Transcript Diapositiva 1

Assessing the ecosystem impacts of fishing in the South
Catalan Sea by developing dynamic simulations
on fishing effort and target species
Marta Coll, Isabel Palomera, Sergi Tudela and Francesc Sardà
Institut de Ciències del Mar (ICM-CSIC)
Barcelona, Spain
1. The South Catalan Sea ecosystem model (SCMEE 2004)
2. The SCS model calibrated with time series of data
3. Temporal dynamic simulations of 5 fishing scenarios
1. The South Catalan Sea ecosystem model
Mass balance model of trophic interactions
Ecopath with Ecosim software version 5.1
40 functional groups from primary producers to main top predators
Includes trawling, purse seining, long lining and troll bait fisheries
6 fishing harbours: Tarragona to Les Cases d’Alcanar
Area modelled of 4300 km2
Tarragona
Cambrils
50-400 m depth: continental shelf and upper slope
Represents the ecosystem in 1994
at
a
la
n
Se
a
L’Ametlla
C
L’Ampolla
Ebro River Delta
Sant Carles de la Ràpita
Les Cases d’Alcanar
Study area
Western Mediterranean
Coll et al., accepted to Journal of Marine Systems
1. The South Catalan Sea model
Ecopath mass balance modelling
Production = Predation + Yield + Net Migration + Biomass accumulation + Other mortality
P
Bi·  i 
B

j
Q
P
Bj·   j·DCij  Yi  Ei  BAi  Bi· i· (1  EEi)
B
B
Basic parameters required per compartment (i):
B: Biomass
P/B: Production per unit of biomass
Q/B: Consumption per unit of biomass
EE: Ecotrophic efficiency (production used within the ecosystem)
1-EE: Other mortality
DCij: Fraction of (i) in the diet of (j)
Y: Catches; E: Net migration; BA: Biomass accumulation
Consumption (i)  Production(i)  Respiration (i)  Unassimilated f ood(i)
Expressed on an annual basis per unit surface area and WW (t·km-2·yr-1)
www.ecopath.org. Pauly et al., 2000. ICES J. Mar. Sci., 57: 697-706; Christensen and Walters. 2004. Ecol. Model., 172(2-4): 109-139.
1. The South Catalan Sea model
Overview of trophic flows and ecosystem structure
Benthic
Demersal
Pelagic
TL
V
Troll bait
Long line
Purse seine
Adult hake
Fin whale
Bonito
Squids
Macro Various small Anchovy
zooplankton pelagics
Jellyfish
Demersal fishes(3)
Demersal sharks
Benthopelagic fishes
Horse
mackerel
III
Sardine
Juvenile
Audouin hake
gull
Blue whiting
Poor cod
Mackerel
Shrimps
Demersal Flatfishes Mullets Octopuses
Demersal fishes(2) fishes(1)
Crabs
Marine turtles
Polychaetes
Seabirds
Zooplankton
Suprabenthos
II
Detritus
Phytoplankton
Discards
I
Anglerfish
Conger eel
Large pelagics
IV
Trawl
Dolphins
By catch
Norway lobster
Benthic
invertebrates
1. The South Catalan Sea model
Impact of fishing activities
Cebo
Palangre
Cerco
Low TLc
Low OI
High PPR
Arrastre
Detr
Desc2
Desc1
Bala
Odon
Aves
Laud
Tmar
Poce
Sard
Scom
23.21
10.43
6.98
1.36
41.99
Trac
15.95
13.78
5.47
1.50
36.70
Ppel
0.13
0.01
0.06
0.06
0.10
Spil
3.16
3.01
4.04
4.16
3.12
Engr
%PPR (pp+det)
Pzoo
%PPR (pp)
Squa
OI
Ppec
TLc
Pinv
Pmix
Micr
Mmer
Mjuv
Tris
Pleu
Loph
Discards
t·km-2·y-1
0.23
0.14
0.01
0.00
0.37
Cong
Mull
Cefm
Cefb
Inve
Rept
Nata
Poli
Supr
Pgel
Mzoo
Zoo
Fito
Trawling
Purse seining
Longlining
Troll bait fishery
Total
Neph
Landings
t·km-2·y-1
2.17
2.61
0.17
0.03
4.98
Fishing feet
1.0
Wide and intense
fishing impact
0.5
0.0
-0.5
Tb
L
P
T
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
1
2
1. Trawling fleet
-1.0
1.0
Target species,
predators
and by-catch
0.5
0.0
-0.5
Tb
L
P
T
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
1
2
2. Purse seine
-1.0
1.0
0.5
0.0
-0.5
1.0
0.5
0.0
-0.5
4. Troll bait
-1.0
Tb
L
P
T
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
