Endangered Varieties of subsp. indica: A Few Thoughts

According to Genesys, of some 1400 Cannabis accessions in global genebanks, as few as five are of subsp. indica.

Let that sink in…

A classification of endangered high-THC cannabis (Cannabis sativa subsp. indica) domesticates and their wild relatives‘ is a crucial new study from John McPartland and Ernest Small, published this April. Since Small’s Cannabis: A Complete Guide, this is the first authoritative work to address the dire situation for Cannabis biodiversity. Its focus is indigenous Central and South Asian varieties of subsp. indica. These face imminent extinction. Preservation in seed banks is urgently needed. McPartland and Small’s case is that classifying subsp. indica varieties is more than a mere dry academic exercisethat formal recognition can bring about decisive action.

Their new taxonomy rests on an analysis of 1100 specimens in 15 herbaria. It splits the old two-part Small and Cronquist lumper taxonomy for subsp. indica into the following four-part system:

  • C. sativa subsp. indica var. indica = True South Asian domesticates, informally “Sativas”
  • C. sativa subsp. indica var. himalayensis = Wild-type South Asian populations
  • C. sativa subsp. indica var. afghanica = True Central Asian domesticates, informally “Indicas”
  • C. sativa subsp. indica var. asperrima = Wild-type Central Asian populations

“Recognizing [these] infraspecific taxa helps to identify populations vulnerable to extinction”, explain the authors. Irreversible loss of biodiversity is occurring through introgressive hybridization (“the infiltration of genes between taxa through the bridge of F1 hybrids.”) Most critically in need of preservation are the wild-type varieties, var. himalayensis and asperrima.

Worth emphasizing if you’ve spent decades caught up in the Cannabis nomenclature nightmare is that the informal classifications Indica and Sativa can be used accurately to refer to some authentic Asian landraces, though not to describe modern hybrid slop. Notable also is that many landraces appear to have arisen as hybrids between var. indica and afghanica, not just in the centuries immediately prior to the Hippie Trail era but as far back as the thirteenth century and no doubt earlier.

Kennedy 1915

Turkestani var. afghanica collected in Xinjiang by Frank Meyer, a germplasm collector for the US Department of Agriculture, Nevada, c. 1912

I’d argue that McPartland and Small understate the extent to which Asian subsp. indica domesticates and wild-type populations are endangered. They relay a few anecdotal examples of introgression involving foreign seed being brought to historic centres of cultivation and biodiversity in the ’70s to ’90s. But they make no mention of the rise of Internet commerce, which today is the gravest danger. The shipping lists of the average Amsterdam or UK Internet ‘seed bank’ would make for horrifying reading now that online retail is arriving in South and Southeast Asia. The largest Amazon campus in the world has just opened in Hyderabad. An unprecedented influx of hybrid seed into Asia is already underway, and it’s only a matter of time before the corporate behemoths get their teeth sunk in to what, historically, was the world’s largest market for Cannabis products.

In essence, what we’re faced with now is Western hybrid slop surging eastward like a tidal wave, wiping out millennia of biodiversity permanently.

McPartland and Small are clear that their current study is provisionalan opening salvo. And as Small himself has emphasized, taxonomy is both a theoretical science and a practical artone that can be repeatedly revised. In that light, I’d note that I have questions concerning how precisely their new classification brings into focus certain crucial regions and populations. To my mind, their treatment of the Himalaya raises questions. Questions, primarily, because I’m not a botanist or a taxonomist. So back up a truck-load of salt with which to take what follows.

Himalaya: I suspect further splitting may be in orderin other words, that typical Himalayan domesticates could be classified as a separate taxon from var. indica. With that, I have doubts about var. himalayensis as defined in the current taxonomic key.

The shortcomings of the study as regards the Himalaya appear mostly to be due to available herbarium specimens not being adequately representative. That said, the historical and modern accounts of Himalayan Cannabis on which the study relies can be quite misleading. There’s a long-standing tendency to claim that Himalayan charas is harvested from wild plants. It can be, but these yield an inferior product known as “jungli”, which sells at a fraction of the price of charas from cultivated fields. An essential Victorian document, the Indian Hemp Drugs Commission (IHDC) is clear on this distinction.


Multipurpose Himalayan landraces, northern Kumaon, 2015. The region had experienced low rainfall during monsoon.

