Grazing Days, Density and the Elsenburg Overstock Reveal

#One hundred ewes, a hundred and ten lambs, two rams and four cattle

Two hundred hectares of Overberg veld carrying 30 ha/LSU. That is the whole operation. Walk onto it and you would call it a smallholding — a handful of cattle and a modest flock, nothing that looks like pressure.

Convert it properly and it is 30.38 LSU on land that carries 6.7.

That is a 4.5-times overstock, and the farmer would never have found it by counting heads. He would have found it eventually — in bare patches, in a lambing percentage that slid three years running, in a winter where the veld ran out in June — but by then the finding costs money to fix instead of an afternoon.

This lesson is the arithmetic that catches it early. It is four numbers, one formula, and one vocabulary trap that can put you on the wrong side of a conversation with a regulator.

#Four numbers, four questions

The ARC's framing is the cleanest: grazing capacity is "the true number of animals the vegetation can sustain"; stocking rate is "the perceived number… or the actual number of animals kept". And the two "sometimes differ by as much as 300%, leading to total degradation" (Mokolobate et al. 2017).

Elsenburg puts it in a sentence you can hold onto: grazing capacity is how many animals you can keep on the farm; stocking rate is how many animals you are really keeping.

#Density has no time term

This is where people get it wrong most often, and the mistake is dimensional rather than conceptual.

Stocking density is simply LSU divided by the hectares of the paddock being grazed, measured at a moment. 75 LSU on 1 ha is 75 LSU/ha. The same 75 LSU on 2 ha is 37.5 LSU/ha — whether they stand there for one day or three.

The days do not enter the calculation. If your density figure has a /day in it, you have computed something else and given it the wrong name.

What does carry a time term is stock days per hectare: LSU × days ÷ hectares. Density answers "how tightly are they packed right now?" Stock days per hectare answers "how much grazing did this hectare supply?" You need both, and they are not substitutes.

#The vocabulary trap

Now the part that can cost you.

A widely used South African regenerative teaching page inverts two of these terms. It defines "stocking rate" as ha/LSU — which is what the department calls grazing capacity — and defines "grazing capacity" as stock days per hectare, computed as 365 ÷ ha/LSU. On that page, 6 ha/LSU becomes 60.8 SDH, then gets rainfall-adjusted as SDH × rainfall ÷ 100 mm (regenagsa.org.za).

The arithmetic is genuinely useful. Stock days per hectare is a better planning currency than ha/LSU because it is additive — you can add days across paddocks and across a season, which you cannot do with an area-per-animal figure.

But the labels conflict with CARA, with the department, and with the journals.

#The one formula this course keeps returning to

Here it is, from the Elsenburg infopak:

Grazing days available = farm size (ha) ÷ grazing capacity (ha/LSU) × 365

That is it. Hectares divided by hectares-per-animal gives you animals; multiply by 365 and you have LSU-grazing-days — a bank balance of feed, denominated in animal-days.

Take 200 ha at 30 ha/LSU:

200 ÷ 30 × 365 = 2 433 LSU-grazing-days.

Now spend it. For 10 LSU that is 243 days — about eight months. For 30 LSU it is 81 days — under three months. Same veld, same year, and the difference between a farm that works and a farm that is buying fodder from April.

This formula is the one that converts "how many cows?" — a question with no honest answer — into "how many days?", which is the question that actually plans a year. And because it is denominated in days, it maps directly onto a calendar: you can see the month your balance runs out.

#The Elsenburg reveal, worked

Back to the farm from the opening. 200 ha, capacity 30 ha/LSU, and the stock:

Now the capacity, and here is the second lesson — the more valuable one.

Against full-year capacity: 200 ÷ 30 = 6.7 LSU. The farm is running 30.38. That is a 4.5-times overstock.

Against an eight-month working figure — because in the Overberg pattern the veld camps are used roughly April to November, 245 days, with crop residues carrying December to March: 200 ÷ (245/365 × 30) = 10 LSU. Elsenburg's own "ideal stocking rate" for this farm sums to 10.03 LSU. Against that denominator, 30.38 LSU is a 3-times overstock.

Both numbers are correct. They answer different questions. Say which denominator you used, every single time — because a farmer who reports "three times overstocked" and a farmer who reports "four and a half times overstocked" may be describing the identical farm, and the gap between them is not a disagreement about the veld. It is a disagreement about how many months of the year the veld is expected to work.

#The same arithmetic, pointing the other way

The Elsenburg case is a farm that looks understocked and is not. Now run the identical sum on a farm that looks recklessly overstocked and may not be — because this is the case the arithmetic alone cannot settle, and pretending otherwise is how the map gets used badly.

A 15-hectare Midvaal smallholding on the Highveld, three years in. Fifty sheep, moved every day. In three years the flock has grazed roughly 6 hectares of the 15; the remaining 9 has never carried an animal and is being built with imported horse manure and straw bedding.

Convert it with the Meissner factors from lesson 1 — and note how much the class assumption moves the answer:

Run the grazing-days formula on the conservative version. Available at the mapped rate: 15 ÷ 6.8 × 365 = 805 LSU-days. Needed for a year: 7.5 × 365 = 2,738 LSU-days. Balance: −1,932 days. On the department's own arithmetic this farm is running at three to four times its polygon, and on the 6 ha actually grazed it is closer to eight.

And yet: three years in, with 60% of the farm never touched and the grazed ground reported as improving.

#Do it on your own farm

If the balance is negative, you have exactly four levers, and only four:

Reduce the demand. Destock — deliberately, in planned tranches, and early. Elsenburg's sequence is castrated animals first, then older animals, then the culls, including ewes that did not lamb last season, applying strict breed standards when choosing. You are protecting the genetics and the veld, in that order.

Buy feed. Legitimate, and expensive. But do it in a feedlot or a defined feeding area — never on the veld. Elsenburg is explicit about why: supplemented animals still graze, so feeding on the veld pays the fodder bill and the degradation bill at the same time.

Shorten the period the veld must carry. Crop residues, planted pasture, a rested block held as standing hay. This is what a fodder flow plan is.

Or accept degradation — which is a decision even when it is made by not deciding.

What you cannot do is hope. The South African evidence on that is unflattering: farmers destock reactively and sell "regardless of the market price", distorting farm planning and income (Nketiah et al., Jàmbá, 2024). The value is not in the destocking decision itself. It is in having made it in writing, with triggers, before the season.

#What changes on Monday

Do the four-minute version tonight: total LSU, hectares, ha/LSU, times 365. Then divide by your LSU total and read off the number of days.

If that number is smaller than the number of days between now and your next reliable green flush, you have a problem that will not improve by waiting. Write the shortfall down, put the date on it, and decide which of the four levers you are pulling — while the market is still normal and the choice is still yours.