SA Soils and What Soil Carbon Really Does Here

#Your soil test will start with a zero

Send a sample from a Highveld camp to a laboratory and the organic carbon figure that comes back will very likely start with a nought. Not 3%. Not 5%. Something like 0.6%, or 0.9%.

That is not a failure of your farming. It is the country. Across 2 380 soil profiles compiled by the Agricultural Research Council's Institute for Soil, Climate and Water, 58% of South African soils hold less than 0.5% organic carbon, 38% hold between 0.5% and 2%, and only 4% exceed 2% (du Preez, van Huyssteen & Mnkeni 2011).

Now read that against the material most regenerative teaching is built from. Overseas regenerative content routinely assumes a starting baseline of 3–5% organic carbon. In most of South Africa that starting point does not exist, and the ceiling above it is lower. Every recommendation built on that assumption arrives here with the arithmetic already wrong.

#Two different losses, and only one of them is yours to argue about

The word "degraded" gets used for two completely different events, and confusing them is what makes farmers spend money in the wrong place.

The first loss is the plough, and it is enormous and settled. Across 27 South African soils with between 1 and 85 years of cropping history, organic carbon in the top 20 cm fell by 10–75%, averaging roughly 45% (du Preez et al. 2011, part 2). It is the least contested soil-carbon fact in the country: turning veld into a land takes away close to half the carbon, and it does not come back on a five-year plan.

The second loss is veld condition, and it is real but modest. On Bloemdal-form soils near Bloemfontein, poor-condition veld held 14% less organic carbon than good-condition veld, and bare soil held 26% less — in the upper 25 millimetres (du Preez et al. 2011).

Look hard at that depth. Twenty-five millimetres is a thumb joint. Most South African grazing effects on carbon are surface effects, and the surface is precisely where measurement error, bulk-density artefacts and last week's dung do their worst. A study sampling the top 25 mm and a study sampling the top 300 mm are not measuring the same thing.

#The number that changes sign when it crosses an ocean

If you want one clean example of why an imported figure is dangerous here, take woody encroachment.

Bush and dwarf shrub thickening into grassland has no single effect on soil carbon. It raises carbon at dry sites and lowers it at wet ones, and the crossover in South Africa sits at roughly 750–900 mm mean annual precipitation. In the top 10 cm, the two driest sites studied gained carbon — +108.9% at 300 mm and +55.9% at 350 mm — while the wettest site, at 1 500 mm, lost 50.5% of its 0–10 cm stock, and close to half of it down to a metre (Zhou et al. 2018). The North American threshold for the same phenomenon is around 400 mm, and the authors of the South African work say plainly that theirs is higher and that you cannot extrapolate across continents.

So a consultant applying the 400 mm rule in the KZN Midlands does not get a slightly wrong answer. He gets the wrong sign: he tells a farmer encroachment is building carbon on a site where it is destroying half of it.

#What a tenth of a tonne actually buys

Here is the arithmetic nobody does before signing up for the carbon story.

Internationally, improved grazing-land management is credited with 0.05–0.5 t C/ha/yr (Lal) — an international figure, and South Africa's arid and semi-arid systems sit at the bottom of that band. A long-term season-of-grazing trial in the arid Eastern Cape is reported to have found the exclosure accumulating about 0.128 Mg C/ha/yr against 0.096–0.105 Mg C/ha/yr under grazing (Talore et al. 2016). That paper sits behind a publisher login, so treat the decimals as indicative. But read the direction: grazing slowed accumulation there. It did not reverse it, and the gap is small next to farm-scale measurement noise.

Take the top of the realistic South African range, 0.1 t C/ha/yr, on a 100 ha Highveld holding:

100 ha × 0.1 t C/ha/yr = 10 t C per year
× 3.67 to convert carbon to carbon dioxide equivalent = about 37 t CO₂e per year

Thirty-seven tonnes. Before you deduct a kilogram of enteric methane. Before you pay for a baseline, annual sampling with bulk density on every core, a verifier, or the buffer pool. On a 100 ha peri-urban Gauteng smallholding, a stand-alone carbon project is not a marginal proposition — it is arithmetic that does not close.

What it costs, what it earns

What the South African market actually shows, as reported in early 2026:

The GRASS project issued 266 255 verified carbon units across more than 95 000 ha in its first monitoring period under Verra VM0042 v2.2 — roughly 2.8 tCO₂e per hectare (Carbon Herald, 26 January 2026).
A separate Endangered Wildlife Trust project issued 805 971 units over more than 90 000 ha under a different methodology, VM0026, on a minimum 40-year contract horizon (EWT, 24 February 2026).

Two things follow. First, these are two projects, not one — do not let anyone add their numbers together. Second, the reported abatement is for sequestration or avoided emissions. A large share of South African grassland crediting is payment for not ploughing: a genuine climate good and a genuine income line, but not evidence that grazing management builds soil carbon.

