A REPORT COMMISSIONED BY SARF AND PREPARED BY

Transcription

1 SARFSP007 - Production of high quality healthy farmed salmon from a changing raw material base with special reference to a sustainable Scottish industry A REPORT COMMISSIONED BY SARF AND PREPARED BY Bluetail Consulting Ltd and The Institute of Aquaculture, University of Stirling

2 Published by the: Scottish Aquaculture Research Forum (SARF) This report is available at: Dissemination Statement This publication may be re-used free of charge in any format or medium. It may only be reused accurately and not in a misleading context. All material must be acknowledged as SARF copyright and use of it must give the title of the source publication. W here third party copyright material has been identified, further use of that material requires permission from the copyright holders concerned. Disclaimer The opinions expressed in this report do not necessarily reflect the views of SARF and SARF is not liable for the accuracy of the information provided or responsible for any use of the content. Suggested Citation: Shepherd C J 1, Monroig O 2 and Tocher D R Production of high quality, healthy farmed salmon from a changing raw material base, with special reference to a sustainable Scottish industry. A study commissioned by the Scottish Aquaculture Research Forum (SARF), 1 Bluetail Consulting Ltd., Richmond TW9 3NL 2 Institute of Aquaculture, School of Natural Sciences, University of Stirling, Stirling FK9 4LA Title: Production of high quality healthy farmed salmon from a changing raw material base with special reference to a sustainable Scottish industry ISBN: First published: July 2015 SARF

3 Report for Scottish Aquaculture Research Forum (SARF) Production of high quality, healthy farmed salmon from a changing raw material base, with special reference to a sustainable Scottish industry Jonathan Shepherd 1, Oscar Monroig 2 and Douglas R Tocher 2 1 Bluetail Consulting Ltd., Richmond TW9 3NL 2 Institute of Aquaculture, School of Natural Sciences, University of Stirling, Stirling FK9 4LA 17 th June

4 CONTENTS Abbreviations (p.5) Executive Summary (p.7) 1. Introduction (p.10) 1.1 Aims of project 1.2 Methodology 2. Farmed salmon industry (p.12) 2.1 Historic growth of international salmon market 2.2 Current UK salmon production and feed supply 2.3 UK salmon consumption, imports and exports 3. Supply and demand for salmon feed ingredients of marine origin (p.15) 3.1 Salmon feed development 3.2 Overview of reduction fisheries 3.3 Supply of fishmeal and fish oil 3.4 Demand from aquaculture, agriculture and human nutrition 3.5 Availability of fishery by-products 4. Effect of dietary innovation in managing cost and availability (p.19) 4.1 Formulating diets to mitigate changing ingredient costs 4.2 Replacement of marine proteins and oils by plant ingredients 4.3 Overview of current salmon diets 5. Managing sustainability concerns by means of certification and standards (p.22) 5.1 Sustainability, food security and means of evaluation 5.2 Demonstrating responsible sourcing of marine feed ingredients 5.3 Demonstrating responsible sourcing of plant feed ingredients 5.4 UK supply chain perspective (salmon farming and feed manufacture) 5.5 UK supply chain perspective (salmon processors and retailers) 6. Are the criticisms of salmon feed sustainability justified? (p.29) 6.1 Overview of the key criticisms and their validity today 6.2 Eco-efficiency, resources use and comparison with land animal production 6.3 Marine resource use and input-output indices 6.4 Effect of changing raw materials on salmon farming sustainability 6.5 Economic, social and ethical issues 6.6 Conclusions and implications for salmon feed sustainability 2

5 7. The issues surrounding omega-3 LC-PUFA in farmed salmon (p.39) 7.1 Public perception 7.2 Recommendations for human intake UK and international recommendations Implications of reducing marine content of salmon feed 7.3 Impact of omega-3 LC-PUFA differentiation by UK retailers Current range of omega-3 label claims The challenge to feed formulation 7.4 Global availability of EPA and DHA Acute shortage in 2014 and Anticipated 2016 global recovery 7.5 Future availability of alternative sources 7.6 Implications for the Scottish salmon industry 8. Novel or emerging alternative feed ingredients (p.50) 8.1 Land animal by-products (LAPs), including insects Current status and availability of land animal by-products in Europe Insect-based feed ingredients Current use of LAPs in salmon feed UK supply chain perspective Conclusions and implications for LAPs/PAP (including insect meals) 8.2 Genetically modified feed ingredients Scope for GM protein ingredients UK use in land animal feeds versus salmon feeds Availability of non-gm proteins Scope for GM oil containing EPA and DHA Conclusions and implications for GM feed ingredients 9. Contaminants and food safety issues (p.57) 9.1 Regulatory background and concern about salmon contaminants 9.2 Beneficial effect of reduced marine ingredient inclusion 9.3 Implications of Marine Harvest initiative to use cleaned fish oil 9.4 Miscellaneous feed-related risks 10. Other salmon feed ingredient issues (p.61) 10.1 Pigments 10.2 Antioxidants 10.3 Other ingredients 10.4 Feed-related quality issues 3

6 11. Industry overview and forward projections (p.63) 11.1 Scottish salmon: current status of salmon farming and projection to 2020 and beyond Current status of salmon farming in Scotland Future projections 11.2 Forward supply of marine ingredients 11.3 Forward plant protein and oil supplies 11.4 Industry feed-related risk exposure 12. Key questions and conclusions (p.69) 12.1 Is Scottish salmon safe and beneficial to consumers? 12.2 Is feeding Scottish salmon a sustainable activity? 12.3 What is the evidence of responsible practice? 12.4 What are the future prospects and challenges? 12.5 What are the industry s strengths, weaknesses, opportunities, and threats ( SWOT analysis)? 13. Main recommendations (p.77) Acknowledgements (p.79) References (p.81) Tables and Figures (p.87) Appendices (p.102) Appendix 1 Main auditable standards and certification schemes currently being used in the farmed salmon supply chain for feed and marine ingredient-related matters (p.102) Appendix 2 Farm and feed-related certification used in Scottish salmon supply chains (p.104) Appendix 3 EU claims for EPA and DHA (p.105) Technical Annexes (p.106) Annex A Atlantic salmon nutrition (p.106) Annex B Human nutrition (p.119) Annex C Alternative feed ingredients (p.128) 4

7 Abbreviations AD, Alzheimer s disease ADHD, attention deficit hyperactivity disorder ASC, Aquaculture Stewardship Council BAP, Best Aquaculture Practice BHT, butylated hydroxytoluene BSE, Bovine Spongiform Encephalopathy CAGR, compound annual growth rate COGPSA, Code of Good Practice for Scottish finfish aquaculture COMA, Committee on Medical Aspects of Food Policy COP, code of practice COX, cyclooxygenase CVD, cardiac and cardiovascular disease CXL, Codex Alimentarius Commission DHA, docosahexaenoic acid DPA, docosapentaenoic acid efcr, economic Feed Conversion Ratio EFSA, European Food Safety Authority ENFEN, Estudio Nacional del Fenómeno El Niño (Peruvian Committee overseeing El Niño ) EPA, eicosapentaenoic acid FAO, Food and Agriculture Organization of the United Nations FCR, feed conversion ratio FFDR, Forage Fish Dependency Ratio FIFO, fish-in fish-out FM, fishmeal FO, fish oil GM, genetically modified GOED, Global Organization for EPA and DHA GRI, Global Reporting Initiative GSI, Global Salmon Initiative GSSI, Global Sustainable Seafood Initiative GVA, Gross Value Added HACCP, Hazard Analysis and Critical Control Points HCB, hexachlorobenzene HOG, head-on-gutted (fish) IBD, Inflammatory Bowel Disease IFFO, International Fishmeal and Fish Oil Organisation IFFO RS, International Fishmeal and Fish Oil Organisation Responsible Supply ISEAL, International Social and Environmental Accreditation and Labeling ISSFAL, International Society for the Study of Fatty Acids and Lipids 5

8 IUCN, International Union for Conservation of Nature IUU, illegal, unreported and unregulated (fish) LAP, land animal by-product LCA, life cycle analysis LC-PUFA, long-chain polyunsaturated fatty acid LNA, α-linolenic acid LOA, linoleic acid MAT, Moving Annual Target MODR, marine oil dependency ratio MPDR, marine protein dependency ratio MPL, maximum permitted level MRL, Maximum Residue Level MSC, Marine Stewardship Council NGO, non-governmental organisation NICE, National Institute for Health and Care Excellence NRC, National Research Council OECD, Organisation for Economic Cooperation and Development PAH, polyaromatic hydrocarbons PAP, Processed Animal Protein/Product PCB, polychlorinated biphenyl PKS, polyketide synthase RTRS, Roundtable on Responsible Soy R&D, Research and Development SACN, Scientific Advisory Committee on Nutrition SARF, Scottish Aquaculture Research Forum SBM, soybean meal SPC, soy protein concentrate SPM, specialised pro-resolving mediator SSC, Sustainable Seafood Coalition SSPO, Scottish Salmon Producers Organisation UFAS, Universal Feed Assurance Scheme USP, Unique Selling Point VKM, Vitenskapskomiteen for mattrygghet (Norwegian Scientific Committee for Food Safety) VO, vegetable oil WFE, whole fish equivalent 6

9 Executive Summary 1. The Scottish salmon industry produced close to 165,000 tonnes of salmon in 2014 and is relatively small in volume compared with Norway and Chile, but holds a distinctive market position. Whereas the Norwegian salmon farming industry is the highly efficient leading world producer and exporter of farmed salmon as a commodity (including to the UK market), the Scottish industry (lacking comparable economies of scale) supplies more differentiated products at higher unit value to offset potentially higher production costs. These range from standard to high performance products, including bespoke supply for niche markets in the UK and overseas (mainly the USA and France). Scottish production therefore relies on a variety of feed products, including specialist formulations to support a premium salmon segment differentiated by label claims on Scottish provenance, the content of omega-3 LC-PUFA in regard to EU intake recommendations, and other feedrelated quality parameters. 2. The growth in Scottish salmon farming has resulted in year-round availability of costefficient, affordable, salmon products on a greatly increased scale, while creating wealth and employment, especially in remote rural locations in the North West Highlands & Islands and in the Northern Isles. The industry has grown to be the leading Scottish food exporter and is sustainable in terms of demonstrating clear economic and social benefits. 3. Salmon feed is the industry s main operating cost and the rapid development of salmon farming has in large part been due to intense innovation, especially in husbandry and nutrition. This has enabled producers and feed manufacturers to achieve greater control over diet costs and to manage availability of feed ingredients by utilising a greater range and more secure supply of raw materials than before. 4. In terms of eco-efficiency, farmed salmon retain feed nutrients, and convert them to nutrients for human consumption, more efficiently than land animals. Life cycle analysis shows salmon farming has a lower carbon footprint and makes better use of resources than farming either pigs or chicken; also that using fishmeal and fish oil from unmarketable wild fish, derived from reduction fisheries, provides more marine protein, energy, and essential long chain fatty acids (EPA & DHA) for human consumption than utilising these marine resources directly as a human food source. 5. Farmed salmon is clearly beneficial for human health and nutrition, with its highly digestible protein, essential amino acid and fatty acid composition, as well as its micronutrient content (including lipid-soluble vitamins A, D, and E). In addition to being nutritious, the virtual absence of environmental contaminants makes Scottish farmed salmon safe to eat. 6. It is shown that criticism of the Scottish salmon industry, alleging that salmon feeds are unsustainable, is largely false and out of date. Despite little, if any, increase in global 7

10 fishmeal and fish oil supplies, due to setting more precautionary fishing limits, growth in fed aquaculture has continued unabated without putting increased pressure on wild fish resources (as some critics had predicted would occur). Instead progressive replacement of marine feed ingredients means that the composition of salmon feed today is dominated by protein and oil from plants rather than fishmeal and fish oil from forage fisheries, with an increasing proportion of fishmeal derived from trimmings and by-products of processing fish for human consumption. However, continuing use of marine ingredients as a minor proportion of the diet is important for salmon health and for maintaining high levels of omega-3 LC-PUFA in the resulting salmon products. 7. To ensure that supplies of fishmeal and fish oil are sustainable, the industry has chosen to source only from well-managed, renewable, fisheries and to evidence this by compliance with relevant standards, chains of custody, etc. At the same time there is growing recognition that sustainability of plant raw materials is not necessarily greater than use of (well managed) marine resources, hence simply substituting marine by plant ingredients in salmon feed is counter-productive in regard to sustainable salmon production, unless the crops are responsibly grown (e.g. by using soya protein concentrate from certified sources with limited environmental impact). In this connection it is argued that the use of marine dependency ratios in rating salmon farming operations has little, if any, relevance to the sustainability of the process. 8. To meet the Scottish industry s planned 2020 salmon output, the required increased volumes of marine and plant proteins and of plant oils will not be a constraint. However, R&D effort should continue to work towards securing a greater proportion of locally-grown plant materials instead of relying mainly on imported processed soya etc. The relatively higher marine content of some Scottish feeds (e.g. compared to Norwegian practice), enabling the production of specialist premium qualities of salmon, directly reinforces a healthy and natural reputation and is a worthwhile trade-off against the potentially greater exposure to feed ingredient cost fluctuations. 9. The Scottish salmon industry s main strength is its well-established, favourable, reputation internationally, reinforced by evidence that the product is safe to eat, sustainable, and benefits from a strong research base. At the same time current salmon supply chain policies strictly avoid the use of both genetically modified feed ingredients and land animal by-products, which serves in turn to restrict formulation flexibility, increase costs, and potentially also conflicts with the industry s focus on sustainability. 10. The industry s main challenges are market related, from losing its premium reputation in the UK, by seeking to compete with more standard and lower cost products, to the possible threat to specific export markets from changing feed ingredient policies. In this regard, genetically modified feed ingredients are now the norm in UK poultry products, but their possible introduction to salmon feed requires due consideration of the potential market costs, benefits, and risks. 8

11 11. Going forward the main priority is to retain and reinforce the strong reputation for Scottish salmon quality in the UK and key export markets, offering quality and high value (including not only healthy and safe, but also third party audited evidence of responsible production and sustainable feeds, etc.). This is against a background of increasing retail discounting and greater focus on more standardised cheaper salmon products, which do not make specific label claims (e.g. on omega-3 content). 12. The most critical priority is to manage the content of EPA and DHA in Scottish salmon, against the background of a world-wide omega-3 LC-PUFA scarcity, in order to maintain the healthy image of Scottish salmon. In addition to recurring El Niño events impacting on the oil-rich anchovy fisheries off Peru and Chile, there is now competition for fish oil from the nutritional supplements industry (e.g. omega-3 capsules). At the same time, replacement of fish oil by plant oil in salmon feeds is inevitably reducing the omega-3 LC-PUFA content of farmed salmon. In the short-term this issue will temporarily resolve itself in Peru, and in the medium-term algal DHA may play a useful role. In the longer term, in view of the food security and consumer health implications, this Report believes it is of strategic interest for the Scottish industry to consider adopting a more flexible attitude to new alternatives to fish oil, including sources of EPA and DHA derived from transgenic oil seed crops, when they become available. 13. The industry s commitment to responsible production includes addressing ethical concerns. For instance, the majority of Scottish salmon farms are certified to the Freedom Food standard for fish welfare, so helping to differentiate the product internationally, while demonstrating the industry s willingness to extend the boundaries of best farming practice. Not only is this ethical focus likely to be better understood and appreciated by UK consumers than technical issues of sustainability, it is also highly supportive of the industry s continuing demonstrable commitment to sustainable development. 14. This Report concludes that the Scottish salmon industry has an excellent story to tell in regard to sustainable feed development and can be confident that the changing raw material base will support continuing production of high quality, healthy farmed salmon, but it needs to address the long term security of supply for omega-3 LC-PUFA. 9

12 1. Introduction 1.1 Aims of project The title of the project (SARF SP007) is The production of high quality healthy farm salmon (Salmo salar) from a changing raw material base, with special reference to a sustainable Scottish industry. The main objective, its relevance to SARF, and the intended use of the results were summarised in the project application, as follows: In order to inform the ongoing debate on the future supply of farmed salmon into home and international markets, this study will review the nutritional, ethical, marketing and sustainability issues surrounding a reduction in our reliance on high levels of marine ingredients, fish meal and fish oil, and their substitution by high quality terrestrial or other ingredients. This study will undertake a critical examination of the implications of changing market and raw material conditions on prospects for maintaining cost-effective supply of Scottish farmed salmon, especially in terms of the nutritional and formulation challenges to meet supply chain and consumer requirements for healthy, sustainable product of good quality. The final report will provide SARF and external audiences with an independent, authoritative, and strategic understanding of the issues surrounding fish feeds in the Scottish salmon farming industry. The background to this project is that during the last decade an increasing proportion of the fishmeal and fish oil in salmon feeds has been replaced by plant-derived (vegetable) raw materials and this has caused changes in flesh composition of farmed salmon, affecting mainly the content of omega-3 (or n-3) long-chain polyunsaturated fatty acids (LC-PUFA). Although algal sources of LC-PUFA are becoming available at high cost and research on LC- PUFA production in genetically modified plants (e.g. Camelina sativa) is very promising, costeffective alternatives to fish oil will not become available on a commercial volume basis in the short-term. Hence the importance of defining the most appropriate strategies for the salmon industry to manage the changing raw materials supply position, while seeking to maintain farmed salmon quality (especially consumer health benefits) and at the same time addressing the question of sustainable use of finite resources and the eco-efficiency of farming salmon. The purpose of this study is to undertake a detailed analysis of the challenges, including the options available to feed formulators and manufacturers going forward and the implications for composition of the resulting farmed salmon. These concern not only human health and nutrition, but also the linked questions of raw material sourcing and final product placement as they relate to ethical, marketing and sustainability considerations. These issues have led to confusion, as well as poorly informed and erroneous comment, and this now requires urgent clarification based on scientific fact and rational argument, so the final report will be supplemented by a clear and concise summary suitable for an intelligent lay audience. This project therefore aims to make available a thorough, independent, and truly credible report, backed up by in-depth analysis of all the relevant issues. 10

13 1.2 Methodology The approaches and research plan were as follows: (i) Literature data collection (ii) Survey and interview key experts and stakeholders in (i) fishmeal and fish oil (ii) alternative raw material quality, availability and sourcing (iii) salmon nutrition and feed formulation (iv) fish carcase quality and assessment (v) salmon supply chains and product marketing (vi) resource allocation, ethics, and eco-efficiency, with reference to farmed salmon production (iii) Human health, nutrition and epidemiology - Meet medical experts and evaluate work on human health effects of fish consumption, especially n-3 LC-PUFA, e.g. Southampton University (iv) Fishmeal and fish oil Discussions with experts at International Fishmeal and Fish Oil Organisation (IFFO), feed supply chain executives, and analysts in order to evaluate trends in demand/supply (v) Protein Analysis of potential alternatives to fishmeal based on discussions with UK and Norwegian Universities and institutes, fish nutritionists, feed formulators, and by-product renderers (vi) Oil Analysis of potential alternatives to fish oil based on discussions with experts in algal and GM plant production sources of n-3 LC-PUFA, including trade organisations (e.g. GOED) (vii) Eco-efficiency and ethics of using marine ingredients for farming fish; consumer perception affecting supply chain policies (e.g. use of Processed (terrestrial) Animal Protein, GM issues). Explore basis of current sensitivities with experts in sustainability and life cycle analysis (especially research being undertaken by Scandinavian workers) and supply chain marketers (viii) The main body of the Report text focuses on the commercial and supply chain issues that are the predominant drivers of raw material use and salmon feed formulation and policies. This is supported by Technical Annexes that provide the scientific background and understanding, which underpin salmon nutrition, raw material choice and human health implications. (ix) Preliminary and final comprehensive Reports to SARF; production of summary pamphlet It should be emphasised that this project was essentially a desk-based study that does not require experimental laboratory research. Given that the Scottish salmon industry is much smaller than the Norwegian industry and reliable industry data is more widely available in Norway, the project involved meetings with researchers both in the UK and Norway. 11

14 2. Farmed salmon industry 2.1 Historical growth of international salmon market Salmon farming is the most highly developed form of large-scale intensive aquaculture owing to its productivity growth and technological change. Figure 1 is a summary of global production, consumption, and trade flows of farmed Atlantic salmon (Kontali, 2013), with Norway and Chile being the large volume producers, compared with smaller industries in Scotland and North America. The 2013 estimated production figures of 1,094,000 tonnes for Norway, Iceland and Faroes Islands and 153,000 tonnes for UK and Ireland are for head-ongutted salmon (HOG, i.e. bled and gutted); this compared with 421,000 tonnes for Chile and 122,000 tonnes in North America. The compound annual growth rate (CAGR) for farmed Atlantic salmon was 6 % on a global basis from but is now reducing (Kontali, quoted by Marine Harvest, 2014). 2.2 Current UK salmon production and feed supply Scottish farmed salmon production in 2013 was 163,234 tonnes whole fish equivalent (WFE, i.e. round bled fish) and is predicted to be c. 165,000 tonnes in This was produced by 27 farming companies and 102 farm sites. By comparison 2013 Norwegian farmed salmon production was 1,143,500 tonnes slaughtered (Kontali, 2013), hence Scottish farmed volume was 12.5 % of combined Scottish and Norwegian production in A study of Scottish aquaculture in 2014 noted that production of marine finfish in 2012 (virtually all salmon) was 164,380 tonnes with a turnover of 550 million from production only; this corresponded to over 800 million turnover and 265 million Gross Value Added (GVA) across the Scottish supply chain, employing over 4,000 individuals (Marine Scotland 2014). As regards feed, according to industry sources UK total salmon feed supply in 2014 was 220,000 tonnes with a claimed overall industry feed conversion rate (FCR) of However, the overall economic FCR (efcr) is probably over 1.2 after taking account of harvest yield, feed consumed by unharvested fish, and health-related production losses. This compares with a global salmon feed market which has grown by 38.7 % in the last four years from 2.70 million tonnes in 2010 to 3.87 million tonnes in 2014 (Ian Carr, Ewos, unpublished). 2.3 UK salmon consumption, imports and exports Assuming Scottish production of farmed salmon in 2014 was c. 165,000 tonnes WFE, this would equate to c. 150,000 tonnes HOG. The Seafish trade summary (based on BTS, HM Customs & Excise) gave total UK salmon imports and exports for 2014 as 88,716 tonnes and 128,611 tonnes, respectively. The UK import figures appear feasible and are dominated by 12

15 Norway, which exported 61,639 tonnes HOG (Jack Møller, Norwegian Seafood Export Council, unpublished), but the export figures are possibly flawed due to international transshipments, etc. (e.g. exports could include UK processed and transformed Norwegian fish). However, if correct, they would equate to a net UK apparent consumption of 110,000 tonnes per annum of all types of salmon on a HOG basis (c. 122,000 WFE). The equivalent data from Nielsen and Kantar (via Seafish) gave the total UK salmon Moving Annual Target (MAT) to end-nov 2014 as only 54,200 tonnes, but this refers to sold product weight. Given the above data and informed market comment, it seems likely that for 2014 the net apparent UK salmon consumption was over 100,000 tonnes per annum. This is reinforced by Asche and Bjørndal (2011), who claimed that UK salmon consumption (excluding tinned salmon) was around 120,000 tonnes WFE and it is likely to be higher now. Total salmon exports in 2014 were 128,611 tonnes worth 640,208 ( 4.98/kg) and the main export markets have been USA (44,183 tonnes), France (28,956 tonnes), and China (14,726 tonnes), according to the Seafish trade summary. It is notable that export value and volume have increased on the back of a deflated /kg. As regards UK fish imports, salmon has taken first position with falling prices driving import value and volume growth. Retail sales are believed to represent around 80 % of the total UK salmon market and in 2014 Tesco, although slipping, still held a dominant 25 % market share of retail, followed by Sainsbury s at 18 %, Morrison at 11 %, Marks & Spencer/Waitrose/Aldi each having c. 9 % and Asda at 8 % (data from Nielsen and Kantar). It is claimed that on-line seafood sales are growing fast, with Ocado the market leader and farmed salmon as a driver. Interesting questions include what volume of Scottish salmon is supplied to the UK market and what proportion represents the Scottish premium market (supplied predominantly by Sainsbury s, Waitrose, and Marks & Spencer; covering a combined 36 % market share). The limited data available to this study and the likelihood of international trans-shipments do not allow Scottish production volumes to be split into home and export destinations, but the sales of premium Scottish salmon in the UK are likely to represent under half of UK total annual salmon sales (with the majority of Scottish salmon going for export), while the balance of UK consumption is comprised of standard product from Norway. The answers to these questions would allow an informed view on how best to optimise the value of premium Scottish sales given different future scenarios; this is not merely about maximising net sales value and product range as between home and export markets, but also how to react to future threats and opportunities (e.g. allocating between markets that are neutral or hostile to purchasing salmon fed on GM ingredients; see 8.2). Summary of Chapter 2 The production of Scottish farmed salmon is estimated to be c. 165,000 tonnes harvest weight in 2014, which is small in the context of the global salmon industry s production 13

16 statistics and trade flows. UK salmon feed volume (excluding exports) in 2014 is c. 220,000 tonnes. The UK salmon import and export figures are calculated to give an estimated UK annual salmon consumption of over 100,000 tonnes HOG (excluding canned salmon etc.). The main Scottish export markets for salmon are the USA, France, and China, in that order. 14

17 3. Supply and demand for salmon feed ingredients of marine origin 3.1 Salmon feed development A good diet is a prerequisite for proper growth and development and optimal health. In animal production, the diet is determined by the composition of the feeds provided and the key to ideal diet formulation is understanding the precise nutritional requirements of the animal, with the aim to produce a feed that supports both optimal growth and health. The feed must provide a balance of nutrients that provides sufficient energy, macronutrients for formation of new tissue (protein and lipid), and micronutrients (minerals and vitamins). Consequently, considerable research has been focused on salmonid fish nutrition, particularly in recent years, and the current nutritional requirements for salmonid species, including Atlantic salmon, were recently updated (NRC, 2011). A brief summary of the nutritional requirements of Atlantic salmon is provided in Annex A. Historically the two most important ingredients in salmon feed have been fishmeal and fish oil. Both had been traditionally used in animal feeds during the 20 th century and so, with the development of intensive salmon farming and the need for artificial feeds, they were obvious ingredients as the major sources of protein and oil/fat in formulations. Indeed, the use of fishmeal and fish oil in salmon feeds was a sensible and logical approach years ago during the initial development of intensive salmon farming in Scotland. Both marine ingredients were readily accepted and digested by salmon, had favourable nutrient compositions, and represented the natural food for salmon (NRC, 2011). However, the use of fishmeal, and more recently of fish oil, has reduced and been progressively replaced by agricultural plant-based commodities, such as soya, sunflower, wheat, corn, beans, rapeseed oil and, in Chile and North America, by poultry by-products. Fishmeal has a more complete amino acid profile compared with protein of plant origin and a higher protein concentration, so it has been a challenge to replace fishmeal completely. During the industry s early phase, salmon feed had high levels of marine protein (c. 60 %) and low levels of fat/oil (c. 10 %). In the 1990s the feed typically consisted of 45 % protein, mostly fishmeal. Today the level of marine protein is lower due to cost optimisation and reduced fishmeal availability. An important development has been the increasingly higher inclusion of oil, made possible by extruded feed technology and driven by the protein sparing effect of oil promoting growth. 3.2 Overview of reduction fisheries A reduction fishery is dedicated to catching fish (usually small pelagic species) for reduction to fishmeal and fish oil; such fisheries are otherwise known as forage or feed fisheries. The species of fish used for fishmeal and fish oil in Norwegian salmon feed in 2012 are detailed in Table 1 with around 50 % purchased from Peru and Chile (Ytrestøyl et al., 2014); the profile of Scottish sourcing is considered to be very similar. The Sustainable Fisheries Partnership (2012) rated the 28 principal reduction fisheries around the Atlantic and South 15

18 America according to sustainability assessment, based on the quality of management and the status of the stock; they concluded that most of the assessed fisheries operate within the limits that would be considered consistent with current good practice in the context of single species management regimes. The largest reduction fishery by far is that for Peruvian anchovy (Engraulis ringens) and it is significant that in 2008 Mondoux et al. ranked Peru the highest out of 53 maritime countries for the sustainability of its fisheries, since when Peru has reduced its fishing overcapacity and further improved its management by the introduction of maximum catch limits per vessel. Some previously negative reports on individual fisheries have also been reconsidered in the light of better modelling; for instance the Atlantic States Marine Fisheries Commission (Sedar, 2015) has now determined that its previous assessment was in error and the Atlantic menhaden (Brevoortia tyrannus) resource is in fact healthy and not overfished. The reduction fish catch has been categorised by species as either industrial-grade unfit for human consumption (e.g. menhaden), or food grade (e.g. Peruvian anchovy), where those willing to purchase them as food are distant from the fishery and cannot normally pay for the preservation and transport costs (FAO/NACA, 2012). It was also noted that prime food fish are diverted for reduction when landings of edible small pelagics (e.g. pilchards, Sardina spp.; Sardinops spp.) are unpredictably large, so that not all can be preserved or processed as human food and downgrades are processed into fishmeal, but this use has markedly reduced. It is obviously in the interest of fishers and processors to maximise the value of the catch, which normally favours its use for human consumption. Indeed thanks to processing development, species such as jack mackerel (Trachurus spp.), which as whole fish were formerly used for reduction purposes, now go instead for human consumption with only their process trimmings going for fishmeal. 3.3 Supply of fishmeal and fish oil Fishmeal and fish oil are globally traded commodities similar to other raw materials or primary agricultural products that have long-since been produced and marketed; aquaculture grew to become the biggest customer for these marine resources (Jackson and Shepherd, 2012). Figs. 2 and 3 detail annual world fishmeal and fish oil production respectively by main producing region from 2003 to 2013, with Peru the dominant world supplier followed by Chile. Over this 11 year period the mean annual fishmeal and fish oil production was 5.23 million tonnes and 0.95 million tonnes, respectively. Fig. 4 compares world fishmeal and fish oil consumption with growth in fed aquaculture from 2000 to It can be seen that rapid growth in fed aquaculture (FAO, 2014) has not been restricted by what is, at best, static supply of marine ingredients since Production is subject to environmental influences and, whereas the potential impact of climate change is not well understood (Callaway et al., 2012), acute phenomena, such as El Niño, have wellknown consequences, especially for the Peruvian anchovy fishery. Figure 4 also clearly shows fishmeal and fish oil production fluctuations during 2009 to 2012 (mainly due to El 16

19 Niño events). Given these supply uncertainties in the dominant southern Pacific anchovy fishery, compounded by increased regulatory focus on precautionary catch quotas, it follows that continuing growth by major users of marine ingredients (notably salmon farming) has only been possible by their increased substitution in feeds by other ingredients (4.2). 3.4 Demand from aquaculture, agriculture, and human nutrition Figure 5 shows the supply chains involved in the marine ingredients industry, serving animal feed (aquaculture and land-animal markets) and human nutrition (fish oil supplements and pharmaceuticals markets). World usage of fish oil from 2003 to 2013 is given in Fig. 6, with emerging demand from direct human consumption (so-called nutraceuticals ) appearing to be at the expense of aquaculture use, which averaged around 800,000 tonnes per annum until 2012, but declined to just below 700,000 tonnes per annum in The pattern of fish oil use in 2013 (Fig. 7), showed that around 75 % of total global supply was used in aquaculture, with 21% going for direct human consumption. Around 83 % of fish oil used in aquaculture feeds in 2013 was consumed by salmonids (60 % mainly salmon) and by marine fish (23 %) (Fig. 8). Whereas prices for fish oil and rapeseed oil were broadly similar a decade ago, demand for fish oil is now largely a derived demand for omega-3 LC-PUFA and priced at a premium above rapeseed oil directly linked to its content of eicosapentenoid acid (EPA) and docosahexaenoic acid (DHA) (Shepherd and Bachis, 2014). Thus rapeseed oil, which has no EPA and DHA, is included as a more cost-effective source of dietary oil for energy purposes only. By contrast, demand for fishmeal is linked to the price of alternative protein ingredients, particularly soybean meal, at a historical price ratio of 3:1 reflecting relative nutritional performance (i.e. if the fishmeal price exceeds three times the price of soy, increased inclusion levels of soy will tend to replace fishmeal, or the opposite if it falls below three times the price of soy). The main competitor of aquaculture for fishmeal is pig feed, especially for young pigs at weaning, but aquaculture is gradually taking market share from land animals, as pig and poultry farmers tend to be more price sensitive than fish farmers (at least than salmon farmers) and substitute with other ingredients when fishmeal prices increase too far in comparison. The opposite is sometimes true for the supply of fish oil for direct human consumption in which the nutraceuticals sector has appeared to be able and willing to pay more for fish oil than aquaculture. An estimated 68 % of global consumption of fishmeal in 2012 was by aquaculture, compared with 30 % by pigs and chicken and 2 % by other sectors, but it is estimated that the use in aquaculture increased to 72 % in 2013 at the expense of land animals (Fig. 9). Of this aquaculture use, it is estimated that 24 % was used by salmonid species, mainly salmon (Fig. 10). 3.5 Availability of fishery by-products 17

20 An increasing proportion of fishmeal, and to a lesser extent fish oil, is derived from seafood and aquaculture by-products, including by-catch and trimmings etc. Ytrestøyl et al. (2014) showed that for % of fishmeal and 27 % of fish oil used in salmon farming in Norway were from seafood industry by-products; it is expected that the Scottish proportions would be broadly similar. Figure 11 projects global use of by-products in marine ingredients from 2000 to 2021 with a figure of 37.5 % in 2014 and reaching 44 % by 2021 (OECD-FAO, 2013). However, there are difficulties reconciling Chinese statistics, in particular, and an alternative view is that global use of by-products in marine ingredients is % currently, with an increasing trend over time (Andrew Jackson, personal communication). This by-product contribution has partially offset some of the reductions in production from whole fish as a result of increased restrictions on the reduction fisheries (Shepherd and Jackson, 2013). Summary of Chapter 3 Until recently salmon feed ingredients were dominated by fishmeal and fish oil. The statistics of fishmeal and fish oil supply are reviewed, showing static production from whole wild fish and an increasing proportion from fishery by-products; this has not limited the global growth of aquaculture. Aquaculture s use of fishmeal increased at the expense of pigs and chicken, but is now on a plateau with feed inclusion rates linked to the relative prices of fishmeal and soya and the ratio varying traditionally around 3:1 soya:fishmeal. Aquaculture s demand for fish oil is mainly from salmon and competes with nutraceuticals on the basis of EPA and DHA levels. The current El Niño has led to a scarcity of Peruvian fish oil and a resulting pressure to reduce inclusion of fish oil in feeds 18

21 4. Effect of dietary innovation in managing cost and availability concerns 4.1 Formulating diets to mitigate changing ingredient costs The pattern of stagnating wild fisheries and the fast growth of aquaculture risked overdependence on a limited range of feed ingredients, especially fishmeal and fish oil, while the issues involved in replacement by alternative proteins and oils were soon appreciated (Sargent and Tacon, 1999; Naylor et al., 2000). Wild ( forage ) fish for reduction purposes are finite resources with production necessarily limited and declining slightly, due to strict regulation of fishing, resulting in reduced catch quotas, especially for the largest producers of fishmeal and fish oil, ie. Peru and Chile (Jackson and Shepherd, 2012). As global commodities, the prices of fishmeal and fish oil show marked fluctuations reflecting supply and demand factors in the market. In the case of salmon feed there is a clear effect of increased marine ingredient substitution by plant meals and oils following supply interruptions, notably El Niño (Shepherd and Bachis, 2014). The price volatility of fishmeal and fish oil is illustrated by Figs. 12 and 13 with widely fluctuating price ratio data during 2010 to 2013 for fishmeal (compared to soybean meal) and fish oil (compared with rapeseed oil), respectively. Apart from a need for low levels of essential fatty acids for salmon growth, supplied via small quantities of (the oil present in) fishmeal or via fish oil, fishmeal itself is not an essential feed ingredient for aquaculture (or land-animal feeds) per se, but it provides a near optimal complete feed in a convenient, cost-effective, and highly digestible product form (Tacon and Metian, 2008). For reasons of cost reduction and security of supply, cheaper alternative ingredients have been researched and progressively substituted in commercial feed formulations, where technically and economically feasible, and after further processing as required. 4.2 Replacement of marine proteins and oils by plant ingredients Information about salmon feed formulation is more widely available in Norway than Scotland due to reporting requirements and policy issues. Despite the same three feed companies being involved in both Norway and Scotland and the Scottish salmon farming industry being mostly owned by Norwegian companies, the smaller Scottish industry is less transparent. However, the two salmon farming industries are similar with some key differences in response to UK market focus, e.g. on omega-3. Thus feed formulation trends are broadly comparable between the two countries, with Scottish salmon feeds being more conservative in terms of rate of change and more varied in reflecting bespoke customer requirements. Fig. 14 shows the large changes in composition of Norwegian salmon feed since 1990 when c. 90 % of the diet came from marine ingredients compared with 29.3 % in 2013; the marine content reduced 15 % between 2010 and 2013, by which year protein represented 55 % of the diet (with 67 % plant/33 % marine) and oil represented 30.1 % of the diet (with 64 % plant/36 % marine). Table 2 details the different plant ingredients used in Norwegian salmon feed production in 2012 and 2013 by the 3 main feed companies, plant protein accounting for 55 % of the total plant ingredients (with soy protein 19

22 concentrate accounting for 62 % of the plant protein); plant oil as rapeseed oil accounting for 29 %; and the balance of 16 % being binders, mainly wheat (Ytrestøyl et al., 2014). A broadly similar range of plant ingredients is used in Scotland. 4.3 Overview of current salmon diets As detailed above, the average Norwegian salmon feeds in 2013 were comprised of 70 % plant and 30 % marine ingredients and market comment suggests this has been maintained in The feed comparison with Scotland is complicated by the proliferation of bespoke diets, some of which are driven by external standards (e.g. Organic, Label Rouge etc.), but mostly by farm customer requirements linked to retail requirements. As a result average Scottish salmon diets for 2013/14 are believed to have been comprised of 60 % plant and 40 % marine ingredients. However, this obscures the large range of different Scottish formulations; for instance in 2014 c. 7,000 tonnes (4 %) of production was made for export to the Label Rouge salmon specification with a marine content of 51 %, while at the other extreme there is significant salmon volume being produced approximating to Norwegian diet specifications. At the same time there has been a move in Scotland during 2013/2014 towards use of High Performance diets of variable specification, but generally higher use of fishmeal. Nearly all Scottish feeds are being currently formulated to give not less than 22 MJ/kg in order to achieve faster growth and improved feed conversion. Typically this entails diets with 37 % lipid and 36 % protein and follows the recent Norwegian dietary regime with high salmon prices allowing farmers to invest in higher specification diets at higher up-front feed cost in order to yield higher marginal productivity. As regards use of marine ingredients, at the present time this means that the overall average marine protein and oil content of salmon diets is probably around 25 % and 15 %, respectively in Scotland, compared with around 20 % and 10 %, respectively in Norway, and with higher average EPA and DHA content of Scottish farmed salmon. In neither country is there any use of land animal by-products, which accounts for over 20 % dietary inclusion in Chile and Canada, or of genetically modified ingredients (e.g. Chile and Canada use GM soya). The specific feed ingredients used in Norway in 2010 and 2012 are compared in Fig. 15. Plant proteins used to supplement soy protein concentrate include wheat and maize gluten, rapeseed protein concentrate, pea protein concentrate, sunflower expeller, faba beans etc. These are included according to the results of computerised least cost formulation in order to optimise recipe cost (depending on price in relation to specified constraints reflecting nutritional value), provided the farm/product contract allows it. There is also significant usage of wheat starch to aid pellet manufacture rather than for significant nutritional value. With feed now being formulated to a target EPA and DHA inclusion, the marine content can differ significantly when Peruvian oil is used versus North Sea oil, with the latter falling from 52 % of oil origin in 2010 to 29 % of oil origin in 2013 at the same time as dietary marine oil content fell from c. 16 % to c. 11 % (Ytrestøyl et al., 2014), as a consequence of using less marine oil but with a higher concentration of EPA and DHA. Fig. 16 illustrates the changing 20

