Safe and Nutritious Aquaculture produce: Effects of alternative feeds replacing fishmeal and fish oil (FM & FO) with plant-based alternatives in Atlantic salmon (Salmo salar) Prof. Gordon Bell Institute of Aquaculture, University of Stirling, Stirling FK9 4LA, Scotland, UK.

Overview • Replacement of fish oil and fish meal-

consequences for growth & product quality. • How much Omega 3 fatty acids do we need to keep us healthy or how much farmed fish should we eat to provide optimal fatty acid nutrition? • Effect of alternative feeds on organic contaminants. • New dietary raw materials

The historical perspective: The Scientific and Practical Basis for using Fish Meal and Fish Oil in aquafeeds • Fish Meal is rich in essential amino acids, is readily •

•

accepted and digested by fish and, until recently, has been readily available and relatively cheap. Fish Oil meets the requirements of fish for n-3 PUFA; freshwater fish and salmonids can convert 18:3n-3 to 20:5n-3 and 22:6n-3 but thrive on 20:5n-3 and 22:6n-3; marine fish have an absolute requirement for 20:5n-3 and 22:6n-3. Fish eat other fish in nature; fish farming is a means of converting low value fish meal and oil to high value food for humans.

FO & FM replacement-the current position • 100% of FO can be replaced with a vegetable

•

•

oil (VO) mix without compromising overall performance – When fish meal is used as the protein source (RAFOA/Torstensen et al 2005, J. Agric Food Chem.). www.rafoa.stir.ac.uk 90% of FM can be replaced by a plant protein mix without any negative effects on growth – when FO is used as the oil source (Espe et al, 2006;2007) What happens when FM & FO are replaced at maximum levels?

Why do we need new alternative aquafeeds? • Fish oil and meal production from feed grade fisheries is static or even • • • • • •

in decline Since 1985, world FM production has fluctuated between 5.2 and 7.5m tonnes (average 6.5 mt). FO production has varied between 0.8 and 1.6 m tonnes (ave. 1 mt). In 2006, aquaculture used 88.5% of global fish oil production and ~68% of fish meal production (Tacon & Metian, 2008). Of this, salmonid culture uses 54 and 23% of FO and FM production, respectively (Jackson, IFFO, 2007). At current usage, by 2012 at least 88% of the fish oil will still be consumed in aquafeeds to counter excessive exploitation pressure and thereby ensure proper conservation of marine resources Current & future EU restrictions on dioxins/PCBs in animal feeds could limit FO use in aquafeeds.

Should we feed fish oil over the whole production cycle? • Feeding high levels of FO results in poorer retention • • •

efficiency of EPA & DHA. This results in EPA and DHA being deposited in lipid stores, especially in visceral fat, where the n-3 HUFA are used for energy production via b-oxidation. Feeding lower levels of FO, in conjunction with VOs, gives improved n-3 HUFA retention efficiencies. It is therefore wasteful to over-supply with FO as valuable EPA & DHA is either lost in the viscera, at processing or used for energy production which could be reduced in diets containing VO rich in monounsaturated fatty acids and PUFA.

Aquamax trial 1: (Lean/Fat Trial)

• Two groups of Atlantic salmon smolts were selected as “fat” or

• • • • • •

“lean”, based on muscle adiposity, by Landcatch Natural Selection (LNS). A third non-pedigreed commercial strain was obtained from Marine Harvest that was similar to those used in the EU RAFOA project. 2000 fish of each strain were transferred to the Marine Harvest FTU at Ardnish, Scotland and the feeding trial commenced in May 2006. 500 fish were stocked into each of 12 x 125m3 net pens. Duplicate pens of each of the 3 strains were fed diets containing either 100% Northern fish oil (NFO) or vegetable oil (VO) blend comprising rapeseed, palm and Camelina oils). Both diets contained 25% fish meal and ~40% plant meals. Fish were fed 3, 6 and 9 mm diets from smolt to harvest at ~3kg. After harvest fish were combined into 3 pens (one pen per strain), the VO fish marked by fin clipping, and all fish fed a diet containing a decontaminated northern fish oil finishing diet for 24 weeks.

