Which practices co‐deliver food security, climate change mitigation and adaptation, and combat land degradation and desertification?

Abstract There is a clear need for transformative change in the land management and food production sectors to address the global land challenges of climate change mitigation, climate change adaptation, combatting land degradation and desertification, and delivering food security (referred to hereafter as “land challenges”). We assess the potential for 40 practices to address these land challenges and find that: Nine options deliver medium to large benefits for all four land challenges. A further two options have no global estimates for adaptation, but have medium to large benefits for all other land challenges. Five options have large mitigation potential (>3 Gt CO2eq/year) without adverse impacts on the other land challenges. Five options have moderate mitigation potential, with no adverse impacts on the other land challenges. Sixteen practices have large adaptation potential (>25 million people benefit), without adverse side effects on other land challenges. Most practices can be applied without competing for available land. However, seven options could result in competition for land. A large number of practices do not require dedicated land, including several land management options, all value chain options, and all risk management options. Four options could greatly increase competition for land if applied at a large scale, though the impact is scale and context specific, highlighting the need for safeguards to ensure that expansion of land for mitigation does not impact natural systems and food security. A number of practices, such as increased food productivity, dietary change and reduced food loss and waste, can reduce demand for land conversion, thereby potentially freeing‐up land and creating opportunities for enhanced implementation of other practices, making them important components of portfolios of practices to address the combined land challenges.

practices are described in Tables 1-3 for land management-based, value chain management-based, and risk management-based practices, respectively, with context caveats and supporting references provided in Tables S1-S3.
Practices often overlap, so are not additive. For example, increasing food productivity will involve changes to cropland, grazing land, and livestock management, which in turn may include increasing soil carbon stocks. The practices cannot therefore be summed and are not mutually exclusive (e.g., cropland management might also increase soil organic matter stocks), and some of the practices considered comprise a few potential management interventions (e.g., improved cropland management is a collection of management interventions). Enabling conditions and strategies such as use of indigenous and local knowledge, attention to gender issues, appropriate governance, etc., are not categorized as practices, so are not included in this analysis. Some suggested methods to address land challenges are better described as overarching frameworks than as practices. For example, climate smart agriculture is a collection of practices aimed at delivering mitigation and adaptation in agriculture, including improved cropland management, grazing land management, and livestock management. Similarly, policy goals, such as land degradation neutrality, include a number of practices. For this reason, policy goals or overarching frameworks (see Table S4) are not treated as practices in this study, but their component practices are.
The IPCC SR1.5 (2018) considered a range of practices (from a mitigation/adaptation perspective only). Table S5 shows how the IPCC SR1.5 options map on to the practices considered in this study. Note that this study excludes most of the energy-related options from IPCC SR1.5, as well as green infrastructure and sustainable aquaculture.
A comprehensive literature review was conducted to gather evidence on the quantitative impact of the practices on each land challenge. The quantified global potential of each practice was then compared to thresholds for each land challenge to assess whether the positive or negative potential was large, moderate, or small. The thresholds for categorization of potentials are shown in Table 4. No equivalence is implied in terms of positive or negative impacts, either in the number or in the magnitude of the impact, that is, one benefit does not equal one adverse side effect. As a consequence, (a) large benefits for one land challenge might outweigh relatively minor negative impacts in addressing another land challenge; (b) some practices may deliver mostly benefits with few negative impacts, but the benefits might be small in magnitude, that is, the practices do no harm, but present only minor co-benefits; and (c) the lack of global estimates of potential does not imply there is no evidence of impact; regional studies often show impacts of the practices, but if the global impact is not available of the literature or cannot be inferred from published studies, no value is given.

| RE SULTS
In the sections below, we provide the quantitative estimates/ranges for the global potential for each practice to address the land challenges of climate change mitigation (Section 3.1), climate change adaptation (Section 3.2), land degradation and desertification (Section 3.3), and food security (Section 3.4) arising from the extensive literature review, before summarizing these potentials in relation to the thresholds in Table 4, across all land challenges.

Practice Description
Increased food productivity Increased food productivity arises when the output of food commodities increases per unit of input, for example, per unit of land or water. It can be realized through many other practices such as improved cropland, grazing land, and livestock management

Improved cropland management
Improved cropland management is a collection of practices consisting of (a) management of the crop: including high carbon input practices, for example, improved crop varieties, crop rotation, use of cover crops, perennial cropping systems, integrated production systems, crop diversification, agricultural biotechnology; (b) nutrient management: including optimized fertilizer application rate, fertilizer type (organic manures, compost, and mineral), timing, precision application, nitrification inhibitors; (c) reduced tillage intensity and residue retention; (d) improved water management: including drainage of waterlogged mineral soils and irrigation of crops in arid/ semiarid conditions; (e) improved rice management: including water management such as mid-season drainage and improved fertilization and residue management in paddy rice systems; and (f) biochar application Improved grazing land management Improved grazing land management is a collection of practices consisting of (a) management of vegetation: including improved grass varieties/sward composition, deep rooting grasses, increased productivity, and nutrient management; (b) animal management: including appropriate stocking densities fit to carrying capacity, fodder banks, and fodder diversification; and (c) fire management: improved use of fire for sustainable grassland management, including fire prevention and improved prescribed burning (see also fire management as a separate practice below)

Improved livestock management
Improved livestock management is a collection of practices consisting of (a) improved feed and dietary additives (e.g., bioactive compounds, fats), used to increase productivity and reduce emissions from enteric fermentation; (b) breeding (e.g., breeds with higher productivity or reduced emissions from enteric fermentation); (c) herd management, including decreasing neonatal mortality, improving sanitary conditions, animal health and herd renewal, and diversifying animal species; (d) emerging technologies (of which some are not legally authorized in several countries) such as propionate enhancers, nitrate and sulfate supplements, archaea inhibitors and archaeal vaccines, methanotrophs, acetogens, defaunation of the rumen, bacteriophages and probiotics, ionophores/antibiotics; and (e) improved manure management, including manipulation of bedding and storage conditions, anaerobic digesters; biofilters, dietary change and additives, soil-applied and animal-fed nitrification inhibitors, urease inhibitors, fertilizer type, rate and timing, manipulation of manure application practices, and grazing management Agroforestry Agroforestry involves the deliberate planting of trees in croplands and silvopastoral systems Agricultural diversification Agricultural diversification includes a set of agricultural practices that aim to improve the resilience of farming systems to climate variability and climate change and to economic risks posed by fluctuating market forces. In general, the agricultural system is shifted from one based on low-value agricultural commodities to one that is more diverse, composed of a basket of higher value-added products Reduced grassland conversion to cropland Grasslands can be converted to croplands by plowing of grassland and seeding with crops. Since croplands have a lower soil carbon content than grasslands and are also more prone to erosion than grasslands, reducing conversion of grassland to croplands will prevent soil carbon losses by oxidation and soil loss through erosion. These processes can be reduced if the rate of grassland conversion to cropland is reduced

Integrated water management
Integrated water management is the process of creating holistic strategies to promote integrated, efficient, equitable, and sustainable use of water for agroecosystems. It includes a collection of practices including water-use efficient irrigation in arid/semiarid areas, improvement of soil water holding capacity through increases in soil organic matter content, and improved cropland management, agroforestry, and conservation agriculture. Increasing water availability, and reliability of water for agricultural production, achieved by using different techniques of water harvesting, storage, and its judicious utilization through farm ponds, dams, and community tanks in rainfed agriculture areas can benefit adaptation Improved and sustainable forest management Improved forest management refers to management practices in forests for the purpose of climate change mitigation. It includes a wide variety of practices affecting the growth of trees and the biomass removed, including improved regeneration (natural or artificial) and a better schedule, intensity, and execution of operations (thinning, selective logging, final cut; reduced impact logging, etc.). Sustainable forest management is the stewardship and use of forests and forest lands in a way, and at a rate, that maintains their biodiversity, productivity, regeneration capacity, vitality, and their potential to fulfill, now and in the future, relevant ecological, economic, and social functions, at local, national, and global levels, and that does not cause damage to other ecosystems

Reduced deforestation and degradation
Reduced deforestation and forest degradation include conservation of existing carbon pools in forest vegetation and soil by controlling the drivers of deforestation (i.e., commercial and subsistence agriculture, mining, urban expansion) and forest degradation (i.e., overharvesting including fuelwood collection, poor harvesting practices, overgrazing, pest outbreaks, and extreme wildfires), also through establishing protected areas, improving law enforcement, forest governance and land tenure, supporting community forest management, and introducing forest certification

Reforestation and forest restoration
Reforestation is the conversion to forest of land that has previously contained forests but that has been converted to some other use. Forest restoration refers to practices aimed at regaining ecological integrity in a deforested or degraded forest landscape. As such, it could fall under reforestation if it were reestablishing trees where they have been lost, or under forest management if it were restoring forests where not all trees have been lost. For practical reasons, here forest restoration is treated together with reforestation Afforestation Afforestation is the conversion to forest of land that historically has not contained forests (see also reforestation) (Continues) (Continues)

Increased soil organic carbon content
Practices that increase soil organic matter content include (a) land use change to an ecosystem with higher equilibrium soil carbon levels (e.g., from cropland to forest); (b) management of the vegetation: including high carbon input practices, for example, improved varieties, rotations and cover crops, perennial cropping systems, biotechnology to increase inputs and recalcitrance of below ground carbon; (c) nutrient management and organic material input to increase carbon returns to the soil: including optimized fertilizer and organic material application rate, type, timing, and precision application; (d) reduced tillage intensity and residue retention; and (e) improved water management: including irrigation in arid/semiarid conditions | 1537 SMITH eT al.

