Food

Introduction

The Food Sector includes agricultural production (crops and livestock) as well as food preparation, consumption, and waste. This essential human activity is responsible for a major share of greenhouse gas emissions today: crop and livestock production is the source of about 1/8 of anthropogenic emissions. Land clearing (which is mostly for agriculture) is the source of another 1/8 of emissions (IPCC, 2014). Many of Project Drawdown’s supply-side agricultural solutions reduce emissions from farming and ranching, while also sequestering significant amounts of carbon. Demand-side solutions like a plant-based diet and reduced food waste reduce the need for land clearing.

Solutions

Supply-Side Solutions:

• Biochar – a biosequestration process for converting biomass to long-lived charcoal (and energy) which can be used as a soil amendment.
• Conservation agriculture – an annual crop production system that provides biosequestration via crop rotation, cover cropping, and reduced tillage.
• Farmland irrigation – a set of energy-efficient irrigation practices that increase crop yields while reducing emissions.
• Farmland restoration – a set of processes for restoring degraded, abandoned land to productivity and biosequestration.
• Improved rice cultivation – a set of practices to reduce methane emissions from paddy rice production using alternate wet and dry periods and other strategies.
• Managed grazing – a set of practices that sequester carbon in grassland soils by adjusting stocking rates, timing, and intensity of grazing.
• Multistrata agroforestry – a perennial cropping system featuring multiple layers of trees and other perennial crops, with high biosequestration impacts.
• Nutrient management – a reduction in the overuse of synthetic nitrogen fertilizers, resulting in reduced emissions of nitrous oxide.
• Regenerative agriculture – an annual crop production system that includes at least four of the following practices: green manure, compost application, organic production, cover crops, crop rotation, and/or reduced tillage.
• Silvopasture – the addition of trees to pastures for increased productivity and biosequestration.
• System of Rice Intensification – an improved smallholder rice production technique that uses wider spacing non-flooded periods, compost application, and other strategies for emissions reduction and improved yields.
• Tree intercropping – an annual crop production system that integrates trees for increased yields, ecosystem services, and biosequestration.
• Tropical staple trees – the production of trees that produce staple crops (starch, protein, oils), to replace some annual cropping with trees providing biosequestration.

Demand-Side Solutions:

• Clean cookstoves – the use of efficient cookstoves that reduce deforestation and improve health in regions where firewood is the main cooking fuel.
• Composting – the conversion of biodegradable waste to a useful soil amendment, while avoiding emissions from landfills.
• Plant-rich diet – reduced emissions associated with reduced livestock production by emphasizing plant-based foods in wealthy countries, while increasing food security and healthy diets. Avoids emissions from land clearing for agriculture by reducing demand.
• Reduced food waste – reducing emissions from agriculture by using its products more efficiently, including redistribution of food before it is wasted. Avoids emissions from land clearing for agriculture by reducing demand.

Methodology and Integration

Modeling Methodology

Each solution in the Food Sector was modeled individually, and then integration was performed to ensure consistency across the sector and with the other sectors. Information gathered and data collected are used to develop solution-specific models that evaluate the potential financial and emission-reduction impacts of each solution when adopted globally from 2020-2050. Models compare a Reference Scenario, that assumes current adoption remains at a constant percent of the current total land area, with high adoption scenarios assuming a reasonably vigorous global adoption path. In doing so, the results reflect the full impact of the solution, i.e. the total 30-year impact of adoption when scaled beyond the solution’s current status.

Figure 1: Project Drawdown Food Sector Framework

Total Land Area

Supply-side Food Sector models define the Total Land Area as the area of land (in hectares) suitable for adoption by solutions. Data on global land is acquired from Global Agro-Ecological Zones database, developed by the Food and Agriculture Organization of the United Nations (FAO) and the International Institute for Applied Systems Analysis (IIASA). The Drawdown Land-Use Model categorizes and allocates land according to agro-ecological zones based on the following factors: thermal climate, moisture regimes, soil quality, slope, cover type, and degradation status. These characteristics influence the suitability of different practices, and solution adoption scenarios are restricted by one or more of these factors.

