1 AGROECOLOGY: TRANSITION PATHWAYS TOWARDS SUSTAINABLE FOOD SYSTEMS Agroecology is a dynamic concept that has gained prominence in scientific, agricultural and political discourse in recent years (IAASTD, 2009; IPES-Food, 2016). During its historical evolution, agroecology has expanded beyond the field, farm and agroecosystem scales to encompass, over the last decade, the whole food system. Agroecological approaches explicitly aim at transforming food and agriculture systems, addressing the root causes of problems and providing holistic and long-term solutions (FAO, 2018a) that consider the complexity of farming systems within their social, economic and ecological contexts (Petersen and Arbenz, 2018). Agroecological approaches are increasingly considered as possible alternatives to the industrial model of agricultural improvement,14 representing concrete transition pathways towards SFSs that enhance FSN (De Schutter, 2010; HLPE, 2016, 2017a,b). In September 2014, FAO organized an International Symposium on Agroecology for FSN, followed in 2015 by three regional meetings in Latin America, Africa and Asia (FAO, 2015a,b; 2016a), a further three regional meetings in 2016 in Latin America, China and Europe, and the most recent in 2017 in North Africa (FAO, 2018b). A second International Symposium was convened by FAO in April 2018, the key outcomes of which are documented in Chapter 4 and have informed the development of some of the recommendations in this report. This chapter begins by describing how the concept of agroecology has emerged from constituent elements of agriculture and ecology to embrace a transdisciplinary science, a set of practices and a social movement. It then presents the definition and development of agroecological principles over time, analysing how these principles contribute to FSN and the achievement of the Sustainable Development Goals (SDGs). Finally, some contested areas in current debates around agroecology and major knowledge gaps are also highlighted. 1.1 Agroecology: a science, a set of practices and a social movement Multiple definitions of agroecology exist as different institutions and countries adopt definitions that reflect their own concerns and priorities. This report aims at defining and characterizing agroecological approaches for sustainable agriculture and food systems that enhance FSN. Historically, traditional agricultural systems in many parts of the world could be considered to be agroecological. These include traditional agroforestry, incorporation of organic material into soils, mixed cropping systems with livestock and the use of a wide variety of edible crops (Altieri, 2004a). Dynamic local knowledge systems developed complex approaches to managing pests, diseases and ensuring culturally appropriate, nutrient-rich food supplies (Altieri 2004a; Oteros-Rozas et al., 2013). Modern agroecological science, as a response to the social and ecological impacts of the so-called “industrial” agriculture model, draws on many locally derived concepts and practices, but is also a dynamic and active area of scientific research (Migliorini et al., 2018l; Montalba et al., 2017; Vandermeer and Perfecto, 2013). In its report on sustainable agricultural development and the role of livestock, the HLPE (2016b)) described agroecology from a scientific and technical perspective as the application of ecological concepts and principles to farming systems, focusing on the interactions between plants, animals, humans and the environment, to foster sustainable agricultural development in order to ensure FSN for all, now and in the future. The report acknowledges that “today’s more transformative visions of agroecology integrate transdisciplinary knowledge, farmers’ practices and social movements while recognizing their mutual dependence”, and calls for looking at a broader conception of the term. This is in line with agroecological approaches having broadened in recent years, to focus on whole agrifood systems, not only farming systems (Thompson and Scoones, 2009), and to go beyond ¹⁴ The industrial model of agricultural improvement refers to intensive agricultural systems, dominated by large-scale specialized farms, relying in certain cases heavily on fossil fuel and purchased, non-renewable and synthetic inputs. These systems are criticized by agroecology proponents who point to their negative social, health and environmental impacts (IPES-Food, 2016; HLPE, 2016). separating scientific and technical dimensions of agroecology from the social and political dimensions, by embracing a transdisciplinary outlook. The notion of agroecology as the application of ecological principles in agriculture, while apparently simple, hides complex realities because ecology and agriculture are dynamic concepts. Ecology refers to the branch of biology dealing not only with interactions among organisms and with their environment (Tansley, 1935) but also to social movements concerned with the protection of the environment (Sills, 1974). Although ecological science began as a subdivision of biology, it has more recently emerged as an interdisciplinary field with many different branches, including political ecology (Robbins, 2004), many of which link biological, physical and social sciences. Agriculture is basically the set of practices through which people produce food (Spedding, 1996). Agriculture, as a concept, is also evolving, with increasing awareness that agriculture is multi- functional (Caron et al., 2008; IAASTD, 2009), and that agricultural production cannot be separated from the other aspects of food systems, such as food supply chains, the food environment and consumption (Jones and Street, eds,1990; HLPE, 2017b). These trends in ecology and agriculture come together in an emerging transdisciplinary focus on understanding and managing coupled social–ecological systems (Berkes and Folke, eds,1998) in a context of growing concerns about human activities, and agriculture in particular, leading to planetary boundaries being exceeded (Steffen et al., 2015; Campbell, 2017). A key reason that agroecology is gaining traction in the discourse on achieving FSN is because it is perceived to bridge ecological and social dimensions associated with the development of resilient food systems in the face of climate change and other global challenges (Caron et al., 2014). Agroecology is increasingly seen as a transdisciplinary, participatory and action-oriented approach (Méndez et al., 2013; Gliessman, 2018) that embraces three dimensions: a transdisciplinary science (Definition 1), a set of practices and a social movement (Wezel et al., 2009; Wezel and Silva, 2017; Agroecology Europe, 2017) (Box 3). These three dimensions of agroecology, their articulation and co-evolution together constitute a holistic approach (e.g. Agroecology Europe, 2017; Gliessman, 2018). Box 3 Multiple definitions of agroecology As a science, agroecology is: (i) the integrative study of the ecology of the entire food system, encompassing ecological, economic and social dimensions or, in brief, the ecology of the food system (Francis et al., 2003); (ii) the application of ecological concepts and principles to the design and management of sustainable food systems (Gliessman, 2007); and, more recently, (iii) the integration of research, education, action and change that brings sustainability to all parts of the food system: ecological, economic and social (Gliessman, 2018). Agroecological practices aim at improving agroecosystems by harnessing natural processes, creating beneficial biological interactions and synergies among their components (Gliessman, ed, 1990) and using, in the best way, ecological processes and ecosystem services for the development and implementation of practices (Wezel et al., 2014). As a social movement, agroecology is seen as a solution to current challenges such as climate change and malnutrition, contrasting with the so-called “industrial” model and transforming it to build locally relevant food systems that strengthen the economic viability of rural areas based on short marketing chains, and fair and safe food production. It supports diverse forms of smallholder food production and family farming, farmers and rural communities, food sovereignty, local knowledge, social justice, local identity and culture, and indigenous rights for seeds and breeds (Altieri and Toledo, 2011; Rosset et al., 2011; Nyéléni, 2015). This dimension of agroecology as a political movement is becoming increasingly prominent (Gonzalez de Molina, 2013; Toledo and Barrera-Bassols, 2017). Sources: FAO (2017a), Agroecology Europe (2017). Definition 1 Transdisciplinary science Transdisciplinary science transcends disciplinary boundaries and seeks to generate transformative outcomes by having: (i) a problem focus (research originates from and is contextualized in ”real-world” problems); (ii) an evolving methodology (the research involves iterative, reflective processes that are responsive to the particular questions, settings and research groupings involved); and (iii) collaboration (including among transdisciplinary researchers, disciplinary researchers and external actors with interests in the research) (Russel et al., 2008). This has been interpreted in agroecology to involve integration of different academic disciplines as well as diverse forms of knowledge, including experiential, cultural and spiritual (Méndez et al., 2015). Transdisciplinary science differs from ”multidisciplinary” science, where people from different disciplines work together, each drawing on their disciplinary knowledge in an additive rather than integrative way, and from ”interdisciplinary” science, where knowledge and methods from different disciplines are integrated, encompassing a synthesis of approaches but not necessarily involving other stakeholders or focus on generating transformative outcomes (Petrie, 1992). 