Land use and global food security in 2050
Summary
Uncertainties about the capacity of the planet to feed a growing population in a context of climate change and on-going debates within the scientific community regarding land-use change trajectories led CIRAD and INRAE to conduct a new foresight study on ‘’Land use and food security in 2050’’. Using a morphological analysis approach, 5 contrasting scenarios of land use and food security by 2050 were developed.
The first three scenarios are based on current competing trends identified in most world regions.
- ‘Metropolization’ links the development of megacities with a nutrition transition led by global agri-food companies selling ultra-processed foods and characterized by an increase of animal products consumption, in a global context of development through market forces and rapid climate change. Small farmers are marginalized.
- ‘Regionalization’ relates the increase of medium-size cities and their networking with rural areas to the emergence of regional food systems based on family farming and traditional food diets, and a set of regional agreements, notably trade agreements.
- ‘Households’ links strong individual mobility between rural and urban areas and the development of non-farm employment to the emergence of hybrid diets based on both traditional and modern value chains, in a globalized world where family farms and cooperatives are major actors in land use.
The last two scenarios involve potential breaks that could change the entire land use and food security system.
- ‘Healthy’ assumes that due to the increasing cost of malnutrition, a radical move towards healthy diets occurs driven by global cooperation and public policies in a context of climate change stabilization, involving a re-configuration of agricultural systems backed by new alliances between stakeholders.
- ‘Communities’ assumes that in a context of recurrent crises, development based on small towns and rural communities occurs, focusing on managing common property in agriculture in order to ensure food security.
For each scenario, variants have been applied reflecting either alternative diet transition (‘Metropolization’) or alternative production technology transition (‘Regionalization’, ‘Healthy’, ‘Communities’).
Agrimonde-Terra foresight method
Agrimonde-Terra’s foresight method is a coupled approach combining scenario building and quantitative simulations.
Scenario building
In contrast to most existing sets of global scenarios, Agrimonde-Terra’s scenarios were not developed along two axes that characterize the major and most uncertain driving forces or outcomes regarding the future of the system concerned. Instead, we used a morphological analysis that provides a multidimensional systemic representation.
With the morphological approach, first the system under study and its main drivers are defined, then alternative assumptions of change are elaborated for each driver. The morphological table sets together these alternative assumptions per driver and thereby helps visualize and explore combinations of driver assumptions. Scenarios are retained combinations of assumptions.
As far as Agrimonde-Terra is concerned, the land use and food security system and its main drivers were first defined. Agrimonde-Terra considers that land-use changes result from complex interactions between direct and external drivers and affect food security. The direct and external drivers considered are, respectively: cropping systems, livestock systems, farm structures, and urban-rural relationships (including urbanization); global (political, economic and social, including demography) context, climate change and food diets.
Several expert groups were involved in the scenario building process. They provided knowledge and assessments about current trends and possible changes, and built collective intelligence about alternative futures. The whole study involved around 80 international experts, including scientific expert groups in the early stages (thematic workshops) and a scenario committee composed of 19 members (either scientists or stakeholders from international and national institutions as well as civil society) that provided guidance on scenario building (see Contributors).
Five contrasting scenarios were built based on extensive discussions between academic researchers and stakeholders in the scenario committee. Each scenario describes the situation in land use and food security in 2050 and is developed into a narrative. Scenario narratives were drafted by the project team and discussed among experts of the scenario committee.
Scenario simulation
The quantitative impacts of scenarios in terms of land use and agricultural production and trade have been analyzed and discussed through an iterative process with the scenario committee.
The project team developped the GlobAgri platform and a database and a biomass balance model were specifically designed (named GlobAgri-AgT). The database and model include 33 agri-food products and 5 forage products, and divided the world into 14 broad regions (see definitions). The reference year (named “2010”) is the 2007-2009 average and the simulation horizon is 2050.
The starting point is the morphological table, which reports the alternative assumptions of change for each driver. These qualitative assumptions are first translated into quantitative model inputs. This involves establishing detailed translation matrices between global qualitative assumptions and model input levels for each agri-food product and each world region. These translation matrices are based on experts’ opinions and existing literature. Once all qualitative assumptions for all the drivers have been translated into quantitative inputs for the model, scenarios may be simulated. Several assumptions of change patterns may co-exist for several drivers in our scenarios. When relevant, the same scenario was simulated using alternative assumptions of change for one or two drivers. Such scenarios variants concern either food diets or cropping and livestock systems.
