Rooibos Waited Ninety Years for a Second Variety

Nearly every commercial rooibos field alive today still descends from one selection made in the 1930s, the Nortier type, gathered from a single wild population and propagated by seed ever since. What is genuinely strange is not that this happened. It is that, ninety years on, it is still essentially the whole story. No second named variety followed it. Only in the last few years has anyone mounted a serious attempt to breed one, and the person leading it is a familiar name, decades removed from the last crop he domesticated.

A crop that will not hold a copy

Ask why rooibos never got a proper sequel to Nortier's selection and the answer starts with the plant itself. Aspalathus linearis is what botanists call an obligate reseeder: burn it or cut it back hard and it does not resprout from the root the way many fynbos shrubs do. It depends entirely on a fresh crop of seedlings, a trait already covered in full. That single fact rules out the shortcut most tree and vine crops use to lock in a good selection. A wine grape or an apple gets grafted or cloned from cuttings, so every vine of a named cultivar is genetically identical to the one that impressed a breeder. Rooibos cannot be propagated that way in commercial practice. Every field is grown from seed, and seed reshuffles genes every generation.

The consequence shows up as soon as anyone measures it. A 2008 review in the Journal of Ethnopharmacology cites an earlier study of 97 rooibos samples taken from 24 different plantations: the aspalathin and nothofagin content, the two compounds that most define the tea's character, varied substantially not just between farms but between bushes within the same field, a spread the paper attributes directly to seed-based propagation. Plant a hundred seeds from the "same" variety and you get a hundred subtly different plants, some more productive, some less, and no clean way to fix the good ones in place the way a grafted orchard does. Rooibos's genetic diversity problem has a deeper layer too: the whole cultivated crop traces to one wild population out of at least four distinct types, and a 2014 climate-modelling study in Ecology and Evolution warns that expanding cultivation into new ground risks further narrowing that base by cross-pollinating with, and homogenising, the wild stock nearby. What that same modelling means for rooibos's own protected geography is its own story, and a reason a hardier bred line matters for more than just die-back.

What an unfixed variety costs a farmer

None of this is academic to the person growing the crop. Earlier rooibos plantations ran a commercial lifespan of seven years or longer, the 2008 review states, but it says plainly that die-back disease now "severely limits" that lifespan. (The fungus behind most of that die-back has its own story, and a working plantation today runs closer to five years.) A grower cannot know, from one seed packet, how the resulting field will perform against drought, root rot, or a bad fire year, because no one has ever bred and fixed a line specifically selected for that. The 2008 review's own list of future research needs states the gap plainly: "improved plant material through selection and breeding, with regards to productive lifespan, composition, and bioactivity, is required." That sentence was written as a call for something that did not yet exist.

The sister crop that got the program rooibos never did

The contrast that makes this genuinely surprising sits one tea over. Honeybush (Cyclopia species), rooibos's much smaller fynbos cousin, got exactly the treatment rooibos did not. Starting in 1992, a Kirstenbosch botanist named J.H. "Hannes" de Lange, working for what is now the South African National Biodiversity Institute, planted more than 200,000 honeybush seedlings across a dozen species at 64 sites on the West Coast and in the Boland, the founding legwork of turning a wild-harvested plant into a farmed one. Within a decade the Agricultural Research Council had formalised what de Lange started into the ARC Honeybush Breeding and Selection Programme, co-founded by de Lange and P. Botma, explicitly aimed at improving biomass yield and, unusually for either crop, running deliberate cross-breeding between species. A 2016 paper by the ARC's Cecilia Bester and colleagues, titled plainly "A breeding strategy for South African indigenous herbal teas," set that program down on paper.

Rooibos got no equivalent. A 2022 Stellenbosch University dissertation on both crops' production systems notes, almost in passing, that despite the rooibos and honeybush sectors' heavy dependence on plant genetic resources, neither has a formal seed genebank in place, and that the ARC's own "Fynbos genebank" at Elsenburg, despite its name, banks cutflower species only. Calls from researchers for a proper rooibos and honeybush genebank have been on record for years without one materialising. A separate 2023 soil-science study out of Stellenbosch, working in a young rooibos plantation, notes that research attention on the crop has gone disproportionately toward its health chemistry rather than the agronomy of growing it well, which is as good an explanation as any for why the breeding side lagged so far behind the marketing.

The same name, a second crop, decades later

Here is the detail that makes the gap worth noticing rather than just lamenting. Cape Natural Tea Products, a Cederberg rooibos processor, now runs a rooibos seed-improvement project of its own, and names its lead as "experienced botanist Dr Hannes De Lange," working with the company's agricultural division. By the company's own account, seed from wild rooibos populations across the growing region was collected and planted, and the resulting plants have now matured enough to move into a selection phase: researchers assess each one on simple, visible traits, counting flowers as a proxy for seed-producing capacity and leaves as a proxy for yield, before the best performers go toward producing seed that ordinary growers can plant. It is, point for point, the same method de Lange used to turn wild honeybush into a farmed crop three decades earlier: gather wild material with good traits, grow it out, select on what you can actually observe, then multiply the winners.

Whether this is literally the same de Lange the honeybush record names is not something Cape Natural Tea Products' own materials confirm outright. What can be said plainly is that the name, the specialty, and the method are the same. What it is not, yet, is a registered cultivar in the formal sense a wine grape or an apple variety gets. South Africa's plant breeders' rights law lets a new, distinct, uniform, and stable variety of a listed crop be registered and named, the same protection that covers an apple cultivar like Cripps Pink. No registered rooibos denomination turned up anywhere in the trade press or industry material the Ministry could find, though only the official plant breeders' rights journal itself could confirm that for certain. What is actually moving from trial toward a farmer's field right now, by every account available, is improved bulk seed, not a certified clone, exactly the low-tech mechanism that has carried nearly the entire history of this crop.

How a new line earns a farmer's trust

There is no marketing rollout waiting at the end of this. A rooibos grower does not switch seed sources because a company issues a press release; growers watch what a neighbour's field is doing three and five years in, whether it stayed ahead of die-back, whether it kept yielding through a dry run of winters, and they buy seed from whoever's fields keep answering those questions well. That is, as far as the record shows, close to how Nortier's own selection actually spread through the Cederberg in the first place: farm to farm, not by decree.

None of this will change what is in your cup next year. But it is the reason there should still be a cup at all in twenty years, grown in a warming, drying Cederberg from fields that, until very recently, nobody was trying to make any hardier than the ones Nortier happened to find in the 1930s. Ninety years on, rooibos is finally getting a second candidate to judge the same patient way.