1
2
3. Longline
-1.0
Target species,
preys and bycatch
1. The South Catalan Sea model
In summary…
The mass balance modelling is a good tool to summarize and integrate the
available information in a coherent way, identifying critical gaps and describing
the ecosystem structure and functioning:
* Quantification of trophic flows, globally or by components
* Estimation of TLs, OI, Mortalities: M2, M0, F
* Indices related with network and information analysis
* Quantification of fishing impact through the MTI, PPR, TLc, GE…
The starting point from where to develop dynamic simulations with the
temporal dynamic module Ecosim:
* Assessing the impact of fishing trough time by changing fishing mortalities or
fishing effort by gear (from an initial value of Ecopath)
* Fitting the model to available data, searching for trophic interactions parameters
and environmental anomaly
* Applying optimization routines to include economic and social data
Walters et al. 1997. Rev. Fish Biol. and Fish., 7: 139-172, Christensen and Walters. 2004. Ecol. Model., 172(2-4): 109-139.
2. Temporal dynamic modelling and calibration process
Ecosim takes the Ecopath master equation and sets up a series of
differential equations of biomass dynamics to calculate changes of each
group over time:
dBi  P 
  i·  Qji   Qij  Ii  (MOi  Fi  ei)·Bi
dt  Q  j
j
dBi/dt: growth rate during time dt of group (i) in terms of its biomass
P/Q: net gross efficiency
MOi: other non-predation natural mortality
Fi: fishing mortality
Ii: immigration rate; ei: emigration rate; Ii-ei·Bi: net migration rate
 Qji
 Qij
Total consumption by group i
j
Total consumption on group i by all predators j
j
Walters et al. 1997. Rev. Fish Biol. and Fish., 7: 139-172, Christensen and Walters. 2004. Ecol. Model., 172(2-4): 109-139.
2. Dynamic modeling and calibration process
Qs are calculated based on the Foraging arena where Bi is divided
into vulnerable and non-vulnerable components and the transfer
rate vij determines the flow control: top-down, bottom-up or
intermediate
v ij·aij·Bi·Bj·Ti·Tj·Sij·Mij/Dj
Qij 
v ij v ij·Ti·Mij aij·Mij·Bj·Sij·Tj/Dj
* vij is expressing the rate with which B move between being vulnerable
and not vulnerable
* Bi is prey biomass; Bj is predator biomass
* aij is the effective search rate for i by j
* Ti and Tj is relative feeding time for prey and predator
* Dj represents effects of handling time as a limit to consumption rate
* Sij are seasonal or long term forcing effects
* Mij are mediation forcing effects
Walters et al. 1997. Rev. Fish Biol. and Fish., 7: 139-172, Christensen and Walters. 2004. Ecol. Model., 172(2-4): 109-139.
2. Dynamic modeling and calibration process
Fitting the model to data …
EwE now includes an iterative process to fit the model and calibrate it with
empirical data
To explore how changes in functional groups can be attributed
* to internal ecosystem factors: feeding interactions and population factors
* to external ecosystem factors: fishing activity and environmental forcing
- From an ecosystem model of a past situation
- Using to force the model changes in:
* fishing effort, fishing mortality, total mortality
- Using available information on biomasses and catches to modify model variables
(mainly vulnerability factor vij) based on the reduction of the goodness of fit
measure that it is the summed-squared residuals (SS) of a predicted from an
observed value
Walters et al. 1997. Rev. Fish Biol. and Fish., 7: 139-172, Christensen and Walters. 2004. Ecol. Model., 172(2-4): 109-139.
