Cannabis conservation

Part of a series on

Conservation

Seed kept in a bottle by a farmer in Uva Province, Sri Lanka. Farmers who save seed hold populations outside any collection.

Cannabis conservation·Grow-out priorities·Populations

Threat Categories

Extinction

Genetic Threats

Genetic erosion·Genetic contamination·Introgression·Gene flow·Hybridisation·Population bottleneck·Genetic drift·Inbreeding depression

Habitat & Environmental Threats

Habitat loss

Eradication Tracking & Analysis

Eradication·Herbicide spraying of cannabis·Herbicide spraying in Mexico·Herbicide spraying in the United States·Herbicide spraying in Colombia·Herbicide spraying in Belize·Herbicide spraying in South Africa·US eradication abroad·Yukthiya·Cannabis eradication in Kirivong·Cannabis eradication in Fiji·Campaigns by country

International Enforcement

Cannabis prohibition·US eradication abroad·INCB

Legal & Policy

Farmers' rights·Access and benefit-sharing·Intellectual property·Single Convention on Narcotic Drugs·1988 Convention·Cannabis legalisation movement

Population Assessment

Minimum viable population·Population viability analysis·Effective population size

Conservation Monitoring

Population monitoring·Genetic monitoring

Conservation Methods - Ex Situ

Ex situ conservation·Gene bank·Cryopreservation·In vitro conservation·Core collection·Seed storage

Conservation Methods - In Situ

In situ conservation·On-farm conservation·Genetic reserve·Community seed bank·Feral and ruderal populations

Genetic Conservation

Germplasm·Genetic resources·Crop wild relative·Conservation genetics

Sampling & Collection

Germplasm collecting·Germplasm sampling

Preservation Strategies

Conservation prioritisation·Complementary conservation

Recovery & Restoration

Reintroduction·Genetic rescue

Conservation Science

Landrace conservation

Portal·Category

Cannabis conservation is the keeping of the genetic variation of Cannabis in living populations and in stored samples. The plants concerned are landraces, feral and wild-growing stands, older hemp cultivars and clonal drug-type lines. Holding samples of germplasm in collections is ex situ conservation.^: 163 Keeping a population going where it grows is in situ conservation.^{: art. 2}

Several reviews report heavy losses and call for action. John McPartland and Ernest Small wrote in 2020 that four taxa of Central and South Asian cannabis are "critically endangered reservoirs of germplasm" and "in urgent need of conservation". A 2016 review by Matthew Welling and colleagues concluded that the international drug conventions had been influential in the "fractionation and erosion" of publicly accessible collections. In the same year, in the words of Robert Clarke and Mark Merlin, "the vast majority of landraces may already be extinct".^: 318 Other authors frame the same plants differently. Canavan and colleagues treat feral cannabis as a weed and possible invader, and Pierre-Arnaud Chouvy warns against fixing a landrace in place through strict conservation.

Some of the reported losses are recorded events: seed collections gathered for Ernest Small's studies were destroyed in 1980 under Canadian government policy, and an American accession received in 1900 no longer exists. Statements about landraces in the field, such as Clarke and Merlin's, are judgements given without a count. The genetic surveys of Moroccan and Iranian populations compare populations with one another, which gives a level of diversity and no trend.

What is being conserved

Feral hemp in a roadside ditch in Buffalo County, Nebraska, in 2017. Authors differ on whether such stands are a genetic resource, a weed or both.

McPartland and Small divide C. sativa subsp. indica into four varieties: two domesticates, var. indica from South Asia and var. afghanica from Central Asia, and two wild-growing relatives, var. himalayensis and var. asperrima. The European subspecies lies outside their study. In their account breeders have crossed the two domesticates since the 1970s, and the plants now sold as "Sativa" and "Indica" have lost the old differences, so they ask for collecting in the landraces' centres of diversity.

Whether any of these plants are wild in the strict sense is unsettled. McPartland and Small allow that truly aboriginal populations may not have survived, and few if any intact wild populations are thought to exist in the review by Welling and colleagues. From whole genomes, Guangpeng Ren and colleagues concluded that the feral plants they sampled were historical escapes from cultivation and suggested that pure wild progenitors have gone extinct. Divergent regions in the genomes of Asian hemp, by contrast, led Ryan Lynch and colleagues to write of "undiscovered wild relatives".

Other authors give other priorities. East Asian germplasm comes first for Welling and colleagues, for its long cultivation across many latitudes, its apparent diversity and its small share of published collections. Clarke and Merlin knew of no North American hemp cultivar still alive, "only their feral offspring",^: 304 and called drug cultivars "sorely under-represented" in seed banks.^: 318 Midwestern feral populations "should be a top priority for conservation efforts" in the view of Tori Ford and colleagues. Writing on the risk of invasion, Canavan and colleagues advise that volunteer plants near fields be removed as soon as they are detected.

