Get Involved
About Us
Our Members

Cannabis Taxonomy: The ‘Sativa’ Vs. ‘Indica’ Debate

Marijuana and hemp (Cannabis) and the closely related hop genus (Humulus) are the only widely known genera included in the small, but economically valuable, Cannabaceae family. Swedish botanist Carl Linnaeus, the “father of modern taxonomy,” first published the scientific name Cannabis sativa in his seminal Species Plantarum of 1753. The Latin name Cannabis derives from Greek (kannabis) and may have been originally derived from Scythian. The term sativa simply means “cultivated” and describes the common hemp plant that was widely grown across Europe in Linnaeus’ time. We, the authors, consider C. sativa to be native to western Eurasia and especially Europe, where, for millennia, the plant has been grown for its strong fibers and nutritious seeds, and from where it was introduced to the New World multiple times during early European colonization. Cannabis sativa plants also produce very small amounts of the compound delta-9-tetrahydrocannabinol (THC), the medically valuable and primary psychoactive cannabinoid found only in Cannabis. Since C. sativa evolved within the geographical limits of western Eurasia, it represents only a small portion of the genetic diversity seen in the genus Cannabis worldwide.1

In 1785, European naturalist Jean-Baptiste Lamarck described and named a second species, Cannabis indica, meaning “Cannabis from India,” after the origin of the first samples of this highly psychoactive plant that reached Europe. Cannabis indica has the genetic potential to produce relatively large amounts of THC. The species is used for marijuana and hashish production, but in many regions of eastern Asia it also has a long history of cultivation for fiber and seed. Humans make cloth out of C. indica fibers and eat the seeds, but this native eastern Eurasian species is more commonly used today as a drug plant with widespread social and medicinal importance reaching well beyond its original geographical range.1

While Karl Hillig, PhD, was a doctoral student at Indiana University, he used morphological and chemical characteristics to investigate the diversity of the Cannabis genus and proposed taxonomic groupings (subspecies) that support the original two-species concept.2-5 Hillig recognized European cultivated Cannabis as a separate species (C. sativa). Because this species typically has narrow leaflets and is primarily used for hemp fiber and seed production, we refer to it as narrow leaflet hemp (NLH). European C. sativa NLH populations are much less genetically diverse than those found in many other regions.

Hillig assigned the remainder of the world’s cultivated varieties to C. indica and divided them into three subspecies. One of these subspecies, C. indica subsp. indica, includes varieties that span the Indian subcontinent from Southeast Asia to western India and into Africa. These traditional drug varieties produce abundant amounts of THC with little if any cannabidiol (CBD). CBD is the second most common cannabinoid, and is non-psychoactive, but it has been shown to be medicinally effective for a variety of indications. By the 19th century, high-THC C. indica subsp. indica reached the Caribbean region and steadily spread throughout Central and South America. Since the 1960s, most of the drug Cannabis that reached North America and Europe was C. indica subsp. indica. Marijuana users commonly call domestically grown plants of these varieties “sativas” because their leaflets are relatively narrow, and therefore exhibit a superficial resemblance to those of European NLH plants. However, in our construct, this is a misnomer as C. sativa plants produce little if any THC. Based on Hillig’s research, we now refer to members of C. indica subsp. indica as narrow leaflet drug (NLD) varieties. Although they have relatively narrow leaflets like NLH (C. sativa) plants, the NLD plants can produce an abundance of THC and are most commonly used for their psychoactive effects. Based on taxonomic tradition, these plants are properly called “indicas” rather than “sativas.”

