Models of Cannabis Taxonomy, Cultural Bias, and Conflicts between Scientific and Vernacular Names

Abstract

Debates over Cannabis sativa L. and C. indica Lam. center on their taxonomic circumscription and rank. This perennial puzzle has been compounded by the viral spread of a vernacular nomenclature, “Sativa” and “Indica,” which does not correlate with C. sativa and C. indica. Ambiguities also envelop the epithets of wild-type Cannabis: the spontanea versus ruderalis debate (i.e., vernacular “Ruderalis”), as well as another pair of Cannabis epithets, afghanica and kafirstanica. To trace the rise of vernacular nomenclature, we begin with the protologues (original descriptions, synonymies, type specimens) of C. sativa and C. indica. Biogeographical evidence (obtained from the literature and herbarium specimens) suggests 18th–19th century botanists were biased in their assignment of these taxa to field specimens. This skewed the perception of Cannabis biodiversity and distribution. The development of vernacular “Sativa,” “Indica,” and “Ruderalis” was abetted by twentieth century botanists, who ignored original protologues and harbored their own cultural biases. Predominant taxonomic models by Vavilov, Small, Schultes, de Meijer, and Hillig are compared and critiqued. Small’s model adheres closest to protologue data (with C. indica treated as a subspecies). “Sativa” and “Indica” are subpopulations of C. sativa subsp. indica; “Ruderalis” represents a protean assortment of plants, including C. sativa subsp. sativa and recent hybrids.

Appendix 1: Protologues

Appendix 1a. Cannabis sativa L.

Diagnosis or description

Linnaeus (1753) classified C. sativa in Class Dioecia (dioecious plants), and Order Tetrandria (male plants have four stamens). These aspects of morphology are the only descriptive elements he provided in Species Plantarum. Even by Linnaeus’s standards, this is an extremely brief description. For genera with only one species, such as Cannabis, Linnaeus provided a longer description in Genera Plantarum (Linnaeus, 1754). For this reason, the ICN explicitly links taxa coined in Species Plantarum with their descriptions in the fifth edition of Genera Plantarum. The Cannabis description in Genera Plantarum is limited to flower and seed morphology. Translated from Latin:

MALE. Calyx: Perianth five-parted: leaves oblong, acuminate-obtuse, concave. Corolla: absent. Stamens: filaments five in number, thread-like, short. Anthers: oblong, quadrangular.

FEMALE. Calyx: Perianth single-leaved, oblong, acuminate, opposite at the base, dehiscing longitudinally, persistent. Corolla: absent. Pistil: Ovary minute, Style divided, sharply pointed, long. Stigma acute. Perianth minimal. Calyx tightly closed. Seed: Nut globose-depressed, bivalvate.

Note a discrepancy: Species Plantarum Linnaeus (1753) described males with four stamens, whereas Genera Plantarum (Linnaeus, 1754) described plants with five stamens. Five stamens is correct. The correct version was passed down verbatim from the first edition of Genera Plantarum, written in 1737. Hence the 1753 error was typographic. Linnaeus corrected his mistake in the second edition of Species Plantarum in 1763.

Synonymy and references

Linnaeus (1753) cited five C. sativa synonyms by five authors, presented in his typical telegraphic style. We provide a translation of Linnaeus’s abbreviations in Box 1. Note a typographic error: Linnaeus said C. foliis digitatis appeared in Materia Medica on page 475. In fact the taxon appears in Materia Medica on page 162 (Linnaeus, 1749).

Box 1 Translation of author and citation abbreviations by Linnaeus (1753)

Cannabis foliis digitatis Linnaeus, Hortus Cliffortianus (1738: p. 457); Linnaeus, Hortus Upsaliensis (1748: p. 297); Linnaeus, Materia Medica (1749: p. 162, no. 457); Dalibard, Florae Parisiensis Prodromus (1749: p. 200); van Royen, Florae Leydensis prodromus (1740: p. 221).

Cannabis sativa Bauhin, Pinax Theatri Botanici (1623: p. 320). ♀

Cannabis mas Daléchamps (d’Aléchamps), Historia generalis Plantarum (1587: p. 497). ♀

Cannabis erratica Bauhin, Pinax Theatri Botanici (1623, p. 320). ♂

Cannabis femina Daléchamps (d’Aléchamps), Historia generalis Plantarum (1587: 497). ♂

Linnaeus notably excluded taxa of Asian Cannabis from the C. sativa protologue. Sixteen years earlier, in his previous synopsis of Cannabis, Linnaeus (1737) discussed two taxa assigned to South Asian Cannabis by van Rheede tot Drakestein (1690), and two taxa assigned to East Asian fiber by Kaempfer (1712). Furthermore, Linnaeus (1737) cited three other authors, Bauhin, Ray, and Morison, who also coined taxa for psychoactive Asian Cannabis, distinct from European hemp. Linnaeus knew about those taxa—he cited taxa that Bauhin, Ray, and Morison assigned to European hemp—but he did not address them. In his Materia Medica, Linnaeus (1749) listed narcotica and inebrians as some of the attributes of Cannabis, which are adjectives used by the aforementioned authors who described Asian Cannabis.

Type specimen

Linnaeus’s type specimen of C. sativa is stored in the Linnaean Herbarium (herb. LINN), specimen no. 1177.2, a lectotype designated by Villiers (1973). As noted by Jarvis (2007), Villiers holds a one-year priority over Schultes et al. (1974), who designated a different lectotype, Hortus siccus Cliffortianus (herb. BM) p. 457. Stearn (1974) dedudced from the numbering system that Linnaeus collected 1177.2 in 1753. No type location is given, but in 1753 Linnaeus lived in Uppsala.

Lectotype 1177.2 consists of the uppermost 28 cm of a fruiting staminate plant. Leaves branch alternately from the stalk, palmately compound, light green, mostly with three leaflets, and with long petioles. Central leaflets are narrowly lanceolate, acuminate, with sharp serrations, up to 78–85 × 7–9.5 mm. Inflorescences are loose, not dense; subtending floral leaves have a sparse covering of capitate-sessile glandular trichomes; perigonal bracts (i.e., bracteole, calyx) have a relatively sparse covering of capitate-stalked glandular trichomes. Styles and stigmas are not prominent, pale green in color. The fruit includes a non-persistent perianth, achene large (4.8 × 2.5 mm), oblong in outline, pale green with a fine reticulated pattern, base short and pointed with a simple articulation.

The previously-recognized lectoype, Clifford Herbarium p. 457.1, is illustrated by Stearn (1974). The flowering top is somewhat denser than 1172.2, with slightly larger leaves, 3–5 leaflets per leaf, leaflets broadly lanceolate, and with larger achenes (5.0 × 3.5 mm). The specimen likely dates to Linnaeus’s time in Holland, and represents “the old cultivated hemp stock of northern Europe” (Stearn, 1974).

Appendix 1b. Cannabis indica Lam.

Diagnosis or description

Lamarck (1785) provided a full description of C. sativa (not just its flower parts, like Linnaeus). Lamarck then separated C. indica from C. sativa by “very distinct” morphological and phytochemical differences. We translate from French: “It is smaller, more ramifications of the stalks, which are tough [woody] and more cylindrical, and distinguished particularly by alternating branches. Their leaflets are very narrow, linear-lanceolate, and very acuminate. The male individuals bear five to seven leaflets, but those that are female may exhibit as few as three for each petiole, and leaves at the top [of the plant] are even simple [not compound]. The female flowers have a velous calyx, and long styles that are alike.”

We interpret “vellous calyx” to mean “velvety perigonal bract,” due to a dense pubescence of capitate-stalked glandular trichomes. Lamarck noted that C. indica does not provide a good source of fiber, and it produces a strong odor “resembling somewhat that of tobacco.” He mentioned the intoxicating properties of C. indica when smoked in a pipe. “The principal effect of this plant consists of going to the head, disrupting the brain, where it produces a sort of drunkenness that makes one forget ones sorrows, and produces a strong gaiety.”

Synonymy and references

Lamarck cited six synonyms and their authors:

• 1. Cannabi similis exotica Bauhin (Pinax Theatri Botanici 1623: p. 320); Bauhin cited descriptions of Bangue by da Orta and Acosta, in Goa, India;

• 2. Kalengi-cansjava Rheede (Hortus Malabaricus 1690; 10: 119); assigned to male plants in Kochi, India;

• 3. Tsjeru-cansjava Rheede (Hortus Malabaricus 1690; 10: 121); assigned to male plants in Kochi, India;

• 4. Cannabis peregrina gemmis fructuum longioribus Morison (Plantarum historiæ universalis Oxoniensis 1699; 3:433); Morison cited Bauhin and Rheede;

• 5. Cannabis indica Rumph (Herbarium Amboinense 1747; 5: 208), a “pre-Linnaean” (pre-1753) taxon assigned to plants in Indonesia;

• 6. Dakka ou Bangua Prévost (1768), an editorialized version of Kolb (1719) who described plants in South Africa.

Four authors in Lamarck’s synonymy (1, 2, 3, 4) previously appeared in Linnaeus (1737). Lamarck omitted two synonyms cited by Linnaeus: Bangue cannabi Ray 1686, and Bangue cannabi simile J. Bauhin and Cherler 1651–2. Both Linnaeus and Lamarck omitted other “pre-Linnean” (pre-1753) authors who differentiated European hemp from psychoactive Cannabis. Going back in time: Linder (1739), Burman (1737), Drury (1729), Hooke (1726), Kaempfer (1712), Petiver (1703), Plukenet (1696), Fryer (1698), Ray (1693), Chardin (1686), Knox (1681), de Flacourt (1661), Parkinson (1640), Sennert (1629), dos Santos (1609), de Gouvea (1606), Van Linschoten (1601), Alpini (1591), Acosta (1578), de L’Ecluse (1567), da Orta (1563), Fuchs (1542), and Bock (1539). The first European to assign separate names to European and Asian Cannabis was Ibn-al-Baitār, around 1240, in Arabic. He differentiated qinnab (plants he knew in Spain) from plants he encountered in Egypt, qinnab hindī, “Indian hemp” (Ibn al-Baitār, 1985).