The study, in my view, misrepresents how the main Cannabis plants harvested in the Himalaya are undoubtedly landraces. These are markedly distinct from the var. indica domesticates of the rest of India, meaning the region south of the Himalaya, excluding Punjab. Wild varieties of some species can be sown as a crop. But typical cultivated Himalayan Cannabis plants are certainly domesticates, albeit part of cropweedwild complexes. As McPartland and Small statethough of their wild-type variety, var. himalayensisthey are harvested for “bast fiber (stalks), bhāng (leaves), hand-rubbed charas (hashīsh), or achenes (seeds).” More accurate would be “and seeds”. Nothing is wasted in the alpine Himalaya, because nothing can be wasted. Large leaves unsuitable for bhang or charas production are often used as a vegetable (“sabzi”).

But this is about more than mere traditions of use. In Kumaonwhich the IHDC rightly takes to be representative of Himalayan Cannabis botany, culture, and economies—typical seeds are markedly large. And eastward through Nepal to Kathmandu, their conspicuous size is evident. Seeds of landraces from the Dhorpatan region of Midwest Nepal are some of the largest I have seen, like marrow fat peas, comparable to those of Chinese hemp landraces. This is undoubtedly evidence of domestication for use as food. Cannabis seeds are an essential source of protein and fat in the Himalayan diet, as is also the case in regions of the Hindu Kush such as Chitral.

Similarly, high levels of bast fibre (which, again, the study notes as a trait of var. himalayensis) surely relate to Himalayan landraces being selected over millennia for use as rope and textiles. Historically, the Khas (Pahari) castes of the alpine Himalaya—then, as now, the people who cultivate Cannabis—wore “hemp” clothing, which was taken as a mark of their low status in the excruciatingly stratified Hindu social system. In Pahari villages, the fibres continue to be a crucial source of rope for tethering livestock, though their use for textiles is being revived, largely driven by Western demand, and sometimes they are sold to brokers for industrial uses. Likewise, Himalayan domesticates are exceptionally tall, given good conditions hitting over five metres.

My guess—though it is nothing more than a guess—is that further study will show the typical ratio of THC to CBD in Himalayan landraces to be closer to that of Central Asian domesticates. Their terpene profiles may also mark them out from var. indica ganja landraces and Central Asian populations.

In other words, a combination of morphological (and phytochemical) characteristics could well be defined to separate Himalayan domesticates reliably as a distinct formal variety, as is necessary in conventional plant taxonomic identification keys.

In following the misleading “wild mountain Cannabis” trope, the study inadvertently distorts some of the better historical sources such as the IHDC. In Table S4, it’s claimed, wrongly, that the IHDC states Urgam, Almora, and Kullu are regions where charas is harvested from wild plants. In fact, the IHDC merely reports this of Urgam, a valley in the Garhwal Himalaya. The Victorians were probably correct on this point, but only in so far as some charas is harvested from jungli stands, as is the case throughout the Himalaya. I have visited Urgam twice, and the roadhead is a point from which cultivated charas from surrounding valleys is exported. Most Garhwali charas today is produced from domesticates, as it almost certainly was in the past. Importantly, the IHDC records that the bulk of charas from Almora and the rest of the Kumaon Himalaya (which sold at several times the price of that from Yarkand) was harvested from cultivated fields. For Kullu, then part of the Native States neighbouring the Punjab Himalaya, the IHDC states as follows:

“There is more or less evidence of cultivation of the [Cannabis] plant all over the Himalayan portion of the province, including the smaller Native States. The cultivation is in small patches. A report from Kulu in 1880 says: ‘Almost every house has a small patch near it, a long strip beside a hedge, or a small bed a few square yards long (sic) in area.’ The only other detail of cultivation furnished is that the season of growth is from April and May to October and November. It may be safely assumed that the method of cultivation does not materially differ from that practised in Kumaon, which has been fully described.”

Ostensible var. himalayensis traits such as high bast fibre appear to be signs of these wild-type plants being part of cropweedwild complexes. Essentially, across northern India from Manipur westward through the alpine Himalaya and north of the Ganges, all the way through Punjab and up into the Hindu Kush and Pamirs of Central Asia, Cannabis exists as one big cropweedwild complex.


Wild-growing var. himalayensis, northern Pakistan. Photo by Landrace Genetics, 2019.

If a pristine aboriginal var. himalayensis population is ever found that’s unaffected by domesticates, perhaps in Arunachal Pradesh or some remote corner of Burma, I strongly doubt it will show a markedly high ratio of THC to CBD or high levels of bast fibre.

My guess is that Himalayan landraces are representativein practice, if not necessarily geneticallyof a millennia-old Central Asian tradition of multipurpose use that long predates ganja, a comparatively recent South Asian innovation for which true var. indica landraces were domesticated.