2. Dynamic modeling and calibration process
South Catalan Sea calibrated ecosystem model: from 1978-2003
We developed an ecosystem model representing 1978-1979
We used changes in fishing mortality and nominal fishing effort of trawling,
purse seining and longline fishery ► best fit: Cv > TRB > nº boats > days fishing
Absolute and relative biomasses to fit the model
Corrected catches from 1978-2003 to compare results
Change vulnerabilities of most sensible interactions (vij) and prediction of an
environmental anomaly
25
20
Sardine biomass
-2
Biomass (t·km )
Anchovy biomass
15
10
5
0
1975
1980
1985
1990
Year
1995
2000
2005
2. Dynamic modeling and calibration process
Predicted and empirical biomasses and
catches1
Fitting summary results
Total
Fishing
SS values
215.22
194.48
57.75/95.68
98.27/108.67
50.07/71.30
Vulnerabilities
Environmental annomaly*
Combined
* Probability occuring by chance: 0.001
Contribution (%)
9.64
63.53-45.91
3.57-11.32
76.74-66.87
* Trophic flow control for the
most sensible prey-predator
interactions: e.g. sardine
* Anomaly function linked with
primary production correlated
with NAO indexes (annual and
winter values) and time series of
temperature
* Identification of compensation
in recruitment of hake when
adult stock is low (commonly
defined in many stocks, Myers
and Cadigan, 1993)
1. Relative and absolute values (y) over time (x); Myers and Cadigan, 1993. Can. J. Fish. Aquat. Sci., 50: 1576-1590.
2. Dynamic modeling and calibration process
Summary: what seems to have happened in these 26 years?
1. Trophic interactions play a key role in explaining the variability
2. Fishing dynamics: adult hake, sardine, anchovy and demersal sharks
3. Environmental forcing paying its role in the pelagic compartment
The model predicts important changes in the
ecosystem structure and functioning, highly
exploited from 1978 and overexploited in 2003
* Biomass decrease of top predators like adult
hake and demersal sharks
* Biomass decrease of target low TL organism:
like small pelagics and juv. hake
* Increase of benthopelagic fishes, jellyfish,
conger eel and small demersal fishes: preys and
competitors
* Lower biomass of sardine than anchovy in
2003
* Lowest levels of anchovy in the late 1990s and
showing modest recovering…
Functional groups
Jellyfish
Shrimps
Crabs
Norway lobster
Benthic cephalopods
Benthopelagic cephalopods
Mullets
Conger eel
Anglerfish
Flatfishes
Juvenile hake
Adult hake
Demersal fishes (1)
Demersal fishes (2)
Demersal fishes (3)
Demersal sharks
Benthopelagic fishes
European anchovy
European pilchard
Other small pelagic fishes
Horse mackerel
Mackerel
Total
B2003/B1978
1.28
0.83
0.58
0.47
0.60
0.47
0.24
1.54
0.76
0.58
0.63
0.08
0.59
1.48
1.20
0.06
4.28
0.72
0.19
0.87
0.64
0.55
0.78
C2003/C1978
1.44
1.02
0.81
1.12
0.84
0.57
3.50
1.32
1.79
1.08
1.65
1.27
2.99
2.22
0.15
7.30
0.55
0.79
1.39
1.41
1.28
1.61
3. Dynamic simulations
3. Development of dynamic simulations of fishing options
* Changing the fishing mortality (F) by group or fishing effort by fleet
* Assuming constant the predicted environmental anomaly
* Making simulations of 20 years from 2003
* Assessing the impact of changing fishing activity
* Comparing predicted values of Bf/Bi and Cf/Ci (1978-2003-2023)
5 Simulations
- If nothing changes…
- If global fishing effort decreases 20% (≈ one fishing day)
- If demersal fishery or purse seine fishing effort decreases 20%
- How to recover high levels of hake, anchovy and sardine
3. Dynamic simulations
Simulation 1: If nothing changes….
