Introduction

Fish is critically important to food security and good nutrition (Allison 2011; Thilsted 2012; Beveridge et al. 2013). Fish and other aquatic foods are high in protein and contain many essential micronutrients. The fishery and aquaculture sectors are the source of income for millions of women and men in low-income families (Béné 2006), thus contributing directly and indirectly to their food security (Béné et al. 2007; Allison 2011; World Bank/FAO/WorldFish 2012). Yet, the potential contributions of fish to food security and nutrition (FSN) are all but ignored in the international debate, as if a firewall stood between the discussions about the role of fish and the broader debates about FSN issues. A recent review of international development and research agencies working on FSN revealed for instance that “fish is strikingly missing from strategies for reduction of micronutrient deficiency, precisely where it could potentially have the largest impact” (Allison et al. 2013: 45).

The purpose of this paper is to make a case for a closer integration of fish into the overall FSN debate. As such the discussion is directed not to those in the fisheries and aquaculture communities who have been relentlessly advocating for fish as an entry point in this debate (see e.g., Béné et al. 2007; Hall et al. 2011; Thilsted 2012), but to the vast majority of the international experts who are influencing the wider debate and policies on FSN.

The World Committee on Food Security (CFS) took an important step in the right direction in 2012 by requesting that the High Level Panel on Food Security (HLPE) undertook an in-depth study of the role of sustainable fisheries and aquaculture for FSN. The report (HLPE 2014) presents the strongest case yet for incorporating fish into the debates on how to achieve FSN for all and suggests a set of activities and policies to be pursued.

Building on this report and a large body of literature, our ambition is to bring fish onto the table. A number of papers have succeeded in presenting a very comprehensive evidence-based argument for supporting fish as a central element in the FSN debate (see e.g., Prein and Ahmed 2000; Roos et al. 2003; 2007; Kawarazuka and Béné 2011; Beveridge et al. 2013). Here, we aim to review the evidence from the contemporary and emerging debates and controversies around fisheries and aquaculture. In doing this we do not go into the micro-level issues concerning social, gendered and micro-economic aspects of fish, FSN at the local levels – for those we refer readers to the HLPE report (HLPE 2014).

Some of the underlying questions in the present study are closely linked to the issues discussed in the wider agriculture/farming literature. The latest estimate suggests for instance that, in 2009, fish accounted for 17 % of the global population’s intake of animal protein and 6.5 % of all protein consumed (FAO 2014). In the context of this Special Issue on “Feeding 9 Billion”, the key questions are: a) Can we maintain these consumption rates, given the projected growth in human population and the growing environmental challenges facing the earth? and b) Could fish - one of the most efficient converters of feed into high quality food – be a substitute for other sources of animal protein? Finally, in the context of the current discourse about future resource scarcity, should fish be considered a more environmentally friendly source of protein than the other livestock production systems?

Fish is more than just a source of animal protein. Fish contains several essential amino acids, especially lysine and methionine. The lipid composition of fish, with the presence of long-chain, poly-unsaturated fatty acids (LC-PUFAs), is unique. Fish is also an important source of essential micronutrients – vitamins D, A and B, and minerals (calcium, phosphorus, iodine, zinc, iron and selenium), which makes it particularly attractive in the current fight against malnutrition in low income and food deficient countries (LIFDCs). Some countries (e.g., Zambia, Brazil, or Chile) have already recognized this potential and have included fish in their national school-feeding programmes. Should this be more systematically considered in countries with high levels of malnutrition to complement or substitute the technology-dependant (and expensive) bio-fortification programmes?

In addressing these issues we must first acknowledge that fish production sectors and associated value chains have been remarkably dynamic in the last three decades – what some refer to as the ‘blue revolution’. Every second fish we consume is now produced in aquaculture. This new situation is a result of the stagnation of capture fisheries following decades of expansion, some of it carried out in an unsustainable manner (FAO 2012), combined with impressive growth rates in the aquaculture industry, subject to some major adjustments to overcome the early challenges such as fish disease and negative environmental impacts. Critics of aquaculture have also pointed to the use of fishmeal and fish oil produced from wild fish, to feed farmed fish. These concerns are valid and need to be taken (along with their solutions) into account in this discussion.

