modelling of aquaculture impact and carrying capacity in the philippines using tropomod
DESCRIPTION
Modelling of the SABBAC zones has the following objectives: • to test scenarios which encourage careful feeding, so waste feed and nutrient input to the environment is minimised; farmers will also save money • to encourage use of better quality feed, where better digestibility of feed means less feed is needed; better quality feed also breaks up less, so more goes to growth The modelling approach also aims to: • maintain enough spacing between cage rows so that remediation of sediments can take place – impact should be low between rows in each zone • maintain enough space between cage rows to prevent reduction of currents by high aggregation of cages – although not predicted by TROPOMOD, this effect is known to exist and has been shown by MSI models • prevent overlap of zones by predicting the extent of the zones and recommending minimum spacing between zones The TROPOMOD model was therefore set up to evaluate the following: • How severe is the impact – what is the maximum impact underneath cages? • How far to the boundary of the impact? • How can husbandry practices be optimised to use the zone most productively?TRANSCRIPT
© www.akvaplan.niva.no 1
TROPOMOD modelling of six SABBAC aquaculture zones
PHILMINAQ projectChris CromeyPatrick White
Cesar Villanoy, Evangeline MandongAll of the PHILMINAQ team made
contributions to this modelling effort
© www.akvaplan.niva.no 2
Particles settling under a fish cage
result in a ‘footprint’ of deposition of solids on the sea bed
© www.akvaplan.niva.no 3
Currents move in different speeds and direction at different depths. Faeces settle more slowly and so are transported further away from cages
0 C u rre n t V e lo c ity
S ou rce
C o a rse M ed iu m
F in e
Waste feed particles settle quickly
Milkfish waste faeces settle very slowly
© www.akvaplan.niva.no 4
Benthic community
Severe impact
(no animals)
High impact (some effect)
Moderate impact
Contour map of waste flux
(grams waste feed and faeces m-2 d-1)
75
15
1
grams solids m bed d-2 -1
1
1 5
7 5
© www.akvaplan.niva.no 5
PHILMINAQ project modelling approach with TROPOMOD
3 important aspects:
1. How severe is the impact – what is the maximum impact underneath cages?
2. How far to the boundary of the impact? (Scotland = Allowable Zone of Effect)
3. How can husbandry practices be optimised to use the zone most productively?
Objectives
Predict if impact is SEVERE underneath cages
as shown by this deposition footprint
Zone colour
Predict distance to boundary of MODERATE impact
Zone colour
Edge of zone
© www.akvaplan.niva.no 6
PHILMINAQ project modelling approach with TROPOMOD
Maintain enough spacing between cage rows so that remediation of sediments can take place – impact should be LOW between rows in each zone
Maintain enough space between cage rows to prevent reduction of currents by high aggregation of cages
Zone colour
Not predicted by TROPOMOD, but this effect is known to exist and has been shown by MSI models
© www.akvaplan.niva.no 7
PHILMINAQ project modelling approach with TROPOMOD
Encourage careful feeding, so that there is less waste feed and less wastage of money
Encourage better quality feed: Feed digestibility is increased Less feed is needed Better quality feed also breaks up less, so more goes to growth
Test these scenarios
Prevent overlap of zones by predicting the extent of the zones – e.g. 600 m between zones for this site
0 200 400 600 800Scale (m )
© www.akvaplan.niva.no 8
Method - common model input data between scenarios
Model input data Value
Zone size 600 m by 200 m (12 ha)
Current (modelled by MSI) Modelled
Feed settling rate for different scenarios
FCR 2.8 – pellet break up (estimated)
FCR 2.0 – intact pellets (measured by PHILMINAQ)
8.9 cm/s (5%), 4.5 cm/s (65%), 1.6 cm/s (30 %)
8.9 cm/s (100%)
Faeces settling rate – measured by PHILMINAQ for Milkfish
0.84 (cm/s)
© www.akvaplan.niva.no 9
Model input data Scenario 1
Poor feeding
Low digest.
Scenario 3
Careful feeding
Better digest.
Feed wasted 27% 10%
Feed digestibility 49% 56%
FCR 2.8:1 2.0:1
Feed input Between 0 and 323 kg/cage/d
Between 0 and 231 kg/cage/d
Scenarios tested
A range of feed inputs are used between 0 and maximum to simulate fish at different stages of growing cycle in each zone
Therefore, these are not worse case scenarios
Method – what is varied between scenarios?