10
17
6.7
8
2
3.4
20
18
1
14
1
0
1978
* Low biomasses of adult hake, sardine
* Decreasing biomass of juv. hake and
several demersal species
* High biomasses of benthopelagic
fishes, conger eel, other small pelagics
(mainly round sardinella), jellyfish,
shrimps and horse mackerel
* Anchovy shows a recovery trend
* Biomasses are maintained and
catches don’t increase
2023
2003
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Functional groups
Jellyfish
Shrimps
Crabs
Norway lobster
Benthic cephalopods
Benthopelagic cephalopods
Mullets
Conger eel
Anglerfish
Flatfishes
Juvenile hake
Adult hake
Demersal fishes (1)
Demersal fishes (2)
Demersal fishes (3)
Demersal sharks
Benthopelagic fishes
European anchovy
European pilchard
Other small pelagic fishes
Horse mackerel
Mackerel
Total
B2023/B1978
1.62
1.35
0.66
0.56
0.86
0.82
0.94
3.25
0.68
0.85
0.77
0.09
0.93
1.68
0.63
0.00
8.61
1.49
0.05
2.66
1.11
0.74
1.05
C2023/C1978
2.34
1.16
0.96
1.60
1.48
2.18
7.40
1.18
2.66
1.34
2.02
1.99
3.40
1.17
0.01
14.73
1.14
0.22
4.25
2.44
1.73
0.89
3. Dynamic simulations
Simulation 2: If fishing effort is globally reduced by 20%
10
17
6.7
8
2
3.4
20
18
14
1
1
0
1978
2023
2003
* Some partial recovery on biomass of
demersal and pelagic depleted species
* Still high biomasses of benthopelagic
fishes, conger eel, other small pelagics,
jellyfish
* Increasing catches of anglerfish,
conger eel, demersal fishes, sardine,
horse mackerel
* Global biomass maintained, non clear
recovery of global catches
Functional groups
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Jellyfish
Shrimps
Crabs
Norway lobster
Benthic ceph.
Benthop. ceph.
Mullets
Conger eel
Anglerfish
Flatfishes
Juvenile hake
Adult hake
Demersal fishes (1)
Demersal fishes (2)
Demersal fishes (3)
Demersal sharks
Benthopelagic fishes
European anchovy
European pilchard
Other small pel. fishes
Horse mackerel
Mackerel
Total
B2023/B1978
If nothing
changes…
1.62
1.01
1.30
0.76
1.05
0.80
0.96
4.17
1.09
0.96
0.80
0.10
0.97
1.68
1.19
0.03
8.38
1.47
0.07
2.41
1.08
0.74
1.05
►
▼
▲
▲
▲
▼
▲
▲
▲
▲
▲
▲
▲
▲
▲
▲
▼
▼
▲
▼
▼
▲
►
C2023/C1978
If nothing
changes…
1.79
1.08
0.63
1.57
1.16
1.78
7.59
1.50
2.40
1.11
1.67
1.66
2.73
1.77
0.06
11.45
0.90
0.23
3.09
1.91
1.39
0.76
▼
▼
▼
▼
▼
▼
▲
▲
▼
▼
▼
▼
▼
▲
▲
▼
▼
▲
▼
▲
▼
▼
3. Dynamic simulations
Simulation 4: If fishing effort is
reduced by 20% for the demersal
fishery
Simulation 3: If fishing effort is
reduced by 20% for purse seine
Functional groups
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Jellyfish
Shrimps
Crabs
Norway lobster
Benthic cephalopods
Benthop. ceph.
