Ten years of no-till moved carbon in the topsoil, not the subsoil

After ten years of a field trial on a Haplic Plinthosol in South Africa, no-till plots held more soil organic carbon than conventionally tilled plots, and the gain sat in the top 20 cm [1]. Crop rotation and residue retention raised particulate organic matter, mostly in the top 40 cm [1]. Between 20 and 60 cm, the authors report limited change [1].

The headline is not that conservation agriculture built carbon through the soil profile. A decade of it changed the topsoil and left the subsoil largely where it was. The authors say so directly: longer-term implementation and monitoring are required to achieve substantial subsoil carbon sequestration and stabilization [1].

That distinction matters if you are planning a low-disturbance system and expecting deep carbon gains on a schedule.

What the trial actually did

The trial was established in 2012, using a randomized complete block design with a strip-split-split plot arrangement [1]. It is a field trial, not a pot study or a simulation. Treatments were no-till (NT) versus conventional tillage (CT); four crop rotations labelled MFM, MFS, MWM and MWS; residue retention (R+) versus removal (R−); and three sampling depths, 0−20, 20−40 and 40−60 cm [1]. Soil was sampled during the 2022 cropping season [1].

The measured set was broad: soil organic carbon, carbon stocks, particulate organic matter, permanganate-oxidizable carbon, total extractable carbon, humic acid, fulvic acid and related humification indices [1]. Soil carbon is not one substance. A review of soil carbon research describes the major pools — particulate organic carbon, mineral-associated organic carbon, dissolved organic carbon, microbial biomass carbon and aggregate-associated carbon — as differing in how they form, stabilize and persist [7]. Measuring several fractions at once is how you tell whether carbon is arriving and staying, or arriving and cycling straight back out.

The numbers, and what they do and do not say

No-till increased soil organic carbon by 0.64%, which the paper describes as a 14.06% improvement over conventional tillage, with greater carbon sequestration in the 0−20 cm layer [1]. The abstract does not state the units for the 0.64 figure, so the relative comparison is the more usable part of that sentence. It reports no variance, p-values or confidence intervals, so from the abstract alone you cannot judge how tight those differences are.

Crop rotation and residue retention enhanced particulate organic matter, particularly at 0−20 and 20−40 cm [1]. Particulate organic matter is the coarse, relatively fresh plant-derived fraction, so this is consistent with more residue entering the system and not being lost immediately.

Then there is a result that runs against the usual story. Conventional tillage increased humic acid concentrations by 25.40%, while no-till increased fulvic acid concentrations by 41.67% in the surface layer [1]. Humic acids are conventionally treated as the more processed, more stable fraction and fulvic acids as the more mobile one, so a tillage system producing more humic acid is not the direction most conservation agriculture writing predicts. The abstract reports the finding without offering a mechanism for it. It is a single-site result and should be read as one.

Principal component analysis identified soil organic carbon, HR and total extractable carbon as the major drivers of variation in carbon stocks [1]. That is a statement about which measured variables co-vary most strongly in this dataset. It is not a causal claim, and the abstract does not present it as one.

What this does not establish

It does not establish that no-till fails to build subsoil carbon anywhere. It establishes that on this soil, at this site, over this period, the subsoil signal was weak [1]. A Haplic Plinthosol is a highly weathered, iron-rich soil, and the result is tied to it.

It does not establish a general rule about conservation agriculture. The trial's own framing is that most South African studies looked at short-term effects and surface layers, which is why this one went deeper and longer [1]. That is a gap being filled, not a question being closed.

It does not tell you what happens at year fifteen or twenty. The authors explicitly call for longer-term implementation and monitoring before expecting substantial subsoil sequestration [1].

And it does not isolate management from climate. A separate study across agroforestry landscapes in Ghana and Burkina Faso, using 1,440 soil samples from protected areas, fallows and croplands across four climatic zones, found climate was the dominant driver of topsoil soil organic carbon variation, explaining 26.3% of it, while the interaction between climate and land use accounted for 34% of the variation in total carbon stocks [14]. Management is not the only lever, and in that dataset it was not the largest one. That study also notes its land-use patterns may partly reflect non-random land selection across heterogeneous landscapes rather than direct management effects [14].

Where the evidence runs out

Start with the shape of the study. One site, one soil type, one sampling season in 2022, and an abstract that reports no statistical detail [1]. The direction of the topsoil result is clear enough. The size of it is not something you can pin down from what is published here.

The humic acid finding is the weakest link in the story. Conventional tillage producing more humic acid than no-till is reported but not explained in the abstract [1]. Until someone works out whether that reflects real humification chemistry, a sampling artefact, or something about this particular Plinthosol, it should not be used to argue that tillage improves carbon quality.

Deep soil carbon is a known blind spot well beyond this trial. A review of soil carbon research in India lists limited long-term field datasets and inadequate representation of deep-soil carbon among the major research gaps, along with uncertainty about carbon permanence [7]. This South African trial is one of the studies starting to fill that gap. One study does not close it.

If you farm perennials rather than annuals, the relevant evidence is different and also incomplete. A review of soil organic carbon in perennial orchards reports that cover crops increased soil organic carbon by 21% and labile carbon by 28% in apple orchards, and that long-term grass cover raised soil organic carbon in pear orchards from 12.70 to 17.65 g per kg [5]. Those are orchard systems, not an annual-crop Plinthosol, and the review itself flags that the sensitivity of labile carbon to future warming is an open question needing urgent attention [5].

Biochar is often proposed as the fix for slow carbon accumulation. A review of biochar in conservation agroecosystems concludes that significant hurdles remain in producing environmentally friendly biochar, selecting the optimal type and amount, and applying it according to local soil characteristics, and notes that biochar's own contaminants and long-term persistence could pose chronic ecological and environmental problems [2]. It is not a settled answer.

What you can take from this trial is narrow and useful. On a weathered South African Plinthosol, ten years of no-till plus rotation plus residue retention improved topsoil carbon and the particulate fraction. It did not move the subsoil. If your expectations run deeper than 20 cm, this paper is a reason to keep measuring rather than a reason to stop.