An agenda for integrated system-wide interdisciplinary agri-food research

This paper outlines the development of an integrated interdisciplinary approach to agri-food research, designed to address the ‘grand challenge’ of global food security. Rather than meeting this challenge by working in separate domains or via single-disciplinary perspectives, we chart the development of a system-wide approach to the food supply chain. In this approach, social and environmental questions are simultaneously addressed. Firstly, we provide a holistic model of the agri-food system, which depicts the processes involved, the principal inputs and outputs, the actors and the external influences, emphasising the system’s interactions, feedbacks and complexities. Secondly, we show how this model necessitates a research programme that includes the study of land-use, crop production and protection, food processing, storage and distribution, retailing and consumption, nutrition and public health. Acknowledging the methodological and epistemological challenges involved in developing this approach, we propose two specific ways forward. Firstly, we propose a method for analysing and modelling agri-food systems in their totality, which enables the complexity to be reduced to essential components of the whole system to allow tractable quantitative analysis using LCA and related methods. This initial analysis allows for more detailed quantification of total system resource efficiency, environmental impact and waste. Secondly, we propose a method to analyse the ethical, legal and political tensions that characterise such systems via the use of deliberative fora. We conclude by proposing an agenda for agri-food research which combines these two approaches into a rational programme for identifying, testing and implementing the new agri-technologies and agri-food policies, advocating the critical application of nexus thinking to meet the global food security challenge.


1
Conventionally defined as when 'all people, at all times, have physical, social and economic access to sufficient, 2 safe and nutritious food that meets their dietary needs and food preferences for an active and healthy life' (FAO 3 2002), food security is generally acknowledged to be one of the 'grand challenges' currently facing humanity.

4
The challenge is neatly summarised as a 'perfect storm' of converging global issues (Beddington 2010) as the 5 world's population is set to reach 9.6 billion by 2050 (UN 2013) with a quadrupling in the global economy, a 6 doubling in the demand for food and fuel, and a more than 50% increase in the demand for clean water 7 (Foresight 2011). This challenge is amplified by the need to stay within the safe operating space for humanity 8 and avoid catastrophic climate change (Rockström et al. 2009). The 5 th IPCC report (IPCC 2014) notes the 9 weight of studies that predict a decline in agricultural production by 2050 due to climate change impacts and 10 summarises the substantial risk evidence that Europe, Africa, Asia and Central and South America will 11 experience water shortages driven by changing climate, leading to declining agricultural production and 12 increased rural poverty during the coming few decades.

13
We acknowledge the long track-record of work establishing the links between food security and global

20
'scientific and policy attention has … mainly focused on increasing total production through increases in yield 21 [which] arguably risks ignoring people's anxieties about sustaining access to food … and the other nutritional,

22
social and economic aspects of food security' (Ingram et al. 2013). Thus, we conclude that achieving adequate 23 food production whilst ensuring environmental and economic sustainability and promoting human health and 24 social equity will require changes in all parts of the food system.

25
Following the work of Soussana (2014) and a recent comprehensive report from the US National 26 Academies (IOM and NRC 2015), this paper charts the development of an integrated approach to agri-food 27 research, working across the food supply chain rather than isolated researchers working on separate parts of the 28 problem. It demonstrates the need for interdisciplinary research that addresses the operation of both 29 environmental and social systems (and their effective integration). While many others are working on these 30 challenges, including the governance and management issues that arise when working across scales (Cash et al. 31 2006), this paper outlines an interdisciplinary and system-wide approach that seeks to overcome many of the 32 key methodological and epistemological challenges faced by existing agri-food research. 1 In doing so, this 33 paper also locates a number of initial successes in implementing this approach as well as offering insights about 34 how a system-wide agenda could be moved forward.

35
A system-wide approach to agri-food research enables questions of the following type to be answered: 36 what might be the effect of a change in a particular consumer habit on crop production, resource use, nutrition 37 1 Framing our argument in terms of the 'agri-food' system should not be taken to imply an undue emphasis on terrestrial cropping systems. We also acknowledge the importance of livestock farming and fisheries, using 'agri-food' as a short-hand for the broader food system. and health? What would be the implications for the food producer, retailer and consumer of a change to more 1 sustainable and resilient crop production, through a new plant variety or agronomic practice? What are the 2 implications for the food security of farmers in poorer countries of changes to markets, consumption and trade 3 across global production networks and value chains? How can changes to land tenure, input pricing, credit, 4 financing and sales improve the food security of the poorest farmers internationally? How can food waste be 5 reduced to ensure the most efficient functioning of the agri-food system? Where are the pressure points or sites 6 of greatest sensitivity to change? Where are the 'hotspots' in terms of resource use, environmental effects or 7 waste? How do we adapt agri-food systems to climate change? How do we present the different solutions