Type specimen

Lamarck’s type specimen of C. indica is deposited at the Muséum National d’Histoire Naturelle in Paris (herb. P). It consists of the uppermost 30 cm of a flowering, seedless pistillate plant. Lamarck’s specimen shows more compact branching with shorter internodes than Linnaeus’s specimen. Branching density may reflect differences in cultivation: Lamarck’s specimen was presumably open-grown, and Linnaeus’s specimen grown in typically dense hemp fields. However, the effects of sowing density upon branching density are minimal in the uppermost 30 cm of a flowering top. Leaves branch alternately from the stalk, palmately compound, medium green, mostly with three leaflets, and short petioles. Central leaflets are narrowly linear-lanceolate, acuminate, with sharp serrations, 55–58 × 5–7 mm. Inflorescence are somewhat compact; subtending floral leaves have an abundant covering of capitate-sessile glandular trichomes; perigonal bracts express a moderate density of capitate-stalked glandular trichomes. Styles and stigmas are prominent, agglutinized with trichome exudate, and light brown. The type specimen is a sample of sinsemilla, lacking fruits.

No type location is given. Lamarck (1785) wrote, “This plant, which Mr. Sonnerat gave us pieces he brought back from India…” According to Schultes et al. (1974), “we are at a loss to indicate a definite area, partly because of vagueness of geographical terminology in that period.” In fact, we know when and where Pierre Sonnerat made his collections. He explored India and China between 1774 and 1781. In India he made extensive collections around Pondicherry (Ly-Tio-Fane, 1976), so Sonnerat likely collected C. indica near Pondicherry. Unfortunately he did not mention Cannabis in his travelogue (Sonnerat, 1782).

Discussion

Lamarck erroneously described C. indica with “alternately branching leaves” and C. sativa with opposite branches (i.e., decussate, the succeeding pairs turned 180°). In fact, both C. indica and C. sativa do both—opposite branching during vegetative growth, switching to alternate branching during flowering. Lamarck may have erred because the specimen he received from Sonnerat was a flowering top, with alternate branching. Lamarck cited authors who illustrations of Asian Cannabis also depicted flowering tops, with alternate branching habits (Acosta 1619, Rheede 1690, Rumpf 1747). Interestingly, the same year that Lamarck (1785) described C. indica, his co-worker André Thouin was growing Chinese hemp. Thouin (1785) described chanvre de Chine with alternate branching.

Note that we cite 1785 as the year that Lamarck coined C. indica. See Breistroffer (1948) for evidence that the commonly cited date of 1783 is inapplicable. Each volume of Encyclopédie Méthodique was printed in two parts; the second part of each was printed later than the date on the title page. We use the date specified by Stafleu and Cowan (1979).

Appendix 1c. Cannabis sativa var. spontanea Vav.

Diagnosis or description

Vavilov (1922) collected wild-type germplasm near Saratov, Samara, Astrakhan, and Tsaritsyn (Volograd). “The study of this hemp in our laboratory by Varvara F. Antropova compels one to consider it to be, without doubt, wild.” He described a series of forms that transition from the cultivated variety to the wild-type variety, “differing in the coloration of the seeds and in the size and form.”

Plants at the wild end of the spectrum expressed anthocyanin in their stalks, and produced seeds that were dull (not shiny), elongated (not rounded), and smaller than seed from cultivated plants. Seed color ranged from light or white forms with a mosaic, to dark gray and light brown with marbling.

Vavilov provided a brief diagnosis, translated from Russian: “At the bases of the seeds there are formations that resemble ‘horseshoes’ (analogous to wild oats), along which there occurs a breaking off and a shedding of seeds at maturation. Frequently this horseshoe is clearly defined and undoubtedly is a morphological and biological feature clearly distinguishing it from the cultivated form. At maturation, the wild variety with the horseshoe is easily distinguishable fro the ordinary cultivated variety by the shedding of the seeds which have not yet reached full maturity. As was mentioned by Professor D. E. Janischevsky, wild forms of hemp appear primarily as dark-seeded marbled forms.”

Synonymy and references

Vavilov (1922) cited no synonyms. He apparently overlooked Czernajew (1859), who assigned the name C. sativa var. spontanea to wild-type plants near Kharkiv, Ukraine. Czernajew’s taxon has no standing, because he did not describe the variety, and therefore did not meet a basic provision of the ICN (Small & Cronquist, 1976). Vavilov mentioned earlier work on wild hemp by Janischevsky, but did not cite a publication.

Type specimen

Several specimens labeled C. sativa var. spontanea are stored in Vavilov’s herbarium at WIR. Small and Cronquist (1976) chose one as the lectotype, specimen Antropova 121. The label states the plant was grown in 1925 at the Kamenno-Stepnaya experiment station in Voronezh, from seeds collected near Saratov in 1921, by Vavilov’s assistant, Varvara F. Antropova.

The specimen consists of the top 42 cm of what appears to be a large plant, with internode spacing similar to Linnaeus’s type specimen. Leaves with 3–5 leaflets, petioles short, leaflets narrow lanceolate, up to 130 × 10 mm. Inflorescences are loose; subtending floral leaves have a sparse covering of capitate-sessile glandular trichomes; perigonal bracts have relatively few capitate-stalked glandular trichomes and many cystolith trichomes. Achenes are medium-sized (3.8–4.0 mm long), oblong in outline, pale green with a fine reticulated pattern overlaid by irregular dark marbling, and a weakly protuberant base.

Discussion

Vavilov (1926) debated whether C. sativa var. spontanea was truly wild, or a spontaneous escape of formerly cultivated plants. This question was also raised by Janischevsky regarding Cannabis ruderalis (see below). Both botanists were unaware that this debate arose earlier, regarding plants in the same location—Saratov. Three naturalists travelling together in 1768 encountered wild-type plants about 10 km downriver from Saratov. Lepechin (1774) said the plants did not differ from cultivated plants. He considered them feral escapes, sown by former inhabitants. “This Indian plant species can not be native here.” Nearby lay the ruins of Uvek, a trading center of the Kipchak Khanate, destroyed by Tīmūr in 1395. Pallas (1793) said the wild-type hemp looked like cultivated plants, which he attributed to former inhabitants. Falck (1786) wrote, “On the Volga one finds the wild hemp especially on the sites of former cities.” He noted branchiness in the wild-type plants, “These give no straight and uniform fibers.”

Vavilov and Janischevsky both decided that plants around Saratov were truly wild. Small (1975a) throws a variable into the debate. He discovered that domesticated Cannabis reverted to a wild-type phenotype within 50 generations (years) of prohibition in Canada. This plasticity makes it difficult to distinguish truly wild plants from formerly cultivated plants that have reverted to wild-type phenotypes. Doebley et al. (2006) also emphasized that domesticated C. sativa easily escapes cultivation and reverts to a wild-type phenotype. Small and Cronquist (1976) suggested that truly indigenous, never-domesticated, wild-type Cannabis may be extinct. “If unaltered wild populations exist (which we doubt) they cannot be clearly distinguished from those contaminated by the influence of domestication.”

Was there any reproductive isolation between wild-type and cultivated plants? Their geographic distributions overlapped (sympatry). Hybridization and introgression was limited by temporal (allochronic) isolation—Janischevsky noted that ruderalis matured in mid-June, while cultivars were still in the vegetative stage. There may have been some ecological inviability as well—a condition where hybrid offspring are normal but suffer lower viability because they cannot find an appropriate ecological niche. However, Vavilov and Janischevsky noted intermediate forms, which suggested some hybridization occurred between the populations.

Appendix 1d. Cannabis ruderalis Janisch.

Janischevsky (1924) coined the species name C. ruderalis, but added parenthetically an alternative taxonomic rank: C. sativa var. ruderalis. “I am inclined to consider it a well marked variety.” We provide key details of his large protologue—a 13 page article.

Diagnosis or description

Translated from Latin: “Fruits rather small, rather hard, narrowly ovate, base attenuate, perianth marbled-spotted, persistent and enclosing [the fruit] at maturity, [the fruit] quickly dropping off [the plant].”

In addition to the Latin diagnosis, Janischevsky provided additional characters: Male flower produced shorter perianths and anthers than those in cultivated plants. Average seed size was 3.5 mm long × 2.5 mm wide, and rarely reached 4.5 × 3.0 mm. The female flower’s perianth was persistent and stuck to the surface of the seed—not unlike hops (Humulus lupulus). The seed had an elongated base in the form of a short pedicle. Within the pedicle Janischevsky detected cells with oily inclusions—so he considered the structure an elaiosome—an oil-rich body that attracts insects, which acts as a seed dispersal agent.

Vegetative characters included relatively short height, usually 0.7–1.1 m but up to 2.1 m, with a strongly branched habitus. Stalks were usually ribbed, with thick xylem at the base (i.e., woody), and with the upper stalk colored by anthocyanin. Leaves had 5–7 leaflets, colored deep green on the top side, and gray-white on the underside due to cystolith trichomes.

Janischevsky conducted common garden experiments, which showed that wild-type plants matured more rapidly than cultivars. They came into flower mid-June, and first seeds were collected July 10th (while cultivars were still in vegetative stage). Wild-type plants compared to cultivars showed greater drought tolerance, and more shade tolerance in forested areas. Wild-type seeds readily disarticulated from plants. The seeds demonstrated a prolonged period of dormancy and slow germination, whereas seeds of cultivated hemp had no period of dormancy.

