What Is a Landrace? African Cannabis, Evidence and Provenance
A landrace is a historically developed, locally maintained population of a crop, not an ancient pure strain, and for African cannabis the label is only as strong as the provenance and evidence behind it.
Why the question matters
Calling a cannabis population an African landrace makes a claim about history, and a claim about history needs evidence that a name, a photograph or a reputation cannot supply. In agricultural science a landrace is a historically developed population shaped by local cultivation and selection. In cannabis culture the word is often used as though it named a clear botanical category, and plants sold as African landraces are usually tied to a country, a look, a cultivation tradition or a reputation for certain effects. Those ties may be historically meaningful. They do not, on their own, show that a population has the genetic or agricultural history attributed to it.
The question also carries commercial and ethical weight. Duvall (2023) notes that profitable cannabis industries in Europe and North America rely on seed taken from Africa, where genetic diversity is high largely because of farmers’ own plant-breeding skills, and Duvall (2019a) argues that the crop’s history has let actors in the Global North extract more value from African resources than African farmers can. A provenance claim is therefore also a claim about whose knowledge and material a product rests on.
This article asks two things: what a landrace is, as agricultural science has defined it, and what historical, ethnobotanical and genetic research can and cannot establish about cannabis populations in Africa. It does not assume that every regional name marks a distinct lineage, or that every locally grown plant belongs to a historically continuous population. The value of a landrace lies in the processes that produced it and in the quality of the evidence available to reconstruct them, not in an exotic name or an unbroken line back to an imagined origin.
What a landrace is
There is no single definition of landrace accepted across agricultural research, and the disagreement is not only about wording: landraces differ with the biology of the crop, how it is propagated, the practices of the farmers who keep it and the environment it grows in.
Three reference points frame the field. Zeven (1998) reviewed how the term had been used since the first references to landraces as genetic resources around 1890. He separated autochthonous landraces, grown for a long time in the farming system concerned, from allochthonous ones brought in from elsewhere, and he stressed that landraces keep evolving and keep mixing with other landraces and cultivars. Villa et al. (2005), working from a literature review, a workshop and interviews with key informants, described a landrace as a dynamic population of a cultivated plant with a historical origin and a distinct identity, lacking formal crop improvement, and often genetically diverse, locally adapted and tied to traditional farming systems. Casañas et al. (2017), in a Perspective article, argued that the concept should not exclude evolution under modern breeding: a landrace may keep evolving “using conventional or modern breeding techniques”, in traditional or new environments, and the most conservative definitions would turn landraces into museum pieces.
A workable definition for cannabis follows from these sources. A landrace is a historically developed population of a cultivated plant, with a recognizable identity, maintained and shaped by cultivation in a particular geographical and cultural setting. Such a population often holds substantial genetic variation and may be adapted to local conditions or practices. These are lines of evidence to investigate, not a checklist in which every box must be ticked.
The lines of evidence are related but separate, and no single one identifies a landrace:
Feature
What it suggests
What it does not prove
Local name
A recognized agricultural identity
A genetically uniform or distinct population
Distinctive appearance
Possible shared ancestry or selection
Historical continuity; appearance also shifts with environment
Long local cultivation
A history worth documenting
Local adaptation, or isolation from other populations
Genetic differentiation
The population differs from those sampled
How it arose, or who maintained it
Local adaptation
Selection under local conditions
That the population stayed in one place
The term therefore describes a historical and agricultural relationship, not simply a genetic result. Genetic evidence can show whether a population is differentiated, how much variation it holds and how it relates to other sampled material. It cannot establish how farmers maintained it, what its local name means or how long it has been grown there.
A population, not a pure strain
The belief that a landrace is an ancient, unchanged and genetically pure strain is the most persistent misunderstanding in commercial cannabis culture, and it contradicts both the definitions above and the biology of the plant.
A landrace is a population, and its members can differ in flowering time, height, branching, leaf shape and chemistry. That variation is usually part of the population’s identity, not a flaw in it. Cannabis sharpens the point: it is normally dioecious, with separate male and female plants (a few cultivars are monoecious), and its pollen travels on the wind, so seed-grown populations stay heterogeneous and are hard to keep isolated. Seed exchange between farmers, introductions from elsewhere and deliberate crossing add further movement of genes.
Four processes explain what happens in such populations. Gene flow moves genes between populations through seed or pollen. Admixture is the resulting mix of ancestries. Selection shifts the frequency of inherited traits, and genetic drift changes variation by chance, most strongly in small populations. None of them alone means a landrace has lost its identity: a population can receive outside genes and keep a recognizable local history, and it can keep its traditional name and practices while changing genetically. Cultural continuity and genetic continuity have to be investigated separately.
For the same reason “purity” is the wrong question. The useful ones concern genetic continuity, population structure, gene flow and selection. Casañas et al. (2017) make a related point about conservation: protecting landraces as frozen original genotypes ignores the processes by which farmers keep adapting them, though this does not make every modern hybrid a landrace. The distinction is between ongoing evolution inside a historically recognizable population and the replacement of that population by unrelated material.
Landrace, cultivar, strain. A cultivar is a cultivated variety selected and maintained for particular characteristics; depending on the crop and breeding method it may be uniform or diverse. A commercial cannabis “strain” may be a clone, a seed line, a hybrid or only a market category. A landrace can feed into a cultivar, and a cultivar can be derived from a local population, so the terms describe different aspects of agricultural identity rather than exclusive kinds of plant.
Wild, feral, landrace. A wild population grows without cultivation; a feral population consists of escapes from cultivation. In cannabis the boundaries blur, because plants growing spontaneously often descend from cultivated material and repeated introductions obscure the source. Balant et al. (2025) note that modern cultivars are now being reintroduced into native areas and mixing with local landraces and wild-growing populations. A plant growing spontaneously in a region should not be called indigenous to it, and a cultivated population should not be called a landrace without evidence of how it arose and was maintained.
Keeping landraces alive is itself a source of change. Regenerating seed from a small number of plants narrows diversity and invites drift and accidental crossing, which is why Welling et al. (2016) suggest that seed cryopreservation could reduce the hybridization and drift that occur when germplasm is regenerated.