Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Table of Contents 2 Chapter 7: Agriculture, Forestry and Other Land Uses (AFOLU) ................................................ 7-1 3 Executive summary.......................................................................................................................... 7-4 4 7.1 Introduction .............................................................................................................................. 7-8 5 7.1.1 Key findings from previous reports ............................................................................. 7-8 6 7.1.2 Boundaries, scope and changing context of the current report .................................... 7-9 7 7.2 Historical and current trends in GHG emission and removals; their uncertainties and 8 implications for assessing collective climate progress................................................................... 7-11 9 7.2.1 Total net GHG flux from AFOLU ............................................................................. 7-12 10 7.2.2 Flux of CO2 from AFOLU, and the non-anthropogenic land sink ............................. 7-15 11 Cross-Chapter Box 6 Implications of reconciled anthropogenic land CO2 fluxes for assessing 12 collective climate progress in the global stocktake .................................................................... 7-21 13 7.2.3 CH4 and N2O flux from AFOLU................................................................................ 7-23 14 7.2.4 Biophysical effects and short-lived climate forcers ................................................... 7-27 15 7.3 Drivers.................................................................................................................................... 7-28 16 7.3.1 Anthropogenic direct drivers – Deforestation, conversion of other ecosystems, and land 17 degradation................................................................................................................................. 7-29 18 7.3.2. Anthropogenic direct drivers – Agriculture ............................................................... 7-34 19 7.3.3. Indirect drivers ........................................................................................................... 7-36 20 7.4. Assessment of AFOLU mitigation measures including trade-offs and synergies ................. 7-39 21 7.4.1. Introduction and overview of mitigation potential..................................................... 7-39 22 7.4.2. Forests and other ecosystems ..................................................................................... 7-48 23 7.4.3. Agriculture ................................................................................................................. 7-61 24 7.4.4. Bioenergy and BECCS............................................................................................... 7-77 25 7.4.5. Demand-side measures .............................................................................................. 7-81 26 7.5. AFOLU Integrated Models and Scenarios ............................................................................. 7-86 27 7.5.1. Regional GHG emissions and land dynamics ............................................................ 7-86 28 7.5.2. Marginal abatement costs according to integrated assessments................................. 7-89 29 7.5.3. Interaction between mitigation in the AFOLU sector and other SDGs in the context of 30 integrated assessments ............................................................................................................... 7-92 31 7.5.4. Regional AFOLU abatement for different carbon prices ........................................... 7-92 32 7.5.5. Illustrative mitigation pathways ................................................................................. 7-93 33 7.6. Assessment of economic, social and policy responses .......................................................... 7-95 34 7.6.1. Retrospective in policy efforts and achieved mitigation within AFOLU .................. 7-95 35 7.6.2. Review of observed policies and policy instruments ................................................. 7-99 36 7.6.3. Assessment of current policies and potential future approaches.............................. 7-108 37 7.6.4. Barriers and opportunities for AFOLU mitigation................................................... 7-111 Do Not Cite, Quote or Distribute 7-2 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.6.5. Linkages to ecosystem services, human well-being and adaptation (incl. SDGs) ... 7-116 2 7.7. Knowledge gaps ................................................................................................................... 7-121 3 Frequently asked questions .......................................................................................................... 7-122 4 References .................................................................................................................................... 7-125 5 Do Not Cite, Quote or Distribute 7-3 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Executive summary 2 The Agriculture, Forestry and Other Land Uses1 (AFOLU) sector encompasses managed 3 ecosystems and offers significant mitigation opportunities while delivering food, wood and other 4 renewable resources as well as biodiversity conservation, provided the sector adapts to climate 5 change. Land-based mitigation measures represent some of the most important options currently 6 available. They can both deliver carbon dioxide removal (CDR) and substitute for fossil fuels, thereby 7 enabling emissions reductions in other sectors. The rapid deployment of AFOLU measures is essential 8 in all pathways staying within the limits of the remaining budget for a 1.5◦C target (high confidence). 9 Where carefully and appropriately implemented, AFOLU mitigation measures are uniquely positioned 10 to deliver substantial co-benefits and help address many of the wider challenges associated with land 11 management. If AFOLU measures are deployed badly then, when taken together with the increasing 12 need to produce sufficient food, feed, fuel and wood, they may exacerbate trade-offs with the 13 conservation of habitats, adaptation, biodiversity and other services. At the same time the capacity of 14 the land to support these functions may be threatened by climate change itself (high confidence). {WGI, 15 Figure SPM7; WGII, 7.1, 7.6} 16 The AFOLU (managed land) sector, on average, accounted for 13-21% of global total 17 anthropogenic greenhouse gas (GHG) emissions in the period 2010-2019 (medium confidence). At 18 the same time managed and natural terrestrial ecosystems were a carbon sink, absorbing around 19 one third of anthropogenic CO2 emissions (medium confidence). Estimated anthropogenic net CO2 20 emissions from AFOLU (based on bookkeeping models) result in a net source of +5.9 ± 4.1 GtCO2 yr- 1 21 between 2010 and 2019 with an unclear trend. Based on FAOSTAT or national GHG inventories, the 22 net CO2 emissions from AFOLU were 0.0 to +0.8 GtCO2 yr-1 over the same period. There is a 23 discrepancy in the reported CO2 AFOLU emissions magnitude because alternative methodological 24 approaches that incorporate different assumptions are used. If the managed and natural responses of all 25 land to both anthropogenic environmental change and natural climate variability, estimated to be a gross 26 sink of -12.5 ± 3.2 GtCO2 yr-1 for the period 2010–2019, are included with land use emisisons, then 27 land overall, constituted a net sink of -6.6 ± 5.2 GtCO2 yr-1 in terms of CO2 emissions (medium 28 confidence). {WGI; 7.2, 7.2.2.5, Table 7.1} 29 AFOLU CO2 emission fluxes are driven by land use change. The rate of deforestation, which 30 accounts for 45% of total AFOLU emissions, has generally declined, while global tree cover and 31 global forest growing stock levels are likely increasing (medium confidence). There are substantial 32 regional differences, with losses of carbon generally observed in tropical regions and gains in temperate 33 and boreal regions. Agricultural CH4 and N2O emissions are estimated to average 157 ± 47.1 MtCH4 34 yr-1 and 6.6 ± 4.0 MtN2O yr-1 or 4.2 ± 1.3 and 1.8 ± 1.1 GtCO2-eq yr-1 (using IPCC AR6 GWP100 values 35 for CH4 and N2O) respectively between 2010 and 2019. AFOLU CH4 emissions continue to increase 36 (high confidence), the main source of which is enteric fermentation from ruminant animals (high 37 confidence). Similarly, AFOLU N2O emissions are increasing, dominated by agriculture, notably from 38 manure application, nitrogen deposition, and nitrogen fertiliser use (high confidence). In addition to 39 being a source and sink for GHG emissions, land plays an important role in climate through albedo 40 effects, evapotranspiration and volatile organic compounds (VOCs) and their mix, although the FOOTNOTE 1 For the AFOLU Sector, anthropogenic greenhouse gas emissions and removals by sinks are defined as all those occurring on ‘managed land’. Managed land is land where human interventions and practices have been applied to perform production, ecological or social functions. Do Not Cite, Quote or Distribute 7-4 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 combined role in total climate forcing is unclear and varies strongly with bioclimatic region and 2 management type. {2.4.2.5, 7.2, 7.2.1, 7.2.3, 7.3} 3 The AFOLU sector offers significant near-term mitigation potential at relatively low cost but 4 cannot compensate for delayed emission reductions in other sectors. (high evidence, medium 5 agreement). The AFOLU sector can provide 20–30% (interquartile range) of the global mitigation 6 needed for a 1.5 or 2oC pathway towards 2050 (robust evidence, medium agreement), though there are 7 highly variable mitigation strategies for how AFOLU potential can be deployed for achieving climate 8 targets. The estimated likely economic (< USD100 tCO2-eq-1) AFOLU sector mitigation potential is 8 9 to 14 GtCO2-eq yr-1 between 2020-2050, with the bottom end of this range representing the mean from 10 integrated assessment models (IAMs) and the upper end representing the mean estimate from global 11 sectoral studies. The economic potential is about half of the technical potential from AFOLU, and about 12 30-50% could be achieved under USD20 tCO2-eq-1. The implementation of robust measurement, 13 reporting and verification processes is paramount to improving the transparency of net-carbon-stock- 14 changes per land unit to prevent misleading assumptions or claims on mitigation. {7.1, 7.4, 7.5} 15 Between 2020 and 2050, mitigation measures in forests and other natural ecosystems provide the 16 largest share of the economic (up to USD100 tCO2-eq-1) AFOLU mitigation potential, followed by 17 agriculture and demand-side measures (high confidence). In the global sectoral studies, the 18 protection, improved management, and restoration of forests, peatlands, coastal wetlands, savannas and 19 grasslands have the potential to reduce emissions and/or sequester 7.3 mean (3.9–13.1 range) GtCO2- 20 eq yr-1. Agriculture provides the second largest share of the mitigation potential, with 4.1 (1.7–6.7) 21 GtCO2-eq yr-1 (up to USD100 tCO2-eq-1) from cropland and grassland soil carbon management, 22 agroforestry, use of biochar, improved rice cultivation, and livestock and nutrient management. 23 Demand-side measures including shifting to sustainable healthy diets, reducing food waste, and 24 building with wood and biochemicals and bio-textiles have a mitigation potential of 2.2 (1.1–3.6) 25 GtCO2-eq yr-1 . Most mitigation options are available and ready to deploy. Emissions reductions can be 26 unlocked relatively quickly, whereas CDR needs upfront investment. Sustainable intensification in 27 agriculture, shifting diets, and reducing food waste could enhance efficiencies and reduce agricultural 28 land needs, and are therefore critical for enabling supply-side measures such as reforestation, 29 restoration, as well as decreasing CH4 and N2O emissions from agricultural production. In addition, 30 emerging technologies (e.g., vaccines or inhibitors) have the potential to substantially increase CH4 31 mitigation potential beyond current estimates. AFOLU mitigation is not only relevant in countries with 32 large land areas. Many smaller countries and regions, particularly with wetlands, have 33 disproportionately high levels of AFOLU mitigation potential density. {7.4, 7.5} 34 The economic and political feasibility of implementing AFOLU mitigation measures is hampered 35 by persistent barriers. Assisting countries to overcome barriers will help to achieve significant 36 short-term mitigation (medium confidence). Finance forms a critical barrier to achieving these gains 37 as currently mitigation efforts rely principally on government sources and funding mechanisms which 38 do not provide sufficient resources to enable the economic potential to be realised. Differences in 39 cultural values, governance, accountability and institutional capacity are also important barriers. 40 Climate change could also emerge as a barrier to AFOLU mitigation, although the IPCC WGI 41 contribution to AR6 indicated that an increase in the capacity of natural sinks may occur, despite 42 changes in climate (medium confidence). The continued loss of biodiversity makes ecosystems less 43 resilient to climate change extremes and this may further jeopardise the achievement of the AFOLU 44 mitigation potentials indicated in this chapter (WGII and IPBES) (high confidence). {WGI Figure 45 SPM7; 7.4, 7.6} Do Not Cite, Quote or Distribute 7-5 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Bioenergy and other biobased options represent an important share of the total mitigation 2 potential. The range of recent estimates for the technical bioenergy potential when constrained 3 by food security and environmental considerations is 5–50 and 50–250 EJ yr-1 by 2050 for residues 4 and dedicated biomass production system respectively. These estimates fall within previously 5 estimated ranges (medium agreement). Poorly planned deployment of biomass production and 6 afforestation options for in-forest carbon sequestration may conflict with environmental and social 7 dimensions of sustainability (high confidence). The global technical CDR potential of BECCS by 2050 8 (considering only the technical capture of CO2 and storage underground) is estimated at 5.9 mean (0.5- 9 11.3) GtCO2 yr-1, of which 1.6 (0.8-3.5) GtCO2 yr-1 is available at below USD100 tCO2-1 (medium 10 confidence). Bioenergy and other bio-based products provide additional mitigation through the 11 substitution of fossil fuels fossil based products (high confidence). These substitution effects are 12 reported in other sectors. Wood used in construction may reduce emissions associated with steel and 13 concrete use. The agriculture and forestry sectors can devise management approaches that enable 14 biomass production and use for energy in conjunction with the production of food and timber, thereby 15 reducing the conversion pressure on natural ecosystems (medium confidence). {7.4} 16 The deployment of all land-based mitigation measures can provide multiple co-benefits, but there 17 are also risks and trade-offs from misguided or inappropriate land management (high 18 confidence). Such risks can best be managed if AFOLU mitigation is pursued in response to the 19 needs and perspectives of multiple stakeholders to achieve outcomes that maximize synergies 20 while limiting trade-offs (medium confidence). The results of implementing AFOLU measures are 21 often variable and highly context specific. Depending on local conditions (e.g., ecosystem, climate, 22 food system, land ownership) and management strategies (e.g., scale, method), mitigation measures 23 have the potential to positively or negatively impact biodiversity, ecosystem functioning, air quality, 24 water availability and quality, soil productivity, rights infringements, food security, and human 25 wellbeing. Mitigation measures addressing GHGs may also affect other climate forcers such as albedo 26 and evapotranspiration. Integrated responses that contribute to mitigation, adaptation, and other land 27 challenges will have greater likelihood of being successful (high confidence); measures which provide 28 additional benefits to biodiversity and human well-being are sometimes described as ‘Nature-based 29 Solutions’. {7.1, 7.4, 7.6} 30 AFOLU mitigation measures have been well understood for decades but deployment remains 31 slow and emissions trends indicate unsatisfactory progress despite beneficial contributions to 32 global emissions reduction from forest-related options (high confidence). Globally, the AFOLU 33 sector has so far contributed modestly to net mitigation, as past policies have delivered about 0.65 34 GtCO2 yr-1 of mitigation during 2010–2019 or 1.4% of global gross emissions (high confidence). The 35 majority (>80%) of emission reduction resulted from forestry measures (high confidence). Although 36 the mitigation potential of AFOLU measures is large from a biophysical and ecological perspective, its 37 feasibility is hampered by lack of institutional support, uncertainty over long-term additionality and 38 trade-offs, weak governance, fragmented land ownership, and uncertain permanence effects. Despite 39 these impediments to change, AFOLU mitigation options are demonstrably effective and with 40 appropriate support can enable rapid emission reductions in most countries. {7.4, 7.6} 41 Concerted, rapid and sustained effort by all stakeholders, from policy makers and investors to 42 land owners and managers is a pre-requisite to achieving high levels of mitigation in the AFOLU 43 sector (high confidence). To date USD0.7 billion yr-1 is estimated to have been spent on AFOLU 44 mitigation. This is well short of the more than USD400 billion yr-1 that is estimated to be necessary to 45 deliver the up to 30% of global mitigation effort envisaged in deep mitigation scenarios (medium 46 confidence). This estimate of the global funding requirement is smaller than current subsidies provided Do Not Cite, Quote or Distribute 7-6 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 to agriculture and forestry. Making this funding available would require a change in flows of money 2 and determination of who pays. A gradual redirection of existing agriculture and forestry subsidies 3 would greatly advance mitigation. Effective policy interventions and national (investment) plans as part 4 of Nationally Dertermined Contributions (NDCs), specific to local circumstances and needs, are 5 urgently needed to accelerate the deployment of AFOLU mitigation options. These interventions are 6 effective when they include funding schemes and long-term consistent support for implementation with 7 governments taking the initiative together with private funders and non-state actors. {7.6} 8 Realizing the mitigation potential of the AFOLU sector depends strongly on policies that directly 9 address emissions and drive the deployment of land-based mitigation options, consistent with 10 carbon prices in deep mitigation scenarios (high confidence). Examples of successful policies and 11 measures include establishing and respecting tenure rights and community forestry, improved 12 agricultural management and sustainable intensification, biodiversity conservation, payments for 13 ecosystem services, improved forest management and wood chain usage, bioenergy, voluntary supply 14 chain management efforts, consumer behaviour campaigns, private funding and joint regulatory efforts 15 to avoid e.g., leakage. The efficacy of different policies, however, will depend on numerous region- 16 specific factors. In addition to funding, these factors include governance, institutions, long-term 17 consistent execution of measures, and the specific policy setting (high confidence). {7.6} 18 There is a discrepancy, equating to 5.5 GtCO2 yr-1 between alternative methods of accounting for 19 anthropogenic land CO2 fluxes. Reconciling these methods greatly enhances the credibility of 20 AFOLU-based emissions offsetting. It would also assist in assessing collective progress in a global 21 stocktake (high confidence). The principal accounting approaches are National GHG inventories 22 (NGHGI) and global modelling approaches. NGHGI, based on IPCC guidelines, consider a much larger 23 area of forest to be under human management than global models. NGHGI consider the fluxes due to 24 human-induced environmental change on this area to be anthropogenic and are thus reported. Global 25 models2, in contrast, consider these fluxes to be natural and are excluded from the total reported 26 anthropogenic land CO2 flux. To enable a like-with-like comparison, the remaining cumulative global 27 CO2 emissions budget can be adjusted (medium confidence). In the absence of these adjustments, 28 collective progress would appear better than it is. {Cross-Chapter Box 6 in this Chapter, 7.2} 29 Addressing the many knowledge gaps in the development and testing of AFOLU mitigation 30 options can rapidly advance the likelihood of achieving sustained mitigation (high confidence). 31 Research priorities include improved quantification of anthropogenic and natural GHG fluxes and 32 emissions modelling, better understanding of the impacts of climate change on the mitigation potential, 33 permanence and additionality of estimated mitigation actions, and improved (real time & cheap) 34 measurement, reporting and verification. There is a need to include a greater suite of mitigation 35 measures in IAMs, informed by more realistic assessments that take into account local circumstances 36 and socio-economic factors and cross-sector synergies and trade-offs. Finally, there is a critical need 37 for more targeted research to develop appropriate country-level, locally specific, policy and land 38 management response options. These options could support more specific NDCs with AFOLU 39 measures that enable mitigation while also contributing to biodiversity conservation, ecosystem 40 functioning, livelihoods for millions of farmers and foresters, and many other Sustainable Development 41 Goals (SDGs) (high confidence). {7.7} 42 FOOTNOTE 2 Book keeping models and dynamic global vegetation models Do Not Cite, Quote or Distribute 7-7 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.1 Introduction 2 7.1.1 Key findings from previous reports 3 Agriculture, Forestry and Other Land Uses (AFOLU) is unique due to its capacity to mitigate climate 4 change through greenhouse gas (GHG) emission reductions, as well as enhance removals (IPCC 2019). 5 However, despite the attention on AFOLU since early 1990s it was reported in the SRCCL as 6 accounting for almost a quarter of anthropogenic emission (IPCC (2019a), with three main GHGs 7 associated with AFOLU; carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O). Overall 8 emission levels had remained similar since the publication of AR4 (Nabuurs et al. 2007). The diverse 9 nature of the sector, its linkage with wider societal, ecological and environmental aspects and the 10 required coordination of related policy, was suggested to make implementation of known and available 11 supply- and demand-side mitigation measures particularly challenging (IPCC 2019a). Despite such 12 implementation barriers, the considerable mitigation potential of AFOLU as a sector on its own and its 13 capacity to contribute to mitigation within other sectors was emphasised, with land-related measures, 14 including bioenergy, estimated as capable of contributing between 20 and 60% of the total cumulative 15 abatement to 2030 identified within transformation pathways (IPCC 2018). However, the vast 16 mitigation potential from AFOLU initially portrayed in literature and in Integrated Assessment Models 17 (IAMs), as explored in SR1.5, is being questioned in terms of feasibility (Roe et al. 2021) and a more 18 balanced perspective on the role of land in mitigation is developing, while at the same time, interest by 19 private investors in land-based mitigation is increasing fast. 20 The SRCCL (IPCC 2019a) outlined with medium evidence and medium agreement that supply-side 21 agriculture and forestry measures had an economic (at USD100 tCO2-eq-1) mitigation potential of 7.2- 22 10.6 GtCO2-eq-1 in 2030 (using GWP100 and multiple IPCC values for CH4 and N2O) of which about a 23 third was estimated as achievable at < USD20 tCO2-eq-1. Agricultural measures were reported as 24 sensitive to carbon price, with cropland and grazing land soil organic carbon management having the 25 greatest potential at USD20 tCO2-eq-1 and restoration of organic soils at USD100 tCO2-eq-1. Forestry 26 measures were less sensitive to carbon price, but varied regionally, with reduced deforestation, forest 27 management and afforestation having the greatest potential depending on region. Although demand- 28 side measures related to food could in theory make a large contribution to mitigation, in reality the 29 contribution has been very small. Overall, the dependency of mitigation within AFOLU on a complex 30 range of factors, from population growth, economic and technological developments, to the 31 sustainability of mitigation measures and impacts of climate change, was suggested to make realisation 32 highly challenging (IPCC 2019a). 33 Land can only be part of the solution alongside rapid emission reduction in other sectors (IPCC 2019a). 34 It was recognised that land supports many ecosystem services on which human existence, wellbeing 35 and livelihoods ultimately depend. Yet over-exploitation of land resources was reported as driving 36 considerable and unprecedented rate of biodiversity loss, and wider environmental degradation (IPCC 37 2019a;IPBES 2019a). Urgent action to reverse this trend was deemed crucial in helping to accommodate 38 the increasing demands on land and enhance climate change adaptation capacity. There was high 39 confidence that global warming was already causing an increase in the frequency and intensity of 40 extreme weather and climate events, impacting ecosystems, food security, disturbances and production 41 processes, with existing (and new) carbon stocks in soils and biomass at serious risk. The impact of 42 land cover on regional climate (through biophysical effects) was also highlighted, although there was 43 no confidence regarding impacts on global climate. 44 Since AR5, the share of AFOLU to anthropogenic GHG emissions had remained largely unchanged at 45 13-21% of total GHG emissions (medium confidence), though uncertainty in estimates of both sources Do Not Cite, Quote or Distribute 7-8 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 and sinks of CO2, exacerbated by difficulties in separating natural and anthropogenic fluxes, was 2 emphasised. Models indicated land (including the natural sink) to have very likely provided a net 3 removal of CO2 between 2007 and 2016. As in AR5, land cover change, notably deforestation, was 4 identified as a major driver of anthropogenic CO2 emissions whilst agriculture was a major driver of the 5 increasing anthropogenic CH4 and N2O emissions. 6 In terms of mitigation, without reductions in overall anthropogenic emissions, increased reliance on 7 large-scale land-based mitigation was predicted, which would add to the many already competing 8 demands on land. However, some mitigation measures were suggested to not compete with other land 9 uses, while also having multiple co-benefits, including adaptation capacity and potential synergies with 10 some Sustainable Development Goals (SDGs). As in AR5, there was large uncertainty surrounding 11 mitigation within AFOLU, in part because current carbon stocks and fluxes are unclear and subject to 12 temporal variability. Additionally, the non-additive nature of individual measures that are often inter- 13 linked and the highly context specific applicability of measures, causes further uncertainty. Many 14 AFOLU measures were considered well-established and some achievable at low to moderate cost, yet 15 contrasting economic drivers, insufficient policy, lack of incentivisation and institutional support to 16 stimulate implementation among the many stakeholders involved, in regionally diverse contexts, was 17 recognised as hampering realisation of potential. 18 None the less, the importance of mitigation within AFOLU was highlighted in all IPCC reports, with 19 modelled scenarios demonstrating the considerable potential role and land-based mitigation forming an 20 important component of pledged mitigation in Nationally Determined Contributions (NDCs) under the 21 Paris Agreement. The sector was identified as the only one in which large-scale Carbon Dioxide 22 Removal (CDR) may currently and at short term be possible (e.g. through afforestation/reforestation or 23 soil organic carbon management). This CDR component was deemed crucial to limit climate change 24 and its impacts, which would otherwise lead to enhanced release of carbon from land. However, the 25 SRCCL emphasised that mitigation cannot be pursued in isolation. The need for integrated response 26 options, that mitigate and adapt to climate change, but also deal with land degradation and 27 desertification, while enhancing food and fibre security, biodiversity and contributing to other SDGs 28 has been made clear (IPCC 2019a; Díaz et al. 2019; IPBES-IPCC 2021). 29 7.1.2 Boundaries, scope and changing context of the current report 30 This chapter assesses GHG fluxes between land and the atmosphere due to AFOLU, the associated 31 drivers behind these fluxes, mitigation response options and related policy, at time scales of 2030 and 32 2050. Land and its management has important links with other sectors and therefore associated chapters 33 within this report, notably concerning the provision of food, feed, fuel or fibre for human consumption 34 and societal wellbeing (Chapter 5), for bioenergy (Chapter 6), the built environment (Chapter 9), 35 transport (Chapter 10) and industry (Chapter 11). Mitigation within these sectors may in part, be 36 dependent on contributions from land and the AFOLU sector, with interactions between all sectors 37 discussed in Chapter 12. This chapter also has important links with IPCC WGII regarding climate 38 change impacts and adaptation. Linkages are illustrated in Figure 7.1. 39 Do Not Cite, Quote or Distribute 7-9 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Figure 7.1 Linkage between Chapter 7 and other chapters within this report as well as to WGII. 3 Mitigation potential estimates in this chapter consider potential emission reductions and removals only 4 within the AFOLU sector itself, and not the substitution effects from biomass and biobased products in 5 sectors such as Energy, Transport, Industry, Buildings, nor biophysical effects of e.g. cooling of cities. 6 These are covered in their respective chapters. 7 8 9 Figure 7.2 Summarised representation of interactions between land management, its products in terms of 10 food and fibre, and land - atmospheric GHG fluxes. For legibility reasons only a few of the processes and 11 management measures are depicted. Do Not Cite, Quote or Distribute 7-10 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 As highlighted in both AR5 and the SRCCL, there is a complex interplay between land management 2 and GHG fluxes as illustrated in Figure 7.2, with considerable variation in management regionally, as 3 a result of geophysical, climatic, ecological, economic, technological, institutional and socio-cultural 4 diversity. The capacity for land-based mitigation varies accordingly. The principal focus of this chapter 5 is therefore, on evaluating regional land-based mitigation potential, identifying applicable AFOLU 6 mitigation measures, estimating associated costs and exploring policy options that could enable 7 implementation. 8 Mitigation measures are broadly categorised as those relating to (1) forests and other ecosystems (2) 9 agriculture (3) biomass production for products and bioenergy and (4) demand-side levers. Assessment 10 is made in the context that land-mitigation is expected to contribute roughly 25% of the 2030 mitigation 11 pledged in Nationally Determined Contributions (NDCs) under the Paris Agreement (Grassi et al. 12 2017), yet very few countries have provided details on how this will be achieved. In light of AR5 and 13 the SRCCL findings, that indicate large land-based mitigation potential, considerable challenges to its 14 realisation, but also a clear nexus at which humankind finds itself, whereby current land management, 15 driven by population growth and consumption patterns, is undermining the very capacity of land, a 16 finite resource, to support wider critical functions and services on which humankind depends. 17 Mitigation within AFOLU is occasionally and wrongly perceived as an opportunity for in-action within 18 other sectors. AFOLU simply cannot compensate for mitigation shortfalls in other sectors. As the 19 outcomes of many critical challenges (UN Environment 2019), including biodiversity loss (Díaz et al. 20 2019) and soil degradation (FAO and ITPS 2015), are inextricably linked with how we manage land, 21 the evaluation and assessment of AFOLU is crucial. This chapter aims to address three core topics; 22 1. What is the latest estimated (economic) mitigation potential of AFOLU measures according to 23 both sectoral studies and integrated assessment models, and how much of this may be realistic 24 within each global region? 25 2. How do we realise the mitigation potential, while minimising trade-offs and risks and 26 maximising co-benefits that can enhance food and fibre security, conserve biodiversity and 27 address other land challenges? 28 3. How effective have policies been so far and what additional policies or incentives might enable 29 realisation of mitigation potential and at what costs? 30 This chapter first outlines the latest trends in AFOLU fluxes and the methodology supporting their 31 estimation (Section 7.2). Direct and indirect drivers behind emission trends are discussed in Section 32 7.3. Mitigation measures, their costs, co-benefits, trade-offs, estimated regional potential and 33 contribution within integrated global mitigation scenarios, is presented in Sections 7.4 and 7.5 34 respectively. Assessment of associated policy responses and links with SDGs are explored in Section 35 7.6. The chapter concludes with gaps in knowledge (Section 7.7) and frequently asked questions. 36 37 7.2 Historical and current trends in GHG emission and removals; their 38 uncertainties and implications for assessing collective climate progress 39 The biosphere on land and in wetlands is a source and sink of CO2 and CH4, and a source of N2O due 40 to both natural and anthropogenic processes that happen simultaneously and are therefore difficult to 41 disentangle (IPCC 2010; Angelo and Du Plesis 2017; IPCC 2019a). AFOLU is the only GHG sector to 42 currently include anthropogenic sinks. A range of methodological approaches and data have been 43 applied to estimating AFOLU emissions and removals, each developed for their own purposes, with 44 estimates varying accordingly. Since the SRCCL (Jia et al. 2019), emissions estimates have been Do Not Cite, Quote or Distribute 7-11 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 updated (Sections 7.2.2 and 7.2.3), while the assessment of biophysical processes and short-lived 2 climate forcers (Section 7.2.4) is largely unchanged. Further progress has been made on the implications 3 of differences in AFOLU emissions estimates for assessing collective climate progress (Section 7.2.2.2, 4 Cross-Chapter Box 6 in this Chapter). 5 7.2.1 Total net GHG flux from AFOLU 6 National Greenhouse Gas Inventory (NGHGI) reporting following the IPCC 1996 guidelines (IPCC 7 1996), separates the total anthropogenic AFOLU flux into: (i) net anthropogenic flux from Land Use, 8 Land-Use Change, and Forestry (LULUCF) due to both change in land cover and land management; 9 and (ii) the net flux from Agriculture. While fluxes of CO2 (Section 7.2.2) are predominantly from 10 LULUCF and fluxes of CH4 and N2O (Section 7.2.3) are predominantly from agriculture, fluxes of all 11 three gases are associated with both sub-sectors. However, not all methods separate them consistently 12 according to these sub-sectors, thus here we use the term AFOLU, separate by gas and implicitly include 13 CO2 emissions that stem from the agriculture part of AFOLU, though these account for a relatively 14 small portion. 15 16 Table 7.1 Net anthropogenic emissions (annual averages for 2010–2019a) from Agriculture, Forestry and 17 Other Land Use (AFOLU). For context, the net flux due to the natural response of land to climate and 18 environmental change is also shown for CO2 in column E. Positive values represent emissions, negative 19 values represent removals. Anthropogenic Natural Natural + Response Anthropogenic Gas Units AFOLU Net Non- Total net AFOLU as Natural land Net-land anthropogenic AFOLU anthropogenic a % of total sinks including atmosphere CO2 emissions h anthropog emissions net natural response flux (i.e. enic GHG (AFOLU + anthropoge of land to anthropogenic emissions non-AFOLU) nic anthropogenic AFOLU + d, f by gas emissions environmental natural fluxes by gas change and across entire land climate surface variability e A B C = A+B D = (A/C) E F=A+E *100 CO2 GtCO2-eq yr-1 5.9 ± 4.1 b, f 36.2 ± 2.9 42.0 ± 29.0 14% -12.5 ± 3.2 -6.6 ± 4.6 (bookkeeping models only). 0 to 0.8 (NGHGI/ FAOSTAT data) MtCH4 yr-1 157.0 ± 47.1 c 207.5 ± 364.4 ± 109.3 -i CH4 62.2 GtCO2-eq yr-1 4.2 ± 1.3 g 5.9 ± 1.8 10.2 ± 3.0 41% -1 MtN2O yr 6.6 ± 4.0 c 2.8 ± 1.7 9.4 ± 5.6 N2O GtCO2-eq yr-1 1.8 ± 1.1 g 0.8 ± 0.5 2.6 ± 1.5 69% Tota GtCO2-eq yr-1 11.9 ± 4.4 44 ± 3.4 55.9 ± 6.1 21% lj (CO2 component considers bookkeeping models only) 20 21 a Estimates are given until 2019 as this is the latest date when data are available for all gases, consistent with 22 Chapter 2, this report. Positive fluxes are emission from land to the atmosphere. Negative fluxes are removals. 23 b Net anthropogenic flux of CO2 are due to land-use change such as deforestation and afforestation and land 24 management, including wood harvest and regrowth, peatland drainage and fires, cropland and grassland 25 management. Average of three bookkeeping models (Hansis et al. 2015; Houghton and Nassikas 2017; Gasser 26 et al. 2020), complemented by data on peatland drainage and fires from FAOSTAT (Prosperi et al. 2020) and Do Not Cite, Quote or Distribute 7-12 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 GFED4s (Van Der Werf et al. 2017). This number is used for consistency with WGI and Chapter 2, this report. 2 Comparisons with other estimates are discussed in 7.2.2. Based on NGHGIs and FAOSTAT, the range is 0 to 3 0.8 Gt CO2 yr-1. 4 c CH4 and N2O emission estimates and assessed uncertainty of 30 and 60% respectively, are based on EDGAR 5 data (Crippa et al. 2021) in accordance with Chapter 2, this report (Sections 2.2.1.3 and 2.2.1.4). Both 6 FAOSTAT (FAO 2021a; Tubiello 2019; USEPA 2019) and the USA EPA (USEPA 2019) also provide data 7 on agricultural non-CO2 emissions, however mean global CH4 and N2O values considering the three databases 8 are within the uncertainty bounds of EDGAR. EDGAR only considers agricultural and not overall AFOLU 9 non-CO2 emissions. Agriculture is estimated to account for approximately 89 and 96% of total AFOLU CH4 10 and N2O emissions respectively. See Section 7.2.3 for further discussion. 11 d Total non-AFOLU emissions are the sum of total CO2-eq emissions values for energy, industrial sources, waste 12 and other emissions with data from the Global Carbon Project for CO 2, including international aviation and 13 shipping, and from the PRIMAP database for CH 4 and N2O averaged over 2007-2014, as that was the period 14 for which data were available. 15 e The modelled CO2 estimates include natural processes in vegetation and soils and how they respond to both 16 natural climate variability and to human-induced environmental changes i.e. the response of vegetation and 17 soils to environmental changes such as increasing atmospheric CO2 concentration, nitrogen deposition, and 18 climate change (indirect anthropogenic effects) on both managed and unmanaged lands. The estimate shown 19 represents the average from 17 Dynamic Global Vegetation Models with 1SD uncertainty (Friedlingstein et 20 al. 2020) 21 f The NGHGIs take a different approach to calculating “anthropogenic” CO2 fluxes than the models (Section 22 7.2.2). In particular the sinks due to environmental change (indirect anthropogenic fluxes) on managed lands 23 are generally treated as anthropogenic in NGHGIs and non-anthropogenic in models such as bookkeeping and 24 IAMs. A reconciliation of the results between IAMs and NGHGIs is presented in Cross-Chapter Box 6 in this 25 Chapter. If applied to this table, it would transfer approximately -5.5 GtCO2 y-1(a sink) from Column E (which 26 would become --7.2 GtCO2 yr-1) to Column A (which would then be 0.4 GtCO2 yr-1). 27 g All values expressed in units of CO2-eq are based on IPCC AR6 100-year Global Warming Potential (GWP100) 28 values with climate-carbon feedbacks (CH4 = 27, N2O = 273) (Chapter 2, Supplementary Material SM2.3 and 29 IPCC WGI AR6 Section 7.6). 30 h For assessment of cross-sector fluxes related to the food sector, see Chapter 12, this report. 31 i While it is acknowledged that soils are a natural CH4 sink (Jackson et al. 2020) with soil microbial removals 32 estimated to be 30 ± 19 MtCH4 yr-1 for the period 2008-2017 (according to bottom-up estimates), natural CH4 33 sources are considerably greater (371 (245-488) MtCH4 yr-1) resulting in natural processes being a net CH4 34 source (IPCC WGI AR6 Section 5.2.2). The soil CH4 sink is therefore omitted from Column E. 35 j. Total GHG emissions concerning non-AFOLU sectors and all sectors combined (Columns B and C) include 36 fluorinated gases in addition to CO2, CH4 and N2O. Therefore, total values do not equal the sum of estimates 37 for CO2, CH4 and N2O. 38 Do Not Cite, Quote or Distribute 7-13 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Figure 7.3 Subdivision of the total AFOLU emissions from Table 7.1 by activity and gas for the period 3 1990 to 2019. Positive values are emissions from land to atmosphere, negative values are 4 removals. Panel A shows emissions divided into major activity and gases. Note that ‘biomass burning’ 5 is only the burning of agriculture residues in the fields. The indicated growth rates between 1990-2000, 6 2000-2010, 2010-2019 are annualised across each time period. Panel B illustrates regional emissions in 7 the years 1990, 2000, 2010, 2019 AFOLU CO 2 (green shading) represents all AFOLU CO2 emissions. It 8 is the mean from three bookkeeping models (Hansis et al. 2015; Houghton and Nassikas 2017; Gasser 9 et al. 2020) as presented in the Global Carbon Budget (Friedlingstein et al. 2020) and is not directly 10 comparable to LULUCF in NGHGIs (Section 7.2.2). Data onCH4 and N2O emissions are from the 11 EDGAR database (Crippa et al. 2021). See Sections 7.2.2 and 7.2.3 for comparison of different 12 datasets. All values expressed are asCO2-eq with GWP100 values: CH4 = 27, N2O = 273. 13 Total global net anthropogenic GHG emissions from AFOLU were 11.9 ± 4.4 GtCO2-eq yr-1 on average 14 over the period 2010-2019, around 21% of total global net anthropogenic GHG emissions (Table 7.1, 15 Figure 7.3, using the sum of bookkeeping models for the CO2 component). When using 16 FAOSTAT/NGHGIs CO2 flux data, then the contribution of AFOLU to total emissions amounts to 13% 17 of global emissions. 18 This AFOLU flux is the net of anthropogenic emissions of CO2, CH4 and N2O, and anthropogenic 19 removals of CO2. The contribution of AFOLU to total emissions varies regionally with highest in Latin 20 America and Caribbean with 58% an lowest in Europe and North America with each 7% (Chapter 2, 21 Section 2.2.3). There is a discrepancy in the reported CO2 AFOLU emissions magnitude because 22 alternative methodlogical approaches that incorporate different assumptions are used (see 7.2.2.2). 23 While there is low agreement in the trend of global AFOLU CO2 emissions over the past few decades 24 (7.2.2), they have remained relatively constant (medium confidence) (Chapter 2). Average non-CO2 25 emission (aggregated using GWP100 IPCC AR6 values) from agriculture have risen from 5.2 ± 1.4 26 GtCO2-eq yr-1 for the period 1990 to 1999, to 6.0 ± 1.7 GtCO2-eq yr-1 for the period 2010 to 2019 (Crippa 27 et al. 2021), Section 7.2.3). Do Not Cite, Quote or Distribute 7-14 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 To present a fuller understanding of land-atmosphere interactions, Table 7.1 includes an estimate of the 2 natural sink of land to atmospheric CO2 (IPCC WGI Chapter 5 and (Jia et al. 2019). Land fluxes respond 3 naturally to human-induced environmental change (e.g. climate change, and the fertilising effects of 4 increased atmospheric CO2 concentration and nitrogen deposition), known as “indirect anthropogenic 5 effects”, and also to “natural effects” such as climate variability (IPCC 2010) (Table 7.1, Section 7.2.2). 6 This showed a removal of -12.5 ± 3.2 GtCO2 yr-1 (medium confidence) from the atmosphere during 7 2010-2019 according to global DGVM models (Friedlingstein et al. 2020) 31% of total anthropogenic 8 net emissions of CO2 from all sectors. It is likely that the NGHIs and FAOSTAT implicitly cover some 9 part of this sink and thus provide a net CO2 AFOLU balance with some 5 GtCO2 lower net emissions 10 than according to bookkeeping models, with the overall net CO2 value close to being neutral. Model 11 results and atmospheric observations concur that, when combining both anthropogenic (AFOLU) and 12 natural processes on the entire land surface (the total “land-atmosphere flux”), the land was a global net 13 sink for CO2 of -6.6 + 4.6 GtCO2 yr-1 with a range for 2010 to 2019 from -4.4 to -8.4 GtCO2 yr-1. (Van 14 Der Laan-Luijkx et al. 2017; Rödenbeck et al. 2003, 2018; Chevallier et al. 2005; Feng et al. 2016; 15 Niwa et al. 2017; Patra et al. 2018). The natural land sink is highly likely to be affected by both future 16 AFOLU activity and climate change (IPCC WGI Box 5.1 and IPCC WGI SPM Figure 7), whereby 17 under more severe climate change, the amount of carbon stored on land would still increase although 18 the relative share of the emissions that land takes up, declines. 19 20 7.2.2 Flux of CO2 from AFOLU, and the non-anthropogenic land sink 21 7.2.2.1 Global net AFOLU CO2 flux 22 23 24 Figure 7.4 Global net CO2 flux due to AFOLU estimated using different methods for the period 1960 to 25 2019 (GtCO2 yr-1). Positive numbers represent emissions. (Grey line) The mean from 17 DGVMs all using 26 the same driving data under TrendyV9 used within the Global Carbon Budget 2020 and including 27 different degrees of management (Bastos et al. 2020; Friedlingstein et al. 2020). (Orange line) Data 28 downloaded 6th June 2021 from FAOSTAT (FAO 2021b; http://www.fao.org/faostat/) comprising: net 29 emissions from (i) forest land converted to other land, (ii) net emissions from organic soils in cropland, 30 grassland and from biomass burning (including peat fires and peat draining (Prosperi et al. 2020) and (iii) 31 net emissions from forest land remaining forest land, which includes managed forest lands (Tubiello et al. 32 2020). (Dark blue line) Net flux estimate from National Greenhouse Gas Inventories (NGHGI) based on Do Not Cite, Quote or Distribute 7-15 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 country reports to the UNFCCC for LULUCF (Grassi et al. 2021) which include land-use change, and 2 flux in managed lands. (Red (EO) line) The 2001 – 2019 average net CO2 flux from non-intact forest- 3 related emissions and removals based on ground and Earth Observation data (EO) (Harris et al. 2021). 4 Data to mask non-intact forest were used in the tropics (Turubanova et al. 2018) and extra-tropics 5 (Potapov et al. 2017). 6 Light blue line: the mean estimate and minimum and maximum (blue shading) from three bookkeeping 7 models (Hansis et al. 2015; Houghton and Nassikas 2017; Gasser et al. 2020). These include land cover 8 change (e.g. deforestation, afforestation), forest management including wood harvest and land 9 degradation, shifting cultivation, regrowth of forests following wood harvest or abandonment of 10 agriculture, grassland management, agricultural management. Emissions from peat burning and 11 draining are added from external data sets (see text). Both the DGVM and Bookkeeping global data is 12 available at: https://www.icos-cp.eu/science-and-impact/global-carbon-budget/2020 (Accessed on 13 04/010/2021). Data consistent with IPCC WGI Chapter 5. Dotted lines denote the linear regression from 14 2000 to 2019. Trends are statistically significant (P < 0.05) with exception for the NGHGI trend (P< 0.01). 15 16 Comparison of estimates of the global net AFOLU flux of CO2 from diverse approaches (Figure 7.4) 17 show differences on the order of several GtCO2 yr-1. When considering the reasons for the differences, 18 and an approach to reconcile them (Section 7.2.2.3; Grassi et al. 2021), there is medium confidence in 19 the magnitude of the net AFOLU CO2 flux. There is a discrepancy in the reported CO2 AFOLU 20 emissions magnitude because alternative methodological approaches that incorporate different 21 assumptions are used (see 7.2.2.2). While the mean of the bookkeeping and DGVM model’s show a 22 small increase in global CO2 net emissions since year 2000, individual models suggest opposite trends 23 (Friedlingstein et al. 2020). The latest FAOSTAT and NGHGI estimates show a small reduction in net 24 emission. Overall, the trends are unclear. 25 Do Not Cite, Quote or Distribute 7-16 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Figure 7.5 Regional net flux of CO2 due to AFOLU estimated using different methods for the period 1990- 3 2019 (GtCO2 yr-1). Positive numbers represent emissions. The upper-central panel depicts the world map 4 shaded according to the IPCC AR6 regions corresponding to the individual graphs. For each regional 5 panel; (Orange line) Total net flux data from FAOSTAT (Tubiello et al. 2020), (Dark blue line) Net 6 emissions estimates from National Greenhouse Gas Inventories based on country reports to the UNFCCC 7 for LULUCF (Grassi et al. 2021), (Light blue line) The mean estimate and minimum and maximum 8 (blue shading) from three bookkeeping models. (Hansis et al. 2015; Houghton and Nassikas 2017; Gasser 9 et al. 2020). Regional estimates from bookkeeping models are available at: 10 https://zenodo.org/record/5548333#.YVwJB2LMJPY (Minx et al. 2021). See the legend in Figure 7.4 for 11 a detailed explanation of flux components for each dataset. 12 13 Regionally (Figure 7.5), there is high confidence of net emissions linked to deforestation in Latin 14 America, Africa and South-East Asia from 1990 to 2019. There is medium confidence in trends 15 indicating a decrease in net emissions in Latin America since 2005 linked to reduced gross deforestation 16 emissions, and a small increase in net emissions related to increased gross deforestation emissions in Do Not Cite, Quote or Distribute 7-17 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Africa since 2000 (Figure 7.5). There is high confidence regarding the net AFOLU CO2 sink in Europe 2 due to forest regrowth and known other sinks in managed forests, and medium confidence of a net sink 3 in North America and Eurasia since 2010. 4 5 7.2.2.2 Why do various methods deliver difference in results? 6 The processes responsible for fluxes from land have been divided into three categories (IPCC 2006, 7 2010): (1) the direct human-induced effects due to changing land cover and land management; (2) the 8 indirect human-induced effects due to anthropogenic environmental change, such as climate change, 9 CO2 fertilisation, nitrogen deposition, etc.; and (3) natural effects, including climate variability and a 10 background natural disturbance regime (e.g. wildfires, windthrows, diseases or insect outbreaks). 11 Global models estimate the anthropogenic land CO2 flux considering only the impact of direct effects, 12 and only those areas that were subject to intense and direct management such as clear-cut harvest. It is 13 important to note, that DGVMs also estimate the non-anthropogenic land CO2 flux (Land Sink) that 14 results from indirect and natural effects (Table 7.1). In contrast, estimates of the anthropogenic land 15 CO2 flux in NGHGIs (LULUCF) include the impact of direct effects and, in most cases, of indirect 16 effects on a much greater area considered “managed” than global models (Grassi et al. 2021). 17 The approach used by countries follows the IPCC methodological guidance for NGHGIs (IPCC 2006, 18 2019a). Since separating direct, indirect and natural effects on the land CO2 sink is impossible with 19 direct observation such as national forest inventories (IPCC 2010), upon which most NGHGIs are 20 based, the IPCC adopted the ‘managed land’ concept as a pragmatic proxy to facilitate NGHGI 21 reporting. Anthropogenic land GHG fluxes (direct and indirect effects) are defined as all those occurring 22 on managed land, that is, where human interventions and practices have been applied to perform 23 production, ecological or social functions (IPCC 2006, 2019a). GHG fluxes from unmanaged land are 24 not reported in NGHGIs because they are assumed to be non-anthropogenic. Countries report NGHGI 25 data with a range of methodologies, resolution and completeness, dependent on capacity and available 26 data, consistent with IPCC guidelines (IPCC 2006, 2019a) and subject to an international review or 27 assessment processes. 28 The FAOSTAT approach is conceptually similar to NGHGIs. FAOSTAT data on forests are based on 29 country reports to FAO-FRA 2020 (FAO 2020a), and include changes in biomass carbon stock in 30 “forest land” and “net forest conversions” in five-year intervals. “Forest land” may include unmanaged 31 natural forest, leading to possible overall overestimation of anthropogenic fluxes for both sources and 32 sinks, though emissions from deforestation are likely underestimated (Tubiello et al. 2020). FAOSTAT 33 also estimate emissions from forest fires and other land uses (organic soils), following IPCC methods 34 (Prosperi et al. 2020). The FAO-FRA 2020 (FAO 2020b) update leads to estimates of larger sinks in 35 Russia since 1991, and in China and the USA from 2011, and larger deforestation emissions in Brazil 36 and smaller in Indonesia than FRA 2015 (FAO 2015;Tubiello et al. 2020). 37 The bookkeeping models by Houghton and Nassikas (2017), Hansis et al. (2015), and Gasser et al. 38 (2020) and the DGVMs used in the Global Carbon Budget (Friedlingstein et al. 2020) use either the 39 LUH2 data set (Hurtt et al. 2020) HYDE (Goldewijk et al. 2017) FRA 2015 (FAO 2015) or a 40 combination. The LUH2 dataset includes a new wood harvest reconstruction, new representation of 41 shifting cultivation, crop rotations, and management information including irrigation and fertilizer 42 application. The area of forest subject to harvest in LUH2 is much less than the area of forest considered 43 “managed” in the NGHGIs (Grassi et al. 2018). The model datasets do not yet include the FAO FRA 44 2020 update (FAO 2020a). The DGVMs consider CO2 fertilization effects on forest growth that are Do Not Cite, Quote or Distribute 7-18 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 sometimes confirmed from the groundbased forest inventory networks (Nabuurs et al. 2013) and 2 sometimes not at all (van der Sleen et al. 2015). 3 Further, the DGVMs and bookkeeping models do not include a wide range of practices which are 4 implicitly covered by the inventories; for example: forest dynamics (Pugh et al. 2019; Le Noë et al. 5 2020) forest management including wood harvest (Nabuurs, et al. 2013; Arneth et al. 2017) agricultural 6 and grassland practices (Pugh et al. 2015; Sanderman et al. 2017; Pongratz et al. 2018); or e.g. fire 7 management (Andela et al. 2017; Arora and Melton 2018). 8 Increasingly higher emissions estimates are expected from DGVMs compared to bookkeeping models, 9 because DGVMs include a loss of additional sink capacity of 3.3 ± 1.1 GtCO2 yr-1 on average over 10 2009-2018, which is increasing with larger climate and CO2 impacts (Friedlingstein et al. 2020). This 11 arises because the DGVM methodological setup requires a reference simulation including climate and 12 environmental changes but without any land use change such as deforestation, so DGVMs implicitly 13 include the sink capacity forests would have developed in response to environmental changes on areas 14 that in reality have been cleared (Gitz and Ciais 2003; Pongratz et al. 2014)(IPCC WGI Chapter 5). 15 Carbon emissions from peat burning have been estimated based on the Global Fire Emission Database 16 (GFED4s; Van Der Werf et al. 2017). These were included in the bookkeeping model estimates and 17 added 2.0 Gt Carbon over 1960-2019 (e.g. causing the peak in South-East Asia in 1998, Figure 7.5). 18 Within the Global Carbon Budget (Friedlingstein et al. 2020), peat drainage from agriculture accounted 19 for an additional 8.6 Gt Carbon from 1960-2019 according to FAOSTAT (Conchedda and Tubiello, 20 2020) used by two of the bookkeeping models, (Hansis et al. 2015; Gasser et al. 2020). 