Diet formulations (g/100g) Component

6mm FO

6mm VO

9mm FO

9mm VO

LT fish meal (South American) Soya (extracted) Wheat Wheat gluten Soya concentrate Maize gluten Fish oil (Nordic) Rapeseed oil Palm oil Camelina oil Premix Astaxanthin (10%)

25.00

25.00

25.00

25.00

10.00 7.70 15.00 6.56 5.72 27.54 2.52 0.05

10.00 7.70 15.00 6.56 5.00 14.04 8.00 5.50 2.52 0.05

11.00 7.88 9.00 7.00 5.00 30.00 5.12 0.05

11.00 7.88 9.00 7.00 5.00 15.00 9.00 6.00 5.12 0.05

Total plant meals

44.98

39.88

39.88

44.26

Fatty acid compositions of 9 mm diets including finishing diet (weight %) Fatty acid 16:0 Total saturates 18:1n-9 20:1n-9 22:1n-11 Total monoenes 18:2n-6 20:4n-6 Total n-6 18:3n-3 20:5n-3 22:6n-3 Total n-3

FO 15.0 26.1 10.1 8.5 13.7 42.6 3.5 0.4 4.6 1.6 8.5 10.1 25.3

VO 15.9 21.7 37.6 4.0 1.6 47.5 16.5 0.1 17.1 10.0 1.2 1.5 13.5

Finishing diet 17.1 31.1 11.9 4.3 6.3 34.5 3.3 0.8 5.0 0.9 12.1 8.9 26.3

Family phenotypes (45), flesh lipid (g/100g), n-3 HUFA (% TFA & absolute (mg/100g flesh) in fish fed very low EPA + DHA levels (2.7% of TFA). Family lipid% 30 99 8 27 47 59 16 97 24 188 44 158 21 75 43 58 15 102 12 121 3 79

n-3 HUFA % TFA mg FA/100g

Family

n-3 HUFA lipid% %TFA

mgFA/100g

3.5 ± 0.5 15.6 ± 2.9

451 ± 86

1

5.1 ± 0.7 13.7 ± 1.3

591 ±

3.6 ± 0.3 15.0 ± 0.6

442 ± 48

17

5.1 ± 0.1 14.1 ± 1.2

663 ±

3.8 ± 0.9 13.8 ± 1.6

428 ± 43

9

5.3 ± 0.9 12.6 ± 1.4

610 ±

4.0 ± 0.9 15.6 ± 1.4

572 ± 71

20

5.4 ± 1.1 13.1 ± 1.5

637 ±

4.1 ± 0.4 14.3 ± 1.1

526 ± 74

7

5.4 ± 1.5 12.9 ± 1.3

611 ±

4.3 ± 0.3 13.3 ± 0.8

481 ± 16

45

5.4 ± 1.4 13.8 ± 1.6

627 ±

4.3 ± 0.3 14.6 ± 0.3

572 ± 49

10

5.4 ± 0.6 13.8 ± 0.9

659 ±

4.3 ± 0.1 15.2 ± 1.1

560 ± 38

26

5.5 ± 0.5 11.7 ± 1.0

564 ±

4.4 ± 0.0 13.0 ± 0.4

521 ± 28

29

5.5 ± 1.0 12.9 ± 1.9

613 ±

4.5 ± 0.6 14.2 ± 1.4

516 ± 15

48

5.5 ± 1.2 11.3 ± 0.9

557 ±

4.5 ± 0.5 13.3 ± 1.0

529 ± 28

14

5.5 ± 0.3 11.9 ± 1.5

588 ±

Relationship between lipid content and n-3 LC-PUFA content of flesh

16

650 600

14 550 12

500 450

10 400

y = 52.728x + 301.08 2

R = 0.4249

.

absolute relative

y = -1.4131x + 20.23

.

2

R = 0.6179

350

8 3

4

5

Lipid content (Percentage of flesh wet weight)

6

7 .

n -3H U F A con ten t (p ercen tage of total fatty

n -3H U F A con ten t (m g/100g fles h )

700

Summary-Lean/Fat trial • SGRs were significantly higher in the MH strain compared to the

• • • • •

two LNS strains and FCRs (range 1.1-1.4) were highest in the Lean strain compared to the other 2 treatments. Diet did not affect SGR or FCR. Flesh lipid content was in order Fat > MH > Lean in fish fed FO but was Lean = Fat > MH in fish fed VO. Flesh DHA levels, and those of other LC-PUFA, were reduced in fish fed 100% VO by 65-75% but could be restored using a FO finishing diet. Flesh dioxin and PCB levels were reduced in fish fed VO (~0.5 in VO & 2 ng TEQ/kg in FO) and generally low compared to the EU permitted maximum of 8 ng TEQ/kg. Different salmon strains respond differently to high VO diets in terms of growth, tissue lipid distribution, lipid concentration and retention and deposition of fatty acids. Different salmon phenotypes have very different retention values for n-3 HUFA when fed n-3 depleted diets-the heritability of the n-3 retention trait is high and could be exploited by selective breeding!

Aquamax trial 2: NIFES, Norway Novel salmon protein production using combined replacement of FM & FO with plant meals & oils

Atlantic salmon fed diets with different plant ingredient inclusion levels Finishing diet period 12 months

Sept 06

June 2006

n=3 100 % Fish oil 100% Fish prot.

Feb 07

n=3 35 % Veg oil 80 % Plant prot.

June 2007

n=3

n=3

70 % Veg oil 40 % Plant prot.

70 % Veg oil 80 % Plant prot.