Finally, given that GlobAgri-AgT cannot deal with the main specificities of the ‘Households’ scenario (mobility, multi-activity and networking), no quantitative results were provided for this scenario.
Selecting above tabs, you can explore assumptions and results produced by the GlobAgri-AgT model. The map below presents the composition of the 14 broad world regions covered by this study.
Source: INRAE / CIRAD
For each input variable of the model (population, diets, cultivable area, area devoted to energy crops and yields), the graph shows the state of the variable in “2010” and its projection in 2050 for the 14 regions of the world, according to the assumptions made for the different scenarios. This visualization allows you to compare the regional situations in “2010” or in 2050 for each scenario.
Population
Diets
Cultivable_Area
Energy_Area
Yield
All Products
Animal Products
Vegetal Products
Oil Products
Soy
Graph title (please don't delete)
Legend items
| Aquatic animal | Freshwater, demersal, pelagic and other marine fish; crustaceans, cephalopods and other molluscs; meat aquatic mammals and other aquatic animals |
| Aquatic feed | (Definition ?) |
| Bovine | Bovine meat |
| Small ruminant | Sheep and goats meat |
| Pork | Pork meat |
| Poultry | Poultry meat |
| Poultry (other) | (Definition ?) |
| Eggs | Eggs |
| Dairy | Dairy products |
| Grass | Permanent meadows and pastures |
| Grass-like forage | Temporary meadows and pastures (mixed grass and ray-grass) |
| Other forages | Cultivated forages (alfalfa, beets, legumes, maize, etc.). |
| Fibers | Jute, jute-like fibres, soft-fibres other, sisal, abaca, hard fibres other, tobacco, rubber and seed cotton |
| Roots and Tuber | Potatoes, cassava, sweet potatoes, yams and other roots |
| Fruits & vegetables | Tomatoes, onions, vegetables other, oranges, mandarines, lemons, limes, grapefruit, citrus other, bananas, plantains, apples, pineapples, dates, grapes and other fruits |
| Maize | Maize |
| Wheat | Wheat |
| Rice | Rice, paddy equivalent |
| Other cereals | Barley, rye, oats, millet, sorghum and other cereals |
| Pulses | Beans, peas and other pulses |
| Soyabeans | Soyabeans |
| Soyabean Cake | Soyabean cake |
| Soyabean Oil | Soyabean oil |
| Sunflowerseed | Sunflowerseed |
| Sunflowerseed Cake | Sunflowerseed cake |
| Sunflowerseed Oil | Sunflowerseed oil |
| Rape and Mustardseed | Rape and mustardseed |
| Rape and Mustard Cake | Rape and mustard cake |
| Rape and Mustard Oil | Rape and mustard oil |
| Other Oilcrops | Groundnuts (shelled eq), coconuts – incl copra, sesameseed, olives and other oilcrops |
| Cake Other Oilcrops | Other oilcrops cake |
| Oil Other Oilcrops | Other oilcrops oil |
| Oilpalm fruit | Oilpalm fruit |
| Palmkernel Cake | Palm kernel cake |
| Palm Products Oil | Palm oil and palmkernel oil |
| Sugar | Sugar cane, sugar beet (sugar in equivalent sugar cane and beet) |
| Other plant products | Nuts, coffee, cocoa beans, tea, pepper, pimento, cloves, spices, other |
| Crop residues | (Definition ?) |
| Other products | Meat other, offals edible, fats animals raw, honey, meat meal, aquatic plants |
| Occasional | (Definition ?) |
Population in million land area in million hectares
Source: from FAOSTAT 2017
Definitions
The 4 simulated scenarios and their variants are described below.
| Init | Initial |
| Metr_A | Metropolization with transition to diets based on animal products |
| Metr_U | Metropolization with transition to diets based on ultraprocessed products |
| Heal_C | Healthy with agricultural technology C (sustainable intensification for cropping systems + agroecological livestock) |
| Heal_D | Healthy with agricultural technolgy D (agroecology for cropping systems + agroecological livestock) |
| Reg_A | Regionalization with agricultural technology A (sustainable intensification for cropping systems + conventional intensification with local resources for livestock systems) |
| Reg_B | Regionalization with agricultural technology B (agroecology for cropping systems + agroecological livestock) |
| Com_AE | Communities with Agroecology (agroecology for cropping systems + agroecological livestock) |
| Com_col | Communities with Collapse (collapse of crop production + backyard livestock) |
The Agrimonde-Terra uses a regional breakdown of 14 regions, detailed below.