21 Remote-sensing products provide valuable spatial and temporal land-use and biomass data globally 22 (including in remote areas), at potentially high spatial and temporal resolutions, that can be used to 23 calculate CO2 fluxes, but have mostly been applied only to forests at the global or even regional scale. 24 While such data can strongly support monitoring reporting and verification, estimates of forest carbon 25 fluxes directly from Earth Observation (EO) data vary considerably in both their magnitude and sign 26 (i.e. whether forests are a net source or sink of carbon). For the period 2005 – 2017, net tropical forest 27 carbon fluxes were estimated as -0.4 GtCO2 yr-1 (Fan et al. 2019); 0.58 GtCO2 yr-1 (Grace et al. 2014); 28 1.6 GtCO2yr-1 (Baccini et al. 2017) and 2.87 GtCO2 yr-1 (Achard et al. 2014). Differences can in part 29 be explained by spatial resolution of the data sets, the definition of “forest” and the inclusion 30 of processes and methods used to determine degradation and growth in intact and secondary forests, or 31 the changes in algorithm over time (Palahí et al. 2021). A recent global study integrated ground 32 observations and remote sensing data to map forest-related GHG emissions and removals at a high 33 spatial resolution (30m spatial scale), although it only provides an average estimate of annual carbon 34 loss over 2001–2019 (Harris et al. 2021). The estimated net global forest carbon sink globally was - 35 7.66 GtCO2 yr−1, being -1.7 GtCO2yr−1 in the tropics only. 36 Remote sensing products can help to attribute changes to anthropogenic activity or natural inter-annual 37 climate variability (Fan et al. 2019; Wigneron et al. 2020). Products with higher spatial resolution make 38 it easier to determine forest and carbon dynamics in relatively small-sized managed forests (e.g. Wang 39 et al. 2020; Heinrich et al. 2021; Reiche et al. 2021). For example secondary forest regrowth in the 40 Brazilian Amazon offset 9 to 14% of gross emissions due to deforestation 1 (Silva Junior et al. 2021; 41 Aragão et al. 2018). Yet disturbances such as fire and repeated deforestation cycles due to shifting 42 cultivation over the period 1985 to 2017, were found to reduce the regrowth rates of secondary forests 43 by 8 to 55% depending on the climate region of regrowth (Heinrich et al. 2021). Do Not Cite, Quote or Distribute 7-19 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.2.2.3 Implications of differences in AFOLU CO2 fluxes between global models and National 2 Greenhouse Gas Inventories (NGHGIs), and reconciliation 3 There is about 5.5 GtCO2 yr-1 difference in the anthropogenic AFOLU estimates between NGHGIs and 4 global models (this number relates to an IAMs comparison for the period 2005-2015 - see Cross-Chapter 5 Box 6 in this Chapter; for comparison with other models see Figure 7.4). Reconciling the differences 6 i.e. making estimates comparable, can build confidence in land-related CO2 estimates, for example for 7 the purpose of assessing collective progress in the context of the Global Stocktake (Cross-Chapter Box 8 6 in this Chapter). The difference largely results from greater estimated CO2 in NGHGIs, mostly 9 occurring in forests (Grassi et al. 2021). This difference is potentially a consequence of: (i) simplified 10 and/or incomplete representation of management in global models (Popp et al. 2017; Pongratz et al. 11 2018), e.g. concerning impacts of forest management in biomass expansion and thickening (Nabuurs et 12 al. 2013; Grassi et al. 2017) (ii) inaccurate and/or incomplete estimation of LULUCF fluxes in NGHGIs 13 (Grassi et al. 2017), especially in developing countries, primarily in non-forest land uses and in soils, 14 and (iii) conceptual differences in how global models and NGHGIs define ‘anthropogenic’ CO2 flux 15 from land (Grassi et al. 2018). The impacts of (i) and (ii) are difficult to quantify and result in 16 uncertainties that will decrease slowly over time through improvements of both models and NGHGIs. 17 By contrast, the inconsistencies in (iii) and its resulting biases were assessed as explained below. 18 Since changing the NGHGIs’ approach is impractical, an interim method to translate and adjust the 19 output of global models was outlined for reconciling a bookkeeping model and NGHGIs (Grassi et al. 20 2018). More recently, an improved version of this approach has been applied to the future mitigation 21 pathways estimated by IAMs (Grassi et al. 2021), with the implications for the Global Stocktake 22 discussed in Cross-Chapter Box 6 in this Chapter. This method implies a post-processing of current 23 global models’ results that addresses two components of the conceptual differences in the 24 “anthropogenic” CO2 flux; (i) how the impact of human-induced environmental changes (indirect 25 effects) are considered, and (ii) the extent of forest area considered ‘managed’. Essentially, this 26 approach adds DGVM estimates of CO2 fluxes due to indirect effects from countries’ managed forest 27 area (using non-intact forest area maps as a proxy) to the original global models’ anthropogenic land 28 CO2 fluxes (Figure 7.6). 29 Do Not Cite, Quote or Distribute 7-20 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Figure 7.6 Main conceptual differences between global models (bookkeeping models, IAMs and 3 DGVMs) and NGHGIs definitions of what is considered the ‘anthropogenic’ land CO 2 flux, and 4 proposed solution (from Grassi et al. 2021). (Panel a) Differences in defining the anthropogenic 5 land CO2 flux by global models (‘Land Use’) and NGHGIs (‘LULUCF’), including the attribution 6 of processes responsible for land fluxes (IPCC 2006; 2010) in managed and unmanaged lands. The 7 anthropogenic land CO2 flux by global models typically includes only the CO 2 flux due to ‘direct 8 effects’ (land-use change, harvest, regrowth). By contrast, most NGHGIs consider anthropogenic 9 all fluxes occurring in areas defined as ‘managed’, including also the sink due to ‘indirect effects’ 10 (climate change, atmospheric CO2 increase, N deposition etc.) and due to ‘natural effects’ (climate 11 variability, background natural disturbances). (Panel b) Proposed solution to the inconsistency, via 12 disaggregation of the ‘Land Sink’ flux from DGVMs into CO2 fluxes occurring in managed and in 13 unmanaged lands. The sum of ‘Land Use’ flux (direct effects from bookkeeping models or IAMs) 14 and the ‘Land Sink’ (indirect effects from DGVMs) in managed lands produces an adjusted global 15 model CO2 flux which is conceptually more comparable with LULUCF fluxes from NGHGIs. Note 16 that the figure may in some cases be an oversimplification, e.g. not all NGHGIs include all recent 17 indirect effects. 18 19 START CROSS-CHAPTER BOX 6 HERE 20 Cross-Chapter Box 6 Implications of reconciled anthropogenic land CO2 fluxes for assessing 21 collective climate progress in the global stocktake 22 Authors: Giacomo Grassi (Italy), Joeri Rogelj (Belgium/Austria), Joanna House (United Kingdom), 23 Alexander Popp (Germany), Detlef van Vuuren (The Netherlands), Katherine Calvin (The United States 24 of America), Shinichiro Fujimori (Japan), Petr Havlik (The Czech Republic), Gert-Jan Nabuurs (The 25 Netherlands) 26 The Global Stocktake aims to assess countries’ collective progress towards the long-term goals of the 27 Paris Agreement in the light of the best available science. Historic progress is assessed based on 28 NGHGIs, while expectations of future progress are based on country climate targets (e.g., NDCs for 29 2025 or 2030 and long-term strategies for 2050). Scenarios consistent with limiting warming well- Do Not Cite, Quote or Distribute 7-21 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 below 2°C and 1.5°C developed by IAMs (Chapter 3) are expected to play a key role as benchmarks 2 against which countries’ aggregated future mitigation pledges will be assessed. This, however, implies 3 that estimates by IAMs and country data used to measure progress are comparable. 4 In fact, there is ~5.5 GtCO2 yr-1 difference during 2005-2015 between global anthropogenic land CO2 5 net flux estimates of IAMs and aggregated NGHGIs, due to different conceptual approaches to what is 6 “anthopogenic”. This approach and its implications when comparing climate targets with global 7 mitigation pathways are illustrated in this Box Figure 1a-e. 8 By adjusting the original IAM output (Cross-Chapter Box 6, Figure 1a) with the indirect effects from 9 countries’ managed forest (Cross-Chapter Box 6, Figure 1b, estimated by DGVMs, see also Figure 7.6), 10 NGHGI-comparable pathways can be derived (Cross-Chapter Box 6, Figure 1c). The resulting apparent 11 increase in anthropogenic sink reflects simply a reallocation of a CO2 flux previously labelled as natural, 12 and thus does not reflect a mitigation action. These changes do not affect non-LULUCF emissions. 13 However, since the atmosphere concentration is a combination of CO2 emissions from LULUCF and 14 from fossil fuels, the proposed land-related adjustments also influence the NGHGI-comparable 15 economy-wide (all sector) CO2 pathways (Cross-Chapter Box 6 Figure 1d). 16 This approach does not imply a change in the original decarbonisation pathways, nor does it suggest 17 that indirect effects should be considered in the mitigation efforts. It simply ensures that a like-with- 18 like comparison is made: if countries’ climate targets use the NGHGI definition of anthropogenic 19 emissions, this same definition can be applied to derive NGHGI-comparable future CO2 pathways. This 20 would have an impact on the NGHGI-comparable remaining carbon or GHG budget (i.e. the allowable 21 emissions until net zero CO2 or GHG emissions consistent with a certain climate target). For example, 22 for SSP2-1.9 and SSP2-2.6 (representing pathways in line with 1.5°C and well-below 2°C limits under 23 SSP2 assumptions), carbon budget is lower by -170 carbon GtCO2-eq than the original remaining 24 carbon budget according to the models’ approach (Cross-Chapter Box 6, Figure 1e). Similarly, the 25 remaining carbon (or GHG) budgets in Chapter 3 (this report), as well as the net zero carbon (or GHG) 26 targets, could only be used in combination with the definition of anthropogenic emissions as used by 27 the IAMs (Cross-Chapter Box 3 in Chapter 3). In the absence of these adjustments, collective progress 28 would appear better than it is. 29 The UNEP’s annual assessment of the global 2030 ‘emission gap’ between aggregated country NDCs 30 and specific target mitigation pathways (UNEP 2020), is only affected to a limited degree. This is 31 because some estimates of global emissions under the NDCs already use the same land-use definitions 32 as the IAM mitigation pathways (Rogelj et al. 2017), and because historical data of global NDC 33 estimates is typically harmonised to the historical data of global mitigation pathway projections (Rogelj 34 et al. 2011). This latter procedure, however, is agnostic to the reasons for the observed mismatch, and 35 often uses a constant offset. The adjustment described here allows this mismatch to be resolved by 36 drawing on a scientific understanding of the underlying reasons, and thus provides a more informed and 37 accurate basis for estimating the emission gap. 38 The approach to deriving a NGHGI-comparable emission pathways presented here can be further 39 refined with improved estimates of the future forest sink. Its use would enable a more accurate 40 assessment of the collective progress achieved and of mitigation pledges under the Paris Agreement. 41 Do Not Cite, Quote or Distribute 7-22 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Cross-Chapter Box 6, Figure 1. Impact on global mitigation pathways of adjusting the modelled 3 anthropogenic land CO2 fluxes to be comparable with National Greenhouse Gas Inventories (NGHGIs) 4 (from Grassi et al. 2021). Panel a: The mismatch between global historical LULUCF CO 2 net flux from 5 NGHGIs (black), and the original (un-adjusted) modelled flux historically and under future mitigation 6 pathways for SSP2 scenarios from Integrated Assessment Models (IAMs, Chapter 3). Panel b: fluxes due 7 to indirect effects of environmental change on areas equivalent to countries’ managed forest (i.e. those 8 fluxes generally considered ‘anthropogenic’ by countries and ‘natural’ by global models). Panel c: 9 original modelled (solid line) LULUCF mitigation pathways adjusted to be NGHGI-comparable (dashed 10 line) i.e. by adding the indirect effects in panel b. The indirect effects in panel b decline over time with 11 increasing mitigation ambition, mainly because of the weaker CO 2 fertilisation effect. In Panel c, the 12 dependency of the adjusted LULUCF pathways on the target becomes less evident after 2030, because the 13 indirect effects in countries’ managed forest (which are progressively more uncertain with time, as 14 highlighted by the grey areas) compensate the effects of the original pathways. Panel d: NGHGI- 15 comparable pathways for global CO2 emissions from all sectors including LULUCF (obtained by 16 combining global CO2 pathways without LULUCF - where no adjustment is needed - and the NGHGI- 17 comparable CO2 pathways for LULUCF (Gütschow et al. 2019; Grassi et al. 2017). Panel e: Cumulative 18 impact of the adjustments from 2021 until net zero CO 2 emissions or 2100 (whatever comes first) on the 19 remaining carbon budget. 20 21 END CROSS-CHAPTER BOX 6 HERE 22 23 7.2.3 CH4 and N2O flux from AFOLU 24 Trends in atmospheric CH4 and N2O concentrations and the associated sources, including land and land 25 use are discussed in Sections 5.2.2 and 5.2.3 of the IPCC WGI sixth assessment report. Regarding 26 AFOLU, the SRCCL and AR5 (Jia et al. 2019; Smith et al. 2014) identified three global non-CO2 27 emissions data sources; EDGAR (Crippa et al. 2021), FAOSTAT (FAO 2021a; Tubiello, 2019) and the 28 USA EPA (USEPA 2019). Methodological differences have been previously discussed (Jia et al. 2019). Do Not Cite, Quote or Distribute 7-23 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 In accordance with Chapter 2, this report, EDGAR data are used in Table 7.1 and Figure 7.3. It is 2 important to note that in terms of AFOLU sectoral CH4 and N2O emissions, only FAOSTAT provides 3 data on AFOLU emissions, while EDGAR and USEPA data consider just the agricultural component. 4 However, the mean of values across the three databases for both CH4 and N2O, fall within the assessed 5 uncertainty bounds (30 and 60% for CH4 and N2O respectively, Section 2.2.1, this report) of EDGAR 6 data. NGHGIs annually submitted to the UNFCCC (Section 7.2.2.3) provide national AFOLU CH4 and 7 N2O data, as included in the SRCCL (Jia et al. 2019). Aggregation of NGHGIs to indicate global 8 emissions must be considered with caution, as not all countries compile inventories, nor submit 9 annually. Additionally, NGHGIs may incorporate a range of methodologies for CH4 and N2O 10 accounting (e.g. Thakuri et al. 2020; Ndung’U et al. 2019; Van der Weerden et al. 2016), making 11 comparison difficult. The analysis of complete AFOLU emissions presented here, is based on 12 FAOSTAT data. For agricultural specific discussion, analysis considers EDGAR, FAOSTAT and 13 USEPA data. 14 7.2.3.1 Global AFOLU CH4 and N2O emissions 15 Using FAOSTAT data, the SRCCL estimated average CH4 emissions from AFOLU to be 161.2 ± 43 16 Mt CH4 yr-1 for the period 2007-2016, representing 44% of total anthropogenic CH4 emissions, with 17 agriculture accounting for 88% of the AFOLU component (Jia et al. 2019). The latest data (FAO 2021a, 18 2020b) highlight a trend of growing AFOLU CH4 emissions, with a 10% increase evident between 1990 19 and 2019, despite year-to-year variation. Forestry and other land use (FOLU) CH4 emission sources 20 include biomass burning on forest land and combustion of organic soils (peatland fires) (FAO 2020c). 21 The agricultural share of AFOLU CH4 emissions remains relatively unchanged, with the latest data 22 indicating agriculture to have accounted for 89% of emissions on average between 1990 and 2019. The 23 SRCCL reported with medium evidence and high agreement that ruminants and rice production were 24 the most important contributors to overall growth trends in atmospheric CH4 (Jia et al. 2019). The latest 25 data confirm this in terms of agricultural emissions, with agreement between databases that agricultural 26 CH4 emissions continue to increase and that enteric fermentation and rice cultivation remain the main 27 sources (Figure 7.7). The proportionally higher emissions from rice cultivation indicated by EDGAR 28 data compared to the other databases, may result from the use of a Tier 2 methodology for this source 29 within EDGAR (Janssens-Maenhout et al. 2019). 30 The SRCCL also noted a trend of increasing atmospheric N2O concentration, with robust evidence and 31 high agreement that agriculture accounted for approximately two-thirds of overall global anthropogenic 32 N2O emissions. Average AFOLU N2O emissions were reported to be 8.7 ± 2.5 Mt N2O yr-1 for the 33 period 2007-2016, accounting for 81% of total anthropogenic N2O emissions, with agriculture 34 accounting for 95% of AFOLU N2O emissions (Jia et al. 2019). A recent comprehensive review 35 confirms agriculture as the principal driver of the growing atmospheric N2O concentration (Tian et al. 36 2020). The latest FAOSTAT data (FAO 2020b, 2021a) document a 25% increase in AFOLU N2O 37 emissions between 1990 and 2019, with the average share from agriculture remaining approximately 38 the same (96%). Agricultural soils were identified in the SRCCL and in recent literature as a dominant 39 emission source, notably due to nitrogen fertiliser and manure applications to croplands, and manure 40 production and deposition on pastures (Jia et al. 2019; Tian et al. 2020). There is agreement within latest 41 data that agricultural soils remain the dominant source (Figure 7.7). 42 Aggregation of CH4 and N2O to CO2 equivalence (using GWP100 IPCC AR6 values), suggests that 43 AFOLU emissions increased by 15% between 1990 and 2019, though emissions showed trend 44 variability year to year. Agriculture accounted for 91% of AFOLU emissions on average over the period 45 (FAO 2020b, 2021a). EDGAR (Crippa et al. 2021), FAOSTAT (FAO 2021a) and USEPA (USEPA 46 2019) data suggest aggregated agricultural emissions (CO2-eq) to have increased since 1990, by 19 Do Not Cite, Quote or Distribute 7-24 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 (1990-2019), 15 (1990-2019) and 21 (1990-2015) % respectively, with all databases identifying enteric 2 fermentation and agricultural soils as the dominant agricultural emissions sources. 3 4 5 Figure 7.7 Estimated global mean agricultural CH4 (Top), N2O (Middle) and aggregated CH4 and 6 N2O (using CO2-eq according to GWP100 AR6 values) (Bottom) emissions for three decades 7 according to EDGARv6.0 (Crippa et al. 2021), FAOSTAT (FAO 2021a) and USEPA (USEPA 2019) 8 databases. Latest versions of databases indicate historic emissions to 2019, 2019 and 2015 9 respectively, with average values for the post-2010 period calculated accordingly. For CH4, 10 emissions classified as ‘Other Ag.’ within USEPA data, are re-classified as ‘Agricultural Biomass 11 Burning’. Despite CH4 emissions from agricultural soils also being included, this category was 12 deemed to principally concern biomass burning on agricultural land and classified accordingly. For 13 N2O, emissions classified within EDGAR as direct and indirect emissions from managed soils, and 14 indirect emissions from manure management are combined under ‘Agricultural Soils’. Emissions 15 classified by FOASTAT as from manure deposition and application to soils, crop residues, drainage 16 of organic soils and synthetic fertilisers are combined under ‘Agricultural Soils’, while emissions 17 reported as ‘Other Ag.’ under USEPA data are re-classified as ‘Agricultural Biomass Burning’. 18 19 7.2.3.2 Regional AFOLU CH4 and N2O emissions 20 FAOSTAT data (FAO 2020b, 2021a) indicate Africa (+ 44%), followed by Southern Asia (+ 29%) to 21 have the largest growth in AFOLU CH4 emissions between 1990 and 2019 (Figure 7.8). Eurasia was 22 characterised by notable emission reductions (--58%), principally as a result of a sharp decline (--63%) Do Not Cite, Quote or Distribute 7-25 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 between 1990 and 1999. The average agricultural share of AFOLU emissions between 1990 and 2019 2 ranged from 66% in Africa to almost 100% in the Middle East. 3 In agreement with AR5 (Smith et al. 2014), the SRCCL identified Asia as having the largest share 4 (37%) of emissions from enteric fermentation and manure management since 2000, but Africa to have 5 the fastest growth rate. Asia was identified as responsible for 89% of rice cultivation emissions, which 6 were reported as increasing (Jia et al. 2019). Considering classification by ten IPCC regions, data 7 suggest enteric fermentation to have dominated emissions in all regions since 1990, except in South- 8 east Asia and Developing Pacific, where rice cultivation forms the principal source (FAO 2021; USEPA 9 2019). The different databases broadly indicate the same regional CH4 emission trends, though the 10 indicated absolute change differs due to methodological differences (Section 7.2.3.1). All databases 11 indicate considerable emissions growth in Africa since 1990 and that this region recorded the greatest 12 regional increases in emissions from both enteric fermentation and rice cultivation since 2010. 13 Additionally, FAOSTAT data suggest that emissions from agricultural biomass burning account for a 14 notably high proportion of agricultural CH4 emissions in Africa (Figure 7.8). 15 The latest data suggest growth in AFOLU N2O emissions in most regions between 1990 and 2019, with 16 Southern Asia demonstrating highest growth (+ 74%) and Eurasia, greatest reductions (- 51%), the latter 17 mainly a result of a 61% reduction between 1990 and 2000 (FAO 2020b, 2021a). Agriculture was the 18 dominant emission source in all regions, its proportional average share between 1990 and 2019 ranging 19 from 87% in Africa, to almost 100% in the Middle East (Figure 7.8). 20 The SRCCL provided limited discussion on regional variation in agricultural N2O emissions but 21 reported with medium confidence that certain regions (North America, Europe, East & South Asia) were 22 notable sources of grazing land N2O emissions (Jia et al. 2019). AR5 identified Asia as the largest 23 source and as having the highest growth rate of N2O emissions from synthetic fertilisers between 2000 24 and 2010 (Smith et al. 2014). Latest data indicate agricultural N2O emission increases in most regions, 25 though variation between databases prevents definitive conclusions on trends, with Africa, Southern 26 Asia, and Eastern Asia suggested to have had greatest growth since 1990 according to EDGAR (Crippa 27 et al. 2021), FAOSTAT (FAO 2021a) and USEPA (USEPA 2019) data respectively. However, all 28 databases indicate that emissions declined in Eurasia and Europe from 1990 levels, in accordance with 29 specific environmental regulations put in place since the late 1980s (Tubiello 2019; European 30 Environment Agency 2020; Tian et al. 2020), but generally suggest increases in both regions since 31 2010. 32 Do Not Cite, Quote or Distribute 7-26 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Figure 7.8 Estimated average AFOLU CH4 (Top) and N2O (Bottom) emissions for three decades 3 according to FAOSTAT data by ten global regions, with disaggregation of agricultural emissions 4 (FAO 2020b; 2021a). Note for N2O, emissions from manure deposition and application to soils, 5 crop residues and synthetic fertilisers are combined under ‘Agricultural Soils’. 6 7 7.2.4 Biophysical effects and short-lived climate forcers 8 Despite new literature, general conclusions from the SRCCL and WGI-AR6 on biophysical effects and 9 short-lived climate forcers remain the same. Changes in land conditions from land cover change or land 10 management jointly affect water, energy, and aerosol fluxes (biophysical fluxes) as well as GHG fluxes 11 (biogeochemical fluxes) exchanged between the land and atmosphere (high agreement, robust 12 evidence) (Erb et al. 2017; Alkama and Cescatti 2016; Naudts et al. 2016; O’Halloran et al. 2012; 13 Anderson et al. 2011). There is high confidence that changes in land condition do not just have local 14 impacts but also have non-local impacts in adjacent and more distant areas (Mahmood et al. 2014; 15 Pielke et al. 2011) which may contribute to surpassing climate tipping points (Brando et al. 2014; 16 Nepstad et al. 2008). Non-local impacts may occur through: GHG fluxes and subsequent changes in 17 radiative transfer, changes in atmospheric chemistry, thermal, moisture and surface pressure gradients 18 creating horizontal transport (advection) (De Vrese et al. 2016; Davin and de Noblet-Ducoudre 2010) 19 and vertical transport (convection and subsidence) (Devaraju et al. 2018). Although regional and global 20 biophysical impacts emerge from model simulations ( Devaraju et al. 2018; De Vrese et al. 2016; Davin 21 and de Noblet-Ducoudre 2010), especially if the land condition has changed over large areas, there is 22 very low agreement on the location, extent and characteristics of the non-local effects across models. Do Not Cite, Quote or Distribute 7-27 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Recent methodological advances, empirically confirmed changes in temperature and precipitation 2 owing to distant changes in forest cover (Meier et al. 2021; Cohn et al. 2019). 3 Following changes in land conditions, CO2, CH4 and N2O fluxes are quickly mixed into the atmosphere 4 and dispersed, resulting in the biogeochemical effects being dominated by the biophysical effects at 5 local scales (high confidence) (Alkama and Cescatti 2016; Li et al. 2015). Afforestation/reforestation 6 (Strandberg and Kjellström 2019; Lejeune et al. 2018), urbanisation (Li and Bou-Zeid 2013) and 7 irrigation (Thiery et al. 2017; Mueller et al. 2016) modulate the likelihood, intensity, and duration of 8 many extreme events including heatwaves (high confidence) and heavy precipitation events (medium 9 confidence) (Haberlie et al. 2015). There is high confidence and high agreement that afforestation in 10 the tropics (Perugini et al. 2017), irrigation ( Mueller et al. 2016; Alter et al. 2015) and urban greening 11 result in local cooling, high agreement and medium confidence on the impact of tree growth form 12 (deciduous vs. evergreen) ( Schwaab et al. 2020; Luyssaert et al. 2018; Naudts et al. 2016), and low 13 agreement on the impact of wood harvest, fertilisation, tillage, crop harvest, residue management, 14 grazing, mowing, and fire management on the local climate. 15 Studies of biophysical effects have increased since AR5 reaching high agreement for the effects of 16 changes in land condition on surface albedo (Leonardi et al. 2015). Low confidence remains in 17 proposing specific changes in land conditions to achieve desired impacts on local, regional and global 18 climates due to: a poor relationship between changes in surface albedo and changes in surface 19 temperature (Davin and de Noblet-Ducoudre 2010), compensation and feedbacks among biophysical 20 processes (Kalliokoski et al. 2020; Bonan 2016), climate and seasonal dependency of the biophysical 21 effects (Bonan 2016), omittance of short-lived chemical forcers (Kalliokoski et al. 2020; Unger 2014), 22 and study domains often being too small to document possible conflicts between local and non-local 23 effects (Hirsch et al. 2018; Swann et al. 2012). 24 25 7.3 Drivers 26 Since AR5 several global assessments (IPBES 2018; NYDF Assessment Report. 2019; UN 27 Environment 2019; IPCC 2019) and studies (e.g. Tubiello 2019; Tian et al. 2020) have reported on 28 drivers (natural and anthropogenic factors that affect emissions and sinks of the land use sector) behind 29 AFOLU emissions trends, and associated projections for the coming decades. The following analysis 30 aligns with the drivers typology used by (IPBES (2019) and the Global Environmental Outlook (UN 31 Environment 2019). Drivers are divided into direct drivers resulting from human decisions and actions 32 concerning land use and land-use change, and indirect drivers that operate by altering the level or rate 33 of change of one or more direct drivers. Although drivers of emissions in Agriculture and FOLU are 34 presented separately, they are interlinked, operating in many complex ways at different temporal and 35 spatial scales, with outcomes depending on their interactions. For example, deforestation in tropical 36 forests is a significant component of sectorial emissions. A review of deforestation drivers’ studies 37 published between 1996 and 2013, indicated a wide range of factors associated with deforestation rates 38 across many analyses and studies, covering different regions (Figure 7.9; Busch and Ferretti-Gallon 39 2017). Higher agricultural prices were identified as a key driver of deforestation, while law 40 enforcement, area protection, and ecosystem services payments were found to be important drivers of 41 reduced deforestation, while timber activity did not show a consistent impact 42 Do Not Cite, Quote or Distribute 7-28 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Figure 7.9 Association of driver variables with more or less deforestation 3 Source: Busch and Ferretti-Gallon (2017) 4 7.3.1 Anthropogenic direct drivers – Deforestation, conversion of other ecosystems, and 5 land degradation 6 The global forest area in 2020 is estimated at 4.1 billion ha, representing 31% of the total land area 7 (FAO 2020a). Most forests are situated in the tropics (45%), followed by boreal (27%), temperate (16%) 8 and subtropical (11%) domains. Considering regional distribution of global forest area, Europe and the 9 Russian Federation accounts for 25%, followed by South America (21%), North and Central America 10 (19 %), Africa (16%), Asia (15%) and Oceania (5%). However, a significant share (54%) of the world’s 11 forest area concerns five countries – the Russian Federation, Brazil, Canada, the USA and China (FAO 12 2020a). Forest loss rates differ among regions though the global trend is towards a net forest loss (UN 13 Environment 2019). The global forest area declined by about 178 Mha in the 30 years from 1990 to 14 2020 (FAO 2020a). The rate of net forest loss has decreased since 1990, a result of reduced deforestation 15 in some countries and forest gains in others. The annual net loss of forest area declined from 7.8 Mha 16 in 1990–2000, to 5.2 Mha in 2000–2010, to 4.7 Mha in 2010–2020, while the total growing stock in 17 global forests increased (FAO 2020a). The rate of decline in net forest loss during the last decade was 18 due mainly to an increase in the rate of forest gain (i.e. afforestation and the natural expansion of 19 forests). 20 Globally, the area of the more open, other wooded land is also of significant importance, with almost 1 21 billion hectares (FAO 2020a). The area of other wooded land decreased by 30.6 Mha between 1990 22 and 2020 with larger declines between 1990–2000 (FAO 2020a). There are still significant challenges 23 in monitoring the area of other wooded land, largely associated with difficulties in measuring tree- 24 canopy cover in the range of 5–10%.The global area of mangroves, one of the most productive terrestrial 25 ecosystems (Neogi 2020a), has also experienced a significant decline (Thomas et al. 2017; Neogi 26 2020b), with a decrease of 1.0 Mha between 1990 and 2020 (FAO 2020a) due to agriculture and Do Not Cite, Quote or Distribute 7-29 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 aquaculture (Bhattarai 2011; Ajonina et al. 2014; Webb et al. 2014; Giri et al. 2015; Fauzi et al. 2019; 2 Thomas et al. 2017). Some relevant direct drivers affecting emissions and removal in forests and other 3 ecosystems are discussed in proceeding sections. 4 7.3.1.1. Conversion of natural ecosystems to agriculture 5 Previous IPCC reports identify land use change as an important driver of emissions and agriculture as 6 a key driver of land use change, causing both deforestation and wetland drainage (Smith et al. 2019d). 7 AR5 reported a trend of declining global agricultural land area since 2000 (Smith et al. 2014). The latest 8 data (FAO 2021b) indicate a 2% reduction in the global agricultural area between 2000 and 2019 9 (Figure 7.10). This area includes (though is not limited to) land under permanent and temporary crops 10 or pasture, temporary fallow and natural meadows and pasture utilized for grazing or agricultural 11 purposes (FAO 2021b), although the extent of land used for grazing may not be fully captured (Fetzel 12 et al. 2017). Data indicate changes in how agricultural land is used. Between 2000 and 2019, the area 13 classified as permanent meadow and pasture decreased (- 6%) while cropland area (under arable 14 production and temporary crops) increased (+ 2%). A key driver of this change has been a general trend 15 of intensification, including in livestock production (Barger et al. 2018; OECD/FAO 2019; UN 16 Environment 2019), whereby less grazing land is supporting increasing livestock numbers in 17 conjunction with greater use of crops as livestock feed (Barger et al. 2018). The share of feed crops, 18 such as maize and soybean, of global crop production is projected to grow as the demand for animal 19 feed increases with further intensification of livestock production (OECD/FAO 2019). Despite 20 increased demand for food, feed, fuel and fibre from a growing human population (FAO 2019b), global 21 agricultural land area is projected to remain relatively stable during the next decade, with increases in 22 production expected to result from agricultural intensification (OECD/FAO 2019). 23 Despite a decline in global agricultural area, the latest data document some regional expansion between 24 2000 and 2019, specifically in Africa (+ 3%) and Asia and the Developing Pacific (+ 1%). Agricultural 25 area declined in all other regions, notably in developed countries (- 9%), due to multiple factors 26 including among others, urbanisation (see Section 7.3.1.2). 27 7.3.1.2. Infrastructure development and urbanisation 28 Although built-up areas (defined as cities, towns, villages and human infrastructure) occupy a relatively 29 small fraction of land (around 1% of global land), since 1975 urban clusters (i.e. urban centres as well 30 as surrounding suburbs) have expanded approximately 2.5 times (UN Environment 2019; Chapter 8, 31 this report). Regional differences are striking. Between 1975 and 2015, built-up areas doubled in size 32 in Europe while urban population remained relatively constant. In Africa built-up areas grew 33 approximately fourfold, while urban population tripled (UN Environment 2019). Trends indicate that 34 rural-to-urban migration will continue and accelerate in developing countries increasing environmental 35 pressure in spite of measures to mitigate some of the impacts (e.g. by preserving or enhancing natural 36 systems within cities for example lakes or natural and urban green infrastructures (UN Environment 37 2019). If current population densities within cities remain stable, the extent of built-up areas in 38 developed countries is expected to increase by 30% and triple in developing countries between 2000 39 and 2050 (Barger et al. 2018). 40 Urban expansion leads to landscape fragmentation and urban sprawl with effects on forest resources 41 and land use (Ünal et al. 2019) while interacting with other drives. For example, in the Brazilian 42 Amazon, the most rapid urban growth occurs within cities that are located near rural areas that produce 43 commodities (minerals or crops) and are connected to export corridors (Richards and VanWey 2015). 44 Urbanisation, coastal development and industrialisation also play crucial roles in the significant loss of 45 mangrove forests (Richards and Friess 2016; Hirales-Cota 2010; Rivera-Monroy et al. 2017). Among 46 infrastructural developments, roads are one of the most consistent and most considerable factors in Do Not Cite, Quote or Distribute 7-30 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 deforestation, particularly in tropical frontiers (Pfaff et al. 2007; Rudel et al. 2009; Ferretti-Gallon and 2 Busch 2014). The development of roads may also bring subsequent impacts on further development 3 intensity due to increasing economic activities (see Chapter 8) mostly in the tropics and subtropics, 4 where the expansion of road networks increases access to remote forests that act as refuges for 5 biodiversity (Campbell et al. 2017) (Box 7.1). Logging is one of the main drivers of road construction 6 in tropical forests (Kleinschroth and Healey 2017) which leads to more severe long term impacts that 7 include increased fire incidence, soil erosion, landslides, and sediment accumulation in streams, 8 biological invasions, wildlife poaching, illicit land colonisation, illegal logging and mining, land 9 grabbing and land speculation (Laurance et al. 2009; Alamgir et al. 2017). 10 11 [START BOX 7.1 HERE] 12 Box 7.1 Case study: Reducing the impacts of roads on deforestation 13 Summary 14 Rapidly expanding roads, particularly in tropical regions, are linked to forest loss, degradation, and 15 fragmentation because the land becomes more generally accessible. Increase of land values of areas 16 adjacent to roads also drives speculation and deforestation related to land tenure (Fearnside 2015). If 17 poorly planned, infrastructure can facilitate fires, illegal mining, and wildlife poaching with 18 consequences for GHG emissions and biodiversity conservation. However, some initiatives are 19 providing new approaches for better planning and then limit environmental and societal impacts. 20 Background 21 Although the number and extent of protected areas has increased markedly in recent decades (Watson 22 et al. 2014), many other indicators reveal that nature is in broad retreat. For example, the total area of 23 intact wilderness is declining rapidly worldwide (Watson et al. 2016), 70% of the world’s forests are 24 now less than 1 km from a forest edge (Haddad et al. 2015), the extent of tropical forest fragmentation 25 is accelerating exponentially (Taubert et al. 2018). One of the most direct and immediate driver of 26 deforestation and biodiversity decline is the dramatic expansion of roads and other transportation 27 infrastructure (Laurance et al. 2014a; Laurance and Arrea 2017; Alamgir et al. 2017). 28 Case description 29 From 2010 to 2050, the total length of paved roads is projected to increase by 25 million km (Dulac 30 2013) including large infrastructure-expansion schemes in Asia (Lechner et al. 2018; Laurance and 31 Arrea 2017) and in South America (Laurance et al. 2001; Killeen 2007)—as well as widespread illegal 32 or unplanned road building (Barber et al. 2014; Laurance et al. 2009). For example, in the Amazon, 33 95% of all deforestation occurs within 5.5 km of a road, and for every km of legal road there are nearly 34 three km of illegal roads (Barber et al. 2014). 35 Interactions and limitations 36 More than any other proximate factor, the dramatic expansion of roads is determining the pace and 37 patterns of habitat disruption and its impacts on biodiversity (Laurance et al. 2009; Laurance and Arrea 38 2017). Much road expansion is poorly planned. Environmental Impact Assessments (EIAs) for roads 39 and other infrastructure are typically too short-term and superficial to detect rare species or assess long- 40 term or indirect impacts of projects (Flyvbjerg 2009; Laurance and Arrea 2017). Another limitation is 41 the consideration of each project in isolation from other existing or planned developments (Laurance et 42 al. 2014b). Hence, EIAs alone are inadequate for planning infrastructure projects and assessing their Do Not Cite, Quote or Distribute 7-31 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 broader environmental, social, and financial impacts and risks (Laurance et al. 2015a; Alamgir et al. 2 2018, 2017). 3 Lessons 4 The large-scale, proactive land-use planning is an option for managing the development of modern 5 infrastructure. Approaches such as the “Global Roadmap” scheme (Laurance and Balmford 2013; 6 Laurance et al. 2014a) Strategic Environmental Assessments (Fischer 2007) can be used to evaluate the 7 relative costs and benefits of infrastructure projects, and to spatially prioritise land-uses to optimise 8 human benefits while limited new infrastructure in areas of intact or critical habitats. For example, the 9 Global Roadmap strategy has been used in parts of Southeast Asia (Sloan et al. 2018), Indochina 10 (Balmford et al. 2016), and sub-Saharan Africa (Laurance et al. 2015b) to devise land-use zoning that 11 can help optimise the many risks and rewards of planned infrastructure projects. 12 [END BOX 7.1 HERE] 13 14 7.3.1.3. Extractive industry development 15 The extent and scale of mining is growing due to increased global demand (UN Environment 2019). 16 Due to declining ore grades, more ore needs to be processed to meet demand, with extensive use of 17 open cast mining. A low-carbon future may be more mineral intensive with for example, clean energy 18 technologies requiring greater inputs in comparison to fossil-fuel-based technologies (Hund et al. 2020). 19 Mining presents cumulative environmental impacts, especially in intensively mined regions (UN 20 Environment 2019). The impact of mining on deforestation varies considerably across minerals and 21 countries. Mining causes significant changes to the environment, for example through mining 22 infrastructure establishment, soil erosion, urban expansion to support a growing workforce and 23 development of mineral commodity supply chains (Sonter et al. 2015). The increasing consumption of 24 gold in developing countries, increased prices, and uncertainty in financial markets is identified as 25 driving gold mining and associated deforestation in the Amazon region (Alvarez-Berrios and Mitchell 26 Aide 2015; Dezécache et al. 2017; Asner and Tupayachi 2017; Espejo et al. 2018). The total estimated 27 area of gold mining throughout the region increased by about 40% between 2012 and 2016 (Asner and 28 Tupayachi 2017). In the Brazilian Amazon, mining significantly increased forest loss up to 70 km 29 beyond mining lease boundaries, causing 11,670 km2 of deforestation between 2005 and 2015, 30 representing 9% of all Amazon forest loss during this time (Sonter et al. 2015). 31 Mining is also an important driver of deforestation in African and Asian countries. In the Democratic 32 Republic of Congo, where the second-largest area of tropical forest in the world occurs, mining-related 33 deforestation exacerbated by violent conflict (Butsic et al. 2015). In India, mining has contributed to 34 deforestation at a district level, with coal, iron and limestone having had the most adverse impact on 35 forest area loss (Ranjan 2019). Gold mining is also identified as a driver of deforestation in Myanmar 36 (Papworth et al. 2017). 37 7.3.1.4. Fire regime changes 38 Wildland fires account for approximately 70% of the global biomass burned annually (Van Der Werf 39 et al. 2017) and constitute a large global source of atmospheric trace gases and aerosols (Gunsch et al. 40 2018; IPCC WGI AR6). Although fires are part of the natural system, the frequency of fires has 41 increased in many areas, exacerbated by decreases in precipitation, including in many regions with 42 humid and temperate forests that rarely experience large-scale fires naturally. Natural and human- 43 ignited fires affect all major biomes, from peatlands through shrublands to tropical and boreal forests, 44 altering ecosystem structure and functioning (Argañaraz et al. 2015; Engel et al. 2019; Mancini et al. 45 2018; Remy et al. 2017; Nunes et al. 2016; Aragão et al. 2018; (Rodríguez Vásquez et al. 2021). Do Not Cite, Quote or Distribute 7-32 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 However, the degree of incidence and regional trends are quite different and a study over 14 year 2 indicated, on average, the largest fires in Australia, boreal North America and Northern Hemisphere 3 Africa (Andela et al. 2019). More than half of the terrestrial surface of the Earth has fire regimes outside 4 the range of natural variability, with changes in fire frequency and intensity posing major challenges 5 for land restoration and recovery (Barger et al. 2018). In some ecosystems, fire prevention might lead 6 to accumulation of large fuel loads that enable wildfires (Moreira et al. 2020a). 7 About 98 Mha of forest and savannahs are estimated to have been affected by fire in 2015 (FAO and 8 UNEP 2020). Fire is a prevalent forest disturbance in the tropics where about 4% of the total forest and 9 savannah area in that year was burned and more than two-thirds of the total area affected was in Africa 10 and South America; mostly open savanna types (FAO and UNEP 2020). Fires have many different 11 causes, with land clearing for agriculture the primary driver in tropical regions, for example, clearance 12 for industrial oil-palm and paper-pulp plantations in Indonesia (Chisholm et al. 2016), or for pastures 13 in the Amazon (Barlow et al. 2020). Other socioeconomic factors are also associated with wildfire 14 regimes such as land-use conflict and socio-demographic aspects (Nunes et al. 2016; Mancini et al. 15 2018). Wildfire regimes are also changing by the influence of climate change, with wildfire seasons 16 becoming longer, wildfire average size increases in many areas and wildfires occurring in areas where 17 they did not occur before (Jolly et al. 2015; Artés et al. 2019). Human influence has likely increased 18 fire weather in some regions of all inhabited continents (IPCC WGI AR6 Technical Summary) and, in 19 the last years, fire seasons of unprecedented magnitude occurred in diverse regions as California (Goss 20 et al. 2020), the Mediterranean basin (Ruffault et al. 2020), Canada (Kirchmeier‐Young et al. 2019) 21 with unprecedented fires in British Columbia in 2021, the Arctic and Siberia (McCarty et al. 2020), 22 Brazilian Amazon (Silva et al. 2021b) and Pantanal (Leal Filho et al. 2021), Chile (Bowman et al. 2019) 23 and Australia (Gallagher et al. 2021; Ward et al. 2020). Lightning plays an important role in the ignition 24 of wildfires, with the incidence of lightning igniting wildfires predicted to increase with rises in global 25 average air temperature (Worden et al. 2017). 26 7.3.1.5. Logging and fuelwood harvest 27 The area of forest designated for production has been relatively stable since 1990. Considering forest 28 uses, about 30% (1.2 billion ha) of all forests is used primarily for production (wood and non-wood 29 forest products), about 10% (424 Mha) is designated for biodiversity conservation, 398 Mha for the 30 protection of soil and water, and 186 Mha is allocated for social services (recreation, tourism, education 31 research and the conservation of cultural and spiritual sites) (FAO and UNEP 2020). While the rate of 32 increase in the area of forest allocated primarily for biodiversity conservation has slowed in the last ten 33 years, the rate of increase in the area of forest allocated for soil and water protection has grown since 34 1990, and notably in the last ten years. Global wood harvest (including from forests, other wooded land 35 and trees outside forests) was estimated to be almost 4.0 billion m3 in 2018 (considering both industrial 36 roundwood and fuelwood) (FAO, 2019). Overall, wood removals are increasing globally as demand 37 for, and the consumption of wood products grows annually by 1% in line with growing populations and 38 incomes with this trend expected to continue in coming decades. When done in a sustainable way, more 39 regrowth will occur and is stimulated by management, resulting in a net sink. However illegal and 40 unsustainable logging (i.e. harvesting of timber in contravention of the laws and regulations of the 41 country of harvest) is a global problem with significant negative economic (e.g. lost revenue), 42 environmental (e.g. deforestation, forest degradation, GHG emissions and biodiversity losses) and 43 social impact (e.g. conflicts over land and resources, disempowerment of local and indigenous 44 communities) (World Bank 2019). Many countries around the world have introduced regulations for 45 the international trade of forest products to reduce illegal logging, with significant and positive impacts 46 (Guan et al. 2018). Do Not Cite, Quote or Distribute 7-33 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Over-extraction of wood for timber and fuelwood) is identified as an important driver of mangrove 2 deforestation and degradation (Fauzi et al. 2019; Bhattarai 2011; Ajonina et al. 2014; Webb et al. 2014; 3 Giri et al. 2015; Thomas et al. 2017; Bhattarai 2011; Ajonina et al. 2014; Webb et al. 2014; Giri et al. 4 2015; Thomas et al. 2017; Fauzi et al. 2019). Unsustainable selective logging and over-extraction of 5 wood is a substantial form of forest and mangrove degradation in many tropical and developing 6 countries, with emissions associated with the extracted wood, incidental damage to the surrounding 7 forest and from logging infrastructure (Pearson et al. 2014, (Fauzi et al. 2019; Bhattarai 2011; Ajonina 8 et al. 2014; Webb et al. 2014; Giri et al. 2015; Thomas et al. 2017).). Traditional fuelwood and charcoal 9 continue to represent a dominant share of total wood consumption in low-income countries (Barger et 10 al. 2018). Regionally, the percentage of total wood harvested used as fuelwood varies from 90% in 11 Africa, 62 % in Asia, 50% in South America to less than 20 % in Europe, North America and Oceania. 12 Under current projections, efforts to intensify wood production in plantation forests, together with 13 increases in fuel-use efficiency and electrification, are suggested to only partly alleviate the pressure on 14 native forests (Barger et al. 2018). Nevertheless, the area of forest under management plans has 15 increased in all regions since 2000 by 233 Mha (FAO-FRA 2020). In regions representing the majority 16 of industrial wood production, forests certified under sustainable forest management programs 17 accounted for 51% of total managed forest area in 2017, an increase from 11% in 2000 (ICFPA 2021). 18 7.3.2. Anthropogenic direct drivers – Agriculture 19 7.3.2.1. Livestock populations and management 20 Enteric fermentation dominates agricultural CH4 emissions (Section 7.2.3) with emissions being a 21 function of both ruminant animal numbers and productivity (output per animal). In addition to enteric 22 fermentation, both CH4 and N2O emissions from manure management (i.e. manure storage and 23 application) and deposition on pasture, make livestock the main agricultural emissions source (Tubiello 24 2019). AR5 reported increases in populations of all major livestock categories between the 1970s and 25 2000s, including ruminants, with increasing numbers directly linked with increasing CH4 emissions 26 (Smith et al. 2014). The SRCCL identified managed pastures as a disproportionately high N2O 27 emissions source within grazing lands, with medium confidence that increased manure production and 28 deposition was a key driver (Jia et al. 2019). The latest data (FAO 2021c) indicate continued global 29 livestock population growth between 1990 and 2019 (Figure 7.10), including increases of 18% in cattle 30 and buffalo numbers, and 30% in sheep and goat numbers, corresponding with CH4 emission trends. 31 Data also indicate increased productivity per animal for example, average increases of 16% in beef, 32 17% in pig meat and 70% in whole (cow) milk per respective animal between 1990 and 2019 (FAO 33 2021c). Despite these advances leading to reduced emissions per unit of product (calories, meat and 34 milk) (FAO 2016; Tubiello 2019), increased individual animal productivity generally requires increased 35 inputs (e.g. feed) and this generates increased emissions (Beauchemin et al. 2020). Manipulation of 36 livestock diets, or improvements in animal genetics or health may counteract some of this. In addition, 37 the production of inputs to facilitate increased animal productivity, may indirectly drive further absolute 38 GHG emissions along the feed supply chain. 39 Although there are several potential drivers (McDermott et al. 2010; Alary V. 2015), increased livestock 40 production is principally in response to growth in demand for animal-sourced food, driven by a growing 41 human population (FAO, 2019) and increased consumption resulting from changes in affluence, notably 42 in middle-income countries (Godfray et al. 2018). Available data document increases in total meat and 43 milk consumption by 24 and 22% respectively between 1990 and 2013, as indicated by average annual 44 per capita supply (FAO 2017a). Updated data indicate that trends of increasing consumption continued 45 between 2014 and 2018 (FAO 2021d). Sustained demand for animal-sourced food is expected to drive Do Not Cite, Quote or Distribute 7-34 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 further livestock sector growth, with global production projected to expand by 14% by 2029, facilitated 2 by maintained product prices and lower feed prices (OECD/FAO 2019). 3 7.3.2.2. Rice cultivation 4 In addition to livestock, both AR5 and the SRCCL identified paddy rice cultivation as an important 5 emissions source (Smith et al. 2014), with medium evidence and high agreement that its expansion is a 6 key driver of growing trends in atmospheric CH4 concentration (Jia et al. 2019). The latest data indicate 7 the global harvested area of rice to have grown by 11% between 1990 and 2019, with total paddy 8 production increasing by 46%, from 519 Mt to 755 Mt (FAO 2021c). Global rice production is projected 9 to increase by 13% by 2028 compared to 2019 levels (OECD/FAO 2019). However, yield increases are 10 expected to limit cultivated area expansion, while dietary shifts from rice to protein as a result of 11 increasing per capita income, is expected to reduce demand in certain regions, with a slight decline in 12 related emissions projected to 2030 (USEPA 2019). 13 Between 1990 and 2019, Africa recorded the greatest increase (+160%) in area under rice cultivation, 14 followed by Asia and the Developing Pacific (+6%), with area reductions evident in all other regions 15 (FAO 2021c) broadly corresponding with related regional CH4 emission (Figures 7.3 and 7.8). Data 16 indicate the greatest growth in consumption (average annual supply per capita) between 1990 and 2013 17 to have occurred in Eastern Europe and West Central Asia (+ 42%) followed by Africa (+ 25%), with 18 little change (+ 1%) observed in Asia and the Developing Pacific (FAO 2017a). Most of the projected 19 increase in global rice consumption is in Africa and Asia (OECD/FAO 2019). 20 7.3.2.3. Synthetic fertiliser use 21 Both AR5 and the SRCCL described considerable increases in global use of synthetic nitrogen fertilisers 22 since the 1970s, which was identified to be a major driver of increasing N2O emissions (Jia et al. 2019). 23 The latest data document a 41% increase in global nitrogen fertiliser use between 1990 and 2019 (FAO 24 2021e) corresponding with associated increased N2O emissions (Figure 7.3). Increased fertiliser use has 25 been driven by pursuit of increased crop yields, with for example, a 61% increase in average global 26 cereal yield per hectare observed during the same period (FAO 2021c), achieved through both increased 27 fertiliser use and varietal improvements. Increased yields are in response to increased demand for food, 28 feed, fuel and fibre crops which in turn has been driven by a growing human population (FAO, 2019), 29 increased demand for animal-sourced food and bioenergy policy (OECD/FAO 2019). Global crop 30 production is projected to increase by almost 15% over the next decade, with low income and emerging 31 regions with greater availability of land and labour resources expected to experience the strongest 32 growth, and account for about 50% of global output growth (OECD/FAO 2019). Increases in global 33 nitrogen fertiliser use are also projected, notably in low income and emerging regions (USEPA 2019). 34 Do Not Cite, Quote or Distribute 7-35 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Figure 7.10 Trends in average global and regional land area under specific land uses (FAO 2021b), 3 inorganic nitrogen fertiliser use (FAO 2021e) (Top) and number of livestock (FAO 2021c) (Bottom) for 4 three decades. For land use classification ‘cropland’ represents the FAOSTAT category ‘arable land’ 5 which includes land under temporary crops, meadow, pasture and fallow. ‘Forest’ and ‘permanent 6 meadow and pasture’ follow FAOSTAT categories. 7 7.3.3. Indirect drivers 8 The indirect drivers behind how humans both use and impact natural resources are outlined in Table 9 7.2, specifically; demographic, economic and cultural, scientific and technological, and institutional 10 and governance drivers. These indirect drivers not only interact with each other at different temporal 11 and spatial scales but are also subject to impacts and feedbacks from the direct drivers (Barger et al. 12 2018). 