Understanding fish production in relation to food security and nutrition

The fastest growing food-supply industry in the world

In 2011, 173 Mt of fish were extracted from the global marine and inland water ecosystems, of which 7 to 10 Mt were discarded prior to landing and 12 Mt were lost at the post-harvest stage. The production from capture fisheries and aquaculture available amounted therefore to 154 Mt, of which about 131 Mt were utilized directly for human consumption (Fig. 1) (HLPE 2014). Thanks to aquaculture and fisheries, the global supply of fish has grown by a factor of 8 since 1950. By comparison, even after the Green Revolution the world rice production increased only by a factor of 3. In effect, fish production has been the fastest growing food industry in the world for the last 40 years and is expected to remain so in the near future. The fish supply per capita has more than tripled in the last half century, from 6 kg/year in 1950 to 18.8 kg/year in 2011 – Fig. 2. In fact, the world fish supply has effectively been growing faster than the world’s population (FAO 2012). These global figures mask, however, some important regional variations: fish consumption is the lowest in Africa (9.1 kg per capita in 2009), while Asia accounts for almost two-thirds of total consumption (20.7 kg per capita).

Fig. 1
figure 1

World fish utilization. Notes: a 2011 estimates. b 2010 estimates. Source data from FAO (2012) and Shepherd and Jackson (2013)

Fig. 2
figure 2

Relative contribution of aquaculture and capture fisheries to production and food fish supply. Source: FAO Statistics and Information Branch of the Fisheries and Aquaculture Department

Fish as the largest source of animal protein

Fish is a major source of animal protein, overshadowing most other sources. In 2010 it represented a source twice as important as poultry, and three times larger than cattle (Fig. 3). Today capture fisheries and aquaculture provide 3 billion people with almost 20 % of their average per capita intake of animal protein, and a further 1.3 billion people with about 15 % of their per capita intake (HLPE 2014). This share can exceed 50 % in some countries. In West Africa, Asian coastal countries, and many small island states, the proportion of total dietary protein from fish can reach 60 % or more (e.g., Gambia, Sierra Leone, Ghana, Cambodia, Bangladesh, Indonesia, Sri Lanka, or the Maldives) (FAO 2014).

Fig. 3
figure 3

World Production of the main sources of animal protein over the period 1960–2010. Source: FAO Stat

The geography of fish as a source of protein is also significant in the FSN discussion. Twenty-two of the 30 countries where fish contribute more than one-third of the total animal protein supply were officially referred to as LIFDCs in 2010 (Kawarazuka and Béné 2011). In other words, almost three-quarters of the countries where fish is an important source of animal protein are poor (income-wise) and food-deficient. Yet in these LIFDCs, even if fish is a substantial proportion of the food intake, undernourishment can still occur as total food intake is often insufficient.

Fish - beyond protein…

Some point out, however, that the main contribution of fish to FSN may not be in relation to its protein content, but its lipid and micro-nutrient content.

The lipid composition of fish is unique, having LC-PUFAs, with many beneficial effects for child development and adult health (Thilsted et al. 1997; Larsen et al. 2011; Richardson and Montgomery 2005). Among fish species that are cheaper in developing countries, small pelagic fish such as anchovy and sardine are some of the richest in LC-PUFAs (USDA 2011), especially compared to large freshwater fish such as carp and tilapia. When its rich nutrient content is preserved, fish can provide protective effects against a wide range of health issues. LC-PUFAs for instance provide protection against diseases such as stroke, high blood pressure or coronary heart disease (Miles and Calder 2012; Rangel-Huerta et al. 2012).

Complementing its fatty acid content, fish is also known to be an important source of essential micronutrients – vitamins D and B, and minerals (Roos et al. 2003; 2007; Bonham et al. 2009). Lipid-rich fish also contain vitamin A. Recent research showed that fish species consumed whole with bones, heads, and viscera play a critical role in micronutrient intakes of people as these parts are where most micronutrients are concentrated. Some of these small fish (such as mola (Amblypharyngodon mola), darkina (Esomus danricus), sardines and pilchards, anchovy, seabass, tilapia) contain high levels of minerals such as calcium, phosphorus, iodine, zinc, iron and selenium, which are low in other foods. The potential contribution that fish (even in small quantity) can therefore offer to address multiple micronutrient deficiencies is now being recognized (e.g., Roos et al. 2007; Kawarazuka and Béné 2011; Thilsted 2012). For instance, the high level of iodine found in fish can help prevent iodine deficiency which is known to cause cretinism (stunted growth and mental retardation).