1.1.1 Agroecology as a science The term “agroecology” appeared for the first time in the scientific literature at the beginning of the twentieth century to designate the application of ecological methods and principles in agricultural sciences, including zoology, agronomy and crop physiology (Figure 2a) (Bensin, 1928, 1930; Friederichs, 1930; Klages, 1942; Gliessman, 1997; Dalgaard et al., 2003; Wezel et al., 2009; Wezel and Soldat, 2009). In the 1950s and 1960s, Tischler published several articles on agroecological research, analysing different components (plants, animals, soils, climate) and their interactions, as well as the impact of human management on them. His book was probably the first book entitled Agroecology (Tischler, 1965). The concept of an “agroecosystem”, considered as a domesticated, human-managed ecosystem, was introduced by Odum (1969). Two decades later, agroecology began to move beyond the field and farm scales to embrace whole agroecosystems (Altieri, 1987, 1989; Conway, 1987; Marten, 1988; Wezel et al., 2009; Wezel and Soldat, 2009). Important contributions also came from Mexican scientists emphazing intercultural processes for constructing agroecological knowledge that combines ecological science with local peoples’ knowledge (e.g. Hernández Xolocotzi, 1977). Building on these reflections, Altieri (1995) defined agroecology as “the application of ecological concepts and principles to the design and management of sustainable agroecosystems”. FAO (FAO, 2016d) further refined this definition, stating that: “Agroecological innovations apply ecological principles - such as recycling, resource use efficiency, reducing external inputs, diversification, integration, soil health and synergies -, for the design of farming systems that strengthen the interactions between plants, animals, humans and the environment for food security and nutrition.” In the 2000s, the transdisciplinary nature of agroecological science, combining natural and social sciences, became increasingly important (Wezel et al., 2015). Agroecology was identified as “an integrated discipline that includes elements from agronomy, ecology, sociology and economics” (Dalgaard et al., 2003). The focus of agroecological science was broadened to encompass the whole agrifood system (Francis et al., 2003; Doré et al., 2006; Gliessman, 2007; Wezel and David, 2012; Côte et al., eds, 2019) and to cover various topics such as: alternative and local food networks; consumer–producer relationships; social agricultural networks; food markets; and public food procurement. This food systems’ approach also includes the relationships between rural and urban areas, leading to the development of urban agroecology (AS PTA, 2011; Almeida and Biazo, 2017; Renting, 2017; Morales et al., 2018; see also Box 4). Box 4 Urban agriculture Potentially, urban and peri-urban agriculture (UPA) can play a role in enhancing social and environmental conditions in cities through food security and poverty alleviation, although some people caution that this should not be overemphasized (Zezza and Tasciotti, 2010). In Equatorial Africa, Lee- Smith (2010) found that UPA increases as urban areas expand and that it favours improvement in human health as well as alleviation of hunger and poverty. At a global level, Mok et al. (2014) found that UPA has potential to make significant contributions to food security, although more research must be done around issues such as urban sprawl. On the other hand, some authors concluded that UPA has only a limited potential to contribute to urban food security in developing countries given constraints related to access to land, water and financial resources to invest in productive areas in the urban space (Badami and Ramankutty, 2015). UPA is also valued for the environmental benefits that it might promote, such as biodiversity conservation, reduction of food miles and therefore carbon emissions, and increasing the green areas in urban landscapes. UPA in its many forms – allotment gardens, rooftop gardens, home orchards, urban arborization and community gardens, among others – can contribute to a number of ecosystem services such as pollination, pest control, climate resilience and water regulation (Lin et al., 2015). In fact, locally produced food in urban areas can contribute to create short circuits of commercialization, reducing transportation and also helping to develop direct selling schemes. Finally, UPA has historically contributed to improve living conditions, increase income and alleviate poverty in cities, strengthening their resilience (Barthel and Isendahl, 2013). In many African countries, where agriculture is the main source of income for the majority of families, UPAs may provide a substantial share of income in addition to promoting a substantial improvement in household diets, contributing to food security and nutrition (Zezza and Tasciotti, 2010). In Mexico City, around 20 percent of all food consumed is produced in urban and peri-urban areas; however, the recognition of the importance of UPA in economic terms and as a source of employment is quite limited. A symbolic dimension of UPAs in Mexico is the pre-Hispanic chinampas system (floating gardens) developed by the Aztecs that were greatly reduced following European colonization (Dieleman, 2017). The machambas are small agricultural plots in urban and peri-urban areas in Mozambique where small entrepreneurs, in general women, cultivate vegetables to sell in the cities; they are an important source of food and income for many households in cities such as Maputo (Sheldon, 1999). In the historical evolution of agroecology as a science (Figure 2b), the scale and dimension of research in agroecology have been enlarged from (i) the plot, field or animal scale to (ii) the farm or agroecosystem scale and, finally, to (iii) the whole food system, which is increasingly becoming a focus for agroecology (Wezel and Soldat, 2009). Figure 2 Historical evolution of Agroecology 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s 2010s A Elements Policies for agroecology Policies and laws Indigenous knowledge and family farms Agrobiodiversity and rights to food Food systems, rural and territorial development food sovereignty Agroecology as a social movement Indigenous agriculture knowledge for natural resource management Agroecological practices are introduced or further developed (conservation agriculture, intercropping, biological control, etc.) Agroecological practices as an alternative paradigm to conventional agriculture Agroecology as a set of practices Scale: field/plot agro-ecosystem Scope: biology, zoology, ecology, crop physiology From descriptive to analytical Scale: field and agro-ecosystem Scope: ecology, agronomy Analytical nature Conceptual framework to design and manage agro-ecosystems; from analytical to prescriptive Further increases in disciplines, scope and scale: agroecology as the ecology of food systems Agroecology as a scientific discipline B Scale/dimension Food system Farm, agroecosystem Plot, field C Disciplines Agricultural, environmental Political Economic Social and cultural 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s 2010s Sources: (A) adapted from Silici (2014), based on Wezel et al. (2009) and Wezel and Soldat (2009); (B) adapted from Wezel et al. (2009). Note: (C) illustrates the disciplinary basis of the principles of agroecology articulated in Section 1.2. 