A second C. indica subspecies originated in Afghanistan where crops were traditionally grown to manufacture sieved hashish, a mechanically concentrated Cannabis drug. From 1974, when descriptions and photos of Afghan Cannabis were published by Harvard professor Richard E. Schultes, PhD, it became readily apparent that it represented a type of drug Cannabis previously unknown outside of Eurasia, belonging neither to Linnaeus’ C. sativa nor Lamarck’s C. indica.6 Its shorter, more robust stature, and broad, dark green leaves easily distinguish it from the taller, lighter green, and more laxly branched NLD varieties. Because of its limited geographic range and restricted usage, the Afghan genome is less diverse than the NLD genome. By the late 1970s, seeds of Afghan hashish varieties reached Europe and North America and were rapidly disseminated among marijuana growers. At this time, all Cannabis varieties were commonly considered to be members of a single species, C. sativa, and the familiar NLD marijuana varieties were called “sativas” to differentiate them from the newly introduced and quite different looking Afghan varieties commonly called “indicas.” Hillig named them C. indica subsp. afghanica, which we now refer to as broad leaflet drug (BLD) varieties to differentiate them from NLD varieties. On average, populations of BLD plants contain approximately equal amounts of THC and CBD. Although BLD varieties are also considered by us to be members of C. indica, it is more correct to distinguish them from subspecies indica from India by calling them subspecies afghanica, or simply “Afghans.”

Hillig’s third grouping within C. indica is subspecies chinensis, which comprises the traditional East Asian fiber and seed varieties and associated feral populations. We refer to this group as broad leaflet hemp (BLH). Like other subspecies of C. indica, varieties of C. indica subsp. chinensis possess the genetic potential to produce psychoactive THC, but East Asian cultural traditions, such as Confucianism, have long encouraged the selection of these varieties for their economically valuable fiber and seed, rather than their psychoactive potential. As a result, total cannabinoid production is lower than in subspecies indica and afghanica.



Evolutionary theory predicts that, at some point in time, there must have been a putative ancestor of the two modern species, C. sativa and C. indica. This ancient ancestor is often referred to as C. ruderalis, which may have originated somewhere in Central Asia. However, by now it is probably extinct, and seemingly ancestral populations are more likely descendants of feral plants that escaped from cultivation long ago. Evolutionary hypotheses based on plant distribution studies, paleoclimate modeling, archaeological evidence, and the historical record propose that C. sativa NLH most likely originated in a temperate region of western Eurasia, possibly in the foothills of the Caucasus Mountains, from a putative hemp ancestor with diminished biosynthetic potential to produce THC. Cannabis indica likely originated in the Hengduan Mountains, in present-day southwestern China, from a putative drug ancestor that had evolved an enhanced ability to produce THC. Early C. indica populations diversified as they were introduced by humans to different geographical regions where they may have further evolved into the three subspecies, all of which produce THC.1

Cultivated plant varieties are called cultivars, and when cultivars are grown and maintained by local farmers over generations, we refer to them as landrace cultivars, or landraces. Landraces evolve in a balance between natural selective pressures exerted by the local environment favoring survival, and human selections favoring a cultivar’s ability to both thrive under cultivation and produce particular culturally preferred products. Early humans spread Cannabis into many new regions as they moved, and at each new camp or settlement they selected seed from superior plants that were suited to their own uses and processing methods. By sowing seeds from the most favorable individuals, traditional farmers developed and maintained the landraces upon which present-day hybrid hemp and drug cultivars were founded.

Hemp cultivars were derived from crosses between different European NLH landraces and East Asian BLH landraces. Traditional Asian, African, and New World drug landraces were, until relatively recently, all pure NLD types. Before the introduction of BLD landraces from Afghanistan in the late 1970s, hybrids between imported NLD landraces formed the core genome of domestically produced drug Cannabis in both North America and Europe. It is through crossing NLD and BLD landraces from such geographically isolated populations that modern hybrid sinsemilla (Spanish for “seedless”) cultivars were created.1

Unfortunately, we cannot return today to a region previously known for its fine Cannabis and expect to find the same landraces that were growing there decades before. Cannabis is open-pollinated, with male and female flowers borne on separate plants, and, therefore, two plants are usually required to produce a seed. Random combinations of alleles (forms of a gene) and accompanying variation are to be expected. Cannabis landrace varieties are best maintained by repeated natural and human selection in situ — nature selecting for survival and humans selecting for beneficial traits. Without persistent human selection and maintenance, these landrace varieties will tend to drift back to their atavistic, naturally selected survival mode.