Plant protein: Wheatgluten, corn gluten, soya concentrate, krill meal Vegetable oil: Linseed oil, palm oil, rapeseed oil

Estimated safe max replacement to meet all nutrient requirements

Crystalline Lys, Met and His was added to meet requirement levels

(NRC, 1993)

Lower final weight in the two high plant protein groups June 2007

June 2006

3.9 kg

0.35 kg 6

Final weight (kg)

5

a

a b

b

4

After 12 months, lower final fish weight in both fish groups fed 80% plant protein, with up to 17 % lower final weights

3

2

Lower growth and feed intake during the first 3 months in 80PP70VO group and no compensatory growth

1

0 FMFO

80PP35VO

40PP70VO

80PP70VO

lower growth in both plant protein groups during the final period

Torstensen et al 2008, Aquaculture, 285, 193-200

Producing fish protein from plant protein? Amount of dietary protein used to produce 1 kg Atlantic salmon protein 2,5

kg protein

2

1 kg high quality fish protein obtained from plant protein and 0.5 kg fish meal proteinSalmon can be a net producer of marine protein!

1,5

1

0,5

0 FMFO

80PP35VO Fish meal protein

40PP70VO

80PP70VO

Plant protein

Torstensen et al 2008, Aquaculture, 285, 193-200.

Summary-NIFES • Lower final weights were seen in fish fed the highest plant • •

•

protein (PP) levels (80PP/35VO & 80PP/70VO) probably due to lower feed intake in first 3 months and final growth phase. However, the two high PP diets produced 1kg of marine protein from 0.5kg of FM and PP blend so were net marine protein producers. Lipid storage was altered in viscera and liver with highest deposition in 80PP/70VO diet. There were no differences in b-oxidation capacity but cholesterol was reduced in high VO groups. Plasma TAG was correlated with increased visceral adiposity in the high VO groups.

Reduction of organic contaminants

Removal of contaminants from fish oils • Triplicate groups of Atlantic salmon (initial weight 0.78 ± •

• • •

0.01 kg) were reared in 125m3 net pens for 11 weeks All fish were fed the same basal diet coated with either: 1. Northern Hemisphere fish oil – high POPs levels (NFO) 2. same oil but stripped of contaminants using a two-step decontamination process (dNFO) or 3. a blend of Southern hemisphere fish oil, rapeseed and soya oils (SFVO) in ratio 40:30:30 by weight Dioxin/furan, DL-PCBs and PBDE levels were measured in the diets, and flesh from initial and final sampling. Tissue fatty acid compositions were also measured in fish at final sampling. Sprague, M., Bendiksen, E.Å., Dick, J.R., Strachan, F., Pratoomyot, J., Berntssen, M.H.G., Tocher, D.R., Bell, J.G. 2010. Effects of decontaminated fish oil or a fish and vegetable oil blend on persistent organic pollutant and fatty acid compositions in diet and flesh of Atlantic salmon (Salmo salar). British Journal of Nutrition 103, 1442-1451.

Dietary POPs Levels 20

n g T E Q /k g (d io x in /P C B s ) n g /g (P B D E s )

NFO dNFO 15

SFVO

10

EU permitted max 7ng TEQ/kg (Dioxin+ DL-PCBs)

67%

5

97%

95%

97%

0 D io x in s

D L -P C B s

D io x in s + P C B s

C o n ta m in a n t

PBDEs

Flesh POPs Levels EU maximum value (Dioxins + DL-PCBs) n g T E Q /k g (d io x in /P C B s ) n g /g (P B D E s )