1.1.2 Agroecology as a set of practices In the 1960s, in particular following the publication of Rachel Carson’s book Silent spring (Carson, 1962), concerns emerged about unexpected impacts of the intensive use of synthetic inputs in agriculture on the environment, particularly about the concentration of pesticide residues through food chains impacting birds of prey. Partly in response to this, a set of agroecological practices emerged over the next few decades (see Section 1.5, Figure 3) aiming at moving away from what has been called an “industrial agriculture model” towards more environmentally friendly and sustainable agricultural systems, optimizing the use of biological processes and ecosystem functions (Hernández Xolocotzi, 1977; Rosset and Altieri, 1997; Wezel et al., 2009; Vanloqueren and Baret, 2009; Altieri et al., 2012a; Wibbelmann et al., 2013; Pimbert, 2015; IPES-Food, 2016; FAO, 2016b; Wezel et al., 2014; Deguine et al., eds, 2017; Wezel, 2017). Agroecology, as a set of practices, aims at designing complex and resilient agroecosystems that, by “assembling crops, animals, trees, soils and other factors in spatially and temporally diversified schemes, favour natural processes and biological interactions that optimize synergies so that diversified farms are able to sponsor their own soil fertility, crop protection and productivity” (Altieri, 2002). Attempts to define which specific practices can be qualified as agroecological are only recently emerging. For example, Wezel et al. (2014) describe agroecological practices as “agricultural practices aiming to produce significant amounts of food while valuing ecological processes and ecosystem services by integrating them as fundamental elements”. For Shiming and Gliessman (2016), “agroecological practices are those ecologically sound methods which can balance and enhance all ecosystem services provided by agroecosystems and hence benefit to the sustainable development of agriculture”. However, there is no definitive set of practices that can be labelled as agroecological, nor clear, consensual boundaries between what is agroecological and what is not (Wezel, 2017). On the contrary, agricultural practices can be classified along a spectrum and qualified as more or less “agroecological”, depending on the extent to which: (i) they rely on ecological processes as opposed to the use of agrochemical inputs; (ii) they are equitable, environmentally friendly, locally adapted and controlled; and (iii) they adopt a systemic approach, rather than focusing only on specific technical measures. Agroecological practices involve processes such as: nutrient cycling; biological nitrogen fixation; improvement of soil structure and health; water conservation; biodiversity conservation and habitat management techniques for crop-associated biodiversity; carbon sequestration; biological pest control and natural regulation of diseases; diversification, mixed cultivation, intercropping, cultivar mixtures; and waste management, reuse and recycling as inputs to the production process, for example use of manure and compost (Reijntjes et al., 1992; Altieri 1995; Nicholls et al., 2016; Wezel et al., 2014; Wezel, 2017). Agroecological practices include, for instance, agroecological responses to new pest epidemics such as the recent spread of fall armyworm in Africa (Box 5) or crop–animal integration in traditional systems such as the rice–duck–fish system in Asia (Box 6). Some of these practices have been applied to varying extents in different parts of the world for decades, while others have emerged more recently with as yet limited levels of adoption (Wezel et al., 2014; Wezel and Silva, 2017). For example, organic fertilization, split fertilization, reduced tillage, drip irrigation, biological pest control, integrated pest management and choice of cultivars resistant/tolerant to biotic stresses (diseases, insect pests and parasitic weeds) are already widely integrated into temperate agriculture, in small- and larger-scale farms. Biofertilizers, natural pesticides and biopesticides, diversified rotations, intercropping and relay intercropping, agroforestry, allelopathic plants, direct seeding into living cover crops or mulch, and integration of semi-natural landscape elements at field, farm and landscape scales are less present in temperate agriculture, but prevalent in some tropical contexts (Leakey, 2014). Some agroecological practices, such as organic fertilization and intercropping, came into use with the development of organic agriculture in the 1940s. Box 5 Agroecological practices to control fall armyworm in Africa Fall armyworm (FAW), a voracious agricultural pest native to North and South America, was first detected on the African continent in 2016 (Goergen et al., 2016). Since then it has spread across sub- Saharan Africa affecting thousands of hectares of cropland, causing up to USD13 billion per annum in crop losses (Abrahams et al., 2017) and threatening the livelihoods of millions of farmers. In their haste to respond to FAW, governments have sometimes relied heavily on agrochemicals that, beyond the risks they can pose on human health and the environment, are likely to undermine biological pest management strategies (Abate et al., 2000; van Huis and Meerman, 1997; Wyckhuys and O’Neil, 2010). Agroecological approaches can offer locally adapted, low-cost, biological pest control options, including: • sustainable soil and land management (e.g mulching), which improves crop health and resilience to pest attack (Altieri and Nicholls, 2003; Clark et al., 1993; Rivers et al., 2016); • intercropping, which can reduce egg-laying by pests through deterrence by volatile chemicals released by intercrop plants (Midega et al., 2018), trapping emerging FAW larvae, increasing their mortality (van Huis, 1981) and providing habitat for natural enemies within the field (Rivers et al., 2016); • crop rotation, which improves soil fertility and diversifies the farm environment (Wyckhuys and O’Neil, 2007; Meagher et al., 2016; Rivers et al., 2016); • weeds, shrubs, trees and (semi-) natural habitats managed at multiple spatial scales, in fields or at field margins, which can provide habitat for a variety of natural pest enemies (Bàrberi et al., 2010; Maas et al., 2013, 2016; Meagher et al., 2016; Wyckhuys and O’Neil, 2007; Bàrberi et al., 2010; Sisay et al., 2018; Leakey, 2014; Morris et al., 2015; van Huis, 1981; Offenberg, 2015); • regular scouting by the farmer to identify pests and assess damage that informs pest management decisions (McGrath et al., 2018); Agroecological practices are now being advocated as a core component of integrated pest management programmes for FAW in sub-Saharan Africa in combination with crop breeding, classical biological control and selective use of chemical pesticides (Harrison et al., 2019; Thierfelder et al., 2018). Box 6 Traditional rice–fish–duck system in Hani terraces, Southwest China The rice–fish–duck system is an important traditional agroecosystem in Hani terraces in Yunnan Province, Southwest China. Integration of crops and animals and circular economy are at the heart of this system. Fish and ducks eat weeds and pests and loosen the soil to improve the growing environment for rice, while rice provides food, shade and shelter for fish and ducks. Pesticides and herbicides cannot be used in this system because of their toxicity to fish and ducks. Therefore, products from rice–fish–duck systems are very popular in consumer markets. Their prices are usually several times higher than the prices of conventional products. For example, the prices of red rice, fish and ducks raised in paddies in Hani terraces are respectively 5, 3 and 2.5 times higher than conventional prices. An improved rice–fish–duck system has been experimented with in Hani terraces and is now popularized. The agroecosystem efficiently exploits the three-dimensional space (and seasonality) of paddies for developing rice–fish co-culture during the crop growing season while ducks are reared in winter during the fallow period. Its economic value is estimated to be 7.8 times that of the current conventional model that only grows the hybrid rice monoculture in summer for half the year and fallows the field in winter (Zhang et al., 2017). This is an example of a Globally Important Agricultural Heritage System (GIAHS) that combines agricultural biodiversity, resilient ecosystems, local communities and a valuable cultural heritage.15 There is a network of 50 GIAHS sites in 20 countries of the world (FAO, 2002; Koohafkan and Altieri, 2010; Koohafkan and Cruz, 2011; HLPE, 2017b). ¹⁵ See: http://www.fao.org/giahs/en/ 1.1.3 Agroecology as a social movement Traditional agricultural systems, in their diversity, are the result of the co-evolution of ecosystems and human communities across many generations. Therefore, agroecosystems cannot be separated from the human communities living in them: social and political dynamics are at the heart of agroecology (Altieri, 2004b; Wibbelmann et al., 2013; Ploeg and Ventura, 2014). Agroecological approaches often arise in response to agrarian crises, and along with broader efforts of social movements to initiate widespread changes (Mier y Terán et al., 2018 ; Box 7). These social movements advocate for a strong connection to be made between agroecology, the right to adequate food and food sovereignty. The concept of food sovereignty was first introduced in international discussions during the World Food Summit in 1996, Rome, by La Via Campesina, an international movement of peasants. In 2007, civil society organizations (CSOs) and social movements gathered in Nyéléni, Mali, defined food sovereignty as “the right of peoples to healthy and culturally appropriate food produced through ecologically sound and sustainable methods, and their right to define their own food and agriculture systems” (Nyéléni, 2007). The initial set of seven principles of food sovereignty included: (i) food as a basic human right; (ii) the need for agrarian reform; (iii) protection of natural resources; (iv) reorganization of food trade to support local food production; (v) reduction of multinational concentration of power; (vi) fostering of peace; and (vii) increasing democratic control of the food system (La Via Campesina, 1996). In February 2015, eight years after this first International Forum for Food Sovereignty, diverse social movements and organizations representing small-scale