| Practices based on land management
Increasing the productivity of land used for food production can deliver significant mitigation by avoiding emissions that would occur if increased food demand were met through expansion of the agricultural land area (Burney, Davis, & Lobell, 2010). If pursued through injudicious use of agrochemical inputs, numerous adverse impacts on greenhouse gas (GHG) emissions and other aspects of environmental sustainability can occur (Table 5), but if pursued sustainably and with appropriate governance and other measures to prevent rebound effects, for example, through sustainable intensification (e.g., Pretty et al., 2018), increased food productivity could provide high levels of mitigation. For example, yield improvement has been estimated to have contributed to emissions' savings of >13 Gt CO 2 eq/year since 1961 (Burney et al., 2010; Table 5). If the considerable remaining global yield gaps (Mueller et al., 2012) could be closed through sustainable intensification, mitigation of a similar magnitude could be realized. This can also reduce the GHG intensity of products (Bennetzen, Smith, & Porter, 2016a, 2016b which means a smaller environmental footprint of production, since demand can be met using less land and/or with fewer animals.
The global mitigation potential for improved grazing land management is moderate (1.4-1.8 Gt CO 2 eq/year), with the lower value in the range for technical potential taken from Smith et al. (2008), which includes only grassland management measures, and the upper value in the range from Herrero et al. (2016), which also includes indirect effects and some components of livestock management, and soil carbon sequestration, so there is overlap with these practices. Conant, Paustian, Del Grosso, and Parton (2005) Zomer et al. (2016) reported that trees in agroforestry landscapes have increased carbon stocks equivalent to 0.7 Gt CO 2 eq/year.
Estimates of global potential range from 0.1 to 5.7 Gt CO 2 eq/year (from an "optimum implementation" scenario of Hawken, 2017), Agricultural diversification mainly aims to increase climate resilience, but it may have a small (but globally unquantified) mitigation

Practice Description
Restoration and reduced conversion of peatlands Peatland restoration involves restoring degraded/damaged peatlands, through rewetting, which both increases carbon sinks, but also avoids ongoing CO 2 emissions from degraded peatlands, so it both prevents future emissions and creates a sink, as well as protecting biodiversity

Biodiversity conservation
Biodiversity conservation refers to practices aiming at maintaining components of biological diversity. It includes conservation of ecosystems and natural habitats, maintenance and recovery of viable populations of species in their natural surroundings (in situ conservation) and, in the case of domesticated or cultivated species, in the surroundings where they have developed their distinctive properties outside their natural habitats (ex situ conservation). Examples of biodiversity conservation measures are establishment of protected areas to achieve specific conservation objectives, preservation of biodiversity hotspots, land management to recover natural habitats, practices to expand or control targeted plant or animal species in productive lands or rangelands (e.g., rewilding), sustainable harvest of native species

Enhanced weathering of minerals
The enhanced weathering of minerals that naturally absorb CO 2 from the atmosphere has been proposed as a CDR technology with a large mitigation potential. The rocks are ground to increase the surface area and the ground minerals are then applied to the land where they absorb atmospheric CO 2

Bioenergy and BECCS
Bioenergy production can mitigate climate change by delivering an energy service, therefore avoiding combustion of fossil energy. It is the most common renewable energy source used today in the world and has a large potential for future deployment. BECCS entails the use of bioenergy technologies (e.g., bioelectricity or biofuels) in combination with CO 2 capture and storage. BECCS simultaneously provides energy and can reduce atmospheric CO 2 concentrations; thus, BECCS is considered a CDR technology. While several BECCS demonstration projects exist, it has yet to be deployed at scale. Bioenergy and BECCS are widely used in many future scenarios as a climate change mitigation option in the energy and transport sector, especially those scenarios aimed at a stabilization of global climate at 2°C or less above pre-industrial levels Note: Context and supporting references are provided in Table S1.
Reducing conversion of grassland to cropland could provide significant climate change mitigation by retaining soil carbon stocks that might otherwise be lost. When grasslands are converted to croplands, they lose on average 36% of their soil organic carbon stocks after 20 years (Poeplau et al., 2011). Assuming an average starting soil organic carbon stock of grasslands of 115 t C/ha (Poeplau et al., 2011), this is equivalent to a loss of 41.5 t C/ha on conversion to cropland. Mean annual global cropland conversion rates  have been around 47,000 km 2 /year (Krause et al., 2017) or 940,000 km 2 over a 20 year period. The equivalent loss of soil organic carbon over 20 years would, therefore, be 14 Gt CO 2 eq = 0.7 Gt CO 2 eq/year. Griscom et al. (2017) estimate a cost-effective mitigation potential of 0.03 Gt CO 2 eq/ year (Table 5).
Integrated water management provides moderate benefits for climate change mitigation through interactions with other land management strategies. For example, promoting soil carbon conservation (e.g., reduced tillage) can improve the water retention capacity of soils. Jat et al. (2015) found that improved tillage practices and residue incorporation increased water-use efficiency by 30%, ricewheat yields by 5%-37%, income by 28%-40%, and reduced GHG emissions by 16%-25%. While irrigated agriculture accounts for only 20% of the total cultivated land, the energy consumption from groundwater irrigation is significant. However, current estimates TA B L E 2 Value chain management-based practices considered in this study

Practice Description
Dietary change Sustainable healthy diets represent a range of dietary changes to improve human diets, to make them healthy in terms of the nutrition delivered, and also (economically, environmentally, and socially) sustainable. A "contract and converge" model of transition to sustainable healthy diets would involve a reduction in overconsumption (particularly of livestock products) in overconsuming populations, with increased consumption of some food groups in populations where minimum nutritional needs are not met. Such a conversion could result in a decline in undernourishment, as well as reduction in the risk of morbidity and mortality due to overconsumption Reduced postharvest losses Approximately one-third of the food produced for human consumption is wasted in post-production operations. The key drivers for post-harvest waste in developing countries are structural and infrastructure deficiencies, requiring responses that process, preserve, and, where appropriate, redistribute food to where it can be consumed immediately

Reduced food waste
Food loss in developed countries mostly occurs at the retail/consumer stage, and practices that focuses on consumer or retailer waste (ranging from better use by date labeling to consumer education campaigns) can reduce pressure on land (see also reducing post-harvest losses above)

Material substitution
Material substitution involves the use of wood or agricultural biomass (e.g., straw bales) instead of fossil fuel-based materials (e.g., concrete, iron, steel, aluminum) for building, textiles, or other applications Sustainable sourcing Sustainable sourcing includes approaches to ensure that the production of goods is done in a sustainable way, such as through low-impact agriculture, zero deforestation supply chains, or sustainably harvested forest products. Currently around 8% of global forest area has been certified in some manner, and 25% of global industrial roundwood comes from certified forests. Sustainable sourcing can also enable producers to increase their percentage of the final value of commodities through improved innovation, coordination, and efficiency in supply chains, as well as labeling to ensure consumer demands. Promoting sustainable and value-added products can reduce the need for compensatory extensification of agricultural areas and is a specific commitment of some sourcing programs (such as forest certification programs)

Management of supply chains
Management of supply chains include improving efficiency and sustainability to reduce climate risk and profitably reduce emissions and can include: (a) increasing the economic value through improved production processes; (b) adopting emission accounting tools (e.g., carbon and water footprinting); (c) improved policies for stability of food supply to minimize food price volatility

