Assessment of Pacific (Clupea pallasii) populaons in the northeast Pacific

Jaclyn Cleary Fisheries and , Canada Sherri Dressel Alaska Department of Fish and Game Overview of Brish Columbia (Canada) and Alaska (USA) Pacific herring

• Populaons and fisheries • Surveys and stock assessment methods • Biggest challenges • Collaboraon and advancement

2 Global distribuon of herring stocks throughout eastern North Atlanc, Arcc and Pacific oceans

ASIA N. AMERICA

EUROPE *

Hay 1985. CJFAS

3 stocks in the northeast Pacific Ocean 3,000

AK BC BC

Alaska

2,000 things

No r (km) British Columbia

Northeast 1,000 Pacific Ocean

−2,000 −1,000 0 1,000 * Eastings (km) Projection: BC Albers (NAD 1983) Pacific herring stocks in the northeast Pacific Ocean that are assessed with age-structured assessment models Pacific herring stocks in the northeast Pacific Ocean

2,000 Alaska AK Kamishak Bay Togiak Bay BC Prince William Sound

1,500 Tenakee Inlet British Seymour Canal Columbia Sitka Sound

Craig

Northeast Northings (km) 1,000 Pacific Prince Rupert District

Ocean Haida Gwaii

Central Coast

Strait of Georgia 500 West Coast Vancouver Island

−1,000 −500 0 500 1,000 * Eastings (km) 5 Projection: BC Albers (NAD 1983) Significance of herring in the Northeast Pacific …

• Key forage fish

NMML

hp://en.wikipedia.org/wiki/Coho_salmon

6 Significance of herring in the Northeast Pacific …

• Key forage fish • High cultural and socio-ecological value for BC Coastal First Naons and Nave Alaskans

Photo credits: Sitka Tribe of Alaksa, H.Kitka

7 Significance of herring in the Northeast Pacific …

• Key forage fish • High cultural and socio-ecological value for BC Coastal First Naons and Nave Alaskans • Important commercial industry

8 Surveys

Stock assessment models

Biomass trends

9 Available data: Pacific herring stocks in the northeast Pacific Ocean

2,000 Alaska AK Kamishak Bay Togiak Bay BC Prince William Sound

1,500 Tenakee Inlet Data rich stocks British Seymour Canal 30 to 60+ year me series of: Columbia Sitka Sound • Validated catch Craig • Age, length, weight, sex

• Survey indices (egg, milt, Northeast Northings (km) 1,000 biomass) Pacific Prince Rupert District -4 different survey methods Ocean Haida Gwaii

Central Coast Periodic studies to examine: • Maturity, fecundity Strait of Georgia 500 West Coast Vancouver Island

−1,000 −500 0 500 1,000 * Eastings (km) Projection: BC Albers (NAD 1983) 10 Survey methods and locaons: Survey type

2,000 Alaska Aerial surveys Egg deposition (spawn) surveys Egg index + Milt survey + Acoustic Index

1,500 Togiak, Kamishak: British Aerial surveys of biomass Columbia

PWS: 3 survey indices: Northeast

Northings (km) Egg, mile-days of milt, acousc 1,000 Pacific SE AK and BC stocks: Ocean Egg deposion (spawn) surveys

500

−1,000 −500 0 500 1,000 Eastings (km) 11 Projection: BC Albers (NAD 1983) Feb-June: Aerial photos of herring spawn acvity

Typical colour of Pacific ocean

Herring milt turning water aqua blue

Photo credit: K. Daniel Herring egg surveys: 1) Idenfy spawn from flights 2) Determine shoreline length of spawn 3) Transect and quadrat placement 4) Collect quadrat data (egg layers/numbers, vegetaon type and percent) 5) Extrapolate to total survey area (shoreline length, spawn width (i.e. transect length), egg density)

Photo credit: DFO

Photo credit: DFO

Photo credit: ADFG, S.Dressel Photo credit: DFO, K.Daniel Eggs to fish

Number of eggs Fecundity relaonship

Assume 50:50 sex rao

Biomass of fish Approaches to stock assessment are similar among regions BC BC and Alaska Alaska Catch age models, fied to: Minimize negave log • Survey index likelihood for catch data, • Proporons at age (obs) Least squares to minimizaon differences between obs. and relave abundance index, • Weight at age (obs) esmated total or relave age-composion, and • Catch (obs/ validated) stock-recruit relaonships abundance index, age-

(given priors distribuons composions, and stock- and penalty funcons) Objecve funcon recruit relaonships (not all)

Models esmate: Esmated using logisc Fixed maturity at age • Selecvity funcon over all years or in • Recruitment me blocks Esmated deviaons in M • Maturity at age Esmated over all years or in using random walk • Natural mortality me blocks