For Welling and colleagues the reason is breeding: landraces and wild relatives are a repository from which alleles can be brought back into domesticated lines. For Ademola Aina and colleagues, American feral plants are new tools for breeding improved cultivars. McPartland and Small add that formally named varieties may help to prevent extinction and provide prior art against patents. Chouvy argues from a different ground: protecting a landrace through its place of origin protects the farming system and rural economy around it. He holds that landraces go on changing and should not be "musealized".

Reports of loss

Main article: Genetic erosion

The Food and Agriculture Organization defines genetic erosion as "the loss of individual genes and the loss of particular combinations of genes" and notes a wider use of the term for the loss of varieties. Reports on cannabis cover both senses.

Reports of loss in cannabis, with what each report stands on

Report	What it stands on	Source
Hundreds of seed collections gathered for Small's studies were destroyed in 1980	A recorded event, told by one of those involved
The American accession PI 5021 "no longer exists and can not be requested"	The holder's own database entry
"Considerable loss of genetic diversity and purity" in the collection of the Vavilov Institute	A review citing a genetic study by Hillig (2004)	: 318
"The majority of traditional fiber, seed, and drug landraces are no longer grown"	The authors' judgement, with no count	: 317
Central and South Asian landraces "face extinction through introgressive hybridization"	Published reports of foreign seed reaching their centres of diversity, and a comparison of THC to CBD ratios across 47 studies, which sets different samples of different dates side by side
The kif landrace has largely given way to hybrids in the Rif Mountains	Field observation and a dozen interviews in July 2013, in the authors' words "only a guesstimate"
The local Moroccan population needs protection from "potential genetic erosion"	Six microsatellite markers in two populations: more diversity in the local one than in the introduced cultivar Critical, with gene flow between them
Decades of eradication "have exterminated many of the naturalized populations in North America"	The authors' statement, with no count; feral populations are still present
A broad loss of suitable area in Eurasia and the United States by 2050	A climate model that samples no genes

Elsewhere in crop science, erosion has been measured by genotyping collections made at two dates, as with maize landraces in central Italy and barley landraces in Sardinia. None of the reports in the table follows one cannabis population in that way. The Food and Agriculture Organization judged in 2010 that "general statements purporting to quantify the overall amount of genetic erosion" over the previous decade were "not warranted".

Crossing with introduced plants

Main article: Genetic contamination

Cannabis is pollinated by wind and all its populations can cross. In McPartland and Small's account, foreign seed reached Afghanistan, Nepal, Jamaica, Thailand, Kashmir, South Africa, Morocco and Kullu between about 1972 and 1988. Central Asian genes raise yields in hybrids, bring earlier maturity at high latitudes and shorten the plants, which are then easier to conceal outdoors and to grow indoors. Clarke and Merlin wrote that the spread of broad-leaflet genes into narrow-leaflet regions "has nearly brought an end to pure NLD landraces".^: 320 They also expected that clones and rules protecting breeders would narrow the gene pool to the most commercially viable cultivars.^: 317

Crossing can end a population's pure genome while its alleles and traits live on in the hybrids. In the Rif, Fatima Bachir and colleagues describe the local varieties as genetically contaminated and rapidly disappearing. Chouvy accepts that foreign landraces probably changed kif and holds that it remains distinct. Accessions regrown near one another in a collection can cross in the same way.

Eradication

Main article: Eradication

The Single Convention on Narcotic Drugs of 1961 requires a party that prohibits cultivation to destroy illicitly grown plants, "except for small quantities required by the Party for scientific or research purposes".^{: art. 22(2)} Eradication campaigns have reached crops and also wild and feral stands, as in the Issyk-Kul province of Kyrgyzstan. Their results are counts of plants, hectares or tonnes destroyed.

Clarke and Merlin quote a list by Watson and Clarke in which "suppression and eradication" stand among the causes of lost landraces.^: 318 Chouvy and Kenza Afsahi wrote that it is "not clear what role forced eradication might have played" in the Rif's turn to hybrids, since it mostly took place outside the historic growing zone. The American programme claimed more than 1.9 billion feral plants destroyed between 1998 and 2006, on Canavan and colleagues' sum of its reports, and feral populations remain. Neither Chouvy and Afsahi nor Canavan and colleagues measure diversity before and after a campaign.