Enhanced urban food systems
Urban areas are becoming the principal territories for practice in improving food access through innovative strategies that aim to reduce hunger and improve livelihoods, including support for urban and peri-urban agriculture, green infrastructure (e.g., green roofs), local markets, enhanced social (food) safety nets, development of alternative food sources and technologies, such as vertical farming, and local food policy and planning initiatives. Such systems have created nutritious food supplies for the city, while improving the health status of urban dwellers, reducing pollution levels, adapting to and mitigating climate change, and stimulating economic development Improved food processing and retailing Improved food processing and retailing involves several practices related to improving packaging, processing, cooling, drying, and extracting, and reducing agri-food GHG emissions from processing and transportation and reducing waste in retailing Improved energy use in food systems Energy efficiency of agriculture can be improved to reduce the dependency on nonrenewable energy sources either by decreased energy inputs, or through increased outputs per unit of input. In some countries, managerial inefficiency (rather than a technology gap) is the main source for energy efficiency loss. Heterogenous patterns of energy efficiency exist at the national scale and promoting energy efficient technologies along with managerial capacity development can reduce the gap and provide large benefits for climate adaptation. Improvements in carbon monitoring and calculation techniques such as the footprinting of agricultural products can enhance energy efficiency transition management and uptake in agricultural enterprises Note: Context and supporting references are provided in Table S2. of mitigation potential are limited to reductions in GHG emissions mainly in cropland and rice cultivation (Smith et al., 2008. Li, Xu, Tiwari, and Ji (2006) estimated a 0.52-0.72 Gt CO 2 eq/year reduction using alternate wetting and drying practices. Current estimates of N 2 O release from terrestrial soils and wetlands account for 10%-15% of anthropogenically fixed nitrogen on the Earth system (Wang et al., 2017).
Improved and sustainable forest management could potentially contribute to moderate mitigation benefits globally, up to about 2 Gt CO 2 eq/year (Table 5). For managed forests, the most effective forest carbon mitigation strategy is the one that, through increasing biomass productivity, optimizes the carbon stocks (in forests and in long-lived products) as well as the wood substitution effects for a given time frame (Erb et al., 2018;Kurz, Smyth, & Lemprière, 2016;Nabuurs, Pussinen, Brusselen, & Schelhaas, 2007;Smyth et al., 2014). Estimates of the mitigation potential also vary depending on the counterfactual, such as businessas-usual management (e.g., Grassi, Pilli, House, Federici, & Kurz, 2018) or other assumptions. Climate change will affect the mitigation potential of forest management due to an increase in extreme events such as fires, insects, and pathogens (Seidl et al., 2017).
More detailed estimates are available at regional or biome level.
For instance, according to Nabuurs et al. (2017), the implementation of Climate-Smart Forestry (a combination of improved forest management, expansion of forest areas, energy substitution, establishment of forest reserves, etc.) in the European Union has the potential to contribute an additional 0.4 Gt CO 2 eq/year mitigation by 2050. In tropical forests, adoption of reduced impact logging and wood processing technologies along with financial incentives can reduce forest fires, forest degradation, maintain timber production, and retain carbon stocks (Sasaki et al., 2016). Forest certification may support sustainable forest management, helping to prevent forest degradation and over-logging (Rametsteiner & Simula, 2003). Community forest management has proven a viable model for sustainable forestry, including for carbon sequestration (Chhatre & Agrawal, 2009).
Reducing deforestation and forest degradation rates represents one of the most effective and robust options for climate change mitigation, with large mitigation benefits globally (up to 5.8 Gt CO 2 eq/year; A large range of estimates exist in the scientific literature for the mitigation potential of reforestation and forest restoration, and they often overlap with estimates for afforestation. At a global level, the overall potential for these options is large (Bastin et al., 2019;Griscom et al., 2017), reaching about 10 Gt CO 2 eq/year ( Table 5). The greatest potential for these options is in tropical and subtropical climates (Houghton & Nassikas, 2018;Lewis, Wheeler, Mitchard, & Koch, 2019).
The climate change mitigation benefits of afforestation and reforestation are reduced at high latitudes owing to surface albedo feedback.

Management of urban sprawl
Unplanned urban expansion of cities along the rural-urban fringe (especially strong in emerging towns and cities in Asia and Africa) has been identified as a driver of forest and agricultural land loss and a threat to food production around cities and may result in a 1.8%-2.4% loss of global croplands by 2030. Policies to prevent urban sprawl have included integrated land use planning, agricultural zoning ordinances and agricultural districts, urban redevelopment, arable land reclamation, and transfer/purchase of development rights or easements

Livelihood diversification
Livelihood diversification (drawing from a portfolio of dissimilar sources of livelihood as a tool to spread risk) has been identified as one option to increase incomes and reduce poverty, increase food security, and promote climate resilience and risk reduction Use of local seeds Using local seeds (also called seed sovereignty) refers to use of non-improved, non-commercial seed varieties. These can be used and stored by local farmers as low-cost inputs and can often help contribute to the conservation of local varieties and landraces, increasing local biodiversity, and often require no pesticide or fertilizer use, leading to less land degradation

Disaster risk management
Disaster risk management encompasses many approaches to try to reduce the consequences of climate and weatherrelated disasters and events on socioeconomic systems through proactive prevention; timely response; quick and effective recovery; and sustainable development. Other options include using early warning systems that can encompass (a) education systems; (b) hazard and risk maps; (c) hydrological and meteorological monitoring (such as flood forecasting or extreme weather warnings); and (d) communication systems

Risk sharing instruments
Risk sharing instruments can encompass a variety of approaches, including intra-household risk pooling, community rotating credit associations (ROSCAs) and other formal and informal credit services, as well as insurance of various kinds. Commercial crop insurance can involve both traditional indemnity-based insurance that reimburses clients for estimated financial losses from shortfalls, or index insurance that pays out the value of an index (such as weather events) rather than actual losses Note: Context and supporting references are provided in Table S3.
The management and control of soil erosion may prevent losses of organic carbon in water-or wind-transported sediments. However, since the final fate of eroded material is still debated, ranging from a source of 1.36-3.67 Gt CO 2 eq/year (Jacinthe & Lal, 2001;Lal, 2004) to a sink of 0.44-3.67 Gt CO 2 eq/year (Stallard, 1998;Smith, Renwick, Buddemeier, & Crossland, 2001;Smith, Sleezer, Renwick, & Buddemeier, 2005;Van Oost et al., 2007;Table 5), the overall impact of erosion control on mitigation is context specific and highly uncertain at the global level (Hoffmann et al., 2013).
Salt-affected soils are highly constrained environments that require permanent prevention of salinization. Their mitigation potential is likely to be small, though prevention of salinization has more potential, though the global mitigation potential is not quantified (Dagar, Sharma, Sharma, & Singh, 2016;UNCTAD, 2011;Wong, Greene, Dalal, & Murphy, 2010 Table 5).
For biochar, a global analysis of technical potential, in which biomass supply constraints were applied to protect against food insecurity, loss of habitat and land degradation, estimated potential abatement of 3.7-6.6 Gt CO 2 eq/year (including 2.6-4.6 Gt CO 2 eq/year carbon stabilization). Considering all published estimates, the estimates of potential range from 0.03 to 6.6 Gt CO 2 eq/year with the lowest estimate from the "plausible" scenario of Hawken (2017; Table 5).  Table 5).
Management of landslides and natural hazards is a key climate adaptation option, but due to limited global areas vulnerable to landslides and natural hazards, its mitigation potential is likely to be modest (Noble et al., 2014). Forest regeneration stabilizes hillsides and reduces landslides (Robledo, Fischler, & Patiño, 2004).
In terms of management of pollution, including acidification, UNEP and WMO (2011) and Shindell et al. (2012)  Negatively impacts more than around 3 million km 2 Negatively impacts more than around 100 million people Note: Magnitudes are for the technical potential of practices globally. For each land challenge, magnitudes are set relative to a marker level as follows. For mitigation, potentials are set relative to the approximate potentials for the mitigation options with the largest individual impacts (~3 Gt CO 2 eq/year; Pacala and Socolow, 2004). The threshold for the "large" magnitude category is set at this level.
For adaptation, magnitudes are set relative to the 100 million lives estimated to be affected by climate change and a carbon-based economy between 2010 and 2030 (DARA, 2012). The threshold for the "large" magnitude category represents 25% of this total. For desertification and land degradation, magnitudes are set relative to the lower end of current estimates of degraded land, 10-60 million km 2 (Gibbs & Salmon, 2015). The threshold for the "large" magnitude category represents 30% of the lower estimate. For food security, magnitudes are set relative to the approximately 800 million people who are currently undernourished (HLPE, 2017). The threshold for the "large" magnitude category represents 12.5% of this total.
TA B L E 5 Summary of global mitigation effects of practices based on land management
There are no global data on the impacts of the management of invasive species/encroachment on mitigation.
Coastal wetland restoration could provide high levels of climate mitigation, with avoided coastal wetland impacts and coastal wet-  Table 5).
Peatland restoration could provide moderate levels of climate mitigation, with avoided peat impacts and peat restoration estimated to deliver 0.6-2 Gt CO 2 eq/year from all global estimates published in Couwenberg, Dommain, and Joosten (2010) Table 5).
Mitigation potential from biodiversity conservation varies depending on the type of practice and specific context. Protected areas are estimated to store over 300 Gt carbon, roughly corresponding to 15% of terrestrial carbon stocks (Campbell, Lobell, Genova, & Field, 2008;Kapos et al., 2008). At global level, the potential mitigation resulting from protection of these areas for the period 2005-2095 is on average about 0.9 Gt CO 2 eq/year relative to a reference scenario (Calvin et al., 2014). The potential effects on the carbon cycle of the management of wild animal species are context dependent. For example, moose browsing in boreal forests can decrease the carbon uptake of ecosystems by up to 75% (Schmitz et al., 2018), and reducing moose density through active population management in Canada is estimated to be a carbon sink equivalent to about 0.37 Gt CO 2 eq/year (Schmitz et al., 2014).
The mitigation potential for bioenergy and bioenergy with carbon capture and strorage (BECCS) derived from bottom-up models is large (IPCC SR1.5, 2018), with technical potential estimated at 100-300 EJ/year (IPCC, 2011) or up to ~11 Gt CO 2 eq/year. These estimates, however, exclude N 2 O associated with fertilizer application and land-use change emissions. Those effects are included in the modeled scenarios using bioenergy and BECCS, with the magnitude depending on where the bioenergy is grown (Wise et al., 2015), at what scale, and whether N fertilizer is used.

| Practices based on value chain management
Dietary change and waste reduction can provide large benefits for mitigation, with potentials of 0.7-8 Gt CO 2 eq/year for dietary change and 0.7-4.5 Gt CO 2 eq/year for food waste reduction  (Table 6).
Some studies indicate that material substitution has the potential for significant mitigation, with one study estimating a 14%-31% reduc- Abbreviation: BECCS, bioenergy with carbon capture and strorage.
While sustainable sourcing presumably delivers a mitigation ben- Efficient use of energy and resources in food transport and distribution can contribute to a reduction in GHG emissions, estimated to be 1% of global CO 2 emissions (James & James, 2010;Vermeulen, Campbell, & Ingram, 2012). Given that global CO 2 emissions in 2017 were 37 Gt CO 2 eq, this equates to 0.37 Gt CO 2 eq/year (covering food transport and distribution, improved efficiency of food processing and retailing, and improved energy efficiency; Table 6).