Data source: EMMA and PHILMINAQ projects
© www.akvaplan.niva.no 10
Days Size Number Biomass(kg)
Feed rate (%/day)
Feed/day(kg)
April 0 0 0 0 0 0
May 25 20 27247 545 8.5 46
June 30 41 26873 1091 8.9 97
July 31 91 26498 2406 7.1 171
August 31 162 26124 4224 5.1 214
September 30 247 25749 6358 4.4 278
October 31 386 25375 9799 3.3 323
November 19 433 25000 10825 1.8 193
Square cages – 12m * 12m * 8m
Zones 1 – 3, 5 and 6: feed ration used in the simulations – each cage randomly allocated a feed ration to simulate fish at different stages of cycle
FCR = 2.8:1
Data source: EMMA project
© www.akvaplan.niva.no 11
Days Size Number Biomass(kg)
Feed rate (%/day)
Feed/day(kg)
April 0 0 0 0 0 0
May 25 20 59444 1189 8.5 101
June 30 41 58628 2404 8.9 214
July 31 91 57810 5261 7.1 374
August 31 162 56994 9233 5.1 471
September 30 247 56176 13875 4.4 611
October 31 386 55360 21369 3.3 705
November 19 433 54542 23616 1.8 425
Large circular cages – 20m diameter * 8m means 2.18 times more
biomass can be contained in cages if stocking density is maintained
Zones 4: feed ration used in the simulations – each cage randomly allocated a feed ration to simulate fish at different stages of cycle
FCR = 2.8:1
Data source: EMMA project data on previous slide scaled by 2.18
© www.akvaplan.niva.no 12
Current used in TROPOMOD were taken from the MSI hydrodynamic model of the area (left) – 30 days of current were used
Zones 1 to 6 are shown in the map
Aquazones identified by the hydrodynamic model
© www.akvaplan.niva.no 13
-35
-25
-15
-5
5
15
25
35
-35 -25 -15 -5 5 15 25 35
Zone 1
-35
-25
-15
-5
5
15
25
35
-35 -25 -15 -5 5 15 25 35
Zone 2
-35
-25
-15
-5
5
15
25
35
-35 -25 -15 -5 5 15 25 35
Zone 3
-35
-25
-15
-5
5
15
25
35
-35 -25 -15 -5 5 15 25 35
Zone 4
-35
-25
-15
-5
5
15
25
35
-35 -25 -15 -5 5 15 25 35
Zone 5
-35
-25
-15
-5
5
15
25
35
-35 -25 -15 -5 5 15 25 35
u (cm/s)v
(cm
/s)
Zone 6
Current velocity distribution
e.g. At zone 1, current is flowing east and west in the Bolinao Narrows
For each site, lines of cages were orientated parallel with the main axis of current e.g. cages in zone 4 were aligned NE-SW
North
© www.akvaplan.niva.no 14
Zone Depth (m) Depth rank
Mean speed (cm/s)
Max speed (cm/s)
Current rank
1 20 1 10.8 28.6 3
2 14 2 11.5 30.6 2
3 13 3 8.9 23.7 4
4 10 4 7.9 24.8 5*
5 10 5 6.0 18.8 6
6 9 6 13.3 39.9 1
Modelled current at the zones from MSI Bolinao model showing a rank of the most dispersive and deepest
This table shows tidal currents – site 4 is exposed to waves from the east, which will increase dispersion
© www.akvaplan.niva.no 15
Definitions Zone colour Predicted flux (g m-2 d-1)
Impact areas:
Low/None
Moderate
High
Severe
<1
1 – 15
15 – 75
75 +
% of zone area HIGH and SEVERE impact
>15
Is more than 1 % of zone SEVERE impact? Yes or No?
> 75
Distance to boundary of zone of effect - 1 g m-2 d-1 contour
1
Measured sediment flux at Bolinao during workshops by PHILMINAQ project - 114.0 g m-2 d-1 (0 m) and 148.7 (25 m) - Both stations devoid of fauna
Zone colouring
© www.akvaplan.niva.no 16
FCR 2.8:1
Less feed wasted, higher digestibilityFCR 2.0:1
Aquaculture Zone 1 – Bolinao Narrows
Impact
Severe
High
Moderate
1
15
75
0 200 400 600 800Scale (m )
N
M odel waste flux - gram s m d-2 -1
© www.akvaplan.niva.no 17
FCR 2.8:1
Less feed wasted, higher digestibilityFCR 2.0:1
Aquaculture Zone 2 – ??