Janischevsky described mutualism between wild hemp and an insect, Pyrrhocoris apterus. The insect sucked oil out of the elaiosome. The rest of the seed was left intact and capable of germination. In the process of feeding, the bug carried the seed “far distances,” and facilitated the spread of wild hemp. The following year, Janischevsky (1925) wrote a whole article about wild hemp and P. apterus.

Synonymy and references

Janischevsky did not list any explicit synonyms under C. ruderalis. He did not mention C. sativa var. spontanea Czernajew (1859), or C. sativa var. spontanea Vavilov (1922). He noted that taxa other than C. sativa were coined by other Russian botanists, such as Messerschmidt (in Amman, 1739), Sievers (1796), Ledebour (1847,–49), and Korshinskii (1898). Yet others suggested that C. sativa, of Asiatic origin, may also be indigenous to Europe—but they expressed this opinion with great caution. Janischevsky took five pages to recount observations by 22 authors: Erndtel, 1730, Messerschmidt (in Amman, 1739), Sievers, 1796, Rochel, 1835, Claus, 1838, Ledebour (1847–49), Christison, 1850, de Candolle, 1855, Becker, 1858, Heuffel, 1858, Dukerley 1866, de Candolle 1869, Kanitz, 1881, Zinger, 1885, Shmal’gauzen, 1897, Korshinskii, 1898, Velenovský 1898, Prain, 1904, Adamović, 1907, Bogdan, 1908, Ascherson & Graebner, 1911, and Vavilov 1922.

Type specimen and illustrations

Janischevsky’s lectotype specimen of C. sativa var. ruderalis, collected near Saratov, was stored at the Komarov herbarium (LE) in St. Petersburg (Small & Cronquist 1976). It may be lost (McP., pers. observ., St. Petersburg, 2010), and no suitable isolectotype was identified at LE. Small (1975b) provided a photograph of the lectotype.

Janischevsky provided two plates of line drawings, reproduced here as third-generation photocopies (Fig. 6, figure legend translated from Russian).

Fig. 6

Line drawings of C. ruderalis by Janischevsky (1924))

Рис. 1. A fruit of hemp collected in the vicinity of Saratov (Dudakov ravine).

Рис. 2. Two diagrammatic drawings of the longitudinal section of the fruits. Рис. 2a. A section in the plane that goes through the edge of the fruit. Рис. 2b. A section in the plane that goes between the edges of the fruit. I- perianth; II- pericarp, e- epicarp, E- endocarp, 5- point of insertion of perianth, 6-receptical transformed into elaiosome which juts out into the space among the carpels (see diagram B) in the form of rounded mases of tissue, v- vascular bundle.

Рис. 3. A bug Pyrrhocoris apterus on a Carex leaf carrying a hemp fruit.

Discussion

Janischevsky (1924) said he found hemp growing in “wild circumstances” in 1897, near Buzuluk (Orenburg Oblast). It flourished along the Tok River, a right tributary of the Samara, in Bashkir tribal land where hemp was not cultivated. Janischevsky added that he began growing specimens of wild hemp in the botanical garden at Saratov University in 1914. Vavilov arrived at Saratov University in 1917, joining his senior colleague. Janischevsky and Vavilov took field trips together down the Volga (Korotkova, 1978). It seems likely that Janischevsky pointed Vavilov’s attention to wild hemp. In Vavilov’s archives, the subject of wild hemp first appeared in a 1920 letter to his assistant Evdokia A. Teplykh, “Collect wild or feral hemp on the borders and the ravines near Saratov from selected plants, at least 500 plants, in separate bags. Keep in mind that wild hemp seed falls off easily, so it must be carefully collected” (Korotkova, 1978).

Vavilov (1926) elaborated upon C. sativa var. spontanea. His review of the literature borrowed from Janischevsky (1924) in a few plagiarized sentences. Vavilov cited six of Janischevsky’s 22 references (Erndtel, 1730, Ledebour, 1847–49, Korshinskii, 1898, Velenovský, 1898, Adamović, 1907, Bogdan, 1908). However, Vavilov added information about wild-type hemp from Mak & Regel, 1862, Boissier 1879, Hooker, 1890, Britton 1899, Dimo & Keller, 1907, Semenov (in Krylov, 1909), Knorring and Minkvitz Knorring & Minkvits, 1912, Vysotsky, 1915, Oganovsky, 1922, Sinskaya 1925, and Khrebtov, 1925.

Vavilov (1926) details work on wild-type hemp by his assistant Tatiana Ya. Serebriakova. She compared germplasm collected by Antropova in Saratov with germplasm collected by Eugeniya N. Sinskaya in the Altai (Sinskaya, 1925). Plants from both accessions grew 60–150 cm tall, with moderately thick stalks that were strongly branched. Leaves were medium sized, with 5–9 narrow leaflets. Seeds were small (2.7–3.0 mm long), dark, orbicular-elongated in shape, with horseshoes and a persistent, marbled perianth. Plants matured before cultivated plants, seeds rapidly dehisced from plants, and seed germination was usually slow and unequal.

Vavilov (1931) extended the range of C. sativa var. spontanea to Central Asia. While traversing the Tiān Shān to Lake Issyk Kul via the Bedel Pass, he wrote, “Cannabis sativa var. spontanea is a very common plant in northern Tiān Shān, especially on its north-facing slopes and valley.” Herbarium specimens collected on the north-facing slopes of the Tiān Shān look very different than Vavilov’s type specimen (shorter, with more branches, and broad oblanceolate leaflets). Zhukovsky (1962) proposed that C. ruderalis had not been sufficiently explored in Central Asia, and some populations may be undescribed forms.

Lastly we come to the question of priority. Vavilov (1922) scooped his senior collegue by coining C. sativa var. spontanea. Janischevsky (1924) erected another Latin name for the same wild-type hemp growing near Saratov. The taxonomic name became a point of contention between the two authors: Janischevsky politely mentioned Vavilov’s research, but always used the epithet ruderalis (Janischevsky, 1924, Janischevsky, 1925). Vavilov politely mentioned Janischevsky’s work, but always used spontanea (Vavilov, 1926, 1931; Vavilov & Bukinich, 1929).

Two taxonomic names that represent contaxic entities, based on different type specimens, are facultative synonyms. Vavilov’s earlier epithet has priority at the rank of variety. Janischevsky’s taxon has priority at the rank of species.

Since then, botanists have altered the ranks of spontanea and ruderalis. Serebriakova and Sizov (1940) elevated spontanea from a variety to a subspecies. The subspecies rank has been adopted by several authors (e.g., Bócsa and Karus Bòcsa & Karus, 1997; Hanelt, 2001; McPartland & Guy, 2004, Clarke & Merlin, 2013). Chu (1959) reduced ruderalis to a sub-varietal form, as C. sativa f. ruderalis. Liou (1988) brought Chu’s taxon back to a variety, C. sativa var. ruderalis. Liou mistakenly believed he coined a var. novo. Soják (1960) used Janischevsky’s taxon to describe hybrids between wild hemp and cultivated hemp, which he named X C. intersita. He later changed the taxon to C. sativa subsp. intersita (Soják, 1980).

A troika of influential Soviet texts set precedence by choosing C. ruderalis over C. sativa var. spontanea: Nekrasova (1934), Yarmolenko (1936), and Mal’tsev (1939). Soviet science at that time writhed under the rise of T. D. Lysenko. Lysenko became the pet scientist of Joseph Stalin. He fabricated genetic theories using the rhetoric of Marx and Michurin, and methodically annihilated his academic opponents. Lysenko labeled Vavilov a Trotskyite, which led to Vavilov’s arrest (Popovsky, 1984). After Vavilov was arrested, his assistant Tatiana Serebriakova coauthored her final Cannabis paper with Ivan A. Sizov. He was a Lysenkoite who “began energetically to liquidate the remnants of Vavilov traditions” (Medvedev, 1969). Serebriakova and Sizov (1940) elevated Vavilov’s taxon from a variety to a subspecies, but without his name in the basionym: C. sativa subsp. spontanea Serebriakova. C. ruderalis was synonymized under that taxon.

Schultes et al. (1974) treated C. ruderalis as a species separate from C. sativa. They expanded Janischevsky’s taxon to Central Asia, and erroneously typified it with a specimen from Tajikistan. Furthermore they described ruderalis as a very short, unbranched plant with broad leaflets. This departs from the concepts of Vavilov and Janischevsky. Schultes et al. (1974) cited Schultes, and described ruderalis as a short plant. Anderson (1980) cited Schultes, and described ruderalis as a short plant with broad, oblanceolate leaflets.

Small and Cronquist (1976) kept Vavilov’s taxon at the rank of variety. They recombined it as C. sativa ssp. sativa var. spontanea, typified by a Vavilov specimen from Saratov. They synonymized C. ruderalis under that taxon. They extended the range of spontanea to wild plants from Central Asia.

Schoenmakers (1986) introduced “Ruderalis” to underground Cannabis breeders. He collected germplasm near the Hungary-Ukraine border. His photos of “roadside ruderalis” illustrated an escape from cultivation, and not a wild-type plant. “The effect after smoking was stoney, but not high.” Other Cannabis breeders assigned the name “ruderalis” to feral plants in North America—obviously escapes from cultivation, and not wild, indigenous plants (e.g., Panik, 2012. Frank (1988) stated that potent, psychoactive “Ruderalis” invariably turned out to be misidentified “Indica.” This confusion persists: a 2013 web video about “Charlotte’s Web” by CNN medical correspondent Sanjay Gupta portrayed “Ruderalis” plants that clearly have Afghanistan heritage.

Appendix 1e. Cannabis sativa f. afghanica Vav.