food producers gathered again in Nyéléni for an International Forum on Agroecology (Nyéléni, 2015). In their final declaration, they consider “agroecology as a key element in the construction of food sovereignty”. For them, agroecology is not only “a narrow set of technologies” but, above all, a political struggle, requiring people to “challenge and transform structures of power in society”, addressing power imbalances and conflicts of interest, in order to “generate local knowledge, promote social justice, nurture identity and culture, and strengthen the economic viability of rural areas”. Agroecology has thus become the political framework under which many social movements and peasant organizations around the world defend their collective rights and advocate for a diversity of locally adapted agriculture and food systems practised by small-scale food producers in different territories (Anderson et al., 2015; Nyéléni, 2015). Agroecology is seen as a bottom-up pathway to food sovereignty, building on traditional knowledge systems, supported rather than led by science, where small producers, their communities and organizations, rather than agrifood business, play a central role. Agroecological approaches aim at building resilient and sustainable local food systems, strongly linked and adapted to their territories and ecosystems (Varghese and Hansen-Kuhn, 2013; Nyéléni, 2015; Anderson et al., 2015). Some national governments have adopted policies embracing the principles of agroecology and food sovereignty in order to transform food systems (Altieri et al., 2012b; Wezel et al., 2009; Lambek et al., 2014). Box 7 Rede Ecovida in Southern Brazil The Rede Ecovida or ”Ecolife Network” is a decentralized system of cooperatives, farmer groups and non-profit organizations that practise agroecology in 150 municipalities in three southern Brazilian states. The network developed in the 1970s as part of broader social movements mobilizing around issues of environmental damage from agriculture, of high social inequalities and uneven land distribution. Ecovida currently comprises 29 farmers’ organizations, 2 700 farming households, 10 cooperatives, 25 associations, 180 farmers’ markets and 30 agrifood private companies. Beyond profit, this network promotes a solidarity economy between producers and consumers in local markets (including door-to- door sales, community canteens, farmers’ markets and restaurants). It uses participatory certification to ensure that farming practices are rooted in agroecology and strengthen the relationships/links/trust among farmers and with urban consumers. Overall, this network promotes horizontal learning methods, solidarity, justice and care for nature. Sources: Perez-Cassarino (2012); Mier y Terán et al. (2018). 1.1.4 Agroecology as an innovative approach to sustainable food systems for food security and nutrition As discussed above, there are an increasing number of definitions for agroecology provided in recent years that have different nuances depending on the authors, institutions or CSOs that provide them. What they have in common is the goal to develop SFSs. Regarding these different definitions and the specific focus of this report on FSN, rather than presenting yet another definition of agroecology per se, a definition of an agroecological approach to SFSs for FSN is offered based on the analysis and information presented in this chapter (Definition 2). Definition 2 Agroecological approach to sustainable food systems for food security and nutrition Agroecological approaches favour the use of natural processes, limit the use of purchased inputs, promote closed cycles with minimal negative externalities and stress the importance of local knowledge and participatory processes that develop knowledge and practice through experience, as well as more conventional scientific methods, and address social inequalities. Agroecological approaches recognize that agrifood systems are coupled social–ecological systems from food production to consumption and involve science, practice and a social movement, as well as their holistic integration, to address FSN. 1.2 Principles of agroecology Scientists have developed different sets of agroecological principles (Reijntjes et al., 1992; Altieri, 1995; Altieri and Nicolls, 2005; Stassart et al., 2012; Dumont et al., 2013, 2016; Nicholls et al., 2016; Peeters and Wezel, 2017; all summarized in Migliorini and Wezel, 2018). Civil society networks also conducted the same exercise (e.g. Nyéléni, 2015; CIDSE, 2018). Today, agroecology is associated with a set of principles for agricultural and ecological management of agrifood systems as well as some wider ranging socio-economic, cultural and political principles (e.g. CIDSE, 2018). These latter principles have emerged only recently in the literature, arising from the activity of agroecological social movements (Figure 2c). FAO (2018c) identified ten elements of agroecology to guide the transition towards sustainable agriculture and food systems.16 These consolidated FAO ten elements are based upon seminal scientific literature on agroecology (in particular: Altieri, 1995; Gliessman, 2007) and upon the extensive and inclusive multi-stakeholder dialogues, gathering states and intergovernmental organizations, CSOs and private actors, held at global, regional and national levels since the first FAO International Symposium on Agroecology (September 2014). Building on all these efforts, the HLPE elaborated a consolidated list of 13 principles, combining and reformulating principles from the three principal sources (Nicholls et al., 2016; CIDSE, 2018; FAO, 2018d) to produce a minimum, non-repetitive but comprehensive set of agroecological principles. These are organized around the three operational principles for SFSs set out in the introduction – improve resource efficiency, strengthen resilience and secure social equity/responsibility (see Table 1). Each agroecological principle was allocated to the operational principle to which it most clearly contributes. However, given the interlinkages among these three categories, this classification is not fully discrete. For example, principles 3, 5 and 6 contribute not only to resilience but also to resource efficiency. Principles are also related to the FAO ten elements.17 ¹⁶ Diversity; co-creation and sharing of knowledge; synergies; efficiency; recycling; resilience; human and social values; culture and food traditions; responsible governance; circular and solidarity economy. ¹⁷ See: http://www.fao.org/3/i9037en/I9037EN.pdf Different principles can be implemented at or impact different scales, from local to global, from the field to the whole food system. At the agroecosystem or landscape scale, some ecological processes, such as water flows, operate over large distances so that what farmers do in one location may impact other people positively (clean water supply) or negatively (flooding or contaminated water) many kilometres away, across administrative and national boundaries (Jackson et al., 2013). Soil eroded from one place may be deposited and support food production elsewhere. Recent research has shown that not only surface water flow but also atmospheric transfers across continents are important, so that change in vegetation cover in the East African highlands impacts rainfall and hence agricultural productivity in the Sahel (van Noordwijk et al., 2014). This means that concepts of resource cycles and flows (principles 1 and 5) need to be related to the scales at which they operate, and many ecosystem services, such as pollination, quantity and quality of water provision and habitat provision for biodiversity conservation, only manifest at landscape scale and, hence, can only be managed by collective action of farmers and other stakeholders (Pagella and Sinclair, 2014). Application of agroecological principles often aims at reducing externalities associated with current models of agricultural production. Measuring and valuing ecosystem service provision at a range of scales is a key area of innovation required to measure performance of food systems in ways that address their sustainability. This is further developed in Chapters 2 and 3. All these agroecological principles contribute, in different direct and indirect ways, to FSN. For instance, principle 2 (reducing the dependency on purchased inputs) can reduce food insecurity in particular for smallholders and for poor farmers because less money is spent on buying inputs and so there is less reliance on credit and, therefore, potentially more resources to buy food (Snapp et al., 2010; Kangmennaang et al., 2017; Hwang et al., 2016). This is a primary motivation for the Zero Budget Natural Farming (ZBNF) agroecological movement in India (Box 8). Principle 9 (social values and diets) together with 5 (biodiversity) impact nutrition directly (Jones et al., 2014b; Powell et al., 2015; Bellon et al., 2016; Demeke et al., 2017; Lachat et al., 2018; HLPE, 2017a,b). Co-creation of knowledge (principle 8) can also have indirect positive impacts on FSN (Box 9). Principle 11 (connectivity) may contribute to strengthen local economies, increasing the proportion of value added remaining on farms and enabling producers to better meet the food needs and demands of local consumers. This latter point can be supported by strong social organizations, which foster greater participation of local food producers and consumers in decision-making (principle 13). Table 1 Consolidated set of 13 agroecological principles Principle FAO’s ten elements Scale application* Improve resource efficiency 1. Recycling. Preferentially use local renewable resources and close as far as possible resource cycles of nutrients and biomass. 