23 ingredient prices for five key raw materials in salmon feed during the period from 2006 to 2013, with a rising trend for fishmeal, fish oil, and soya, compared with more stable prices for rapeseed oil and wheat. The 2013 cost of salmon feed in Norway and Scotland of Marine Harvest (the largest global salmon producer) was estimated at 50 % and 46 % of production costs, respectively, compared with Canada and Chile where it was lower at 41 % and 43 %, respectively, mainly due to lower feed costs linked to use of land animal by-products (Marine Harvest, 2014); in any event it is clear that feed costs dominate production costs for farmed salmon. Summary of Chapter 4 Dietary innovation has enabled feed formulators to use a wider range of ingredients, hence reducing costs and improving security and stability of raw material supply. In particular, marine ingredients have been progressively replaced by plant proteins and oils. It is estimated that on average Scottish salmon diets now include c. 40 % marine ingredients, which is more than Norwegian salmon feed, but reflects focus on omega-3 LC- PUFA content as well as Label Rouge products. Whereas feed is the main operating cost in salmon farming, the absence of land animal by-products (LAPs) from European salmon feeds means that costs are somewhat higher compared with salmon farming in Chile and North America. 21

24 5. Managing sustainability concerns via certification and standards 5.1 Sustainability, food security and means of evaluation Aquaculture, and especially salmon farming, has received much criticism from NGOs and environmental commentators, who question whether it is a sustainable form of food production. In seeking to take an informed view it is important to define the relevant terms. A generally accepted definition of sustainable development from the United Nations Brundtland Commission is that it meets the needs of the present generation without compromising the ability of future generations to meet their own needs (WCED, 1987). Included in this definition is not only a development that secures the global resource basis and the environment, but also includes a social and economic aspect with responsibility for securing the basic needs of the present and future population. It is clear that access to sufficient food with a satisfactory nutritional quality is a basic human need, and one of the major challenges in the next years will be to increase the world s food production to support a population of 9-11 billion people in 2050 despite changing climate conditions. The United Nations 2005 World Summit noted that this requires reconciling three sustainability pillars, i.e. Environmental, Social, and Economic demands (see Fig. 17), which have become adopted by many sustainability standards and certification systems, especially for the food industry. Thus fish and aquaculture supply chains, especially in Western Europe (including the UK) and North America, increasingly require producers to provide evidence of responsible practice, not only in regard to food safety, but also adherence to specified sustainability goals. At the same time it should be noted that sustainable development, as defined by the United Nations, should be seen as an aspirational journey and, realistically, supply chains should strive to move in the direction of sustainability and monitor progress towards achieving that goal rather than claiming they have succeeded in arriving at it. It is acknowledged that this interpretation is inconvenient for supply chains seeking to make straightforward consumer-facing sustainability claims, but reflects the difficulty of making such a claim except at a moment in time within a continually changing landscape; this in turn highlights the need for continual innovation in order to adapt to changing circumstances (economic, social, and environmental). Methods for comparing the environmental cost of aquatic and terrestrial production include the analysis of material and energy flow, risk analysis, life cycle analysis, ecological footprint, environmental impact assessment, as well as environmental (or ecological) cost-benefit analysis etc. No single method is robust enough to capture all relevant environmental impacts and costs, hence these tools should be used together and interpreted in a complementary way to properly appreciate the system s eco-efficiency (e.g. for estimating nutrient retention efficiency, mass balance models are more useful than life cycle analysis) (Ytrestøyl et al., 2014). Also some environmental aspects of aquaculture are unique (e.g. 22

25 using pelagic species as feed) and it is important to consider the management of reduction fisheries Demonstrating responsible sourcing of marine feed ingredients Standards and certification schemes have become an important tool by which supply chains seek to manage concern issues, such as sustainability. As a generalisation, when considering food and feed, Asia is most concerned about safety and purity of food products, while Europe and North America take that as a given and are more concerned about long-term sustainability of the activity, especially from an environmental perspective. Appendix 1 profiles the main auditable standards and certification schemes currently being used in the farmed salmon supply chain for marine ingredient and feed-related matters (including web addresses for each). It may be noted that the Code of Good Practice for Scottish finfish aquaculture (COGPSA) seeks to enhance the industry s reputation for respecting the environment through adoption of best practice and greener technologies and reducing the impact on wild fisheries by increasing use of alternative feed sources ( In addition to recommendations on feed formulation and use (e.g. source from suppliers participating in the Universal Feed Assurance Scheme (UFAS)), the code therefore requires that fish-catch supplies used in the manufacture of fishmeal are sourced from fisheries that are properly and responsibly managed; it specifies that this should be either by reference to the FAO (Food and Agriculture Organisation of the United Nations) Code of Conduct for Responsible Fisheries ( or the IFFO RS scheme ( or by another globally recognised standard for responsible operation. In the area of marine feed ingredients used in aquaculture there are six commonly used standards. Apart from GlobalGap (which is more focused on food safety), they all claim to be based on the key principles underlying the FAO Code of Conduct for Responsible Fisheries. The MSC (Marine Stewardship Council) standard certifies a fishery; IFFO RS certifies a fishmeal factory; ASC (Aquaculture Stewardship Council), BAP (Best Aquaculture Practice) and GlobalGap all certify fish farms, although BAP has a separate feed mill standard. Friend of the Sea certifies fisheries, fishmeal plants, feed mills and fish farms. The MSC aims to promote sustainable fishing practices, has tightened its rules on certifying low trophic level fisheries, and has virtually no certified reduction fish stocks. This will undermine the strategy of its sister organisation ASC to insist long-term on MSC certified material (or another fishery standard, which is ISEAL-compliant) for whole fish, unless it changes when the new ASC salmon feed standard is developed and implemented. Meanwhile, as an alternative to MSC, the ASC salmon standard is permitting the use of raw material derived from fisheries with a high rating by the FishSource scoring system ( devised by Sustainable Fisheries Partnership. The IFFO RS certifies that a factory manufacturing 23

26 fishmeal or fish oil is responsibly managed, including the sourcing of fisheries raw material, and is being increasingly adopted as a step in a certified supply chain via an associated chain of custody. Increasing adoption and implementation of these standards will further improve environmental sustainability, but could also impact negatively on the available supply of fishmeal and fish oil for salmon farming, unless standards development takes account of scientific and economic criteria rather than political dogma. 5.3 Demonstrating responsible sourcing of plant feed ingredients The Roundtable on Responsible Soy (RTRS) Association approved Version 1.0 of its RTRS Standard for Responsible Soy Production versions in 2010 ( The new ProTerra Foundation Standard Version becomes effective in January 2015 and includes soya ( Cert ID is focused on Non- GMO certification and has become the benchmark for Non-GMO identity preservation and is now on Version 5.1 ( The use of these three (and similar schemes) in plant ingredient sourcing for fish feed, primarily for soya, is still seen as work in progress. Given the growing recognition of potential environmental problems (e.g. rain forest degradation) associated with uncontrolled soya production, especially in South America, it seems illogical for environmental NGOs to have focused on use of marine ingredients in aquaculture and to have largely ignored plant ingredients until recently. 5.4 UK supply chain perspective (salmon farming and feed manufacture) Unlike Norway, UK salmon production is strongly influenced by retail supply contracts with much bespoke production (noting that most salmon is grown under contract to a particular specification, but a significant proportion of such fish will not necessarily be sold on that basis by the retailer in question). Appendix 2 summarises the salmon farm and feed-related certifications used currently by Scottish salmon farms and their associated supply chains. All but one farming company (Grieg Seafood Hjaltland) are members of the Scottish Salmon Producers Organisation (SSPO) and subscribe to the COGPSA. Label Rouge ( has a premium trademarked French salmon label; despite being a niche export market from Scotland, it is currently undergoing a mini-revival. Small quantities of organic salmon are also produced. Widespread adoption of the Freedom Food certification, with 60 % - 70 % Scottish production represented ( is an international differentiator and goes beyond welfare in its responsible practice provisions to advocate sustainable feed and in requiring IFFO RS certification of marine ingredients. The most widely (but not exclusively) used Scottish salmon farming certification is GlobalGap. The extent to which this continues will partly depend on the success of a new version being developed which incorporates a more balanced approach to welfare, the 24

27 environment and food safety. GlobalGap is focused on food safety and has its own feed assurance scheme, which is covered off by UFAS in the case of feed manufacturers. The ASC salmon standard was launched in 2012 and has received global endorsement by Marine Harvest, which has now committed to be 100 % ASC certified by This trend seems likely to be followed by other salmon producers belonging to the Global Salmon Initiative of leading global producers ( which is concentrating its efforts on three areas, including feed sourcing. During late 2014 two Marine Harvest Scottish salmon farms acquired ASC certification. The ASC salmon feed module is currently in development and will need to modify some current provisions to enable widespread adoption in the UK. For example, it includes a requirement for fishmeal and fish oil to come from MSC certified fisheries by 2017 (or an ISEAL-compliant fishery standard), whereas it seems most unlikely that significant quantities of such material will be available by then. Although IFFO RS certification is near-universal as regards marine ingredient use in Scottish salmon, it is currently only accepted in a chain of custody role by the ASC salmon standard (although it is acceptable in other species standards, such as Pangasius), while the global leading BAP certification is more inclusive, but not currently adopted by any Scottish producers. Going forward there is a need for greater harmonisation between different standards; the Global Sustainable Seafood Initiative plans to benchmark these in order to facilitate sourcing decisions ( This could result in greater reciprocal recognition based on the different roles individual schemes can play and lead potentially to a Russian doll approach to aquaculture/feed certification ((Andrew Jackson, personal communication), depending on supply chain priorities (e.g. MSC addresses ecosystem management; Sustainable Fishing Partnership also seeks to do this via its scoring matrix; GlobalGap address food safety; IFFO RS certifies fishmeal factories and the downstream chain of custody, etc.). It is clear that sustainability issues are a strategic concern for the owners of the three principal UK salmon feed companies and lie at the heart of the day-to-day focus of their UK feed businesses. Increasingly major salmon purchasing companies oblige their supply chain partners to provide evidence demonstrating responsible management and use of renewable resources. Skretting ( has a SEA programme for developing sustainable feed solutions for aquaculture that takes account of long term environmental considerations, with the ambition to significantly reduce our own environmental footprint based on use of sustainable raw materials, developing sustainable nutritional solutions, and an ambitious commitment to reducing the carbon footprint of its operations (50 % reduction from 2009 by 2015). In this context a sustainable nutritional solution is defined as a nutritional solution that addresses a major sustainability issue within the industry through optimising the ingredients, the production, and/or the effects on business, animals, people and planet (economy, society and/or ecology). Their Sustainable Procurement Strategy has 25

28 the objective of moving Skretting to purchase only demonstrably sustainable raw materials. For BioMar ( their approach to sustainability entails that the industry is run on a commercial basis which meets the needs of the present without compromising the needs of the future and a specific development program (Bio-Sustain) has the focus on increasing sustainability of fish farming. EWOS ( ) are committed to transparent sustainability reporting; an independently reviewed Global Reporting Initiative (GRI) level B+ report details strategy, profile, management approach, and performance on all aspects of sustainability: economic, social and environmental. 5.5 UK supply chain perspective (salmon processors and retailers) In order to ascertain the positions on salmon sourcing of leading multiple retailers and processors of salmon in the UK, the following were contacted (and all replied at direct personal interviews, telephone discussions, correspondence, or a combination of the above): Aldi, Asda, the Co-operative, Marks & Spencer, Morrison s, Sainsbury s, Tesco, and Waitrose. Young s Seafood and the Icelandic Seachill group were consulted as leading UK seafood processors and distributors, with a strong presence in the salmon supply chain. The overall picture of UK retail salmon standards gained is that of a complex and differentiated set of competing codes aimed at increasing the retailers level of control over the supply chain. With two exceptions, each major retailer consulted has its own specific code of practice. Sainsbury s value campaign has three categories ( Basics, By Sainsbury s and Taste the Difference ), but has no boundaries on ethical issues and therefore does not apply tiering to welfare, sustainability, or any other aspect of sourcing policy. Thus Sainsbury s specification for all its farmed salmon requires that it be responsibly sourced, with IFFO RS as the default position for marine ingredients, but also requiring Freedom Food certification, compliance with COGPSA, GlobalGap certification and ISO certification for all tiers, while in addition their in-house specification covers requirements for omega-3, vitamins, pigment, etc. By comparison, Tesco has two salmon tiers ( Standard and Finest ) both supported by in-house codes of good practice (COPs). Marks & Spencer s salmon COP includes the specification for Lochmuir salmon, requiring that all marine ingredients are certified, either to the IFFO RS or the MSC standard, with the farms being audited for compliance against both the COP and Freedom Food certification, while their Organic Smoked Salmon meets the requirements of EU-recognised organic accreditation bodies. Whereas Appendix 1 gives the overview on salmon supply chain certification practice, (excluding the salmon feed issues of omega-3 content, GM feed ingredients and use of land animal products, which are discussed elsewhere in this report), other points of individual processor and retailer difference on sustainability and feed areas include the following: Sainsbury s: Fishery assessed against their sustainability rating system with IFFO RS or MSC as a default green on the decision tree for fishmeal and fish oil (and with IFFO RS as a minimum); requirement for GlobalGap and Freedom Food certification. 26

29 Morrison s: Aquaculture sourcing policy; risk access via decision tree, e.g. feed suppliers to comply with GAA, BAP, ASC marine targets; IFFO RS requirement for sourcing marine ingredients for land animal and fish farming by 2020, whereas their salmon feed suppliers already buy only from IFFO RS-certified plants. Waitrose: Decision tree has aquaculture certification scheme with focus on Waitrose Aquaculture Points of Difference (POD s) used with its responsible fish and shellfish policy, requiring GlobalGap certification and Freedom Food of salmon farms, etc. Wild caught fish used to provide fish feed ingredients are IFFO RS certified. No LAPs permitted in fish feed. Vegetable protein ingredients only from non-gm sources at levels with no prejudice to fish welfare, eating quality. The Co-operative: Responsible Farmed Fish Standard defines preferred feed status as IFFO RS (if unavailable refer to Fish Forum); special interest in social/ethical concerns; Co-op s fresh and smoked salmon is all Scottish and is all Freedom Food certified. Asda: is a strong supporter of farm assurance schemes (BAP or GlobalGap). Tesco: Requires all salmon producers to satisfy Tesco aquaculture requirements, including Tesco specific requirements and GlobalGap certification; Finest tier has Freedom Food in addition to this. Feed producers are assessed against UFAS and Tesco requirements, including sustainability of feed ingredients. Aldi: aims for Scottish provenance (with Freedom Food certification) in the case of farmed salmon supply within Scotland, but is less prescriptive for sales elsewhere. Salmon processors: Icelandic Seachill and Young s Seafood both rely on GlobalGap certification and Young s Seafood has a wild capture fishery decision tree. As regards sourcing of fishmeal and fish oil, the following requirements appear to be common for the processors and retailers interviewed: Traceability to species and country of origin No endangered species used, as defined by the International Union for Conservation of Nature (IUCN) Red List Preference for feed manufacturers to provide evidence of responsible sourcing Avoidance of illegal, unreported and unregulated (IUU) fish As regards product labelling, the Sustainable Seafood Coalition (SSC) codes of conduct on fish sourcing and environmental claims ( are intended to bring consistency, but they are somewhat generalised and of less relevance to salmon retailers, which are largely compliant (i.e. can claim their farmed salmon is responsibly sourced ). The SSC focus is on risk assessment and traceability, which allows retailers to have their own standards; in practice retailers buy from farm suppliers, which not only meet their own retail codes of practice, but also meet the independent third party standards. 27

30 In conclusion, UK supply chains aquaculture focus on sustainable feed issues is partly driven by competing UK retailers, whose policies can in turn be heavily influenced by pressure from environmental NGOs, which themselves are often competing for support from consumers and supply chains. It is claimed that the existence of different private standards in the marketplace can cause confusion, especially to retail fish buyers. As regards consumers, the evidence is that such standards are poorly understood, but the challenge is to provide consistent and trusted information to consumers that helps foster sustainability throughout the production and supply chain. To the extent that sustainability issues are being used as a tool of competition by allowing firms to underpin their responsible sourcing credentials, standards for salmon farming and feed are continuing to be driven up, but this has implications in restricting choice among different raw material sources in a manner that may not always be rational in terms of overall eco-efficiency. Summary of Chapter 5 There has been concern about the sustainability of salmon farming and feed and in this context it is relevant to consider social, economic, and environmental sustainability. In turn this has led to intense focus on the use of certification against third party standards to demonstrate responsible behaviour, as detailed for different links in the farmed salmon supply chain. This sustainability focus is partly driven by competition between retailers, including profiling values by leaders and followers, which has helped to drive improvement in feed sourcing, use of sustainable feed ingredients, and responsible behaviour. At the same time continuing standards development needs to address the key criteria and avoid becoming unnecessarily restrictive. 28

31 6. Are the criticisms of salmon feed sustainability justified? 6.1 Overview of the key criticisms and their validity today There are a number of concerns and misunderstandings about the use of marine ingredients in aquaculture, especially salmon farming; these have been critically reviewed by Shepherd (2013) and by Shepherd and Little (2014); they include the following: (i) (ii) (iii) (iv) Salmon are strictly carnivorous fish and it is a wasteful use of scarce marine resources to farm them. This view has been discredited by the success of R&D in developing cost-effective commercial diets mostly comprised of plant proteins and oils. Since fishmeal and fish oil can be largely substituted in salmon feeds, market forces (as influenced by cost, availability, certification status etc.) have resulted in the majority use of non-marine raw materials as feed ingredients for what appear to be omnivorous rather than strictly carnivorous fish. It is claimed there is a fishmeal trap meaning that use of wild ( forage ) fish in aquaculture feed leads to unsustainable damage as such stocks are harvested for fishmeal and fish oil, with the result that reduction fisheries will therefore inevitably become overfished (Naylor et al., 2009). However, with the exception of so-called trash fishing in Asia (which does not supply the salmon farming industry), this has not occurred due to: dietary innovation, increased use of fish processing trimmings, and to the increasingly responsible management of reduction fisheries (e.g. Shepherd and Little, 2014). At the same time, as already noted in Fig. 4, rapid growth in fed aquaculture has not been restricted by the static supply of marine ingredients. The suggestion that reduction fisheries are poorly managed is not borne out by the available information, as detailed in 3.2. To be sustainable, all fisheries have to be well managed and regulated (FAO, 2005; Worm et al., 2009), as they have been shown to be for most feed-grade species in the Americas and North West Europe, which are the usual origins of fishmeal and fish oil used in salmon farming in Scotland and Norway. It has been argued that aquaculture production is consuming large amounts of pelagic fish for feed that could have been used as human food and the salmon industry is therefore reducing the amount of marine protein available for human consumption (Naylor et al., 2000, 2009; Naylor and Burke, 2005). However, fish routinely used for reduction are either unfit for direct human consumption or there is no demand in practice for the fish in question (see 6.5). At the same time, Tacon and Metian (2008) have claimed that using fish landed by industrial fisheries in the Americas and Europe as aquaculture feed in the long run significantly expands the effective supply of fish for human consumption to an estimated net increment of 11 million tonnes per annum. 29

32 (v) (vi) (vii) (viii) Some critics believe that all fish should be processed for human food rather than for terrestrial or aquatic livestock feed. When it is argued that there is little or no consumer demand for forage fish species, the reply has sometimes been that such fish should be given to the poor free of charge, whereas experience from Peru, for example, is that local communities show limited interest unless they are processed into added value product at additional expense. As regards the ethics of using Peruvian anchovy for reduction purposes, FAO/NACA (2012) reported that if 8 million tonnes (for which there is a lack of effective domestic demand) were processed and supplied instead to distant consumers as a canned product, the annual cost would be in the order of US$25 billion per year and such subsidies might well be challenged under World Trade Organisation rules. As regards the view that forage fish are more valuable in the water than in the net, this ignores the conversion ratio in the wild which is of the order of 10 kg of prey to 1 kg of food fish, whereas the aquaculture option is far more productive. Ytrestoyl et al. (2011) showed that harvesting fish higher in the marine food web (e.g. cod) is far less efficient in providing marine nutrients for human consumption compared to harvesting capelin (on which cod predate) and producing capelin meal and oil for salmon production, as detailed later in 6.2. As pointed out by Welch et al. (2010) and Bibus (2015), high trophic level predators may require hundreds of times more marine resources than lower trophic level fish, such as menhaden or anchovy used for reduction. Hence the importance of taking into account primary productivity when comparing the eco-efficiency of reduction fishing with fishing for direct human consumption of higher trophic species. It has been said that use of fishmeal and fish oil is unsustainable on public health grounds as it results in the concentration of marine contaminants, which can then enter the food chain via aquaculture products. Today s situation is that statutory regulatory controls require constant scrutiny of contaminant levels and in practice farmed salmon generally have lower contaminants levels than wild salmon, at least in Europe (EFSA, 2012) as detailed in 9.1. It is claimed that conventional management of low trophic level fisheries, such as those for small pelagic fish used in reduction, can be risky because it does not adequately account for their wide population swings and high catchability, nor does it include the critical role of forage fish as food for marine mammals, birds and predator fish (e.g. Lenfest, 2012). Although small pelagic populations do fluctuate widely and are easily caught and reduced, recoverability is equally important, as evidenced by the predictably sharp rebound in the Peruvian anchovy fishery following an El Niño event. Rather than questioning the harvesting of low trophic level species (e.g. in the case of MSC s apparent shift in policy on Peruvian anchovy (see 6.6 (viii)), Garcia et al. (2012) have argued for 30

33 (ix) balanced harvesting in order to achieve better fisheries and conservation objectives. Salmon farming in Scotland continues to be subject to criticism for its perceived impact on the environment, despite the fact that the regulations in place are probably the most stringent in the world (Scott, 2010) and also there is voluntary regulation via the COGPSA. A scientific review of the environmental impacts of aquaculture (SAMS, 2002) concluded that salmon farming in Scotland has no significant impact on nutrient enrichment except in extreme cases where there is poor water exchange. Since the SAMS report was published, the position has improved still further due to the use of improved regulatory models and to the near-universal adoption of management and zoning agreements between adjacent sites. 6.2 Eco-efficiency, resource use and comparison with land animal production Compared with warm-blooded terrestrial livestock, fish are better suited for farming, being ectotherms with neutral buoyancy in water, and having lower maintenance and respiratory costs. They are therefore more efficient converters of feed energy to bodyweight. Compared with wild fish, farmed fish show improved growth and nutrient retention since they are protected from predators and need less energy to access food (Smith, 1992; Bergheim and Åsgård, 1996). Also species like salmon have a shorter digestive tract than strictly vegetarian species and this means a greater proportion of the carcase is edible. But the fundamental questions are how farmed salmon compares to other food production systems and what are the prospects for its improved sustainability. Table 3 is a comparison of product yield, energy, and protein retention in the edible parts of Atlantic salmon, pig, chicken and lamb (Bjørkli, 2002). Not only are they superior in terms of harvest and edible yield, but salmon retain the most protein (31 %) relative to pig, chicken and lamb; salmon also retain the most energy (23 %) in their edible parts. The feed conversion data clearly show salmon to be the most efficient converters at about 1.15, compared with land animals. In comparison wild salmon are among the least efficient converters with an FCR of around 10 (Marine Harvest, 2012); adjusted to make this comparison valid on a like-for-like dry weight basis, the FCR of wild salmon is around Hognes et al. (2011) analysed a comprehensive dataset from Norwegian farmed salmon in 2010 and showed that salmon s carbon footprint was 2.6 CO 2 e/kg edible product, compared with values of 3.4 and 3.9 for Swedish chicken and pig, respectively, and they also outperform land animals in terms of efficiency of land and freshwater usage, and the use of non-renewable phosphorus resources etc. This is reinforced by Welch et al. (2010), who found that farming salmon increased the production of animal protein at generally lower costs in terms of land, water, nitrogen, and agricultural chemical costs than terrestrial livestock. In addition Ytrestøyl et al. (2011) showed that salmon farming is a far more efficient way of producing nutrients for human consumption (e.g. EPA and DHA) than the 31

34 most efficient land animal farming systems, since salmon retain nutrients more efficiently and also convert feed nutrients to nutrients for human consumption more efficiently. In fact it has been demonstrated that Atlantic salmon can be a net producer of both marine protein (Crampton et al., 2010; Bendiksen et al., 2011) and of n-3 LC-PUFA (Sanden et al., 2011; Turchini et al., 2011), although for the latter it should be stressed that the levels of n-3 LC- PUFA in the products are far lower than those in fish fed fish oil. This reinforces the fundamentals of salmon farming and it is of interest that theoretical calculations by Ytrestøyl et al. (2011) indicate that harvesting the capelin resource to produce fishmeal and fish oil to feed farmed salmon gave nearly 10 times more marine protein, 15 times more energy and 6 times more essential omega-3 LC-PUFA (including cod liver oil) compared to harvesting the cod resource, which would otherwise predate on capelin. 6.3 Marine resource use and input/output indices Although the main driver for replacement of marine ingredients by plant raw materials has been the combination of nutritional innovation and market forces in the face of fluctuating supply and highly volatile price, replacement has in turn impacted on the different indicators used to relate marine ingredient input to farmed salmon output. These marine sustainability indicators include fish-in fish-out ratio (FIFO), marine protein dependency ratio (MPDR), forage fish dependency ratio (FFDR), etc. Although analyses of FIFO ratios to assess the use of marine ingredients in aquaculture production showed the ratio for all fed species combined has decreased as aquaculture developed (from 1.0 in 1995 to about 0.7 in 2006), with a further projected decrease towards 0.2 by 2020 (Tacon and Metian, 2008, 2009), whereas FIFO has been used by environmental and anti-fish farming groups to imply that aquaculture, particularly salmon farming, was an inefficient practice wasting precious marine resources; this interpretation has been criticised by Byelashov and Griffin (2014) and by Bibus (2015). The flawed methodology for calculating FIFO (Kaushik and Troell, 2010) and the initial high values of 4.1 or -5:1 for salmon farming (Tacon and Metian, 2008) were challenged by Jackson (2009), who proposed instead a mass balance approach. Crampton et al. (2010) have suggested that the FIFO concept should rather be replaced by marine nutrient dependency ratios for which the amount of each marine nutrient used in feed is divided by the amount of that nutrient in the farmed product, producing separate ratios for protein (MPDR) and oil (MODR). Irrespective of how the FIFO ratio is calculated, FIFO is not an indicator of sustainable use of marine resources because sustainability must be based on responsible harvesting of fish that are used for fishmeal and fish oil according to international regulations (backed up by private standards if so desired). In any event the FIFO ratio has no nutritional basis and is therefore in no way a measure of production efficiency. ASC has included the FFDR as one of its indicators, although again it cannot be claimed that this is an indicator of sustainability. FFDR calculates the quantity of forage fish required to produce the amount of fishmeal and fish oil to produce a unit of farmed fish. For 2013 the 32

35 corresponding FFDRs for the Norwegian salmon industry were 1.54 and 0.69 for fish oil and meal, respectively, which are well below the ASC standards of < 2.95 for fish oil and < 1.35 for fishmeal for forage fish dependency in regard to both fishmeal and fish oil (Ytrestøyl et al, 2014); by comparison the equivalent (unpublished) values for one Scottish producer in 2013 of 2.09 for fish oil and 0.76 for fishmeal were higher than Norway, but well below the ASC minimum values. 6.4 Effect of changing raw materials on salmon farming sustainability From the standpoint of salmon nutrition, it is important to recognise that it is not possible to change the amino acid composition of salmon, while the fatty acid composition can readily be changed; this has implications for sustainability as currently there are no alternative sources of EPA and DHA to fish oil, while there are many alternatives for amino acids. In other words salmon protein content is largely fixed as a function of genetic makeup, while its fat content is essentially a storage concept lacking functionality. As regards protein retention, Torstensen et al. (2008) and Bendiksen et al. (2011) have found similar levels in Atlantic salmon fed either mainly marine diets or mainly plant-based diets, explaining why salmon can be produced with feeds containing high inclusions of plant ingredients and only low inclusions of marine ingredients. Calculations by Nofima have shown that changing the salmon diet composition from 85 % plant ingredients to 88 % marine ingredients resulted in almost the same carbon footprint (unless poultry by-products are used when it will increase close to the carbon footprint of Swedish chicken due to allocating the carbon footprint from poultry production to that of its by-products according to their mass), while excluding marine ingredients from South America and the Mexican Gulf from the 2010 diet increased the carbon footprint by 7 % (Ytrestøyl et al., 2011). These observations support the view of Torrissen et al. (2011) that increased utilisation of plant ingredients may not be as sustainable as often claimed. At the same time it is important to recognise the increased use of fishery by-products (25 % of fishmeal and 27 % of fish oil in Norway by 2013), although such material has little, if any, alternative use, the effect on contribution to carbon footprint of the salmon diet will depend on whether the life cycle analysis (LCA) calculation adopts economic or mass allocation. Economic allocation looks at the relative value of inputs and outputs, thus using cod viscera with zero opportunity cost to feed salmon of high value is clearly a rational resource use; by contrast mass allocation would penalise use of cod viscera by arguing that the major proportion of the carbon footprint of cod is due to the weight of its trimmings, hence if the latter is used as feed for salmon, the carbon footprint of the cod viscera should be reflected in the farmed salmon s footprint. Micronutrients also play an important role, for example the use of bacteriallyfermented pigment sources has been adopted by the Scottish industry for marketing reasons as it is perceived to be more natural than synthetic astaxanthin mainly used in Norway; however, bacterial fermentation uses sugar and is thus energy demanding with negative implications for the environmental footprint. 33

36 6.5 Economic, social and ethical issues In narrow economic terms, Scottish salmon farming appears a profitable and sustainable industry, which also confers significant social benefits, especially in generating wealth and employment within remote rural communities. Although a comprehensive and up-to-date cost benefit analysis is not available, the economic benefits of salmon farming to the Scottish economy are substantial in terms of added value and jobs across the whole supply chain (Marine Scotland, 2014). Social and ethical issues around aquaculture mainly concern the fishing industry and touch on feed supply issues. Increasingly processors subscribe to the exchange of ethical data via membership of Sedex ( Also the Responsible Fishing Scheme is currently being strengthened and extended to cover safety and welfare issues; this scheme may in the future allow linkage with responsible sourcing certification for aquaculture feeds detailed above ( The use of natural resources potentially raises social, ethical and economic questions. According to FAO (2014), 14 % of world fish production in 2012 was destined for non-food uses of which 75 % was reduced to fishmeal and fish oil. Although the FAO encourages using more fish directly for human consumption, they are of the opinion that it is more efficient in a protein-hungry world to harvest the unmarketable species for animal and fish feed, subsequently consumed by man, than not to harvest the fish at all. Consumers are not ordinarily being denied direct fish consumption as a result of fishmeal and fish oil production and their use in salmon diets results in net edible production of fish protein today. Hence it is justifiable to claim that the use of fishmeal and fish oil from reduction fisheries and fishery by-products is rational in economic terms and an ethically acceptable use of resources, provided they are sustainably managed. To go further and say that in any event such use should be minimised only makes sense if using alternative non-marine materials is a more efficient use of resources; this is not necessarily the case for plant materials when relative carbon footprint, land use etc. are taken into account. However, in making such comparisons it should be noted that analytical problems can be important, such as giving appropriate credit for the increased use of recycled materials (e.g. fishery byproducts; see 6.4) and making valid judgements on the use of land animal by-products, given continuing academic debate about the use of economic versus mass allocation in LCA. 6.6 Conclusions and implications for salmon feed sustainability (i) Salmon farming is a very efficient way of producing nutrients for human consumption and inherently more efficient than the most efficient land animal production systems, with clearly superior food conversion efficiency and nutrient retention. At the same time it has a high production yield of edible meat and 34

37 (ii) (iii) (iv) there is a large demand for its rendered by-products. It follows that using fishmeal for salmon farming is a more efficient allocation of this resource than using 26 % of global fishmeal for feeding pigs or chickens, as took place in Analysis of carbon footprint and nutrient flows from Norwegian salmon farming reinforces its overall sustainability. It is reasonable to draw the same general conclusions for Scottish salmon farming, although the existence of bespoke diets in Scotland, the recent trend to high performance diets, and its use of energyintensive fermented bacterial pigments, all tend to complicate like-for-like comparison with Norway. Although life cycle analysis suggests salmon farming and cod fishing are comparable, Nofima found catching cod prey species (e.g. capelin) to use as salmon feed ingredients is a far more efficient way of supplying nutrients for human consumption than leaving such prey in the sea and harvesting the resulting cod; thus using capelin to produce farmed salmon gave nearly 10 times more marine protein and more than 6 times more omega-3 LC-PUFA compared with leaving them in the sea as prey for cod and then harvesting the wild cod resource (Ytrestøyl et al., 2014). Much of the criticism of salmon farming sustainability is out-of-date or misses the key feed-related issues. Firstly there has been a radical change in ingredient use during the last 10 years with marine ingredients being largely replaced by vegetable ingredients. But the environmental movement is dominated by marine conservation organisations which pressure fish supply chains to continue lowering the use of marine ingredients for aquaculture without considering sustainability issues in regard to alternative ingredients (e.g. rain forest impacts of increased soya cultivation in Brazil). The sustainability of substituting marine ingredients with plant material should be assessed thoroughly since production of plant ingredients requires freshwater, fertilizers, phosphorus, pesticides, land area, and it contributes to soil depletion; since most plant ingredients can be used for human consumption, the benefits of substituting marine ingredients produced from well managed fisheries are not obvious (Ytrestøyl et al., 2014). Considering past trends and current predictions, the sustainability of the aquaculture sector is more likely to be closely linked to the sustained supply of terrestrial and plant ingredients for aquaculture feeds than to marine ingredients (FAO, 2012). The real issue with marine ingredients, especially with fish oil, is supply and demand. What was not sustainable in the long term was the increasing use of high volumes of finite resources (especially fish oil), which is not to say that fishmeal and fish oil are themselves inherently unsustainable resources (Jackson and Shepherd, 2012). Although fishmeal and fish oil supplies to the Scottish salmon feed industry are required by the supply chains to show that they are 35

38 (v) (vi) (vii) responsibly managed and renewable, these resources will always be finite and limiting, hence require careful stewardship. This has in turn given rise to the use of marine indicators in an attempt to score feed sustainability and such indicators can be useful as a means to judge the extent of reliance of the production system on what are finite, scarce and fluctuating resources. Irrespective of what method is used, ratios such as FIFO and FFDR are not indicators of sustainable use of marine resources, since sustainability must be based on responsibly managed fish stocks according to international fishery regulations, backed up, where appropriate, by use of third party certification to help manage the sourcing choices to be made. Evidence of responsible salmon farming practice and sourcing of feed ingredients is now a consistent requirement by UK supply chains and acts as a competitive tool in the salmon market. As a result retail market behaviour is helping to drive improvement by means of certification standards specified in supply contracts, with special focus on the provenance of marine ingredients. Thus a requirement that fishmeal be sourced from IFFO RS certified factories is now becoming universal for Scottish salmon feed, although use of MSC certification as a more detailed indicator of the sustainability of forage fisheries is not currently practicable due to the lack of such certified material. Although there is growing interest in ASC certification of salmon farming in Norway and Scotland, this remains a work in progress pending clarification about the future feed standard rules (e.g. eligibility of other standards). There is a clear risk that future certification requirements, which are partly driven by NGO agendas, will become over-restrictive so as to severely limit sourcing options and reduce further the flexibility of feed formulation, hence increasing costs, without a commensurate improvement in sustainability. The available evidence supports the view that the forage fisheries providing certified material to the salmon farming industry in Scotland and Norway are responsibly managed. Although the largest such fishery for Peruvian anchovy is subject to periodic reduction linked to El Niño events, prompt government intervention by Peru minimises the risk of overfishing. In general a more precautionary approach by state regulators, coupled with adoption of third party certification schemes, is enabling marine feed ingredients to be sourced sustainably. The overall result has been to curb the availability of fishmeal and fish oil from forage fisheries. This being the case, short-term interruptions in supply due to El Niño-related events (as described in 7.4 for fish oil) demonstrate that management measures are in place and working properly to avoid overexploitation of the fishery resource. Longer term they place much greater focus on the use of by-products from fish processed for human consumption and on the use of plant sources of protein and oil. At the same time certification 36

39 (viii) (ix) schemes are now starting to recognise that alternative plant ingredients also pose sustainability issues and need to be sourced responsibly. As regards the question of single species stocks versus ecosystem assessments, there is potential concern about how to include ecosystem effects in fish stock evaluations and it is true that striking an appropriate balance between seabird or marine mammal stocks and pelagic fish stocks implies making a similar judgement as between food security and biodiversity (c.f. the set-aside problem in agriculture). Striking the right balance between the level of reduction catch and leaving fish in the water for predatory fish, birds and mammals is as much down to subjective judgement as to scientific method. In this connection the unforeseen introduction of new tighter MSC rules on assessing low trophic level fisheries (immediately following a successful pre-assessment of the Peruvian anchovy fishery) caused dismay to the Peruvian fisheries society and suspicion as to the willingness of MSC to certify forage fisheries on principle. At the same time it is probably true that most types of fishing, either for human consumption or for reduction purposes, would benefit from adopting more ecosystem management. Alongside a picture of growing environmental sustainability, the Scottish salmon farming industry is able to point to clear social and economic benefits, including employment and wealth creation in remote regions. Provided that alternative sources of omega-3 LC-PUFA can be developed, there is no reason to believe that the industry cannot continue to develop sustainably. Nor is there justification for criticising Scottish salmon farming on feed grounds today. Summary of Chapter 6 Criticisms of the sustainability of salmon feed have been largely overtaken by events and are unjustified today. For example marine ingredients for salmon feed are derived from demonstrably well-managed reduction fisheries and use of the resulting fishmeal and fish oil is not increasing, despite the continuing growth in aquaculture, because of large scale substitution by plant ingredients. In terms of eco-efficiency and resource use, Norwegian data show that salmon farming is clearly superior to land animal farming, with a lower carbon footprint, and is an extremely efficient way of producing nutrients for human consumption. Also feeding forage fish to salmon instead of leaving them for cod to consume is far superior in energy flow terms than harvesting cod and leaving forage fish in the sea without harvesting. As regards the effect on sustainability of moving to predominantly plant-based diets, the current marine sustainability indicators have limited relevance; nor will this dietary change necessarily result in more sustainable feed as plant ingredients have their own resource issues. While it is recognised that marine resources are finite and forage fisheries need responsible management to be renewable, the longer 37

40 term sustainability of aquaculture is more likely to be linked to the sustained supply of terrestrial and plant ingredients than marine feed ingredients. However, new omega-3 sources will need developing to ensure long term sustainability and this prospect is now in sight. Diet innovations and supply chain focus mean that Scottish salmon feed is continuing along a sustainable development path with clear evidence of responsible sourcing of feed ingredients. Not only is the Scottish salmon industry demonstrating environmental responsibility, but clear evidence of a positive social and economic profile contributes to its sustainable development. 38