| Practices based on risk management
In general, because these options are focused on adaptation and other co-benefits, the mitigation benefits are modest, and mostly unquantified. Extensive and less dense urban development tends to have higher energy usage, particularly from transport (Liu, Zhou, & Wu, 2015), such that a 10% reduction of very low density urban fabrics is correlated with 9% fewer emissions per capita in Europe (Baur, Förster, & Kleinschmit, 2015). However, the exact contribution to mitigation from the prevention of urban sprawl through land conversion in particular has not been well quantified (Thornbush, Golubchikov, & Bouzarovski, 2013). Suggestions from selected studies in the United States are that biomass decreases by half when forest is converted to urban land (Briber et al., 2015), and a study in Bangkok found a decline by half in carbon sinks in the urban area in the past 30 years (Ali, Pumijumnong, & Cui, 2018).
There is no literature specifically on the linkages between livelihood diversification and climate mitigation benefits, although some forms of diversification that include agroforestry would likely result in increased carbon sinks (Altieri, Nicholls, Henao, & Lana, 2015;Descheemaeker et al., 2016). There is no literature exploring linkages between use of local seeds and GHG emission reductions.
While disaster risk management can presumably have mitigation co-benefits, as it can help reduce food loss on-farm (e.g., crops destroyed before harvest or avoided animal deaths during droughts and floods, meaning reduced production losses and wasted emissions), there is no quantified global estimate for this potential (Table 7).
Risk sharing instruments could have some mitigation co-benefits if they buffer household losses and reduce the need to expand agricultural lands after experiencing risks. However, the overall impacts of these are unknown. Furthermore, commercial insurance may induce producers to bring additional land into crop production, particularly marginal or land with other risks that may be more environmentally sensitive (Claassen, Cooper, & Carriazo, 2011).   (2010), Vermeulen et al. (2012) 3.2 | Potential of the practices for delivering adaptation

| Practices based on land management
Increasing food productivity by practices such as sustainable intensification improves farm incomes and allows households to build assets for use in times of stress, thereby improving resilience . By reducing pressure on land and increasing food production, increased food productivity could be beneficial for adaptation . Pretty et al. (2018) report that 163 million farms occupying 4.53 Mkm 2 have passed a redesign threshold for application of sustainable intensification, suggesting the minimum number of people benefiting from increased productivity and adaptation benefits under sustainable intensification is >163 million, with the total likely to be far higher (Table 8). Around 30% of the world's rural population use trees across 46% of all agricultural landscapes (Lasco, Delfino, Catacutan, Simelton, & Wilson, 2014), meaning that up to 2.3 billion people benefit from agroforestry, globally (Table 8).
Agricultural diversification is key to achieving climatic resilience Cohn et al., 2017). Crop diversification is an important climate change adaptation option (Vermeulen et al., 2012), which can improve resilience by engendering a greater ability to suppress pest outbreaks and dampen pathogen transmission, as well as by buffering crop production from the effects of greater climate variability and extreme events (Lin, 2011).
Reduced conversion of grassland to cropland may lead to adaptation benefits by stabilizing soils in the face of extreme climatic events, since grasslands are more resilient than cropping systems (Lal, 2001), thereby increasing resilience, but since it would likely have a negative impact on food production/security (since croplands produce more food per unit area than grasslands), the wider adaptation impacts would likely be negative. However, there is no literature quantifying the global impact of avoidance of conversion of grassland to cropland on adaptation.
Integrated water management provides large co-benefits for adaptation (Dillon & Arshad, 2016) by improving the resilience of crop production systems to future climate change (Porter et al., 2014; Table 8). Improving irrigation systems and integrated water resource management, such as enhancing urban and rural water supplies and reducing water evaporation losses (Dillon & Arshad, 2016), are significant options for enhancing climate adaptation. Many technical innovations (e.g., precision water management) can lead to beneficial adaptation outcomes by increasing water availability and the reliability of agricultural production, using different techniques of water harvesting, storage, and its judicious utilization through farm ponds, dams, and community tanks in rainfed agriculture areas. Integrated water management practices that use freshwater would be expected to have few adverse side effects in regions where water is plentiful, but large adverse side effects in regions where water is scarce (Grey & Sadoff, 2007;Liu et al., 2017;Scott et al., 2011).
Improved and sustainable forest management positively impacts adaptation by limiting the negative effects associated with pollution (of air and fresh water), diseases, exposure to extreme weather  (2016)  events and natural disasters, and poverty (e.g., Smith et al., 2014).
Furthermore, sustainable forest management has a number of potential co-benefits for adaptation, ecosystem services, biodiversity conservation, microclimatic and water regulation, soil erosion protection and coastal area protection (Locatelli, 2011).
There is high agreement that reduced deforestation positively affects adaptation and resilience of coupled human-natural systems and the stability of the water cycle. Based on the number of people affected by natural disasters (CRED, 2015), the number of people depending to varying degrees on forests for their livelihoods (World Bank et al., 2009), the area of managed forest and the current annual deforestation rate (Keenan et al., 2015), the estimated global potential effect for adaptation is largely positive for forest management, and moderately positive for reduced deforestation when accumulated until the end of the century ( Table 8). The uncertainty of these global estimates is high.
More robust qualitative and some quantitative estimates are available at local and regional level. According to Karjalainen, Sarjala, and Raitio (2009), reducing deforestation and habitat alteration contribute to limiting infectious diseases such as malaria in Africa, Asia, and Latin America, thus lowering the expenses associated with healthcare treatments. Bhattacharjee and Behera (2017)  Forest restoration may facilitate the adaptation and resilience of forests to climate change by enhancing connectivity between forest areas and conserving biodiversity hotspots (Dooley & Kartha, 2018;Ellison et al., 2017;Locatelli, Catterall, et al., 2015;Locatelli, Evans, Wardell, Andrade, & Vignola, 2011;Locatelli, Pavageau, Pramova, & Di Gregorio, 2015). Furthermore, forest restoration may improve ecosystem functionality and services, provide microclimatic regulation for people and crops, wood and fodder as safety nets, soil erosion protection and soil fertility enhancement for agricultural resilience, coastal area protection, water and flood regulation (Locatelli, Catterall, et al., 2015;Locatelli, Pavageau, et al., 2015).
Soil organic carbon increase is promoted as an action for climate change adaptation. Since increasing soil organic matter content is a measure to address land degradation, and restoring degraded land helps to improve resilience to climate change, soil carbon increase is an important option for climate change adaptation. With around 120 thousand km 2 land lost to degradation every year, and over 3.2 billion people negatively impacted by land degradation globally (IPBES, 2018), practices designed to increase soil organic carbon have a large potential to address adaptation needs (Table 8).
Since soil erosion control can prevent land degradation and desertification, it improves the resilience of agriculture to climate change and increases food production (IPBES, 2018;Lal, 1998) Prevention and/or reversal of topsoil salinization requires the combined management of groundwater, irrigation techniques, drainage, mulching, and vegetation, with all of these considered relevant for adaptation (Dagar et al., 2016;Qadir, Noble, & Chartres, 2013;UNCTAD, 2011). Taking into account the widespread diffusion of salinity problems, many people can benefit from its implementation by farmers. The relation between compaction prevention and/or reversion and climate adaption is less evident, and can be related to better hydrological soil functioning (Chamen et al., 2015;Epron et al., 2016;Tullberg et al., 2018).
Biochar has potential to benefit climate adaptation by improving the resilience of crop production systems to future climate change by increasing yield in some regions and improving water holding capacity (Sohi, 2012;Woolf, Amonette, Street-Perrott, Lehmann, & Joseph, 2010). By increasing yield by 25% in the tropics (Jeffery et al., 2017), this could increase food production for 3.2 billion people affected by land degradation (IPBES, 2018), thereby potentially improving their resilience to climate change shocks (Table 8). The use of large areas of land to provide feedstock for biochar could adversely impact adaptation by occupying land that could be used for food production, though the impact has not been quantified globally.
In terms of fire management, Doerr and Santín (2016) showed that globally the average number of people killed by wildfire was 1940, and the total number of people affected was 5.8 million from 1984 to 2013. Johnston et al. (2012) showed the average mortality attributable to landscape fire smoke exposure was 339,000 deaths annually. The regions most affected were sub-Saharan Africa (157,000) and Southeast Asia (110,000). Estimated annual mortality during La Niña was 262,000, compared with around 100,000 excess deaths across Indonesia, Malaysia, and Singapore (Table 8).