8

In itia l NFO

6

dNFO SFVO

4

2

57%;34 % 96%

48%

3%; 64%

93%

73%; 89%

0 D io x in s

D L -P C B s

D io x in s + P C B s

C o n ta m in a n t

PBDEs

Flesh Fatty Acid Composition (%) NFO

dNFO

SFVO

16:0

17.1 ± 0.6a

16.8 ± 0.3a

12.2 ±0.4b

Total Saturated

25.3 ± 0.8a

25.1 ± 0.5a

19.0 ± 0.6b

18:1n-9

25.1 ± 0.6a

21.9 ± 0.3b

28.9 ± 0.4c

Total monoenes

40.8 ± 0.8a

39.3 ± 0.6b

38.1 ± 0.3c

18:2n-6

5.3 ± 0.2a

5.8 ± 1.0a

20.9 ± 0.6b

20:4n-6

0.6 ± 0.0a

0.6 ± 0.0a

0.5 ± 0.0b

Total n-6 PUFA

7.3 ± 0.3a

7.7 ± 1.0a

23.0 ± 0.6b

13:3n-3

2.3 ± 0.1a

2.3 ± 0.1a

4.5 ± 0.2b

18:4n-3

1.5 ± 0.1a

1.8 ± 0.1b

0.8 ± 0.0c

20:4n-3

1.2 ± 0.1a

1.1 ± 0.1b

0.7 ± 0.0c

20:5n-3

5.9 ± 0.4a

6.9 ± 0.3b

4.9 ± 0.1c

22:5n-3

2.1 ± 0.2a

2.0 ± 0.1a

1.8 ± 0.1b

22:6n-3

13.1 ± 0.6a

13.6 ± 0.6a

6.7 ± 0.2b

Total n-3 PUFA

26.6 ± 1.2a

28.0 ± 1.0b

20.0 ± 0.3c

Total PUFA

34.0 ± 1.4a

35.7 ± 0.7b

43.0 ± 0.7c

DHA-rich algal meal (Schizocytrium/C. conhii) as a replacement for FO: effects on feed & flesh POPs and fatty acids

Flesh fatty acid compositions Initial

NHFO

SHFO

11% AM

5% AM

16:0

13.2 ± 0.41a

13.47 ± 0.16a

14.62 ± 0.55b

10.71 ± 0.47c

10.78 ± 0.38c

Total Saturated

20.21 ± .064a

20.88 ± 0.41a

22.72 ± 0.92b

16.44 ± 0.85c

16.54 ± 0.53c

18:1n-9

32.75 ± 0.61a

20.87 ± 0.55b

19.68 ± 0.63c

38.00 ± 0.37d

38.09 ± 0.51d

Total monoenes

45.03 ± 0.91a

42.92 ± 0.57b

39.66 ± 0.35c

47.34 ± 0.46d

49.08 ± 0.51e

18:2n-6

11.41 ± 0.22a

7.67 ± 0.22b

7.18 ± 0.29b

13.36 ± 0.34c

13.31 ± 0.24c

20:4n-6

0.51 ± 0.03a

0.57 ± 0.02a

0.64 ± 0.04b

0.57 ± 0.03a

0.47 ± 0.02c

Total n-6 PUFA

13.24 ± 0.22

9.56 ± 0.25

9.09 ± 0.35

16.74 ± 0.44

15.99 ± 0.24

18:3n-3

3.78 ± 0.15

2.10 ± 0.11

1.84 ± 0.14

4.61 ± 0.18

4.54 ± 0.10

18:4n-3

0.82 ± 0.03

1.58 ± 0.05

1.60 ± 0.04

0.46 ± 0.02

0.52 ± 0.04

20:4n-3

0.83 ± 0.06

1.50 ± 0.03

1.32 ± 0.01

0.55 ± 0.02

0.60 ± 0.03

20:5n-3 20:5n-3

4.94 ± 0.33

5.85 ± 0.19

8.14 ± 0.25

2.62 ± 0.17

2.89 ± 0.19

22:5n-3

2.10 ± 0.09

2.78 ± 0.05

3.29 ± 0.06

1.26 ± 0.06

1.37 ± 0.09

22:6n-3 22:6n-3

7.57 ± 0.64

11.65 ± 0.32

10.35 ± 0.32

8.92 ± 0.53

7.38 ± 0.28

Total n-3 PUFA

20.36 ± 1.22

25.68 ± 0.48

26.72 ± 0.63

18.88 ± 0.83

17.72 ± 0.61

Total PUFA

34.76 ± 1.32

36.20 ± 0.52

37.62 ± 0.93

36.21 ± 1.23

34.38 ± 0.82

Fish oil availability is a concern in the short to medium term: can new GM products be realistic replacements?

Vegetarian EPA and DHA by metabolic engineering of the yeast, Yarrowia lipolytica (DuPont)

 Safe - FDA approved as a secondary ingredient     

in food-grade citric acid production. Produced by fermentation Grown on glucose or invert sugar High oil content Rapid growth Easily to manipulate

Fatty acid compositions of SHFO and Yarrowia oil Yarrowia meal/oil _______________________________________________________________ Fatty acid

Lipid % 14:0 16:0 Total saturates 18:1n-9 Total monoenes 18:2n-6 20:3n-6 20:4n-6 Total n-6 18:3n-3 20:4n-3 20:5n-3 22:6n-3 Total n-3LC-PUFA Total PUFA

Southern fish oil 100 8 16 30 12 20 2 1 1 3 0 1 18 10 30 33

? 1 2 5 3 5 18 2 1 4 2 1 55 0 56 80

New crop platforms Producing omega-3 LC-PUFAs in Camelina sativa.

Acknowledgements • Stirling: J. Pratoomyot, F. Strachan, R.J. Henderson, Dick, J.R., M. Sprague & D.R. Tocher • NIFES: Bente Torstensen, Ingunn Stubhaug, Marit Espe, Gro-Ingunn Hemre, Ernst Morten Hevrøy and Marc Berntssen. • BioMar • Landcatch Natural Selection • Aquamax: www.aquamaxip.eu