time the decision was made from the stock assessment (exploitable biomass; million lb), the setline survey (WPUE; lb/skate) and the commercial fi shery (WPUE; lb/skate).


Fi gure 4. Time series fi shery targets harvests (Fishery CEY), and total removals from all sources (million lb).


Fi gure 5. Trend in setline survey WPUE, 1997-2012, colors indicate the contributions from each regulatory area to the geographically-weighted total.


Fi gure 6. Trends in setline survey WPUE by regulatory area, percentages below the area labels indicate the percent change from 2011 to 2012 observations.


Figure 7. Trends in setline survey NPUE by regulatory area, percentages below the area labels indicate the percent change from 2011 to 2012 observations.


Fi gure 8. Observed proportions-at-age from the setline survey, 1997-2012. The area of each circle is scaled relative to the legend value in the upper left. Age-20 (prior to 2002) and age-25 (thereafter) represent plus-groups containing that age and all older ages. The 1987 year-class is identifi ed by the diagonal line for visual reference.


Fi gure 9. Coastwide aggregate age distributions from the 2012 setline survey (upper panel) and commercial fi shery (lower panel).


Fi gure 10. Trends in average age (upper panel) and average fi sh weight (lower panel) observed in the setline survey and commercial fi shery. Values represent coastwide weighted averages across all regulatory areas.


Fi gure 11. Trends in smoothed female setline survey length-at-age (upper panel) and weight-at-age (lower panel), 1996-2012.


Fi gure 12. Trends in smoothed male setline survey length-at-age (upper panel) and weight-at-age (lower panel), 1996-2012.


Fi gure 13. Comparison of area-swept estimated total biomass from the NOAA Bering Sea bottom trawl survey for all sizes of halibut (upper line, with approximate 95% confi dence intervals), the portion of that biomass believed to be available to the IPHC setline survey (lower line, with approximate 95% confi dence intervals) and the calibrated survey WPUE estimate based on comparison data from 2006 (inverted triangles; numbers above points indicate annual estimates).


Fi gure 14. Comparison of area-swept estimated total numbers from the NOAA Bering Sea bottom trawl survey for all sizes of halibut (middle line at the end of the time-series, with approximate 95% confi dence intervals), the portion of those numbers believed to be available to the IPHC setline survey (upper line at the end of the time-series, with approximate 95% confi dence intervals) and the portion of those numbers estimated to be available to the commercial fi shery (lower line at the end of the time-series, with approximate 95% confi dence intervals).


Fi gure 15. Recent area-swept estimates of total numbers of fi sh from the NOAA Bering Sea bottom trawl survey by 10-cm size-bin (30-160 cm), 2004-2012. Values reported below the year labels indicate the total biomass and total numbers estimates for that year.


Fi gure 16. Comparison of area-swept estimated total numbers from the NOAA Aleutian Islands bottom trawl survey for all sizes of halibut (points, with approximate 95% confi dence intervals), and the portion of those numbers believed to be available to the IPHC setline survey (crosses, with approximate 95% confi dence intervals).


Fi gure 17. Comparison of area-swept estimated total numbers from the NOAA Gulf of Alaska bottom trawl survey for all sizes of halibut (points, with approximate 95% confi dence intervals), and the portion of those numbers believed to be available to the IPHC setline survey (crosses, with approximate 95% confi dence intervals).


Fi gure 18. Trends in commercial fi shery WPUE by regulatory area, percentages below the area labels indicate the percent change from 2011 to 2012 observations. The shaded portion in each panel indicates historical data not currently included in the stock assessment model.


Fi gure 19. Trends in commercial fi shery WPUE by regulatory area. The shaded portion in each panel indicates historical data not currently included in the stock assessment model.


Fi gure 20. Age distributions from the 2012 commercial fi shery by regulatory area.


Fi gure 21. Observed numbers-at-age from the commercial fi shery, 1996-2012. The area of each circle is scaled relative to the legend value in the upper left. Age-20 (prior to 2002) and age-25 (thereafter) represent plus-groups containing that age and all older ages. The 1987 year-class is identifi ed by the diagonal line for visual reference.


Fi gure 22. Trends in average fi sh weight observed in the commercial fi shery by regulatory area.


Fi gure 23. Trends in smoothed female commercial fi shery length-at-age (upper panel) and weight-at-age (lower panel), 1996-2012.


Fi gure 24. Trends in smoothed male commercial fi shery length-at-age (upper panel) and weight-at-age (lower panel), 1996-2012.


Fi gure 25. Maturity curve used to calculate the proportion of female stock contributing to the spawning biomass.

Fi gure 26. Retrospective analysis among recent stock assessments.


Fi gure 27. Retrospective analysis for the 2011 “wobblesq” model.

Fi gure 28. Retrospective analysis of cohort-strength estimates from the 2011 wobblesq model (fi gure from the 2011 stock assessment document).


Fi gure 29. Comparison of three alternative approaches solving the retrospective pattern in the 2011 wobblesq model.

Fi gure 30. Retrospective analysis for the model allowing time-varying availability, and using data through 2011.


Fi gure 31. Comparison of the 2011 and revised stock assessment models using data updated through 2011.

Fi gure 32. Comparison of the 2011 and revised stock assessment model estimates of recruitment using data updated through 2011.


Fi gure 33. Comparison of the relative spawning biomass estimates from the 2011 (wobblesq) model (upper panel) and the current assessment (lower panel) using data through 2011.


Fi gure 34. Biomass time-series results from the model allowing time-varying availability, and using data through 2011.


Fi gure 35. Estimated annual availability curves for the setline survey (upper two panels) and the commercial fi shery (lower two panels) by sex. Note that the upper curve for the 60-80 cm range on each panel represents the most recent year, indicating a shift toward smaller fi sh.


Fi gure 36. Fixed selectivity curve assigned to bycatch mortality.

Fi gure 37. Estimated trends in sex-specifi c catchability for the commercial fi shery (females are represented by the upper line, males the lower line).


Fi gure 38. Fit to the commercial fi shery and setline survey WPUE and NPUE indices of abundance.


Fi gure 39. Fit to setline survey total proportions-at-age (points indicate the observed data, lines the model predictions.


Fi gure 40. Fit to setline survey female proportions-at-age (points indicate the observed data, lines the model predictions.


Fi gure 41. Fit to setline survey male proportions-at-age (points indicate the observed data, lines the model predictions.


Fi gure 42. Fit to commercial fi shery total catch-at-age (points indicate the observed data, lines the model predictions.


Fi gure 43. Fit to commercial fi shery female catch-at-age (points indicate the observed data, lines the model predictions.


Fi gure 44. Fit to commercial fi shery male catch-at-age (points indicate the observed data, lines the model predictions.


Fi gure 45. Standardized residuals (observed minus expected values) from the fi t to setline survey total catch-at-age. Circle areas are scaled relative to the legend value in the upper left, fi lled circles indicate positive values.


Fi gure 46. Standardized residuals (observed minus expected values) from the fi t to setline survey female catch-at-age. Circle areas are scaled relative to the legend value in the upper left, fi lled circles indicate positive values.


Fi gure 47. Standardized residuals (observed minus expected values) from the fi t to setline survey male catch-at-age. Circle areas are scaled relative to the legend value in the upper left, fi lled circles indicate positive values.


Fi gure 48. Standardized residuals (observed minus expected values) from the fi t to commercial fi shery total catch-at-age. Circle areas are scaled relative to the legend value in the upper left, fi lled circles indicate positive values.


Fi gure 49. Standardized residuals (observed minus expected values) from the fi t to commercial fi shery female catch-at-age. Circle areas are scaled relative to the legend value in the upper left, fi lled circles indicate positive values.


Fi gure 50. Standardized residuals (observed minus expected values) from the fi t to commercial fi shery male catch-at-age. Circle areas are scaled relative to the legend value in the upper left, fi lled circles indicate positive values.


Fi gure 51. Time series of biomass results.

Fi gure 52. Time series of age-8 recruitments with estimation uncertainty.


Fi gure 53. Illustration of the current IPHC harvest control rule for determining the relative target harvest rate as a function of relative spawning biomass, consistent with the IPHC’s overall harvest policy.

Fi gure 54. Time-series of spawning biomass relative to harvest policy reference points.


Fi gure 55. Illustration of the method for calculating the Effective Coastwide Harvest Rate (ECHR), consistent with the IPHC’s overall harvest policy.

Fi gure 56. Time series of realized coastwide harvest rates (bars) and hindcast harvest rate targets (horizontal dashes). Note that hindcast harvest rate targets represent the current perception of exploitable biomass, not the perception in that year.


Fi gure 57. Phase plot of relative stock size and fi shing intensity.

Fi gure 58. Results of bridging analysis indicating the spawning biomass estimated by the 2011 (wobblesq) model updated with data through 2012 and the current assessment model (2012 Base-case).


F igure 59. Sensitivity analysis for the 2011 (wobblesq) model illustrating the effect of time-varying vs. time-invariant setline survey catchability.

Fi gure 60. Sensitivity analysis to the value used for female natural mortality; the results from the best-estimate (0.15/year) are represented by the middle line.


Fi gure 61. Sensitivity analysis to hypothetical very strong effects of the bait experiment on standard survey skates deployed.

Fi gure 62. Results of the retrospective analysis on spawning biomass estimates.


Fi gure 63. Comparison of exploitable biomass for three-year forecasts assuming size-at-age remains constant at 2012 observed values (upper line) or follows recent (and generally declining) trends by age (lower line).

Fi gure 64. One-year forecasts under alternative removal scenarios.


Fi gure 65. Three-year forecasts under alternative removal scenarios and assuming recent trends in size-at-age continue.


Appendix A: Data summaries by regulatory area.

Table A1. Time-series of total removals by regulatory Area (million lb, net wt.).

Year 2A 2B 2C 3A 3B 4 4A 4B 4CDE Total1974 1.00 6.43 6.17 13.50 5.10 8.33 NA NA NA 40.541975 0.94 9.18 6.93 13.85 4.65 4.28 NA NA NA 39.841976 0.72 9.51 6.28 14.64 5.20 5.29 NA NA NA 41.631977 0.70 7.39 3.87 13.02 5.12 4.14 NA NA NA 34.241978 0.59 6.20 4.82 13.75 3.17 6.38 NA NA NA 34.901979 0.54 6.84 5.56 17.62 1.33 6.79 NA NA NA 38.681980 0.52 7.16 4.12 18.44 1.53 9.95 NA NA NA 41.721981 0.70 7.01 4.87 19.85 2.02 7.62 NA NA NA 42.061982 0.74 6.60 4.33 18.16 7.04 6.21 NA NA NA 43.081983 0.81 6.63 7.30 18.15 9.80 8.72 NA NA NA 51.411984 1.03 10.55 6.86 23.10 8.30 7.89 NA NA NA 57.731985 1.17 12.32 10.51 24.17 11.85 8.69 NA NA NA 68.711986 1.40 13.25 12.21 37.74 9.78 11.54 NA NA NA 85.921987 1.52 14.83 12.28 37.49 9.11 12.98 NA NA NA 88.211988 1.22 15.27 13.11 46.55 7.39 13.70 NA NA NA 97.231989 1.29 12.69 11.73 41.97 9.01 12.42 NA NA NA 89.101990 0.95 11.06 12.39 38.19 11.13 -- 4.67 2.34 7.33 88.061991 0.94 9.76 12.28 34.44 14.44 -- 4.87 2.69 9.11 88.531992 1.15 9.98 12.81 37.05 11.10 -- 5.35 3.56 8.85 89.861993 1.22 13.23 14.35 33.44 9.24 -- 4.52 2.87 6.99 85.861994 1.01 12.02 13.44 34.97 5.46 -- 4.12 3.12 7.92 82.071995 1.17 12.56 10.02 26.32 5.00 -- 3.74 2.66 7.28 68.731996 1.16 11.25 11.50 27.77 5.76 -- 3.81 3.05 7.22 71.511997 1.41 14.11 12.66 33.71 10.79 -- 4.88 4.24 7.82 89.611998 1.94 14.90 13.42 33.76 12.86 -- 5.39 3.79 8.01 94.071999 1.80 14.37 12.74 33.11 15.98 -- 6.36 4.48 9.17 98.002000 1.68 12.63 11.43 28.00 17.39 -- 7.08 5.57 9.11 92.892001 1.99 12.06 11.02 29.82 18.52 -- 6.84 5.32 8.88 94.452002 1.92 14.20 11.38 30.26 19.83 -- 6.97 4.93 8.51 97.992003 1.52 13.89 11.83 32.24 19.62 -- 6.78 4.66 7.90 98.442004 1.69 14.71 14.47 35.54 17.39 -- 5.31 3.49 7.48 100.052005 1.88 15.24 14.65 36.17 14.94 -- 5.38 2.84 8.70 99.802006 1.98 14.80 14.24 35.13 12.78 -- 5.25 2.43 8.33 94.922007 1.73 12.52 12.69 36.96 10.88 -- 4.71 2.24 8.81 90.532008 1.63 10.12 10.50 34.23 12.85 -- 4.74 2.51 8.35 84.932009 1.50 8.60 8.41 30.73 12.92 -- 4.20 2.30 7.57 76.242010 1.18 8.71 7.48 29.07 12.22 -- 3.91 2.54 7.45 72.542011 1.07 8.75 4.32 23.20 9.30 -- 3.71 2.66 7.03 60.042012 1.15 7.74 4.61 18.73 7.21 -- 3.45 2.37 6.11 51.36


Table A2. Time-series of fi shery removals by regulatory Area (million lb, net wt.).

Year 2A 2B 2C 3A 3B 4 4A 4B 4C 4D 4E Total1974 0.52 4.62 5.60 8.19 1.67 0.71 NA NA NA NA NA 21.311975 0.46 7.13 6.24 10.60 2.56 0.63 NA NA NA NA NA 27.621976 0.24 7.28 5.53 11.04 2.73 0.72 NA NA NA NA NA 27.541977 0.21 5.43 3.19 8.64 3.19 1.22 NA NA NA NA NA 21.881978 0.10 4.61 4.32 10.30 1.32 1.35 NA NA NA NA NA 22.001979 0.05 4.86 4.53 11.34 0.39 1.37 NA NA NA NA NA 22.541980 0.02 5.65 3.24 11.97 0.28 0.71 NA NA NA NA NA 21.871981 0.20 5.66 4.01 14.23 0.45 NA 0.49 0.39 0.30 0.01 0.00 25.741982 0.21 5.54 3.50 13.52 4.80 NA 1.17 0.01 0.24 0.00 0.01 29.011983 0.27 5.44 6.38 14.13 7.76 NA 2.50 1.34 0.42 0.15 0.01 38.391984 0.43 9.05 5.87 19.77 6.69 NA 1.05 1.10 0.58 0.39 0.04 44.971985 0.50 10.49 9.42 21.77 11.09 NA 1.78 1.28 0.64 0.70 0.04 57.701986 0.59 11.43 11.04 34.66 9.22 NA 3.56 0.28 0.72 1.29 0.05 72.831987 0.60 12.42 11.05 32.89 8.10 NA 3.83 1.56 0.91 0.73 0.12 72.201988 0.49 12.91 11.57 39.41 7.20 NA 1.96 1.62 0.72 0.46 0.01 76.341989 0.48 10.48 9.73 35.19 8.04 NA 1.05 2.72 0.59 0.69 0.01 68.981990 0.34 8.69 10.06 29.96 8.91 NA 2.61 1.39 0.55 1.05 0.06 63.621991 0.36 7.26 9.03 24.07 12.35 NA 2.35 1.58 0.71 1.50 0.11 59.311992 0.44 7.68 10.06 27.43 8.80 NA 2.75 2.36 0.81 0.74 0.07 61.151993 0.51 10.72 11.48 23.08 7.92 NA 2.61 2.00 0.85 0.85 0.07 60.081994 0.37 9.98 10.61 25.69 3.90 NA 1.84 2.06 0.73 0.73 0.12 56.021995 0.30 9.66 7.82 18.46 3.13 NA 1.63 1.69 0.67 0.65 0.13 44.141996 0.30 9.58 8.92 19.87 3.69 NA 1.72 2.10 0.69 0.72 0.12 47.691997 0.42 12.46 9.96 24.71 9.12 NA 2.93 3.35 1.13 1.16 0.25 65.491998 0.46 13.23 10.24 25.85 11.22 NA 3.44 2.92 1.26 1.32 0.19 70.121999 0.46 12.75 10.21 25.43 13.91 NA 4.40 3.60 1.77 1.91 0.27 74.702000 0.49 10.84 8.48 19.33 15.47 NA 5.18 4.72 1.75 1.94 0.35 68.552001 0.68 10.33 8.44 21.60 16.37 NA 5.05 4.50 1.66 1.86 0.48 70.972002 0.86 12.11 8.63 23.27 17.35 NA 5.11 4.10 1.22 1.76 0.56 74.952003 0.82 11.82 8.44 22.82 17.26 NA 5.04 3.88 0.89 1.96 0.42 73.362004 0.88 12.20 10.27 25.24 15.48 NA 3.58 2.73 0.96 1.66 0.32 73.312005 0.81 12.37 10.66 26.19 13.20 NA 3.42 1.98 0.54 2.59 0.37 72.112006 0.83 12.04 10.51 25.77 10.80 NA 3.34 1.59 0.49 2.37 0.37 68.122007 0.79 9.80 8.50 26.55 9.27 NA 2.84 1.42 0.55 2.73 0.58 63.032008 0.68 7.78 6.22 24.58 10.75 NA 3.03 1.77 0.73 2.56 0.60 58.702009 0.49 6.66 4.97 21.80 10.80 NA 2.54 1.60 0.65 2.22 0.46 52.182010 0.42 6.76 4.50 20.52 10.13 NA 2.33 1.84 0.79 2.12 0.41 49.832011 0.55 6.72 2.46 14.70 7.33 NA 2.36 2.06 0.79 2.19 0.46 39.612012 0.59 5.93 2.70 11.96 5.08 NA 1.58 1.71 0.58 1.42 0.32 31.87


Table A2. Time-series of setline survey WPUE by regulatory Area (O32; net lb/skate).

Year 2A 2B 2C 3A 3B 4A 4B 4D 4IC 4ID 4S 4N 4CDE Total1974 NA NA NA NA NA NA NA NA NA NA NA NA NA NA1975 NA NA NA NA NA NA NA NA NA NA NA NA NA NA1976 NA NA NA NA NA NA NA NA NA NA NA NA NA NA1977 NA 13.7 NA 58.4 NA NA NA NA NA NA NA NA NA NA1978 NA 19.1 NA 26.9 NA NA NA NA NA NA NA NA NA NA1979 NA NA NA 41.0 NA NA NA NA NA NA NA NA NA NA1980 NA 25.5 NA 76.2 NA NA NA NA NA NA NA NA NA NA1981 NA 16.5 NA 131.4 NA NA NA NA NA NA NA NA NA NA1982 NA 20.6 113.7 130.3 NA NA NA NA NA NA 5.5 0.0 NA NA1983 NA 18.0 142.2 119.0 NA NA NA NA NA NA 4.3 0.2 NA NA1984 NA 57.4 259.6 361.2 NA NA NA NA NA NA 6.4 0.5 NA NA1985 NA 41.7 260.5 377.5 NA NA NA NA NA NA 5.8 0.7 NA NA1986 NA 37.8 282.6 305.1 NA NA NA NA NA NA 7.3 0.2 NA NA1987 NA NA NA NA NA NA NA NA NA NA 8.1 0.3 NA NA1988 NA NA NA NA NA NA NA NA NA NA 17.3 0.2 NA NA1989 NA NA NA NA NA NA NA NA NA NA 10.6 0.2 NA NA1990 NA NA NA NA NA NA NA NA NA NA 12.3 0.7 NA NA1991 NA NA NA NA NA NA NA NA NA NA 11.0 2.0 NA NA1992 NA NA NA NA NA NA NA NA NA NA 8.6 0.8 NA NA1993 NA 95.7 NA 261.1 NA NA NA NA NA NA 19.4 4.8 NA NA1994 NA NA NA 253.9 NA NA NA NA NA NA 14.7 3.7 NA NA1995 31.0 159.1 NA 300.1 NA NA NA NA NA NA 16.3 3.5 NA NA1996 33.7 165.9 306.2 317.3 352.2 NA NA NA NA NA 23.5 18.1 NA NA1997 36.4 144.1 410.6 330.6 413.9 245.4 281.6 111.2 111.2 111.2 19.3 4.0 23.2 138.21998 37.3 83.3 234.8 281.2 434.9 299.0 215.6 299.0 299.0 299.0 25.6 6.5 44.5 133.91999 38.3 88.1 208.9 240.7 437.9 290.3 203.1 290.3 290.3 290.3 25.8 0.0 42.1 126.12000 40.6 91.2 240.1 271.7 373.1 275.8 216.3 212.9 212.9 212.9 18.7 2.9 31.5 120.62001 43.0 101.1 244.1 256.1 357.1 198.8 171.3 196.8 196.8 196.8 20.0 4.5 31.3 112.32002 34.5 91.8 267.8 299.4 297.2 168.4 119.1 262.5 262.5 262.5 11.6 2.1 31.1 108.82003 22.8 72.7 228.3 229.3 261.6 154.1 104.1 194.8 194.8 194.8 17.0 3.5 29.0 91.62004 27.9 85.8 176.0 269.7 236.3 137.4 73.3 131.9 131.9 131.9 17.0 3.1 23.3 88.42005 29.0 71.9 174.7 275.9 211.2 106.8 86.2 69.2 69.2 69.2 16.2 3.4 17.4 82.12006 16.8 58.7 146.7 232.5 181.2 84.9 95.5 54.4 60.5 111.2 16.8 3.3 17.0 71.12007 19.4 57.2 142.6 211.6 191.3 66.5 87.2 58.6 46.1 50.9 11.9 2.7 13.3 65.82008 19.1 89.8 106.4 189.1 126.0 84.1 103.3 77.5 69.7 15.4 8.2 2.5 12.1 60.22009 8.3 86.3 115.6 148.8 113.0 84.1 106.8 78.4 53.1 20.4 11.6 3.3 14.5 55.42010 17.3 88.8 110.3 117.1 91.4 73.0 68.4 48.0 55.3 57.8 12.2 3.2 13.1 47.02011 27.0 79.8 136.3 120.5 79.8 58.4 67.9 33.5 51.5 14.3 9.7 3.2 10.1 44.72012 28.5 103.4 161.4 137.0 86.8 63.6 48.3 36.2 37.0 1.4 11.5 3.7 11.2 49.9


Table A2. Time-series of fi shery WPUE by regulatory Area (net lb/skate).

Year 2A 2B 2C 3A 3B 4A 4B 4C 4D 4E Total1974 59 64 57 65 57 NA NA NA NA NA NA1975 59 68 53 66 68 NA NA NA NA NA NA1976 33 53 42 60 65 NA NA NA NA NA NA1977 83 61 45 61 73 NA NA NA NA NA NA1978 39 63 56 78 53 NA NA NA NA NA NA1979 50 48 80 86 37 NA NA NA NA NA NA1980 37 65 79 118 113 NA NA NA NA NA NA1981 33 67 144 142 160 158 99 110 NA NA NA1982 22 69 146 168 203 103 NA 91 NA NA NA1983 NA NA NA NA NA NA NA NA NA NA NA1984 63 147 284 502 474 366 161 NA 197 NA 3411985 62 139 345 500 592 337 234 594 330 NA 3801986 55 118 290 506 506 260 238 427 218 NA 3421987 53 130 260 498 478 342 220 384 241 NA 3441988 134 137 281 503 654 453 224 371 201 NA 3871989 113 133 258 457 590 409 268 333 432 NA 3811990 168 176 270 354 484 418 209 288 381 NA 3321991 158 149 233 319 466 471 329 223 399 NA 3331992 117 171 230 397 440 372 280 249 412 NA 3391993 147 208 256 393 514 463 218 257 851 NA 3991994 93 215 207 354 377 463 197 167 480 NA 3281995 116 219 234 417 476 349 189 286 475 NA 3511996 159 227 239 473 557 515 269 297 543 NA 4151997 226 241 246 458 563 483 275 335 671 NA 4231998 194 232 236 452 611 525 287 287 627 NA 4291999 342 213 199 437 538 497 310 271 535 NA 3982000 263 229 187 443 579 548 320 223 556 NA 4172001 171 227 196 469 431 474 270 203 511 NA 3822002 181 223 244 508 399 402 245 148 503 NA 3802003 173 221 233 485 365 355 196 105 388 NA 3462004 143 203 240 486 328 315 202 120 445 NA 3382005 137 195 203 446 293 301 238 91 379 NA 3142006 156 201 170 403 292 241 218 72 280 NA 2832007 96 198 160 398 257 206 230 65 237 NA 2682008 69 174 161 370 234 206 193 94 247 NA 2492009 98 188 155 318 211 234 189 88 249 NA 2362010 149 222 158 285 173 182 142 82 188 NA 2102011 92 240 175 280 140 189 165 75 166 NA 2092012 120 259 208 269 134 191 154 60 162 NA 209


IPHC Staff harvest advice and regulatory proposals: 2013

Bruce M. Leaman, Ian J. Stewart, Raymond A. Webster, and Heather L. Gilroy

Introduction

In providing advice on stock management for 2013, the staff has been cognizant of direction from the Commission concerning the incorporation of uncertainty in advice, the recommendations from the 2012 Performance Review (http://www.iphc.int/meetings-and-events/review.html), and the results of the 2012 external peer review of the IPHC stock assessment and management framework (Stewart et al. 2013a). The 2012 stock assessment (Stewart et al. 2013b) provides a new format for the staff’s management advice, compared with that for previous years. Previous staff advice was provided as a single recommended catch limit for each Regulatory Area. Staff advice for 2013 takes the form of a risk-benefi t decision table, wherein risks of negative impacts on stock and fi shery performance are associated with the benefi ts of particular choices of harvest level. This format for decision making more fully refl ects uncertainty and allows the Commission to weigh the risks and benefi ts of management choices, as well as overall harvest policy, when deciding on catch limits. A signifi cant aspect of this advice table is that it is structured at the coastwide level, rather than at the level of individual regulatory areas. This orientation results from the necessary defi nition of harvest policy reference points (threshold and limit) at the coastwide level. Detailed results of the stock assessment and the 2012 peer review are reported in the 2012 Report of Assessment and Research Activities (http://www.iphc.int/library/raras/308-rara-2012.html).

Updated assessments

Late season data from 2011, which were unavailable for the 2011 assessment, revised the initial estimate of 2012 exploitable biomass (Ebio) (260 Mlb) from the selected 2011 assessment model downward by approximately 3% to 252 Mlb. This adjustment is similar to that seen in recent years. In the absence of a new version of the assessment model, free of retrospective bias, the beginning of 2012 estimated Ebio as estimated using the 2011 model in 2012 would be lower by an additional 13%, to 219 Mlb. However, we know these estimates to be incorrect because of the retrospective bias. The 2012 stock assessment corrects this bias and the revised estimate of Ebio for the beginning of 2012 is 179 Mlb. Updating the corrected assessment with 2012 survey and catch data results in an estimated Ebio for the beginning of 2013 of 186 Mlb. These changes and estimates are detailed in Table 1.

Coastwide commercial fi shery weight per unit effort (WPUE) in 2012 was unchanged from the 2011 value and has been stable for the past three years. However, there were notable positive changes in commercial WPUE in Area 2 and modest negative changes in Area 3 and most of Area 4. The 2012 unadjusted IPHC stock assessment survey WPUE showed a coastwide increase of 12%, the fi rst increase since the current expanded survey began in 1996. Only Area 4B showed a decline in survey WPUE for 2012.