Current issues in relation to fish contribution to food security

Fishmeal and fish oil in aquaculture

In addition to being used directly as human food, fish also contributes indirectly to human nutrition when it is used as fishmeal for aquaculture and poultry/livestock feeds (Tacon and Metian 2009). In 2011, 23 Mt of fish – essentially small pelagic fish species such as anchovy, herring, mackerel and sardine – have been destined to non-direct human consumption, of which 75 % (17 Mt) was reduced to fishmeal and fish oil for aquaculture, poultry and other livestock feeding (cf. Fig. 1). In 2010, 73 % of the total world fishmeal was used to feed farmed fish, followed by pigs (20 %), poultry (5 %) and others (2 %) (Shepherd and Jackson 2013).

From a FSN perspective, the use of fishmeal for farmed fish (and livestock) raises important issues. Leaving aside the debate on the role of small pelagic fish in supporting larger fish, birds and marine mammals in the ecosystem (Smith et al. 2010), is fishmeal the most efficient way to use fish (especially low-cost small pelagic fish rich in LC-PUFA) or would these fish contribute more to food security if a larger share of them was eaten directly by people? Indeed, despite some substantial improvement in the last decade, the rate of conversion of fishmeal to fish is still a source of concern (Troell et al. 2014). On average, for every kg of farmed fish produced, 0.7 kg of wild fish is needed (Tacon and Metian 2008). This average figure, however, masks important differences: while for omnivorous farmed fish, the rate is down to an acceptable 0.2 to 1.41 kg of wild fish per 1.0 kg of farmed fish, for carnivorous farmed fish, the figure is higher: 1.35 to 5.16 kg to produce one kg of farmed fish (Boyd et al. 2007).

Fish losses and implications on food security and nutrition

The global discards of fish (fish caught but dumped overboard due to low quality, damage or spoilage, non-targeted species or below regulation size) were estimated to be around 7.3 Mt in 2005, 80 % of which coming from industrial fleets. In contrast, small-scale fisheries generate less wastage in the form of discards (about 2 Mt a year, that is, 4 % of their estimated landingFootnote 1) (Kelleher 2005).

With increasing fish scarcity and increasing fish prices, species previously deemed commercially inferior are progressively integrated into consumer eating habits and markets. Most shrimp trawlers, which used to discard up to 95 % of their catch, are now landing more bycatch for human consumption (Béné et al. 2007). In small-scale fisheries where discarding fish is rare, substantial quantities and quality are lost due to post-harvest mishandling during transport, storage, processing, on the way to markets and waiting to be sold. Especially in developing countries, where access to electricity and cold chain can still be an issue, fish post-harvest losses can be significant. Estimated at 10–12 Mt in 2005, the total accounts for 10 % of global capture and culture fisheries’ production (Béné et al. 2007). While physical losses in small-scale fisheries are less than 5 %, economic losses can be substantial. In Africa, some estimates (FAO Focus nd) put post-harvest losses in some cases to levels as high as 20–25 %. In aquaculture, waste streams in value chains have been large in the past but are tending to decline rapidly as competitive pressures force innovation (Arthur et al. 2013).

Sustainability of fisheries and implications for food security

Because total fish production (availability) is an important dimension of food security, a key issue in this debate is the environmental sustainability of capture fisheries. The extent to which capture fisheries have exceeded sustainable levels has generated strong expert and public opinion debates (Worm et al. 2006; Pauly et al. 2005; Hilborn 2013) and many media headlines, scientific papers and environmental campaigns have been in the last two decades framed around the idea that world fisheries resources are in crisis due to overfishing (see e.g., Pauly et al. 1998; Myers and Worm 2003). FAO has expressed a more nuanced but nonetheless concerned view about the state of world marine fisheries (e.g., FAO 2012), acknowledging the granularity in the state of resources worldwide. The current consensus is that global fisheries would be more productive if the levels of overfishing were reduced (Srinivasan et al. 2010), and the environmental sustainability of fisheries were recognized to be a sine qua none condition for FSN (HLPE 2014).

