Cannabis taxonomy
More actions
Cannabis taxonomy is the classification of plants in the genus Cannabis within the family Cannabaceae. The taxonomy of the genus has been disputed since the eighteenth century and remains unresolved. Three principal treatments coexist in the modern literature: a monotypic concept in which the genus contains a single polymorphic species, Cannabis sativa L.; a two-species concept in which Cannabis indica Lam. is treated as a separate species; and a three-species concept in which a further species, Cannabis ruderalis Janisch., is also recognised.[1][2]
The dispute is further compounded by a vernacular nomenclature, drawn from the late-twentieth-century drug trade, in which the labels "Sativa" and "Indica" came to be applied to plants in ways that do not correspond to the formal scientific epithets of the same names. McPartland and Guy described this divergence in 2017, after a survey of herbarium specimens and primary descriptions, and showed that the vernacular and formal taxonomies refer to substantially different sets of plants.[3]
Forty years of intentional hybridisation between drug-type populations from different geographic origins, combined with prohibition-era barriers to systematic study, has left the taxonomy of the modern commercial accession landscape only loosely resolved. Conservation taxonomy of pre-hybridisation source populations is a distinct and active line of work, most fully developed in the four-variety treatment of landrace populations proposed by McPartland and Small in 2020.[4]
History of Cannabis classification
Pre-Linnaean and early Linnaean treatments
Before 1753, Cannabis was classified with various unrelated plants on the basis of leaf shape and human use; eighteenth-century botanists shifted the genus between the families now called Urticaceae, Moraceae and Cannabaceae. The history of these family-level placements is treated in the article on Cannabaceae.[1]
Linnaeus, Lamarck and polytypicism
The starting point for the modern nomenclature is the publication of Linnaeus's Species Plantarum in 1753, in which he described a single species, Cannabis sativa L., based on European cultivated hemp.[1] Thirty-two years later, Lamarck described a second species, Cannabis indica Lam., based on plants from India that he distinguished from European hemp by eight morphological characters together with a strong odour and a psychoactive effect.[3]
Through the nineteenth century, field botanists applied the names C. sativa and C. indica inconsistently. A herbarium survey by McPartland and Guy mapped 18th- and 19th-century specimens against the floristic regions they came from and found no clean geographic separation: Linnaean-school botanists from Scandinavia and Britain tended to apply C. sativa regardless of provenance, while Lamarckian-school botanists from France and Francophone Russia tended to apply C. indica or to coin new species names. The same plants in the same regions were given different names by different observers, producing a record in which the formal nomenclature reflected the affiliations of the field botanist as much as the plant.[3]
Janischewsky and the three-species concept
A third species, Cannabis ruderalis Janisch., was described by Janischewsky in 1924 from wild-growing plants of the lower Volga basin in southeastern European Russia, distinguished by short stature, autoflowering habit and a small thick-walled achene.[1] Vavilov and Bukinich accepted the three-species treatment in their 1929 work on Afghan agriculture, and the Soviet tradition continued to recognise three species through the mid-twentieth century.[3]
Mid-twentieth-century rationalisation
Two influential treatments published two years apart shaped the taxonomy that dominated the late-twentieth-century literature. Schultes and colleagues argued in 1974, in a paper titled "Cannabis: an example of taxonomic neglect", for retaining three species on grounds of consistent morphological, chemical and geographic differentiation.[1] Small and Cronquist proposed the opposite in 1976: a single polymorphic species, Cannabis sativa L., with two subspecies (subsp. sativa and subsp. indica) each containing both cultivated and wild varieties, on the grounds of full interfertility and continuous morphological variation.[5] Small's monotypic treatment was widely adopted in agronomy, regulatory frameworks and the drug-policy literature through the 1980s and 1990s, and was reaffirmed by Small in a 2015 review.[6]
Taxonomic position
Cannabis is a genus of annual herbaceous plants in the family Cannabaceae, order Rosales. Its sister genus is Humulus (hops); the two are nested within a clade of otherwise woody Cannabaceae genera. Molecular clock analyses place the Cannabis–Humulus divergence at approximately 27.8 million years ago, with a centre of origin inferred from fossil pollen on the northeastern Tibetan Plateau.[1] Family- and genus-level history are treated in the Cannabaceae article.