What a participating farmer actually nets per hectare per year, after aggregator fees, verification and buffer-pool deductions, is not published anywhere this course could verify as at July 2026. Ask an aggregator for a signed indicative term sheet, and ask for the sequestration-versus-avoided-conversion split, before you believe a number.

Contested — here is both sides

Does improved grazing build meaningful soil carbon in South Africa?

The optimistic case, at its strongest. Rangeland is the majority land use and the Grassland biome already carries about 31% of national stocks, so the scale is national. South African soils are so depleted that the deficit — and therefore the headroom — is enormous. And there is a real local result behind it: Kotzé (2015) sampled a clayey grassland and a sandy savanna and found soil degradation less pronounced under rotational than under continuous grazing, with rest periods compensating for high grazing pressure in a way continuous grazing could not. That is a South African, peer-reviewed, rest-and-recovery result about soil.

The sceptical case, at its strongest. Franke & Kotzé (2022) argue that it is questionable how far high-density grazing can be called more representative of natural ecosystems than any other system, that associating it with nature has polarised and blurred the discussion, and that the relationship may have led to an overselling of high-density grazing principles, taken up by policy makers and development agencies without sufficient empirical basis. Add the arid Eastern Cape result above, where grazing lowered the accrual rate.

Where this course lands. Improved grazing in South Africa has reasonably evidenced benefits for cover, rest-driven recovery, forage supply, drought resilience and stocking capacity, and weak, context-dependent benefits for soil carbon. Sell the first. Do not sell the second as though it were the first.

#The ceiling nobody can draw for you

A soil cannot hold unlimited carbon. Protective capacity is set largely by clay and silt content: sandy, low-clay Highveld and Kalahari soils have a low ceiling that can be approached within a couple of decades, while high-clay vertic soils have a higher one and are often already close to it.

That is the principle, and it is sound. What does not exist is the number. There is no published South African map of soil carbon saturation deficit, and no national set of saturation thresholds by soil form or clay content that this course could find. So when someone quotes you a ceiling percentage for your Avalon footslope, they are stating something the South African literature has not established. Treat saturation as a principle, not as a figure.

#What you may honestly say at the farmers' day

You may say you are protecting a national stock rather than rebuilding one. You may say that not ploughing your veld avoids a roughly 45% carbon withdrawal, proven on South African soils rather than argued about. You may say that lifting basal cover and litter raises carbon inputs, that the local evidence for rest-driven recovery is genuine, and that your surface millimetres are where you will see it first.

You may not say your grazing will build 1% soil carbon, or that doing so holds an extra 150 000 litres per hectare. On a soil at 0.4% organic carbon — where most of the country starts — a 1% absolute gain is a multi-decade proposition if it is achievable at all, and the water-holding gain from organic matter, while real, is far smaller and slower than that slogan implies.

And say the timeframe out loud: detecting a real change against South African background variability generally takes five years or more and a lot of cores. Anyone offering a verified soil carbon gain in three seasons is selling you measurement noise.

So stop treating soil carbon as the reason to change your grazing, and treat it as a slow byproduct of changes worth making anyway. The next lesson takes the same spade and points it at something that pays back inside one season: what your soil form says about where the water goes.

#Check yourself

4 questions — answers explained as you go

1A neighbour returns from an overseas regenerative course and says his plan is to lift his Highveld camp from 3% soil organic carbon to 5% in a decade. What is the first thing wrong with the plan?
Why: The plan fails at the baseline, not at the ambition. Imported material assumes a 3–5% starting point most South African soils do not have, so every downstream calculation — water holding, nutrient release, credit volume — is built on a soil that is not there.
2On a farm with 40 ha of old lands under maize and 300 ha of veld in poor condition, where is the larger soil carbon prize?
Why: Per hectare the plough is the far bigger event: 10–75% loss, mean near 45%, against 14% under poor veld condition and 26% under bare ground in the upper 25 mm. Veld condition is still worth fixing for cover, water and forage — but ranked as a carbon intervention, the old lands come first.
3Woody plants are thickening on two farms — one at 400 mm rainfall in the Northern Cape, one at 1 200 mm in the KZN Midlands. What does the South African evidence predict for soil carbon?
Why: Encroachment flips sign across a rainfall gradient, and South Africa's crossover is roughly double North America's, so importing the 400 mm figure gives the wrong direction on a mesic farm. Note what this does not license: at a dry site the carbon may rise while grazing value falls, and the forage case still says control the bush.
4A grassland carbon project offers to pay you for credits. Which single question best tells you what you are actually being paid for?
Why: The other three are useful but do not distinguish the two products. Much South African grassland crediting pays for avoided conversion — real value, and a real income stream, but it does not scale with how well you graze and it is not evidence that grazing builds soil carbon. Ask for the split before you sign, and read the contract length while you are there.