Vavilov encountered wild and weedy plants in Afghanistan. First he described C. sativa f. afghanica (Vavilov, 1926). His description is transcribed below, where we combine the best of two English translations (Vavilov, 1926, 1992):

Diagnosis or description

“We observed wild hemp… in districts where hemp cultivation is entirely unknown. Belts of black hemp follow the sowings of corn and other cereals along the Kunar River (on the border between Afghanistan and India [now Pakistan]) from Chekosarai to Jalālābād, a distance of 160–200 km. In between Chekosarai and Jalālābād we discovered among the wild hemp a peculiar race with light-coloured, small seeds, and with a thin perianth easily taken away (f. afganica Vav.). The seeds of this race were very small (1000 seeds weighed 2.1–2.7 g), ten times smaller than the large-fruited races of the Far East (1000 seeds 26.0 g), the common Central Russian races (Orel, Kurst) show a weight of 17–19 g.

“The wild hemp collected on the Kunar River approached the cultivated type with respect to seed color and the slightly splitting pericarp, but is distinguished by readily shedding its seeds and in the development of “horse-shoes.” When sown, the seeds germinate very slowly and unequally, i.e., the plants show features of a typical wild plant. The wild hemp displayed other new characteristics. The leaflets of these plants were distinguished their obovate narrow shape, not observed by us among the European, Siberian and Turkestani forms… The wild Afghani races found along the frontier with Pakistan, with light-colored and easily splitting pericarps, have spread into Pakistan as well. The plants constitute a morphological link between the wild and the cultivated races of hemp with respect to the most important differentiating characteristics.”

In an accompanying table, Vavilov provided results of a common-garden experiment of C. sativa f. afghanica by his assistant Serebriakova. Plants matured late in the season, and seed germination was slow and unequal. Plants grew 60–150 cm tall, with moderately thick stalks, and strongly branched. Leaves were small, with 5–9 leaflets, and narrowly obovate in shape. Seed was described in an either-or duality: Seed size was either small (2.7–3.0 mm) or medium (3.0–4.0 mm), seed color was either dark or light, seed shape was either orbicular-elongate or orbicular, the seed perianth was either persistant or easily removed, the seed base either with or without a horseshoe. Vavilov seemed to be describing two kinds of plants. He wrote, “the plants show features of a typical wild plant,” followed by “the plants constitute a morphological link between the wild and the cultivated races of hemp.”

Synonymy and references

Vavilov provided no synonymy and cited no references.

Type specimen

No specimen labeled afghanica is deposited in Vavilov’s herbarium at WIR. Serebriakova grew germplasm of Afghan provenance at the Kamenno-Stepnaya Experiment Station (Voronezh Oblast); these herbarium specimens are labelled C. sativa var. spontanea, and annotated “like ruderal” (WIR specimens 4031, 4032, 4034, 4036, 4038, 4044, 4046). All are immature, without seeds, and could be either afghanica or kafirstanica.

Discussion

The either-or duality in Vavilov (1926) got sorted 3 years later, when Vavilov and Bukinich (1929) named a second taxon of Afghan plants: C. indica var. kafirstanica. The either-or descriptions in Vavilov (1926) referred to either afghanica or kafirstanica. Vavilov and Bukinich described kafirstanica as truly wild, with tiny achenes that were dark-colored and marbled, with an elaiosome. They described and illustrated C. indica f. afghanica, with larger achenes that were light colored (“white”), and a colorless involcre (perianth?), with little or no “horseshoe” (Fig. 7).

Fig. 7

Seeds illustrated by Vavilov and Bukinich (1929). Original caption: Left to right: 1. from northern Afghanistan—cultivated form—Cannabis sativa L.; 2. ordinary Russia hemp from Orel; 3. wild hemp from Saratov; 4., Cannabis indica f. kafiristanica Vav.; 5. Cannabis indica var. afghanica Vav. The upper row enlarged 6 times, the lower row showing the bases of achenes enlarged 10 times

Vavilov and Bukinich considered afghanica a “specialized form” of kafirstanica, “transitional” between wild kafiristanica and cultivated C. indica. Their nomenclature was not consistent—they used C. indica var. afghanica on page 380 (see caption in Fig. 7), f. afghanica on page 381, and C. indica var. kafirstanica f. afghanica on page 382. Small and Cronquist (1976) argue that the caption (Fig. 7) erroneously reversed the ranks of two taxa, Cannabis indica f. kafiristanica and C. indica var. afghanica, so they are typographical errors.

We propose that Vavilov erred by describing afghanica as “transitional” between wild and cultivated plants. The afghanica was probably domesticated—a feral escape, or recently naturalized. Its seed lacked a “horseshoe” (elongated base in the form of a short pedicle), and lacked a persistent perianth. It resembled a small version of the Afghani “cultivated form” (Fig. 7). The size of the cultivated seed would have been augmented by irrigation and fertilizer. Vavilov and Bukinich described afghanica growing in “abandoned lots and neglected plots of land, in fertilized agricultural soil, and between agricultural fields.” Arable land in Afghanistan is rarely abandoned and neglected.

Emboden (1974) assigned a varietal rank to the taxon, as C. indica var. afghanica. He gave seed size as 2.8 × 1.9 mm, without stating the provenance of his material. Small and Cronquist (1976) synonymized C. sativa f. afghanica Vav. under Lamark’s taxon—a domesticated variety. Vavilov’s taxon at the rank of species, as C. afghanica, was employed by McPartland (1996), McPartland et al. (2000), and Clarke and Watson (2002). Clarke (1998) used a varietal rank, C. indica var. afghanica. The taxon C. sativa var. afghan appeared in a U.S. patent (No. 6,403,530). Cervantes (2006) offered two ranks, C. afghanica and C. sativa var. afghanica. A subspecies rank, C. indica subsp. afghanica, was employed by McPartland and Hillig (2004) and Clarke and Merlin (2013). Recently the species rank, C. afghanica, has been applied by others (Tennstedt & Saint, 2009, Brownjohn & Ashton, 2012, Macedo et al., 2013, Henry, 2015, Laursen, 2015).

Appendix 1 f. Cannabis indica var. kafiristanica Vav.

Diagnosis or description

The description by Vavilov and Bukinich (1929) is translated from Russian: “Races of wild hemp in eastern Afghanistan have extremely small fruits with mosaic (1000 fruits weigh 2.1-2.7 g), i.e., 6-8 times smaller than small-seeded Central Russian cultivated hemp (Orel and Kursk hemp weighs 17-19 g). Characteristic for them is ready shattering of fruits due to the presence of a horseshoe, slow and uneven germination, i.e., the usual attributes of a wild plant. As regards to vegetative features, Afghan wild-weedy hemp is distinguished by small leaves with obovate leaflets of narrowed shape. In general, it is characterized by short stature, profuse branching from the first internode, and by short internodes.”

Vavilov and Bukinich also mention its early ripening (90–100 days in Voronezh Oblast). They provided a drawing of the seed (Fig. 8). Although Vavilov and Bukinich described kafiristanica plants as short in stature, a photograph of plants identified as C. indica var. kafirstanica (Fig. 267 in Vavilov and Bukinich) shows plants that were equal in height to maize plants with tassels.

Fig. 8

Type specimen of C. sativa var. kafiristanica Vav. (McPartland photo, taken at herb. WIR)

Synonymy and references

Vavilov provided no synonymy and cited no references.

Type specimen

Several specimens labeled C. sativa var. kafiristanica are deposited in Vavilov’s herbarium at WIR in St. Petersburg. Small and Cronquist (1976) chose one as the lectotype: WIR 3952, germplasm collected by Vavilov at Chekhosarai (now Asadābād) in 1924 and cultivated in 1927 at Pushkin Experiment Station (Detskoye Selo, St. Petersburg). A photograph of the lectotype appeared in Vavilov and Bukinich (1929), and appears here (Fig. 8). The entire plant is mounted on the herbarium sheet, about 30 cm tall. It is an immature pistillate plant, with tight internode spacing; nine pairs of opposite branches and only three alternate branches near the apex. Leaves with 5–7 overlapping leaflets, petioles long and thick, leaflets broad, oblanceolate, dark green, with coarse serrations, up to 46 × 18 mm. The inflorescences are immature but nevertheless compact and leafy, with an abundant covering of capitate-sessile glandular trichomes.

Discussion

Vavilov said afghanica and kafirstanica grew in the Kunar River valley—the border between Afghanistan and Pakistan—between Chighasaray (now Asadābād) and Jalālābād. We note a geographical contradiction: The Kunar River valley is not part of Kāfiristān (present-day Nuristān Province), it lies in Kunar Province, populated by Pushtan Afghanis, not ethnic Kāfirs. The place where Vavilov collected kafirstanica was not in Kāfiristān.

Serebriakova and Sizov (1940) made no mention of afghanica and kafirstanica; they placed drug plants from Afghanistan in the taxon C. sativa subsp. culta prol. Asiatica var. narcotica. Emboden (1974) gave seed size of C. indica var. kafiristanica as 3.0 × 2.2 mm (larger than var. afghanica), without stating the provenance of his material.

Small and Cronquist (1976) recombined Vavilov’s taxon as C. sativa subsp. indica var. kafiristanica. They considered it the wild-type associate of domesticated C. sativa subsp. indica var. indica. As such, they expanded the taxon’s concept beyond Afghanistan to wild-type plants from India, Nepal, China, South Africa, and Colombia. Hillig and Mahlberg (2004) also extended kafiristanica beyond Afghanistan to feral populations from India and Nepal. The rank of kafiristanica has bounced around a bit, like that of afghanica. Chrtek (1981) elevated it to a species rank, C. kafiristanica. A subspecies rank, C. indica subsp. kafiristanica, was employed by McPartland and Guy (2004) and Clarke and Merlin (2013).

Appendix 2: Taxonomic bias and personality cults surrounding Linnaeus and Lamarck in the 18th–19th centuries

Carl Linnaeus (1707–1778) is the most famous naturalist of all time. “Linnaeus became the subject of hero worship after his death to an extent previously unknown in botany” (Stafleu, 1971a). Linnaeus’s impact on biological thought and his renown among peers is difficult to convey today (Blunt, 2001). Linnaeus’s four-point “revolution” developed in stages: the sexual system (Linnaeus, 1735), hierarchical taxonomic ranks (Linnaeus, 1735), generic reform as a set of theoretical axioms (Linnaeus, 1737a, and their practical application Linnaeus (1737b), and lastly binomial nomenclature (Linnaeus, 1753).