2. Input reduction. Reduce or eliminate dependency on purchased inputs and increase self-sufficiency Recycling FI, FA Efficiency FA, FO Strengthen resilience 3. Soil health. Secure and enhance soil health and functioning for improved plant growth, particularly by managing organic matter and FI enhancing soil biological activity. 4. Animal health. Ensure animal health and welfare. 5. Biodiversity. Maintain and enhance diversity of species, functional diversity and genetic resources and thereby maintain overall agroecosystem biodiversity in time and space at field, farm and landscape scales. 6. Synergy. Enhance positive ecological interaction, synergy, integration and complementarity among the elements of agroecosystems (animals, crops, trees, soil and water). 7. Economic diversification. Diversify on-farm incomes by ensuring that small-scale farmers have greater financial independence and value addition opportunities while enabling them to respond to demand from consumers. Part of diversity Synergy Part of diversity FI, FA FI, FA FI, FA FA, FO Secure social equity/responsibility 8.Co-creation of knowledge. Enhance co-creation and horizontal sharing of knowledge including local and scientific innovation, especially through farmer-to-farmer exchange. 9. Social values and diets. Build food systems based on the culture, identity, tradition, social and gender equity of local communities that provide healthy, diversified, seasonally and culturally appropriate diets. 10. Fairness. Support dignified and robust livelihoods for all actors engaged in food systems, especially small-scale food producers, based on fair trade, fair employment and fair treatment of intellectual property rights. 11. Connectivity. Ensure proximity and confidence between producers and consumers through promotion of fair and short distribution networks and by re-embedding food systems into local economies. 12. Land and natural resource governance. Strengthen institutional arrangements to improve, including the recognition and support of family farmers, smallholders and peasant food producers as sustainable managers of natural and genetic resources. 13. Participation. Encourage social organization and greater participation in decision-making by food producers and consumers to support decentralized governance and local adaptive management of agricultural and food systems. *Scale application: FI = field; FA = farm, agroecosystem; FO = food system Source: derived from from Nicholls et al., 2016; CIDSE, 2018; FAO, 2018c. Co-creation and sharing of knowledge Parts of human and social values and culture and food traditions Circular and solidarity economy Responsible governance FA, FO FA, FO FA, FO FA FA, FO FO It has been suggested that for agroecology to significantly impact FSN and generate sustainable diets,18 power inequalities must be addressed within the food system at multiple scales and in different dimensions (HLPE, 2017a; Mier y Teran et al., 2018; Pimbert and Lemke, 2018). Horizontal teaching methods (principle 8) are options for agroecology to address social inequalities; principles 10–13 articulate how other inequalities can be addressed as part of an agroecological approach. Potential trade-offs must also be considered in each specific context. For instance, depending on quantity and type of inputs, reduced input use (principle 2) could lead to lower productivity, lower income and thus higher food insecurity. In addition, agroecological methods, if more labour-intensive, could increase women’s workload, leading to worsening the nutritional status of children if gender relationships within households are not changed (principle 9). Box 8 Zero Budget Natural Farming – Scaling-up agroecology in India Zero Budget Natural Farming (ZBNF) is both a set of farming methods and a grassroots peasant movement in India born in Karnataka. It is estimated that ZNBF methods are used by 100 000 farming families in Karnataka, and by millions of families at the national level. In 2015 the Government of Andhra Pradesh announced its objective to reach 500 000 farmers with ZBNF by 2020. Interest in ZBNF methods arose partly because of the high rates of farmers’ debt, originating from the costs of fertilizers, seeds, energy and equipment (mechanization and irrigation), which have been linked to high suicide rates. More than a quarter of a million farmers have committed suicide in India in the last two decades. ”Zero Budget”, which means not relying on credit, and not buying inputs, promises to put an end to heavy debt, by drastically reducing production costs. “Natural Farming” means farming with nature and without purchased chemical inputs. ZBNF methods include: mulching; intercropping; controlled irrigation; contour bunds; use of local earthworm species and fermented microbial culture; combined seed treatment with cow dung, sugar, pulse flour, urine and soil. At the local level, ZBNF operates mainly through volunteers, members of farmer organizations and community leaders, motivated by the founder of the movement, Subhash Palekar, an agricultural scientist who has written many publications on ZNBF methods. At the state level, intensive five-day training camps are held, with support from volunteers and allied organizations. A survey of 97 ZBNF farmers reported increased yield, seed diversity, product quality, household food autonomy, income and health, along with reduced farm expenses and credit needs. The following strategic elements were critical for the successful implementation of ZBNF in India: • Charismatic leadership. A highly charismatic teacher, Subhash Palekar has played a key role in motivating and promoting ZBNF methods through books, training courses and other public appearances. • Horizontal pedagogical practices. While Palekar teaches in a more vertical manner, most of the teaching is done through farmer-to-farmer exchanges and mentoring. • Supportive public policy. Training is provided at the state level in several Indian states. • Local and favourable markets. At least eight shops exclusively retail ZBNF products in cities such as Bangalore and Mysore, but marketing remains a challenge. • Strong social organization. States organize training camps and informal networks support training and ongoing support for ZBNF with links to allied organizations. • Efficient farming practices. Farmers report improved yields, food quality and income, and reduced farm expenses and credit. • Cultural relevance. ZBNF methods address the credit and debt concerns of farmers in socially and culturally adapted ways. Sources: Khadse et al. (2018) ; Kumar (2018); La Via Campesina (Undated) ¹⁸ “Sustainable diets are those with low environmental impacts which contribute to food and nutrition security and to healthy life for present and future generations. Sustainable diets are protective and respectful of biodiversity and ecosystems, culturally acceptable, accessible, economically fair and affordable; nutritionally adequate, safe and healthy; while optimizing natural and human resources” (FAO, 2012a). Box 9 Participatory agroecology research to address food security and nutrition in Malawi Using participatory education and agroecology in Malawi, thousands of rural families have seen dramatic improvements in maternal and child nutrition, food security, crop diversity, land management practices and gender equality. Central to the success of this long-term programme has been iterative, participatory, transdisciplinary research methods that used multiple measures to assess and improve farming and social change with participating farmers (Bezner Kerr and Chirwa, 2004; Nyantakyi- Frimpong et al., 2017). Agroecology education was integrated with nutrition and social equity issues through interactive, dialogue-based methods, such as recipe days, discussion groups and theatre (Satzinger et al. 2009; Bezner Kerr et al., 2016a; Bezner Kerr et al., 2018a). Peer-driven farmer-led methods mobilized communities to test and use agroecological practices such as legume intercrops, compost, agroforestry and crop diversification (Bezner Kerr et al., 2007; Bezner Kerr et al. 2018b; Owoputi et al., 2018). When farmers used more agroecological practices, such as the incorporation of nutrient-rich legumes into maize-based cropping systems, yields stabilized, fertilizer costs fell and soil cover increased (Snapp et al., 2010; Kangmennaang et al., 2017; Owoputi et al., 2018). Households using agroecological practices who participated in community education programmes had significant improvements in child growth, food security, maternal dietary diversity and self-reported health (Bezner Kerr et al., 2010; Nyantakyi-Frimpong et al., 2016a; Owoputi et al,. 