This document develops Regulatory Area harvest options consistent with three of the row-specifi c harvest choices in the coastwide decision table (Table 2; Stewart et al. 2013). The three rows of the decision table upon which the Commission requested additional details at its 2012 Interim Meeting correspond to Total Removals of 35 Mlb, 40.2 Mlb (the Blue Line), and 45 Mlb.


Some elements of these options (e.g., commercial fi shery wastage) are based on scaled 2012 values corresponding to the relationship of 2012 removals and those identifi ed in the harvest options. For example, if the chosen harvest option results in a Fishery Constant Exploitation Yield (FCEY) that is 70% of that for 2012, the projected removal for commercial fi shery wastage would be scaled to 70% of that estimated for 2012.

Biomass apportionment among Regulatory Areas and harvest scenarios

Staff recommends continuing with the same basis for apportionment of the coastwide Ebio estimate into regulatory area Ebio estimates as was used in 2012. That is, the coastwide estimate is apportioned using survey WPUE for each regulatory area, adjusted for hook competition (catchability) and survey timing differences, averaged using a three-year, reverse weighting formula, and multiplied by the 0-400 fm bottom area for each regulatory area (Webster and Stewart 2013, Webster 2011, Webster and Hare 2010).

The combination of hook competition (Fig. 1) and survey timing adjustments (Fig. 2), together with the reverse-weighting of the three most recent index values, results in adjusted survey WPUE values that differ, sometimes appreciably, from the unadjusted values (Fig. 3). This distinction is important because the apportionment formula that is used to estimate exploitable biomass distribution among regulatory areas uses the adjusted survey WPUE index (weighted by bottom area). The areas at the extreme ends of the halibut range (Areas 2A, 4B) generally show the greatest differences between raw and adjusted WPUE values (Fig. 3). Compared with the biomass distribution at the beginning of 2012, Areas 4B and 3B had lower proportions of the coastwide total while Areas 2C, 3A, and 4A had higher proportions (Fig. 4). We also present the historical apportionment-based estimates of Ebio distribution from Webster and Stewart (2013) for reference (Fig. 5). Of note in this fi gure are the increasing estimates of Ebio in Area 2 since harvest rates were reduced under the coastwide assessment approach.

Following apportionment of the coastwide Ebio to regulatory areas under each harvest scenario, a Total Constant Exploitation Yield (TCEY) is obtained by applying the target harvest rate for each regulatory area (21.5% for Areas 2 and 3A; 16.125% for Areas 3B and 4) to the Ebio estimates. For alternative levels of coastwide harvest, the TCEY is adjusted based on the relative change from the Blue Line in Table 2, uniformly across all areas (Table 3). Area-specifi c harvest rates from the adjusted TCEY values are shown in Table 4.

Other removals (Table 5) are subtracted from the TCEY to obtain the FCEY values by regulatory area, assuming all other removals (except wastage coastwide, plus sport catch in Areas 2A and 2B, and treaty subsistence in Area 2A, which are scaled proportional to the FCEY because of catch sharing plans) remain the same as in 20121. One set of FCEY values for 2013, consistent with the current harvest policy, corresponds to the “Blue Line” results from the stock assessment decision-making table. These area specifi c values are reported in Table 6.

At the request of the Commissioners during the Interim Meeting, alternatives to the Blue Line were added to the preliminary decision-making table, with smaller increments of total mortality. Fully apportioned results for these rows from the decision table with 35 and 45 Mlb of total mortality are presented in Tables 7 and 8. The Blue Line apportionment table is also presented assuming the full sport-charter Guideline Harvest Level (GHL) is removed in 2013 rather than

1 Historical tables of estimated FCEY, staff catch limit recommendations, and Commission-adopted Catch Limits are presented in Tables A1 and A2.


the estimated catch from 2012 which was used in the initial calculations (Table 9). The only change from Table 3 is increased sport removals in Area 2C consistent with a 0.788 Mlb GHL. The apportionment results for the 35 Mlb mortality row with full GHL removals are presented in Table 10. The GHL for Area 2C in this alternative is 0.788 Mlb, and it is 2.01 Mlb for Area 3A. At the 45 Mlb total mortality level, the GHL for 2C remains unchanged, but the GHL for Area 3A increases to 2.73 Mlb (Table 11). The North Pacifi c Fishery Management Council has indicated (Fig. 6) that similar management measures for charter fi shing will be in place for 2013 as were employed in 2012.

While the choice of harvest options from the risk-benefi t decision table is the purview of the Commission, the staff notes that an understanding of future stock and fi shery behavior is dependent on the degree to which a consistent harvest policy is implemented. As such, attention is drawn to the impact of harvest choices if catch limit decisions are made uniquely at the level of Regulatory Areas, rather than at a coastwide level in conformance with a harvest policy that has been evaluated. Additional analyses can be conducted to determine the coastwide harvest rates and the implied harvest policy that would result under such scenarios.

Fishing seasons

As in the past years, the staff recommends March 15 to November 15 opening and closing dates for the quota share fi shing season. This recommendation is a compromise between minimizing interceptions of migrating fi sh and providing opportunity for market presence of wild halibut. All Area 2A commercial fi sheries should also occur within this period.

For the Area 2A directed commercial fi shery, the staff recommends an opening pattern similar to 2012, starting the last week of June with a series of 10-hour periods, with fi shing period limits. Therefore we recommend the following series for 2013: June 26, July 10, July 24, August 7, August 21, September 4, and September 18. The size of the fi shing period limits will be determined when more information is available on fl eet participation.

Catch sharing plans: Areas 2A, 2B, and 4CDE

The Commission does not make allocative decisions within regulatory areas or among dif-ferent user groups. However, for Areas 2A and 4CDE the staff recommends that the Commission endorse the catch sharing plans developed by the Pacifi c and North Pacifi c Fishery Management Councils for these areas, respectively. Similarly, the staff recommends that the catch sharing al-location of 85% commercial and 15% recreational, developed by DFO for Area 2B, be endorsed.

Proposed changes to the IPHC regulations

Area 2C sport fi shing regulations for the charter vesselsThe IPHC has received a request from the North Pacifi c Fishery Management Council

(Appendix 1) concerning management measures to restrict the charter halibut harvest in Area 2C, in order to stay within the Council’s (GHL). The request is to maintain the current regulation of the reverse slot limit allowing the retention of one fi sh, ≤45 inches or ≥68 inches in length, with head on. In addition, as in the past, if the halibut was fi lleted the entire carcass must be retained on board the vessel until all fi llets are offl oaded.


References

Stewart, I.J., Martell, S., Webster, R.A., Forrest, R., Ianelli, J., and Leaman, B.M. 2013a. Assessment review team meeting, October 24-26, 2012. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2012: 239-266.

Stewart, I.J., Leaman, B.M., Martell, S. and Webster, R.A. 2013b. Assessment of the Pacifi c halibut stock at the end of 2012. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2012: 93-186.

Webster, R.A. 2011. Weighted averaging of recent survey indices. Int. Pac. Halibut Comm. Report of Assessmnent and Research Activities 2010: 241-250.

Webster, R.A. and Hare, S.R. 2010. Adjusting IPHC setline survey WPUE for survey timing and hook competition. Int. Pac. Halibut Comm. Report of Assessmnent and Research Activities 2010: 251-260.

Webster, R.A. and Stewart, I. J. 2013. Apportionment and regulatory area harvest calculations. Int. Pac. Halibut Comm. Report of Assessmnent and Research Activities 2012: 187-205.


Table 1. Bridging analysis showing relationship of original 2011, updated 2011, and updated 2012 estimates using the 2011 model variant, and new assessment estimates using the 2012 model variant which corrects retrospective bias. All weights in Mlb (net wt.).

Model 2011 (wobblesq) 2012

End year 2011 2011 2012 2012

Data fi nalized in November2011

November2012

November2012

November2012

Quantity

2012 Spawning biomass 319 309 272 197

2012 Relative spawning bio-mass 42% 41% 38% 34%

2013 Spawning biomass -- -- 324 201

2013 Relative spawning bio-mass -- -- 46% 35%

2012 Exploitable biomass 260 252 219 179

2013 Exploitable biomass -- -- 258 186

2012 Coastwide harvest rate 19.4% 18.9% 21.8% 26.7%


Table 2. Risk-benefi t decision table. Values indicate the probability of the outcome in each column given the level of removals for that row.

Coastwide Fishery CEY


Fishing intensity Stock status Stock trend

Catch trend

Effective coastwide

HR Spawning biomass

Fishery CEY

2013 2014 2016 2014is

greater than

target

isless than30%

isless than20%

isless than2013

is 5% less than2013

isless than2013

isless than2013

0.0 (0.0) 0% 25% <1% 23% <1% 41% 0%0.0 (16.5) <1% 25% <1% 76% 2% 95% 0%3.4 (20.0) <1% 25% <1% 77% 2% 96% <1%

12.9 (30.0) 1% 25% <1% 79% 2% 97% 1%17.7 (35.0) 23% 25% <1% 80% 2% 97% 19%22.7 (40.2) 50% 25% <1% 82% 3% 97% 48%27.3 (45.0) 75% 25% <1% 83% 3% 98% 75%32.1 (50.0) 84% 25% <1% 84% 3% 98% 85%36.2 (54.3) 97% 25% <1% 85% 4% 98% 97%41.6 (60.0) >99% 25% <1% 86% 4% 99% >99%

a b c d e f g


Table 3. Apportioned distribution of total CEY (M lb) for alternative levels of mortality (from Table 2). The middle row represents the Blue Line results (Stewart et al. 2013). All biomass values are reported in millions of net pounds.

2A 2B 2C 3A 3B 4A 4B 4CDETotal CEY

Coastwide Har-vest Rate

0.29 1.86 1.77 5.34 1.48 0.68 0.38 1.12 12.93 6.9%0.37 2.37 2.25 6.80 1.89 0.87 0.49 1.43 16.46 8.8%0.59 3.81 3.61 10.92 3.03 1.39 0.79 2.30 26.43 14.2%0.82 5.28 5.00 15.13 4.20 1.93 1.09 3.18 36.63 19.6%1.04 6.67 6.33 19.13 5.30 2.44 1.38 4.02 46.30 24.8%1.13 7.30 6.92 20.92 5.80 2.67 1.50 4.40 50.65 27.2%1.26 8.11 7.70 23.27 6.45 2.97 1.67 4.90 56.34 30.2%

Table 4. Area-specifi c harvest rates (%) based on alternative levels of mortality (from Table 2). The middle row represents the Blue Line results (Stewart et al. 2013), and the current harvest policy rates by management area.

2A 2B 2C 3A 3B 4A 4B 4CDETotal CEY

Coastwide Harvest Rate

7.6 7.6 7.6 7.6 5.7 5.7 5.7 5.7 12.93 6.99.7 9.7 9.7 9.7 7.2 7.2 7.2 7.2 16.46 8.8

15.5 15.5 15.5 15.5 11.6 11.6 11.6 11.6 26.43 14.221.5 21.5 21.5 21.5 16.1 16.1 16.1 16.1 36.63 19.627.2 27.2 27.2 27.2 20.4 20.4 20.4 20.4 46.30 24.829.7 29.7 29.7 29.7 22.3 22.3 22.3 22.3 50.65 27.233.1 33.1 33.1 33.1 24.8 24.8 24.8 24.8 56.34 30.2

Table 5. “Other removals” (026; Mlb, net) observed in 2012. Note that these differ from the values deducted from the TCEY in apportionment calculations due to the scaling of commercial fi shery wastage relative to 2012 values.

2A 2B 2C 3A 3B 4A 4B 4CDE TotalBycatch 0.10 0.18 0.01 1.26 1.11 0.99 0.44 2.25 6.34Wastage 0.01 0.17 0.08 0.56 0.47 0.08 0.03 0.08 1.48Sport NA NA 1.41 3.94 0.01 0.02 0.00 0.00 5.38Personal/subsistence NA 0.41 0.43 0.31 0.02 0.02 0.00 0.03 1.21Total 0.11 0.75 1.91 6.07 1.61 1.11 0.48 2.36 14.40


Table 6. Apportionment results for the Blue Line current harvest policy approach. All biomass values are reported in millions of net pounds.

2A 2B 2C 3A 3B 4A 4B 4CDE TotalExploitable biomass 3.81 24.54 23.28 70.38 26.02 11.97 6.75 19.74 186.49Percent of total 2.0 13.2 12.5 37.7 14.0 6.4 3.6 10.6 100.0Harvest rate (%) 21.5 21.5 21.5 21.5 16.1 16.1 16.1 16.1 19.6Total CEY 0.82 5.28 5.00 15.13 4.20 1.93 1.09 3.18 36.63Other removals (O26) 0.11 0.69 1.89 5.89 1.46 1.08 0.47 2.33 13.93Fishery CEY 0.71 4.58 3.12 9.24 2.73 0.85 0.62 0.85 22.70

Table 7. Apportionment results for the Table 2 row with 35 Mlb total mortality, assuming 2011 levels of Other removals. All biomass values are reported in millions of net pounds.

2A 2B 2C 3A 3B 4A 4B 4CDE TotalExploitable biomass 3.81 24.54 23.28 70.38 26.02 11.97 6.75 19.74 186.49Percent of total 2.0 13.2 12.5 37.7 14.0 6.4 3.6 10.6 100.0Harvest rate (%) 18.4 18.4 18.4 18.4 13.8 13.8 13.8 13.8 16.8Total CEY 0.70 4.52 4.29 12.97 3.60 1.65 0.93 2.73 31.40Other removals 0.11 0.67 1.88 5.81 1.39 1.07 0.46 2.32 13.71Fishery CEY 0.59 3.85 2.41 7.16 2.20 0.58 0.47 0.41 17.69

Table 8. Apportionment results for the Table 2 row with 45 Mlb total mortality, assuming 2011 levels of Other removals. All biomass values are reported in millions of net pounds.



Table 9. Apportionment results for the Blue Line, assuming the full GHL is removed in Areas 2C and 3A. All biomass values are reported in millions of net pounds.


Table 10. Apportionment results for the Table 2 row with 35 Mlb total mortality, assuming the full GHL is removed in Areas 2C and 3A. All biomass values are reported in millions of net pounds.


Table 11. Apportionment results for the Table 2 row with 45 Mlb total mortality, assuming the full GHL is removed in Areas 2C and 3A. All biomass values are reported in millions of net pounds.



Figure 1. Hook competition adjustment factors that are applied to setline survey WPUE for biomass apportionment.


Figure 2. Survey timing adjustment factors that are applied to setline survey WPUE for biomass apportionment, by area and year.


Figure 3. Comparison of survey WPUE without adjustments (Raw WPUE), with both timing and hook competition adjustments applied (Adj WPUE), and with adjustments and a 75:20:5 weighting of the three most recent years’ values (Adj with KF).


Figure 4. Distribution of exploitable biomass, as estimated in the 2012 stock assessment, at the beginning of 2012 (left) and 2013 (right).

Figure 5. Apportionment-based exploitable biomass estimates by year and area, calculated by applying the proportions in Table 1 of Webster and Stewart (2013) to the 2012 stock assessment estimates of coastwide Ebio.


Figure 6. Letter concerning management measures for charter vessel recreational fi shing in Areas 2C and 3A.


Tabl

e A1.

Est

imat

e d fi

sher

y C

EY,

staf

f rec

omm

ende

d ca

tch

limits

, and

app

rove

d ca

tch

limits

of P

acifi

c hal

ibut

by

IPH

C re

gula

tory

ar

ea (i

n th

ousa

nds o

f pou

nds,

net w

eigh

t), 2

003-

2012

.

Reg

ulat

ory

Are

aEs

timat

ed F

isher

y C

EY20

0320

041

2005

2006

2007

220

0820

0920

1020

1120

122A

31,

290

1,81

01,

170

1,49

066

065

050

057

01,

120

1,15

02B

11,3

2015

,780

412

,700

413

,200

46,

2204

4,65

044,

9204

5,55

047,

9404

6,63

04

2C9,

110

17,0

3011

,800

10,3

304,

980

3,92

02,

860

2,39

02,

330

3,21

03A

34,2

2029

,980

26,3

0024

,940

27,6

3022

,250

20,8

4018

,280

14,3

6011

,920

3B29

,190

15,6

0010

,700

8,57

016

,770

14,2

7013

,200

8,91

07,

510

5,07

04A

11,2

203,

470

3,40

03,

250

5,23

03,

510

2,20

02,

120

2,57

01,

570

4B7,

760

2,81

01,

700

1,07

02,

560

2,70

02,

090

2,75

02,

210

1,87

04C

DE

13,8

203,

390

4,40

03,

110

3,85

03,

680

1,97

03,

820

3,99

02,

470

Tota

l11

7,93

089

,870

72,1

7065

,960

67,9

0055

,630

48,5

8044

,390

42,0

3033

,890

Reg

ulat

ory

Are

aSt

aff R

ecom

men

datio

ns20

0320

0420

0520

0620

072

2008

2009

2010

2011

2012

2A3

1,31

01,

480

1,33

01,

380

1,02

01,

000

860

760

910

990

2B11

,750

13,8

004

13,2

504

13,2

204

9,72

048,

0604

6,96

046,

5904

7,65

046,

6304

2C8,

500

11,3

1010

,930

10,6

307,

810

6,21

04,

540

3,71

02,

330

2,62

03A

22,6

3025

,060

25,4

7025

,200

26,0

1024

,220

22,5

3019

,990

14,3

6011

,920

3B17

,130

15,6

0013

,150

10,8

6012

,830

10,9

0011

,670

9,90

07,

510

5,07

04A

4.97

03,

470

3,44

03,

350

3,98

03,

100

2,65

02,

330

2,41

01,

570

4B4,

180

2,81

02,

260

1,67

01,

970

1,86

01,

940

2,16

02,

180

1,87

04C

DE

4,45

03,

390

3,99

03,

550

3,65

03,

890

2,93

03,

580

3,72

02,

470

Tota

l74

,920

76,9

2073

,820

69,8

6066

,990

59,2

4054

,080

49,0

2041

,070

33,1

40Re

gula

tory

Are

aC

atch

Lim

its20

0320

0420

0520

0620

072

2008

2009

2010

2011

2012

2A3

1,31

01,

480

1,33

01,

380

1,34

01,

220

950

810

910

989

2B11

,750

13,8

004

13,2

504

13,2

204

11,4

704

9,00

047,

6304

7,50

047,

6504

7,03

84

2C8,

500

10,5

0010

,930

10,6

308,

510

6,21

05,

020

4,40

02,

330

2,62

43A

22,6

3025

,060

25,4

7025

,200

26,2

0024

,220

21,7

0019

,990

14,3

6011

,918

3B17

,130

15,6

0013

,150

10,8

609,

220

10,9

0010

,900

9,90

07,

510

5,07

04A

4.97

03,

470

3,44

03,

350

2,89

03,

100

2,55

02,

330

2,41

01,

567

4B4,

180

2,81

02,

260

1,67

01,

440

1,86

01,

870

2,16

02,

180

1,86

94C

DE

4,45

03,

785

3,98

93,

550

4,10

03,

890

3,46

03,

580

3,72

02,

465

Tota

l74

,920

76,5

0573

,819

69,8

6065

,170

60,4

0054

,080

50,6

7041

,070

33,5

401 S

taff

catc

h lim

it re

com

men

datio

ns re

vise

d af

ter B

lueb

ook

base

d on

CEY

har

vest

polic

y2 E

stim

ated

fi sh

ery

CEY

and

staff

reco

mm

enda

tions

from

coas

twid

e sto

ck as

sess

men

t with

surv

ey p

artit

ioni

ng to

area

as p

rese

nted

in th

e 200

7 Ann

ual M

eetin

g H

ando

ut. T

he cl

osed

area

stoc

k as

sess

men

t pro

duce

d sta

ff re

com

men

datio

ns b

y ar

ea in

mill

ions

of p

ound

s of:

Are

a 2A

= 1.

34; A

rea 2

B =

11.4

7; A

rea 2

C =

8.51

; Are

a 3A

= 26

.20;

Are

a 3B

= 9.

22; A

rea 4

A =

2.89

; and

4B

= 1.

44 an

d Are

a 4CD

E =

4.10

. The

Com

miss

ion

dete

rmin

ed th

e 20

07 c

atch

lim

its u

sing

reco

mm

enda

tions

bas

ed o

n th

e cl

osed

are

a sto

ck a

sses

smen

t.

3 Are

a 2A

incl

udes

spor

t cat

ch a

nd tr

eaty

Indi

an c

atch

4 A

rea

2B in

clud

es sp

ort c

atch


Tabl

e A2.

Est

imat

ed fi

sher

y C

EY,

staf

f rec

omm

ende

d ca

tch

limits

, and

app

rove

d ca

tch

limits

of P

acifi

c hal

ibut

by

IPH

C re

gula

tory

ar

ea (i

n th

ousa

nds o

f pou

nds,

net w

eigh

t), 1

993

- 200

2.

Reg

ulat

ory

Est

imat

ed F

ishe

ry C

EY

Are

a19

9319

941

1995

219

9619

9719

9819

9920

0020

0120

022A

346

049

052

093

01,

050

690

830

1,14

041,

310

2B9,

810

8,32

09,

520

15,9

9015

,380

11,2

107,

850

9,99

0411

,750

2C10

,410

12,6

608,

540

Skip

ped

11,4

1015

,480

10,4

906,

310

8,78

08,

500

3A23

,130

27,0

2016

,870

Betw

een

33,5

5038

,710

24,6

7011

,940

21,8

9024

,140

3B4,

070

3,58

03,

660

Mod

els

11,4

9030

,990

26,8

3018

,360

25,4

6028

,560

4A11

,110

8,42

06,

420

9,82

011

,960

4BA

rea

4 =

Are

a 4

=A

rea

4 =

Are

a 4

= 10

,210

6,71

06,

770

10,0

607,

510

4CD

E5,

590

5,00

05,

920

25,2

9013

,280

9,80

04,

130

7,63

011

,810

Tota

l53

,470

57,0

7045

,030

98,6

6013

6,21

098

,820

62,6

1094

,770

105,

540

Reg

ulat

ory

Staf

f Rec

omm

enda

tion

Are

a19

9319

9419

9519

9619

9719

9019

9920

0020

0120

022A

346

050

045

052

070

042

069

083

01,

1404

1,31

02B

9,81

09,

500

8,50

09,

520

12,5

007,

800

11,2

109,

970

9,99

0411

,750

2C10

,410

12,0

008,

500

9,00

010

,000

8,00

010

,490

8,40

08,

780

8,50

03A

23,1

3026

,000

20,0

0020

,000

25,0

0033

,000

24,6

7018

,310

21,8

9022

,630

3B4,

070

4,00

03,

700

3,70

09,

000

6,40

013

,370

15,0

3118

,500

17,1

304A

2,02

01,

800

2,00

01,

950

3,00

01,

300

4,24

04,

970

4,97

04,

970

4B2,

020

2,10

01,

600

2,31

03,

200

1,30

03,

980

4,91

04,

910

3,44

04C

DE

1,52

01,

500

2,30

01,

660

2,80

080

54,

130

4,13

04,

450

4,45

0To

tal

53,4

4057

,400

47,0

5048

,660

66,2

0059

,025

72,7

8066

,550

74,6

3074

,180

Regu

lato

ryC

atch

lim

itsA

rea

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2A3

600

550

520

520

700

820

760

830

1,14

01,

310

2B10

,500

10,0

009,

520

9,52

012

,500

13,0

0012

,100

10,6

0010

,510

11,7

502C

10,0

0011

,000

9,00

09,

000

10,0

0010

,500

10,4

908,

400

8,78

08,

500

3A20

,700

26,0

0020

,000

20,0

0025

,000

26,0

0024

,670

18,3

1021

,890

22,6

303B

6,50

04,

000

3,70

03,

700

9,00

011

,000

13,3

7015

,030

16,5

3017

,130

4A2,

020

1,80

01,

950

1,95

02,

940

3,50

04,

240

4,97

04,

970

4,97

04B

2,30

02,

100

2,31

02,

310

3,48

03,

500

3,98

04,

910

4,91

04,

180

4CD

E1,

720

1,50

01,

660

1,66

02,

580

3,50

04,

450

4,45

04,

450

4,45

0To

tal

54,3

4056

,950

48,6

6048

,660

66,2

0071

,820

74,0

6067

,500

73,1

8074

,920

1 Av

erag

e of

stan

dard

and

alte

rnat

ive

(con

serv

ativ

e) a

sses

smen

ts (1

994)

2 Fro

m 1

995

on, C

EY b

ased

on

proj

ecte

d ra

ther

than

lagg

ed e

bio

3 Are

a 2A

incl

udes

spor

t cat

ch a

nd tr

eaty

Indi

an c

atch

4 W

ith th

e low

er se

ries o

f Are

a 2B

spor

t cat

ch es

timat

es (i

nclu

ding

.887

Mlb

in 2

001)

, Are

a 2A

B ex

ploi

tabl

e bio

mas

s is 6

6.71

. With

11%

of t

he to

tal i

n A

rea 2

A, t

he 2

001

setli

ne C

EY is

1.1

2 M

lb in

2A

and

10.5

1 M

lb in

2B.

With

a sp

ort c

atch

esti

mat

e of

1.5

8 M

lb in

200

1, th

e ex

ploi

tabl

e bi

omas

s for

2A

B is

67.6

2 M

lb a

nd th

e 20

01 se

tline

CEY

is 1

.14

M lb

in 2

A an

d 9.

99 M

lb in

2B

Sour

ce –

IPH

C A

nnua

l Mee

ting

Briefi n

g,H

Gilr

oy, J

an 2

013.


Optimal harvest rates for Pacifi c halibut

Steven Martell, Ian J. Stewart, and Bruce M. Leaman

Abstract

Key to determining the optimal harvest rate that maximizes the long-term sustainable yield is having knowledge of stock productivity and what age, or size, of fish are being harvested by all fishing sectors. Recent size-at-age data and changes in the stock assessment model for Pacific halibut have resulted in changes in estimated productivity and selectivity for both the directed fishery and the setline survey. Previous estimates of optimal exploitation rates are outdated and need to be revised. An age- sex- and size-structured equilibrium model is used to investigate how variation in growth, changes in size-selectivity, size-limits, and bycatch impact estimates of optimal exploitation rates in each of the Regulatory Areas for Pacific halibut. Differences in growth rates between regulatory areas translate into as much as a 40% difference in yield per recruit. A maximum size limit of 140 cm results in little to no benefit of protecting spawning stock biomass under the current assumed discard mortality rate of 16%. In the absence of size limits, over all yield per recruit increases but at the expense of reducing spawning biomass and changing the composition of the catch towards smaller fish. Optimal harvest rates would have to be adjusted downwards to compensate for the loss of spawning biomass. Undetected shifts in commercial selectivity towards capturing smaller/younger fish results in an upward bias in the estimates of optimal harvest rates. A reduction in bycatch rates in non-directed fisheries can lead to desirable increases in optimal exploitation rates. Conversely, unreported bycatch results in an upward bias in optimal exploitation rates. The optimal harvest rate is very sensitive to estimates of natural mortality, steepness of the stock-recruit relationship, and fisheries selectivity. The harvest control rule should be updated with changes in model structure and updated parameter estimates to ensure the harvest policy is kept up to date. Results presented are preliminary and include a number of simplifying assumptions. Recommendations concerning revisions to the optimum harvest rate will be developed over the coming year.

