One Amaranthus plant, 400,000 seeds: herbicide resistance tightens its grip on SA farms

A single Amaranthus plant can produce between 300 000 and 400 000 seeds, and roughly 80% of that seed germinates in the first year. That arithmetic, laid out by Professor Juan Vorster of the University of Pretoria and co-leader of the South African Herbicide Resistance Initiative, explains why a few weeds standing after a spray is never a minor setback — it's the opening move in a battle that gets more expensive every season.

Across South Africa, herbicide resistance is now a documented problem in no-till grain operations that lean almost entirely on chemical weed control. Common pigweed (Amaranthus hybridus), Palmer amaranth (Amaranthus palmeri), ryegrass (Lolium species) and fleabane (Conyza bonariensis) are the headline species. Others, such as feathertop grass (Chloris virgata), are queuing up behind them.

How does resistance actually develop?

Palesa Mmereki, a molecular biologist at the Agricultural Research Council (ARC), frames it as pure natural selection. "When a herbicide is repeatedly applied, most susceptible weeds die, but individual plants with genetic characteristics that allow them to survive may remain," she said. Those survivors reproduce, and with each generation a larger share of the population withstands a dose that would once have killed it.

Vorster adds that farming practice accelerates the process: overreliance on glyphosate, limited herbicide rotation, and production systems depending almost entirely on chemical control.

The true scale of the problem is unknown. Of the 49 countries in sub-Saharan Africa, only three have reported herbicide resistance cases — the first almost 40 years ago, with sporadic reports since. Vorster blames a shortage of specialised weed scientists, saying available data probably understates reality.

Where are the hotspots?

Summer rainfall regions. Glyphosate-resistant common pigweed has been confirmed around Winterton and Bergville in KwaZulu-Natal, with the problem extending beyond. In Gauteng, Mpumalanga and the Free State, populations resistant to both glyphosate (HRAC Group 9) and ALS-inhibiting herbicides (Group 2 — group numbers identify herbicides by the plant process they disrupt) are tightening their hold in Delmas, Standerton, Reitz, Heilbron and Bethlehem, particularly in maize, soya bean, dry bean and sunflower fields. Atrazine resistance, a Group 5 herbicide, first surfaced in 1990 and is reappearing.

Palmer amaranth dominates in the Northern Cape, North West and Limpopo, with glyphosate resistance the primary concern in cotton — though it also hits maize and soya beans. Vorster warns these resistant Amaranthus populations are spreading fast.

Winter rainfall regions. In the Western Cape's cereal belt, ryegrass troubles wheat, canola and other crops, showing resistance to glyphosate and paraquat (Group 22), and spreading beyond established areas. Mmereki points to wheat monoculture inflating the seedbank. Ryegrass seed drops to the soil, some germinating, some staying dormant — so removing visible weeds doesn't clear the field.

How fast can it happen, and how does it travel?

Vorster cites resistance developing in as few as three generations under certain conditions, referencing a US experiment where Amaranthus developed resistance after exposure to sublethal doses via spray drift. Similar conditions exist along field boundaries. From there, harvesters move seed between fields and farms, mud on vehicles carries it further, and birds, livestock and wind all contribute. Mmereki notes South Africa has no phytosanitary regulations covering machinery movement between farms.

The biggest risk is in-field reproduction. With 80% first-year germination and viability up to three years, a handful of survivors rapidly rebuilds the seedbank.

What does testing look like now?

The clearest warning sign, Vorster says, is patchy survival — some weeds of one species dying while neighbours live. But rule out application errors first: "Spraying weeds at the correct growth stage is essential, as plants that have grown too large may survive an application even when they are not resistant."

Traditional confirmation means collecting seed, growing it under greenhouse conditions and applying different herbicide doses — a six-to-eight-week process during which surviving weeds keep producing seed. Molecular testing is faster for known mechanisms, but Mmereki cautions that reliable genetic markers for non-target-site resistance (mechanisms other than changes to the herbicide's target within the plant) aren't yet available in South Africa, so greenhouse testing remains necessary.

What is HARP changing?

The Herbicide Action and Resistance Platform (HARP), launched in South Africa in 2026, unites Grain SA, the University of Pretoria, Stellenbosch University, the University of the Free State, the ARC and CenGen. Elsabet Wessels, CenGen's acting chief scientist, says the organisation is developing DNA-based testing to identify known resistance mutations.

Phase one implements a University of Pretoria panel for common pigweed and Palmer amaranth that distinguishes between Amaranthus species and flags mutations for glyphosate and ALS resistance. A second phase, subject to funding, covers ryegrass, including glyphosate, ALS and ACCase inhibitor (Group 1) resistance, with fleabane also under consideration.

Molecular results can come back within two weeks when a known mutation is responsible. Young plants can be tested before they set seed, and a single leaf sample screens for resistance to multiple herbicide groups. But Wessels is clear about limits: "New or previously unidentified mutations, as well as resistance arising through entirely different mechanisms, such as non-target-site resistance, will not be detected by the genetic tests and will need to be confirmed phenotypically by other HARP partners."

Combined genetic and high-throughput greenhouse screening could identify emerging hotspots before they spread, eventually feeding a national database and resistance map to inform weed-control decisions.

How do farmers break the cycle?

Vorster's advice is blunt: inspect fields after every application, remove and burn survivors rather than leaving them to seed. Rotate herbicides by HRAC group number, not product name — different brands can share active ingredients acting on the same target. Mixtures help.

Mmereki recommends integrating crop rotation, herbicide rotation, pre- and post-emergence control and consistent scouting. "Diversifying cropping systems can reduce reliance on the same herbicide groups, whereas repeated maize and soya bean rotations or monocultures can create difficulties, particularly where glyphosate and ALS resistance are already present," she said. Cover crops, including sorghum-based cereals, suppress weeds, though Vorster cautions Amaranthus competes hard for nitrogen — extra N can feed the weed more than the crop.

For no-till farmers, whose systems trade herbicide reliance for erosion control and moisture retention, the stakes are highest. Early detection makes control possible, though rarely cheap; the goal is stopping survivors from replenishing the seedbank. Mmereki argues South Africa needs better local data on seed longevity, dormancy and emergence, so farmers can plan multi-season depletion strategies. Watch for HARP's pigweed and amaranth panel results — two weeks from leaf to answer — as the next practical tool to land.