13 14 Table 7.2 Indirect drivers of anthropogenic land and natural resource use patterns Global and regional trends in population growth: There was a 43% increase in global population between 1990 and 2018. The greatest growth was observed in Africa and the Demography Do Not Cite, Quote or Distribute 7-36 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII Middle East (+ 104%) and least growth in Eastern Europe and West-Central Asia (+ 7%) (FAO 2019b). Global and regional projections: Population is projected to increase by 28% between 2018 and 2050 reaching 9.7 billion (FAO 2019). The world’s population is expected to become older, more urbanised and live in smaller households (UN Environment 2019). Human migration: Growing mobility and population are linked to human migration, a powerful driver of changes in land and resource use patterns at decadal timescales, with the dominant flow of people being from rural areas to urban settlements over the past few decades, notably in the developing world (Adger et al. 2015; Barger et al. 2018). Changes in land use and management come from individual and social responses to economic opportunities (e.g. demand for a particular commodity or improved market access), mediated Economic by institutions and policies (e.g. agricultural subsidies and low-interest credit or government- development led infrastructure projects) (Barger et al. 2018). and cultural Projections on consumption: If the future global population adopts a per capita consumption factors rate similar to that of the developed world, the global capacity to provide land-based resources will be exceeded (Barger et al. 2018). Economic growth in the developing world is projected to double the global consumption of forest and wood products by 2030, with demand likely to exceed production in many developing and emerging economies in Asia and Africa within the next decade (Barger et al. 2018). Global trade: Market distorting agricultural subsidies and globalisation increases pressure on land systems and functions, with global trade and capital flow influencing land use, notably in developing countries (Yao et al. 2018; Furumo and Aide 2017; Pendrill et al. 2019a; (UN Environment 2019), OECD/FAO 2019). Estimates suggest that between 29 and 39% of emissions from deforestation in the tropics resulted from the international trade of agricultural commodities (Pendrill et al. 2019a). Technological factors operates in conjunction with economic drivers of land use and management, whether through intensified farming techniques and biotechnology, high-input Science and approaches to rehabilitating degraded land (e.g. Lin et al. 2017; Guo et al. 2020) or through technology new forms of data collection and monitoring (e.g. Song et al. 2018; Thyagharajan and Vignesh 2019; Arévalo et al. 2020). Changes in farming and forestry systems: Changes can have both positive and negative impacts regarding multiple factors, including GHG emission trends. Fast advancing technologies shape production and consumption, and drive land-use patterns and terrestrial ecosystems at various scales. Innovation is expected to help drive increases in global crop production during the next decade (OECD/FAO 2019). For example, emerging gene editing technologies, may advance crop breeding capabilities, though are subject to biosafety, public acceptance and regulatory approval (Jaganathan et al. 2018; Chen et al. 2019; Schmidt et al. 2020). Technological changes were significant for the expansion of soybean in Brazil by adapting to different soils and photoperiods (Abrahão and Costa 2018). In Asia, technological development changed agriculture with significant improvements in production and adaptation to climate change (Thomson et al. 2019; Giller and Ewert 2019; Anderson et al. 2020; Cassman and Grassini 2020). Developments such as precision agriculture and drip irrigation have facilitated more efficient agrochemical and water use (UN Environment 2019). Research and development are central to forest restoration strategies that have become increasingly important around the world as costs vary depending on methods used, from natural regeneration with native tree species to active restoration using site preparation and planting (Löf et al. 2019). In addition, climate change poses the challenge about tree species selection in the future. Innovations in the forest sector innovations also form the basis of a bioeconomy associated with bioproducts and new processes (Verkerk et al. 2020; Cross-Working Group Box 3 in Chapter 12). Do Not Cite, Quote or Distribute 7-37 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII Emerging mitigation technologies: Chemically synthesised methanogen inhibitors for ruminants are expected to be commercially available in some countries within the next two years and have considerable CH4 mitigation potential (McGinn et al. 2019; Melgar et al. 2020; Beauchemin et al. 2020; Reisinger et al. 2021) (Section 7.4.3). There is growing literature (in both academic and non-academic sphere) on the biological engineering of protein. Although in its infancy and subject to investment, technological development, regulatory approval and consumer acceptance, it is suggested to have the potential to disrupt current livestock production systems and land use (Stephens et al. 2018; Ben-Arye and Levenberg 2019; Post et al. 2020; RethinkX 2019). The extent to which this is possible and the overall climate benefits are unclear (Lynch and Pierrehumbert 2019; Chriki and Hocquette 2020). Institutional factors often moderate the relevance and impact of changes in economic and demographic variables related to resource exploitation and use. Institutions encompass the Institutions and rule of law, legal frameworks and other social structures (e.g. civil society networks and governance movements) determining land management (e.g. formal and informal property rights, regimes and their enforcement); information and knowledge exchange systems; local and traditional knowledge and practice systems (Barger et al. 2018). Land rights: Land tenure often allows communities to exercise traditional governance based on traditional ecological knowledge, devolved and dynamic access rights, judicious use, equitable distribution of benefits (Mantyka-Pringle et al. 2017; Wynberg 2017; Thomas et al. 2017), biodiversity (Contreras-Negrete et al. 2014) and fire and grazing management (Levang et al. 2015; Varghese et al. 2015). Agreements and Finance: Since AR5, global agreements were reached on climate change, sustainable development goals, and the mobilisation of finance for development and climate action. Several countries adopted policies and commitments to restore degraded land (Barger et al. 2018). The UN Environment Programme (UNEP) and the Food and Agriculture Organization of the UN (FAO), launched the UN Decade on Ecosystem Restoration (https://www.decadeonrestoration.org/). Companies have also made pledges to reduce impacts on forests and on the rights of local communities as well as eliminating deforestation from their supply chains. The finance sector, a crucial driver behind action (Section 7.6, Box 7.12), has also started to make explicit commitments to avoiding environmental damage (Barger et al. 2018) and net zero targets (Forest Trends Ecosystem Marketplace 2021), though investment is sensitive to market outlook. 1 2 Do Not Cite, Quote or Distribute 7-38 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.4. Assessment of AFOLU mitigation measures including trade-offs and 2 synergies 3 AFOLU mitigation or land-based climate change mitigation (used in this chapter interchangeably) are 4 a variety of land management or demand management practices that reduce GHG emissions and/or 5 enhance carbon sequestration within the land system (i.e. in forests, wetlands, grasslands, croplands and 6 pasturelands). If implemented with benefits to human well-being and biodiversity, land-based 7 mitigation measures are often referred to as nature-based solutions and/or natural climate solutions 8 (Glossary). Measures that result in a net removal of GHGs from the atmosphere and storage in either 9 living or dead organic material, or in geological stores, are known as CDR, and in previous IPCC reports 10 were sometimes referred to as greenhouse gas removal (GGR) or negative emissions technologies 11 (NETs) (Rogelj et al. 2018a; Jia et al. 2019). This section evaluates current knowledge and latest 12 scientific literature on AFOLU mitigation measures and potentials, including land-based CDR 13 measures. Section 7.4.1 provides an overview of the approaches for estimating mitigation potential, the 14 co-benefits and risks from land-based mitigation measures, estimated global and regional mitigation 15 potential and associated costs according to literature published over the last decade. Subsequent 16 subsections assess literature on 20 key AFOLU mitigation measures specifically providing: 17 ● A description of activities, co-benefits, risks and implementation opportunities and barriers 18 ● A summary of conclusions in AR5 and IPCC Special Reports (SR15, SROCCC and SRCCL) 19 ● An overview of literature and developments since the AR5 and IPCC Special Reports 20 ● An assessment and conclusion based on current evidence 21 Measures are categorised as supply-side activities in: (1) forests and other ecosystems (Section 7.4.2), 22 (2) agriculture (Section 7.4.3), (3) bioenergy and other land-based energy technologies (Section 7.4.4); 23 as well as (4) demand-side activities (Section 7.4.5) (Figure 7.11). Several information boxes are 24 dispersed within the section and provide supporting material, including case studies exploring a range 25 of topics from climate-smart forestry in Europe (Box 7.2), agroforestry in Brazil (Box 7.3), climate- 26 smart village approaches (Box 7.4), farm systems approaches (Box 7.5), mitigation within Indian 27 agriculture (Box 7.6), and bioenergy and BECCS mitigation calculations (Box 7.7). Novel measures, 28 including enhanced weathering and novel foods are covered in Chapter 12, this report. In addition, as 29 mitigation within AFOLU concerns land management and use of land resources, AFOLU measures 30 impact other sectors. Accordingly, AFOLU measures are also discussed in other sectoral chapters 31 within this report, notably demand-side solutions (Chapter 5), bioenergy and Bioenergy with Carbon 32 Capture and Storage (BECCS) (Chapter 6), the use of wood products and biomass in buildings (Chapter 33 9), and CDR measures, food systems and land related impacts, risks and opportunities of mitigation 34 measures (Chapter 12). 35 7.4.1. Introduction and overview of mitigation potential 36 7.4.1.1. Estimating mitigation potentials 37 Mitigation potentials for AFOLU measures are estimated by calculating the scale of emissions 38 reductions or carbon sequestration against a counterfactual scenario without mitigation activities. The 39 types of mitigation potential estimates in recent literature include: (1) technical potential (the 40 biophysical potential or amount possible with current technologies), (2) economic potential (constrained 41 by costs, usually by a given carbon price (Table 7.3), (3) sustainable potential (constrained by 42 environmental safeguards and/or natural resources, e.g. limiting natural forest conversion), and (4) 43 feasible potential (constrained by environmental, socio-cultural, and/or institutional barriers), however, 44 there are no set definitions used in literature. In addition to types of mitigation estimates, there are two Do Not Cite, Quote or Distribute 7-39 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 AFOLU mitigation categories often calculated: supply-side measures (land management interventions) 2 and demand-side measures (interventions that require a change in consumer behaviour). 3 Two main approaches to estimating mitigation potentials include: 1) studies on individual measures 4 and/or sectors – henceforth referred to as sectoral assessments, and 2) integrated assessment models 5 (IAM). Sectoral assessments include studies focusing on one activity (e.g. agroforestry) based on spatial 6 and biophysical data, as well as econometric and optimisation models for a sector, e.g. the forest or 7 agriculture sector, and therefore cover a large suite of practices and activities while representing a broad 8 body of literature. Sectoral assessments however, rarely capture cross-sector interactions or impacts, 9 making it difficult to completely account for land competition, trade-offs, and double counting when 10 aggregating sectoral estimates across different studies and methods (Smith et al. 2014; Jia et al. 2019). 11 On the other hand, IAMs assess the climate impact of multiple and interlinked practices across sectors 12 and therefore, can account for interactions and trade-offs (including land competition, use of other 13 resources and international trade) between them. However, the number of land-based measures used in 14 IAMs are limited compared with the sectoral portfolio (Figure 7.11). The resolution of land-based 15 measures in IAMs are also generally coarser compared to some sectoral estimates, and as such, may be 16 less robust for individual measures (Roe et al. 2021). Given the differences between and strengths and 17 weaknesses of the two approaches, it is helpful to compare the estimates from both. We combine 18 estimates from both approaches to establish an updated range of global land-based mitigation potential. 19 For the 20 land-based mitigation measures outlined in this section, the mitigation potential estimates 20 are largely derived from sectoral approaches, and where data is available, are compared to IAM 21 estimates. Integrated assessment models and the emissions trajectories, cost-effectiveness and trade- 22 offs of various mitigation pathways are detailed in Section 7.5. It should be noted that the underlying 23 literature for sectoral as well as IAM mitigation estimates consider GWP100 IPCC AR5 values (CH4 = 24 28, N2O = 265) as well as GWP100 IPCC AR4 values (CH4 = 25, N2O = 298) to convert CH4 and N2O 25 to CO2-eq. Where possible, we note the various GWP100 values (in IAM estimates, and the wetlands and 26 agriculture sections), however in some instances, the varying GWP100 values used across studies 27 prevents description of non-CO2 gases in native units as well as conversion to AR6 GWP100 (CH4 = 27, 28 N2O = 273) CO2-eq values to aggregate sectoral assessment estimates. 29 7.4.1.2. Co-benefits and risks 30 Land interventions have interlinked implications for climate mitigation, adaptation, food security, 31 biodiversity, ecosystem services, and other environmental and societal challenges (Section 7.6.5). 32 Therefore, it is important to consider the net effect of mitigation measures for achieving both climate 33 and non-climate goals (Section 7.1). 34 While it is helpful to assess the general benefits, risks and opportunities possible for land-based 35 mitigation measures (Smith et al. 2019a), their efficacy and scale of benefit or risk largely depends on 36 the type of activity undertaken, deployment strategy (e.g. scale, method), and context (e.g. soil, biome, 37 climate, food system, land ownership) that vary geographically and over time (Smith et al. 2019a,b; 38 Hurlbert et al. 2019; Chapter 12, Section 12.5) (robust evidence, high agreement). Impacts of land- 39 based mitigation measures are therefore highly context specific and conclusions from specific studies 40 may not be universally applicable. If implemented at appropriate scales and in a sustainable manner, 41 land-based mitigation practices have the capacity to reduce emissions and sequester billions of tonnes 42 of carbon from the atmosphere over coming decades, while also preserving or enhancing biodiversity, 43 water quality and supply, air quality, soil fertility, food and wood security, livelihoods, resilience to 44 droughts, floods and other natural disasters, and positively contributing to ecosystem health and human 45 wellbeing (high confidence) (Toensmeier 2016; Karlsson et al. 2020). 46 Overall, measures in the AFOLU sector are uniquely positioned to deliver substantial co-benefits. Do Not Cite, Quote or Distribute 7-40 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 However, the negative consequences of inappropriate or misguided design and implementation of 2 measures may be considerable, potentially impacting for example, mitigation permanence, longevity, 3 and leakage, biodiversity, wider ecosystem functioning, livelihoods, food security and human well- 4 being (Section 7.6; WGII, Box 2.2. ‘Risks of maladaptive mitigation’. Land-based mitigation may also 5 face limitations and trade-offs in achieving sustained emission reductions and/or removals due to other 6 land challenges including climate change impacts. It is widely recognised that land-use planning that is 7 context-specific, considers other sustainable development goals, and is adaptable over time can help 8 achieve land-based mitigation that maximises co-benefits, avoids or limits trade-offs, and delivers on 9 international policy goals including the SDGs, Land Degradation Neutrality, and Convention on 10 Biological Diversity (Section 7.6; Chapter 12). 11 Potential co-benefits and trade-offs are outlined for each of the 20 land-based mitigation measures in 12 the proceeding sub-sections and summarised in Figure 7.12. Section 7.6.5. discusses general links with 13 ecosystem services, human well-being and adaptation, while Chapter 12 (Section 12.5) provides an in- 14 depth assessment of the land related impacts, risks and opportunities associated with mitigation options 15 across sectors, including positive and negative effects on land resources, water, biodiversity, climate, 16 and food security. 17 7.4.1.3. Overview of global and regional technical and economic potentials in AFOLU 18 IPCC AR5 (2014). In the AR5, the economic mitigation potential of supply-side measures in the 19 AFOLU sector was estimated at 7.18–10.60 GtCO2-eq yr-1 in 2030 with carbon prices up to USD100 20 tCO2-eq-1, about a third of which could be achieved at < USD20 tCO2-eq-1 (medium evidence; medium 21 agreement) (Smith et al. 2014). AR5 provided a summary table of individual AFOLU mitigation 22 measures, but did not conduct a detailed assessment for each. 23 IPCC SRCCL (2019). The SRCCL assessed the full range of technical, economic and sustainability 24 mitigation potentials in AFOLU for the period 2030-2050 and identified reduced deforestation and 25 forest degradation to have greatest potential for reducing supply-side emissions (0.4–5.8 GtCO2-eq yr– 1 26 ) (high confidence) followed by combined agriculture measures, 0.3–3.4 GtCO2-eq yr–1 (medium 27 confidence) (Jia et al. 2019). For the demand-side estimates, shifting towards healthy, sustainable diets 28 (0.7–8.0 GtCO2-eq yr–1) (high confidence) had the highest potential, followed by reduced food loss and 29 waste (0.8–4.5 GtCO2-eq yr–1) (high confidence). Measures with greatest potential for CDR were 30 afforestation/reforestation (0.5–10.1 GtCO2-eq yr–1) (medium confidence), soil carbon sequestration in 31 croplands and grasslands (0.4–8.6 GtCO2-eq yr–1) (medium confidence) and BECCS (0.4–11.3 GtCO2- 32 eq yr–1) (medium confidence). The SRCCL did not explore regional potential, associated feasibility nor 33 provide detailed analysis of costs. 34 IPCC AR6. This assessment concludes the likely range of global land-based mitigation potential is 35 approximately 8 – 14 GtCO2-eq yr-1 between 2020-2050 with carbon prices up to USD100 tCO2-eq-1, 36 about half of the technical potential (medium evidence; medium agreement). About 30-50% could be 37 achieved < USD20 tCO2-eq-1 (Table 7.3). The global economic potential estimates in this assessment 38 are slightly higher than the AR5 range. Since AR5, there have been numerous new global assessments 39 of sectoral land-based mitigation potential (Fuss et al. 2018; Griscom et al. 2017, 2020; Roe et al. 2019; 40 Jia et al. 2019; Griscom et al. 2020; Roe et al. 2021) as well as IAM estimates of mitigation potential 41 (Frank et al. 2019; Johnston and Radeloff 2019; Riahi et al. 2017; Baker et al. 2019; Popp et al. 2017; 42 Rogelj et al. 2018a), expanding the scope of AFOLU mitigation measures included and substantially 43 improving the robustness and spatial resolution of mitigation estimates. A recent development is an 44 assessment of country-level technical and economic (USD100 tCO2-eq-1) mitigation potential for 20 45 AFOLU measures, including for demand-side and soil organic carbon sequestration in croplands and 46 grasslands, not estimated before (Roe et al. 2021). Estimates on costs, feasibility, sustainability, Do Not Cite, Quote or Distribute 7-41 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 benefits, and risks have also been developed for some mitigation measures, and they continue to be 2 active areas of research. Developing more refined sustainable potentials at a country-level will be an 3 important next step. Although most mitigation estimates still do not consider the impact of future 4 climate change, there are some emerging studies that do (Doelman et al. 2019; Sonntag et al. 2016). 5 Given the IPCC WG1 finding that the land sink is continuing to increase although its efficiency is 6 decreasing with climate change, it will be critical to better understand how future climate will affect 7 mitigation potentials, particularly from CDR measures. 8 Across global sectoral studies, the economic mitigation potential (up to USD100 tCO2-eq-1) of supply- 9 side measures in AFOLU for the period 2020-2050 is 11.4 mean (5.6–19.8 full range) GtCO2-eq yr–1, 10 about 50% of the technical potential of 24.2 (4.9 - 58) GtCO2-eq yr–1 (Table 7.3). Adding 2.1 GtCO2-eq 11 yr–1 from demand-side measures (accounting only for diverted agricultural production to avoid double 12 counting with land-use change effects), total land-based mitigation potential up to USD100 tCO2-eq-1 13 is 13.6 (6.7 – 23.4) GtCO2-eq yr–1. This estimate aligns with the most recent regional assessment (Roe 14 et al. 2021), which found the aggregate global mitigation potential of supply and demand-side measures 15 to be 13.8 ± 3.1 GtCO2-eq yr–1 up to USD100 tCO2-eq-1 for the period 2020-2050. Across integrated 16 assessment models (IAMs), the economic potential for land-based mitigation (Agriculture, LULUCF 17 and BECCS) for USD100 tCO2-eq-1 is 7.9 mean (4.1–17.3 range) GtCO2-eq yr–1 in 2050 (Table 7.3). 18 We add the estimate for BECCS here to provide the full land-based potential, as IAMs optimize land 19 allocation based on costs, which displaces land-based CDR activities for BECCS. Combining both IAM 20 and sectoral approaches, the likely range is therefore 7.9–13.6 (rounded to 8–14) GtCO2-eq yr-1 up to 21 USD100 tCO2-eq-1 between 2020-2050. Considering both IAM and sectoral economic potential 22 estimates, land-based mitigation could have the capacity to make the AFOLU sector net negative GHG 23 emissions from 2036 (Figure 7.12), although there are highly variable mitigation strategies for how 24 AFOLU potential can be deployed for achieving climate targets (Illustrative Mitigation Pathways in 25 7.5.5). Economic potential estimates, which reflect a public willingness to pay, may be more relevant 26 for policy making compared with technical potentials which reflect a theoretical maximum that may 27 not be feasible or sustainable. 28 Among the mitigation options, the protection, improved management, and restoration of forests and 29 other ecosystems (wetlands, savannas and grasslands) have the largest potential to reduce emissions 30 and/or sequester carbon at 7.3 (3.9–13.1) GtCO2-eq yr-1 (up to USD100 tCO2-eq-1), with measures that 31 ‘protect’ having the single highest total mitigation and mitigation densities (mitigation per area) in 32 AFOLU (Table 7.3, Figure 7.11). Agriculture provides the second largest share of mitigation, with 4.1 33 (1.7–6.7) GtCO2-eq yr-1 potential (up to USD100 tCO2-eq-1), from soil carbon management in croplands 34 and grasslands, agroforestry, biochar, rice cultivation, and livestock and nutrient management Table 35 7.3, Figure 7.11. Demand-side measures including shifting to sustainable healthy diets, reducing food 36 waste, and improving wood products can mitigate 2.2 (1.1 - 3.6) GtCO2-eq yr-1 when accounting only 37 for diverted agricultural production from diets and food waste to avoid double counting with measures 38 in forests and other ecosystems (Table 7.3, Figure 7.11). The potential of demand-side measures 39 increases three-fold, to 6.5 (4 – 9.5) GtCO2-eq yr-1 when accounting for the entire value chain including 40 land-use effects, but would overlap with other measures and is therefore not additive. 41 Most mitigation options are available and ready to deploy. Emissions reductions can be unlocked 42 relatively quickly, whereas CDR need upfront investment to generate sequestration over time. The 43 protection of natural ecosystems, carbon sequestration in agriculture, sustainable healthy diets and 44 reduced food waste have especially high co-benefits and cost efficiency. Avoiding the conversion of 45 carbon-rich primary peatlands, coastal wetlands and forests is particularly important as most carbon lost 46 from those ecosystems are irrecoverable through restoration by the 2050 timeline of achieving net zero 47 carbon emissions (Goldstein et al. 2020). Sustainable intensification, shifting diets, reducing food waste Do Not Cite, Quote or Distribute 7-42 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 could enhance efficiencies and reduce agricultural land needs, and are therefore critical for enabling 2 supply-side measures such as reduced deforestation, restoration, as well as reducing N2O and CH4 3 emissions from agricultural production - as seen in the Illustrative Mitigation Pathway IMP-SP (Section 4 7.5.6). Although agriculture measures that reduce non-CO2, particularly of CH4, are important for near- 5 term emissions reductions, they have less economic potential due to costs. Demand-side measures may 6 be able to deliver non-CO2 emissions reductions more cost efficiently. 7 Regionally, economic mitigation potential up to USD100 tCO2-eq-1 is estimated to be greatest in tropical 8 countries in Asia and developing Pacific (34%), Latin America and the Caribbean (24%), and Africa 9 and the Middle East (18%) because of the large potential from reducing deforestation and sequestering 10 carbon in forests and agriculture (Figure 7.11). However, there is also considerable potential in 11 Developed Countries (18%) and more modest potential in Eastern Europe and West-Central Asia (5%). 12 These results are in line with the IAM regional mitigation potentials (Figure 7.11). The protection of 13 forests and other ecosystems is the dominant source of mitigation potential in tropical regions, whereas 14 carbon sequestration in agricultural land and demand-side measures are important in Developed 15 Countries and Asia and developing Pacific. The restoration and management of forests and other 16 ecosystems is more geographically distributed, with all regions having significant potential. Regions 17 with large livestock herds (Developed Countries, Latin America) and rice paddy fields (Asia and 18 developing Pacific) have potential to reduce CH4. As expected, the highest total potential is associated 19 with countries and regions with large land areas, however when considering mitigation density (total 20 potential per hectare), many smaller countries, particularly those with wetlands have disproportionately 21 high levels of mitigation for their size (Roe et al. 2021). As global commodity markets connect regions, 22 AFOLU measures may create synergies and trade-offs across the world, which could make national 23 demand-side measures for example, important in mitigating supply-side emissions elsewhere (Kallio & 24 Solberg 2018 ). 25 Although economic potentials provide more realistic, near-term climate mitigation compared to 26 technical potentials, they still do not account for feasibility barriers and enabling conditions that vary 27 by region and country. For example, according to most models, including IAMs, avoided deforestation 28 is the cheapest land-based mitigation option (Table 7.3, Sections 7.5.3 and 7.5.4), however 29 implementing interventions aimed at reducing deforestation (including REDD+) often have higher 30 transaction and implementation costs than expected due to various barriers and enabling conditions 31 (Luttrell et al. 2018; Section 7.6). The feasibility of implementing AFOLU mitigation measures, 32 including those with multiple co-benefits, depends on varying economic, technological, institutional, 33 socio-cultural, environmental and geophysical barriers (high confidence) (Smith et al. 2019a). The 34 section for each individual mitigation measure provides an overview of co-benefits and risks associated 35 with the measure and Section 7.6.6 outlines key enabling factors and barriers for implementation. 36 37 Table 7.3 Estimated annual mitigation potential (GtCO2-eq yr-1) in 2020-2050 of AFOLU mitigation 38 options by carbon price. Estimates reflect sectoral studies based on a comprehensive literature review 39 updating data from (Roe et al. 2019) and integrated assessment models using the IPCC AR6 database 40 (Section 7.5). Values represent the mean, and full range of potential. Sectoral mitigation estimates are 41 averaged for the years 2020-2050 to capture a wider range of literature, and the IAM estimates are given 42 for 2050 as many model assumptions delay most land-based mitigation to mid-century. The sectoral 43 potentials are the sum of global estimates for the individual measures listed for each option. IAM 44 potentials are given for mitigation options with available data; e.g., net land-use CO2 for total forests & 45 other ecosystems, and land sequestration from A/R, but not reduced deforestation (protect). Sectoral 46 estimates predominantly use GWP100 IPCC AR5 values (CH4 = 28, N2O = 265), although some use 47 GWP100 IPCC AR4 values (CH4 = 25, N2O = 298); and the IAMs use GWP100 IPCC AR6 values (CH4 = Do Not Cite, Quote or Distribute 7-43 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 27, N2O = 273). The sectoral and IAM estimates reflected here do not account for the substitution effects 2 of avoiding fossil fuel emissions nor emissions from other more energy intensive resources/materials. For 3 example, BECCS estimates only consider the carbon dioxide removal (CDR) via geological storage 4 component and not potential mitigation derived from the displacement of fossil fuel use in the energy 5 sector. Mitigation potential from substitution effects are included in the other sectoral chapters like 6 energy, transport, buildings and industry. The total AFOLU sectoral estimate aggregates potential from 7 agriculture, forests & other ecosystems, and diverted agricultural production from avoided food waste 8 and diet shifts (excluding land-use impacts to avoid double counting). Because of potential overlaps 9 between measures, sectoral values from BECCS and the full value chain potential from demand-side 10 measures are not summed with AFOLU. IAMs account for land competition and resource optimization 11 and can therefore sum across all available categories to derive the total AFOLU potential. Key: ND = no 12 data; Sectoral = as assessed by sectoral literature review; IAM = as assessed by integrated assessment 13 models; EJ = ExaJoule primary energy. Estimate < USD20 < USD50 < USD100 Mitigation option Technical type tCO2-eq-1 tCO2-eq-1 tCO2-eq-1 Sectoral 0.9 (0.5 - 1.4) 1.6 (1 – 2.4) 4.1 (1.7 - 6.7) 11.2 (1.6 - 28.5) Agriculture total IAM 0.9 (0 - 3.1) 1.3 (0 - 3.2) 1.8 (0.7 - 3.3) ND Agriculture - Carbon sequestration Sectoral 0.5 (0.4 - 0.6) 1.2 (0.9 - 1.6) 3.4 (1.4 - 5.5) 9.5 (1.1 - 25.3) (soil carbon management in croplands and grasslands, agroforestry, and biochar) IAM ND ND ND ND Agriculture - Reduce CH4 and N2O Sectoral 0.4 (0.1 - 0.8) 0.4 (0.1 - 0.8) 0.6 (0.3 - 1.3) 1.7 (0.5 - 3.2) emissions (improve enteric fermentation, manure management, nutrient management, and rice IAM 0.9 (0 - 3.1) 1.3 (0 - 3.2) 1.8 (0.7 - 3.3) ND cultivation) Sectoral 2.9 (2.2 - 3.5) 3.1 (1.4 - 5.1) 7.3 (3.9 - 13.1) 13 (5 - 29.5) Forests & other ecosystems total IAM 2.4 (0 - 10.5) 3.3 (0 - 9.9) 4.2 (0 - 12.1) ND Forests & other ecosystems - Protect Sectoral 2.3 (1.7 - 2.9) 2.4 (1.2 - 3.6) 4.0 (2.5 - 7.4) 6.2 (2.8 - 14.4) (reduce deforestation, loss and degradation of peatlands, coastal wetlands, and grasslands) IAM ND ND ND ND Forests & other ecosystems - Restore Sectoral 0.15 0.7 (0.2 - 1.5) 2.1 (0.8 - 3.8) 5 (1.1 - 12.3) (afforestation, reforestation, peatland IAM restoration, coastal wetland restoration) 0.6 (0.2 - 6.5) 0.6 (0.01 - 8.3) 0.7 (0.07 - 6.8) ND (A/R) Forests & other ecosystems - Manage Sectoral 0.4 (0.3 - 0.4) ND 1.2 (0.6 - 1.9) 1.8 (1.1 - 2.8) (improve forest management, fire management) IAM ND ND ND ND Demand-side measures Sectoral ND ND 2.2 (1.1 - 3.6)* 4.2 (2.2 - 7.1)* (shift to sustainable healthy diets, reduce food waste, and enhanced and improved use of wood products) * for all three only the direct avoided IAM ND ND ND ND emissions; land use effects are in measures above Sectoral ND ND 1.6 (0.5 - 3.5) 5.9 (0.5 - 11.3) Do Not Cite, Quote or Distribute 7-44 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII BECCS (only the CDR component, i.e the geological IAM 0.08 (0 - 0.7) 0.5 (0 - 6) 1.8 (0.2 - 9.9) ND storage. Substitution effects are accounted in other sectoral chapters: energy, transport) Bioenergy from residues Sectoral ND ND ND Up to 57 EJ yr-1 TOTAL AFOLU (agriculture, forests & other ecosystems, Sectoral 3.8 (2.7 - 4.9) 4.3 (2.3 - 6.7) 13.6 (6.7 - 23.4) 28.4 (8.8 - 65.1) diverted ag production from demand-side) TOTAL AFOLU (agriculture, forests & other ecosystems, IAM 3.4 (0 - 14.6) 5.3 (0.6 - 19.4) 7.9 (4.1 - 17.3) ND BECCS) Do Not Cite, Quote or Distribute 7-45 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 3 Figure 7.11 Global and regional mitigation potential (GtCO2-eq yr-1) in 2020–2050 for 20 land-based 4 measures. (a) Global estimates represent the mean (bar) and full range (error bars) of the economic Do Not Cite, Quote or Distribute 7-46 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 potential (up to USD100 tCO2-eq-1) based on a comprehensive literature review of sectoral studies 2 (references are outlined in the sub-section for each measure in 7.4.2–7.4.5). Potential co-benefits and 3 trade-offs for each of the 20 measures are summarized in icons. (b) Regional estimates illustrate the 4 mean technical (T) and economic (E) (up to USD100 tCO2-eq-1 ) sectoral potential based on data from 5 (Roe et al. 2021). IAM economic potential (M) (USD100 tCO2-eq-1) data is from the IPCC AR6 6 database. 7 8 9 10 Figure 7.12 Historic land sector GHG flux estimates and illustrative AFOLU mitigation pathways to 11 2050, based on data presented in Sections 7.2, 7.4 and 7.5. Historic trends consider both (A) 12 anthropogenic (AFOLU) GHG fluxes (GtCO2-eq yr-1) according to FAOSTAT (FAO 2021a; 2021b) and 13 (B) the estimated natural land CO2 sink according to (Friedlingstein et al. 2020). Note that for the 14 anthropogenic net land CO2 flux component, several approaches and methods are described within the 15 literature (Section 7.2.2) with a wide range in estimates. For clarity, only one dataset (FAOSTAT) is 16 illustrated here. It is not intended to indicate preference for one particular method over others. Historic 17 flux trends are illustrated to 2019, the latest year for which data is available. Projected economic 18 mitigation potential (at costs of up to USD100 tCO2-eq-1) includes estimates from IAMs and sectoral 19 studies (Table 7.3). The sectoral estimates are disaggregated into agriculture + forests & other 20 ecosystems, + demand-side measures (only accounting for diverted agricultural production to avoid 21 double counting), and + BECCS (illustrating that there may be additional potential, with the caveat that 22 there is likely overlap with other measures). Projected mitigation assumes adoption of measures to 23 achieve increasing, linear mitigation, reaching average annual potential in 2050, although this does not 24 reflect deployment rates for most measures. For illustrative purposes, a pathway to projected emissions in 25 2050 according to a scenario of current policy (C7 - Above 3.0oC - Model: GCAM 5.3) is additionally 26 included for reference. 27 Do Not Cite, Quote or Distribute 7-47 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.4.2. Forests and other ecosystems 2 7.4.2.1. Reduce deforestation and degradation 3 Activities, co-benefits, risks and implementation opportunities and barriers. Reducing deforestation 4 and forest degradation conserves existing carbon pools in forest vegetation and soil by avoiding tree 5 cover loss and disturbance. Protecting forests involves controlling the drivers of deforestation (such as 6 commercial and subsistence agriculture, mining, urban expansion) and forest degradation (such as 7 overharvesting including fuelwood collection, poor harvesting practices, overgrazing, pest outbreaks, 8 and extreme wildfires), as well as by establishing well designed, managed and funded protected areas 9 (Barber et al. 2020), improving law enforcement, forest governance and land tenure, supporting 10 community forest management and introducing forest certification (Smith et al. 2019b). Reducing 11 deforestation provides numerous and substantial co-benefits, preserving biodiversity and ecosystem 12 services (e.g. air and water filtration, water cycling, nutrient cycling) more effectively and at lower costs 13 than afforestation/reforestation (Jia et al. 2019). Potential adverse side effects of these conservation 14 measures include reducing the potential for agriculture land expansion, restricting the rights and access 15 of local people to forest resources, or increasing the dependence of local people to insecure external 16 funding. Barriers to implementation include unclear land tenure, weak environmental governance, 17 insufficient funds, and increasing pressures associated to agriculture conversion, resource exploitation 18 and infrastructure development (Sections 7.3 and 7.6). 19 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 20 potential, costs, and pathways. Reducing deforestation and forest degradation represents one of the 21 most effective options for climate change mitigation, with technical potential estimated at 0.4–5.8 22 GtCO2 yr–1 by 2050 (high confidence) (SRCCL, Chapters 2 and 4, and Table 6.14). The higher technical 23 estimate represents a complete halting of land use conversion in forests and peatland forests (i.e., 24 assuming recent rates of carbon loss are saved each year) and includes vegetation and soil carbon pools. 25 Ranges of economic potentials for forestry ranged in AR5 from 0.01–1.45 GtCO2 yr-1 for USD20 tCO2- 1 26 to 0.2–13.8 GtCO2 yr-1 for USD100 tCO2-1 by 2030 with reduced deforestation dominating the forestry 27 mitigation potential LAM and MAF, but very little potential in OECD-1990 and EIT (IPCC AR5). 28 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Since the 29 SRCCL, several studies have provided updated and convergent estimates of economic mitigation 30 potentials by region (Busch et al. 2019; Griscom et al. 2020; Austin et al. 2020; Roe et al. 2021). 31 Tropical forests and /savannas in Latin America provide the largest share of mitigation potential (3.9 32 GtCO2 yr-1 technical, 2.5 GtCO2 yr-1 at USD100 tCO2-1) followed by Southeast Asia (2.2 GtCO2 yr-1 33 technical, 1.5 GtCO2 yr-1 at USD100 tCO2-1) and Africa (2.2 GtCO2 yr-1 technical, 1.2 GtCO2 yr-1 at 34 USD100 tCO2-1) (Roe et al. 2021). Tropical forests continue to account for the highest rates of 35 deforestation and associated GHG emissions. While deforestation shows signs of decreasing in several 36 countries, in others, it continues at a high rate or is increasing (Turubanova et al. 2018). Between 2010- 37 2020, the rate of net forest loss was 4.7 Mha yr-1 with Africa and South America presenting the largest 38 shares (3.9 Mha and 2.6 Mha, respectively) (FAO 2020a). 39 A major uncertainty in all studies on avoided deforestation potential is their reliance on future reference 40 levels that vary across studies and approaches. If food demand increases in the future, for example, the 41 area of land deforested will likely increase, suggesting more technical potential for avoiding 42 deforestation. Transboundary leakage due to market adjustments could also increase costs or reduce 43 effectiveness of avoiding deforestation (e.g., Ingalls et al. 2018; Gingrich et al. 2019). Regarding forest 44 regrowth, there are uncertainties about the time for the secondary forest carbon saturation (Zhu et al. 45 2018; Houghton and Nassikas 2017). Permanence of avoided deforestation may also be a concern due 46 to the impacts of climate change and disturbance of other biogeochemical cycles on the world’s forests 47 that can result in future potential changes in terrestrial ecosystem productivity, climate-driven Do Not Cite, Quote or Distribute 7-48 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 vegetation migration, wildfires, forest regrowth and carbon dynamics (Ballantyne et al. 2012; Kim et 2 al. 2017b; Lovejoy and Nobre 2018; Aragão et al. 2018). 3 Critical assessment and conclusion. Based on studies since AR5, the technical mitigation potential for 4 reducing deforestation and degradation is significant, providing 4.5 (2.3 - 7) GtCO2 yr-1 globally by 5 2050, of which 3.4 (2.3 – 6.4) GtCO2 yr-1 is available at below USD100 tCO2-1 (medium confidence) 6 (Figure 7.11). Over the last decade, hundreds of subnational initiatives that aim to reduce deforestation 7 related emissions have been implemented across the tropics (Section 7.6). Reduced deforestation is a 8 significant piece of the NDCs in the Paris Agreement (Seddon et al. 2020) and keeping the temperature 9 below 1.5oC (Crusius 2020). Conservation of forests provides multiple co-benefits linked to ecosystem 10 services, biodiversity and sustainable development (Section 7.6.). Still, ensuring good governance, 11 accountability (e.g. enhanced monitoring and verification capacity; Bos 2020), and the rule of law are 12 crucial for implementing forest-based mitigation options. In many countries with the highest 13 deforestation rates, insecure land rights often are significant barriers for forest-based mitigation options 14 (Gren and Zeleke 2016; Essl et al. 2018). 15 7.4.2.2. Afforestation, reforestation and forest ecosystem restoration 16 Activities, co-benefits, risks and implementation opportunities and barriers. Afforestation and 17 reforestation (A/R) are activities that convert land to forest, where reforestation is on land that has 18 previously contained forests, while afforestation is on land that historically has not been forested (Box 19 7.2). Forest restoration refers to a form of reforestation that gives more priority to ecological integrity 20 as well, even though it can still be a managed forest. Depending on the location, scale, and choice and 21 management of tree species, A/R activities have a wide variety of co-benefits and trade-offs. Well- 22 planned, sustainable reforestation and forest restoration can enhance climate resilience and biodiversity, 23 and provide a variety of ecosystem services including water regulation, microclimatic regulation, soil 24 erosion protection, as well as renewable resources, income and livelihoods (Ellison et al. 2017; Locatelli 25 et al. 2015; Verkerk et al. 2020; Stanturf et al. 2015). Afforestation, when well planned, can help address 26 land degradation and desertification by reducing runoff and erosion and lead to cloud formation 27 however, when not well planned, there are localised trade-offs such as reduced water yield or 28 biodiversity (Teuling et al. 2017; Ellison et al. 2017) . The use of non-native species and monocultures 29 may have adverse impacts on ecosystem structure and function, and water availability, particularly in 30 dry regions (Ellison et al. 2017). A/R activities may change the surface albedo and evapotranspiration 31 regimes, producing net cooling in the tropical and subtropical latitudes for local and global climate and 32 net warming at high latitudes (Section 7.4.2). Very large-scale implementation of A/R may negatively 33 affect food security since an increase in global forest area can increase food prices through land 34 competition (Kreidenweis et al. 2016). 35 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 36 potential, costs, and pathways. AR5 did not provide a new specification of A/R potential, but referred 37 to AR4 mostly for forestry measures (Nabuurs et al. 2007). AR5 did view the feasible A/R potential 38 from a diets change scenario that released land for reforestation and bioenergy crops. AR 5 provided 39 top-down estimates of costs and potentials for forestry mitigation options - including reduced 40 deforestation, forest management, afforestation, and agroforestry, estimated to contribute between 1.27 41 and 4.23 GtCO2 yr-1 of economically viable abatement in 2030 at carbon prices up to USD100/t CO2- 42 eq (Smith et al. 2014). 43 The SRCCL remained with a reported wide range of mitigation potential for A/R of 0.5–10.1 GtCO2 44 yr-1 by 2050 (medium confidence) (SRCCL Chapters 2 and 6; Roe et al. 2019; Fuss et al. 2018; Griscom 45 et al. 2017; Hawken 2017; Kreidenweis et al. 2016). The higher estimate represents a technical potential 46 of reforesting all areas where forests are the native cover type (reforestation), constrained by food Do Not Cite, Quote or Distribute 7-49 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 security and biodiversity considerations, considering above and below-ground carbon pools and 2 implementation on a rather theoretical maximum of 678 Mha of land (Roe et al. 2019; Griscom et al. 3 2017). The lower estimates represent the minimum range from an Earth System Model and a sustainable 4 global CDR potential (Fuss et al. 2018). Climate change will affect the mitigation potential of 5 reforestation due to impacts in forest growth and composition, as well as changes in disturbances 6 including fire. However, none of the mitigation estimates included in the SRCCL account for climate 7 impacts. 8 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Since SRCCL, 9 additional studies have been published on A/R mitigation potential by Bastin et al. (2019), Lewis et al. 10 (2019), (Doelman et al. 2019), (Favero et al. 2020) and (Austin et al. 2020). These studies are within 11 the range reported in the SRCCL stretching the potentials at the higher range. The rising public interest 12 in nature-based solutions, along with high profile initiatives being launched (UN Decade on Restoration 13 announced in 2019, the Bonn challenge on 150 million ha of restored forest in 2020 and e.g. the trillion- 14 tree campaign launched by the World Economic Forum in 2020), has prompted intense discussions on 15 the scale, effectiveness, and pitfalls of A/R and tree planting for climate mitigation (Anderegg et al. 16 2020; Bond et al. 2019; Heilmayr et al. 2020; Holl and Brancalion 2020; Luyssaert et al. 2018). The 17 sometimes sole attention on afforestation and reforestation suggesting it may solve the climate problem 18 to large extent in combination with the very high estimates of potentials have led to polarisation in the 19 debate, again resulting in a push back to nature restoration only (Lewis et al. 2019). Our assessment 20 based on most recent literature produced regional economic mitigation potential at USD100 tCO2-1 21 estimate of 100-400 MtCO2 yr-1 in Africa, 210-266 MtCO2 yr-1 in Asia and developing Pacific, 291 22 MtCO2-eq yr-1 in Developed countries (87% in North America), 30 MtCO2-eq yr-1 in Eastern Europe 23 and West-Central Asia, and 345-898 MtCO2-eq yr-1 in Latin America and Caribbean (Roe et al. 2021), 24 which totals to about 1200 MtCO2 yr-1, leaning to the lower range of the potentials in earlier IPCC 25 reports. A recent global assessment of the aggregate costs for afforestation and reforestation suggests 26 that at USD100 tCO2-1, 1.6 GtCO2 yr-1 could be sequestered globally for an annual cost of USD130 27 billion (Austin et al. 2020). Sectoral studies that are able to deal with local circumstances and limits 28 estimate A/R potentials at 20 MtCO2 yr-1 in Russia (Eastern Europe and West-Central Asia) 29 (Romanovskaya et al. 2020) and 64 MtCO2 yr-1 in Europe (Nabuurs et al. 2017). (Domke et al. 2020) 30 estimated for the USA an additional 20% sequestration rate from tree planting to achieve full stocking 31 capacity of all understocked productive forestland, in total reaching 187 MtCO 2 yr-1 sequestration. A 32 new study on costs in the USA estimates 72-91 MtCO2 yr-1 could be sequestered between now and 2050 33 for USD100/t CO2 (Wade et al. 2019). The tropical and subtropical latitudes are the most effective for 34 forest restoration in terms of carbon sequestration because of the rapid growth and lower albedo of the 35 land surface compared with high latitudes (Lewis et al. 2019).. Costs may be higher if albedo is 36 considered in North America, Russia, and Africa (Favero et al. 2017). In addition, a wide variety of 37 sequestration rates have been collected and published in e.g. IPCC Good Practice Guidance for the 38 AFOLU sector (IPCC 2006). 39 Critical assessment and conclusion. There is medium confidence that the global technical mitigation 40 potential of afforestation and reforestation activities by 2050 is 3.9 (0.5–10.1) GtCO2 yr-1, and the 41 economic mitigation potential (< USD100 tCO2-1) is 1.6 (0.5 – 3.0) GtCO2 yr-1 (requiring about 200 42 Mha). Per hectare a long (about 100 year) sustained effect of 5-10 t(CO2) ha-1 yr-1 is realistic with ranges 43 between 1-20 t(CO2) ha-1 yr-1. Not all sectoral studies rely on economic models that account for leakage 44 (Murray et al. 2004; Sohngen and Brown 2004), suggesting that technical potential may be 45 overestimated. 46 7.4.2.3. Improved forest management 47 Activities, co-benefits, risks and implementation opportunities and barriers. Do Not Cite, Quote or Distribute 7-50 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Improved sustainable forest management of already managed forests can lead to higher forest carbon 2 stocks, better quality of produced wood, continuously produce wood while maintaining and enhancing 3 the forest carbon stock, and can also partially prevent and counteract the impacts of disturbances (Kurz 4 et al. 2008; Marlon et al. 2012; Abatzoglou and Williams 2016; Tian et al. 2018; Seidl et al. 2017; 5 Nabuurs et al. 2017; Ekholm 2020). Furthermore it can provide benefits for climate change adaptation, 6 biodiversity conservation, microclimatic regulation, soil erosion protection and water and flood 7 regulation with reduced lateral C fluxes (Ashton et al. 2012; Verkerk et al. 2020; Martínez-Mena et al. 8 2019). Often, in existing (managed) forests with existing C stocks, large changes per hectare cannot be 9 expected, although many forest owners may respond to carbon price incentives (Favero et al. 2020; 10 Ekholm 2020). The full mitigation effects can be assessed in conjunction with the overall forest and 11 wood use system i.e., carbon stock changes in standing trees, soil, harvested wood products (HWPs) 12 and its bioenergy component with the avoided emissions through substitution. Forest management 13 strategies aimed at increasing the biomass stock may have adverse side effects, such as decreasing the 14 stand-level structural complexity, large emphasis on pure fast growing stands, risks for biodiversity and 15 resilience to natural disasters. 16 Generally measures can consist of one or combination of longer rotations, less intensive harvests, 17 continuous-cover forestry, mixed stands, more adapted species, selected provenances, high quality 18 wood assortments, etc. Further, there is a trade-off between management in various parts of the forest 19 product value chain, resulting in a wide range of results on the role of managed forests in mitigation 20 (Agostini et al. 2013; Braun et al. 2016; Gustavsson et al. 2017; Erb et al. 2017; Soimakallio et al. 2016; 21 Hurmekoski et al. 2020; Favero et al. 2020). Some studies conclude that reduction in forest carbon 22 stocks due to harvest exceeds for decades the joint sequestration of carbon in harvested wood product 23 stocks and emissions avoided through wood use (Soimakallio et al. 2016; Seppälä et al. 2019), whereas 24 others emphasise country level examples where investments in forest management have led to higher 25 growing stocks while producing more wood (Cowie et al. 2021; Schulze et al. 2020; Ouden et al. 2020). 26 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 27 potential, costs, and pathways. In the SRCCL, forest management activities have the potential to 28 mitigate 0.4–2.1 GtCO2-eq yr–1 by 2050 (medium confidence) (SRCCL: Griscom et al. 2017; Roe et al. 29 2019). The higher estimate stems from assumptions of applications on roughly 1.9 billion ha of already 30 managed forest which can be seen as very optimistic. It combines both natural forest management as 31 well as improved plantations, on average with a small net additional effect per hectare, not including 32 substitution effects in the energy sector nor the buildings sector. 33 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). The area of 34 forest under management plans has increased in all regions since 2000 by 233 Mha (FAO-FRA 2020). 35 The roughly 1 billion ha of secondary and degraded forests would be ideal to invest in and develop a 36 sustainable sector that pays attention to biodiversity, wood provision and climate mitigation at the same 37 time. This all depends on the effort made, the development of expertise, know-how in the field, nurseries 38 with adapted provenances, etc as was also found for Russian climate smart forestry options (Leskinen 39 et al. 2020). Regionally, recently updated economic mitigation potential at USD100 tCO2-1 have 179- 40 186 MtCO2-eq yr-1 in Africa, 193-313 MtCO2-eq yr-1 in Asia and developing Pacific, 215-220 MtCO2- 41 eq yr-1 in Developed countries , 82-152 MtCO2-eq yr-1 in Eastern Europe and West-Central Asia, and 42 62-204 MtCO2-eq yr–1 in Latin America and Caribbean (Roe et al. 2021). 43 Regional studies can take into account the local situation better: Russia Romanovskaya et al. (2020) 44 estimate the potential of forest fires management at 220–420 MtCO2 yr-1, gentle logging technology at 45 15–59, reduction of wood losses at 61–76 MtCO2 yr-1. In North America, (Austin et al. 2020) estimate 46 that in the next 30 years, forest management could contribute 154 MtCO2 yr-1 in the USA and Canada 47 with 81 MtCO2 yr-1 available at less than USD100 tCO2-1. In one production region (British Columbia) Do Not Cite, Quote or Distribute 7-51 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 a cost-effective portfolio of scenarios was simulated that directed more of the harvested wood to longer- 2 lived wood products, stopped burning of harvest residues and instead produced bioenergy to displace 3 fossil fuel burning, and reduced harvest levels in regions with low disturbance rates. Net GHG emissions 4 were reduced by an average of -9 MtCO2-eq yr-1 (Smyth et al. 2020). In Europe, climate smart forestry 5 could mitigate an additional 0.19 GtCO2 yr-1 by 2050 (Nabuurs et al. 2017), in line with the regional 6 estimates in (Roe et al. 2021). 7 In the tropics, estimates of the pantropical climate mitigation potential of natural forest management (a 8 light intensity management in secondary forests), across three tropical regions (Latin America, Africa, 9 Asia), is around 0.66 GtCO2-eq yr-1 with Asia responding for the largest share followed by Africa and 10 Latin America (Roe et al. 2021). Selective logging occurs in at least 20% of the world’s tropical forests 11 and causes at least half of the emissions from tropical forest degradation (Asner et al. 2005; Blaser and 12 Küchli 2011; Pearson et al. 2017). Reduced-impact logging for climate (RIL-C; promotion of reduced 13 wood waste, narrower haul roads, and lower impact skidding equipment) has the potential to reduce 14 logging emissions by 44% (Ellis et al. 2019), while also providing timber production. 15 Critical assessment and conclusion. There is medium confidence that the global technical mitigation 16 potential for improved forest management by 2050 is 1.7 (1–2.1) GtCO2 yr-1, and the economic 17 mitigation potential (< USD100 tCO2-1) is 1.1 (0.6–1.9) GtCO2 yr-1. Efforts to change forest 18 management do not only require e.g. a carbon price incentive, but especially require knowledge, 19 institutions, skilled labour, good access etc. These requirements outline that although the potential is of 20 medium size, we estimate a feasible potential towards the lower end. The net effect is also difficult to 21 assess, as management changes impact not only the forest biomass, but also the wood chain and 22 substitution effects. Further, leakage can arise from efforts to change management for carbon 23 sequestration. Efforts e.g. to set aside large areas of forest may be partly counteracted by higher 24 harvesting pressures elsewhere (Kallio and Solberg 2018). Studies such as (Austin et al. 2020) implicitly 25 account for leakage and thus suggest higher costs than other studies. We therefore judge the mitigation 26 potential at medium potential with medium agreement. 