Introduction

The current harvest objectives for Pacific halibut are to achieve a high level of yield while at all times maintaining healthy female spawning stock biomass to ensure long-term sustainability. The current harvest policy intends to achieve a desired exploitation rate of 21.5% in regulatory Areas 2A, 2B, 2C, and 3A, and a lower exploitation rate of 16.125% in Areas 3B, 4A, 4B, and 4CDE. These harvest rates were determined using a simulation modeling approach based on the results of closed area assessments in Areas 2B, 2C and 3A (Clark and Hare, 2006) and subsequent considerations of surplus production and yield per recruit calculations in western regulatory areas, as well as, a revised definition of exploitable biomass. Since that time, there have been several significant changes to the assessment model and regulatory area management that warrant a review and update to the current harvest policy for Pacific halibut. Most significant of these aforementioned changes was the transition from several area-based assessments to a single coast-wide assessment model and apportionment. There have also been changes in mean size-at-age and fisheries selectivity for Pacific halibut, both of which have implications for optimal harvest rate calculations.


The definition of optimal harvest rate can be broadly defined as the harvest rate that would lead to maximizing the long-term yield. Optimal harvest rates can be defined in terms of specific metrics; for example, the harvest rate that maximizes the yield per recruit (FYPR), or the harvest rate that reduces the spawning biomass per recruit to some specified level (FSPR), or the harvest rate that maximizes the long-term sustainable yield (FMSY). Each of these alternative harvest rate metrics requires similar biological and fisheries selectivity information, with the exception of the FMSY which requires additional information about the relationship between spawning stock biomass and recruitment. For the purpose of this report, the harvest rate that maximizes long-term sustainable yield is what is implied with the term optimal harvest rate and is denoted herein by FMSY.

Factors that affect the optimal harvest rate calculation fall into two general categories: 1) biological properties that define the underlying productivity of the stock, and 2) fishing related properties associated with target and non-target fisheries that catch halibut. Important biological parameters that define the optimal harvest rate include: growth parameters and the variance in size-at-age, maturity- and fecundity-at-age, natural mortality rates, the steepness of the stock-recruitment relationship, and migration rates among the various fishing grounds (Beddington and Kirkwood, 2005). Although dispersal and movement is important for determining optimal harvest rates, it is often ignored under the assumption of a unit stock (area is sufficiently large enough to ignore dispersal and migration). The vulnerability of fish to fishing gear, also known as selectivity, is also extremely important in determining optimal harvest rates (Hilborn and Walters, 1992). In general, the age-at-entry to the fishery relative to the age-at-maturity defines the optimal fishing mortality rate; as the age-at-entry to the fishery increases relative to the age-at-maturity, the higher the value of FMSY.

The current harvest strategy for Pacific halibut also includes the use of a minimum size limit in the directed commercial longline fishery. Halibut that are below the minimum legal size of 81.3 cm total length are required to be released and the IPHC assumes a 16% mortality rate for released fish based on previous tagging studies. Changes in the size limit also have implications for the FMSY calculations, especially if release mortality rates are significant. In general, if halibut less than the minimum size limit are routinely captured, then reducing the minimum size limit will also result in reductions in the optimal harvest rate. Note that the EBio calculation is based on all halibut greater than 66 cm (or 26 inches) so the definition of Ebio does not change with changes in size limits.

Previous studies have examined the impacts of alternative size limits on the yield per recruit and spawning biomass per recruit using steady-state or equilibrium models (Clark and Parma, 1995). Previous halibut studies have not considered the cumulative impacts of size-selective fishing on optimal harvest rate calculations for Pacific halibut. Using a minimum size limit, faster growing individuals from a given cohort would experience higher total mortality rates over their lifetime as they recruit to the legal size at a younger age in comparison to slower growing individuals. The cumulative effect of fishing mortality imposes a higher total mortality rate on faster growing halibut; remaining individuals left in the population consist of slower growing individuals (Taylor et al., 2005). The cumulative effects of size-selective fishing can, therefore, give the appearance of declining mean weight-at-age in the population.

Further, previous studies have not addressed how different release mortality rates affect yield per recruit and optimal harvest rates for Pacific halibut. Coggins et al. (2007) and Pine et al.


(2008) demonstrated that for a given size limit, increases in discard mortality can lower the realized yield per recruit. This is also the case if both lower and upper size limits (i.e., slot-limit) are in place, where the assumption is that an upper size limit would add additional protection to large sexually mature females.

Maximizing yield per recruit, or maximizing the total landed biomass is not always desirable from an economic perspective. In the cases where there is a differential price structure for the size of fish landed (e.g., large fish fetch a higher price per pound than small fish), it may be more desirable to fish at rates lower than FMSY in order to maximize value of all fish landed. Moreover, fishing mortality rates associated with Maximum Economic Yield (MEY) are generally lower than fishing mortality rates associated with MSY (Gordon, 1954).

In this paper, I use an equilibrium model for Pacific halibut to examine how changes in biological components and size-specific fishing mortality impact estimates of optimum harvest rate calculations. In addition to the directed fishery, the impacts of other constant removals from non-directed fisheries (e.g., bycatch) on estimates of optimal fishing mortality rates in the directed fishery are also examined. The cumulative effects of size selective fishing on changes in mean weight-at-age are also explored. Size-at-age data from the 2011 setline survey are used to estimate growth curves for each regulatory area, and these growth curves are used to illustrate how relative differences in growth affect estimates of optimal harvest rates in each regulatory area.

Methods

Equilibrium model The optimal fishing mortality rate for Pacific halibut is based on the assumption that we wish

to maximize the overall yield coast wide. To determine the appropriate harvest rate, the equilibrium yield is calculated over a range of alternative effective coast wide harvest rates, and the harvest rate that maximizes this yield is set equal to FMSY. Key inputs to this equilibrium model include: natural mortality rate, the steepness of the stock recruitment relationship, the maturity-at-age information and the size-based selectivity function estimated from the most recent stock assessment model.

Including release mortality and cumulative effects of size-selective fishing The equilibrium model described in the previous section also takes into consideration that all

fish captured are either retained or discarded, and those that are discarded are subject to an additional 16% mortality. In the model we assume that all fish less than the minimum size limit (or greater than the maximum size limit) are discarded.

To represent the cumulative effects of size-selective fishing, the equilibrium model is partitioned into 11 individual sub-populations that differ by growth rate. Populations with fast growth recruit to the fishery at a younger age and therefore are subject to higher fishing mortality rates. The cumulative effect of continuous fishing over all of the populations results in higher overall total mortality rates for faster growth populations and this gives an appearance of declining size-at-age with increasing fishing mortality rates.


Life-history parameters For this paper, the assumed natural mortality and selectivity parameters are listed in Table 1.

Estimated growth parameters for each regulatory area are summarized in Table 6 in Appendix A. Sexual maturity for female halibut was assumed to be a logistic function of age, where the age-at-50% maturity is 10.91 years and the standard deviation is 1.406 years. Relative fecundity-at-age is assumed to be proportional to mature female weight-at-age. The allometric length-weight relationship (a,b parameters in W=aLb) was assumed to be the same for both sexes (Table 1).

Estimates of growth parameters for Pacific halibut were based on size-at-age data ontained from the 2011 set line survey. Estimated growth parameters by sex are listed in Table 6. Estimates of asymptotic lengths for females ranged from 126.1 cm for Area 2A to 199.5 cm for Area 4C, for males asymptotic lengths ranged from 97.0 cm in are 3B to 126.8 cm in Area 4B. Growth coefficients for females averaged 0.0775 and for males averaged 0.0811. Growth rates in Area 2A are greatly accelerated relative to the standard von Bertalanffy growth curve at younger ages (p<1) in Area 2A, whereas growth rates for females are reduced in all other areas (p>1). Also note that estimates of the p parameter are heavily confounded with estimates of asymptotic length, especially in areas where data from large–older fish are lacking.

The two key parameters that define the underlying stock-recruitment relationship in this model are: Bo, the unfished spawning stock biomass, and steepness (h) which is defined as the fraction of unfished recruitment that is obtained when the spawning stock biomass is reduced to 20% of its unfished state. Normally these two parameters are obtained by fitting a stock-recruitment relationship to the historical estimates of spawning biomass and recruitment numbers (usually integrated within the stock assessment model). The current assessment model for Pacific halibut has no built in stock recruitment relationship at this time, so these parameters are not readily available. In the absence of Bo and h estimates, Bo was arbitrarily set at 100 pounds, and a steepness value of 0.75 was chosen somewhat arbitrarily because estimates of FMSY were fairly similar to those obtained for Areas 3A, 2B and 2C by Clark and Hare (2006). Note also, that in arriving at a value of 0.75 for steepness, no bycatch was assumed for the non-directed fisheries (i.e., bycatch for the trawl fishery was set equal to 0). If bycatch was included in the initial development of the model, the assumed value of steepness would likely be higher to compensate for recruitment loss associated with bycatch of juvenile halibut.

Optimal fishing rates To determine the fishing mortality rate that would maximize the relative yield in each

regulatory area, a discrete range of equilibrium fishing mortality rates was used to calculate FMSY for that regulatory area. The relationship between fishing mortality and yield is then plotted for each regulatory area (also referred to as equilibrium yield curves).

The equilibrium yield curves obtained for each regulatory area assume no migration between each regulatory area, and for the purposes of this paper, are effectively treated as closed populations. Therefore, a single value of Bo is used for all regulatory areas and we only report the relative yields per 100 pounds of unfished spawning biomass. The other reason for assuming the same scaling and steepness parameter is that it also allows for direct comparisons of yield-per-recruit, spawning biomass per-recruit, discards per-recruit etc. in response to differences in size-at-age (growth) in each regulatory area.


Scenarios Combinations of policy parameters in the equilibrium model were explored to examine the

implications of changing fishing regulations on optimal harvest rate calculations. Also the sensitivity of optimal harvest rates to alternative assumptions about discard mortality rates, steepness, or the effects of bycatch non-directed fisheries was also examined. In addition to the base scenario (S1, Table 2), eight additional scenarios were examined to explore the effects of minimum and maximum size limits (S2, S3), a 10 cm shift in commercial selectivity towards smaller fish (S4), other mortality associated with non-directed fisheries that remove a constant catch (S5), and sensitivity to size-dependent natural mortality (S6, S7) and steepness (S8, S9).

The intention of scenarios 2 and 3 is to examine how the overall equilibrium yield, yield-per-recruit, spawning biomass per-recruit, and estimates of FMSY would change with changes in size limits. Similarly, how would these same variables change if the fishery targeted smaller fish (S4)? In the case of Scenario 4, the same size-based selectivity coefficients are used, but the liner interpolation over size is shifted to 50–120 cm from the status quo of 60–130 cm. In other words, if a 100 cm female had a selectivity of 0.535 in the base scenario, is S4 it has a selectivity of 0.535 at 90 cm.

In scenario S5, the bycatch from a non-directed trawl fishery is assumed to be constant, irrespective of the density of halibut on the trawl grounds (a worst-case scenario). Also, bycatch is not expected to result in additional compensation in juvenile survival rates for new recruits (i.e., there is no impact on steepness of the stock-recruit relationship). In the case of constant bycatch, the fishing mortality rate is expected to decrease exponentially with increasing halibut density. For example, if the bycatch fisheries discard a fixed amount of 1 million pounds of dead halibut each year and the equilibrium biomass is 10 million pounds, then the corresponding fishing mortality rate of the bycatch fishery is proportional to 1/10, or 0.1. If, however, the equilibrium biomass is at a lower level, e.g., 2.5 million pounds, then the equilibrium fishing mortality rate is proportional to 1/2.5, or 0.40.

It was also assumed that bycatch from the trawl fisheries selected fish of small and intermediate sizes. This selectivity was approximated with a double logistic function with the size at 50% selectivity at 61 cm for the ascending limb and 81.3 for the descending limb, and a standard deviation of 0.1 cm (knife-edge) for both ascending and descending portions of the curve. In reality, the actual selectivity curves could differ markedly, and the appropriate size-composition data would have to be integrated into the assessment model to estimate selectivity parameters for a discard fishery.

In scenarios 6 and 7, natural mortality is assumed to be size-dependent where small halibut have a higher natural mortality rate than larger halibut (S6), or natural mortality rates increase with increasing size (S7). In both scenarios 6 and 7, the average natural mortality rate is approximately 0.15 when integrated over all size classes.

Lastly, scenarios 8 and 9 are intended to demonstrate how sensitive the reference fishing mortality rate calculation is to the assumed value of steepness in the stock assessment model. This is akin to the range of recruitment values used in the Clark and Hare (2006) simulation study where no density-dependent effects on recruitment were assumed.


Results

Growth Size-at-age data in each of the regulatory areas has very marked differences in both the mean

length-at-age, and the distribution of age-classes (Fig. 1). Male and female halibut in Area 2A are fast growing but tend to have a much smaller asymptotic size than halibut sampled in other regulatory areas. Moreover, the age-composition in Area 2A is truncated relative to other regions, with very few fish older than 17 years. The coefficient of variation in length-at-age is much higher in Areas 2B and 2C, especially for females. Estimated growth rates for female halibut in these two areas is nearly linear for younger ages, and on average older halibut in these regions are much larger in comparison to other regions with old female halibut. Area 4B is also another anomaly in that the age-distribution is much older, especially for males, with many sampled individuals beyond age 20. Additional details about the growth model and estimated model parameters are found in Appendix A.

We note here that estimated growth parameters in Figure 1 are biased due to size-based selectivity in the setline survey and potentially contaminated due to the cumulative effects of size-selective fishing. Nonetheless, the relative differences in growth rates in each of the regulatory areas, is what is important in this analysis.

Area-specific estimates of FMSYThe relative equilibrium yield versus fishing mortality rate in each of the regulatory areas

(Fig. 2) demonstrate the relative differences in expected yield based only on differences in halibut growth (size-at-age) in each of the regulatory areas. For each of the regulatory areas shown in Figure 2, a minimum size limit of 81.3 cm exists, steepness is fixed at an arbitrary value of 0.75, size-selectivity is the same in each area, and natural mortality rates are the same (M=0.15) for all areas. The only biological difference between regulatory areas is the growth rate. For each recruiting halibut in a specific regulatory area, the maximum yield per recruit would be obtained in Area 2A. Halibut in Area 2A are very large at younger ages and more vulnerable to the fishing gear (selectivity) at an age when they are numerically more abundant.

In contrast, in Area 4C halibut obtain very large sizes, but the growth rate is much slower in comparison to 2A so fewer individuals are available to be harvested and less yield per recruit is obtained in this area. The net result of this difference in growth rates is that estimates of FMSYare lower in Area 4C relative to Area 2A (Fig. 2 and Table 3). The difference in early growth between 2A and 4C translates into roughly 40% more yield per recruit in Area 2A.

Equilibrium yield curves for each regulatory area under each of the 9 alternative scenarios are shown in Figure 3 and the corresponding estimates of optimal exploitation rates are summarized in Table 3. Relative to the current harvest rate policy of 21.5% and 16.125%, estimates of optimal exploitation rates for Areas 2B and 2C are below the current 21.5% value. However, recall that this is based on the assumption of a Beverton-Holt stock recruitment relationship with a steepness value set at an arbitrary level of 0.75. The utility of S1 is to serve as a baseline in which to compare impacts of alternative size limits and model assumptions on the estimates of optimal exploitation rates that would maximize the average long-term yield in each of the statistical areas.

The maximum size limit scenario results in slight increases in the estimates of FMSY in areas where halibut grow to a sufficiently large size to benefit from such protection (Table 3).


Imposing a maximum size limit does not result in any yield benefits in any of the regulatory areas (Scenario S2, Table 4). In fact, in Areas 2A and 3B, there is a very small probability that an individual halibut would survive and grow to surpass the upper legal size limit of 140 cm. There is only a minor improvement in the relative spawning biomass in Areas 2A and 3B associated with a maximum size limit of 140 cm (Table 5). Whereas, there is a further reduction in the spawning biomass depletion in areas where halibut grow beyond the 140 cm maximum size limit and fishing at FMSY, and the amount of spawning biomass reduction is proportional to the discard mortality rates.

If size-limits were removed altogether, and there is no change in the size-selectivity of the commercial fishery, then estimates of FMSY would have to be reduced (S3, Table 3) in order to compensate for the increased total mortality rate associated with retaining fish smaller than 81.3 cm. Relative increases in overall yield do occur with the removal of the minimum size limits, as the yield per recruit in each area increases, with the exception of Area 4B. However, this modest increase in overall yield does come at the expense of reducing spawning stock biomass, as well as, reducing the average size of landed fish.

Scenario 4 represents a shift in the commercial selectivity towards smaller fish, and the net impact of this shift is a reduction in the FMSY values for each regulatory area, as well as, decreases in the overall landed yield (Tables 3 and 4). Recall that this scenario was run with the current minimum size limit of 81.3 cm in place and serves to show that minimum size-limits alone does not afford protection of spawning stock biomass if discard mortality rates are greater than 0. Although corresponding increases in spawning biomass are observed in Table 5. for scenario 4, this increase owes to the reduction in FMSY that would be required to maximize yield if selectivity were to shift towards smaller fish.

The effects of other discard mortality, from non-directed fisheries, also plays a role in harvest policy calculations. In scenario 5, a constant total harvest of 2 Mlb (i.e., bycatch from a trawl fishery) was imposed as an increasing fishing mortality rate with increasing directed Fe. In this scenario, a constant level of bycatch results in a dramatic reduction in overall yield in the directed fishery (Table 4) and a reduction in the optimal harvest rate (FMSY) that would produce the maximum sustainable yield in the directed fishery. For example, in Area 2A, 2 lb of bycatch would reduce the directed yield by 1.41 lb if fishing at the optimal fishing mortality rate (Table 4, scenario S5).

In the case where bycatch fisheries remove a fixed amount, the mortality rate increases with declining stock size. Whereas, in the directed fishery, annual catches would scale down with reductions in exploitable biomass, and fishing mortality rates would scale down if the spawning biomass falls below B30%. Note also that FMSY estimates would increase if efforts were made to reduce bycatch in non-directed fisheries.

Sensitivity to assumed parameter values Scenarios 6 and 7 examine how sensitive MSY-based reference points are to estimates of

natural mortality rates. The current assessment model assumes natural mortality is independent of size/age. These two scenarios examine a size-effect in natural mortality. In general, if M is size/age independent, the increasing M results in increases in the estimates of FMSY, and vice versa (Walters and Martell, 2004). Also increases in M results in a decrease in MSY and the


spawning biomass at FMSY. If M is size-dependent and decreases with size, estimates of FMSYalso decrease, and vice versa. Natural mortality also plays a role in the general scaling of MSY; as fewer older fish are available for harvest due to high natural mortality rates, then the value of MSY decreases. Hence mis-specification of M can lead to biased estimates of other reference points (e.g., unfished spawning biomass Bo).

Lastly, estimates of FMSY are very sensitive to the steepness of the stock-recruitment relationship. With increasing steepness the corresponding estimates of FMSY also increase, and vice versa. The resilience of the stock to over-fishing decreases with decreasing values of steepness, the overall yield declines and there are fewer recruits per unit of spawning biomass (i.e., lower stock productivity).

Wastage in the directed fishery In all scenarios where size limits exist, wastage in the directed fishery increases with

increasing fishing mortality. Under the current status quo scenario (S1), the long-term average wastage in the directed fishery is estimated to be less than 5% of the total landed yield in each of the regulatory areas if fishing mortality rates are less that 0.25 (Fig. 4). Note that for the purposes of this paper, as well as for the wastage estimates that go into the stock assessment model, it is assumed that the selectivity of the commercial fishery is the same as the estimated selectivity curve in the setline survey.

The use of an upper size limit (S2), results in increased wastage over the status quo scenario, but only in areas where halibut attain sufficiently large sizes. At low equilibrium fishing mortality rates wastage in Areas 2B, 2C, 4B and 4C are greater than 10% of the landed catch due to large halibut (greater than 140 cm) in these regions (Fig. 4). As fishing mortality rates increase and erode the size structure, wastage rates decline and effectively become the same as that of a minimum size limit only.

In scenario 4, where the commercial selectivity curve was shifted by 10 cm towards smaller fish, the long-term average wastage in the directed fishery more than doubles what is currently assumed under the status quo scenario (Fig. 4). Note that a minimum size limit of 81.3 cm is also maintained in the S4 calculations. Lastly, I also note here that under Scenario 3 (not shown in Fig. 4), there is no wastage, as it is assumed that all fish harvested are landed.

Changes in mean weight-at-age Mean size-at-age is predicted to decline with increasing size-selective fishing mortality,

especially for age-classes that are fully recruited to the fishing gear (Fig. 5). Ages less than 8 years are not expected to show much of a change in the mean weight-at-age because they are only partially recruited to the gear, and have not been subjected to intense fishing mortality.

There are substantial differences in how the predicted mean weight-at-age would change with increasing fishing mortality across regulatory areas. This is a result of differences in growth rates among regulatory areas. Despite these differences, the general pattern of cumulative size-selective fishing results larger changes in mean size for older individuals and little to no change for age classes that are not subject to intensive fishing. Similar patterns are also evident in the raw size-at-age data collected from the setline survey (Fig. 6). In recent years, exploitation rates in Area 2B are estimated to be greater than the target rate of 21.5%. Mean weight-at-age for age-6 fish in Area 2B have varied little between 2002 and 2006; whereas there has been a decline in mean weight-at-age for ages 10 and 14 between 2007 and 2009 (Fig. 6). There is considerable variability in the observed size-at-age data from the set line survey in all of the regulatory areas.


Discussion

Factors that affect estimates of optimal harvest rates come in two general forms: (1) biological components that define the underlying productivity of the stock, and (2) fishery components that affect size-at-entry and size-specific mortality. The former cannot be directly managed but must be taken into consideration in harevst policy, especially if there are temporal changes in stock productivity. The latter can be directly controlled through a variety of tools that limit gear specifications, size limits, or even areas fished, to control size-at-entry into the fishery and reduce post release mortality rates. In the case of Pacific halibut, there have been recent changes in size-at-age and there is also considerable variation in the size-at-age among the regulatory areas. Optimal harvest rate calculations for each of the regulatory areas will have to be updated frequently owing to continued changes, and differences, in size-at-age among regulatory areas. Also, any changes in fisheries operations or changes in fishery regulations (e.g., change in size limits or bycatch) will also affect optimal harvest calculations. In general, the underlying parameters that define the harvest control rule should be updated on a routine, or even annual, basis to ensure that biological and technological changes are taken into account to ensure harvest policy is kept up to date.

In this paper, we examined how changes in stock productivity and size-selectivity interact with estimates of fishing mortality rates that maximize long-term sustainable yields. This was approximated using an age- and sex-structured equilibrium model conditioned on regulatory area size-at-age data from the 2011 setline survey and parameters from the most recent stock assessment. The previous harvest policy was based on a stochastic simulation model for Pacific halibut that is now dated because of: (a) continued changes in size-at-age for Pacific halibut, (b) a transition from area-based assessments to a coast-wide model, (c) recent recognition that fishery and survey selectivity must vary over time in the coast-wide assessment due to changes in the distribution of the stock, and (d) reduction in the bycatch levels in non-directed fisheries. There have been previous equilibrium-based models for the development of harvest policy for Pacific halibut (e.g., Clark and Parma, 1995), but these models did not explicitly consider the effects of wastage and bycatch on optimal harvest rates. Moreover, previous analyses were also limited to the core halibut areas (2B, 2C and 3A). This study attempts to address some of these shortcomings, but is still incomplete.

At this point, this analysis should be considered a work in progress for the sole reason that a key population parameter (i.e., steepness) that defines the underlying stock productivity is not yet available for the new coast-wide assessment. The steepness parameter was arbitrarily set at a value of 0.75 and was chosen because it resulted in estimates of FMSY that are similar to those obtained by Clark and Hare (2006). Nevertheless, the relative changes in FMSY among regulatory areas based on differences in growth rates would not differ if a reliable estimate of steepness was available. Also critical to optimal harvest rate calculations is the area-based selectivity. At present, the coast-wide assessment model explicitly assumes size-based selectivity for each fishing gear does not vary by regulatory area, yet it is allowed to vary over time due to shifts in the distribution of the stock. In comparison to differences in size-at-age, regulatory area differences in selectivity-at-age relative to maturity-at-age will also have a large impact on estimates of area specific optimal exploitation rates. The previous closed-area assessment models estimated marked differences in selectivity among regulatory Areas 2B and 3A (Clark and Hare, 2006). The transition to a coast-wide model introduced a new assumption that size-based selectivity is the same for all regulatory areas.


One issue, that has not been examined here, and has very important harvest policy implications is the initial recruitment and movement of halibut among regulatory areas. Area-specific optimal harvest rates are sensitive to movement of halibut among regulatory areas. There has been a considerable effort in this regard to understand the movement of halibut (e.g., Loher and Seitz, 2006; Webster, 2009) and what the potential implications are for harvest policy (Valero and Hare, 2009, 2010). In general, areas with a net migration loss are nearly equivalent to having a higher natural mortality rate in a closed area model. The harvest policy implications in such a case would be to harvest at a higher rate, but the total removals would would scale down. The opposite is true for an area that has a net migration increase, harvest at a lower rate, but the scale of the harvest increases due to immigration to the area. If however, the objective is to maximize the yield from all areas combined, then the optimal harvest rate calculations are much more complex involving dispersal kernels for new recruits and age-specific migration transition matrices. Under such circumstances it is not possible to make generalized statements about how optimal harvest rates would change because the answer depends on the relative migration coefficients among the regulatory areas and the initial distribution of new recruits. Valero and Hare (2010) had made progress in this area and their early conclusions suggested that harvest policies to the north of Area 2 would have fairly severe implications for Area 2 itself due to downstream migration of halibut into this area.

It is intuitive to think that imposing a maximum size limit would afford protection to sexually mature fish that grow beyond the size limit and that this would lead to an increase in spawning biomass (or reduce the level of depletion). This does occur, but only if there is a very low discard mortality rate associated with releasing fish. In the case examined here, with 140 cm size limit and fishing at FMSY, the spawning biomass in each regulatory area remains nearly the same or declines in comparison to no maximum size limit. The reason for this decline is related to a discard mortality rate of 0.16 per year, the relatively low number of halibut currently growing to this size, and under MSY-based harvest policies, values of FMSY would increase in areas where halibut grow to sufficient size and a maximum size limit is used in the harvest policy.

Shifts in the directed commercial selectivity schedule towards smaller halibut pose a conservation concern if the discard mortality rate is greater than 0, even if minimum size limits are in place. If individual IFQ holders are not accountable for their discard mortality of undersized fish, then fishing can continue until their quota is filled with legal-sized halibut. If there is a shift towards catching smaller fish, or the probability of capturing a legal-size fish in a given area is low, then the corresponding increase in discard mortality results in a higher overall total mortality rate that may not be accounted for if the shift in selectivity goes undetected. This is of particular concern in the current assessment of Pacific halibut, where the wastage calculation assumes the commercial fishery selectivity is the same as the top 33% of the setline survey WPUE (Gilroy and Hare, 2009). Moreover, estimated commercial selectivity is based on composition data from port samples, not fish sampled on the boats at sea when the gear is being retrieved. In other words, the wastage calculation in the directed fishery is based on a tenuous assumption about how the commercial gear selects fish less than 81.3 cm.

The harvest policy implications of undetected changes in selectivity and estimates of optimal exploitation rates are somewhat insensitive if the discard mortality rates are low or even negligible. If under-sized fish are handled with extreme care such that survival rates are near 100%, then the previous discussion about uncertainty in commercial selectivity for under-size fish is moot. Moreover, the estimate of wastage would consist only of lost or abandoned gear.