Who holds the material

Main article: Gene bank

Public gene banks that hold cannabis include the hemp collection of the National Plant Germplasm System at Geneva, New York, the Vavilov Institute in Saint Petersburg, INRA Maroc in Rabat, the Suceava Genebank in Romania and the Leibniz Institute of Plant Genetics and Crop Plant Research in Germany. Aina and colleagues wrote in 2025 that only a handful of the world's repositories have documented deliberate conservation of cannabis, and that federal prohibition kept hemp out of the American system until the Geneva collection began in 2021. In November 2026 most of Geneva's accessions were marked not available. Clarke and Merlin reported 43 cannabis accessions at the Svalbard Global Seed Vault in 2016, duplicated in three other seed banks. Given impure samples and low seed numbers, they wrote that "there really is no reliable reserve of Cannabis seeds".^: 318

Outside the public collections, the University of Mississippi keeps indoor-grown living plants. Company collections in 2016 included Ecofibre's in Australia and HortaPharm's in Amsterdam. Jason Sawler and colleagues described "only a patchwork of private collections of marijuana strains worldwide". In the cannabis trade a "seed bank" is a business selling seed, which is not a gene bank in the sense of the Food and Agriculture Organization's standards.^: 2–3 Farmers who save their own seed hold populations with no collection behind them: in the Tai'an district of Shandong in 1995, farmers kept back seed of the local landrace each year.

Legal limits

Main article: Genetic resources

The Convention on Biological Diversity defines genetic resources as "genetic material of actual or potential value" and leaves access to them to national governments.^{: arts. 2, 15} The plant treaty of 2001 leaves farmers' rights to national governments. Its multilateral system of access and benefit-sharing covers the crops in its Annex I, and cannabis is not one of them.^{: 12, 46–49}

The Single Convention defines cannabis as the flowering or fruiting tops of the plant, which leaves out seed. It does not apply to growing the plant exclusively for fibre, seed or horticulture.^{: arts. 1, 28(2)} Welling and colleagues still counted limits on moving seed between parties to the 1961 convention and the Convention on Psychotropic Substances of 1971 among the barriers to a shared physical collection. National law can go further: the federal definition of marijuana in the United States takes in the seed, with exceptions for hemp seed and sterilised seed. The Geneva collection sends seed for research but not for home gardening.

For intellectual property, Chouvy argues that plant variety protection does not suit landraces and that an appellation of origin is the only form suited to them, though the illegality of the product stands in the way for now.

Stored samples

Main article: Ex situ conservation

Seed samples in moisture-proof foil bags in the cold vault at Fort Collins, Colorado (2013), where the Geneva hemp collection planned its backups.

Cannabis seed appears to tolerate drying and cold, and for storage of more than three years it is commonly held at −20 °C. Regrowing is harder. Plots may need up to 5 km between them to avoid crossing, accessions from different latitudes need different day lengths to flower and indoor rooms hold few plants, so that alleles may be lost at each cycle. Clarke and Merlin put the need at 1,000 plants for a monoecious accession and 2,000 for a dioecious one to reproduce 99 per cent of its alleles, and wrote that many accessions hold fewer than 1,000 viable seeds.^: 318

Clonal lines are kept as living plants, and frozen shoot tips have regrown in trials at the University of Mississippi. Welling and colleagues proposed a "virtual" core collection of up to 400 accessions, left in the existing gene banks and described in one shared database, as a way round the limits on moving seed.

In the field

Main article: In situ conservation

McPartland and Small judged in situ conservation "much preferable" for crops and their wild relatives in general and called seed banks "an immediate priority" for cannabis. The reviews by Welling and colleagues, Clarke and Merlin and McPartland and Small describe no genetic reserve or on-farm project set up for the plant, which does not show that no local effort exists.

Farmers who go on growing a landrace keep it with no outside programme, a process Stephen Brush called de facto conservation. Chouvy credits Berber tribes in the Rif with making the conservation of kif possible despite the drug laws and state repression. Clarke and Merlin called the spread of introduced genes a "degradation of in situ conservation".^: 320 In the terms of Nigel Maxted and colleagues, a taxon in a protected area with no plan for it is conserved only passively. Police in Sri Lanka reported destroying 150,000 plants inside Yala National Park in February 2021. Mauricio Bellon and colleagues treat storage and field conservation as complementary, since stored seed can restock a lost population and gives a past sample to set beside the living one.

Known gaps

A field crop of cannabis collected by P. H. Davis near Ketama in 1970. Herbarium sheets cannot be regrown, but they can be sampled.

Public passport and provenance data for cannabis accessions are absent or of low quality, which in Welling and colleagues' view rules out the mapping methods other crops use to choose where to collect next. The long-term viability of stored seed had not, in their review, been shown systematically.

Global diversity remains undersampled, with Asian material notably underrepresented, in the genome study by Lynch and colleagues. A good representation of unhybridised plants is not available for analysis, McPartland and Small wrote, herbarium collections from Asia are limited and curators have often been unable to allow sampling. The plant's legal status and economic importance make wild-growing populations hard to sample, according to Manica Balant and colleagues.

Bachir and colleagues suggest sampling older populations, above all from herbaria, to work out the history of Moroccan cannabis. For on-farm work in general, Bellon and colleagues found little evidence that interventions did more than farmers were already doing, and none yet of effects on genetic diversity. Their cases come from other crops.

See also

• Conservation genetics
• Effective population size
• Feral and ruderal Cannabis
• Gene flow
• Genetic monitoring
• Germplasm
• Introgression
• Landrace Cannabis
• Landrace conservation