- ➔ Land is the core resource of agrifood systems. It underpins food security, biodiversity, livelihoods, ecosystem services, cultural heritage, and mitigation of and adaptation to climate change.
- ➔ Land is a finite resource, increasingly stressed in both quantity and quality. Competing demands – ranging from feed, fibre and the production of biofuels to the expansion of urban areas – require that agriculture be efficient and productive, but degradation further strains land’s potential.
- ➔ Human-induced land degradation is not a new phenomenon; it dates to the beginning of agriculture. However, its accelerated pace and intensified impacts make addressing land degradation and related abandonment more urgent than ever.
- ➔ Action against land degradation can be costly. While it brings private benefits to land users, most benefits are enjoyed by the broader society. This makes such action a public good, requiring public policies and investment.
- ➔ Farmers’ ability and their incentive to adapt to and restore land depend on farm size, land conditions and socioeconomic context. Tailored solutions that align incentives with public benefits are essential for progress towards sustainable production.
Since the invention of agriculture 12 000 years ago, land has played a central role in sustaining civilizations. As the fundamental resource of agrifood systems, it interacts with natural systems in complex ways, influencing soil quality, water resources and biodiversity, while securing global food supplies and supporting the achievement of the Sustainable Development Goals (SDGs). Biophysically, it consists of a range of components including soil, water, flora and fauna, and provides numerous ecosystem services including nutrient cycling, carbon sequestration and water purification, all of which are subject to climate and weather conditions. Socioeconomically, land supports many sectors such as agriculture, forestry, livestock, infrastructure development, mining and tourism. Land is also deeply woven into the cultures of humanity, including those of Indigenous Peoples, whose unique agrifood systems are a profound expression of ancestral lands and territories, waters, non-human relatives, the spiritual realm, and their collective identity and self-determination.1 Land, therefore, functions as the basis for human livelihoods and well-being.2
At its core, land is an essential resource for agricultural production, feeding billions of people worldwide and sustaining employment for millions of agrifood workers. Healthy soils, with their ability to retain water and nutrients, underpin the cultivation of crops, while pastures support livestock; together they supply diverse food products essential to diets and economies. Currently, more than 95 percent of the global food supply is grown or raised on land.3 In 2022, the agricultural sector also employed 892 million people worldwide (accounting for 26.2 percent of total employment), with an additional 13 percent of the global workforce engaged in non-agricultural agrifood systems jobs, providing livelihoods, generating incomes and supporting food security.4
Land also plays a critical role in maintaining the ecological balance and providing indispensable ecosystem services for agrifood systems. Forest lands and wetlands regulate water cycles, prevent floods and recharge aquifers, ensuring that agriculture has access to a reliable supply of water. Healthy soils store vast amounts of organic carbon, mitigating the impacts of climate change, while also serving as reservoirs of biodiversity, housing countless organisms that support nutrient cycling and pest control. These services are essential for both food production and environmental sustainability.2, 5
Yet, land is a finite resource. Various demands – driven by policy decisions, market trends and consumer preferences – add significant pressure to already scarce land resources. These include demand for biofuels, which requires large areas of land for crops, often competing with food production. Emissions trading schemes also influence land use, prioritizing carbon sequestration projects like reforestation over other uses. Additionally, the growing need for feed crops further strains land resources. Urbanization exacerbates these pressures by converting agricultural and natural lands into urban areas, reducing the availability of land for food production and other essential uses. At the same time, urban lifestyles and rising incomes are reshaping consumption patterns, with demand growing for more diverse and resource-intensive diets, including higher consumption of meat, dairy and processed foods.6, 7 Balancing these competing demands is essential to ensure land remains able to support sustainable food production and maintain ecosystem services.
The expansion of agriculture and the accompanying growth of the human population have long exerted pressure on land systems, triggering changes over time that result in land degradation.8–12 These cumulative impacts have compromised not only the productive capacity of land but also its ecological integrity and the many services it provides. As a long-term trend driven by unsustainable agricultural practices, land degradation exerts significant stress on agricultural production systems, undermining their stability, increasing their vulnerability and reducing their resilience,13, 14 ultimately jeopardizing food security and livelihoods.