However, if the release mortality rates are appreciable, then estimates of optimum harvest rates must also include release mortality associated with size-limits (Goodyear, 1993; Coggins et al., 2007). In general, as the size limit increases the optimum fishing rate that would maximize yield increases exponentially. This relationship is also the same for the fishing mortality rate that would deplete the spawning biomass to some target level. The exponential increase in FMSYoccurs when individuals have had at least one chance to spawn before they become vulnerable to fishing. Pine et al. (2008) demonstrated that as the release mortality rates increase, the potential of a minimum size limit to hedge against overfishing decreases as the release mortality rates increase.

If the observed changes in size-at-age are a result of cumulative size-selective fishing, then the largest changes in size-at-age would be expected in areas with higher fishing mortality rates. In closed populations, the variance in size-at-age for older fish is expected to decrease with increasing fishing mortality rates. Areas 2B and 2C are thought to have fairly high fishing mortality rates based on the results of the stock assessments and biomass apportionment (Hare, 2012). Based on the size-at-age data from the setline survey, the largest variance in size-at-age is found in Areas 2B and 2C, suggesting that these areas either have a low exploitation rates or are heavily influenced by migration from adjacent areas that have lower exploitation rates.

In the very near future, the Pacific halibut assessment model is likely to evolve to a more implicit spatial representation where estimated selectivities by regulatory area may differ. Given that steepness is unknown, and selectivity likely differs by regulatory area, it is not recommended to change the current harvest policy until, at a minimum, these two issues have been addressed. Preferably, the full suite of biological factors, including dispersal of recruits and migration, and factors that affect selectivity would be included in the harvest policy analysis.

Acknowledgements

The authors would like to thank the following persons for discussions about harvest policy for Pacifi c halibut: Steven Hare, Juan Valero, Robyn Forrest and Jim Ianelli. Also, comments from several IPHC staff including Ray Webster, Claude Dykstra, Heather Gilroy, Stephen Keith and Gregg Williams during earlier presentations of this work are greatly appreciated and have led to many refi nements in this work. Special thanks to Aaron Ranta and Jay Walker for helping with database queries and technical support.

References

Beddington, J. R. and Kirkwood, G. P. 2005. The estimation of potential yield and stock status using life-history parameters. Philosophical Transactions of the Royal Society B-Biological Sciences 360(1453):163–170.

Clark, W. and Hare, S. 2006. Assessment and management of Pacifi c halibut: data, methods, and policy. Int. Pac. Halibut Comm. Sci. Rep. No. 83

Clark, W. and Parma, A. 1995. Re-evaluation of the 32-inch Commercial Size Limit. Int. Pac. Halibut Comm. Tech. Rep. No. 33.

Coggins, L., Catalano, M., Allen, M., Pine, W., and Walters, C. 2007. Effects of cryptic mortality and the hidden costs of using length limits in fi shery management. Fish and Fisheries 8(3):196–210.


Gilroy, H. and Hare, S. 2009. Wastage of halibut in the commercial halibut fi shery. Int. Pac. Halibut Comm. Report of Assessment and Reseach Activities 2008:59–62.

Goodyear, C. 1993. Spawning stock biomass per recruit in fi sheries management: foundation and current use. Canadian Special Publication of Fisheries and Aquatic Sciences, pages 67–82.

Gordon, H. 1954. The economic theory of a common-property resource: the fi shery. The Journal of Political Economy, 62(2):124–142.

Hare, S. 2012. Assessment of the pacifi c halibut stock at the end of 2011. Int. Pac. Halibut Comm. Report of Assessment and Research Activities, 2011:91–194.

Hilborn, R. and Walters, C. J. 1992. Quantitative fi sheries stock assessment: choice, dynamics and uncertainty, volume 2. Springer.

Loher, T. and Seitz, A. 2006. Seasonal migration and environmental conditions of pacifi c halibut hippoglossus stenolepis, elucidated from pop-up archival transmitting (pat) tags. Mar. Ecol. Prog. Series 317:259.

Martell, S. J. D., Pine, W. E., and Walters, C. J. 2008. Parameterizing age-structured models from a fi sheries management perspective. Can. J. Fish. Aquat. Sci., 65:1586–1600.


Table 1. Natural mortality rate and size-specifi c selectivity coeffi cients (scaled to the maximum estimated coeffi cient) used in the equilibrium model.

Sex-specifi cParameter Symbol Value Female MaleUnfi shed spawning biomass 100Steepness h 0.75Natural mortality M 0.15 0.1439

145 110k 0.10 0.12

Age-at-50% maturity 10.91Std Age-at-50% maturity 1.406Length weight scale a 6.821e-6Length weight power b 3.24

Selectivity coeffi cients Female Male60 0.000 0.00070 0.153 0.13280 0.310 0.24390 0.441 0.367

100 0.535 0.535110 0.610 0.694120 0.695 0.848130 0.785 1.000


Table 2. Parameter settings for alternative model scenarios. See text for description of scenarios; h is the steepness of the stock-recruitment relationship, M is the natural mortality rate for females, SL corresponds to size limit, DM is discard mortality rate, size shift in selectivity (cm), and other mortality is the range of instantaneous mortality rates from non-directed fi sheries.

Scenario h M Min SL Max SL DMSelectivity shift

Other Mortality

S1 0.75 0.15 81.3 ∞ 0.16 0 0S2 0.75 0.15 81.3 140 0.16 0 0S3 0.75 0.15 0 ∞ 0.16 0 0S4 0.75 0.15 81.3 ∞ 0.16 -10 cm 0S5 0.75 0.15 81.3 ∞ 0.16 0 0.02–

0.153S6 0.75 0.23–0.12 81.3 ∞ 0.16 0 0S7 0.75 0.12–0.23 81.3 ∞ 0.16 0 0S8 0.85 0.15 81.3 ∞ 0.16 0 0S9 0.65 0.15 81.3 ∞ 0.16 0 0

Table 3. Estimates of optimal exploitation rates for each regulatory area and scenario combination. Scenario descriptions are found on page 7.

Scenario 2A 2B 2C 3A 3B 4A 4B 4C 4DS1 0.248 0.197 0.180 0.297 0.318 0.260 0.180 0.176 0.300S2 0.248 0.221 0.221 0.300 0.321 0.271 0.225 0.237 0.304S3 0.221 0.180 0.163 0.256 0.275 0.229 0.168 0.163 0.260S4 0.201 0.163 0.151 0.229 0.245 0.209 0.155 0.151 0.233S5 0.168 0.137 0.133 0.172 0.185 0.168 0.137 0.133 0.176S6 0.180 0.133 0.120 0.193 0.205 0.163 0.124 0.111 0.193S7 0.213 0.176 0.168 0.252 0.275 0.233 0.172 0.172 0.256S8 0.260 0.209 0.197 0.289 0.311 0.264 0.197 0.189 0.293S9 0.155 0.129 0.120 0.180 0.193 0.163 0.124 0.120 0.180


Table 4. Relative change in yield by regulatory area in comparison to S1 (status quo) for each of the alternative scenarios while fi shing at rates defi ned in Table 3 (MSY based fi shing mortality).

Scenario 2A 2B 2C 3A 3B 4A 4B 4C 4DS1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00S2 0.00 -0.17 -0.27 -0.01 0.00 -0.04 -0.33 -0.41 -0.01S3 0.65 0.30 0.11 0.85 0.78 0.28 -0.01 0.04 0.45S4 -0.04 -0.03 -0.02 -0.14 -0.15 -0.09 -0.04 -0.04 -0.14S5 -1.41 -0.84 -0.69 -1.41 -1.44 -1.12 -0.71 -0.61 -1.47S6 -0.72 -0.83 -1.00 -0.86 -0.89 -0.98 -1.00 -1.00 -0.91S7 0.32 0.38 0.48 0.24 0.24 0.40 0.45 0.44 0.26S8 1.10 0.72 0.66 0.57 0.55 0.63 0.69 0.49 0.60S9 -1.13 -0.76 -0.69 -0.87 -0.88 -0.82 -0.78 -0.60 -0.92

Table 5. Relative spawning biomass while fi shing at for each scenario and regulatory area.

Scenario 2A 2B 2C 3A 3B 4A 4B 4C 4DS1 23.8 25.5 25.8 26.0 26.8 26.9 27.6 28.3 27.1S2 23.9 24.4 24.2 25.7 26.6 26.2 25.5 25.4 26.9S3 22.6 24.5 25.4 23.0 23.5 25.0 27.0 27.6 24.4S4 24.8 26.1 26.7 27.5 28.3 27.7 27.8 28.8 28.5S5 26.0 27.7 26.8 29.4 30.1 29.0 28.4 29.2 30.4S6 24.1 25.9 25.7 26.0 26.6 27.2 26.7 28.2 27.3S7 25.7 27.7 28.0 28.7 29.3 29.0 28.7 30.0 29.7S8 20.5 22.4 22.4 24.3 25.1 24.7 24.1 25.6 25.6S9 29.2 29.5 29.7 30.4 31.3 30.8 30.7 31.8 31.9


Table 6. Estimated growth parameters for each regulatory area based on fi tting a 4 parameter growth model to the 2011 size-at-age data from the setline survey.

Reg. L∞ k to pArea Female Male Female Male Female Male Female Male2A 126.1 99.3 0.105 0.095 −2.30 −4.53 0.983 0.8182B 165.3 107.6 0.076 0.069 −4.62 −7.45 1.499 0.9862C 192.2 104.3 0.063 0.110 −6.08 −3.65 1.736 1.2963A 147.6 97.8 0.057 0.073 −5.81 −4.25 1.167 0.7333B 135.6 97.0 0.073 0.083 −4.33 −4.90 1.252 0.9954A 153.6 107.0 0.073 0.086 −5.16 −3.94 1.601 1.2024B 167.3 126.8 0.109 0.080 −4.75 −3.60 2.830 1.2034C 199.5 125.7 0.079 0.068 −5.86 −11.21 2.717 2.1854D 153.4 114.6 0.064 0.065 −6.23 −4.61 1.530 0.953


●

●●

●

●●

●

●●●

●●

●●

●●●●

●

●

●●

●●●●●

●

●

●●

●

●

●●

●

●

●

●●●●

●

●

●

●●

●

●●

●

●

●

●●

●

●●

●●

●●●

●●

●

●

●●

●

●●●●●

●

●●

●

●●●

●

●

●

●

●●

●

●●

●

●●●●

●

●

●

●●●

●

●

●

●●

●

●●

●

●●

●

●

●

●●

●

●●

●●●●

●●

●

●

●●

●

●

●

●

●

●●●

●●

●

●●●●●●

●

●●●●

●

●●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●●

●

●●

●

●

●

●

●

●●

●

●

●

●●

●●

●

●

●●●

●

●

●●

●●

●●

●●●●

●●

●●●●●

●●

●

●●

●●

●

●●

●

●●

●

●

●

●●

●

●

●

●

●●●

●

●

●●

●●

●

●

●

●

●●

●●

●

●●

●●

●●

●

●

●

●

●●●●●●

●

●

●

●

●

●●●

●●

●●

●

●●

●

●●

●

●

●

●●

●●

●

●

●

●●

●●

●●

●●

●

●

●

●●

●

●●

●●

●

●

●

●

●●

●

●

●

●●

●

●

●●

●

●

●●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●●

●

●●

●

●

●●●

●●●

●●●●●

●

●●

●●●

●●

●

●

●●

●

●●

●●●

●●●

●

●●

●

●

●

●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●●●

●

●

●

●

●●

●

●●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●

●●

●

●●

●●●●●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●●●

●

●

●●

●●

●

●●

●

●

●●●

●

●●

●

●●

●●

●

●

●●

●

●

●

●

●●●

●●

●●

●

●●

●

●

●

●

●●

●

●●

●●

●●

●●●

●

●

●

●

●

●●

●

●

●●

●●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●●

●

●●

●●●

●

●

●

●

●

●

●

●

●●

●

●

●

●●●●

●●

●

●●

●

●

●

●●●●

●

●●

●

●

●

●

●

●●

●●

●

●

●●

●

●

●

●

●

●

●●

●

●●

●

●●

●

●

●

●

●●

●

●

●

●

●

●●

●

●

●

●●

●●●

●

●

●●●

●

●

●

●●

●

●●●

●

●●●

●

●

●

●

●

●●

●●

●

●

●●

●

●

●

●●

●

●

●

●●

●

●●

●

●●

●●

●

●

●

●●

●

●

●

●

●●●●

●●

●●●

●

●

●

●

●●●

●●

●●●●

●●

●●

●●

●

●●●●●●

●

●

●●●

●

●

●●●

●

●●

●●

●

●●

●

●

●

●●●

●

●●

●●

●

●●

●●

●●

●

●●

●●●

●

●

●●

●

●

●●

●

●●

●●

●

●

●

●●

●

●

●●

●●

●●

●●●●

●●●●

●●

●●

●●

●

●●

●●

●

●

●

●●

●

●

●

●

●

●●

●

●

●

●

●●●

●

●●●

●●

●●

●●

●

●

●●●●●●●

●●

●

●●

●●●

●

●

●●●

●

●●

●●●

●

●

●

●●

●●

●●

●

●●

●●

●●

●●

●

●●●

●●●

●

●●

●

●●

●●

●●

●

●●

●

●●

●●

●

●

●●

●●

●

●

●

●

●

●

●●●●●●●

●

●

●●

●●

●●

●

●

●

●●●

●

●●

●

●

●●

●●

●

●

●

●●

●●●

●

●

●●●

●●●

●●

●

●●

●

●●

●●

●

●●

●●

●●

●

●●

●●

●●

●●●●

●●

●

●●

●

●

●

●

●

●

●●

●●

●

●

●●

●

●

●●

●

●●

●

●

●●

●●

●

●●

●

●

●●●

●

●●●

●

Fem

ale

Mal

e

5075100

125

150

175 5075100

125

150

175

1020

3040

50A

ge (y

ears

)

Fork length (cm)

Are

a●

2A 2B 2C 3A 3B 4A 4B 4C 4D

Figu

re1.

Est

imat

ed g

row

th cu

rves

by

regu

lato

ry a

rea

and

sex

base

d on

fi tt

ed m

odel

s to

size

-at-

age d

ata

colle

cted

in th

e 201

1 se

tline

su

rvey

.


Figu

re 2

. Rel

ativ

e eq

uilib

rium

yie

ld v

ersu

s fi s

hing

mor

talit

y in

the

dire

cted

fi sh

ery

assu

min

g st

eepn

ess h

=0.7

5, a

min

imum

size

lim

it of

81.

3 cm

, and

no

othe

r so

urce

s of

add

ition

al m

orta

lity.

Ver

tical

arr

ows

poin

ting

to th

e ax

is in

dica

te th

e es

timat

ed F

for

each

reg

ulat

ory

area

.

●

●

●

●●

●

●

●

●

●

●

●

●

●

0.0

2.5

5.0

7.5

10.0

0.00

0.25

0.50

0.75

1.00

Fish

ing

mor

talit

y ra

te

Relative yield

Are

a●

2A 2B 2C 3A 3B 4A 4B 4C 4D


Figu

re 3

. Rel

ativ

e eq

uilib

rium

yie

ld v

ersu

s fi s

hing

mor

talit

y fo

r al

l 9 s

cena

rios

des

crib

ed in

Tab

le 2

, whe

re t

he d

iffer

ent

line

colo

rs d

enot

es R

egul

ator

y ar

ea.

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●●

●●

●

●

●●

●

●

●

●

●

●

●

●●

●●

●

●●

●

●

●

●

●●

●●

●●

●

●

●●

●

●

●

●

●

●●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●●

●

●

●

●●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●●

●●

●●

●

12

3

45

6

78

9

0.0

2.5

5.0

7.5

10.0 0.0

2.5

5.0

7.5

10.0 0.0

2.5

5.0

7.5

10.0

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

Fish

ing

mor

talit

y ra

te

Relative yield

Are

a●

2A 2B 2C 3A 3B 4A 4B 4C 4D


Figu

re 4

. Per

cent

was

tage

in th

e di

rect

ed fi

sher

y ve

rsus

long

-ter

m e

quili

briu

m fi

shin

g m

orta

lity

rate

for

each

reg

ulat

ory

area

. Sc

enar

io 1

con

sist

s of

a m

inim

um s

ize

limit

of 8

1.3

cm, s

cena

rio

2 in

clud

es a

n ad

ditio

nal m

axim

um s

ize

limit

of 1

40 c

m, a

nd

scen

ario

4 c

onsi

sts o

f a 8

1.3

cm m

inim

um si

ze li

mit

and

com

mer

cial

sele

ctiv

ity sh

iftin

g 10

cm

tow

ards

smal

ler fi s

h.

●

●

●

●

●

●

●

●

●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●●

●●

●●

●

●

●

●

●

2A2B

2C

3A3B

4A

4B4C

4D

010203040 010203040 010203040

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

Fish

ing

mor

talit

y ra

te

Percent wastage (dead discards/landed catch)

Scen

ario

●1 2 4


Figu

re 5

. Exp

ecte

d ch

ange

in m

ean

wei

ght-

at-a

ge fo

r fe

mal

e ha

libut

by

regu

lato

ry a

rea

for

ages

6, 1

2, 1

4, a

nd 2

0 ye

ars.

2A2B

2C

3A3B

4A

4B4C

4D

0204060 0204060 0204060

0.0

0.2

0.4

0.6

0.0

0.2

0.4

0.6

0.0

0.2

0.4

0.6

Fish

ing

mor

talit

y ra

te

Mean weight−at−age (lb)

Age

6 12 14 20


Figu

re 6

. Obs

erve

d di

stri

butio

n of

wei

ght a

t age

s 6,1

2 an

d 14

yea

rs fo

r sam

pled

fem

ale P

acifi

c hal

ibut

in th

e set

line s

urve

y be

twee

n 19

98 to

201

1. In

terp

olat

ion

base

d on

loes

s sm

ooth

ing.

● ●

● ● ●●● ● ●● ● ●

●● ●● ●● ●●●● ● ●● ●

●●

●●●

●●●●●●● ● ●●

●●●●●●●●● ●●●●●●●●●●●●● ●●●●●● ●●●●●●●●●●●●●●

●●

●●●●

●●●●●● ●●●● ●●●●●●●●●●●●●●●●●●● ●●●●●●●

● ●●●●●● ●●●●

●

●● ●●●●●●

●●●

●

● ●●

●● ● ●●●● ●●

● ●●●●●

●●● ● ●

●●● ●●

●

●● ●●●●●●●●●

●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● ●●●●●●●●●●●●●●

● ● ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●

●●●●●● ●●●●●●● ●●● ●●●●●

●●● ●●●●●●●●●●●●●●●●●●● ●●●

●

● ● ● ●●

●●

●●

●●● ●●●●●●

●●●

●●●●

●

●●● ●●●●●●

●

●●●●● ●●

●●●●●

●●● ●●●●●●

●

●●

● ●●●●

●

●●●

●●●●●

●●●●●●●●●●●●●●●● ●●●●●●●●●●

●

●●●●●● ●●●●●●●●●●●●●●

●●●●●● ●●●●

●●

● ●●

●

●●●●

● ●●● ●●● ●●●

●●●●● ●

● ●●● ●●

●●●●

●●

●● ●●●●●●● ●●●

●●●●●●●●●●●●●●●● ●●●●●●●●●●●● ●●●●●●●●●●●●●●●●

●●● ●●●●

●

2B2C

3A

3B4A

4B

0255075 0255075

2000

2004

2008

2000

2004

2008

2000

2004

2008

Year

Net weight (lb)

Age ●●

6 10 14


Figu

re 7

. Obs

erve

d an

d pr

edic

ted

leng

th-a

t-ag

e by

reg

ulat

ory

area

and

sex

for

the

2011

setli

ne su

rvey

dat

a.

2A2B

2C

3A3B

4A

4B4C

4D

50100

150

200 50100

150

200 50100

150

200

1020

3040

5010

2030

4050

1020

3040

50A

ge (y

ears

)

Fork length (cm)

Obs

erve

dFe

mal

e

Mal

e

Pred

icte

dFe

mal

e

Mal

e


Coastwide comparison of alternative setline survey baits

Raymond A. Webster, Steven M. Kaimmer, Claude L. Dykstra and Bruce M. Leaman

Abstract

A coastwide comparison of the current setline survey bait, chum salmon, with two potential replacements, pollock and pink salmon, showed differences in O32 halibut WPUE, U32 catch rates, rates of bait return, bycatch rates, and rates of baits missing on setting. There was also evidence for differences among the length and age distributions of halibut captured using the three baits. These differences often varied greatly among regulatory areas. Any transition to a new survey bait must allow for the collection of suffi cient information to account for these differences and their effects on the IPHC stock assessment.

Introduction

Since the stock assessment survey was standardised in its current layout in 1998, the International Pacifi c Halibut Commission (IPHC) has been using chum salmon (Oncorhynchus keta) exclusively as bait for its annual setline survey. The IPHC’s minimum quality requirement is #2 semi-bright, Alaska Seafood Marketing Institute (ASMI) grade A through E, H&G (headed and gutted), IQF (individually quick frozen) chum with meat-colored fl esh. With the price of chum increasing and availability decreasing, we wish to consider alternative baits. Before replacing the current bait, it is important that we compare it with possible alternatives to ensure that the survey index will not be affected by the change, or to estimate a correction factor to apply if there is an effect on the index.

In 2011, a small-scale pilot study was undertaken in two setline survey regions (Webster et al. 2012) comparing two alternative designs. The results of the study led us to select a randomised block design that deployed all three baits on a single set. In 2012, the study was conducted coastwide in conjunction with the annual IPHC setline survey.

Methods

Experimental designThis study compared three baits using a modifi cation of an experimental design selected in the

2011 pilot study (Webster et al. 2012). The baits were chum salmon, pink salmon (Oncorhynchus gorbuscha), and walleye pollock (Theragra chalcogramma). The chum salmon were of or exceeded the minimum quality requirement as detailed above. The majority of pollock was incidentally captured during the 2012 Pacifi c cod A-season (from the F/V Aleutian Lady, a freezer longliner), and was J-cut, frozen at sea, and supplied in 40 lb fi bre bags. Due to a distribution error at the storage facility in Dutch Harbor, a second pollock product type was used in some areas: whole, round, frozen pollock (200-800 gm/fi sh, frozen at sea, provided in 60 lb fi ber bags, and sourced from the F/V Victory – a cod longliner), were supplied to two boats and used in all of 4B (Adak and Attu regions), 55 stations out of 66 for Unalaska (F/V Van Isle trips 5-7), and 18 stations out of 46 for Portlock (F/V Van Isle trip 8). Vessels fi shing with the whole, round pollock were instructed to


remove and discard the heads when baiting up. The pink salmon used on all vessels were #1 IQF (3 lbs up) sourced from multiple providers in multiple locations. All baits were purchased frozen and were then partially to fully thawed before cutting and baiting. All gear was hand baited. The crew was responsible for cutting the bait into pieces weighing between 0.25 and 0.33 pounds each. Pollock were generally cut in halves or thirds to achieve the bait weight standard on each hook. Care was taken to keep the bait size consistent across all sets. IPHC staff monitored bait size during the charter to ensure compliance with charter standards.

We used a randomised block design, with all three baits in a random order on a single set. In this design, sets of eight skates were used as experimental blocks, with each set having baited skates of each bait type alternating, and separated by full length unbaited skates. There was concern that the bait plume of each baited skate could extend across adjacent skates, potentially affecting their catch. However, we found no evidence for this in the 2011 pilot study when using the unbaited intermediate skates, hence we proceeded with this design in the full coastwide study.

In order to simultaneously run the standardised setline survey and conduct this study, four consecutive chum salmon skates were used on each set, along with one each of pollock and pink salmon. Typically, the setline survey has used from fi ve to eight skates at each station, but to accommodate the experimental skates on the same set, it was decided that a maximum of four skates of chum salmon was feasible.

A previous analysis (Webster and Hare 2011) showed that the outermost (end) skates of a set have higher WPUE of halibut on average than the interior skates, an observation known as the ‘end skate effect’. In order to ensure a fair comparison of chum salmon with the two baits used on single 100-hook skates, survey staff recorded which halibut were caught on the fi rst and last 50 hooks of the chum skates section. Data from only these halibut were used in the treatment comparisons. For other comparisons, such as missing baits, bycatch, and returned baits, data from the fi rst and fi nal chum skates were used averaged.

As in the pilot study and the stock assessment survey, a standard skate of gear was defi ned as an 1800-foot skate with one hundred #3 (16/0) circle hooks spaced 18 feet apart.

Statistical analysisWe fi tted linear models to test for differences among baits, and whether those differences

varied among regulatory areas or not (i.e., a bait by area interaction effect). Variables examined were O32 halibut WPUE, proportions of hooks catching U32 halibut, proportion of hooks with returned baits, and proportion of baits falling off hooks during setting (missing baits). All analyses included survey set as a blocking variable.

WPUE was computed by dividing the halibut catch by the number of effective skates. The number of effective skates was computed as the number of skates set, adjusted for missing hooks (hook counts less than 100). Missing baits have not previously been deducted from hook totals when computing setline survey WPUE, and we follow that protocol here. However, in the pilot study, pollock had a much higher proportion of missing baits than chum salmon (around 0.03 compared with 0.005 or less for chum). Similar differences would have the potential to affect comparisons among bait types in this study, and therefore an analysis of missing baits upon setting was also performed.

The O32 WPUE data were analysed by fi tting general linear models and testing for signifi cant effects through an analysis of variance (ANOVA). WPUE was fi rst transformed using a Box-Cox transformation with λ = 0.4 to more closely satisfy the normality and constant variance assumptions


of the model. Counts on U32 fi sh (those with fork lengths less than 32 inches), returned baits, and missing baits were analysed using a generalized linear model (GLIM) with binomial family and logit link function (logistic regression), using the same independent variables as for the O32 WPUE analyses. That is, counts of U32 fi sh per skate were assumed to have a binomial distribution, with population size being the number of effective hooks on each skate. Effects were tested using F tests that allow for over- or under-dispersion often present in such binomial count data.

We also tested for two effects of interactions among adjacent baits, as we did in the 2011 pilot study. Each baited skate could potentially affect the catch of one or both of its neighbours, and so the two effects each accounted for one direction: the fi rst was an effect on the baited skates set immediately after, and the second an effect on the baited skate set before. As discussed in Mead (1988), two tests must be done: the fi rst is a test of the interaction effect after accounting for the bait effect, and the second is a test of the bait effect after accounting for the interaction effect. The test of interaction effects presented here is of the combined effect of the two possible directions the effect could have.

For the major variables of interest, we also performed post-hoc pairwise comparison t-tests between chum salmon and each of the alternative baits for each regulatory area. In doing these tests, one approach is to fi x the global signifi cance level, α, so that the probability of getting a signifi cant result on at least one test equals α (=0.05, for example). This would be the approach of methods such as Tukey’s pairwise comparisons, and ensures a low probability of incorrectly concluding there is any statistically signifi cant difference when there is none (Type 1 error). The drawback of this is that it also increases the probability of concluding that there is no difference when there is one (Type II error). When considering the switch to a new bait, we felt that it was important to avoid concluding the change would have no effect when it in fact it would, and for this reason we did not adjust the p-values of the pairwise comparisons in the manner of methods such as Tukey’s tests.

The proportion of hooks returned with baits is used to adjust survey WPUE for hook competition (Clark 2008, Etienne et al. 2010). We computed adjusted WPUE for the three baits to assess the effect of this adjustment on bait differences.

Preliminary comparisons of the length and age distributions of halibut captured on the alternative baits with those caught on chum baits were also undertaken using simulated versions of Fisher’s Exact Test for each regulatory area. This test assumes independence of all sampled halibut, and therefore does not account for the nested nature of the sampling design. Nevertheless, the results can still be used to highlight potentially important differences among baits.