2018). There was also evidence of improved gender and other forms of social equity in communities for households with HIV-positive family members (Bezner Kerr et al., 2016b, 2019; Nyantakyi-Frimpong et al., 2016b). In households where spouses began discussing farming practices with each other, there were higher levels of food security and dietary diversity. Farmers began to take more pride in their own experimentation, traditional knowledge and ability to mentor others (Bezner Kerr et al., 2018b). Some communities organized the sharing of seeds and agroecological knowledge, and reported greater resilience under conditions of poor rainfall due to improved soil quality (Bezner Kerr et al., 2018b, 2019). Key findings from the case study: • Farmer-to-farmer teaching and experimentation were the primary teaching approaches and were effective at sharing knowledge. • Unequal social relationships including gender inequalities were assessed, discussed and improved over time. • Relevant educational strategies were developed by local communities to address these inequalities in an iterative way. • Linking agroecology to FSN outcomes took at least two years before such outcomes were realized, and required transdisciplinary and participatory approaches. 1.3 Contribution of agroecological approaches to food security and nutrition for rural consumers in low-income countries Not only do agroecological practices contribute to FSN, but they also contribute to 10 of the 17 SDGs (UN, 2015) through integrated practices that cut across many areas (FAO, 2018a) and help address poverty and hunger, education, gender equality, decent work and economic growth, reduced inequalities, responsible consumption and production, climate action, life on land, and peace and justice. Along with the SDGs, agroecology can also contribute to the Koronivia Joint Work on Agriculture (KJWA) (St-Louis et al., 2018) on adaptation, soils, nutrient use, manure management and livestock systems (see points 2.c, 2.d, and 2.e of the KJWA), and help realize the aims of the Paris Climate Agreement, the CBD and the United Nations Convention to Combat Desertification (FAO, 2018a). Beyond yield and production, an assessment of the contribution of agroecological approaches to FSN needs to incorporate multiple metrics that take into account social, economic and environmental impacts of agriculture. Agroecological approaches could play an important role in securing sustainable diets for all now and in the future as part of a transition towards more sustainable food systems that enhance FSN (De Schutter, 2011, 2012; IPES-Food 2016, DeLonge et al., 2016). Numerous studies have found positive relationships between diversified farming systems (a key principle of agroecology), household dietary diversity and nutrition (Talukder et al., 2000; De Clerck, 2013; Oyarzun et al., 2013; Jones et al., 2014b; Khoury et al., 2014; Carletto et al., 2015; Kumar et al., 2015; Olney et al., 2015; Shively and Sununtnasik, 2015; Jones, 2017). Bliss et al. (2017) examined the diversified farming systems of 30 Nicaraguan households. For the farmers, dietary diversity was a driver for diversification on fields, and the related higher crop diversity, with difference in harvest time, meaning greater food availability throughout the year. In Southern Benin, Bellon et al. (2016) found a positive correlation between on-farm diversity and women’s dietary diversity score (DDS),19 as most of the food grown on-farm was for consumption rather than for sale. Jones et al. (2018) also found that on-farm agrobiodiversity was associated with more diverse and micronutrient-adequate diets among women in the Peruvian Andes. In a survey of 390 households in Mexico, Becerril (2013) found improved body mass indexes in households using the diversified traditional “milpa” system (intercropping of maize, beans and squash) compared to other households with less diversified farming systems. In their study on Guatemala's Mayan Achí people, Luna-González and Sørensen (2018) found that nutritional functional diversity and DDSs were positively correlated with higher crop and animal species diversity (derived from traditional intercropped milpa systems, home gardens, local market, wild gathering) but higher DDDs were not correlated with better child anthropometric status. Other factors, such as limited access to health care or safe water, may have prevented improved child growth. In northern Malawi, studies have shown that legume intercropping, along with a participatory approach sensitive to cultural values and promoting gender equality, enhanced both food and nutritional security (Bezner Kerr et al., 2016c; Nyantakyi-Frimpong et al., 2016b; see Box 9). These results are especially significant as many Malawian households experience food insecurity and malnutrition (Ecker and Qaim, 2011) leading to poor health outcomes including stunting in young children (FAO, 2014a). In Uzbekistan, Gotor et al. (2018), studying a programme related to conservation and use of fruit species, showed that families growing more fruit species consumed a greater proportion of fruits in their diets, increasing their dietary diversity. Dawson et al. (2013) showed that agroforestry practices exploit differences in the phenology of fruit-tree species to provide critical nutritional supplements (particularly Vitamins A, C and B6) and maintain dietary diversity throughout the year. They highlighted that their extensive root systems allow trees to store water, be productive and contribute to dietary diversity, even in dry environments, in seasons when herbaceous vegetation cannot survive without irrigation. In Machakos (Kenya), an average household can achieve year-round dietary diversity with 20 trees of ten species either dispersed throughout their farm (on borders, around the home and in fields) or in a 8×18 m² (0.015 ha) fruit orchard (Kehlenbeck and McMullin, 2015). In a survey of 368 coffee-producing households, Bacon et al. (2017) found that food security was enhanced for agroforestry farmers who grew more of their own food and incorporated more diversified production elements including fruit trees and red bean crops. However, the authors cautioned that not all diversification is equally beneficial to different parameters of FSN. A meta-analysis revealed a significant positive relationship between indicators of dietary quality of children under five and landscape-scale tree cover in Africa, associated with maximum fruit and vegetable consumption at an intermediate level of tree cover (45 percent), after which it declines (Ickowitz et al., 2014). Diversified production in home gardens with application of agroecological practices provides an avenue for FSN for poor households with limited access to food. Home gardeners in Ghana, using intercropping, seed-saving, organic manures and crop residues as well as domestic waste, contributed to greater food availability, food access and nutrient supply (Bagson and Naamwintome, 2012). Vijayalakshmi and Thooyavathy (2012) found similar results in a study looking at the impact of home gardens on women's nutrition. In a small-scale study of 12 households in Bangladesh, Ferdous et al. (2016) highlighted a drastic increase in vegetable consumption among households trained in the Rangpur model (a home-garden strategy based on seven production niches, 14 vegetables selected for year-round cultivation, fruit and locally adapted crops). After the intervention, vegetable intake almost doubled, with 55–79 kg/person/year produced compared to 21–30 kg/person/year before the intervention. Several studies revealed a positive impact of organic farming practices on FSN (Miyashita and Kayunze, 2016; da Silva et al., 2018; Kamau et al., 2018). Miyashita and Kayunze (2015), for example, found significant differences in terms of FSN when they compared organic and conventional farming in the United Republic of Tanzania. On the other hand, the study of 139 farming households by Kaufman (2015) in northern Thailand showed variable impact of organic agricultural systems on food security when compared to conventional counterparts. While organic farmers had slightly higher mean levels of food security and lower debt levels compared to conventional farmers, the findings ¹⁹ For more information about DDS see http://www.fao.org/3/a-i1983e.pdf were not statistically significant (Kaufman, 2015). The author concludes that greater support for viable markets for organic products is needed to translate into significant differences in FSN for organic producers (Kaufman, 2015). In contrast, some studies also showed no significant relationship with the application of agroecological practices and measured parameters related to FSN. For instance, on-farm diversification in Nigeria had no impact on household DDS for the poorest surveyed households, although greater diversity was evident with middle- and high-income households (Ayenew et al., 2018). In Kenya, Ng’endo et al. (2015). also found no significant correlation between agrobiodiversity and households’ FSN. 1.4 Contested areas and knowledge gaps in agroecology There is no common, consensual definition of what constitutes an agroecological approach shared by all the actors involved (practitioners, scientists, social activists). Complete agreement on all the aspects embedded in the plurality if approaches or on how it should contribute to transformation of food systems is also lacking. While this makes it hard to pin down exactly what is agroecology and what is not, it also provides a flexibility that allows agroecological approaches to develop in locally adapted ways. It is thus necessary to examine key contested areas and knowledge gaps, which is the aim of this section. 