41 7. The issues surrounding omega-3 LC-PUFA in farmed salmon 7.1 Public perception Ozonaka et al. (2012) surveyed the position of farmed salmon in five European countries (including the UK) and found that the relative rating of quality dimensions of salmon compared to meat categories showed some common patterns, specifically that salmon is regarded as superior in healthiness compared to other meats. This reinforces the feedback from certain UK retailers that some of their customers take healthiness of salmon into account in their purchasing decisions, suggesting in turn that negative publicity on the health benefits of farmed salmon might reduce demand. Public awareness of omega-3 issues is promoted by retailers, human nutrition companies and also by Seafish, which has introduced a fully integrated omega-3 campaign that was launched in the first week of February Elements of the Seafish campaign include advertising in digital and print publications, tastings through street food venders and supermarket sampling, as well as sharing campaign messages and cookery tips through peer-to-peer marketing with family and health bloggers; they claim to have had a reach of 37.2 million people, with Facebook engagement up by 200 % and Twitter up by 50 %, as detailed in At the same time it should be noted that the health benefits of farmed salmon consumption are not restricted to its content of omega-3 LC-PUFA. Salmon comprises highly digestible protein and essential amino acids and marine lipids, as well as vitamins and minerals, etc. It therefore follows that consumption of salmon is more nutritious compared with consuming omega-3 supplements alone. 7.2 Recommendations for human intake Recommendations on the dietary intake of omega-3 LC-PUFA for humans are almost exclusively based on evidence of their beneficial effects on cardiac health and cardiovascular disease. However, there is a vast literature on the effects of dietary n-3 LC-PUFA in many other pathological conditions and in neural development, as well as considerable physiological, biochemical and molecular data on the mechanisms whereby EPA and DHA may elicit their effects. A brief summary of the current state-of-the-art regarding current understanding of dietary omega-3 LC-PUFA and human health is provided in Annex B. In terms of permissible European label claims, it is necessary to distinguish between health and nutritional claims (separate but inter-connected), and to recognise that the detailed regulatory framework is changing over time UK and international recommendations Current UK dietary recommendations for omega-3 LC-PUFAs were proposed in 1994 by the UK Committee on Medical Aspects of Food Policy (COMA), which was disbanded in 2000 and 39

42 then replaced by the Scientific Advisory Committee on Nutrition (SACN), ( In 2004 SACN then revised the recommendations as follows: their recommendation is to eat at least two portions (with a portion being 140 g) of fish per week, one of which should be oily fish (with two portions of fish per week, one white and one oily, providing approximately 450 mg per day of omega-3 LC-PUFA), including for pregnant women. The EU claims for EPA and DHA are summarised in Appendix 3. The recommendation for consumption of EPA + DHA is a daily intake of 250 mg (as per the sources listed in the Appendix to Regulation (EC) 1924/2006). In 2010 EFSA published its recommendation that a daily intake of 250 mg of omega-3 LC-PUFA for adults may reduce the risk of heart disease. Other international recommendations include the International Society for the Study of Fatty Acids and Lipids (ISSFAL, which recommends a minimum daily intake of 500 mg of combined EPA and DHA for healthy adults and for cardiovascular health; and the American Heart Association ( which recommends that everyone should eat at least two servings (c. 230 g) of oily fish (high in EPA and DHA) per week. In 2009 the Seafish Industry Authority of the UK published a useful guide to these health claims, including guides to consumption, the legislative framework, and interpretation (as per Implications of reducing marine content of salmon feed Shepherd and Bachis (2014) showed that, as a result of the combined EPA and DHA content of added oil in Norwegian feeds falling from 20 % in 2002 to 7.2 % in 2012, the number of days recommended requirement met by one portion of Norwegian salmon had fallen proportionately to approximately 6 days (EFSA at 250 mg/day), 3.4 days (SACN at 450 mg/day), and 3 days (ISSFAL at 500 mg/day). The average UK consumer eats less fish than in Norway and a survey carried out on behalf of Public Health England and the Food Standards Agency on diet and nutrition in the UK showed that, for oily fish, average consumption was well below the recommended one (140 g) portion per week in all age groups, with men and women aged 19 to 64 eating just 52 g and 54 g/week of oily fish, respectively ( As regards Norway, Ytrestøyl et al. (2014) found on average salmon fillets in 2013 contained 1.36 % of EPA and DHA, hence a consumer eating one (130 g) portion of farmed salmon/week would achieve the EFSA recommended intake of 1.75 g/week combined EPA and DHA. The Norwegian Scientific Committee for Food Safety (VKM, 2014) concluded that current concentrations of EPA, DHA, and DPA (docosapentaenoic acid, 22:5n-3) in farmed salmon are about 50 % of the corresponding levels in 2006 because of the further replacement of fish protein and fish oil with plant protein and vegetable oils in feed for farmed salmon, while the concentration 40

43 of vitamin D appears unchanged. The level of omega-6 fatty acids is about 4-fold higher than in Although the omega-3 LC-PUFA levels in farmed salmon have decreased, the VKM report concluded that, with current average consumption of fish (in Norway), the contribution of EPA and DHA from fish will reach the European recommended intake of EPA + DHA for adults and 2-year olds, but for pregnant women the average sum of EPA and DHA intake is not sufficient to meet the European recommendation for this group, although the average intake of DHA is just sufficient to meet the national intake recommendation for pregnant women. However, it will be shown that due to continuing replacement of fish oil in salmon feed, these results are no longer necessarily valid for farmed salmon being produced in Norway and the UK and they are likely to be compromised by global availability of EPA and DHA during In a recent study of different farmed and wild salmon products available in retail outlets in Scotland, Henriques et al. (2014) showed that, although wild salmon products had higher relative values of EPA + DHA, farmed salmon products generally delivered a higher dose of EPA + DHA compared to the wild salmon products, due to their higher lipid content, and were therefore better able to deliver recommended dietary intake levels than the wild salmon products. At the same time the study confirmed the elevated levels of omega-6 PUFA, specifically linoleic acid (18:2n-6, LOA) (whereas previously reported values were c. 10 % for farmed fish and usually under 3 % for farmed salmon); but it also showed that 18:2n-6 does not have a major impact on the nutritional quality of the farmed salmon and certainly does not outweigh the considerable benefits of the high doses of omega-3 LC- PUFA recorded. Also Byelashov (2014) has pointed out that, while the α-linolenic acid (18:3n-3, LNA) content of wild salmon is below 1 %, its current content in farmed salmon is only slightly below DHA and higher than that of EPA. In the short time since the Henriques study was performed, however, EPA + DHA levels in farmed salmon may have declined further and omega-6 levels may have increased further. It seems likely that this will have come to the attention of wild salmon suppliers (e.g. in Alaska) raising the possibility that it might be used by them to promote the health benefits of wild salmon consumption and undermine the equivalent claims for farmed salmon. 7.3 Impact of omega-3 differentiation by UK retailers Current range of omega-3 label claims and specifications There is clear segmentation in UK farmed salmon supply between those retailers willing to claim that one portion provides the weekly required intake, formulated to deliver high levels of healthy omega-3 long-chain fatty acids (or similar wording) and those simply stating variable omega-3 content (or similar wording). The labels of those claiming the former (so-called premium suppliers ) claim that consuming the following portions will deliver the recommended intake, as follows: Sainsbury s (130 g portion); Marks & Spencer ( Lochmuir Salmon : 100 g portion); Waitrose fillets (100 g portion); and Tesco ( Finest : 100 g portion). EFSA would appear to be the (minimum) target reference of choice (i.e g 41

44 EPA + DHA). Waitrose policy requires finished product to contain 1.75 g/100 g of EPA + DHA. However, the Marks & Spencer target content for Lochmuir Salmon is claimed to be 3.1 g of omega-3 LC PUFA in 110 g fillet (based on a 15 % fat content) and to deliver an adult s weekly recommended intake of EPA + DHA in one salmon portion. If this is so, Lochmuir Salmon is likely to have levels of omega-3 LC PUFA significantly higher than those other premium suppliers targeting the EFSA recommendation of 1.75 g /100 g of EPA + DHA. For these premium suppliers/products, there is a clear link being made between the health of the consumer and the content of omega-3 LC-PUFA in the product on offer and in general the premium suppliers specify Scottish farmed provenance. In other words, omega-3 has become something of a Unique Selling Point (USP) linked to Scottish production; with Marks & Spencer appearing to take a leading position in terms of the omega-3 LC PUFA content in its Lochmuir product, suggesting differentiation may exist even within the premium segment. As regards provenance, other farmed salmon retailers in the UK do not focus on Scottish sourcing by comparison with the above premium suppliers (except Aldi and Co-op in the case of Scottish sales), nor do they ordinarily claim any relation between portion size and intake requirements or recommendations. This differentiation focus is clearly used as a competitive tool, but does raise related questions, such as: For non-premium salmon, portion size is unlikely to be related to international recommendations on omega-3 LC-PUFA intake. The term omega-3 fatty acids may justifiably include more than EPA and DHA, e.g. DPA. In evaluating competing product claims, care must therefore be taken to ensure valid comparison on a like-for-like basis (e.g. using total LC-PUFA content would give higher flesh values when compared with content of only EPA + DHA). There is no indication whether premium salmon is formulated with an overage of omega-3 LC-PUFA to take account of the reduction in levels due to cooking the product (which will vary with cooking method and fat loss, etc.), while intake recommendation levels are based on an assumption of what is actually consumed. There is no reference to omega-6 fatty acid levels which have increased in farmed salmon due to vegetable oil inclusion levels The challenge to feed formulation The existence of a range of bespoke salmon products in response to retail market differentiation represents a complex supply chain challenge for farms and feed requirements; this is, of course, true for various aspects of feed and farming specification (e.g. welfare and husbandry), but is well illustrated by the challenge of supplying different levels of omega-3 LC-PUFA via the feed. For those products which aim to target meeting the EFSA recommendation, the relevant variables include: The level of EPA and DHA in the fish oil component of the added oil fraction 42

45 The bodyweight and fat level of the salmon The target flesh level of EPA and DHA in a given portion size The need to control how the fillet flesh level is sampled (e.g. by use of Scottish quality cut, SQC) The choice of laboratory method and analytical tolerances in measuring EPA and DHA Scottish feed manufacturers appear to use different models of how to add EPA and DHA in the diet to yield the required fillet fat level in harvested salmon depending on the circumstances. The preferred strategy is to source the highest level of EPA and DHA (usually Peruvian anchovy oil) and then to dilute down with rapeseed oil to achieve the required feed specification. As a rule of thumb, feed mills appear to add between 6.5 % and 8.5 % of EPA + DHA in the oil fraction of the feed to achieve 1.75 g in a 130 g fillet portion of salmon (EFSA target), depending on salmon bodyweight and fat level, or proportionately more assuming a 100 g portion. In other words, to achieve the target of 1.75 g in a fillet portion of 130 g (i.e g/100 g) requires on average 7.5 % of combined EPA and DHA in the oil fraction of the feed or around 10 % in the case of a 100 g portion. From supply chain discussions, it appears that the current range of omega-3 inclusion in the dietary oil in Scotland normally varies from around 6 % to 13 % but one farmer may add significantly more for a specific product. This is to meet specific salmon contract requirements laid down by retailers and can be largely explained by the premium suppliers seeking to meet the EFSA recommendations. The lower end of the Scottish farmed range of around 6 % is in the normal range of standard Norwegian practice during 2014 and is claimed to be typical of current UK imports of Norwegian salmon (supplying an estimated 60 % of the UK market). It follows that at these lower levels there is no certainty that the EFSA claim will be met consistently (unless consumers eat more salmon each week) and this lower cost range corresponds to the majority of UK (non-premium) salmon volume labelled on the basis of variable omega-3 levels. 7.4 Global availability of EPA and DHA Acute shortage in 2014 and 2015 Peru's industrial fleet was given low anchovy fishing quotas in 2014 and was then restricted further by intermittent fishing bans as the Government responded to El Niño-like conditions. According to the Peruvian Committee overseeing El Niño (ENFEN), the current climatic event is technically a succession of so-called Kelvin Waves and so far has not met their definition of a true El Niño, but for simplicity this report will refer to it as an El Niño. This caused the anchovy to move into deeper inaccessible waters, or to move south where industrial fishing is prohibited within 10 miles of the coast. Given Peru s global dominance of fishmeal and fish oil production, coupled with the high levels of EPA and DHA in Peruvian anchovy, this severely impacted on the global availability of EPA and DHA resulting in record price levels reaching close to US$ 3000/tonne for Peruvian fish oil. Although precise data are 43

46 not available, market estimates suggest 2014 global production was only 140,000 tonnes of EPA + DHA (from c. 800,000 tonnes of crude fish oil) compared with a normal year of up to a maximum of 200,000 tonnes. It also seems possible that stocks of EPA and DHA held in fish oil stores reduced by c. 30,000 tonnes during 2014, hence total consumption may have reached around 170,000 tonnes by the year-end. Fig. 18 (Holtermann, unpublished) estimates supply and demand for EPA and DHA in 2014 and assumes that this total consumption would equate to an inclusion rate in the oil fraction of global salmon feed of around 6 % or 50,000 tonnes; these are ballpark estimates assuming non-salmonid aquaculture also consumed 50,000 tonnes and human nutrition consumed 61,000 tonnes (with the balance consumed by technical uses, such as hardening, tanning, and energy). It should be noted that the model s sensitivity assumptions focus on catch variability rather than oil yield or EPA/DHA ratios, etc., whereas EPA and DHA levels will vary for a given fishery and do not have a fixed ratio for a given species. The model assumed 27 % EPA + DHA in Peruvian anchovy oil for 2014 and 28 % EPA + DHA in 2015 (closer to historic average values) and focused on the total sum of EPA + DHA for each species in each origin. In Peru EPA levels are normally higher in the north and DHA levels are higher in the south; also the oil content of whole anchovies is normally in the 3-6% range of bodyweight depending on nutritional status (average 4.5%), with EPA and DHA being around the 25% level in the oil, this would make the EPA and DHA content of whole anchovies around 1.1%, with lean fish tending to have higher levels. This variability further complicates projections for 2015 and beyond. In early December 2014, following a scientific survey, the Peruvian authorities confirmed there would be no quota for the second fishing season. At the time of writing it is impossible to estimate fish oil production in 2015 pending follow-up surveys in Peru, although the lack of a second quota and some current ENFEN forecasts are expected to have a positive effect on biomass recovery during 2015 ( Fig. 19 (Holtermann, unpublished) illustrates the sensitivity of global EPA and DHA availability to different Peruvian quotas. For instance, if the 2015 Peruvian anchovy quota is in the range of 3.0 to 3.5 million tonnes (compared with around 2.3 million tonnes in 2014) and the quotas are all caught, this could correspond to global supply of 150, ,000 tonnes of EPA + DHA, other things being equal. Under these circumstances, assuming no stock carryover is available from 2014 and the consumption by other segments remain unchanged, this corresponds to 30,000-37,000 tonnes for salmon feed (down from an estimated 50,000 tonnes in 2014) corresponding to an average of 4-5 % EPA + DHA content in the oil fraction of global salmon feed. In practice it is unlikely there will be no 2014 stock carry over, but on the other hand some market growth in demand can be expected. The potential implications of this situation for 2015 salmon feed consumption include the following: 44

47 Although unknowable right now due to the current uncertainty over the timing of the recovery in Peruvian anchovy fishing, it seems likely that the 2015 EPA and DHA available for salmon will be less than in 2014, possibly corresponding to 4-5 % of the oil on average in global salmon feed. In practice, depending on joint purchasing arrangements, Scandinavian feed companies might be willing to allocate more stock to their Scottish subsidiaries in order to allow them to meet contractual obligations. The buffer from fish oil being held in stock from 2014 is unknown, but thought to be lower than the 2013 year-end stocks carried over to With fish oil prices having recently reached record levels, approachingus$3,000/tonne, it is of no surprise that non-traditional sources are now appearing on world markets; these include Indian sardine (Sardinella spp.) oil for which there is weak domestic demand (biomass is estimated at c. 4 million tonnes with a 10 % oil yield and c. 24 % EPA and DHA), as well as oil from Chinese anchovy (Engraulis spp.) and from north-west Africa (e.g. Sardinella spp.) now being exported to Chile. High prices will tend to dampen demand from price sensitive segments. Other things being equal, demand growth for human nutrition and non-salmon aquaculture would be expected to be around 2.5 % - 5 % per annum; however, human nutrition growth has stalled since 2013 and fish oil inclusion in non-salmonid feeds can be expected to fall at current prices. In a worst case scenario, if the available EPA and DHA for global salmon feed falls below 4 % of the oil fraction (equivalent to c. 1.2 % of total feed), this then approaches the 1 % risk area for essential fatty acid deficiency in salmon and could predispose to health problems or reduced performance assuming no fish oil contribution from fishmeal present in the diet. Perhaps the critical question is how the largest global salmon producers respond to this scarcity situation. For example, there is a view some might forego perceived claims on human health benefits on the basis that consumers still have the option of eating salmon twice a week (if it now has only half the EPA and DHA content previously added to feed) in order to meet the EFSA intake target. It is claimed that standard average Norwegian feed inclusions have very recently reduced to 5 % of the oil fraction at which level they will already be insufficient to deliver the EFSA recommendation in one weekly portion. The logical extension of this policy is to dilute to 4 % in order to conserve stocks; if this floor level is then threatened, the only option would be to outbid other demand segments for the remaining oil, most likely at the expense of other aquaculture species. Even in the event of these more pessimistic supply scenarios being realised in 2015, it should be noted that feed manufacturers are likely to have bought forward cover to meet their EPA and DHA obligations for the premium segment of the Scottish salmon farming industry for a major proportion of their 2015 contract requirements; not that this necessarily guarantees supply. 45

48 7.4.2 Anticipated 2016 global recovery Historical precedent provides confidence of a likely strong rebound in Peruvian anchovy stocks following the current oceanic conditions, most probably starting in the first half year of 2015 as the high current biomass of juvenile anchovy grows through to adulthood. In turn this suggests that global fish oil will increase to c. 1 million tonnes of crude fish oil corresponding to up to a maximum of 200,000 tonnes of EPA + DHA, although Christian Meinich (personal communication) is suggesting 171,000 tonnes as a best case for History also suggests that more normal production levels will be sustained until the next El Niño event, which is unlikely to occur within 4-5 years, i.e. not before Assuming that compound annual growth rates for direct human nutrition and salmon farming are each 2.5 % and for non-salmon aquaculture are 5 % annually from 2016, this will permit pro-rata growth in EPA and DHA demand, assuming 170,000 tonnes total consumption in 2016 and reaching 192,000 tonnes consumption in 2020 (including 55,000 tonnes for salmon but with spare capacity to increase potentially up to 63,000 tonnes). There is then a risk of another El Niño occurring by around 2020, but the effect is likely to be offset by an increased supply of EPA and DHA from novel sources (see below). 7.5 Future availability of alternative sources These include genetically modified (GM) sources of EPA and DHA and they are already being used in Chilean salmon feeds with EPA-containing GM yeast, termed Verlasso salmon ( while production of EPA and DHA from transgenic oilseed crops is likely to result in commercial production by around 2020, as detailed later (8.2.2). Other potential sources are as follows: Alongside increased focus from existing small pelagic fisheries, increasing use of fishing industry by-products is expected to continue in order to supply marine ingredients, including fish oil. However, logistic and processing considerations complicate the use of trimmings to manufacture fish oil other than in fishmeal factories. Norwegian experience suggests that ensiling by-products overcomes some of these practical constraints, but this is not suited to oil production. Nor is the EU discard ban likely to result in greatly increased supplies of oily fish for rendering. The recycling of salmon oil for salmon production is prohibited by relevant codes of practice, but is already being used in aquaculture feeds (e.g. for seabass, Dicentrachus labrax, and gilt-head bream, Sparus aurata) and could potentially supply c. 150,000 tonnes of global salmon oil, albeit of 70 % vegetable origin and containing not more than 10,000 tonnes of EPA and DHA. This level of production is somewhat theoretical as it requires processing of visceral fat, but nevertheless reduces the total demand from non-salmonid aquaculture, making more EPA and DHA available for salmon. In conclusion, increased EPA and DHA from fishery sources is unlikely to have major volume significance in the short to medium term. 46

49 DHA Gold ( is sold in a dry form by DSM consisting of whole cell DHA-rich algal biomass; it is a non-gm product, classified as a feed material, and FEMAS-approved for sale in the EU. However, fermentation of heterotrophic organisms (e.g. Schizochytrium) is already supplying limited quantities of DHA oil to the nutraceutical industry at a significant premium over fish oil. This requires fermentation capacity and is becoming a competitive field with developments by companies, such as DSM, ADM, Alltech, etc. Cultivating such organisms requires the use of sugar which currently depends on energy, but ethanol is an alternative while the energy market is highly competitive. The potential applicability of DHA to salmon feed is to blend a fish oil scenario with a similar 1:1 ratio of EPA to DHA as seen in wild Atlantic salmon. As Peruvian anchovy has an 18:12 ratio of EPA to DHA, so the DHA can be added to anchovy oil to achieve a 1:1 EPA to DHA ratio, before balancing the blend with rapeseed oil; hence % of the fish oil can come from pure DHA but still remain in the natural range of fish oil composition of wild Atlantic salmon. Preliminary discussions with DHA producers are now being considered by GSI members for contract supply of such a fish oil equivalent. It is predicted that 100,000 tonnes of 26 % DHA produced by fermentation could be available by In the long term it is recognised that cultivation of autotrophic/phototrophic algae using photosynthesis is the most efficient solution, but it is proving hugely complex and difficult to scale-up. 7.6 Implications for the Scottish salmon industry There is currently an acute shortage of sufficient EPA and DHA to meet the different demand segments, including salmon farming. The evidence suggests that this should start to reverse soon and supplies are likely to be normal in 2016 and thereafter until the next El Niño occurs around By that time it is predicted that alternative novel sources of EPA and DHA will be available from GM and non-gm sources; otherwise the current acute shortage becomes a chronic shortage. At the moment it is impossible to forecast fish oil supply in 2015, but the expectation is that global salmon feed inclusion will fall below the 6 % EPA + DHA level in the feed oil and below the levels needed to achieve sufficient in salmon to meet the EFSA weekly intake recommendation. The worst scenario would be if the oil inclusion level of EPA + DHA falls below c. 4 %, which is close to the minimum level to ensure salmon health; in this case it is predicted that the salmon industry would be willing to pay a sufficient price to take share from other segments, such as non-salmon aquaculture. The supply position is complicated by salmon producers with preferential raw material access being able to source additional EPA and DHA at the expense of others. However, it may be that Scottish feed manufacturers have taken the precaution of forward cover for 2015 to mitigate the shortfall in world supplies. 47

50 Potentially the largest Scottish industry problem arising from the current shortage is for the premium segment, which will have to drop its claim about one portion meeting weekly recommended requirements unless sufficient cover is available. The Norwegian industry appears unconcerned about taking the pragmatic attitude that consumers may simply increase weekly salmon consumption to two or more portions by way of compensation for lower levels of omega-3 LC-PUFA. Although an estimated 70 % of the UK salmon market makes no claims (other than of variable omega-3 levels) and c. 70,000 tonnes of Norwegian salmon are imported, this pragmatic option is not readily available in the UK and in overseas markets purchasing Scottish premium niche salmon. In the event of relabelling becoming necessary, it is likely that there will be concern and confusion on the part of consumers, with potential damage to brand reputations. In all probability this would result in adverse media comment, for example criticising the supply of a less healthy product with much lower levels of omega-3 LC-PUFA and higher levels of omega-6 PUFA. As highlighted in 7.2, there is a clear risk that farmed salmon in general and the Scottish salmon industry in particular are becoming vulnerable to negative marketing by wild salmon suppliers on the issue of omega-3 and omega-6 levels and related health benefits. At its worst this could bring down the image of the farmed salmon market to the extent that consumers not only believe it no longer confers such benefits, but that it can instead be harmful compared with the wild product. Given the probability that a more normal supply situation will return by late 2015/2016, premium brands with sufficient 2014/2015 omega-3 cover would clearly benefit from being in a position to maintain their label claims during If they are unable to do so, however, they may find it difficult to regain customer confidence and recover their premium niche once supplies normalise, even if they are once again able to make the labels health claims. But for the Scottish farming industry and the UK home and export markets as a whole, this 2015 supply shortfall has the potential for longer term damage to salmon s image as a healthy product. For Scottish supply chains with insufficient forward stock cover of fish oil containing appropriate levels of EPA and DHA to meet 2015 contract obligations, the options are limited until Peruvian fishing recommences. In this connection, fishing has just been authorised in the Southern region of Peru and unconfirmed market rumours suggest that fishing of the larger anchovy stock in North/Central Peru will be authorised imminently. It is assumed that optimisation to conserve oil stocks will already be in place at both farm level (e.g. finisher diets) and feed mill level (e.g. blending with vegetable oil to meet target requirements while taking advantage of legal tolerances and of DPA content if permissible, etc.). The impact of high price on demand from different segments and the reduced inclusion levels in Norwegian and Chilean salmon feed may offer buying opportunities, including from non-traditional fish oil sources, and limited supplies of DHA may be available from fermentation. 48

51 At the same time it could be useful for the industry to consider the marginal cost of additional EPA + DHA in Scottish feed; thus if, for example, the replacement cost of 26 % EPA + DHA in anchovy oil is US$3,000/tonne (i.e. equivalent to US$11,500/tonne for active ingredient), assuming that it is necessary to meet the EFSA claim for salmon by adding an average of 7 % of the oil fraction of the feed as EPA + DHA, this is equivalent at 30 % oil inclusion to 2.1 % of the total feed, hence costing US$240/tonne. On this basis should the market be supplying only 5 % of the oil fraction of the feed to conserve material, the additional 2 % to regain scope for the EFSA claim would cost c US$69 (or US$34.50 for each percent inclusion of EPA and DHA), which seems a minor cost premium set against the reputational cost and risks of doing otherwise (assuming that EPA and DHA are available at this replacement cost). On the same basis, if current salmon feed cost/tonne is c. US$1,250/tonne, an additional 2 % of EPA + DHA at US$69 would increase feed cost by 5.5 %. This calculation must, however, take into account the inclusion of fishmeal since fishmeal contains 9 % oil with high levels of EPA + DHA. Also if some producers are using more EPA + DHA, others will have to use less due to scarcity limits. Summary of Chapter 7 Farmed salmon is a comprehensive nutritional package for consumer nutrition and health, including highly digestible protein and essential amino acids and marine fats, as well as vitamins and minerals, etc. Hence the consumer health benefits of salmon are more than just omega-3 and are more effective than relying on omega-3 supplements. But in practice consumers link the healthiness of eating salmon to omega-3 LC-PUFA, hence the strong focus by supply chains on its content. There is currently an acute shortage of EPA and DHA due to an El Niño event, but more normal conditions are expected to return during Thereafter sufficient supplies are likely to be available until around 2020 when El Niño may return, but by then it is expected that novel sources of EPA and DHA will be commercially available in volume; otherwise there will be a chronic shortage. In the immediate short term there is a potential problem for Scottish salmon products with label claims promising that one portion delivers the recommended weekly intake of omega-3 LC-PUFA. Depending on their stock cover of fish oil, suppliers may risk breaching the EFSA recommendations or having to alter label claims. Along with consequent risks to individual retail brands, there is a growing reputational risk to Scottish salmon on grounds of omega-3 (and omega-6) levels. In addition to conserving fish oil stocks by optimisation at farms and feed mills, and exploring new sourcing opportunities (possibly including DHA produced by fermentation), the salmon industry needs to plan a strategic response to this issue. 49

52 8. Novel and emerging alternative feed ingredients Considerable research has been performed to assess a large range of alternative protein and oil/fat sources as ingredients for aquaculture feeds and there is a substantial body of data specifically on salmon. It is outwith the scope of the present Report to exhaustively review this literature, but a summary of the major issues relevant to the application of alternative ingredients highlighting the main pro s and con s is provided in Annex C. 8.1 Land animal by-products (LAPs) including insect meals Current status and availability of LAPs in Europe The processing of land animals provides a potential source of protein for aquaculture feeds, especially waste materials which may not have a human food-related market. Before 2000, animal proteins were widely used in fish feeds, especially meat and bone meal, bloodmeal, and feathermeal. As a consequence of Bovine Spongiform Encephalopathy (BSE), most animal proteins were banned from terrestrial and aquatic animal feeds, since when there has been a cautious step-by-step relaxation. From June 2013, due to improved testing methods, the use of non-ruminant Processed Animal Protein (PAP) has been approved for use in aquaculture in the EU (regulation 56/2013). Today five main product lines exist in Europe, but in the UK there is insufficient slaughterhouse capacity to produce porcine material. Table 4 details availability of the main non-ruminant PAP products in the UK and EU for 2013, which would be legally available for inclusion in aquaculture feeds (Woodgate, 2014). In principle PAPs could also be imported from the USA, which supplies the majority of such material used in Chilean and Canadian salmon feeds. Segregation and traceability of non-ruminant PAPs has been an area of focus by processors to ensure that individual products are not contaminated with ruminant protein and that a suitable documentation system is in place to ensure traceability from the slaughterhouse via the processing plant to the feed mill. Safety and security of processing of Category 3 animal by-products is the subject of detailed controls and a FEMAS sector note is pending UK approval Insect-based feed ingredients There is increasing recognition that insect meals are a potential substitute for conventional sources of protein in animal feeds. Since wild salmon eat insects during freshwater life, there is current R&D investment on the potential of using insects as safe and healthy ingredients of salmon feed. Thus NIFES in Norway has commenced an Aquafly project, which will explore the potential to tailor the insect nutrient profile to meet the nutritional requirements of farmed salmon ( It is claimed that many insect species are highly nutritious and their production has less environmental impact compared with traditional sources of 50

53 animal protein. At an EU level, DG Sanco has now commissioned EFSA to provide an opinion on the available safety evidence around insect protein; at the same time the European Commission is funding the so-called PROteINSECT project, which is investigating quality, safety, processes, and human acceptance around the use of insects in animal feed. Makkar et al. (2014) reviewed existing research on five major insect species that are claimed to have potential for animal feed (black soldier fly, mealworm, locusts, grasshoppers and crickets, housefly maggots and silkworms) and concluded that black soldier fly larvae have the most promise for replacing soybeanmeal in pig and poultry diets. However, in addition to the need for salmon nutritional studies, it is recognised that there is a need for cost-effective mass insect-rearing facilities and a regulatory framework and sanitary procedures for the safe use of bio-wastes (including managing the risks of diseases, heavy metals and pesticides, etc.). The scope for insect use as a potential protein replacement in salmon feeds is clearly at an early stage of evaluation Current use of LAPs in salmon feed Since 2001 no land animal by-products have been used in salmon feeds in Europe despite the progressive relaxation of the BSE-related ban on their use. Although unconfirmed, it is understood that they have been re-introduced into some non-salmon diets in continental Europe, possibly trout feeds. This underlines the fact that their use in aquaculture feeds under specified conditions, as PAP, is once more legally permissible for non-ruminant by-products, but there is a marked reluctance by supply chains to incorporate animal by-products in feed for farmed salmon, as discussed below. However, the situation outside Europe is very different, due mainly to the lack of any large-scale BSE-related problems in North and South America. This is despite technical formulation benefits (e.g. the histidine content of bloodmeal helps to avoid salmon cataracts). For instance Chilean salmon feeds routinely contain animal by-products, with one standard grower diet currently formulated with 2.8 % poultry oil and 19 % land animal proteins. There is no legal embargo on doing so and one retailer had imported Chilean salmon, but emphasised that it only takes place subject to the salmon having been fed a non-laps based salmon feed UK supply chain perspective The current policy of the UK (and Norwegian) salmon farming industries not to use LAPs in feed was accepted by certain retailers as being somewhat contradictory. This is in view of their purchasing imports of farmed warmwater prawns (Penaeus spp.) and Pangasius (otherwise known as Bassa or River Cobbler, i.e. Pangasius spp.), which may have been fed LAPs. It was noted that canned North American salmon imports are largely based on enhancement programmes involving Pacific salmon (Oncorhynchus spp.) smolts likely to have been fed LAPs during freshwater hatchery production prior to seawater release. 51

54 This voluntary ban is reinforced by the provisions of the COGPSA, which prohibits use of LAPs in Scottish salmon. The rationale of the ban is fear of negative consumer reaction and this has been variously linked to BSE concerns, the horsemeat contamination issue, the supposed unnaturalness of such ingredients, and associations of animal by-products with use of waste material otherwise destined for landfill disposal. Two retailers felt frustrated by this situation, pointing out that it is not a food safety issue and would give access to global salmon production. However, they conceded that their customers do not wish the policy to change. An additional complicating factor cited by all respondents is the constraints of halal (and kosher) market requirements, which exclude the possibility of porcine material being used (e.g. pig blood), hence further supply chain segregation. In practice this means that use could only be made of poultry products, such as poultry offal meal, feather meal and poultry oil. It should also be borne in mind that the activities of anti-salmon lobby groups, such as the Global Alliance against Industrial Aquaculture ( would in all likelihood respond with an emotional consumer campaign and pillory any salmon producers and supply chains which broke ranks and led with a policy change on this issue. Against this it should be emphasised that most NGOs argue strongly for including LAPs in salmon feed on the basis that they represent a sustainable solution, including the World Wildlife Fund and Marine Conservation Society also the ASC standard supports the use of LAPs Conclusions and implications for LAPs/PAP (including insect meals) The use of animal by-products in salmon feeds is routine in the Chilean farming industry and has obvious technical scope as a cost-effective ingredient in Europe, including for Scottish farmed salmon. There is strong support for PAP inclusion in Scottish salmon feed by most NGOs on sustainability grounds and it is supported by the ASC salmon standard. However, supply chain comments indicate there is strong resistance by UK and continental European consumers and it is therefore not under consideration for salmon feeds in either Scotland or Norway and this is unlikely to change in the foreseeable future. The potential for insect-based ingredients has yet to be evaluated and developed, but it is likely to meet the same consumer objections. As things stand, there is no appetite for any change in policy, despite the EU having removed legal obstacles for infeed use of processed non-ruminant material, but in the event of this being considered as a salmon feed ingredient in the future, cast-iron source assurance guarantees would be needed and porcine material would need to be excluded Genetically modified feed ingredients Scope for GM protein ingredients UK use in land animal feeds versus salmon feeds 52

55 Following supply concerns about non-gm soybeans in 2013, all UK retailers, with the exception of Waitrose, changed their sourcing policies to permit the inclusion of GM plant ingredients in terrestrial livestock feeds, especially chicken. Although inclusion of approved GM feed ingredients is legally permissible in the UK, currently salmon producers, and salmon feed companies have a strict non-gm policy. This appears to reflect the following: (i) (ii) (iii) (iv) (v) Negative consumer attitudes to GM in Europe, especially in France, Germany, Austria and Italy, with potentially severe export implications for Scottish salmon. The current lack of a clear commercial disadvantage to sourcing non-gm ingredients, such as reduced availability and substantially increased cost for non- GM soy protein concentrate. It is not logistically feasible for the UK fish feed and fish farming industries to maintain both GM fed and non-gm fed supply lines - it is all or none. The current consensus to maintaining non-gm fed salmon status in Scotland is shared by the SSPO, some NGOs, and the Scottish National Party, as part of a green and clean stance. It is currently also shared by the Norwegian salmon farming industry and its leaders. However, apart from Waitrose, the UK retail respondents privately took a generally pragmatic attitude to the possibility of a more flexible GM policy if circumstances changed. They accepted that UK consumer attitudes were becoming less hostile to GM, but the largest supply chain concern relates to the risk of losing key export markets, especially France. Availability of non-gm proteins Norway s salmon industry requirements for non-gm soy (as Soy Protein Concentrate) were 365,000 tonnes in 2013 and current Scottish protein requirements are estimated at 77,000 tonnes of which over 50,000 tonnes is likely to be soy protein concentrate (Ytrestøyl et al, 2014). Non-GM soya products represent a niche market dominated by GM soya but there is no current problem in sourcing non-gm material. Currently 5.5 million tonnes of non-gm soya are being grown and supplied to world markets (of c. 200 million tonnes total soya production, which is mostly GM), mainly from Brazil, at a hidden premium, currently a little below US$100/tonne (c. 10 %) over GM soya (Christian Meinich, personal communication). Given that GM soya is used in Chilean salmon feeds, total requirement for non-gm soy protein concentrate for salmon feed is just under 1 % of global supply of non-gm soya for Scottish salmon and totals only 7.5 % including Norwegian salmon. It can be concluded that there is no serious availability threat or sufficient price driver for switching to GM soya in salmon feeds at the present time. However, it is prudent for salmon feed producers to be prepared for a switch if circumstances change; thus the Norwegian salmon feed producers have been granted annual exemptions for the 19 EU-approved GM ingredients as a fall-back 53

56 in case of lack of raw materials (Henrik Stenwig (FHL) There have also been discussions in Norway about whether it is logistically possible to supply GM and non-gm feed at the same time by using different mills; the conclusion is that this is more a matter of policy and additional cost than logistics. But it would be more difficult, although not impossible, to segregate the resulting fish streams post-harvest. It is far more difficult to see this as being feasible in Scotland and any move to abandon the non-gm policy would need to be by all three UK feed producers. Under these circumstances it is difficult to see how Waitrose could maintain their policy of non-gm fed salmon without it being fed on imported non-gm feed. Leading NGOs appear divided, with Greenpeace opposed to the use of GM worldwide, whereas WWF is not against in principle on a case by case approach using the precautionary principle. WWF stated that its offices in France and Germany were the most reluctant (on both environmental and health grounds). As regards the long term perspective of food security, it has been argued that aquaculture will only be able to make a significant contribution to feeding 9 billion people if GM feed materials are used (Svejgaard, 2014) Scope for GM oil containing EPA and DHA Commercial production of so-called Verlasso salmon in Chile involves the use of yeast, which has been genetically modified to produce omega-3 LC-PUFA, specifically EPA. The transgenic yeast cells are produced by fermentation using glucose and killed before the whole dead cells are added to salmon feed as a partial replacement for fish oil; as such Verlasso salmon has achieved a favourable sustainability endorsement from the Monterrey Bay Seafood Watch ( Camelina oil (from the oilseed crop False Flax, Camelina sativa) has been suggested as commercially suitable for salmon feeds since it contains about 30 % α-linolenic acid (a precursor for biosynthesis of omega-3 LC-PUFA like EPA and DHA), with relatively low levels of omega-6 PUFA and has been used experimentally to replace fish oil in salmon diets (Hixson et al., 2014). However, it is the recent success in producing EPA and DHA at levels equivalent to those in fish oil from algal DNA inserted into Camelina which has aroused most interest (Ruiz-Lopez et al., 2013) ( Another form of GM Camelina is already grown in North America for biofuel and it has been estimated that 1 hectare of the new transgenic crop would produce about 750 kg of oil containing around 12 % EPA + 8 % DHA and it should be highly scale-able, hence 1 million hectares could produce 750,000 tonnes of oil or 150,000 tonnes of EPA + DHA. It is claimed the next steps are to clarify the intellectual property and seek regulatory approval together with an industrial partner. Although the research has been performed in the UK and the European Parliament agreed on 13 th Jan 2015 to allow individual member states to determine their GM policy, it would seem prudent to seek regulatory approval for growing the transgenic crop commercially outside the EU, e.g. in Canada or the USA. Assuming the resulting oil is to be imported into the EU for use in salmon feed, the product would need 54