Results

Halibut catchThere was very strong evidence that O32 WPUE differed among the three baits, and that these

differences varied among regulatory areas (Table 1). Figure 1 shows that in general, WPUE of pollock was higher than that of other baits in Gulf of Alaska areas (Areas 2A to 3B), but lower in parts of Area 4. WPUE of pink salmon was generally closer to that of chum salmon than pollock, although it was slightly, but signifi cantly lower in Areas 3A and 3B.

Comparison of the proportions of hooks that caught U32 halibut among baits differs from the O32 WPUE comparison, although again there is very strong evidence for bait differences that vary by area (Table 1). In almost all Gulf of Alaska areas and Area 4A, pink salmon caught signifi cantly


fewer halibut than chum salmon (Fig. 2). Catch of U32 halibut using pollock baits was closer to chum salmon in most areas, but was lower in Areas 2B, 2C and most dramatically, in Area 4B.

For neither O32 WPUE nor U32 catch rates was there clear evidence for interactive effects on catch of using different baits together on the same set (Table 1).

For Area 4B, the comparison of WPUE between pollock and other baits was confounded with any effect of using a non-standard pollock bait in that area, i.e., we do not know if the low WPUE using pollock was something we would have observed had we used the standard (j-cut) pollock bait type. Only some sets in Areas 3A (18/369 sets) and 4A (51/107 sets) used the non-standard pollock bait, so at least in these areas, we can compare sets that used each type of pollock bait. Differences among the three baits types are similar for sets that used each type of pollock bait (Table 2), and although these comparisons are also confounded with geographical differences, they imply that the low catch using pollock in Area 4B is not due to the product type of pollock used.

Returned baits and bycatchVery large differences were observed among the mean proportion of hooks returning with

baits still on them (Table 1 and Fig. 3). Most striking is the large fraction of pollock baits returned, with return rates of 0.4-0.7% in many areas, up to eight times the rate of chum salmon returns. Pink salmon return rates were lower on average than those of chum salmon.

A large factor in the high rates of bait return for pollock was much lower catch rates of bycatch species (Fig. 4). Pollock baits have consistently the lowest bycatch rates across all areas, and this appears to be largely due to lower catch rates of the two main bycatch species, spiny dogfi sh (Areas 2A to 3A) and Pacifi c cod (Areas 3A to 4D). Bycatch rates using pink salmon are lower than chum salmon in Areas 2B, 2C, and 3A and similar elsewhere.

We also compared the proportion of hooks with returned baits between sets using standard and non-standard pollock bait. As with WPUE, the patterns were similar for both groups of sets (Table 2). The proportion of hooks with returned baits was much lower on the sets in Area 3A using whole pollock, but as there were only 18 sets here, it is hard to draw clear conclusions from these data.

Adjusted WPUEThe proportion of baits returned was used to compute O32 WPUE adjusted for hook

competition. In previous work (Clark 2008), the area-specifi c adjustments were scaled so that there was no adjustment at the coastwide level. In this analysis, we did not scale the area-specifi c adjustments for each bait, thus enabling adjusted WPUE to be compared among baits. Therefore, the adjusted values in Figure 5 are calculated in almost the same way as that proposed by Etienne et al. (2010), except that we are applying adjustments to WPUE, not numbers per unit effort (NPUE) (see also Webster et al. 2011, pages 235-237).

After adjusting for hook competition, O32 halibut WPUE becomes lower for pollock than chum salmon in all areas except Area 2A, whereas unadjusted values showed higher WPUE when using pollock in the Gulf of Alaska (Fig. 1). WPUE of pink and chum salmon are similar in the Gulf of Alaska, but pink has higher adjusted WPUE than chum salmon in Areas 4B and 4C.

Missing baitsThe proportion of baits that were observed falling off hooks during setting, known as missing

baits, was low for all bait types (Figure 6), generally less than 0.01, and no more than 0.02. Missing


baits are not used in the calculation of effective skates when computing WPUE, and we have not done so here (although we did in the pilot study’s analysis because proportions were much higher for the alternative baits). While there were clear differences in the proportions among baits (Table 1 and Figure 6), with pollock and pink salmon tending to have higher proportions than chum salmon, all proportions are very low, and adjusting the numbers of hooks fi shed for missing baits in the analyses would make little difference to the comparisons of bait types discussed above.

Interestingly, this was the only variable which had evidence for interactive effects from neighbouring baits (Table 1), that is, the proportion of missing baits for a skate of gear depended on the type of bait used on adjacent baited skates. More detailed examination of this effect showed that it was uni-directional: on an average set, a bait type’s rate of missing baits was affected by the bait type set prior to it, not the one that followed. The mean rates (not presented) indicate that missing bait rates are higher when a bait treatment did not follow a baited skate, i.e., for the fi rst skate (or skates, for chum) of each set. We suspect this is related to the drag or lack of it on the different treatments (including the blank skates), which affects sink rate of the gear already deployed and thereby the speed of the hooks leaving the setting chute. The faster the hooks leave the setting chute, the more likely it is for bait to fall off the hooks.

Length and age distributionsFigures 7 and 8 compare the length distributions of halibut captured using pollock and pink

salmon respectively with that of chum salmon, while Figures 9 and 10 make the same comparisons for the age distribution of halibut. There is evidence for differences between the length and age distributions of each alternative bait and chum salmon for at least some regulatory areas. For most Gulf of Alaska areas (Areas 2B to 3B) and Area 4C, pollock catches fewer small fi sh than chum salmon (Fig. 7). There is no evidence of differences between the length distributions of pollock and chum salmon for Areas 4A, 4B and 4D, and only weak evidence for Area 2A. The length distributions of halibut caught using pink and chum salmon are generally very similar, with only Area 3B having statistically signifi cant distributions (Fig. 8). It appears that chum salmon catches more very small fi sh in this area, while pink salmon catches more fi sh above 85 cm in length.

There is less evidence for differences in age distributions. Comparing pollock and chum salmon (Fig. 9), while there appears to be a tendency for younger fi sh to be caught using chum salmon in Area 2, the statistical tests showed only weak evidence for a difference in Areas 2A and 2C. There was evidence for a difference in age distributions for Area 4A (younger fi sh when using chum salmon) and Area 4C (due to the large proportion of age 9-10 fi sh when using pollock). Comparing pink and chum salmon (Fig. 10), as with length, the age distributions are generally close. There is some evidence of differences for Areas 2C (younger fi sh when using chum salmon) and 4B (chum has a stronger mode at 9-10 years), and weak evidence for Area 4C.

Adjustment factors The ratio of O32 halibut WPUE using chum salmon and each alternative bait provides

an estimate of a scaling factor that can be applied to convert WPUE of the alternative bait to WPUE of the standard chum salmon bait. These ratios along with standard errors estimated from bootstrapping are presented in Table 3.


Discussion

This study has found clear differences between the catch of chum salmon and each of the proposed alternative survey baits, walleye pollock and pink salmon. Further, these differences vary greatly among regulatory areas. Any change to an alternative setline survey bait will require careful accounting for these differences. This study does provide scaling factors with which to adjust O32 WPUE for either alternative bait so that it is comparable to WPUE of chum salmon. However, this study does not provide any information on whether there are temporal differences in the relative performance of the three baits. Just as we observe great spatial variability in WPUE, it seems likely that catch rates may also vary with time, depending on factors such as availability of different food sources that could make each bait more or less appealing in a given year relative to other baits.

A change to another survey bait would also require the stock assessment model to estimate different values for catchability and selectivity for length (or age) for chum salmon and the selected replacement bait. Obtaining suffi cient information for estimating differences in these parameters will likely require using chum salmon and the replacement bait together on the setline survey for one or more additional years.

References

Etienne M. P., Obradovich S. Yamanaka L. and McAllister M. 2010. Extracting abundance indices from longline surveys: method to account for hook competition and unbaited hooks. Preprint arXiv:1005.0892v2 (submitted CJFAS).

Clark, W.G. 2008. Effect of hook competition on survey CPUE. Int. Pac. Halibut Comm. Report of Research and Assessment Activities 2007: 211-215.

Mead, R. 1988. The design of experiments: statistical principles for practical application. Cambridge University Press, Cambridge.

Webster, R.A. and Hare, S. R. 2011 Adjusting IPHC setline survey WPUE for survey timing and hook competition. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2010: 251-259.

Webster, R.A., Hare, S. R., Valero, J. L., and Leaman, B. M. 2011. Notes on the IPHC setline survey design, alternatives for estimating biomass distribution, and the hook competition adjustment Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2010: 229-240.

Webster, R.A., Kaimmer, S., Leaman, B. M., and Dykstra, C.L. 2012. Bait comparison pilot study. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2011: 561-576.


Table 1. Statistical tests of effects.

Effect O32 WPUE U32 countReturned

baits Missing baits

Set F1226,2430=5.80 p<0.001

F1226,2430=11.8 p<0.001

F1226,2430=5.10 p<0.001

F1226,2430=2.58 p<0.001

Bait F2,2430=22.8 p<0.001

F2,2430=34.4 p<0.001

F2,2430=1277 p<0.001

F2,2430=321 p<0.001

Area*BaitF16,2430=5.77

p<0.001

F16,2430=5.23

p<0.001

F16,2430=17.5

p<0.001

F16,2430=5.35

p<0.001

Interactions F6,2430=1.88 p=0.081

F6,2430=1.58 p=0.15

F6,2430=1.57 p=0.15

F6,2430=3.91 p=0.001

Table 2. Comparison of mean O32 WPUE (lbs/skate) and proportion of hooks returned with baits on sets that used standard pollock baits and those that used the non-standard, whole pollock baits.

Variable Reg AreaStandard pollock Whole pollock

SA PO PS SA PO PS

WPUE3A 154 192 140 189 207 1254A 71 68 65 81 72 66

Returned baits

3A 0.051 0.241 0.035 0.046 0.089 0.0334A 0.069 0.516 0.078 0.046 0.435 0.025


Table 3. Estimated adjustment factors for converting pollock and pink salmon WPUE to chum salmon WPUE, with standard errors (estimated from 1000 bootstrap resamples) in parentheses.

Reg Area

Raw WPUE WPUE with hook adjustment

Pollock Pink salmon Pollock Pink salmon

2A 0.83 (0.15) 1.05 (0.19) 0.98 (0.18) 0.92 (0.19)

2B 0.76 (0.06) 1.19 (0.11) 1.18 (0.10) 0.88 (0.09)

2C 0.77 (0.07) 0.91 (0.07) 1.45 (0.16) 0.83 (0.10)

3A 0.81 (0.04) 1.12 (0.05) 1.34 (0.07) 1.01 (0.06)

3B 0.89 (0.05) 1.14 (0.06) 1.57 (0.10) 0.84 (0.06)

4A 1.08 (0.13) 1.15 (0.14) 2.32 (0.31) 1.12 (0.17)

4B 2.02 (0.38) 0.74 (0.14) 2.72 (0.55) 0.61 (0.12)

4C 1.74 (0.98) 1.09 (0.22) 2.87 (1.39) 0.69 (0.16)

4D 0.92 (0.29) 1.28 (0.32) 1.58 (0.52) 1.09 (0.26)


Figure 1. Average O32 WPUE by bait type and regulatory area. Symbols above alternative baits indicate the p-value range for pairwise comparisons with chum salmon (see text): *** p≤0.001; ** 0.001<p≤0.01; * 0.01<p≤0.05; • 0.05<p≤0.1; NS p>0.1.

SA PO PS0

10

20

30

40

502A

NSNS

SA PO PS0

50

100

150

2B

***

•

SA PO PS0

50

100

150

200

250

2C

**NS

SA PO PS0

50

100

150

200

3A

***

**

SA PO PS0

20406080

100120

3B

NS

*

SA PO PS0

20

40

60

80

4A

NS NS

SA PO PS0

20

40

60

80

4B

***

NS

SA PO PS0

10

20

30

40

50

4C

•

NS

SA PO PS0

10

20

30

404D

NS

NS

Bait

WPU

E (lb

s/sk

ate)


Figure 2. Average proportion of hooks with U32 halibit, by bait type and regulatory area. Symbols above alternative baits indicate the p-value range for pairwise comparisons with chum salmon (see text): *** p≤0.001; ** 0.001<p≤0.01; * 0.01<p≤0.05; • 0.05<p≤0.1; NS p>0.1.

SA PO PS0.000

0.002

0.004

0.006

0.008

0.010

2A

NS

*

SA PO PS0.00

0.01

0.02

0.03

0.04

2B

** ***

SA PO PS0.00

0.01

0.02

0.03

0.04

0.05

2C

**•

SA PO PS0.00

0.02

0.04

0.06

0.08

0.10

3A

NS***

SA PO PS0.000.020.040.060.080.100.120.14

3B

NS

***

SA PO PS0.000.010.020.030.040.050.060.07

4A

NS***

SA PO PS0.000

0.005

0.010

0.015

0.0204B

***

NS

SA PO PS0.000

0.005

0.010

0.015

0.020

4C

NS

NS

SA PO PS0.000

0.005

0.010

0.015

0.020

4D

• NS

Bait

Prop

ortio

n of

hoo

ks

Bait


Figure 3. Average proportion of hooks with returned baits, by bait type and regulatory area. Symbols above alternative baits indicate the p-value range for pairwise comparisons with chum salmon (see text): *** p≤0.001; ** 0.001<p≤0.01; * 0.01<p≤0.05; • 0.05<p≤0.1; NS p>0.1.

SA PO PS0.00

0.05

0.10

0.15

0.20

0.252A

***

***

SA PO PS0.0

0.1

0.2

0.3

0.4

0.52B

***

***

SA PO PS0.0

0.1

0.2

0.3

0.4

2C

***

NS

SA PO PS0.00

0.05

0.10

0.15

0.20

0.25

3A

***

**

SA PO PS0.0

0.1

0.2

0.3

0.43B

***

***

SA PO PS0.0

0.1

0.2

0.3

0.4

0.5

4A

***

NS

SA PO PS0.00.10.20.30.40.50.60.7

4B

***

***

SA PO PS0.0

0.2

0.4

0.6

0.8

4C

***

***

SA PO PS0.0

0.2

0.4

0.6

4D

***

**

Bait

Prop

ortio

n of

hoo

ks


Figu

re 4

. Av

erag

e pr

opor

tion

of h

ooks

with

byc

atch

, by

spec

ies,

bait

type

and

reg

ulat

ory

area

.

SA

PO

PS

0.00

0.02

0.04

0.06

0.08

2A

SA

PO

PS

0.00

0.05

0.10

0.15

2B

SA

PO

PS

0.00

0.05

0.10

0.15

2C

SA

PO

PS

0.00

0.05

0.10

0.15

0.20

3A

SA

PO

PS

0.00

0.05

0.10

0.15

0.20

3B

SA

PO

PS

0.00

0.05

0.10

0.15

0.20

4A

SA

PO

PS

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

4B

SA

PO

PS

0.00

0.05

0.10

0.15

0.20

0.25

0.30

4C

SA

PO

PS

0.00

0.05

0.10

0.15

4D

Bai

t

Proportion of hooks

Paci

fic C

od

Sabl

efis

h (B

lack

cod)

Spin

y D

ogfis

h

Ala

ska

Skat

e

Long

nose

Ska

te

Yello

w Ir

ish

Lord

Gre

at S

culp

in

Whi

tebl

otch

ed S

kate

Oth

er


Figure 5. O32 WPUE adjusted for hook competition, by bait type and regulatory area.

SA PO PS0

20

40

60

80

100

2A

SA PO PS0

50100150200250300350

2B

SA PO PS0

100200300400500600700

2C

SA PO PS0

100

200

300

400

500

6003A

SA PO PS0

100

200

300

400

3B

SA PO PS0

50

100

150

200

250

4A

SA PO PS0

50

100

150

4B

SA PO PS0

50

100

1504C

SA PO PS0

20

40

60

804D

Bait

Adj

uste

d W

PUE

(lbs/

skat

e)


Figure 6. Average proportion of hooks with baits missing on setting, by bait type and regulatory area.

SA PO PS0.000

0.002

0.004

0.006

0.008

0.010

0.0122A

SA PO PS0.000

0.005

0.010

0.015

0.020

2B

SA PO PS0.0000.0010.0020.0030.0040.0050.006

2C

SA PO PS0.000

0.005

0.010

0.015

0.020

3A

SA PO PS0.0000.0020.0040.0060.0080.0100.012

3B

SA PO PS0.000

0.002

0.004

0.006

0.008

4A

SA PO PS0.0000.0010.0020.0030.0040.0050.006

4B

SA PO PS0.000

0.001

0.002

0.003

0.004

4C

SA PO PS0.000

0.001

0.002

0.003

0.004

4D

Bait

Prop

ortio

n of

hoo

ks


Figure 7. Length distribution of halibut captured using Pollock (green bars) compared with chum salmon (red bars). Fish are grouped by 5 cm length bins from 70 to 130 cm, with <70 cm bin on the left, and a ≥130 cm bin on the right of each fi gure.

<70 80 95 110 1250.00

0.05

0.10

0.15

p = 0.088

2A

<70 80 95 110 1250.00

0.05

0.10

0.15

p < 0.001

2B

<70 80 95 110 1250.00

0.04

0.08

0.12 p < 0.001

2C

<70 80 95 110 1250.00

0.10

p < 0.001

3A

<70 80 95 110 1250.00

0.10

0.20 p < 0.001

3B

<70 80 95 110 1250.00

0.10

0.20 p > 0.1

4A

<70 80 95 110 1250.00

0.10

p > 0.1

4B

<70 80 95 110 1250.00

0.15

0.30 p = 0.002

4C

<70 80 95 110 1250.00

0.10

0.20 p > 0.1

4D

Fork length (cm)

Prop

ortio

n of

sam

pled

hal

ibut


Figure 8. Length distribution of halibut captured using pink salmon (pink bars) compared with chum salmon (red bars). Fish are grouped by 5 cm length bins from 70 to 130 cm, with <70 cm bin on the left, and a ≥130 cm bin on the right of each fi gure.

<70 80 95 110 1250.00

0.05

0.10

0.15

p > 0.1

2A

<70 80 95 110 1250.00

0.05

0.10

0.15

p > 0.1

2B

<70 80 95 110 1250.00

0.04

0.08

0.12 p > 0.1

2C

<70 80 95 110 1250.00

0.10

p > 0.1

3A

<70 80 95 110 1250.00

0.10

0.20 p < 0.001

3B

<70 80 95 110 1250.00

0.10

0.20 p > 0.1

4A

<70 80 95 110 1250.00

0.10

p > 0.1

4B

<70 80 95 110 1250.00

0.15

0.30 p > 0.1

4C

<70 80 95 110 1250.00

0.10

0.20 p > 0.1

4D

Fork length (cm)

Prop

ortio

n of

sam

pled

hal

ibut


Figure 9. Age distribution of halibut captured using Pollock (green bars) compared with chum salmon (red bars). Fish are grouped by 2 year age bins from age 7 to 24, with a <7 years bin on the left, and a ≥25 years bin on the right of each fi gure. The number below each 2-year bin is the lesser year of that bin (e.g., “9” = 9 & 10).

<7 9 13 17 210.00

0.10

0.20

0.30

p = 0.096

2A

<7 9 13 17 210.00

0.10

0.20

0.30

p > 0.1

2B

<7 9 13 17 210.00

0.10

0.20

0.30

p = 0.06

2C

<7 9 13 17 210.00

0.10

0.20

0.30

p > 0.1

3A

<7 9 13 17 210.00

0.10

0.20

p > 0.1

3B

<7 9 13 17 210.00

0.10

0.20

0.30

p = 0.019

4A

<7 9 13 17 210.00

0.10

0.20

0.30

p > 0.1

4B

<7 9 13 17 210.0

0.2

0.4

0.6

p < 0.001

4C

<7 9 13 17 210.0

0.1

0.2

0.3

0.4

p > 0.1

4D

Age

Prop

ortio

n of

sam

pled

hal

ibut


Figure 10. Age distribution of halibut captured using pink salmon (pink bars) compared with chum salmon (red bars). Fish are grouped by 2 year age bins from age 7 to 24, with a <7 years bin on the left, and a ≥25 years bin on the right of each fi gure. The number below each 2-year bin is the lesser year of that bin (e.g., “9” = 9 & 10).

<7 9 13 17 210.00

0.10

0.20

0.30

p > 0.1

2A

<7 9 13 17 210.00

0.10

0.20

0.30

p > 0.1

2B

<7 9 13 17 210.00

0.10

0.20

0.30

p = 0.038

2C

<7 9 13 17 210.00

0.10

0.20

0.30

p > 0.1

3A

<7 9 13 17 210.00

0.10

0.20

p > 0.1

3B

<7 9 13 17 210.00

0.10

0.20

0.30

p > 0.1

4A

<7 9 13 17 210.00

0.10

0.20

0.30

p = 0.006

4B

<7 9 13 17 210.0

0.2

0.4

0.6

p = 0.075

4C

<7 9 13 17 210.0

0.1

0.2

0.3

0.4

p > 0.1

4D

Age

Prop

ortio

n of

sam

pled

hal

ibut


Halibut size at age research1

Bruce M. Leamana, Steven Martella, Timothy Lohera, Kirstin K. Holsmanb, Bruce S. Millerc, Kerim Y. Aydind, and Gordon H. Krusee

aInternational Pacifi c Halibut Commission, bAlaska Fisheries Science Center/Joint Institute for Study of Atmosphere and Ocean (UW), cUniversity of Washington, dAlaska Fisheries Science Center, eUniversity of Alaska

Introduction

The Commission staff is initiating projects to examine the processes and impacts of changes in halibut size at age (SAA) over the history of the halibut fi shery. Causes of variation in size at age are legion and can originate from both natural and anthropogenic causes. It is important to differentiate SAA from individual growth. Most fi shery-based data concern SAA rather than growth because the sizes of the same individuals are not sampled over time. Rather, size at age data for individuals from the same and different cohorts are obtained from sampling. As such, we can only infer lifetime growth for individuals from the same cohort and generally have no knowledge of the growth trajectory of these individuals prior to sampling. A detailed review of mechanisms and potential avenues of investigation for understanding changing SAA and changes in individual growth is provided in Loher (2013). Both that paper and the ongoing concerns about the impacts of changing SAA expressed in IPHC stock assessments and research papers (Clark and Hare 2008, Hare 2011, 2012, Stewart et al. 2013, Valero 2012) have prompted detailed investigation of this issue.

As an internal project, the staff is developing new estimates of age and SAA composition of halibut for the period 1996-2002, a period over which there were changes in IPHC age determination methodology. Age estimates for the 1996-2002 period were originally estimated using a surface ageing methodology, now known to produce biased estimates of age for older fi sh. The IPHC stock assessment currently uses estimates of age composition for this period that are derived from the relationship between surface and section (break and burn, B/B) age estimates (Clark 2004, Clark and Hare 2006). After 2002, all ages were derived from either B/B or break and bake (B/Bk) methodologies, which are considered reliable. The data from this internal project will be used initially within the current IPHC stock assessment to enable assessment of commercial fi shery and survey age data derived using a common methodology for the period after 1996.

A second project aims to employ an integrated approach to examining SAA changes and involves investigators from several agencies and disciplines. This project, summarized below, has been submitted as a proposal to the North Pacifi c Research Board, which has identifi ed halibut SAA as a priority area of investigation.

Background

Pacifi c halibut (Hippoglossus stenolepis) are widespread throughout the north Pacifi c Ocean and have been actively managed by the International Pacifi c Halibut Commission (IPHC) in 1Portions of this report are extracted from proposal #650: Fishery, Climate, and Ecological Effects on Pacifi c Halibut Size-at-age (SAA), submitted to the North Pacifi c Research Board, December 2012.


the Northeast Pacifi c since 1923. Indigenous fi sheries for halibut date back millennia and the commercial fi shery off the west coast of North America began in 1888 (IPHC 1998). Today, the fi shery remains one of the most important commercial fi sheries in the region. Historical IPHC sampling programs have created a record of both statistical and biological data, including a long time series of observations of halibut SAA.

In the past two decades, the SAA of halibut has undergone an extensive reduction (Figure 1; Hare 2011). For some ages, realized SAA in halibut is roughly half that of two decades ago. The ageing methodology used by the IPHC transitioned in 2002 from surface readings to B/B and eventually to B/Bk reading of otoliths. While the reduction in SAA is exaggerated by the evolution of ageing methodology, hence assigned ages, the reduced SAA is also evident in younger age groups (≤ 12 y) for which ageing methodology biases are minimal. On its own, this pattern of change is of considerable biological interest but is of equal signifi cance to the management of halibut fi sheries, because it affects available estimates of halibut biomass and resultant fi shery yield. Moreover, failure to understand and account for this pressing issue could lead to conservation concerns.

The observed changes in SAA over recent years are remarkable but they are of even greater interest when considered over the longer period of observation available in IPHC records. The SAA currently observed is of approximately the same magnitude as that observed in the 1920s. Records of SAA from IPHC research cruises (pre-1960) and surveys (post-1960) show a SAA pattern of small values in the 1920s, rising relatively steadily to historical high values in the 1990s, and the decline to present values. So, while there is considerable interest in the causes of the recent decades-long decline in halibut SAA, it is also pertinent to consider the causes of increased SAA in the earlier period. In recent years, lower recruitment of cohorts spawned in the early and late 1990s combined with decreasing SAA have led to steady declines in the exploitable biomass of halibut stocks, with consequent effects on the yield from the fi shery.

The cause or causes of the changes in halibut SAA are poorly understood. The decrease observed over the past two decades has been posited to result from both intrinsic (e.g., density dependence) and extrinsic (e.g., competition, environmental forcing, bioenergetic changes, anthropogenic impacts) factors; ideally, mechanism(s) advanced to account for change should address both the decrease seen in recent years and the long period of increasing SAA observed historically.

Processes affecting SAA

Fish growth is the integrated outcome of environmental conditions, trophic conditions, ecological interactions, and genetics that infl uence somatic production and size-selective mortality over the lifespan of an individual fi sh. Generally, most fi sh meet metabolic demands before allocating assimilated energy towards somatic or gonadal tissue growth, and exhibit metabolic rates that increase rapidly with water temperature. Thus, slight variability in environmental conditions, food availability, or fi sh behavior can have compound effects on subsequent SAA. Although primarily driven by growth, variation in SAA can also refl ect annual variation in size-selective mortality as increased predation pressure may preferentially remove the smallest individuals from a population or cohort (resulting in an apparent increase of mean size), or intensifi ed harvest pressure may selectively remove the largest or fastest growing fi sh (resulting in an apparent decrease in mean SAA). Hence, changes in growth due to variability in temperature and food availability need to be quantifi ed in order to differentiate fi shery and environmental mechanisms affecting mean SAA of Pacifi c halibut.


Fishing can further infl uence apparent growth rates and SAA of exploited fi sh populations by a variety of direct and indirect processes. First, when harvest commences on an unfi shed stock, total mortality rate increases above baseline natural mortality rates resulting in a shortening of mean lifespan and loss of the oldest or largest individuals in a population. This decline in mean size of fi sh in the catch is a normal response to fi shing, which stabilizes when fi shing mortality rate is held at sustainable levels. Fisheries can be size selective as a result of differences in the area or time of fi shing, size-specifi c gear selectively, and size-specifi c probability of fi sh retention (Pope et al. 1975). Size-selective harvest can alter the demographics of the population if certain sizes are systematically removed or size restrictions are in place. Fishing can also indirectly impact growth and SAA by altering ecological interactions. If prey resources are limited, fi shing can relieve intra-specifi c competition and increase growth rates, age-at-maturity, and recruitment. Density-dependent growth, coupled to a similar reaction of recruitment in response to fi shing (e.g., Ricker 1954), is an important compensatory response that allow fi sh stocks to sustain fi sheries. Conversely, fi shery removals can intensify predation risk from predatory species or cannibalistic older age-classes, either reducing foraging opportunities (which reduces SAA) and/or intensifying size-dependent predation risk (which increases SAA). The degree to which any of these factors infl uence growth is also further modifi ed by environmental conditions that impact the distribution and abundance of prey, predator, and competitive species. There are numerous examples of both increased SAA after intensifi ed fi shing (e.g., Millner and Whiting 1996) as well as decreased SAA from over-harvest (e.g., Helser and Martell 2007, Schweigert et al. 2002). Thus, the magnitude of fi shing effects on individual growth and population SAA is often ecosystem- and species-dependent and subject to non-linear processes that may attenuate or amplify growth.

