Talk:Landrace Cannabis

Talk

Discussion for: Landrace Cannabis • View • Links • Related changes

Use this page for verification notes, disputes, and action items. Keep narrative content on the article.

Page scope

Landrace Cannabis covers cannabis landraces: what the term means and how its definition has changed, how far farmers' selection shapes them, how they differ from cultivars and hybrids, and their conservation. The method of mass selection and its genetics belong to Mass selection; the making and keeping of named cultivars to Cultivar.

Data verification

Record source conflicts, dating problems and sources not used here. Keep narrative content on the article.

Discussion

Definition of mass selection

Mass selection is the choice of individual plants on their own phenotype, with the seed of the chosen plants bulked to sow the next generation. It was the earliest method of improving cross-pollinated crops: farmers chose the most desirable maize ears each year and sowed the bulked seed of those ears (Fehr, Principles of Cultivar Development, vol. 1, 1987; 1991 printing, pp. 174–175). Sudika and Soemeinaboedhy (2022, Journal für Kulturpflanzen 74: 134–141) define it the same way. The bulked seed comes from individually chosen plants. The lead uses this definition, which the Mass selection and Cultivar pages share.
Some breeders keep the name "mass selection" for seed plants chosen after open pollination, when the pollen comes from chosen and unchosen plants alike. They call selection of both parents "recurrent phenotypic selection" (Fehr, pp. 174–175).

How much selection a landrace receives

Zeven (Euphytica 104, 1998) reviews definitions across crops and does not make mass selection what defines a landrace. Banga (1944) has a population developed "without or with only little mass selection" (p. 130). Prospéri and colleagues (1994) define a landrace by "more or less directed mass selection during several generations" (p. 132).
Zeven's own text pulls two ways. On p. 129: "As already stated no, or only limited, human selection is carried out to maintain a landrace". On p. 136 Banga "wrongly writes that no selection or only mild mass selection is carried out", intensity "will have varied from nil to quite strong selection" and "little has been recorded". On p. 137 human selection "has had a great influence on the development of landraces". His preferred definition, Mansholt's as amended (p. 137), has no selection in it. The article reports both sides of this.
For cannabis, Barcaccia and colleagues (2020) write that "hemp-type and drug-type varieties have been bred mainly through mass selection", and Clarke and Merlin (2016, p. 304) that "Early hemp and drug landraces were developed through recurrent mass selection". Neither calls it the defining mechanism. Clarke (1995) reports a hemp landrace in Shandong kept with "no intentional selection by the farmers, except possibly for the large-seed characteristic".
Selection and diversity: Clarke and Merlin (2016, pp. 304–305) have farmers favouring homozygosity in selected traits and natural selection favouring heterozygosity in the rest. Barcaccia and colleagues use mass selection cycles to "increase homozygosity and uniformity". The article therefore does not say that mass selection sustains diversity.

First use of the term

Casañas and colleagues (2017) write that "von Rümker (1908) introduced the term". Zeven (p. 127) dates the first reference to landraces as genetic resources to the Vienna congress of 1890, where E. von Proskowetz and F. Schindler proposed discussing their conservation, and (p. 129) names von Rümker's paper in the Jahrbuch der Deutschen Landwirtschafts-Gesellschaft 23: 137–167 as the first definition found. The quoted German of 1890 speaks of "primitiver Kulturformen", so the two accounts can both stand.

Landraces grown outside their home region

Von Rümker held that a landrace grown outside its native region "will continue to preserve its original characters" (Zeven's wording, p. 129). Mansholt (1909) held that its genetic composition changes there (pp. 129, 135). Fruwirth (1930) wrote of the capacity of a landrace to adapt to a new region by changing (p. 130). Zeven's classification (p. 137) follows the second view.
Zeven (p. 137) defines an allochthonous landrace as an autochthonous landrace of a foreign region "recently introduced into the region concerned", which may become autochthonous after generations of growing and seed change. Maxted, Magos Brehm and Kell (2013, PDF pp. 270–271) define it as an introduced crop that is locally adapted and note that many were "introduced a significant time ago".