• Thermal climate – includes tropical, temperate, and boreal (high latitude or high elevation). Several of the agricultural solutions with the most powerful mitigation impact, like tropical staple trees and multistrata agroforestry, are limited to tropical or even tropical humid climates. Temperate forest and tropical forest are likewise limited by climate.
• Moisture regime – includes humid (1000 millimeters of rainfall or more per year), semi-arid (250-1000 millimeters per year), and arid (0-250 millimeters of rainfall per year). Several solutions are constrained by rainfall; for example, bamboo is only suited to humid climates, whether tropical or temperate.
• Soil quality – includes prime, good, and marginal. While no solutions are limited to soils of a given quality, yields decrease in good and marginal soils.
• Slope types – include minimal, moderate, and steep. Moderate and steep slopes are more vulnerable to erosion due to tillage, and are also difficult to work with mechanized equipment due to risk of rollovers. Thus, fully perennial solutions, while applicable on all slopes, are particularly advisable for steeper slopes.
• Current cover – includes forest, grassland, rainfed cropland, and irrigated cropland. Certain practices like farmland irrigation are only suited to irrigated cropland. Forest protection is only suited to currently forested areas.
• Degradation status – each agro-ecological zone is designated degraded or non-degraded. Degraded zones are suitable for restoration and show lower yields. For example, non-degraded grassland is ideal for managed grazing and silvopasture, while degraded grassland may be restored via afforestation or perennial biomass production. Degraded forest is suited to restoration via temperate forests or tropical forests.

Demand-side solutions utilized other model structures. Composting uses the Drawdown Reduction and Replacement Solutions (RRS) Model to evaluate the global organic municipal solid waste stream; clean cookstoves uses the RRS model to evaluate terawatt hours (therms) of heating energy for cooking. Reduced food waste and plant-rich diet use fully customized models due to the inherent complexity in measuring food demand. These models use over 450,000 data points relating to country-specific consumption patterns of different commodities (see reduced food waste and plant-rich diet for more details).

Adoption Scenarios

Three general Project Drawdown scenario were developed for the Food Sector:

• Plausible Scenario: this scenario represents incremental growth of food solutions using a high adoption trajectory to 2050.
• Drawdown Scenario:  this scenario is optimized to reach drawdown by 2050 using more ambitious projections.
• Optimum Scenario: this scenario represents the most optimistic case, in which supply-side solutions capture nearly 100 percent of allocated land. The scenario assumes that 100 percent of the global population adopts a plant-rich diet and food waste is reduced to zero.

Each solution model uses unique adoption trajectories evaluated based on meta-analyses of existing prognostications of solutions, extrapolations from historical data, or scenario analyses depending on the availability of global and regional data.

Integration

Agricultural Production Clusters

Drawdown’s approach seeks to model integration between and within sectors, and avoid double counting. Several tools were developed to assist in this effort. The Agroecological Zone (AEZ) model categorizes the world’s land by: current cover (e.g. forest, grassland, cropland), thermal climate, moisture regime, soil quality, slope, and state of degradation.  Both Food (supply-side) and Land Use solutions were assigned to AEZs based on suitability. Once current solution adoption was allocated for each zone (e.g. semi-arid cropland of minimal slopes), zone priorities were generated and available land was allocated for new adoption. Priorities were determined based on an evaluation of suitability, consideration of social and ecological co-benefits, mitigation impact, yield impact, etc. For example, Indigenous peoples’ land management is given a higher priority than forest protection for AEZs with forest cover, in recognition of indigenous peoples’ rights and livelihoods. Multistrata agroforestry is highly prioritized in tropical humid climates due to its high sequestration rate, food production, and highly limited climate constraints.