1.4.1 Political and social dimensions of food production Some scientists, food-system actors and social movements have diverging views about whether social and political dimensions of food production should be considered as an integral and indivisible part of agroecology, critical for agroecology to be transformative (Méndez et al., 2013; Rosset and Altieri, 2017; Sanderson Bellamy and Ioris, 2017; Giraldo and Rosset, 2018). De Molina (2013) argues that not recognizing the social and political implications of agroecology could lead to negative social, environmental and FSN consequences for marginalized groups who may be disadvantaged with a “business as usual” model of agricultural improvement. This argument is in line with research on how political, social and economic context determines how technology is employed in addressing FSN (Bezner Kerr, 2012; Gómez et al., 2013; Stone and Glover, 2017). Some authors have suggested distinguishing a political or transformative agroecology, considering political and social factors to address FSN at a broader scale, from a technically-focused agroecology at the field scale (Méndez et al. 2013; Sanderson Bellamy and Ioris, 2017). Attention has in particular been drawn to the importance of addressing context-specific gender and social inequalities and related labour and economic dimensions through agroecological approaches (Batello et al., 2019; Bezner Kerr et al., 2019). Other authors have noted that agroecology, when embedded within a larger food systems policy intervention or food sovereignty initiative, can have a positive impact on FSN (Kanter et al., 2015; Wittman and Blesh, 2017). A just food system (Pimbert and Lemke, 2018) addresses wages and working conditions within it (principle 10) creating a direct link to FSN. Improved livelihoods for farm labourers, producers, market intermediaries, entrepreneurs and processors may enable them to achieve higher incomes and, therefore, purchase food. Increased proximity of producers and consumers and re-embedded local food systems (principle 11) may contribute to improving local economies. For example, producers can profit by receiving a higher share of revenue if less is taken by intermediaries or actors over a long supply chain for marketing and distribution of produce. Also, local food enterprises and retailers can increase their price margins and become better linked and known to local consumers. An important point here is also that producers can respond more effectively to the real food needs and demands of local consumers. This latter point is strongly supported by social organizations, which foster greater participation and decision-making of food producers and consumers (principle 13). 1.4.2 Difficulty in providing labels: illustration through the convergence with organic agriculture As agreeing on a generic definition proves to be difficult, so it is to provide universal label and labelling mechanisms. Yet, some initiatives by stakeholder groups and companies are under way. One form of certification that is suggested is the participatory guarantee system (PGS), in which certification is done using a democratic process involving producers, scientists and consumers (see Box 32 in Appendix 1). This may also generate difficulties when addressing distinction with other innovative approaches. There is, for example, a growing debate about similarities, difference and convergence of organic agriculture and agroecology (Migliorini and Wezel, 2018). There is also related debate about whether synthetic pesticides and chemical fertilizers should be excluded from agroecological production, as under organic agriculture (with a few exceptions), or be acceptable to a certain degree or in defined situations. 1.4.3 Can agroecology feed the world? Some people think that farmers cannot feed the world with agroecology, others contend that it is impossible to feed the world in the future without agroecology. These echo divergent views on whether organic farming could feed the global population (De Ponti et al., 2012, Muller et al., 2017). It is commonly estimated that an increase in agricultural production will be required to feed a growing global population, expected to reach 9.7 billion by 2050, unless major changes are made in global food systems (HLPE, 2016; Berners-Lee et al., 2018; Le Mouël et al., eds, 2018), especially in Africa (van Ittersum et al., 2016). Estimates vary depending on whether food losses and waste, urbanization and changing diets, non-food uses (animal feed, biofuels and others) are considered in the modelling (HLPE, 2013b, 2014; Kahane et al., 2013; Keating et al., 2014; Berners-Lee et al., 2018; Le Mouël et al., 2018; Keating and Carberry, 2010; Alexandratos and Bruinsma, 2012; Valin, 2014). FAO (2017b) estimated that global agricultural production would have to increase by almost 50 percent between 2012 and 2050. However, the need for such an increase in agricultural production is contested, as the earlier assumptions are challenged: some estimates indicate that enough food is produced today to potentially feed 9 billion people (IPES-Food, 2016; HLPE, 2014, 2017b; Chappell, 2018) or even 9.75 billion (Berners-Lee et al., 2018). The debate as to whether or not agroecology can feed the world may be based on a false premise since, despite high levels of production, food insecurity and malnutrition still persist today (Chappell, 2018; HLPE, 2016, 2017b), even in food-exporting countries such as Brazil and South Africa (FAO et al., 2017). Today, almost one-third of food produced for human consumption is either lost or wasted, yet different forms of malnutrition coexist in most countries (HLPE, 2014, 2017b). Globally, around 820 million people are still hungry (FAO et al., 2018), about 2 billion are overweight or obese (Ng et al., 2014) and an estimated 2 billion people suffer from malnutrition caused by micronutrient deficiencies (iron, iodine, vitamin A, folate and zinc) (HLPE, 2017b). FAO (2018e) found that a “business as usual” scenario is likely to lead to significant undernourishment by 2050 even if gross agricultural output increases by 50 percent. On the contrary, alternative “towards sustainability” scenarios, through more balanced diets, more sustainable production and consumption patterns, as well as through fairer food and income distribution, in line with agroecological approaches, could lead to a drastic reduction in undernourishment and improvement of nutritional security even if agricultural production increases only by around 40 percent. Increasing production alone might thus not be sufficient to achieve FSN in its four dimensions (availability, access, utilization and stability) (FAO, 2018b). There is a growing awareness that hunger and malnutrition may not be only a matter of food production, but mainly of different entitlements, leading to unequal access to food, to natural resources (land, water, genetic resources), inputs, markets and services (Sen, 1981; Smith and Haddad, 2015; HLPE, 2017b). Previous HLPE reports have extensively discussed the issues raised by inequalities in access to food and resources (see in particular: HLPE, 2011a,b, 2012, 2013a, 2015, 2016, 2017c). Therefore, agroecological approaches are presented as promising avenues to achieve FSN, since they do not consider productivity alone and suggest addressing social inequalities and power asymmetries (Massett et al., 2011 ; Kanter et al., 2015; HLPE, 2018), including gender and ethnic minority inequalities (Massicotte, 2014; Bezner Kerr et al., 2019). Furthermore, ”feeding the world” is sometimes framed as a question of calories or production, and includes debates about the nutritional implications of different farming systems (HLPE, 2017b). Meeting kilocalorie energy requirements, however, does not translate automatically into nutritional security (Pingali, 2015; Traore et al., 2012; Keating et al., 2014), as some forms of calorie consumption (e.g. foods with high sugar, salt or fat content) can worsen nutritional status (HLPE, 2017b). FSN indicators now go beyond calorie count, and include measures of child growth, diet quality and reported experience of food insecurity at the individual and household level (Arimond et al., 2010; Carletto et al., 2012). In many parts of the world, the so-called “industrial” agriculture model relying on the intensive use of fossil fuel and chemical inputs has resulted in increased agricultural productivity at the expense of loss of biodiversity, land degradation, loss of soil fertility and chemical contamination of soil and water, with major consequences on human, animal and planetary health (Kremen and Miles, 2012). A number of recent studies suggest that industrial agriculture cannot ensure sustainable food systems for FSN in the long term because of these negative impacts (Campbell et al., 2017; Frison et al., 2011; IPES-Food, 2016; Mahon et al., 2017; Kremen