Farming has already caused the loss of an estimated 300 billion tonnes of stored carbon and put around 14% of species at risk of extinction through agricultural land use. A new study from the University of Oxford, Leiden University, and China Agricultural University, and published in Nature Food , provides one of the most detailed analyses yet of where these losses occur, which foods drive them, and how the two types of damage overlap. The 300 billion tonnes of carbon loss is equivalent to around 140 years of agricultural emissions at current rates.
The study finds that carbon loss and biodiversity loss are strongly correlated: two-thirds of both total losses are concentrated on just one-third of agricultural land, mainly spanning parts of South America, sub-Saharan Africa and South-East Asia. Half of all carbon loss occurs on just 12% of farmland. Because the losses overlap so heavily, targeting these areas could deliver gains for climate and nature simultaneously.
Animal-sourced foods drive 59% of food-related carbon loss and 71% of biodiversity loss, while providing less than 15% of global calories (excluding calories from fish). The disparity is stark: calorie for calorie, animal-sourced food causes eight times more carbon loss and 14 times more biodiversity loss than plant-sourced food. Beef and dairy alone account for 29% of the total carbon loss and 41% of the total biodiversity loss from farmland, driven by the sheer amount of land their production requires.
Around 18% of farmland’s carbon and biodiversity impacts are linked to internationally traded food, with animal products making up the majority of these trade flows. Brazil is the largest net exporter of both carbon and biodiversity loss; China is the largest net importer. The largest trade flows of environmental damage include exports of beef and pigmeat from Brazil to China.
The study also models what targeted land restoration could achieve. Restoring the 30% of farmland with the highest both losses could sequester around 170 billion tonnes of carbon — equivalent to roughly 15 years of global fossil fuel emissions — and prevent around 60% of the extinction risk currently caused by farmland.
Professor Paul Behrens, Oxford Martin School, University of Oxford, said: ‘Climate and nature have been treated as separate problems for too long, with separate targets and separate negotiations. Our findings show they’re often the same problem on the same piece of land, with two-thirds of the carbon damage and two-thirds of the biodiversity damage from farming happening on just a third of agricultural land. There is a huge opportunity to target the specific places and foods. The same actions can deliver for the climate and for nature at once.’
Researcher Baoxiao Liu, Institute of Environmental Sciences (CML), Leiden University, said; ’Food production on farmland has ongoing environmental costs that are often overlooked. We link long-term damage to both climate (carbon loss) and nature (biodiversity loss) to foods we currently consume. If farmland were no longer used for food, it could potentially be used for nature, helping to restore carbon stocks and provide habitat for more species. Our daily eating choices therefore have a real impact.’
-ENDS-
Notes to Editors:
Publication details
Journal: Nature Food
Title: High-resolution carbon and biodiversity mapping shows correlated losses across space and agricultural products
Publication Date: 10:00 am BST 6 August 2026
DOI/Link: 10.1038/s43016-026-01387-0 https://www.nature.com/articles/s43016-026-01387-0
Authors and affiliations:
Baoxiao Liu (Institute of Environmental Sciences (CML), Leiden University)
Paul Behrens (Institute of Environmental Sciences (CML), Leiden University; Oxford Martin School, University of Oxford)
Zhongxiao Sun (College of Land Science and Technology, China Agricultural University)
Laura Scherer (Institute of Environmental Sciences (CML), Leiden University)
Media contact:
Sara Davis
Email: sara.davis@oxfordmartin.ox.ac.uk
Phone: 01865 287365
About Professor Paul Behrens
Professor Paul Behrens is a British Academy Global Professor at the Oxford Martin School, University of Oxford. His research focuses on food systems and environmental change, and has been featured in outlets including the BBC, The Guardian and Nature journals.
Background information
FAQ
1. What is new about this study?
Previous global studies have tended to do one of two things well: map environmental impacts at high spatial resolution, or trace impacts through international supply chains to specific food products. But they rarely do both at once, and rarely for carbon and biodiversity together. This study combines all of these. It assesses carbon and biodiversity loss at the grid-cell level (around 5.6 km at the equator) across 123 agricultural commodities, and links those impacts through global trade flows from the farms where production happens to the consumers who ultimately drive demand. That combination is what allows the researchers to identify both the foods and the specific places where action could deliver the largest joint benefits for climate and nature.
2. Why are only a few foods responsible for such a large share of the impact?
The main reason is land use. Some foods - especially beef and dairy - require far more land to produce than others. This means more natural ecosystems are converted into farmland, leading to both carbon loss and declines in biodiversity.
3. Where are the biggest impacts happening?
The study finds that impacts are concentrated in specific regions, including parts of South America, Africa and Asia. In general, carbon losses are higher in parts of the northern hemisphere, while biodiversity losses are greater in many regions of the south.
4. How does global trade influence these impacts?
Around 18% of impacts are linked to internationally traded food, meaning that consumption in one country can drive environmental pressures elsewhere. Brazil is the largest net exporter of both carbon and biodiversity loss, and China the largest net importer, with animal products dominating the biggest trade flows between them.
5. What are the most effective ways to reduce these impacts?
Because impacts are concentrated in a small number of foods and locations, targeted changes could have a large effect. The study models that restoring the 30% of farmland with the highest both losses could sequester carbon equivalent to roughly 15 years of global fossil fuel emissions and prevent around 60% of the extinction risk currently caused by farmland. Shifts in diets, improvements in agricultural production, and better coordination across supply chains all have a role to play.
6. What are the main limitations of the research?
The study is based on global datasets and modelling, which involve uncertainties. While the overall patterns are robust, the authors note that results at very local levels should be interpreted with caution and would benefit from more detailed data.
Nature Food
High-resolution carbon and biodiversity mapping shows correlated losses across space and agricultural products