Breeds, Indigenous Chickens and Worked Economics

#The most expensive decision you will make is a logistics decision

In SAPA's own survey of small producers, laying mash bought in bags cost R8 572 a tonne. The identical feed bought in bulk cost R6 091 a tonne, and large commercial buyers paid R5 954 (SAPA Industry Profile 2023, Table 18 — 2H 2023 prices, excluding VAT and delivery). Ninety-two percent of small egg producers bought in bags.

That is roughly a 40% feed-cost handicap, paid every month, by almost everybody. On ten tonnes a year, at those 2H 2023 prices, it is R24 810 — more than the entire annual margin on a 200-hen layer flock, as you will see below.

So before we argue about Boschveld against Lohmann, know what the argument is worth. Genetics moves your feed conversion by a fraction. How you buy the feed moves your feed price by forty percent.

#The one South African breed with a properly published record

Most of what circulates about indigenous South African chickens is repetition. One breed is different. Boschveld — a Venda × Matabele × Ovambo composite developed for rural free-range production — has a peer-reviewed intensive production record worth reading line by line.

Okoro and colleagues followed 1 432 layers, drawn from 2 467 day-old chicks, from 21 to 44 weeks of lay (Okoro et al. 2017, S. Afr. J. Anim. Sci. 47(2)):

Mortality over the whole period ran between 0.03% and 3.97% — the paper reports it as a range, not as a trend.

Read the top row again. Hen-day production rose only to 56.56% — at its best, the flock laid an egg on 56 days out of every 100 hen-days. This was an intensive trial on a full ration, not a scavenging village flock, so it is the breed's upper case, not its typical one.

Potchefstroom Koekoek, the ARC-developed dual-purpose composite, is the other bird everyone names — and here the evidence is much weaker. The frequently repeated figures for first lay, annual egg numbers, average egg weight and hatchability could not be confirmed against a primary source: the review carrying them is paywalled beyond its abstract, and the ARC Irene data reproduced by FAO is not machine-readable. Treat every Koekoek annual-production number in circulation as unsourced until ARC Irene confirms it. Do not put one into a spreadsheet you intend to plan on.

What is published is a single comparison, and it is stark. At 24 weeks, a Potchefstroom Koekoek egg averaged 42.57 g ± 3.13 against a Lohmann Brown's 68.30 g ± 5.23 — 300 eggs per breed, from 100 birds of each, P < 0.001 (Hlokoe, Tyasi & Mbazima 2026, PLoS One 21(1):e0340483, University of Limpopo).

Carry two caveats with that pair of numbers wherever you repeat them. It is one timepoint at 24 weeks, and the paper records that the Lohmann began laying at 18 weeks and the Koekoek at 20 — so one bird is six weeks into lay and the other four, and early eggs are small eggs. This is not a peak-production comparison. And 68.30 g is a high figure for a Lohmann Brown at 24 weeks against published breed standards, so the gap is probably at the wide end. The direction is not in doubt; the magnitude is one study.

#Feed is the enterprise, and it is a bet on SAFEX

Feed is 65–70% of broiler production cost — SAPA uses both figures in different places, and Agribook gives about 70%. Work with the range; anyone quoting a single decimal is inventing precision. Either way, a poultry enterprise is a geared bet on the grain price.

The Q4 2025 SAFEX spot averages — and NAMC's "Q4 2025" means January to March 2026, aligned to the financial year — were yellow maize R3 398/t, down 29.86% year on year, and soybeans R6 814/t, down 23.11% (NAMC Poultry Products Price Monitor, Issue 25, June 2026). Both fell sharply — a feed-cost model built in a high grain year will mislead you in a low one. Rebuild the budget every season.

#The broiler batch, backed out and then re-modelled

SAPA reports that small-scale producers sold dressed birds at R40.37/kg against a commercial R31/kg (2H 2023). That number is useless until you know what bird stands behind it, so back it out. R40.37/kg equates to R79.12 per bird sold, so the carcass is about 1.96 kg. At SAPA's dressing assumption of 72%, the live bird was about 2.72 kg. At SAPA's FCR assumption of 1.7, that bird ate about 4.6 kg of feed. Only now can you build a budget.

Now the correction almost every farm budget gets wrong. It is tempting to scale that R18.87 by the survival rate — 95% survival, so about R17.90 a bird on those same 2H 2023 numbers. That is wrong, because a bird that dies still cost you its chick price and most of its feed. Model it per bird placed: revenue and the abattoir toll scale with survivors, chick cost does not scale at all, and feed scales only partially.

Done properly, on those 2H 2023 numbers, 5% mortality gives roughly R16.30 per bird placed and 8% gives roughly R14.50 — not the R17.40 to R17.90 naive scaling suggests. A 300-bird batch therefore returns roughly R4 400 to R4 900 per cycle at those 2023 prices, before any overhead, and you need six or more cycles a year for that to resemble a wage. SAPA publishes no small-producer mortality distribution, so 5–8% is a working assumption, and its own respondents named "poor growth and high mortality from disease or cold" as a leading complaint. Re-run the model at 15%. Selling live at the survey's weighted average of R76.32 avoids the abattoir toll and lifts the margin to roughly R24 — the arithmetic behind the live-bird market you met in lesson 1.

#Layers, and the scale conversation nobody wants to have

Now scale it. A 200-hen flock laying 25 dozen per hen per year produces about 5 000 dozen, and — on those 2H 2023 prices — about R26 000 of margin after feed and pullet cost, before labour.

Twenty-six thousand rand, for daily collection, feeding, water, washing, grading, packing, marketing and delivery, seven days a week, all year. Two hundred hens is a side income. It is not a livelihood, and no amount of regenerative framing changes that arithmetic. The livelihood number sits in the low thousands of hens, or at a direct-sale price far above that 2H 2023 R21.28 a dozen — which is why the channel work in lesson 1 matters more than the breed work in this one.

Set that R26 000 beside the R24 810 you would burn buying ten tonnes in bags. Group-buying a bulk load with your neighbours is very likely the highest-return decision open to a small South African poultry producer, and it is a phone call, not a capital project.

#What to do on Monday

Phone two neighbours and price a shared bulk feed delivery; it is worth more than every other decision in this lesson combined. Then take whichever breed paper you were about to believe, find its raw feed intake and its raw production rate, and divide one by the other yourself. And before you place another batch, rebuild your margin per bird placed, at 5%, 8% and 15% mortality.