Today, nearly all inhabited parts of the world are subject to some form of human-induced land degradation,15, 16 with producers largely bearing the immediate burden of impacts on croplands. For example, they suffer revenue losses as a result of low yields or of costly compensation measures; the latter include soil amendments with inorganic and organic fertilizers and agricultural lime, some of which may not be accessible to all. However, the impacts of land degradation extend beyond producers, as society at large bears the externalized costs through climate change, loss of biodiversity and ecosystem services, and diminished future agricultural potential, making land degradation a problem that requires both local and global solutions.17 Nonetheless, land degradation is not an inevitable consequence of agricultural production. When managed sustainably, agricultural systems can maintain – and even enhance – land health, supporting productivity while preserving ecosystem functions.
Trends and challenges in agricultural land use
Understanding contemporary agriculture requires an examination of key trends and challenges in how land is used globally. These range from large-scale shifts in land cover type to the structure of farm holdings. The way humanity organizes, manages and utilizes land – land use18 – has undergone significant transformations, accelerated in particular by the invention of nitrogen fertilizer in the early twentieth century and the introduction of new agricultural technologies during the green revolution.
Shifting patterns in agricultural land use
Globally, agricultural land, encompassing both croplands and permanent meadows and pastures, spans nearly 4.8 billion hectares (ha), representing more than one-third of the world’s land area.19 Figure 1 presents the world agricultural land area by category as of 2023 and shows that only 12 percent of the total land area (i.e. arable land and permanent crops) sustains global crop production. Permanent meadows and pastures make up one-quarter of the total land, while forests cover one-third. Between 2001 and 2023, global agricultural land area experienced a net decrease of 75 million hectares (Mha) (−2 percent), with cropland area increasing by 78 Mha and permanent meadows and pastures decreasing by 151 Mha. However, the changes in land use were not geographically uniform.
Figure 1 World Agricultural Land Area by Main Category, 2023
Figure 2 presents the net changes in cropland – encompassing arable land and permanent crops – against changes in forest land. In general, in regions with cropland expansion, deforestation was also observed. Notably, in sub-Saharan Africa, cropland expanded by 69 Mha while 72 Mha of forests were lost; similarly, in Latin America and the Caribbean, an expansion of 25 Mha of cropland coincided with deforestation spanning 85 Mha. Globally, forest area declined by 109 Mha. Conversely, regions that saw a decrease in cropland, such as Eastern and South-eastern Asia and Northern America, recorded afforestation.
Figure 2 Land-Use Change In Cropland And Forest Land By Region And Subregion, 2001–2023
SOURCE: Figure 6 in FAO. 2025. Land statistics 2001–2023. FAOSTAT Analytical Briefs, No. 107. Rome. https://doi.org/10.4060/cd5765en
Remote sensing evidence further reinforces this link, showing that agricultural expansion – particularly cropland development – is the primary driver of global deforestation. Nearly 90 percent of global deforestation is driven by agriculture. Of all agricultural activities, cropland expansion is the single largest contributor, accounting for almost half of the total deforested area, followed by livestock grazing. While the former was the main driver in Asia and Africa, the latter was the largest contributor to deforestation in the Americas and Oceania.20 These findings underscore the central role of agriculture in shaping land-use change and the urgent need to balance food production with forest conservation.