Management of landslides and natural hazards are usually listed
among planned adaptation options in mountainous and sloped hilly areas, where uncontrolled runoff and avalanches may cause climatic disasters, affecting millions of people from both urban and rural areas. Landslide control requires both increasing plant cover and engineering practices (see Table 8). Enhanced weathering of minerals has been proposed as a mechanism for improving soil health and food security (Beerling et al., 2018), but there is no literature estimating the global adaptation benefits.

| Practices based on value chain management
Decreases in pressure on land and decreases in production intensity associated with sustainable healthy diets or reduced food waste could also benefit adaptation. For example, Westhoek et al. (2014) estimate a 23% reduction of cropland in Europe through halving meat consumption. However, the size of this effect is not well quantified globally (Muller et al., 2017).
Reducing food waste and losses can relieve pressure on the global freshwater resource, thereby aiding adaptation. Food losses account for 215 km 3 /year of freshwater resources, which Kummu et al. (2012) report to be about 12%-15% of the global consumptive water use.
Given that 35% of the global population is living under high water stress or shortage (Kummu, Ward, Moel, & Varis, 2010), reducing food waste could benefit 320-400 million people (12%-15% of the 2,681 million people affected by water stress/shortage).
While no studies report quantitative estimates of the effect of material substitution on adaptation, the effects are expected to be similar to reforestation and afforestation if the amount of material substitution leads to an increase in forest area. Additionally, some studies indicate that wooden buildings, if properly constructed, could reduce fire risk compared to steel, which softens when burned (Gustavsson et al., 2006;Ramage et al., 2017).
It is estimated that 500 million smallholder farmers depend on agricultural businesses in developing countries (World Bank, 2017), meaning that better promotion of value-added products and improved efficiency and sustainability of food processing and retailing could potentially help up to 500 million people to adapt to climate change. However, how sustainable sourcing in general could help farmers and forest management is mostly unquantified. More than 1 million farmers have currently been certified through various schemes (Tayleur et al., 2017), but how much this has helped them prepare for adaptation is unknown.
Management of supply chains has the potential to reduce vulnerability to price volatility. Consumers in lower income countries are most affected by price volatility, with sub-Saharan Africa and South Asia at highest risk (Fujimori et al., 2019;Regmi & Meade, 2013). However, understanding the stability of food supply is one of the weakest links in global food system research (Wheeler & von Braun, 2013) as instability is driven by a confluence of factors (Headey & Fan, 2008 led to reduced staple food prices for 2 billion people (Timmer, 2009).
Spending less on food frees up money for other activities, including adaptation, but it is unknown by how much (Zezza et al., 2009;Ziervogel & Ericksen, 2010). Another example of a reduction in staple food prices occurred in Bangladesh with food stability policies saving rural households US$887 million in total (Torlesse, Kiess, & Bloem, 2003).
Food supply stability through improved supply chains also potentially reduces conflicts (by avoiding food price riots, which occurred in countries with over 100 million total population in 2007-2008), and thus increases adaptation capacity (Raleigh, Choi, & Kniveton, 2015).
There are no global estimates of the contribution of urban food systems, in contributing to adaptation, but since the urban population in 2018 was 4.2 billion people, this sets the upper limit on those who could benefit.
Improved energy use in food systems in agriculture could benefit 65% (760 million people) of poor working adults who make a living through agriculture (World Bank, 2017).

| Practices based on risk management
Reducing urban sprawl is likely to provide adaptation co-benefits via improved human health (Anderson, 2017;Frumkin, 2002), as sprawl contributes to reduced physical activity, worse air pollution, and exacerbation of urban heat island effects and extreme heat waves (Stone, Hess, & Frumkin, 2010). The most sprawling cities in the United States have experienced extreme heat waves more than double those of denser cities (Stone et al., 2010). Other adaption cobenefits are less well understood; there are likely to be cost savings from managing or planning growth, as one study found 2% savings in metropolitan budgets, which could then be spent on adaptation planning (Deal & Schunk, 2004). Thornton & Herrero, 2014). Surveys of farmers in climate variable areas find that livelihood diversification is increasingly favored as an adaptation option (Bryan et al., 2013), although it is not always successful, since it can increase exposure to climate variability (Adger et al., 2011). There are over 570 million small farms in the world (Lowder, Skoet, & Raney, 2016); it is not clear, however, how many farmers have not yet practiced diversification and thus how many would be helped by supporting this practice (Rigg, 2006).
It has been estimated that currently more than half of smallholder farmers in the developing world still rely to some degree on use of local seeds (Altieri, Funes-Monzote, & Petersen, 2012;McGuire & Sperling, 2016). Use of local seeds can potentially facilitate adaptation, as moving to use of commercial seeds can increase costs for farmers (Howard, 2015). Local seed networks and banks also protect local agrobiodiversity and landraces, which are important to facilitate adaptation, as they may be more resilient to some forms of climate change (Coomes et al., 2015;van Niekerk & Wynberg, 2017;Vasconcelos et al., 2013). States have been less likely to adapt to extreme weather events than those not holding insurance (Annan & Schlenker, 2015). It is unclear how many people worldwide use insurance as an adaptation strategy; Platteau, De Bock, and Gelade (2017) suggest less than 30% of smallholders take out any form of insurance, but it is likely in the millions (Tables 9 and 10). Burney et al. (2010) estimated that an additional global cropland area of 11.11-15.14 Mkm 2 would have been needed if productivity had not increased between 1961 and 2000. Given that agricultural expansion is the main driver of land degradation and desertification, increased food productivity could have prevented this area from exploitation and land degradation (Table 11). to rangeland, increases in ground cover by vegetation, and protection against wind erosion (Bestelmeyer et al., 2015;Schwilch, Liniger, & Hurni, 2014). In many drylands, land cover is threatened by overgrazing, so management of stocking rates and grazing can help to prevent the advance of land degradation (Smith, House, et al., 2016).

| Practices based on land management
Considering the widespread distribution of degraded and desertified lands globally, more than 10 Mkm 2 could benefit from improved management techniques.
Agroforestry can help stabilize soils to prevent land degradation and desertification, so given that there is around 10 Mkm 2 of land with more than 10% tree cover (Garrity, 2012), agroforestry could benefit up to 10 Mkm 2 of land.
Agricultural diversification usually aims to increase climate and food security resilience, for example, through "climate smart agriculture" (Table S4;

Risk sharing instruments
Unquantified but likely to be several million or in degraded/desertified areas, but this value sets the maximum contribution of preventing the conversion of grasslands to croplands, a small global benefit for land degradation and desertification control (Table 11).
Most land degradation processes that are sensitive to climate change pressures (e.g., erosion, decline in soil organic matter, salinization, waterlogging, drying of wet ecosystems) benefit from integrated water management. Integrated water management options include management to reduce aquifer and surface water depletion, and to prevent over extraction, and provide direct co-benefits for prevention of land degradation. Strategies such as water-use efficiency and irrigation improve soil health through increases in soil organic matter content, thereby delivering benefits for prevention Forests are important in helping to stabilize land and regulate water and microclimate (Locatelli, Catterall, et al., 2015;Locatelli, Pavageau, et al., 2015). Based on the extent of forests exposed to degradation (Gibbs & Salmon, 2015) and dry forests at risk of desertification (Bastin et al., 2017;Núñez et al., 2010), the estimated global potential effect for reduced land degradation and avoided desertification is large for both forest management and for reduced deforestation and forest degradation when accumulated until the end of the century (Table 11). Uncertainty in these global estimates is high. More robust estimates are available at regional levels. For example, land management may have contributed to
In Thailand, desertification risk was reduced when bare lands were converted to agriculture and forests, and from non-forests to forests (Wijitkosum, 2016).
Forest restoration is a key option in achieving the overarching aim of reducing land degradation globally, such as through land degradation neutrality (Table S4), not only in drylands (Safriel, 2017 (Melo et al., 2013). The Y Ikatu Xingu campaign in Brazil (launched in 2004) aims to contain deforestation and degradation processes by reversing forest loss on 3,000 km 2 in the Xingu Basin (Durigan, Guerin, & Costa, 2013).
Afforestation, reforestation, and forest restoration are also used to prevent desertification. Forests tend to maintain water and soil quality by reducing runoff and trapping sediments and nutrients (Idris Medugu, Majid, Johar, & Choji, 2010;Salvati, Sabbi, Smiraglia, & Zitti, 2014), but planting of non-native species in semiarid regions can deplete soil water resources if they have high evapotranspiration rates (Feng, Gong, Mei, & Cui, 2016). Afforestation and reforestation programs can be deployed over large areas of the Earth, so can create synergies in areas prone to desertification. Global estimates of land potentially available for afforestation are up to 25.8 Mkm 2 by the end of the century, depending on a variety of assumptions on socioeconomic developments and climate policies (Griscom et al., 2017;Kreidenweis et al., 2016;Popp et al., 2017). The higher end of this range is achieved under the assumption of a globally uniform reward for carbon uptake in the terrestrial biosphere, and is halved by considering tropical and subtropical areas only to minimize albedo feedbacks (Kreidenweis et al., 2016). When safeguards are introduced (e.g., excluding existing cropland for food security, boreal areas, etc.), the area available declines to about 6.8 Mkm 2 (95% confidence interval of 2.3 and 11.25 Mkm 2 ), of which about 4.7 Mkm 2 is in the tropics and 2.1 Mkm 2 is in temperate regions (Griscom et al., 2017;Table 11). These estimates largely overlap with those for forest restoration.
Increasing soil organic matter content is a measure to address land degradation. With around 120 thousand km 2 lost to degradation every year, and over 3.2 billion people negatively impacted by land degradation globally (IPBES, 2018), practices designed to increase soil organic carbon have a large potential to address land degradation (Lal, 2004). With over 2.7 billion people affected globally by desertification (IPBES, 2018), practices to increase soil organic carbon content could be applied to an estimated 11.37 Mkm 2 of desertified land (Lal, 2001; Table 11).  (Table 11). Oldeman, Hakkeling, and Sombroek (1991) estimated the global extent soil affected by salinization is 0.77 Mkm 2 /year, which sets the upper limit on the area that could benefit from measures to address soil salinization. The global extent of chemical soil degradation (salinization, pollution, and acidification) is about 1.03 Mkm 2 (Oldeman et al., 1991) giving the maximum extent of land that could benefit from the management of pollution and acidification (  (Table 11).