Linnaeus’s genius was soon recognized. He became a Full Professor at age 34. The King of Sweden knighted him at age 46. A generation of devoted students trained under Linnaeus, and he served as supervisor (“praeses”) for 186 Uppsala doctorates. Seventeen of his most committed scholars became known as the “apostles.” Thanks to Linnaeus’s connections with the Swedish East India Company, his apostles explored the world and spread Linnaean taxonomy. Seven apostles died on expedition. Five wrote about Cannabis (Hasselquist, Forsskål, Falk, Sparrman, Thunberg).

Linnaeus’s innovations competed with other taxonomic systems at that time, by Ray, Morison, Rivinus, Tournefort, Herman, Boerhaave, Ludwig, and Magnol. The Linnaean system was embraced quickly in Holland (where he published 14 of his early works), and of course Sweden. Next came Great Britain. English literature soon dominated botany—particularly concerning Cannabis in India—so the reasons why British botanists became the staunchest supporters of Linnaeus are worth reviewing:

Linnaeus visited England in 1736, and impressed four leading botanists—Hans Sloane, Philip Miller, Peter Collinson, and John Dillenius. Dillenius (1741) introduced Linnaeus’s work to English literature. Collinson organized the election of Linnaeus as a Fellow of the Royal Society in 1754. He asked Linnaeus to send an apostle to London; Linnaeus sent Daniel Solander, who gained a post at the British Museum. Solander soon proselytized Joseph Banks, and they botanized together with James Cook on the Endeavour.

Stillingfleet (1759) translated into English six essays from Linnaeus’s “unrivalled school of natural history.” Coxe (1811) referred to Stillingfleet as “one of the body-guards of Linnaeus.” Stillingfleet encouraged Hudson (1762) to organize Flora Anglica along Linnaean lines, which “marks the establishment of Linnean principles of botany in England” (Smith, 1824). Philip Miller, the influential author of Gardeners Dictionary and keeper of the Chelsea Physic Garden, announced that he “adopted in a great measure the system of Linnaeus” (Miller, 1768). James Edward Smith founded the Linnaean Society of London in 1788. Four years earlier, Smith had purchased Linnaeus’s herbarium and library from his widow. Smith (1821) published a tome of Linnaeus’s correspondence, and cemented London as the center of Linnaean studies.

The anti-Linnaeans

Stafleu (1971a) states that Linnaeus’s authoritative pragmatism appealed strongly to the practical and bourgeois British, whereas French philosopher-naturalists, under sway of the Enlightenment, considered Linnaeus an authoritarian of the past.

Comte de Buffon (1707–1788) was a French naturalist and a proponent of the Enlightenment. Buffon became Linnaeus’s primary scientific rival. He criticized three aspects of Linnaeus’s four-point revolution. Buffon mocked Linnaeus’s sexual system for being artificial. He sharply criticized Linnaeus for overthrowing Tournefort’s generic reform based on arbitrary principles.

Linnaeus’s hierarchical system of taxonomic ranks particularly rankled compte de Buffon (1749). He was under the influence of John Locke, an Enlightenment author, who argued against essentialism in favor of nominalism. Linnaeus’s taxonomic system was essentialist to its core—he treated species as typological and immutable entities created by God. Buffon argued that we name a creature by its collection of observed qualities (nominalism), rather than classify a creature by its similarity to an eternal essence or prototype. Buffon repeated the argument in Locke (1690) that individuals can be observed, but “species” and “genera” are unobservable abstractions, unless they consist of a single individual.

Buffon clashed with Linnaeus over “species concepts”—an argument that continues today regarding Cannabis. Linnaeus saw sharp-cut delineations between species. Buffon argued that “Nature progresses by unknown gradations,” and organisms were “not tied up in neat and orderly parcels.” compte de Buffon (1749) asserted, “In Nature there actually exist only individuals; genera and orders and classes exist only in our imaginations.” Subsequently, compte de Buffon (1753) recognized species, as a group that could interbreed, “a constant succession of similar individuals that can reproduce together.” This concept makes him the founder of the biological species concept, usually attributed to Mayr (1942).

Jean-Baptiste Lamarck (1744–1829) was a protégé of Buffon. Lamarck’s supporters hailed him as “the French Linnaeus.” He deviated from Linnaean orthodoxy in many ways. Lamarck (1778) published a flora that departed from Linnaeus’s approach in two ways—it was written in vernacular French instead of Latin, and it used dichotomous keys for plant identification. Subsequently, Lamarck (1783) launched a new format for writing flora texts. He devised the new format as an alternative to the telegraphic Linnaean style. The format remains in use today (Frodin, 2001).

Lamarck criticized three aspects of Linnaeus’s four-point revolution. Lamarck (1783) critiqued Linnaeus’s systéme sexuel as defective. Lamarck (1788) presented a lengthy critique of Linnaeus’s generic reform, “the so-called axioms and extremely laconic maxims with which he filled his Philosophia botanica and Critica botanica, rather than by solid proof which alone could convince those who were not impressed by mere authority.” Lamarck (1778) repeated Buffon’s nominalist stance that all hierarchical ranks above species were subjective creations of the human mind. He argued that the local European flora was too riddled with species gaps to enable a valid perception of higher ranks. Lamarck (1791) again asserted that higher ranks (Class, Family, Genera) were surely not the work of Nature, but artificial, the result of arbitrary human judgment. Lamarck (1792) enlarged upon this critique as a response to pushback from Linnaeans.

Lamarck’s theory of evolution totally alienated Linnaeans. Lamarck (1809) recognized evolution as a gradual process. This made it difficult to find well-established, concrete lines of separation between taxonomic groups. In contrast, Linnaeus treated species as entities with fixed forms given to them by God. “Every genus is natural, created as such in the beginning, hence not to be rashly split up or stuck together by whim or according to anyone’s theory” (Linnaeus, 1735). “We count as many species as there were forms created in the beginning” (Linnaeus, 1751). This dispute continues to echo in the modern quarrel between creationists and evolutionists.

The anti-Lamarckians

Despite Lamarck’s departures from Linnaean orthodoxy, he praised Linnaeus as the greatest botanist in history. Late in his career he even regarded Linnaeus’s artificial classification system as a useful device for identifying plants (Lamarck & Mirbel, 1803). Nevertheless, Lamarck’s criticisms “were neither forgotten nor forgiven” by other botanists (Williams, 2001). “Those who had enjoyed Buffon’s support and patronage, and those who accepted his model of natural history, underwent several institutional setbacks and bore the brunt of a massive theoretical offensive” (Corsi, 1988).

In Lamarck among the Anglos, Hull (1984) wrote about Lamarck’s critics. “Few British or American scientists got their views of Lamarck by reading translations of Lamarck, let alone Lamarck’s original works in French. In most instances, Lamarck’s reputation was formed by descriptions of his views in secondary sources.” Primary among those secondary sources was Georges Cuvier, an opponent of Buffon. Cuvier (1836) wrote a biting, ironic eulogy of Lamarck that was translated into English.

Carl Ludwig Willendow, a disciple of Linnaeus, updated Species Plantarum after Linnaeus died. Willendow did his best to dismantle Lamarckian taxonomy and nomenclature. A perusal of volume one (in two parts) reveals that Willdenow accepted 55 of Lamarck’s taxa, but rejected 61 as synonyms. Willdenow synonymized Lamarck’s taxa for the flimsiest reasons. Sometimes he just annotated a question mark after them (e.g., Lobelia nummularia). Willdenow reduced Coffea mauritiana to a variety of Coffea arabica, “it seems to me hardly different.” He rejected several Eriocaulon species because they weren’t illustrated. He rejected Fagara heterophylla because “it seems kind of mixed.” He rejected Jasione perennis because it was “an imperfectly dried specimen.”

Willdenow (1805) rejected Lamarck’s C. indica on one morphological character—Lamarck described C. indica with alternate branching. Because C. sativa also showed alternate branching, Willdenow argued no differences existed between them. He ignored Lamarck’s other morphological and phytochemical differences. Willdenow found one distinction in C. indica, “leaflets have a more tapered base, which is known to be very inconsistent.” Willdenow incorporated C. indica into C. sativa, rather than name it a separate variety.

Willdenow’s six-volume Species Plantarum was the most influential botanical work of its day, although it came under “severe criticisms for many manifest errors” (Long, 1843). Wight and Walker-Arnott (1834) initiated the process of “clearing up many doubtful synonyms in Willdenow,” although they neglected C. indica. Botanists have reinstated nearly half of Willdenow’s 61 rejected species names (as well as two genera,

Canthium and Nuxia), many reinstated as basionyms.

British botanists in India displayed a strong Linnaean bias. Their opinions were shaped by Johann König, a pupil of Linnaeus. König organized a botanical society in India, the “United Brotherhood,” which included two influential British botanists, Fleming and Roxburgh. Fleming (1810) wrote, “Dela Marck [de Lamarck] is of the opinion that the Indian Gánja is a different species of Cannabis from the Cannabis sativa. But Willdenow assures us that…he could not perceive any difference between them.” Roxburgh (1832) confirmed, “I perfectly agree with Willdenow in thinking all the varieties, if even such they can be called, centre in one species.” In Flora Indica, Hooker and Thomson (1855) wrote about “bhang and chirris” obtained from Cannabis sativa. “With regard to nomenclature, we shall not alter names established by Linnaeus.” They degregate “hair-splitters,” and criticize Lamarck’s theory of evolution.

