An overview of the key issues to be discussed relating to ... · MARAM International Stock...
Transcript of 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
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
• Fishery and Data• Stock Structure Hypotheses• Sardine Two Stock Assessment Model Fits• Considering the Parasite Data
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)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
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)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
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)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
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
• Fishery and Data• Stock Structure Hypotheses• Sardine Two Stock Assessment Model Fits• Considering the Parasite Data
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)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
West: 20100
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Length Class (cm)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
South: 2010
0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Length Class (cm)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
South: 20110
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Length Class (cm)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
West: 2011
0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Length Class (cm)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
South: 20120
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Length Class (cm)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
West: 2012
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)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
West: 2013 0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Length Class (cm)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
South: 2013
0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25
Length Class (cm)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
South: 2014
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)
ObservedAge1 20%Age1 40%Age1+ 20%Age1+ 40%
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