An important point in this debate, however, is that improving FSN through fisheries would depend on stock recovery and also on access to and distribution of the harvest, as well as gender consideration (de Schutter 2012; Williams 2010). Indeed overfishing per se is only one aspect of the problem. Other economic activities, such as oil drilling, coastal development, pollution, or dams and water flow management have significant negative impact on aquatic habitats (Halpern et al. 2008).

Indirect contribution to food security and nutrition through livelihood support

A critical pathway to enhance FSN is through the income that people generate from engaging in remunerated activities. In this respect aquaculture and fisheries (especially through the number of small-scale operators engaged in fishing, aquaculture, processing, and trading businesses – see e.g., Béné et al. 2010) play a critical role in low income and emergent countries. Altogether it is estimated that between 660 and 820 million people (fishers, fish-farmers, fish traders, workers in fish processing factories, and their families) depend on fish-related activities as a source of income (Allison et al. 2013; HLPE 2014). This represents more than 10 % of the world population. For most of these households, the revenues generated may not necessarily be very high (Neiland and Béné 2004; Allison et al. 2011; Béné and Friend 2011), but it is often the main component of their livelihood, which allows them to secure accessibility to food (Heck et al. 2007; Béné et al. 2009; Eide et al. 2011).

Feeding 9 billion by 2050: where does fish stand?

The global human population is expected to exceed 9 billion by 2050 (UN estimates), increasing the pressure on the food sectors to maximize production and reduce waste. Production increase must occur in a sustainable way and in a context where key resources, such as land and water, are likely to be scarcer and where climatic change impact will intensify. The fish-production sector is no exception.

In this context two key questions emerge. First, will fisheries and aquaculture be able to maintain the current global fish consumption rate of 18 kg per capita per year, and the equivalent regional values, and if not, how will society address the needs of expected winners and losers (Barange et al. 2014)? So far technological and institutional innovations have ensured that the combined production of fish through fisheries and aquaculture has been faster than the world population’s demand of fish. The question is now whether we can keep up this pace with another 2 billion people added in the next 35 years, and how the four dimensions of food security (availability, accessibility/affordability, utilization and stability) will be balanced to ensure that fish go to those who need them most (Merino et al. 2012).

The second key question is whether sustainable fisheries and aquaculture will be able to help address the bigger food security issue that will affect the world in the coming decades? In particular, could aquaculture become a substitute form of protein for some of the less efficient food production systems, or even be used to compensate for the decline in farming systems’ productivity that is predicted as a consequence of the impact of climate change?

How much fish do we need?

Global drivers of fish demand

World population is often presented as a key driver for the growth in seafood demand and for fisheries development. In reality, a more important driver for fish (and other animal-source food) consumption is income (Speedy 2003). Demand for fish as food is particularly high in the wealthier strata of societies, including in the low-income countries, and as income will continue to increase in highly populated countries such as China and India, demand levels are likely to rise more strongly (Garcia and Rosenberg 2010). Overall, a large increase in the number of people moving into the middle class, particularly but not exclusively in Asia, is likely to result in a very large expansion in the demand for fish.

Income is, however, not the only driver of fish demand. It is recognized that urbanization is also an important factor increasing animal-source food consumption in general and fish consumption in particular. Delgado et al. (1997) suggest that changes in food preference driven by urbanization alone have, in the past, accounted for an extra 5.7–9.3 kg per capita consumption of meat and fish per year. Similarly, Betru and Kawashima (2009) and Toufique and Belton (2014) present data from Ethiopia and Bangladesh, indicating that urbanization affects animal food consumption rates independently of income.

These different factors explain the rapid increase in demand for meat, milk, and fish in the emerging economies of Asia. In China, for example, the demand for fish is likely to increase from 24.4 kg per person per year in 2000 to 41 kg per person per year by 2030 (World Bank 2014).