| US-CA | Canada/USA |
| BR-AR | Brazil/Argentina |
| RoAm | Rest of America |
| EU | EU27 |
| FSU | Former Soviet Union |
| N-ME | Near and Middle East |
| N-AF | North Africa |
| W-AF | West Africa |
| ECSA | ECS Africa |
| CN | China |
| IN | India |
| RoAs | Rest of Asia |
| OCEA | Oceania |
| RoW | Rest of the World |
The model establishes a balance (in tonnes) for 33 agri-food products and 5 forage products. Considered products are reported below.
| Aquatic animal | Freshwater, demersal, pelagic and other marine fish; crustaceans, cephalopods and other molluscs; meat aquatic mammals and other aquatic animals |
| Bovine | Bovine meat |
| Small ruminant | Sheep and goats meat |
| Pork | Pork meat |
| Poultry | Poultry meat |
| Eggs | Eggs |
| Dairy | Dairy products |
| Grass | From permanent meadows and pastures |
| Grass-like forage | From temporary meadows and pastures (mixed grass and ray-grass) |
| Other forages | Cultivated forages (alfalfa, beets, legumes, maize, etc.). |
| Fibers | Jute, jute-like fibres, soft-fibres other, sisal, abaca, hard fibres other, tobacco, rubber and seed cotton |
| Roots and Tuber | Potatoes, cassava, sweet potatoes, yams and other roots |
| Fruits & vegetables | Tomatoes, onions, vegetables other, oranges, mandarines, lemons, limes, grapefruit, citrus other, bananas, plantains, apples, pineapples, dates, grapes and other fruits |
| Maize | Maize |
| Wheat | Wheat |
| Rice | Rice, paddy equivalent |
| Other cereals | Barley, rye, oats, millet, sorghum and other cereals |
| Pulses | Beans, peas and other pulses |
| Soyabeans | Soyabeans |
| Soyabean Cake | Soyabean cake |
| Soyabean Oil | Soyabean oil |
| Sunflowerseed | Sunflowerseed |
| Sunflowerseed Cake | Sunflowerseed cake |
| Sunflowerseed Oil | Sunflowerseed oil |
| Rape and Mustardseed | Rape and mustardseed |
| Rape and Mustard Cake | Rape and mustard cake |
| Rape and Mustard Oil | Rape and mustard oil |
| Other Oilcrops | Groundnuts (shelled eq), coconuts – incl copra, sesameseed, olives and other oilcrops |
| Cake Other Oilcrops | Other oilcrops cake |
| Oil Other Oilcrops | Other oilcrops oil |
| Oilpalm fruit | Oilpalm fruit |
| Palmkernel Cake | Palm kernel cake |
| Palm Products Oil | Palm oil and palmkernel oil |
| Sugar | Sugar cane, sugar beet (sugar in equivalent sugar cane and beet) |
| Other plant products | Nuts, coffee, cocoa beans, tea, pepper, pimento, cloves, spices, other |
| Crop residues | Stover |
| Other products | Meat other, offals edible, fats animals raw, honey, meat meal, aquatic plants |
| Occasional | Food leftovers, cut-and-carry, forages and legumes, roadside grasses |
Key finding
Land-use changes (1)
Most Agrimonde-Terra scenarios lead to an expansion of the world’s agricultural land area. Hence, according to our hypotheses, the land-using effects of increased world food consumption (following increased world population, either reinforced or alleviated by changing food diets) tend to exceed the land-saving effects of increased performances
in world agricultural production (following increased crop yields and improved livestock feed-to-output ratios).
