An eight-model ensemble estimates no-till plus cover crops could store 16.4 million tonnes of carbon a year in Midwest cropland

July 2025 — Combining eight crop simulation models, a Michigan State University-led team estimated that switching Midwest cropland to no-till with cover crops would add 0.36 ± 0.12 tonnes of soil carbon per hectare a year, or 16.4 million tonnes across the region. Here's what happened, what it cost, and what went wrong.

Cost and outcome (as of July 11, 2025)

Outcome (modeled): +16.4 Tg C a year for no-till plus cover crops vs business as usual. (Basso et al., Scientific Reports, July 11, 2025)
Net climate (modeled): 1.2 t CO₂e per hectare a year on average after nitrous oxide. (Basso et al., Scientific Reports, July 11, 2025)
Cost: the paper doesn't estimate what farmers would pay or earn.
Conflicts: the lead author co-founded CIBO; two co-authors have Indigo Ag interests. (Basso et al., Scientific Reports, July 11, 2025)

What the study did

The team ran eight widely used cropping-system models (APSIM, ARMOSA, CropSyst, DayCent, DSSAT, EPIC, SALUS and STICS) across 40,000 locations in 934 counties in 12 Midwest states, covering about 46 million hectares. Each simulation cell covered about 4 square kilometres. (Basso et al., Scientific Reports, July 11, 2025)

They compared regenerative scenarios against dynamic business-as-usual baselines and validated the ensemble against long-term field trials. It showed greater accuracy and lower uncertainty than any single model, the authors wrote. (Basso et al., Scientific Reports, July 11, 2025)

Results

No-till plus cover crops raised modeled soil carbon by 0.36 ± 0.12 Mg per hectare a year, or 16.4 Tg a year regionally, while raising nitrous oxide emissions by 0.16 kg per hectare a year on average. (Basso et al., Scientific Reports, July 11, 2025)

Each practice alone gave about half the gain: 8.9 Tg C for no-till and 8.5 Tg C for cover crops on conventionally tilled land. Gains are only fully realized where starting carbon is low, the authors said. (Basso et al., Scientific Reports, July 11, 2025)

After netting out nitrous oxide, no-till with cover crops gave an average climate benefit of 1.2 t CO₂e per hectare a year, and up to 2.5 t less than conventional tillage, varying by place. Cutting nitrogen fertilizer to 75% barely changed soil carbon but lowered nitrous oxide by 15%. (Basso et al., Scientific Reports, July 11, 2025)

Why more models

When the results were scaled up to the Midwest, the coefficient of variation in soil carbon change fell from 99% with one model to 36% with eight. The authors call model choice a "quantitative lottery." (Basso et al., Scientific Reports, July 11, 2025)

Checking individual long-term trial results against meta-analysis ranges, they found that 38% (20 of 52) of peer-reviewed data points fell outside the range, some with unrealistic rates such as 8 Mg C per hectare a year. (Basso et al., Scientific Reports, July 11, 2025)

They warned that short-term sampling, typical of carbon-market contracts that often last only a year, raises the odds of spurious results. (Basso et al., Scientific Reports, July 11, 2025)

Who's behind it

Bruno Basso of Michigan State is lead author, with co-authors including G. Philip Robertson and Keith Paustian. Funders include the U.S. Department of Energy, the National Science Foundation, USDA NIFA, the Walton Family Foundation and the United Soybean Board. (Basso et al., Scientific Reports, July 11, 2025)

The paper discloses that Basso is a co-founder of CIBO Technologies, and that Paustian and Yao Zhang have a financial interest in Indigo Ag. (Basso et al., Scientific Reports, July 11, 2025)

What went wrong

Commercial ties. The lead author co-founded CIBO, and two co-authors have a financial interest in Indigo Ag. (Basso et al., Scientific Reports, July 11, 2025)
Simulated, not measured. Every carbon figure here is a model estimate.
Unreliable trial data. 38% of the peer-reviewed trial data it checked fell outside realistic ranges. (Basso et al., Scientific Reports, July 11, 2025)
Low-carbon soils only. Full gains depend on low starting carbon. (Basso et al., Scientific Reports, July 11, 2025)
What we don't know yet: how the ensemble compares with soil sampling on enrolled farms.

Our follow-up question for Michigan State University

Is the ensemble being used in any CIBO or Indigo crediting methodology?

Sources

Basso B., Tadiello T., Millar N., Robertson G.P., Paustian K. et al., “A multi model ensemble reveals net climate benefits from regenerative practices in US Midwest croplands,” Scientific Reports, published July 11, 2025. nature.com

Disclosure: [not in copy] Editor placeholder