27 28 [START BOX 7.2 HERE] 29 Box 7.2 Climate Smart Forestry in Europe 30 Summary 31 European forests have been regarded as prospering and increasing for the last 5 decades. However, 32 these views also changed recently. Climate change is putting a large pressure on mono species and high 33 stocked areas of Norway spruce in Central Europe (Hlásny et al. 2021; Senf and Seidl 2021) with 34 estimates of mortality reaching 200 million m3, biodiversity under pressure, the Mediterranean area 35 showing a weak sector and harvesting pressure in the Baltics and north reaching maxima achievable. A 36 European strategy for unlocking the EU’s forests and forest sector potential was needed at the time of 37 developing the LULUCF regulation and was based on the concept of “Climate Smart Forestry” (CSF) 38 (Nabuurs et al. 2017; Verkerk et al. 2020). 39 Background 40 The idea behind CSF is that it considers the whole value chain from forest to wood products and energy, 41 illustrating that a wide range of measures can be applied to provide positive incentives for more firmly 42 integrating climate objectives into the forest and forest sector framework. CSF is more than just storing 43 carbon in forest ecosystems; it builds upon three main objectives; (i) reducing and/or removing GHG 44 emissions; (ii) adapting and building diverse forests for forest resilience to climate change; and (iii) Do Not Cite, Quote or Distribute 7-52 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 sustainably increasing forest productivity and incomes. These three CSF objectives can be achieved by 2 tailoring policy measures and actions to regional circumstances in Member States forest sectors. 3 Case description 4 The 2015 annual mitigation effect of EU-28 forests via contributions to the forest sink, material 5 substitution and energy substitution is estimated at 569 MtCO2 yr-1, or 13% of total current EU 6 emissions. With the right set of incentives in place at EU and Member States levels, it was found that 7 the EU-28 has the potential to achieve an additional combined mitigation impact through the 8 implementation of CSF of 441 MtCO2 yr-1 by 2050. Also, with the Green Deal and its Biodiversity and 9 Forest Strategy more emphasis will be placed on forests, forest management and the provision of 10 renewables. It is the diversity of measures (from strict reserves to more intensively managed systems 11 while adapting the resource) that will determine the success. Only with co-benefits in e.g. nature 12 conservation, soil protection, and provision of renewables, wood for buildings and income, the 13 mitigation and adaptation measures will be successful. 14 Interactions, limitations and lessons 15 Climate Smart Forestry is now taking shape across Europe with various research and implementation 16 projects (Climate Smart Forest and Nature Management, 2021). Pilots and projects are being 17 implemented by a variety of forest owners, some with more attention on biodiversity and adaptation, 18 some with more attention on production functions. They establish examples and in longer term the 19 outreach to the 16 million private owners in Europe. However, the right triggers and incentives are often 20 still lacking. E.g. adapting the spruce forest areas in Central Europe to climate change requires 21 knowledge about different species, biodiversity and different management options and eventually use 22 in industry. It requires alternative species to be available from the nurseries, as well improved 23 monitoring to assess the success and steer activities. 24 [END BOX 7.2 HERE] 25 26 7.4.2.4. Fire management (forest and grassland/savanna fires) 27 Activities, co-benefits, risks and implementation opportunities and barriers. Fire management 28 objectives include safeguarding life, property, and resources through the prevention, detection, control, 29 restriction, and management of fire for diverse purposes in natural ecosystems (SRCCL Chapter 6). 30 Controlled burning is an effective economic method of reducing fire danger and stimulating natural 31 regeneration. Co-benefits of fire management include reduced air pollution compared to much larger, 32 uncontrolled fires, prevention of soil erosion and land degradation, biodiversity conservation in 33 rangelands, and improvement of forage quality (Hurteau et al. 2019; Hurteau and Brooks 2011; Falk 34 2017). Fire management is still challenging because it is not only fire suppression at times of fire, but 35 especially proper natural resource management in between fire events. Furthermore, it is challenging 36 because of legal and policy issues, equity and rights concerns, governance, capacity, and research needs 37 (Russell-Smith et al. 2017; Goldammer 2016 ; Wiedinmyer and Hurteau 2010). It will increasingly be 38 needed under future enhanced climate change. 39 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 40 potential, costs, and pathways. In the SRCCL, fire management is among the nine options that can 41 deliver medium-to-large benefits across multiple land challenges (climate change mitigation, 42 adaptation, desertification, land degradation, and food security) (high confidence). Total emissions from 43 fires have been on the order of 8.1 GtCO2-eq yr-1 in terms of gross biomass loss for the period 1997– 44 2016 (SRCCL, Chapter 2 and Cross-Chapter Box 3 in Chapter 2). Reduction in fire CO2 emissions was Do Not Cite, Quote or Distribute 7-53 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 calculated to enhance land carbon sink by 0.48 GtCO2-eq yr-1 for the 1960–2009 period (Arora and 2 Melton 2018) (SRCCL, Table 6.16). 3 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). 4 Savannas. Savannas constitute one of the most fire-prone vegetation types on Earth and are a significant 5 source of GHG emissions. Savanna fires contributed 62% (4.92 PgCO2-eq yr-1) of gross global mean 6 fire emissions between 1997 and 2016. Regrowth from vegetation postfire sequesters the CO2 released 7 into the atmosphere, but not the CH4 and N2O emissions which contributed an approximate net of 2.1 8 PgCO2-eq yr−1 (Lipsett-Moore et al. 2018). Therefore, implementing prescribed burning with low 9 intensity fires, principally in the early dry season, to effectively manage the risk of wildfires occurring 10 in the late dry season is associated with reducing emissions (Whitehead et al. 2014). Considering this 11 fire management practice, estimates of global opportunities for emissions reductions were estimate at 12 69.1 MtCO2-eq yr−1 in Africa (29 countries, with 20 least developed African countries accounting for 13 74% of the mitigation potential), 13.3 MtCO2-eq yr−1 in South America (six countries), and 6.9 MtCO2- 14 eq yr-1 in Australia and Papua New Guinea (Lipsett-Moore et al. 2018). In Australia, savanna burning 15 emissions abatement methodologies have been available since 2012, and abatement has exceeded 4 16 MtCO2-eq mainly through the management of low intensity early dry season fire (Lynch et al. 2018). 17 Until August 2021, 78 were registered (Australian Government, Clean Energy Regulator, 2021). 18 Forests. Fire is also a prevalent forest disturbance (Scott et al. 2014; Falk et al. 2011; Andela et al. 19 2019). About 98 Mha of forest were affected by fire in 2015, affecting about 4% of the tropical (dry) 20 forests, 2% of the subtropical forests, and 1% of temperate and boreal forests (FAO 2020a). Between 21 2001-2018, remote sensing data showed that tree-covered areas correspond to about 29% of the total 22 area burned by wildfires, most in Africa. Prescribed fires are also applied routinely in forests worldwide 23 for fuel reduction and ecological reasons (Kalies and Yocom Kent 2016). Fire resilience is increasingly 24 managed in southwestern USA forest landscapes, which have experienced droughts and widespread, 25 high-severity wildfires (Keeley et al. 2019). In these forests, fire exclusion management, coupled with 26 a warming climate, has led to increasingly severe wildfires (Hurteau et al. 2014). However, the impacts 27 of prescribed fires in forests in reducing carbon emissions are still inconclusive. Some positive impacts 28 of prescribed fires are associated with other fuel reduction techniques (Loudermilk et al. 2017; Flanagan 29 et al. 2019; Stephens et al. 2020), leading to maintaining C stocks and reducing C emissions in the 30 future where extreme fire weather events are more frequent (Krofcheck et al. 2018, 2019; Hurteau et 31 al. 2019). (Bowman et al. 2020b,a; Goodwin et al. 2020; Hurteau et al. 2019). Land management 32 approaches will certainly need to consider the new climatic conditions (e.g., the proportion of days in 33 fire seasons with the potential for unmanageable fires more than doubling in some regions in northern 34 and eastern boreal forest) (Wotton et al. 2017). 35 Critical assessment and conclusion. There is low confidence that the global technical mitigation 36 potential for grassland and savanna fire management by 2050 is 0.1 (0.09–0.1) GtCO2 yr-1, and the 37 economic mitigation potential (< USD100 tCO2-1) is 0.05 (0.03–0.07) GtCO2 yr-1. Savanna fires produce 38 significant emissions globally, but prescribed fires in the early dry season could mitigate emissions in 39 different regions, particularly Africa. Evidence is less clear for fire management of forests, with the 40 contribution of GHG mitigation depending on many factors that affect the carbon balance (e.g., 41 Simmonds et al. 2021). Although prescribed burning is promoted to reduce uncontrolled wildfires in 42 forests, the benefits for the management of carbon stocks are unclear, with different studies reporting 43 varying results especially concerning its long term effectiveness. (Bowman et al. 2020b; Wotton et al. 44 2017). Under increasing climate change however, an increased attention on fire management will be 45 necessary. Do Not Cite, Quote or Distribute 7-54 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.4.2.5. Reduce degradation and conversion of grasslands and savannas 2 Activities, co-benefits, risks and implementation opportunities and barriers. Grasslands cover 3 approximately 40.5 % of the terrestrial area (i.e., 52.5 million km 2) divided as 13.8% woody savanna 4 and savanna; 12.7% open and closed shrub; 8.3 % non-woody grassland; and 5.7% is tundra (White et 5 al. 2000). Sub-Saharan Africa and Asia have the most extensive total area, 14.5 and 8.9 million km 2, 6 respectively. A review by Conant et al. (2017) reported based on data on grassland area (FAO 2013) 7 and grassland soil carbon stocks (Sombroek et al. 1993) a global estimate of about 343 Pg C (in the top 8 1 m), nearly 50% more than is stored in forests worldwide (FAO 2007). Reducing the conversion of 9 grasslands and savannas to croplands prevents soil carbon losses by oxidation, and to a smaller extent, 10 biomass carbon loss due to vegetation clearing (SRCCL, Chapter 6). Restoration of grasslands through 11 enhanced soil carbon sequestration, including a) management of vegetation, b) animal management, 12 and c) fire management, was also included in the SRCCL and is covered in Section 7.4.3.1. Similar to 13 other measures that reduce conversion, conserving carbon stocks in grasslands and savannas can be 14 achieved by controlling conversion drivers (e.g., commercial and subsistence agriculture, see Section 15 7.3) and improving policies and management. In addition to mitigation, conserving grasslands provide 16 various socio-economic, biodiversity, water cycle and other environmental benefits (Claassen et al. 17 2010; Ryals et al. 2015; Bengtsson et al. 2019). Annual operating costs, and opportunity costs of income 18 foregone by undertaking the activities needed for avoiding conversion of grasslands making costs one 19 of the key barriers for implementation (Lipper et al. 2010). 20 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 21 potential, costs, and pathways. The SRCCL reported a mitigation potential for reduced conversion of 22 grasslands and savannas of 0.03–0.12 GtCO2-eq yr-1 (SRCCL: Griscom et al. 2017) considering the 23 higher loss of soil organic carbon in croplands (Sanderman et al. 2017). Assuming an average starting 24 soil organic carbon stock of temperate grasslands (Poeplau et al. 2011), and the mean annual global 25 cropland conversion rates (1961–2003) (Krause et al. 2017), the equivalent loss of soil organic carbon 26 over 20 years would be 14 GtCO2-eq, i.e. 0.7 GtCO2 yr-1 (SRCCL, Chapter 6). IPCC AR5 and AR4 did 27 not explicitly consider the mitigation potential of avoided conversion of grasslands-savannas but the 28 management of grazing land is accounted for considering plant, animal, and fire management with a 29 mean mitigation potential of 0.11-0.80 tCO2-eq ha-1 yr-1 depending on the climate region. This resulted 30 in 0.25 GtCO2-eq yr-1 at USD20 tCO2-1 to 1.25 GtCO2-eq yr-1 at USD100 tCO2-1 by 2030. 31 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Unlike most of 32 the measures covered in Section 7.4, there are currently no global, spatially explicit mitigation potential 33 estimates for reduced grassland conversion to generate technical and economic potentials by region. 34 Literature developments since AR5 and SRCCL are studies that provide mitigation estimates in one or 35 a few countries or regions. Modelling experiments comparing Californian forests and grasslands found 36 that grasslands resulted in a more resilient C sink than forests to future climate change (Dass et al. 37 2018). However, previous studies indicated that precipitation is a key controller of the carbon storage 38 in these grasslands, with the grassland became a carbon sink in 2005, when the region received 39 relatively high spring precipitation (Ma et al. 2007). In North America, grassland conversion was the 40 source for 77% of all new croplands from 2008-2012 (Lark et al. 2015). Avoided conversion of North 41 American grasslands to croplands presents an economic mitigation potential of 0.024 GtCO2-eq yr-1 and 42 technical potential of 0.107 GtCO2-eq yr-1 (Fargione et al. 2018). This potential is related mainly to root 43 biomass and soils (81% of emissions from soils). Estimates of GHG emissions from any future 44 deforestation in Australian savannas also point to the potential mitigation of around 0.024 GtCO2-eq 45 yr-1 (Bristow et al. 2016). The expansion of the Soy Moratorium (SoyM) from the Brazilian Amazon 46 to the Cerrado (Brazilian savannas) would prevent the direct conversion of 3.6 Mha of native vegetation 47 to soybeans by 2050 and avoid the emission of 0.02 GtCO2-eq yr-1 (Soterroni et al. 2019). Do Not Cite, Quote or Distribute 7-55 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Critical assessment and conclusion. There is low confidence that the global technical mitigation 2 potential for reduced grassland and savanna conversion by 2050 is 0.2 (0.1–0.4) GtCO2 yr-1, and the 3 economic mitigation potential (< USD100 tCO2-1) is 0.04 GtCO2 yr-1. Most of the carbon sequestration 4 potential is in belowground biomass and soil organic matter. However, estimates of potential are still 5 based on few studies and vary according the levels of soil carbon, and ecosystem productivity (e.g. in 6 response to rainfall distribution). Conservation of grasslands presents significant benefits for 7 desertification control, especially in arid areas (SRCCL, Chapter 3). Policies supporting avoided 8 conversion can help protect at-risk grasslands, reduce GHG emissions, and produce positive outcomes 9 for biodiversity and landowners (Ahlering et al. 2016). In comparison to tropical rainforest regions that 10 have been the primary target for mitigation policies associated to natural ecosystems (e.g. REDD+), 11 Conversion grasslands and savannas has received less national and international attention, despite 12 growing evidence of concentrated cropland expansion into these areas with impacts of carbon losses. 13 7.4.2.6. Reduce degradation and conversion of peatlands 14 Activities, co-benefits, risks and implementation barriers. Peatlands are carbon-rich wetland 15 ecosystems with organic soil horizons in which soil organic matter concentration exceeds 30% (dry 16 weight) and soil carbon concentrations can exceed 50% (Page and Baird 2016, Boone Kauffman et al. 17 2017). Reducing the conversion of peatlands avoids emissions of above- and below-ground biomass 18 and soil carbon due to vegetation clearing, fires, and peat decomposition from drainage. Similar to 19 deforestation, peatland carbon stocks can be conserved by controlling the drivers of conversion and 20 degradation (e.g. commercial and subsistence agriculture, mining, urban expansion) and improving 21 governance and management. Reducing conversion is urgent because peatland carbon stocks 22 accumulate slowly and persist over millennia; loss of existing stocks cannot be easily reversed over the 23 decadal timescales needed to meet the Paris Agreement (Goldstein et al. 2020). The main co-benefits 24 of reducing conversion of peatlands include conservation of a unique biodiversity including many 25 critically endangered species, provision of water quality and regulation, and improved public health 26 through decreased fire-caused pollutants (Griscom et al. 2017). Although reducing peatland conversion 27 will reduce land availability for alternative uses including agriculture or other land-based mitigation, 28 drained peatlands constitute a small share of agricultural land globally while contributing significant 29 emissions (Joosten 2009). Mitigation through reduced conversion of peatlands therefore has a high 30 potential of avoided emissions per hectare (Roe et al. 2019). 31 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 32 potential, costs, and pathways. In the SRCCL (Chapters 2 and 6), it was estimated that avoided peat 33 impacts could deliver 0.45–1.22 GtCO2-eq yr-1 technical potential by 2030-2050 (medium confidence) 34 (Griscom et al. 2017; Hawken 2017; Hooijer et al. 2010). The mitigation potential estimates cover 35 tropical peatlands and include CO2, N2O and CH4 emissions. The mitigation potential is derived from 36 quantification of losses of carbon stocks due to land conversion, shifts in GHG fluxes, alterations in net 37 ecosystem productivity, input factors such as fertilisation needs, and biophysical climate impacts (e.g., 38 shifts in albedo, water cycles, etc). Tropical peatlands account for only ~10% of peatland area and about 39 20% of peatland carbon stock but about 80% of peatland carbon emissions, primarily from peatland 40 conversion in Indonesia (about 60%) and Malaysia (about 10%) (Page et al. 2011; Leifeld and 41 Menichetti 2018; Hooijer et al. 2010). While the total mitigation potential of peatland conservation is 42 considered moderate, the per hectare mitigation potential is the highest among land-based mitigation 43 measures (Roe et al. 2019). 44 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Recent studies 45 continue to report high carbon stocks in peatlands and emphasize the vulnerability of peatland carbon 46 after conversion. The carbon stocks of tropical peatlands are among the highest of any forest, 1,211- 47 4,257 tCO2-eq ha-1 in the Peruvian Amazon (Bhomia et al. 2019) and 1,956-14,757 tCO2-eq ha-1 in Do Not Cite, Quote or Distribute 7-56 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Indonesia (Novita et al. 2021). Ninety percent of tropical peatland carbon stocks are vulnerable to 2 emission during conversion and may not be recoverable through restoration; in contrast, boreal and 3 temperate peatlands hold similar carbon stocks (1,439-5,619 tCO2-eq ha-1) but only 30% of northern 4 carbon stocks are vulnerable to emission during conversion and irrecoverable through restoration 5 (Goldstein et al. 2020). A recent study shows global mitigation potential of about 0.2 GtCO2-eq yr-1 at 6 costs up to USD100 tCO2-1 (Roe et al. 2021). Another study estimated that 72% of mitigation is achieved 7 through avoided soil carbon impacts, with the remainder through avoided impacts to vegetation (Bossio 8 et al. 2020). Recent model projections show that both peatland protection and peatland restoration 9 (Section 7.4.2.7) are needed to achieve a 2ºC mitigation pathway and that peatland protection and 10 restoration policies will have minimal impacts on regional food security (Leifeld et al. 2019, 11 Humpenöder et al. 2020). Global studies have not accounted for extensive peatlands recently reported 12 in the Congo Basin, estimated to cover 145,500 km2 and contain 30.6 Pg C, as much as 29% of total 13 tropical peat carbon stock (Dargie et al. 2017). These Congo peatlands are relatively intact; continued 14 preservation is needed to prevent major emissions (Dargie et al. 2019). In northern peatlands that are 15 underlain by permafrost (roughly 50% of the total peatlands north of 23º latitude, (Hugelius et al. 2020), 16 climate change (i.e. warming) is the major driver of peatland degradation (e.g. through permafrost thaw) 17 (Schuur et al. 2015, Goldstein et al. 2020). However, in non-permafrost boreal and temperate peatlands, 18 reduction of peatland conversion is also a cost-effective mitigation strategy. Peatlands are sensitive to 19 climate change and there is low confidence about the future peatland sink globally (SRCCL, Chapter 20 2). Permafrost thaw may shift northern peatlands from a net carbon sink to net source (Hugelius et al. 21 2020). Uncertainties in peatland extent and the magnitude of existing carbon stocks, in both northern 22 (Loisel et al. 2014) and tropical (Dargie et al. 2017) latitudes limit understanding of current and future 23 peatland carbon dynamics (Minasny et al. 2019). 24 Critical assessment and conclusion. Based on studies to date, there is medium confidence that peatland 25 conservation has a technical potential of 0.86 (0.43–2.02) GtCO2-eq yr-1 of which 0.48 (0.2–0.68) 26 GtCO2-eq yr-1 is available at USD100 tCO2-1 (Figure 7.11). High per hectare mitigation potential and 27 high rate of co-benefits particularly in tropical countries, support the effectiveness of this mitigation 28 strategy (Roe et al. 2019). Feasibility of reducing peatland conversion may depend on countries’ 29 governance, financial capacity and political will. 30 7.4.2.7. Peatland restoration 31 Activities, co-benefits, risks and implementation barriers. Peatland restoration involves restoring 32 degraded and damaged peatlands, for example through rewetting and revegetation, which both increases 33 carbon accumulation in vegetation and soils and avoids ongoing CO2 emissions. Peatlands only account 34 for about 3% of the terrestrial surface, predominantly occurring in boreal ecosystems (78%), with a 35 smaller proportion in tropical regions (13%), but may store about 600 Gt Carbon or 21% of the global 36 total soil organic Carbon stock of about 3000 Gt (Leifeld and Menichetti 2018; Page et al. 2011). 37 Peatland restoration delivers co-benefits for biodiversity, as well as regulating water flow and 38 preventing downstream flooding, while still allowing for extensive management such as paludiculture 39 (Tan et al. 2021). Rewetting of peatlands also reduces the risk of fire, but may also mobilize salts and 40 contaminants in soils (van Diggelen et al. 2020) and in severely degraded peatlands, restoration of 41 peatland hydrology and vegetation may not be feasible (Andersen et al. 2017). At a local level, 42 restoration of peatlands drained for agriculture could displace food production and damage local food 43 supply, although impacts to regional and global food security would be minimal (Humpenöder et al. 44 2020). Collaborative and transparent planning processes are needed to reduce conflict between 45 competing land uses (Tanneberger et al. 2020b). Adequate resources for implementing restoration 46 policies are key to engage local communities and maintain livelihoods (Ward et al. 2021; Resosudarmo 47 et al. 2019). Do Not Cite, Quote or Distribute 7-57 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 2 potential, costs, and pathways. Large areas (0.51 Mkm2) of global peatlands are degraded of which 3 0.2 Mkm2 are tropical peatlands (Griscom et al. 2017; Leifeld and Menichetti 2018). According the 4 SRCCL, peatland restoration could deliver technical mitigation potentials of 0.15 - 0.81GtCO2-eq yr-1 5 by 2030-2050 (low confidence) (Chapter 2 and 6 of the SRCCL; (Couwenberg et al. 2010; Griscom et 6 al. 2017), though there could be an increase in methane emissions after restoration (Jauhiainen et al. 7 2008). The mitigation potential estimates cover global peatlands and include CO2, N2O and CH4 8 emissions. Peatlands are highly sensitive to climate change (high confidence), however there are 9 currently no studies that estimate future climate effects on mitigation potential from peatland 10 restoration. 11 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). The most recent 12 literature and reviews indicate with high confidence that restoration would decrease CO2 emissions and 13 with medium confidence that restoration would decrease net GHG emissions from degraded peatlands 14 (Wilson et al. 2016; Ojanen and Minkkinen 2020; van Diggelen et al. 2020). Although rewetting of 15 drained peatlands increases CH4 emissions, this effect is often outweighed by decreases in CO2 and N2O 16 emissions but depends very much on local circumstances (Günther et al. 2020). Restoration and 17 rewetting of almost all drained peatlands is needed by 2050 to meet 1.5-2ºC pathways which is unlikely 18 to happen (Leifeld et al. 2019); immediate rewetting and restoration minimises the warming from 19 cumulative CO2 emissions (Nugent et al. 2019). 20 According to recent data, the technical mitigation potential for global peatland restoration is estimated 21 at 0.5-1.3 GtCO2-eq yr-1 (Leifeld and Menichetti 2018; Griscom et al. 2020; Bossio et al. 2020; Roe et 22 al. 2021; Figure 7.11), with 80% of the mitigation potential derived from improvements to soil carbon 23 (Bossio et al. 2020). The regional mitigation potentials of all peatlands outlined in Roe et al. (2021) 24 reflect the country-level estimates from (Humpenöder et al. 2020). 25 Climate mitigation effects of peatland rewetting depend on the climate zone and land use. Recent 26 analysis shows the strongest mitigation gains from rewetting drained temperate and boreal peatlands 27 used for agriculture and drained tropical peatlands (Ojanen and Minkkinen 2020). However, estimates 28 of emission factors from rewetting drained tropical peatlands remain uncertain (Wilson et al. 2016; 29 Murdiyarso et al. 2019). Topsoil removal, in combination with rewetting, may improve restoration 30 success and limit CH4 emissions during restoration of highly degraded temperate peatlands (Zak et al. 31 2018). In temperate and boreal regions, co-benefits mentioned above are major motivations for peatland 32 restoration (Chimner et al. 2017; Tanneberger et al. 2020a). 33 Critical assessment and conclusion. Based on studies to date, there is medium confidence that peatland 34 restoration has a technical potential of 0.79 (0.49–1.3) GtCO2-eq yr-1 (median) of which 0.4 (0.2–0.6) 35 GtCO2-eq yr-1 is available up to USD100 tCO2-1. The large land area of degraded peatlands suggests 36 that significant emissions reductions could occur through large-scale restoration especially in tropical 37 peatlands. There is medium confidence in the large carbon stocks of tropical peat forests (1,956-14,757 38 tCO2-eq ha-1) and large rates of carbon loss associated with land cover change (640-1,650 tCO2-eq ha- 1 39 ) (Novita et al. 2021; Goldstein et al. 2020). However, large-scale implementation of tropical peatland 40 restoration will likely be limited by costs and other demands for these tropical lands. 41 7.4.2.8. Reduce conversion of coastal wetlands 42 Activities, co-benefits, risks and implementation barriers. Reducing conversion of coastal wetlands, 43 including mangroves, marshes and seagrass ecosystems, avoids emissions from above and below 44 ground biomass and soil carbon through avoided degradation and/or loss. Coastal wetlands occur 45 mainly in estuaries and deltas, areas that are often densely settled, with livelihoods closely linked to 46 coastal ecosystems and resources (Moser et al. 2012). The carbon stocks of these highly productive Do Not Cite, Quote or Distribute 7-58 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 ecosystems are sometimes referred to as “blue carbon”. Loss of existing stocks cannot be easily reversed 2 over decadal timescales (Goldstein et al. 2020). The main drivers of conversion include intensive 3 aquaculture, agriculture, salt ponds, urbanisation and infrastructure development, the extensive use of 4 fertilisers, and extraction of water resources (Lovelock et al. 2018). Reduced conversion of coastal 5 wetlands has many co-benefits, including biodiversity conservation, fisheries production, soil 6 stabilisation, water flow and water quality regulation, flooding and storm surge prevention, and 7 increased resilience to cyclones (UNEP 2020; Windham-Myers et al. 2018a). Risks associated with the 8 mitigation potential of coastal wetland conservation include uncertain permanence under future climate 9 scenarios, including the effects of coastal squeeze, where coastal wetland area may be lost if upland 10 area is not available for migration as sea levels rise (IPCC WGII Ch. 3.4.2.5; (Lovelock and Reef 2020). 11 Preservation of coastal wetlands also conflicts with other land use in the coastal zone, including 12 aquaculture, agriculture, and human development; economic incentives are needed to prioritise wetland 13 preservation over more profitable short-term land use. Integration of policies and efforts aimed at 14 coastal climate mitigation, adaptation, biodiversity conservation, and fisheries, for example through 15 Integrated Coastal Zone Management and Marine Spatial Planning, will bundle climate mitigation with 16 co-benefits and optimise outcomes (Herr et al. 2017). 17 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 18 potential, costs, and pathways. Coastal wetlands contain high, yet variable, organic carbon stocks, 19 leading to a range of estimates of the global mitigation potential of reduced conversion. The SRCCL 20 (Chapter 2) and SROCCC (Chapter 5), report a technical mitigation potential of 0.15–5.35 GtCO2-eq 21 yr-1 by 2050 (Lovelock et al. 2017; Pendleton et al. 2012; Howard et al. 2017; Griscom et al. 2017). The 22 mitigation potential is derived from quantification of losses of carbon stocks in vegetation and soil due 23 to land conversion, shifts in GHG fluxes associated with land use, and alterations in net ecosystem 24 productivity. The wide range in estimates mostly relate to the scope (all coastal ecosystems vs. 25 mangroves only) and different assumptions on decomposition rates. Loss rates of coastal wetlands have 26 been estimated at 0.2-3% yr-1, depending on the vegetation type and location (Atwood et al. 2017; 27 Howard et al. 2017). 28 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Global technical 29 mitigation potential for conservation of coastal wetlands from recent literature have focused on 30 protection of mangroves; estimates range from 0.06–2.25 GtCO2-eq yr-1 (Griscom et al. 2020; Bossio 31 et al. 2020) with 80% of the mitigation potential derived from improvements to soil carbon (Bossio et 32 al. 2020). Regional potentials (Roe et al. 2021) reflect mangrove protection; marsh and seagrass 33 protection were not included due to lack of country-level data on marsh and seagrass distribution and 34 conversion. 35 Global estimates show mangroves have the largest per hectare carbon stocks (see IPCC WGII AR6 36 Box 3.4 for estimates of carbon stocks, burial rates and ecosystem extent for coastal wetland 37 ecosystems). Mean ecosystem carbon stock in mangroves is 3131 tCO2-eq ha-1 among the largest carbon 38 stocks on Earth. Recent studies emphasize the variability in total ecosystem carbon stocks for each 39 wetland type, based on species and climatic and edaphic conditions (Kauffman et al. 2020; Bedulli et 40 al. 2020; Ricart et al. 2020; Wang et al. 2021; Alongi et al. 2020), and highlight the vulnerability of soil 41 carbon below 1 m depth (Arifanti et al. 2019). Sea level strongly influences coastal wetland distribution, 42 productivity, and sediment accretion; therefore, sea level rise will impact carbon accumulation and 43 persistence of existing carbon stocks (Macreadie et al. 2019, IPCC WGII AR6 Box 3.4). 44 Recent loss rates of mangroves are 0.16-0.39% y-1 and are highest in Southeast Asia (Friess et al. 2019; 45 Hamilton and Casey 2016). Assuming loss of soil C to 1 m depth after deforestation, avoiding mangrove 46 conversion has the technical potential to mitigate approximately 23.5-38.7 MtCO2-eq y-1 (Ouyang and 47 Lee 2020); note, this potential is additional to reduced conversion of forests (Griscom et al. 2020, Do Not Cite, Quote or Distribute 7-59 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.4.2.1). Regional estimates show that about 85% of mitigation potential for avoided mangrove 2 conversion is in Southeast Asia and Developing Pacific (32 MtCO2-eq yr-1 at USD100 tCO2-1), 10% is 3 in Latin American and the Caribbean (4 MtCO2-eq y-1), and approximately 5% in other regions 4 (Griscom et al. 2020; Roe et al. 2021). 5 Key uncertainties remain in mapping extent and conversion rates for salt marshes and seagrasses 6 (McKenzie et al. 2020). Seagrass loss rates were estimated at 1-2% y-1 (Dunic et al. 2021) with 7 stabilization in some regions (IPCC WGII Ch. 3.4.2.5; (de los Santos et al. 2019); however, loss occurs 8 non-linearly and depends on site-specific context. Tidal marsh extent and conversion rates remains 9 poorly estimated, outside of the USA, Europe, South Africa, and Australia (Mcowen et al. 2017; 10 Macreadie et al. 2019). 11 Critical assessment and conclusion. There is medium confidence that coastal wetland protection has a 12 technical potential of 0.8 (0.06–5.4) GtCO2-eq yr-1 of which 0.17 (0.06–0.27) GtCO2-eq yr-1 is available 13 up to USD100 tCO2-1. There is a high certainty (robust evidence, high agreement) that coastal 14 ecosystems have among the largest carbon stocks of any ecosystem. As these ecosystems provide many 15 important services, reduced conversion of coastal wetlands is a valuable mitigation strategy with 16 numerous co-benefits. However, the vulnerability of coastal wetlands to climatic and other 17 anthropogenic stressors may limit the permanence of climate mitigation. 18 7.4.2.9. Coastal wetland restoration 19 Activities, co-benefits, risks and implementation barriers. Coastal wetland restoration involves 20 restoring degraded or damaged coastal wetlands including mangroves, salt marshes, and seagrass 21 ecosystems, leading to sequestration of ‘blue carbon’ in wetland vegetation and soil (SRCCL Ch 6, 22 SROCCC Ch 5). Successful approaches to wetland restoration include: (1) passive restoration, the 23 removal of anthropogenic activities that are causing degradation or preventing recovery; and (2) active 24 restoration, purposeful manipulations to the environment in order to achieve recovery to a naturally 25 functioning system (Elliott et al. 2016; IPCC WGII Ch 3). Restoration of coastal wetlands delivers 26 many valuable co-benefits, including enhanced water quality, biodiversity, aesthetic values, fisheries 27 production (food security), and protection from rising sea levels and storm impacts (Barbier et al. 2011; 28 Hochard et al. 2019; Sun and Carson 2020; Duarte et al. 2020). Of the 0.3 Mkm2 coastal wetlands 29 globally, 0.11 Mkm2 of mangroves are considered feasible for restoration (Griscom et al. 2017). Risks 30 associated with coastal wetland restoration include uncertain permanence under future climate scenarios 31 (IPCC WGII AR6 Box 3.4), partial offsets of mitigation through enhanced methane and nitrous oxide 32 release and carbonate formation, and competition with other land uses, including aquaculture and 33 human settlement and development in the coastal zone (SROCCC, Chapter 5). To date, many coastal 34 wetland restoration efforts do not succeed due to failure to address the drivers of degradation (van 35 Katwijk et al. 2016). However, improved frameworks for implementing and assessing coastal wetland 36 restoration are emerging that emphasize the recovery of ecosystem functions (Cadier et al. 2020; Zhao 37 et al. 2016). Restoration projects that involve local communities at all stages and consider both 38 biophysical and socio-political context are more likely to succeed (Brown et al. 2014; Wylie et al. 2016). 39 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 40 potential, costs, and pathways. The SRCCL reported that mangrove restoration has the technical 41 potential to mitigate 0.07 GtCO2 yr-1 through rewetting (Crooks et al. 2011) and take up 0.02–0.84 42 GtCO2 yr-1 from vegetation biomass and soil enhancement through 2030 (medium confidence) (Griscom 43 et al. 2017). The SROCCC concluded that cost-effective coastal blue carbon restoration had a potential 44 of ~0.15-0.18 GtCO2-eq yr-1, a low global potential compared to other ocean-based solutions but with 45 extensive co-benefits and limited adverse side effects (Gattuso et al. 2018). Do Not Cite, Quote or Distribute 7-60 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Recent studies 2 emphasise the timeframe needed to achieve the full mitigation potential (Duarte et al. 2020; Taillardat 3 et al. 2020). The first project-derived estimate of the net GHG benefit from seagrass restoration found 4 1.54 tCO2-eq (0.42 MgC) ha-1 yr-1 10 years after restoration began (Oreska et al. 2020); comparable to 5 the default emission factor in the Wetlands Supplement (IPCC 2014). Recent studies of rehabilitated 6 mangroves also indicate that annual carbon sequestration rates in biomass and soils can return to natural 7 levels within decades of restoration (Cameron et al. 2019; Sidik et al. 2019). A meta-analysis shows 8 increasing carbon sequestration rates over the first 15 years of mangrove restoration with rates 9 stabilising at 25.7 ± 7.7 tCO2-eq (7.0 ± 2.1 MgC) ha-1 yr-1 through forty years, although success depends 10 on climate, sediment type, and restoration methods (Sasmito et al. 2019). Overall, 30% of mangrove 11 soil carbon stocks and 50-70% of marsh and seagrass carbon stocks are unlikely to recover within 30 12 years of restoration, underscoring the importance of preventing conversion of coastal wetlands (7.4.2.8) 13 (Goldstein et al. 2020). 14 According to recent data, the technical mitigation potential for global coastal wetland restoration is 15 0.04-0.84 GtCO2-eq yr-1 (Griscom et al. 2020; Bossio et al. 2020; Roe et al. 2021) with 60% of the 16 mitigation potential derived from improvements to soil carbon (Bossio et al. 2020). Regional potentials 17 based on country-level estimates from Griscom et al. (2020) show the technical and economic (up to 18 USD100 tCO2-1) potential of mangrove restoration; seagrass and marsh restoration was not included 19 due to lack of country-level data on distribution and conversion (but see (McKenzie et al. (2020) for 20 updates on global seagrass distribution). Although global potential is relatively moderate, mitigation 21 can be quite significant for countries with extensive coastlines (e.g., Indonesia, Brazil) and for small 22 island states where coastal wetlands have been shown to comprise 24-34% of their total national carbon 23 stock (Donato et al. 2012). Furthermore, non-climatic co-benefits can strongly motivate coastal wetland 24 restoration worldwide (UNEP 2021a). Major successes in both active and passive restoration of 25 seagrasses have been documented in North America and Europe (Lefcheck et al. 2018; Orth et al. 2020; 26 de los Santos et al. 2019); passive restoration may also be feasible for mangroves (Cameron et al. 2019). 27 There is high site-specific variation in carbon sequestration rates and uncertainties regarding the 28 response to future climate change (Jennerjahn et al. 2017; Nowicki et al. 2017; IPCC WGII AR6 Box 29 3.4). Changes in distributions (Kelleway et al. 2017; Wilson and Lotze 2019), methane release (Al-Haj 30 and Fulweiler 2020), carbonate formation (Saderne et al. 2019), and ecosystem responses to interactive 31 climate stressors are not well-understood (Short et al. 2016; Fitzgerald and Hughes 2019; Lovelock and 32 Reef 2020). 33 Critical assessment and conclusion. There is medium confidence that coastal wetland restoration has a 34 technical potential of 0.3 (0.04–0.84) GtCO2-eq yr-1 of which 0.1 (0.05–0.2) GtCO2-eq yr-1 is available 35 up to USD100 tCO2-1. There is high confidence that coastal wetlands, especially mangroves, contain 36 large carbon stocks relative to other ecosystems and medium confidence that restoration will reinstate 37 pre-disturbance carbon sequestration rates. There is low confidence on the response of coastal wetlands 38 to climate change; however, there is high confidence that coastal wetland restoration will provide a suite 39 of valuable co-benefits. 40 7.4.3. Agriculture 41 7.4.3.1. Soil carbon management in croplands and grasslands 42 Activities, co-benefits, risks and implementation opportunities and barriers. Increasing soil organic 43 matter in croplands are agricultural management practices that include (1) crop management: for 44 example, high input carbon practices such as improved crop varieties, crop rotation, use of cover crops, 45 perennial cropping systems (including agroforestry see Section 7.4.3.3), integrated production systems, Do Not Cite, Quote or Distribute 7-61 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 crop diversification, agricultural biotechnology, (2) nutrient management including fertilization with 2 organic amendments / green manures (Section 7.4.3.6), (3) reduced tillage intensity and residue 3 retention, (4) improved water management: including drainage of waterlogged mineral soils and 4 irrigation of crops in arid / semi-arid conditions, (5) improved rice management (Section 7.4.3.5) and 5 (6) biochar application (Section 7.4.3.2) (Smith et al. 2019d). For increased soil organic matter in 6 grasslands, practices include (1) management of vegetation: including improved grass varieties/sward 7 composition, deep rooting grasses, increased productivity, and nutrient management, (2) livestock 8 management: including appropriate stocking densities fit to carrying capacity, fodder banks, and fodder 9 diversification, and (3) fire management: improved use of fire for sustainable grassland management, 10 including fire prevention and improved prescribed burning (Smith et al. 2014, 2019d). All these 11 measures are recognized as Sustainable Soil Management Practices by FAO (Baritz et al. 2018). Whilst 12 there are co-benefits for livelihoods, biodiversity, water provision and food security Smith et al. 2019b 13 , and impacts on leakage, indirect land-use change and foregone sequestration do not apply (since 14 production in not displaced), the climate benefits of soil carbon sequestration in croplands can be 15 negated if achieved through additional fertiliser inputs (potentially causing increased N 2O emissions; 16 (Guenet et al. 2021), and both saturation and permanence are relevant concerns. When considering 17 implementation barriers, soil carbon management in croplands and grasslands is a low-cost option at a 18 high level of technology readiness (it is already widely deployed globally) with low socio-cultural and 19 institutional barriers, but with difficulty in monitoring and verification proving a barrier to 20 implementation (Smith et al. 2020a). 21 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 22 potential, costs, and pathways. Building on AR5, the SRCCL reported the global mitigation potential 23 for soil carbon management in croplands to be 1.4–2.3 GtCO2-eq yr-1 (Smith et al. 2014), though the 24 full literature range was 0.3-6.8 GtCO2-eq yr-1 (Frank et al. 2017; Sommer and Bossio 2014; Conant et 25 al. 2017; Dickie et al. 2014b; Fuss et al. 2018; Griscom et al. 2017; Hawken 2017; Henderson et al. 26 2015; Herrero et al. 2016; Paustian et al. 2016; Powlson et al. 2014; Sanderman et al. 2017; Zomer et 27 al. 2016; Roe et al. 2019). The global mitigation potential for soil organic carbon management in 28 grasslands was assessed to be 1.4–1.8 GtCO2-eq yr-1, with the full literature range being 0.1-2.6 GtCO2- 29 eq yr-1 (Conant et al. 2017; Herrero et al. 2016;2013; Roe et al. 2019). Lower values in the range 30 represented economic potentials, whilst higher values represented technical potentials – and uncertainty 31 was expressed by reporting the whole range of estimates. The SR1.5 outlined associated costs reported 32 in literature to range from USD -45 to 100 tCO2-1, describing enhanced soil carbon sequestration as a 33 cost-effective measure (IPCC 2018). Despite significant mitigation potential, there is limited inclusion 34 of soil carbon sequestration as a response option within IAM mitigation pathways (Rogelj et al. 2018a). 35 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). No recent 36 literature has been published which conflict with the mitigation potentials reported in the SRCCL. 37 Relevant papers include Lal et al. (2018) which estimated soil carbon sequestration potential to be 0.7- 38 4.1 GtCO2-eq yr-1 for croplands and 1.1-2.9 GtCO2-eq yr-1 for grasslands. Bossio et al. (2020) assessed 39 the contribution of soil carbon sequestration to natural climate solutions and found the potential to be 40 5.5 GtCO2 yr-1 across all ecosystems, with only small portions of this (0.41 GtCO2-eq yr-1 for cover 41 cropping in croplands; 0.23, 0.15, 0.15 GtCO2-eq yr-1 for avoided grassland conversion, optimal grazing 42 intensity and legumes in pastures, respectively) arising from croplands and grasslands. Regionally, soil 43 carbon management in croplands is feasible anywhere, but effectiveness can be limited in very dry 44 regions (Sanderman et al. 2017). For soil carbon management in grasslands the feasibility is greatest in 45 areas where grasslands have been degraded (e.g. by overgrazing) and soil organic carbon is depleted. 46 For well managed grasslands, soil carbon stocks are already high and the potential for additional carbon 47 storage is low (Roe et al. (2021) estimete greatest economic (up to USD100 tCO2-1) potential between Do Not Cite, Quote or Distribute 7-62 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2020 and 2050 for croplands to be in Asia and the developing Pacific (339.7 MtCO2 yr-1) and for 2 grasslands, in Developed Countries (253.6 MtCO2 yr-1). 3 Critical assessment and conclusion. In conclusion, there is medium confidence that enhanced soil 4 carbon management in croplands has a global technical mitigation potential of 1.9 (0.4-6.8) GtCO2 yr- 1 5 , and in grasslands of 1.0 (0.2-2.6) GtCO2 yr-1, of which, 0.6 (04-0.9) and 0.9 (0.3-1.6) GtCO2 yr–1 is 6 estimated to be available at up to USD100 tCO2-1 respectively. Regionally, soil carbon management in 7 croplands and grasslands is feasible anywhere, but effectiveness can be limited in very dry regions, and 8 for grasslands it is greatest in areas where degradation has occurred (e.g. by overgrazing) and soil 9 organic carbon is depleted. Barriers to implementation include regional capacity for monitoring and 10 verification (especially in developing countries), and more widely through concerns over saturation and 11 permanence. 12 7.4.3.2. Biochar 13 Activities, co-benefits, risks and implementation opportunities and barriers. Biochar is produced by 14 heating organic matter in oxygen-limited environments (pyrolysis and gasification) (Lehmann and 15 Joseph 2012). Feedstocks include forestry and sawmill residues, straw, manure and biosolids. When 16 applied to soils, biochar is estimated to persist from decades to thousands of years, depending on 17 feedstock and production conditions (Singh et al. 2015; Wang et al. 2016). Biochar systems producing 18 biochar for soil application plus bioenergy, generally give greater mitigation than bioenergy alone and 19 other uses of biochar, and are recognised as a CDR strategy. Biochar persistence is increased through 20 interaction with clay minerals and soil organic matter (Fang et al. 2015). Additional CDR benefits arise 21 through “negative priming” whereby biochar stabilises soil carbon and rhizodeposits (Archanjo et al. 22 2017; Hagemann et al. 2017; Weng et al. 2015; Han Weng et al. 2017; Weng et al. 2018; Wang et al. 23 2016). Besides CDR, additional mitigation can arise from displacing fossil fuels with pyrolysis gases, 24 lower soil N2O emissions (Cayuela et al. 2014, 2015; Song et al. 2016; He et al. 2017; Verhoeven et al. 25 2017; Borchard et al. 2019), reduced nitrogen fertiliser requirements due to reduced nitrogen leaching 26 and volatilisation from soils (Liu et al. 2019; Borchard et al. 2019), and reduced GHG emissions from 27 compost when biochar is added (Agyarko-Mintah et al. 2017; Wu et al. 2017). Biochar application to 28 paddy rice has resulted in substantial reductions (20-40% on average) in N2O (Awad et al. 2018; Liu et 29 al. 2018; Song et al. 2016) (Section 7.4.3.5) and smaller reduction in CH4 emissions (Kammann et al. 30 2017; Kim et al. 2017a; Song et al. 2016; He et al. 2017; Awad et al. 2018). Potential co-benefits include 31 yield increases particularly in sandy and acidic soils with low cation exchange capacity (Woolf et al. 32 2016; Jeffery et al. 2017); increased soil water-holding capacity (Omondi et al. 2016), nitrogen use 33 efficiency (Liu et al. 2019; Borchard et al. 2019), biological nitrogen fixation (Van Zwieten et al. 2015); 34 adsorption of organic pollutants and heavy metals (e.g. Silvani et al. 2019); odour reduction from 35 manure handling (e.g. Hwang et al. 2018) and managing forest fuel loads (Puettmann et al. 2020). Due 36 to its dark colour, biochar could decrease soil albedo (Meyer et al. 2012), though this is insignificant 37 under recommended rates and application methods. Biochar could reduce enteric CH4 emissions when 38 fed to ruminants (Section 7.4.3.4). Barriers to upscaling include insufficient investment, limited large- 39 scale production facilities, high production costs at small scale, lack of agreed approach to monitoring, 40 reporting and verification, and limited knowledge, standardisation and quality control, restricting user 41 confidence (Gwenzi et al. 2015). 42 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 43 potential, costs, and pathways. Biochar is discussed as a mitigation option in AR5 and CDR strategy 44 in the SR1.5. Consideration of potential was limited as biochar is not included in IAMs. The SRCCL 45 estimated mitigation potential of 0.03-6.6 GtCO2-eq yr-1 by 2050 based on studies with widely varying 46 assumptions, definitions of potential, and scope of mitigation processes included (SRCCL, Chapters 2 Do Not Cite, Quote or Distribute 7-63 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 and 4: (Roberts et al. 2010; Pratt and Moran 2010; Hristov; Lee and Day 2013; Dickie et al. 2014a; 2 Hawken 2017; Fuss et al. 2018; Powell and Lenton 2012; Woolf et al. 2010). 3 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Developments 4 include mechanistic understanding of ‘negative priming’ and biochar-soil-microbes-plant interactions 5 (DeCiucies et al. 2018; Fang et al. 2019). Indirect climate benefits are associated with persistent yield 6 response to biochar (Kätterer et al. 2019; Ye et al. 2020), improved crop water use efficiency (Du et al. 7 2018; Gao et al. 2020) and reduced GHG and ammonia emissions from compost and manure (Sanchez- 8 Monedero et al. 2018; Bora et al. 2020a,b; Zhao et al. 2020). A quantification method based on biochar 9 properties is included in the IPCC guidelines for NGHGIs (IPCC 2019b). Studies report a range of 10 biochar responses, from positive to occasionally adverse impacts, including on GHG emissions, and 11 identify risks (Tisserant and Cherubini 2019). This illustrates the expected variability (Lehmann and 12 Rillig 2014) of responses, which depend on the biochar type and climatic and edaphic characteristics of 13 the site (Zygourakis 2017). Biochar properties vary with feedstock, production conditions and post- 14 production treatments, so mitigation and agronomic benefits are maximised when biochars are chosen 15 to suit the application context (Mašek et al. 2018). A recent assessment finds greatest economic potential 16 (up to USD100 tCO2-1) between 2020 and 2050 to be in Asia and the developing Pacific (793 MtCO2 17 yr-1) followed by Developed Countries (447 MtCO2 yr-1) (Roe et al. 2021). Mitigation through biochar 18 will be greatest where biochar is applied to responsive soils (acidic, low fertility), where soil N 2O 19 emissions are high (intensive horticulture, irrigated crops), and where the syngas co-product displaces 20 fossil fuels. Due to the early stage of commercialisation, mitigation estimates are based pilot-scale 21 facilities, leading to uncertainty. However, the long-term persistence of biochar carbon in soils has been 22 widely studied (e.g. Singh et al. 2012; Fang et al. 2019; Zimmerman and Ouyang 2019). The greatest 23 uncertainty is the availability of sustainably-sourced biomass for biochar production. 24 Critical assessment and conclusion. Biochar has significant mitigation potential through CDR and 25 emissions reduction, and can also improve soil properties, enhancing productivity and resilience to 26 climate change (medium agreement, robust evidence). There is medium evidence that biochar has a 27 technical potential of 2.6 (0.2–6.6) GtCO2-eq yr-1, of which 1.1 (0.3–1.8) GtCO2-eq yr-1 is available up 28 to USD100 tCO2-1. However mitigation and agronomic co-benefits depend strongly on biochar 29 properties and the soil to which biochar is applied (strong agreement, robust evidence). While biochar 30 could provide moderate to large mitigation potential, it is not yet included in IAMs, which has restricted 31 comparison and integration with other CDR strategies. 32 7.4.3.3. Agroforestry 33 Activities, co-benefits, risks and implementation opportunities and barriers. Agroforestry is a set of 34 diverse land management systems that integrate trees and shrubs with crops and/or livestock in space 35 and/or time. Agroforestry accumulates carbon in woody vegetation and soil (Ramachandran Nair et al. 36 2010) and offers multiple co-benefits such as increased land productivity, diversified livelihoods, 37 reduced soil erosion, improved water quality, and more hospitable regional climates (Ellison et al. 2017; 38 Kuyah et al. 2019; Mbow et al. 2020; Zhu et al. 2020). Incorporation of trees and shrubs in agricultural 39 systems, however, can affect food production, biodiversity, local hydrology and contribute to social 40 inequality (Amadu et al. 2020; Fleischman et al. 2020; Holl and Brancalion 2020). To minimise risks 41 and maximise co-benefits, agroforestry should be implemented as part of support systems that deliver 42 tools, and information to increase farmers’ agency. This may include reforming policies, strengthening 43 extension systems and creating market opportunities that enable adoption (Jamnadass et al. 2020, 44 Sendzimir et al. 2011, Smith et al. 2019b). Consideration of carbon sequestration in the context of food 45 and fuel production, as well as environmental co-benefits at the farm, local, and regional scales can 46 further help support decisions to plant, regenerate and maintain agroforestry systems (Miller et al. 2020; 47 Kumar and Nair 2011). In spite of the advantages, biophysical and socioeconomic factors can limit the Do Not Cite, Quote or Distribute 7-64 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 adoption (Pattanayak et al. 2003). Contextual factors may include, but are not limited to; water 2 availability, soil fertility, seed and germplasm access, land policies and tenure systems affecting farmer 3 agency, access to credit, and to information regarding the optimum species for a given location. 4 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 5 potential, costs, and pathways. The SRCCL estimated the global technical mitigation potential of 6 agroforestry, with medium confidence, to be between 0.08 and 5.6 GtCO2-eq yr-1 by 2050 (Griscom et 7 al. 2017; Dickie et al. 2014a; Zomer et al. 2016; Hawken 2017). Estimates are derived from syntheses 8 of potential area available for various agroforestry systems e.g., windbreaks, farmer managed natural 9 regeneration, and alley cropping and average annual rates of carbon accumulation. The cost-effective 10 economic potential, also with medium confidence, is more limited at 0.3-2.4 GtCO2-eq yr-1 (Zomer et 11 al. 2016; Griscom et al. 2017; Roe et al. 2019). Despite this potential, agroforestry is currently not 12 considered in integrated assessment models used for mitigation pathways (Section 7.5). 13 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Updated 14 estimates of agroforestry’s technical mitigation potential and synthesised estimates of carbon 15 sequestration across agroforestry systems have since been published. The most recent global analysis 16 estimates technical potential of 9.4 GtCO2-eq yr-1 (Chapman et al. 2020) of agroforestry on 1.87 and 17 1.89 billion ha of crop and pasture lands below median carbon content, respectively. This estimate is at 18 least 68% greater than the largest estimate reported in the SRCCL (Hawken 2017) and represents a new 19 conservative upper bound as Chapman et al. (2020) only accounted for aboveground carbon. 