57 first to be registered as an EU-approved GM feed ingredient. Unless there are unexpected problems with the above scenario (e.g. a serious intellectual property dispute), given unsatisfied market demand from aquaculture and nutraceuticals, there seems no obvious reason why this transgenic oil should not be commercially available in the UK (and Norway via its EU agreements) by the year A similar timescale seems likely for the production of EPA and DHA from transgenic rapeseed by a joint venture between BASF and Cargill; however, they are indicating that the nutraceutical sector will be their preferred target market. The only other obvious comparable development is in Australia also with transgenic rapeseed (Canola) (Kitessa et al., 2014; with CSIRO/Nuseed developers suggesting that commercial supply could be available by 2017, which seems an ambitious timescale and appears to focus mainly on DHA production. The existence of an entire logistical system for handling oilseed products means that the supply chain is already present; hence the GM oil production cost need be no higher than conventional rapeseed oil. The market implications of a continuing decline in EPA and DHA levels are seen as a potential driver for change in the event of GM oil becoming commercially available. Indeed the lack of any cellular material, protein, or nucleic acid in oil, and the inability to test for GM status of such oil, were noted in this context Conclusions and implications for GM feed ingredients Unless non-gm protein supply becomes less competitive in terms of supply and price, there is little incentive to change today due to the risk to salmon export markets. But given the market importance of omega-3 LC-PUFA content, the commercial availability of costeffective GM oil from recombinant oilseed crops could herald a relaxation of non-gm policy, possibly as a special case linked to consumer health, but perhaps leading to wider use of GM feed ingredients. In view of the brand focus on Scottish salmon as a healthy product, a switch to using GM oil might well be seen as a price worth paying to avoid the risk of fish oil supply interruptions, if and when transgenic material becomes commercially available. It is likely that consumer attitudes will continue to soften and that current negative attitudes by NGOs, trade associations and some Scottish politicians may reverse in the next five years, allowing penetration of salmon fed GM ingredients even to markets traditionally most hostile to GM-fed products (e.g. France). In this connection the lack of any genetic material in GM oil removes at a stroke a principal objection to its use and European consumers might be less concerned about theoretical risks of environmental effects in the case of oil imported from GM crops grown outside Europe. In any event UK retailers, other than Waitrose, emphasised that they would follow, but would be unwilling to lead, a change to GM-fed salmon. The overall conclusion is that a switch to GM feed ingredients is seen as most likely happening by 2020, possibly linked to GM oil, and that it would be likely to take place before any change in LAPs/PAP policy became commercially feasible. 55

58 Summary of Chapter 8 Although there is routine use of LAPs in salmon feeds outside Europe, and the EU is now permitting its use as PAP in fish feeds, there is a supply chain embargo in Scottish (and Norwegian) salmon feeds. This is entirely due to a perceived negative consumer attitude, despite clear technical and sustainability benefits, and strong support for doing so from some NGOs; this is unlikely to change in the foreseeable future. As regards the use of GM feed ingredients, a similar embargo exists today in European salmon farming. This is despite their use in UK chicken feeds, signs of a weakening in UK consumer hostility to the GM issue, and a more flexible approach by the EU and UK governments. Given the availability of non-gm soya at only a modest premium today, there is no obvious incentive to alter this policy on grounds of protein availability. The position with omega-3 LC-PUFA is different as the likely commercial availability of GM oilseed crop sources by 2020 from outside Europe offers a potential technical solution to a chronic shortage, which will otherwise put at risk the brand identity of Scottish farmed salmon. The lack of genetic material in GM oil and the potential for enhancing the healthiness of farmed salmon by its use in feed could help to overcome reluctance of UK and continental supply chains (with salmon exports to France being possibly the main challenge). It is concluded that the current European embargo on using GM salmon feed ingredients is likely to be overturned once GM sources of EPA and DHA become commercially available; and is likely to precede any relaxation of policy on LAPs/PAP. 56

59 9. Contaminants and food safety issues 9.1 Regulatory background and concern about salmon contaminants The presence of undesirable substances in feed is controlled by European Parliament and Council Directive 2002/32/EC of 7 May 2002 (as amended), which sets maximum permitted levels (MPLs) for these substances. Feed that contains a contaminant at a level above the relevant MPL is deemed to be unsafe and must be withdrawn and disposed of outside the feed and food chains. In addition to these controls on fish feed, controls exist on raw materials and on the resulting product (salmon); the MPLs are regularly reviewed and frequently reduced. The area of contaminants is therefore closely monitored and controlled. Nevertheless on grounds of public health, a report about the presence of organic contaminants in farmed salmon raised concerns about the presence of eating farmed fish owing to the presence of contaminants in marine ingredients, which then enter the food chain via aquaculture (Hites et al., 2004). It has since been shown that the potential health risks are extremely small compared to the health benefits of eating salmon products, with the benefits estimated to be at least 100-fold greater than the estimates of harm, which may not exist at all (Cohen et al., 2005; Mozaffarian and Rimm, 2006). In addition recent data (EFSA, 2012) show that farmed salmon and trout contained on average lower levels of dioxins and PCBs than wild-caught salmon and trout, at least for Europe. 9.2 Beneficial effect of reduced marine ingredient inclusion The Norwegian Scientific Committee for Food Safety (VKM, 2014) concluded that the benefits of eating fish clearly outweigh the negligible risk presented by current levels of its contaminants and other known undesirable substances. As regards farmed salmon, concentrations of dioxins and dioxin-like PCBs, as well as mercury, have decreased by about 30 % and 50 %, respectively, compared with the corresponding levels in 2006, due to replacement of marine feed ingredients by plant ingredients. New fish feed contaminants, such as the pesticide endosulfan, polyaromatic hydrocarbons (PAHs), and mycotoxins are unlikely to be a safety issue since the concentrations are very low and not detectable with sensitive analytical methods. The health risk associated with the environmental contaminants known as brominated flame retardants is low, as also stated in the 2011 EFSA assessment. Furthermore it is VKM s opinion that present exposure to residues of veterinary medicinal products, including residues of antibiotics in farmed fish, in the Norwegian diet is of no concern, since the levels are very low and often not detectable even with sensitive analytical methods. The report also concluded that calculated exposures for the synthetic antioxidants from a 300 g portion (more than twice the portion size used to calculate beneficial components such as omega-3 LC-PUFA) of farmed fish fillet are of no concern and are below the relevant acceptable daily intake. In contrast to their conclusion in 2006, VKM state that there is now no reason for specific dietary limitations on fatty fish consumption for pregnant women. Given that sourcing of marine ingredients by Scottish fish feed companies is either undertaken wholly or partly by the Norwegian (Skretting and EWOS) or 57

60 Danish (BioMar) parent companies to take advantage of bulk purchasing, it is likely that the contaminants picture in Scottish salmon feed closely parallels the Norwegian experience. The main differences will be due to local UK sourcing of fishery by-products, the tendency for standard Scottish salmon feed formulations to lag behind Norwegian formulations in terms of the rate of replacement of marine ingredients in Scottish compared to Norwegian diets, and the greater use of bespoke salmon diets in Scotland with higher marine content. As regards the possibility that a switch from marine to predominately plant-based diets will simply mean a greater risk of pesticide contamination of the feed, in practice any such contamination is unlikely to penetrate through the hull which is normally discarded from oil seeds during processing. 9.3 Implications of Marine Harvest initiative to use cleaned fish oil Notwithstanding the low levels of contaminants in farmed salmon and their reduction due to replacement of marine content, it is of interest that in 2014 Marine Harvest entered into an agreement with FF Skagen in Denmark to clean all relevant fish oil used for Marine Harvest salmon farming in order to remove environmental pollutants. According to Marine Harvest s CEO This is a major step in the strategy of Marine Harvest becoming a fully integrated protein producer, able to follow and control all ingredients from our fish feed to the consumer s plate. From time to time the discussion of how many salmon meals the consumer should and could eat appears in the media. We recognize that such debates can confuse the consumer. Our aim with the cleaning of fish oil is to remove all doubts regarding salmon intake. The result is that Marine Harvest salmon becomes even healthier". This initiative is a differentiator for Marine Harvest and will put pressure on its competitors to follow in the face of not infrequent allegations that farmed fish are a potential source of contamination; it should result in farmed salmon being consistently lower in contaminants, not only than wild fish, but also than much land animal produce. It may not be coincidental that this policy is being introduced at a time of generally falling EPA and DHA levels in farmed salmon; insofar as this problem can be mitigated by consumers eating more salmon, a move to ensure minimal salmon contaminant levels overcomes a potential barrier to their being able to do so. At the same time a policy of cleaning all fish oil supplies should give access to wider sources of fish oil than might otherwise be the case, although it will incur increased production costs. There may also be potential benefits to fish health and performance as there is some evidence that salmon fed with diets using cleaned oil may grow and convert better under experimental conditions than salmon which are fed diets using uncleaned oil (Olli et al., 2010). From 2010 to 2012, partly in order to source the highest levels of EPA and DHA, European salmon feed companies switched from mainly North Atlantic to mainly South American sources of fish oil and this has had the additional benefit of reducing their exposure to relatively higher background levels of environmental contamination in North Atlantic fisheries and fish oil. At the same time background contaminant levels in fishmeal and fish 58

61 oil are falling anyway, due in part to lower use of North Atlantic species, such as herring, which now are mainly used for direct human consumption. 9.4 Miscellaneous feed-related risks Although the increased replacement of marine by plant feed ingredients has reduced environmental contaminants in salmon to well below statutory levels, and increasing use of cleaned oil will improve this still further, it remains a potential concern, especially with the use of fishery by-products. The latter may be derived from human consumption fisheries spread over a wide geographical area with differing exposures to industrial runoff causing marine contamination. With increasing use of by-products, this requires continuing vigilance and could represent a particular risk problem for organic salmon production which relies on by-products as feed. In this connection, if dioxin is found in blue whiting by-products, for example, it can be removed at extra cost, whereas PCBs are more difficult to remove. The discovery of hexachlorobenzene (HCB) in fishmeal from one Peruvian factory in 2014 has now been shown to be a one-off accidental occurrence. Following the introduction of new regulations about clean water in Peru, this factory s water treatment plant used a contaminated chemical by-product in the process without following normal Hazard Analysis and Critical Control Points (HACCP) procedures. As a result concentrated sludge containing the contaminant was added back to the fishmeal, which was then exported to Europe and elsewhere before being detected. Procedures have now been put in place to ensure this should never happen again. The prohibition by salmon Codes of Practices of using farmed salmon material in feed ( inter-species recycling ) avoids the theoretical possibility of prion diseases occurring, akin to BSE. By contrast a well-known problem is the transmission of parasites, moulds and toxins to salmon from contaminated feeds. In practice these risks are minimised by good dry storage conditions for feed ingredients and finished feed and by the temperature of feed manufacture. In addition to feed mills adopting HACCP systems for risk awareness and control, UK retailers normally require compliance with the British Retail Consortium s global standard for food safety ( which was created to establish a standard for due diligence and supplier approval. Summary of Chapter 9 Despite an elaborate EU control system to prevent undesirable substances reaching farmed salmon via feeds and feed ingredients, there has been historical concern about the presence of contaminants in farmed salmon, mainly due to marine feed ingredients. However, various detailed studies have now concluded that the benefits of eating farmed salmon clearly outweigh the negligible risk presented by current low (or undetectable) levels of contaminants and other known undesirable substances; Norwegian data partly reflect the reduction in marine ingredients in Norwegian salmon feed and will be mirrored 59

62 also in Scottish salmon feed, albeit at a reduced rate. The Marine Harvest initiative to use cleaned oil takes this a stage further and enables consumers to eat salmon on an unrestricted basis as regards health risks. More care needs to be taken, however, with the increased feed use of by-products from processing of fish for human consumption; this is a potential challenge in producing organic salmon feeds. 60

63 10. Other feed ingredient issues 10.1 Carotenoid pigments The red carotenoid-rich bacterium Paracoccus carotinifaciens is authorised by the EU as a sensory additive for use in salmon and trout. This fermentation product is marketed as Panaferd-AX ( and, due to it being a naturally produced compound, has recently been adopted as the pigmenter of choice for Scottish farmed salmon in place of synthetic astaxanthin used previously. In this regard Scottish salmon is differentiated from Norwegian salmon which continues to rely mainly on synthetic pigment. However fermentation is energy-intensive and using this more natural product in salmon feed has a penalty in terms of carbon footprint (see 6.4) Antioxidants Several synthetic antioxidants are authorised for use as feed additives in the European Union, including ethoxyquin and butylated hydroxytoluene (BHT), which are generally added to fish meal and fish oil, respectively, to limit lipid oxidation. According to Article 12 of Regulation (EC) No 396/2005, EFSA has reviewed the Maximum Residue Levels (MRLs) currently established at European level for ethoxyquin. Although this active substance is no longer authorised within the European Union, an MRL was established by the Codex Alimentarius Commission (CXL). Based on the assessment of the available data, EFSA assessed the CXL, and a consumer risk assessment was carried out. The CXL was found not to be adequately supported by data and a possible risk to consumers was identified, hence further consideration by risk managers is now needed. According to IFFO (2014) the latest request for further information by EFSA received in November 2014 included questions that have not been raised in previous years and stems from the increasing negative publicity around the continued use of ethoxyquin (often described as a pesticide residue). It is clear that EFSA and the Commission are under increasing pressure to revoke the use of ethoxyquin, but this can only be done if EFSA issues a negative opinion and it is clear that they need more scientific evidence upon which to base their opinion. In addition, they expressed their concern of replacing ethoxyquin with an alternative synthetic antioxidant such as BHT. In this connection a study by Lundebye et al. (2010) found that a 300 g portion of farmed Atlantic salmon would contribute up to 75 % of the acceptable daily intake for BHT; although levels will have fallen considerably due to marine ingredient replacement by plant materials. An update on EFSA s assessment of ethoxyquin was included in the European Commission s relevant Standing Committee agenda in December It is currently impossible to guess what the outcome of these deliberations will be. Any authorisation is valid for 10 years and may be renewed, but this is now a live issue at EU regulatory level and of concern to the salmon industry; and the balance of probability is against this renewal happening. If BHT also becomes unavailable, 61

64 the alternative products commercially available (e.g. α-tocopherol-based natural antioxidants ) are less effective and higher in cost than ethoxyquin and BHT Other ingredients Other feed ingredients include the following: synthetic amino acids (e.g. methionine, lysine, histidine, etc.) are referred to in Annex C section 1.2; attractants are referred to in Annex C section 1.4; and enzymes (e.g. phytase, carbohydrases, etc.), binders and process improvers are referred to in Annex A section 3 and Annex C section 1.3. These are required for a variety of reasons, including balancing nutrient contents, increasing bioavailability of nutrients, ensuring feed intake, and improving and maintaining feed pellet stability. Many of these ingredients are specifically required and/or have greater relevance and importance in feeds with increased substitution of marine ingredients by plant-based materials Feed-related product quality issues Formerly it was suggested that flesh quality issues such as gaping were feed-related and as a result some feed formulations were modified to include higher micronutrient levels. However, the incidence of such problems has reduced and current thinking is that this is because of better handling and may therefore be unrelated to feed. Excessive fat remains a recurrent problem with large fish, especially after cold-smoking, for whatever reason (as referred to in Annex A, section A2). Summary of Chapter 10 Scottish salmon feed use of a bacterial fermentation (hence natural ) product as a flesh pigmenter and this differentiates Scottish from Norwegian salmon, but the trade-off is that it has a greater carbon footprint. Attention is drawn to the possible non-renewal by the EU authorities of approval for ethoxyquin as a feed additive to control fat oxidation (e.g. in consignments of fishmeal and fish oil) and the implications are discussed. Physical product quality issues, such as flesh gaping, have reduced and are not now considered likely to be feed-related. 62

65 11. Industry overview and forward projections 11.1 Scottish salmon: current status of salmon farming and projections to 2020 and beyond Current status of salmon farming in Scotland The Scottish salmon farming industry plays a unique economic and social role in the Scottish economy, especially in regard to generating wealth and employment in remote rural west coast highland communities and in the Northern and Western Isles. In terms of technological development, the Scottish salmon industry closely resembles the much larger Norwegian salmon industry, which ultimately owns the majority of Scottish farming companies and is approximately eight times larger in production volume. Scottish farmed salmon production in 2013 was 163,234 tonnes whole fish equivalent (WFE) and is predicted to be c. 165,000 tonnes in Scottish farmed salmon has developed in response to a wide range of individual customer requirements, partly reflecting competition between UK retailers for whom farmed salmon has become the largest volume fish species sold in a variety of different product forms. Although the majority of retail salmon in the UK is currently imported from Norway, having been reared under generally standard dietary regimes to achieve a fairly uniform quality for onward processing, UK retailers look to Scottish production to rear different qualities of salmon under contract enabling differentiated supply with regard to specific features. For example, different retailers vary in terms of the content of omega-3 LC-PUFA and the label claims made on the pack for marketing purposes in order to reinforce their brand identities. This is potentially confusing to the market, but it must be admitted that the need to take account of biological variability, against the background of frequently changing regulations, makes labelling a significant challenge for retailers. Although supply chains also have a differentiated approach to standards and certification, there appear to be some common requirements (e.g. in the case of marine feed ingredients, reference to the FAO Code of Conduct for Responsible Fisheries, including traceability to species and country of origin, evidence of avoidance of IUU fish, etc.). At the same time Scottish production has some distinctive features which do not have the same focus elsewhere, such as a commitment to fish welfare, with % of farms being certified to the Freedom Food standard. Another differentiator is the adoption of a natural approach to pigmentation by means of a bacterial fermentation product in contrast to the predominant Norwegian use of synthetic pigment. However, the most significant points of difference are the focus on omega-3 LC-PUFA content of Scottish farmed salmon, together with the more conservative approach to substituting marine feed ingredients with plant ingredients compared with Norway, despite the individual range of ingredients being similar; these two features of omega-3 and marine content of the feed are linked, especially as regards fish oil. 63

66 The result is a variety of different feed specifications being adopted, usually under contract to supply different markets, ranging from high inclusions of marine ingredients for export to France under the Label Rouge specification, to products virtually identical with more standard Norwegian imported product. In between are specific niche products for specific retailers requiring bespoke feed formulations. The overall picture is of a somewhat differentiated supply of Scottish farmed salmon with a significant proportion occupying a premium market position. It remains to be seen whether the recent UK success of discount retailers will expand consumption of lower cost standard products and tend to undermine this niche Scottish premium position in the UK Future projections In 2013 the Scottish Marine Plan consultation set a sustainable target of 210,000 tonnes salmon production by 2020 ( This represents a compound annual growth rate of just under 3 % from the 2013 production of 163,000 tonnes and would appear realistic, provided new production sites are available, etc. This is also in line with global salmon projections. Although annual growth was 6 % during the period from 2004 to the estimated 2014 production, growth rate has diminished in recent years and is expected to diminish further going forward, resulting in a projected 3 % annual global growth from 2013 to 2020 (Kontali, quoted by Marine Harvest, 2014). At an assumed and possibly conservative (economic) feed conversion ratio (efcr) of 1.2, Scottish salmon production of 210,000 tonnes in 2020 would equate to a feed requirement of 252,000 tonnes. According to the Imani Report (Marine Scotland, 2014), production of 210,000 tonnes of marine finfish in 2020 (nearly all salmon) would yield a turnover of 771 million from production only; this corresponds to an estimated 1.1 billion turnover and 345 million GVA across the Scottish supply chain, employing over 7,000 individuals. If UK total salmon feed supply in 2014 was 220,000 tonnes from salmon with an overall efcr of c. 1.2, and annual salmon harvest volume is expected to increase from c. 165,000 tonnes to 210,000 tonnes in 2020, it seems not unreasonable to assume an improved efcr within the range of This would correspond to a 2020 feed volume requirement in the range of 231, ,000 tonnes Forward supply of marine ingredients Table 5 (OECD-FAO, 2013) is an annual global forecast from 2013 to 2022 of total fishmeal and fish oil production, production from whole fish, consumption, variation in stocks, and price. Using the same fish model, FAO projected total fishery production, aquaculture, fishmeal and fish oil production in Table 6 (FAO, 2014), assuming that about 16 % of capture fishery production will be reduced to fishmeal and fish oil (down 7 % on the average), but total baseline 2022 production will be million tonnes and million tonnes, respectively, i.e. 15 % and 10 % up on the base period, with almost 95 % of the additional gain for fishmeal coming from improved use of fish waste, cuttings and trimmings; (this compares with optimistic scenarios of million tonnes and

67 million tonnes, respectively). These assumptions are based on sustained demand and high prices of fishmeal and growing added value fishery products for human consumption, leading to more by-products being used in fishmeal manufacture; OECD-FAO estimate that fishmeal from by-products should represent 49 % of total fishmeal production in 2022 and that, with global demand stronger than supply, prices of fishmeal and fish oil will increase by 6 % and 23 %, respectively, in nominal terms by Although the OECD-FAO model is the best available, it should be added that this is no guarantee of its predictive ability. Making long-term forecasts of fishmeal and fish oil supply and demand is somewhat speculative and has to take account of a variety of complex and interacting drivers; these include the demand for pelagic fish for direct human consumption, the effect on reduction fisheries of increased regulatory curbs based on the precautionary principle, climatic events (e.g. El Niño), and the increased exploitation of fish processing wastes for reduction purposes. At the same time demand for fishmeal and fish oil is a function of price and availability of alternative proteins and alternative oils, particularly novel sources of omega-3 LC-PUFA (e.g. oils from transgenic oilseed crops), as well as the rate of growth, not only of fed aquaculture species (e.g. salmon and marine fish), but also of the growth and profitability of young pigs and day-old chicks. This is in turn influenced by innovation of feed formulators, plant protein processors, and geneticists, as they seek to reduce fishmeal inclusion rates to save on scarce, fluctuating, and costly ingredients, and also by the efforts of civic society (e.g. NGOs) to penalise use of marine ingredients via certification on the supposed grounds of sustainability. As regards demand for fish oil by salmon, it has been shown that demand has become a derived demand for EPA and DHA fatty acids and, as such, that salmon feed is competing strongly with the human nutrition market, while supply is linked to that of its co-product fishmeal; although this is over-reliant on omega-3 LC-PUFA rich species, such as sardine and anchovy catch, hence subject to the volatility of the Peruvian catch due to El Niño. As regards the limited amount of pelagic fishmeal, Olsen and Mohammad (2012) concluded this will not be a major obstacle for a continued moderate growth in global aquaculture production; it is reasonable to make the same conclusion in regard to global salmon production. On the somewhat conservative assumption of a 30 % marine/70 % plant content of feed ingredients (lower in marine content than today s Scottish average, but the same as current Norwegian diets), 231, ,000 tonnes of salmon feed for Scotland in 2020 would require 69,000-76,000 tonnes of marine ingredients and 162, ,000 tonnes of plant ingredients. If the fishmeal requirement in 2020 is 45,000-50,000 tonnes, given that UK and Irish production of fishmeal in 2014 was c. 39,000 tonnes, which is likely to rise over time, most, if not all, fishmeal requirements could be sourced locally if so desired. However, it should be remembered that increasing volumes of UK production of marine ingredients come from salmon by-products which cannot be recycled back into salmon diets. 65

68 11.3 Forward plant protein and oil supply OECD and FAO secretariats (OECD-FAO, 2013) analysed the price ratios of aquaculture species relative to oilseed products ( and concluded that tight supplies of fishmeal and fish oil are likely to contribute to an increased price ratio between fish and oilseed products over the medium term due to continuing demand from early rearing of pigs and salmon farming and from continuing omega-3 demand. They recognised the price ratios will be exacerbated in El Niño years (forecast to occur in 2015 and 2020), which will further constrain supply and support higher prices. Furthermore they project a 26 % increase in world production of oilseeds and a switch in distribution of land from coarse grains to oilseeds. Global protein meal output is projected to increase by 25 % or 67 million tonnes by 2022; this envisages consumption growth of protein meal slowing somewhat due to slower absolute growth in global livestock production and slower growth in the relative use of protein meal in feed rations. The overall availability and price of terrestrial ingredients will also depend on factors such as freshwater availability; while the model used is based on past behaviour, it is recognised that tipping points may arise in regard to the availability of terrestrial feed sources. In the short-term, it is reassuring for security of supply that, as the principal plant feed ingredient in Scotland, non-gm soy (as soy protein concentrate) is under 1 % of global supply of non-gm soy, which totals 5.5 million tonnes. Also global salmon feed demand for non-gm soya is under 7.5 % of world supplies of non-gm soya. But the non-gm policy means potential vulnerability to continuing availability of non-gm soya, which is essentially a niche product accounting for under 3 % of global soya production. The European salmon industry imports non-gm soya mainly from Brazil as already crushed concentrate, paying nearly US$100/tonne premium over GM soya and there is no guarantee this premium won t increase significantly over time. Hence the strategic logic of (i) avoiding over-reliance on imported soya products by greater focus on locally grown protein substitutes, including legumes, beans and peas (see 12.4); (ii) focussing on novel sources of EPA and DHA, especially from algal fermentation and transgenic oilseed crops Industry feed-related risk exposure The major feed-related risks include the following: The current acute shortage of fish oil, hence EPA and DHA, provokes adverse public criticism about the impaired health benefits of Scottish farmed salmon which is not allayed by the recovery of fish oil stocks; this damages the Scottish salmon market. In the event that premium suppliers are forced to drop their omega-3 LC-PUFA label claims, it may encourage wild salmon suppliers to promote their product against farmed salmon on the grounds of superior health benefit and omega-3/omega-6 ratio. 66

69 The current acute shortage of fish oil, hence EPA and DHA, is likely to return by around 2020 and there could be unanticipated events prior to 2020; these could include another El Niño event, or a large rise in demand from the human nutrition market (currently growing slowly), neither of which is particularly likely, but both possibilities would reduce aquaculture supply. There is also the possibility that alternative novel sources of EPA and DHA, which this report believes will be commercially available by 2020, are unduly delayed in becoming available, or are only available at prohibitively high prices. By around 2020 with the possibility of another El Niño-linked scarcity of fish oil and the likely availability of competitive, approved GM-derived sources of EPA and DHA, the pressure to drop opposition to GM feed ingredients becomes difficult to resist. Such a move would undoubtedly provoke some criticism, although it is believed UK public opposition will continue to weaken and would therefore be muted by comparison with the reaction today, especially given that all retailers except for one have already taken this step for poultry products. However, the main risk will be to export markets, especially France, should opposition to GM there continue to be strong. Given the technical and sustainability arguments for including land animal by-products (LAPs) in salmon feed, coupled with their use in imported farmed products from Asia, there is a small possibility that a UK retailer might break ranks and specify use of such material for salmon. The biggest concern would then be the likelihood of adverse press criticism and antifarmed salmon lobby groups causing damage to the whole UK market (depending on the extent to which the UK has become accustomed to GM fed products, this is a potential risk of using GM feed ingredients as well). The occurrence of a major food safety incident linked to farmed salmon can never be discounted, although it is difficult to envisage how this could be feed-related. Bacterial food contamination (e.g. Listeria spp.) is not uncommon in food, including salmon, but the worst such incidents have occurred due to Clostridium botulinum toxin formation in canned fish products and in vacuum packed smoked fish products, due to processing system irregularities or breakdowns, with fatal consequences on occasion. In conclusion the major feed-related risks to the Scottish salmon industry are believed to be reputational, linked mainly to reducing levels of EPA and DHA in feed and consumer perceptions of the resulting salmon product. Whereas the likely future supply of GM ingredients could help solve the fish oil scarcity problem, it is not without its own market risks. Similar issues arise with the use of LAPs, food safety incidents, etc. 67

70 Summary of Chapter 11 The current profile of the Scottish industry is of a differentiated supply, including the production of premium salmon products, and its planned increase to 2020 is achievable given the assumptions. As regards the required supply of feed ingredients to 2022, this is discussed in relation to the FAO and OECD forecasts. For Scotland it is concluded that sourcing the required volumes of marine and plant proteins will not be problematic, although the ability to use a greater proportion of locally-grown plant materials will be advantageous. However, to maintain adequate supply of omega-3 LC-PUFA will require an increasing reliance on novel sources, especially from algal fermentation and transgenic oil seed crops. It is concluded that the main feed-related risks faced by the industry are reputational, linked mainly to the reducing levels and increasing costs of available omega- 3 LC-PUFA and to the future transition to novel sources without exacerbating such risks. 68

71 12. Key questions and conclusions 12.1 Is Scottish salmon safe and beneficial for consumers? Lack of contaminants and food safety The picture on contaminants in farmed salmon is encouraging despite past adverse publicity. It is clear that concentrations of dioxins, dioxin-like PCBs, and mercury in Norwegian salmon have fallen very significantly in the past six years due to replacement of marine feed ingredients by plant ingredients; newer contaminants, such as endosulfan and PAHs, as well as antimicrobial therapeutants, are unlikely to be a safety issue since the concentrations are very low or undetectable with sensitive analytical methods. At the same time EFSA has confirmed that contaminant levels in wild salmon and trout in Europe are higher than in farmed products; also that the health risk associated with brominated flame retardants is low. As a result the Norwegian food safety authorities conclude the benefits of eating fish clearly outweigh the negligible risks involved. Notwithstanding the relatively higher marine content of some UK salmon diets, broadly similar conclusions may be drawn for Scottish salmon. However, from time to time it is possible that organic farmed salmon may have relatively higher contaminant levels (albeit within legally permissible ranges), reflecting their dependence on feed containing fishmeal derived from trimmings by-products of fish caught for human consumption over a wide geographical area, including areas which may have higher background levels, e.g. North Sea bordering the Baltic. Supply of nutrients for human health and nutrition Salmon are an excellent source of, and delivery system for, the supply of high quality nutrients; they are a healthy component of the human diet offering easily digested, high quality protein, good balance of essential amino acids, and a source of key minerals and vitamins (including Iodine, Selenium, Vitamins D and E, etc.). Although the increased use of plant-based ingredients has reduced the relative proportion of omega-3 LC-PUFA in farmed salmon compared to the levels a decade or more ago, the higher fat content has ensured that farmed salmon still deliver a high dose of EPA and DHA, which is generally sufficient to satisfy the current dietary recommendations for omega-3 LC-PUFA intake in 1-2 portions. Furthermore, the absolute level of EPA and DHA in farmed Scottish salmon has been generally higher than in wild salmon. Several recent reports that have included detailed benefit/risk analyses have confirmed that the health benefits obtained by the consumption of farmed salmon far outweigh any perceived health risks Is feeding Scottish salmon a sustainable activity? Eco-efficiency of salmon farming 69

72 Recent detailed studies of the salmon farming industry in Norway by Nofima have shown that it is a more efficient way of producing nutrients for human consumption than either pig or chicken farming, as demonstrated by its climate impact, area of land occupation, and use of non-renewable phosporus resources. Farmed salmon retain nutrients more efficiently and are better converters of feed nutrients to nutrients for human consumption than the most efficient land animal production. Life cycle analysis suggests salmon farming and cod fishing are comparable, but if the objective is to provide marine nutrients for human consumption, the Norwegian study showed it is far more efficient to harvest pelagic fish for fishmeal and fish oil, than to leave them in the sea. Overall salmon farming is a more efficient use of resources than commercial fishing, especially when taking account of fishing externalities. The oceans provide an underutilised source of nutrients for human consumption and salmon farming offers a highly efficient mechanism for transforming these resources into high quality food which is available on a year round basis. Substitution of marine by plant ingredients Finite marine feed resources are now being replaced by plant-based ingredients (and fishery by-products), hence enabling salmon farming to achieve net fish protein and oil production. The levels of plant and marine content in Norway are 70 % and 30 % respectively, whereas the corresponding figures for Scotland cover a wider range of dietary inclusions (from 50 % and 50 %, respectively, for Label Rouge, to 70 % and 30 %, respectively, in standard diets), but with the recent growth in high performance diets are believed to average around 60 % plant and 40 % marine content today. Despite the greater feed differentiation in Scotland, the continuing overall trend towards feed substitution of marine by vegetable ingredients, together with the increasing evidence that reduction fisheries are not being overexploited to produce more and more feed ingredients, contradicts the traditional view that increased salmon farming relies on the unsustainable use of marine ingredients. In other words rising demand for fish feed ingredients has not increased pressure on wild fish resources, but instead there has been an increased use of plant protein and fat sources. An important conclusion is that there is no evidence that terrestrial agricultural animal and plant feed resources are more sustainable for farming salmon than using feed ingredients based on wild-caught marine resources from sustainably managed stocks (or from fishery by-products); an increased utilisation of plant ingredients may not be as sustainable as generally assumed. At the same time substitution with plant ingredients causes the farmed salmon content of omega-3 LC-PUFA to fall with negative implications for consumers (pending availability and consumer acceptance of cost-effective alternative sources of omega-3 LC-PUFA), unless they compensate by consuming more oil-rich fish. Therefore, Scottish certification focus on sustainable raw materials counters claims of an unsustainable use of marine ingredients, hence the sustainability issues for farmed Scottish salmon are now mainly related to post-smolt survival following seawater 70

73 transfer, lice/disease, and escapement, rather than feed ingredients. Given that fishmeal and fish oil are, and always will be, finite and limiting resources, there is increasing focus on optimising the strategies for their use, such as targeting critical life stages (e.g. broodstock, fry, and smolt diets). Social, ethical and economic considerations Although still a young industry, salmon farming has greatly increased the availability and reduced the cost of supplying salmon to world markets. It has also brought sustainable employment to many remote rural locations. That it has not only survived, but grown to be a highly efficient global industry, is mainly due to innovation and rapid technological change, enabling increased productivity, cost reduction, and close control of the production process. There is growing interest in the use of certification to manage potential areas of social and ethical concern related to fish supply chains (e.g. welfare of fishermen and also farmed fish) and fish processors increasingly subscribe to the exchange of ethical data. As regards the wider issue of using fish as feed, it is more efficient in a protein-hungry world to harvest unmarketable species of fish for indirect human consumption via aquaculture than not to harvest them at all, especially when their use in salmon diets yields net edible production of fish protein What is the evidence of responsible practice? Adopting aquaculture standards The main focus area for salmon farming standards is long-term sustainability from an environmental standpoint, especially as it relates to use of marine feed ingredients. In addition there are well-established food chain standards controlling process quality, food safety, and particular focus on fish welfare in Scotland (via Freedom Food). Of the six commonly used aquaculture standards, four are widely adopted by the Scottish industry, reflecting focus on sustainable feed issues and largely driven by competing brand values on the part of UK retailers, in some cases under pressure from environmental NGOs. The most widely used finfish standard is GlobalGap, but the recent ASC certification of two Marine Harvest sites in Scotland suggests that this may become a future benchmark of responsible practice (although its feed standard is still under development). Sustainability issues have been used as a competitive tool with the result that standards for salmon farming and feed use are demanding (at the risk of becoming disproportionately so in some cases) and continue to be driven up. This will tend to restrict choice among different raw materials and must be based on appropriate sustainability criteria to avoid being irrelevant, or unduly influenced by campaign agendas lacking a rigorous scientific basis, hence driving the need for continuing innovation. 71

74 Use of certified ingredients To demonstrate responsible sourcing of salmon feed to supply chains, the use of certified ingredients has been adopted by Scottish feed suppliers and farmers. In the case of marine ingredients, IFFO RS certification is probably universal in Scotland and provides evidence of traceability back to responsibly managed fish stocks, avoidance of IUU fish, and control of by-product raw material, with an associated chain of custody. In the case of soya supplies, most are certified by ProTerra, RTRS, or Cert ID, but this is a work in progress for other plant ingredients. Looking ahead there is a need for greater reciprocal recognition based on the different roles individual aquaculture/feed certification schemes can play in managing supply chain issues, possibly including a Russian doll approach whereby an aquaculture certificate incorporates other specific ingredient or welfare-related certifications depending on priorities What are the future prospects and challenges? Raw material availability Continuing focus is needed on how best to maintain EPA and DHA levels in farmed salmon given the increased price and reduced availability of fish oil of appropriate certification status and lack of contaminants. The most feasible alternative source in the medium term is likely to be via GM plant oil, while algal and microbial sources and even alternative marine oils need to be kept under review. Interim solutions are needed to manage this situation cost-effectively and DHA from algal fermentation is now available at high cost and being studied as one solution. As regards alternative sources of protein feed ingredients, it is important to have suitable alternative plant protein concentrates to reduce the current high dependence upon imported soy protein concentrate in practical feed manufacture. Current focus is on the use of legumes, faba (field) beans, and peas in salmon diets since both crops could be grown locally. Use of genetically modified feed ingredients So far the price premium for non-gm soya is below US$100/tonne, but should this increase, pressure will mount to use GM soya in salmon feed in the same way that GM feed ingredients are used in UK poultry. In addition, it seems likely that volume supply of EPA and DHA will become commercially available via oil from genetically modified oilseed crops by about 2020 assuming EU approval. 72

75 If (as seems likely) there will be continuing widespread global use of GM feed ingredients, an unwillingness to utilise such ingredients will impose a significant opportunity cost on Scottish production, which is more than simply a price premium for non-gm ingredients (e.g. soy) assuming they are still available. Thus in the case of GM plant sources of EPA and DHA, avoiding GM will mean that the Scottish industry will otherwise have to accept that it is vulnerable to the accusation of supplying a less healthy product than in those countries using GM feed ingredients. Use of by-products from terrestrial livestock production Despite its potential benefits (e.g. cost savings and formulation flexibility), and its widespread use in salmon farming in the Americas, there is strong supply chain resistance to incorporating terrestrial by-products into salmon feeds in the UK. EU approval has been granted for use of PAP in aquaculture feeds, although in commercial practice this would mean poultry material as porcine by-products would be problematic for religious and cultural reasons. However, UK retailers are unwilling to accept the high risk of a negative customer reaction. It is recognised that this policy restricts sourcing and runs counter to sustainability options, but the likelihood is that a more flexible approach to GM feed ingredients would come before acceptance of PAPs. Research bottlenecks Continuing focus is required on how best to maintain EPA and DHA levels in farmed salmon given the increased price and reduced availability of fish oil of appropriate certification status and lack of contaminants. The most feasible alternative source is likely to be GM oil by 2020, while algal and microbial sources need to be kept under review. Interim solutions are needed to manage this situation cost-effectively. Although alternative protein feed ingredients (e.g. beans and peas) are being studied, assuming these are suitable and cost-effective in salmon trials, it is important to be able to concentrate the protein in order for it to replace soy protein concentrate in practical feed manufacture. The technical process and associated logistics will therefore also need to be defined and offer a cost-effective solution What are the industry s strengths, weaknesses, opportunities, and threats ( SWOT analysis)? (i) The principal marketing strength of the Scottish salmon farming industry is that it represents a well-established premium niche, linked to a green and clean coastal environment and a globally respected traditional prime Scottish product, with a healthy image. Salmon marketing is dominated by UK retailers, who require that production is contracted to their own and third party standards; these are exacting and somewhat higher than would normally be expected (e.g. in 73