Mullets
Conger eel
Anglerfish
Flatfishes
Juvenile hake
Adult hake
Demersal fishes (1)
Demersal fishes (2)
Demersal fishes (3)
Demersal sharks
Benthopelagic fishes
European anchovy
European pilchard
Other small pel. fishes
Horse mackerel
Mackerel
Total
B2023/B1978
If nothing
changes…
1.61
1.02
1.35
0.66
0.86
0.84
0.93
3.25
0.68
0.85
0.77
0.09
0.93
1.67
0.70
0.004
8.55
1.51
0.07
2.69
1.10
0.74
1.05
▼
▼
▲
▲
►
▲
▼
►
►
►
►
►
►
▼
▲
►
▼
▲
▲
▲
▼
►
►
C2023/C1978
If nothing
changes…
2.33
1.16
0.96
1.60
1.52
2.17
7.38
1.18
2.66
1.33
2.01
1.98
3.37
1.30
0.01
14.24
0.92
0.24
3.54
2.28
1.61
0.83
▼
►
►
►
▲
▼
▼
►
►
▼
▼
▼
▼
▲
►
▼
▼
▲
▼
▼
▼
▼
Functional groups
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Jellyfish
Shrimps
Crabs
Norway lobster
Benthic ceph.
Benthop. ceph.
Mullets
Conger eel
Anglerfish
Flatfishes
Juvenile hake
Adult hake
Demersal fishes (1)
Demersal fishes (2)
Demersal fishes (3)
Demersal sharks
Benthopelagic fishes
European anchovy
European pilchard
Other small pel. fishes
Horse mackerel
Mackerel
Total
B2023/B1978
If nothing
changes…
1.63
1.01
1.30
0.76
1.06
0.78
0.96
4.18
1.08
0.96
0.81
0.10
0.98
1.69
1.09
0.029
8.43
1.45
0.05
2.41
1.09
0.75
1.04
▲
▼
▲
▲
▲
▼
▲
▲
▲
▲
▲
▲
▲
▲
▲
▲
▼
▼
►
▼
▼
▲
▼
C2023/C1978
If nothing
changes…
1.79
1.08
0.63
1.58
1.13
1.78
7.60
1.50
2.40
1.11
1.67
1.67
2.75
1.62
0.06
11.87
1.11
0.22
3.77
2.07
1.51
0.82
▼
▼
▼
▼
▼
▼
▲
▲
▼
▼
▼
▼
▼
▲
▲
▼
▼
►
▼
▼
▼
▼
* Some recovery on biomass of pelagic
depleted species
* General recovery on biomass of
demersal depleted species
* Still high biomasses of benthopelagic
fishes, conger eel, other small pelagics,
jellyfish
* Still high biomasses of benthopelagic
fishes, conger eel, other small pelagics,
jellyfish
* Increasing catches of sardine and some
demersal fishes
* Increasing catches of some demersal
fishes
3. Dynamic simulations
Simulation 5: how to recover high levels of hake, anchovy and sardine
If we reduce the fishing rate of adult
hake to F/Z <0.81; eliminating
fishing on juv. hake < 25cm
(immature ones) to 80%, reducing
F/Z
for
sardine
<0.52
and
maintaining F/Z for anchovy <0.5
8
5.3
2
17
2.7
11
19
20
12
1
* Recovery of biomasses of adult
0
1978
2003
2023
and juv. hake, sardine and other
benthic and pelagic species
Functional groups
* Lower levels for anchovy
comparing 1978 but higher ones
respect 2003 (25%)
* Lower biomasses of
benthopelagic fishes, jellyfish,
conger and other pelagic fishes
* Higher levels of caches for target
demersal and pelagic species
* Global increase of biomasses
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Jellyfish
Shrimps
Crabs
Norway lobster
Benthic cephalopods
Benthop. cephal.