Bayesian esmaon: • Spawning biomass and

SBt and SBT+1 reflect uncertainty Frequenst esmaon: median values of marginal SBt and SBT+1 reflect point posterior distribuons and esmates, and bootstrapped credible intervals confidence intervals (not all)

15 Overview of stock status

16 Stocks with highest overall biomass in recent 25-years

Pacific herring biomass: spawning (lines) and harvest (bars)

Togiak Bay Kamishak Bay Prince William Sound Tenakee Inlet 300 90 90 90 200 60 60 60 100 30 30 30

0 0 0 0

Seymour Canal Sitka Sound Craig Haida Gwaii ) 3

10 90 90 90 90 ) × 3 60 60 60 60

30 30 30 30

Biomass (t 0 0 0 0

Prince Rupert District Central Coast Strait of Georgia West Coast Vancouver Island

90 90 90 90 Biomass (t x 10 60 60 60 60

30 30 30 30

0 0 0 0 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 Year Years (1970-2016) 25-years 17 Stocks experiencing low biomass trends in recent 25-years Pacific herring biomass: spawning (lines) and harvest (bars)

Togiak Bay Kamishak Bay Prince William Sound Tenakee Inlet 300 12.5 90 10.0 20 200 7.5 60 5.0 10 100 30 2.5 0 0 0 0.0

Seymour Canal Sitka Sound Craig Haida Gwaii

) 25 80 3

) 7.5

10 20 3 75 60 × 15 5.0 50 40 10 2.5 25 5 20

Biomass (t 0.0 0 0 0

Prince Rupert District Central Coast Strait of Georgia West Coast Vancouver Island 80 60 75

Biomass (t x 10 90 60 40 50 40 60 20 20 30 25

0 0 0 0 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 Year Years (1951-2016) 25-years 18 Biomass synchrony among stocks between regions… Pacific herring relative spawning biomass

Togiak Bay Kamishak Bay Prince William Sound Tenakee Inlet 1.00

0.75

0.50

0.25

0.00 Seymour Canal Sitka Sound Craig Haida Gwaii 1.00

0.75

0.50

0.25 Relave biomass

Relative biomass Relative 0.00 Prince Rupert District Central Coast Strait of Georgia West Coast Vancouver Island 1.00

0.75

0.50

0.25

0.00 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 Year Years (1970-2016) 19 Biomass synchrony among stocks between regions… PacificSitka, Seymour, Craig herring relative spawning biomass Prince Rupert, Haida Gwaii, Central AK Coast, West Coast Vancouver Island BC

1.00

0.75

0.50 Relative biomass Relative Relave biomass 0.25

0.00 1970 1980 1990 2000 2010 1970 1980 1990 2000 2010 Year

Seymour Canal Craig Prince Rupert District West Coast Vancouver Island Stock Sitka Sound Haida Gwaii Central Coast

20 Opposing trends between regions… Pacific herring stocks in the northeast Pacific Ocean Seymour Canal Sitka Sound Craig Haida Gwaii 25 2,000 Seymour Canal 60 Sitka Sound AlaskaCraig Haida GwaiiAK 7.5 Kamishak Bay 25

) 20 ) ) 75 ) Togiak Bay 3 3 3 3 60 BC 7.5 10 10 10 10

) 20 ) ) 15 7540 )

3 Prince William Sound 3 3 3 × × × 5.0 50 × 10 10 10 10 10 15 40 × × × 5.0 50 × 2.5 25 Sitka 20

SSB (t SSB (t 10 SSB (t SSB (t 5 2.5 25 20 SSB (t SSB (t SSB (t 0.0 0 0 0 SSB (t 5

0.6 0.0 0 0 0 1.00 1,500 1.00 Tenakee Inlet Seymour Canal Sitka Sound 0.6Craig Haida Gwaii 1.00 1.00 25 British 0.75 0.75 Seymour Canal 0.4 0.4 60 Columbia 7.5 0.75 0.75

) 20 ) ) 75 ) 0.4 Sitka Sound 3 0.4 3 3 3

M M M 0.50 M 0.50 10 10 10 10 15 40 × × × × 5.00.2 500.2 M Haida M M 0.50 Craig M 0.50 0.25100.2 0.250.2 2.5 25 20 0.25 0.25 SSB (t SSB (t SSB (t SSB (t Gwaii 0.0 0.0 0.005 0.00 Northeast 1960 1980 2000 1960 1980 2000 0.0 1960 1980 Northings (km) 2000 0.0 1960 1980 2000 0.00 0.00 0.0 0 0 1,000 0 Prince Rupert District 1960 1980 2000 1960 Pacific1980 2000 1960 1980 2000 1960 1980 2000 0.6 Prince Rupert District Central Coast 1.00 Strait of Georgia 1.00West Coast OceanVancouver Island Prince Rupert District Central Coast Strait of Georgia Haida Gwaii West Coast Vancouver Island 80 60 0.7580 0.7575 ) ) ) 0.4 0.4 ) 90 West 60 3 3 3 60 3 Central75 Coast ) ) ) ) 90 3 3 10 10 3 10 3 40 10 60