Genetic selection is another, potentially deleterious effect of fi shing on growth, maturity, and other fi sh traits. In laboratory experiments, size-selective harvest of largest 90th percentile of Atlantic silversides (Menidia menidia) caused selection for fi sh with genotypes with slower growth over just four generations of harvest (Conover and Munch 2002). Occurrences of genetic selection in wild capture fi sheries depend on factors such as harvest rate, selection differential, and the heritability of the fi sh traits under selection. In fi sh populations, the heritability for traits, such as body size and age at maturity, tend to range around 0.2-0.3, which is suffi cient for selection to occur (Law 2000). Attempts to use phenotypic traits to disentangle genetic from other infl uences on wild fi sh populations have been challenging. In northern cod (Gadus morhua) off the coasts of Newfoundland and Labrador, the age of 50% maturity shifted one year younger from the mid-1980s to the mid-1990s (Olson et al. 2004, 2005). However, slower growth was also experienced; slower growth would be expected to delay maturation. Because this expectation is opposite the maturity observations, this was taken as evidence for genetic selection. Lack of recovery of these traits after a decade of a fi shery moratorium was interpreted as further evidence for genetic selection. The analysis made use of a “probabilistic maturation reaction norm” (PMRN) for age and size at maturity, defi ned by the age- or size-specifi c probabilities at which an individual becomes mature during a time period (Olson et al. 2005). These conclusions received some criticisms, including the observation that decreases in size or age of maturation occurred synchronously with declines in sea-surface temperature and increased sea ice cover (Kuparinen and Merilä 2007).

Notwithstanding the pros and cons of PMRN methodology to consider evidence of genetic selection (Dieckmann and Heino 2007), it is diffi cult to apply this approach in the case of Pacifi c halibut because of the current diffi culty in identifying fi rst maturation in the species. The SAA data alone do not lend themselves to this approach; long-term trends in mean weight of age-8


female halibut – low in the 1920s, high in the 1970s, and low in the 2000s (Clark and Hare 2002) – suggests that growth can be variable and have trends rather than suggesting that genetic selection is occurring. Previous analyses also suggest that halibut recruitment is environmentally driven (Clark et al. 1999) and growth variability may be a density-dependent response to changes in stock size (Clark and Hare 2002), although subsequent investigations also suggested the infl uence of inter-specifi c competition (Hare and Clark 2008). Bioenergetics models can be used to predict the effect of changing environmental and trophic conditions on growth and accordingly calculate the residual impacts of fi shery-induced effects. Fish metabolism scales predictably with body size, temperature, and activity, thus bioenergetics models can use straightforward environmental data (i.e., temperature) and fi eld-based diet observations to predict individual fi sh consumption or growth rates and, when scaled to the population level, SAA distributions. Bioenergetic models have long been used to address a variety of ecological questions (Hartman and Kitchell 2008; Ney 1993) and more recently have been applied to assess fi shery impacts on growth (e.g., Meka and Margraf 2007).

This study proposes a comprehensive investigation and analysis of candidate causes for SAA changes in Pacifi c halibut as well as an integrated approach to incorporating SAA dynamics into the assessment and management of the halibut stock. It builds on existing information (particularly the unique historical archive of halibut otoliths maintained by the IPHC), develops new understanding of ecosystem infl uences on growth, assesses the impact of fi shery-induced changes, and creates a fl exible modeling framework to integrate SAA changes into development of optimum harvest policies for Pacifi c halibut. The proposed study is subdivided into four components to meet our objectives:

1. Re-examination of historical halibut ageing structures (prior to 2002) to derive a 80+ year time series of SAA, as well as length and age at maturity observations, using a less biased break and burn aging method for individuals older than age 12;

2. Retrospective spatial analysis of ecological data based on IPHC and NMFS fi eld surveys of relative abundance, diet, and biological characteristics of halibut and co-occurring species, to assess the relative importance of extrinsic and intrinsic factors by area;

3. Bioenergetic modeling of halibut growth and maturity based on physiological studies to test hypotheses concerning changes in SAA; and

4. Integrated modeling of halibut SAA, maturity, potential fi shery effects, and their impacts on stock assessment, harvest policy, and associated yield.

Residuals of the integrated growth model (i.e., variance in growth not explained by density-dependent or bioenergetics-based and environment-dependent effects) will be examined for possible evidence of fi shery/genetic effects. This approach will provide new insights into the factors affecting growth with updated datasets that we will generate in this project. The participating agencies have a unique capability to integrate their historical data with contemporary observations and analytic techniques to develop an optimum approach to incorporating the impacts of halibut SAA changes on assessment and management. Ultimately, the project will also provide broad insights into factors that affect halibut SAA, some of which may be linked directly to harvest policy and management of the Pacifi c halibut resource.


Project components

i. SAA otolith update of Pacifi c halibut samples prior to 2002 We will re-examine archived halibut ageing structures (prior to 2002) to derive an 80+ year

time series of SAA, as well as length- and age-at-maturity observations, using a less biased break and bake (B/Bk) aging method for individuals older than age 12. IPHC historical otoliths archives contain previously-estimated surface ages with associated georeferencing and biometric data. Data are available from commercial fi shery sampling, research cruises, and some surveys. Representative samples of this otolith collection will be re-aged using the B/Bk technique.

ii. Retrospective spatial analysis of trophic and ecological dataPacifi c halibut is one of three major groundfi sh apex predators that have dominated the Gulf of

Alaska (GOA) shelf demersal community since the late 1970s; the others being Pacifi c cod (Gadus macrocephalus) and arrowtooth fl ounder (Atheresthes stomias). The composition of groundfi sh biomass in the GOA has varied tremendously on a decadal scale since the 1970s, in particular with a rise and decline of walleye pollock (Theragra chalcogramma) in the 1980s, followed by the rise of arrowtooth fl ounder from the 1990s to the present. Explanations for the changing patterns include historical fi shing pressure and decadal climate variability, although causality has been extremely diffi cult to determine in either a conceptual or quantitative modeling framework (Gaichas et al. 2011).

Since the early 1980s, researchers at the NOAA Alaska Fisheries Science Center (AFSC), in partnership with project PI Dr. Bruce Miller, collected and analyzed stomach contents of major groundfi sh species in the GOA and eastern Bering Sea (EBS). To date, over 53,000 samples of the three major predators have been collected (Table 1). This is part of a collection of 250,000 samples including walleye pollock, Pacifi c ocean perch (Sebastes alutus), skates, fl atfi sh, and a range of other species. Some of these samples were analyzed as part of NPRB Projects #525, #622, #628, and portions of the Bering Sea and Gulf of Alaska Integrated Ecosystem Research Programs (BSIERP and GOAIERP). Ongoing collections and the database are under the supervision of NOAA project collaborator Dr. Kerim Aydin.

To date, analyses of patterns and trends in Pacifi c halibut diets have not been a focus of research with these data, particularly with respect to trends in halibut growth. A preliminary examination reveals intriguing patterns with respect to the changes in halibut SAA over decades; for example, walleye pollock declined as a percentage of consumption for halibut >75 cm fork length (Fig. 1). However, from such simple analyses, it is not immediately clear how observed dietary changes may relate to changes in individual halibut growth. First, sample sizes have not been historically uniform across the region, with more focus on the western and Central GOA (west of Prince William Sound) than the east and southeast (Table 1), thus limiting comparative analysis and perhaps biasing broad, pooled observations. Second, the snapshot in Figure 1 does not pinpoint any source of growth limitation; it is possible that supply of pollock has decreased, but it is also possible that competition with arrowtooth fl ounder (a pollock specialist) has increased, or that competition at smaller sizes with Pacifi c cod (a benthic predator) limits growth in smaller halibut, therefore limiting lifetime individual growth trajectories. Finally, straight dietary percentages do not necessarily refl ect changes in total ratio, food competition, or actual growth. To approach these questions, the following three food-habits associated activities are planned as part of this project. First, in 2013, the AFSC is planning to conduct its biennial groundfi sh trawl survey of the Gulf


of Alaska, including the collection of groundfi sh stomachs for laboratory analysis. Collections, already planned as part of standard AFSC cruise operations, will include up to 800 additional Pacifi c halibut, focusing on low sample-size areas such as the east and southeast portions of the survey, and potentially including Pacifi c cod and arrowtooth fl ounder collections in these low sample areas. These additional collections will be analyzed in the laboratory as part of this project. Should survey contingency planning limit 2013 collections, analysis will instead focus on previously-collected GOA samples of halibut, cod, and arrowtooth that have not yet been analyzed.

Second, a set of multivariate statistical analyses will be performed both on the diet data and on haul-scale community composition metrics from the historical AFSC GOA groundfi sh survey biomass data (1990-2013) and associated data, such as surface and bottom water temperature measured during the survey. A particular interest is to determine the spatial scale over which variation in diet composition is associated with variation in groundfi sh community composition, especially with respect to local, sub-regional, regional, or interannual changes in prey supply (particularly walleye pollock), or the presence of apex predator competitors. Emphasis will also be placed on comparing the western and central GOA with the eastern and southeastern GOA where trends in SAA have been more stable over time.

Finally, while the diet data can be used to index overall stomach content weight, inferences on whether detected shifts in diet composition also represent shifts in consumption or growth rates will be made by using the historical trends in these data in conjunction with our planned halibut bioenergetics modeling.

iii. Bioenergetics Modeling We propose a 3-step bioenergetics modeling approach:

1. Use bioenergetics models to retrospectively estimate regionally- and annually-specifi c growth and foraging effi ciencies from otolith-based estimates of SAA and fi eld-based observations of temperature and diet;

2. Use generalized linear regressions to quantify the relationships between a variety of fi shing and environmental factors (i.e., harvest pressure, fl eet distribution, inter- and intraspecifi c competition, circulation, stratifi cation, PDO and ENSO indices) and variability in growth and foraging effi ciency (defi ned as the ratio of energy consumed to maximum potential energy consumed); and

3. Use results of these analyses to inform integrative population growth modeling.

Bioenergetic estimates of growth and foraging effi ciencyWe will use the broadly-applied Wisconsin bioenergetics modeling approach (Ney 1990;

Kitchell et al. 1977) developed for Pleuronectes species (Holsman and Aydin in prep.), which estimates temperature- and weight-specifi c maximum daily consumption rates for individual fi sh. Realized individual consumption rates based on in situ fi sh are typically much lower than those observed in the laboratory because inter- and intra-species competition, mismatched prey phenology or distributions, and predator avoidance behaviors by prey species often limit capture and consumption rates. Realized consumption is modeled as a mass balance equation where energy consumed is allocated to growth or lost to metabolism and waste. Temperature will be extracted from trawl surveys temperatures and available environmental data for the EBS and GOA and diet composition data from bottom trawl surveys will be used to estimate predator/prey densities for


each region (e.g., southern EBS and three sub areas of northern GOA). All metabolic and waste functions are parameterized from laboratory experiments or derived from literature values for the species or similar animals.

Foraging effi ciency, defi ned as the ratio of observed to maximum consumption over time, is iteratively fi t through minimization of observed weight and weight by the mass balance equation to predict daily growth. Using weight-at-age derived from otolith-based observations of SAA of individual fi sh, we will estimate individual foraging effi ciencies.

Trophic and environmental effects on foraging effi ciency Foraging effi ciency provides a temperature- and weight-corrected index of growth; e.g., low

values indicate that observed growth is lower than that expected based only on changes in prey composition or water temperature. By exploring spatial and temporal patterns in age-specifi c foraging effi ciency, we can identify demographic and distributional bottlenecks in growth (i.e., areas or ages with low foraging effi ciencies, and explore environmental or anthropogenic drivers of changes in growth and resultant SAA. To evaluate if variability in foraging effi ciency is associated with spatio-temporal patterns in environmental conditions, we will build a full set of generalized mixed effects models, where foraging effi ciency is a function one or more environmental factors (i.e., inter- and intraspecifi c competition, circulation, stratifi cation, PDO and ENSO indices). This approach will allow us to evaluate the strength of temporal trends in foraging effi ciency and resultant SAA and the degree to which declines may be attributed to measurable factors (such as temperature or indices of fi shing pressure). We also plan to include null models (overall intercept–only model), and area specifi c intercept only models. We will rank all models with Akaike’s information criterion (AIC; Burnham & Anderson 2002) and use AIC model selection and averaging to identify the relative weight of each environmental factor in predicting age-specifi c and region-specifi c foraging effi ciencies.

Simulate SAA for integrative modelingUsing a similar approach to step 1, we will use otolith-based estimates of weight at age to

derive expected SAA distributions from retrospective growth trajectories of individual fi sh chosen to represent key geographic regions and time periods within the recent SAA time series. Otolith increment analyses will be conducted on these individuals to generate proxies of somatic growth at age (sensu Chambers and Miller 1995), from age 1 through age at collection. Initially (i.e., during project year 1), a total of 30 age-15 fi sh collected within each of three geographic regions (EBS, central GOA, eastern GOA) will be analyzed, representing the 1977 and 1992 year-classes (i.e., collected in 1992 and 2007, respectively). These year-classes represent periods of relatively high and low observed mean SAA, and correspond to regions for which corresponding stomach samples are available (Table 1). In year 2, we will endeavor to expand sample sizes within these groupings and to add the western GOA. We will develop time- and area-specifi c somatic-otolith length relationships, permitting conversion of individual increment measurements into estimates of somatic growth in length, and subsequently to weight via published length-weight relationships for the species. These growth data will be used in the integrated model. The resulting trajectories will be combined to generate annual region-specifi c distributions for estimated SAA and annual overall population SAA distributions (i.e., regionally-biomass weighted average SAA). Deviations between estimated and observed SAA will inform integrative modeling and provide additional information to aid in model convergence.


iv. Integrated Growth ModelThe objectives of developing an integrated growth model for Pacifi c halibut are two-fold: (1)

hypotheses testing, and (2) fully specify the growth model into the halibut assessment to better quantify uncertainty associated with changes in size-at-age. Model specifi cation will include elements that attempt to explain additional variation in the size-at-age data collected by the IPHC over the past few decades. Possible covariates for growth include intra- and interspecifi c competition (density effects), environmental covariates (changes in temperature, upwelling, etc.). Also, as explained earlier, another hypothesis that could partially explain observed changes in size-at-age is the cumulative effects of size-selective fi shing. In this case, faster growing individuals recruit to the fi shery at younger ages relative to slower growing members of the same cohort. Over time, remaining individuals in the population will consist of slower growing individuals that are less vulnerable to size selective fi shing gear. Such cumulative effects could be signifi cant and give the appearance of declining size-at-age with increasing fi shing mortality rates over time.

Models will be fi t to SAA data by sex in each of the IPHC regulatory areas. We might expect interesting contrasts in SAA models developed for areas of heavier (e.g., Areas 2B and 2C) versus lighter fi shing intensity (e.g., 3A, 3B, 4A, 4B, 4CDE), which may help to elucidate cause and effect. Although the constant harvest rate policy for halibut has been 20% of exploitable biomass, realized harvest rate has exceeded 50% (particularly in 2B and 2C) in some recent years.

To address intraspecifi c competition, we will include halibut biomass estimates. To address interspecifi c competition, we will use biomass estimates of competitors determined from stomach analyses (e.g., arrowtooth fl ounder). A suite of environmental factors thought to affect halibut growth will include the GAK1 time series of subsurface temperatures off Seward, AK (http://www.ims.uaf.edu/gak1/), the Pacifi c Decadal Oscillation (refl ecting broad environmental variability derived from sea surface temperature), and other variables. The model developed will be akin to a generalized additive modeling approach.

A change in the average SAA of Pacifi c halibut has implications for the harvest policy. In short, with declining rates of growth, the optimal fi shing mortality rate that maximizes long-term sustainable yields also declines, and vice versa. Integrating the growth model, along with variables that are correlated with growth, into the annual stock assessment will allow for immediate updates of the harvest policy and ensure that this policy is commensurate with the current growth regime. Moreover, uncertainty in growth can also be integrated into the decision-making framework for providing annual advice on catch quotas.

References

Burnham, K.P. and Anderson, D.R. 2002 Model selection and inference: a practical information-theoretic approach, 2nd ed. Springer, New York, NY

Chambers, C.C. and Miller, T.J. 1995. Evaluating fi sh growth by means of otolith increment analysis: special properties of individual-level longitudinal data. Pp. 155–175 in D.H. Secor, J.M. Dean, and S. Campana (eds.), Recent Developments in Otolith Research, University of South Carolina Press, Columbia.

Clark, W.G. 2004. Statistical distribution of IPHC age readings. Inter. Pac. Hal. Comm. Report of Assessment and Research Activities 2003:99-110.


Clark, W.G. and Hare, S.R. 2008. Assessment of the Pacifi c halibut stock at the end of 2007. Inter. Pac. Hal. Comm. Report of Assessment and Research Activities 2007: 177-204.

Clark, W.G. and Hare, S.R. 2006. Assessment and management of Pacifi c halibut: Data, methods and policy. Inter. Pac. Hal. Comm. Sci. Rep. 83.

Clark, W.G. and Hare, S.R. 2002. Effects of climate and stock size on recruitment and growth of Pacifi c halibut. N. Amer. J. Fish. Mgmt. 22:852-862.

Clark, W.G., Hare, S.R., Parma, A.M., Sullivan, P.J, and Trumble, R.J. 1999. Decadal changes in growth and recruitment of Pacifi c halibut (Hippoglossus stenolepis). Can. J. Fish. Aquat. Sci. 56:242-252.

Conover, D.O. and Munch, S.B. 2002. Sustaining fi sheries yields over evolutionary time scales. Science 297:94-96.

Dieckmann, U. and Heino, M. 2007. Probabilitistic maturation reaction norms: Their history, strengths, and limitations. Mar. Ecol. Progr. Ser. 335:253-269.

Gaichas, S.K., Aydin, K.Y., and Francis, R.C. 2011. What drives dynamics in the Gulf of Alaska? Integrating hypotheses of species, fi shing, and climate relationships using ecosystem modeling. Can. J. Fish. Aquat. Sci. 68: 1553-1578.

Hanson, P.C., Johnson, T.B., Schindler, D.E., and Kitchell, J.F. 1997. Fish Bioenergetics 3.0. University of Wisconsin Sea Grant Institute, Technical Report WISCU- T-97-001, Madison.

Hare, S.R. 2011. Assessment of the Pacifi c halibut stock at the end of 2010. Inter. Pac. Hal. Comm. Report of Assessment and Research Activities 2010:85-176.

Hare, S.R. 2012. Assessment of the Pacifi c halibut stock at the end of 2011. Inter. Pac. Hal. Comm. Report of Assessment and Research Activities 2011:91-194.

Hare, S.R. and Clark, W.G. 2008. 2007 IPHC harvest policy analysis: past, present, and future considerations. Int. Pac. Halibut Comm. Report of Assessment and Research Activities 2007: 275-295.

Helser, T. and Martell, S. 2007. Stock Assessment of Pacifi c Hake (whiting) in US and Canadian Waters in 2007. Pac. Fish. Manage. Counc., Portland, Oreg, 362

Holsman, K.H. and Aydin, K.Y. (in prep.). Field and bioenergetic based daily ration estimates for walleye pollock (Theragra chalcogramma), Pacifi c cod (Gadus macrocephalus), and arrowtooth fl ounder (Atheresthes stomias) from Alaska (USA). In prep.

International Pacifi c Halibut Commission. 1988. The Pacifi c halibut : Biology, fi shery, and management. Int. Pac. Halibut Comm. Tech. Rep. 40. 64p.

Kitchell, J.F., Stewart, D.J., and Weininger, D. 1977. Applications of a bioenergetics model to perch (Perca fl avescens) and walleye (Stizostedion vitreum). J. Fish. Res. Board Can. 34:1922–1935.

Kuparinen, A. and Merilä, J. 2007. Detecting and managing fi sheries-induced evolution. TREE 22:652-659.


Hartman, K.J. and Kitchell, J.F. 2008: Bioenergetics Modeling: Progress since the 1992 Symposium, Trans. Am. Fish. Soc. 137: 216-223

Law, R. 2000. Fishing, selection, and phenotypic evolution. ICES J. Mar. Sci. 57:659-668.

Loher, T. 2013. Potential mechanisms, research avenues, and management action associated with size at age and growth of Pacifi c halibut. Int. Pac. Hal. Comm. Report of Assessment and Research Activities 2012: 457-486.

Lorenzen, K. and Enberg, K. 2002. Density-dependent growth as a key mechanism in the regulation of fi sh stocks: Evidence from among-population comparisons. Proc. Roy. Soc. London B 269:49-54.

Meka, J.M. and Margraf, F.J. 2007. Using a bioenergetic model to assess growth reduction from catch and release fi shing and hooking injury in rainbow trout, Oncorhynchus mykiss. Fish. Manag. Ecol. 14:131 – 139.

Millner, R.S. and Whiting, C.L. 1996. Long-term changes in growth and population abundance of sole in the North Sea from 1940 to the present. ICES J. Mar. Sci. 53:1185-1195.

Ney, J.J. 1993. Bioenergetics today: growing pains on the cutting edge. Trans. Amer. Fish. Soc. 122:736–748.

Ney, J.J. 1990. Trophic economics in fi sheries: assessment of demand–supply relationships between predators and prey. Rev. Aquat. Sci. 2:55–81.

Olsen, E.M, Heino, M., Lilly, G.R., Morgan, M.J., Brattey, J., Ernande, B., and Dieckmann, U. 2004. Maturation trends indicative of rapid evolution preceded the collapse of northern cod. Nature 428:932-935.

Olsen, E.M., Lilly, G.R., Heino, M., Morgan, M.J., Brattey, J., and Dieckmann, U.2005. Assessing changes in age and size at maturation in collapsing populations of Atlantic cod (Gadus morhua). Can. J. Fish. Aquat. Sci. 62:811-823.

Pope, J.A., Margette, A.R., Hamley, J.M., and Akyüz, E.F. 1975. Manual of methods for fi sh stock assessment. Part III. Selectivity of fi shing gear. FAO, United Nations, Rome.

R Development Core Team. 2010 R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. www.R-project.org

Ricker, W.E. 1954. Stock and recruitment. J. Fish. Res. Board Can. 11:559-623.

Schweigert, J., Funk, F., Oda, K., and T., M. (2002). Herring size-at-age variation in the North Pacifi c. In Peterson, W. and Hay, D., editors, REX workshop on temporal variations in size-at-age for fi sh species in coastal areas around the Pacifi c Rim, number 20, pages 47–57. PICES Sci. Rep.

Skaug, H. and Fournier, D. 2011. Random Effects in AD Model Builder ADMB-RE User Guide Version 10.0 (2011-01-18). http://admb-project.org/documentation/manuals/admb-user-manuals


Stewart, I.J., Leaman, B.M., Martell, S.J.D., Webster, R.A. 2013. Assessment of the Pacifi c halibut at the end of 2012. Inter. Pac. Hal. Comm. Report of Assessment and Research Activities 2012: 93-186.

Swain, D.P., Sinclair, A.F., and Hanson, J.M. 2007. Evolutionary response to size-selective mortality in an exploited fi sh population. Proc. Roy. Soc. B 274:1015-1022.

Valero, J.L. 2012. Harvest policy considerations on retrospective bias and biomass projections. Inter. Pac. Hal. Comm. Report of Assessment and Research Activities 2011:311-330.


Table 1. Sample sizes of stomach collections of three major groundfi sh predators in the NOAA Alaska Fisheries Science Center food habits database. Gulf of Alaska Species Decade Bering Sea West Central East & SE TotalHalibut 1980-1989 601 - - -

1990-1999 3,191 449 1,344 57 2000-2009 1,259 958 1,614 173Halibut Total 5,051 1,407 2,958 230 9,646Arrowtooth 1980-1989 2,926 48 105 -

1990-1999 4,567 1,348 6,336 1892000-2009 8,930 1,861 4,036 423

Arrowtooth Total 16,423 3,257 10,477 612 30,769Cod 1980-1989 12,008 170 365 -

1990-1999 20,320 1,161 3,130 882000-2009 12,640 1,296 2,362 43

Cod Total 44,968 2,627 5,857 131 53,583


Figure 1. Weight (net lbs) at age for Pacifi c halibut 1926-2011 (Hare 2011). Note that estimated ages are not corrected for changes in ageing methodology. Surface ages 1926-1996, mixture of surface and break and burn ages 1997-2001, break and burn or break and bake ages after 2002.

05

1015

1920 1960 2000

Age 6 05

1015

1920 1960 2000

Age 7 05

1015

2025

30

1920 1960 2000

Age 8 010

2030

40

1920 1960 2000

Age 9 010

2030

4050

1920 1960 2000

Age 10

010

3050

70

1920 1960 2000

Age 11 020

4060

80

1920 1960 2000

Age 12 020

4060

8010

0

1920 1960 2000

Age 13 020

4060

8010

0

1920 1960 2000

Age 14 020

4060

8010

0

1920 1960 2000

Age 15

020

4060

8012

0

1920 1960 2000

Age 16 050

100

150

1920 1960 2000

Age 17 050

100

150

1920 1960 2000

Age 18 050

100

150

1920 1960 2000

Age 19 050

100

150

1920 1960 2000

Age 20

Wei

ght (

net l

bs)

Year (1926 to 2011)


Figure 2. Percent weight in diet of major prey items of Pacifi c halibut in the GOA, by decade and length.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

15 25 35 45 55 65 75 85 95 100+

Perc

entw

eigh

tin

diet

Halibut fork length (cm)

OffalW. PollockOther pelagicsSandlanceOther demersalsFlatfishBairdiMisc. CrabEpifaunaShrimpPlankton

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

15 25 35 45 55 65 75 85 95 100+

Perc

entw

eigh

tin

diet

Halibut fork length (cm)

OffalW. PollockOther pelagicsSandlanceOther demersalsFlatfishBairdiMisc. CrabEpifaunaShrimpPlankton

1990-1999

2000-2009


Appendix I. IPHC 2013 Annual Research Plan (Preliminary)November 2012IPHC Staff

Executive summary

The fi shery for Pacifi c halibut (Hippoglossus stenolepis) is one of the most valuable and geographically largest in the north Pacifi c Ocean. Industry participants from Canada and the United States have prosecuted the fi shery and depended upon the resource since the turn of the 20th century. Annual removals have been as high as 100 million pounds; the long term average is 64 million pounds.

Management of the Pacifi c halibut resource and fi shery has been the responsibility of the International Pacifi c Halibut Commission (IPHC) since its creation in 1923. To assess, forecast, and manage the resource and fi shery requires accurate assessments, continuous monitoring, and research responsive to the needs of managers and stakeholders.

The IPHC staff has developed and compiled a series of research studies for 2013, which it recommends to the Commission for adoption through this Annual Research Plan. The recommended studies require a funding level of $0.45 million during fi scal year 2013. Several studies will contribute towards greater understanding of two issues currently facing the Commission and industry stakeholders, notably the decline in size at age and further development of advanced geomagnetic tags for describing the scope and timing of migration.

