Designing the Grazing Plan: Paddocks, Water and the Piosphere

#The camp with no water in it

The arithmetic in the last lesson produces a number — 21 camps, say — and the number produces a temptation. Peg it out, string the wire, admire the map. Then in January the mob walks into camp 14 and finds no trough, because camp 14 was drawn with a ruler and the borehole is in camp 9. The animals walk back to camp 9's fence, stand there, and graze camp 14 at a radius from somewhere else.

That is not a fencing problem. Water is what makes a camp usable.

#The paddock count is also a water-point count

Take the arithmetic straight through. Last lesson's 180 ha block wanted a 60-day late-season recovery on a 3-day graze, so (60 ÷ 3) + 1 gave 21 camps of about 8.6 ha each. Every one of those 21 needs drinking water inside its own fence. The paddock count is therefore also a watering-point count — and that is the line people leave out of the budget. Going from 8 camps to 21 does not multiply the fencing and leave the water alone; it multiplies the pipe, the troughs, the float valves and the leak-hunting too. Reticulation cost scales with camp count, not with hectares.

Two ways out, both legitimate. Shape camps so several meet at one point and share a trough through the fence. Or accept fewer camps and a longer graze period, and lose some recovery precision. What is not legitimate is drawing 21 camps, budgeting 21 fences and one borehole, and finding out in January.

#Sizing water for a mob, not for a herd

Continuous grazing spreads demand across the day and across the farm. A single mob in a single camp does not: everything arrives at once. Cattle drinking at about 32 °C take roughly 48 litres a day for a 270 kg animal, 68 litres for a 410 kg cow and 78 litres for a 725 kg bull; in winter as much as 90% of the requirement can come from the pasture itself (Farmer's Weekly). Run the 30-LSU mob from the last lesson on a hot January afternoon and you need roughly 2 040 litres a day, most of it drawn in two or three short peaks. The failure mode nobody costs is the recovery rate of the trough, not the daily total — a 2 000-litre reservoir refilling at 200 litres an hour empties in one peak and leaves the tail of the mob waiting. Size delivery to the peak.

Then test what is in the water. South African target water quality ranges for livestock watering sit at TDS 0–1 000 mg/ℓ, sulphate 0–1 000 mg/ℓ, sodium 0–2 000 mg/ℓ, magnesium 0–500 mg/ℓ, fluoride 0–2 mg/ℓ and nitrate 0–100 mg/ℓ (DWAF Volume 5). Salinity, sulphate and nitrate are the common borehole offenders; blue-green algae is the seasonal one on surface dams, and the guideline is blunt — visible blue-green scum means a high risk of acute toxicity, and access must be denied.

#Put the piosphere where you want it

Every water point creates a piosphere — the trampled, dunged, hammered ring where animals concentrate. You do not decide whether you have one. You decide where it is.

This is the best-evidenced South African soil finding in the area. Sampling communal, commercial and land-reform systems across a clayey grassland and a sandy savanna, Kotzé (2015) found degradation driven by aggregate breakdown, organic-matter loss and nutrient loss in the sacrifice area around the water point — and found early deterioration of aggregate structure near water points even in rotational camps that showed little degradation overall (Kotzé 2015, University of the Free State). Rotation did not abolish the piosphere; it made the rest of the camp better.

So moving a trough is a soil intervention — which is where the soil form map from Module 2 stops being a classroom exercise.

The design move almost nobody makes: site the trough on ground you have already damaged, or on ground that can take the punishment, and where you want the fertility to end up. Old lands, bare scalds, a sheet-eroded patch — that is where a concentration ring is a gift rather than a wound.

#Fencing: the specifications are published, the price is not

The electric fencing specifications for South African agriculture are public and worth following (KZN DARD): cattle on 3 strands at 290 / 600 / 900 mm; sheep on even terrain on 4 strands at 150 / 335 / 600 / 900 mm; minimum voltage of 3 000 V for wooled sheep, 2 000 V for cattle and goats and 1 500 V for thin-skinned animals; energiser sizing of about 1.0 J per 15 km, with a 20 J unit running up to about 175 km. Then the part farmers skimp on: earthing on 1 m galvanised 20 mm pegs in permanently moist soil, at least 20 of them interconnected for a 20 J unit.

What is not published is what any of it costs on a farm today, and this course will not invent it.

#Theft is a design constraint, not an afterthought

In the January to March 2025 quarter alone, just over 19 000 sheep, nearly 12 000 cattle and 12 000 goats — about 43 000 head — were reported stolen nationally across 5 671 stock-theft cases, with direct losses of at least R200 million; more than 25 600 counts of stock theft were recorded over the preceding year. AfriForum is explicit that the R200 million is direct loss only, excluding lost breeding potential, security spending and higher insurance premiums (AfriForum, 6 June 2025).

Now put that next to the design you have drawn. Adaptive multi-paddock grazing puts your whole herd in one small camp at one known location, on a schedule, often far from the house. A single mob in a single small paddock is a single point of failure for a thief. No imported curriculum mentions this, because where those curricula were written it is not the binding constraint. In peri-urban Gauteng it is.

That does not argue against subdivision. It argues for building the plan so the mob is reachable — the tightest camps nearest the homestead, a night kraal or a herded return kept practical, the far blocks used in the season when you can watch them, and your identification marks kept current so a recovered animal is provably yours.

#Infrastructure is the last money you spend

The cost structure of an extensive South African beef enterprise, from the Land Bank's own budget book: roughly three-quarters of variable cost is bought feed, and 55% is winter lick alone. Repairs to troughs, fences, windmills and dams are under 1% of the annual variable cost (Land Bank Livestock Enterprise Budget 2023/24 v2). The capital cost of infrastructure is separate, and it is large.

Read that and the spending order writes itself. Every intervention that genuinely extends the grazing season — deferred camps carrying standing forage, a rested block held as winter reserve, calving timed so peak cow demand meets spring growth — attacks the largest cost line on the farm. Subdividing camps attacks a line already under 1%, in the hope of an ecological return the next lesson will show is contested. The published South African survey work is direct: increasing grazing density "requires large monetary investment in fencing and watering infrastructure with little to no gain in productivity to offset these costs" (Venter, Cramer & Hawkins 2019).

So spend in this order, and be willing to stop at any step:

1. Fodder flow. Use the grazing-days arithmetic — farm size ÷ grazing capacity × 365 — for the LSU-grazing-days you hold against the days you need to the next reliable green flush (Elsenburg). Fix the gap with stocking rate and reserve camps before you fix it with wire.
2. Breeding season. A controlled season, pregnancy diagnosis and culling the empties move more money than any fence.
3. Water. Then, and only then, troughs and reticulation — water is what makes existing camps usable.
4. Fencing. Last, and financed on the right tenor: the spend is front-loaded and the return back-loaded into years three to seven, so a twelve-month facility against a seven-year return is how a drought in year two forecloses on your fence.

The next lesson decides how much of this you should build at all — because the evidence that subdivision produces more grass and more meat is weaker than the people selling it say, and more interesting than the people dismissing it admit.