M M M 0.50 M 0.50 × × × × Coast 50 10 10 10 40 60 40 10 × × × 0.2 0.2 × 50 40 60 20 0.25 0.25 30 Vancouver 25

20 SSB (t SSB (t SSB (t SSB (t 20 30 25 Strait of Georgia

20 SSB (t SSB (t SSB (t 0.0 0.0 0.00 0.00 SSB (t 0 0 0 Island 500 0 1960 1980 2000 1960 1980 2000 0 1960 1980 2000 0 1960 1980 2000 0 West Coast Vancouver0 Island 1.00 1960 1980 2000 Prince Rupert District Central Coast Strait of Georgia West Coast Vancouver Island 1.00 0.75 0.9 80 0.75 −1,000 0.75 −500 0 500 0.9 1,000 * 0.4 60 75 0.75 ) ) ) ) 900.4 Eastings (km) 3 3 3 60 0.50 3 0.6

M M M 0.50 M 10 10 0.50 10 40 10 Projection: BC0.6 Albers (NAD 1983) M M M M

× × 0.50 21 ×

× 50 400.2 60 0.25 0.250.2 0.3 20 0.25 0.25 0.3 30 25

20 SSB (t SSB (t SSB (t SSB (t 0.0 0.00 0.00 0.0 0.0 0.00 0.00 0.0 0 1960 1980 2000 0 1960 1980 2000 0 1960 1980 2000 0 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 1.00 0.75 0.9 0.4 0.75 0.50 0.6

M M M 0.50 M 0.2 0.25 0.25 0.3

0.0 0.00 0.00 0.0 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 What are the biggest challenges to assessing (AK-BC) Pacific herring stocks, and, What can we learn from each other?

22 What are the biggest challenges to assessing (AK-BC) Pacific herring stocks, and, What can we learn from each other?

• Stability in data collecon programs: securing funding for annual surveys and stock assessment • Nonstaonarity: coping with me varying changes in growth, maturity and natural mortality in stock assessment and in the esmaon of biological reference points

23 Example: Esmang natural mortality

• According to Clark (1999, CJFAS): Natural mortality is the most influenal parameter in fisheries stock assessments, relang directly to esmated stock producvity, sustainable catch, opmal exploitaon rates and calculaon of reference points.

24 Example: Esmang natural mortality

Kamishak Bay SitkaSitka Sound WestWCVI Coast Vancouver Island ) 3

75 75 ) ) ) 3 3 3 20 10 10 10 × × × 50 50

10 25 25 Biomass (t Biomass (t Biomass (t

Biomass (t x 10 0 0 0

0.9 0.9 0.9

0.6 0.6 0.6 M M M M

0.3 0.3 0.3

0.0 0.0 0.0 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 Esmang constant M Esmang M in me blocks Esmang me varying across me according to changes in deviaons in M using a PDO if it results in random walk, with cubic improved model fits spline interpolang (defined by AICc criterion)

25 Example: Esmang natural mortality

Kamishak Bay SitkaSitka Sound WestWCVI Coast Vancouver Island ) 3

75 75 ) ) ) 3 3 3 20 10 10 10 × × × 50 50

10 25 25 Biomass (t Biomass (t Biomass (t

Biomass (t x 10 0 0 0

0.9 0.9 0.9

0.6 0.6 0.6 M M M M

0.3 0.3 0.3

0.0 0.0 0.0 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 1960 1980 2000 Addressing crical challenges through collaboraons: • Can we further invesgate nonstaonarity (e.g., in M) within stock assessment models using a joint simulaon study to idenfy best approaches for esmang M for herring • Adopng similar approaches for esmang M could provide consistent basis for

esmang biological reference points (e.g., B0) that are comparable across regions

26 What are the biggest challenges to assessing (AK-BC) Pacific herring stocks, and, What can we learn from each other? Other areas for collaboraon: Catch age models, fied to: • Survey index Joint studies to • Proporons at age (obs) invesgate changes in • Weight at age (obs) weight-age, fecundity, • Catch (obs/ validated) and growth

Objecve funcon

Models esmate: • Selecvity • Recruitment Field study • Maturity at age • Natural mortality Simulaon study • Spawning biomass and uncertainty

27 What are the biggest challenges to assessing (AK-BC) Pacific herring stocks, and, What can we learn from each other?

• We’re keen on developing on-going collaboraons to help us address current challenges such that both countries can advance our respecve herring stock assessment programs with more synchrony and improve assessment approaches for this key forage species

28 Acknowledgements: • Ma Grinnell • Greg Buck • Ted Os • Kyle Hebert • Steve Moffi

29