Soil Carbon Credits Need a Plot, Not a Factor: Measuring Organic Matter for a Hostile Market

Cracked dry soil in submerged village of Aceredo during early 2022's drought in northern Portugal and Spain. Photograph by o_andras, CC BY-SA 4.0, via Wikimedia Commons.

01The macro challenge: a real stock, a hostile instrument

Raising soil organic matter is agronomy — regenerative practice that holds water, nutrients and living roots in the landscape. Selling a credit for that raise is a claim about additionality, permanence, leakage and exclusive title to a tonne. Those are different trades. The first can be done with a spade, a lab and a resample. The second requires a registry, a methodology, an auditor and a buyer who believes all four. The second is the one that has been failing in public. The statute and the market are the entry point. The subject is soil degradation where organic matter collapses, and capture where the serial, not the plot, owns the climate story.

FAO’s GSOC-MRV protocol is the methods document the first trade actually needs. It is a conceptual framework and a set of standard methodologies for monitoring stock change at farm level after the adoption of sustainable soil management, intended for application across agricultural lands including food, fibre, forage and bioenergy crops [14]. That is measurement — climate stewardship as observation. Zhou, Wang and colleagues, working on cropland carbon budgets in the US Midwest, showed why the measurement still has to carry an uncertainty term: SOC stock uncertainty propagates directly into carbon-budget and credit calculations [13]. A point estimate without that term is not conservative. It is incomplete.

The instrument waiting to receive the number is the problem. Probst and co-authors assessed 2,346 carbon-crediting projects covering roughly a fifth of credits issued to date and estimated that more than 84 percent of those credits did not reflect real emission reductions [12]. A 2025 study of 95 Verra-registered projects already flagged for overstatement found that the auditors who had signed them off had not caught the flaws; two-thirds of the Verra-accredited auditors involved were on those files [7][9].

"Farmers are increasingly being challenged to balance productivity, profitability and sustainability. This partnership with Orizon is about supporting our agricultural clients navigate that shift in a practical way, by enabling access to carbon markets that are linked to real farming practices and independently verified outcomes." Bill Blackie, chief executive of business and commercial banking, Standard Bank Group, speaking to Food For Mzansi, 16 May 2026 [5]

Blackie’s May 2026 framing is the one that should stop a cooperative treasurer: the market pitch now has to promise independently verified outcomes linked to real farming practice [5]. Soil MRV that feeds an issuance machine without that link is not a climate service. It is throughput for a serial — and an enclosure of origin ecological work into someone else’s claim. Background integrity studies remain load-bearing method evidence [7][9][12]; they are not current expert colour inside the six-month window.

02The transitional opportunity: sell the measurement, not the serial

The useful product is the time series: soil organic carbon at stated depth, with method, date, coordinates and a named sampler, repeated. That product satisfies a CSRD land-sector line, a buyer’s regenerative-agriculture claim, and a development-finance result. It does not require Verra. It does require the discipline FAO already wrote down [14]. The protocol is written for farm-level projects that adopt sustainable soil management; it is intended for annual and perennial crops, paddy rice and grazing, and it treats GHG emissions and removals together with stock change [14]. That breadth is the point. A cocoa agroforestry plot, a cereal rotation and a pasture can sit in the same methods family. They cannot sit in the same emission factor. Regenerative practice is the work. The time series is how origin keeps the proof without surrendering title.

Uncertainty is not a footnote to that family. Zhou and co-authors showed that SOC stock uncertainty propagates into both the carbon budget and any credit calculated from it [13]. A buyer who accepts a point estimate is accepting a number whose error bar they have not priced. A cooperative that publishes the error bar is selling a more honest product, and a more defensible one when an assurer asks how the figure was made.

If a credit is later issued, it should be issued against that time series, not instead of it, and the tonne should be uniquely allocated. Double-use — inventory reduction plus credit retirement on the same stock change — is the agrifood version of the joint-production problem Caro et al. identified in supply-chain footprinting [15]. A cooperative that keeps the measurement can refuse the second claim. A cooperative that only holds a registry login cannot. Isa Mulder, of Carbon Market Watch, told Mongabay that if verifiers cannot be treated as independent, “the whole thing comes crashing down,” because the market “relies on a lot of trust” in standards, verifiers and intermediaries [7]. A farm-level SOC time series does not require that trust stack. A serial does.

Michael Lilje, speaking on the same May 2026 programme launch, put farmer caution at the centre of any credit pathway.

"Trust and simplicity are essential for farmer participation in carbon markets. There is clear interest from farmers, but also understandable caution. Our role is to ensure carbon projects are robust, transparent and farmer-focused. Working with Standard Bank allows farmers to engage through a trusted relationship, while we ensure the integrity of the carbon credits generated." Michael Lilje, CEO and founder, Orizon Agriculture, speaking to Food For Mzansi, 16 May 2026 [5]

That is not an argument against raising soil carbon. It is an argument against using an opaque issuance layer as the farm’s route to being paid for it. Payment can travel through an offtake premium, a public MRV grant, or a Scope 3 primary-data licence. Those channels do not need the serial to exist first — and they are the channels that leave stewardship nearer the land.

03The Institute analyst take: who owns the tonne — and who captures the claim

Samuel Nguiffo’s warning about forest carbon applies, with less poetry and the same mechanics, to agricultural soil. When a price is attached to a slice of the ecosystem rather than to the people who manage it, outsiders arrive to harvest the slice. “This commodification of nature is likely to disconnect communities from the forest, its traditional benefits and the decision-making processes” [10]. Soil carbon projects that vest issuance in an aggregator, with the farmer as a contracted practice-follower, repeat that disconnect one horizon down. Corporate “regen” programmes that harvest soil data as a condition of offtake and then book the tonne as the buyer’s inventory progress are the same capture under a friendlier label.

Alondra Cerdes Morales, president of the Bribri and Cabecar Indigenous Network in Costa Rica, put the counter-condition without rejecting markets as such: it is “unacceptable that the value of the land is being weighed solely in monetary terms while the community voices and rights are ignored,” and any participation has to be “a more participatory way of working with these project organizers” [10]. A soil-credit contract that the cooperative cannot read, and cannot exit without losing the data, fails that test — and fails the legitimacy test for climate stewardship.

For investors, a portfolio company that reports “insetting via soil carbon” without showing unique title, uncertainty and a resample schedule is holding a claim the Probst and Giles papers have already marked as the default failure mode [7][9][12]. The valuation question is not how many tonnes. It is whether anyone else is allowed to count them, and whether the farmer who raised the organic matter still holds the primary observation. Cannon’s reporting on the voluntary forest-carbon market in 2024 already framed that market as one whose future depends on integrity, not on volume [11]. Soil credits inherit the same test. Volume without unique title is not a climate position. It is inventory risk — and origin enclosure.

04The proactive resolution: stewardship closes at the plot

Implement the FAO protocol at the plots that would have been enrolled anyway [14]. Publish the uncertainty [13]. Licence the time series into buyer Scope 3 and public-programme MRV. Treat a voluntary serial as an optional overlay with a one-claim rule, not as the business model. That sequence survives a hostile credit market. The reverse sequence does not. It also survives a CSRD restatement, because the underlying observation does not depend on a registry remaining credible. If the serial is later withdrawn, the soil measurement is still there. If the measurement was never taken, withdrawal leaves a hole in the climate file and a hole in the offtake story at the same time. Stewardship closes where organic matter is raised and proven — on the plot, with origin holding the record.

The soil can still be improved in a broken market. What cannot be done honestly is to treat that market as the proof that the soil improved, or as climate stewardship. Proof is a resample the producer holds. The serial is, at best, a receipt.