Impact
Severe
High
Moderate1
15
75
0 200 400 600 800Scale (m )
M odel w aste flux - gram s m d-2 -1
N
© www.akvaplan.niva.no 18
1
15
75
0 200 400 600 800Scale (m )
M odel waste flux - gram s m d-2 -1
NFCR 2.8:1
Less feed wasted, higher digestibilityFCR 2.0:1
Aquaculture Zone 3 – ??
Impact
Severe
High
Moderate
© www.akvaplan.niva.no 19
FCR 2.8:1 Less feed wasted, higher digestibilityFCR 2.0:1
Aquaculture Zone 4 – 2 rows of 12 large cages
Impact
Severe
High
Moderate
1
15
75
0 200 400 600 800Scale (m )
M odel waste flux - gram s m d-2 -1
N
Zone 4 is exposed to waves from the east, so the model will overpredict impact
© www.akvaplan.niva.no 20
FCR 2.8:1 Less feed wasted, higher digestibilityFCR 2.0:1
Aquaculture Zone 5 – ??
Impact
Severe
High
Moderate
1
15
75
0 200 400 600 800Scale (m )
M odel waste flux - gram s m d-2 -1
N
© www.akvaplan.niva.no 21
FCR 2.8:1 Less feed wasted, higher digestibilityFCR 2.0:1
Aquaculture Zone 6 – Pens
Impact
Severe
High
Moderate1
15
75
0 200 400 600 800Scale (m )
M odel waste flux - gram s m d-2 -1
N
© www.akvaplan.niva.no 22
Prohibited Aquaculture zones
119.88 119.9 119.92 119.94 119.96 119.98 120
Longitude(E)
16.28
16.29
16.3
16.31
16.32
16.33
16.34
16.35
16.36
16.37
16.38
16.39
16.4
Lat
itu
de(
N)
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
Reference Vectors(m/s)0.10.30.60.93
Dep
th (
m)
ANDA
BOLINAO
Guigiwanen
CaquiputanStrait
Hr 61
Siapar
© www.akvaplan.niva.no 23
Avoiding impact overlap between aqua zones
© www.akvaplan.niva.no 24
119.88 119.9 119.92 119.94 119.96 119.98 120
Longitude(E)
16.28
16.29
16.3
16.31
16.32
16.33
16.34
16.35
16.36
16.37
16.38
16.39
16.4
Lat
itu
de(
N)
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
Reference Vectors(m /s)0.10.30.60.93
Dep
th (
m)
ANDA
BOLINAO
Guigiwanen
CaquiputanStrait
Hr 61
Siapar
1
7
6
3
2
4
Recommended cage culture zones
© www.akvaplan.niva.no 25
Recommended cage culture zones
119.88 119.9 119.92 119.94 119.96 119.98 120
Longitude(E)
16.28
16.29
16.3
16.31
16.32
16.33
16.34
16.35
16.36
16.37
16.38
16.39
16.4
Lat
itu
de(
N)
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
Reference Vectors(m/s)0.10.30.60.93
Dep
th (
m)
ANDA
BOLINAO
Guigiwanen
CaquiputanStrait
Hr 61
Siapar
1
7
6
3
2
4
© www.akvaplan.niva.no 26
Zones – what area is impacted
Zone ScenarioArea impacted - % of zone HIGH/SEVERE
SEVERE (% of zone) Rank
1 FCR 2.8 54 6.1 1
2 FCR 2.8 53 6.5 2
3 FCR 2.8 50 9.3 4
4 FCR 2.8 53 20.3 6(Exposed)
5 FCR 2.8 45 13.5 5
6 FCR 2.8 52 7.2 3
1 FCR 2.0 36 0.0
2 FCR 2.0 35 0.0
3 FCR 2.0 36 0.0
4 FCR 2.0 44 5.9 Exposed
5 FCR 2.0 35 0.0
6 FCR 2.0 35 0.0
Zone 4 is exposed and therefore model probably over estimates impact because the model does not take account of waves
© www.akvaplan.niva.no 27
Zones – how severe is the impact
Zone Scenario
Average flux in HIGH/SEVERE zone
(g/m2/d) Rank
1.0 FCR 2.8 43 1
2.0 FCR 2.8 44 2
3.0 FCR 2.8 50 4
4.0 FCR 2.8 73 6 (exposed)
5.0 FCR 2.8 58 5
6.0 FCR 2.8 46 3
1.0 FCR 2.0 24
2.0 FCR 2.0 26
3.0 FCR 2.0 29
4.0 FCR 2.0 45
5.0 FCR 2.0 33
6.0 FCR 2.0 30
Zone 4 is exposed and therefore model probably over estimates impact because the model does not take account of waves