Demand projection

Modelling exercises have been conducted recently with the objective of estimating the projections of fish demand and supply. These modelling exercises include the World Bank-FAO-IFPRI Fish 2030 analysis elaborated on the IMPACT model developed by the International Food Policy Research Institute (IFPRI) (World Bank 2014); the OECD-FAO Agricultural outlook model built on the combined multimarket, partial equilibrium AgLink-CoSiMo model (OECD-FAO 2013; Lem et al. 2014); and a series of peer-reviewed articles relying on various types of modelling and projection tools (Rice and Garcia 2011; Merino et al. 2010, 2012; Barange et al. 2014). The time horizon of these different analyses is not always the same. For instance, the OECD-FAO outlook model works over a 10 year-projection period (i.e., up to 2023 as per the last iteration of the model), while the World Bank-FAO-IFPRI model runs until 2030; Merino et al. 2010 used a 20 year simulation, calibrated on a 1997–2004 data set (meaning technically that their projection runs until 2024), while Merino et al. (2012) and Barange et al. (2014) proposed a projection up to 2050. A strict comparison of the different projections is therefore difficult.

A bigger issue is that very few of these studies integrated information on the drivers of changes (the combination of urbanization and increase in income), to estimate with accuracy the future demand for fish across the world. Many studies assume constant consumption rates in the future (e.g., Barange et al. 2014) or fixed nutritional targets (Rice and Garcia 2011). Others worked directly with projected fish consumption (e.g., OECD-FAO 2013; Merino et al. 2012), that is, by dividing the projected supply by the projected population. None of these approaches therefore offers an appropriate basis for estimating the actual demand for fish. The World Bank-FAO-IFPRI study uses regional fish consumption rates to estimate the global demand but the report does not clarify how these figures are constructed.Footnote 2 Aggregating the regional figures at the global level, the report estimates that the world demand for fish will be around 152 Mt in 2030 (World Bank 2014).

Merino et al. (2012) estimated the expected production capacity of marine ecosystems exploited under maximum sustainable yield principles, and projected aquaculture production requirements to achieve a range of food consumption targets. They concluded that between 125 and 210 Mt of fish by 2050 will be necessary to maintain fish consumption at around 15–20 kg per capita per year. Starting from a different angle, Rice and Garcia (2011) sought to estimate the need for additional fish necessary to supply 20 % of the dietary protein requirement to feed a 9 billion population by 2050. On this basis, an additional production of 75 Mt of fish from fisheries and aquaculture would be needed above the 2006 production level (144 Mt), that is, approximately 215 Mt. This represents an almost 50 % increase in production with respect to the 2006 level. While this figure is above other projections, it has the advantage that it starts from an actual estimate of the future needs (as opposed to the estimate of apparent consumption rates), but without factoring in future production potentials or market responses to the fishmeal/fish oil demand (see below).

In summary, the current understanding of the global drivers of fish demand (urbanization and increase in living standards in developing and emerged countries) is relatively well established but not all of the current models have integrated these drivers comprehensively. In comparison, the efforts to better understand the ability of the world to produce fish in the future (i.e., the supply side of the equation) have been more elaborate.

The future of fisheries production and the impact of climate change

A consensus has emerged in the literature that the doom-and-gloom rhetoric that had driven the discussion surrounding the state of marine fisheries in the late 2000s (Garcia and Grainger 2005; Caddy and Seijo 2005) was exaggerated (Grafton et al. 2010; Hilborn 2010) and that although the situation remains concerning in respect to many stocks, we are not likely to face the global collapse that had been announced by some biologists (e.g., Myers and Worm 2003; Worm et al. 2006; Pauly 2009). Instead, the downward trend of overfished stocks may have been reined in (Fig. 4).Footnote 3 Reflecting this, most of the projections proposed in the recent literature estimate that the global fisheries’ landings are likely to be stable in the short to medium term. The OECD-FAO model for instance estimates that capture fisheries will be 5 % higher by 2024 than it is was in 2013, that is, around 96 Mt (OECD-FAO 2013) while the World Bank-FAO-IFPRI model estimates that this will be around 93 Mt in 2030. These figures are at a global scale however, and some regional outlooks are for good stock rebuilding, while others are for a worsening in overfishing.