76 (ii) (iii) (iv) comparison with standard Norwegian production); this is a double-edged sword but helps to ensure the economic, social, and ecological efficiency of the industry. Despite a relatively higher use of marine ingredients than elsewhere, the Scottish industry can justifiably claim to be sustainable and safe (c.f. Marine Harvest s Blue Revolution message). At the same time the industry focus on research and innovation covers a wide range of salmon feed and quality-related issues, from nutrition to genetics, and should not be under-estimated as a strategic strength. The potential weaknesses of the industry are that its sustainability focus and policy on prohibiting the use of GM feed and land animal by-products mean that it has a relatively inflexible feed formulation policy and potentially higher costs compared with overseas competitors. It is also vulnerable to scarcity of omega-3 LC-PUFA which underpins the healthy image of salmon. There is currently an acute shortage of omega-3 LC-PUFA-rich fish oil which is expected to reverse very shortly, but may only be resolved longer term once GM-derived omega-3 LC- PUFA becomes available and current industry hostility to GM feed is reversed. Following from the above, the main threats are market-related. Firstly, the likely longer term pressure to accept GM feed ingredients in order to maintain healthy omega-3 LC-PUFA levels in Scottish salmon would risk losing the French export market which is anti-gm. In this connection the UK feed industry structure means that it could not realistically produce GM and non-gm feed at the same time, representing a potential competitive threat from Norway where mill logistics would allow this, offering greater cost flexibility and export flexibility versus Scotland. Also there is a risk of losing the industry s premium niche by a less differentiated commodity approach in order to compete with lower cost, more standard products, including Norwegian imports. It is suggested that the opportunity for Scottish salmon going forward is to seek to retain its healthy image, aligned to naturalness in terms of dietary focus (e.g. marine content) and fish welfare. In this connection the main challenges will be to continue to supply a specialist product range which appears differentiated from standard commodity product (for which the Norwegian industry has advantages of scale) and is beneficial to consumers; the latter is more than just about omega-3, but in any event it will require managing the limited availability of omega-3 LC-PUFA, including how to adapt policy and respond best to the likely availability of GM oil. Summary of Chapter 12 Farmed salmon is highly beneficial for human health and nutrition with its highly digestible protein, essential amino acid and fatty acid (including omega-3 LC-PUFA) composition, and micronutrient content (including lipid-soluble vitamins A, D and E). 74

77 In addition to being nutritious, its freedom from contaminants makes farmed salmon safe to eat. The growth in salmon farming has resulted in year-round availability of cost-efficient, affordable salmon products on a greatly increased scale, while creating wealth and sustainable employment, including to remote rural locations; the industry therefore demonstrates clear economic and social benefits. In terms of eco-efficiency, farmed salmon retain feed nutrients, and convert them to nutrients for human consumption, more efficiently than land animals. Life cycle analyses showed salmon farming has a lower carbon footprint and makes better use of resources than farming either pigs or chicken; also that using fishmeal and fish oil from capture fisheries provides more marine protein, energy, and EPA and DHA for human consumption than utilising these marine resources directly as a human food source. For the above reasons, ethical objections to using fish as feed are invalid since it represents a more efficient use of resources to harvest well managed but unmarketable species of fish for indirect human consumption via aquaculture than not to harvest them at all. As pointed out by FAO, in a protein-hungry world it is more efficient to utilise such fish for aquaculture feed; in the same way farmed salmon production is a highly efficient way of delivering essential marine fatty acids to human consumers. Continuing global growth in fed aquaculture has not increased pressure on wild fish resources, but instead there has been increased substitution of fishmeal and fish oil with plant protein and oil sources. Not that plant feed resources are necessarily more sustainable for farming salmon than using feed ingredients based on wild-caught marine resources, provided they are from sustainably managed stocks or from fishery by-products. Although fishmeal and fish oil supplies to the Scottish salmon feed industry are required by the supply chains to show that they are responsibly managed and renewable, these resources will always be finite and limiting, hence require careful stewardship. However, replacement of fish oil by plant oil in salmon feeds means a reduction in the omega-3 LC-PUFA content of farmed salmon. Evidence of environmental sustainability has become a method by which UK salmon supply chains can profile their credentials; this in turn has tended to encourage other firms to follow, leading over time to more demanding standards for farming and feed use being adopted by the industry. The result is that standards are continuing to be driven up across the industry, including the need to demonstrate responsible sourcing of feed ingredients. Looking ahead there is a need to ensure the sustainability criteria are appropriate for greater harmonisation between different standards and certification schemes. Nonetheless the evidence of responsible practice supports the view that the Scottish salmon industry is developing sustainably from the feed standpoint. 75

78 As regards future prospects and challenges, first and foremost is how best to maintain the omega-3 LC-PUFA content of salmon so as not to undermine its healthy reputation (possibly by using GM plant oil longer term, and possibly DHA produced by fermentation in the short term); also a supply of locally-grown protein concentrates would be beneficial. It is believed the advantages of embracing GM feed ingredients outweigh the disadvantages, while recognising that consumer attitudes may continue to prohibit use of LAPs/PAP. The feed research priorities follow logically from the above considerations. A SWOT analysis of the Scottish salmon industry identifies the main strength as being a well-established, international market identity; this is reinforced by evidence that it is safe and sustainable and benefits from a strong research base. The industry s main challenge is the scarcity of omega-3 LC-PUFA, which potentially undermines its reputation for healthy product. At the same time supply chain policies prohibiting GM feed ingredients and LAPs restrict formulation flexibility, increase costs, and conflict with the industry s focus on sustainability. The industry challenges are also market related, from losing its premium reputation in the UK by seeking to compete with more standard and lower cost products, to the specific threat of losing export markets (especially France) due to the GM issue. By contrast the key opportunity is to maintain and extend its reputation, including the benefits of husbandry and diet composition in regard to fish welfare and human health, supported by a continuing demonstrable commitment to sustainable development. 76

79 13. Main recommendations 1) The most critical priority is the omega-3 LC-PUFA problem. If possible top supply chain managers should examine all the options and strive for a shared supply chain consensus on how best to manage EPA and DHA scarcity. In the short-term this issue will temporarily resolve itself in Peru; in the medium-term algal DHA may play a useful role; in the long-term, this Report believes that in view of the food security and consumer health implications, it is becoming increasingly urgent for the Scottish industry to consider adopting a more flexible attitude to GM sources of EPA and DHA. 2) However, promoting the health attributes of Scottish salmon is not only about EPA and DHA. It should also take account of there being negligible contamination and the content of amino acids, vitamins and minerals, etc. At the same time more meaningful pack labelling on omega-3 LC-PUFA content, would be helpful to the market, but may be difficult to achieve in practice given biological, regulatory, and market issues. 3) Despite being technically complicated, attention should be drawn to the evidence of Scotland s responsible sourcing of those feed ingredients certified as sustainable (of both marine and plant origin) as this could help to balance the current misguided focus of some commentators and standards on the use of marine ingredients without recognising their benefits, while largely ignoring potential issues with plant ingredients. 4) Viewed in this light, the relatively higher marine content of some Scottish feeds (e.g. compared with Norwegian farmed product or standard Scottish product) directly reinforces a healthy and natural reputation (e.g. via omega-3 LC-PUFA, use of natural pigments, etc.) and is a worthwhile trade-off against the resulting greater exposure to feed ingredient cost fluctuations. 5) Indirectly feed-linked, the Scottish ethical commitment to fish welfare (e.g. via Freedom Food status) is a global differentiator (e.g. not only sustainable, safe, and healthy, but also compassionate ) and it demonstrates the industry s willingness to step beyond traditional boundaries of best practice. 6) Although forward supply of proteins is not seen as a major issue, there should be continuing focus on cost-effective local alternatives to non-gm soy protein concentrate in particular. 7) The overall market challenge is how best to retain and reinforce the wellestablished, valuable niche identity for Scottish salmon, offering quality and high value (including consumer health, responsible production, etc.) in the context of increased retail discount focus based largely on bulk standard salmon. Despite the omega-3 difficulty referred to, farmed salmon has an excellent story to tell and needs to do so more coherently. 77

80 8) For Scottish salmon to gain optimal commercial leverage in the longer term requires a better understanding of the UK salmon market size, structure, and demand dynamics in comparison with its key export markets, especially the USA, France, and China. 78

81 Acknowledgements We gratefully acknowledge the support of SARF, which commissioned this study (Project SP007), generously funded by SSPO and Scottish Government, as well as the encouragement and advice given by Richard Slaski and Sandra Gray (SARF) and by John Webster (SSPO).The project could not have been undertaken without the cooperation and openness of those interviewed in various countries, including a willingness to disclose unpublished data, and comment on interim drafts. For making available tables and figures to this Report, the authors are grateful to Andrew Jackson and IFFO, Lesley Jenkins and SSPO; Christian Meinich and Chr. Holtermann ANS; Niels Alsted and BioMar Group; Julia Brooks and Seafish; Petter Arnesen and Marine Harvest ASA; Ragnar Nystøyl and Kontali Analyse AS. Recent analyses of resource allocation by the Norwegian salmon farming industry have been highly relevant when compiling a Report focused on the Scottish salmon industry and we are pleased to acknowledge the work of Trine Ytrestøyl, Torbjørn Åsgård and their colleagues of Nofima s, Sunndalsøra laboratory, and of Erik Hognes and his colleagues of SINTEF s office in Bergen, with its implications for salmon feed and farming in Scotland. At the same time we have had strong support from many individuals and stakeholder organisations linked to the Scottish salmon farming industry, drawn from the associated salmon supply chains, trade organisations, academic research, government, NGO community etc. Although the following is not meant to be a comprehensive list of contributors, the authors wish to express their appreciation of contributions made to this study by the following individuals and organisations in particular: Farming and feed supply: Alan Sutherland and Dougie Hunter (Marine Harvest Scotland); Petter Arnesen (Marine Harvest ASA); John Barrington (Scottish Sea Farms); John Williamson and Richard McKinney (Skretting); Douglas Low (Ewos Ltd), Louise Buttle and Vivjan Crampton (Ewos Innovation); Guy Mace and Nick Bradbury (BioMar Ltd); Marie Hillestad, Vidar Gundersen, Kjell Måsøval, Trygve Sigholt (BioMar AS); Niels Alsted (BioMar Group); John Webster, Lesley Jenkins, and James Smith (SSPO). Ingredients/raw materials and supply chain trade organisations: Adam Ismail (GOED omega-3); Stephen Woodgate (Fabra); Jon Tarlebo, Andreas Nordgreen and George- Harald Lerøy (Norsildmel AL); Jorge Mora (Marine Ingredients Association - see IFFO/Peru); Richie Flynn (IFFA); Henrik Stenwig (FHL); Judith Nelson (AIC); Toby Parker (United Fish Products); Hazel Curtis, Julia Brooks and Richard Watson (Seafish Industry Authority); Andrew Mallison, Andrew Jackson, and Enrico Bachis (IFFO); Christian Meinich (Chr. Holtermann ANS) Downstream supply chain, including processing and retail: Ally Dingwall (Sainsbury s); Gareth Bennell (Aldi); Charlotte Maddocks (Tesco); Hannah Macintyre and Patrick Blow (Marks & Spencer); Huw Thomas (Morrison s); Stuart Smith (The Co-operative Group); Chris Brown (Asda); Jeremy Langley (Waitrose); Nigel Edwards 79

82 (Icelandic Seachill); Mike Mitchell, David Parker, and Iain Michie (Youngs Seafood); Jack-Robert Møller (Norwegian Seafood Council). Government agencies, Universities, NGOs, etc: Neil Auchterlonie (CEFAS); Renaud Wilson (Defra); Johnathan Napier (BBSRC, Rothamsted); Phillip Calder (University of Southampton); Bente Ruyter and Trond Storbakken (Norwegian University of Life Sciences); David Little and Richard Newton (University of Stirling); Piers Hart (WWF- UK); Joséphine Labat (WWF-France); Dawn Purchase (MCS); Hazel Curtis and Richard Watson (Seafish Industry Authority); Daniel Lee (Bangor University). 80

83 References Asche, F., Bjørndal, T The economics of salmon aquaculture. Second Edition, Blackwell Publishing Ltd., Hoboken, 248 pp. Bergheim, A., Åsgård, T., Waste production from aquaculture, in: Baird, D.J., Beveridge, M.C.M., Kelly, L.,A, Muir, J.F. (Eds.), Aquaculture and water resource management. Oxford, Blackwell Science, Oxford, pp Bendiksen, E.A., Johnsen, C.A., Olsen, H.J., Jobling, M., Sustainable aquafeeds: Progress towards reduced reliance upon marine ingredients in diets for farmed Atlantic salmon (Salmo salar L.). Aquaculture 314, Berntssen, M.H.G., Julshamn, K., Lundebye, A.-K., Chemical contaminants in aquafeed and Atlantic salmon (Salmo salar) following the use of traditional- versus alternative feed ingredients. Chemosphere 78, Bibus, D M Long chain omega-3s from forage fish provides value to farmed seafood. Lipid Technology, in press. DOI /lite Bjørkli, J., Protein og energiregnskap hos laks, kylling, gris og lam. (Protein and energy account in chicken, pig and lamb). Master s thesis at Agricultural University of Norway (in Norwegian). Byelashov, O.A., Griffin, M.E., Fish in, fish out: Perception of sustainability and contribution to public health. Fisheries 39, Callaway, R., Shinn, A.P., Grenfell, S.E., Bron, J.E., Burnell, G., Cook, E.J., Crumlish, M., Culloty, S., Davidson, K., Ellis, R.P., Flynn, K.J., Fox, C., Green, D. M., Hays, G., Hughes, A., Johnston, E., Lupatsch, I., Malham, S., Mendzil, A.F., Nickell, T., Pickerell, T., Andrew F. Rowley, A.F., Stanley, M., Tocher, D.R., Turnbull, J.F., Webb, G., Wootton, E., Shields, R., Review of climate change impacts on marine aquaculture in the UK and Ireland. Aquatic Conservation: Marine and Freshwater Ecosystems 22, Cohen, J.T., Bellinger, D.C., Connor, W.E., A quantitative risk-benefit analysis of changes in population fish consumption. American Journal of Preventive Medicine, 29, Crampton, V.O., Nanton, D.A., Ruohonen, K., Skjervold, P.-O., El-Mowafi, A., Demonstration of salmon farming as a net producer of fish protein and oil. Aquacult. Nutr. 16, European Food Safety Authority (EFSA) Panel on Dietetic Products, Nutrition and Allergies (NDA), Scientific opinion related to the tolerable upper intake level of eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA) and docosapentaenoic acid (DPA). EFSA J. 10, EFSA, Update of the monitoring of levels of dioxin and PCBs in food and feed. Scientific Report of European Food Safety Authority. EFSA J., 10, FAO, Code of conduct for responsible fisheries. Food and Agriculture Organization of the United Nations, Rome, 41 pp. FAO, The state of world fisheries and aquaculture Food and Agriculture Organization of the United Nations, Rome, 196 pp. 81

84 FAO, The state of world fisheries and aquaculture Food and Agriculture Organization of the United Nations, Rome, 209 pp. FAO, The state of world fisheries and aquaculture Food and Agriculture Organization of the United Nations, Rome, 223 pp. FAO/GIEWS, Food outlook, global information and early warning system. no. 10/11/12. October/November/December Food and Agriculture Organization of the United Nations, Rome. FAO/NACA, Farming the Waters for People and Food. Subasinghe, R.P., Arthur, J.R., Bartley, D.M., De Silva, S.S., Halwart, M., Hishamunda, N., Mohan, C.V., Sorgeloos P., (eds.) Proceedings of the Global Conference on Aquaculture 2010, Phuket, Thailand September FAO, Rome and NACA, Bangkok. 896 pp. FAO/WHO, Report of the joint FAO/WHO expert consultation on the risks and Benefits of fish consumption. Food and Agriculture Organization of the United Nations, Rome / World Health Organization, Geneva, 50 pp. Garcia, S.M., Kolding, J., Rice, J., Rocher, M-J., Zhou, S., Arimoro, T., Beyer, J.E., Borges, L., Bundy, A., Dunn, D., Fulton, E.A., Hall, M., Heino, M., Law, R., Makino, M., Rijmsdorp, A.D., Simard, F., Smith, A.D.M., Reconsidering the consequence of selective fisheries. Science 235, GOED, GOED (Global Organization for EPA and DHA) disputes JAMA Omega-3 meta- Analysis (Press Release published on 2 pp. GOED, Global recommendations for EPA and DHA intake (Rev 16 April 2014), pp. 20. Henriques, J., Dick, J.R., Tocher, D.R., Bell, J.G., Nutritional quality of salmon products available from major retailers in the UK: Content and composition of n-3 long-chain polyunsaturated fatty acids. Br. J. Nutr. 112, Hites, R.A., Foran, J.A., Carpenter, D.O., Hamilton, M.C., Knuth, B.A., Schwager, S.J., Global assessment of organic contaminants in farmed salmon. Science 303, Hixson, S.M., Parrish C.C., Anderson D.M., Full substitution of fish oil with camelina (Camelina sativa) oil, with partial substitution of fish meal with camelina meals, in diets for farmed Atlantic salmon (Salmo salar) and its effect on tissue lipids and sensory quality. Food Chem. 157, Hognes, E.S., Ziegler, F., Sund, V., Carbon footprint and area use of farmed Norwegian salmon. SINTEF report no F SINTEF Fisheries & Aquaculture, Trondheim, 30 pp. IFFO, The Marine Ingredients Organisation: Global standard for Responsible supply (IFFO-RS). In: IFFO [online]. IFFO, The Marine Ingredients Organisation: Fishmeal and Fish Oil Statistical Yearbook In IFFO [online]. IFFO, The Marine Ingredients Organisation: Fishmeal and Fish Oil Statistical Yearbook In IFFO [online]. IFFO, Ethoxyquin: the current situation. December 2014, Unpublished report. 82

85 ISSFAL, Report of the sub-committee on recommendations for intake of polyunsaturated fatty acids in healthy adults. International Society for the Study of Fatty Acids and Lipids, Brighton. Jackson, A.J., Fish in Fish out ratios explained. Aquacult. Europe 34, Jackson, A.J, Shepherd, C.J., The future of fishmeal and fish oil, in: Ryder, J., Ababouch, L., Balaban, M. (Eds), Second International Congress on Seafood Technology on Sustainable, Innovative and Healthy Seafood FAO Fisheries and Aquaculture Proceedings. No. 22. Food and Agriculture Organisation, Rome, pp Kaushik, S., Troell, M., Taking the fish-in fish-out ratio a step further. Aquaculture Europe 35, Kontali Analyse, Kristiansund NO-6517, Norway. New analysis; salmon world Lenfest, Little fish big impact. A report from the Lenfest Forage Fish Task Force. Lenfest Ocean Program, Washington D.C., 108 pp. Makkar, H.P.S., Tran, G., Heuze, V., Amkers, P., State of the art on use of insects as animal feed. Anim. Feed Sci. Technol. 197, Marine Harvest, Salmon farming industry handbook 2012, Oslo, 72 pp. Marine Harvest, Salmon farming industry handbook Oslo, 84 pp. Marine Scotland, An assessment of the benefits to Scotland of aquaculture. Imani Development, Edinburgh, 164 pp. Mondoux, S., Pitcher, T., Pauly, D., Ranking maritime countries by the sustainability of their fisheries, in: Alder, J., Pauly, D. (Eds), Fisheries Centre Research Report 16, Fisheries Centre, University of British Columbia, pp Mozaffarian, D., Rimm, E.B., Fish intake, contaminants, and human health evaluating the risks and the benefits. J. Am. Med. Assoc. 296, NRC, Nutrient Requirements of Fish and Shrimp. The National Academies Press, Washington DC, 376 pp. Naylor, R.L., Goldburg, R.J., Primavera, J.H., Kautsky, N., Beveridge, M.C.M., Clay, J., Folke, C., Lubchenco, J., Mooney, H. Troell, M., Effect of aquaculture on world fish supplies. Nature 405, Naylor, R., Burke, M., Aquaculture and ocean resources: raising tigers of the sea. Ann. Rev. Environ. Res. 30, Naylor, R.L., Hardy, R.W., Bureau, D.P., Chiu, A., Elliott, M., Farrell, A.P., Feeding aquaculture in an era of finite resources. Proc. Natl. Acad. Sci. USA, 106, Nofima, Resource utilisation of Norwegian salmon farming in 2012 and Tromsø, 35 pp. Nøstbakken, O.J., Hove, H.T., Duinker, A., Lundebye, A.-K., Berntssen, M.H.G., Hannisdal, R., Lunestad, B.T., Maage, A., Madsen, L., Torstensen, B.E., Julshamn, K., Contaminant levels in Norwegian farmed Atlantic salmon (Salmo salar) in the 13-year period from Environment. Int. 74,

86 OECD-FAO, Organisation for Economic Co-operation and Development / Food and Agriculture Organization of the United Nations Agricultural outlook , OECD Publishing and FAO. OECD-FAO, Organisation for Economic Co-operation and Development / Food and Agriculture Organization of the United Nations Agricultural outlook OECD Publishing, Paris, 324 pp. Olli, J.J., Breivik, H., Mørkøre, T., Ruyter, B., Johansen, J., Reynolds, P., Thorstad, O., Berge, G., Removal of persistent organic pollutants from Atlantic salmon (Salmo salar L.) diets: Influence on growth, feed utilization efficiency and product quality. Aquaculture 310, Olsen, R.L., Hasan, M.R., A limited supply of fishmeal: impact on future increases in global aquaculture production. Trends Food Sci Tech, 27, Onozaka, Y., Hansen, H., Tveterås, R., Salmon s position among consumers. Global Aquaculture Advocate, 15, Ruiz-Lopez, N., Haslam, R.P., Usher, S.L., Napier, J.A., Sayanova, O., Reconstitution of EPA and DHA biosynthesis in Arabidopsis: iterative metabolic engineering for the synthesis of n-3 LC-PUFAs in transgenic plants. Metab. Eng. 17, Sanden, M., Stubhaug, I., Berntssen, M.H.G., Lie, Ø., Torstensen, B.E., Atlantic Salmon (Salmo salar L.) as a net producer of long-chain marine ω-3 fatty acids. J. Agric. Fd. Chem. 59, Sargent, J.R., Tacon, A.G.J., Development of farmed fish: a nutritionally necessary alternative to meat. Proc. Nutr. Soc. 58, SCF, Scientific Committee Food Opinion of the SCF on the risk assessment of dioxins and dioxin-like PCBs in food. European Commission, Brussels, Update based on new scientific information available since the adoption of the SCF opinion of 22. November Adopted on 30 May 2001 SAMS, Review and Synthesis of the Environmental Impacts of Aquaculture. Scottish Association for Marine Science and Napier University in association with Scottish Executive SSPO, Sustainable Scottish salmon. Scottish salmon farming annual report 2013, Perth, 20 pp. SEDAR, SEDAR 40 Atlantic menhaden stock assessment report. SEDAR, North Charleston, 643 pp. Shepherd, C.J., Are the criticisms of aquaculture justified, - Feeding fish to fish? World Agriculture 3, Shepherd, C.J., Jackson, A.J., Global fishmeal and fish-oil supply: inputs, outputs and markets. J. Fish Biol. 83, Shepherd, C.J., Little, D.C., Aquaculture: are the criticisms justified? II Aquaculture s environmental impact and use of resources, with special reference to farming Atlantic salmon. World Agricult. 4,

87 Shepherd, C.J., Bachis, E., Changing supply and demand for fish oil. Aquaculture Econ. Manage. 18, Simopoulos, A.P., The importance of the omega 6/omega 3 fatty acid ratio in cardiovascular disease and other chronic diseases. Exp. Biol. Med. 233, Sissener, N.H., Julshamn, K., Espe, M., Lunestad, B.T., Hemre, G.I., Waagbø, R., Måge, A., Surveillance of selected nutrients, additives and undesirables in commercial Norwegian fish feeds in the years Aquacult. Nutr. 19, Smith, R L (1992) Elements of ecology, third edition, Harper Collins Publishers Inc. 617 pp. ISBN Scott, D (2010) A review of the regulation of salmon farming in Scotland. Stirling Aquaculture report to Sustainable Fisheries Partnership. 23pp. Available at: Sustainable Fisheries Partnership, Global sustainability overview of fisheries used for fishmeal and fish oil June 2012, 11 pp. Svejgaard, T., BioMar, Ewos await turning point in acceptance for GM fish feed. Intrafish Tacon, A.G., Metian, M., Global overview on the use of fishmeal and fish oil in industrially compounded aquafeeds: trends and future prospects. Aquaculture 285, Tacon, A.G.J., Metian, M., Fishing for aquaculture: Non-food use of small pelagic forage fish - A global perspective. Rev. Fisheries Sci. 17, Tacon, A.G.J., Metian, M., Fish matters: Importance of aquatic foods in human nutrition and global food supply. Rev. Fisheries Sci. 21, Torrissen, O., Olsen, R.E., Toresen, R., Hemre, G.I., Tacon, A.G.J., Asche, F., Hardy, R.W., Lall, S., Atlantic salmon (Salmo salar): The super-chicken of the sea? Rev. Fisheries Sci. 19, Torstensen, B., Espe, M., Sanden, M., Stubhaug, I., Waagbø, T., Hemre, G-I., Fontanillas, R., Nordgarden, U., Hevrøy, E.M., Olsvik, P., Berntssen, M.H.G., Novel production of Atlantic salmon (Salmo salar) protein based on combined replacement of fish meal and fish oil with plant meal and vegetable oil blends. Aquaculture 285, Turchini, G.M., Ng, W-K., Tocher, D.R., Fish oil replacement and alternative lipid sources in aquaculture feeds. CRC Press, Baton Rouge, 552 pp. VKM, Benefit-risk assessment of fish and fish products in the Norwegian diet an update. Scientific Opinion of the Scientific Steering Committee. VKM Report 15, Oslo, 293 pp. WCED, Our common future. World Commisssion on Envionment and Development. Welch, A., Hoenig, R., Steiglitz, J., Benetti, D., Tacon, A., Sims, N., O Hanlon, B., From fishing to the sustainable farming of carnivorous marine finfish. Rev. Fish. Sci. 18,

88 Woodgate, S., Land animal PAP for aquaculture feeds Briefing note, November Foodchain and Biomass Renewables Association (FABRA), unpublished Worm, B., Hilborn, R., Baum, J.K., Branch, T.A., Collie, J.S., Costello, C., Fogarty, M.J., Fulton, E.A., Hutchings, J.A., Jennings, S., Jensen, O.P., Lotze, H.K., Mace, P.M., McClanahan, T.R., Minto, C., Palumbi, S.R., Parma, A.M., Ricard, D., Rosenberg, A.A., Watson, R., Zeller, D., Rebuilding global fisheries. Science 325, Ytrestøyl, T., Aas, T.S., Berge, G.M., Hatlen, B., Sorensen, M., Ruyter, B., Thomassen, M., Skontorp Hognes, E., Ziegler, F., Sund, V., Åsgård, T., Resource utilisation and ecoefficiency of Norwegian salmon farming in NOFIMA report no. 53/2011. The Norwegian Institute of Food, Fisheries, and Aquaculture Research, Tromsø, 65 pp. Ytrestøyl, T., Aas, T.S., Åasgård, T., Resource utilisation of Norwegian salmon farming in NOFIMA report no. 36/2014. The Norwegian Institute of Food, Fisheries, and Aquaculture Research, Tromsø, 35 pp. 86

89 Tables and Figures Table 1. Species used for fishmeal and fish oil in Norwegian salmon feed in 2012 (Nofima, 2014). Forage fisheries Fish meal (tons) Fish oil (tons) Sum % of marine ingredients Capelin Sprat Sandeel Blue whiting Atlantic herring Atlantic mackerel Norway pout Anchoveta Chilean jack mackerel Gulf menhaden South American pilcard Boar fish Krill other/unknown species Sum forage fisheries % Trimmings/silage Herring Capelin Mackerel Whitefish Fish protein concentrates Unknown/other Sum marine ingredients

90 Table 2. Plant ingredients used in Norwegian salmon feed production in 2010 and 2013 (Nofima, 2014). Plant ingredients (tonnes) Protein sources Soy protein concentrate Wheat gluten Sunflowermeal Peaprotein concentrate Fababeans Dehulled horse beans 4442 Maize Sum plant protein sources Oil sources Rapeseed oil Other plant oils 0 0 Sum plant oil Binders Wheat Pea Tapioca 3396 Sum plant ingredients

91 Table 3. Product yield, energy, and protein retention in edible parts of Atlantic salmon, pig, chicken and lamb (Bjørkli, 2002). Atlantic salmon Pig Chicken Lamb Harvest Yield (%) a Edible Yield (%) b FCR c Energy Retention (%) d Protein Retention (%) e (a) Harvest Yield is yield gutted and bled animal (b) Edible yield is ratio of total body weight that is normally eaten, muscle, body adipose tissue and liver, lung and heart for pig. Skin is excluded for all animals. (c) FCR = (kg feed fed)/(kg body weight gain) (d) Energy retention + (energy in edible parts)/(gross energy fed) (e) Protein retention = (kg protein in edible parts)/(ky protein fed) Table 4. Production of non-ruminant processed land animal proteins in the UK and EU (including the UK) for 2013 (Woodgate, 2014). UK production (tonnes pa) EU production (tonnes pa) Poultry PAP* 75, ,000 Feathermeal PAP 25, ,000 Porcine PAP none 165,000 Porcine bloodmeal PAP 1,000 65,000 Porcine blood products 5,000 45,000 * PAP signifies Processed Animal Protein manufactured according to current EU regulations. 89

92 Table 5. World fishmeal and fish oil projections to 2022 (OECD-FAO, 2013). 90

93 Table 6. FAO fish model: overall trends to 2022 (FAO, 2014). 91

94 Fig. 1. Diagram of global production, consumption, and trade flows of farmed Atlantic salmon (tonnes, head-on gutted) for 2013 (Kontali Analyse, 2014). Global trade 2013: Farmed Atlantic salmon - world wide (In tonnes hog) Source: Kontali Analyse AS Norway, Iceland, Faroe Islands: Harvest: Market: Russia: Harvest: Market: North America: Harvest: Market: EU: Harvest : Market: Japan: Harvest: 0 Market: Other Asia: Harvest: 0 Market: Latin America: Harvest (CL) : Market: Australia & New Zealand: Harvest: Market:

95 Fig. 2. Annual world fishmeal production by major producer from 2005 to 2013 (tonnes x 1000) (IFFO). WORLD FISHMEAL PRODUCTION MAJOR PRODUCER ('000 mt) 7,000 6,000 5,000 6,035 Peru Chile Thailand U.S.A Japan Scandinavia* Others 5,630 5,629 5,286 5,124 4,892 4,562 4,556 4,675 4,000 3,000 2,000 1, *Includes Iceland, Denmark and Norway source: IFFO and ISTA Mielke GmbH, OIL WORLD Fig. 3. Annual world fish oil production by major producer from 2005 to 2013 (tonnes x 1000) (IFFO). WORLD FISH OIL PRODUCTION MAJOR PRODUCERS ('000 mt) 1,200 1, Peru Chile Scandinavia* U.S.A Japan Other 1,052 1,075 1,080 1, * Includes Denmark, Norway & Iceland source: IFFO and ISTA Mielke GmbH, OIL WORLD 93

96 Fig. 4. Global fed aquaculture production and its use of fishmeal and fish oil from 2000 to 2012 (FAO and IFFO). Fig. 5. The fishmeal and fish oil supply chains (Shepherd and Jackson, 2013). 94

97 Fig. 6. World fish oil consumption by aquaculture, direct human consumption, and other uses from 2004 to 2013 (IFFO) Trends of Use for Fish Oil (,000 mt) Aquaculture DHC Other Fig. 7. Use of fish oil by market segment in 2013 (IFFO). Fish Oil Usage by Market % 21% Aquaculture DHC Other 75% 95

98 Fig. 8. Use of fish oil by different categories of farmed fish and crustaceans in 2013 (IFFO). Fishoil use in Aquaculture 2013 Cyprinids Eels Crustaceans Marine fish Other Salmonoids Tilapias 0% 2% 5% 6% 23% 60% 4% Fig. 9. Use of fishmeal by aquaculture, chicken, pigs and other uses in 2013 (IFFO). Fishmeal Usage by Market % 20% Aquaculture Chicken 6% Pig Other 72% 96

99 Fig. 10. Use of fishmeal by different categories of farmed fish and crustaceans in 2013 (IFFO). Use of Fishmeal in Aquaculture 2013 Cyprinids Eels Crustaceans Marine fish Other Salmonoids Tilapias 8% 5% 5% 24% 28% 9% 21% Fig. 11 Projected increase in share of fishmeal from fishery by-products to 2021 (OECD and FAO, 2012). 97

100 Fig. 12. The ratio of weekly fishmeal (65 % FOT) and soyameal (44 % FOT) prices (RMB/mt C&F Shanghai, China) from 2010 to 2013 (Shepherd and Bachis, 2014) Fig.12 The ratio of weekly fishmeal (65% FOT) and soyameal (44% FOT) prices (RMB/mt C&F Fig. 13. The ratio of fish oil to rapeseed oil prices (US$/mt C&F Rotterdam, Holland) from 2010 to 2013 (Shepherd and Bachis, 2014) Fig. 13 The ratio of fish oil to rapeseed oil prices (US$/mt C&F Rotterdam, Holland) from 98

101 Fig. 14. Development of salmon feed composition in Norwegian salmon farming from 1990 to 2013 (Nofima, 2014). 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Diet composition (%) microingredients starch plant oil plant protein marine oil marine protein Fig. 15. Raw material ingredients used in Norwegian salmon feeds in 2010 and 2012 (Nofima, 2014). tons 400, , , , , , ,000 50,000 0 feed ingredients in Norwegian salmon feed

102 Fig. 16. Prices of fishmeal, fish oil, rapeseed oil, soymeal and wheat from 2006 to 2013 (Marine Harvest, 2014, after Holtermann). Fig. 17. Interaction of social, ecological, and economic factors in sustainable development (Nofima, 2014). 100

103 Fig. 18. Supply and demand for EPA and DHA in salmon feed in 2014 (Holtermann, 2014) EPA/DHA Supply/demand epa/dha content in oil fraction of feed - basis respectively 25% and 30% oil fraction in Chilean and Norwegian salmon feed - other consumptions as per actual quality/origin used Production Chile+Norway salmon feed Other Aqua Tech/hard Omega-3 Burning Fig. 19. Supply and demand for EPA and DHA in salmon feed in 2015 (Holtermann, 2014) ,0 mill quota (El Niño) Prod. 4,00 mill Peru N/C q2+q4 quotas EPA/DHA supply/demand epa/dha content in oil fraction of feed - basis respectively 25% and 30% oil frction in Chilean and Norwegian slamon feed - other consumption as per actual quality/origin used 4,5 mill quota % epa/dha 3 % epa/dha 6 % epa/dha in salmon feed fat.. 7,5 % epa/dha Production Chile+Norway salmon feed Other Aqua Tech/hard Omega-3 Burning 101

104 APPENDIX 1 Main auditable standards and certification schemes currently being used in the farmed salmon supply chain for feed and marine ingredient-related matters (modified after IFFO) (i) Marine Stewardship Council (MSC) ( Unit of certification: individual fishery - claim of sustainable Accreditation standard: ISEAL Sponsors: Pew Foundation; WWF but UK based Positive: pioneer; seen as gold standard for assessing sustainable fisheries Negative: high expense and time-consuming process Ecolabel: Offers associated Chain of Custody Best known fishery standard, but very small volumes of whole fish from MSC certified fisheries are being made available for fishmeal and fish oil production (ii) (iii) (iv) Aquaculture Stewardship Council (ASC) ( Unit of certification: individual farm claim of responsible Accreditation standard: ISEAL Sponsors: Pew Foundation; WWF; sister of MSC, but Dutch based Positive: Seen as very demanding; high future credibility likely for salmon farming Negative: No agreed feed module yet Ecolabel: chain of custody Friend of the Sea ( Unit of certification: individual fishery, fish farm or products, etc. claim of sustainable Accreditation standard unknown: verbally claim to be ISO compliant but not certified Privately owned Italian NGO Positive: strong following in Southern Europe and wide-ranging activities Negative: seen globally as not demanding and with poor credibility Ecolabel GAA s Best Aquaculture Practice (BAP) ( Unit of certification: individual fish farm; claim of certified best practice Accreditation standard: ISO guide 65 Owned by industry-led Global Aquaculture Allliance in USA Positive: strong industry following, especially in the Americas (and shrimp) Negative: lack of recognition by European salmon farmers and some retailers Business-to-business scheme; Best aquaculture Practices Certified mark; 5 star approach to integrated chain 102

105 (v) (vi) Five different BAP standards covering production chain; the feed mill standard requires 50% of marine ingredients to be certified and both MSC and IFFO RS are recognised. Global Gap ( Aquaculture is an offshoot from agriculture standard certification: covers entire production chain from brood-stock to post-harvest Accreditation standard: EN45011 or ISO 65 Owned by an NGO (Global partnership for safe and sustainable agriculture) Positive: growing in aquaculture; focus on food safety issues on farms and processing plants Negative: entry level low; less focused on environmental than safety issues Business-to-business scheme with Chain of Custody Marine Ingredients Organisation - IFFO RS ( Unit of certification: fishmeal factory; claim of IFFO assured = Responsible Supply Accreditation standard: ISO guide 65 Owned by International Fishmeal & Fish Oil Organisation (also known as Marine Ingredients Organisation); non-profit trade organisation Positive: tailor made for fishmeal and fish oil industry Negative: Uncertainty about future status related to ASC Business-to-Business scheme with Chain of Custody; responsible sourcing and safe production Recognised by GAA s BAP feed mill standard; partially recognised by ASC 103

106 Appendix 2 Farm and feed-related certification used in Scottish salmon supply chains 1. Aquaculture responsible practice Code of Good Practice for Scottish Finfish aquaculture accepted by c. 95 % of Scottish salmon farms. ASC restricted so far to Marine Harvest (two sites now certified) Label Rouge premium French export; small niche but undergoing mini-revival Organic production under EU rules (e.g. Soil Association; Organic Food Federation) Global Gap focus on food safety; most farms certified, including all Marine Harvest, SSF, Scottish Salmon Co., Cooke Aquaculture, etc. RSPCA Freedom Food % of Scottish salmon biomass is Freedom Food certified (otherwise producers belong to Farm Assurance schemes). This is more than just about welfare, linking into Responsible Practice. Freedom Food requires the use of IFFO RS for marine ingredients sourcing and strives for sustainable feed. 2. General farm and feed assurance schemes Global Gap has its own feed assurance scheme UFAS covers off this requirement in the case of feed suppliers UFAS all UK feed companies; this in turn requires raw material to be FEMAS or GMP Quality management schemes e.g. ISO Feed ingredients 3.1 Marine ingredients IFFO RS near universal now, including Chain of Custody for feed producers MSC very limited raw material, so hardly used in practice (more trimmings-related) ASC and also Sustainable Fisheries Partnership 3.2 Plant ingredients Pro-Terra Soya increasing adoption RTRS increasing adoption for soya but work in progress Cert-ID to certify non-gmo status [in practice used to control specification, sourcing, and traceability, rather than Non-GM status per se) 4. Processor/Retailer The following companies have their own Codes of Practice: Waitrose (100 % bespoke); Marks & Spencer (100 % bespoke); Sainsbury (100 % bespoke); Co-operative (not as well defined but some specific requirements); Tesco (bespoke). 5. Not used currently Global Aquaculture Alliance BAP includes Feed Mill standard Friend of the Sea limited to occasional US menhaden imports; undemanding 104

107 Appendix 3 The following health claims for EPA and DHA have been approved by EFSA (Commission Regulation (EU) 1924/2006 and 432/2012): DHA and EPA contribute to the normal function of the heart (0.25 g per day) DHA and EPA contribute to the maintenance of normal blood pressure (3 g per day) DHA and EPA contribute to the maintenance of normal blood triglyceride levels (2 g per day) DHA contributes to maintenance of normal blood triglyceride levels (2 g per day in combination with EPA) DHA contributes to maintenance of normal brain function (0.25 g per day) DHA contributes to the maintenance of normal vision (0.25 g per day) DHA maternal intake contributes to the normal brain development of the foetus and breastfed infants (0.2 g DHA plus the daily recommended intake of omega-3 fatty acids (EPA+DHA for adults which is 0.25 g per day) DHA maternal intake contributes to the normal development of the eye of the foetus and breastfed infants (0.2 g DHA plus the daily recommended intake of omega-3 fatty acids (EPA+DHA) for adults which is 0.25 g per day 105