Control of soil erosion
Biochar could deliver benefits in efforts to address land degradation and desertification through improving water and nutrient holding capacity (Sohi, 2012;Woolf et al., 2010), and stimulating nutrient cycling and biological activity, but the global effect is not quantified.
Management of landslides and natural hazards aims to control a severe land degradation process affecting sloped and hilly areas, many of them with poor rural inhabitants (Gariano & Guzzetti, 2016), but the global potential has not been quantified.
There are no global data on the impacts of management of invasive species/encroachment on desertification, though the impact is presumed to be positive. There are no global studies examining the potential role of restoration and avoided conversion of coastal wetlands on desertification. However, since degradation of coastal wetlands is widespread, restoration of coastal wetlands could potentially deliver moderate benefits for addressing land degradation, with 0.29 Mkm 2 globally considered feasible for restoration (Griscom et al., 2017;Table 11).
Large areas (0.46 Mkm 2 ) of global peatlands are degraded and so, considered suitable for restoration (Griscom et al., 2017). Thus, peatland restoration could deliver moderate benefits for addressing land degradation (Table 11).
There are no global estimates of the effects of biodiversity conservation on reducing degraded lands. However, at the local scale, biodiversity conservation programs have been demonstrated to stimulate gains in forest cover over large areas over the last three decades (e.g., in China; Zhang et al., 2013). Management of wild animals can influence land degradation processes by grazing, trampling, and compacting soil surfaces, thereby altering surface temperatures and chemical reactions affecting sediment and carbon retention (Cromsigt et al., 2018).
While spreading of crushed minerals onto land as part of enhanced mineral weathering may provide soil/ plant nutrients in nutrient-depleted soils (Beerling et al., 2018), there is no literature reporting on the potential global impacts of this practice in addressing land degradation or desertification.
Large-scale production of bioenergy and BECCS requires significant amounts of land Popp et al., 2017;Smith, Haszeldine, et al., 2016), with as much as 15 Mkm 2 in 2100 in 2°C scenarios (Popp et al., 2017), increasing pressures for land degradation and desertification (Table 11). However, bioenergy production can either increase (Mello et al., 2014;Robertson et al., 2017) or decrease (FAO, 2011;Lal, 2014) soil organic matter, depending on where it is produced and how it is managed. Since no global estimates of these impacts are available, they are not included in the quantification in Table 11.

| Practices based on value chain management
Dietary change and waste reduction both result in decreased cropland and pasture extent (Bajželj et al., 2014;Stehfest et al., 2009;Tilman & Clark, 2014), reducing the pressure for land degradation (Table 12).
Reduced post-harvest losses could spare 1.98 Mkm 2 of cropland globally (Kummu et al., 2012) meaning that land degradation pressure could be relieved from this land area through reduction of postharvest losses. The effects of material substitution on land degradation depend on management practice; some forms of logging can lead to increased land degradation. No studies were found linking material substitution to desertification (Table 13).

| Practices based on risk management
For management of urban sprawl, urban expansion has been identi- Degradation can be a driver of livelihood diversification (Batterbury, 2001;Lestrelin & Giordano, 2007), which can be reversed if diver-  . However, there is conflicting evidence from some areas that that more diverse-income households may also increase land degradation (Palacios et al., 2013;Warren, 2002).
Use of local seeds may play a role in addressing land degradation as they reduce need for inputs such as chemical fertilizers or mechanical tillage (Mousseau, 2015;Reisman, 2017). Some antidesertification programs have also shown more success using local seed varieties (Bassoum & Ghiggi, 2010;Nunes et al., 2016