Lamarck’s phytochemical character—C. indica caused intoxication—was dismissed by British botanists. Linnaeans adhered to Linnaeus (1751), who rejected phytochemical characters, such as fragrance and taste. Lacking knowledge of Mendelian inheritance, most Linnaeans believed that phytochemical traits depended entirely upon the environment—either climate or cultivation. Fiber-type plants “degenerated” into drug-type plants, and vice-versa, depending on how and where they were grown.

O’Shaughnessy (1838–1840) stated, “difference of climate seems to me more than sufficient to account for the absence of the resinous secretion, and consequent want of narcotic power… the Cannabis sativa and Indica are identical.” A cultivation argument was first put forward by Fleming (1801). He believed phenotypic variation was due to crop spacing—plants in Bengal were grown very far apart, often nine or ten feet, compared to closely-sown European hemp fields. Royle (1840) attributed differences between Indian hemp and European hemp to “openness of planting.” The Indian method of growing plants 9–10 ft apart caused a loss in fiber softness and flexibility, and gave rise to “a full secretion of the principles.”

Appendix 3: Details regarding alphanumerical sites in Fig. 2

Figure 2 illustrates the distribution of plants that field botanists in the 18th–19th centuries identified as C. sativa, C. indica, or other Cannabis species. Many more naturalists wrote about “hemp,” but did not assign a Latin name to the plants.

Our narrative is divided in two parts: Part A . Plants assigned the name Cannabis sativa and segregates, and Part B . Plants assigned the name Cannabis indica and segregates. Within each part, the narrative is organized by floristic regions.

The map in Fig. 2 shows boundaries of ten floristic regions by Good (1964) and Takhtajan (1986): ❶ Euro-Siberian region, west; ❷ Euro-Siberian region, east; ❸ Mediterranean region; ❹ West and Central Asia (Irano-Turanian); ❺ Sino-Japanese (East Asiatic); ❻ African-Indian desert (Saharo-Arabian); ❼ Sudano-Zambezian-Sindhia; ❽ South Asia (Indian); ❾ Continental Southeast Asia (Indochina). ❿ Malayian (Malesian).

Part A . Cannabis sativa and segregates

❶ Euro-Siberian region, west:

Western Europe was blanketed by botanists who wrote scores of local floras. Many mentioned C. sativa growing wild, ruderal, or spontaneous. For this region, we present a geographic sample, rather than an exhaustive list. Early mentions include England (s1, Hill, 1760), France and Switzerland (s2, Lamarck, 1785), Germany (s3, Willdenow, 1787), Bohemia (s4, Presl & Presel, 1819), Poland (s5, Wulff, 1765), Hungary (s6, Heuffel, 1858), Galicia (s7, Zawadzki, 1835), Lithuania and Volhynia (s8, Eichwald, 1830), and the Moscow region (s9, Stephan, 1792).

Catherine the Great commissioned six physician-botanists to explore her new dominions—Pallas, Lepechin, Falck, Georgi, Gmelin, and Güldenstädt. Three of them studied under Linnaeaus (Lepechin, Falck, Georgi); Gmelin corresponded with him; Güldenstädt had no connection with him; and Pallas was anti-Linnaean—Pallas (1766) refuted Linnaeus’s scala naturae, and Pallas (1774) laid out an animal classification meant to compete with Linnaeus’s Systema Naturae. Güldenstädt and Pallas departed from Linnaean orthodoxy by coining new Cannabis species.

Pallas, Falck, and Lepechin wrote about wilder hanf near Chetverosvyatsky monastery on the Volga, about 10 km downriver from Saratov (s10). A stone’s throw from the monastery laid the ruins of Uvek, a trading center of the Tartars (i.e., Kipchak Khanate) until 1395. Pallas (1793) said wilder hanf looked like cultivated plants, and he attributed its presence to former inhabitants—the Tartars. Lepechin (1774) said the plants were feral escapes of crops sown by the Tartars. Falck (1786) assigned the plants to C. sativa, but noted a wild-type characteristic, “…only branched stems. These give no straight and uniform fibers.”

Falck (1786) assigned C. sativa to feral plants along the Terek River in Northern Caucasus (s11). Georgi (1800) wrote of C. sativa, “We have indigenous or self-growing hemp in the Taurus mountains (s11), on the Terek (s11), along the Don, and the Dnieper in Novorossiya (s12).” Güldenstädt (1791) encountered wilder hanf in Ukraine and referred to it as “Cannabis” or a new name, Cannabis vulgaris (×1).

Russian exploration continued into the nineteenth century. Marschall von Bieberstein (1808) wrote a flora of the Caucasus and Crimea, and listed ruderal C. sativa, without naming a specific location (s13). Meyer (1831) reported C. sativa growing wild around villages below Mount Elbrus (s11). von Besser (1822) said C. sativa grew spontaneously near Kremenets in northwest Ukraine (s14).

Botanists started using variety names: Czernajew (1859) named wild hemp C. sativa var. spontanea, near Kharkiv in northeast Ukraine (s15). Lindemann (1881) applied Czernajew’s taxon to wild hemp growing near Kherson (s12). Korshinskii (1898) applied C. sativa var. vulgaris to wild-type plants in southern Russia (s16). Vavilov and Janischevsky at Saratov (s10), assigned two taxa to the same population of wild-type plants: C. sativa var. spontanea (Vavilov, 1922) and C. sativa var. ruderalis (Janischevsky, 1924).

❷ Euro-Siberian region, east:

Daniel Messerschmidt and Johann Gmelin explored Siberia, 1720–1727 and 1733–1743, respectively. Messerschmidt, not a Linnaean, coined a prolix polynomial for plants in Transbaikal, east of Lake Baikal: “Cannabis erratica, montana, procera, daurica, folio minore semine lupulino similes, paruulo, guttato” (×2) (Amman, 1739). Gmelin (1768) was a Linnaean, and used Linnaeus’s pre-1753 name, Cannabis foliis digitatis, for wild hemp between Krasnoyarsk and Irkutsk (s17).

Falck (1786) assigned C. sativa to feral plants in Bashkiria (s18). Georgi (1800) wrote of C. sativa, “We have indigenous or self-growing hemp… in the Urals (s19), and the Ufa (s18), in Siberia between the rivers Yenisei and Irkutsk (s17).”

Pallas (1776) used Messerschmidt’s polynomial for “a small wild hemp, which has a strange appearance” near Kyakhta (×2). Sievers (1796) assigned Cannabis erratica to wild-type plants growing in Transbaikal near Kyakhta (×2). Ledebour (1847–49) synonymized C. erratica by Messerschmidt and Sievers under a new varietal name, but tentatively, with a question mark: Cannabis sativa β? davurica. Turczaninow (1856) synonymized Messerschmidt’s taxon under C. sativa. Herder (1892) synonymized C. sativa β davurica and C. erratica under C. sativa.

von Ledebour et al. (1833) assigned C. sativa to wild-type plants growing along the Bukhtarma River in the eastern Kazakh steppe (s20) and along the Ulagan River in the Altai Mountains (s21). Nearly a century later, Sinskaya (1925) discovered a full spectrum of Cannabis morphologies in the Altai, from wild-type forms to fully domesticated plants (s21).

❸ Mediterranean region:

The headwaters region of the Po River in Italy is known as Piemonte (in Italian) and Piémont (in French). Lamarck (1785) described C. sativa as presque naturalisée in Piémont (s22). Other botanists assigned new names to a unique landrace in that region. D’Andrieux (1771) coined Cannabis sativa gigantea for chanvre de Piémont (×3). His son-in-law Vilmorin added cachet to chanvre du Piémont by giving it a new binomial name, C. maxima (×3) (Sénac, 1826). Rey (1835) imported chanvre du Piémont from Carmagnola, and coined the taxon C. gigantea (×3). Vilmorin (1837) countered Rey’s new binomial by coining his own, C. gigantea (×3). His son, Vilmorin (1892) assigned C. sativa excelsior to chanvre du Piémont (×3).

Elsewhere in Italy, Parlatore (1867) simply used C. sativa for feral hemp (s23). Velenovský (1898) described spontaneous C. sativa in Bulgaria (s24). Across the Mediterranean in Algeria, Guyon (1842) wrote about hhachiche. He knew Lamarck’s C. indica from the literature, but stated that the plants in Algeria were “none other than our common hemp, C. sativa” (s25). According to de Candolle (1869), Friedrich Noé, a German botanist working in Istanbul (s26), used the name Cannabis orientalis for plants in his herbarium collection, but did not publish the taxon. Noé’s collaborator Boissier (1879) reduced C. orientalis (along with C. indica and C. chinensis) to a synonym of C. sativa.

❹ West and Central Asia (Irano-Turanian):

Samuel Gmelin corresponded with Linneaus, like his aforementioned uncle. He collected C. sativa near Lankarān and Astārā (s27) in present-day Azerbaijan (Herder, 1892). Hohenacker (1838) also collected ruderal specimens of C. sativa in Azerbaijan, close to Lankarān and Astara on the border with Persia (s27).

Falck (1786) assigned C. sativa to feral plants in “Bukhārā” (the closest he actually got to Bukhārā was Orsk, s28) and “Soongaria” (a.k.a., Dzungaria—Falck’s assistant Bardanes skirted the edge of Dzungaria in Semey, s29). Becker (1873) collected wild C. sativa near Kasumkent in Dagestan.

Chardin (1811), or rather his editor, assigned the taxon C. sativa to bueng plants in Isfahān, Persia (s30). Claus (1838) reported “abundant” C. sativa on islands dotting the Volga River delta (s31). Karelin and Kirilov (1841) assigned C. sativa to plants “in pratensibus ad fl. Irtysch frequens [frequently in meadows near Irtysh River, s29], nec non in deserti Soongoro-Kirghisici [and certainly on the Kirghiz steppe, s32], arenosis ad lacum Noor-Saissan [growing on sand near Lake Zaysan, s29].”