20 Considering both above- and belowground carbon of windbreaks, alley cropping and silvopastoral 21 systems at a more limited areal extent (Griscom et al. 2020), the economic potential of agroforestry was 22 estimated to be only about 0.8 GtCO2-eq yr-1. Variation in estimates primarily result from assumptions 23 on the agroforestry systems including, extent of implementation and estimated carbon sequestration 24 potential when converting to agroforestry. 25 Regional estimates of mitigation potential are scant with agroforestry options differing significantly by 26 geography (Feliciano et al. 2018). For example, multi-strata shaded coffee and cacao are successful in 27 the humid tropics (Somarriba et al. 2013; Blaser et al. 2018), silvopastoral systems are prevalent in 28 Latin American (Peters et al. 2013; Landholm et al. 2019) while agrosilvopastoral systems, shelterbelts, 29 hedgerows, and windbreaks are common in Europe (Joffre et al. 1988; Rigueiro-Rodriguez 2009). At 30 the field scale, agroforestry accumulates between 0.59 and 6.24 t ha -1 yr-1 of carbon aboveground. 31 Belowground carbon often constitutes 25% or more of the potential carbon gains in agroforestry 32 systems (De Stefano and Jacobson 2018; Cardinael et al. 2018). Roe et al. (2021) estimate greatest 33 regional economic (up to USD100 tCO2-1) mitigation potential for the period 2020-2050 to be in Asia 34 and the developing Pacific (368.4 MtCO2-eq yr-1) and Developed Countries (264.7 MtCO2-eq yr-1). 35 Recent research has also highlighted co-benefits and more precisely identified implementation barriers. 36 In addition to aforementioned co-benefits, evidence now shows that agroforestry can improve soil 37 health, regarding infiltration and structural stability (Muchane et al. 2020); reduces ambient 38 temperatures and crop heat stress (Arenas-Corraliza et al. 2018; Sida et al. 2018); increases groundwater 39 recharge in drylands when managed at moderate density (Ilstedt et al. 2016; Bargués-Tobella et al. 40 2020); positively influences human health (Rosenstock et al. 2019); and can improve dietary diversity 41 (McMullin et al. 2019). Along with previously mentioned barriers, low social capital, assets, and labour 42 availability have been identified as pertinent to adoption. Practically all barriers are interdependent and 43 subject to the context of implementation. 44 Critical assessment and conclusion. There is medium confidence that agroforestry has a technical 45 potential of 4.1 (0.3-9.4) GtCO2-eq yr-1 for the period 2020-2050, of which 0.8 (0.4-1.1) GtCO2-eq yr-1 46 is available at USD100 tCO2-1. Despite uncertainty around global estimates due to regional preferences Do Not Cite, Quote or Distribute 7-65 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 for management systems, suitable land availability, and growing conditions, there is high confidence in 2 agroforestry’s mitigation potential at the field scale. With countless options for farmers and land 3 managers to implement agroforestry, there is medium confidence in the feasibility of achieving 4 estimated regional mitigation potential. Appropriately matching agroforestry options, to local 5 biophysical and social contexts is important in maximising mitigation and co-benefits, while avoiding 6 risks (Sinclair and Coe 2019). 7 8 [START BOX 7.3 HERE] 9 Box 7.3 Case study: agroforestry in Brazil – CANOPIES 10 Summary 11 Brazilian farmers are integrating trees into their croplands in various ways, ranging from simple to 12 highly complex agroforestry systems. While complex systems are more effective in the mitigation of 13 climate change, trade-offs with scalability need to be resolved for agroforestry systems to deliver on 14 their potential. The Brazilian-Dutch CANOPIES project (Janssen 2020) is exploring transition 15 pathways to agroforestry systems optimised for local ecological and socio-economic conditions 16 Background 17 The climate change mitigation potential of agroforestry systems is widely recognised (FAO 2017b; 18 Zomer et al. 2016) and Brazilian farmers and researchers are pioneering diverse ways of integrating 19 trees into croplands, from planting rows of eucalyptus trees in pastures up to highly complex agroforests 20 consisting of >30 crop and tree species. The degree of complexity influences the multiple functions that 21 farmers and societies can attain from agroforestry: the more complex it is, the more it resembles a 22 natural forest with associated benefits for its C storage capacity and its habitat quality for biodiversity 23 (Santos et al. 2019). However, trade-offs exist between the complexity and scalability of agroforestry 24 as complex systems rely on intensive manual labour to achieve high productivity (Tscharntke et al. 25 2011). To date, mechanisation of structurally diverse agroforests is scarce and hence, efficiencies of 26 scale are difficult to achieve. 27 Case description 28 These synergies and trade-offs between complexity, multifunctionality and scalability are studied in the 29 CANOPIES (Co-existence of Agriculture and Nature: Optimisation and Planning of Integrated 30 Ecosystem Services) project, a collaboration between Wageningen University (NL), the University of 31 São Paulo and EMBRAPA (both Brazil). Soil and management data are collected on farms of varying 32 complexity to evaluate C sequestration and other ecosystem services, economic performance and labour 33 demands. 34 Interactions and limitations 35 The trade-off between complexity and labour demand is less pronounced in EMBRAPA’s integrated 36 crop-livestock-forestry (ICLF) systems, where grains and pasture are planted between widely spaced 37 tree rows. Here, barriers for implementation relate mostly to livestock and grain farmers’ lack of 38 knowledge on forestry management and financing mechanisms5 (Gil et al. 2015). Additionally, linking 39 these financing mechanisms to C sequestration remains a Monitoring, Reporting and Verification 40 challenge (Smith et al. 2020b). 41 Lessons Do Not Cite, Quote or Distribute 7-66 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Successful examples of how more complex agroforestry can be upscaled do exist in Brazil. For example, 2 on farm trials and consistent investments over several years have enabled Rizoma Agro to develop a 3 citrus production system that integrates commercial and native trees in a large-scale multi-layered 4 agroforestry system. The success of their transition resulted in part from their corporate structure that 5 allowed them to tap into the certified Green Bonds market (CBI 2020). However, different transition 6 strategies need to be developed for family farmers and their distinct socio-economic conditions. 7 [END BOX 7.3 HERE] 8 9 7.4.3.4. Enteric fermentation 10 Activities, co-benefits, risks and implementation opportunities and barriers. Mitigating CH4 emissions 11 from enteric fermentation can be direct (i.e. targeting ruminal methanogenesis and emissions per animal 12 or unit of feed consumed) or indirect, by increasing production efficiency (i.e. reducing emission 13 intensity per unit of product). Measures can be classified as those relating to (1) feeding, (2) 14 supplements, additives and vaccines, and (3) livestock breeding and wider husbandry (Jia et al. 2019). 15 Co-benefits include enhanced climate change adaptation and increased food security associated with 16 improved livestock breeding (Smith et al. 2014). Risks include mitigation persistence, ecological 17 impacts associated with improving feed quality and supply, or potential toxicity and animal welfare 18 issues concerning feed additives. Implementation barriers include feeding/administration constraints, 19 the stage of development of measures, legal restrictions on emerging technologies and negative impacts, 20 such as the previously described risks (Smith et al. 2014; Jia et al. 2019; Smith et al. 2019b). 21 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 22 potential, costs, and pathways. AR5 indicated medium (5-15%) technical mitigation potential from 23 both feeding and breeding related measures (Smith et al. 2014). More recently, the SRCCL estimated 24 with medium confidence, a global potential of 0.12-1.18 GtCO2-eq yr-1 between 2020 and 2050, with 25 the range reflecting technical, economic and sustainability constraints (SRCCL, Chapter 2: Hristov et 26 al. 2013; Dickie et al. 2014a; Herrero et al. 2016; Griscom et al. 2017). The underlying literature used 27 a mixture of IPCC GWP100 values for CH4, preventing conversion of CO2-eq to CH4. Improved livestock 28 feeding and breeding were included in IAM emission pathway scenarios within the SRCCL and SR1.5, 29 although it was suggested that the full mitigation potential of enteric CH4 measures is not captured in 30 current models (Rogelj et al. 2018b; IPCC 2018). 31 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Recent reviews 32 generally identify the same measures as those outlined in the SRCCL, with the addition of early life 33 manipulation of the ruminal biome (Grossi et al. 2019; Eckard and Clark 2020; Thompson and 34 Rowntree 2020; Beauchemin et al. 2020; Ku-Vera et al. 2020; Honan et al. 2021). There is robust 35 evidence and high agreement that chemically synthesised inhibitors are promising emerging near-term 36 measures (Patra 2016; Jayanegara et al. 2018; Van Wesemael et al. 2019; Beauchemin et al. 2020) with 37 high (e.g. 16-70% depending on study) mitigation potential reported (e.g. Hristov et al. 2015; McGinn 38 et al. 2019; Melgar et al. 2020) and commercial availability expected within two years in some countries 39 (Reisinger et al. 2021). However, their mitigation persistence (McGinn et al. 2019), cost (Carroll and 40 Daigneault 2019; Alvarez-Hess et al. 2019) and public acceptance (Jayasundara et al. 2016) or 41 regulatory approval is currently unclear while administration in pasture-based systems is likely to be 42 challenging (Patra et al. 2017; Leahy et al. 2019). Research into other inhibitors/feeds containing 43 inhibitory compounds, such as macroalga or seaweed (Chagas et al. 2019; Kinley et al. 2020; Roque et 44 al. 2019), shows promise, although concerns have been raised regarding palatability, toxicity, 45 environmental impacts and the development of industrial-scale supply chains (Abbott et al. 2020; Vijn 46 et al. 2020). In the absence of CH4 vaccines, which are still under development (Reisinger et al. 2021) Do Not Cite, Quote or Distribute 7-67 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 pasture-based and non-intensive systems remain reliant on increasing production efficiency 2 (Beauchemin et al. 2020). Breeding of low emitting animals may play an important role and is a subject 3 under on-going research (Pickering et al. 2015; Jonker et al. 2018; López-Paredes et al. 2020). 4 Approaches differ regionally, with more focus on direct, technical options in developed countries, and 5 improved efficiency in developing countries (Caro Torres et al. 2016; Mottet et al. 2017b; MacLeod et 6 al. 2018; Frank et al. 2018). A recent assessment finds greatest economic (up to USD100 tCO2-eq-1) 7 potential (using the IPCC AR4 GWP100 value for CH4 ) for 2020-2050 in Asia and the developing Pacific 8 (32.9 MtCO2-eq yr-1) followed by Developed Countries (25.5 MtCO2-eq yr-1) (Roe et al. 2021). Despite 9 numerous country and sub-sector specific studies, most of which include cost analysis (Hasegawa and 10 Matsuoka 2012; Hoa et al. 2014; Jilani et al. 2015; Eory et al. 2015; Pradhan et al. 2017; Pellerin et al. 11 2017; Ericksen and Crane 2018; Habib and Khan 2018; Kashangaki and Ericksen 2018; Salmon et al. 12 2018; Brandt et al. 2019b; Dioha and Kumar 2020; Kiggundu et al. 2019; Kavanagh et al. 2019; Mosnier 13 et al. 2019; Pradhan et al. 2019; Sapkota et al. 2019; Carroll and Daigneault 2019; Leahy et al. 2019), 14 sectoral assessment of regional technical and notably economic (Beach et al. 2015; USEPA 2019) 15 potential is restricted by lack comprehensive and comparable data. Therefore, verification of regional 16 estimates indicated by global assessments is challenging. Feed quality improvement, which may have 17 considerable potential in developing countries (Mottet et al. 2017a; Caro et al. 2016), may have negative 18 wider impacts. For example, potential land use change and greater emissions associated with production 19 of concentrates (Brandt et al. 2019b). 20 Critical review and conclusion. Based on studies to date, using a range of IPCC GWP100 values for 21 CH4, there is medium confidence that activities to reduce enteric CH4 emissions have a global technical 22 potential of 0.8 (0.2–1.2) GtCO2-eq yr-1, of which 0.2 (0.1–0.3) GtCO2-eq yr-1 is available up to USD100 23 tCO2-eq-1 (Figure 7.11). The CO2-eq value may also slightly differ if the GWP100 IPCC AR6 CH4 value 24 was uniformly applied within calculations. Lack of comparable country and sub-sector studies to assess 25 the context applicability of measures, associated costs and realistic adoption likelihood, prevents 26 verification of estimates. 27 7.4.3.5. Improve rice management 28 Activities, co-benefits, risks and implementation opportunities and barriers. Emissions from rice 29 cultivation mainly concern CH4 associated with anaerobic conditions, although N2O emission also occur 30 via nitrification and denitrification processes. Measures to reduce CH4 and N2O emissions include (1) 31 improved water management (e.g. single drainage and multiple drainage practices), (2) improved 32 residue management, (3) improved fertiliser application (e.g. using slow release fertiliser and nutrient 33 specific application), and (4) soil amendments (including biochar and organic amendments) (Pandey et 34 al. 2014; Yagi et al. 2020; Sriphirom et al. 2020; Kim et al. 2017b). These measures not only have 35 mitigation potential but can improve water use efficiency, reduce overall water use, enhance drought 36 adaptation and overall system resilience, improve yield, reduce production costs from seed, pesticide, 37 pumping and labour, increase farm income, and promote sustainable development (Sriphirom et al. 38 2019; Tran et al. 2018; Yamaguchi et al. 2017; Quynh and Sander 2015). However, in terms of 39 mitigation of CH4 and N2O, antagonistic effects can occur, whereby water management can enhance 40 N2O emissions due to creation of alternate wet and dry conditions (Sriphirom et al. 2019), with trade- 41 offs between CH4 and N2O during the drying period potentially off-setting some mitigation benefits. 42 Barriers to adoption may include site-specific limitations regarding soil type, percolation and seepage 43 rates or fluctuations in precipitation, water canal or irrigation infrastructure, paddy surface level and 44 rice field size, and social factors including farmer perceptions, pump ownership, and challenges in 45 synchronising water management between neighbours and pumping stations (Yamaguchi et al. 2019; 46 Yamaguchi et al. 2017; Quynh and Sander 2015). Do Not Cite, Quote or Distribute 7-68 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 2 potential, costs, and pathways. AR5 outlined emissions from rice cultivation of 0.49-0.723 GtCO2-eq 3 yr-1 in 2010 with an average annual growth of 0.4% yr-1. The SRCCL estimated a global mitigation 4 potential from improved rice cultivation of 0.08-0.87 GtCO2-eq yr-1 between 2020 and 2050, with the 5 range representing the difference between technical and economic constraints, types of activities 6 included (e.g. improved water management and straw residue management) and GHGs considered 7 (SRCCL, Chapter 2: Dickie et al. 2014a; Paustian et al. 2016; Beach et al. 2015; Griscom et al. 2017; 8 Hawken 2017). 9 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Since AR5 and 10 the SRCCL, studies on mitigation have principally focused on water and nutrient management practices 11 with the aim of improving overall sustainability as well as measurements of site-specific emissions to 12 help improve the resolution of regional estimates. Intensity of emissions show considerable spatial and 13 temporal variation, dependent on site specific factors including degradation of soil organic matter, 14 management of water levels in the field, the types and amount of fertilisers applied, rice variety and 15 local cultivation practices. Variation in CH4 emissions have been found to range from 0.5-41.8 mg/m2/hr 16 in Southeast Asia (Sander et al. 2014; Chidthaisong et al. 2018; Setyanto et al. 2018; Sibayan et al. 17 2018; Wang et al. 2018; Maneepitak et al. 2019), 0.5-37.0 mg/m2/hr in Southern and Eastern Asia 18 (Zhang et al. 2010; Wang et al. 2012; Oo et al. 2018; Takakai et al. 2020; Wang et al. 2018), and 0.5- 19 10.4 mg/m2/hr in North America (Wang et al. 2018). Current studies on emissions of N2O also showed 20 high variation in the range of 0.13-654 ug/m2/hr (Akiyama et al. 2005; Islam et al. 2018; Kritee et al. 21 2018; Zschornack et al. 2018; Oo et al. 2018). 22 Recent studies on water management have highlighted the potential to mitigate GHG emissions, while 23 also enhancing water use efficiency (Tran et al. 2018). A meta-analysis on multiple drainage systems 24 found that Alternative Wetting and Drying (AWD) with irrigation management, can reduce CH4 25 emissions by 20-30% and water use by 25.7 %, though this resulted in a slight yield reduction (5.4%) 26 (Carrijo et al. 2017). Other studies have described improved yields associated with AWD (Tran et al. 27 2018). Water management for both single and multiple drainage can (most likely) reduce methane 28 emissions by ~35 % but increase N2O emissions by about 20% (Yagi et al. 2020). However, N2O 29 emissions occur only under dry conditions, therefore total reduction in terms of net GWP is 30 approximately 30%. Emissions of N2O are higher during dry seasons (Yagi et al. 2020) and depend on 31 site specific factors as well as the quantity of fertiliser and organic matter inputs into the paddy rice 32 system. Variability of N2O emissions from single and multiple drainage can range from 0.06-33 kg/ha 33 (Hussain et al. 2015; Kritee et al. 2018). AWD in Vietnam was found to reduce both CH4 and N2O 34 emissions by 29-30 and 26-27% respectively with the combination of net GWP about 30% as compared 35 to continuous flooding (Tran et al. 2018). Overall, greatest average economic mitigation potential (up 36 to USD100 tCO2-eq-1) between 2020 and 2050 is estimated to be in Asia and the developing Pacific 37 (147.2 MtCO2-eq yr-1) followed by Latin America and the Caribbean (8.9 MtCO2-eq yr-1) using the 38 IPCC AR4 GWP100 value for CH4 (Roe et al. 2021). 39 Critical assessment and conclusion. There is medium confidence that improved rice management has 40 a technical potential of 0.3 (0.1-0.8) GtCO2-eq yr-1 between 2020 and 2050, of which 0.2 (0.05-0.3) 41 GtCO2-eq yr-1 is available up to USD100 tCO2-eq-1 (Figure 7.11). Improving rice cultivation practices 42 will not only reduce GHG emissions, but also improve production sustainability in terms of resource 43 utilisation including water consumption and fertiliser application. However, emission reductions show 44 high variability and are dependent on site specific conditions and cultivation practices. Do Not Cite, Quote or Distribute 7-69 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.4.3.6. Crop nutrient management 2 Activities, co-benefits, risks and implementation opportunities and barriers. Improved crop nutrient 3 management can reduce N2O emissions from cropland soils. Practices include optimising fertiliser 4 application delivery, rates and timing, utilising different fertiliser types (i.e. organic manures, composts 5 and synthetic forms), and using slow or controlled-released fertilisers or nitrification inhibitors (Smith 6 et al. 2014; Griscom et al. 2017; Smith et al. 2019b). In addition to individual practices, integrated 7 nutrient management that combines crop rotations including intercropping, nitrogen biological fixation, 8 reduced tillage, use of cover crops, manure and bio-fertilizer application, soil testing and comprehensive 9 nitrogen management plans, is suggested as central for optimising fertiliser use, enhancing nutrient 10 uptake and potentially reducing N2O emissions (Bationo et al. 2012; Lal et al. 2018; Bolinder et al. 11 2020; Jensen et al., 2020; Namatsheve et al., 2020). Such practices may generate additional mitigation 12 by indirectly reducing synthetic fertilizer manufacturing requirements and associated emissions, though 13 such mitigation is accounted for in the Industry Sector and not considered in this chapter. Tailored 14 nutrient management approaches, such as 4R nutrient stewardship, are implemented in contrasting 15 farming systems and contexts and supported by best management practices to balance and match 16 nutrient supply with crop requirements, provide greater stability in fertilizer performance and to 17 minimize N2O emissions and nutrient losses from fields and farms (Fixen 2020; Maaz et al. 2021). Co- 18 benefits of improved nutrient management can include enhanced soil quality (notably when manure, 19 crop residues or compost is utilised), carbon sequestration in soils and biomass, soil water holding 20 capacity, adaptation capacity, crop yields, farm incomes, water quality (from reduced nitrate leaching 21 and eutrophication), air quality (from reduced ammonia emissions) and in certain cases, it may facilitate 22 land sparing (Sapkota et al. 2014; Johnston and Bruulsema 2014; Zhang et al. 2017; Smith et al. 2019b; 23 Mbow et al. 2019). 24 A potential risk under certain circumstances, is yield reduction, while implementation of practices 25 should consider current soil nutrient status. There are significant regional imbalances, with some 26 regions experiencing nutrient surpluses from over fertilization and others, nutrient shortages and 27 chronic deficiencies (FAO 2021e). Additionally, depending on context, practices may be inaccessible, 28 expensive or require expertise to implement (Hedley 2015; Benson and Mogues 2018) while impacts 29 of climate change may influence nutrient use efficiency (Amouzou et al. 2019) and therefore, mitigation 30 potential. 31 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 32 potential, costs, and pathways. The SRCCL broadly identified the same practices as outlined in AR5 33 and estimated that improved cropland nutrient management could mitigate between 0.03 and 0.71 34 GtCO2-eq yr-1 between 2020 and 2050 (SRCCL Chapter 2: Dickie et al. 2014a; Beach et al. 2015; 35 Paustian et al. 2016; Griscom et al. 2017; Hawken 2017). 36 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Research since 37 the SRCCL highlights the mitigation potential and co-benefits of adopting improved nutrient 38 management strategies, notably precision fertiliser application methods and nutrient expert systems, 39 and applicability in both large-scale mechanised and small-scale systems (Aryal et al. 2020; USEPA 40 2019; Hijbeek et al. 2019; Griscom et al. 2020; Tian et al. 2020; Sapkota et al. 2021). Improved crop 41 nutrient management is feasible in all regions, but effectiveness is context dependent. Sub-Saharan 42 Africa has one of the lowest global fertiliser consumption rates, with increased fertiliser use suggested 43 as necessary to meet projected future food requirements (Mueller et al. 2012; Adam et al. 2020; ten 44 Berge et al. 2019; Falconnier et al. 2020). Fertiliser use in Developed Countries is already high (Figure 45 7.10) with increased nutrient use efficiency among the most promising mitigation measures (Roe et al. 46 2019; Hijbeek et al. 2019). Considering that Asia and developing Pacific, and Developed Countries 47 accounted for the greatest share of global nitrogen fertiliser use, it is not surprising that these regions Do Not Cite, Quote or Distribute 7-70 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 are estimated to have greatest economic mitigation potential (up to USD100 tCO2-eq-1) between 2020 2 and 2050, at 161.8 and 37.1 MtCO2-eq yr-1 respectively (using the IPCC AR4 GWP100 value for N2O) 3 (Roe et al. 2021). 4 Critical assessment and conclusion. There is medium confidence that crop nutrient management has a 5 technical potential of 0.3 (0.06–0.7) GtCO2-eq yr-1 of which 0.2 (0.05–0.6) GtCO2-eq yr-1 is available 6 up to USD100 tCO2-eq-1. This value is based on GWP100 using a mixture of IPCC values for N2O and 7 may slightly differ if calculated using AR6 values. The development of national roadmaps for 8 sustainable fertilizer (nutrient) management can help in scaling-up related practices and in realising this 9 potential. Crop nutrient management measures can contribute not only to mitigation, but food and 10 nutrition security and wider environmental sustainability goals. 11 12 [START BOX 7.4 HERE] 13 Box 7.4 Case study: the climate-smart village approach 14 Summary 15 The climate-smart villages (CSV) approach aims to generate local knowledge, with the involvement of 16 farmers, researchers, practitioners, and governments, on climate change adaptation and mitigation while 17 improving productivity, food security, and farmers' livelihoods (Aggarwal et al. 2018). This knowledge 18 feeds a global network that includes 36 climate-smart villages in South and Southeast Asia, West and 19 East Africa, and Latin America. 20 Background 21 It is expected that agricultural production systems across the world will change in response to climate 22 change, posing significant challenges to the livelihoods and food security of millions of people (IPCC 23 2014). Maintaining agricultural growth while minimising climate shocks is crucial to building a resilient 24 food production system and meeting sustainable development goals in vulnerable countries. 25 Case description 26 The CSV approach seeks an integrated vision so that sustainable rural development is the final goal for 27 rural communities. At the same time, it fosters the understanding of climate change with the 28 implementation of adaptation and mitigation actions, as much as possible. Rural communities and local 29 stakeholders are the leaders of this process, where scientists facilitate their knowledge to be useful for 30 the communities and learn at the same time about challenges but also the capacity those communities 31 have built through time. The portfolio includes weather-smart activities, water-smart practices, 32 seed/breed smart, carbon/nutrient-smart practices, and institutional/market smart activities. 33 Interactions and limitations 34 The integration of technologies and services that are suitable for the local conditions resulted in many 35 gains for food security and adaptation and for mitigation where appropriate. It was also shown that, in 36 all regions, there is considerable yield advantage when a portfolio of technologies is used, rather than 37 the isolated use of technologies (Govaerts et al. 2005; Zougmoré et al. 2014). Moreover, farmers are 38 using research results to promote their products as climate-smart leading to increases in their income 39 (Acosta-Alba et al. 2019). However, climatic risk sites and socioeconomic conditions together with a 40 lack of resource availability are key issues constraining agriculture across all five regions. 41 Lessons Do Not Cite, Quote or Distribute 7-71 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 1. Understanding the priorities, context, challenges, capacity, and characteristics of the territory and 2 the communities regarding climate, as well as the environmental and socioeconomic dimensions, 3 is the first step. Then, understanding climate vulnerability in their agricultural systems based on 4 scientific data but also listening to their experience will set the pathway to identify climate-smart 5 agriculture (CSA) options (practices and technologies) to reduce such vulnerability. 6 2. Building capacity is also a critical element of the CSV approach, rural families learn about the 7 practices and technologies in a neighbour's house, and as part of the process, families commit to 8 sharing their knowledge with other families, to start a scaling-out process within the communities. 9 Understanding the relationship between climate and their crop is key, as well as the use of weather 10 forecasts to plan their agricultural activities. 11 3. The assessment of the implementation of the CSA options should be done together with community 12 leaders to understand changes in livelihoods and climate vulnerability. Also, knowledge 13 appropriation by community leaders has led to farmer-to-farmer knowledge exchange within and 14 outside the community (Ortega Fernandez and Martínez-Barón 2018). 15 [END BOX 7.4 HERE] 16 17 7.4.3.7. Manure management 18 Activities, co-benefits, risks and implementation opportunities and barriers. Manure management 19 measures aim to mitigate CH4 and N2O emissions from manure storage and deposition. Mitigation of 20 N2O considers both direct and indirect (i.e. conversion of ammonia and nitrate to N2O) sources. 21 According to the SRCCL, measures may include (1) anaerobic digestion, (2) applying nitrification or 22 urease inhibitors to stored manure or urine patches, (3) composting, (4) improved storage and 23 application practices, (5) grazing practices and (6) alteration of livestock diets to reduce nitrogen 24 excretion (Mbow et al. 2019; Jia et al. 2019). Implementation of manure management with other 25 livestock and soil management measures can enhance system resilience, sustainability, food security 26 and help prevent land degradation (Smith et al. 2014; Mbow et al. 2019; Smith et al. 2019d), while 27 potentially benefiting the localised environment, for example, regarding water quality (Di and Cameron 28 2016). Risks include increased N2O emission from the application of manure to poorly drained or wet 29 soils, trade-offs between N2O and ammonia emissions and potential eco-toxicity associated with some 30 measures. 31 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 32 potential, costs, and pathways. AR5 reported manure measures to have high (> 10%) mitigation 33 potential. The SRCCL estimated a technical global mitigation potential between 2020 and 2050 of 0.01- 34 0.26 GtCO2-eq yr-1, with the range depending on economic and sustainable capacity (SRCCL, Chapter 35 2: (Dickie et al. 2014a; Herrero et al. 2016). Conversion of estimates to native units is restricted as a 36 mixture of GWP100 values was used in underlying studies. Measures considered were typically more 37 suited to confined production systems (Jia et al. 2019; Mbow et al. 2019), while improved manure 38 management is included within IAM emission pathways (Rogelj et al. 2018b). 39 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Research 40 published since SRCCL broadly focuses on measures relevant to intensive or confined systems (e.g. 41 (Hunt et al. 2019; Sokolov et al. 2020; Im et al. 2020; Adghim et al. 2020; Mostafa et al. 2020; Kavanagh 42 et al. 2019), highlighting co-benefits and risks. For example, measures may enhance nutrient recovery, 43 fertiliser value (Sefeedpari et al. 2019; Ba et al. 2020; Yao et al. 2020) and secondary processes such as 44 biogas production (Shin et al. 2019). However, the potential antagonistic relationship between GHG 45 and ammonia mitigation and need for appropriate management is emphasised (Aguirre-Villegas et al. Do Not Cite, Quote or Distribute 7-72 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2019; Kupper et al. 2020; Grossi et al. 2019; Ba et al. 2020). In some circumstances, fugitive emissions 2 may reduce the potential mitigation benefits of biogas production (Scheutz and Fredenslund 2019; 3 Bakkaloglu et al. 2021), while high implementation cost is identified as an adoption barrier, notably of 4 anaerobic digestion (Liu and Liu 2018; Niles and Wiltshire 2019; Ndambi et al. 2019; Ackrill and Abdo 5 2020; Adghim et al. 2020). Nitrification inhibitors have been found to be effective at reducing N2O 6 emissions from pasture deposited urine (López-Aizpún et al. 2020), although the use of nitrification 7 inhibitors is restricted in some jurisdictions due to concerns regarding residues in food products (Di and 8 Cameron 2016; Eckard and Clark 2020) while limited evidence suggests eco-toxicity risk under certain 9 circumstances (Kösler et al. 2019). Some forage crops may naturally contain inhibitory substances 10 (Simon et al. 2019, 2020; de Klein et al. 2020), though this warrants further research (Podolyan et al. 11 2020; Gardiner et al. 2020). 12 Country specific studies provide insight into regionally applicable measures, with emphasis on small- 13 scale anaerobic digestion (e.g. dome digesters), solid manure coverage and daily manure spreading in 14 Asia and the developing Pacific, and Africa (Hasegawa et al. 2016; Hasegawa and Matsuoka 2012; Hoa 15 et al. 2014; Jilani et al. 2015; Pradhan et al. 2017; Ericksen and Crane 2018; Pradhan et al. 2019; 16 Kiggundu et al. 2019; Dioha and Kumar 2020). Tank/lagoon covers, large-scale anaerobic digestion, 17 improved application timing, nitrogen inhibitor application to urine patches, soil-liquid separation, 18 reduced livestock nitrogen intake, trailing shoe, band or injection slurry spreading and acidification are 19 emphasised in developed countries (Kaparaju and Rintala 2011; Pape et al. 2016; Liu and Liu 2018; 20 Lanigan et al. 2018; Eory et al. 2015; Jayasundara et al. 2016; Pellerin et al. 2017; Carroll and 21 Daigneault 2019; Eckard and Clark 2020). Using IPCC AR4 GWP100 values for CH4 and N2O, a recent 22 assessment finds 69% (63.4 MtCO2-eq yr-1) of economic potential (up to USD100 tCO2-eq-1) between 23 2020-2050, to be in Developed Countries (Roe et al. 2021). 24 Critical assessment and conclusion. There is medium confidence that manure management measures 25 have a global technical potential of 0.3 (0.1-0.5) GtCO2-eq yr-1, (using a range of IPCC GWP100 values 26 for CH4 and N2O), of which 0.1 (0.09-0.1) GtCO2-eq yr-1 is available at up to USD100 tCO2-eq-1 (Figure 27 7.11). As with other non-CO2 GHG mitigation estimates, values may slightly differ depending upon 28 which IPCC GWP100 values were used. There is robust evidence and high agreement that there are 29 measures that can be applied in all regions, but greatest mitigation potential is estimated in developed 30 countries in more intensive and confined production systems. 31 32 [START BOX 7.5 HERE] 33 Box 7.5 Farming system approaches and mitigation 34 Introduction 35 There is robust evidence and high agreement that agriculture needs to change to facilitate environment 36 conservation while maintaining and where appropriate, increase overall production. The SRCCL 37 identified several farming system approaches, deemed alternative to conventional systems (Olsson et 38 al. 2019; Mbow et al. 2019; Smith et al. 2019a). These may incorporate several of the mitigation 39 measures described in 7.4.3, while potentially also delivering environmental co-benefits. This Box 40 assesses evidence specifically on the mitigation capacity of some such system approaches. The 41 approaches are not mutually exclusive, may share similar principles or practices and can be 42 complimentary. In all cases, mitigation may result from either (1) emission reductions or (2) enhanced 43 carbon sequestration, via combinations of management practices as outlined in Figure 1 within this Box. 44 The approaches will have pros and cons concerning multiple factors, including mitigation, yield and 45 co-benefits, with trade-offs subject to the diverse contexts and ways in which they are implemented. Do Not Cite, Quote or Distribute 7-73 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 3 Box 7.5, Figure 1 Potential mitigation mechanisms and associated management practices 4 Is there evidence that these approaches deliver mitigation? 5 Agroecology (AE) including Regenerative Agriculture (RA) 6 There is limited discussion on the mitigation potential of AE (Gliessman 2013; Altieri and Nicholls 7 2017), but robust evidence that AE can improve system resilience and bring multiple co-benefits 8 (Aguilera et al. 2020; Tittonell 2020; Wanger et al. 2020; Altieri et al. 2015; Mbow et al. 2019) (IPCC 9 WGII AR6 Box 5.10). Limited evidence concerning the mitigation capacity of AE at a system level (Saj 10 et al. 2017; Snapp et al. 2021) makes conclusions difficult, yet studies into specific practices that may 11 be incorporated, suggest AE may have mitigation potential (Section 7.4.3) (medium confidence). 12 However, AE, that incorporates management practices used in organic farming (see below), may result 13 in reduced yields, driving compensatory agricultural production elsewhere. Research into GHG 14 mitigation by AE as a system and impacts of wide-scale implementation is required. Despite absence 15 of a universally accepted definition (see Annex I), RA is gaining increasing attention and shares 16 principles of AE. Some descriptions include carbon sequestration as a specific aim (Elevitch et al. 17 2018). Few studies have assessed mitigation potential of RA at a system level (e.g. Colley et al. 2020). 18 Like AE, it is likely that RA can contribute to mitigation, the extent to which is currently unclear and 19 by its case-specific design, will vary (medium confidence). Do Not Cite, Quote or Distribute 7-74 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Conservation Agriculture (CA) 2 The SRCCL noted both positive and inconclusive results regarding CA and soil carbon, with sustained 3 sequestration dependent on productivity and residue returns (Jia et al. 2019; Mirzabaev et al. 2019; 4 Mbow et al. 2019). Recent research is in broad agreement (Ogle et al. 2019; Corbeels et al. 2020, 2019; 5 Gonzalez-Sanchez et al. 2019; Munkholm et al. 2020) with greatest mitigation potential suggested in 6 dry regions (Sun et al. 2020). Theoretically, CA may facilitate improved nitrogen use efficiency (Lal 7 2015; Powlson et al. 2016) (limited evidence), though CA appears to have mixed effects on soil N2O 8 emission (Six et al. 2004; Mei et al. 2018). CA is noted for its adaptation benefits, with wide agreement 9 that CA can enhance system resilience to climate related stress, notably in dry regions. There is evidence 10 that CA can contribute to mitigation, but its contribution is depended on multiple factors including 11 climate and residue returns (high confidence). 12 Integrated Production Systems (IPS) 13 The integration of different enterprises in space and time (e.g. diversified cropping, crop and livestock 14 production, agroforestry), therefore facilitating interaction and transfer of recourses between systems, 15 is suggested to enhance sustainability and adaptive capacity (Franzluebbers et al. 2014; Lemaire et al. 16 2014; Gil et al. 2017; Peterson et al. 2020; Walkup et al. 2020; Garrett et al. 2020; Hendrickson et al. 17 2008; Weindl et al. 2015; Olseen et al. 2019). Research indicates some mitigation potential, including 18 by facilitating sustainable intensification (Box 7.11), though benefits are likely to be highly context 19 specific (e.g. (Herrero et al. 2013; Carvalho et al. 2014; Piva et al. 2014; de Figueiredo et al. 2017; 20 Guenet et al. 2021; Rosenstock et al. 2014; Weindl et al. 2015; Thornton and Herrero 2015; Lal 2020; 21 Descheemaeker et al. 2016). The other systems outlined within this Box may form or facilitate IPS. 22 Organic Farming (OF) 23 OF can be considered a form of AE (Lampkin et al. 2017) though is discussed separately here as it is 24 guided by specific principles and associated regulations (Annex I). OF is perhaps noted more for 25 potential co-benefits, such as enhanced system resilience and biodiversity promotion, than mitigation. 26 Several studies have reviewed the emissions footprint of organic compared to conventional systems 27 (e.g. Mondelaers et al. 2009; Tuomisto et al. 2012; Skinner et al. 2014; Meier et al. 2015; Seufert and 28 Ramankutty 2017; Clark and Tilman 2017; Meemken and Qaim 2018; Bellassen et al. 2021). 29 Acknowledging potential assessment limitations (Meier et al. 2015; van der Werf et al. 2020), evidence 30 suggests organic production to typically generate lower emissions per unit of area, while emissions per 31 unit of product vary and depend on the product (high agreement, medium evidence). OF has been 32 suggested to increase soil carbon sequestration (Gattinger et al. 2012), though definitive conclusions 33 are challenging (Leifeld et al. 2013). Fewer studies consider impacts of large-scale conversion from 34 conventional to organic production globally. Though context specific (Seufert and Ramankutty 2017), 35 OF is reported to typically generate lower yields (Seufert et al. 2012; De Ponti et al. 2012; Kirchmann 36 2019; Biernat et al. 2020). Large-scale conversion, without fundamental changes in food systems and 37 diets (Muller et al. 2017; Theurl et al. 2020), may lead to increases in absolute emissions from land use 38 change, driven by greater land requirements to maintain production (e.g. Leifeld 2016; Meemken and 39 Qaim 2018). OF may have mitigation capacity in certain instances though impacts of large-scale 40 conversion require further research. 41 [END BOX 7.5 HERE] 42 43 [START BOX 7.6 HERE] 44 Box 7.6 Case study: Mitigation Options and Costs in the Indian Agricultural Sector Do Not Cite, Quote or Distribute 7-75 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Objective 2 To assess the technical mitigation potentials of Indian agriculture and costs under a Business as Usual 3 scenario (BAU) and Mitigation scenario up to 2030 (Sapkota et al. 2019). 4 Results 5 The study shows that by 2030 under BAU scenario GHG emissions from the agricultural sector in India 6 would be 515 MtCO2-eq yr-1 (using GWP100 and IPCC AR4 values) with a technical mitigation potential 7 of 85.5 MtCO2-eq yr-1 through the adoption of various mitigation practices. About 80% of the technical 8 mitigation potential could be achieved by adopting cost-saving measures. Three mitigation options, i.e., 9 efficient use of fertiliser, zero-tillage, and rice-water management, could deliver more than 50% of the 10 total technical abatement potential. Under the BAU scenario the projected GHG emissions from major 11 crop and livestock species is estimated at 489 MtCO2-eq in 2030, whereas under mitigation scenario 12 GHG emissions are estimated at 410 MtCO2-eq implying a technical mitigation option of about 78.67 13 MtCO2-eq yr-1 (Box 7.6, Figure 1). Major sources of projected emissions under the BAU scenario, in 14 order of importance, were cattle, rice, buffalo, and small ruminants. Although livestock production and 15 rice cultivation account for a major share of agricultural emissions, the highest mitigation potential was 16 observed in rice (about 36 MtCO2-eq yr-1) followed by buffalo (about 14 MtCO2-eq yr-1), wheat (about 17 11 MtCO2-eq yr-1) and cattle (about 7 MtCO2-eq yr-1). Crops such as cotton and sugarcane each offered 18 mitigation potential of about 5 MtCO2-eq yr-1 while the mitigation potential from small ruminants 19 (goat/sheep) was about 2 MtCO2-eq yr-1. 20 Sapkota et al. (2019) also estimated the magnitude of GHG savings per year through adoption of various 21 mitigation measures, together with the total cost and net cost per unit of CO2-eq abated. When the 22 additional benefits of increased yield due to adoption of the mitigation measures were considered, about 23 80% of the technical mitigation potential (67.5 out of 85.5 MtCO2-eq) could be achieved by cost-saving 24 measures. When yield benefits were considered, green fodder supplements to ruminant diets were the 25 most cost-effective mitigation measure, followed by vermicomposting and improved diet management 26 of small ruminants. Mitigation measures such as fertigation and micro-irrigation, various methods of 27 restoring degraded land and feed additives in livestock appear to be cost-prohibitive, even when 28 considering yield benefits, if any. The study accounted for GHG emissions at the farm level and 29 excluded emissions arising due to processing, marketing or consumption post farm-gate. It also did not 30 include emissions from feed production, since livestock in India mostly rely on crop by-products and 31 concentrates. Further the potential of laser land levelling seems exaggerated which may also be 32 redundant with already accounted potential from ‘improved water management in rice’. The mitigation 33 potential of agro-ecological approaches/technologies such as natural farming which is picking up in 34 India in recent years has also been overlooked. 35 Do Not Cite, Quote or Distribute 7-76 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Box 7.6, Figure 1 Contribution of various crops and livestock species to total agricultural emission in 2012 3 (baseline) and by 2030 under business as usual (BAU) and mitigation scenarios for Indian Agricultural 4 sector. Source: Sapkota et al. (2019). 5 [END BOX 7.6 HERE] 6 7 7.4.4. Bioenergy and BECCS 8 Activities, co-benefits, risks and implementation opportunities and barriers. Bioenergy refers to 9 energy products (solid, liquid and gaseous fuels, electricity, heat) derived from multiple biomass 10 sources including organic waste, harvest residues and by-flows in the agriculture and forestry sectors, 11 and biomass from tree plantations, agroforestry systems, lignocellulosic crops, and conventional 12 food/feed crops. It may reduce net GHG emissions by displacing the use of coal, oil and natural gas 13 with renewable biomass in the production of heat, electricity, and fuels. When combined with carbon 14 capture and storage (BECCS) and biochar production, bioenergy systems may provide CDR by durably 15 storing biogenic carbon in geological, terrestrial, or ocean reservoirs, or in products, further contributing 16 to mitigation (Section 7.4.3.2, Chapters 3, 4, 6 and 12) (Chum et al. 2011; Hammar and Levihn 2020; 17 Emenike et al. 2020; Cabral et al. 2019; Moreira et al. 2020b; Wang et al. 2020: Johnsson et al. 2020). 18 This section addresses especially aspects related to land use and biomass supply for bioenergy and 19 BECCS. The mitigation potential presented here and in Table 7.3, includes only the CDR component 20 of BECCS. The additional mitigation achieved from displacing fossil fuels is covered elsewhere 21 (Chapters 6, 8, 9, 10, 11 and 12). 22 Modern bioenergy systems (as opposed to traditional use of fuelwood and other low-quality cooking 23 and heating fuels) currently provide approximately 30 EJ yr-1 of primary energy, making up 53% of 24 total renewable primary energy supply (IEA 2019). Bioenergy systems are commonly integrated within 25 forest and agriculture systems that produce food, feed, lumber, paper and other biobased products. They 26 can also be combined with other AFOLU mitigation options: deployment of energy crops, agroforestry 27 and A/R can provide biomass while increasing land carbon stocks (Sections 7.4.2.2 and 7.4.3.3) and 28 anaerobic digestion of manure and wastewater, to reduce methane emissions, can produce biogas and 29 CO2 for storage (Section 7.4.3.7). But ill-deployment of energy crops can also cause land carbon losses 30 (Hanssen et al. 2020) and increased biomass demand for energy could hamper other mitigation 31 measures such as reduced deforestation and degradation (Sections 7.4.2.1). Do Not Cite, Quote or Distribute 7-77 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Bioenergy and BECCS can be associated with a range of co-benefits and adverse side-effects (Section 2 12.5; Jia et al. 2019; Calvin et al. 2021; Smith et al. 2016). It is difficult to disentangle bioenergy 3 development from the overall development in the AFOLU sector given its multiple interactions with 4 food, land, and energy systems. It is therefore not possible to precisely determine the scale of bioenergy 5 and BECCS deployment at which negative impacts outweigh benefits. Important uncertainties include 6 governance systems, future food and biomaterials demand, land use practices, energy systems 7 development, climate impacts, and time scale considered when weighing negative impacts against 8 benefits (SRCCL, Cross-Chapter Box 7; Box 7.7). (Turner et al. 2018b; Daioglou et al. 2019; Kalt et 9 al. 2020; Wu et al. 2019; Robledo-Abad et al. 2017; Hanssen et al. 2020; Calvin et al, 2021;Cowie et 10 al. 2021). The use of municipal organic waste, harvest residues, and biomass processing by-products as 11 feedstock is commonly considered to have relatively lower risk, provided that associated land use 12 practices are sustainable (Cowie et al. 2021). Deployment of dedicated biomass production systems can 13 have positive and negative implications on mitigation and other sustainability criteria, depending on 14 location and previous land use, feedstock, management practice, and deployment strategy and scale 15 (Sections 12.5 and 17.3.3.1;(Popp et al. 2017; Humpenöder et al. 2018; Rulli et al. 2016; Brondizio et 16 al. 2019; Hasegawa et al. 2020; Fujimori et al. 2019; Drews et al. 2020; Schulze et al. 2020; Stenzel et 17 al. 2020; Daioglou et al. 2017; Staples et al. 2017; Carvalho et al. 2017; Mouratiadou et al. 2020; 18 Buchspies et al. 2020; Hanssen et al. 2020). 19 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 20 potential, costs, and pathways. Many more stringent mitigation scenarios in AR5 relied heavily on 21 bioenergy and BECCS. The SR1.5 reported a range for the CDR potential of BECCS (2100) at 0.5 to 5 22 GtCO2-eq yr-1 when applying constraints reflecting sustainability concerns, at a cost of 100-200 USD 23 tCO2-1 (Fuss et al. 2018). The SRCCL reported a technical CDR potential for BECCS at 0.4-11.3 GtCO2 24 yr-1 (medium confidence), noting that most estimates do not include socio-economic barriers, the 25 impacts of future climate change, or non-GHG climate forcing (IPCC. 2019a). The SR1.5 and SRCCL 26 highlighted that bioenergy and BECCS can be associated with multiple co-benefits and adverse side- 27 effects that are context specific. 28 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). The role of 29 bioenergy and BECCS in mitigation pathways has been reduced as IAM-based studies have 30 incorporated broader mitigation portfolios and have explored non-CO2 emissions reduction and a wider 31 variation of underlying assumptions about socio-economic drivers and associated energy and food 32 demand, as well as deployment limits such as land availability for A/R and for cultivation of crops used 33 for bioenergy and BECCS (Grubler et al. 2018; Van Vuuren et al. 2018). 34 Increased availability of spatially explicit data and advances in the modelling of crop productivity and 35 land use, land carbon stocks, hydrology, and ecosystem properties, have enabled more comprehensive 36 analyses of factors that influence the contribution of bioenergy and BECCS in IAM-based mitigation 37 scenarios, and also associated co-benefits and adverse side-effects (Wu et al, 2019, Li et al. 2020, Turner 38 et al 2018a, Hanssen et al. 2020; Ai et al. 2021; Drews et al. 2020; Hasegawa et al. 2021). Yet, IAMs 39 are still coarse in local land use practices. (Daioglou et al. 2019; Wu et al. 2019; Moreira et al. 2020b). 40 Literature complementary to IAM studies indicate opportunities for integration of biomass production 41 systems into agricultural landscapes (e.g., agroforestry, double cropping) to produce biomass while 42 achieving co-benefits (Section 12.5). Similarly, climate-smart forestry puts forward measures (Box 7.3) 43 adapted to regional circumstances in forest sectors, enabling co-benefits in nature conservation, soil 44 protection, employment and income generation, and provision of wood for buildings, bioenergy and 45 other biobased products (Nabuurs et al. 2017). 46 Studies have also investigated the extent and possible use of marginal, abandoned, and degraded lands, 47 and approaches to help restore the productive value of these lands (Elbersen et al. 2019; Awasthi et al. Do Not Cite, Quote or Distribute 7-78 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2017; Chiaramonti and Panoutsou, 2018; Fernando et al. 2018; Rahman et al. 2019; Fritsche et al. 2017; 2 Næss et al. 2021). In the SRCCL, the presented range for degraded or abandoned land was 32 - 1400 3 Mha (Jia et al. 2019). Recent regional assessments not included in the SRCCL found up to 69 Mha in 4 EU-28, 185 Mha in China, 9.5 Mha in Canada, and 127 Mha in the USA (Elbersen et al. 2019; Zhang 5 et al 2020; Emery et al. 2017; Liu et al. 2017;Vera et al. 2021). The definition of 6 marginal/abandoned/degraded land, and the methods used to assess such lands remain inconsistent 7 across studies (Jiang et al. 2019), causing large variation amongst them (Jiang et al. 2021). Furthermore, 8 the availability of such lands has been contested since they may serve other functions (subsistence, 9 biodiversity protection, etc.) (Baka 2014). 10 11 [START BOX 7.7 HERE] 12 Box 7.7 Climate change mitigation value of bioenergy and BECCS 13 Besides emissions, and possible avoided emissions, related to the supply chain, the GHG effects of 14 using bioenergy depend on: (i) change in GHG emissions when bioenergy substitutes another energy 15 source; and (ii) how the associated land use and possible land use change influence the amount of carbon 16 that is stored in vegetation and(Calvin et al. 2021) soils over time. Studies arrive at varying mitigation 17 potentials for bioenergy and BECCS due to the large diversity of bioenergy systems, and varying 18 conditions concerning where and how they are deployed (Cowie et al. 2021; Elshout 2015; Harper et al 19 2018; Kalt et al 2019; Muri 2018; Brandão et al. 2019; Buchspies et al. 2020; Calvin et al. 2021). 20 Important factors include feedstock type, land management practice, energy conversion efficiency, type 21 of bioenergy product (and possible co-products), emissions intensity of the products being displaced, 22 and the land use/cover prior to bioenergy deployment ( Zhu et al. 2017; Hanssen et al. 2020; Staples et 23 al. 2017; Daioglou et al. 2017; Carvalho et al. 2017; Mouratiadou et al. 2020). Studies arrive at 24 contrasting conclusions also when similar bioenergy systems and conditions are analysed, due to 25 different methodologies, assumptions, and parameterisation (Harper et al 2018; Kalt et al 2019; Brandão 26 et al. 2019; Albers et al. 2019; Buchspies et al. 2020; Bessou et al. 2020; Rolls and Forster 2020; Cowie 27 et al. 2021). 28 Box 7.7 Figure 1 shows emissions associated with biomass supply (residues and crops grown on 29 cropland not needed for food) in 2050, here designated emission-supply curves. The curves are 30 constructed assuming that additional biomass supply consistently comes from the available 31 land/biomass resource that has the lowest GHG emissions, i.e., the marginal GHG emissions increase 32 with increasing biomass use for bioenergy. Net negative emissions indicate cases where biomass 33 production increases land carbon stocks. 34 One curve (EMF-33) is determined from stylised scenarios using IAMs (Rose et al. 2020). One of the 35 two curves determined from sectoral models, Constant Land Cover, reflects supply chain emissions and 36 changes in land carbon storage caused by the biomass supply system itself. These two curves are 37 obtained with modelling approaches compatible with the modelling protocol used for the scenarios in 38 the AR6 database, which accounts for the land-use change and all other GHG emissions along a given 39 transformation trajectory, enabling assessments of the warming level incurred. 40 The Natural Regrowth curve attribute additional “counterfactual emissions” to the bioenergy system, 41 corresponding to estimated uptake of CO2 in a counterfactual scenario where land is not used for 42 bioenergy but instead subject to natural vegetation regrowth. This curve does not show actual emissions 43 from the bioenergy system, but it provides insights in the mitigation value of the bioenergy option 44 compared to alternative land-use strategies. To illustrate, if biomass is used instead of a primary energy 45 source with emission factor 75 kg CO2 GJ-1, and the median values in the Natural Regrowth curve are Do Not Cite, Quote or Distribute 7-79 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 adopted, then the curve indicates that up to about 150 EJ of biomass can be produced and used for 2 energy while achieving higher net GHG savings than the alternative to set aside the same land for natural 3 vegetation regrowth (assuming same conversion factor). 4 The large ranges in the bars signify the importance of uncertainties and how the biomass is deployed. 5 Variation in energy conversion efficiencies and uncertainty about magnitude, timing, and permanence 6 of land carbon storage further complicate the comparison. Finally, not shown in Box 7.7 Figure 1, the 7 emission-supply curves would be adjusted downwards if displacement of emission intensive energy 8 was included or if the bioenergy is combined with CCS to provide CDR. 9 10 11 Box 7.7, Figure 1 Emissions associated with primary biomass supply in 2050 (residues and crops grown 12 on cropland not needed for food), as determined from sectoral models (Daioglou et al. 2017; Kalt et al. 13 2020), and stylised scenarios from the EMF-33 project using Integrated Assessment Models (Rose et al. 14 2020). All methods include LUC (direct and indirect) emissions. Emissions associated with Natural 15 Regrowth include counterfactual carbon fluxes (see text). The sectoral models include a more detailed 16 representation of the emissions, including Life-Cycle emissions from fertiliser production. IAM models 17 may include economic feedbacks such as intensification as a result of increasing prices. As an indication: 18 for natural gas the emission factor is around 56, for coal around 95 kg CO 2 GJ-1. 