an overview of the key issues to be discussed relating to ... · maram international stock...

An overview of the key issues to be discussed relating to the South African sardine two-mixing-stock hypothesis

MARAM International Stock Assessment Workshop30th November 2015

Carryn de Moor

Marine Resource Assessment and Management Group (MARAM)Department of Mathematics and Applied MathematicsUniversity of Cape Town

Outline

• Fishery and Data• Stock Structure Hypotheses• Sardine Two Stock Assessment Model Fits• Considering the Parasite Data

Arisk Arisk

Catch History

Small Pelagic Fishery: primarily sardine and anchovy, also round herring, chub mackerel and juvenile horse mackerel

0

100000

200000

300000

400000

500000

600000

700000

800000

1949

1952

1955

1958

1961

1964

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

2009

2012

Annu

al La

ndin

gs (t

ons)

Directed sardine Anchovy

Sardine Bycatch Horse Mackerel

Round Herring Chub Mackerel

Data

• Two annual hydro-acoustic surveys- November total biomass; split at Cape Agulhas- May recruitment; split at Cape Infanta

Arisk Arisk

Hydro-acoustic Surveys

Provided by Janet Coetzee

Cape Agulhas

Cape Infanta

Hydro-acoustic Surveys

0

10

20

30

40

50

60

70

0

500

1000

1500

2000

2500

3000

3500

4000

1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014

Biomass Recruitment

West Coast

0

1

2

3

4

5

6

0

500

1000

1500

2000

2500

3000

3500

4000

1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014

Biomass Recruitment

South Coast

0

10

20

30

40

50

60

70

0

500

1000

1500

2000

2500

3000

3500

4000

1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014

Biomass Recruitment

Full AreaTotal Area

West Coast South Coast

Distributional Shift

Provided by Janet Coetzee

Data

• Two annual hydro-acoustic surveys- November total biomass; split at Cape Agulhas- May recruitment; split at Cape Infanta

• Limited ageing information – excluded from current assessment• Proportion at length data

- Commercial catches (quarterly)- November survey

• Parasite prevalence at length data *NEW*

Arisk Arisk

Outline


Arisk Arisk

Stock Structure Hypotheses

• Single sardine stock- used historically- previous & current OMPs were developed using a single stock operating model

• Two sardine stocks (“west” and “south”) split at Cape Agulhas- a two stock model has been under development since 2009- a two mixing stock hypothesis produced an acceptable fit to the data in time for OMP-14 development, but future projections were heavily dependent on assumption of future stock mixing

• Current OMP includes some spatial management advice (i.e. based qualitatively not quantitatively on two stock hypothesis)

Arisk Arisk

Key Model Structure/Assumptions

• Operating model conditioned over Nov 1983 – Nov 2014• Quarterly time steps• Catch-at-age taken in a pulse each quarter• Maturity ogive that is time-dependent (4 time-periods) *NEW*• Time-invariant juvenile and adult M• Hockey stick S/R relationship• Estimate commercial selectivity-at-length (parametric form used

which was informed by initial work with a non-parametric curve)

• Survey trawl selectivity-at-length logistic for small lengths• Survey biomass is associated with 0+ biomass *NEW*• Weight-at-length relationship, adjusted annually based on

average survey weight *NEW*

Arisk Arisk

Key Model Structure/Assumptions• Two sardine stocks: key assumptions

- mixing : “west” stock recruits move to “south” stock (estimated annually - no assumed relationship)- mixing : no “south” recruits move to “west” stock- mixing : no adult movement- 0+ biomass west/east of Cape Agulhas is from the “west” / “south” stock- recruitment west/east of Cape Infanta is from to the “west” / “south” stock- same Mjuv and Madult for both stocks- allowance for difference in May survey bias east/west of Cape Infanta(proportion of recruits surveyed east of Cape Infanta ≤ that surveyed west of Cape Agulhas (winter spawning))- allowance for difference in weights and growth by stock- S/R curve parameters estimated separately for each stock

Arisk

Key Model Structure/Assumptions• Two sardine stocks: key assumptions

- mixing : “west” stock recruits move to “south” stock (estimated annually - no assumed relationship)- mixing : no “south” recruits move to “west” stock- mixing : no adult movement a lower proportion of 2+ sardine move- 0+ biomass west/east of Cape Agulhas is from the “west” / “south” stock- recruitment west/east of Cape Infanta is from to the “west” / “south” stock- same Mjuv and Madult for both stocks- allowance for difference in May survey bias east/west of Cape Infanta(proportion of recruits surveyed east of Cape Infanta ≤ that surveyed west of Cape Agulhas (winter spawning))- allowance for difference in weights and growth by stock- S/R curve parameters estimated separately for each stock- age- and time-invariant annual infection of “west” stock sardine

Arisk

Fish age Infection Movement

Outline

• Fishery and Data• Stock Structure Hypotheses• Sardine Two Stock Assessment Model Fits – INITIAL RESULTS• Considering the Parasite Data