Schrenk (in Trautvettero, 1867) collected C. sativa at three places in eastern Kazakhstan in 1840–1841: on Lake Alakol near the Dzungarian Gate (s32); the Kyskatsch mountains, between Lake Balkhash and Lake Zaysan (between s29 and s32); and near Khantau, between Lake Balkhash and Bishkek (s33). Griffith (1847) found Cannabis (no species name) at Jegdalek between Kabul and Jalalabad in Afghanistan (s34).

Basiner (1848) assigned C. sativa to a plant called kender in Khiva (s35). Bunge (1851) collected C. sativa near Kulagin on the Ural River, just north of the Caspian Sea (north of s31). Becker (in Herder, 1892) collected wild-type C. sativa along the lower Volga near Sarepta, Bogdo, Saratov, and Astrakhan (between s10 and s31). Semenov (1998) reported wild C. sativa growing near Kurmenty Pass northeast of Lake Issyk-Kul in 1857 (just east of s33). Seminov wasn’t the best botanist—at herb. LE we saw a specimen he called “C. sativa” that was actually Humulus lupulus.

Valikhanov (1865) said hashīsh was extracted from C. sativa in Kāšḡar (s36). Shaw (1880) wrote a Turkic-English dictionary in Kāšḡar, and translated kaindir as the hemp plant, C. sativa (s36). Aitchison (1888) applied the name C. sativa to plants at Rui-Kauf in Persia (s37). Meyer (in USDA, 1912) collected C. sativa at Sanju oasis (s38), “a small-seeded variety of hemp… used for seed oil and hashish made from the young tops.”

❺ Sino-Japanese (East Asiatic)

Thunberg (1784), a student of Linnaeus, extended the C. sativa taxon to Nagasaki, Japan (s39). von Bunge (1833) found C. sativa growing “quasi-feral” in northern China (s40). Martius (1832) gave the name Cannabis sativa gigantea to plants growing 6 m tall in China. He did not give a location, but Canton (i.e., Guǎngzhōu, s41) was the only Chinese port open to foreigners at that time. Bretschneider (1882), a Russian physician in Běijīng (s42), assigned Chinese hemp to C. sativa.

❻ African-Indian desert (Saharo-Arabian)

Forsskål (1775), a student of Linneaus, assigned C. sativa to plants in Egypt (s43). Gastinel (1849) did not recognize C. indica as a species separate from C. sativa in Cairo (s43).

❼ Sudano-Zambezian-Sindhia

Griffith (1847) found “Cannabis” (no species name) in places that later became famous for charas: Dādur Nāla in Sindh (s44), and Burhan in Punjab (s45). Jacquemont (1861) assigned C. sativa to plants in Rajputana and Punjab (between s44 and s45). Aitchison (1864) described C. sativa in Punjab. Duthie (1898) reported wild C. sativa growing in three Chitral locations—Drosh, Dir, and Mirga (s46).

❽ South Asia (Indian)

Many botanists who knew Lamarck’s taxon nevertheless adhered to the Linnaean party line, and applied C. sativa to plants throughout India—in Bengal (s47, Fischer, 1810, Roxburgh, 1832, Butter, 1839), in southern India (s48, Buchanan, 1807), and in the Himalaya (s49, Hardwicke, 1801, Royle, 1839, Hooker and Thomson, 1855). Wallich (1828–1849) applied C. sativa to plants all over South Asia—Calcutta (Bengal), Sudallapur (Karnataka), Bareilly (Uttar Pradesh), Hyderbad (Andhra Pradesh), and Kathmandu (Nepal).

❾ Continental Southeast Asia (Indochina)

de Loureiro (1790) assigned C. sativa to plants cultivated for fiber and drugs in “Cochinchina,” southern Vietnam (s50). Wallich (1828–1849) collected C. sativa in the Toong Dong mountains near Innwa, Burma (s51). Griffith (1847) reported “C. sativa is found here” in Sassi (south of s51). Crévost (1917) coined the taxon Cannabis gigantea for plants grown for fiber and seed oil in the Tonkin highlands (×5).

❿ Malayian (Malesian)

Stickman (1754), a student of Linnaeus, assigned C. sativa to drug plants cultivated in the Moluccas islands, Indonesia (s52). Thunberg (1796), another student, wrote about C. sativa, known as ginje in Jakarta (3000 km west of s52).

Part B . Cannabis indica and segregates

❶ Euro-Siberian region, west

No plants native to this region were named C. indica by field botanists.

❷ Euro-Siberian region, east

No plants native to this region were named C. indica by field botanists.

❸ Mediterranean region

von Maltzan (1869) said hashīsh and kif in Algiers was made from C. indica (i1), whereas De Courtive (1848) and Dukerley (1866) equivocated between C. indica and C. sativa for plants from Algiers. Mongeri (1865) described C. indica cultivation at Izmit and Bursa, south of Istanbul (i2).

❹ West and Central Asia (Irano-Turanian)

Regel (1880) identified plants growing near Lake Issyk-Kul as a new variety, C. sativa γ asperrima (i3). Polak (1865) assigned C. indica to medicinal plants at Tehran (i4). Vámbéry (1868) described beng or bengis as “the poison produced from the Cannabis indica” in Bukhārā and Khoqand (i5).

Henderson and Hume (1873) assigned C. indica to drug-type plants cultivated in Yarkand (i6). However, at the Kew Herbarium, a Yarkand specimen collected by Henderson is labeled “Cannabis sinensis.” Vavilov and Bukinich (1929) found feral and wild Cannabis in Afghanistan, and named them C. indica f. afghanica and C. indica var. kafirstanica, respectively (i7).

❺ Sino-Japanese (East Asiatic)

Fischer (1810) erected a new taxon for Chinese hemp, Cannabis chinensis. He did not state provenance for the germplasm, but Canton (i.e., Guǎngzhōu, ×6) was the only Chinese port open to foreigners at that time. von Humboldt (1811) considered Chinese hemp a type of C. indica (i7). von Siebold (1827) described C. sativa b indica in Japan (i8). Boitard (1839) suggested that Chinese hemp was “probably the same species as Indian bangue.”

Tatarinov (1858), a Russian physician in Běijīng, assigned C. indica to local medical plants (i9). Hedde (1848) assigned C. indica to germplasm obtained from Shanti district, Guǎngzhōu (i7). Itier (1846) assigned C. indica to germplasm obtained from Huángtián, 170 km N.E. of Guǎngzhōu (i7).

Itier gave germplasm to Delile (1849), who grew it in Montpellier. Delile was unaware of Fisher’s publication, and re-coined Cannabis chinensis (×6). Koch (1854) noted that C. chinensis resembled drug plants from India. Vilmorin (1851) gave Chinese hemp a new name, “Cannabis gigantea Delile.” This Latin name was an unfortunate choice, because it had previous been applied to chanvre de Piémont (×3) (Rey, 1835; Vilmorin, 1837).

Jomard (1852) wrote about Chinese hemp, to which he assigned Vilmorin’s taxon C. gigantea. Itier (1853) also reassigned lo-má to C. gigantea. Heuzé (1860) synonymized Vilmorin’s taxon with “Cannabis indica Lam.” Alefeld (1866) assigned Cannabis sativa gigantea to “Chinese, Oberländer, or Piedmontese giant hemp”—in other words, to both C. gigantea Vilmorin, 1837 (Piedmontese hemp) and C. gigantea Delile ex Vilmorin, 1851 (Chinese hemp).

Pabst (1887) erected C. sativa var. chinensis, and described plants as “rich and protruding branches, 3-6 m high.” Hoffmann (1944) coined Cannabis var. indica subvar. Gigantea for Chinese hemp, indicating its kinship with indica. Zhang (1990) noted relatively high levels of THC in some Yúnnán landraces, which he named Cannabis sativa ssp. indica var. yunnanica (i10). These landraces may trace back to Sayyid Ajjal Šams al-Din ʿOmar (1211–1279), a Muslim from Bukhārā, appointed as Yúnnán provincial governor by Kublai Khan. Hillig (2005) and Clarke and Merlin (2013) considered Chinese hemp a biotype of indica.

❻ African-Indian desert (Saharo-Arabian)

Ibn al-Baitār, a Spanish Moor who moved to Cairo around 1240, identified qinnab hindī (Indian hemp) growing in Egypt (i11), distinct from plants he knew in Spain, qinnab and qinnab barrī (Ibn al-Baitār, 1985). Godard (1867) and Mackenzie (1893) describe hashīsh made from C. indica in Egypt (i11).

Hasselquist (1766), a student of Linnaeus, broke from Linnaean orthodoxy and assigned Cannabis vulgaris to plants cultivated in Palestine for chashis (×4). Stokes (1812) synonymized Hasselquist’s taxon under C. indica (i12). Mongeri (1865) described C. indica cultivation at Mosul in present-day Iraq (i13).

❼ Sudano-Zambezian-Sindhia

von Maltzan (1873) lived in Aden, Yemen, where he encountered hashīsh made from C. indica (i14). Honigberger (1852) identified plants in Punjab and Kashmir as C. indica, due to their unique chemistry compared to C. sativa (i15).

❽ South Asia (Indian)

Rheede (1690) gave the names Kalengi-cansjava and Tsjeru-cansjava to male and female Cannabis, respectively, at Kochi on the Malabar coast (×7). O’Shaughnessy (1838–1840) employed the taxon C. indica, yet he considered “the Cannabis sativa and indica are identical” (i16). Kerr (1877) a British-trained Bengali botanist, referred to Bengal gañjā as C. indica, distinct from C. sativa (i16).

Johann Friedrich Metz was a German missionary who collected plants on the Malabar coast (i17). His collection of “Cannabis indica” at herb. LE is a dense, lightly-seeded bud—the oldest specimen of manicured gañjā we have seen. Rottler (1836–7), a Danish missionary at Tranquebar and Madras (i17), translated Tamil kañcam as Cannabis indica. A few botanists in the Himalaya referred to local plants as C. indica (i18, Cleghorn, 1866, Aitchison, 1869, Lawrence, 1895).

