Honey Standards, Pollination and What an Apiary Honestly Earns

#Fifteen hundred rand

In November 2024, DALRRD inspectors went through wholesalers at Crown Mines and Kempton Park. Between the 25th and the 29th they seized 1 059 honey-based syrups and 388 non-compliant honey products under the Agricultural Product Standards Act 119 of 1990. Two wholesalers were fined R1 500 each (SAnews).

Hold those two numbers next to each other. Nearly fifteen hundred products off the shelf, and a penalty that would not cover a decent bee suit.

South Africa produces roughly 2 000 to 2 500 tonnes of honey a year and consumes about 5 000 tonnes. Imports climbed from 476 t in 2001 to 4 206 t in 2017, around 76% of it from China over the seventeen-year average, and the trade has no enforcement body, no self-regulatory body and no ethical trade body behind it (Sihlobo, 15 Dec 2025).

You are not going to win that fight on price. What you can do is be the producer whose jar is exactly what the label says, sold to someone who knows where the farm is. That starts with a gazette from the year 2000.

#What the law says honey is

The regulation sets the limits and says nothing about how you hit them. Beekeeping practice reads two of them as management tests rather than laboratory curiosities, and it is worth knowing which — while remembering that the reading is practice, not law.

Moisture is the one tied to harvest timing: 20% is a ceiling to stay under, not a target to aim at, and it is the limit your own decision about when to pull supers moves most directly.

HMF and diastase are the ones tied to heat and storage. Warm a batch hard to make it flow through a filter and you can move both in the wrong direction at once. Neither is something you can eyeball — if you intend to sell into a market that will test, budget for the analysis rather than assuming.

#The label

Note item 5. "My honey is raw so the labelling rules do not apply to me" is exactly backwards — R.835 defines raw honey and adds a requirement, in specified letters, to everything else you already owe.

Item 9 carries a trap that will catch you if you work from an old checklist. R.835 requires net mass "in accordance with the Trade Metrology Act, 1973". That Act was repealed in whole by section 42(1) of the Legal Metrology Act 9 of 2014 (Legal Metrology Act, Schedule 1). The obligation did not disappear — it moved. Net-mass declaration on a pre-packed retail product now runs under the 2014 Act, administered by the National Regulator for Compulsory Specifications, not the SABS. R.835 predates the repeal and still carries the old cross-reference; whether it has since been amended could not be confirmed, so check the current consolidated regulation before you print a thousand labels.

#Ten hives, honestly

Now the arithmetic. Everything below is dated, and the dates matter more than the rands.

Run it. Ten hives at the planning yield, all sold direct at the March 2025 price: 100 kg × R100 = R10 000 at the bottom, 300 kg × R150 = R45 000 at the top. That is gross revenue, before a single replacement colony, before fuel, before the 30 to 40% security premium from the previous lesson, and before your labour.

Now anchor the yield against history instead of against a brochure. The long series is sobering:

• 8.3 kg per hive per year — the average across 1911 to 1988
• 24 kg per hive — the 1974/5 national census
• about 25 kg per hive — the 2008 national unweighted average

(UCT CSSR WP 221)

At the 1911–1988 average, ten hives produce 83 kg, which is R8 300 to R12 450 gross at 2025 direct prices. At the 2008 average of 25 kg, ten hives produce 250 kg, or R25 000 to R37 500.

And now the number that matters most in this lesson. Commercial beekeeping content claims R40 000 to R100 000 or more a year (March 2025) from ten hives across honey, comb, pollination, swarm removal and wax. An independent report on a ten-hive hobbyist put actual earnings at about R20 000 a year against R15 000 in startup costs, and described a twenty-hive operation earning something comparable to a domestic worker's annual pay (GroundUp, 22 Jan 2021).

R20 000 observed in January 2021, against R40 000 to R100 000 claimed in March 2025. The observation sits almost exactly on the historical yield anchors; the claim requires the top of the yield range, the top of the price range, near-perfect direct sales across five revenue lines, and no colony losses at all. Build your budget on the observation and treat anything above it as upside you have to earn.

#The pollination economy, and why most of it is not yours

South Africa's real bee money is in pollination, and it is concentrated.

Deciduous fruit is planted on 54 294 ha, of which 43 999 ha — 81% — requires pollination, and 70 to 90% of that industry depends on it. Hortgro's own hives-per-hectare figures are apples 2.5, pears 5, apricots 2, plums 6. In 2019 the industry spent R141.3 million on pollination at an average R980 per hive, about 3% of pre-harvest costs (Hortgro, 29 Apr 2021).

For a Highveld farm this is a market you can see but mostly cannot reach. The deciduous fruit is in the Western Cape, the canola with it, and the macadamias are in Mpumalanga, Limpopo and KwaZulu-Natal. Getting hives there means migrating them — which is the documented vector of the capensis clone from lesson 1, and a theft exposure in transit on top. Do not build a Gauteng bee enterprise on pollination revenue.

#When two industry numbers cannot both be true

This is the exercise the module exists for, and it uses real published figures.

Melin et al. (2014) put the Western Cape deciduous-fruit pollination requirement at about 30 000 managed hives (SAJS).

Hortgro (2021) puts the national deciduous-fruit requirement at about 91 000 hives, derived by assuming 1.5 pollinations per hive across those 43 999 hectares (Hortgro).

The Western Cape is the overwhelming majority of South African deciduous fruit. So these two figures describe substantially the same orchards and differ by roughly a factor of three.

Three more figures in this industry deserve the same treatment before you use them:

• No defensible current national hive count could be found. The best available estimate is 90 000 to 110 000, from data around 2004 to 2008. The registered-beekeeper series — 1 246 in 2016, 4 459 in 2021, "fewer than 2 000" in 2025 — is internally inconsistent, almost certainly because registrations lapse.
• The "up to 30% of hives lost annually" figure has no method behind it. It traces to a commercial beekeeping-supplies website with no sample, no method and no date. It is not a survey result.
• No national varroa infestation survey published since 2015 could be located. The resistance finding from lesson 3 is well evidenced and mechanistically explained. Whether it still holds across managed populations is a question nobody appears to have re-measured.

None of that means the industry is dishonest. It means it is under-measured, which puts the burden of measurement on you.

#What to do on Monday

Print R.835's regulation 6(3) table and its marking list, and check one jar of your own honey against both — moisture and letter heights first, because those are the two you control without a laboratory. Then write a ten-hive budget on the R20 000 a year observed in January 2021 and the 8.3 to 25 kg historical yield anchors, not the brochure, and see whether the enterprise still earns its place on your farm. If it does not on paper, it will not in the shed either.