Assessment and stock identifi cationTwo priority topics are addressed: assessment information and migration studies. The staff

is proposing to continue with collection of juvenile abundance data, which is incorporated into the annual assessment, and an improved maturity stage assignment, which contributes to the age-at-maturity schedule. Second, as part of the continued focus on understanding halibut migration, the staff proposes to continue with the development of the geomagnetic archival tag technology, which involves several projects to resolve tag attachment, geomagnetic data resolution, and release locations in Area 4B. A new study into the halibut length/weight relationship is proposed to address the frequent observation that the current relationship is defi cient in some uses. In addition, an expansion of the annual assessment survey into northern California waters is also proposed to evaluate stock density in that area, based on new information on recent removals.

Management strategyOne study is proposed which begins the staff’s investigation into the declining trend in size at

age. The study takes advantage of IPHC’s extensive historical otolith archives, dating to the 1920s.

EcologyThree studies are proposed which are continuations of studies currently underway:

oceanographic monitoring with the profi lers from the survey platforms, mercury and contaminant assessment, and assessment of Ichthyophonus prevalence.


The staff notes that this Annual Research Plan does not describe all research activities conducted by IPHC. Various studies are conducted as a matter of routine work by the staff, some in collaboration with other agencies. The research recommendations included in this Plan are based on identifi ed research gaps, to supplement research already under way and advance the IPHC mission.

Introduction

The Pacifi c halibut (Hippoglossus stenolepis) resource has been the target of exploitation for centuries. Commercial harvests have been recorded since the late 1800s, and more recently a sport fi shery has grown in popularity throughout its range. Annual harvests have varied throughout history and have reached nearly 100 million pounds in 1990 and again in 2004. At the turn of the 20th century, the resource status suffered a serious decline, leading the halibut industry to petition the U.S. and Canadian governments for creation a multilateral management agency to institute fi shery restrictions aimed at stock recovery for the benefi t of fi shers from both countries. The International Pacifi c Halibut Commission (IPHC) was created in 1923 by a convention between the two countries. The 1923 Convention was also noteworthy in that it was the fi rst treaty to be concluded anywhere for the conservation of a depleted deep-sea fi shery. The Commission was charged with studying the life history of halibut and with recommending regulations for the preservation and development of the fi shery.

The IPHC, as an international fi sheries organization, receives monies from both the U.S. and Canadian governments to support an executive director and staff. Annually, the IPHC meets to conduct the business of the Commission. At this annual meeting the budgets, research plans, biomass estimates, catch recommendations, as well as regulatory proposals are discussed and approved then forwarded to the respective governments for implementation. The IPHC staff and offi ces are currently located in Seattle, Washington.

The IPHC conducts numerous projects annually to support both major mandates: stock assessment and basic halibut biology. Current projects include standardized stock assessment fi shing surveys from northern California to the end of the Aleutian Islands, as well as fi eld sampling in major fi shing ports to collect scientifi c information from the halibut fl eet. In conjunction with these ongoing programs, the IPHC conducts numerous biological and scientifi c experiments to further the understanding and information about Pacifi c halibut.

The 2012 IPHC Performance Review recommended the creation of a Five Year Research Plan and an Annual Research Plan (ARP). The plans would provide linkage to Commission objectives, with an accompanying process for input and periodic reviews by the Commission, interested stakeholders, the Research Advisory Board, and a peer review. The IPHC staff is tasked with developing the preliminary ARP for presentation to the Commission at the Interim Meeting, where discussion of overall research priorities, individual studies and associated budgets will occur. The staff will further develop the ARP following the Interim Meeting and present a fi nal ARP at the Annual Meeting for Commission approval.

Research focus and priorities

Nearly all of the research done by the IPHC is directed toward one of three continuing objectives of the Commission: 1) improving the annual stock assessment and quota recommendations; 2) developing information on current management issues; and 3) adding to knowledge of the biology


and life history of halibut. In each of these areas the work program applies the best information and methods available, and the research program aims to improve the information and methods by answering the most important outstanding questions.

IPHC research is conducted within four areas of study as identifi ed within the Five Year Research Plan. These areas, which connect to the IPHC mission and support the assessment and management objectives of the Commission, are 1) assessment and stock identifi cation; 2) management strategy; 3) biology; and 4) ecology.

The ARP is based on management and assessment needs as prioritized by the IPHC staff and Commission. It is the Commission’s long term goal to also obtain the views and advice of its Research Advisory Board (RAB) and external scientifi c input in the formulation and prioritization of the ARP. For 2012, this process is still being developed, so input from those sources will be brought into the process during in the 2013 research development cycle.

For the past several years, two primary topics have been at the forefront of discussions about the halibut resource. The fi rst has been the continuing decline in size at age (SAA), with the resulting effects and impacts on the assessment, harvest policy, and stock status. The second issue has been the migratory behavior of the stock, specifi cally seasonal and ontogenetic migration, including sex- and age-specifi c differences in spawning migration timing and duration. In the following section, studies for 2013 will be proposed which address both topics. Briefl y, the IPHC staff proposes to begin an otolith increment study which would examine growth patterns during earlier time periods (project 2013-01). Understanding migration patterns is the overarching goal of the archival tag program, which has several aspects examining tag type, location, tag shedding and resolution of geomagnetic location data (projects 650.xx).

Proposed for 2013

Research proposed by IPHC staff goes through an internal review process by the staff Science Board. This year, the Board met in early October to review staff proposals for 2013 research. For each proposal, the Board discussed the merits, objectives, design, and coherence with the Commission’s research goals and objectives. The Principal Investigator (PI) subsequently joined the Board for a broad discussion of the project. Concerns, questions and need for refi nements or revisions, if any, about the proposal were communicated to the PI at that time. Following a full review of all proposals, the Board assigned a priority rating to each project, based on the following criteria:

High – Research which has a direct bearing on the assessment or its inputs, harvest policy, or cur-rent management structure. Postponement of a high priority project would have a signifi -cant and immediate impact on management or IPHC operation.

Medium – Research which addresses an assessment issue or management question/need. Post-ponement will not have an immediate signifi cant impact on fi shery management or IPHC operation but may impact future analyses.

Low – Research which addresses current issues of any subject but is not considered having a timely need or being crucial to current IPHC management or operation.

Based on the Science Board discussions and the topics previously outlined, the IPHC staff recommends the following research studies for funding in FY2013.


Assessment and stock identifi cation

Project 604.00: Monitoring juvenile halibut abundance via NMFS trawl surveys

Priority: HighBudget: $ 53,631Start Date: 1996Anticipated ending: ContinuingPersonnel: L. Sadorus, A. Ranta, I. Stewart

The series of NMFS trawl survey data on halibut, parallel to our assessment survey data, is extremely valuable as a second fi shery-independent data source for stock assessment. Trawl data are particularly useful because they include large numbers of juveniles (ages 3-7) that do not appear in large numbers in the setline survey. Otoliths have been collected on the NMFS surveys since 1996 and provide relevant age information. These data are incorporated into IPHC’s database of the NMFS haul data, expanded to estimates of relative abundance and age/size composition, and stored in a database at IPHC. Project cost is comprised of personnel and travel. For 2013, samplers will be deployed in the Bering Sea and Gulf of Alaska surveys.

Project 636.00: Evaluation of Pacifi c halibut macroscopic maturity stage assignments

Priority: HighBudget: $ 18,800Start: 2004Anticipated Ending: ContinuingPersonnel: K. MacTavish, other staff as needed

The staff believes it is necessary to re-evaluate our classifi cation criteria for female gonad maturity stage. The method currently used on the assessment surveys is based on visual criteria established in the early 1990s and modifi ed in 1995. These survey data combined with the age data are important components in the stock assessment model. Four maturity stages are presently assigned to female halibut; immature (F1), maturing (F2), spawning (F3) and resting (F4). Once a female halibut has spawned, the gonad transitions to a resting phase, back to maturing, and then to spawning again. Our criteria for classifi cation also assume that the immature (F1) stage is only seen with immature fi sh but we are seeing anomalies during the survey that question this assumption. Gonad samples were collected in 2004 from which to base this study. In 2013, work will continue on fi nalizing a sampling protocol for measurement of oocyte diameters, and contract slide preparation for gonads. The PI will also begin assessment of archived gonads from a set of previously-prepared slides.


Project 650.13: Archival tags: mounting protocols (OCA)

Priority: MediumBudget: $ 14,800Start Date: 2009Anticipated ending: 2014Personnel: T. Loher

For 2013, the staff intends to continue holding halibut in tanks at the Oregon Coast Aquarium (OCA) in Newport, OR to investigate alternate mounting protocols for the externally-mounted archival tags. A total of 30 halibut were captured via hook-and-line and transported live to the OCA. The fi sh are treated for parasites, examined regularly to assess healing and/or relative infection rates among mounting types, and behavior monitored. At the end of the holding period, fi sh will be measured to assess relative growth among treatment groups, and tags will be removed to examine the effects of the tag mounts on the tissue and musculature at the attachment site, or internal interactions in the case of an internal-external-streamer modifi cation. The results will support the anticipated use of this type of technology in subsequent years. Expenses for 2013 involve the care and feeding of the fi sh at OCA.

Project 650.14: Archival tags: tag attachment protocols

Priority: HighBudget: $ 600Start Date: 2009Anticipated ending: 2014Personnel: T. Loher

External and internal tag recovery rates are being tested in the fi eld release of archival test tags. In August-September 2009, 200 fi sh were tagged off southern Kodiak Island (in Areas 3A and 3B), half with external tags and half with internal implants. Fish were also tagged with a bright pink cheek tag, and rewards of $100 will be given for all tags recovered. Nine fi sh were recovered in 2011. Expenses in 2013 consist of tag rewards. Note that the because of a subsequent decision to focus only on an external mount protocol, this project is proposed to be redone in Area 3A in 2013 (see 2013-04, below).

Project 650.15: Archival tags: coastwide deployment

Priority: HighBudget: $ 125,000Start Date: 2016Anticipated ending: ContinuingPersonnel: T. Loher, B. Leaman, R. Webster, J. Forsberg

In preparation for a coastwide release of archival tags in 2016, the staff has been working with Lotek Wireless (St. John’s, NL) on a specifi c tag design and confi guration for IPHC use. Although


no fi eld activity is planned for 2013, Lotek is continuing their work on our requirements and construction. Results from the 2009 release of dummy archival tags in Area 3A and the examination of several mounting protocols on fi sh being held at the Oregon Coast Aquarium will feed into the design of the tag and its attachment to the fi sh.

Project 650.16: Archival tags: Area 4B site selection

Priority: HighBudget: $ 1,000Start date: 2010Anticipated ending: 2014Personnel: T. Loher, J. Forsberg, survey team

In 2009, 773 fi sh were tagged in Area 4B to evaluate tag recovery rates in preparation of a future release of archival tags in the area. Recovery rates of PIT tags released in the Aleutians were quite low, without evidence of recovery hotspots. This suggests that if archival tags were deployed in the Aleutians, we would likely recover relatively few of those tags. This would result in either too few data to draw any conclusions or require that a very large number of tags be initially deployed. Given that archival tags cost $500-1200 each, resorting to a very large deployment would be fi nancially prohibitive and problematic. The goal is to locate at least two release sites which will yield a suffi cient number of recoveries. The requested budget for 2013 is to cover the rewards for the anticipated recoveries.

Project 650.17: Archival tags: geomag tag performance

Priority: LowBudget: $ 6,000Start Date: 2011Anticipated ending: 2012Personnel: T. Loher, J. Nielsen (UAF Juneau)

In 2011 we deployed both Desert Star and Lotek geomagnetic tags on 30 halibut in two regions of the Gulf of Alaska: in Area 2C, just offshore of southern Prince of Wales Island; and in Area 3A, offshore of southern Kodiak Island. Tagging was restricted to large fi sh (110-150 cm FL), most likely to be mature females and likely to conduct a spawning migration shortly after tagging, and was divided into two deployment locations because the coastline and bathymetry of the areas are largely perpendicular to one another with respect to the magnetic environment. In Area 2C, total magnetic fi eld gradients run largely parallel to shore, whereas in Area 3A around Kodiak that gradient runs perpendicular to shore. As such, we hypothesized that geomagnetic positioning based on total fi eld strength would more accurately detect onshore-offshore movement in 2C and alongshore migration around Kodiak. Recoveries are expected in 2013 to enable testing of the hypothesis; project expenses are for the rewards. Note that the because of a tag internal architecture redesign, this project is proposed to be redone in Area 4A in 2013 (see 2013-05, below).


Project 2013-03: Estimate of length/weight relationship and head/ice/slime adjustment (NEW)

Priority: LowBudget: $ 5,000Start: 2013Anticipated Ending: open endedPersonnel: R. Webster, L. Erickson, K. MacTavish, H. Gilroy

The purpose of this study is to collect data for use in estimating the relationship between fork length and net weight, including the estimate of adjustments necessary to convert head-on weight to net weight. Data will be collected coastwide at sampled ports throughout the season in order to estimate spatial and seasonal variation in the length to weight relationship. In the current length-weight relationship, adjustments are made for head, ice, and slime, and are used when estimating the net weight of commercial offl oads. The current relationship between fork length and net weight includes adjustments for the weight of the head, and of ice and slime: gross weight is assumed to include 12% head weight and 2% ice and slime, which combine to give a multiplier of 0.8624 to convert gross to net weight. In practice, deductions of 12% in Areas 2A and 2B, and 11.8% in Alaska, are applied to commercial landings at the plants to convert from gross to net weight. These both include the 2% deduction for ice and slime assumed in the IPHC length-net weight relationship, but 10% for the head. IPHC port samplers will be tasked to collect data at plants within their port. Therefore in addition, data collected during the study will provide direct estimates of adjustment factors to compare with the currently assumed values, and will allow us to assess variability in the weight of heads and ice and slime. The end result is expected to be new adjustment factors that, if appropriate, can be applied consistently across all ports, or be allowed to vary with regulatory area.

Project 2013-04: Archival tags: tag attachment protocols (NEW)

Priority: HighBudget: $ 46,250Start: 2013Anticipated Ending: 2015Personnel: Loher

This proposal is an update of 650.14, which was a 2009 release of 200 tags – half external tags, and half with internal implants. It is proposed to be redone to fully evaluate external attachment. The Board was supportive of this project, as the results are needed to evaluate three potential tag attachment sites on the fi sh. The release is being designed to occur from the surveys to reduce costs while still achieving a broad distribution of releases. Design issues regarding the number of tags to be deployed and shedding rates are being refi ned. The study was given a high priority because a suitable external tag attachment site is crucial to the success of the coastwide archival tag study.


Project 2013-05: Archival tags: geomag performance (NEW)

Priority: HighBudget: $ 6,400Start: 2013Anticipated Ending: 2016Personnel: Loher

This proposal is an update of 650.17, which was a 2011 release of 30 tags in Area 2C and 3A to examine location resolution of geomag tags. The study is proposed to be redone because an improved geomag design has recently been released, which is expected to perform better than the design used in 2011. The proposed study entails releasing ~30 fi sh on the Area 4A-south assessment survey. The Board gave this a high priority, as it will provide information on the accuracy of the location data provided by the Lotek archival tags.

Project 2013-06: SSA Expansion – California pilot (NEW)

Priority: MediumBudget: $ 49,408Start: 2013Anticipated Ending: currently only planned for 2013Personnel: C. Dykstra, survey team

The IPHC staff is considering extending the assessment survey into the waters off northern California. Currently the survey stops at the Oregon/California border, which has traditionally been the southern end of commercial fi shing in recent years. However, recent reports of previously unknown but signifi cant sport fi shery harvests of halibut from northern California waters, which contributed to exceeding the catch limit for that area, have indicated the potential for a larger share of the resource in this area than has been assumed. Adding this area into the assessment requires a measure of fi sh density, which would be provided by the survey. This issue also has implications to the Pacifi c Fishery Management Council’s Area 2A Catch Sharing Plan, which allocates a portion of the Area 2A catch limit to the area south of Humbug Mountain, Oregon, including California. The current staff proposal would extend the 10 x 10 nm systematic survey grid off northern California, to a terminus of 40o N., based on a review of halibut sport fi shery sampling by California Fish and Game.


Management Strategy

Project 2012-01: Otolith increment analysis (New)

Budget: $ 4,176Priority: HighStart Date: 2013Anticipated ending: 2015Personnel: T. Loher, S. Wischniowski

This study is an internal IPHC project but may be part of a broader, comprehensive study to examine potential causes for the recent changes in halibut size at age (SAA) as well as an integrated approach to incorporating SAA dynamics into halibut assessment and management. The broader study would be funded through a grant application to the North Pacifi c Research Board, in cooperation with National Marine Fisheries Service and the University of Alaska. For the internal IPHC project staff will mine the otolith archives for historical samples which were collected at prescribed time intervals and measure the otolith growth increments. The relation between otolith growth and somatic growth is not well understood in many fi shes, including halibut. But the IPHC otolith archives provide a unique opportunity to potentially examine changes in otolith growth over time and, by extension, halibut growth. Anticipated work in 2013 includes refi ning the study design, otolith selection, cross sectioning, and aging.

Project 2013-02: Genetic sexing techniques (SNPs)

Budget: $ 156,324Priority: noneStart: 2013Anticipated Ending: ContinuingPersonnel: T. Loher, L. Hauser (UW MMBL), other staff as needed

This is a proposal to estimate the sex composition of the commercial landings through the use of genetic markers identifi ed in tissue collections of landed commercial fi sh. This work has potential for a high degree of accuracy, if sex-specifi c markers can be identifi ed, but the staff Science Board had reservations about the practical application of the approach to the coastwide commercial fi shery, given an existing, though imprecise technique for sexing the catch. The assessment team would fi rst like to quantify the gain of having more precise estimates of sex composition in the commercial catch versus the existing technique and fi tting the assessment model to a sex aggregated total catch before making a substantial investment in this technology. Accordingly, the Board recommends placing this proposal on hold for 2013.


Ecology

Project 610.13: Oceanographic monitoring of the north Pacifi c and Bering Sea continen-tal shelf with water column profi lers

Priority: MediumBudget: $ 53,676Start date: 2009Anticipated ending: ContinuingPersonnel: L. Sadorus, P. Stabeno (NMFS PMEL)

The IPHC maintains one of the most extensive sampling platforms in the north Pacifi c. This plat-form provides enormous potential for collection of valuable oceanographic data. In particular, understanding the dynamics of the structure of the mixed layer depth – a major Global Ocean Eco-system Dynamics (GLOBEC) goal - requires in situ vertical profi ling. Since 2001, IPHC has suc-cessfully deployed a SeaBird SBE-19 water column profi ler during the annual stock assessment survey. A second profi ler was added to the program in 2007. In 2009, a NOAA grant provided for the complete outfi tting of all chartered survey vessels, resulting in a complete coastwide deploy-ment. Annual costs are directed towards maintenance and calibration of the profi lers, and data preparation necessary for submission to the National Ocean Data Center.

Project 642.00: Assessment of mercury and contaminants in Pacifi c halibut

Priority: MediumBudget: $ 4,000Start Date: 2002Anticipated ending: ContinuingPersonnel: C. Dykstra, B. Gerlach (ADEC)

The staff proposes to continue our collaboration with the Alaska Department of Environmental Conservation (ADEC) in 2013, collecting halibut tissue samples for analysis of heavy metal and organic pollutant loading. This work has been ongoing since 2002. Results from a 2002 collection of halibut samples led the Alaska Division of Public Health in 2003 to conclude that the concentra-tions of heavy metals in Alaskan Pacifi c halibut were not a public health concern. In 2004 the fi rst results regarding organic pollutants (PCB’s, pesticides) were released demonstrating that halibut had the lowest concentrations of the fi ve species (including salmon and sablefi sh) examined. The Alaska Division of Public Health updated their advice on fi sh consumption in 2007 with some restrictions on the number of meals of halibut for women of child bearing age and young children. Since 2002 the IPHC has submitted 1,293 samples for testing by ADEC. The IPHC and ADEC are continuing to qualify the data with physical parameters (age, size, and weight) and additional analyses will be done on the samples. ADEC and the Environmental Protection Agency planned on going ahead with this study regardless of IPHC input. Our involvement in the project has al-lowed us to provide input on study design, sampling protocols in the fi eld, etc., which will make the resultant information much more robust.


Project 661.11: Ichthyophonus prevalence in halibut

Priority: MediumBudget: $ 49,239Start Date: 2012Anticipated ending: ongoingPersonnel: C. Dykstra, G. Williams, J. Gregg (USGS), P. Hershberger (USGS)

Ichthyophonus is a protozoan parasite from the class Mesomycetozoea, a highly diverse group of organisms having characteristics of both animals and fungi. It has been identifi ed in many marine fi sh, and is considered a causative agent in herring fi shery collapses world-wide and there is concern over its effects on the success of salmon spawning on major rivers such as the Yukon.

In 2011 the IPHC ran a small pilot project looking at Ichthyophonus prevalence in Pacifi c halibut in response to some initial test results from a 2010 U.S. Geological Survey (USGS) study which found high incidence rates in sport caught halibut in Cook Inlet, AK. The 2011 pilot took place in three geographically disparate areas (Oregon, Prince William Sound proper, and the northern Bering Sea). Results from this study found some of the highest incidence rates for any marine species in the Prince William Sound region (76.7% incidence), with lower, but still signifi cant levels in Oregon (33.8%) and the northern Bering Sea (26.6%). USGS defi nes the Prince William Sound result as an epizootic event as the incidence rate is much higher than background rates seen in other halibut studies.

In 2012, tissue samples were collected in all survey areas to further describe the spatial nature of the prevalence. In addition, samples were collected from smaller juveniles caught on the NMFS trawl survey in the Bering Sea. Prevalence of infection measured at ten longline survey sites ranged from 15% near Attu Island to over 70% in Prince William Sound, with a mean overall prevalence (Bering Sea to Oregon Coast) of 47%. Prevalence in smaller halibut (<50 cm) captured by trawl in the Bering Sea and Aleutian Island was 2.4%, indicating infections establish after some ontogenetic shift in diet, habitat, or behavior.

The prevalence of infection reported here is higher than that which has been observed in studies of other sympatric fi sh species, including other pleuronectids, suggesting that either susceptibility and/or infection pressures are higher in halibut. While ichthyophoniasis has been shown to reduce growth rate, decrease swimming stamina, and cause mortality in other fi sh hosts, its effects on Pacifi c halibut are unknown. For this reason, the proposal being developed for 2013 involves studying the effect of infection intensity on growth and survival on fi sh held in the USGS lab. Study specifi cs are currently being worked out with USGS; a fi nal design will be provided in advance of the Annual Meeting.

Alternative research packages

The Commission requested to see options for research which would be contingent on various funding levels. The full suite of projects outlined by the IPHC staff is summarized in Table 1, with a total budget of $606,293. The staff Science Board recommended not proceeding with the proposal for the development of production-scale genetic sexing techniques in 2013, which had a proposed budget of $156,324. By excluding this project, the total budget for the remaining proposed projects is $449,969.


One approach to constructing the alternative research packages is to consider the proposals based on priority and/or area of study. Project budgets can be summed up within these strata for evaluation, which is shown in the following table. Note, however, that the cells within a column are not additive, as some projects span more than one area of study. For example, the Otolith increment study, assigned a high priority, was shown in Table 1 as being applicable to the Management Strategy and Biology areas of study. Thus it would appear in the totals for both study areas in the High priority column, yet appear just once in the column total for High priority projects. This type of presentation enables the Commission to consider those groups of projects of a prescribed priority and/or area of study which could be conducted if budgetary limitations were present.

Area of Study High Medium Low TotalAssessment/stock identifi cation $ 251,681 $ 64,208 $ 22,989 $ 338,878Management strategy $ 4,176 -- $ 14,989 $ 19,165Biology $ 22,976 -- $ 14,989 $ 37,965Ecology -- $ 106,915 $ 14,989 $ 121,904Total $ 255,857 $ 171,123 $ 22,989 $ 449,969

Note that many of the studies proposed for 2013 were initiated in previous years and are considered as ongoing research initiatives. Some have been structured as multi-year studies of a fi xed duration, whereas others have more of a long term focus. For reference, the total proposed research package for 2013 has the following split between ongoing and new studies:

Study type No. of studies Proposed budgetOngoing 11 $ 338,735New for 2013 5 $ 111,234Total 16 $ 449,969

Future research directions

The staff anticipates additional future studies on SAA, perhaps in collaboration with other agencies. In conjunction with those studies, the staff will be considering additional otolith studies which would include re-aging selected time series which would benefi t both SAA studies and the assessment. Another potential research avenue will be the application of environmental (profi ler) data to recruitment scenarios and year class strength.

Recommendations for changes to the Five Year Research Plan

The staff has no recommendations for changes to the Five Year Plan at this time. The Commission and staff are currently developing the Annual and Five Year Plan, including the process to support their development. As these are further developed, the staff expects to recommend any changes necessary to support the Commission’s needs.


Tabl

e 1.

Su

mm

ary

of p

ropo

sed

rese

arch

for

2013

.

Proj

ect #

Stud

y N

ame

Prio

rity

Assessment & Stock ID

Manage-mentStrategy

Biology

Ecology

Bud

get

(US$

)L

ead/

PIB

egin

/End

Dat

es

604.

00M

onito

ring

juve

nile

hal

ibut

abu

ndan

ce v

ia N

MFS

tra

wl s

urve

ysH

igh

●53

,631

Sado

rus

Ong

oing

610.

13O

cean

ogra

phic

mon

itorin

g of

the

north

Pac

ifi c

and

Ber

ing

Sea

cont

inen

tal s

helf

with

wat

er c

olum

n pr

ofi le

rsM

ediu

m●

53,6

76Sa

doru

sO

ngoi

ng

618.

00IP

HC

Inte

rn P

rogr

amLo

w●

●●

●14

,989

Sado

rus

Ong

oing

636.

00Ev

alua

tion

of P

acifi

c ha

libut

mac

rosc

opic

mat

urity

st

age

assi

gnm

ents

Hig

h●

●18

,800

Mac

Tavi

sh20

08/2

014

642.

00A

sses

smen

t of m

ercu

ry a

nd c

onta

min

ants

in P

a-cifi c

hal

ibut

Med

ium

●4,

000

Dyk

stra

Ong

oing

650.

13A

rchi

val t

ags:

mou

ntin

g pr

otoc

ols (

OC

A)

Med

ium

●14

,800

Lohe

r20

1665

0.14

Arc

hiva

l tag

s: t

ag a

ttach

men

t pro

toco

lsH

igh

●60

0Lo

her

2016

650.

15A

rchi

val t

ags:

coa

stw

ide

depl

oym

ent

Hig

h●

125,

000

Lohe

r20

1665

0.16

Arc

hiva

l tag

s: A

rea

4B si

te se

lect

ion

Hig

h●

1,00

0Lo

her

Ope

n en

ded

650.

17A

rchi

val t

ags:

geo

mag

tag

perf

orm

ance

Low

●3,

000

Lohe

r20

1466

1.11

Icht

hyop

honu

s pre

vale

nce

in h

alib

utM

ediu

m●

49,2

39D

ykst

ra20

12/2

013

2013

-01

Oto

lith

incr

emen

t ana

lysi

sH

igh

●●

4,17

6Lo

her

Ope

n en

ded

2013

-02

Gen

etic

sexi

ng te

chni

ques

(SN

Ps)

---

●15

6,32

4Lo

her

On

hold

for 2

013

2013

-03

Leng

th/w

eigh

t and

hea

d/ic

e/sl

ime

fact

ors

Low

●5,

000

Web

ster

Ope

n en

ded

2013

-04

Arc

hiva

l tag

s: t

ag a

ttach

men

t pro

toco

ls (6

50.1

4 up

date

)H

igh

●46

,250

Lohe

r20

16

2013

-05

Arc

hiva

l tag

s: g

eom

ag ta

g pe

rfor

man

ce (6

50.1

7 up

date

)H

igh

●6,

400

Lohe

r20

14+

2013

-06

SSA

exp

ansi

on:

Cal

iforn

ia p

ilot

Med

ium

●49

,408

Dyk

stra

2013

Tota

l (w

ithou

t 201

3-02

)$

449,

969

Ove

rall

Tota

l (al

l pro

ject

s)$

606,

293


Review of 2012 - Project summaries

This section is provides a brief recap of projects conducted in 2012. Full reports on each project can be found in the 2012 RARA.