❾ Continental Southeast Asia (Indochina)

Hedde (1848) assigned C. indica to germplasm obtained from Tourane, now Đà Nẵng, Vietnam (i19).

❿ Malayian (Malesian)

Rumpf (1747) assigned C. indica to drug plants cultivated in the Moluccas islands, Indonesia (i20). Blume (1825), a German botanist who moved to Jakarta, used the taxon Cannabis sativa var. indica for plants with the local name ginji (3000 km west of i20).

Appendix 4: Taxonomic models in the 20th-21st century

The “species debate” became a cause célèbre in 1970s court cases. The primary adversaries in this debate were Richard Evans Schultes (1915–2001), a Harvard ethnobotanist, and Ernest Small (1940-), a Canadian taxonomist who specialized in Cannabis. Their taxonomic debates in courtrooms often involved “an appreciable irrational emotive component” (Small, 1979). One taxonomic paper from that era made reference to “the Honorable Harry Anslinger,” and the “menace” posed by Cannabis (Quimby et al., 1973).

Schultes’s taxonomic work is limited to two publications. Initially, Schultes (1970) considered Cannabis a monotypic genus. Subsequently, Schultes et al. (1974) recognized C. sativa, C. indica, and C. ruderalis. Schultes changed his opinion based on an analysis of herbarium specimens, a survey of the Mississippi Cannabis plantation, and “preliminary field work” in Afghanistan.

Schultes collaborated with Loren C. Anderson at Florida State University. Anderson published two papers. Anderson (1974) compared wood anatomy in C. sativa (a feral hemp plant in Kansas, likely of Chinese provenance) and C. indica (a plant that Schultes collected in Afghanistan). Anderson (1980) compared leaf variation in four populations of Cannabis: 1. C. sativa (fiber varieties—tall, laxly branched plants), 2. C. sativa SS (Small-Seeded drug plants from India and Pakistan—relatively tall, with narrow leaflets), 3. C. indica (drug plants from Afghanistan—short, densely branched, with broad leaflets), and 4. C. ruderalis (wild plants, probably from Central Asia—very short and unbranched). Anderson illustrated his species concept in a line drawing (Fig. 5).

Small began Cannabis taxonomic research in 1971, under common-garden conditions on a three-acre farm near Ottawa, Canada. He grew nearly 400 Cannabis accessions obtained from the USDA, the UN, the University of Mississippi, and from botanical gardens around the world. Small and colleagues conducted cross-breeding experiments (Small, 1972, 1984). They assessed many taxonomic characters: cannabinoid content (Small & Beckstead, 1973a, b; Small et al., 1975; Small & Marcus, 2003), seed chemistry (Small et al., 1976), morphological variation in seeds (Small 1975), vegetative characters (Small et al., 1976), flower characters (Small & Naraine, 2015a, b), developmental characters (Small et al., 2003), and biogeographical aspects (Small, 2015a, b).

This work was summarized (Small and Cronquist, 1976), and elaboratored in a two-volume text (Small, 1979). His work continues to evolve—Small (2015a) has taken on aspects of Hillig’s work regarding Chinese hemp. Small’s classification and nomenclature is widely accepted around the world (e.g., Hanelt, 2001; Yang, 2003; Kojoma et al., 2006; Mukherjee et al., 2008; Mabberley, 2008; Shipunov, 2010; Chandra et al., 2013). Small’s taxonomic scheme is attested on high-profile websites. Examples:

• http://frps.eflora.cn, ru.wikipedia.org/wiki

• http://plants.usda.gov/java

• www.ars-grin.gov

• www.theplantlist.org

• www.tropicos.org

• www.ncbi.nlm.nih.gov/taxonomy

Google Scholar (https://scholar.google.com) computes citation metrics for authors and their publications. We applied the “advanced search” algorithm, using the words “taxonomy,” “cannabis,” and author name. After screening for appropriate hits, we obtained the following results: Small: 18 publications, cited collectively by 675; Hillig: 7 publications, cited collectively by 299; de Meijer: 5 publications, cited collectively by 207; Schultes: 2 publications, cited collectively by 193; Gilmore: 5 publications, cited collectively by 179. No other authors were cited more than 150 times.

Hillig and colleagues cultivated 157 accessions under common-garden conditions at Indiana University. They obtained accessions from Small, the CPRO, Dutch seed banks, the Vavilov Institute, law enforcement agencies, field stations, and botanical gardens across Eurasia. Hillig analyzed genetic characters (Hillig, 2004a, 2005), cannabinoid profiles (Hillig & Mahlberg, 2004), terpenoid profiles Hillig, 2004b, seed morphology (Hillig, 2005b, c), leaflet characters (Hillig, 2005b, c), stalk morphology (Hillig, 2005b, c), developmental characters (Hillig, 2005b, c), host-parasite characters (Hillig, 2005b, c, McPartland & Hillig, 2003, 2004, 2006), and biogeographical aspects (Hillig, 2005a, b).

De Meijer and colleagues systematically investigated Cannabis under common-garden conditions at Wageningen University. They collected over 150 Cannabis accessions—now the CPRO collection. They obtained germplasm from the Vavilov Institute, IPK-Gatersleben Institute, Dutch seed banks, European breeders of fiber-type plants, and botanical gardens across Eurasia. De Meijer has continued Cannabis studies at GW Pharmaceuticals in the United Kingdom. De Meijer analyzed cannabinoid content (De Meijer & van Soest, 1992; De Meijer & Keizer, 1996), the inheritance of cannabinoid phenotypes (De Meijer et al., 2003; Mandolino et al., 2003; De Meijer & Hammond, 2005. De Meijer et al., 2009a, b), stalk morphology and chemistry (De Meijer, 1993a, 1994b, 1995), seed morphology and chemistry De Meijer & Keizer, 1996, leaflet characters (De Meijer et al., 1992; De Meijer & Keizer, 1996), host-parasite relationships (De Meijer, 1993b; De Meijer & Keizer, 1996), developmental characters (de Meijer & Keizer1996), sequence heterogeneity in THCA-synthase genes (Onofri et al., 2015), and biogeographical aspects (De Meijer, 1999, De Meijer, 2004, 2014).

Gilmore and colleagues researched Cannabis genetics at Australian National University. Their early work analyzed police-confiscated materials of unknown origin, lacking taxonomic inferences. They developed primers for microsatellite DNA loci, also called short tandem repeats (STRs), and validated their work for forensic purposes (Gilmore & Peakall, 2003; Gilmore et al., 2003, 2007; Howard et al., 2008, Howard et al., 2009). Gilmore et al. (2007) switched to polymorphic cpDNA and mtDNA loci, which they sequenced from known CPRO accessions.

“Strain” names

The International Code of Nomenclature for Cultivated Plants (ICNCP, Brickell, 2009) regulates the naming of plants whose recent evolution has been influenced by human selection. The ICNCP’s basic unit of classification is the cultivar, “an assemblage of plants that (a) has been selected for a particular character or combination of characters, (b) is distinct, uniform, and stable in these characters, and (c) when propagated by appropriate means, retains those characters.”

The ICNCP’s rules for naming and describing cultivars involve valid publication, typification (“Nomenclatural Standard”), and priority. The ICNCP is available online for consultation regarding these provisions.

Cultivar names meeting ICNCP provisions are placed in single quotation marks. The best known Cannabis cultivar, judging from the number of Google search hits, is ‘FINOLA’. The ICNCP also recognizes “Group” names, “All members of a Group must share the character(s) by which that Group is defined.” Small (2015a) proposed six Group names for cultivated kinds of Cannabis and their hybrids.

Article 2.2 in ICNCP stipulates that the words “variety,” “form,” and “strain” must not be used for the word “cultivar” (Brickell, 2009). Notwithstanding Article 2.2, some national and international plant registries use “variety” interchangeably with “cultivar.” The words “variety” and “form” also designate subspecies ranks in the ICN. “Strain” is not formally recognized by the ICN or ICNCP.

Segregates within “Sativa” and “Indica” are called “strains.” Watson (1985) coined several well-known “strain” names, such as “Original Haze”, “Skunk #1”, “California Orange”, “Afghani #1”, and “Early Girl”. Small (2015a) pointed out that strains are conceptually identical to cultivars, but almost no strains have met ICNCP requirements for cultivar recognition.

Clarke and Merlin (2015) took umbrage, and argued that “the bigger picture” justifies using “strain” and “cultivar” as equivalents. Small (2015b) retorted that they showed insufficient respect for codes of nomenclature, which hold pragmatic and moral status. “The botanical codes are by and large adhered to by scientists, commercial interests and editors because they provide stability and reliability to names that otherwise would result in confusion.”

“Strain” names have proliferated exponentially. Watson (1985) named 10 strains. Within 15 years, Dutch seed companies offered 150 strains for sale (Clarke, 2001). A decade later, the number of strain names reached 900 (Cannabis Strain Database, 2010). Most recently, Leafly (2015) listed 1535 strain names, and Seedfinder (2015) listed 6510 strain names. Doyle (2007) referred to strains as ganjanyms. Most strains are recognized hybrids, characterized as “Sativa-dominant” or “Indica-dominant.” In today’s largely illicit market, strain names are swapped and counterfeited, and generally unreliable (Lee, 2013; Sawler et al., 2015; Pierson, 2016).

The strain “AK-47” examples the arbitrariness of vernacular classification: “AK-47” won “best Sativa” in the 1999 Cannabis Cup, and won “best Indica” 4 years later. The ancestry of “Super Silver Sour Diesel Haze” offers a window into this warren of questionable pedigrees (Box 2).

Box 2 Putative ancestry of “Super Silver Sour Diesel Haze”, from Seedfinder (2015)