© www.akvaplan.niva.no 28
Zone Cages Spacing between cages Zone biomass modelled
Average situation with all different fish sizes in zone (EMMA data)
Zone biomass if all
fish 386 grams in all cages
Maximum biomass in zone
1, 2, 3, 5, 6
2 rows of 18 20 m between cages
120 m between rows
137 tonnes 353 tonnes A
4 2 rows of 12
30 m between cages
120 m between rows
277 tonnes 514 tonnes B
Zone 4 cages are large circular cages (20 m diameter* 8m deep)
Zones 1,2,3,5,6, are square cages (12m* 12m*8m deep)
A 386 gram fish require highest feed, 9.8 tonnes per cage *36 = 353 tonnes in zone (square cages)
B 386 gram fish require highest feed, 21.4 tonnes per cage *24 = 514 tonnes in zone (large circular cages)
© www.akvaplan.niva.no 29
Scenario – for zone 5 which has the lowest currents, can the zone support a higher biomass if large circular cages are used instead of small square cages?
Large, circular cages versus small, square cages
Scenario Configuration
Area impacted - % of zone
HIGH/SEVERESEVERE
(% of zone)Zone peak biomass
FCR 2.8 2 * 18 square cages 45 13.5 353 tonnes
FCR 2.02 * 18 square cages
(353 tonnes) 35 0.0 353 tonnes
FCR 2.8
FCR 2.0
YES - ? tonnes can be at peak biomass as opposed to 353 tonnes with small cages ( a x % increase)
© www.akvaplan.niva.no 30
Overall rank Zone Rank - magnitude of
SEVERE impact
Rank - extent of SEVERE
impact
Rank - depth Rank - current
1 1 1 1 1 3
2 2 2 2 2 2
3 6 3 3 6 1
4 3 4 4 3 4
5 4 6 6 4 5
6 5 5 5 5 6
However, zone 4 is exposed to waves which are not modelled so zone 4 is likely to be more dispersive in reality
Zones ranked in terms of dispersiveness
© www.akvaplan.niva.no 31
Summary
TROPOMOD model predicted impact with 2 scenarios, FCR of 2.8:1 and 2.0:1.
Results are for different sized fish throughout the zone, which is realistic, not worse case.
Improvement of FCR of 2.8 to 2.0:1 resulted in:
• reduced feed needed by 29 %
• there was no/little SEVERE impact under cages
• HIGH impact areas were less than 50 % of the total zone area
© www.akvaplan.niva.no 32
Recommendations
Zone Cage configuration
Peak biomass
(tonnes)
Spacing
1, 2, 3, 5, 6 2 * 18 square cages
353 20 m between cages, 120 m between rows
4 2 * 12 large circular cages
514 30 m between cages, 120 m between rows
As the deposition footprints extend between 200 and 400 m from the edge of the zone, it is recommended the distance between zones should be a minimum of 600 m.
© www.akvaplan.niva.no 33
TROPOMOD has been developed using a similar concept to the following models:
MERAMOD – Sea bass/bream in the Mediterranean – scientifically tested at several sites with sediment traps and benthic fauna sampling
DEPOMOD – Salmon/cod for North Atlantic – used in the Scottish regulatory system to consent farms in Scotland
DEPOMOD and MERAMOD have 100+ users worldwide
Tests with TROPOMOD at Bolinao were satisfactory – the model predicted on average 104 g m-2 d-1, which corresponded to an average value measured in traps of 131 g m-2 d-1 in May 2007.
ECASA toolbox - a tool box of environmental impact and socioeconomic indicators, including descriptions of models and where to find them
www.ecasatoolbox.org.uk