Species concepts
Monotypic concept
The monotypic concept treats Cannabis as a single polymorphic species, Cannabis sativa L. Most contemporary genomic work resolves the genus as one species on grounds of full interfertility, broadly continuous morphological variation across populations and the absence of stable genetic discontinuities at most loci.[6][2] Within this concept, the most-developed treatment is two subspecies: subsp. sativa (hemp-type, low THC, mostly broad-leaflet European cultivars and their derivatives) and subsp. indica (drug-type, including all populations selected for high cannabinoid content).[5][4]
Polytypic revival
A polytypic revival in the early 2000s reopened the species question on chemotaxonomic and allozyme grounds. Hillig and Mahlberg analysed cannabinoid variation in 157 accessions of known geographic origin in 2004 and showed that the dry-weight ratio of tetrahydrocannabinol to cannabidiol falls into three discrete chemotypes governed by alleles at the gene encoding the cannabinoid synthases; the chemotype distribution divided the accessions into two principal gene pools corresponding to C. sativa and C. indica.[7] Hillig extended the analysis to allozyme variation at 17 loci in 2005 and proposed a polytypic treatment recognising three species and seven putative taxa.[8]
The Hillig treatment has continued to be cited in chemotaxonomic and breeding-focused work, but has not displaced the monotypic concept in regulatory or genomic contexts.
Modern molecular framing
Genome-wide molecular work has not converged on a single classification. Sawler and colleagues analysed 14,031 single-nucleotide polymorphisms in 81 marijuana and 43 hemp accessions in 2015 and found significant genome-wide differentiation between the two groups, consistent with either two species or two subspecies but not resolving the rank question.[9] Reimann-Philipp and colleagues reviewed the contemporary literature in 2024 and identified three frameworks in current scientific use: monotypic with subspecies, polytypic with up to three species and a single phenotypically diverse species without formal subdivision; none had achieved consensus.[2]
Lynch and colleagues published a wild Cannabis pangenome in 2025 and showed that the loci controlling cannabinoid synthase isoforms are themselves the product of recent domestication imposed on an older wild mosaic, complicating any classification that draws species boundaries from cannabinoid chemistry alone.[10] Cyto-nuclear discordance and pervasive hybridisation in post-1970s drug-type accessions have made strict species circumscription progressively more difficult in the molecular record.
Subspecies, varieties and conservation taxonomy
The most-developed contemporary subspecific treatment was proposed by McPartland and Small in 2020 specifically to give formal names to wild and landrace drug-type populations endangered by displacement and hybridisation.[4] The treatment sits within the monotypic species concept and recognises Cannabis sativa subsp. indica (the drug-type subspecies) with four varieties:
- var. indica: narrow-leaflet drug-type populations from South Asia, the source of the plants known in the drug trade as "Sativa".
- var. afghanica: broad-leaflet drug-type populations from Central Asia, the source of the plants known in the drug trade as "Indica".
- var. himalayensis: narrow-leaflet wild-growing relatives of var. indica from the South Asian Himalayan foothills.
- var. asperrima: broad-leaflet wild-growing relatives of var. afghanica from Central Asia and the Hindu Kush.
The treatment is the most-developed conservation-oriented taxonomy in the recent literature, but is not universally adopted; many breeders, regulators and authors continue to use the Hillig three-species framework, the unmodified Small 1976 two-subspecies treatment or vernacular labels.[3][2]
Vernacular taxonomy
The vernacular labels "Sativa" and "Indica" in widespread use among consumers, breeders and dispensaries do not correspond to the formal scientific epithets of the same names. McPartland and Guy traced the vernacular usage to drug-trade and breeder literature of the 1970s and 1980s, in which the term "Indica" came to denote short, broad-leaflet, fast-flowering plants of Central Asian origin, and "Sativa" to denote tall, narrow-leaflet, late-flowering plants of South Asian origin.[3]
In the formal McPartland and Small 2020 treatment, these vernacular categories map onto formal taxa as follows: vernacular "Indica" corresponds to C. sativa subsp. indica var. afghanica; vernacular "Sativa" corresponds to C. sativa subsp. indica var. indica.[4] Both vernacular groups therefore sit within a single formal subspecies, and neither corresponds to Cannabis sativa in the formal sense (which refers to the hemp-type subspecies). The mismatch has persisted in commercial and popular usage and is treated in detail in the article on Sativa vs Indica.
Morphological classification
Anderson introduced the morphological categories narrow-leaflet drug (NLD) and broad-leaflet drug (BLD) in 1980, on the basis of leaflet shape, plant habit and flowering time in cultivated drug-type accessions.[3] Hillig formalised the framework in 2005 by mapping the same categories onto allozyme gene pools.[8] NLD and BLD biotypes correspond closely to C. sativa subsp. indica var. indica and var. afghanica respectively under the McPartland and Small 2020 treatment, and to vernacular "Sativa" and "Indica" respectively under common usage.