The above-discussed trends in land use highlight broader changes in land cover – the physical cover of the Earth’s surface including natural or planted vegetation and human construction.18 Between 1992 and 2019, natural and semi-natural types of land cover lost over 20 percent more area than they gained, mostly due to conversion to cropland, as well as desertification and urban expansion.21 In the twentieth century, approximately 400 Mha of land were abandoned globally, including not only areas affected by land degradation but also those left idle due to socioeconomic shifts such as rural depopulation, changing labour markets, and evolving land-use priorities.22 Globally, the majority of remaining natural land cover is located in close proximity to areas of intensive land use, increasing the risk of habitat fragmentation in areas rich in ecosystem services that underpin agricultural productivity.23
Farm structures and their implications
Beyond these large-scale changes in land use and cover, the structure of agricultural production itself varies significantly worldwide, particularly concerning the size of landholdings. Farming operations encompass a wide spectrum of land sizes, frequently referred to as small, medium and large. The distinction is important because farm size influences not only how land is managed, but also the adoption of agricultural technologies and the ecological outcomes of farming practices. Smaller farms often encounter constraints in accessing mechanization and inputs,24, 25 and present greater diversity;26 larger farms, on the other hand, may contribute to landscape homogenization with implications for biodiversity and ecosystem services.27
When attempting to define farm size categorization, it is best to consider landholding size relative to the full distribution of holdings in a given country; what is considered small in one country may be perceived as large in another. In this regard, SDG Indicator 2.3.2 defines small-scale producers using nationally relevant distributions. Farmers located in the bottom 40 percent of the national distribution of physical land size (and/or livestock herd size) and in the bottom 40 percent of national total on-farm revenue distribution are considered small-scale producers.28
Adopting such a relative national definition of small-scale or large-scale producers is relevant for national policymaking. Depending on the distribution of land, livestock heads and revenues, thresholds that identify small-scale food producers might be 2 ha in one country and 50 ha in another, or annual revenues of USD 1 500 in one country and USD 250 000 in another.28, 29
While this country-relative approach captures national distributions, a globally consistent threshold is also useful for identifying common resource constraints and scale-specific technological interventions. Farms under 2 ha face similar challenges – such as limited mechanization, restricted input access and weaker market participation – regardless of national context. This makes the threshold highly relevant for global agricultural development strategies. At the other end of the spectrum, the definition of a large landholding also varies significantly by region, with 50 ha or 100 ha usually used as the lower threshold.30–32 The thresholds cited in Chapter 3 of this report consider holdings between 2 ha and 50 ha as medium-sized, while those exceeding 50 ha and 1 000 ha are considered large and very large, respectively. Almost 500 million farms in the world cover less than 2 ha, falling within the category of smallholding, while very large farms control vast areas of farmland.33
Although farm size is an important element in the link between land degradation and agricultural production, it cannot capture the full diversity of farming systems. Tenure security, market access, gender dynamics and agroecological conditions all play pivotal roles in shaping agricultural outcomes. To understand the development, characteristics and heterogeneity of farming systems, with the goal of eventually guiding policies, research and policymaking rely on many different farm classification systems.34 Box 1 presents a brief overview of different types of farm classification and their uses.
Box 1Beyond farm size: matching policy with farm classification
Categorizing farms according to a single dimension, such as land size or revenue, risks occluding broader structural and environmental determinants from view. Above all, the choice of farm classification should be guided by the specific research questions and policy objectives at hand. Aligning classification design with the stage of the policy process under discussion strengthens the link between research and intervention. Failure to do so may result in typologies that lack transferability, validity or acceptance among stakeholders, particularly if they are seen as unfair or overly reductive.34
Tenure-based classifications may be relevant when considering outcomes linked to land degradation, because tenure security impacts land management and investment decisions. Farmers with secure tenure are more likely to invest in soil conservation practices and long-term land improvements, as they can be more confident that they will reap the rewards of their investments.35
Market-oriented classifications, which distinguish between farms producing primarily for subsistence needs and those producing to meet commercial demand, are also useful. Farms integrated into commercial value chains typically have greater access to inputs, credit and extension services, and may therefore be better positioned to adopt sustainable practices that reduce land degradation.36
Gender-based classifications can illuminate important inequalities in access to land, resources and decision-making power. For instance, female-headed households often face greater constraints in accessing land and credit, but are also found to employ distinct and often more conservation-oriented land management strategies.37
Holistic classifications, integrating a range of biophysical and socioeconomic factors – such as land quality, degree of mechanization, market integration and value of productive assets – may enable researchers to acquire deeper insights into the determinants of land outcomes.38 These may be data-driven, combining cluster analysis, machine learning or other statistical methods to group farms or regions based on multiple variables simultaneously.39 Such multidimensional frameworks must strike a balance between comprehensiveness and parsimony: while richer classifications can capture the full complexity of farming systems, they may become unwieldy for broad application and policy communication.
Farm size-based classifications, such as the one adapted in this report, are generally used as a practical approach to capture multiple overlapping dimensions of vulnerability and capacity, which are relevant to land degradation. For example, insecure tenure, market constraints, subsistence farming strategies and female management are all associated with smaller farm sizes.40–42
The farm size-based classification used in this report highlights vast disparities in land use and farm structures across regions. This underscores the need for context-specific agricultural policies that can address the varying challenges to technology adoption, resource access and environmental management to ensure sustainable agricultural productivity growth.