Each unit of land was allocated to a separate solution to avoid overlap between practices. The exception to this are farmland irrigation, nutrient management, and women smallholders, which can be implemented in addition to other practices. The constraint of limited available land meant that many solutions could not reach their technical adoption potential. The AEZ model thus prevents double-counting for adoption of agricultural and land use solutions.

Food Demand Cluster

Total food demand is calculated based on global population (from the family planning and educating girls solutions from the Women and Girls Sector) and dietary trends (from the plant-rich diet and reduced food waste solutions from within the Food Sector). The adoption of plant-rich diet intersects with reduced food waste by restricting the total potential food wasted. Both plant-rich diet and reduced food waste are adopted at 50 percent, 75 percent, and 100 percent in the Plausible, Drawdown, and Optimum Scenarios, respectively. Reduced food waste restricts available organic waste feedstocks, impacting composting, waste-to-energy, large-scale methane digesters and landfill methane capture in the Energy Sector.

Yield Model

Drawdown’s yield model calculates total annual global supply of crops and livestock products based on their area of adoption in each of the three scenarios, and global yield impacts of each solution (including both gains due to increased productivity per hectare and losses due to reduction of productive area due to adoption of non-agricultural solutions, e.g., loss of grazing area due to afforestation of grasslands). Grain surpluses in the yield model were also used to set a ceiling for the amount of crops available for use as feedstock for the bioplastic Materials solution.

The yield model matches demand and supply as an integrated system. Both Reference Scenarios showed a food deficit in the high and medium population scenarios (see family planning and educating girls solutions). This would require the clearing of forest and grassland for food production, with associated emissions from land conversion.

All three Drawdown scenarios show agricultural production sufficient to meet food demand and provide a surplus that can be used in bio-based industry, for example as feedstock for bioplastic production. Due to this surplus, no land clearing is necessary, resulting in impressive emissions reduction from avoided deforestation.  Because population change (resulting from educating girls and family planning), plant-rich diet, and reduced food waste are the principal drivers of this effect, Drawdown allocates the resulting reduction in emissions from land clearing to these solutions. However, as the impacts of population on yield and food demand are highly complex, we do not include avoided land conversion emissions associated with population change in the final emissions calculations for those solutions.

Modeling Saturation

Biosequestration does not have limitless potential. In most cases, there is a maximum amount of carbon that can be stored in soils and aboveground perennial biomass before they become saturated. Biosequestration continues after saturation but is offset by more or less equal emissions. In most cases soils, and biomass can return to their approximate pre-agricultural or pre-degradation levels of carbon. This takes anywhere between 10-50 years in agricultural cases, and sometimes somewhat longer in the case of ecosystems like forests. Data about saturation time is very limited.

The Drawdown land model takes the conservative approach that all land units currently adopted for agricultural solutions like conservation agriculture or silvopasture have already achieved saturation, and will not be contributing additional sequestration. New adopted land is assumed to sequester for at least 30 years before achieving saturation.

Note that there are some important exceptions to saturation. Certain ecosystems continue to sequester soil carbon for centuries, notably peatlands and coastal wetlands. Some scientists argue that tropical forests can continue to sequester carbon at a slower rate after saturation. The addition of biochar to saturated soils may be able to overcome this constraint, as does the use of biomass from bamboo or afforestation in long-term products like buildings.

Results

Mitigation Impact

In Drawdown’s Plausible Scenario, Food Sector solutions are responsible for 30.6 percent of total emissions mitigation impact. In the Drawdown and Optimum Scenarios, they contribute 29.3 percent and 31.7 percent, respectively.