19 [END BOX 7.7 HERE] 20 21 Critical assessment and conclusion. Recent estimates of technical biomass potentials constrained by 22 food security and environmental considerations fall within previous ranges corresponding to medium 23 agreement, (e.g., Turner et al. 2018b; Daioglou et al. 2019; Wu et al. 2019, Hansen et al 2020; Kalt et 24 al. 2020) arriving at 4-57 and 46-245 EJ yr-1 by 2050 for residues and dedicated biomass crops, 25 respectively. Based on studies to date, the technical net CDR potential of BECCS (including LUC and 26 other supply chain emissions, but excluding energy carrier substitution) by 2050 is 5.9 (0.5-11.3) GtCO2 27 yr-1 globally, of which 1.6 (0.5-3.5) GtCO2 yr-1 is available at below USD100 tCO2-1 (medium 28 confidence) (Figure 7.11) (Lenton 2010; Koornneef et al. 2012; McLaren 2012; Powell and Lenton 29 2012; Fuss et al. 2018; Turner et al. 2018a; Hanssen et al. 2020; Roe et al. 2021). The equivalent 30 economic potential as derived from IAMs is 1.8 (0.2 - 9.9) GtCO2 yr-1 (Table 7.3). 31 Do Not Cite, Quote or Distribute 7-80 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Technical land availability does not imply that dedicated biomass production for bioenergy and BECCS 2 is the most effective use of this land for mitigation. Further, implications of deployment for climate 3 change mitigation and other sustainability criteria are context dependent and influenced by many 4 factors, including rate and total scale. While governance has a critical influence on outcome, larger 5 scale and higher expansion rate generally translates into higher risk for negative outcomes for GHG 6 emissions, biodiversity, food security and a range of other sustainability criteria (Rochedo et al. 2018; 7 Daioglou et al. 2019; Junginger et al. 2019; Galik et al. 2020; Searchinger 2017; Vaughan et al. 2018; 8 de Oliveira Garcia et al. 2018; Stenzel et al. 2020). 9 However, literature has also highlighted how the agriculture and forestry sectors may respond to 10 increasing demand by devising management approaches that enable biomass production for energy in 11 conjunction with supply of food, construction timber, and other biobased products, providing climate 12 change mitigation while enabling multiple co-benefits including for nature conservation (Parodi et al. 13 2018; Springmann et al. 2018; Nabuurs et al. 2017; Rosenzweig et al. 2020; Clark et al. 2020; Favero 14 et al. 2020; Hanssen et al. 2020; Section 7.4 and Cross-Working Group Box 3 in Chapter 12). 15 Strategies to enhance the benefits of bioenergy and BECCS include (i) management practices that 16 protect carbon stocks and the productive and adaptive capacity of lands, as well as their environmental 17 and social functions (van Ittersum et al. 2013, Gerssen-Gondelach et al. 2015; Moreira et al. 2020b) (ii) 18 supply chains from primary production to final consumption that are well managed and deployed at 19 appropriate levels (Donnison et al. 2020; Fajardy et al. 2018); and (iii) development of a cross-sectoral 20 agenda for biobased production within a circular economy, and international cooperation and 21 governance of global trade in products to maximize synergies while limiting trade-offs concerning 22 environmental, economic and social outcomes (very high confidence). Finally, the technical feasibility 23 of BECCS depends on investments in and the roll-out of advanced bioenergy technologies currently not 24 widely available (Daioglou et al. 2020b, Baker et al 2015). 25 7.4.5. Demand-side measures 26 7.4.5.1. Shift to sustainable healthy diets 27 Activities, co-benefits, risks and implementation opportunities and barriers. The term ‘Sustainable 28 healthy diets’ refers to dietary patterns that ‘promote all dimensions of individuals’ health and 29 wellbeing; have low environmental pressure and impact; are accessible, affordable, safe and equitable; 30 and are culturally acceptable’ (FAO and WHO 2019). In addition to climate mitigation gains, a 31 transition towards more plant-based consumption and reduced consumption of animal-based foods, 32 particularly from ruminant animals, could reduce pressure on forests and land used for feed, support the 33 preservation of biodiversity and planetary health (Theurl et al. 2020; FAO 2018c), and contribute to 34 preventing forms of malnutrition (i.e. undernutrition, micronutrient deficiency, and obesity) in 35 developing countries (Chapter 12, Section 12.4.). Other co-benefits include lowering the risk of 36 cardiovascular disease, type 2 diabetes, and reducing mortality from diet-related non-communicable 37 diseases (Toumpanakis et al. 2018; Satija and Hu 2018; Faber et al. 2020; Magkos et al. 2020). 38 However, transition towards sustainable healthy diets could have adverse impacts on the economic 39 stability of the agricultural sector (MacDiarmid 2013; Aschemann-Witzel 2015; Van Loo et al. 2017). 40 Therefore, shifting toward sustainable and healthy diets requires effective food-system oriented reform 41 policies that integrate agriculture, health, and environment policies to comprehensively address 42 synergies and conflicts in co-lateral sectors (agriculture, trade, health, environment protection etc.) and 43 capture spill-over effects (climate change, biodiversity loss, food poverty) (Galli et al. 2020; FAO and 44 WHO 2019). 45 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 46 potential, costs, and pathways. According to the AR5, changes in human diets and consumption Do Not Cite, Quote or Distribute 7-81 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 patterns can reduce emissions 5.3 to 20.2 GtCO2-eq yr-1 by 2050 from diverted agricultural production 2 and avoided land-use change (Smith et al. 2014). In the SRCCL, a “contract and converge” model of 3 transition to sustainable healthy diets was suggested as an effective approach, reducing food 4 consumption in over-consuming populations and increasing consumption of some food groups in 5 populations where minimum nutritional needs are not met (Smith et al. 2019b). The total technical 6 mitigation potential of changes in human diets was estimated as 0.7 - 8 GtCO2-eq yr–1 by 2050 (SRCCL, 7 Chapter 2 and 6; (Springmann et al. 2016; Hawken 2017; Tilman and Clark 2014), ranging from a 50% 8 adoption of healthy diets (<60g of animal-based protein) and only accounting for diverted agricultural 9 production, to the global adoption of a vegetarian diet. 10 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Since the 11 SRCCL, global studies continue to find high mitigation potential from reducing animal-source foods 12 and increasing proportions of plant-rich foods in diets. Springmann et al. (2018) estimated that diet 13 changes in line with global dietary guidelines for total energy intake and consumption of red meat, 14 sugar, fruits, and vegetables, could reduce GHG emissions by 29% and other environmental impacts by 15 5–9% compared with the baseline in 2050. Poore and Nemecek (2018) revealed that shifting towards 16 diets that exclude animal-source food could reduce land use by 3.1 billion ha, decrease food-related 17 GHG emissions by 6.5 GtCO2-eq yr–1, acidification by 50%, eutrophication by 49%, and freshwater 18 withdrawals by 19% for a 2010 reference year. Frank et al. (2019) estimated non-CO2 emissions 19 reductions of 0.4 GtCO2-eq yr–1 at a carbon price of USD100 tCO2-1 and 0.6 GtCO2-eq yr–1 at USD20 20 tCO2-1 in 2050 from shifting to lower animal-source diets (430 kcal of livestock calorie intake) in 21 developed and emerging countries. From a systematic literature review, Ivanova et al. (2020) found 22 mitigation potentials of 0.4-2.1 tCO2-eq capita-1 for a vegan diet, of 0.01-1.5 for a vegetarian diet, and 23 of 0.1-2.0 for Mediterranean or similar healthy diet. 24 Regionally, mitigation potentials for shifting towards sustainable healthy diets (50% convergence to 25 <60g of meat-based protein, only accounting for diverted production) vary across regions. A recent 26 assessment finds greatest economic (up to USD100 tCO2-1) potential for 2020-2050 in Asia and the 27 developing Pacific (609 MtCO2-eq yr-1) followed by Developed Countries (322 MtCO2-eq yr-1) based 28 on IPCC AR4 GWP100 values for CH4 and N2O) (Roe et al. 2021). In the EU, (Latka et al. 2021) found 29 that moving to healthy diets through price incentives could bring about annual reductions of non-CO2 30 emissions from agriculture of 12-111 MtCO2-eq yr-1. At the country level, recent studies show that 31 following National Dietary Guidelines (NDG) would reduce food system GHG emissions by 4–42%, 32 confer large health gains (1.0–1.5 million quality-adjusted life-years) and lower health care system costs 33 (NZD 14–20 billion) in New Zealand Drew et al. (2020); reduce 28% of GHG emissions in Argentina 34 Arrieta and González (2018); about 25% in Portugal Esteve-Llorens et al. (2020) and reduce GHG 35 emissions, land use and blue water footprint by 15–60% in Spain Batlle-Bayer et al. (2020). In contrast, 36 Aleksandrowicz et al. (2019) found that meeting healthy dietary guidelines in India required increased 37 dietary energy intake overall, which slightly increased environmental footprints by about 3–5% across 38 GHG emissions, blue and green water footprints and land use. 39 Critical assessment and conclusion. Shifting to sustainable healthy diets has large potential to achieve 40 global GHG mitigation targets as well as public health and environmental benefits (high confidence). 41 Based on studies to date, there is medium confidence that shifting toward sustainable healthy diets has 42 a technical potential including savings in the full value chain of 3.6 (0.3-8.0) GtCO2-eq yr-1 of which 43 2.5 (1.5-3.9) GtCO2-eq yr-1 is considered plausible based on a range of GWP100 values for CH4 and N2O. 44 When accounting for diverted agricultural production only, the feasible potential is 1.7 (1 – 2.7) GtCO2- 45 eq yr-1. A shift to more sustainable and healthy diets is generally feasible in many regions (medium 46 confidence). However, potential varies across regions as diets are location- and community- specific, 47 and thus may be influenced by local production practices, technical and financial barriers and associated Do Not Cite, Quote or Distribute 7-82 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 livelihoods, everyday life and behavioural and cultural norms around food consumption (Meybeck and 2 Gitz 2017; Creutzig et al. 2018; FAO 2018b). Therefore, a transition towards low-GHG emission diets 3 and achieving their mitigation potential requires a combination of appropriate policies, financial and 4 non-financial incentives and awareness-raising campaigns to induce changes in consumer behaviour 5 with potential synergies between climate objectives, health and equity (Rust et al. 2020). 6 7.4.5.2. Reduce food loss and waste 7 Activities, co-benefits, risks and implementation opportunities and barriers. Food loss and waste 8 (FLW) refer to the edible parts of plants and animals produced for human consumption that are not 9 ultimately consumed (UNEP 2021b). Food loss occurs through spoilage, spilling or other unintended 10 consequences due to limitations in agricultural infrastructure, storage and packaging (Parfitt et al. 2010). 11 Food waste typically takes place at the distribution (retail and food service) and consumption stages in 12 the food supply chain and refers to food appropriate for human consumption that is discarded or left to 13 spoil (HLPE 2014). Options that could reduce FLW include: investing in harvesting and post-harvesting 14 technologies in developing countries, taxing and other incentives to reduce business and consumer- 15 level waste in developed countries, mandatory FLW reporting and reduction targets for large food 16 businesses, regulation of unfair trading practices, and active marketing of cosmetically imperfect 17 products (van Giesen and de Hooge 2019; Sinclair Taylor et al. 2019). Other studies suggested 18 providing options of longer-lasting products and behavioural changes (e.g. through information 19 provision) that cause dietary and consumption changes and motivate consumers to actively make 20 decisions that reduce FLW. Reductions of FLW along the food chain bring a range of benefits beyond 21 GHG mitigation, including reducing environmental stress (e.g. water and land competition, land 22 degradation, desertification), safeguarding food security, and reducing poverty (Galford et al. 2020; 23 Venkat et al. 2020). Additionally, FLW reduction is crucial for achieving SDG 12 which calls for 24 ensuring ‘sustainable consumption and production patterns’ through lowering per capita global food 25 waste by 50% at the retail and consumer level and reducing food losses along food supply chains by 26 2030. In line with these SDG targets, it is estimated that reducing FLW can free up several million km2 27 of land (high confidence). The interlinkages between reducing FLW and food system sustainability are 28 discussed in Chapter 12. Recent literature identifies a range of barriers to climate change mitigation 29 through FLW reduction, which are linked to technological, biophysical, socio-economic, financial and 30 cultural contexts at regional and local levels (Blok et al. 2020; Vogel and Meyer 2018; Gromko and 31 Abdurasalova 2019; Rogissart et al. 2019). Examples of these barriers include infrastructural and 32 capacity limitations, institutional regulations, financial resources, constraining resources (e.g. energy), 33 information gaps (e.g. with retailers), and consumers’ behaviour (Blok et al. 2020; Gromko and 34 Abdurasalova 2019). 35 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 36 potential, costs, and pathways. In AR5, reduced FLW was considered as a mitigation measure that 37 could substantially lower emissions, with estimated mitigation potential of 0.6–6.0 GtCO2-eq yr–1 in the 38 food supply chain (Smith et al. 2014). In the SRCCL, the technical mitigation potential of reducing food 39 and agricultural waste was estimated at 0.76-4.5 GtCO2-eq yr–1 (SRCCL, Chapter 2 and 6: Bajželj et al. 40 2014; Dickie et al. 2014b; Hawken 2017). 41 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Since the 42 SRCCL, there have been very few quantitative estimates of the mitigation potential of FLW reductions. 43 Evidence suggests that reducing FLW together with overall food intake could have substantial 44 mitigation potential, equating to an average of 0.3 tCO2-eq capita-1 (Ivanova et al. 2020). Some regional 45 sectoral studies indicate that reducing FLW in the EU can reduce emissions by 186 MtCO2-eq yr–1, the 46 equivalent of around 15% of the environmental impacts (climate, acidification, and eutrophication) of 47 the entire food value chain (Scherhaufer et al. 2018). In the UK, disruptive low-carbon innovations Do Not Cite, Quote or Distribute 7-83 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 relating to FLW reduction were found to be associated with potential emissions reductions ranging 2 between 2.6 and 3.6 MtCO2-eq (Wilson et al. 2019). Other studies investigated the effect of tax 3 mechanisms, such as ‘pay as you throw’ for household waste, on the mitigation potential of reducing 4 FLW. Generally, these mechanisms are recognised as particularly effective in reducing the amount of 5 waste and increasing the recycling rate of households (Carattini et al. 2018; Rogissart et al. 2019). 6 Technological FWL mitigation opportunities exist throughout the food supply chain; post-harvest 7 opportunities for FLW reductions are discussed in Chapter 12. Based on IPCC AR4 GWP100 values for 8 CH4 and N2O, greatest economic mitigation potential (up to USD100 tCO2-1) for the period 2020-2050 9 from FLW reduction is estimated to be in Asia and developing Pacific (192.3 GtCO2-eq yr-1) followed 10 by Developed Countries (101.6 GtCO2-eq yr-1) (Roe et al. 2021). These estimates reflect diverted 11 agricultural production and do not capture potential from avoided land-use changes. 12 Critical assessment and conclusion. There is medium confidence that reduced FLW has large global 13 technical mitigation potential of 2.1 (0.1–5.8) GtCO2-eq yr–1 including savings in the full value chain 14 and using GWP100 and a range of IPCC values for CH4 and N2O. Potentials at 3.7 (2.2–5.1) GtCO2-eq 15 yr-1 are considered plausible. When accounting for diverted agricultural production only, the feasible 16 potential is 0.5 (0.0–0.9) GtCO2-eq yr-1. See the section above for the joint land use effects of food 17 related demand-side measures which increases three-fold when accounting for the land-use effects as 18 well. But this would overlap with other measures and is therefore not additive. Regionally, FLW 19 reduction is feasible anywhere but its potential needs to be understood in a wider and changing socio- 20 cultural context that determines nutrition (high confidence). 21 7.4.5.3. Improved and enhanced use of wood products 22 Activities, co-benefits, risks and implementation opportunities and barriers. The use of wood products 23 refers to the fate of harvested wood for material uses and includes two distinctly different components 24 t affecting the carbon cycle, including carbon storage in wood products and material substitution. When 25 harvested wood is used for the manufacture of wood products, carbon remains stored in these products 26 depending on their end use and lifetime. Carbon storage in wood products can be increased through 27 enhancing the inflow of products in use, or effectively reducing the outflow of the products after use. 28 This can be achieved through additional harvest under sustainable management (Pilli et al. 2015; 29 Johnston and Radeloff 2019), changing the allocation of harvested wood to long-lived wood products 30 or by increasing products’ lifetime and increasing recycling (Brunet-Navarro et al. 2017; Jasinevičius 31 et al. 2017; Xu et al. 2018; Xie et al. 2021). Material substitution involves the use of wood for building, 32 textiles, or other applications instead of other materials (e.g. concrete, steel which consume more energy 33 to produce) to avoid or reduce emissions associated with the production, use and disposal of those 34 products it replaces. 35 The benefits and risks of improved and enhanced improved use of wood products are closely linked to 36 forest management. First of all, the enhanced use of wood products could potentially activate or lead to 37 improved sustainable forest management that can mitigate and adapt (Verkerk et al. 2020). Secondly, 38 carbon storage in wood products and the potential for substitution effects can be increased by additional 39 harvest, but with the risk of decreasing carbon storage in forest biomass when not done sustainably 40 (Smith et al. 2019b). Conversely, reduced harvest may lead to gains in carbon storage in forest 41 ecosystems locally, but these gains may be offset through international trade of forest products causing 42 increased harvesting pressure or even degradation elsewhere (Pendrill et al. 2019b; Kastner et al. 2011; 43 Kallio and Solberg 2018). There are also environmental impacts associated with the processing, 44 manufacturing, use and disposal of wood products (Adhikari and Ozarska 2018; Baumgartner 2019). 45 See Section 9.6.4 of this report for additional discussion on benefits and risks. Do Not Cite, Quote or Distribute 7-84 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Conclusions from AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL); mitigation 2 potential, costs, and pathways. There is strong evidence at the product level that wood products from 3 sustainably managed forests are associated with less greenhouse emissions in their production, use and 4 disposal over their life-time compared to products made from emission-intensive and non-renewable 5 materials. However, there is still limited understanding of the substitution effects at the level of markets, 6 countries (Leskinen et al. 2018). AR5 did not report on the mitigation potential of wood products. The 7 SRCCL (Chapters 2 and 6) finds that some studies indicate significant mitigation potentials for material 8 substitution, but concludes that the global, technical mitigation potential for material substitution for 9 construction applications ranges from 0.25-1 GtCO2-eq yr–1 (medium confidence) (McLaren 2012; 10 Miner 2010; Roe et al. 2019). 11 Developments since AR5 and IPCC Special Reports (SR1.5, SROCCC and SRCCL). Since the 12 SRCCL, several studies have examined the mitigation potential of the enhanced and improved use of 13 wood products. A global forest sector modelling study (Johnston and Radeloff 2019) estimated that 14 carbon storage in wood products represented a net carbon stock increase of 0.34 GtCO2-eq yr-1 globally 15 in 2015 and which could provide an average mitigation potential (by increasing the HWP pool) of 0.33– 16 0.41 GtCO2-eq yr-1 for the period 2020–2050, based on the future socio-economic development (SSP 17 scenarios) and its effect on the production and consumption of wood products. Traded feedstock 18 provided another 0.071 GtCO2 yr-1 of carbon storage in 2015 and 0.12 GtCO2 yr-1 by 2065. These 19 potentials exclude the effect of material substitution. Another recent study estimated the global 20 mitigation potential of mid-rise urban buildings designed with engineered wood products at 0.04-3.7 21 GtCO2 yr-1 (Churkina et al. 2020). Another study (Oliver et al. 2014) estimated that using wood to 22 substitute for concrete and steel as building materials could provide a technical mitigation potential of 23 0.78-1.73 GtCO2 yr-1 achieved through carbon storage in wood products and through material and 24 energy substitution. 25 The limited availability or absence of estimates of the future mitigation potential of improved use of 26 wood products for many world regions represents an important knowledge gap, especially with regards 27 to material substitution effects. At the product level, wood products are often associated with lower 28 fossil-based emissions from production, use and disposal, compared to products made from emission- 29 intensive and non-renewable materials (Sathre and O’Connor 2010; Geng et al. 2017; Leskinen et al. 30 2018). 31 Critical assessment and conclusion. Based on studies to date, there is strong evidence and medium 32 agreement that the improved use of wood products has a technical potential of 1.0 (0.04–3.7) GtCO2- 33 eq yr-1 and economic potential of 0.4 (0.3–0.5) GtCO2-eq yr-1.There is strong evidence and high 34 agreement at the product level that material substitution provides on average benefits for climate change 35 mitigation as wood products are associated with less fossil-based GHG emissions over their lifetime 36 compared to products made from emission-intensive and non-renewable materials. However, the 37 evidence at the level of markets or countries is uncertain and fairly limited for many parts of the world. 38 There is medium confidence that material substitution and carbon storage in wood products contribute 39 to climate change mitigation when also the carbon balances of forest ecosystems are considered of 40 sustainably managed large areas of forests in medium term. The total future mitigation potential will 41 depend on the forest system considered, the type of wood products that are produced and substituted 42 and the assumed production technologies and conversion efficiencies of these products. 43 Do Not Cite, Quote or Distribute 7-85 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 7.5. AFOLU Integrated Models and Scenarios 3 This section assesses the literature and data available on potential future GHG dynamics in the AFOLU 4 sector, the cost-effectiveness of different mitigation measures, and consequences of climate change 5 mitigation pathways on land-use dynamics as well as relevant sustainable development indicators at the 6 regional and global level based on global integrated models. 7 Land-based mitigation options interact and create various trade-offs, and thus need to be assessed 8 together as well as with mitigation options in other sectors, and in combination with other sustainability 9 goals (Popp et al. 2014; Obersteiner et al. 2016; Roe et al. 2019; Van Vuuren et al. 2019; Prudhomme 10 et al. 2020; Strefler et al. 2021). The assessments of individual mitigation measures or sectoral estimates 11 used to estimate mitigation potential in Section 7.4, when aggregated together, do not account for 12 interactions and trade-offs. Integrative land-use models (ILMs) combine different land-based mitigation 13 options and are partially included in Integrated Assessment Models (IAMs) which combine insights 14 from various disciplines in a single framework and cover the largest sources of anthropogenic GHG 15 emissions from different sectors. Over time, ILMs and IAMs have extended their system coverage 16 (Johnson et al. 2019). However, the explicit modelling and analysis of integrated land-use systems is 17 relatively new compared to other sectoral assessments such as the energy system (Jia et al. 2019). 18 Consequently, ILMs as well as IAMs differ in their portfolio and representation of land-based 19 mitigation options, the representation of sustainability goals other than climate action as well as the 20 interplay with mitigation in other sectors (van Soest et al. 2019; Johnson et al. 2019). These structural 21 differences have implications for the regional and global deployment of different mitigation options as 22 well as their sustainability impacts. 23 As a consequence of the relative novelty of land-based mitigation assessment in ILMs and IAMs, the 24 portfolio of land-based mitigation options does not cover the full option space as outlined in Section 25 7.4. The inclusion and detail of a specific mitigation measure differs across models. Land based 26 mitigation options are only partially included in ILM and IAM analyses, which mostly rely on 27 afforestation/reforestation and bioenergy with CCS (BECCS). Most ILM and IAM scenarios are based 28 on the Shared Socio-economic Pathways (SSPs) (Riahi et al. 2017), which is a set of contrasting future 29 scenarios widely used in the research community such as in the CMIP6 exercise, the SRCCL and the 30 IPBES global assessment. However, the coverage of land-based mitigation options in these scenarios 31 is mostly limited to dietary changes, higher efficiency in food processing (especially in livestock 32 production systems), reduction of food waste, increasing agricultural productivity, methane reductions 33 in rice paddies, livestock and grazing management for reduced methane emissions from enteric 34 fermentation, manure management, improvement of N-efficiency, international trade, first generation 35 of biofuels, avoided deforestation, afforestation, bioenergy and BECCS (Van Meijl et al. 2018; Popp et 36 al. 2017; Frank et al. 2019). Hence, there are mitigation options not being broadly included in integrated 37 pathway modelling as soil carbon, forest management, agroforestry or wetland management 38 (Humpenöder et al. 2020) which have the potential to alter the contribution of land-based mitigation in 39 terms of timing, potential and sustainability consequences (Frank et al. 2017). 40 7.5.1. Regional GHG emissions and land dynamics 41 In most of the assessed mitigation pathways, the AFOLU sector is of great importance for climate 42 change mitigation as it (i) turns from a source into a sink of atmospheric CO 2 due to large-scale 43 afforestation and reforestation, (ii) provides high amounts of biomass for bioenergy with or without 44 CCS and (iii), even under improved agricultural management, still causes residual non-CO2 emissions 45 from agricultural production and (iv) interplays with sustainability dimensions other than climate action Do Not Cite, Quote or Distribute 7-86 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 (Popp et al. 2017; Rogelj et al. 2017; Van Vuuren et al. 2018; Frank et al. 2018; van Soest et al. 2019; 2 Hasegawa et al. 2018). Regional AFOLU GHG emissions in scenarios with <4°C warming in 2100 3 (scenario category C7), as shown in Figure 7.13, are shaped by considerable CH4 and N2O emissions 4 throughout 2050 and 2100, mainly from ASIA and AFRICA. CH4 emissions from enteric fermentation 5 are largely caused by ASIA, followed by AFRICA, while CH4 emissions from paddy rice production 6 are almost exclusively caused by ASIA. N2O emissions from animal waste management and soils are 7 more equally distributed across region. 8 In most regions, CH4 and N2O emission are both lower in mitigation pathways that limit warming to 9 <1.5°C, < 2°C and <3°C (C1-C6) compared to scenarios with <4°C (Popp et al. 2017; Rogelj et al. 10 2018a). In particular, the reduction of CH4 emissions from enteric fermentation in ASIA and AFRICA 11 is profound. Land-related CO2 emissions, which include emissions from deforestation as well as 12 removals from afforestation, are slightly negative (i.e. AFOLU systems turn into a sink) in <1.5°C, < 13 2°C and <3°C mitigation pathways compared to <4°C scenarios. Carbon sequestration via BECCS is 14 most prominent in ASIA, LAM, AFRICA and OECD90+EU, which are also the regions with the highest 15 bioenergy area. 16 17 18 19 Figure 7.13 Land-based regional GHG emissions and removals in 2050 (top) and 2100 (bottom) for 20 scenarios from the AR6 Database with <1.5°C (C1, C2), < 2°C (C3, C4), <3°C (C5, C6) and <4°C 21 (C7) global warming in 2100 (scenario type is indicated by colour). The categories shown include 22 CH4 emissions from enteric fermentation (EntF) and rice production (Rice), N 2O emissions from 23 animal waste management (AWM) and fertilisation (Soil). The category CO 2 Land includes CO2 24 emissions from land-use change as well as removals due to afforestation/reforestation. BECCS 25 reflects the CO2 emissions captured from bioenergy use and stored in geological deposits. The 26 annual GHG emission data from various models and scenarios is converted to CO 2 equivalents 27 using GWP factors of 27 for CH4 and 273 for N2O. The data is summarised in boxplots (Tukey 28 style), which show the median (vertical line), the interquartile range (IQR box) and the range of 29 values within 1.5 x IQR at either end of the box (horizontal lines) across all models and scenarios. Do Not Cite, Quote or Distribute 7-87 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 The number of data points available for each emission category, scenario type, region and year is 2 shown at the edge of each panel. Regional definitions: AFRICA = Sub-Saharan Africa, ASIA = 3 Asia, LAM = Latin America and Caribbean, MID_EAST = Middle East, OECD90+EU = OECD 90 4 and EU, REF = Reforming Economies of Eastern Europe and the Former Soviet Union. 5 6 Figure 7.14 indicates that regional land use dynamics in scenarios with <4°C warming in 2100 are 7 characterised by rather static agricultural land (i.e. cropland and pasture) in ASIA, LAM, OECD90+EU 8 and REF, and increasing agricultural land in AFRICA. Bioenergy area is relatively small in all regions. 9 Agricultural land in AFRICA expands at the cost of forests and other natural land. 10 The overall land dynamics in <1.5°C, < 2°C and <3°C mitigation pathways are shaped by land- 11 demanding mitigation options such as bioenergy and afforestation, in addition to the demand for other 12 agricultural and forest commodities. Bioenergy production and afforestation take place largely in the 13 (partly) tropical regions ASIA, LAM and AFRICA, but also in OECD90+EU. Land for dedicated 14 second generation bioenergy crops and afforestation displace agricultural land for food production 15 (cropland and pasture) and other natural land. For instance, in the <1.5°C mitigation pathway in ASIA, 16 bioenergy and forest area together increase by about 2.1 million km2 between 2020 and 2100, mostly at 17 the cost of cropland and pasture (median values). Such large-scale transformations of land use have 18 repercussions on biogeochemical cycles (e.g. fertiliser and water) but also on the economy (e.g. food 19 prices) and potential socio-political conditions. 20 21 22 23 Figure 7.14 Regional change of major land cover types by 2050 (top) and 2100 (bottom) relative to 24 2020 for scenarios from the AR6 Database with <1.5°C (C1, C2), < 2°C (C3, C4), <3°C (C5, C6) 25 and <4°C (C7) global warming in 2100 (scenario type is indicated by colour). The data is 26 summarised in boxplots (Tukey style), which show the median (vertical line), the interquartile 27 range (IQR box) and the range of values within 1.5 x IQR at either end of the box (horizontal lines) 28 across all models and scenarios. The number of data points available for each land cover type, Do Not Cite, Quote or Distribute 7-88 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 scenario type, region and year is shown at the right edge of each panel. Regional definitions: 2 AFRICA = Sub-Saharan Africa, ASIA = Asia, LAM = Latin America and Caribbean, MID_EAST 3 = Middle East, OECD90+EU = OECD 90 and EU, REF = Reforming Economies of Eastern Europe 4 and the Former Soviet Union. 5 6 7.5.2. Marginal abatement costs according to integrated assessments 7 In this section, Integrated Assessment Model (IAM) results from the AR6 database are used to derive 8 marginal abatement costs which indicate the economic mitigation potential for the different gases (N2O, 9 CH4, CO2) related to the AFOLU sector, at the global level and at the level of five world regions. This 10 review provides a complementary view on the economic mitigation potentials estimated in Section 7.4 11 by implicitly taking into account the interlinkages between the land-based mitigation options 12 themselves as well as the interlinkages with mitigation options in the other sectors such as BECCS. The 13 review systematically evaluates a range of possible economic potential estimates across gases, time, 14 and carbon prices. 15 For different models and scenarios from the AR6 database, the amount of mitigated emissions is 16 presented together with the respective carbon price (Figure 7.15). To determine mitigation potentials, 17 scenarios are compared to a benchmark scenario which usually assumes business-as-usual trends and 18 no explicit additional mitigation efforts. Scenarios have been excluded, if they do not have an associated 19 benchmark scenario or fail the vetting according to the AR6 scenario database, or if they do not report 20 carbon prices and CO2 emissions from AFOLU. Scenarios with contradicting assumptions (for example, 21 fixing some of the emissions to baseline levels) are excluded. Furthermore, only scenarios with 22 consistent3 regional and global level results are considered. Mitigation potentials are computed by 23 subtracting scenario specific emissions and sequestration amounts from their respective benchmark 24 scenario values. This difference accounts for the mitigation that can be credited to the carbon price 25 which is applied in a scenario. A few benchmark scenarios, however, apply already low carbon prices. 26 For consistency reasons, a carbon price that is applied in a benchmark scenario is subtracted from the 27 respective scenario specific carbon price. This may generate a bias because low carbon prices tend to 28 have a stronger marginal impact on mitigation than high carbon prices. Scenarios with carbon prices 29 which become negative due to the correction are not considered. The analysis considers all scenarios 30 from the AR6 database which pass the criteria and should be considered as an ensemble of opportunity 31 (Huppmann et al. 2018). 32 This approach is close to integrated assessment marginal abatement cost curves (MACCs) as described 33 in the literature (Frank et al. 2018, 2019; Harmsen et al. 2019; Fujimori et al. 2016) in the sense that it 34 incorporates besides the technical mitigation options also structural options, as well as behavioural 35 changes and market feedbacks. Furthermore, indirect emission changes and interactions with other 36 sectors can be highly relevant (Daioglou et al. 2019; Kalt et al. 2020) and are also accounted for in the 37 presented potentials. Hereby, some sequestration efforts can occur in other sectors, while leading to less 38 mitigation in the AFOLU sector. For instance, as an integral part of many scenarios, BECCS 39 deployment will lead to overall emissions reductions, and even provision of CDR as a result of the 40 interplay between three direct components i) LULUCF emissions/sinks, ii) reduction of fossil fuel 41 use/emissions, iii) carbon capture and sequestration. Since the latter two effects can compensate for the 42 LULUCF effect, BECCS deployment in ambitious stabilisation scenarios may lead to reduced FOOTNOTE: 3 Scenarios are considered consistent between global and regional results (based on R5 regions), if the sum of regional emissions (or sequestration efforts) does not deviate more than 10% from the reported global total. To take into account that small absolute values have a higher sensitivity, a deviation of 90% is allowed for absolute values below 100. Do Not Cite, Quote or Distribute 7-89 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 sink/increased emissions in LULUCF (Kalt et al. 2020). The same holds for trade-offs between carbon 2 sequestration in forests versus harvested wood products both for enhancing the HWP pool and for 3 material substitution. The strengths of the competition between biomass use and carbon sequestration 4 will depend on the biomass feedstocks considered (Lauri et al. 2019). 5 In the individual cases, the accounting of all these effects is dependent on the respective underlying 6 model and its coverage of inter-relations of different sectors and sub-sectors. The presented potentials 7 cover a wide range of models, and additionally, a wide range of background assumptions on macro- 8 economic, technical, and behavioural developments as well as policies, which the models have been 9 fed with. Subsequently, the range of the resulting marginal abatement costs is relatively wide, showing 10 the full range of expected contributions from land use sector mitigation and sequestration in applied 11 mitigation pathways. 12 13 14 Figure 7.15 Mitigation of CO2, CH4 and N2O emissions (in CO2-eq yr-1 using IPCC AR6 GWP100 values) 15 from the AFOLU sector for increasing carbon price levels for 2030 and 2050. In the left side panels, single 16 data points are generated by comparing emissions between a policy scenario and a related benchmark 17 scenario, and mapping these differences with the respective carbon price difference. Plots only show the 18 price range of up to USD(2010)250 tCO2-eq-1 and the mitigation range between -2,000 and 6,000 MtCO2- 19 eq yr-1 for better visibility. At the right-hand side, based on the same data as left-hand side panels, 20 Boxplots show Medians (vertical line within the boxes), Means (dots), 33%-66% intervals (Box) and 10%- 21 90% intervals (horizontal lines). Numbers on the very right indicate the number of observations falling 22 into the respective price range per variable. A wide range of carbon price induced mitigation options Do Not Cite, Quote or Distribute 7-90 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 (such as technical, structural and behavioural options in the agricultural sector, afforestation, 2 reforestation, natural re-growth or avoided deforestation in the LULUUCF sector, excluding carbon 3 capture and sequestration from BECCS) are reflected in different scenarios. 4 At the global level, the analysis of the economic mitigation potentials from N 2O and CH4 emissions 5 from AFOLU (which mainly can be related to agricultural activities) and CO2 emissions (which mainly 6 can be related to LULUCF emissions) reveals a relatively good agreement of models and scenarios in 7 terms of ranking between the gases. On the right-hand side panels of Figure 7.15, only small overlaps 8 between the ranges (showing the 10-90% intervals of observations) and mainly for lower price levels, 9 can be observed, despite all differences in underlying model structure and scenario assumptions. 10 N2O emissions show the smallest economic potential of the three different gases in 2030 as well as in 11 2050. The mitigation potential increases until a price range of USD150-200 and to a median value of 12 around 0.6 GtCO2-eq yr-1 mitigation in 2030 and 0.9 GtCO2-eq yr-1 in 2050, respectively, while 13 afterwards with higher prices the expansion is very limited. Mitigation of CH4 emissions has a higher 14 potential, also with increasing mitigation potentials until a price range of USD150-200 in both years, 15 with median mitigation of around 1.3 GtCO2-eq yr-1 in 2030 and around 2.4 GtCO2-eq yr-1 in 2050, 16 respectively. The highest mitigation potentials are observed for CO2, but also the highest ranges of 17 observations among the three gases. In 2030, a median of 4 GtCO2-eq yr-1 mitigation potential is 18 reported for the price range of USD200–250. In 2050, for the carbon price range of between USD100 19 and USD200, a median of around 4.8 GtCO2-eq yr-1 can be observed. 20 When compared with the sectoral estimates from Harmsen et al. (2019), the integrated assessment 21 median potentials are broadly comparable for the N2O mitigation potential; Harmsen et al. 2050 22 mitigation potential at USD125 is 0.6 GtCO2-eq yr-1 while the integrated assessment estimate for the 23 same price range is 0.7 GtCO2-eq yr-1. The difference is substantially larger for the CH4 mitigation 24 potential; 0.9 GtCO2-eq yr-1 in Harmsen et al. while 2 GtCO2-eq yr-1 the median integrated assessment 25 estimate. While the Harmsen et al. MACCs consider only technological mitigation options, integrated 26 assessments typically include also demand side response to the carbon price and GHG efficiency 27 improvements through structural change and international trade. These additional mitigation options 28 can represent more than 60% of the total non-CO2 mitigation potential in the agricultural sector, where 29 they are more important in the livestock sector, and thus the difference between sectoral and integrated 30 assessments is more pronounced for the CH4 emissions (Frank et al. 2019). 31 Economic CO2 mitigation potentials from land use change and forestry are larger compared to potentials 32 from non-CO2 gases, and at the same time reveal high levels of variation in absolute terms. The 66th 33 percentile in 2050 goes up to 5.2 GtCO2-eq yr-1 mitigation, while the lowest observations are even 34 negative, indicating higher CO2 emissions from land use in scenarios with carbon price compared to 35 scenarios without (counterintuitive dynamics explained below). 36 Land use is at the centre of the interdependencies with other sectors, including energy. Some models 37 see a strong competition between BECCS deployment with its respective demand for biomass, and CO2 38 mitigation/sequestration potentials in the land sector. Biomass demand may lead to an increase in CO2 39 emissions from land use despite the application of a carbon price when land use expansion for dedicated 40 biomass production, such as energy plantations, comes from carbon rich land use/land cover 41 alternatives, or when increased extraction of biomass from existing land uses, such as forest 42 management, leads to reduction of the carbon sink (Daioglou 2019; Luderer et al. 2018, SI) and can 43 explain the high variety of observations in some cases. Overall, the large variety of observations shows 44 a large variety of plausible results, which can go back to different model structures and assumptions, 45 showing a robust range of plausible outcomes (Kriegler et al. 2015). Do Not Cite, Quote or Distribute 7-91 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 7.5.3. Interaction between mitigation in the AFOLU sector and other SDGs in the 2 context of integrated assessments 3 Besides the level of biomass supply for bioenergy, the adoption of SDGs may also significantly impact 4 AFOLU emissions and the land use sector’s ability for GHG abatement (Frank et al. 2021). Selected 5 SDGs are found to have positive synergies for AFOLU GHG abatement and to consistently decrease 6 GHG emissions for both agriculture and forestry, thereby allowing for even more rapid and deeper 7 emissions cuts. In particular, the decreased consumption of animal products and less food waste 8 (SDG12), and the protection of high biodiversity ecosystems such as primary forests (SDG15) deliver 9 high synergies with GHG abatement. On the other hand, protection of highly biodiverse ecosystems 10 from conversion (SDG15) limits the global biomass potentials for bioenergy (Frank et al. 2021), and 11 while several forestry measures enhancing woody biomass supply for bioenergy may have synergies 12 with improving ecosystems conditions, many represent a threat to biodiversity (Camia et al. 2020). See 13 also Section 7.6.5. and Chapter 17 Section 17.3.3.7, Figure 17.1, Supplementary Material Table 17.1. 14 15 7.5.4. Regional AFOLU abatement for different carbon prices 16 17 Figure 7.16 Regional mitigation efforts for CO2, CH4 and N2O emissions (in CO2-eq yr-1 using IPCC AR6 18 GWP100 values) from the AFOLU sector for increasing carbon price levels for 2030 and 2050. Underlying 19 datapoints are generated by comparing emissions between a policy scenario and a related benchmark Do Not Cite, Quote or Distribute 7-92 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 scenario, mapping these differences with the respective carbon price differences. Boxplots show Medians 2 (vertical line within the boxes), Means (dots), 33%-66% intervals (box) and 10%-90% intervals 3 (horizontal lines) for respective scenarios of carbon prices implemented in intervals of USD50 from a 4 price of USD0 to USD250. Regions: Asia (ASIA), Latin America and Caribbean (LAM), Middle East 5 (MIDDLE_EAST),Africa (AFRICA), Developed Countries (OECD 90 and EU) (OECD+EU) and 6 Reforming Economies of Eastern Europe and the Former Soviet Union (REF). 7 At the regional level (Figure 7.16), the highest potential from non-CO2 emissions abatement, and mostly 8 from CH4, is reported for ASIA with the median of mitigation potential observations from CH4 9 increasing up to a price of USD200 in the year 2050, reaching a median of 1.2 GtCO2-eq yr-1. In terms 10 of economic potential, ASIA is followed by LAM, AFRICA, and OECD+EU, where emission reduction 11 mainly is achieved in the livestock sector. 12 The highest potentials from land-related CO2 emissions, including avoided deforestation as well as 13 afforestation, can be observed in LAM and AFRICA with strong responses of mitigation (indicated by 14 the median value) to carbon prices mainly in the lower range of displayed carbon prices. In general, 15 CO2 mitigation potentials show a wide range of results in comparison to non-CO2 mitigation potentials, 16 but mostly also a higher median value. The most extreme ranges are reported for the regions LAM and 17 AFRICA. A medium potential is reported for ASIA and OECD+EU, while REF has the smallest 18 potential according to model submissions. These estimates reflect techno-economic potentials and do 19 not necessarily include feasibility constraints which are discussed in Chapter 7.6. 20 7.5.5. Illustrative mitigation pathways 21 Different mitigation strategies can achieve the net emission reductions that would be required to follow 22 a pathway limiting global warming, with very different consequences for the land system. Figure 7.17 23 shows Illustrative Mitigation Pathways (IMPs) for achieving different climate targets highlighting 24 AFOLU mitigation strategies, resulting GHG and land use dynamics as well as the interaction with 25 other sectors. For consistency this chapter discusses IMPs as described in detail in chapters 1 and 3 of 26 this report but focusing on the land-use sector. All pathways are assessed by different IAM realizations 27 and do not only reduce GHG emissions but also use CDR options, whereas the amount and timing varies 28 across pathways, as do the relative contributions of different land-based CDR options. 29 The scenario ModAct (below 3.0°C warming, C6) is based on the prolongation of current trends (SSP2) 30 but contains measures to strengthen policies for the implementation of National Determined 31 Contributions (NDCs) in all sectors including AFOLU (Grassi et al. 2018). This pathway shows a strong 32 decrease of CO2 emissions from land-use change in 2030, mainly due to reduced deforestation, as well 33 as moderately decreasing N2O and CH4 emissions from agricultural production due to improved 34 agricultural management and dietary shifts away from emissions-intensive livestock products. 35 However, in contrast to CO2 emissions, which turn net-negative around 2050 due to 36 afforestation/reforestation, CH4 and N2O emissions persist throughout the century due to difficulties of 37 eliminating these residual emissions based on existing agricultural management methods (Stevanović 38 et al. 2017; Frank et al. 2017). Comparably small amounts of BECCS are applied by the end of the 39 century. Forest area increases at the cost of other natural vegetation. 40 IMP Neg-2.0 is similar to ModAct scenario in terms of socio-economic setting (SSP2) but differs 41 strongly in terms of the mitigation target (likely 2°C, C3) and its strong focus on the supply side of 42 mitigation measures with strong reliance on net-negative emissions. Consequently, all GHG emission 43 reductions as well as afforestation/reforestation and BECCS-based CDR start earlier in time at a higher 44 rate of deployment. However, in contrast to CO2 emissions, which turn net-negative around 2030 due 45 to afforestation/reforestation, CH4 and N2O emissions persist throughout the century, similar to 46 ModAct, due to ongoing increasing demand for total calories and animal-based commodities (Bodirsky Do Not Cite, Quote or Distribute 7-93 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 et al. 2020) and difficulties of eliminating these residual emissions based on existing agricultural 2 management methods (Stevanović et al. 2017 ; (Frank et al. 2017). In addition to abating land-related 3 GHG emissions as well as increasing the terrestrial sink, this example also shows the potential 4 importance of the land sector in providing biomass for BECCS and hence CDR in the energy sector. 5 Cumulative CDR (2020-2100) amounts to 502 GtCO2 for BECCS and 121 GtCO2 for land-use change 6 (including afforestation and reduced deforestation). In consequence, compared to ModAct scenario, 7 competition for land is increasing and much more other natural land as well as agricultural land 8 (cropland and pasture land) is converted to forest or bioenergy cropland with potentially severe 9 consequences for various sustainability dimensions such as biodiversity (Hof et al. 2018) and food 10 security (Fujimori et al. 2019). 11 IMP Ren is similar to IMP Neg-2.0 in terms of socio-economic setting (SSP2) but differs substantially 12 in terms of mitigation target and mitigation efforts in the energy sector. Even under the more ambitious 13 climate change mitigation target (1.5°C with no or low OS, C1), the high share of renewable energy in 14 IMP Ren strongly reduces the need for large-scale land-based CDR, which is reflected in smaller 15 bioenergy and afforestation areas compared to IMP Neg-2.0. However, CH4 and N2O emissions from 16 AFOLU persist throughout the century, similar to ModAct scenario and IMP Neg-2.0. 17 In contrast to IMPs Neg-2.0 and Ren, IMP SP (Soergel et al. 2021; 1.5°C with no or low OS, C1) 18 displays a future of generally low resource and energy consumption (including healthy diets with low 19 animal-calorie shares and low food waste) as well as significant but sustainable agricultural 20 intensification in combination with high levels of nature protection. This pathway shows a strong near- 21 term decrease of CO2 emissions from land-use change, mainly due to reduced deforestation, and in 22 difference to all other IMPs described in this chapter strongly decreasing N2O and CH4 emissions from 23 agricultural production due to improved agricultural management but also based on dietary shifts away 24 from emissions-intensive livestock products as well as lower shares of food waste. In consequence, 25 comparably small amounts of land are needed for land demanding mitigation activities such as BECCS 26 and afforestation. In particular, the amount of agricultural land converted to bioenergy cropland is 27 smaller compared to other mitigation pathways. Forest area increases either by regrowth of secondary 28 vegetation following the abandonment of agricultural land or by afforestation / reforestation at the cost 29 of agricultural land. 30 Do Not Cite, Quote or Distribute 7-94 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 2 Figure 7.17 Evolution and break down of (A) global land-based GHG emissions and removals and (B) 3 global land use dynamics under four Illustrative Mitigation Pathways, which illustrate the differences in 4 timing and magnitude of land-based mitigation approaches including afforestation and BECCS. All 5 pathways are based on different IAM realisations: ModAct scenario (below 3.0°C, C6) from IMAGE 3.0; 6 IMP Neg-2.0 (likely 2°C, C3) from AIM/CGE 2.2; IMP Ren (1.5°C with no or low OS, C1) from 7 REMIND-MAgPIE 2.1-4.3; IMP SP (1.5°C with no or low OS, C1) from REMIND-MAgPIE 2.1-4.2; In 8 panel A the categories CO2 Land, CH4 Land and N2O Land include GHG emissions from land-use change 9 and agricultural land use (including emissions related to bioenergy production). In addition, the category 10 CO2 Land includes removals due to afforestation / reforestation. BECCS reflects the CO 2 emissions 11 captured from bioenergy use and stored in geological deposits. CH 4 and N2O emissions are converted to 12 CO2-eq using GWP100 factors of 27 and 273 respectively. 13 14 7.6. Assessment of economic, social and policy responses 15 7.6.1. Retrospective in policy efforts and achieved mitigation within AFOLU 16 Since the establishment of the UNFCCC, international agencies, countries, sub-national units and 17 NGO's have developed policies to facilitate and encourage GHG mitigation within AFOLU (Figure 18 7.18). Early guidance and policies focused on developing GHG inventory methodology with some 19 emphasis on afforestation and reforestation projects, but the Clean Development Mechanism (CDM) in 20 the Kyoto Protocol focused attention on emission reduction projects, mostly outside of AFOLU. As 21 successive IPCC WGIII reports illustrated large potential for AFOLU mitigation, methods to quantify 22 and verify carbon emission reductions emerged within several projects in the early 2000s, through both 23 voluntary (e.g., the Chicago Climate Exchange (CCX)) and regulated (e.g., New South Wales and 24 California) markets. The CDM dedicated large attention to LULUCF, including dedicated 25 methodologies and bodies. The reasons for limited uptake of CDM afforestation/reforestation projects Do Not Cite, Quote or Distribute 7-95 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 were multiple and not limited to the regulatory constraints, but also due to the low abatement potential 2 (poor cost/performance ratio) compared to other mitigation opportunities. 3 Following COP 13 in Bali, effort shifted to advancing policies to reduce deforestation and forest 4 degradation (REDD+) in developing countries. According to Simonet et al. (2019), nearly 65 Mha have 5 been enrolled in REDD+ type programs or projects funded through a variety of sources, including UN 6 REDD, the World Bank Forest Carbon Partnership Facility, and bi-lateral agreements between countries 7 with Norway being the largest donor. While there has been considerable focus on forest and agricultural 8 project-based mitigation actions, national governments were encouraged to incorporate project-based 9 approaches with other sectoral strategies in their Nationally Appropriate Mitigation Strategies 10 (NAMAs) after 2012. NAMAs reflect the country’s proposed strategy to reduce net emissions across 11 various sectors within their economy (e.g. forests or agriculture). More recently, Nationally Determined 12 Contributions (NDCs) indicate whether individual countries plan to use forestry and agricultural 13 policies or related projects amongst a set of measures in other sectors, to reduce their net emissions as 14 part of the Paris Agreement (e.g., Forsell et al. 2016; Fyson and Jeffery 2019). 15 16 17 Figure 7.18 Milestones in policy development for AFOLU measures. 18 19 The many protocols now available can be used to quantify the potential mitigation to date resulting 20 from various projects or programs. For instance, carbon registries issue credits using protocols that 21 typically account for additionality, permanence and leakage, thus providing evidence that the projects 22 are a net carbon benefit to the atmosphere. Protocol development engages the scientific community, 23 project developers, and the public over a multi-year period. Some protocols have been revised multiple 24 times, such as the USA State of California’s forest carbon protocol, which is in its fifth revision, with 25 the latest in 2019 (see http://www.climateactionreserve.org/how/protocols/forest/). Credits from carbon 26 registries feed into regulatory programs, such as the cap and trade program in California, or voluntary 27 offset markets (Hamrick and Gallant 2017a). Although AFOLU measures have been deployed across a Do Not Cite, Quote or Distribute 7-96 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 range of projects and programs globally to reduce net carbon emissions, debate about the net carbon 2 benefits of some projects continues (e.g. Krug 2018). 