The Western world began using imported marijuana and hashish in the 1960s, and all of the remarkable imported varieties available then were traditionally maintained landraces. Within a decade, the demand for quality drug Cannabis exceeded traditional supplies, and mass production in the absence of selection became the rule. Rather than planting only select seeds, farmers began to sow all their seeds in an effort to supply market demand, and the quality of commercially available drug Cannabis began to decline. In addition, travelers returned to the supplying nations and introduced seeds of “improved” Western sinsemilla varieties that interbred with the local landraces and thus contaminated the local genomes. Landraces can no longer be replaced; they can only be preserved. The few remaining pure landrace varieties in existence now, some kept alive for decades as seeds and cuttings, are the keys to future developments in drug Cannabis breeding and evolution. It will be a continuing shame to lose the best results of hundreds of years of selection by local farmers. After all, our role should be as caretakers preserving the legacy of traditional farmers for the future benefit of all.



Cannabis research is a work in progress, and not all researchers agree on a single taxonomy.7 DNA sequencing is currently being used to characterize the diversity of many plant and animal groups, including Cannabis. While our knowledge grows and the evolutionary history of Cannabis is revealed, changes in taxonomic nomenclature will continue to reflect our deepening understanding of this medically valuable, yet controversial, plant. More broadly, whether we discover that Cannabis plants belong to one or more species, we can be sure that humans have long known, used, dispersed, cultivated, and artificially selected these plants to perpetuate a truly wide range of diversity.

Robert C. Clarke is the author of several Cannabis science books and has traveled extensively throughout Eurasia documenting traditional Cannabis production and use. His breeding interests include selection and preservation of landrace varieties, and developing narrow leaflet drug varieties and hashish cultivars. Clarke is the co-founder and director of BioAgronomics Group, an international cannabis industry consultancy, serves as projects manager for the International Hemp Association, and holds a seat on the Phylos Bioscience Cannabis Evolution Project scientific advisory board. He may be contacted at

Mark D. Merlin, PhD, is a professor in the botany department of the University of Hawai`i at Mānoa and an elected member of the Linnean Society of London. A large part of his long-term biological and historical research has involved drug plants and their past and present uses by people. He has authored or co-authored a number of books on this general subject, including Man and Marijuana: Some Aspects of their Ancient Relationships (Fairleigh Dickinson University Press, 1972), On the Trail of the Ancient Opium Poppy (Associated University Presses, 1984), Kava: The Pacific Drug (co-authored with Vincent Lebot and Lamont Lindstrom; Yale University Press, 1992), and, most recently, Cannabis: Evolution and Ethnobotany (co-authored with Robert C. Clarke; University of California Press, 2013). He may be contacted at


  1. Clarke RC, Merlin MD. Cannabis: Evolution and Ethnobotany. Berkeley, CA: University of California Press; 2013.
  2. Hillig KW. A chemotaxonomic analysis of terpenoid variation in Cannabis. Biochem Syst Ecol. 2004;32(10):875-891.
  3. Hillig KW. A multivariate analysis of allozyme variation in 93 Cannabis accessions from the VIR Germplasm Collection. J Indust Hemp. 2004;9(2):5-22.
  4. Hillig KW. Genetic evidence for speciation in Cannabis (Cannabaceae). Genet Resour Crop Ev. 2005;52(2):161-180.
  5. Hillig KW, Mahlberg PG. A chemotaxonomic analysis of cannabinoid variation in Cannabis (Cannabaceae). Am J Bot. 2004;91(6):966-975.
  6. Schultes RE, Klein WM, Plowman T, Lockwood TE. Cannabis: An example of taxonomic neglect. Bot Mus Leafl Harv Univ. 1974;23(9):337-364.
  7. Small E. Evolution and classification of Cannabis sativa (marijuana, hemp) in relation to human utilization. Bot Rev. 2015;81(3):189-294.