NLD/BLD is morphology-and-region-based rather than nomenclature-rank-based, and as a working classification for landrace and wild-derived populations it is in wider use among breeders and conservationists than the formal trinomials. The framework is treated in detail in NLD/BLD classification.
Molecular phylogenetics
Whole-genome and reduced-representation studies have characterised the genetic structure of Cannabis in stages: van Bakel and colleagues published the first draft genome in 2011, Sawler and colleagues established genome-wide hemp-drug differentiation in 2015 and Lynch and colleagues published a pangenome of wild and cultivated accessions in 2025.[9][10] The picture from molecular work is partial: historical hemp-drug differentiation and the South Asian / Central Asian split among drug-type populations are recoverable from genome data, but recent hybridisation has obscured taxonomic structure in modern commercial accessions, and no molecular study to date has provided unambiguous evidence for one species concept over the others.[2] Genomic work specific to Cannabis is treated in Phylogenetics of Cannabis.
Hybridisation and the modern accession landscape
A practical complication for any contemporary Cannabis classification is the scale of intentional hybridisation between drug-type populations of different geographic origin since the late 1970s. Breeders working in North America and Europe crossed vernacular "Sativa" and "Indica" lines extensively from the 1980s onward, producing the hybrid cultivar populations that now dominate the legal and illegal markets. McPartland and Guy concluded that recently hybridised accessions blur taxonomic boundaries to the point that morphological and molecular methods alike struggle to assign them reliably to the underlying gene pools.[3]
This complication is one of the principal reasons that conservation taxonomy focuses on pre-hybridisation source populations: the formal four-variety treatment of McPartland and Small 2020 was prepared to give names to landrace and wild populations whose displacement was happening faster than they could be characterised under the existing nomenclature.[4] The relationship between formal Cannabis taxonomy and the conservation status of landrace populations is treated in Landrace Cannabis.
See also
References
- ↑ 1.0 1.1 1.2 1.3 1.4 1.5 McPartland, J.M. (2018). Cannabis systematics at the levels of family, genus, and species. Cannabis and Cannabinoid Research, 3(1), 203–212. doi:10.1089/can.2018.0039
- ↑ 2.0 2.1 2.2 2.3 2.4 Reimann-Philipp, U., Speck, M., Orser, C., Johnson, S., Hilyard, A., Turner, H., Stokes, A.J. & Small-Howard, A.L. (2024). Genomics-based taxonomy to clarify cannabis classification. Genome, 67(1), 1–13. doi:10.1139/gen-2023-0005
- ↑ 3.0 3.1 3.2 3.3 3.4 3.5 3.6 3.7 McPartland, J.M. & Guy, G.W. (2017). Models of cannabis taxonomy, cultural bias, and conflicts between scientific and vernacular names. Botanical Review, 83(4), 327–381. doi:10.1007/s12229-017-9187-0
- ↑ 4.0 4.1 4.2 4.3 4.4 McPartland, J.M. & Small, E. (2020). A classification of endangered high-THC cannabis (Cannabis sativa subsp. indica) domesticates and their wild relatives. PhytoKeys, 144, 81–112. doi:10.3897/phytokeys.144.46700
- ↑ 5.0 5.1 Small, E. & Cronquist, A. (1976). A practical and natural taxonomy for Cannabis. Taxon, 25(4), 405–435. doi:10.2307/1220524
- ↑ 6.0 6.1 Small, E. (2015). Evolution and classification of Cannabis sativa (marijuana, hemp) in relation to human utilization. Botanical Review, 81(3), 189–294. doi:10.1007/s12229-015-9157-3
- ↑ Hillig, K.W. & Mahlberg, P.G. (2004). A chemotaxonomic analysis of cannabinoid variation in Cannabis (Cannabaceae). American Journal of Botany, 91(6), 966–975. doi:10.3732/ajb.91.6.966
- ↑ 8.0 8.1 Hillig, K.W. (2005). Genetic evidence for speciation in Cannabis (Cannabaceae). Genetic Resources and Crop Evolution, 52(2), 161–180. doi:10.1007/s10722-003-4452-y
- ↑ 9.0 9.1 Sawler, J., Stout, J.M., Gardner, K.M., Hudson, D., Vidmar, J., Butler, L., Page, J.E. & Myles, S. (2015). The genetic structure of marijuana and hemp. PLOS ONE, 10(8), e0133292. doi:10.1371/journal.pone.0133292
- ↑ 10.0 10.1 Lynch, R.C., Padgitt-Cobb, L.K., Garfinkel, A.R., Knaus, B.J., Hartwick, N.T., et al. (2025). Domesticated cannabinoid synthases amid a wild mosaic cannabis pangenome. Nature. doi:10.1038/s41586-025-09065-0