13
Developing a more interdisciplinary and system-wide approach would involve five steps: 1, describing the agri-14 food ecosystem; 2, identifying the research themes that emerge; 3, defining a quantitative methodology for 15 analysing and modelling agri-food ecosystems and thereby integrating these research themes; 4, establishing a 16 complementary methodology to address the political, ethical and legal tensions within the ecosystem; and 5,

17
setting out an agenda for agri-food research that exploits the ecosystem concept to develop innovative ways to 18 combine these two approaches into an analytical framework for determining, evaluating and implementing new 19 agri-food policies and technologies. The remainder of this paper outlines this approach in more detail,

23
The first step in developing a system-wide approach to agri-food research is to describe what the system is, what 24 processes to include and where to set boundaries. From first principles, the agri-food system comprises all of the 25 processes involved in producing and consuming food from the capture of sunlight by photosynthesis in plants, 26 harnessing the ecosystem services provided by the agricultural landscape that are central to food production, 27 through the conversion of plants and animal feed into human food, to the purchase, preparation, consumption 28 and metabolism of foodstuffs by humans. Our increasingly globalized agri-food system is characterised by a 29 growing separation between production and consumption with a range of corporations and institutions playing 30 an increasingly important intermediary role.

31
Previous attempts to describe the complete system of agricultural production have included the idea of 32 the 'agro-ecosystem' (Conway 1987). Under this model, after establishing a suitable ecosystem boundary, all of 33 the processes and participants in crop production were defined, allowing material flows, interactions, inputs and 34 outputs to be described and analysed. This model was found to be suitable for describing the whole agri-food 35 system and in previous work we expanded the range of processes and stakeholders to create an agri-food  improving resource efficiency and sustainability in food supply chains. This model went through a large number 38 of modifications arising from its exposure to multidisciplinary experts including university academics and 39 leaders from research funding bodies and industry. The updated model is outlined in Figure 1: Figure 1A shows 40 the actors involved, the external influences, and more detail of the inputs and outputs involved in food 1 production and consumption; and Figure 1B shows the sources of loss and waste, the environmental and health 2 penalties than can ensue and the environmental and socioeconomic benefits of the agri-food system. The 3 unifying definition of waste across the entire system should be noted in Figure 1A, which includes inefficiencies 4 at the farm level as well excess eating as a part of such waste (Horton et al. 2016). The contemporary agri-food 5 system is subject to many external influences including the actions of NGOs and pressure groups, innovations in 6 science and technology, labour unrest and geopolitical events, together with natural hazards such as flooding 7 and drought, which can have a significant impact on the resilience of agri-food systems as was demonstrated by 8 the 2007-8 'price shock' (Mittal, 2009).

9
This conceptualization of the agri-food system seeks to integrate: agricultural and land-use strategy; 10 crop production and harvesting; corporate and farmers' means for managing labour, credit, technology and 11 sales; food processing, storage and distribution; retailing; and purchasing, preparation and consumption. It 12 demonstrates how losses and waste occur at all points in the system, illustrating the environmental impacts of 13 food production and consumption and highlighting the human consequences of the agri-food system in terms of 14 the health-related outcomes of dietary decisions (often highly constrained by socio-economic circumstances).

15
The model is presented in linear terms but, in practice, agri-food systems are usually complex networks  (indicated by dotted arrows). Figure 1B includes the important feedback from environmental impacts, which can 20 lead to further losses in crop yield, increase in food waste and amplification of health effects (dotted arrows).

21
We also show that the agri-food system has numerous other outputs besides food for human consumption,

22
including food waste, animal waste, non-food biomass and human sewage. The importance of representing 23 them in this way is that they can be viewed as a resource which can be utilised and even fed back into the 24 system (dotted arrows). Thus, waste can be converted to energy via anaerobic digestion or processed to recover 25 valuable resources, such as fertiliser (Li et al. 2015).

26
The ecosystem model in Figure 1 is genericit can be used to describe any agri-food system, in any    in each of these five domains and it is clear that, due to the highly interconnected food supply system, the 6 answers to many of these questions depend on understanding events and processes taking place in other 7 domains. Asking questions within the framework proposed in Figure 1 also has a transformative impact on the 8 framing of questions within each domain as we now seek to illustrate.