Fig. 4
figure 4

Global trends in the state of world marine fish stocks (1974–2011). Source: FAO (2012)

Yet another key factor for which much uncertainty remains is the impact of climate change. Unlike most terrestrial animals, aquatic animal species are poikilothermic (cold-blooded) and changes in aquatic habitat temperatures will more rapidly and significantly influence distribution, prey availability, metabolism, growth and reproduction, with stronger impact on fishing and aquaculture distribution and productivity (Cheung et al. 2009). At the same time however, the interconnectedness of aquatic systems allows many species to change spatial distribution more easily as ecosystems shift, to remain in their zones of preference. Clearly, therefore, the impact of global climate change on ocean capture fisheries will be important. Biological predictions based on ocean-atmosphere general circulation models (OA-GCMs) have already demonstrated that the physical and chemical properties of the oceans will be modified, affecting the productivity, distribution, seasonality and efficiency of food webs, from primary producers to fish (Steinacher et al. 2010; Cheung et al. 2009, 2011). Some of these earlier GCM models however, were limited by their coarse resolution, too low to capture the processes that dominate the dynamics of the world’s coastal and shelf regions, where most fisheries operate. More advanced models are now available (e.g., Merino et al. 2012; Blanchard et al. 2012; Barange et al. 2014).

Overall, and with few exceptions, the conclusion of all these models is that although climate change will alter the present geographical distribution of shelf-sea ecosystems productivity, in most of the regions and EEZs, the overall potential impact is projected to be low to moderate. Barange et al. (2014), for instance, used a high resolution shelf-sea physical-biological model that allowed them to scale down the analysis and gave greater confidence in predicting the consequences at national scales. They conclude that by 2050, estimates of national fish production should remain on average within ±10 % of the present yields.

The contribution of aquaculture to future fish supply

The second element on the supply side is aquaculture. Discussion of the rise of aquaculture has so far largely focused on its contribution to global aquatic animal food supplies, ignoring the resultant changes in species composition of the fishes consumed, how it is farmed, and the implications for food and nutrition security (Kawarazuka and Béné 2011; UNHRC 2012). As a consequence, our understanding mainly concerns the question of the ability of the aquaculture industry to maintain its rate of growth. In this regard, most of the recent analyses agree that the era of exponential growth is over and while the sector will still continue to grow, the projected rate of growth is expected to decelerate. The main causes of this slower growth are likely to be freshwater scarcity, lower availability of locations for optimal production, and high costs of fishmeal, fish oil and other feeds (FAO 2012). Nonetheless, the World Bank-FAO-IFPRI model suggests this rate will still be sufficient to maintain a steady rise, reaching the point where it will equal global fisheries production by 2030 – around 93 Mt (World Bank 2014) (Fig. 5). Technical innovation, improved farm and animal health management, and improved and more efficient germplasm will be responsible for this increased growth, in combination with the continued expansion of fish-farming. Combined with a projected capture fish production that will remain fairly stable over the 2000–2030 period (see above), the global fish supply is projected to rise to 187 Mt by 2030 (World Bank 2014). These figures are consistent with the projections proposed by OECD-FAO in which the global fish production will reach 181 Mt in 2022, of which 161 Mt would be destined for direct human consumption (OECD-FAO 2013).

Fig. 5
figure 5

Global fish production: data and projection 1984–2030 from the IMPACT model (World Bank 2014). Data source: FAO FishStat and IMPACT projection model

A key element in this discussion is the importance of fishmeal and fish oil and how markets and technological innovations will respond to price signals (Merino et al. 2010, 2012; FAO 2012; World Bank 2014; Troell et al. 2014).Footnote 4 The consensus is that the use of fishmeal in aquaculture feeds is expected to decrease in percentage with time, thanks to increasingly effective replacements, including plant proteins, waste products from fish and terrestrial animals and use of better/improved breeds of aquatic animals with better feed conversion (Tacon et al. 2011). Formulated feeds are a significant factor in production costs, and this is a strong incentive to develop technology that will make feeds more affordable and sustainable. Overall, the proportion of fish used for non-direct human consumption has decreased from 30 % in the early 1990s to 15 % in 2010 and the World Bank analysis concluded that the projected expansion of aquaculture will be achieved with only an 8 % increase in the global fishmeal supply during the 2010–2030 period (World Bank 2014).