and Merenlender, 2018). The consequences of such production systems in terms of diet imbalance are increasingly a polemical issue, which calls for increased attention from consumers (HLPE, 2017b). Moreover, several studies challenged the common idea that agroecological systems are less productive than more “conventional” or “industrial” agricultural models (intensive and specialized), and thus cannot make a large contribution to feeding the world. For example, Poux and Aubert (2018) have recently modelled the potential for agroecological approaches (including elimination of pesticides and synthetic fertilizers, shifting to healthier diets, and development of hedges, trees, ponds and other habitats for increased biodiversity) to feed Europe. They estimated that production would decline by 35 percent but that food requirements for Europe and for the export market in cereals, dairy and wine would be maintained, greenhouse gas emissions would be reduced by 45 percent and biodiversity and natural resources improved. Pretty et al. (2003), De Schutter (2010, 2012, Ponisio et al. (2015) and Reganold and Wachter (2016) summarized many examples, mainly from tropical and subtropical countries, showing significant yield increases associated with agroecological or organic farming. Pretty et al. (2003) showed that the weighted average increases were 37 percent per farm and 48 percent per hectare, while d’Annolfo et al. (2017) showed in their meta analysis that, following the adoption of agroecological practices, yields increased in 61 percent of the cases analysed and decreased in 20 percent, while farm profitability increased in 66 percent of cases. Given the underinvestment in agroecological research noted below, it remains unclear how representative the cases so far documented are and which aspects of the agroecological approaches adopted were responsible for yield and profit improvements. 1.4.4 Knowledge systems There are debates around the role and contribution of indigenous and local food producers in knowledge generation and the significance of cultural context for shaping this knowledge, including the role of women, elders, ceremonies, community organizations and opportunities for interaction with scientists (IAASTD, 2009; Etkin, 2006; Méndez et al., 2013; Snapp and Pound, eds, 2017; IIED, 2018). Local knowledge is used here to refer to the knowledge held by a defined group of people (Sinclair and Walker, 1999). It embraces traditional knowledge (passed down from one generation to the next), indigenous knowledge that is culturally bound and locally derived knowledge from contemporary learning based on local observation and experimentation (Sinclair and Joshi, 2004). Some argue that traditional knowledge is deep, but narrow, while scientific knowledge is broad but shallow, and that agroecology involves the co-production of knowledge through the mutual shaping of different knowledge streams (Vandermeer and Perfecto, 2013). Scholars and indigenous groups also debate the notion that local knowledge is ”new” scientific knowledge and caution about the dangers of such knowledge being separated out from other social–ecological knowledge (Barthel et al., 2013; Massicotte, 2014; IIED, 2018). There is mounting evidence that much local agroecological knowledge is dynamic, based on contemporary observation and experimentation by farmers and comparable with and largely complementary to global scientific knowledge (Richards, 1985; Sinclair and Walker, 1999; Thorne et al., 1999; IAASTD, 2009;Cerdán et al., 2012; Kuria et al., 2018). While some agroecological knowledge is widely held by people living in a particular locality (Joshi et al., 2004), in other cases different people within communities may have different interests and opportunities to observe agroecological processes leading to marked differences in knowledge according to gender or other forms of social differentiation (Crossland et al., 2018). Debates about the roles of farmers and social movements in agroecological knowledge and agroecological research relate to the potential to ”scale-out” agroecology effectively (Pimbert, ed, 2018a). Several scholars and social movements in the ”political agroecology” stream have emphasized the significance of democratic processes in agroecological knowledge generation, with the process of small food producer-led, decentralized, autonomous knowledge generation being as important as the specific technical knowledge being generated through more formal scientific approaches (Massicotte, 2014). Researchers have also pointed to the need for agroecology to explicitly address gender, ethnic minorities and other social inequalities in order to effectively impact FSN (Massicotte, 2014; Bezner Kerr et al., 2019). These issues can create tensions between social movements and scientists. This may happen when the way science generates knowledge and judges its validity is not respected, when ethics and social control of scientific production are not addressed, when the contribution of non-academic actors in knowledge production is not considered. This is especially the case when investment decisions are being made and power imbalances exist. Consideration of these situations has led to explicit attempts to bridge across different knowledge systems (Mendez et al., 2013; Tengö et al., 2014). 1.4.5 Knowledge gaps The severely limited public investment in agroecological approaches, estimated at between 1 percent and 1.5 percent of total agricultural and aid budgets, partly explains the remaining knowledge gaps (DeLonge et al., 2016; Miles et al., 2017; Pimbert and Moeller, 2018). Most private and public investments in agricultural research over the last 50 years were primarily based on “Green Revolution” technologies (including agrochemicals and mechanization) and, in particular, on genetics (Vanloqueren and Baret, 2009; DeLonge et al., 2016; Miles et al., 2017; Pimbert and Moeller 2018). For example, in the United Kingdom, aid for agroecological projects represents less than 5 percent of agricultural aid and less than 0.5 percent of its total aid budget since 2010 (Pimbert and Moeller, 2018). In the United States of America, research and development related to diversified systems – a major avenue for agroecological systems – amounts to less than 2 percent of public agricultural research funding (Carlisle and Miles, 2013). FAO estimates that 8 percent of their 2018–2019 work contributes to agroecological transitions (FAO, 2018f). In addition, the majority of teaching and research institutions, and extension services, have been oriented to the so-called “industrial” agriculture rather than to the promotion of agroecological technologies. Typical education programmes in agronomy are mostly oriented towards single solution problem solving in conventional agriculture. There is now a growing number of education programmes that bring more systemic and holistic approaches, as well as experiential learning, into focus (Francis et al., 2011, 2017). Therefore, comparisons between agroecological approaches and the dominant “industrial” agriculture model need to consider the funding bias skewed against agroecological research, education and extension (Delonge et al., 2016; Pimbert and Moeller, 2018). Two key knowledge gaps are how to effectively link agroecology to public policies to address FSN (Sabourin et al., 2018), and what are the economic and social impacts of agroecology for different groups in communities, including labour costs and FSN (Sanderson Bellamy and Ioris, 2017; Bezner Kerr et al., 2019). Assessing the yield gap between “industrial” and agroecological systems is an active area for research. Although several studies suggest that there are comparable yields, higher yield stability, particularly under extreme weather conditions, and increased profitability for those using agroecological methods, further research is required, in a wider range of socio-ecological conditions (d’Annolfo et al., 2017; Sanderson Bellamy and Ioris, 2017). How to scale out agroecological approaches in ways that foster democratic processes and address the needs of marginalized groups is also lacking, with some evidence for context-specific methods being effective at addressing FSN and SFSs, if political and economic barriers are addressed (IPES- Food, 2016; Mier y Terán et al., 2018; Sinclair and Coe, 2019). The design of resilient agricultural systems is an imperative to cope with climate change and increased climate variability. Resilience is particularly important in areas most likely to be affected by extreme climate events such as prolonged droughts, floods and heavy winds (Ching et al., eds, 2011; Koohafkan et al., 2012; Rhodes, 2013; Scialabba and Müller-Lindenlauf, 2010; Altieri et al., 2015). Holt-Giménez (2002) indicated that agroecological systems are more adapted to such context, and might even help mitigate the effects of climate change. However, further research is needed to better understand the processes that support more resilient systems in different contexts. Many gaps remain in terms of how to support such transitions and what are the key barriers to overcome (Gliessman, 2016; Côte et al., 2019). Several ”lock-ins” that may prevent the transition towards agroecological