3 A new assessment of projects over the last two decades finds emission reductions or offsets of at least 4 7.9 GtCO2-eq (using GWP100 and a mix of IPCC values for CH4 and N2O) over the last 12 years due to 5 agricultural and forestry activities (Table 7.4). More than 80% of these emission reductions or offsets 6 have been generated by forest-based activities. The total amounts to 0.66 GtCO2 yr-1 for the period 7 2010-2019, which is 1.2% of total global, and 5.5% of AFOLU emissions reported in Table 7.1, over 8 the same time period (high confidence). 9 The array of activities in Table 7.4 includes the Clean Development Mechanism, REDD+ activities 10 reported in technical annexes of country biennial update reports to the UNFCCC, voluntary market 11 transactions, and carbon stored as a result of carbon markets in Australia, New Zealand and California 12 in the USA. Although other countries and sub-national units have developed programs and policies, 13 these three regions are presented due to their focus on forest and agricultural carbon mitigation, their 14 use of generally accepted protocols or measures and the availability of data to quantify outcomes. 15 The largest share of emission reductions or carbon offsets in Table 7.4 has been from slowing 16 deforestation and REDD+, specifically from efforts in Brazil (86% of total), which substantially 17 reduced deforestation rates between 2004 and 2012 (Nepstad et al., 2014), as well as other countries in 18 Latin America. With the exception of Roopsind et al. (2019), estimated reductions in carbon emissions 19 from REDD+ in Table 7.4 are measured relative to a historical baseline. As noted in Brazil's Third 20 Biennial Update Report (Ministry of Foreign Affairs 2019), estimates are made in accordance with 21 established methodologies to determine the benefits of results-based REDD+ payments to Brazil. 22 REDD+ estimates from other countries also have been derived from biennial update reports. 23 Regulatory markets provide the next largest share of carbon removal to date. Data from the Australia 24 Emissions Reduction Fund is an estimate of carbon credits in agriculture and forestry purchased by the 25 Australian government. In the case of California, offset credits from forest and agricultural activities, 26 using methods approved by a third-party certification authority (Climate Action Reserve), have been 27 allowed as part of their state-wide cap and trade system. Transaction prices for forest and agricultural 28 credits in California were around USD13 tCO2-1 in 2018, and represented 7.4% of total market 29 compliance. By the end of 2018, 80 MtCO2 had been used for compliance purposes. 30 For New Zealand, the carbon reduction in Table 7.4 represents forest removals that were surrendered 31 from post-1989 forests between 2008 and the 2020. Unlike offsets in voluntary markets or in California, 32 where permanence involves long-term contracts or insurance pools, forests in the New Zealand market 33 liable for emissions when harvested or following land use change. This means sellers account for future 34 emission risks related to harvesting when they enter forests into carbon contracts. Offset prices were 35 around USD13 tCO2-1 in 2016 but have risen to more than USD20 tCO2-1 in 2020. 36 The voluntary market data in Table 7.4 are offsets developed under the major standard-setting 37 organizations, and issued from 2008-2018 (e.g., Hamrick and Gallant 2018). Note that there is some 38 potential for double counting of voluntary offsets that may have been transacted in the California 39 compliance market, however this would only have applied to transactions of US-issued offsets, and the 40 largest share of annual transactions of voluntary AFOLU credits occurs with credits generated in Latin 41 America, followed by Africa, Asia and North America. Europe and Oceania have few voluntary carbon 42 market transactions. Within forestry and agriculture, most of the voluntary offsets were generated by 43 forestry projects. Using historical transaction data from various Forest Trends reports, the offsets 44 generated were valued at USD46.9 million yr-1. Prices for voluntary offset transactions in the period 45 2014-2016 ranged from USD4.90 to 5.40 tCO2-1 (Hamrick and Gallant 2017a). Do Not Cite, Quote or Distribute 7-97 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Voluntary finance has amounted to USD0.5 billion over a 10-year period for development of forest and 2 agricultural credits. The three regulatory markets quantified amount to USD2.7 billion in funding from 3 2010 to 2019. For the most part, this funding has focused on forest projects and programs, with 4 agricultural projects accounting for 5-10% of the total. In total, reported funding for AFOLU projects 5 and programs has been USD4.4 billion over the past decade, or about USD569 million yr -1 (low 6 confidence). The largest share of the total carbon includes efforts in the Amazon by Brazil. Government 7 expenditures on regulatory programs and business expenditures on voluntary programs in Brazil (e.g., 8 the soy or cattle moratoriums) were not included in financing estimates due to difficulties obtaining that 9 data. If Brazil and CDM (for which we have no cost estimates) are left out of the calculation, average 10 cost per ton has been USD3.20 tCO2-1. 11 12 Table 7.4 Estimates of achieved emission offsets or reductions in AFOLU through 2018. Data include 13 CDM, voluntary carbon standards, compliance markets, and reduced deforestation from official 14 UNFCCC reports. Carbon sequestration due to other government policies not included. Fund / Mechanism Total Emission Time Frame Mt CO2-eq Financing Reductions or yr-1 (Million USD yr-1) Offsets (Mt CO2-eq) CDM-forest a 11.3 2007-2015 1.3 - CDM-agriculture a 21.8 2007-2015 2.4 - REDD + (Guyana) b 12.8 2010-2015 2.1 33.0 Reduced 6,894.5 2006-2017 574.5 49.2 Deforestation/ REDD + Brazil c REDD + Indonesia c 244.9 2013-2017 49.0 13.4 REDD + Argentina c 165.2 2014-2015 55.1 1.4 REDD + Others c 211.8 2010-2017 26.5 46.0 Voluntary Market d 95.3 2009-2018 9.5 46.9 Australia ERF e 33.7 2012-2018 4.8 50.5 California f 122.2 2013-2018 20.4 227.1 New Zealand Carbon 83.9 2010-2019 8.4 101.7 Trading g Total 7,897.4 2007-2018 658.1 8 569.1 15 a Clean Development Mechanism Registry: https://cdm.unfccc.int/Registry/index.html (accessed 16 22/06/2021) 17 b Roopsind et al. 2019. 18 c UNFCCC REDD+ Web Platform (https://redd.unfccc.int/submissions.html) and UNFCCC Biennial Update 19 Report database (https://unfccc.int/BURs) 20 d (Hamrick and Gallant 2017a). State of Forest Carbon Finance. Forest Trends Ecosystem Marketplace. 21 Washington, DC. Do Not Cite, Quote or Distribute 7-98 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 e Data for Australia carbon credit units (ACCUs) from Australia Emission Reduction Fund Registry for 2 forest agricultural and savanna practices through FY2018/19 (downloaded on 24/10/2019): 3 (http://www.cleanenergyregulator.gov.au/ERF/project-and-contracts-registers/project-register). 4 f Data from the California Air Resources Board Offset Issuance registry (https://ww2.arb.ca.gov/our- 5 work/programs/compliance-offset-program) for forestry and agricultural early action and compliance 6 credits. 7 g Surrendered forest carbon credits from post-1989 forests in New Zealand. Obtained from New Zealand 8 Environmental Protection Authority. ETS Unit Movement interactive report (Excel based). 9 https://www.epa.govt.nz/industry-areas/emissions-trading-scheme/ets-reports/unit-movement/. Obtained 10 13/08/2020.8 All non-CO2 gases are converted to CO2-eq using IPCC GWP100 values recommended at the 11 time the project achieved approval by the relevant organisation or agency. 12 13 The large number of policy approaches described in Table 7.4 combined with efforts by other 14 international actors, such as the Global Environmental Facility (GEF), as well as non-state actors (e.g., 15 eco-labelling programs and corporate social responsibility initiatives) illustrate significant policy 16 experimentation over the last several decades. Despite widespread effort, AFOLU measures have thus 17 far failed to achieve the large potential for climate mitigation described in earlier IPCC WG III reports 18 (high confidence). The limited gains from AFOLU to date appear largely to result from lack of 19 investment and other institutional and social barriers, rather than methodological concerns (high 20 confidence). 21 7.6.2. Review of observed policies and policy instruments 22 7.6.2.1. Economic incentives 23 Emissions Trading/Carbon Taxes. While emissions trading programs have been developed across the 24 globe, forest and agriculture have not been included as part of the cap in any of the existing systems. 25 However, offsets from forestry and agriculture have been included in several of the trading programs. 26 New Zealand has a hybrid program where carbon storage in forests can be voluntarily entered into the 27 carbon trading program, but once entered, forests are counted both as a sink for carbon if net gains are 28 positive, and a source when harvesting occurs. New Zealand is considering rules to include agricultural 29 GHG emissions under a future cap (Henderson et al. 2020; see: https://www.agmatters.nz/topics/he- 30 waka-eke-noa/). 31 The state of California has developed a formal cap and trade program that allows a limited number of 32 forest and agricultural offset credits to be used under the cap. All offsets must meet protocols to account 33 for additionality, permanence and leakage. Forest projects used as offsets in California currently are 34 located in the USA, but the California Air Resources Board adopted a tropical forest carbon standard, 35 allowing for avoided deforestation projects from outside the USA to enter the California market (CARB 36 2019). 37 Canadian provinces have developed a range of policy options that can include carbon offsets. Quebec 38 has an emissions trading program that plans to allow forest and agricultural offsets generated within the 39 province to be utilised. Alberta also allows offsets to be utilised by regulated sectors while British 40 Columbia allows offsets to be utilised by the government for its carbon neutrality goals (Government 41 of Alberta, 2021) . Over 20 countries and regions have adopted explicit carbon taxes on carbon emission 42 sources and fossil fuels, however, the charges have not been applied to non-CO2 agricultural emissions 43 (OECD 2021a). California may implement regulations on methane emissions from cattle, however, 44 regulations if approved, will not go into effect until 2024. Institutional and trade-related barriers (e.g., 45 leakage) likely will limit widespread implementation of taxes on emissions in the food sector globally. 46 Many countries exempt purchases of fuels used in agricultural or fishery production from fuel or carbon 47 taxes, thus lowering the effective tax rate imposed on those sectors (OECD 2021a). Furthermore, 48 bioenergy, produced from agricultural products, agricultural waste, and wood is often exempted from Do Not Cite, Quote or Distribute 7-99 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 explicit carbon taxes. Colombia recently implemented a carbon tax on liquid fuels but allowed 2 domestically produced forestry credits to offset the tax. Colombia also is in the process of developing 3 an emissions trading scheme (OECD 2021a). 4 REDD+/Payment for Ecosystem Services (PES). PES programs for a variety of ecosystem services 5 have long been utilised for conservation (e.g. Wunder 2007) and may now be as large as USD42 billion 6 yr-1 (Salzman et al. 2018). REDD+ emerged in the early 2000s and is a widely recognized example of 7 PES program focused on conservation of tropical forests (Table 7.4). However, our summation of 8 actually paid funds in Table 7.4 is much smaller than what is portrayed by Salzman et al. (2018). 9 REDD+ may operate at the country level, or for specific programs or forests within a country. As with 10 other PES programs, REDD+ has evolved towards a results-based program that involves payments that 11 are conditioned on meeting certain successes or milestones, such as rates of deforestation (Angelsen 12 2017). 13 A large literature has investigated whether PES programs have successfully protected habitats. Studies 14 in the USA found limited additionality for programs that encouraged conservation tillage practices, but 15 stronger additionality for programs that encouraged set-asides for grasslands or forests (Woodward et 16 al. 2016; Claassen et al. 2018;), although the set-asides led to estimated leakage of 20 up to 100% (Wu 17 2000; Pfaff and Robalino 2017; Kallio and Solberg 2018). Evidence from the EU similarly suggests 18 that payments for some agro-environmental practices may be additional, while others are not (Chabé- 19 Ferret and Subervie 2013). Other studies, in particular in Latin America where many PES programs 20 have been implemented, have found a wide range of estimates of effectiveness (e.g. Honey-Rosés et al. 21 2011; Robalino and Pfaff 2013; Mohebalian and Aguilar 2016; Jayachandran et al. 2017; Börner et al. 22 2017; Alix-Garcia et al. 2015; Robalino et al. 2015; Burivalova et al. 2019). Despite concerns, the 23 many lessons learned from PES program implementation provide critical information that will help 24 policymakers refine future PES programs to increase their effectiveness (medium confidence). 25 While expectations that carbon-centred REDD+ would be a simple and efficient mechanism for climate 26 mitigation have not been met (Turnhout et al. 2017; Arts et al. 2019), progress has nonetheless occurred. 27 Measuring, monitoring and verification systems have been developed and deployed, REDD readiness 28 programs have improved capacity to implement REDD+ on the ground in over 50 countries, and a 29 number of countries now have received results-based payments. 30 Empirical evidence that REDD+ funding has slowed deforestation is starting to emerge. Simonet et al. 31 (2019) showed that a REDD+ project in Brazil reduced deforestation certainly until 2018, while 32 Roopsind et al. (2019) showed that country-level REDD+ payments to Guyana encouraged reduced 33 deforestation and increased carbon storage. Although more impact evaluation (IE) analysis needs to be 34 conducted on REDD+ payments, these studies support the country-level estimates of carbon benefits 35 from REDD+ shown in Table 7.4. Nearly all of the analysis of PES and REDD+ to date has focused on 36 the presence or absence of forest cover, with little to no analysis having been conducted on forest 37 degradation, conservation, or enhancement of forest stocks. 38 Agro-environmental Subsidy Programs/PES. Climate policy for agriculture has developed more 39 slowly than in other sectors due to concerns with food security and livelihoods, political interests, and 40 difficulties in coordinating diffuse and diverse activities and stakeholders (e.g. nutritional health, rural 41 development, and biodiversity conservation) (Leahy et al. 2020). However, a review of the National 42 Adaptation Programme of Action (NAPAs), National Adaptation Plans (NAPs), NAMAs, and NDCs 43 in the Paris Agreement suggest an increasing focus of policy makers on agriculture and food security. 44 The vast majority of parties to the Paris Agreement recognise the significant role of agriculture in 45 supporting a secure sustainable development pathway (Richards and VanWey 2015) with the inclusion 46 of agriculture mitigation in 103 NDCs from a total of 160 NDC submissions. Livestock is the most Do Not Cite, Quote or Distribute 7-100 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 frequently cited specific agricultural sub-sector, with mitigation activities generally focusing on 2 increasing efficiency and productivity. 3 Agriculture is one of the most subsidised industries globally, especially in the European Union and the 4 USA. While subsidy payments over the last 20 years have shifted modestly to programs designed to 5 reduce the environmental impact of the agricultural sector, only 15-20% of the more than USD700 6 billion spent globally on subsidies are green payments (OECD 2021b). Under the Common Agricultural 7 Policy in the EU, up to 30% of the direct payments to farmers (Pillar 1) have been green payments 8 (Henderson et al. 2020), including some actions that could increase carbon storage or reduce emissions. 9 Similarly, at least 30% of the rural development payments (Pillar 2) are used for measures that reduce 10 environmental impact, including reduction of GHG emissions and carbon storage. There is limited 11 evidence that these policies contributed to the 20% reduction in GHG emissions from the agricultural 12 sector in the EU between 1990 and 2018 (Baudrier et al. 2015) and Eurostat 2020). 13 The USA spends USD4 billion yr-1 on conservation programs, or 12% of net farm income (Departament 14 Of Agriculture 2020). In real terms, this expenditure has remained constant for 15 years, supporting 12 15 Mha of permanent grass or woodland cover in the Conservation Reserve Program (CRP), which has 16 increased soil carbon sequestration by 3 tCO2 ha-1 yr-1 (Paustian et al. 2019; Conant et al. 2017), as well 17 as other practices that could lower net emissions. Gross GHG emissions from the agricultural sector in 18 the US, however, have increased since 1990 (US-EPA 2020) due to reductions in the area of land in the 19 US CRP program and changes in crop rotations, both of which caused soil carbon stocks to decline 20 (US-EPA 2020). When combined with increased non-CO2 gas emissions, the emission intensity of US 21 agriculture increased from 1.5 to 1.7 tCO2 ha-1 between 2005 and 2018 (high confidence). 22 China has implemented large conservation programs that have influenced carbon stocks. For example, 23 the Sloping Land Conversion Program, combined with other programs, has increased forest cover and 24 carbon stocks, reduced erosion and increased other ecosystem services in China in recent years (Ouyang 25 et al. 2016). Despite increased forest area in China, however, land use change and management 26 potentially were net contributors to carbon emissions from 1990-2010 (Lai et al. 2016). As part of 27 Brazil's national strategy, numerous practices to reduce GHG emissions from agriculture, and in 28 particular from the animal agriculture industry, have been subsidized. Estimates by Manzatto et al. 29 (2020) suggest that the program may have reduced agricultural emissions by 169 MtCO2 between 2010 30 and 2020. Given the large technical and economic potential for agroforestry to be deployed in Africa, 31 subsidy approaches could be deployed along with other polices to enhance carbon through innovative 32 practices such as regreening (Box 7.10). 33 7.6.2.2. Regulatory approaches 34 Regulations on land use include direct controls on how land is used, zoning, or legally set limits on 35 converting land from one use to another. Since the early 2000s, Brazil has deployed various regulatory 36 measures to slow deforestation, including enforcement of regulations on land use change in the legal 37 Amazon area. Enforcement of these regulations, among other approaches is credited with encouraging 38 the large-scale reduction in deforestation and associated carbon emissions after 2004 (Nepstad et al. 39 2014). Empirical evidence has found that regulations reduced deforestation in Brazil (Arima et al. 2014) 40 but over time, reversals occurred when enforcement was not consistent (Azevedo et al. 2017) (Box 7.9). 41 Many OECD countries have strong legal frameworks that influence agricultural and forest management 42 on both public and private land. These include for example, legal requirements to protect endangered 43 species, implement conservation tillage, protect riparian areas, replant forests after harvest, maintain 44 historical species composition, forest certification, and other approaches. Increasingly, laws support 45 more widespread implementation of nature-based solutions for a range of environmental issues (e.g. 46 see European Commission 2021) The extent to which the combined influence of these regulations has Do Not Cite, Quote or Distribute 7-101 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 enhanced carbon storage in ecosystems is not quantified although they are likely to explain some of the 2 persistent carbon sink that has emerged in temperate forests of OECD countries (high confidence). In 3 the least developed and developing countries, regulatory approaches face challenges in part because 4 environmental issues are a lower priority than many other socioeconomic issues (e.g., poverty, 5 opportunity, essential services), and weak governance (Mayer Pelicice 2019; Walker et al. 2020); Box 6 7.2). 7 Set asides and protected areas have been a widely utilised approach for conservation, and according to 8 (FAO 2020d), 726 Mha (18%) of forests are in protected areas globally. A review of land sparing and 9 land sharing policies in developing countries indicated that most of them follow land sparing models, 10 sometimes in combination with land sharing approaches. However, there is still no clear evidence of 11 which policy provides the best results for ecosystem services provision, conservation, and livelihoods 12 (Mertz and Mertens 2017). The literature contains a wide range of results on the effectiveness of 13 protected areas to reduce deforestation (Burivalova et al. 2019), with studies suggesting that protected 14 areas provide significant protection of forests (e.g., Blackman 2015), modest protection (Andam et al. 15 2008), as well as increases in deforestation (Blackman 2015) and possible leakage of harvesting to 16 elsewhere (Kallio and Solberg 2018). An estimate of the contributions of protected areas to mitigation 17 between 2000 and 2012, showed that in the tropics, PAs reduced carbon emissions from deforestation 18 by 4.88 Pg C, or around 29%, when compared to the expected rates of deforestation (Bebber and Butt 19 2017). In that study, the tropical Americas (368.8 TgC y−1) had the largest contribution, followed by 20 Asia (25.0 TgC y−1) and Africa (12.7 TgC y−1). The authors concluded that local factors had an 21 important influence on the effectiveness of protected areas. For example, in the Brazilian Amazon, 22 protected area effectiveness is affected by the government agency controlling the land (federal 23 indigenous lands, federal PAs, and state PAs) (Herrera et al. 2019). Because protected areas limit not 24 just land use change, but also logging or harvesting non-timber forest products, they may be relatively 25 costly approaches for forest conservation (medium confidence). 26 Community forest management (CFM) allows less intensive use of forest resources, while at the same 27 time providing carbon benefits by protecting forest cover. Community forest management provides 28 property rights to communities to manage resources in exchange for their efforts to protect those 29 resources. In many cases, the local communities are indigenous people who otherwise would have 30 insecure tenure due to an advancing agricultural frontier or mining activity. Other examples are forest 31 owner associations like those discussed in Box 7.8. According to the Rights and Resources Initiative 32 (2018), the area of forests under community management increased globally by 152 Mha from 2002 to 33 2017, with over 500 Mha under community management in 2017. Studies have now shown that 34 improved property rights with community forest management can reduce deforestation and increase 35 carbon storage (Bowler et al. 2012; Alix-Garcia et al. 2005; Blackman 2015; Fortmann et al. 2017; 36 Burivalova et al. 2019; Alix-Garcia 2007; Deininger and Minten 2002). Efforts to expand property 37 rights, especially community forest management, have reduced carbon emissions from deforestation in 38 tropical forests in the last two decades (high confidence), although the extent of carbon savings has not 39 been quantified globally. 40 41 [START BOX 7.8 HERE] 42 Box 7.8 Management of native forests by the Menominee people in North America and lessons 43 from forest owner associations 44 Summary of the case – Indigenous peoples include more than 5 000 different peoples, with over 370 45 million people, in 70 countries on five continents (UNIPP 2012). Forests cover more than 80% of the 46 area occupied by indigenous peoples (330 million hectares) point to their critical for forest governance Do Not Cite, Quote or Distribute 7-102 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 (Garnett et al. 2018; Fa et al. 2020). The Menominee people (Wisconsin, USA) practice sustainable 2 forestry on their reservation according to a land ethic integral to the tribal identity. The Tribe calls 3 themselves “The Forest Keepers,” recognizing that the connection of their future to the sustainable 4 management of the forest that allowed the forest volume standing today to be higher than when timber 5 harvesting began more than 160 years ago. Management practices are based on continuous forest 6 inventories (Mausel et al. 2017). 7 Introduction to the case - Forest management and timber harvesting operations began shortly after the 8 Menominee Indian Reservation was created by treaty in 1854. The Menominee reservation sits on ca. 9 95000 ha of land in Wisconsin that spans multiple forest types and is more diverse than adjacent forests. 10 The collectively maintained reservation has 87% of its land under sustained yield forestry. 11 Case description - The Tribe, in the 19th century, had already mastered vegetation manipulation with 12 fire, sustainable forestry, multiple-use, ecosystem, and adaptive management. The centerpiece of the 13 Tribe’s economy has been its forest product industry, Menominee Tribal Enterprises (MTE) (Pecore 14 1992). A balance between growth and removals and continuous forest inventories (CFI) are central for 15 forest management for the past 160 years, aiming not at very large volumes, but at very high quality 16 trees. During this same period, more than 2.3 billion board feet have been harvested from the same area, 17 equivalent to 0.3 m3 ha-1y-1). 18 Interactions and limitations –In 2013, the Menominee Tribe started a collaboration with the US Forest 19 Service to implement climate adaptation measures. The Tribe actively works to reduce the risk of forest 20 damage and decided to further promote diversity by planting tree seedlings adapted to a warming 21 climate (https://toolkit.climate.gov/case-studies/and-trees-will-last-forever). However, new challenges 22 are related to increasing pressures on forest ecosystems such as non-native insects, pathogens, weed 23 invasions, and the costs for continuous forest inventories to support long-term forest management. 24 Identified lessons - The elements of sustainability are intertwined with Menominee history, culture, 25 spirituality, and ethics. The balance between the environment, community, and economy for the short 26 term as well as future generations is an example of protecting the entire environment as the Menominee 27 land is a non-fragmented remnant of the prehistoric Lake States forest which has been dramatically 28 reduced all around the reserve (Schabel and Pecore 1997). These and other types of community forest 29 owner associations exist all over the world. Examples are Södra in Sweden (with 52,000 forest owners) 30 (Södra, 2021) or Waldbauernverband in North-Rhine Westphalia (with 150,000 forest owners and 31 covering 585,000 ha) (AGDW-The Forest Owners, 2021). These are ways for small forest owners 32 to educate, jointly put wood on the market, employ better forest management, use machinery together, 33 and apply certification jointly. In this manner and with all their diversity of goals, they manage to 34 maintain carbon sinks and stocks, while preserving biodiversity and producing wood. 35 [END BOX 7.8 HERE] 36 37 Bioenergy targets. Multiple policies have been enacted at national and supra-national levels to promote 38 the use of bioenergy in the transport sector, and for bioelectricity production. Existing policies mandate 39 or subsidize the production and use of bioenergy. In the past few years, policies have been proposed, 40 put in place or updated in Australia (Renewable Energy Target), Brazil (RenovaBio, Nationally 41 Determined Contribution), Canada (Clean Fuel Standard), China (Biodiesel Industrial Development 42 Policy, Biodiesel Fuel Blend Standard), the European Union (Renewable Energy Directive II), the USA 43 (Renewable Fuel Standards), Japan (FY2030), Russia (Energy Strategy Bill 2035), India (Revised 44 National Policy on Biofuels), and South Africa (Biofuels Regulatory Framework). 45 While current policies focus on bioenergy to decarbonise the energy system, some also contain 46 provisions to minimise the potential environmental and social trade-offs from bioenergy production. 47 For instance, the EU Renewable Energy Directive (EU-REDII) and US Renewable Energy Standard 48 (US-RFS) assign caps on the use of biofuels, which are associated with indirect land-use change and Do Not Cite, Quote or Distribute 7-103 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 food-security concerns. The Netherlands has a stringent set of 36 sustainability criteria to which the 2 certified biomass needs to comply. The EU-REDII also sets a timeline for the complete phase-out of 3 high-risk biofuels (Section 7.4.4). 4 7.6.2.3. Voluntary actions and agreements 5 Forest certification programs, such as Forest Sustainability Council (FSC) or Programme for the 6 Endorsement of Forest Certification (PEFC), are consumer driven, voluntary programs that influence 7 timber harvesting practices, and may reduce emissions from forest degradation with reduced impact 8 logging and other approaches (medium confidence). Forest certification has expanded globally to over 9 440 Mha (Kraxner et al. 2017). As the area of land devoted to certification has increased, the amount 10 of timber produced from certified land has increased. In 2018, FSC accounted for harvests of 427 11 million m3 and jointly FSC and PEFC accounted for 689 million m3 in 2016 or around 40% of total 12 industrial wood production (FAO 2018c). There is evidence that reduced impact logging can reduce 13 carbon losses in tropical regions (Pearson et al. 2014); (Ellis et al. 2019). However, there is conflicting 14 evidence about whether forest certification reduces deforestation (e.g., Tritsch et al. 2020; Blackman et 15 al. 2018). 16 Supply chain management in the food sector encourages more widespread use of conservation 17 measures in agriculture (high confidence). The number of private commitments to reduce deforestation 18 from supply chains has greatly increased in recent years, with at least 865 public commitments by 447 19 producers, processors, traders, manufacturers and retailers as of December, 2020 (New York 20 Declaration on Forests 2021). Industry partnerships with NGOs, such as the Roundtable on Sustainable 21 Palm Oil (RSPO), have become more widespread and visible in agricultural production. For example, 22 RSPO certifies members all along the supply chain for palm oil and claims around 19% of total 23 production. Similar sustainability efforts exist for many of the world's major agricultural products, 24 including soybeans, rice, sugar cane, and cattle. 25 There is evidence that the Amazon Soy Moratorium (ASM), an industry-NGO effort whereby large 26 industry consumers agreed voluntarily not to purchase soybeans grown on land deforested after 2006, 27 reduced deforestation in the legal Amazon (Nepstad et al. 2014; Gibbs et al. 2015). However, recent 28 studies have shown that some deforestation from the Amazon was displaced to the Cerrado (Brazilian 29 savannas) region (Moffette and Gibbs 2021). which is a global hotspot for biodiversity, and has 30 significant carbon stocks. These results illustrate the importance of broadening the scope of supply 31 chain management to minimize or eliminate displacement (Lima et al. 2019). In addition, while 32 voluntary efforts may improve environmental outcomes for a time, it is not clear that they are sufficient 33 to deliver long-term reductions in deforestation, given the increases in deforestation that have occurred 34 in the Amazon in recent years (Box 7.9). Voluntary efforts would be more effective at slowing 35 deforestation if they present stronger linkages to regulatory or other approaches (Lambin et al. 2018). 36 37 [START BOX 7.9 HERE] 38 Box 7.9 Case study: Deforestation control in the Brazilian Amazon 39 Summary 40 Between 2000 and 2004, deforestation rates in the Brazilian Legal Amazon (is a socio-geographic 41 division containing all nine Brazilian states in the Amazon basin) increased from 18,226 to 27,772 km2 42 yr-1 2008 (INPE, 2021) . A set of public policies designed in participatory process involving federal 43 government, states, municipalities, and civil society successfully reduced deforestation rates until 2012. 44 However, deforestation rates increased after 2013, and particularly between 2019 and 2020. Successful Do Not Cite, Quote or Distribute 7-104 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 deforestation control policies are being negatively affected by changes in environmental governance, 2 weak law enforcement, and polarisation of the national politics. 3 Background 4 In 2004, the Brazilian federal government started the Action Plan for Prevention and Control of 5 Deforestation in the Legal Amazon (PPCDAm) (Ministry of Environment, Government of Brazil, 6 2018) 7 . The PPCDAm was a benchmark for the articulation of forest conservation policies that included central 8 and state governments, prosecutor offices, and the civil society. The decline in deforestation after 2008 9 is mostly attributed to these policy options. In 2012, deforestation rates decreased to 4,571 km2 yr-1. 10 Case description 11 Combating deforestation was a theme in several programs, government plans, and projects not being 12 more restricted to the environmental agenda. This broader inclusion resulted from a long process of 13 insertion and articulation in the government dating back to 2003 while elaborating on the Sustainable 14 Amazon Plan. In May 2003, a historic meeting took place in an Amazonian city, with the President of 15 the Republic, State Governors, Ministers, and various business leaders, civil institutions, and social 16 movements. It was presented and approved the document entitled "Sustainable Amazonia - Guidelines 17 and Priorities of the Ministry of Environment for the Sustainable Development of the Amazon 18 Brazilian," containing several guidelines for conservation and sustainable use in the region. At the 19 meeting, the Union and some states signed a Cooperation Agreement aiming to elaborate a plan for the 20 Amazon, to be widely discussed with the various sectors of the regional and national society (Ministerio 21 do Meio Ambiente. MMA 2013). 22 Interactions and limitations 23 The PPCDAm had three main lines of action: 1. territorial management and land use; 2. command and 24 control; and 3. promotion of sustainable practices. During the execution of the 1st and 2nd phases of 25 the PPCDAm (2004-2011), important results in the territorial management and land use component 26 included, for example, the creation of 25 Mha of federal Protected Areas (PAs) located mainly in front 27 of the expansion of deforestation, as well as the homologation of 10 Mha of Indigenous Lands. Also, 28 states and municipalities created approximately 25 Mha, so that all spheres of government contributed 29 to the expansion of PAs in the Brazilian Amazon. In the Command and Control component, agencies 30 performed hundreds of inspection operations against illegal activities (e.g., illegal logging) under 31 strategic planning based on technical and territorial priorities. Besides, there was a significant 32 improvement of the environmental monitoring systems, involving the analysis of satellite images to 33 guide actions on the ground. Another policy was the restriction of public credit to enterprises linked to 34 illegal deforestation following a resolution of the Brazilian Central Bank (2008) (Ministerio do Meio 35 Ambiente. MMA 2013). Also, in 2008, Brazil created the Amazon Fund, a REDD+ mechanism 36 (Government of Brazil, n.d). 37 However, the country's political polarisation has gradually eroded environmental governance, 38 especially after the Brazilian Forest Code changes in 2012 (major environmental law in Brazil), the 39 presidential impeachment in 2016, presidential elections in 2018, and the start of the new federal 40 administration in 2019. Successful deforestation control policies are being negatively affected by 41 critical changes in the political context, and weakening the environmental rule of law, forest 42 conservation, and sustainable development programs (for example, changes in the Amazon Fund 43 governance in disagreement with the main donors). In 2019, the annual deforestation rate reached 44 10,129 km2 being the first time it surpassed 10,000 km2 since 2008 (INPE, 2021) . Besides, there has Do Not Cite, Quote or Distribute 7-105 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 been no effective transition from the historical economic model to a sustainable one. The lack of clarity 2 in the ownership of land is still a major unresolved issue in the Amazon. 3 Lessons 4 The reduction of deforestation in the Brazilian Amazon was possible due to effective political and 5 institutional support for environmental conservation. The initiatives of the Action Plan included the 6 expansion of the protected areas network (conservation unities and indigenous lands), improvement of 7 deforestation monitoring to the enforcement of environmental laws, and the use of economic 8 instruments, for example, by cutting off public credit for municipalities with higher deforestation rates 9 (Souza et al. 2013; Ricketts et al. 2010; Blackman and Veit 2018; Nepstad et al. 2014; Arima et al. 10 2014). 11 The array of public policies and social engagement was a historical and legal breakthrough in global 12 protection. However, the broader political and institutional context and actions to reduce the 13 representation and independent control of civil society movements in decision-making bodies weaken 14 this structure with significant increases in deforestation rates, burnings, and forest fires. 15 [END BOX 7.9 HERE] 16 17 [START BOX 7.10 HERE] 18 Box 7.10 Regreening the Sahel, Northern Africa 19 Case description 20 More than 200 million trees have regenerated on more than 5 Mha in the Sahel (Sendzimir et al. 2011). 21 The Maradi/Zinder region of Niger is the epicentre of experimentation and scale up. This vast 22 geographic extent generates significant mitigation potential despite the relatively modest per unit area 23 increase in carbon of about 0.4 Mg C ha-1 a-1 (Luedeling and Neufeldt 2012). In addition to the carbon 24 benefits, these agroforestry systems decrease erosion, provide animal fodder, recharge groundwater, 25 generate nutrition and income benefits and act as safety nets for vulnerable rural households during 26 climate and other shocks (Bayala et al. 2014, 2015; Binam et al. 2015; Sinare and Gordon 2015; Ilstedt 27 et al. 2016). 28 Lessons 29 A mélange of factors contributed to regreening in the Sahel. Increased precipitation, migration, 30 community development, economic volatility and local policy reform have all likely played a role 31 (Haglund et al. 2011; Sendzimir et al. 2011; Brandt et al. 2019a; Garrity and Bayala 2019); the easing 32 of forestry regulations has been particularly critical in giving farmers greater control over the 33 management and use of trees on their land (Garrity et al. 2010). This policy shift was catalysed by 34 greater regional autonomy resulting from economic decline and coincided with successful pilots and 35 NGO-led experimentation, cash-for-work, and training efforts to support changes in land management 36 (Sendzimir et al. 2011). Participation of farmers in planning and implementation helped align actions 37 with local knowledge and goals as well as market opportunities. 38 Regreening takes place when dormant seed or tree stumps sprout and are cultivated through the 39 technique, called Farmer Managed Natural Regeneration (FMNR). Without planting new trees, FMNR 40 is presumed to be cheaper than other approaches to restoration, though comparative economic analysis 41 has yet to be conducted (Chomba et al. 2020). Relatively lower investment costs are believed to have 42 contributed to the replication across the landscape. These factors worked together to contribute to a 43 groundswell of action that affected rights, access, and use of local resources (Tougiani et al. 2009). Do Not Cite, Quote or Distribute 7-106 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Regreening in the Sahel and the consequent transformation of the landscape has resulted from the 2 actions of hundreds of thousands of individuals responding to social and biophysical signals (Hanan 3 2018). This is an example for climate change mitigation, where eliminating regulations – versus 4 increasing them - has led to carbon dioxide removal. 5 [END BOX 7.10 HERE] 6 7 7.6.2.4. Mitigation Effectiveness: Additionality, Permanence and Leakage 8 Additionality, permanence and leakage have been widely discussed in the forestry and agricultural 9 mitigation literature (Murray et al. 2007), including in AR5 (Section 11.3.2 of the WGIII report) and 10 earlier assessment reports. Since the earlier assessment reports, new studies have emerged to provide 11 new insights on the effect of these issues on the credibility of forest and agricultural mitigation. This 12 assessment also provides additional context not considered in earlier assessments. 13 Typically, carbon registries will require that project developers show additionality by illustrating that 14 the project is not undertaken as a result of a legal requirement, and that the project achieves carbon 15 reductions above and beyond a business as usual. The protocols developed by the California Air 16 Resources Board to ensure permanence and additionality are strong standards and may even limit 17 participation (e.g. Ruseva et al. 2017). The business as usual is defined as past management actions by 18 the same entity that can be verified. Additionality can thus be observed in the future as a difference 19 from historical actions. This approach has been used by several countries in their UNFCCC Biennial 20 Update Reports to establish reductions in carbon emissions from avoided deforestation (e.g., Brazil and 21 Indonesia). 22 However, alternative statistical approaches have been deployed in the literature to assess additionality 23 with a quasi-experimental method that rely on developing a counterfactual (e.g. Andam et al. 2008; 24 Blackman 2015; Fortmann et al. 2017; Roopsind et al. 2019; Sills et al. 2015). In several studies, 25 additionality in avoided deforestation was established after the project had been developed by 26 comparing land-use change in treated plots where the policy or program was in effect with land use 27 change in similar untreated plot. Alternatively, synthetic matching statistically compares trends in a 28 treated region (i.e., a region with a policy) to trends in a region without the policy, and has been applied 29 in a region in Brazil (e.g., Sills et al. 2015), and at the country level in Guyana (Roopsind et al. 2019). 30 While these analyses establish that many projects to reduce deforestation have overcome hurdles related 31 to additionality (high confidence), there has not been a systematic assessment of the elements of project 32 or program design that lead to high levels of additionality. Such assessment could help developers 33 design projects to better meet additionality criteria. 34 The same experimental methods have been applied to analyse additionality of the adoption of soil 35 conservation and nutrient management practices in agriculture. Claassen et al. (2018) find that programs 36 to promote soil conservation are around 50% additional across the USA (i.e. 50% of the land enrolled 37 in soil conservation programs would not have been enrolled if not for the programme), while Woodward 38 et al. (2016) find that adoption of conservation tillage is rarely additional. Claassen et al. (2018) find 39 that payments for nutrient management plans are nearly 100% additional, although there is little 40 evidence that farmers reduce nutrient inputs when they adopt plans. It is not clear if the same policy 41 approaches would lead to additionality in other regions. 42 Permanence focuses on the potential for carbon sequestered in offsets to be released in the future due 43 to natural or anthropogenic disturbances. Most offset registries have strong permanence requirements, 44 although they vary in their specific requirements. VCS/Verra requires a pool of additional carbon credits 45 that provides a buffer against inadvertent losses. The Climate Action Reserve (CAR) protocol for forests Do Not Cite, Quote or Distribute 7-107 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 requires carbon to remain on the site for 100 years. The carbon on the site will be verified at pre- 2 determined intervals over the life of the project. If carbon is diminished on a given site, the credits for 3 the site have to be relinquished and the project developer has to use credits from their reserve fund 4 (either other projects or purchased credits) to make up for the loss. Estimates of leakage in forestry 5 projects in the AR5 suggest that it can range from 10% to over 90% in the USA (Murray et al. 2004), 6 and 20-50% in the tropics (Sohngen and Brown 2004) for forest set-asides and reduced harvesting. 7 Carbon offset protocols have made a variety of assumptions. The Climate Action Reserve (CAR) 8 assumes it is 20% in the USA. One of the voluntary protocols (Verra) uses specific information about 9 the location of the project to calculate a location specific leakage factor. 10 More recent literature has developed explicit estimates of leakage based on statistical analysis of carbon 11 projects or programs. The literature suggests that there are two economic pathways for leakage (e.g. 12 (Roopsind et al. 2019), either through a shift in output price that occurs when outputs are affected by 13 the policy or program implementation, as described in (Gan and McCarl 2007; Wear and Murray 2004; 14 Murray et al. 2004; Sohngen and Brown 2004), or through a shift in input prices and markets, such as 15 for labour or capital, as analysed in (Alix-Garcia et al. 2012; Andam et al. 2008; Fortmann et al. 2017; 16 Honey-Rosés et al. 2011). Estimates of leakage through product markets (e.g. timber prices) have 17 suggested leakage of up to 90% (Sohngen and Brown 2004; Murray et al. 2004; Gan and McCarl 2007; 18 I. Kallio and Solberg 2018), while studies that consider shifts in input markets are considerably smaller. 19 The analysis of leakage for the Guyana program by Roopsind et al. (2019) revealed no statistically 20 significant leakage in Suriname. A key design feature for any program to reduce leakage is to increase 21 incentives for complementary mitigation policies to be implemented in areas where leakage may occur. 22 Efforts to continue to draw more forests into carbon policy initiatives will reduce leakage over time 23 Roopsind et al. (2019), suggesting that if NDCs continue to encompass a broader selection of policies, 24 measures and forests over time, leakage will decline. 25 7.6.3. Assessment of current policies and potential future approaches 26 The Paris Agreement encourages a wide range of policy approaches, including REDD+, sustainable 27 management of forests, joint mitigation and adaptation, and emphasises the importance of non-carbon 28 benefits and equity for sustainable development (Martius et al. 2016). Around USD 0.7 billion yr-1 has 29 been invested in land-based carbon offsets (Table 7.4), but as noted in Streck (2012), there is a large 30 funding gap between these efforts and the scale of efforts necessary to meet 1.5 or 2.0°C targets outlined 31 in SR1.5. As Box 7.12 discusses, forestry actions could achieve up to 5.8 GtCO2 yr-1 with costs rising 32 from USD178 billion yr-1 to USD400 billion yr--1 by 2050. Over half of this investment is expected to 33 occur in Latin America, with 13% in SE Asia and 17% in Sub-Saharan Africa (Austin et al. 2020). 34 Other studies have suggested that similar sized programs are possible, although they do not quantify 35 total costs (e.g. Griscom et al. 2017; Busch et al. 2019). The currently quantified efforts to reduce net 36 emissions with forests and agricultural actions are helpful, but society will need to quickly ramp up 37 investments to achieve carbon sequestration levels consistent with high levels of mitigation. Only 2.5% 38 of climate mitigation funding goes to land-based mitigation options, an order of magnitude below the 39 potential proportional contribution (Buchner et al. 2015). 40 To date, there has been significantly less investment in agricultural projects than forestry projects to 41 reduce net carbon emissions (Table 7.4). For example, the economic potential (available up to USD100 42 tCO2-1) for soil carbon sequestration in croplands is 1.9 (0.4–6.8) GtCO2 yr-1 (Section 7.4.3.1), however, 43 less than 2% of the carbon in Table 7.4 is derived from soil carbon sequestration projects. While 44 reductions in CH4 emissions due to enteric fermentation constitute a large share of potential agricultural 45 mitigation reported in Section 7.4, agricultural CH4 emission reductions so far have been relatively 46 modest compared to forestry sequestration. The protocols to quantify emission reductions in the Do Not Cite, Quote or Distribute 7-108 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 agricultural sector are available and have been tested, and the main limitation appears to be the lack of 2 available financing or the unwillingness to re-direct current subsidies (medium confidence). 3 Although quantified emission reductions in agricultural projects are limited to date, a number of OECD 4 and economy in transition parties have reduced their net emissions through carbon storage in cropland 5 soils since 2000. These reductions in emissions have typically resulted from policy innovations outside 6 of the climate space, or market trends. For example, in the USA, there has been widespread adoption of 7 conservation tillage in the last 30 years as a labour-saving crop management technique. In Europe, 8 agricultural N2O and CH4 emissions have declined due to reductions in nutrient inputs and cattle 9 numbers (Henderson et al. 2020).These reductions may be attributed to mechanism within the Common 10 Agricultural Policy (Section 7.6.2.1), but could also be linked to higher nutrient prices in the 2000-2014 11 period. Other environmental policies could play a role, for example, efforts to reduce water pollution 12 from phosphorus in The Netherlands, may ultimately reduce cattle numbers, also lowering CH 4 13 emissions. 14 Numerous developing countries have established policy efforts to abate agricultural emissions or 15 increase carbon storage. Brazil, for instance, developed a subsidy program in 2010 to promote 16 sustainable development in agriculture, and practices that would reduce GHG emissions. Henderson et 17 al. (2020) report that this program reduced GHG emission in agricultural by up to 170 MtCO2 between 18 2010 and 2018. However, the investments in low-carbon agriculture in Brazil amounted only 2% of the 19 total funds for conventional agriculture in 2019. Programs on deforestation in Brazil had successes and 20 failures, as described in Box 7.9. Indonesia has engaged in a wide range of programs in the REDD+ 21 space, including a moratorium implemented in 2011 to prevent the conversion of primary forests and 22 peatlands to oil palm and logging concessions (Henderson et al. 2020) (Tacconi and Muttaqin 2019; 23 Wijaya et al. 2017). Efforts to restore peatlands and forests have also been undertaken. Indonesia reports 24 that results based REDD+ programs have been successful and have led to lower rates of deforestation 25 (Table 7.4). 26 Existing policies focused on GHG management in agriculture and forestry is less advanced in Africa 27 than in Latin American and Asia, however, Henderson et al. (2020) report on 10 countries in Sub 28 Saharan Africa that have included explicit policy proposals for reducing AFOLU GHG emissions 29 through their NDCs. These include efforts to reduce N2O emission, increase implementation of 30 conservation agriculture, improve livestock management, and implement forestry and grassland 31 practices, including agroforestry (Box 7.10). Within several of the NDCs, countries have explicitly 32 suggested intensification as an approach to reduce emission in the livestock sector. However, it is 33 important to note caveats associated with pursuing mitigation via intensification (Box 7.11) 34 The agricultural sector throughout the world is influenced by many policies that affect production 35 practices, crop choices and land use. It is difficult to quantify the effect of these policies on reference 36 level GHG emissions from the sector, as well as the cost estimates presented in Sections 7.4 and 7.5. 37 The presence of significant subsidy programs intended to improve farmer welfare and rural livelihoods 38 makes it more difficult to implement regulatory programs aimed at reducing net emissions in 39 agriculture, however, it may increase the potential to implement new subsidy programs that encourage 40 practices aimed at reducing net emissions (medium confidence). For instance, in the USA, crop 41 insurance can influence both crop choices and land use (Claassen et al. 2017; Miao et al. 2016), both of 42 which will affect emission trends. Regulations to limit nutrient applications have not been widely 43 considered, however, federal subsidy programs have been implemented to encourage farmers to conduct 44 nutrient management planning. 45 46 [START BOX 7.11 HERE] Do Not Cite, Quote or Distribute 7-109 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Box 7.11 Sustainable intensification within agriculture: evidence and caveats 2 Introduction 3 Sustainable intensification (SI) has received considerable attention as a suggested means of pursuing 4 increased overall production, reducing associated environmental externalities, and potentially releasing 5 agricultural land for alternative uses, such as forestry or rewilding (Godfray and Garnett 2014; Pretty 6 2018). The concept was explored within the SRCCL (SRCCL Section 5.6.4.4 and Cross-Chapter Box 7 6 in Chapter 5; (Mbow et al. 2019)). SI is context specific and dynamic, with no universally prescribed 8 methodology (HLPE 2019). Equal importance is given to enhancing sustainability as to achieving 9 agricultural intensification. The former aspect is often challenging to realise, measure and maintain. 10 The extent of sustainable intensification 11 Total global agricultural land area has remained relatively stable while overall production has increased 12 in recent decades (Section 7.3), indicating that agricultural intensification, as judged by production per 13 unit of land (OECD/FAO 2019; Petersen and Snapp 2015) has taken place. Changes in agricultural land 14 use and degradation of natural resources (UN Environment 2019; IPBES 2019) suggests however that 15 not all of this intensification is sustainable. Although agricultural intensification has led to less GHG 16 emissions compared to a scenario where that intensification had not taken place (Burney et al. 2010), 17 absolute agriculture related emissions have continued to increase (Section 7.2). Active pursuit of SI was 18 found to be expanding, with implementation on an increasing area, notably in developing countries 19 (Pretty et al. 2018), yet regional agricultural area expansion at the expense of native habitat also 20 continues in such regions (Section 7.3). Although there are specific examples of SI (Box 7.13) global 21 progress in achieving SI is acknowledged as slow (Cassman and Grassini 2020) with potentially 22 multiple, context specific geophysical and socio-economic barriers to implementation (Silva et al. 23 2021a; Firbank et al. 2018). 24 Preconditions to ensure sustainable intensification 25 Increasing the total amount of product produced by improving production efficiency (output per unit of 26 input) does not guarantee SI. It will only be successful if increased production efficiency translates into 27 reduced environmental and social impacts as well as increased production. For example, AR5 28 highlighted a growing emphasis on reducing GHG emissions per unit of product via increasing 29 production efficiency (Smith et al. 2014), but reductions in absolute GHG emissions will only occur 30 when production efficiency increases at a greater rate than the rate at which production increases (Clark 31 et al. 2005). 