Putting all the pieces together

Given the very rough projections for fish demand and the more elaborate projections for fish supply just reviewed above, will fisheries and aquaculture be able to maintain their current contribution to food security in the future? The answer that emerges from the literature is that as far as food availability and demands are concerned, it is a conditional yes. All the projection models currently available seem to agree that the overall fish consumption rate could be maintained, in other words, that the fisheries and aquaculture aggregated growth will keep up with population growth rates. In fact, the OECD-FAO even estimates that the world fish consumption will increase by another 10 % and reach 20.9 kg per capita per year by 2023. The World Bank-FAO-IFPRI report is slightly more conservative and estimates that the per capita fish consumption will remain at around 18 kg per year in 2030. Merino et al. (2012) and Barange et al. (2014) reached the same conclusion for 2050, but this is essentially due to the underlying assumptions of their models.

All these analyses, however, stress the fact that for this outcome to occur, several specific conditions must be satisfied: capture fisheries will need to be exploited according to sustainable principles; very significant technological development will need to take place in aquaculture feeds to reduce fishmeal dependency and in farm management and germplasm to improve the overall efficiency of aquaculture; and discards, waste and losses will need to be reduced. Some of these conditions could be particularly challenging and a ‘business-as-usual’ approach is expected to fail.Footnote 5

In addition to whether or not the world (and the markets) will be responsible enough to ensure that these conditions are satisfied, the question of who will be the winners and losers is vital to the other key FSN issues, which are: access to livelihoods in fish value chains and affordability of fish. Unfortunately, the consensus is that in the coming decades, the current situation regarding the imbalance between consumers in developed and developing countries is unlikely to change. While the present figures indicate the lowest fish consumption per capita in Africa (9.1 kg in 2009), the different models project that this imbalance will deteriorate further. The OECD expects that per capita consumption in Africa will decrease by a further 10 % by 2024, while that of Asia will show the highest growth rate (+14 %, OECD-FAO 2013). The World Bank-FAO-IFPRI projection (World Bank 2014) is even more alarming with per capita fish consumption expected to decline in sub-Saharan Africa by 1 % annually to 5.6 kg in 2030.Footnote 6 Yet at the same time developing countries will account for more than 91 % of the total increase in fish consumption. Even so, their annual per capita fish consumption will still remain below that of more developed regions (19.8 vs. 24.2 kg) (World Bank 2014).

A last important point needs to be mentioned. In view of the importance of income growth and urbanization as drivers of fish demand, the objective of merely maintaining fish production growth rate at the same level as the growth in world population would not be enough to prevent fish price from increasing. The urban population with the highest income growth will increase fish consumption while low-income people will experience reductions in their fish consumption. This is an important distributional and food security consequence. To avoid that, fish production needs to expand faster than population growth.

Can fish contribute further to the 2050 global food security challenge?

The arguments presented above are mainly constructed around a sectorial perspective. At least three other trans-sectorial (or systemic) arguments should be considered in relation to the wider debate of ‘feeding 9 billion by 2050’.

First, in terms of animal protein availability, with 18.2 kg per capita per year, fish is providing 115, 133, and 189 % more protein per capita than pig, poultry and beef respectively. In fact, fish (combining capture fisheries and aquaculture) has been the main contributor to the 61 % increase in the world per capita consumption of animal protein for the period 1969–2009 (Table 1). As economic development is expected to continue driving an increasing trend in animal protein demand, (OECD-FAO 2013) and aquaculture is projected to remain the fastest growing food commodity sector, this sector will soon become even more central in the future food security of the world population.