9
In Land Use and Resource Management research a principal objective is to understand the pressure 10 on global land and soil from the demographic drivers of increasing human population and wealth as well as 11 related pressures on other resources such as water. Providing space for building puts pressure on the land 12 available for agriculture, and both squeeze out land needed to maintain habitats and biodiversity (Blum 2006

37
(http://www.bbsrc.ac.uk/funding/filter/food-system-resilience/, accessed 9 December 2016). The N8 agri-food programme has a £8m budget from the HEFCE Catalyst fund (with matched funding from the eight partner universities), organised in three research strands on sustainable food production, resilient supply chains and improved consumption and health (http://n8agrifood.ac.uk/, accessed 9 December 2016). management strategies that restore soil ecosystem function (Cameron et al. 2013). An important element of this 1 research is the collaboration between scientists and farmers, deploying scientific knowledge about soil 2 conservation in farming practices (MacMillan & Benton 2014).

3
In poorer parts of the world food security of smaller farmers reflects not just lack of land, but lack of 4 access to credit, farm inputs such as fertilisers and adequate labour. These can be intensified by their occurring 5 at key times of the year in crop production cycles. Therefore research needs to explore how small-scale farmers 6 manage labour, credit and social networks to improve farm productivity, as well as examining how they    Informed by an integrated agri-food perspective, research on Food Processing, Distribution and Sales 8 has two aspects. In wealthier countries, the effects of retail concentration and the increasing complexity of food 9 businesses and their lengthening supply chains are key priorities. In poorer countries many of these also apply,

27
An integrated approach to agri-food systems demonstrates how research on Food Consumption should 28 seek to connect the behaviour of consumers, as individuals and groups, to the systems of provision that make 29 food available to them and to explore the consequences of their (often highly constrained) food choices in terms The Food Standards Agency's recent summit on Our Food Future (February 2016) highlighted a link between convenience and connection where it was argued that an increasing reliance on processed food led to a growing sense of disconnection between food producers and consumers (https://www.food.gov.uk/sites/default/files/our-food-future-full-report.pdf, 9 December 2016). 5 The discovery of horsemeat in processed beef products sold by a number of UK supermarket firms drew media attention to the length and complexity of food supply chains (http://www.bbc.co.uk/news/uk-21335872, accessed 9 December 2016).
initiatives, based on the assumption that increased consumer knowledge will lead to desirable changes in 1 attitudes and behaviour. 6 But, as the Foresight report on 'Tackling Obesity' recognised, 'policies aimed solely 2 at individuals will be inadequate', emphasising the need for 'wider cultural changes' involving coordinated

20
The above discussion clearly shows that sustainable food security solutions will depend upon knowledge that 21 drives a step-change in innovation that spreads throughout agri-food systems. To achieve this goal requires a

2
This approach will only succeed if there are equally high levels of input from all the parts of the agri-

21
In order to achieve a truly integrated analysis of agri-food systems, a method of quantitative analysis and a 22 modelling tool as described in the previous section is necessary but insufficient. Understanding the ethical, 23 legal and political issues that shape agri-food systems is also required. Integrating insights from the political and 24 social sciences into agri-food research is crucial because food security will require more than the examination of 25 food production and consumption from a purely scientific or technological point of view. This is because 26 questions regarding the distribution of the 'goods' associated with food systems involve inherently political 27 decisions necessitating research on complex decision-making processes. Understanding the inherently political 28 dimensions of the agri-food system is also required because various aspects associated with food security,

38
An integrated approach to agri-food research is ultimately concerned with justice, since theories of 39 social justice offer us first principles by which to determine 'who gets what and why' in any socio-economic or political system (Allen 2008;Clapp 2012). Research on global food security must also address larger ethical 1 and practical questions about substantive and procedural justice (both domestically and globally) and a resulting 2 just distribution of food system-related benefits and burdens.

3
As Figure 1 illustrates, every level of the agri-food system is subject to political influence. This is true 4 in terms of agricultural regulation, public health policy, environmental standards, food waste programmes and 5 policy incentives. It also applies to the political-economic dimensions of food security including the capitalist 6 structures that govern global food production and distribution (Morgan et al. 2006). Decisions about how to 7 respond to food security concerns will have considerable moral/ethical implications. Such ethical considerations 8 must be taken into account within any heuristically viable approach to agri-food research. Long-term, 9 politically legitimate solutions will necessarily involve better understandings of existing food-related political 10 structures, processes and alternatives.