| Practices based on land management
Increased food productivity has fed many millions of people. Erisman, Sutton, Galloway, Klimont, and Winiwarter (2008), for example, estimated that over 3 billion people worldwide could not have been fed without increased food productivity arising from N fertilization (Table 14).
Improved cropland management to achieve food security aims to close yield gaps by increasing use efficiency of essential inputs such as water and nutrients. Large production increases (45%-70% for most crops) are possible from closing yield gaps to 100% of attainable yield, by optimizing fertilizer use and irrigation, although overuse of nutrients causes adverse environmental impacts (Mueller et al., 2012). This improvement could affect 1,000 million people.
Improved grazing land management includes grasslands, rangelands, and shrublands, and all sites on which pastoralism is practiced.
In general terms, continuous grazing may cause severe damage to topsoil quality through, for example, compaction. This damage may be reversed by short grazing exclusion periods under rotational TA B L E 1 3 Summary of effects on land degradation and desertification of practices based on risk management  Goodwin and Smith (2003), Wright and Wimberly (2013) Abbreviation: BECCS, bioenergy with carbon capture and strorage.
grazing systems (Drewry, 2006;Greenwood & McKenzie, 2001;Taboada et al., 2011). Due to the widespread diffusion of pastoralism, improved grassland management may potentially affect more than 1,000 million people, many of them under subsistence agricultural systems.
Meat, milk, eggs, and other animal products, including fish and other seafoods, will play an important role in achieving food security (Reynolds, Wulster-Radcliffe, Aaron, & Davis, 2015). Improved livestock management with different animal types and feeds may also impact one million people . Ruminants are efficient converters of grass into human edible energy and protein and grassland-based food production can produce food with a comparable carbon footprint to mixed systems (O'Mara, 2012). However, in the future, livestock production will increasingly be affected by competition for natural resources, particularly land and water, competition between food and feed and by the need to operate in a carbon-constrained economy (Thornton, Steeg, Notenbaert, & Herrero, 2009 Agricultural diversification is not always economically viable; technological, biophysical, educational, and cultural barriers may emerge that limit the adoption of more diverse farming systems. Nevertheless, diversification could benefit 1,000 million people, many of them under subsistence agricultural systems (Birthal, Roy, & Negi, 2015;Massawe, Mayes, & Cheng, 2016;Waha et al., 2018).
Integrated water management provides direct benefits to food security by improving agricultural productivity (Godfray & Garnett, 2014;Tilman, Balzer, Hill, & Befort, 2011), thereby potentially affecting the livelihood and well-being of >1,000 million people (Campbell et al., 2016) suffering hunger and highly vulnerable by climate change. Increasing water availability through reliable supply of water for agricultural production using different techniques of water harvesting, storage, and its judicious utilization through farm ponds, dams, and community tanks in rainfed agriculture areas (Rao, Rejani, et al., 2017;Rivera-Ferre et al., 2016), thereby potentially affecting the livelihood and well-being of >1,000 million people (Campbell et al., 2016) suffering hunger and highly vulnerable by climate change.
Forests play a major role in providing food to local communities (non-timber forest products, mushrooms, fodder, fruits, berries, etc.), and diversify daily diets directly or indirectly through improving productivity, hunting, diversifying tree-cropland-livestock systems, and grazing in forests. Based on the extent of forest contributing to food supply, considering the people undernourished (FAO, IFAD, & WFP, 2015;Rowland, Ickowitz, Powell, Nasi, & Sunderland, 2017) and the annual deforestation rate (Keenan et al., 2015), the global potential to enhance food security is moderate for improved forest management and large for reduced deforestation (Table 14).
Deforestation could reduce local precipitation by 20%, severely impacting non-irrigated agricultural lands (Lawrence & Vandercar, 2015). A 20% decrease in water availability close to tropical forests could impact 100s of millions of people. For example, if 50% of the Amazon and Congo Basins were deforested, 115 million people would be impacted given that the population of people within ~1,000 km of these basins is 578 million people, if only 20% of the population is negatively impacted. Impacts on people in other countries affected by teleconnections or exports are not included in this conservative estimate, which is also conservative since 60% of population in Congo Basin are farmers, most on unirrigated farms with large poor population centers, and 10% of people in South America work in the agriculture sector with large population centers relying on food produced close to the Amazon region. Reduced deforestation and degradation could therefore deliver benefits for food security for many more than 100 million people (Table 14).
The uncertainty of these global estimates is high. More robust qualitative and some quantitative estimates are available at the regional level. For example, managed natural forests, shifting cultivation, and agroforestry systems are demonstrated to be crucial to food security and nutrition for hundreds of million people in rural landscapes worldwide (Sunderland et al., 2013;Vira, Wildburger, & Mansourian, 2015). According to Erb et al. (2016), deforestation would not be needed to feed the global population by 2050, in terms of quantity and quality of food. At the local level, Cerri et al. (2018) suggested that reduced deforestation, along with integrated cropland-livestock management, would positively affect more than 120 million people in the Cerrado, Brazil. In sub-Saharan Africa, where population and food demand are projected to continue to rise substantially, reduced deforestation may have strong positive effects on food security .
Afforestation and reforestation may negatively affect food security (Boysen, Lucht, & Gerten, 2017;Frank et al., 2017;Kreidenweis et al., 2016). It is estimated that large-scale afforestation plans could increase food prices by 80% by 2050 (Kreidenweis et al., 2016), and more general mitigation measures in the agriculture, forestry and other land use sector could cause undernourishment in 80-300 million people (Frank et al., 2017;Table 14). For reforestation, the potential adverse side affects with food security are smaller than afforestation, because forest regrows on recently deforested areas, and its impact would be felt mainly through impeding possible expansion of agricultural areas.
On a smaller scale and when implemented sustainably, forested land also offers benefits in terms of food supply, especially when forest is established on degraded land and other land that cannot be used for agriculture. For example, food from forests represents a safety net during times of food and income insecurity (Wunder, Angelsen, & Belcher, 2014), and wild-harvested meat and fish provide 30%-80% of protein intake from many rural communities (McIntyre, Liermann, & Revenga, 2016;Nasi, Taber, & Vliet, 2011). An example of how an afforestation/reforestation program has improved food security for >100 million people is the "Grain for Green" program in China. The results indicate that the area of land affected by heavy and severe soil erosion has decreased by 55.2% and 53.6%, respectively, while the water holding capacity was 25.2% higher in 2009 than that in 1990. Increased grain yields and agricultural productivity have been recorded following Grain for Green (Yao & Li, 2010), and the results strongly indicate a positive impact of cropland conversion on soil C stocks (which can increase fertility and soil water retention; Deng, Liu, & Shangguan, 2014). Most studies concur that the physical properties of the soil, including soil fertility, porosity, and nutrients, have improved, and soil erosion and river sedimentation have slowed down (Delang & Yuan, 2015).  Table 14).
Increasing soil organic matter stocks can increase yield and improve yield stability (Lal, 2006;Pan, Smith, & Pan, 2009;Soussana et al., 2019), though this is not universally seen (Hijbeek et al., 2017). Lal (2006) concludes that crop yields can be increased by 20-70, 10-50, and 30-300 kg/ha for maize for wheat, rice, and maize, respectively, for every 1 t C/ha increase in soil organic carbon in the root zone. Increasing soil organic carbon by 1 t C/ha could increase food grain production in developing countries by 32 Mt/year (Lal, 2006). Frank et al. (2017) estimate that soil carbon sequestration could reduce calorie loss associated with agricultural mitigation measures by 65%, saving 60-225 million people from undernourishment compared to a baseline without soil carbon sequestration (Table 14). Lal (1998) (Table 14).
Although there are biophysical barriers, such as access to appropriate water sources and limited productivity of salt-tolerant crops, prevention/reversal of soil salinization could benefit 1-100 million people (Qadir et al., 2013). Soil compaction affects crop yields, so prevention of soil compaction could benefit an estimated 1-100 million people globally (Anderson & Peters, 2016).
Biochar, on balance, could provide moderate benefits for food security by improving yields by 25% in the tropics, but with more limited impacts in temperate regions (Jeffery et al., 2017), or through improved water holding capacity and nutrient use efficiency (Sohi, 2012). These benefits could, however, be tempered by additional pressure on land if large quantities of biomass are required as feedstock for biochar production, thereby causing potential conflicts with food security (Smith, 2016). Smith (2016) estimated that 0.4-2.6 Mkm 2 of land would be required for biomass feedstock to deliver 2.57 Gt CO 2 eq/year of CO 2 removal. If biomass production occupied 2.6 Mkm 2 of cropland, equivalent to around 20% of the global cropland area, this could potentially have a large effect on food security, although Woolf et al. (2010) argue that abandoned cropland could be used to supply biomass for biochar, thus avoiding competition with food production. Similarly, Woods et al. (2015) estimate that 5-9 Mkm 2 of land is available for biomass production without compromising food security and biodiversity, considering marginal and degraded land and land released by pasture intensification (Table 14). Assuming annual grain consumption per capita to be 300 kg/year (estimated based on data included in FAO, 2018;FAO et al., 2018;Pradhan et al., 2013;World Bank, 2018), the loss of 18.6 Mt/year would remove cereal crops equivalent to that consumed by 62 million people, providing an estimate of the potential of fire management to contribute to food security (Table 14).
Landslides and other natural hazards affect 1-100 million people globally, so preventing them could provide food security benefits to this many people.
In terms of measures to tackle pollution, including acidification, Shindell et al. (2012) considered about 400 emission control measures to reduce ozone and BC. This strategy increases annual crop yields by 30-135 Mt due to ozone reductions in 2030 and beyond.
If annual grain consumption per capita is assumed as 300 kg/year (estimated based on data included in FAO, 2018;FAO et al., 2018;Pradhan et al., 2013;World Bank, 2018), increase in annual crop yields by 30-135 Mt feeds 100-450 million people.
There are no global data on the impacts of management of invasive species/encroachment on food security.
Since large areas of converted coastal wetlands are used for food production (e.g., mangroves converted for aquaculture; (Naylor et al., 2000), restoration of coastal wetlands could potentially displace food production and damage local food supply, potentially leading to adverse impacts on food security, though these effects are likely to be very small given that a small proportion of human food comes from the oceans and other aquatic ecosystems (Pimentel, 2006). These impacts could be offset by careful management, such as the careful siting of ponds within mangroves (Naylor et al., 2000; (Smith et al., 2015).
However, at the same time, some options to preserve biodiversity, such as protected areas, may potentially conflict with food production by local communities (Molotoks, Kuhnert, Dawson, & Smith, 2017).
The spreading of crushed minerals on land as part of enhanced mineral weathering on nutrient-depleted soils can potentially increase crop yield by replenishing plant available silicon, potassium, and other nutrients (Beerling et al., 2018), but there are no estimates of the potential magnitude of this effect for global food production.
Although Woods et al. (2015) estimate that 5-9 Mkm 2 of land could be available for bioenergy feedstock production without compromising food security or biodiversity, competition for land between bioenergy and food crops could lead to adverse side effects for food security. Many studies indicate that bioenergy/BECCS could increase food prices (Calvin et al., 2014;Popp et al., 2017;Wise et al., 2009). Only three studies were found that link bioenergy to the population at risk of hunger, but they estimate an increase in this population of between 2 million and 150 million people (Table 14).

| Practices based on value chain management
Dietary change can free up agricultural land for additional production (Bajželj et al., 2014;Stehfest et al., 2009;Tilman & Clark, 2014) and reduce the risk of some diseases (Aleksandrowicz et al., 2016;Tilman & Clark, 2014), with large positive impacts on food security (Table 15). Kummu et al. (2012) estimate that an additional 1 billion people could be fed if food waste was halved globally. This includes both post-harvest losses and retail and consumer waste, and measures such as improved food transport and distribution (Table 15).
While no studies quantified the effect of material substitution on food security, the effects are expected to be similar to reforestation and afforestation if the amount of material substitution leads to an increase in forest area.
Since 821 Kummu et al. (2012) Enhanced urban food systems Up to 1,260 million people Low confidence Benis andFerrão (2017), de Zeeuw andDrechsel (2015), Padgham, Jabbour, and Dietrich (2014), Specht et al. (2014) Improved food processing and retailing 500 million people Low confidence World Bank (2017) Improved energy use in food systems Up to 2,500 million people Low confidence IEA (2014) (Tayleur et al., 2017). Supply chain management has a direct effect on food security; for example, food price spikes affect food security and health, with clearly documented effects of stunting among young children as a result of the 2007-2008 food supply crisis (Arndt, Hussain, & Østerdal, 2012;Brinkman et al., 2009;de Brauw, 2011;Darnton-Hill & Cogill, 2010) with a 10% increase in wasting attributed to the crisis in South Asia alone (Vellakkal et al., 2015). There is conflicting evidence on the impacts of different food price stability options for supply chains, and little quantification of these (Alderman, 2010;Byerlee, Jayne, & Myers, 2006;del Ninno, Dorosh, & Subbarao, 2007;von Braun, Algieri, & Kalkuhl, 2014). Reduction in staple food prices due to price stabilization resulted in more expenditure on other foods and increased nutrition (e.g., oils, animal products), leading to a 10% reduction in malnutrition among children in one study (Torlesse et al., 2003), while protectionist policies (food price controls) and safety nets to reduce price instability resulted in a 20% decrease in risk of malnutrition in another (Nandy, Daoud, & Gordon, 2016). Models using policies for food aid and domestic food reserves to achieve food supply and price stability showed the highest effectiveness of all options in achieving climate mitigation and food security goals (e.g., more effective than carbon taxes) as they did not exacerbate food insecurity and did not reduce ambitions for achieving temperature goals (Fujimori et al., 2019).
For urban food systems, increased food production in cities combined with governance systems for distribution and access can improve food security, with a potential to produce 30% of food consumed in cities. The urban population in 2018 was 4.2 billion people, so 30% represents 1,230 million people who could benefit in terms of food security from improved urban food systems (Table 15).
It is estimated that 500 million smallholder farmers depend on agricultural businesses in developing countries (World Bank, 2017), which set the maximum number of people who could benefit from improved food processing and retailing.
Up to 2,500 million people could benefit from improved energy efficiency in agriculture, based on the estimated number of people worldwide lacking access to clean energy and instead relying on biomass fuels for their household energy needs (IEA, 2014).

