Soil across much of the world is being lost faster than nature can replace it, according to a new report from the Food and Agriculture Organization of the United Nations.
The report, "Status of the World's Soil Resources 2026", synthesises a decade of research on the condition of soils worldwide and the threats to them. It identifies erosion as the critical one.
Among its lead authors is Isaac Larsen, an associate professor in the Department of Earth, Geographic, and Climate Sciences at the University of Massachusetts Amherst, who was a lead author of its regional assessment for North America.
The central concern of the FAO assessment is that soil can be lost far faster than it can naturally regenerate. Larsen's own earlier work put a figure on that gap in the American Midwest: a 2022 study in the journal Geology, on which he was senior author, found erosion across the region running at ten to a thousand times the pre-agricultural rate.
The comparison is between two rates: how fast soil leaves a field and how fast new soil forms in place beneath it. The 2022 study measured the second for the first time in the region, using beryllium-10 left in remnant prairie soils by cosmic rays to put the median pre-agricultural erosion rate, which reflects the rate at which soils form, at 0.04 millimetres a year. Against that baseline the modern rate runs from ten times faster to a thousand times faster, depending on the site.
Research he co-authored in 2021 put numbers to what that has already cost one farming region. In the US Corn Belt, about 35 per cent of cultivated land had lost its A-horizon topsoil, the dark upper layer that carries most of a field's organic matter. The study estimated the loss had cut crop yields by 6 per cent in the affected region and was costing $2.8 billion a year.
"Soils are fundamental to the well-being of humanity," Larsen said. "The way we treat soils impacts our health, water, climate, jobs and our ability to feed ourselves - the future of food depends on the future of soil."
The report's wider finding is that the knowledge required to manage soils better has grown substantially while practice has not followed. Most of the world's soils, Larsen said, are being "managed in ways that are unsustainable".
The recommended measures are long established and unglamorous: reduced tillage, cover crops, crop diversification and agroforestry. The report's emphasis is on translating the science into policy, on the argument that agricultural productivity and food security depend on it.
None of the four is novel, and that is the report's point. The measures are designed to keep soil where it is, by disturbing it less, covering it for more of the year and giving it roots to hold on to, and the obstacle to using them more widely is not scientific.
It is the ratio that rewards a second look. A system drawing down between ten and a thousand units for every one it puts back is not, stated in those terms, an environmental crisis. It is a direct debit.
The Department of Unnecessary Research has worked the figures through on that basis. At the upper end of the range, a single year of farming is funded by roughly a millennium of deposits. At the lower end it is a decade. Either way the account is in credit on the day, which is the only day most accounts are examined.
Double-entry conventions are satisfied, and unusually well. The debit side is documented to a standard most ledgers never reach: 35 per cent of the Corn Belt's cultivated area, 6 per cent of its yield, $2.8 billion a year. The credit side is also known, and arrives at a rate best described as geological.
On that reading, the four recommended practices are not conservation measures but instalments. Reduced tillage, cover crops, crop diversification and agroforestry are a repayment schedule, and at the prevailing ratio the first of them roughly services the interest on 2026.
No overdraft facility has been arranged. The field is the balance, and the account closes itself.