The Carbon Beneath the Plough

ANo climate measure sells itself more easily than the idea of paying farmers to put carbon back into their fields. The world's soils hold more carbon than the atmosphere and all living vegetation combined, the plough has been draining that store for centuries, and the remedies on offer, chiefly leaving the plough in the shed, cost a fraction of what industrial capture costs and repay part of their expense in retained moisture. Between 2015 and 2023 the area enrolled in soil carbon schemes worldwide rose from under two million hectares to more than thirty million. The same years produced a second and far less comfortable literature, written by the people sent back to core the fields afterwards. Its finding is monotonous. Much of the carbon credited was never added, and much of what was added does not stay. Whether the fault lies in the way the soil is measured or in the bargains struck with the people who farm it has become the central dispute of the field.

BStefan Traore, a soil biogeochemist at the Nordic Soil Observatory in Uppsala, holds that the bargains are beside the point until the measurement is honest. Between 2013 and 2022 her group re-cored 96 fields in four countries that had been farmed without the plough for at least a decade, taking 1,180 profiles in all, and sampled each field to a full metre instead of to the customary thirty centimetres. At thirty centimetres those fields had gained 0.28 tonnes of carbon per hectare per year. Sampled to a metre the very same fields had gained 0.04 tonnes, a figure her group could not distinguish from zero. Across the same period the area enrolled in the region roughly quadrupled, an expansion that reads in buyer reports as progress. What did not change was the share of fields that could be shown to hold more carbon than when they entered. The reason, on her account, lies in the sampling protocol: with the plough withdrawn, roots and residue concentrate carbon in the top layer while the layers beneath give up a comparable amount. Depths, she notes, had been chosen for convenience rather than for the reach of roots, since a core a technician can push by hand is a core a budget can absorb.

CThe account yields a prediction, and Traore has tested it. If the shallow gain is redistribution working itself out against a ceiling, then fields that gave up the plough long ago should by now have stopped gaining anything. At 23 sites left unploughed for between 28 and 41 years, accumulation measured to a metre had ceased by roughly the nineteenth year, after which the profile held steady. Gains of any size were confined to soils that had already lost a large share of their original carbon; where soils had never been badly depleted, four decades without the plough produced no measurable gain at all. Traore nonetheless presents deep sampling as a safeguard for the crediting system rather than as a verdict on the practice itself, and where soils have genuinely been drained she accepts that what happens next is not a question soil physics can answer.

DRafael Duarte, who studies land economics at the Cerrado Institute in Goiania, answers that carbon lost from suitable ground is telling you about tenure, not about physics. His survey covered 214 farms whose soils decades of ploughing had left far below their original carbon content, precisely the ground on which a gain is physically available, and it recorded, five years after the crediting contract expired, carbon retained on 71 per cent of owner-operated fields against 24 per cent of fields held on leases of under five years. A store of carbon that nobody expects to own in a decade, Duarte argues, is a fallow waiting for a good grain price. He concedes two things. His farms lie within a single state, and tillage was judged indirectly, from the residue cover visible in satellite imagery, because no independent record of field operations exists.

EThe argument left the seminar room some time ago. Since 2022 the standards used to certify soil carbon credits have required sampling to at least one metre and re-verification every five years, and projects unable to produce the deeper figures have had issuance suspended. Traore points out, uncomfortably for her own case, that coring to a metre costs more on a small holding than the credits that holding can earn, so the measure she trusts is the one that prices out the farms most in need of the money. Duarte replies that this appetite for whatever a probe can measure is exactly what keeps the length of a tenancy out of the accounts, since an agreement between landlord and tenant cores no better than a promise does. Both regret one habit above all: that the sector goes on reporting hectares enrolled instead of carbon still held, a practice that flatters every party to it.

FA settlement of sorts is emerging in which the two accounts are assigned to different stages rather than ranked against each other. Where the carbon has merely shifted down the profile, no arrangement between buyer and farmer will conjure a net gain and the field should never have been credited; where the soil was genuinely depleted and a gain is physically available, what survives turns instead on who expects to be farming that field in ten years' time. The unresolved question is whether carbon returned to a soil stays there across the century that permanence rules are written against, and no series of measurements yet runs long enough to settle it. What both researchers reject is the assumption that produced the targets in the first place, namely that a climate can be repaired by counting the hectares that have signed up to repair it.