Arisk Arisk

Four Alternative Initial AssumptionsArisk Arisk

Model

AnnualProportion of

West Stock Infected

Age-1 fish move Age 2-5+ fish move

a) 20% Yes No

b) 40% Yes No

c) 20% Yes Yes

d) 40% Yes Yes

0

500

1000

1500

2000

2500

3000

3500

1983 1988 1993 1998 2003 2008 20130+

Bio

mas

s ('0

00t)

Year

Age1 20%Age1 40%Age1+ 20%Age1+ 40%

South

0

500

1000

1500

2000

2500

3000

3500

1983 1988 1993 1998 2003 2008 2013

0+ B

iom

ass

('000

t)

Year

Age1 20%Age1 40%Age1+ 20%Age1+ 40%

West

Model Fits – Nov BiomassArisk Arisk

0123456789

10

1983 1988 1993 1998 2003 2008 2013R

ecru

its (i

n bi

llion

s)

Year

Age1 20%Age1 40%Age1+ 20%Age1+ 40%

South

0

20

40

60

80

100

1983 1988 1993 1998 2003 2008 2013

Rec

ruits

(in

billi

ons)

Year

Age1 20%Age1 40%Age1+ 20%Age1+ 40%

West

Model Fits – May RecruitmentArisk Arisk

• Note difference in y-axis• No south coast survey estimates for early part of time series

• Note difference in y-axis; but west stock maximum recruitment lower than previous model

• West stock σR larger than previous model

• Similar plots for alternative models

0

5

10

15

20

0 500 1000 1500R

ecru

its (i

n bi

llion

s)

Spawner Biomass ('000t)

South

0

25

50

75

100

125

0 200 400 600 800 1000

Rec

ruits

(in

billi

ons)

Spawner Biomass ('000t)

West

Model Fits – Stock RecruitmentArisk Arisk

• Increase in “south” stock biomass primarily a result of movement from “west” stock rather than “south” stock productivity

• Proportion of 2+ sardine moving is ~25% of age-1 sardine

0

0.2

0.4

0.6

0.8

1

1.2

1983 1988 1993 1998 2003 2008 2013

Prop

ortio

n m

ovin

g

Year

a)

0

0.2

0.4

0.6

0.8

1

1.2

1983 1988 1993 1998 2003 2008 2013Pr

opor

tion

mov

ing

Year

b)

Model Fits – Estimated MovementArisk Arisk

20% Infection Proportion 40% Infection Proportion

• Increase in “south” stock biomass primarily a result of movement from “west” stock rather than “south” stock productivity

• Proportion of 2+ sardine moving is ~25% of age-1 sardine

0

2

4

6

8

10

12

1983 1988 1993 1998 2003 2008 2013Bi

omas

s mov

ing

('000

t)Year

Model Fits – Estimated MovementArisk Arisk

20% Infection Proportion 40% Infection Proportion

0

2

4

6

8

10

12

1983 1988 1993 1998 2003 2008 2013

Biom

ass m

ovin

g ('0

00t)

Year

00.20.40.60.8

11.2

0 5 10 15 20 25

Surv

ey S

elec

tivity

length (cm)

Age1 20%Age1 40%Age1+ 20%Age1+ 40%

West

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0 5 10 15 20 25

Length Class (cm)

ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%

South

Model Fits – Survey Prop at LengthAverage predicted is much closer to average observed than previous model

00.010.020.030.040.050.060.070.08

0 5 10 15 20 25

Length Class (cm)


West

00.20.40.60.8

11.2

0 5 10 15 20 25

Surv

ey S

elec

tivity

length (cm)

Age1 20%Age1 40%Age1+ 20%Age1+ 40%

South

South coast winter spawning?

00.020.040.060.080.1

0.120.140.16

0 5 10 15 20 25

Length Class (cm)


South

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0 5 10 15 20 25

Length Class (cm)


West

00.20.40.60.8

11.2

0 5 10 15 20 25Com

mer

cial

Sel

ectiv

ity

length (cm)

Age1 20%Age1 40%Age1+ 20%Age1+ 40%

South

00.20.40.60.8

11.2

0 5 10 15 20 25Com

mer

cial

Sel

ectiv

ity

length (cm)

Age1 20%Age1 40%Age1+ 20%Age1+ 40%

West

Model Fits – Commercial Prop at Length

Outline


Arisk Arisk

Comparison of Model Predicted Parasite Prevalence With Observations

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


West: 20100

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


South: 2010

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


South: 20110

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


West: 2011

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


South: 20120

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


West: 2012


0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


West: 2013 0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


South: 2013

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


South: 2014


0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

Length Class (cm)


West: 2014

Some Discussion Points

Comment on the ongoing development of the two mixing stock hypothesis, and the use of parasite data as a ‘biological tag’:• The revised model structure• Fitting the model to parasite prevalence data (and/or parasite

intensity data)• Assuming a time- and age-invariant infection rate• Estimating annual age-1 movement proportions, with ages 2+ a

time-invariant proportion of those estimates• Modelling two south stock co-horts (1 Nov, 1 May) to account for

winter spawning• The chosen stock-recruitment relationship

An overview of the key issues to be discussed relating to the South African sardine two-mixing-stock hypothesis

Thank you for your attention

an overview of the key issues to be discussed relating to ... · maram international stock...

Documents