Indian botanists have long wondered whether the Iranian-speaking Central Asian nomads known as the Scythians brought Cannabis cultivation to the Indian Himalaya. Based on accounts of the Pontic Scythians by Herodotus (c. 440 BCE), it seems likely that Scythian Cannabis was a multipurpose crop. It was certainly used for textiles (as fine as the best European linen, says Herodotus) and smoked. It’s highly likely it was also cultivated for food use. Current studies on Central Asia are revealing how nomadic cultures were inextricably involved in cultivation, and Scythians themselves spanned settled agricultural and urban communities such as Shache (Yarkand) to true nomads such as the Royal Scythians, a point this study overlooks. We now know that around 500 BCE the first unequivocal evidence of cultivation south of the Indus appears in the Uttarakhand Himalaya in the form of a sudden unprecedented rise in pollen at a retting lake in Garhwal. This is very likely to be connected to economic changes brought about by the Achaemenid conquest of northwest India under the Persian Emperor Darius I, who subdued the eastern Scythians in 518 BCE, brought them south to India as the core of his war machine in 513 BCE, and stationed them across the newly conquered territories as part of Persia’s satrapies.


Scythians—history’s original dope fiends and the inventors of cavalry warfare.

Whether production of fibre in Garhwal was merely answering demand from the Achaemenids, actively instigated by their Scythian military, or begun by other Central Asian Scythian-like tribes fleeing the new empire, we will never know. Perhaps it was none of these things. But notably, in Nepal today, some Khas (i.e., Pahari) communities claim that they are descended from Central Asian nomads, a subject I looked at in an earlier post here. Relatedly, all the Pahari dialects such as Garhwali, Kumaoni, and Nepali probably developed independently of other Indo-Iranian languages of lowland India from a separate northern branch of Khas Prakrits following the Achaemenid era.

Suffice to say, my guess is that the earliest Central Asian subsp. indica domesticates were multipurpose plants, and that var. indica and afghanicatrue Sativas and Indicasare comparitively recent specialized creations that have arisen, respectively, from the custom of smoking ganja with tobacco and the significantly older (probably pre-850 CE) Central Asian practice of sieving resin glands through silk to make charas.

Ganja & Bhang: If this is all getting too speculative for your tastes, then strap in, because what follows deals with the dating of ancient Indian texts, which is a notoriously imprecise and arbitrary business.

Based on what little circumstantial evidence is available, it would appear that true var. indica domesticates, i.e. ganja landraces, developed and diffused across South and Southeast Asia in conjunction with the popularization of tobacco smoking and cultivation. This would fit with the earliest date for the term ganja, which is the sixteenth century.

McPartland and Small have misread a study on Indian names for Cannabis by Gerrit Meulenbeld. They claim the Anandakanda, a treatise on alchemy, describes the rogueing process (i.e., the sinsemilla technique) through which ganja is produced. In fact, the footnotes indicate this description is from a twentieth-century commentary on the Anandakanda by Bhagwan Das. They cite G. K. Sharma, a botanist, who speculated the text may date from the tenth to thirteenth centuries.

But Meulenbelda scholar of Sanskirt and Indian medicineargues that the Anandakanda substantially postdates the thirteenth century and gives the presence of the term ganja as evidence for that. He states that “The name ganja appears for the first time in the sixteenth century, in the Bhaavapraakashaniganthu and Todara’s Aayurvedasaukhya.” Recent studies also indicate that a text by the Kashmiri scholar Narahari Pandit, in which the term ganja appears, very likely dates not to the thirteenth centuryas is claimed in the IHDCbut to the seventeeth century.

var indica IHDC

Bengali var. indica: ‘Ganja plant almost ready to cut, Naogaon’. IHDC, 16th February 1894

Traditional Asian terminology for subsp. indica is very fluid, which can confuse the uninitiated. Bhang can mean subsp. indica and all its products, whether charas, ganja, or leaf, and any preparation thereof. But more narrowly, it’s also a distinct type of cultigen employed for a specific product (in practice, crude seeded inflorescences and small leaves) such as is still produced under license in India at sites such as Hosphiarpur in the Punjab. From there, bhang is exported to regions such as Rajasthan. Historically, Punjab has only been a bhang producing region, in part due to Sikh culture. By contrast, to the east in regions such as Uttar Pradesh and Madhya Pradesh, production was focused on ganja, a more sophisticated crop and product, sown and harvested at different times, and requiring significant skill to raise. Importantly, while ganja can be called bhang, bhangin this narrow sensecannot be called ganja.