Project 604.00: Monitoring juvenile halibut abundance via NMFS trawl surveys

Budget: $ 55,661Start Date: 1996Anticipated ending: ContinuingPersonnel: L. Sadorus, A. Ranta, I. Stewart

The series of NMFS trawl survey data on halibut, parallel to our assessment survey data, is extremely valuable as a second fi shery-independent data source for stock assessment. Trawl data are particularly useful because they include large numbers of juveniles (ages 3-7) that do not appear in large numbers in the setline survey. Otoliths have been collected on the NMFS surveys since 1996 and provide relevant age information. These data are incorporated into IPHC’s database of the NMFS haul data, expanded to estimates of relative abundance and age/size composition by IPHC area (NMFS calculates estimates by INPFC area), and stored in a database at IPHC. Project cost is comprised of personnel and travel. In 2011, samplers were deployed on the NMFS Gulf of Alaska and Bering Sea surveys. For 2012, samplers were deployed in the Bering Sea and Aleutian Island surveys.

Project 610.13: Oceanographic monitoring of the north Pacifi c and Bering Sea continen-tal shelf with water column profi lers

Budget: $ 73,893Start date: 2009Anticipated ending: ContinuingPersonnel: L. Sadorus, P. Stabeno (NMFS PMEL)

The IPHC maintains one of the most extensive sampling platforms in the north Pacifi c. This platform provides enormous potential for collection of valuable oceanographic data. In particular, understanding the dynamics of the structure of the mixed layer depth – a major GLOBEC goal - requires in situ vertical profi ling. Since 2001, IPHC has successfully deployed a SeaBird SBE-19 water column profi ler du ring the annual stock assessment survey. A second profi ler was added to the program in 2007. In 2009, a NOAA grant provided for the complete outfi tting of all chartered survey vessels, resulting in a complete coastwide deployment. Annual costs are directed towards maintenance and calibration of the profi lers, and data preparation necessary for submission to the National Ocean Data Center. Over 1,200 casts were made in 2012.


Project 618.00: Undergraduate Internship

Budget: $ 13,933Start Date: 2002Anticipated duration: ContinuingPersonnel: L. Sadorus, other staff support as needed

One undergraduate will be selected through the intern/co-op programs at regional universities and colleges to do a combination of offi ce and at-sea work based out of the Commission offi ces during the summer months. The program includes various pre-determined offi ce tasks as well as being assigned a research project then designing and executing said project. A fi nal report and presentation are given at the conclusion of the employment term. The report is usually included in the RARA. Unfortunately, in 2012 we were unsuccessful in hiring an intern to develop an image-based technology solution for fi sh sampling.

Project 636.00: Evaluation of Pacifi c halibut macroscopic maturity stage assignments

Budget: $ 16,800Start: 2004Anticipated Ending: ContinuingPersonnel: K. MacTavish, other staff as needed

The staff believes it is necessary to re-evaluate our classifi cation criteria for female gonad maturity stage. The method currently used on the assessment surveys is based on visual criteria established in the early 1990s and modifi ed in 1995. These survey data combined with the age data are important components in the stock assessment model. Four maturity stages are presently assigned to female halibut; immature (F1), maturing (F2), spawning (F3) and resting (F4). Once a female halibut has spawned, the gonad transitions to a resting phase, back to maturing, and then to spawning again. Our criteria for classifi cation also assume that the immature (F1) stage is only seen with immature fi sh but we are seeing anomalies during the survey that question this assumption. Gonad samples were collected in 2004 from which to base this study. In 2012, work was undertaken to look for a size gradient for oocytes dependent on their location within the gonad, determine the maximum precision for oocyte diameter measurements by oocyte maturation stage, and to begin the design of a sampling protocol for measurement of oocyte diameters.

Project 642.00: Assessment of mercury and contaminants in Pacifi c halibut

Budget: $ 4,000Start Date: 2002Anticipated ending: ContinuingPersonnel: C. Dykstra, B. Gerlach (ADEC)

The staff continued in our collaboration with the Alaska Department of Environmental Conservation (ADEC) in 2012, collecting halibut tissue samples for analysis of heavy metal and organic pollutant loading. This work has been ongoing since 2002. Results from a 2002 collection of


halibut samples led the Alaska Division of Public Health in 2003 to conclude that the concentrations of heavy metals in Alaskan Pacifi c halibut were not a public health concern. In 2004 the fi rst results regarding organic pollutants (PCB’s, pesticides) were released demonstrating that halibut had the lowest concentrations of the fi ve species (including salmon and sablefi sh) examined. The Alaska Division of Public Health updated their advice on fi sh consumption in 2007 with some restrictions on the number of meals of halibut for women of child bearing age and young children. Since 2002 the IPHC has submitted 1,293 samples for testing by ADEC. The IPHC and ADEC are continuing to qualify the data with physical parameters (age, size, and weight) and additional analyses will be done on the samples. ADEC and EPA planned on going ahead with this study regardless of IPHC input. Our involvement in the project has allowed us to provide input on study design, sampling protocols in the fi eld, etc., which will make the resultant information much more robust. Sampling continued in 2012 with a targeted collection of 70 samples (15 fi sh between 10-20 lbs., 15 fi sh between 20-40 lbs., 30 fi sh between 40-100 lbs., and 10 fi sh greater than 100 lbs.) from each of four sites. Fifty-fi ve (55) samples were obtained from Semidi, 67 from Seward, 56 from northern Washington, and 50 from lower Oregon.

Project 650.13: Archival tags: tag mounting protocols (OCA)

Budget: $ 14,800Start Date: 2009Anticipated ending: 2014Personnel: T. Loher

For 2012, the staff intends to continue holding halibut in tanks at the Oregon Coast Aquarium (OCA) in Newport, OR to investigate alternate mounting protocols for the externally-mounted archival tags. The 2008 releases in Area 2B were our fi rst experience with using an external mount, and that process suggested some revisions and improvements could be possible which would reduce any effect the tags may have on the fi sh’s behavior. Additional improvements to tag design may also be helpful in creating a different mounting device. A total of 30 halibut were captured via hook-and-line and transported live to the OCA. The fi sh are treated for parasites, examined regularly to assess healing and/or relative infection rates among mounting types, and behavior monitored. At the end of the holding period, fi sh will be measured to assess relative growth among treatment groups, and tags will be removed to examine the effects of the tag mounts on the tissue and musculature at the attachment site, or internal interactions in the case of an internal-external-streamer modifi cation. The results will support the anticipated use of this type of technology in subsequent years. Expenses for 2012 involve the care and feeding of the fi sh at OCA.


Project 650.14: Archival tags: tag attachment protocols

Budget: $ 4,000Start Date: 2009Anticipated ending: 2014Personnel: T. Loher

External and internal tag recovery rates are being tested in the fi eld release of archival test tags. In August-September 2009, 200 fi sh were tagged off southern Kodiak Island (in Areas 3A and 3B), half with external tags and half with internal implants. Fish were also tagged with a bright pink cheek tag, and rewards of $100 will be given for all tags recovered. Nine fi sh were recovered in 2011; 3 in 2012.

Project 650.15: Archival tags: coastwide deployment

Budget: $ 130,000Start Date: 2016Anticipated ending: ContinuingPersonnel: T. Loher, B. Leaman, R. Webster, J. Forsberg

In preparation for a coastwide release of archival tags in 2016, the staff has been working with Lotek Wireless (St. John’s, NL) on a specifi c tag design and confi guration for IPHC use. Although no fi eld activity is planned for 2013, Lotek is continuing their work on our requirements and construction. Results from the 2009 release of dummy archival tags in Area 3A and the examination of several mounting protocols on fi sh being held at the Oregon Coast Aquarium will feed into the design of the tag and its attachment to the fi sh.

Project 650.16: Archival tags: Area 4B site selection

Budget: $ 2,600Start date: 2010Anticipated ending: 2014Personnel: T. Loher, J. Forsberg, survey team

In 2009, 773 fi sh were tagged in Area 4B to evaluate tag recovery rates in preparation of a future release of archival tags in the area. Recovery rates of PIT tags released in the Aleutians were quite low, without evidence of recovery hotspots. This suggests that if archival tags were deployed in the Aleutians, we would likely recover relatively few of those tags. This would result in either too few data to draw any conclusions or require that a very large number of tags be initially deployed. Given that archival tags cost $500-1200 each, resorting to a very large deployment would be fi nancially prohibitive and problematic. Our goal is to locate at least two release sites which will yield a suffi cient number of recoveries. Eleven tags were recovered in 2011. In 2012, only fi ve tags were recovered.


Project 650.17: Archival tags: geomag tag performance

Budget: $ 6,000Start Date: 2011Anticipated ending: 2012Personnel: T. Loher, J. Nielsen (UAF Juneau)

In 2011 we deployed both Desert Star and Lotek geomagnetic tags on 30 halibut in two regions of the Gulf of Alaska: in Area 2C, just offshore of southern Prince of Wales Island; and in Area 3A, offshore of southern Kodiak Island. Tagging was restricted to large fi sh (110-150 cm FL), most likely to be mature females and likely to conduct a spawning migration shortly after tagging, and was divided into two deployment locations because the coastline and bathymetry of the areas are largely perpendicular to one another with respect to the magnetic environment. In Area 2C, total magnetic fi eld gradients run largely parallel to shore, whereas in Area 3A around Kodiak that gradient runs perpendicular to shore. As such, we hypothesized that geomagnetic positioning based on total fi eld strength would more accurately detect onshore-offshore movement in 2C and alongshore migration around Kodiak. Only one tag has been recovered to date.

Project 661.11: Ichthyophonus prevalence in halibut

Budget: $ 50,739Start Date: 2012Anticipated ending: 2012Personnel: C. Dykstra, G. Williams, J. Gregg (USGS), P. Hershberger (USGS)

This study will further characterize Ichthyophonus prevalence across the Pacifi c halibut’s range, to determine overall prevalence rates and to see if the Prince William Sound results are repeatable. Twelve sites will be targeted, spread out over the assessment survey ranging from Oregon to the northern Bering Sea. As there is knowledge regarding herring infection rates in Prince William Sound, Sitka Sound area, and Lynn Canal, these areas are likely to be included in the primary target areas for sample collection. The study was modifi ed to do a more intensive sampling (stratifi ed by age or size) in the Bering Sea and/or Aleutian Islands where we were able to source samples from smaller fi sh caught during the NMFS trawl survey. IPHC collected the samples, and the USGS lab in Marrowstone will conduct the culture and testing component. Sample analysis and results are currently being fi nalized, and will be reported in the RARA.


Project 662.11: Hook modifi cation study to reduce rockfi sh bycatch on circle hooks

Budget: $ 29,676Start Date: 2012Anticipated ending: 2012Personnel: S. Kaimmer, S. Wischniowski

A study to examine a potential hook modifi cation designed to reduce rockfi sh bycatch was conducted in southeastern Alaska during May 11-21. Rockfi sh bycatch continues to be a problem in hook & line fi sheries targeting Pacifi c halibut. This is particularly so for yelloweye and quillback rockfi sh which in some areas are under varying degrees of protection. Previous camera observation of halibut and rockfi sh (Kaimmer 1998) suggested that hook attacks by halibut are much more forceful than those made by rockfi sh, and we hypothesized that spring wires across the hook gap might be able to reduce rockfi sh hooking while not impeding halibut captures. We observed no decrease in hooking success for yelloweye rockfi sh over 4.5 kg using various spring wire confi gurations. Too few quillback rockfi sh were encountered to make a defi nitive statement regarding this species. However, the charter vessel’s crew commented that the wired hooks were very awkward to work with, both in baiting and in removing fi sh from the hook, and it would take a major improvement in bycatch avoidance to make these hooks feasible in a fi shery.

Project 421.11: Examination of potential alternative bait for the assessment survey

Budget: $ 400,000Start Date: 2012Anticipated ending: 2012Personnel: R. Webster, S. Kaimmer, C. Dykstra, survey team

A coastwide comparison of alternative baits for the assessment survey was conducted in 2012. A 2011 pilot study conducted to refi ne the experimental design also led to the decision to examine pollock and pink salmon as alternatives to the standard #2 chum salmon, in this year’s study. There were signifi cant differences in O32 WPUE between chum salmon and the two alternative baits that varied by regulatory area. Most notably, WPUE for pollock was somewhat higher in general than WPUE for the two salmon baits in the Gulf of Alaska, but much lower in parts of Area 4. There was also evidence for differences in catch rates of U32 halibut and bycatch species among the three baits. We also compared the length and age distributions of halibut caught using the three baits. The results will be further analyzed in 2013, with the expectation that any change would not be implemented until 2014, at the earliest.


Appendix II. IPHC regulation proposals submitted for the 2013 Annual Meeting

IPHC Staff

The following are regulation proposals received from the industry.

Number Received From Title / Topic Type1 James Whitethorn-

West Brothers Group Harvest ticket for Alaska halibut and black cod

Management

2 Shawn Crittenden- Not just a Sport Fisherman

Statewide charter tag (for Oregon) Management

3 Ronn Buschmann- Individual

Adoption of circle hook as the legal gear in the directed halibut fi sheries

Regulatory

4 Don Martin-Individual

Account for preserved fi sh on board sport fi shing vessels

Regulatory

5 Ronn Buschmann-Individual

Halibut catch and (careful) release regulation

Regulatory

Proposal numbers 6 and 7 were forwarded on to the North Pacifi c Fishery Management Coun-cil (NPFMC) by letter on December 1, 2012. They do not need to be discussed as part of IPHC deliberations as they are not relevant to IPHC regulations. They are provided for information only.6 Robert Mosher-

IFQ holderIn Alaska, change from blocks to percent of catch limit

NA

7 James Whitethorn-West Brothers Group

In Alaska, allow D-class halibut on C-class vessels

NA

8 Letter to NPFMC Letter referring to props 6 and 7


#1


#2


IPHC Regulations Proposal Submission Form Proposal Title:Adoption of the Circle hook as the only legal gear in directed halibut fisheries Year Proposed For: 2013 Submission Information (Please print or type) Name: Ronn Buschmann Affiliation: self Address:Box 1367 City:Petersburg State/Prov: AK Postal/ZIP Code: 99833 Telephone(907) 723-1642 Fax: Email:[email protected] Signature: Ronn Buschmann 1. What is the definition and objective of the proposal? This proposal is to adopt the Circle hook or the soft circle type hooks used commonly with mechanical baiting machines as the sole legal gear in directed commercial, charter(both guided and unguided), recreational, and subsistence halibut fishing. This would outlaw “J” hooks which are still used on some commercial boats and treble hooks which some charter and recreational users utilize. The traditional Native American halibut hook would be legal gear in the personal use and subsistence halibut fishery. 2. Impacts: Describe who you think this proposed change might affect (include fishers,

processors, agencies, and the public). 2a. Who might benefit from the proposed change? The resource. Presently the

commission has adopted a slot limit for 2C charter operations and there is also an influx of small fish into Southeast Alaska as well as other areas. This will lead to more sorting in both the commercial and recreational sectors as the commercial fishermen release <32” fish and the recreational fishermen deal with regulations like the one fish charter limit by releasing smaller fish in favor of larger fish. Halibut tend to swallow J hooks which can be impossible to remove without damaging the fish and treble hooks are very difficult to remove from a live fish and tend to catch in the gills and cause heavy bleeding. The circle hook is most likely to hook in the harder parts of the mouth just inside the jaw, cause little bleeding and is easily removed. The purpose of this regulation is to increase survivability among fish that are caught and released for whatever reason.

2b. Who might suffer hardships or be worse off? Some operators will be required to purchase circle hooks to replace their existing gear. Those who are fishing multiple species, i.e. salmon and halibut, will need to carry two sets of gear. 3. Are there other solutions to the problem described above? If so, why were they rejected? None

#3


#4


IPHC Regulations Proposal Submission FormProposal Title: Halibut Catch and Release Regulation

Year Proposed For: _2013

Name: Ronn Buschmann

Affiliation: self

Address:P. O. Box 1367

City:Petersbur~ StatelProv: AK Postal/ZIP Code: 99833

Signature:Ronn Buschmann

Telephone(907) 723-1642 Fax: (' (u ,

1. What is the definition and objective of the proposal? The objective of the proposal is toestablish a guideline for the catch and release of halibut which are not to be retained for anyreason but generally because they are too small or do not fit slot limit requirements. Thisregulation specifically prohibits gaffing or harpooning any fish which may not be retained. Afish which is to be measured to establish that it is >32" in the commercial fishery or which maybe released in the recreational or charter fishery must be released unharmed. A small fish maybe removed from the water and placed on deck for a short period of time by lifting it with thehook and leader used to catch it. Larger fish must be measured in the water. Any fish which isgaffed or harpooned must be taken and reported.

3. Are there other solutions to the problem described above? H so, why were they rejected?

none

2. Impacts: Describe who you think this proposed change might affect (include fishers,processors, agencies, and the public).2a. Who might benefit from the proposed change? The resource and future fishers in all

sectors. We are at as point in this fishery where we have a declining biomass, increasingdemand, and unmeasured impacts from the cod and pollock trawl fishery. It is necessary thatwe take whatever measures we can to increase survivability among those fish which arereleased.

#5


#6


#7


December 1, 2012

VIA EMAIL

Mr. Eric Olson, Chair North Pacific Fishery Management Council 605 West 4th Avenue, Suite 306 Anchorage, AK 99501-2252

Dear Eric, The International Pacific Halibut Commission completed its Interim Meeting this week. Regulation Proposals were reviewed at the meeting, and two of the proposals recommend changes to the Alaska IFQ program. As such, we are forwarding the proposals to you for the Council review process. The proposals are attached and include the following: Change from Blocks to Percent of Catch Limit; and To Be Able to Fish D Class Halibut on C Class Boats.

Sincerely,

Bruce M. Leaman Executive Director

attachments

#8


Appendix III. Catch limit, research, stock assessment and apportionment comments submitted for the 2013 Annual Meeting

IPHC Staff

The following proposals received from the industry were completed on the Regulation Proposal and Catch Limit Forms and are relating to the stock assessment and apportionment process, research, and catch limits.

Number Received From Title / Topic Area1 James Whitethorn-

West Brothers Group Survey based and historical catch share plan

Coastwide

2 Rex Murphy-Alaska Charter Association

Abundance based management for all halibut removals

Coastwide

3 Greg Sutter-Alaska Charter Association

Sport and Commercial (Setline) hook and release mortality study

Research

4 Rob Jones-Northwest Indian Fisheries Commission

Improving understanding of halibut biomass in Area 2A

Area 2A

5 James Whitethorn-West Brothers Group

Area 2C total removals of 8.4 M lb and commercial catch (limit) of 6.6 M lb.

Area 2C

6 Melvin B. Grove Jr.-Prince William Sound (PWS) Charter Boat Association

Closed Area for longline between Me-morial Day to Labor Day in IPHC stat area 230-240

Area 3A – PWS

7 Dave Fraser-Adak Community Develop-ment Corporation

Recommendation of harvest rate and percent of coastwide total biomass

Area 4A


#1


IPHC Regulations Proposal Submission Form Proposal Title: Abundance based management of all halibut removals Year Proposed For: 2016 Submission Information (Please print or type) Name: Rex Murphy Affiliation: Alaska Charter Association Address: PO Box 478 City: Homer State/Prov: AK Postal/ZIP Code: 99603 Telephone: 907-235-8282 Fax:907-235-8282 Email: [email protected]

Signature: 1. What is the definition and objective of the proposal? The objective is abundance-based management of all halibut removals by the IPHC. Article III of the Protocol allows the IPHC to limit the size and quantity of halibut taken by directed fisheries, and additionally allows the IPHC during both open and closed seasons, to permit, limit, regulate or prohibit the incidental catch of halibut that may be taken, retained, possessed, or landed from each area or portion of an area, by vessels fishing for other species of fish1. Removals of halibut under 32 inches (U32) represent future lost yield for directed fisheries. The IPHC currently does not manage total U32 removals. While the total CEY represents the sustainable amount of O32 removals, the IPHC has deducted U32 recreational and subsistence catch from the total CEY for years, and in 2011 also began deducting O26/U32 bycatch and wastage from the total CEY. In an attempt to compensate directed fisheries for the O26/U32 removals from the total CEY, the IPHC increased the harvest rates in all areas. These changes amount to a wash in some but not all areas. In areas 3A and 3B, directed fisheries suffered a decrease in commercial catch limits of approximately 310,000 and 630,000 pounds in each area respectively relative to the 2010 methodology. (See attached Tables 1 and 2.) The deduction of O26/U32 bycatch from the O32 total CEY is in fact a reallocation of directed fishery TAC to compensate for non-directed removals. Adjustments to harvest rates and deductions of U32 removals that are not part of the O32 total CEY are confusing and non-transparent methods of managing the resource. This proposal suggests that IPHC actively manage all halibut removals by establishing a U32 Total CEY and U32 catch limits in addition to the current O32 Total CEY and catch limits. This approach maintains the historical continuity of the O32 harvest data, while transparently managing all removals on an abundance basis. Upon

1 Article III of the Protocol amending the Convention between Canada and the United States of America for the Preservation of the Halibut Fishery of the Northern Pacific Ocean and Bering Sea. Available at: http://www.lexum.com/ca_us/en/cts.1980.44.en.html

#2


implementation, only O32 removals would be deducted from the O32 Total CEY, U32 removals would be deducted from the U32 Total CEY and the allocation and management of both O32 and U32 removals would be handled domestically. This approach would require fundamental changes in how U32 removals are managed by the IPHC as well as domestically; therefore an implementation year of 2016 is suggested. A reasonable first step would be to task IPHC staff to prepare a discussion paper on managing U32 removals for review at the 2014 meeting. 2. Impacts: Describe who you think this proposed change might affect (include fishers,

processors, agencies, and the public).

The proposed change will affect fishermen, processors, agencies, the resource and the public, with the hopeful results of maximizing sustainable yield and minimizing bycatch and wastage.

2a.Who might benefit from the proposed change?

A comprehensive approach to resource management is obviously preferable to an approach that does not actively manage all removals according to abundance. The future of the directed fishery depends on the health of the entire stock. The IPHC is charged with maintaining stocks at optimal yield, so it makes perfect sense that all halibut removals must be actively managed to attain this goal.

Directed fisheries will benefit in areas where the implementation of O26/U32 deductions from the total CEY has negatively impacted their total allowable catch.

2b.Who might suffer hardships or be worse off? Depending on the amount of U32 removals that the IPHC determines is sustainable, and depending on how U32 removals are allocated and managed domestically, directed or non-directed fisheries with U32 removals could be affected.

3. Are there other solutions to the problem described above? If so, why were they rejected? Please attach any other supporting materials. All items submitted by November 2, 2012 will be considered at the IPHC Annual Meeting. Remember to include contact information and signature.


Table 1: 2012 Catch Limit Calculations including O26/U32 Bycatch and Wastage in Other Removals2

Reg Area eBio

Harvest Rate

Total CEY

Other Removals including O26/U32

2012 Fishery

CEY

Slow Up- Full

Down

2012 Staff Catch Rec

2A 6.15 0.215 1.32225 0.17 1.15225 -0.16 0.99 2B 34.9 0.215 7.5035 0.87 6.6335 0.00 6.63 2C 27.28 0.215 5.8652 2.65 3.2152 -0.59 2.63 3A 92 0.215 19.78 7.86 11.92 0.00 11.92 3B 41.17 0.161 6.62837 1.57 5.05837 0.00 5.06 4A 14.86 0.161 2.39246 0.83 1.56246 0.00 1.56 4B 14.25 0.161 2.29425 0.43 1.86425 0.00 1.86

4CDE 29.4 0.161 4.7334 2.28 2.4534 0.00 2.45 Table 2: 2012 Catch Limit Calculations not including O26/U32 Bycatch and Wastage in Other Removals3

Reg Area eBio

Harvest Rate

Total CEY

Other Removals not

including O26/U32

2012 Fishery

CEY

Slow Up- Full

Down

2012 Catch Rec

2A 6.15 0.2 1.23 0.13 1.10 -0.12 0.98 2B 34.9 0.2 6.98 0.58 6.41 0.00 6.41 2C 27.28 0.2 5.456 2.50 2.96 -0.46 2.50 3A 92 0.2 18.4 6.17 12.23 0.00 12.23 3B 41.17 0.15 6.1755 0.49 5.69 0.00 5.69 4A 14.86 0.15 2.229 0.49 1.74 0.00 1.74 4B 14.25 0.15 2.1375 0.31 1.83 0.00 1.83

4CDE 29.4 0.15 4.41 1.41 3.00 0.00 3.00

2 eBIo, other removals sourced from 2012 IPHC Bluebook, page 207. Does not factor in non-scientific “other considerations” applied in 2012. 3 eBIo, other removals sourced from 2012 IPHC Bluebook, page 207. O26/U32 bycatch and wastage sourced from 2011 IPHC RARA pages 185 and 187. Assumes 2011 catch limits were also calculated not including bycatch and wastage.


IPHC Regulations Proposal Submission FormProposal Title:__Sport Halibut and Commercial Setline Hook and Release Mortality Study_

Year Proposed For: __2013____________

Submission Information (Please print or type)

Name: Greg Sutter

Affiliation: Alaska Charter Association

Address:

City: Homer State/Prov: AK Postal/ZIP Code: 99603

Telephone Fax: Email:

Signature: pp. 1. What is the definition and objective of the proposal? A research study should be conducted to establish current hook and release mortality rates for both the sport halibut and commercial setline fisheries. This study should be conducted prior to pursuing any policy for separate accountability for the sport and commercial sectors.

2. Impacts: Describe who you think this proposed change might affect (include fishers, processors, agencies, and the public).

2a. Who might benefit from the proposed change? All users of the halibut fishery will benefit when there is accurate accountability for all removals, including wastage.

2b. Who might suffer hardships or be worse off? Sport fishers and setline fishermen who might be assigned an arbitrary high rate1 of hook and release mortality will be denied their full utilization of their harvest allocation (Optimal Yield).

3. Are there other solutions to the problem described above? If so, why were they rejected? Nohook and release mortality has ever been applied to the sport halibut fishery in the past. The only scientific studies actually performed on Pacific halibut were conducted in 1958 and 1960 using J-hooks, and the number of samples was very small due to problems with high water temperature2.With the predominant use of circle hooks in both sectors and the growing use of barbless circle hooksin the sport sector, it is believed that sport fish release mortality should be much lower than 6%. Likewise, a commercial release mortality of 16% is most likely higher than actual.

Please attach any other supporting materials. All items submitted by November 2, 2012 will be considered at the IPHC Annual Meeting. Remember to include contact information and signature.

1 Existing data for sport hook and release mortality is outdated and flawed by current standards of research.2 Viability of tagged Pacific halibut. Gordon J. Peltonen. 25 p. (1969). Available at: http://www.iphc.int/publications/scirep/Report0052.pdf

#3


#4


#5


#6


#7

IPHC_bluebook_2013.pdf

Documents

Transcript of IPHC_bluebook_2013.pdf