| Practices based on risk management
Unregulated urban sprawl can affect food security; highly productive soils have experienced the highest rate of conversion of any soil type in the United States (Nizeyimana et al., 2001). Specific types of agriculture are often practiced in urban-influenced fringes, such as fruits, vegetables, and poultry and eggs, the loss of which can have an impact on the types of nutritious foods available in urban areas (Francis et al., 2012). China experienced a loss of 30 Mt of grain production from 1998 to 2003 attributed to urbanization (Chen, 2007).
However, overall global quantification has not been attempted (Table 16).
Livelihood diversification is associated with increased welfare and incomes and decreased levels of poverty in several country studies (Arslan et al., 2018;Asfaw, Pallante, & Palma, 2018). These are likely to have large food security benefits (Barrett, Reardon, & Webb, 2001;Niehof, 2004), but there is little global quantification.
Use of local seeds can provide considerable benefits for food security because of the increased ability of farmers to revive and strengthen local food systems (McMichael & Schneider, 2011); studies have reported more diverse and healthy food in areas with strong food sovereignty networks (Bisht et al., 2018;Coomes et al., 2015).
Women in particular may benefit from seed banks for low value, but nutritious crops (Patnaik, Jongerden, & Ruivenkamp, 2017).
However, there may be lower productivity yields from local and unimproved seeds, so the overall impact on food security is ambiguous (McGuire & Sperling, 2016).
Disaster risk management approaches can have important impacts on reducing food insecurity, and current systems for drought warning and other storms currently reach over 100 million people.
When these early warning systems help farmers harvest crops in advance of impending weather events, or make agricultural decisions to prepare for adverse events, they are likely to have positive impacts on food security (Fakhruddin, Kawasaki, & Babel, 2015).

Famine early warning systems have been successful in Sahelian
Africa to alert authorities of impending food shortages so that food acquisition and transportation from outside the region can begin, potentially helping millions of people (Genesio et al., 2011;Hillbruner & Moloney, 2012 (Claassen et al., 2011;Goodwin, Vandeveer, & Deal, 2004).
3.5 | Summary of the potentials of practices across mitigation, adaptation, desertification, land degradation, and food security Table 17 provides a summary of the potentials of practices across mitigation, adaptation, desertification, land degradation, and food security, using the thresholds given in Table 4.

| D ISCUSS I ON
Understanding the potential of practices to address the land challenges is extremely important in supporting ongoing, near-term, future policy-making (e.g., Paris Agreement) and to attempt to bridge the gap between science, policy makers, and the general public.
Moreover, the main findings are obtained by an extended literature review, which makes the study comprehensive (40 options across four land challenges) and as robust as possible (thousands of items of information). Indeed, such a wide-ranging and inclusive assessment has not previously been conducted. The main findings, limitations, and conclusions are presented below.

| Co-delivery across the land challenges
Nine options deliver medium to large benefits for all four land chal- Most agricultural land management practices (except for reduced grassland conversion to cropland, which potentially adversely affects food security), deliver benefits across the four land challenges.
Among the forest land management options, afforestation and reforestation have the potential to deliver large co-benefits across all land challenges except potentially for food security, where the evidence is mixed. Some studies suggest possible adverse impacts of afforestation/reforestation on food security due to adverse impacts on TA B L E 17 Summary of the global potentials of practices across mitigation, adaptation, desertification, land degradation, and food security, using the thresholds given in Table 4  Note: Cell colors correspond to the large, moderate, and small categories shown in Table 4. Dark blue = large positive; mid-blue = moderate positive; light blue = small positive; no color = no effect; light red = small negative; mid-red = moderate negative; dark red = large negative; green = variable.
Hatching for the cell showing land degradation and desertification impacts of Bioenergy and BECCS indicates uncertainty in the magnitude of the negative impact; while large-scale production of bioenergy could require up to 15 Mkm 2 in 2100 in 2°C scenarios, it is not known how much of this land would be degraded/desertified by such land use change. Letters in cells: l, m, and h correspond to low, medium, and high confidence that the largest estimated potential is within the indicated magnitude category. ND = no data on global impact (even though regional data may exist).
food prices (Kreidenweis et al., 2016), while others suggest that food productivity can be increased by reducing soil erosion and increasing agricultural productivity (Yao & Li, 2010). Among the soil-based practices, some global data are missing, but none except biochar (if large areas are dedicated to feedstock production) shows any potential for negative impacts. Potential negative impacts could arise from additional pressure on land if large quantities of biomass feedstock are required for biochar production (Smith, 2016), through land competition can be minimized by sustainable location and management (Woolf et al., 2010), and biochar addition to soils can improve productivity (Jeffery et al., 2017). Where global data exist, most practices in other/all ecosystems deliver benefits except for a potential moderate negative impact on food security by restoring peatlands currently used for agriculture. Of the two practices specifically targeted at CDR, there are missing data for enhanced weathering of minerals for three of the land challenges, but large-scale bioenergy and BECCS show a potential large benefit for mitigation, but small to large adverse impacts on the other three land challenges, mainly driven by increased pressure on land due to feedstock demand, though again, this could be managed by sustainable location and management of the land used for feedstock production (Woods et al., 2015).
While data allow the impact of material substitution to be assessed only for mitigation, the three other demand-side practices: dietary change, reduced post-harvest losses, and reduced food waste provide large or moderate benefits across all land challenges for which data exist. Data are lacking to assess the impact of the supply-side practices on more than three of the land challenges, but there are large to moderate benefits for all those for which data are available. Data are not available to assess the impact of risk management-based practices on all of the land challenges, but there are small to large benefits for all of those for which data are available.

| Study limitations and data/knowledge gaps
The analysis presented here is based on an aggregation of information from studies with a wide variety of assumptions about how response options are implemented and the contexts in which they occur. Response options implemented differently at local to global scales could lead to different outcomes. The potential magnitude of impacts of each practice is assessed using values from the literature, many of which may consider potentials in isolation of other practices. While some practices may be compatible with others, it is not possible to add the potentials together, since many are known not to be additive. Furthermore, a number of practices are mutually exclusive since they cannot be practiced on the same land, for example, afforestation cannot be practiced on the same land as cropland management. In addition, the potentials of practices quoted in literature overlap between options. For example, a component of the potential of cropland management for mitigation or adaptation may arise from soil carbon sequestration, for which there are separate estimates of potential. As a result of these issues, the potentials for each practice cannot be simply summed to get a total global potential for any of the land challenges. Assessing the combined potential requires that the practices be considering in the same framework that conserves land, excludes mutually exclusive practices on the same land area, and considers nonoverlapping practices, so cannot be done with a purely literature-based approach.
Assessing the magnitude at global scale means that many important, context-specific interactions, for example, by location, ecosystem type, administrative unit, cannot be accounted for, and that important regional data have not been condoered. In terms of knowledge gaps, most of the practices for which information was available have medium to high positive potential for addressing land challenges (see Table 17). However, many of the estimates have low to medium confidence and many options have no data, showing that there are considerable knowledge gaps. Knowledge of the impacts of some practice-land challenge relationships is more robust and well established in the scientific literature or other information sources (statistics, inventory data) than others (e.g., high confidence: "h" in Table 17), such as increased food productivity with food security, and reduced deforestation and forest degradation with mitigation).
The low to medium confidence may also derive from some flexibility related to the criteria used to define magnitude of impact of each practice (see Table 4). For example, magnitude criteria needed to be defined to be comparable across options from different sectors (agriculture, forestry, soil), but in defining them in this way, the interpretation of the effects of each contribution to specific land challenges may be oversimplified, (see e.g., "low confidence" for forest management and reduced deforestation and forest degradation for all land challenges except for mitigation). Furthermore, the magnitude of contribution (low, medium, high) and trend (positive, negative) may have been affected by the selected criteria (see Table 4; e.g., relevant information not found for missing thresholds).
Many practices are known to be important for at least one land challenge by lack global estimates of potential across the other land challenges, even if an impact has been demonstrated at regional level (hence the large number of "no data on global impact" cells in Table 17). This particularly affects the supply chain and risk management practices but also affects some land management practices. For example, there are no global estimates of the potential for management of invasive species/encroachment for any of the land challenges, despite its acknowledged benefits for preventing land degradation and desertification locally. We have retained it in the list of practices to acknowledge its potential importance, and to highlight the knowledge gap of its impact at global scale.

| Conclusions
Most mitigation practices can be applied without competing for available land and have the potential to provide multiple co-benefits. Although most practices can be applied without competing for available land, some, such as land to provide feedstock for bioenergy/ BECCS (and under some circumstances, large-scale afforestation), could potentially increase demand for land conversion. If applied at scales necessary to remove CO 2 from the atmosphere at the scales of several Gt CO 2 eq/year, this increased demand for land could lead to adverse side effects for adaptation, food security, and potentially on land degradation and desertification, so safeguards are required to ensure that expansion of energy crops does not impact natural systems and food security. If applied on a limited share of total land and integrated into sustainably managed landscapes, there will be fewer adverse side effects and some positive co-benefits could be realized.
Reduced grassland conversion to croplands, restoration and reduced conversion of peatlands, and restoration and reduced conversion of coastal wetlands affect smaller land areas globally, so the impacts of these options are smaller globally, but could be locally significant.
Further scientific efforts are thus needed to provide policy with robust, comprehensive, and transparent approaches, models, and tools for land use forecasting, incorporating multiple side effects, that is, biophysical, economic, and social. While policies and respective support from the scientific community remain sectoral, cross-linkages between sustainable land management and human well-being may be missed.

ACK N OWLED G EM ENTS
The input of P.S. contributes to the following UKRI-funded projects: