Navin Ramankutty

Fruit and vegetable supply is essential for a healthy diet, yet the exposure of supply chains to climate-related conditions is often overlooked, particularly when sourcing crosses national borders. We develop provincial-scale food flows for 18 fruits and 16 vegetables using average 2018–2022 production and trade data, including estimated interprovincial flows and derived products. We then link these recent sourcing patterns to changes in maximum five-day rainfall, consecutive dry days, days below 0 °C, and days above 35 °C from +1 °C to +3 °C global mean warming relative to 1850–1899. Across Canada’s supply, 43 % of fruit and 19 % of vegetables are sourced from regions projected to experience at least 15 additional days above 35 °C, 19 % and 8 % from regions with at least five additional consecutive dry days, 31 % and 9 % from regions with at least a 5 % decrease in maximum five-day rainfall, and 7 % and 2 % from regions with at least 10 fewer days below 0 °C, respectively. Exposure is concentrated in specific supply chains, including lettuce sourced from California and Arizona and avocados from Mexico, and varies across provinces because of differences in sourcing portfolios. This exposure assessment holds recent sourcing patterns constant and does not estimate crop-yield losses, future trade adjustments, prices, or supply disruptions. It provides a basis for identifying the most exposed commodity, province, and source regions that may warrant more detailed impact and adaptation assessment.

Farm size is a key farming system characteristic at the heart of debates on productivity, land reform, smallholder production and livelihoods, and sustainability outcomes. Several global or regional datasets on farm size exist that form the foundation of research applications (e.g. estimating the contribution of smallholders to global food production) and practice (e.g. informing rural development strategies). However, existing datasets have crucial limitations as they are built using national-level statistics—which mask important subnational variability, have limited geographic coverage, and/or are based on modeled rather than empirical data. We address these limitations by developing the first global dataset of mean farm size that is both entirely empirical and at predominantly subnational scale. Our dataset is based on census or survey data for 101 countries at the smallest subnational administrative level for which data was publicly available, and at the national level for 99 additional countries, for the target year 2000. Comparing our product to existing (quasi) global farm size datasets highlights differences in spatial coverage and scale and in subnational variability, which result in important differences in farm size patterns across the globe. Our dataset thus highlights the importance of spatially-explicit high-resolution datasets at global scale, which enable more accurate and nuanced analyses of farm size as an important driver of land and food systems for policy and research.

As climate change intensifies, extreme weather is becoming a major threat to global food security. Yet we still lack a good understanding of how these extremes will be distributed across the world’s agricultural lands—particularly across small versus large croplands and pasturelands. In this study, we assess their exposure to extreme weather in a warming world. In a world that is 2 °C warmer than today, 25% (11 million km ² ) of present-day agricultural lands will face over two months of extreme heat, up from 16% today (7 million km ² ), and another ∼2% (5 million km ² ) will be exposed to a combination of two or more extremes, up from 10% (4 million km ² ). The total area exposed to prolonged dry conditions and extreme precipitation will remain unchanged (less than 1% or 5 million km ² , with increases in some regions balanced by decreases elsewhere), while ∼2% less area (2 million km ² ), down from 7% (3 million km ² ), will experience a month of frost. Future exposure to extreme weather varies by land use type. Pasturelands will experience prolonged exposure to heat stress, whereas croplands will be exposed to higher excessive rains and heat stress combined. Spatial correlations between farm size and geography indicate potential differences in exposure. Exposure to extreme precipitation and heat stress is highest in small (1–2 ha) and medium (2–4 ha) cropland and pastureland, respectively. These findings offer a preliminary global assessment of how exposure to extreme weather varies by farm size and land use, underscoring the need for tailored adaptation strategies to safeguard food security in a warming world.

The global agrifood system is central to many challenges humanity faces today. Despite significant growth in total production, it fails to ensure food security for all, drives biodiversity decline, and majorly contributes to climate change. Research on agrifood-system burdens often focusses on the national level and isolated burdens, ignoring their systemic complexity. We address this knowledge gap by combining global subnational datasets proxying four key dimensions of agrifood-system burdens: environmental footprint, climate change, income poverty, and malnutrition. We map global hotspots of co-occurring agrifood-system burdens for 2017. We overlay these with data on ambient population counts, agricultural areas, farm size distributions, and lands inhabited by Indigenous peoples to identify spatial correspondence between people in vulnerable contexts and food production regions facing these burdens. We further assess countries’ relative burden against their inequality and governance indicators. Burden hotspots occupy many regions worldwide, especially in low-income, (sub)tropical regions. Single burdens occupy regions harbouring about 5 billion people (∼66% of the global population) and 1.8 billion ha of agricultural lands (∼40%), while multiple burdens occupy regions with about 1.9 billion people and 470 million ha agricultural lands. Environmental footprint is the strongest contributor to these burden profiles. Regions with traditionally marginalised communities (i.e. small-scale farmers and Indigenous peoples) disproportionally face multiple burdens. Agrifood-system burdens are more prevalent in countries with higher economic inequality and poorer governance. Burden profiles vary substantially within and between countries, necessitating regionalised and context-specific policies for effective, bundled, and targeted solutions. Addressing agrifood-system burdens can also synergise with tackling other current global challenges, like biodiversity loss and environmental justice.

Our current global food system is failing to feed the world while simultaneously emitting between 26%–34% of greenhouse gases (GHGs) that alone could preclude us from meeting the Paris climate agreement goal of limiting warming to 1.5 °C or 2 °C above pre-industrial levels. But emissions from food consumption are not uniform amongst the world’s inhabitants and thus, we estimate those differentiated responsibilities. As expected, the emissions from those barely eating enough to survive are among the lowest. But the interplay of production practices, trade, dietary preferences, the nutrition transition, and within-country inequality in access to food shape the variations in global food system emissions. By combining the most recent estimates of access to food by income decile with trade-adjusted GHG emissions data for food, we present estimates of the inequality in emissions from food consumption on a global scale. We find that the top 15% of emitters account for 30% of the total emissions, equalling the contribution of the bottom 50%. Furthermore, we assess the reductions required from the top emitters to achieve two goals. First, to yield space for increased emissions to those not meeting basic dietary requirements to thrive; we find that only an additional 0.4% of the population in 2012 would need to cap their emissions so that 8.8% of the population can increase their emissions and be able to thrive. Second, to reduce agricultural GHG emissions to meet the 2 °C goal, we find that, between 40%–45% of the world’s population in 2012 consumed diets above a target per capita cap, while 89%–91% consumed diets above a target per capita cap calculated using a future 2050 population. This means that efforts to reduce emissions from the food system will be part of almost everyone’s life up to 2050 but for at least 40% that responsibility starts now.

As climate change intensifies, extreme weather is becoming a major threat to global food security. Yet we still lack a good understanding of how these extremes will be distributed across the world’s agricultural lands—particularly across small versus large crop and pasturelands. In this study, we assess the exposure of global agricultural lands to extreme weather in a warming world, focusing on small and large crop and pasturelands. In a world that is 2°C warmer than today, 25% (11 million hectares, Mha) of present-day agricultural lands will face over two months of extreme heat, up from 16% today (7 Mha), and another ~2% (5 Mha) will be exposed to a combination of two or more extremes, up from 10% (4 Mha). The total area exposed to prolonged dry conditions and extreme precipitation will remain unchanged (less than 1% or 5 Mha, with increases in some regions balanced by decreases elsewhere), while ~2% less area (2 Mha), down from 7% (3 Mha), will experience a month of frost. Future exposure to extreme weather varies by land use type. Pasture lands will experience prolonged exposure to heat stress, whereas croplands will be exposed to higher excessive rains and heat stress combined. Spatial correlations between farm size and geography indicate potential differences in exposure. Exposure to extreme precipitation and heat stress is highest in small (1-2 ha) and medium (2-4 ha) crop and pastureland, respectively. These findings offer a preliminary global assessment of how exposure to extreme weather varies by farm size and land use, underscoring the need for tailored adaptation strategies to safeguard food security in a warming world.

While there are many global geospatial datasets representing the extent of agriculture, they predominantly represent croplands. Only a couple of global data products represent the full global agricultural footprint, including both cropland and pastures. Our own research team's most recent complete publicly available agricultural land cover dataset, including both croplands, and pastures, represents circa 2000. These data, distributed on a graticule of 5 arcmin (∼ 10 km² at the Equator), have been integrated into a considerable number and diversity of research studies, modelling, data science, and media applications. Further, users of these data have been interested in them for studying a variety of issues such as land use, food security, climate change, and biodiversity loss. Here we present an updated dataset on the global distribution of agricultural lands (cropland and pasture) circa 2015 (15 years on since the initial study). Past studies that have constructed such datasets have been one-off exercises that have been infrequently repeated due to the amount of effort required. Therefore, in this work, we developed a transparent and reproducible approach to update our data product while also enabling easier reproduction of future datasets. We distribute our 2015 product at the same resolution and formats as the prior product, and accompany it with a full set of replicable code and data for reconstruction. In this article we explain how the data was constructed, with links to the permanent DOIs where the data can be readily downloaded by the user community (Mehrabi et al., 2024; https://doi.org/10.5281/zenodo.11540553).

In West Africa, increasing rural–urban flows of food, driven by growing urban populations, require functional, efficient links between cities and production areas. However, underlying mechanisms of urban food sourcing in West Africa are poorly understood. This study deepens understanding of spatial interactions between cities and production areas by examining the effects of settlement size, geographical distance, and agricultural suitability on food inflows to four West African cities. The analysis was informed by theoretical spatial models and data on food flows, road network, agricultural suitability, and settlements. Results showed that food travelled further from larger supplying settlements, and towards the two larger destination cities. This supports the idea of a hierarchical system, where food provisioning area and upstream supply chain length increase with settlement size. Overall, towns with fewer than 100,000 inhabitants, often representing aggregation centres, were among the major suppliers to the cities. Complementary agricultural suitability between origin and destination shaped food flow direction and length, but poor road access and international borders impeded trade. Spatial models did not fully explain food flows: they were also influenced by historical factors shaping certain settlements’ importance as sources. Study cities were supplied by a diversity of more and less concentrated food sources, representing production sites or aggregating markets, which should theoretically support food supply resilience. Improvements to storage and road infrastructure, and removing trade barriers, could improve food supply to cities, and producer and trader livelihoods. Emerging research on urban food systems governance could support understanding of how best to do this.

While there are many global geospatial datasets representing the extent of agriculture, they predominantly represent croplands. Only a couple of global data products represent the full global agricultural footprint, including pastures. Our own research team’s most recent complete publicly available agricultural land cover dataset, including both croplands and pastures, represent circa 2000. These data, distributed on a graticule of 5 arcminutes (~10 km² at the equator), have been integrated into a considerable number and diversity of research studies, modeling, data science and media applications. Further, users of these data have been interested in them for studying a variety of issues such as land use, food security, climate change and biodiversity loss. Here we present an updated dataset on the global distribution of agricultural lands (cropland and pasture) circa 2015 (15 years on since the initial study). Past studies that have constructed such datasets have been one-off exercises that have been infrequently repeated due to the amount of effort required. Therefore, in this work, we developed a transparent and reproducible approach to update our data product while also enabling easier reproduction of future datasets. We distribute our 2015 product at the same resolution and formats as the prior product, and accompany it with a full set of replicable code and data for reconstruction. In this article we explain how the data was constructed, with links to the permanent DOIs where the data can be readily downloaded by the user community (Mehrabi et al. 2024; DOI: 10.5281/zenodo.11540554).