108 Technical Annexes Annex A. Atlantic Salmon Nutrition Atlantic salmon like other essentially carnivorous species have poor ability to digest carbohydrate as this macronutrient group is not a major component of natural the diet. As a consequence, efficient feeds must be based almost exclusively on protein and lipid and, consequently, salmon can be considered to have a relatively high requirement for these macronutrients. However, it is generally not appropriate to specify definitive requirement levels for either protein and lipid as these will depend on the overall energy content of the diet. However, in addition to energy, the major macronutrients, protein and lipid, also supply several key nutrients that cannot be synthesised endogenously and so are essential components of the diet, the essential amino acids (EAA) and essential fatty acids (EFA), for which there are specific requirement levels. In addition the diet must contain several key minerals and other components that cannot be synthesised endogenously such as vitamins. It is perhaps surprising considering the size of the industry globally, that the requirement levels for many of these essential components of the diet have not been properly empirically determined in Atlantic salmon and that many are extrapolated from studies on other species, particularly rainbow trout (NRC, 2011). While this has not been a major problem while feeds were formulated with high levels of FM and FO, that closely reflected the natural diet, it has become an increasingly important issue as levels of the marine ingredients have decreased to be replaced by terrestrially-sourced ingredients with different nutrient compositions (Glencross et al., 2007). This has prompted considerable national and, especially international research into fish nutrition including the major EU FP5 (PEPPA, ARRAINA, FORM etc., individual FM and FO replacement), FP6 (AQUAMAX, dual FM/FO replacement) and FP7 (ARRAINA, micronutrients) projects. In addition to the effects of changing raw materials on aquaculture production there is increasing awareness of the importance of these dietary changes on the fish itself and therefore the link between fish nutrition and fish health has been the subject of several recent reviews (Verlhac-Trichet, 2010; Montero and Izquierdo, 2011; Kiron, 2012; Oliva-Teles, 2012; Pohlenz and Gatlin, 2014; Tocher and Glencross, 2015). A.1 Protein content Fish, like other animals do not have a requirement for protein per se. Protein is required in the diet to provide EAA and amino groups for synthesis of non-essential amino acids (NEAA) and deficiencies in EAA lead to reduced growth and feed efficiency. In feeds, protein generally refers to crude protein (CP), calculated as N 6.25, based on the assumption that proteins contain 16 % N. The nutritional value of proteins is a function of digestibility and amino acid composition, which vary markedly between different protein sources. In addition, amino acids play an important role in supplying energy and are generally an efficient fuel for most fish species. Therefore, the digestible protein (DP) to digestible energy (DE) ratio has been suggested as a more rational means of expressing protein requirement than percentage CP of the diet, although this is limited 106

109 as amino acid composition of the protein and amino acid requirements are not a function of diet DE (Bureau and Encarnaçao, 2006). Thus, dietary DE content varies with different energy-yielding nutrients (e.g. lipids and carbohydrates) that do not have the same ability to be utilised by fish and spare dietary amino acids (Encarnacao et al., 2006). However, increasing the lipid content of the diet can reduce dietary protein (i.e. amino acid) catabolism and requirement, commonly referred to as protein-sparing. Within the context of the above, the dietary digestible energy and protein requirements of Atlantic salmon, that is the minima required to achieve maximum growth, have been estimated at 4,400 (kcal/kg diet) and 36 % digestible CP, respectively (Table A1). However, this is over the whole life/growth cycle and protein requirements are greater in smaller (younger) fish, with recommended dietary protein levels decreasing from 48 % in salmon under 20 g to 34 % in fish greater than 1.5 kg (NRC, 2011). It should be noted that while the protein requirement in terms of dietary concentration (% of diet) is higher in fish than in mammals, the absolute requirement is generally not (g/kg body weight gain) as salmon/fish have a lower absolute energy requirement, which results in similar g body weight gain/g protein ingested as mammals, but much greater feed efficiency (gain:feed). Protein requirement as a concentration (% protein) or as a function of the DE content of the diet (DP/DE) is often higher in smaller fish as described above, but higher absolute protein levels (g protein/kg biomass gain) can also be required by some species as fish grow larger. However, reduced protein retention efficiency in larger fish was not reported in Atlantic salmon (Azevedo et al., 2004). Although feeding and growth functions generally increase with water temperature, digestible protein requirement is largely unaffected by temperature or other environmental factors such as salinity. A.1.1 Amino acid requirements. Over the years there has been great debate on the ideal way of expressing EAA requirements with both per unit of DE (g/mj DE) and as a proportion of dietary protein (g/16g N) being advocated (NRC, 2011). However, both ways have their limitations and so the recent NRC Committee recommended expressing EAA requirements as a percentage of diet on a dry matter basis (NRC, 2011). On this basis, the current best estimates of the quantitative EAA requirements of Atlantic salmon as reported in the recent NRC report are shown in Table A1. It can be clearly seen that these requirements and those for Pacific salmon (Oncorhynchus spp.) almost totally parallel those of rainbow trout showing their origin as largely extrapolation from empirical studies in trout. However, specific requirement studies on arginine, lysine, methionine and threonine have been performed in Atlantic salmon (see NRC, 2011). Histidine requires special mention at it is generally required at about 0.8 % in ongrowing fish, but is required at higher levels of around 1.4 % in post-smolts undergoing particularly rapid growth just after seawater transfer to prevent ocular pathology, specifically bilateral cataracts. Taurine is not included in Table A1 as it is not a normal amino acid but an amino-sulphonic acid derivative of cysteine, however it has been reported to be required by rainbow trout fry fed diets formulated with 100 % plant proteins and is conditionally essential for many marine fish (NRC, 2011). 107

110 A.2 Lipid content Although an optimum level of dietary lipid cannot be truly defined, there is a range within which dietary lipid should be supplied. The lower limit represents the amount of dietary lipid required to supply the requirements for essential lipid components such as EFA that will depend upon the precise lipid source(s) used and their corresponding lipid class and fatty acid profiles. Increasing dietary lipid above this minimum level will support higher growth rates, partly based on protein sparing as described above, and this is the biological mechanism underpinning the increasing use of high energy (oil) feeds in aquaculture and salmon farming in particular. However, dietary lipid will reach an upper limit where the lipid level exceeds the capacity of the fish to effectively digest all the dietary lipid resulting in reduced lipid digestibility and, in effect, wasting space in the feed that could be occupied by other nutrients (e.g. protein). However, even if the lipid level does not exceed digestive capacity, it must still be balanced with DE requirements. If not, and dietary lipid exceeds DE requirements, this leads to unwanted deposition of lipid in adipose tissue stores in the peritoneal cavity, liver or other tissues (Company et al., 1999; Craig et al., 1999; Gaylord and Gatlin III, 2000). This is inefficient as supplying an energy-yielding nutrient that is simply deposited unused in tissue stores is, in effect, wasted energy. Although lipid deposits contribute to increased weight, as this is not flesh (muscle), it is not contributing to yield. This is highlighted in species such as Atlantic cod (Gadus morhua) that deposit lipid in the liver or other species with large perivisceral stores. However as an oily fish, Atlantic salmon deposit significant amounts of lipid in the flesh and this partly underpins the ability of salmon to tolerate and utilise the higher dietary lipid levels associated with high energy feeds. However, excess lipid deposition in the flesh can also cause problems in salmon resulting in gaping, oil leakage and other issues for processors (Bell et al., 1998; Hillestad et al., 1998). A.2.1 Essential fatty acids. In addition to providing energy, dietary lipid also supplies EFA. All organisms can synthesise saturated and monounsaturated fatty acids such as 16:0, 18:0 and 18:1n-9 but vertebrates, including fish, cannot synthesise C 18 polyunsaturated fatty acids (PUFA) such as linoleic acid (LOA, 18:2n-6) or α-linolenic acid (LNA, 18:3n-3) because they lack the Δ12 (or ω6) and Δ15 (or ω3) desaturases responsible for their production from 18:1n-9. However, long-chain (LC)-PUFA, most commonly defined as PUFA with C20 and 3 double bonds, including eicosapentaenoic acid (EPA, 20:5n-3), docosahexaenoic acid (DHA, 22:6n-3) and arachidonic acid (ARA, 20:4n-6) are the main biologically active PUFA (Tocher, 2003). However, C 18 PUFA, LNA and LOA, can be converted to the biologically active LC-PUFA in vertebrates although this varies with species dependent upon the presence and expression of genes of fatty acid desaturation and elongation (Tocher, 2003) (Fig. A1). In consequence, which PUFA can satisfy EFA requirements varies similarly with species (Glencross, 2009; Tocher, 2010). Salmon express all the enzyme activities necessary for the production of LC- PUFA from C 18 PUFA and so LNA can satisfy n-3 (or omega-3 ) EFA requirements (~ 1.0 % of diet dry weight), although EPA and DHA can satisfy requirements at a lower level (Table A1). The specific requirement for n-6 PUFA or ARA has so far not been defined. 108

111 Like all fatty acids, LC-PUFA have important structural roles as constituents of phospholipids that are the major components of cellular biomembranes, and confer various functional properties by affecting both physico-chemical properties of the membrane (e.g. fluidity) as well as influencing membrane protein (e.g. receptors, carriers and enzymes) functions (Tocher, 1995). Similarly, as components of triacylglycerols (TAG) and other storage lipids such as wax esters, LC-PUFA can function as an energy reserve, with energy recovered through fatty acid β- oxidation in mitochondria (Tocher, 2003). In addition, however, underpinning the essentiality of PUFA, LC-PUFA have more specific functional roles as important regulators of metabolism either as themselves or their derivatives. Regulation can be extracellular at the level of tissues through LC-PUFA derivatives including eicosanoids such as prostaglandins, leukotrienes, lipoxins, resolvins and protectins, or intracellular as ligands for transcription factors that control gene expression (Tocher, 2010). These regulatory roles underpin the importance of LC- PUFA as belonging to either the n-6 or n-3 series, the concept of balance in dietary intake of the two PUFA series that cannot be interconverted in vertebrates (see Fig. A1), and the impact that dietary LC-PUFA can have on both health and disease (Lands, 2014). A.2.2 Phospholipids. It has been known for over 30 years that inclusion of phospholipids in the diet could improve performance of many freshwater and marine fish species. The primary beneficial effect is improved growth in both larvae and early juveniles, but also increased survival rates and decreased incidence of malformation in larvae, and perhaps increased stress resistance (Tocher et al., 2008). In Atlantic salmon, studies showed that the phospholipid requirement was 4-6 % of diet in fish of initial size 180 mg, 4 % in fish of g initial weight, whereas in fish of 7.5 g initial weight, no requirement was observed (Poston 1990a,b, 1991). A requirement for dietary phospholipids has not been established for adult salmon, although this has been virtually unstudied. The mechanism underpinning the role of the phospholipids in larval and early juvenile fish is not fully understood but appears to be independent of EFA requirements and not related to the delivery of other essential dietary components such choline, inositol or phosphorus (Tocher et al., 2008). Studies also suggested that the phospholipid effect was not due to generally enhanced emulsification and digestion of lipids. Current evidence suggests that early stages of fish have impaired ability to transport dietary lipids away from the intestine possibly through limitations in lipoprotein synthesis and the current hypothesis is that the enzymatic location of the limitation is in phospholipid biosynthesis (Tocher et al., 2008). Thus, dietary phospholipids increase the efficiency of transport of dietary fatty acids and lipids (energy) from the gut to the rest of the body through enhanced lipoprotein synthesis. A.2.3 Cholesterol. Although there have been a few early studies investigating dietary supplementation of cholesterol in salmonids, primarily in relation to the development of coronary arteriosclerotic lesions (Farrell and Munt, 1981; Farrell et al., 1986), there are few reports on the effects of dietary cholesterol on growth or metabolism in fish. However, the effects of dietary cholesterol supplementation on growth, organ indices, digestibility of 109

112 cholesterol and macronutrients, and tissue fatty acid compositions have been investigated in Atlantic salmon (Bjerkeng et al., 1999). This study showed that dietary cholesterol level had no significant effect on specific growth rate (SGR), mortality, apparent digestibility coefficients of macronutrients, and total lipid content. However, both hepatic cholesterol concentration and hepatosomatic index were increased by the dietary cholesterol supplement. In a recent study, it was shown that genes of the cholesterol biosynthesis pathway in liver were up-regulated in salmon fed a vegetable oil blend compared to fish fed a diet containing fish oil (Taggart et al., 2008). A.3 Carbohydrate Carbohydrates include sugars, mono- and disaccharides, oligosaccharides and polysaccharides including starch (storage carbohydrate) and cellulose (structural and the major component of crude fibre in diets). Energy is obtained from carbohydrates by the oxidation of monosaccharides through the glycolytic pathway and the tri-carboxylic cycle. The enzymes for these pathways and also those of gluconeogenesis (glucose production from amino acid and lipid catabolism), glycogenesis (glucose storage as glycogen) and glycogenolysis (glucose mobilisation from glycogen) are all present in most fish species and so glucose can function as an energy source in fish as it does in mammals. However, based on glucose tolerance tests that show the clearance rate of glucose from the blood is relatively slow, teleost fish are considered as glucose intolerant compared to mammals although the precise biochemical/physiological reason is not clear. Furthermore, sugars are highly water-soluble and so cannot be efficient components of fish feeds and so carbohydrate can only be included in feeds as water-insoluble starch or partially digested/hydrolysed starch products. These products are cheaper than protein or lipid but their ability to effectively supply energy varies greatly with species dependent upon digestibility (reviewed Krogdahl et al., 2005), with herbivorous and omnivorous fish having higher digestibility for raw starch (e.g. 90 % and % in carp and tilapia, respectively) than carnivorous species with salmon generally showing values of < 60% (Thodesen and Storebakken, 1998) with digestibility decreasing with increasing inclusion level (Arnesan et al., 1995; Hillestad et al., 2001). Thus, the effectiveness of dietary carbohydrate (starch) to promote protein-sparing varies with species and is limited in carnivorous species based on factors that may include digestive amylase level, poor blood clearance, number of insulin receptors and regulation of gluconeogenesis (NRC, 2011). Non-starch polysaccharides (NSP, e.g. cellulose) have very limited digestibility in all fish and so play only a very minor role as energy sources in fish although some can be added to feeds as binders or stabilisers of pellets (NRC, 2011). High levels of NSP can have detrimental effects in fish and, as higher levels of can be introduced into fish diets as a result of the increasing use of plant products and so there has been greater focus on dietary NSP in recent years. Overall, the maximum recommended level for digestible starch in salmonids is around % with the level of dietary fibre restricted to no more than 10 % or lower (NRC, 2011). The application of exogenous enzymes, including carbohydrases, as feed additives to improve nutrient digestibility of plantbased feedstuffs has been studied extensively in poultry and swine and is now used throughout 110

113 the world (Adeola and Cowieson, 2011). However, the use of carbohydrase enzymes to reduce the anti-nutritional effects of NSP and increase the utilisation of carbohydrates in feeds has not been extensively applied in aquaculture, despite their positive effects on nutrient digestibility (Adeola and Cowieson, 2011). The use of carbohydrase enzymes as feed supplements in aquaculture has been the subject of a recent review (Castillo and Gatlin, 2015). A.4. Minerals and vitamins Minerals are elements that are essential for life and that constitute the inorganic component of the diet. Determination of the mineral requirements of fish is complicated by the fact that some can be present in the water and be absorbed through the gills or taken up by drinking (in fish in seawater). Different levels of minerals in fresh and sea water and the different osmotic regulatory physiology in fish depending upon water salinity are particularly pertinent issues in anadromous species such as Atlantic salmon. Furthermore, difficulty in formulating suitable test diets with low mineral contents, leaching from diets prior to consumption and limited data on bioavailability all complicate the determination of accurate quantitative requirements for minerals. Vitamins are organic compounds that are also essential for life and that generally cannot be produced endogenously by fish although some can be derived from other essential nutrients or from gut microorganisms and this can spare a portion of dietary requirements. Growth performance is not the only or, in some cases, best parameter for quantifying requirements for minerals and vitamins as absence of known deficiency signs, tissue contents, development (skeletal deformities), enzyme activities, lipid content and oxidation and other functional assays have all been used. A.4.1 Macrominerals. The macrominerals are those required at relatively high concentrations for formation of skeletons, electron transfer and membrane potentials, acid-base regulation and osmoregulation including calcium, phosphorus, potassium, sodium, chlorine (chloride) and magnesium. Requirement levels of Atlantic salmon for macrominerals as reported in the recent NRC report (NRC, 2011) are shown in Table A1. Other than phosphorus and, to some extent, magnesium (required in freshwater but likely not in seawater that has a high level of magnesium), the current data are largely extrapolated from rainbow trout data. A.4.2 Microminerals. The microminerals or trace minerals function as components of enzymes and hormones including cofactors or activators of enzyme activity, participating in a wide range of metabolic processes. As such, they are required in much lower concentrations than macrominerals and include copper, iodine, iron, manganese, selenium, zinc, cobalt, chromium and molybdenum although requirement levels for only the first six are contained in the recent NRC report (NRC, 2011) and listed for salmonids in Table A1. A.4.3 Fat-soluble vitamins. The fat-soluble vitamins including vitamins A (retinol), D (1,25- dihydrocholecalciferol), E (α-tocopherol) and K (napthoquinone) can be stored in the body and 111

114 the time to deplete them varies with species, which can complicate defining quantitative requirements. Most of these vitamins can occur in different chemical forms that can be present at different levels in different feed ingredients. For instance, natural sources of vitamin D are ergocalciferol (vitamin D2) in plants and cholecalciferol (vitamin D3) in animals with D3 generally having higher bioavailability (Lock et al., 2010). Carotenoids such as b-carotene and others may be able to be converted into vitamin A in some fish species although this has not been quantified. Deficiency signs have been described in salmonids, often including Atlantic salmon, for all the fat-soluble vitamins (NRC, 2011). However, quantitative requirements for vitamins A and D have not been accurately defined in Atlantic salmon, but minimum requirement levels have been reported for vitamins E and K (Table A1). A.4.4 Water-soluble vitamins. The water-soluble vitamins are not stored in the body and so a constant dietary supply is required. While deficiency signs for all the water-soluble vitamins have been reported for salmonids, occasionally including salmon, there are few empirically derived data on the quantitative minimum dietary requirements for Atlantic salmon (NRC, 2011). Consequently, other than vitamin B6 (Pyridoxine; Lall and Weerakoon, 1990) and vitamin C (ascorbic acid; Sandnes et al., 1992) for which dietary requirements have been defined, the levels currently used in Atlantic salmon feeds have been extrapolated from data on other salmonids including rainbow trout and Pacific salmon (Table A1). A.4.5. Astaxanthin (carotenoid pigment). In salmonids, two oxycarotenoids, astaxanthin (3,3'- dihydroxy-4,4'-diketo-β-carotene) and canthaxanthin (4-4'-diketo-β-carotene) are responsible for the red to orange pigmentation of the flesh, skin, and fins. Astaxanthin, the main carotenoid pigment of wild salmonids including Atlantic salmon (Christiansen et al., 1995), is derived mainly from zooplankton. Carotenoids may have several biochemical and physiological functions including antioxidant and singlet oxygen quenching, protection against photosensitisation, provitamin A activity, effects on gene expression and enhancement of immune functions (Bendich, 1993). Dietary astaxanthin improved growth of Atlantic salmon (Torrissen, 1984; Christiansen et al., 1994) and therefore it has been suggested to be considered as a vitamin (Torrissen and Christiansen, 1995). However, further qualitative and, especially quantitative, data and evidence of deficiency symptom associated with absence of astaxanthin is required before it can be considered to be an essential dietary nutrient. 112

115 References Adeola, O., Cowieson, A.J., Opportunities and challenges in using exogenous enzymes to improve non-ruminant animal production. J. Anim. Sci. 89, Arnesen, P., Krogdahl, Å., Sundby, A., Nutrient digestibilities, weight gain and plasma and liver levels of carbohydrate in Atlantic salmon (Salmo salar, L) fed diets containing oats and maize. Aquacult. Nutr. 1, Azevedo, P.A., Leeson, S., Cho, C.Y., Bureau, D.P., Growth and feed utilization of large size rainbow trout (Oncorhynchus mykiss) and Atlantic salmon (Salmo salar) reared in freshwater: Diet and species effects, and responses over time. Aquacult. Nutr. 10, Bell, J.G., McEvoy, J., Webster, J.L., McGhee, F., Millar, R.M., Sargent, J.R., Flesh lipid and carotenoid composition of Scottish farmed Atlantic salmon (Salmo salar). J. Agric. Food Chem. 46, Bendich, A., Biological functions of dietary carotenoids. Ann. N. Y. Acad. Sci. 691, Bjerkeng, B., Storebakken, T., Wathne, E., Cholesterol and short-chain fatty acids in diets for Atlantic salmon Salmo salar (L.): effects on growth, organ indices, macronutrient digestibility, and fatty acid composition. Aquaculture Nutr. 5, Bureau, D.P., Encarnaçao, P.M., Adequately determining the amino acid requirements of fish: The case example of lysine, in: Cruz-Suarez, L.E., Ricque-Marie, D., Tapia-Salazar, M., Nieto- Lopez, M.G., Villareal Cavazos, D.A., Puella Cruz, A.C., Garcia-Ortega, A. (Eds.), Avances en Nutricion Acuicola, VIII Simposium Internacional de Nutricion Acuicola. Mazatlan, Sinoloa, Mexico, November. Universidad Autonoma de Nuevo Leon, Monterrey, Nuevo Leon, Mexico, pp Castillo, S., Gatlin III, D.M., Dietary supplementation of exogenous carbohydrase enzymes in fish nutrition: A review. Aquaculture 435, Christiansen, R., Lie, Ø., Torrissen, O.J., Effect of astaxanthin and vitamin A on growth and survival during first feeding of Atlantic salmon Salmo salar L. Aquacult. Res. 25, Christiansen, R., Struksnæs, G., Estermann, R., Torrissen, O.J., Assessment of flesh colour in Atlantic salmon, Salmo salar L. Aquacult. Res. 26, Company, R., Calduch-Giner, J. A., Kaushik, S., and Perez-Sanchez, J. (1999). Growth performance and adiposity in gilthead sea bream (Sparus aurata): risks and benefits of high energy diets. Aquaculture 171, Craig, S.R., Washburn, B.S., Gatlin III, D.M., Effects of dietary lipids on body composition and liver function in juvenile red drum, Sciaenops ocellatus. Fish Physiol. Biochem. 21, Encarnaçao, P., de Lange, C.F.M., Bureau, D.P., Diet energy source affects lysine utilization for protein deposition in rainbow trout (Oncorhynchus mykiss). Aquaculture 261, Farrell, A.P., Munt, B., Dietary induced hypercholesterolemia and arteriosclerosis in brook trout (Salvelinus fontinalis). Physiologist 24, 43. Farrell, A.P., Saunders, R.L., Freeman, H.C., Mommsen, T.P., Arteriosclerosis in Atlantic salmon. Effects of dietary cholesterol and maturation. Arteriosclerosis 6, Gaylord, T.G., Gatlin III, D.M., Dietary lipid level but not L-carnitine affects growth performance of hybrid striped bass (Morone chrysops x M. saxatilis). Aquaculture 190,

116 Glencross, B.D., Exploring the nutritional demand for essential fatty acids by aquaculture species. Rev. Aquacult. 1, Glencross, B.D., Booth, M., Allan, G.L., A feed is only as good as its ingredients a review of ingredient evaluation strategies for aquaculture feeds. Aquacult. Nutr. 13, Hillestad, M., Johnsen, F., Asgard, T., Protein to carbohydrate ratio in high-energy diets for Atlantic salmon (Salmo salar L.). Aquacult. Res. 32, Hillestad, M., Johnsen, F., Austreng, E., Asgard, T., Long-term effects of dietary fat level and feeding rate on growth, feed utilization and carcass quality of Atlantic salmon. Aquacult. Nutr. 4, Kiron, V., Fish immune system and its nutritional modulation for preventive health care. Anim. Feed Sci. Technol. 173, Krogdahl, Å., Hemre, G.-I., Mommsen, T.P Carbohydrates in fish nutrition: digestion and absorption in postlarval stages. Aquacult. Nutr. 11, Lall, S.P., Weerakoon, D.E.M., Vitamin B 6 requirement of Atlantic salmon (Salmo salar). FASEB J. 4, Lands, B., Historical perspectives on the impact of n-3 and n-6 nutrients on health. Prog. Lipid Res. 55, Lock, E.-J., Waagbø, R., Wendelaar Bonga, S., Flik, G., The significance of vitamin D for fish: a review. Aquacult. Nutr. 16, Montero, D., Izquierdo, M., Welfare and health of fish fed vegetable oils as alternative lipid sources to fish oil, in: Turchini, G.M., Ng, W.-K., Tocher, D.R. (Eds.), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp National Research Council (NRC) Nutrient Requirements of Fish and Shrimp. The National Academies Press, Washington DC. Oliva-Teles, A., Nutrition and health of aquaculture fish. J. Fish Dis. 35, Pohlenz, C., Gatlin III, D.M., Interrelationships between fish nutrition and health. Aquaculture 431, Poston, H.A., 1990a. Performance of rainbow trout fed supplemental soybean lecithin and choline. Prog. Fish Cult. 52, Poston, H.A., 1990b. Effect of body size on growth, survival and chemical composition of Atlantic salmon fed soy lecithin and choline. Prog. Fish Cult. 52, Poston, H.A., Response of Atlantic salmon fry to feed-grade lecithin and choline. Prog. Fish Cult. 53, Sandnes, K., Torrissen, O., Waagbø, R., The minimum dietary requirement of vitamin C in Atlantic salmon Salmo salar fry using Ca ascorbate-2-monophosphate as dietary source. Fish Physiol. Biochem. 10, Taggart, J.B., Bron, J.E., Martin, S.A.M., Seear, P.J., Hoyheim, B., Talbot, R., Villeneuve, L., Sweeney, G.E., Houlihan, D.F., Secombes, C.J., Tocher, D.R., Teale, A.J., A description of the origins, design and performance of the TRAITS/SGP Atlantic salmon (Salmo salar L.) cdna microarray. J. Fish Biol. 72,

117 Thodesen, J., Storebakken, T., Digestibility of diets with precooked rye or wheat by Atlantic salmon, Salmo salar L. Aquacult. Nutr. 4, Tocher, D.R., Glycerophospholipid metabolism, in: Hochachka, P.W., Mommsen, T.P. (Eds.), Biochemistry and Molecular Biology of Fishes, Vol. 4 Metabolic and Adaptational Biochemistry. Elsevier Press, Amsterdam, pp Tocher, D.R., Metabolism and functions of lipids and fatty acids in teleost fish. Rev. Fisheries Sci. 11, Tocher, D.R., Fatty acid requirements in ontogeny of marine and freshwater fish. Aquacult. Res. 41, Tocher, D.R., Glencross, B.D., Lipids and fatty acids, in: Lee, C.-S., Lim, C., Webster, C., Gatlin III, D.M. (Eds.), Dietary Nutrients, Additives, and Fish Health. Wiley-Blackwell, New Jersey, in press. Tocher, D.R., Bendiksen, E.Å., Campbell, P.J., Bell, J.G., The role of phospholipids in nutrition and metabolism of teleost fish. Aquaculture 280, Torrissen, O.J., Pigmentation of salmonids-effect of carotenoids in eggs and start feeding diet on survival and growth rate. Aquaculture 43, Torrissen, O.J., Christiansen, R., Requirements for carotenoids in fish diets. J. Appl. Ichthyol. 11, Verlhac-Trichet, V., Nutrition and immunity: an update. Aquacult. Res. 41,

118 Tables and Figures - Annex A TABLE A1. Nutrient Requirements of salmonids on a dry matter basis a,b Atlantic salmon Rainbow trout Pacific salmon Digestible energy (kcal/kg diet) c 4,400 4,200 4,200 Digestible protein (%) c Nutrient Requirements c Amino acids (%) Arginine Histidine Isoleucine Leucine Lysine Methionine Methionine + cystine Phenylalanine Phenylalanine + tyrosine Threonine Tryptophan Valine Fatty acids (%) 18:3n n-3 LC-PUFA d n-6 PUFA NT Cholesterol NT NT NT Phospholipids NT ( ) e NT ( ) e NT Macrominerals (%) Calcium NR NR NR Chlorine NT NT NT Magnesium NT Phosphorous Potassium NT NT 0.80 Sodium NR NR NT Microminerals (mg/kg) Copper 5 3 NT Iodine R Iron NT NT Manganese NT Selenium NT 0.15 R Zinc NT 116

119 Fat-soluble vitamins f A (mg/kg) NT 0.75 R D ( g/kg) NT 40 NR E (mg/kg) K (mg/kg) < 10 R R Water-soluble vitamins (mg/kg) Biotin NT Choline g NT Folacin NT 1 2 Myoinsitol g NT Niacin NT Pantothenic acid NT Riboflavin NT 4 7 Thiamin NT 1 10 Vitamin B Vitamin B 12 NT R 0.02 Vitamin C h NT a Requirements determined with highly purified ingredients in which the nutrients are highly digestible and therefore values represent near 100 % bioavailability. b R, required in diet but quantity not determined; NR, not required under practical conditions; NT, not tested. c Typical energy and digestible crude protein concentrations (digestible N x 6.25) in commercial diets. d EPA (20:5n-3) and/or DHA (22:6n-3) e Values in parentheses represent requirements reported for fry/early juvenile stages f Conversion factors for fat-soluble virtamins are: 10,000 IU = 3,000 g vitamin A (retinol), 1 IU = g vitamin D (cholecalciferol). g Diet without phospholipid h As L-ascorbyl-2-monophosphate or L-ascorbly-2-polyphosphate Taken with permission from NRC (2011). 117

120 Fig. A1. The pathways of long-chain PUFA biosynthesis from α-linolenic (18:3n-3) and linoleic (18:2n-6) acids in Atlantic salmon. Horizontal arrows represent desaturations catalysed by Δ6 and Δ5 fatty acyl desaturases and vertical down arrows represent elongations catalysed by Elovl5, Elovl2 and Elovl4 fatty acid elongases. β-ox, limited peroxisomal β-oxidation. 118

121 Annex B. Human Nutrition Fish and seafood are generally universally recognised as important components of a healthy diet as they supply not only high quality, easily digested protein, but also essential micronutrients including the minerals selenium and iodine and vitamins, particularly fat-soluble vitamins such as cholecalciferol (vitamin D 3 ) (De Roo et al., 2012; Tacon and Metian, 2013). However, although all these nutrients are beneficial effects in the human diet, the nutrients most associated with the beneficial effects of eating fish are the n-3 (or omega-3 ) LC-PUFA, EPA and DHA (Lands, 2014). These key conditionally-essential fatty acids have been the most studied and the physiological, biochemical and molecular mechanisms underpinning their critical roles in human health are increasingly being elucidated and understood. The so-called oily fish, such as salmon, are arguably the most effective foods in delivering physiologically effective doses of n-3 LC-PUFA to human consumers (Tur et al., 2012). Our understanding of the beneficial effects of dietary EPA and DHA on human health have been largely based on two main lines of evidence, epidemiological studies and randomised controlled (intervention) trials, although laboratory studies investigating biochemical and molecular mechanisms have also provided mechanistic support to these in vivo approaches (Gil et al., 2012). Based on all the evidence, many recommendations for EPA and DHA intake for humans have been produced by a large number of global and national health agencies and associations, and government bodies. The recommendations of over 50 organisations were recently compiled by the Global Organisation for EPA and DHA Omega 3s (GOED, 2014). B.1. Cardiac and cardiovascular disease (CVD) Epidemiological studies, that are normally large and long, looking at n-3 LC-PUFA intake (fish/fish oil) and/or status (blood/tissue fatty acid compositions) almost uniformly show a protective effect, decreasing the risk of developing CVD (Delgado-Lista et al., 2012). When all confounders are controlled the data are robust albeit they show only an association and not cause and effect. However, randomised controlled trials, normally designed for patients with some disease already (i.e. in at-risk patients), have also been generally positive and show that patients benefit from dietary n-3 LC-PUFA therapy (Calder and Yaqoob, 2012; Delgado-Lista et al., 2012). Until recently, meta-analyses were also largely positive (e.g. Casula et al., 2013; see Calder, 2014) although there have been some recent studies questioning the efficacy of n-3 PUFA supplements (Rizos et al., 2012; Chowdhury et al., 2014). The biochemical mechanisms underlying the beneficial effects of n- 3 LC-PUFA are primarily based on their lowering of known risk factors for CVD (blood cholesterol and TAG, hypertension) and effects on platelet function (Uauy and Valenzuela, 2000). Molecular mechanisms are less well understood but include effects on hepatic lipid metabolism such as decreasing TAG synthesis, that involve, among other mechanisms, the important role of PUFA in transcriptional regulation of genes of lipid metabolism (Deckelbaum et al., 2006; Georgiadi and Kersten, 2012). Based on their effects on CVD, many health agencies worldwide recommend up to 500 mg/d of EPA and DHA for reducing CVD risk or 1 g/d for secondary prevention in existing CVD patients, with a dietary strategy for achieving 500 mg/d being to consume two fish meals per week with at least one of oily fish (ISSFAL, 2004; Gebauer et al., 2006; Aranceta and Perez-Rodrigo, 2012; EFSA, 2012). 119

122 B.2. Inflammatory disease The most robust evidence for a beneficial effect of dietary n-3 LC-PUFA in inflammatory disease has been obtained in rheumatoid arthritis (Miles and Calder, 2012). The doses required to gain benefit are much higher than for CVD, at around 3 g/day and so dietary n-3 LC-PUFA are still not part of classical treatment plans. There is also increasing evidence for beneficial effects of dietary n-3 LC-PUFA in Inflammatory Bowel Disease (IBD) such as Crohn s disease and ulcerative colitis (Cabré et al., 2012). Although there are studies showing some beneficial effects of n-3 LC-PUFA in various other inflammatory diseases there are too few studies to be conclusive. In contrast, there is a very large body of research into the mechanisms of action of n-3 LC-PUFA on inflammatory pathways (Rangel-Huerta et al., 2012). Until recently, the major mechanism was thought to be the anti-ara effect of n-3 LC-PUFA. In brief, inflammatory responses are largely driven by eicosanoids produced from the major n-6 LC-PUFA, ARA, but other LC-PUFA, particularly EPA, compete with ARA at level of the cyclooxygenase (COX) and lipoxygenase enzymes responsible for the production of eicosanoids, both reducing the production of pro-inflammatory ARA derivatives and producing less inflammatory EPA-derived mediators (Calder, 2007). However, there is considerable research showing a wide range of mechanisms potentially underpinning the effects of n-3 LC-PUFA on inflammatory responses and immune pathways (Chapkin et al., 2009; Calder, 2013). In particular in recent years, it has been established that the n-3 LC-PUFA, EPA and DHA, are also precursors of other COX-2 derived non-classical eicosanoid derivatives called resolvins, maresins and protectins, which are specialised pro-resolving mediators (SPM) that bring about resolution of the inflammatory response and a return to homeostasis (Serhan and Petasis, 2011; Weylandt et al., 2012). In addition, as ligands of various transcription factors, n-3 LC-PUFA and their derivatives can also have affects on expression of genes in inflammatory and immune pathways (Chapkin et al., 2009; Schmuth et al., 2014). Therefore, n-3 LC-PUFA have a combination of effects acting to reduce the respiratory burst and increase resolution (Calder, 2013). While ARA- EPA competition and the effects of n-3 LC-PUFA on eicosanoid production (Montero and Izquierdo, 2011; Martinez-Rubio et al., 2013) and inflammatory and immune gene expression have also been studied in fish (Montero and Izquierdo, 2011; Martinez-Rubio et al., 2012, 2014), the role of EPA and, especially, DHA in the production of SPM has not been studied in fish. B.3. Neural development There is good robust evidence that decreased DHA status can lead to cognitive and visual impairment and that DHA supplements have positive beneficial outcomes in pre-term infants (Carlson et al., 1993). In term infants, DHA supplementation appears less effective suggesting that it is rectifying low DHA status that is effective rather than increasing normal DHA status (Campoy et al., 2012). There have also been several reports of potential beneficial effects of dietary DHA supplementation in a number of psychological/behavioural/psychiatric disorders including attention deficit hyperactivity disorder (ADHD) and depression and, although there are some reports of benefits, there are insufficient studies and data to draw firm conclusions (Ortega et al., 2012). However, it is becoming generally recognised that n-3 LC-PUFA are potential key nutrients 120

123 to prevent various pathological conditions associated with the normal aging process (Ubeda et al., 2012). This has prompted research into the effects of n-3 LC-PUFA on dementia, including Alzheimer s disease (AD) and other age-related cognitive impairments (Dangour et al., 2012). In general, there is a lack of agreement in findings from intervention studies to support a benefit of n-3 LC-PUFA on cognitive function but this may also reflect intrinsic limitations in the design of published studies (Dangour et al., 2012). However, DHA supplementation trials in patients with some pre-diagnosed cognitive impairment indicated that this appeared to slow AD progression (Quinn et al., 2010). In long-term animal AD model trials, n-3 LC-PUFA improved cognitive function and diminished the amount of neuronal loss (Hooijmans et al., 2012). As chronic inflammation is observed in AD, the mechanism of n-3 LC-PUFA is postulated to be due to down-regulation of inflammation and promotion of resolution of the inflammatory response (Hjorth et al., 2013). B.4. Cancer Epidemiological studies have indicated that, in general, consumption of oily fish or taking n-3 LC- PUFA supplements may have a protective effect (i.e. decreasing risk) in colo-rectal, breast and prostate cancers (Gerber, 2012). Further evidence is difficult to obtain as randomised, controlled trials similar to those carried out in CVD are not possible in cancer, but there are studies indicating some beneficial effects of n-3 LC-PUFA supplementation in chemotherapy (Bougnoux et al., 2009). In small lung carcinoma, n-3 LC-PUFA appear to sensitise cancerous cells to chemotherapeutants thus increasing efficiency and perhaps decreasing side effects, enabling the patient to undergo more cycles (Murphy et al., 2011a). In addition, n-3 LC-PUFA may protect against the muscle mass loss likely due to hyper-metabolism promoting proteolysis that is often associated with cancer (Murphy et al., 2011b). The mechanisms underlying the latter two beneficial effects are not understood, but are possibly based on the effects of n-3 LC-PUFA on inflammatory processes. B.5. Negative reports Two recent studies have associated n-3 LC-PUFA status with prostate cancer (Brasky et al., 2011, 2013). The interpretation of these data has been criticised on a number of grounds including the fact that n-3 LC-PUFA intake was not monitored in these studies and, as n-3 LC-PUFA supplementation is not uncommon in cancer patients (see above), the reason for the observed association between n-3 LC-PUFA and prostate cancer cannot be established. However, these negative reports have impacted on sales of FO-derived n-3 LC-PUFA products for direct human consumption and halted the ever-increasing trend over the last few years that peaked at 25 % of total FO in , but that slipped back to 22 % in 2012 (IFFO, 2013). More generally, negative effects of dietary n-3 LC-PUFA have been attributed with the pro-oxidant effect of LC-PUFA (Garrido et al., 1989; Tsuduki et al., 2011). This can occur through the use of dietary sources that already contain oxidised lipids if not prepared/stored properly with appropriate anti-oxidant protection, but also through pro-oxidant effects in vivo (Albert et al., 2013; Garcia-Hernandez et al., 2013). Increased lipid oxidation products and anti-oxidant defence mechanisms (enzymes etc.) 121