32 Defined indicators are required. Measurement of SI requires multiple indicators and metrics. It can be 33 assessed at farm, regional or global scales and temporal aspect must be considered. SI may warrant 34 whole system redesign or land reallocation (Garnett et al. 2013; Pretty et al. 2018). Accordingly, there 35 is high agreement concerning the need to consider multiple environmental and social outcomes at wider 36 spatial scales, such as catchments or regions (Weltin et al. 2018; Bengochea Paz et al. 2020; Cassman 37 and Grassini 2020). Impacts may be considered in relative terms (per area or product unit), with 38 relationships potentially antagonistic. Both area- and product unit-based metrics are valid, relevant 39 under different contexts and useful in approaching SI, but do not capture overall impacts and trade-offs 40 (Garnett 2014). To reduce the risk of unsustainable intensification, quantitative data and selection of 41 appropriate metrics to identify and guide strategies are paramount (Garnett et al. 2013; Gunton et al. 42 2016; Cassman and Grassini 2020). 43 Avoiding unsustainable intensification Do Not Cite, Quote or Distribute 7-110 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 It is critical that intensification does not drive expansion of unsustainable practices. Increased 2 productivity with associated economic reward could incentivise and reward agricultural land expansion, 3 or environmentally and socially damaging practices on existing and former agricultural land (Phalan 4 2018; Ceddia et al. 2013). Accordingly, coordinated policies are crucial to ensuring desired outcomes 5 (Reddy et al. 2020; Kassam and Kassam 2020; Godfray and Garnett 2014; ). Barretto et al. (2013) found 6 that in agriculturally consolidated areas, land-use intensification coincided with either a contraction of 7 both cropland and pasture areas, or cropland expansion at the expense of pastures, both resulting in a 8 stable farmed area. In contrast, in agricultural frontier areas, land-use intensification coincided with 9 expansion of agricultural lands. 10 In conclusion, SI within agriculture is needed given the rising global population and the need to address 11 multiple environmental and social externalities associated with agricultural activities. However, 12 implementation requires strong stakeholder engagement, appropriate regulations, rigorous monitoring 13 and verification and comprehensive outreach and knowledge exchange programmes. 14 [END BOX 7.11 HERE] 15 16 A factor that will influence future carbon storage in so-called land-based reservoirs involves considering 17 short- and long-term climate benefits, as well as interactions among various natural climate solution 18 options. The benefits of various natural climate solutions depend on a variety of spatially dependent 19 issues as well as institutional factors, including their management status (managed or unmanaged 20 systems), their productivity, opportunity costs, technical difficulty of implementation, local willingness 21 to consider, property rights and institutions, among other factors. Biomass energy, as described 22 elsewhere in this chapter and in (Cross-Working Group Box 3 in Chapter 12), is a potential example of 23 an option with trade-offs that emerge when policies favour one type of mitigation strategy over another. 24 Bioenergy production needs safeguards to limit negative impacts on carbon stocks on the land base as 25 is already in place in the EU Renewable Energy Directive and several national schemes in Netherlands, 26 UK and Denmark. (DeCicco and Schlesinger 2018; (Favero et al. 2020; Buchholz et al. 2016; Khanna 27 et al. 2017). It is argued that a carbon tax on only fossil fuel derived emissions, may lead to massive 28 deployment of bioenergy, although the effects of such a policy can be mitigated when combined with 29 policies that encourage sustainable forest management and protection of forest carbon stocks as well as 30 forest management certification (Favero et al. 2020 and Nabuurs et al. 2017, Baker et al, 2019) (high 31 confidence). 32 If biomass energy production expands and shifts to carbon capture and storage (e.g. BECCS) during the 33 century, there could be a significant increase in the area of crop and forestland used for biomass energy 34 production (Sections 7.4 and 7.5). BECCS is not projected to be widely implemented for several 35 decades, but in the meantime, policy efforts to advance land based measures including reforestation and 36 restoration activities (Strassburg et al. 2020) combined with sustainable management and provision of 37 agricultural and wood products are widely expected to increase the terrestrial pool of carbon (Cross- 38 Working Group Box 3 in Chapter 12). Carbon sequestration policies, sustainable land management 39 (forest and agriculture), and biomass energy policies can be complementary (Favero et al. 2017; Baker 40 et al. 2019). However, if private markets emerge for biomass and BECCS on the scale suggested in the 41 SR1.5, policy efforts must ramp up to substantially value, encourage, and protect terrestrial carbon 42 stocks and ecosystems to avoid outcomes inconsistent with many SDGs (high confidence). 43 7.6.4. Barriers and opportunities for AFOLU mitigation 44 The AR5 and other assessments have acknowledged many barriers and opportunities to effective 45 implementation of AFOLU measures. Many of these barriers and opportunities focus on the context in Do Not Cite, Quote or Distribute 7-111 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 developing countries, where a significant portion of the world's cost-effective mitigation exists, but 2 where domestic financing for implementation is likely to be limited. The SSPs capture some of this 3 context, and as a result, IAMs (Section 7.5) exhibit a wide range of land-use outcomes, as well as 4 mitigation potential. Potential mitigation, however, will be influenced by barriers and opportunities that 5 are not considered by IAMs or by bottom-up studies reviewed here. For example, more efficient food 6 production systems, or sustainable intensification within agriculture, and globalised trade could enhance 7 the extent of natural ecosystems leading to lower GHG emissions from the land system and lower food 8 prices (Popp et al. 2017), but this (or any) pathway will create new barriers to implementation and 9 encourage new opportunities, negating potential benefits (Box 7.11). It is critically important to 10 consider the current context in any country. 11 7.6.4.1. Socio-economic barriers and opportunities 12 Design and coverage of financing mechanisms. The lack of resources thus far committed to 13 implementing AFOLU mitigation, income and access to alternative sources of income in rural 14 households that rely on agriculture or forests for their livelihoods remains a considerable barrier to 15 adoption of AFOLU (high confidence). Section 7.6.1 illustrates that to date only USD0.7 billion yr-1 16 has been spent on AFOLU mitigation, well short of the more than USD400 billion yr -1 that would be 17 needed to achieve the economic potential described in Section 7.4. Despite long-term recognition that 18 AFOLU can play an important role in mitigation, the economic incentives necessary to achieve AFOLU 19 aspirations as part of the Paris Agreement or to maintain temperatures below 2.0 o C have not emerged. 20 Without quickly ramping up spending, the lack of funding to implement projects remains a substantial 21 barrier (high confidence). Investments are critically important in the livestock sector, which has the 22 highest emissions reduction potential among options because actions in the sector influence agriculture 23 specific activities, such as enteric fermentation, as well as deforestation (Mayberry et al. 2019). In many 24 countries with export-oriented livestock industries, livestock farmers control large swaths of forests or 25 re-forestable areas. Incentive mechanisms and funding can encourage adoption of mitigation strategies, 26 but funding is currently too low to make consistent progress. 27 Scale and accessibility of financing. The largest share of funding to date has been for REDD+, and 28 many of the commitments to date suggest that there will be significant funding in this area for the 29 foreseeable future. Funding for conservation programs in OECD countries and China affects carbon, 30 but has been driven by other objectives such as water quality and species protection. Considerably less 31 funding has been available for agricultural projects aimed at reducing carbon emissions, and outside of 32 voluntary markets, there do not appear to be large sources of funding emerging either through 33 international organisations, or national programs. In the agricultural sector, funding for carbon must be 34 obtained by redirecting existing resources from non-GHG conservation to GHG measures, or by 35 developing new funding streams (Henderson et al. 2020). 36 Risk and uncertainty. Most approaches to reduce emissions, especially in agriculture, require new or 37 different technologies that involve significant time or financial investments by the implementing 38 landholders. Adoption rates are often slow due to risk aversion among agricultural operators. Many 39 AFOLU measures require carbon to be compensated to generate positive returns, reducing the 40 likelihood of implementation without clear financial incentives. Research to show costs and benefits is 41 lacking in most parts of the world. 42 43 [START BOX 7.12 HERE] 44 Box 7.12 Financing AFOLU mitigation; what are the costs and who pays? Do Not Cite, Quote or Distribute 7-112 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 Achieving the large contribution to mitigation that the AFOLU sector can make requires public and 2 private investment. Austin et al. (2020) estimate that in forestry, USD178 billion yr-1 is needed over the 3 next decade to achieve 5 GtCO2 yr-1, and investments need to ramp up to USD400 billion yr-1 by 2050 4 to expand effort to 6 GtCO2 yr-1. Other land-based options, such as mangrove protection, peatland 5 restoration, and agricultural options would increase this total cost estimate, but have smaller to 6 negligible opportunity costs. 7 Financing needs in AFOLU, and in particular in forestry, include both the direct effects of any changes 8 in activities – costs of planting or managing trees, net revenues from harvesting, costs of thinning, costs 9 of fire management, etc. – as well as the opportunity costs associated with land use change. Opportunity 10 costs are a critical component of AFOLU finance, and must be included in any estimate of the funds 11 necessary to carry out projects. They are largest, as share of total costs, in forestry because they play a 12 prominent role in achieving high levels of afforestation, avoided deforestation, and improved forest 13 management. In case of increasing soil carbon in croplands through reduced tillage, there are often cost 14 savings associated with increased residues because there is less effort tilling, but the carbon effects per 15 hectare are also modest. There could, however, be small opportunity costs in cases where residues may 16 otherwise be marketed to a biorefinery. The effect of reduced tillage on yields varies considerably across 17 sites and crop types, but tends to enhance yields modestly in the longer-run. 18 Opportunity costs are a direct financing costs for activities that require land uses to change. For instance 19 a government can encourage planting forests on agricultural land by (a) requiring it, (b) setting up a 20 market or market-based incentives, or (c) buying the land and doing it themselves. In each case, the 21 required investment is the same – the planting cost plus the net foregone returns of agricultural rents – 22 even though a different entity pays the cost. Private entities that pay for carbon credits will also bear 23 the direct costs of planting plus the opportunity costs. In the case of avoided deforestation, opportunity 24 costs similarly must be paid to individual actors to avoid the deforestation. 25 [END BOX 7.12 HERE] 26 27 Poverty. Mitigation and adaptation can have important implications for vulnerable people and 28 communities, for example, mitigation activities consistent with scenarios examined in the SR1.5 could 29 raise food and fiber prices globally (Section. 7.5). In the NDCs, 82 Parties included references to social 30 issues (e.g. poverty, inequality, human well-being, marginalisation), with poverty the most cited factor 31 (70 Parties). The number of hungry and food insecure people in the world is growing, reaching 821 32 million in 2017, or one in every nine people (FAO 2018b), and two-thirds live in rural areas (Laborde 33 Debucquet et al. 2020). Consideration of rural poverty and food insecurity is central in AFOLU 34 mitigation because there are a large number of farms in the world (about 570 million), and most are 35 smaller than 2 hectares. It is important to better understand how different mitigation policies affect the 36 poor. 37 Cultural values and social acceptance. Barriers to adoption of AFOLU mitigation will be strongest 38 where historical practices represent long-standing traditions (high confidence). Adoption of new 39 mitigation practices, however, may proceed quickly if the technologies can be shown to improve crop 40 yields, reduce costs, or otherwise improve livelihoods (Ranjan 2019). AR6 presents new estimates of 41 the mitigation potential for shifts in diets and reductions in food waste, but given long-standing dietary 42 traditions within most cultures, some of the strongest barriers exist for efforts to change diets (medium 43 confidence). Furthermore, the large number of undernourished who may benefit from increased calories 44 and meat will complicate efforts to change diets. Regulatory or tax approaches will face strong 45 resistance, while efforts to use educational approaches and voluntary measures have limited potential 46 to slow changes in consumption patterns due to free-riders, rebound effects, and other limitations. Food Do Not Cite, Quote or Distribute 7-113 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 loss and waste occurs across the supply chain, creating significant challenges to reduce it (FAO 2019c). 2 Where food loss occurs in the production stage, i.e. in fields at harvest, there may be opportunities to 3 align reductions in food waste with improved production efficiency, however, adoption of new 4 production methods often requires new investments or changes in labour practices, both of which are 5 barriers. 6 7.6.4.2. Institutional barriers and opportunities 7 Transparent and accountable governance. Good governance and accountability are crucial for 8 implementation of forest and agriculture mitigation. Effective nature-based mitigation will require 9 large-scale estimation, modelling, monitoring, reporting and verification of GHG inventories, 10 mitigation actions, as well as their implications for sustainable development goals and their interactions 11 with climate change impacts and adaptation. Efforts must be made to integrate the accounting from 12 projects to the country level. While global datasets have emerged to measure forest loss, at least 13 temporarily (e.g. Hansen et al., 2013), similar datasets do not yet exist for forest degradation and 14 agricultural carbon stocks or fluxes. Most developing countries have insufficient capacity to address 15 research needs, modelling, monitoring, reporting and data requirements (e.g. Ravindranath et al. 2017), 16 compromising transparency, accuracy, completeness, consistency and comparability. 17 Opportunity for political participation of local stakeholders is barrier in most places where forest 18 ownership rights are not sufficiently documented (Essl et al. 2018). Since incentives for self- 19 enforcement can have an important influence on deforestation rates (Fortmann et al. 2017), weak 20 governance and insecure property rights are significant barriers to introduction of forest carbon offset 21 projects in developing countries, where many of the low-cost options for such projects exist (Gren and 22 Zeleke 2016). Governance challenges exist at all levels of government, with poor coordination, 23 insufficient information sharing, and concerns over accountability playing a prominent role within 24 REDD+ projects and programs (Ravikumar et al. 2015). In some cases, governments are increasingly 25 centralising REDD+ governance and limiting the distribution of governance functions between state 26 and non-state actors (Zelli et al. 2017; Phelps et al. 2010). Overlap and duplication in FLEGT and 27 REDD+ also limits governance effectiveness (Gupta et al. 2016). 28 Clear land tenure and land-use rights. Unclear property rights and tenure insecurity undermine the 29 incentives to improve forest and agricultural productivity, lead to food insecurity, undermine REDD+ 30 objectives, discourage adoption of farm conservation practices, discourage tree planting and forest 31 management, and exacerbate conflict between different land users (Sunderlin et al. 2018; Antwi-Agyei 32 et al. 2015; Borras and Franco 2018; Felker et al. 2017; Riggs et al. 2018; Kansanga and Luginaah 33 2019). Some positive signs exist as over 500 million hectares of forests have been converted to 34 community management with clear property rights in the past two decades (Rights and Resources 35 Initiative 2018), but adoption of forest and agricultural mitigation practices will be limited in large 36 remaining areas with unclear property rights (Gupta et al. 2016). 37 Lack of institutional capacity. Institutional complexity, or lack thereof, represents a major challenge 38 when implementing large and complex mitigation programs (e.g., REDD+) in agriculture, forest and 39 other land uses (Bäckstrand et al. 2017). Without sufficient capacity, many synergies between 40 agricultural and forest programs, or mitigation and adaptation opportunities, may be missed (Duguma 41 et al. 2014). Another aspect of institutional complexity is the different biophysical and socio-economic 42 circumstances as well as the public and private financial means involved in the architecture and 43 implementation of REDD+ and other initiatives (Zelli et al. 2017). 44 7.6.4.3. Ecological barriers and opportunities 45 Availability of land and water. Climate mitigation scenarios in the two recent special reports (SR1.5C 46 and SRLCC) that aim to limit global temperature increase to 2°C or less involve carbon dioxide (CO2) Do Not Cite, Quote or Distribute 7-114 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 removal from the atmosphere. To support large-scale CDR, these scenarios involve significant land-use 2 change, due to afforestation/reforestation, avoided deforestation, and deployment of Biomass Energy 3 with Carbon Capture and Storage (BECCS). While a considerable amount of land is certainly available 4 for new forests or new bioenergy crops, that land has current uses that will affect not only the costs, but 5 also the willingness of current users or owners, to shift uses. Regions with private property rights and a 6 history of market-based transactions may be the most feasible for land use change or land management 7 change to occur. Areas with less secure tenure or a land market with fewer transactions in general will 8 likely face important hurdles that limit the feasibility of implementing novel nature-based solutions. 9 Implementation of nature-based solution may have local or regionally important consequences for other 10 ecosystem services, some of which may be negative (high confidence). Land use change has important 11 implications for the hydrological cycle, and the large land use shifts suggested for BECCS when not 12 carried out in a carefully planned manner, are expected to increase water demands substantially across 13 the globe (Stenzel et al. 2019; Rosa et al. 2020). Afforestation can have minor to severe consequences 14 for surface water acidification, depending on site-specific factors and exposure to air pollution and sea- 15 salts (Futter et al. 2019). The potential effects of coastal afforestation on sea-salt related acidification 16 could lead to re-acidification and damage on aquatic biota. 17 Specific soil conditions, water availability, GHG emission-reduction potential as well as natural 18 variability and resilience. Recent analysis by (Cook-Patton et al. 2020) illustrates large variability in 19 potential rates of carbon accumulation for afforestation and reforestation options, both within 20 biomes/ecozones and across them. Their results suggest that while there is large potential for 21 afforestation and reforestation, the carbon uptake potential in land-based climate change mitigation 22 efforts is highly dependent on the assumptions related to climate drivers, land use and land management, 23 and soil carbon responses to land-use change. Less analysis has been conducted on bioenergy crop 24 yields, however, bioenergy crop yields are also likely to be highly variable, suggesting that bioenergy 25 supply could exceed or fall short of expectations in a given region, depending on site conditions. 26 The effects of climate change on ecosystems, including changes in crop yields, shifts in terrestrial 27 ecosystem productivity, vegetation migration, wildfires, and other disturbances also will affect the 28 potential for AFOLU mitigation. Climate is expected to reduce crop yields, increase crop and livestock 29 prices, and increase pressure on undisturbed forest land for food production creating new barriers and 30 increasing costs for implementation of many nature-based mitigation techniques (IPCC WGII AR6 31 Chapter 5) (medium confidence). 32 The observed increase in the terrestrial sink over the past half century can be linked to changes in the 33 global environment, such as increased atmospheric CO2 concentrations, N deposition, or changes in 34 climate (Ballantyne et al. 2012), though not always proven from ground-based information 35 (Vandersleen et al.2015). While the terrestrial sink relies on regrowth in secondary forests (Houghton 36 and Nassikas 2017), there is emerging evidence that the sink will slow in the northern hemisphere as 37 forests age (Nabuurs et al. 2013), although saturation may take decades (Zhu et al. 2018). Forest 38 management through replanting, variety selection, fertilisation, and other management techniques, has 39 increased the terrestrial carbon sink over the last century (Mendelsohn and Sohngen 2019). Saturation 40 of the sink in situ may not occur when e.g. substitution effects of timber usage are also 41 considered. 42 Increasing concentrations of CO2 are expected to increase carbon stocks globally, with the strongest 43 effects in the tropics (Kim et al. 2017a; Schimel et al. 2015; AR6 WGI, Fig SPM7)) and economic 44 models suggest that future sink potential may be robust to the impacts of climate change (Tian et al. 45 2018). However, it is uncertain if this large terrestrial carbon sink will continue in the future (e.g. 46 Aragão et al. 2018), as it is increasingly recognized that gains due to CO2 fertilization are constrained Do Not Cite, Quote or Distribute 7-115 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 by climate and increasing disturbances (Schurgers et al. 2018; Duffy et al. 2021; IPCC WGII AR6 2 Chapter 5). Further, negative synergies between local impacts like deforestation and forest fires may 3 interact with global drivers like climate change and lead to tipping points (Lovejoy and Nobre 2018; ). 4 Factors that reduce permanence or slow forest growth will drive up costs of forest mitigation measures, 5 suggesting that climate change presents a formidable challenge to implementation of nature-based 6 solutions beyond 2030 (high confidence). 7 In addition to climate change, Dooley and Kartha (2018) also note that technological and social factors 8 could ultimately limit the feasibility of agricultural and forestry mitigation options, especially when 9 deployed at large-scale. Concern is greatest with widespread use of bioenergy crops, which could lead 10 to forest losses (Harper et al. 2018).Deployment of BECCS and forest-based mitigation can be 11 complementary (Favero et al. 2017; Baker et al. 2019), but inefficient policy approaches could lead to 12 net carbon emissions if BECCS replaces high-carbon content ecosystems with crops. 13 Adaptation benefits and biodiversity conservation. Biodiversity may improve resilience to climate 14 change impacts as more-diverse systems could be more resilient to climate change impacts, thereby 15 maintaining ecosystem function and preserving biodiversity (Hisano et al. 2018). However, losses in 16 ecosystem functions due species shifts or reductions in diversity may impair the positive effects of 17 biodiversity on ecosystems. Forest management strategies based on biodiversity and ecosystems 18 functioning interactions can augment the effectiveness of forests in reducing climate change impacts on 19 ecosystem functioning (high confidence). In spite of the many synergies between climate policy 20 instruments and biodiversity conservation, however, current policies often fall short of realising this 21 potential (Essl et al. 2018). 22 7.6.4.4. Technological barriers and opportunities 23 Monitoring, reporting, and verification. Development of satellite technologies to assess potential 24 deforestation has grown in recent years with the release of 30 m data by Hansen et al. (2013), however, 25 this data only captures tree cover loss, and increasing accuracy over time may limit its use for trend 26 analysis (Ceccherini et al. 2020; Palahí et al. 2021). Datasets on forest losses are less well developed 27 for reforestation and afforestation. As Mitchell et al. (2017) point out, there has been significant 28 improvement in the ability to measure changes in tree and carbon density on sites using satellite data, 29 but these techniques are still evolving and improving and they are not yet available for widespread use. 30 Ground-based forest inventory measurements have been developed in many countries, most 31 prominently in the northern hemisphere, but more and more countries are starting to develop and collect 32 national forest inventories. Training and capacity building is going on in many developing countries 33 under UNREDD and FAO programmes. Additional efforts to harmonize data collection methods and 34 to make forest inventory data available to the scientific community would improve confidence in forest 35 statistics, and changes in forest statistics over time. To some extent the Global Forest Biodiversity 36 Initiative fills in this data gap (https://gfbi.udl.cat/). 37 7.6.5. Linkages to ecosystem services, human well-being and adaptation (incl. SDGs) 38 The linkage between biodiversity, ecosystem services, human well-being and sustainable development 39 is widely acknowledged (Millenium Ecosystem Assesment 2005; UN Environment 2019). Loss of 40 biodiversity and ecosystem services will have an adverse impact on quality of life, human well-being 41 and sustainable development (Díaz et al. 2019). Such losses will not only affect current economic 42 growth but also impede the capacity for future economic growth. 43 Population growth, economic development, urbanisation, technology, climate change, global trade and 44 consumption, policy and governance are key drivers of global environmental change over recent 45 decades (Kram et al. 2014; UN Environment 2019; WWF 2020). Changes in biodiversity and ecosystem Do Not Cite, Quote or Distribute 7-116 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 services are mainly driven by habitat loss, climate change, invasive species, over-exploitation of natural 2 resources, and pollution (Millenium Ecosystem Assesment 2005). The relative importance of these 3 drivers varies across biomes, regions, and countries. Climate change is expected to be a major driver of 4 biodiversity loss in the coming decades, followed by commercial forestry and bioenergy production 5 (OECD 2012; UN Environment 2019). Population growth along with rising incomes and changes in 6 consumption and dietary patterns, will exert immense pressure on land and other natural resources 7 (IPCC. 2019a). Current estimates suggest that 75% of the land surface has been significantly 8 anthropogenically altered, with 66% of the ocean area experiencing increasing cumulative impacts and 9 over 85% of wetland area lost (Díaz et al. 2019). As discussed, in section 7.3, land-use change is driven 10 amongst others by agriculture, forestry (logging and fuelwood harvesting), infrastructural development 11 and urbanisation, all of which may also generate localised air, water, and soil pollution (Díaz et al. 12 2019). Over a third of the world’s land surface and nearly three-quarters of available freshwater 13 resources are devoted to crop or livestock production (Díaz et al. 2019). Despite a slight reduction in 14 global agricultural area since 2000, regional agricultural area expansion has occurred in Latin America 15 and the Caribbean, Africa and the Middle East (FAO 2019; (OECD/FAO 2019). The degradation of 16 tropical forests and biodiversity hotspots, endangers habitat for many threatened and endemic species, 17 and reduces valuable ecosystem services. However, trends vary considerably by region. As noted in 18 Section 7.3, global forest area declined by roughly 178 Mha between 1990 and 2020 (FAO 2020a), 19 though the rate of net forest loss has decreased over the period, due to reduced deforestation in some 20 countries and forest gains in others. Between 1990 to 2015, forest cover fell by almost 13% in Southeast 21 Asia, largely due to an increase in timber extraction, large-scale biofuel plantations and expansion of 22 intensive agriculture and shrimp farms, whereas in Northeast Asia and South Asia it increased by 23% 23 and 6% respectively, through policy instruments such as joint forest management, payment for 24 ecosystem services, and restoration of degraded forests (IPBES 2018b). It is lamenting that the area 25 under natural forests which are rich in biodiversity and provide diverse ecosystem services decreased 26 by 301 Mha between 1990 and 2020, decreasing in most regions except Europe and Oceania with largest 27 losses reported in Sub-Saharan Africa (FAO 2020a). The increasing trend of mining in forest and coastal 28 areas, and in river basins for extracting has had significant negative impacts on biodiversity, air and 29 water quality, water distribution, and on human health (Section 7.3). Freshwater ecosystems equally 30 face a series of combined threats including from land-use change, water extraction, exploitation, 31 pollution, climate change and invasive species (Díaz et al. 2019). 32 7.6.5.1. Ecosystem services 33 An evaluation of eighteen ecosystem services over the past five decades (1970-2019) found only four 34 (agricultural production, fish harvest, bioenergy production and harvest of materials) to demonstrate 35 increased performance, while the remaining fourteen, mostly concerning regulating and non-material 36 contributions, were found to be in decline (Díaz et al. 2019). The value of global agricultural output 37 (over USD3.54 trillion in 2018) had increased approximately threefold since 1970, and roundwood 38 production (industrial roundwood and fuelwood) by 27%, between 1980 to 2018, reaching some 4 39 billion m3 in 2018. However, the positive trends in these four ecosystem services does not indicate long- 40 term sustainability. If increases in agricultural production are realised through forest clearance or 41 through increasing energy-intensive inputs, gains are likely to be unsustainable in the long run. 42 Similarly, an increase in fish production may involve overfishing, leading to local species declines 43 which also impacts fish prices, fishing revenues, and the well-being of coastal and fishing communities 44 (Sumaila and Lam 2020). Climate change and other drivers are likely to affect future fish catch 45 potential, although impacts will differ across regions (Sumaila et al. 2017; IPCC 2019b). 46 The increasing trend in aquaculture production especially in South and Southeast Asia through intensive 47 methods affects existing food production and ecosystems by diverting rice fields or mangroves Do Not Cite, Quote or Distribute 7-117 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 (Bhattacharya and Ninan 2011). Although extensive traditional fish farming of carp in central Europe 2 can contribute to landscape management, enhance biodiversity and provide ecosystem services, there 3 are several barriers to scale up production due to strict EU environmental regulations, vulnerability to 4 extreme weather events, and to avian predators that are protected by EU laws, and disadvantages faced 5 by small-scale enterprises that dominate the sector (European-Commission 2021). Bioenergy 6 production may have high opportunity costs in land-scarce areas and compete with land used for food 7 production which threatens food security and affects the poor and vulnerable. But these impacts will 8 differ across scale, contexts and other factors. 9 Currently, land degradation is estimated to have reduced productivity in 23% of the global terrestrial 10 area, and between USD235 billion and USD577 billion in annual global crop output is at risk because 11 of pollinator loss (Díaz et al. 2019). The global trends reviewed above are based on data from 2,000 12 studies. It is not clear whether the assessment included a quality control check of the studies evaluated 13 and suffer from aggregation bias. For instance, a recent meta-analysis of global forest valuation studies 14 noted that many studies reviewed had shortcomings such as failing to clearly mention the methodology 15 and prices used to value the forest ecosystem services, double counting, data errors, etc, (Ninan and 16 Inoue 2013). Furthermore the criticisms against the paper by (Costanza et al. 1997), such as ignoring 17 ecological feedbacks and non-linearities that are central to the processes that link all species to each 18 other and their habitats, ignoring substitution effects may also apply to the global assessment (Smith 19 1997); Bockstael et al. 2000; Loomis et al. 2000). Land degradation has had a pronounced impact on 20 ecosystem functions worldwide (Scholes et al. 2018). Net primary productivity of ecosystem biomass 21 and of agriculture is presently lower than it would have been under a natural state on 23% of the global 22 terrestrial area, amounting to a 5% reduction in total global net primary productivity (Scholes et al. 23 2018). Over the past two centuries, soil organic carbon, an indicator of soil health, has seen an estimated 24 8% loss globally (176 GtC) from land conversion and unsustainable land management practices 25 (Scholes et al. 2018). Projections to 2050 predict further losses of 36 GtC from soils, particularly in 26 Sub-Saharan Africa. These losses are projected to come from the expansion of agricultural land into 27 natural areas (16 GtC), degradation due to inappropriate land management (11 GtC) and the draining 28 and burning of peatlands (9 GtC) and melting of permafrost (Scholes et al. 2018). Trends in biodiversity 29 measured by the global living planet index between 1970 to 2016 indicate a 68% decline in monitored 30 population of mammals, birds, amphibians, reptiles, and fish WWF 2020). FAO’s recent report on the 31 state of the world’s biodiversity for food and agriculture points to an alarming decline in biodiversity 32 for food and agriculture including associated biodiversity such as pollination services, micro-organisms 33 which are essential for production systems (FAO 2019d). These suggest that overall ecosystem health 34 is consistently declining with adverse consequences for good quality of life, human well-being, and 35 sustainable development. 36 Although numerous studies have estimated the value of ecosystem services for different sites, 37 ecosystems, and regions, these studies mostly evaluate ecosystem services at a single point in time See 38 (Costanza et al. 1997; Nahuelhual et al. 2007; de Groot et al. 2012; Ninan and Kontoleon 2016; Xue 39 and Tisdell 2001). The few studies that have assessed the trends in the value of ecosystem services 40 provided by different ecosystems across regions and countries indicate a declining trend (Costanza et 41 al. 2014; Kubiszewski et al. 2017). Land use change is a major driver behind loss of biodiversity and 42 ecosystem services in most regions (Archer et al. 2018; Rice et al. 2018; IPBES 2018b; M. Fischer et 43 al. 2018). Projected impacts of land use change and climate change on biodiversity and ecosystem 44 services (material and regulating services) between 2015 to 2050 were assessed to have relatively less 45 negative impacts under global sustainability scenarios as compared to regional competition and 46 economic optimism scenarios (Díaz et al. 2019). The projected impacts were based on a subset of 47 Shared Socioeconomic Pathway (SSP) scenarios and GHG emissions trajectories (RCP) developed in Do Not Cite, Quote or Distribute 7-118 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 support of IPCC assessments. There are synergies, trade-offs and co-benefits between ecosystem 2 services and mitigation options with impacts on ecosystem services differing by scale and contexts 3 (high confidence). Measures such as conservation agriculture, agroforestry, soil and water conservation, 4 afforestation, adoption of silvopastoral systems, can help minimise trade-offs between mitigations 5 options and ecosystem services (Duguma et al. 2014). Climate smart agriculture (CSA) is being 6 promoted to enable farmers to make agriculture more sustainable and adapt to climate change (Box 7 7.4). However, experience with CSA in Africa has not been encouraging. For instance, a study of 8 climate smart cocoa production in Ghana shows that due to lack of tenure (tree) rights, bureaucratic and 9 legal hurdles in registering trees in cocoa farms, and other barriers small cocoa producers could not 10 realise the project benefits (Box 7.13). Experience of CSA in some other Sub-Saharan African countries 11 and other countries such as Belize too has been constrained by weak extension systems and policy 12 implementation, and other barriers (Arakelyan et al. 2017; Kongsager 2017). 13 14 START BOX 7.13 HERE 15 Box 7.13 Case study: climate smart cocoa production in Ghana 16 Policy Objectives 17 1. To promote sustainable intensification of cocoa production and enhance the adaptive capacity of 18 small cocoa producers. 19 2. To reduce cocoa-induced deforestation and GHG emissions. 20 3. To improve productivity, incomes, and livelihoods of smallholder cocoa producers. 21 Policy Mix 22 The climate smart cocoa (CSC) production programme in Ghana involved distributing shade tree 23 seedlings that can protect cocoa plants from heat and water stress, enhance soil organic matter and water 24 holding capacity of soils, and provide other assistance with agroforestry, giving access to extension 25 services such as agronomic information and agro-chemical inputs. The shade tree seedlings were 26 distributed by NGOs, government extension agencies, and the private sector free of charge or at 27 subsidised prices and was expected to reduce pressure on forests for growing cocoa plants. The CSC 28 programme was mainly targeted at small farmers who constitute about 80% of total farm holdings in 29 Ghana. Although the government extension agency (Cocobod) undertook mass spraying or pruning of 30 cocoa farms they found it difficult to access the 800,000 cocoa smallholders spread across the tropical 31 south of the country. The project brought all stakeholders together i.e., the government, private sector, 32 local farmers and civil society or NGOs to facilitate the sustainable intensification of cocoa production 33 in Ghana. Creation of a community-based governance structure was expected to promote benefit 34 sharing, forest conservation, adaptation to climate change, and enhanced livelihood opportunities. 35 Governance Context 36 Critical enablers 37 The role assigned to local government mechanisms such as Ghana’s Community Resource Management 38 Area Mechanisms (CREMAs) was expected to give a voice to smallholders who are an important 39 stakeholder in Ghana’s cocoa sector. CREMAs are inclusive because authority and ownership of natural 40 resources are devolved to local communities who can thus have a voice in influencing CSC policy 41 thereby ensuring equity and adapting CSC to local contexts. However, ensuring the long-term 42 sustainability of CREMAs will help to make them a reliable mechanism for farmers to voice their Do Not Cite, Quote or Distribute 7-119 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 concerns and aspirations, and ensure their independence as a legitimate governance structure in the long 2 run. The private sector was assigned an important role to popularise climate smart cocoa production in 3 Ghana. However, whether this will work to the advantage of smallholder cocoa producers needs to be 4 seen. 5 Critical barriers 6 The policy intervention overlooks the institutional constraints characteristic of the cocoa sector in 7 Ghana where small farmers are dominant and have skewed access to resources and markets. Lack of 8 secure tenure (tree rights) where the ownership of shade trees and timber vests with the state, 9 bureaucratic and legal hurdles to register trees in their cocoa farms are major constraints that impede 10 realisation of the expected benefits of the CSC programme. This is a great disincentive for small cocoa 11 producers to implement CSC initiatives and nurture the shade tree seedlings and undertake land 12 improvement measures. The state marketing board has the monopoly in buying and marketing of cocoa 13 beans including exports which impeded CREMAs or farming communities from directly selling their 14 produce to MNCs and traders. However, many MNCs have been involved in setting up of CREMA or 15 similar structures, extending premium prices and non-monetary benefits (access to credit, shade tree 16 seedlings, agro-chemicals) thus indirectly securing their cocoa supply chains. A biased ecological 17 discourse about the benefits of climate smart agriculture and sustainable intensive narrative, 18 complexities regarding the optimal shade levels for growing cocoa, and dependence on agro-chemicals 19 are issues that affect the success and sustainability of the project intervention. Dominance of private 20 sector players especially MNCs in the sector may be detrimental to the interests of smallholder cocoa 21 producers. 22 Source: Nasser et al. (2020) 23 END BOX 7.13 HERE 24 25 7.6.5.2. Human well-being and Sustainable Development Goals 26 Conservation of biodiversity and ecosystem services is part of the larger objective of building climate 27 resilience and promoting good quality of life, human well-being and sustainable development. While 28 two of the 17 SDGs directly relate to nature (SDGs 14 and 15 covering marine and terrestrial ecosystems 29 and biodiversity), most other SDGs relating to poverty, hunger, inequality, health and well-being, clean 30 sanitation and water, energy, etc., are directly or indirectly linked to nature (Blicharska et al. 2019). A 31 survey among experts to assess how 16 ecosystem services could help in achieving the SDGs relating 32 to good environment and human well-being suggested that ecosystem services could contribute to 33 achieving about 41 targets across 12 SDGs (Wood et al. 2018). They also indicated cross-target 34 interactions and synergetic outcomes across many SDGs. Conservation of biodiversity and ecosystem 35 services is critical to sustaining the well-being and livelihoods of poor and marginalised people, and 36 indigenous communities who depend on natural resources (high confidence). Nature provides a broad 37 array of goods and services that are critical to good quality of life and human well-being. Healthy and 38 diverse ecosystems can play an important role in reducing vulnerability and building resilience to 39 disasters and extreme weather events (SCBD) Secretariat of the Convention on Biological Diversity 40 2009; The Royal Society Science Policy Centre 2014; Ninan and Inoue 2017). 41 Current negative trends in biodiversity and ecosystem services will undermine progress towards 42 achieving 80% (35 out of 44) of the assessed targets of SDGs related to poverty, hunger, health, water, 43 cities, climate, oceans and land (Díaz et al. 2019). However, Reyers and Selig (2020) note that the 44 assessment by (Díaz et al. 2019) could only assess the consequences of trends in biodiversity and Do Not Cite, Quote or Distribute 7-120 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 ecosystem services for 35 out of the 169 SDG targets due to data and knowledge gaps, and lack of 2 clarity about the relationship between biodiversity, ecosystem services and SDGs. 3 Progress in achieving the 20 Aichi Biodiversity targets which are critical for realising the SDGs has 4 been poor with most of the targets not being achieved or only partially realised (SCBD 2020). There 5 could be synergies and trade-offs between ecosystem services and human well-being. For instance, a 6 study notes that although policy interventions and incentives to enhance supply of provisioning services 7 (e.g., agricultural production) have led to higher GDP, it may have an adverse effect on the regulatory 8 services of ecosystems (Kirchner et al. 2015). However, we are aware of the inadequacies of traditional 9 GDP as an indicator of well-being. In this context the Dasgupta Biodiversity Review argues for using 10 the inclusive wealth approach to accurately measure social well-being by tracking the changes in 11 produced, human and natural capital (Dasgupta 2021). Targets for nature (biodiversity and ecosystem 12 services) should be refined so as to fit in with the metrics tracked by the SDGs (Ferrier et al. 2016; Rosa 13 et al. 2017). 14 7.6.5.3. Land-based mitigation and adaptation 15 Combined mitigation and adaptation approaches have been highlighted throughout Section 7.4 16 regarding specific measures. Land-based mitigation and adaptation to the risks posed by climate change 17 and extreme weather events can have several co-benefits as well as help promote development and 18 conservation goals. Land-based mitigation and adaptation will not only help reduce GHG emissions in 19 the AFOLU sector, but measures are required to closely link up with adaptation. In the central 2°C 20 scenario, improved management of land and more efficient forest practices, a reduction in deforestation 21 and an increase in afforestation, would account for 10% of the total mitigation effort over 2015–2050 22 (Keramidas et al. 2018). If managed and regulated appropriately, the Land sector could become carbon- 23 neutral as early as 2030–2035, being a key sector for emissions reductions beyond 2025 (Keramidas et 24 al. 2018). Nature-based solutions (NbS) with safeguards has immense potential for cost-effective 25 adaptation to climate change; but their impacts will vary by scale and contexts (high confidence). 26 Griscom et al. 2017 estimate this potential to provide 37% of cost-effective CO2 mitigation until 2030 27 needed to meet 2°C goals with likely co-benefits for biodiversity. However, due to the time lag for 28 technology deployment and natural carbon gain this mitigation potential of NbS by 2030 or 2050 can 29 be delayed or much lower than the estimated potential (Qin et al. 2021). 30 7.7. Knowledge gaps 31 Closing knowledge gaps and narrowing uncertainties are crucial to advance AFOLU mitigation. 32 Knowledge gaps exist across a range of areas, from emissions accounting and mitigation measure 33 development to integration of scientific and traditional knowledge and development and sustainable 34 implementation strategies. The following are identified as priorities: 35 • Uncertainty in contemporary emissions and sinks within AFOLU is still high. There is on-going 36 need to develop and refine emission factors, improve activity data and facilitate knowledge 37 exchange, concerning inventories and accounting. For example, insufficient knowledge on CO2 38 emissions relating to forest management and burning or draining of organic soils (wetlands and 39 peatlands), limits certainty on CO2 and CH4 fluxes. 40 • Improved monitoring of the land CO2 balance is urgently needed, including impacts of land 41 degradation and restoration efforts (e.g., in tropical and boreal regions), making use of 42 combined remote sensing, artificial intelligence, ground-based and modelling tools (Grassi et 43 al. 2021). Improved estimates would provide more reliable projections of nationally determined Do Not Cite, Quote or Distribute 7-121 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 contributions to emissions reduction and enhancement of sinks, and reconciliation of national 2 accounting and modelling results (Nabuurs et al. 2019). 3 • The future impacts of climate change on land systems are highly uncertain, for example, the 4 role of permafrost thaw, tipping points, increased disturbances and enhanced CO2 fertilization 5 (Friedlingstein et al. 2020). Further research into these mechanisms is critical to better 6 understand the permanence of mitigation measures in land sector. 7 • There is need to understand the role of forest management, carbon and nitrogen fertilization 8 and associated interactions in the current forest carbon sink that has emerged in the last 50 to 9 70 years. These aspects are likely to explain much of the difference between bookkeeping 10 models and other methodologies. 11 • Continued research into novel and emerging mitigation measures and associated cost efficiency 12 (e.g. CH4 inhibitors or vaccines for ruminants) is required. In addition to developing specific 13 measures, research is also needed into best practice regarding implementation and optimal 14 agricultural land and livestock management at regional and country level. Further research into 15 the feasible mitigation potential of sustainable intensification in terms of absolute GHG 16 emissions and appropriate policy mechanisms, is required to implement and advance this 17 strategy. 18 • Research into accounting systems and policy options that will enable agricultural soil and forest 19 carbon to be utilized as offsets (voluntary or regulatory) is needed to increase financing for 20 land-based CDR. Design of incentives that consider local institutions and novel frameworks for 21 cooperation between private finance and public governance can encourage investment. Equally, 22 research to adjust or remove regulations and subsidy schemes that may hamper land-based 23 mitigation efforts, is urgently required. 24 • Improving mitigation potential estimates, whether derived from sectoral studies or IAMs to 25 account for biophysical climate effects, and impacts of future climate change (e.g. mitigation 26 permanence), biodiversity loss and corresponding feedbacks is needed. IAM ‘usability’ can be 27 enhanced by integrating a wider set of measures and incorporating sustainability considerations. 28 • Research into the feasibility of improving and enhancing sustainable agricultural and forestry 29 value chains, provision of renewable products (building with wood) and the sustainability of 30 bioenergy is critically important. Modelled scenarios do not examine many poverty, 31 employment and development trade-offs, which are highly context specific and vary 32 enormously by region. Trade-off analysis and cost-benefit analysis can assist decision making 33 and policy. 34 • In-depth understanding of mitigation-SDG interactions is critical for identifying mitigation 35 options that maximize synergies and minimize trade-offs. Mitigation measures have important 36 synergies, trade-offs and co-benefits, impacting biodiversity and resource-use, human well- 37 being, ecosystem services, adaptation capacity and many SDGs. In addition to exploring 38 localised economic implementation costs, studies are needed to understand how measures will 39 impact and interact with wider environmental and social factors across localities and contexts. 40 41 Frequently Asked Questions (FAQs) 42 FAQ 7.1 Why is the Agriculture, Forestry and Other Land Uses (AFOLU) sector unique when 43 considering GHG mitigation? Do Not Cite, Quote or Distribute 7-122 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 There are three principal reasons that make the AFOLU sector unique in terms of mitigation; 2 1. In contrast to other sectors, AFOLU can facilitate mitigation in several different ways. 3 Specifically, AFOLU can (a) reduce emissions as a sector in its own right, (b) remove 4 meaningful quantities of carbon from the atmosphere and relatively cheaply, and (c) provide 5 raw materials to enable mitigation within other sectors, such as energy, industry or the built 6 environment. 7 2. The emissions profile of AFOLU differs from other sectors, with a greater proportion of non- 8 CO2 gases (N2O and CH4). The impacts of mitigation efforts within AFOLU can vary according 9 to which gases are targeted, as a result of the differing atmospheric lifetime of the gases and 10 differing global temperature responses to the accumulation of the specific gases in the 11 atmosphere. 12 3. In addition to tackling climate change, AFOLU mitigation measures have capacity, where 13 appropriately implemented, to help address some critical, wider challenges, as well as 14 contributing to climate change adaptation. AFOLU is inextricably linked with some of the most 15 serious challenges that are suggested to have ever faced humanity, such as large-scale 16 biodiversity loss, environmental degradation and the associated consequences. As AFOLU 17 concerns land management and utilises a considerable portion of the Earth’s terrestrial area, the 18 sector greatly influences soil, water and air quality, biological and social diversity, the provision 19 of natural habitats, and ecosystem functioning, consequently impacting many SDGs. 20 FAQ 7.2 What AFOLU measures have the greatest economic mitigation potential? 21 Economic mitigation potential refers to the mitigation estimated to be possible at an annual cost of up 22 to USD100 tCO2-1 mitigated. This cost is deemed the price at which society is willing to pay for 23 mitigation and is used as a proxy to estimate the proportion of technical mitigation potential that could 24 realistically be implemented. Between 2020 and 2050, measures concerning forests and other ecosystem 25 are estimated to have an average annual mitigation potential of 7.3 (3.9 - 13.1) GtCO2-eq yr-1 at 26 USD100 tCO2-1. At the same cost, agricultural measures are estimated to have a potential of 4.1 (1.7- 27 6.7) GtCO2-eq yr-1. Emerging technologies, such as CH4 vaccines and inhibitors, could sustainably 28 increase agricultural mitigation potential in future. The diverted production effects of changes in 29 demand (reduced food losses, diet changes and improved and enhanced wood products use), is 30 estimated to have an economic potential of 2.2 (1.1–3.6) GtCO2-eq yr-1. However, cost forms only one 31 constraint to mitigation, with realization of economic potential dependent on multiple context-specific 32 environmental and socio-cultural factors. 33 FAQ 7.3 What are potential impacts of large-scale establishment of dedicated bioenergy 34 plantations and crops and why is it so controversial? 35 The potential of bioenergy with carbon capture and storage (BECCS) remains a focus of debate with 36 several studies evaluating the level at which BECCS could be sustainably implemented, published since 37 AR5. BECCS involves sequestering carbon through plant growth (i.e. in trees or crops) and capturing 38 the carbon generated when this biomass is processed for power or fuel. The captured carbon then 39 requires long-term storage in for example, geological, terrestrial or ocean reservoirs, or in products. 40 While appearing to create a net removal of carbon from the atmosphere, BECCS requires land, water 41 and energy which can create adverse side-effects at scale. Controversy has arisen because some of the 42 models calculating the energy mix required to keep the temperature to 1.5°C have included BECCS at 43 very large scales as a means of both providing energy and removing carbon from the atmosphere to 44 offset emissions from industry, power, transport or heat. For example, studies have calculated that for 45 BECCS to achieve 11.5 GtCO2-eq per year of carbon removal in 2100, as envisaged in one scenario, 46 380-700 Mha or 25-46% of all the world’s arable and cropland would be needed. In such a situation, Do Not Cite, Quote or Distribute 7-123 Total pages: 185 Final Government Distribution Chapter 7 IPCC AR6 WGIII 1 competition for agricultural land seriously threatens food production and food security, while also 2 impacting biodiversity, water and soil quality, and landscape aesthetic value. More recently however, 3 the scenarios for BECCS have become much more realistic, though concerns regarding impacts on food 4 security and the environment remain, while the reliability of models is uncertain due to methodological 5 flaws. Improvements to models are required to better capture wider environmental and social impacts 6 of BECCS in order to ascertain its sustainable contribution in emissions pathways. Additionally, the 7 opportunity for other options that could negate very large-scale deployment of BECCS, such as other 8 carbon dioxide removal measures or more stringent emission reductions in other sectors, could be 9 explored within models. Do Not Cite, Quote or Distribute 7-124 Total pages: 185