Agriculture is the largest driver of deforestation globally, and this conversion of land from forests to agriculture, results in emissions which are contributing to climate change. This thesis focuses on exploring agriculture-driven deforestation at the country level, from the perspective of quantifying emissions, estimating the potential for mitigation, including identifying potential barriers to success, and highlighting enabling conditions for mitigation of these emissions. Efforts to reduce deforestation are being undertaken, for example through the mechanism REDD+; reducing emissions from deforestation and forest degradation and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks in developing countries. At the same time, efforts are underway to try to reduce hunger by increasing food security (for example through the sustainable development goals (SDGs)). Competition for land can result when both these goals are pursued at the same time, because forested land is protected for carbon storage, while agricultural land is expanded (often into forests) to provide sufficient land for growing food. There are several ways in which both goals, forest protection and food security might be achieved together, and we focus on assessing the potential of two approaches which can potentially spare forested land. These approaches are: increasing production on existing agricultural land, and expanding agriculture onto non-forested available land. Emerging phenomena such as Large Scale Land Acquisitions (LSLA, otherwise known as land grabs) add to the complexity of the challenge, and we discuss the potential threat which LSLA has on forested land, and how to avoid LSLA for agriculture in forested land. A transformational change of the land sector is proposed to ensure that both goals can be met. Several ingredients are required to achieve a transformational change, and linking REDD+ to Climate Smart Agriculture (CSA) approaches is discussed. CSA interventions are those which are able to reduce emissions or store carbon while increasing the adaptive capacity of agriculture to climate change and increasing food production.
Chapter 2 provides new estimates of emissions from agriculture-driven deforestation in 91 countries using a data-driven approach. Latin America was found to have the highest emissions, and these emissions peaked between 2000 and 2005 and then declined. Emissions in Africa has been rising since 1990, with the countries in the Congo Basin being particular contributors to this rise in emissions. Uncertainties of these country emission estimates are ±62.4% (average for 1990-2015), and emissions from Asia are the most uncertain. The uncertainty of the input datasets was used to estimate the uncertainty of the emissions estimate, and the area of deforestation, and fraction which agriculture is driving deforestation were found to be the largest contributors to uncertainty of the emissions estimates. Increasing the certainty of these two data types should be a priority, and will lead to an increased certainty for the emissions estimates.
Chapter 3 compares direct and indirect emissions from agriculture at the national level, where direct are emissions from existing agricultural land, and indirect emissions are those from agriculture-driven deforestation. A decision tree was produced which can be used to guide decision making by identifying priority countries for mitigation initiatives. The decision tree uses several indicators related to the potential for mitigation, enabling environment, and associated risks to livelihoods to identify countries which have the most potential for the mitigation of either direct or indirect agricultural emissions. Six priority countries are highlighted as having a good mitigation potential for agriculture-driven deforestation while having a good enabling environment (in this case engagement in REDD+) and which also have low risks to livelihoods from the implementation of interventions in the agriculture sector. They are: Panama, Paraguay, Ecuador, Mexico, Malaysia and Peru.
Chapter 4 focusses on LSLA, and their potential impacts on forests. A country level analysis was carried out, and the characteristics which are typically found in countries which have LSLA were described. Countries which have these characteristics and which do not yet have LSLA are for example considered to be at risk from LSLA. Countries which have LSLA or are at risk from LSLA were assessed for the risk of LSLA-driven deforestation. Other key targets for interventions to reduce deforestation are highlighted, such as those countries with large numbers of LSLA and which already have a lot of agriculture-driven deforestation. The potential conflicts between LSLA and REDD+ are discussed, and investor-side policies such as zero deforestation pledges from commodity producers, green procurement policies, and initiatives such as the Roundtable For Sustainable Palm Oil are highlighted as potential solutions to these conflicts. Lessons learned from implementing REDD+, which has a number of shared characteristics with LSLA, can be applied in order to reduce the negative impacts of LSLA.
Chapter 5 discusses the potential for forest-land sparing interventions to be implemented in the agriculture sector. A transformative change which incorporates multiple interventions and brings together the forest and agriculture sectors is proposed. Climate Smart Agriculture approaches should be considered, but only when they do not lead to expansion of agriculture into forests. The need for supporting policies to avoid this occurring is discussed. Policy coherence is a barrier to this change as policies favouring both conversion to agriculture (including those which enable LSLA), and forest protection can occur in the same place. The use of the landscape approach as a platform to address this challenge is discussed. Landscape-level emissions accounting, which takes into consideration both direct and indirect emissions from agriculture, can be used to evaluate the impact of mitigation interventions across sectors. The need for transparency in the land sector, in relation to emissions reporting in particular is introduced, and is a key requirement for access to carbon finance which can potentially support forest land-sparing interventions.
Chapter 6 concludes the thesis, and discusses the wider implications for this work. The link between the findings in this thesis and the SDGs is explored. The SDGs may lead to future competition for land due to goals which focus on reducing hunger, protecting forests and increasing the proportion of renewable energy unless action is taken. Future data needs are discussed, as although we provide (in chapter 2) new data on agriculture-driven deforestation, they are still uncertain and data on potential future trends in agriculture-driven deforestation are not available. The need for consideration of emissions related to the impact of agriculture on forest degradation and on carbon losses in soils is another data gap, and relates to recent efforts to restore degraded land – which could be one of the most promising mitigation efforts which can also support the production of more food for growing global populations. The urgent need to address climate change highlights the opportunities in the land sector, not only to mitigate emissions, but also to promote food security.