Mullets
Conger eel
Anglerfish
Flatfishes
Juvenile hake
Adult hake
Demersal fishes (1)
Demersal fishes (2)
Demersal fishes (3)
Demersal sharks
Benthopelagic fishes
European anchovy
European pilchard
Other small pelagic fishes
Horse mackerel
Mackerel
Total
B2023/B1978
1.00
1.11
1.41
0.84
0.94
1.35
1.01
0.10
0.47
0.73
1.40
1.15
0.91
1.08
0.65
0.001
2.37
0.90
1.65
1.38
0.82
0.88
1.10
In nothing
changes…
▼▼
▼
▲
▲
▲
▲
▲
▼▼
▼
▼
▲▲
▲▲
▼
▼
▲
▲
▼▼
▼
▲▲
▼
▼
▲
▲
B2023/B2003
C2023/C1978
In nothing
changes…
C2023/C2003
0.78
1.41
1.47
1.67
1.57
2.89
4.14
0.07
0.62
1.27
2.24
15.13
1.53
0.73
0.54
0.01
0.55
1.25
8.67
1.58
1.28
1.61
1.41
2.43
1.49
1.35
1.76
2.43
2.35
0.24
0.82
2.28
0.48
2.01
1.94
2.20
1.20
0.002
4.04
0.69
1.31
2.20
1.81
2.06
1.24
▲
▲
▲
▲
▲
▲
▼
▼
▼
▼
▲
▼
▼
▲
▼
▼
▼
▲
▼
▼
▲
▲
1.69
1.47
1.67
1.57
2.89
4.14
0.07
0.62
1.27
0.44
1.22
1.53
0.73
0.54
0.01
0.55
1.25
1.65
1.58
1.28
1.61
1.60
and catches
1. As recommended for ground fished stocks: Mertz and Myersl, 1998. Can. J. Fish. Aquat. Sci., 55: 478-484;
2. As recommended for small pelagic fishes: Patterson. 1992. Rev. Fish Biol. Fish, 2: 321-338.
Conclusions
Simulation examples are showing interesting results:
* Target species are driven by fishing activity and we need to lower fishing
impact to recover them, probably preventing as well the proliferation of other
species (jellyfishes and benthopelagic fishes: trophic cascades)
* A reduction of 20% of effort would imply some improvement of the ecosystem
respect the actual state
* To recover the system we need an intervention in both pelagic and demersal
fisheries to increase top predator biomasses and relax the impact on target
small pelagic fishes, while increasing the predation on preys of top predators
and competitors of low TL target species
Ecological modeling in the Mediterranean context is shown as an appropriated
tool to investigate fishing management options
To answer important ecological questions, to pose new ones and to assess the
ecosystem effects of fishing
This is especially relevant in the Mediterranean because it take into account the
multispecific nature of ecosystem and fisheries: essential under the EAF
Conclusions
The ecosystem modeling approach is a tool to improve the strategic nature of
the management: where we are, where we are going?
Complementing the tactical management from stock assessment and
evaluation tools
This can contribute to evolve the reactive management of fishing resources
into a more adaptive and strategic one, in line with recommendations of GFCM
Ecological modeling is nourished by conventional assessment methods,
information that we already have and we organize into an ecosystem context
We need:
to continue collecting this essential information
to increase it: some critical gaps (diet of key specie, ontogeny)
to collect new data to monitor model predictions (validate or refuse)
Are benthopelagic fishes increasing in Mediterranean exploited ecosystems?
Models are always under construction in the sense that when new data or new
ideas are available, they can be improved
Conclusions
EwE Ecological modeling shows an essential improvement with the ability to fit
models to data:
* From calibrated models we can derive ecosystem indicators like L index
(presented by S. Libralato and collaborators1)
* They can be used to derive classical indicators as fishing mortalities (F),
predator mortalities (M2), maximum sustainable catches (MSY) from an
ecosystem context
* They can also include socioeconomic data to assess the optimum equilibrium
of different fishing options taking into account social, economic and ecological
criteria (example in Venice lagoon 2)
* They are the baseline from where to develop spatial simulations3
We suggest to the Sub-Committee of Stock Assessment:
To foment the ecosystem modeling application in the Mediterranean
by implementing EwE and other tools
by implementing them to different scales
We are also working in the Adriatic Sea (1970s to 2000s): This will enable us to
have another example to compare observed patterns and different model scales
1. Libralato et al., Submitted to Journal of Applied Ecology; 2. Granzotto et al., 2004. Chemistry and Ecology, 20(1): 435-449;
3. Walters et al., 1999. Ecosystems, 2: 539-554.
THANKS!