17
Section 4 then established that understanding how to devise and deliver sustainable agri-food systems is wholly 18 dependent on resolving the competing political and ethical influences upon it. In this section we ask whether 19 these latter two research approaches can be brought together to provide a means for more fully integrated agri-20 food research.

21
One potentially viable method is to examine the socio-economic, political and ethical factors at each

34
We propose that this analytical approach should be combined with the systems analysis approach that new technology or policy would be formulated within the whole agri-food ecosystem context, mapping its 1 components, processes and boundaries (as outlined above). It would then be subject to LCA. The data and 2 evidence emerging from this analysis would be made available to all stakeholders for further analysis. This 3 would involve two further stages of analysis: simulation modelling and experimental testing to fine tune the 4 technology or policy; and debate and discussion, through deliberative fora and public engagement.

5
As suggested above, a promising mechanism for generating reflection and consensus between 6 stakeholders in cases of evidence complexity and entrenched interests is through targeted 'deliberative fora', 7 where multisectoral stakeholders and representatives within the agri-food system can be guided through a series   food research are to be realised. Taking the example of GM discussed above, the processes of scientific analysis 37 and testing have previously often been divorced from the public discourse about risk and ethics. If the two 38 processes were brought together as represented in Figure 3, the conflict might be resolvedor at least the competing interests would be rendered more transparent such that trust between science, technology, 1 government and public might be restored.

2
A second key barrier is the major conflict embedded in the agri-food system. The primary purpose of 3 the food producing sectors is to make money not to provide sustainable global food security, the definition of 4 which includes access to nutritious food (Trudge 2016). For example, high agricultural productivity, necessary 5 for farmers, agri-businesses and food retailers to make a profit, whilst also keeping prices low for consumers, 6 currently requires environmentally unsustainable farming practices. The drive to increase yields of corn and 7 sugar cane leads to increased use of sweeteners, with consequent health effects. The environmental and health 8 impacts of these practices are not costed within the system and thus, there are currently no effective incentives 9 to implement the required improvement. For the reasons given above, regulations are often ineffective and have 10 unforeseen consequences. Thus, even if rational, evidence-based solutions could be generated from the research 11 approaches we are advocating, would they be implemented?

12
Research is therefore urgently needed to find ways to incentivise all sectors of the agri-food system

24
This paper has outlined the development of an integrated approach to agri-food research in order to address the 25 complex challenge of food security. It has sought to map the agri-food system, to identify its component parts 26 and to argue the case for approaching the system in an integrated way rather than as a series of separate 27 domains. We have shown how taking this approach transforms the framing of research within each domain and 28 we have proposed two ways of taking this agenda forward, through the application of quantitative analysis 29 (using LCA and related methods) and through the recognition of the ethical, political and legal tensions that 30 characterise the system (using deliberative fora). We have also identified some of the methodological and 31 epistemological challenges of taking these ideas forward, acknowledging some of the barriers to their practical 32 implementation.

33
Our approach might also be thought of in terms of the critical deployment of 'nexus thinking' (Leck et 34 al. 2015), an approach that is bring advocated in the UK through parallel research programmes from the ESRC 35 and EPSRC and in a range of international initiatives. 8 Rather than seeing energy, food and water resources as 36 8 ESRC has invested £1.8m in its Nexus Network programme (http://www.thenexusnetwork.org/ ), while EPSRC has invested £4.5m on as similar programme, focused on safeguarding the UK's food, water and energy security (https://www.epsrc.ac.uk/newsevents/news/ukwaterenergyfood/). Similar programmes are being developed in the US by the National Science Foundation (https://www.nsf.gov/pubs/2016/nsf16524/nsf16524.htm). There is a Future Earth Network on the nexus separate systems, nexus thinking addresses the inter-dependencies, tensions and trade-offs between these 1 different domains, similar to the approach taken in this paper, moving beyond national, sectoral, policy and 2 disciplinary silos to identify more efficient, equitable and sustainable ways of using scarce resources. While 3 some have criticised the concept as little more than a contemporary 'buzzword' (Cairns & Krzywoszynska 4 2016) and others have promoted the value of nexus thinking in methodological terms (Stirling 2015), we are 5 keen to put the concept to work through practical applications that explore the links between food, energy and 6 water security at a range of geographical scales. 9 Consistent with the idea of nexus thinking, this paper has 7 sought to outline an integrated agenda for system-wide interdisciplinary agri-food research, capable of 8 addressing the global challenges of enhanced food security.                   Ethical, legal, and political tensions in agri-food systems