Where the starting material comes from
A breeding programme needs variation to select from, and it gets it in one of three ways. It selects out of an existing seed-grown population, which is how most of the earliest cultivars everywhere arose — Longjing 43 out of the Longjing population, the Wuyi ming cong off the cliffs, the first Assam and Ceylon clones out of estate seedling fields. It makes a deliberate cross between two chosen parents and raises the seedlings. Or it introduces material from elsewhere and selects within it, which is how assamica genetics reached Japan and Burmese material reached Taiwan. All three depend on the same underlying resource: a varied population somebody kept. This is the practical reason the disappearance of seed-grown stands is treated as a problem by breeders rather than only by traditionalists. A programme with nothing but its own previous releases to work with is recombining a narrowing set of material, and the traits it will need in forty years are, by definition, ones nobody is currently selecting for.
What a breeder is actually looking at
The candidate plants are grown out and assessed over years, and the criteria are unromantic. Yield, first: how much harvestable shoot per plant per season. Budding date, because a few days early or late is worth money in a seasonal market. Vigour and rooting ability, since a plant that will not root from cuttings cannot be distributed. Hardiness against the local limiting factor, whether that is frost, drought or heat. Disease and pest resistance. And only then quality, assessed by making tea from the candidate and cupping it, which is slow, subjective and cannot be done at all until the plant is old enough to yield enough leaf to process. That ordering explains a lot about the world’s planted area. Cultivars are selected predominantly for agronomic performance, because that is what can be measured early and repeatedly, and the aroma-led cultivars that speciality drinkers care about are a small and comparatively recent branch of the enterprise.
Clonal trials and multi-site testing
A promising individual is multiplied by cutting and planted out in replicated trials, usually at several sites with different climates and soils. This stage is where most candidates fail, and it is where the time goes. A tea plant has to be established, trained and brought into bearing before it can be evaluated, so each round of assessment is years rather than seasons, and a single site is not enough because a plant that excels in one district may be mediocre in the next. Sri Lanka’s practice of recommending particular TRI clones for particular elevation bands is the visible output of exactly this testing. Two consequences follow. Programmes release in batches, because the trial infrastructure is shared and the candidates move through it together — which is why cultivar names so often come with a neighbouring number, LDP1 and LDP2, Fuyun 6 and Fuyun 7. And the release recommendation is site-specific, so planting a cultivar outside the conditions it was tested in is doing something the programme did not evaluate.
Release, registration and naming
A surviving candidate is released, which in most producing countries means a formal registration with a national or provincial authority, a designated name or number, and a published description. The naming conventions differ sharply and are covered separately, but the underlying act is the same: the plant acquires a public identity, and mother-bush material is made available so that nurseries can propagate it. Japan gives poetic names, Taiwan and Sri Lanka give numbers, Kenya gives selection codes, China gives descriptive names plus a national or provincial approval status. Registration is what makes a cultivar name meaningful rather than local, and it is also the point at which a plant stops being the programme’s and becomes the industry’s. Everything that happens afterwards — propagation drift, misapplication of the name, plantings that never get recorded — happens outside the institution that made it, which is why a cultivar name on a smallholding several nursery generations downstream is a weaker claim than the same name in a research station’s field book.
Adoption, which is the stage that usually fails
A released cultivar reaches fields only if growers plant it, and growers replant slowly for reasons that have nothing to do with the plant: replanting costs several years of income, capital is scarce, and an existing field is still producing something. So the gap between a national cultivar register and a national planted area is enormous everywhere. Japan has released dozens of cultivars and remains dominated by one. Taiwan has a numbered series running well past twenty and its planted area is dominated by a plant selected long before the series began. Sri Lanka, Kenya and Indonesia have all had substantial extension programmes and all report ageing planted areas. The lesson for reading a cultivar catalogue is that a register describes what is available, not what is grown. A cultivar can be well regarded, extensively trialled and formally released, and be present in the field in quantities that round to nothing.
Changing a field without replanting it
There is a partial route around the replanting cost, which is to graft new material onto an established root system rather than starting again — cutting an existing bush back and grafting scions of the desired cultivar onto the stump. The attraction is obvious: an established root system means a much faster return to bearing than a nursery plant would give. The practice is used in Chinese and East Asian tea cultivation to convert fields to a more marketable cultivar, and TeaHQ records it as a known technique rather than quantifying how widespread it is, which varies by country and has not been verified here. It matters to a reader for one specific reason. A grafted field is a field whose roots are one plant and whose harvested shoots are another, which makes any story linking the age of the roots to the character of the tea considerably more complicated than it sounds — and it is one more way a cultivar claim can be true of the leaf and misleading about the field.
How the targets have shifted
TeaHQ's own reading of well-attested general knowledge, with no single source behind it.
The trait priorities of tea breeding have moved with the industries that funded it. The mid-twentieth-century programmes were built around yield and uniformity for an expanding commodity market. Later work in several countries added disease resistance, drought tolerance and mechanisation suitability as labour became scarce and expensive. More recently a small strand has selected explicitly for aroma and for distinctiveness — the Japanese fragrance cultivars, the Taiwanese and Fujian aroma crosses — which is a response to a speciality market that did not exist when the programmes started. There is a slower change underneath all of it. A cultivar planted now will still be in the ground when the climate it was selected under has moved, so breeding for tolerance of variability rather than for peak performance in average conditions is increasingly what programmes say they are doing. Whether the plants released reflect that yet is not something TeaHQ can assess, and it would take decades to find out.