Figure 2: Mitigation Impact by Sector, 2020-2050 (in Gigatons of Carbon Dioxide-Equivalent)

© 2017 Project Drawdown

Figure 3: Food Sector Plausible Scenario Emissions and Adoption Results, 2020-2050

Food Sector solutions include supply-side solutions (i.e. agricultural production) and demand-side solutions (i.e. diet, cooking, and waste). Demand-side solutions account for 48.0 percent, 44.7 percent, and 40.7 percent of Food Sector emissions reductions in the Plausible, Drawdown, and Optimum Scenarios, respectively.

Figure 4: Mitigation Impacts by Solution—Plausible Scenario, 2020-2050 (in Gigatons of Carbon Dioxide-Equivalent)

© 2017 Project Drawdown

Table 1: Mitigation Impact of Food Sector Solutions Under the Three Studied Scenarios

Per-hectare impacts were calculated using meta-analysis. Tables 2 and 3 show the ranges, which were determined based on one standard deviation above and below the mean of all data collected. Sequestration types include soil organic carbon (SOC), aboveground biomass (AGB), or both. Emissions reduction includes carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O).

Table 2: Sequestration Rates (in Metric Tons of Carbon per Hectare per Year)

Table 3: Emissions Reduction Rates (in Metric Tons of Carbon Dioxide Equivalent per Hectare per Year)

Financial Impact

Taken as a whole, the food sector offers considerable financial savings. Some solutions have no net cost as they require no new inputs or equipment. Instead, they involve a different system for managing the same farm or ranch elements.

Table 4: Net Cost and Net Savings by Solution (in US2014$ Billion) – Plausible Scenario Only

Sector-Level Benchmarks

The Intergovernmental Panel on Climate Change (IPCC) publication Climate Change 2014: Mitigation of Climate Change, Table 11.4, reports the impact of diet change and food waste reduction at 1.3-13.3 gigatons of carbon dioxide-equivalent emissions reductions by 2050. Drawdown’s combined plant-rich diet and reduced food waste solutions mitigate 3.5, 5.0, and 6.5 gigatons per year by 2050 for the Plausible, Drawdown, and Optimum Scenarios, respectively. This is right in the center of the IPCC’s range.

A recent study in Nature by Paustian, et al estimates agricultural biosequestration at up to 8 gigatons of emissions per year. Drawdown’s model calculates 7.3 gigatons per year for agricultural biosequestration in the Plausible Scenario by 2050, in line with Paustian, et al. Annual mitigation impacts are 10.0 and 12.2 gigatons per year in the Drawdown and Optimum Scenarios, respectively. Drawdown’s results are higher in these aggressive scenarios as a result of our emphasis on high-carbon solutions like tree intercropping, silvopasture, multistrata agroforestry, and tropical staple trees (not central to Paustian’s approach). This is somewhat offset by our greatly reduced adoption of afforestation in the Land Use Sector, compared to the IPCC and others.

Conclusions and Limitations

Conclusions

The food sector is of critical importance to achieving drawdown. Demand-side solutions avoid emissions from land clearing for agriculture and provide additional emissions reductions as well. Supply-side solutions sequester substantial carbon and reduce emissions from agriculture. Particularly good news – few of these solutions were developed for climate mitigation. Instead, most were created to increase agricultural productivity and resilience while improving ecosystem services from farms and ranches. Mitigating climate change through the food system thus results in multiple co-benefits, from healthier diets to increased water-holding capacity on farmland. Drawdown’s emphasis on high-carbon strategies including agroforestry and perennial crops places these strategies in the limelight where they belong.

Limitations

Many farmers and land managers implement more than one of these solutions on the same land (e.g., conservation agriculture with tree intercropping). Insufficient data is available to permit modeling of the impacts of adopting more than one solution per site, so Drawdown allocated only one biosequestration solution per given land unit. However, the Research Team determined that emissions reduction strategies could be applied to the same land as biosequestration solutions, so that nutrient management, for example, can be on the same land as regenerative agriculture. It would be desirable to model the impacts of adoption of multiple sequestration solutions on the same land units in the future. Albedo impacts are not modeled but would be useful to include in future upgrades.