Land-use change (2)
The extent of world agricultural land expansion
varies widely across scenarios. It is particularly high for scenarios involving either stagnation of crop yields and deterioration
of livestock production performances (+41% for “Communities with collapse”) or a huge increase in animal product consumption (+27% for “Metropolization with animal products”). It is far more limited, even close to zero, for scenarios involving either reduced calories availability in food regimes (+3% for “Communities with agroecology”) or a limited increase in consumption of animal products joined with the substitution from ruminant to monogastric meat in meat consumption (+0.6% for “Healthy with agricultural technology C” and -1% for “Metropolization
with ultra-processed products”)
Land-use change (3)
Only the scenarios “Metropolization with ultra-processed products” and “Healthy with agricultural technology C” are able to produce sufficient food for the expected growing population up to 2050 without further major deforestation at the world scale. Under all the other scenarios, ensuring food availability in 2050 would require large areas of deforestation all over the world. In some cases, deforestation is so huge that the corresponding scenarios may be considered
clearly unsustainable in 2050: the Com_Col and Metr_A scenarios for example.
Agricultural productivity
Ensuring world food availability in 2050 without significant further deforestation will imply a major increase in the performances of agricultural systems in some regions, notably in India, West Africa and ECS Africa. Indeed, these are the three regions where the agricultural land area expands significantly under all scenarios. This is not the case for other regions where agricultural land may expand or decrease according to the scenarios.
Two scenarios are clearly not able to ensure world food security
Metropolization
– overweight, obesity and diet-related diseases
– degradation of resources and sensitive to climate change
– increased instability on world agricultural markets
– increased spatial and economic inequalities
Communities
– serious tensions on land and degradation of resources
– rebuilding local food systems based on agroecological cropping and livestock systems could be en option but …
- -10% decrease in daily calories availability per capita
- food access difficulties in urban areas
Two scenarios are likely able to ensure world food security, but under conditions
Healthy
– healthier diets, less overnutrition and related deseases, but also undernutrition
– limited agricultural land area expansion at world level
– protection/restoration of natural resources
– but potential tensions between food security and CC mitigation objectives
Regionalization
– contributes to reducing overnutrition and related deseases
– development of agri-food industries in small and medium-sized cities, contributes positively to rural development, rural employment, rural incomes
– but significant land expansion
– clearly unsustainable for some regions
Authors
Chantal Le Mouël (INRAE, SMART), Marie Delattre-Gasquet (CIRAD), Olvier Mora (INRAE, DEPE), Patrice Dumas (CIRAD, CIRED), Agneta Forslund (INRAE, SMART) Stéphane Manceron (INRAE, DEPE), Catherine Donnars (INRAE, DEPE), Olivier Rechauchère (INRAE, DEPE), with the collaboration of Thierry Brunelle (CIRAD, CIRED), Elodie Marajo-Petitzon (INRAE, SMART), Clémence Moreau (CIRAD), Marco Barzman (INRAE, DEPE) and Pauline Marty (INRAE, DEPE)
Contributors
Agrimonde-Terra has benefitted from the knowledge, experience and support of:
Members of the Scenario Advisory Committee who followed the work throughout the foresight process:
Agnes Andersson Djurfeldt (Lund University, Sweden), Leïth Ben Becher (Synagri, Tunisia), Mohamed Elloumi (INRAT, Tunisia), Adama Faye (IPAR, Senegal), Richard Guissou/Yves-Gérard Bazie (Ministry of Agriculture and Food Security, Burkina Faso), Holger Kray (World Bank, USA), John Lewis (Terra Global Capital, USA), Patrick Meyfroidt (University of Louvain, Belgium), Marc Mueller (FAO, Italy), Siwa Msangui (IFPRI, USA), Antonio Onorati (International Planning Committee for Food Sovereignty, Italy), Stéphane Parmentier (Oxfam, Belgium), Alain Retière (Cap 2100, France), Roberta Sonnino (Cardiff University, UK), Sébastien Treyer (IDDRI, France), Dominique van der Mensbrugghe (AgMIP, Purdue University, USA), Joost Vervoort (CCAFS and Oxford University, UK), Hisham Zehni (IFAD, Italy).
Scientific coordinators of the thematic workshops:
Francis Aubert (AgroSup Dijon), Frédéric Lançon (CIRAD), Jacques Marzin (CIRAD), Laurent Piet (INRAE), David Makowski (INRAE), Eric Malézieux (CIRAD), Florent Maraux (CIRAD), Philippe Lecomte (CIRAD), Alexandre Ickowitz (CIRAD), Philippe Lescoat (AgroParisTech).
80 experts who participated in the thematic workshops.
Partners
INRAE, CIRAD