Ganja first appears in English in the work of a merchant, Thomas Bowrey (16591713), who also has the honour of providing us with among the first English-language accounts of getting stoned. Earlier colonial accounts of Cannabis such as Garcia de Orta’s from Goa, southwest India, mention only the term bhang and say nothing of smoking. Bowrey had extensive experience of Bengal, the Coromandel Coast, and Sumatra. Like Herodotus, he’s proved to be a strikingly reliable witness. This is to be expected. He was in Asia to make money through trading commodities. Unlike his recent editors, he understood that ganja and bhang are not merely distinct products but also (typically) come from distinct types of plant. Ganja, according to Bowrey, was from a smaller-leaved, more refined crop. Bhang plants had larger leaves and comparatively “grosse seed”. Crucially, he records that the only Cannabis crop then cultivated throughout Bengal and along the coast of southeast India was bhang. Ganja was an imported product and was shipped from “the Island Sumatra”, where Bowrey himself lived for several years at Achin (i.e., Aceh). According to contemporary Acehnese, Cannabis itself was introduced to Sumatra by Indians, who may well have dominated cultivation and export of ganja.


Illustration of Goan bhang by Garcia de Orta (1501–1568), published 1563

Bowrey tells us that “beinge of a more pleasant Operation, much addictinge to Venery, [ganja] is sold at five times the price [bhang] is.” It was often smoked mixed with tobacco, “a very Speedy way to be besotted.” It was also commonly consumed by chewing. A search for the term in the MonierWilliams Sanksrit dictionary yields gṛñjana (गृञ्जन), defined as “the tops of [Cannabis] chewed to produce an inebriating effect.” Conceivably, the sinsemilla technique may have originated as a means of producing seedless inflorescences for chewing. Perhaps ganja was an ingredient in paan mixes, along with betel leaf and areca nuts. Regardless, by the time Pierre Sonnerat arrived in India in the 1770s, ganja cultivation had established itself along the Coromandel Coast, and presumably the rest of tropical India, as tobacco cultivation certainly had.

Ganja domesticatestrue var. indicaappear to have diverged from bhang domesticates due to selective pressure for potent and refined aromatic inflorescences—their strong odor “resembling somewhat that of tobacco”, as Lamarck noted of Sonnerat’s specimen. The ganja landraces cultivated in contemporary Aceh are overwhelmingly more “intoxicating” than the crop cultivated for bhang in regions such as Punjab. My guess is that bhang domesticates represent an older form of Indian Cannabis akin toand probably genetically related tothose of Central Asia. Studies already indicate the cannabinoid ratios of crops from Punjab may be like those of Central Asia. Meulenbeld persuasively argues that in Indian texts the term bhang can only be taken to refer unequivocally to Cannabis from the eleventh century, and that it only features writ large from the thirteenth century. As discussed here, the early thirteenth century was when Central Asian Cannabis culture exploded southward and westward out of Khorasan following the Mongol annihilation of Silk Road cities such as Balkh, Samarkand, and Bukhara. From the Scythians to the qalandars, Iranian influences were pivotal in the evolution of Indian Cannabis culture. Ganja—true var. indicaon the other hand, is a uniquely Indic innovation of what was once known as the East Indies, a region encompassing India, Indonesia, and much of mainland Southeast Asia.

McPartland and Small’s new paper is a landmark in Cannabis studies and a crucial document for any collector, whether amateur or academic. Future studies, in my view as an aficionado, would do well to note the significance of bhang crops from regions such as Punjab, Sindh, and Uttar Pradesh, a substantial majority of which may exhibit distinct morphological and phytochemical traits from classic var. indica ganja domesticates. Collection efforts could also focus on Himalayan landraces from Far West Nepal, and on regions of Northeast India where pristine aboriginal var. himalayensis populations may yet exist. As McPartland and Small emphasize, wild-type populations are the essential reservoir of biodiversity.

In shortstop hybridizingstart collecting!

3 responses to “Endangered Varieties of subsp. indica: A Few Thoughts

  1. Pingback: Nanda Devi: Landraces and Tall Tales from the Himalayas | The Real Seed Company: The Honest Online Source for Cannabis Landraces Since 2007·

  2. Pingback: Do Landrace Farmers Need Modern Hybrid Strains? | The Real Seed Company: The Honest Online Source for Cannabis Landraces Since 2007·

  3. Pingback: Extinction by Hybridization | The Real Seed Company: The Honest Online Source for Cannabis Landraces Since 2007·

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s