124 are both generally observed in n-3 LC-PUFA supplementation trials, and the need to also increase anti-oxidant intake when on supplements is well known. However, a certain level of peroxide tone is required for eicosanoid biosynthesis (Smith, 2005), and there is also evidence that n-3 LC- PUFA can have anti-oxidant actions (Giordano and Visoli, 2014). B.6 Dietary recommendations for fish and n-3 LC-PUFA It is evident that many health agencies worldwide recognise the importance of increasing dietary intake of EPA and DHA from fish or other sources to decrease the risk of CVD. Quantitative recommendations on dietary intake of EPA and DHA were first proposed around a quarter a century ago in a NATO report from 1989 (Simopoulos, 1989). Based on the accumulating evidence many national, international, and professional organisations have since made specific recommendations for the consumption of fish and/or EPA and DHA (Table B1). For example, in 2005, the Dietary Guidelines for Americans report stated that evidence suggests consuming approximately two servings of fish per week (approximately 227 g) may reduce the risk of mortality from CHD and that consuming EPA and DHA may reduce the risk of mortality from CVD in people who have already experienced a cardiac event (USDA, US Department of Health and Human Services. 2005). The current advice and guidelines issued by Governments and international health organisations was comprehensively reviewed recently (Aranceta and Perez- Rodrigo, 2012). In July 2014, the UK National Institute for Health and Care Excellence (NICE) revised their recommendations for treatment for the secondary prevention of cardiovascular disease [NICE Clinical guideline 181 (2014)]. In this revision, NICE changed their guidelines for Lipid Modification therapy by removing 1 g omega-3 fatty acids (along with some other treatments) from the advice for primary and secondary prevention of cardiovascular disease, essentially leaving only statins for the majority of patients. This decision was based on the recent publication of meta-analyses studies that questioned the efficacy of omega-3 supplements for the secondary prevention of cardiovascular disease (Rizos et al., 2012; Chowdhury et al., 2014; Wen et al., 2014). Indeed, NICE now advise clinicians to tell people that there is no evidence that omega-3 fatty acid compounds help to prevent cardiovascular disease. However, these meta-analyses studies have been hotly debated and disputed particularly over the influence of one or two null studies that can skew the data (GOED, 2012). In addition, recent work has suggested that, with concentrated omega-3 ethyl ester supplements such as Omacor (Abbott Laboratories) or Lovaza (GlaxoSmithKline), the method of delivery has an impact and that these supplements should be taken with a fatty meal in order to stimulate a full digestive response and ensure proper digestion and maximum bioavailability of the EPA and DHA (Davidson et al., 2012). It should be emphasised that the NICE decision pertains to the prescription of a pharmacological drug and not, apparently, omega-3 itself as NICE still contain the recommendation to eat at least two portions of fish per week, including a portion of oily fish in their Lifestyle Changes for persons at risk of cardiovascular disease. Furthermore, it should be emphasised that no other major health advice providers such as the American Heart Association etc. have changed their recommendations. Clearly, the NICE 2014 revision is not a welcome development however, the continued recommendation for oily 122

125 fish emphasises the value of fish itself, especially species such as salmon, rather than dietary supplements, as a healthy component of the diet and the ideal source of omega-3 fatty acids, EPA and DHA. References Albert, B.B., Cameron-Smith, D., Hofman, P.L., Cutfield, W.S., Oxidation of marine omega-3 supplements and human health. Biomed. Res. Int. 2013, Aranceta, J., Pérez-Rodrigo, C., Recommended dietary reference intakes, nutritional goals and dietary guidelines for fat and fatty acids: a systematic review. Br. J. Nutr. 107, S8-22. Bougnoux, P., Hajjaji, N., Ferrasson, M.N., Giraudeau, B., Couet, C., Le Floch, O., Improving outcome of chemotherapy of metastatic breast cancer by docosahexaenoic acid: a phase II trial. Br. J. Cancer 101, Brasky, T.M., Darke, A.K., Song, X., Tangen, C.M., Goodman, P.J., Thompson, I.M., Meyskens Jr, F.L., Goodman, G.E., Minasian, L.M., Parnes, H.L., Klein, E.A., Kristal, A.R., Plasma phospholipid fatty acids and prostate cancer risk in the SELECT trial. J. Natl. Cancer Inst. 105, Brasky, T.M., Till, C., White, E., Neuhouser, M.L., Song, X., Goodman, P., Thompson, I.M., King, I.B., Albanes, D., Kristal, A.R., Serum phospholipid fatty acids and prostate cancer risk: results from the prostate cancer prevention trial. Am. J. Epidemiol. 173, Cabré, E., Mañosa, M., Gassulla, M.A., Omega-3 fatty acids and inflammatory bowel diseases a systematic review. Br. J. Nutr. 107, S240-S252. Calder, P.C., Immunomodulation by omega-3 fatty acids. Prostaglandins Leukotrienes Essent. Fatty Acids 77, Calder, P.C., n-3 Fatty acids, inflammation and immunity: new mechanisms to explain old actions. Proc. Nutr. Soc. 72, Calder, P.C., Very long chain omega 3 (n 3) fatty acids and human health. Eur. J. Lipid Sci. Technol. 116, Calder, P.C., Yaqoob, P., Marine omega-3 fatty acids and coronary heart disease. Curr. Opin. Cardiol. 27, Campoy, C., Escolano-Margarit, V., Anjos, T., Szajewska, H., Uauy, R., Omega 3 fatty acids on child growth, visual acuity and neurodevelopment. Br. J. Nutr. 107, S85-S106. Carlson, S.E., Werkman, S.H., Rhodes, P.G., Tolley, E.A., Visual- acuity development in healthy preterm infants: effect of marine-oil supplementation. Am. J. Clin. Nutr. 58, Casula, M., Soranna, D., Catapano, A.L., Corrao G., Long-term effect of high dose omega-3 fatty acid supplementation for secondary prevention of cardiovascular outcomes: A metaanalysis of randomized, double blind, placebo controlled trials. Atherosclerosis Suppl. 14, Chapkin R.S., Kim, W., Lupton, J.R., McMurray, D.N., Dietary docosahexaenoic and eicosapentaenoic acid: Emerging mediators of inflammation. Prostaglandins Leukotrienes Essent. Fatty Acids 81,

126 Chowdhury, R., Warnakula, S., Kunutsor, S., Crowe, F., Ward, H.A., Johnson, L., Franco, O.H., Butterworth, A.S., Forouhi, N.G., Thompson, S.G., Khaw, K.-T., Mozaffarian, D., Danesh, J., Di Angelantonio, E., Association of dietary, circulating, and supplement fatty acids with coronary risk: A systematic review and meta-analysis. Ann. Intern. Med. 160, Dangour, A.D., Andreeva, V.A., Sydenham, E., Uauy, R., Omega 3 fatty acids and cognitive health in older people. Br. J. Nutr. 107, S152-S158. Davidson, M.H., Johnson, J., Rooney, M.W., Kyle, M.L., Kling, D.F., A novel omega-3 free fatty acid formulation has dramatically improved bioavailability during a low-fat diet compared with omega-3-acid ethyl esters: The ECLIPSE (Epanova compared to Lovaza in a pharmacokinetic single-dose evaluation) study. J. Clin. Lipidology 6, Deckelbaum, R.J., Worgall, T.S., Seo, T., n-3 Fatty acids and gene expression. Am. J. Clin. Nutr. 83(suppl), 1520S 1525S. Delgado-Lista, J., Perez-Martinez, P., Lopez-Miranda, J., Perez-Jimenez, F., Long chain omega-3 fatty acids and cardiovascular disease: a systematic review. Br. J. Nutr. 107, S201- S213. De Roo, B., Sneddon, A., Macdonald, H., Fish as adietary source of ghealthy liong chain n-3 polyunsaturated fatty acids (LC n-3 PUFA and vitamin D: A review of current literature June (Accessed November 2014). European Food Safety Authority (EFSA) Panel on Dietetic Products, Nutrition and Allergies (NDA), Scientific opinion related to the tolerable upper intake level of eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA) and docosapentaenoic acid (DPA). EFSA J. 10, Garcia-Hernandez, V.M., Gallar, M., Sanchez-Soriano, J., Micol, V., Roche, E., Garcia-Garcia, E., Effect of omega-3 dietary supplements with different oxidation levels in the lipidic profile of women: a randomized controlled trial. Int. J. Food Sci. Nutr. 64, Garrido, A., Garrido, F., Guerra, R., Valenzuela, A., Ingestion of high doses of fish oil increases the susceptibility of cellular membranes to the induction of oxidative stress, Lipids 24, Gebauer, S.K., Psota, T.L., Harris, W.S., Kris-Etherton, P.M., n-3 fatty acid dietary recommendations and food sources to achieve essentiality and cardiovascular benefits. Am. J. Clin. Nutr. 83, 1526S-1535S. Georgiadi, A., Kersten, S., Mechanisms of gene regulation by fatty acids. Adv. Nutr. 3, Gerber, M., Omega-3 fatty acids and cancers: a systematic update review of epidemiological studies. Br. J. Nutr. 107, S Gil, A., Serra-Majem, L., Calder, P. C., Uauy, R., Systematic reviews of the role of omega-3 fatty acids in the prevention and treatment of disease. Br. J. Nutr. 107, S1-S2. Giordano, E., Visioli, F., Long-chain omega 3 fatty acids: Molecular bases of potential antioxidant actions. Prostaglandins, Leukotrienes Essent. Fatty Acids 90, 1-4. GOED, GOED (Global Organization for EPA and DHA) Disputes JAMA Omega-3 Meta- Analysis. (posted 12 September 2012; accessed 27 October 2014) 124

127 GOED, Global recommendations for EPA and DHA intake (Rev 16 April 2014). Accessed 27 October Hjorth, E., Zhu, M., Cortés Toro, V., Vedin, I., Palmblad, J., Cederholm, T., Freund-Levi, Y., Faxen- Irving, G., Wahlund, L.-O., Basun, H., Eriksdotter, M., Schultzberg, M., Omega-3 fatty acids enhance phagocytosis of Alzheimer's disease-related amyloid-β42 by human microglia and decrease Inflammatory markers. J. Alzheimer s Dis. 35, Hooijmans, C.R., Pasker-de Jong, P.C., de Vries, R.B., Ritskes-Hoitinga, M., The effects of long-term omega-3 fatty acid supplementation on cognition and Alzheimer's pathology in animal models of Alzheimer's disease: a systematic review and meta-analysis. J. Alzheimers Dis. 28, IFFO, The Marine Ingredients Organisation: Fishmeal and Fish Oil Statistical Yearbook In IFFO [online]. ISSFAL, International Society for the Study of Fatty Acids and Lipids, Report of the Sub- Committee on Recommendations for intake of polyunsaturated fatty acids in healthy adults. ISSFAL, Brighton. Lands, B., Historical perspectives on the impact of n-3 and n-6 nutrients on health. Prog. Lipid Res. 55, Martinez-Rubio, L., Morais, S., Evensen, Ø., Wadsworth, S., Vecino, J.L.G, Ruohonen, K., Bell, J.G. Tocher, D. R., Functional feeds reduce heart inflammation and pathology in Atlantic salmon following experimental challenge with Atlantic salmon reovirus (ASRV). PLoS ONE 7, e Martinez-Rubio, L., Morais, S., Evensen, Ø., Wadsworth, S., Vecino, J.L.G., Ruohonen, K., Bell, J.G., Tocher, D.R., Effect of functional feeds on fatty acid and eicosanoid metabolism in liver and head kidney of Atlantic salmon (Salmo salar L.) with experimentally induced Heart and Skeletal Muscle Inflammatory disease. Fish Shellfish Immunol. 34, Martinez-Rubio, L., Evensen, Ø., Krasnov, A., Wadsworth, S., Ruohonen, K., Vecino, J.L.G., Bell, J.G., Tocher, D.R., Functional feeds reduce heart inflammation and pathology in Atlantic salmon (Salmo salar) following experimental challenge with Piscine Myocarditis Virus (PMCV). BMC Genomics 15, 462. Miles, E.A., Calder, P.C., Influence of marine n-3 polyunsaturated fatty acids on immune function and a systematic review of their effects on clinical outcomes in rheumatoid arthritis. Br. J. Nutr. 107, S171-S184. Montero, D., Izquierdo, M., Welfare and health of fish fed vegetable oils as alternative lipid sources to fish oil, in: Turchini, G.M., Ng, W.-K., Tocher, D.R. (Eds.), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp Murphy, R.A., Mourtzakis, M., Chu, Q.S.C., Baracos, V.E., Reiman, T., Mazurak, V.C., 2011a. Supplementation with fish oil increases first-line chemotherapy efficacy in patients with advanced nonsmall cell lung cancer. Cancer 117, Murphy, R.A., Mourtzakis, M., Chu, Q.S.C., Baracos, V.E., Reiman, T., Mazurak, V.C., 2011b. Nutritional intervention with fish oil provides a benefit over standard of care for weight and skeletal muscle mass in patients with nonsmall cell lung cancer receiving chemotherapy. Cancer 117,

128 Ortega, R.M., Rodriguez-Rodriguez, E., Lopez-Sobaler, A.M., Effects of omega 3 fatty acids supplementation in behavior and non-neurodegenerative neuropsychiatric disorders. Br. J. Nutr. 107, S261-S270. Quinn, J.F., Raman, R., Thomas, R.G., Yurko-Mauro, K., Nelson, E.B., Van Dyck, C., Galvin, J.E., Emond, J., Jack, C.R., Weiner, M., Shinto, L., Aisen, P.S., Docosahexaenoic acid supplementation and cognitive decline in Alzheimer disease: a randomized trial. JAMA 304, Rangel-Huerta, O.D., Aguilera, C.M., Mesa, M.D., Angel Gil, A., Omega-3 long-chain polyunsaturated fatty acids supplementation on inflammatory biomarkers: a systematic review of randomised clinical trials. Br. J. Nutr. 107, S159-S170. Rizos, E.C., Ntzani, E.E., Bika, E., Kostapanos, M.S., Elisaf, M.S., Association Between omega- 3 fatty acid supplementation and risk of major cardiovascular disease events: A systematic review and meta-analysis. J. Am. Med. Assoc. 308, Schmuth, M., Moosbrugger-Martinz, V., Blunder, S., Dubrac, S., Role of PPAR, LXR, and PXR in epidermal homeostasis and inflammation. Biochim. Biophys. Acta 1841, Serhan, C.N., Petasis, N.A., Resolvins and protectins in inflammation resolution. Chem. Rev. 111, Simopoulos, A.P., Summary of the NATO advanced research workshop on dietary omega 3 and omega 6 fatty acids: biological effects and nutritional essentiality. J. Nutr. 119, Smith, W.L., Cyclooxygenases, peroxide tone and the allure of fish oil. Curr. Opin. Cell Biol. 17, Tacon, A.G.J., Metian, M., Fish matters: Importance of aquatic foods in human nutrition and global food supply. Rev. Fisheries Sci. 21, Tsuduki, T., Honma, T., Nakagawa, K., Ikeda, I., Miyazawa, T., Long-term intake of fish oil increases oxidative stress and decreases lifespan in senescence-accelerated mice. Nutrition 27, Tur, J.A., Bibiloni, M.M., Sureda, A., Pons, A., Dietary sources of omega 3 fatty acids: public health risks and benefits. Br. J. Nutr. 107, S23-S52. Uauy, R., Valenzuela, A., Marine oils: The health benefits of n-3 fatty acids. Nutrition 16, Úbeda, N., Achóna, M., Varela-Moreirasa, G., Omega 3 fatty acids in the elderly. Br. J. Nutr. 107, S137-S151. US Department of Agriculture, US Department of Health and Human Services Dietary Guidelines Advisory Committee report Internet: Accessed 20 November Wen, Y.T., Dai, J.H., Gao, Q Effects of Omega-3 fatty acid on major cardiovascular events and mortality in patients with coronary heart disease: A meta-analysis of randomized controlled trials. Nutr. Metab. Cardiovasc. Dis. 24, Weylandt, K.H., Chiua, C.-Y., Gomolkaa, B.,Waechtera, S.F., Wiedenmann, B., Omega-3 fatty acids and their lipid mediators: Towards an understanding of resolvin and protectin formation. Omega-3 fatty acids and their resolvin/protectin mediators. Prostaglandins Other Lipid Mediat. 97,

129 TABLE B1. International recommendations for long-chain n-3 LC-PUFA Recommendation Source and reference Date Total n-3 ALA EPA+DHA n-3:n-6 % of energy g mg NATO workshop :1 UK Committee on Medical Aspects of Food Policy ISSFAL workshop ANC (France) (DHA, ) 5:1 Eurodiet Health Council of the Netherlands :1 American Heart Association 2002 ~1 g/d (2ry prev CHD) US National Academics of Science, Institute of Medicine ~140 European Society of Cardiology 2003 ~1 g/d (2ry prev CHD) WHO FA (1-2 fish meals/wk) ISSFAL UK Scientific Advisory Committee on Nutrition 2004 ~450 (Min 2 portions fish/wk ; 1 oily) UK National Institute for Health and Care Excellence 200x ~1 g/d (2ry prev CHD) UK National Institute for Health and Care Excellence portions fish/wk ; 1 oily) ALA, -linolenic acid; ANC, Apports Nutritionnels Conseilles; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; FAO, Food and Agriculture Organization; ISSFAL, International Society for the Study of Fatty Acids and Lipids; NATO, North Atlantic Treaty Organization; WHO, World Health Organization. 127

130 Annex C. Alternative feed ingredients Feeds for carnivorous fish including salmonids have to be based largely on protein and fat/oil, as carbohydrate has no major role as a supplier of energy in carnivorous fish. This is most likely simply an evolutionary adaptation to a natural diet that contains very little carbohydrate. The consequence of this is that efficient feeds for Atlantic salmon must have high protein contents and this limits the usefulness of many traditional fish feed ingredients. Current formulations of salmon feeds in Norway contain around % protein and % oil although absolute levels can vary depending upon raw ingredient costs with feeds formulated to a specific FCR that can be equally achieved with different protein/oil combinations (Fig. C1). The oil content of salmon feeds has increased over the years with the development of high-energy feeds that exploit the fact that twice as much energy per unit mass is provided by lipid compared to protein (or carbohydrate) and this enables DE:DP ratio and growth potential to be maximised through the concept of protein sparing (Sargent et al., 2002). The lipid level has now stabilised with maximum levels of around 35 %, which has been shown to provide greatest growth while minimising excess lipid deposition in the flesh that can have adverse effects during down-stream processing (Einen and Roem, 1997; Sargent and Tacon, 1999). Modern salmon feeds in Norway and Scotland have sustainability built in to their formulations and therefore they are all based on a blend of terrestrial plant meals and vegetable oils with marine ingredients, essentially fishmeal (FM) and fish oil (FO), although krill products are also used in small amounts in some feeds. The levels of FM and FO have progressively decreased over the last decade and this trend continues (Fig. C1). The main alternative protein source is soy protein concentrate along with smaller amounts of other plant protein concentrates such as wheat or corn/maize gluten and pea protein concentrate. Other plant meals such as sunflower expeller, wheat and beans are included as ingredients that supply a mix of protein and starch. Starch is included in salmon feeds primarily as a nutritional binder and for its physical properties in extruded pellets, enabling pellets to expand via starch gelatinisation during extrusion, rather than as a energy nutrient. Figure C1 shows the levels of the major plant protein ingredients used in salmon feeds in Norway in recent years and the trends in the proportions. Rapeseed oil is the vegetable oil (VO) used almost exclusively in combination with FO both in Norway and Scotland, although small amounts of palm oil and, perhaps, linseed oil, may be included. The situation with salmon feeds in Scotland is generally similar to Norway although slightly more complicated. The Scottish salmon market is highly consumer driven via the major retailers and each retailer also has its particular niche market and, within a retailer, niche products. Therefore, feeds for Scottish salmon consist of a rather diverse range of bespoke products tailored to meet individual customer/niche markets. The differences in formulation between the feeds are based on the particular specifications dictated by salmon producers that, in return are responding to retailer specifications. Commonly, although not exclusively, the specifications include the amount of marine ingredients (FM and FO) and may or may not include reference to specific (minimum) levels of n-3 LC-PUFA (EPA and DHA) in feed or flesh. Pigment source (e.g. the red carotenoid-rich bacterium Paracoccus carotinifaciens or the natural extract of the red yeast, Phaffia rhodozyma, versus synthetic astaxanthin) is another possible differentiator. The range and relative proportions 128

131 of alternative ingredients to FM and FO that are then used in each formulation will be dictated by these specifications within the overall nutritional requirements of salmon. For producers of Scottish salmon not targeting specific niche markets but rather the mainstream market, the nutritional profile of the feed and the ingredients used are determined by least cost formulation more similar to Norwegian salmon feeds. However, the levels of FM and FO in salmon feeds in Scotland are generally higher than they are in Norway although, like Norway, they are continuing to decrease. Although an over-simplification, it is not unreasonable to suggest that Scottish producers take a cautious approach and, as a result, the levels in Scotland lag those in Norway by approximately a year. In this way, Scottish salmon maintains its position as a premium product while continuing to embrace increasing replacement and enhanced sustainability. The range of alternatives to FM are generally similar in Scotland and Norway although some ingredients can vary such as faba beans, which tend to be used to a greater degree in Scotland simply based on the greater match between availability and supply for this ingredient in Scotland compared to Norway. Finally, although modern, sustainable feeds for salmon in Scotland contain far lower levels of both FM and FO, a recent study showed that farmed salmon remain an excellent source of EPA and DHA for human consumers (Henriques at al., 2014). Although the relative proportions of EPA and DHA were lower in farmed salmon products than in wild salmon products (Fig. C2A, C), in absolute terms the farmed salmon delivered, on average, double the dose of EPA and DHA than the wild salmon (Fig. C2B, C). Therefore, the recommended dose (500 mg/d or 3.5 g/week) of EPA and DHA for lowering risk of CVD could be achieved by eating two 150 g portions of farmed salmon, whereas 4-5 portions of wild salmon would be required (Henriques et al., 2014). C.1. Plant protein ingredients In terms of supply, cost and sustainability, the most favourable alternatives to FM as protein ingredients for salmon feeds are products of terrestrial agriculture, namely, plant-based meals (Hardy, 2010). The plant protein products currently being used in salmon feeds can be placed into three broad categories including the oilseed meals (soybean, rapeseed/canola, sunflower expeller, cottonseed etc.), pulses/legumes (pea, bean, lupin etc.), and grains (wheat and corn/maize etc.). However, compared to FM, plant protein ingredients have essentially similar differences in terms of nutrient contents and composition that affect and potentially limit the range and levels of plant proteins in salmon feeds. The main issues that must be taken into consideration when formulating feeds are related to energy density/protein content, amino acid composition, anti-nutritional factors and palatability. In addition, much of the phosphorus in plant proteins is present as phytic acid, which has low digestibility in most fish species and, consequently, there has been much research into the use of phytase to improve phosphorus utilisation from plant feedstuffs (Kumar et al., 2012). C.1.1. Protein content. The first major issue with most plant meals is energy density as some meals have high carbohydrate content (including non-starch polysaccharides, NSP) that is of 129

132 little nutritional value for piscivorous/carnivorous species such as Atlantic salmon. This translates into protein contents in most plant meals being too low to be effective replacements for FM. The protein contents of the most commonly used FM are %, whereas plant meals are commonly in the range of % (Tables C1 and C2). Some solvent-extracted oilseed meals (soy, sunflower) can have higher protein contents of % but, to be considered as primary replacers for FM, plant products must have protein contents of > 55 % and ideally > 60 %. Therefore, each of the main groups of plant protein ingredients listed above also have associated protein concentrates including soy protein concentrate (SPC, ~ 64 % protein), pea protein concentrate (~ 55 %), wheat (~ 80 %) and corn (~ 64 %) glutens (Tables C1 and C2). These plant protein concentrates are currently the primary protein replacers for FM in salmon feeds. C.1.2. Amino acid composition. The ideal protein has been defined as the amino acid profile that meets exactly the requirement of the animal with no excess or deficit and with all essential amino acids (EAA) balanced to be equally limiting (Wang and Fuller, 1989; Emmert and Baker, 1997). The ideal protein concept is not perfect as the need to balance amino acid levels in diets and the effects of amino acid balance/imbalance, whatever that means exactly, are not very well understood in any fish species. However, the ideal protein concept has become of more practical importance with the increasing substitution of FM with other proteins and may provide a reasonable starting point for new diet formulation. Thus, some plant proteins have generally sub-optimal amino acid compositions for animals with specific EAA such as methionine, lysine and, perhaps, arginine and threonine being found in lower levels than ideal for many fish species, including salmon (Gatlin III et al., 2007). Specifically, corn and wheat glutens (and distillers dried grains) are low in lysine, and arginine is also low in wheat with soybean meal being low in methionine. The amino acid composition can sometimes be improved by processing with EAA levels being increased in SPC compared to soybean meal (Gatlin III et al., 2007). Imbalances in EAA can also be managed by a combination of blending different protein ingredients and further balanced by supplementation with crystalline amino acids, especially methionine and lysine and, in the case of salmon, histidine. C.1.3. Anti-nutritionals and other undesirable components. Anti-nutritional factors are endogenous substances in foods or feed ingredients that can produce negative effects on health and nutrient balance via mechanisms including effects on digestion, metabolic regulation and immune function. Plant products in particular contain a range of antinutritionals including enzyme (especially proteinase) inhibitors, glycosides such as saponins, phytic acid/phytate, glucosynolates, lectins, phytoestrogens, phytosterols, antivitamins (e.g. thiaminase, chitin and fluorine), allergens, tannins, alkaloids, and cyanogens among others (Table C3) (Gatlin III et al., 2007). The precise chemical nature of these compounds and their potential effects in fish including salmonids have been the subject of several comprehensive reviews (Francis et al., 2001; Hemre et al., 2009; Krogdahl et al., 2010; NRC, 2011). Currently, the major replacer for FM is SPC that does not result in the significant health problems associated with the inclusion of full-fat or solvent extracted soybean meal (SBM) in 130

133 salmonids, namely the enteropathy affecting the distal intestine (Baeverfjord and Krogdahl, 1996). This non-infectious sub-acute enteritis may be due to the presence of high levels of soyasaponins in SBM (Chikwati et al., 2012). Sunflower meal and wheat gluten are the next most quantitatively important replacers for FM and relatively low levels of anti-nutritionals are key characteristics. Thus, sunflower expeller is generally lower in the anti-nutritionals common in other oilseed meals such as soybean and rapeseed. Legume products such as pea protein concentrate have been reported to have potential negative effects including enteropathy, similar to SBM, likely associated with the presence of specific anti-nutritionals (Penn et al., 2011). However, this was only observed when pea protein concentrate was fed in high concentrations of around 35 % as lower levels of legumes do not appear to have negative effects (Penn et al., 2011). In addition to restricting inclusion levels, anti-nutritional factors can be controlled using heat treatments and phytase, and genetic improvement of plants in the longer term (Gatlin III et al., 2007). Adventitious toxins are compounds that are natural or accidental contaminants derived from human sources (other than pathogenic agents) present in/on feed ingredients that could have potential health or environmental effects. The main examples in FM and FO are biogenic amines (e.g. histamine) produced by the action of anaerobic bacteria on amino acids, and the environmental pollutants dioxins and PCBs discussed earlier. In plant ingredients they include organochlorine pesticide residues (e.g. aldrin, dieldrin, toxaphene), polycyclic aromatic hydrocarbons (PAHs), metals and other mineral elements (e.g. arsenic, cadmium, lead and mercury), and mycotoxins including aflatoxin, ochratoxin and fumonisin among others produced by fungal species (e.g. Aspergillus, Penicillium and Fusarium) that can be produced under processing and/or storage. The old pesticides are those that are considered as environmental contaminants (DDT, HCB etc) associated with FO, whereas the new contaminants/pesticides are those currently used on crops are being introduced through the increased use of plant ingredients in feeds. A recent study reported a widescope screening for detection and identification of pesticides and polycyclic aromatic hydrocarbons (PAHs) in feeds and fish tissues (Nácher-Mestre et al., 2014). Qualitative validation was performed for over 133 representative pesticides and 24 PAHs and subsequent application of the screening method to aquaculture samples enabled the detection several compounds from the target list, including chlorpyrifos-methyl, pirimiphosmethyl and ethoxyquin (Nácher-Mestre et al., 2014). However, in its recent comprehensive report, the Norwegian Scientific Committee for Food Safety concluded that for new contaminants in fish feed such as the pesticide endosulfan, PAHs and mycotoxins, the concentrations in farmed Norwegian salmon fish were likely not to be a food safety issue since the concentrations were very low and often not detectable even with sensitive analytical methods (VKM, 2014). One further concern in Europe regarding plant proteins is the presence of geneticallymodified (GM) products, especially those derived from soybean, canola and maize (Pusztai and Bardocz, 2006). However, a study in Atlantic salmon showed that, although short transgenic sequences were detected in gut tissues, no transgenic fragments were found in liver, muscle or brain (Sanden et al., 2004). Although there is no scientific evidence that the inclusion of GM ingredients in fish feeds would be in any way harmful either to the fish or 131

134 human consumers and, despite the fact that almost all major retailers have openly acknowledged, without significant negative public reaction, that it is impossible to guarantee their poultry products are fed GM-free ingredients, salmon feeds in Scotland are still GMfree. Recent advances in the development of GM oilseed crops producing EPA and DHA (see below) will potentially pose the most significant challenge to the GM debate in the near future. C.1.4. Palatability and attractants. It has often been observed that feed intake can be negatively affected when high levels of FM is replaced with plant protein products albeit this effect may be temporary and the fish subsequently adapt to the new feedstuffs and feed intake increases (Torstensen et al., 2005). Palatability and stimulation of fish feed ingestion is the sum of many dietary characteristics including physical stimuli, such as colour and movement, as well as chemical stimuli, such as smell and taste of food particles (Kolkovski et al. 1997). Various metabolic compounds including free amino acids, betaine, nucleotides, nucleosides, amines, sugars, and organic acids/alcohols/aldehydes can act as attractants stimulating feeding activity and food consumption (Kasumyan and Døving, 2003). The precise mechanism responsible for decreased feed intake sometimes observed in salmon when fed diets with high levels of FM replacement is not known. However, beneficial effects of inclusion of small amounts of other marine meals such as krill meal or squid meal (see below) may include stimulation of feed intake and thus be related to the provision of feed attractants that these products likely contain. C.2. Other alternative protein ingredients C.2.1 Processed Animal Proteins/Products (PAPs). Formerly more commonly referred to as Land Animal Products (LAPs), PAPs are essentially, by-products of the terrestrial livestock rendering industry and include poultry/porcine blood meal, poultry feather meal, poultry byproducts, meat and bone meals. Generally, PAP ingredients compare favorably cost-wise with many other types of protein sources commonly used in fish feeds. The PAPs are produced using a variety of raw materials and cooking and drying equipment and conditions. In some respects, PAPs are the most directly comparable substitute for fishmeal as their amino acid compositions is generally more similar to fishmeal than most plant products as well as also containing minerals, phospholipids and cholesterol depending upon the type of product. They also do not have high levels of anti-nutritional factors. However, more attention needs to be paid to accurately characterising of the nutritive value of the different types of animal proteins available on the market. Furthermore, the use of PAPs is not without other complications, at least in some regions of the world. Quality variability is often cited as one of the issues limiting the use of PAPs in markets where they are allowed, although production practices have greatly improved and many PAPs are of consistent quality and high nutritive value for fish and shrimp. However, the outbreak of bovine spongiform encephalopathy (BSE) in the UK in the 1980 s led to a ban on the use of animal products in aquaculture feeds within the European Union for many years. Although this ban has been relaxed on some products such as blood meals, there is still only limited use of 132

135 PAPs in aquaculture feeds in Europe, mainly due to consumer and retailer resistance based on perceived risks of including animal by-products in animal feeds. Consequently, there are currently no PAPs used in Atlantic salmon feeds in Scotland or Norway. In addition, porcine products are probably unlikely to ever be used due to different religious belief systems (halal and kosher) that prohibit the consumption of pig products leaving essentially only poultry by-products as potential future ingredients. Currently, salmon feed formulations in Chile include between % PAPs (predominantly feather meal?) and around 3 % poultry oil (industry data). Despite this, consumer and retailer resistance suggests that there is little prospect for the use of PAPs in feeds for Scottish salmon in the near future. C.2.2. Insects and insect meal. Insects can be the natural food for many fish species, including salmon in freshwater. There is currently considerable interest in developing insects and insect meals as new ingredients in feeds for animals, particularly poultry and fish. The EU initiative PROteINSECT have suggested that one hectare of land could produce at least 150 tons of insect protein per year compared to under a ton of protein per year for soy planted over the same area. Insect species under investigation include black soldier fly (Hermetia illucins), mealworms/mealworm beetle (Tenebrio molitor), houseflys (Musca domestica) and silkworm (Bombyx mori) with the larval forms (maggots/pupae) being the source of the protein. European companies such as Protix biosystems (Netherlands) and Ynsect (France) produce insect meals (technically just another form of PAP) albeit their cost (e.g. ~ 3000/ton for the Ynsect product) is still not competitive with FM. However, according to FAO the price of FM will reach 2,000-2,200 by 2020 and so optimising insect products and scalable technology with associated increased production capacity could result in decreased prices for insect meal that may become competitive with FM by Ironically, the high cost of insect meal is partly based on the costs of the insects feed. Currently in EU regulations forbid the use of biowaste, such as manure, retail or municipal wastes, as feed and so insect meal produced in Europe must use plant based by-products from agrifood industries. In addition, the downstream processing technology (drying, grinding and defatting etc.) is expensive. Together these limitations drive up cost and therefore limit application, which in turn is probably hampering the expansion in volume required both to reduce prices and to reliably supply the aquaculture feed industry. Current production is in the hundreds to low thousands of tonnes and even Ynsects planned (construction starting 2016) large-scale facility will only produce 10,000 tonnes of insect products. Biowaste is, of course, cheap and, in other regions such as South East Asia this source can be used to feed insects, which in turn could help to make insect products more cost effective such that they could be used to replace FM in these area that, indirectly, could ease some pressure on global FM supplies. Regulation hurdles also remain. In the EU, regulators are waiting for more extensive human health and safety data, although discussions are ongoing about opening up the aquaculture industry to insect feedstock. 133

136 C.2.3 Other marine meals. Krill meal contains almost 60 % protein and around 9 % oil (NRC, 2011), and is currently the most commonly used alternative marine protein ingredient in salmon feeds in Scotland. However, it is generally used at relatively low levels (< 5 %) and is included more for the provision of micronutrients and other bioactive molecules (e.g. attractants etc.) that can have beneficial health effects in fish rather than as a source of protein per se. Krill meal is produced alongside krill oil that is arguably the more important product driving the krill fishery as an important additional source of n-3 LC-PUFA, which in addition is largely esterified to phospholipid increasing bioavailability and increasing its value as a supplement for direct human consumption. Other crustacean meals include shrimp and crab meals, which are dried recovered fishery wastes and are, therefore, by-products of the seafood processing industry. Shrimp and crab meals can contain between % protein but they also contain very high levels of ash (27-41 %) (NRC, 2011), and therefore have generally been tested in fish feed formulations as sources of minerals and, especially, the natural pigment, astaxanthin, and possibly feed attractants due to the presence of free amino acids. Two commercial animal feed ingredients are produced from squid; squid meal, which is a high protein (76.5 %) product and squid liver meal, which has lower protein (45 %) and higher lipid (15 %) (NRC, 2011). Production volumes are relatively low and so costs are relatively high, which limits the use of squid meals to more specialised feeds for broodstock or as live feed enrichers or supplements to larval marine fish. C.3 Lipid/oil sources Global oil and fat production totalled more than 185 million tonnes in 2012, with total production of VO at over 160 million tonnes, and animal fats including tallow, lard and butter totalling another 25 million tonnes (Table C4). Therefore alternatives to FO are plentiful but they all suffer from the same major drawback in contrast to FO in that they are not sources of the n-3 LC-PUFA, EPA and DHA (Table C4). Higher plants generally do not produce fatty acids with chain length greater than C 18 and so C 20 and C 22 LC-PUFA are not components of any VO, and animal fats are dominated by saturated and monounsaturated fatty acids with only low levels of PUFA in general (Table C5). C.3.1. Issues with alternative oils. Replacement of FO generally poses no problem in terms of energy supply to support optimal growth and production, and there are no major problems in terms of general product quality or, in many cases, consumer acceptance in many parts of the world (Torstensen et al., 2005; Rosenlund et al., 2011). As alluded to above, the main issue is fatty acid composition and the problem in replacing FO reduces to the problem of alternative sources on n-3 LC-PUFA (Tocher, 2015). The fatty acid compositions of the majority of VO are dominated by only four fatty acids, 16:0, 18:0, 18:1n-9 and 18:2n-6 (Table C5) (Gunstone, 2011). Saturated fatty acid-rich oils include palm oil (16:0) and the shortchain or lauric (12:0)-rich oils, coconut and palm kernal oils, monoenoic acid (18:1n-9)-rich oils include rapeseed and olive oils, and linoleic acid (18:2n-6)-rich oils include soybean, sunflower, cottonseed, corn and sesame. Linolenic acid (18:3n-3) is much less common and 134

137 the only major VO rich in this fatty acid is linseed (flax), although it still only represented 0.6 % of world VO production in 2012 (Table C4). Soybean and rapeseed oils contain reasonable amounts of 18:3n-3, but with 18:2n-6 : 18:3n-3 ratios of almost 7 and over 2, respectively. Some oilseeds contain Δ6-desaturated fatty acids such as 18:3n-6 in borage and evening primrose oils, and stearidonic acid (18:4n-3) in echium and bugglossoides oils (Table C5) (Gunstone, 2011). The same issues currently restricting the use of animal-derived protein sources also pertain to the use of animal fats in salmon feeds. In any case, lard and tallow have physical properties (lack of fluidity) based upon their high contents of saturated fatty acids that limit their use in cold-water species such as salmon, particularly in winter (Bureau and Meeker, 2011). The same argument applies to any high-saturated oil, such as palm oil, although this can be included at low levels. Currently, VO are the only readily available, cost effective, sustainable alternative lipid source for salmon feeds and so, in the last 15 years, considerable research has been performed to investigate the effects of complete and partial replacement of dietary FO with VO in a range of species including Atlantic salmon (Turchini et al., 2009, 2011a). The VOs most commonly used include rapeseed/canola, soybean, palm, sunflower, corn, olive, linseed, camelina and echium oils. The scientific literature is now extensive and several reviews are also available (Bell and Waagbø, 2008; Bell and Tocher, 2009; Tocher, 2009; Turchini et al., 2009, 2011a). The overarching conclusion is that with dietary oils, fish including salmon are what they eat and so compositions of farmed products are substantially influenced by the fatty acid composition of the feeds, and thus the oil used in the formulations (Table C5). Therefore, as VOs do not contain EPA or DHA, increasing replacement of FO with VO inevitably impacts on the content of these fatty acids in salmon flesh (Table C5) (Turchini et al., 2009, 2011a; Sales and Glencross, 2011; Tocher et al., 2011). However, these studies have also provided the information for the most effective use of VOs to minimise the effects of sustainable feed formulations. In terms of which oil to use, studies have indicated the characteristics that are desirable in a VO for it to be an effective replacement for FO; these include a high monoene content to effectively supply energy, relatively low 18:2n-6 as this is already in excess in the human diet, and a reasonable level of 18:3n-3 to exploit the endogenous ability of salmon to produce some EPA and, perhaps DHA, from the C 18 precursor (Torstensen and Tocher, 2011; Turchini et al., 2011b; Emery et al., 2013; Lewis et al., 2013). Rapeseed oil (RO) is the single VO that best fits these criteria having > 60 % 18:1n-9, 10 % 18:3n-3, and moderate 18:2n-6 and an n-6:n-3 PUFA ratio of only 2 and therefore RO is the preferred choice in the Scottish salmon industry as well as being locally produced and readily available in Scotland. A further strategy to minimise the impact of sustainable feeds on levels of n-3 LC-PUFA in salmon flesh is to utilise finishing feeds. These feeds contain higher levels of FO and can be fed for variable times prior to harvest to increase EPA and DHA to satisfy particular retail specifications. C.3.2 Supply of n-3 LC-PUFA. The issue in FO replacement is based on the characteristic of FO as a unique source of n-3 LC-PUFA, EPA and DHA. As described above, n-3 LC-PUFA are important nutrients with key metabolic and functional roles in salmon and there are physiological requirement levels, albeit these are relatively low and can be possibly satisfied 135