Table 1 World per capita meat and fish food supply (kg per capita per year) 1969–2009

Second, in term of efficiency, fish in aquaculture systems are very efficient converters of feed into protein – in fact far more efficient than most terrestrial livestock system (Fig. 6a). For instance, poultry converts about 18 % of their consumed food and pigs about 13 %, as compared with 30 % in the case of fish (Hasan and Halwart 2009). Production of 1 kg of beef protein requires 61.1 kg of grain, production of 1 kg of pork protein requires 38 kg of grain, while fish only requires 13.5 kg (Hall et al. 2011). Most of this difference comes from two biological characteristics of fish which give them great advantages over land-based livestock in growth performance: (i) the fact that fish are poikilotherms and therefore do not expend energy maintaining a constant body temperature; and (ii) the fact that, because finfish are physically supported by the aquatic medium, fewer resources are used on bony skeletal tissues, and a larger part of the food they eat is effectively allocated to body growth.

Fig. 6
figure 6

a Feed and protein conversion efficiency of beef, pork and fish; b Nitrogen and phosphorous emissions for animal production systems. Source: Data for fish are derived from Hall et al. (2011). Data for beef, pork and chicken are derived from Flachowsky (2002) in Poštrk (2003)

Third, in terms of carbon footprint, aquatic animal production systems have a lower carbon footprint per kilogram of output compared with other terrestrial animal production systems (Hall et al. 2011). As a consequence, nitrogen and phosphorous emissions (kg of nitrogen and phosphorus produced per tonne of protein produced) from aquaculture systems are much lower than those in beef and pork production systems and slightly higher than that of poultry (Fig. 6b). In fact, some aquaculture production systems such as farming of bivalves absorb nitrogen and phosphorous emissions from other systems.

All these reasons are important arguments for giving fish far greater attention in the food security debate and in the current discussion about how to feed 9 billion by 2050. The CFS has started to recognize this and the report that was commissioned in 2012 indeed recommends that fish (a) becomes an integral element in inter-sectoral national FSN policies and programmes and (b) should be more systematically included in countries’ nutritional programmes and interventions aiming at tackling micronutrient deficiencies especially among children and women (HLPE 2014: 18 Recommendation 1a and 1b).

Conclusion

Fish is already making a major contribution to human food supply and to the support of FSN for more than 660 million fish-workers and their families. It also provides more than 4.5 billion consumers with at least 15 % of their average per capita intake of animal protein. In addition, because fish is more nutritious than staple foods such as cereals, providing in particular essential fatty acids and micronutrients, it can play an extremely important role in improving the nutritional status of individuals, in particular those at risk such as children and women.

Yet limited attention has been given so far to fish as a key element in FSN strategies at national level and in wider development discussions and interventions. Part of the problem might have been that specialists in fisheries debates have concentrated predominantly on questions of biological sustainability and on the economic efficiency of fisheries, neglecting issues linked to its contribution to reducing hunger and malnutrition and to supporting livelihoods (Kawarazuka and Béné 2010, 2011). On the other end of the spectrum, and with too few exceptions, most (non-fishery) food security experts and decision-makers seem unfamiliar with these facts and, therefore, unaware of the critical role that aquaculture is likely to play in the future. The problem is particularly pronounced in the current debate about how to make the food system more nutrition sensitive, i.e., how to change and improve the food systems in order to advance nutrition. As a consequence, fish has so far been only marginally included in the international FSN debate. Many nutritional programmes are still not aware of, or not recognizing and building on the potential of fish for the reduction of micronutrient deficiency.

In this paper, we make the case that fish deserves more attention in food policies due to its importance in the food basket, its unique nutritional properties, its higher efficiency of production and carbon footprint compared to other forms of animal production systems. We acknowledge some challenges, however, especially in making fish more affordable for the poor or in maintaining – or restoring – the environmental sustainability of the sectors. We also stress that the ability to meet the potential average fish demands of 9 billion people masks inequalities and inequities in who eats the fish and who benefits from the value chains. At present, people in Africa, the poor in many societies, and women and minority groups, including small-scale fishing and aquaculture communities are in tension with large corporations and production units over access to fish and fish-related employment (HLPE 2014). We have shown that the best available projections for fish supply and demand are relatively positive in terms of the capacity to meet future demands, although more solid modelling is still needed to better incorporate demand projections. Climate change impacts on fish production also create uncertainty in the projections.

In conclusion, fish should certainly be on the menu.