138 by supplying dietary 18:3n-3 (Glencross, 2009; Tocher, 2003, 2010). The major problem in replacing FO in salmon feeds is maintaining n-3 LC-PUFA in farmed fish at the high levels that have salmon recognised as a beneficial and healthy part of the human diet. As FO and, to a lesser extent, FM are the primary sources, this implies that the current global supply of n-3 LC-PUFA is similarly finite and limited, and the gap between supply and demand can be estimated. Based on the most commonly recommended dose for cardiac health (500 mg/day; see GOED, 2014), the total demand for n-3 LC-PUFA is over 1.25 million tonnes whereas total supply is optimistically estimated at just over 0.8 million tonnes indicating a shortfall of over 0.4 million tonnes (Table C7). The majority of supply (almost 90 %) is from capture fisheries, whether as food fish or via FO and FM, with relatively small additional amounts realistically estimated from seafood by-products and recycling, unfed aquaculture and algal sources. While it is acknowledged that the calculations in Table C7 contain some assumptions and estimates, and the precise extent of the difference can be the argued, the fact that the gap exists is not in question irrespective of how it is calculated (Naylor et al., 2009). There is a fundamental, global lack of n-3 LC-PUFA to supply all human needs, whether by direct consumption or via aquaculture. C.3.3. Alternative sources of n-3 LC-PUFA. There is an urgent need to find alternative sources of EPA and DHA. The following discussion assesses all other possible current or future sources of n-3 LC-PUFA as potential alternatives to FO in feeds for Atlantic salmon. C Lower trophic levels and mesopelagic fish. Possible alternative marine sources for oils include utilising lower trophic levels, specifically zooplankton such as krill and calanoid copepods in the southern and northern hemispheres, respectively, and mesopelagic fish. The biomass at lower trophic levels is large, but there are inherent dangers associated with fishing down the marine food web (Pauly et al., 1998). The utilisation of krill and copepods has been studied and, in general, zooplankton can be potentially good lipid sources (Olsen et al., 2011). However the harvesting of krill and copepods poses significant technological challenges and cost. For most species, lack of schooling behaviour makes harvest by traditional trawling technology an expensive economic option (Olsen et al., 2011). Antarctic krill, which do form schools, is the only species being targeted for commercial harvest, apart from a small scientific quota (~1000 mt) of the calanoid copepod, Calanus finmachicus (Croxall and Nicol, 2004). MSC-certified krill meal up to 10 % is currently being used in some salmon feeds during the seawater phase in both Norway and Scotland as the higher growth rate and consequent shorter cycle compensates for the increased cost. Other feeds can have krill meal at even higher levels, but these are premium feeds focussed on health benefits and are used sparingly. Therefore these krill products are not being used as primary sources of n-3 LC-PUFA. Currently, krill lipid products are used almost exclusively for the human nutraceutical market. Although there may be evidence that harvesting krill, and potentially copepods, could be sustainable, there are still significant environmental and ecological concerns 136

139 (Olsen et al., 2011). For instance, Antarctic krill are near the base of a food chain that includes whales and penguins that would suggest there could be opposition to greatly increased exploitation (Hill et al., 2006). Copepods have an additional problem, as the oil is rich in wax esters rather than TAG, which may limit its widespread use (Olsen et al., 2004; Bogevik et al., 2010). Mesopelagic fish that inhabit the intermediate pelagic water masses between the euphotic zone at 100 m depth and the deep bathypelagic zone at 1000 m are available in potentially large quantities (1-6 billion tonnes), with lantern fish, myctophids, constituting about 60 % of biomass (Irigoien, 2014). Different species can contain between 16 and 60 % of dry weight (Falk-Petersen et al., 1986; El-Mowafi et al., 2010) as lipid that can be in the form of TAG or, in some species, wax esters and, regardless of species, they are all good sources of n-3 LC-PUFA (Olsen et al., 2011). On the positive side, they are resources that, so far, have not been the subject of commercial exploitation, and they do not compete with existing or potential human feed production. Negative points include biological (seasonal variation), ecological (mixed fishery difficult to manage), technical (capture methods and on-board processing), and nutritional (wax esters) issues. C By-catch and seafood processing by-products. In almost all fisheries there are nontarget catch and/or discarded target catch that, together, make up the by-catch. However, both the precise definition and resultant estimates of by-catch can be controversial (FAO, 2009). In 2005, the discard rate was estimated at around 7 million tonnes/year or 8 % of global catch (Kelleher, 2005). By its nature, by-catch is a diffuse resource (Batista, 2007) and this imposes a major limitation to its usefulness as a source of FO, although processing of by-products, including oil production, at sea is an increasing trend (Falch et al., 2007). Another limitation is that by-catch includes a multitude of species, not necessarily oily, which limits the quantity and quality (lipid class and fatty acid compositions, lipid soluble nutrients and contaminants) of the oils produced (Batista, 2007). Another potential source of fish/marine oils is seafood industry by-products including viscera, heads, carcasses and trimmings, particularly those produced from pelagic fisheries and aquaculture. Whereas these by-products can produce significant amounts of FM (Lekang and Gutierrez, 2007), the production of oil is largely dictated by species. Thus, by-products from oily species including salmon, herring and mackerel can be a source of substantial FO whereas by-products from other pelagic (white fish) fisheries have generally lower lipid contents (Lekang and Gutierrez, 2007). Liver from species like cod and halibut have traditionally been used for FO production, but production is now relatively small (~ 40,000 tonnes) and goes mostly for direct human consumption as vitamin A and D supplements as much as sources of n-3 LC-PUFA (Hertrampf and Piedad- Pascual, 2000). Oil is now being actively recovered from aquaculture species waste, particularly salmon farming, with around 20,000 tonnes reportedly recovered in Norway (Rubin, 2009), and 50,000 tonnes in Chile in 2006 (Tacon and Metian, 2008). 137

140 Other limitations to the use of oils from aquaculture by-products include regulatory issues preventing intra-species use to prevent recycling of contaminants and disease transfer, and so oils from aquaculture are currently used either for human nutrition or other farmed species depending on quality (Skåra et al., 2004). In addition, recycled oils from freshwater aquaculture species including carps and tilapia can be of limited value due to the generally low levels of n-3 LC-PUFA (Borghesi et al., 2008). Overall, by-catch and seafood by-products are potential sources of FO and, although there is currently some production, the contribution of these sources to overall FO supply is not well quantified (Jackson and Shepherd, 2012; Shepherd and Jackson, 2013). C Single cell oils (microalgae). Potentially, culture of the main primary producers, marine microalgae, could offer the ideal long-term, sustainable solution to the problem of n-3-lc-pufa supply (Apt and Behrens, 1999). Various photosynthetic microalgae are already commonly used in hatcheries to supply both EPA (e.g. diatoms) and DHA (e.g. flagellates) to live feeds, rotifers (Brachionus spp.) and Artemia nauplii, for the rearing of larval marine crustacean and fish species (Muller-Feuga, 2004). Production usually employs medium- to high-density batch, semi-continuous or continuous culture in relatively small volumes (Stottrup and McEvoy, 2003). Up-scaling of production to the volumes required for algal oil and/or algal biomass to supply the amount of n-3 LC-PUFA required to replace FO in commercial aquafeed production has very significant biological and technological challenges (Chauton et al., 2015). Economic production of n-3 LC-PUFA would require algae to demonstrate simultaneous high growth and high lipid content with a high proportion of EPA and DHA. These can be almost exclusive traits as lipid deposition is often associated with conditions when growth is limited (e.g. N limitation) (Richmond, 2008). Technical challenges include efficient capture of light energy in highdensity culture with effective temperature control. These issues remain to be solved but there are several research strategies targeting their solution. These include exploiting culture conditions to direct metabolism towards lipid production, to improve biomass productivity or lipid yield by mutagenesis and selective breeding, and to improve strains by genetic modifications to optimise light absorption and increase biosynthesis of EPA and DHA (Chauton et al., 2015). Therefore, in the future, algal species with favourable biological characteristics may be found and/or developed, and photo-bioreactor technologies may improve considerably enabling economically sustainable production of microalgae rich in EPA and DHA for use in aquafeed. In contrast, heterotrophic microalgal species including Crypthecodinium and thraustochytrids such as Schizochytrium are already being utilised for the commercial production of DHA using large-scale biofermentor technology (Raghukumar, 2008). Even so, the high production costs are generally limiting the use of these products to direct human consumption mainly in the form of DHA supplementation of infant formulae (Ward and Singh, 2005). Research showed that substituting FO with thraustochytrid oil in the diet of Atlantic salmon parr significantly increased muscle DHA without any detriment 138

141 to growth (Miller et al., 2007). Similarly, biomass from Schizochytrium and C. cohnii have been used to substitute for FO with no deleterious effects in larval microdiets and starter feeds for gilthead seabream (Sparus aurata) (Atalah et al., 2007; Ganuza et al., 2008). Currently, a thraustrochytrid algal biomass product (DHAgold, DSM, Maryland) containing around 18 % DHA (by weight) is commercially available and apparently being used in aquaculture (Miller et al., 2011). Therefore, these DHA-rich products from heterotrophic microalgae may have niche markets in marine hatcheries, particularly for high-value marine species. Production volumes would have to be increased and costs reduced before these products could be viable for wider application in aquaculture and, particularly, salmon farming. Biochemically, Schizochytrium sp. are particularly interesting as they appear to have two alternative pathways for n-3 LC-PUFA biosynthesis (Lippmeier et al., 2009). Primary production of DHA and 22:5n-6 in Schizochytrium is via a PUFApolyketide synthase (PKS) (Metz et al., 2001), a series of three genes similar to those found in LC-PUFA-producing marine bacteria (Yazawa, 1996; Morita et al., 2000; Gong et al., 2014). In addition, a series of aerobic PUFA desaturase and elongase genes are also present in Schizochytrium (Lippmeier et al., 2009). C Transgenic oils as sources of omega-3. As discussed above, it is unlikely that microalgal biomass can be produced on the scale necessary and at an economic cost to satisfy the demands of aquaculture for n-3 LC-PUFA, at least in the short- to mediumterm (Miller et al., 2011; Chauton et al., 2015). However, microalgae represent a highly valuable source of genes encoding for the biosynthetic enzymes required for n-3 LC-PUFA production (Venegas-Caleron et al., 2010). The overall strategy being to genetically modify existing organisms that have oil deposition as a major trait and thus combine this with the n-3 LC-PUFA biosynthesis trait. Potential candidates include other oleaginous microorganisms or conventional oilseed crops to produce entirely novel sources of de novo n-3 LC-PUFA (Zhu et al., 2010; Sayanova and Napier, 2011). Progress into the metabolic engineering of oleaginous microorganisms to produce n-3 LC- PUFA has been recently reviewed (Gong et al., 2014). The most successful to date has been the metabolic engineering of the yeast Yarrowia lipolytica, which resulted in a strain that produced EPA at 15 % of dry weight (Xue et al., 2013). It was shown that the EPA- Yarrowia cell mass was suitable as a feed ingredient for Atlantic salmon (Hatlen et al., 2012) although disruption was required to increase the bioavailability of lipid and EPA (Berge et al., 2013). To the best of the author s knowledge, the transgenic yeast oil or cell biomass are not yet fully commercially available although they appear to be used by a DuPont/AquaChile venture (Verlasso ) to produce a niche salmon product ( However, as the production of the transgenic yeast uses biofermentor technology similar to that used for Schizochytrium, it appears unlikely that it could be produced in volumes and at a cost that would make it viable as a large-scale alternative to FO in aquaculture, at least in the short- to medium-term. 139

142 Oilseed crops dominate world oil production and there is a highly organised and wellestablished infrastructure for the cultivation, harvest, processing, distribution, marketing and utilisation of vegetable oils (Salunkhe et al., 1992). Therefore, oilseed crops are highly practical platforms from which to develop a novel, renewable supply of n-3 LC-PUFA. However, conventional plant breeding strategies cannot be used as the genes required for LC-PUFA synthesis are simply not present in higher plants, leaving transgenesis as the only option for modification of oilseeds to contain LC-PUFA. Therefore, the only currently viable approach to developing a novel, renewable supply of EPA and DHA is the metabolic engineering of oilseed crops with the capacity to synthesise n-3 LC-PUFA in seeds (Haslam et al., 2013). Production of n-3 LC-PUFA in terrestrial plant seeds was demonstrated in the model plant Arabidopsis (Petrie et al., 2012; Ruiz-Lopez et al., 2013), and very recently reported in an oilseed crop, Camelina sativa (Petrie et al., 2014; Ruiz-Lopez et al., 2014). C. sativa or false flax, is a member of the Brassicaceae family and an ancient crop that, in the wild-type, produces an oil with α-linolenic acid (18:3n-3) at up to 45 % of total fatty acids (Gunstone and Harwood, 2007). Transgenic C. sativa lines have now been developed by transformation with algal genes encoding the n-3 LC-PUFA biosynthetic pathway and expression restricted to the seeds via seed-specific promoters to produce oils with over 20 % of total fatty acids as n-3 LC-PUFA, either as EPA alone or as EPA + DHA (12 % + 8 %) (Ruiz-Lopez et al., 2014). As well as being easily transformable by Agrobacterium floral infiltration, Camelina has additional desirable traits including modest input requirements (water and pesticides) and ability to thrive in semi-arid conditions (Tocher et al., 2011). In the US, several states are actively growing Camelina as a biofuels crop, indicating the wide acceptance of this crop platform. Furthermore, wild-type Camelina oil has already been shown to be suitable for inclusion in fish feeds and contains no anti-nutritional factors detrimental to fish growth (Petropoulos et al., 2009; Morais et al., 2012; Hixson et al., 2014). Ultimately, all animal production will depend on terrestrial plants/agriculture and this requires land. However, it is pertinent to emphasise that the production of n-3 LC-PUFA in terrestrial oilseed crops should not require additional arable land as the ideal solution would be to switch some VO production from n-6 PUFA-rich crops to the new n-3 LC-PUFA crops. However, irrespective of the advances in science, one of the greatest challenges, at least in Europe, will be to change public opinion towards acceptance of genetically modified products before these transgenic oils can be used commercially on a global scale. 140

143 References Apt, K.E., Behrens, P.W., Commercial developments in microalgal biotechnology. J. Phycol. 35, Atalah, E., Cruz, C.M.H., Izquierdo, M.S., Rosenlund, G., Caballero, M.J., Valencia, A., Robaina, L., Two microalgae Crypthecodinium cohnii and Phaeodactylum tricornutum as alternative source of essential fatty acids in starter feeds for seabream (Sparus aurata). Aquaculture 270, Baeverfjord, G., Krogdahl, Å., Development and regression of soybean meal induced enteritis in Atlantic salmon, Salmo salar L., distal intestine: a comparison with the intestines of fasted fish. J. Fish Dis. 19, Batista, I., By-catch, underutilized species and underutilized fish parts as food ingredients, in: Shahidi, F. (Ed.), Maximizing the value of marine by-products. Woodhead Publishing Limited, Cambridge, pp Bell, J.G., Henderson, R.J., Tocher, D.R., McGhee, F., Dick, J.R., Porter, A., Smullen, R., Sargent, J.R Substituting fish oil with crude palm oil in the diet of Atlantic salmon (Salmo salar) affects muscle fatty acid compositions and hepatic fatty acid metabolism. J. Nutr. 132, Bell, J.G., Henderson, R.J. Tocher, D.R., Sargent, J.R., Replacement of dietary fish oil with increasing levels of linseed oil; modification of flesh fatty acid compositions in Atlantic salmon (Salmo salar) using a fish oil finishing diet. Lipids 39, Bell, J.G., McEvoy, J., Tocher, D.R., McGhee, F., Campbell, P.J., Sargent, J.R Replacement of fish oil with rapeseed oil in diets of Atlantic salmon (Salmo salar) affects tissue lipid compositions and hepatocyte fatty acid metabolism. J. Nutr. 131, Bell, J.G., Tocher, D.R., Farmed Fish: The impact of diet on fatty acid compositions, in: Rossell, B. (Ed.), Oils and Fats Handbook volume 4; Fish Oils. Leatherhead Food International, Leatherhead, pp Bell, J.G., Waagbø, R., Safe and nutritious aquaculture produce: benefits and risks of alternative sustainable aquafeeds, in: Holmer, M., Black, K., Duarte, C.M., Marba, N., Karakassis, I. (Eds.), Aquaculture in the Ecosystem. Springer Verlag BV, London, pp Berge, G.M., Hatlen, B., Odom, J.M., Ruyter, B., Physical treatment of high EPA Yarrowia lipolytica biomass increases the availability of n-3 highly unsaturated fatty acids when fed to Atlantic salmon. Aquacult. Nutr. 19, Bogevik, A.S., Henderson, R.J., Mundheim, H., Waagbø, R., Tocher, D.R., Olsen, R.E., The influence of temperature on the apparent lipid digestibility in Atlantic salmon (Salmo salar) fed Calanus finmarchicus oil at two dietary levels. Aquaculture 309, Borghesi, R., Arruda, L.F., Oetterer, M., Fatty acid composition of acid, biological and enzymatic fish silage. Boletim do Centro de Pesquisa de Processamento de Alimentos 26, Bransden, M.P., Carter, C.G., Nichols, P.D Replacement of fish oil with sunflower oil in feeds for Atlantic salmon (Salmo salar L.): effect on growth performance, tissue fatty acid composition and disease resistance. Comp. Biochem. Physiol. 135B, Bureau, D.P., Meeker, D.L., Terrestrial animal fats, in: Turchini, G.M., Ng, W.-K., Tocher, D.R. (Eds.), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp

144 Chauton, M.S., Reitan, K.I., Norsker, N.H., Tveterås, R., Kleivdal, H.T., A techno-economic analysis of industrial production of marine microalgae as a source of EPA and DHA-rich raw material for aquafeed: Research challenges and possibilities. Aquaculture 436, Chikwati, E.M., Venold, F.F., Penn, M.H., Rohloff, J., Refstie, S., Guttvik, A., Hillestad, M., Krogdahl, Å., Interaction of soyasaponins with plant ingredients in diets for Atlantic salmon (Salmo salar L.). Br. J. Nutr. 107, Croxall, J.P., Nicol, S., Management of Southern Ocean fisheries: global forces and future sustainability. Antarctic Sci.16, Einen, O., Roem, A.J., Dietary protein/energy ratios for Atlantic salmon in relation to fish size: growth, feed utilisation and slaughter quality. Aquacult. Nutr. 3, El-Mowafi, A., Nanton, D., Berntssen, M., Evaluation of lantern fish (Benthosema Pterotum) as marine source in fish feeds: nutrient composition and contaminants assessment. Proc. 3rd Global Fisheries Aquacult. Res. Conf. Foreign Agricultural Relations (FAR), Egypt, 29 November - 1 December 2010 pp Emery, J.A., Hermon, K., Hamid, N.K.A., Donald, J.A., Turchini, G.M., 2 substrate competition: Dietary linoleic acid (18:2n- (18:3n-3) bioconversion in the teleost rainbow trout. PLoS ONE 8, e Desaturase -linolenic acid Emmert, J.L., Baker, D.H., Use of the ideal protein concept for precision formulation of amino acid levels in broiler diets. J. Appl. Poul. Res. 6, Falch, E., Sandbakk, M., Aursand, M., On-board handling of marine by-products to prevent microbial spoilage, enzymatic reactions and lipid oxidation, in: Shahidi, F. (Ed.), Maximizing the value of marine by-products. Woodhead Publishing Limited, Cambridge, pp Falk-Petersen, I.B., Falk-Petersen, S., Sargent, J.R., Nature, origin and possible roles of lipid deposits in Maurolicus muelleri (Gmelin) and Benthosema glaciale (Reinhart) from Ullsfjorden, Northern Norway. Polar Biol. 5, FAO, The state of world fisheries and aquaculture Food and Agriculture Organization of the United Nations, Rome, 196 pp. FAO, The state of world fisheries and aquaculture Food and Agriculture Organization of the United Nations, Rome, 223 pp. Francis, G., Makkar, H.P.S., Becker, K., Antinutritional factors in plant-derived alternate fish ingedients and their effects in fish. Aquaculture 199, Ganuza, E., Benitez-Santana, T., Atalah, E., Vega-Orellana, O., Ganga, R., Izquierdo, M.S., Crypthecodinium cohnii and Schizochytrium sp as potential substitutes to fisheries-derived oils from seabream (Sparus aurata) microdiets. Aquaculture 277, Gatlin III, D.M., Barrows, F.T., Brown, P., Dabrowski, K., Gibson, G.T., Hardy, R.W., Elliot, H., Hu, G., Krogdahl, A.,Nelson, R., Overturf, K., Rust, M., Sealey, W., Skonberg, D., Souza, E.J., Stone, D., Wilson, R., Wurtele, E., Expanding the utilization of sustainable plant products in aquafeeds: a review. Aquacult. Res. 38, Glencross, B.D., Exploring the nutritional demand for essential fatty acids by aquaculture species. Rev. Aquacult. 1, GOED, Global recommendations for EPA and DHA intake (Rev 16 April 2014). Accessed 27 October

145 Gong, Y., Wan, X., Jiang, M., Hu, C., Hu, H., Huang, F., Metabolic engineering of microorganisms to produce omega-3 very long-chain polyunsaturated fatty acids. Prog. Lipid Res. 56, Gunstone, F.D., The world s oils and fats, in: Turchini, G.M., Ng, W.-K., Tocher, D.R. (Eds), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp Gunstone, F.D., Harwood, J.L., Occurrence and characteristics of oils and fats, in: Gunstone, F.D., Harwood, J.L., Dijkstra, A.J. (Eds.), The lipid handbook, 3rd edition. Taylor & Francis, CRC Press, Boca Raton, pp Hardy, R.W., Utilization of plant proteins in fish diets: effects of global demand and supplies of fishmeal. Aquacult. Res. 41, Haslam, R.P., Ruiz-Lopez, N., Eastmond, P., Moloney, M., Sayanova, O., Napier, J.A., The modification of plant oil composition via metabolic engineering better nutrition by design. Plant Biotechnol. J. 11, Hatlen, B., Berge, G.M., Odom, J.M., Mundheim, H., Ruyter, B., Growth performance, feed utilisation and fatty acid deposition in Atlantic salmon, Salmo salar L., fed graded levels of highlipid/high EPA Yarrowia lipolytica biomass. Aquaculture, , Hemre, G.I., Amlund, H., Aursandf, A., Bakke, A.M., Olsen, R.E., Ringo, E., Svihus, B., Criteria for safe use of plant ingredients in diets for aquacultured fish, Opinion of the Panel on Animal Feed of the Norwegian Scientific Committee for Food Safety. Oslo, Norway:VKM. Henriques, J., Dick, J.R., Tocher, D.R., Bell, J.G., Nutritional quality of salmon products available from major retailers in the UK: Content and composition of n-3 long-chain polyunsaturated fatty acids. Br. J. Nutr. 112, Hertrampf, J. W., Piedad-Pascual, F., Handbook of ingredients for aquaculture feeds. Kluwer Academic Publishers. Dordrecht. Hill, S.L., Murphy, E.J., Reid, K., Trathan, P.N., Constable, A.J., Modelling Southern Ocean ecosystems: krill, the food-web, and the impacts of harvesting. Bio. Rev. 81, Hixson, S.M., Parrish C.C., Anderson D.M., Full substitution of fish oil with camelina (Camelina sativa) oil, with partial substitution of fish meal with camelina meals, in diets for farmed Atlantic salmon (Salmo salar) and its effect on tissue lipids and sensory quality. Food Chem. 157, Irigoien, X., Klevjer, T.A., Røstad, A., Martinez, U., Boyra, G., Acuña, J.L., Bode, A., Echevarria, F., Gonzalez-Gordillo, J.I., Hernandez-Leon, S., Agusti, S., Aksnes, D.L., Duarte, C.M., Kaartvedt, S., Large mesopelagic fishes biomass and trophic efficiency in the open ocean. Nature Commun. 5:3721, DOI: /ncomms4271 p.10. Jackson, A.J, Shepherd, C.J., The future of fishmeal and fish oil, in: Ryder, J., Ababouch, L., Balaban, M. (Eds), Second International Congress on Seafood Technology on Sustainable, Innovative and Healthy Seafood FAO Fisheries and Aquaculture Proceedings. No. 22. Food and Agriculture Organisation, Rome, 238 pp. Kasumyan A.O., Døving K.B., Taste preferences in fishes. Fish Fisheries 4, Kelleher, K., Discards in the world's marine fisheries: An update. FAO Fisheries Technical Paper 470, Rome,

146 Kolkovski, S., Arieli, A., Tandler, A., Visual and chemical cues stimulate microdiet ingestion in seabream larvae. Aquacult. Internat. 5, Krogdahl, Å., Penn, M.H., Thorsen, J., Refstie, S., Bakke, A.M., Important antinutrients in plant feedstuffs for aquaculture: An update on recent findings regarding responses in salmonids. Aquacult. Res. 41, Kumar, V., Sinha, A.K., Makkar, H.P.S., De Boeck, G., Becker, K., Phytate and phytase in fish nutrition. J. Anim. Physiol. Anim. Nutr. 96, Lekang, O.I., Gutierrez, M.A., Sources for omega 3 oils (In Norwegian). Rubin, Trondheim, Norway, pp Lewis, M.J., Hamid, N.K.A., Alhazzaa, R., Hermon, K., Donald, J.A., Sinclair, A.J., Turchini, G.M., Targeted dietary micronutrient fortification modulates n 3 LC-PUFA pathway activity in rainbow trout (Oncorhynchus mykiss). Aquaculture , Lippmeier, J.C., Crawford, K.S., Owen, C.B., Rivas, A.A., Metz, J.G., Apt, K.E., Characterization of both polyunsaturated fatty acid biosynthetic pathways in Schizochytrium sp. Lipids 44, Metz, J.G., Roessler, P., Facciotti, D., Levering, C., Dittrich, F., Lassner, M., Valentine, R., Lardizabal, K., Domergue, F., Yamada, A., Yazawa, K., Knauf, V., Browse, J., Production of polyunsaturated fatty acids by polyketide synthases in both prokaryotes and eukaryotes. Science 293, Miller, M.R., Nichols, P.D., Carter, C.G., Replacement of fish oil with thraustochytrid Schizochytrium sp. L. oil in Atlantic salmon parr (Salmo salar L) diets. Comp. Biochem. Physiol. 148A, Miller, M.R., Nichols, P.D., Carter, C.G., New alternative n-3 long chain polyunsaturated fatty acid-rich oil sources, in: Turchini, G.M., Ng, W.-K., Tocher, D.R. (Eds.), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp Morais, S., Edvardsen, R.B., Bell, J.G., Tocher, D.R., Transcriptomic analyses of genes of lipid metabolism in intestine of Atlantic cod (Gadus morhua) fed diets with increasing proportions of Camelina oil as replacement for fish oil. Comp. Biochem. Physiol. 161B, Morita, N., Tanaka, M., Okuyama, H., Biosynthesis of fatty acids in the docosahexaenoic acid-producing bacterium Moritella marina strain MP-I. Biochem. Soc. Trans. 28, Muller-Feuga, A., Microalgae for aquaculture. The current global situation and future trends, in: Richmond, A. (Ed.), Handbook of microagal culture. Blackwell, Oxford, pp Nácher-Mestre, J., Serrano, R., Portolés, T., Berntssen, M.H.G., Pérez-San chez, J., Hernández, F., Screening of pesticides and polycyclic aromatic hydrocarbons in feeds and fish tissues by gas chromatography coupled to high-resolution mass spectrometry using atmospheric pressure chemical ionization. J. Agricult. Fd. Chem. 62, NRC, Nutrient Requirements of Fish and Shrimp. The National Academies Press, Washington DC, 376 pp. Naylor, R.L., Hardy, R.W., Bureau, D.P., Chiu, A., Elliott, M., Farrell, A.P., Forster, I., Gatlin, D.M., Goldburg, R.J., Hua, K., Nichols, P.D., Feeding aquaculture in an era of finite resources. Proc. Natl. Acad. Sci. USA 106, Nofima, Today s and tomorrow's feed ingredients in Norwegian aquaculture. Tromsø, 69 pp. 144

147 Nofima, Resource utilisation of Norwegian salmon farming in 2012 and Tromsø, 35 pp. Olsen, R.E., Henderson, R.J., Sountama, J., Hemre, G.-I., Ringø, E., Melle, W., Tocher, D.R., Atlantic salmon, Salmo salar, utilizes wax ester-rich oil from Calanus finmarchicus effectively. Aquaculture 240, Olsen, R.E., Waagbø, R., Melle, W., Ringø, E., Lall, S.P., Alternative marine resources, in: Turchini, G.M., Ng, W.-K., Tocher, D.R. (Eds.), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp Pauly, D., Christensen,V., Dalsgaard, J., Froese, R., Torres, F., Fishing down marine food webs. Science 279, Penn, M.H., Bendiksen, E.Å., Campbell, P., Krogdahl, Å., High level of dietary pea protein concentrate induces enteropathy in Atlantic salmon (Salmo salar L.). Aquaculture 310, Petrie, J.R., Shrestha, P., Belide, S., Kennedy, Y., Lester, G., Liu, Q., Divi, U.K., Mulder, R.J., Mansour, M.P., Nichols, P.D., Singh, S.R., Metabolic engineering Camelina sativa with fish oil-like levels of DHA. PLoS ONE 9, e Petrie, J.R., Shrestha, P., Zhou, X.-R., Mansour, M.P., Liu, Q., Belide, S., Nichols, P.D., Singh, S.R., Metabolic engineering plant seeds with fish oil-like levels of DHA. PLoS ONE 7, e Petropoulos, I.K., Thompson, K.D., Morgan, A., Dick, J.R., Tocher, D.R. Bell, J.G., Effects of substitution of dietary fish oil with a blend of vegetable oils on liver and peripheral blood leukocyte fatty acid composition, plasma prostaglandin E 2 and immune parameters in three strains of Atlantic salmon (Salmo salar). Aquacult. Nutr. 15, Pusztai, A., Bardocz, S., GMO in animal nutrition: potential benefits and risks, in: Mosenthin, R., Zentek, J., Zebrowska, T. (Eds.), Biology of Growing Animals, Elsevier press, pp Raghukumar, S., Thraustochytrid marine protists: Production of PUFAs and other emerging technologies. Mar. Biotechnol. 10, Richmond, A. (Ed.), Handbook of Microalgal Culture: Biotechnology and Applied Phycology. John Wiley and Sons, New York, 584 pp. Rosenlund, G., Corraze, G., Izquierdo, M., Torstensen, B.E., The effects of fish oil replacement on nutritional and organoleptic qualities of farmed fish, in: Turchini, G.M., Ng, W.- K.,Tocher, D.R. (Eds.), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp Rubin, By-products; RUBIN, Trondheim, Norway. Ruiz-Lopez, N., Haslam, R.P., Usher, S.L., Napier, J.A., Sayanova, O., Reconstitution of EPA and DHA biosynthesis in Arabidopsis: iterative metabolic engineering for the synthesis of n-3 LC-PUFAs in transgenic plants. Metab. Eng. 17, Ruiz-Lopez, N., Haslam, R.P., Napier, J.A., Sayanova, O., Successful high-level accumulation of fish oil omega-3 long-chain polyunsaturated fatty acids in a transgenic oilseed crop. Plant J. 77, Sales, J., Glencross, B.D., A meta-analysis of the effects of dietary marine oil replacement with vegetable oils on growth, feed conversion and muscle fatty acid composition of fish species. Aquacult. Nutr. 17, e271-e287. Salunkhe, D.K., Adsule, R.N., Chavan, J.K., Kadam, S.S., World Oilseeds: Chemistry, Technology and Utilization. Van Nostrand Reinhold, New York. 145

148 Sanden, M., Bruce, I.J., Rahman, M.A., Hemre, G.I., The fate of transgenic sequences present in genetically modified plant products in fish feed, investigating the survival of GM soybean DNA fragments during feeding trials in Atlantic salmon, Salmo salar L. Aquaculture 237, Sargent, J.R., Tacon, A.G.J., Development of farmed fish: a nutritionally necessary alternative to meat. Proc. Nutr. Soc. 58, Sargent, J.R., Tocher, D.R., Bell, J.G., The lipids, in: Halver, J.E., Hardy, R.W. (Eds.), Fish Nutrition, 3 rd Edition. Academic Press, San Diego, pp Sayanova, O., Napier, J.A., Transgenic oilseed crops as an alternative to fish oils. Prostaglandins Leukot. Essent. Fatty Acids 85, Shepherd, C.J., Jackson, A.J., Global fishmeal and fish-oil supply: inputs, outputs and markets. J. Fish Biol. 83, Skåra, T., Sivertsik, M., Birkeland, S., Production of salmon oil from filleting by-products - effects of storage conditions on lipid peroxidation and content of n-3 polyunsaturated fatty acids. J. Food Sci. 69, E417-E421. Stottrup, J.G., McEvoy, L.A. (Eds.), Live Feeds in Marine Aquaculture. Blackwell Science, Oxford, 318 pp. Tacon, A.G.J., Metian, M., Global overview on the use of fishmeal and fish oil in industrially compounded aquafeeds: Trends and future prospects. Aquaculture 285, Tocher, D.R., Metabolism and functions of lipids and fatty acids in teleost fish. Rev. Fisheries Sci. 11, Tocher, D.R., Issues surrounding fish as a source of omega-3 long-chain polyunsaturated fatty acids. Lipid Technol. 21, Tocher, D.R., Fatty acid requirements in ontogeny of marine and freshwater fish. Aquacult. Res. 41, Tocher, D.R., Omega-3 long-chain polyunsaturated fatty acids and aquaculture in perspective. Aquaculture, in press. Tocher, D.R., Francis, D.S., Coupland, K., n-3 Polyunsaturated fatty acid-rich vegetable oils and blends, in: Turchini, G.M., Ng, W.-K.,Tocher, D.R. (Eds.), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp Torstensen, B.E., Tocher, D.R., The Effects of fish oil replacement on lipid metabolism of fish, in: Turchini, G.M., Ng, W.-K., Tocher, D.R. (Eds.), Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, pp Torstensen, B.E., Bell, J.G., Rosenlund, G., Henderson, R.J., Graff, I.E., Tocher, D.R., Lie, Ø., Sargent, J.R., Tailoring of Atlantic salmon (Salmo salar L.) flesh lipid composition and sensory quality by replacing fish oil with a vegetable oil blend. J. Agric. Food Chem. 53, Turchini, G.M., Francis, D.S., Keast, R.S.J., Sinclair, A.J., 2011b.Transforming salmonid aquaculture from a consumer to a producer of long chain omega-3 fatty acids. Fd. Chem. 124, Turchini, G.M., Ng, W.-K., Tocher, D.R. (Eds), 2011a. Fish Oil Replacement and Alternative Lipid Sources in Aquaculture Feeds. Taylor & Francis, CRC Press, Boca Raton, p

149 Turchini, G.M., Torstensen, B.E., Ng, W.K., Fish oil replacement in finfish nutrition. Rev. Aquacult. 1, Venegas-Caleron, M., Sayanova, O., Napier, J.A., An alternative to fish oils: Metabolic engineering of oil-seed crops to produce omega-3 long chain polyunsaturated fatty acids. Prog. Lipid Res. 49, VKM, Benefit-risk assessment of fish and fish products in the Norwegian diet an update. Scientific Opinion of the Scientific Steering Committee. VKM Report 15 [293 pp], ISBN: , Oslo, Norway. Available online: Wang, T.C., Fuller, M.F The optimum dietary amino acid pattern for growing pigs. 1. Experiments by amino acid deletion. Brit. J. Nutr. 62, Ward, O.P., Singh, A., Omega-3/6 fatty acids: Alternative sources of production. Process Biochem. 40, Xue, Z., Sharpe, P.L., Hong, S.P., Yadav, N.S., Xie, D., Short, D.R., Damude, H.G., Rupert, R.A., Seip, J.E., Wang, J., Pollak, D.W., Bostick, M.W., Bosak, M.D., Macool, D.J., Hollerbach, D.H., Zhang, H., Arcilla, D.M., Bledsoe, S.A., Croker, K., McCord, E.F., Tyreus, B.D., Jackson, E.N., Zhu, Q., Production of omega-3 eicosapentaenoic acid by metabolic engineering of Yarrowia lipolytica. Nat. Biotechnol. 31, Yazawa, K., Production of eicosapentaenoic acid from marine bacteria. Lipids 31, S297-S300. Zhu, Q., Xue, Z.X., Yadav, N., Damude, H., Pollak D.W., Rupert, R., Seip, J., Hollerbach, D., Macool, D., Zhang, H.X., Bledsoe, S., Short, D., Tyrues, B., Kinney, A., Picataggio, S., Metabolic engineering of an oleaginous yeast for the production of omega-3 fatty acids, in: Cohen, Z., Ratledge, C. (Eds.), Single Cell Oils: Microbial and Algal Oils. AOCS press, Urbana, pp

150 Tables and Figures Annex C 148

151 149

152 150

153 151

154 152

155 153

156 154

157 Fig. C1. Ingredients in Norwegian salmon feeds. Percentages of main ingredients used in feeds for Atlantic salmon in Norway in 2010 and 2013 along with the average nutrient compositions in 2010 and Grey bars, 2010; Black bars, 2012/2013. Figure prepared from data extracted from reports on resource utilisation of Norwegian salmon farming (Nofima, 2011, 2014)

158 EPA + DHA (g.100g -1 flesh) EPA + DHA (g.100g -1 flesh) EPA + DHA (Percentage) EPA + DHA (Percentage ) Fig. C2. Levels of EPA + DHA in in farmed (grey bars) and wild (black bars) salmon products obtained from major UK retailers. Panel A represents the relative proportions (percentage of total fatty acids) and Panel B the absolute contents (g / 100 g flesh) of EPA + DHA. Each bar represents a specific salmon product. Data are means ± SD (n = 2-4). Panel C shows the consolidated comparison of EPA + DHA levels in relative (percentage) and absolute (g/100 g flesh) terms. Data are means ± SD (n = 34 and 12 for farmed and wild products, respectively). Letters denote that the differences between wild and farmed were significant (P < 0.05). 40 A Farmed Wild Fish products B Farmed Wild Fish products C b a Farmed Wild a b Fish products 0 156

159 Charity Registration: SC Company Registration: SC267177