Kif in the Rif

Chouvy and Afsahi (2014) qualify their field estimate: "While this is only a guesstimate based on visual observations carried across the Rif it is clear that kif cultivation is quickly receding". The yield of three to five times more hashish comes from the cultivators they interviewed.
The main text says kif "is said by many to be a landrace". Their note 4 adds that "it can be safely assume that it became an inbred line (IBL) long ago".
The two accounts date the change differently. Chouvy and Afsahi have hybrids starting slowly in the early 2000s. Clarke and Merlin (2016, p. 316) have modern seed taken to Morocco in the 1980s, with traits of Western varieties "seen in nearly all the Rif populations" within ten years.

Sources for the characteristics

Camacho Villa and colleagues (2005) built their working definition from a literature review, workshop discussion and interviews. Its features are "commonly present" but "not always all present", with exceptions by propagation method, breeding system, length of formal improvement, seed management and use.
Harlan's description of landraces, quoted at length by Zeven (p. 130), comes from "Our vanishing genetic resources", Science 188: 618–621 (Zeven's bibliography, p. 138). The article cites it through Zeven. It no longer cites Crops and Man (1975), which carried no page number, or Harlan's "Agricultural Origins: Centers and Noncenters" (1971) for these points.
The glossary in Friis-Hansen and Sthapit (2000, p. 199) has landraces as "Farmer-developed varieties of crop plants that are heterogeneous, adapted to local environmental conditions and have their own local names". Its next sentence excludes improvement by "formal or private/NGO breeding programmes".
Zeven's statements on yield (pp. 128, 136–137) cover crops generally. His paper has no cannabis in it.

Isolation and movement

An earlier sentence here had genetic isolation "principally a matter of geography acting upon whatever plants happened to be brought by humans to a given area". Wikipedia's Landrace has the same sentence about animals, cited there to Sponenberg. It is no longer in the article.
Zeven (p. 128) has landraces "commonly considered as endemic to a particular region" and yet always moved, with a newcomer often forming a new landrace with the local one. Clarke and Merlin (2016) write of "geographically isolated regions" (p. 310) and of landraces founded by small introductions (pp. 297, 307–308). Both stand in the section.

Sinsemilla and seed

In Clarke and Merlin (2016, p. 308) a sinsemilla field has all male plants removed, leaving "unfertilized (therefore seedless)" females, and seed comes from pollen collected on purpose. An earlier sentence had this roguing and the retention of resin-heavy females concentrating "the breeding population". It is no longer in the article.
For hashish, Afghan sieved resin "was collected from entire populations, so individual plant selection for potency was difficult and rarely practiced" (Clarke and Merlin 2016, p. 302).
The same authors have sinsemilla growing as what "essentially gave birth to the intentional breeding of potent hybrid drug Cannabis" (p. 308), so the article no longer contrasts these practices with line-breeding.
The sentence on seed from hand-rubbed plants still needs a citation.

Animal breeds and other crops

The FAO "Glossary of Selected Terms" in In Vivo Conservation of Animal Genetic Resources (2013, p. xviii) defines a landrace as "a breed" and belongs to an animal guideline. The article no longer cites it. Crop statements now rest on Camacho Villa and colleagues (2005) and Maxted, Magos Brehm and Kell (2013), and cannabis statements on Clarke and Merlin (2016). The sentence on landraces anchoring local economies across four continents is no longer in the article.
Andreani and colleagues (2022) report a seven-year project to rescue the squash landrace Berrettina di Lungavilla by participative breeding. It concerns no cannabis, and the article no longer mentions it.
The FAO resource book's Section B and Maxted, Magos Brehm and Kell (2013) are the same work, so they share one citation.

Intellectual property and seed law

An earlier sentence had landraces "often free from intellectual property and other regulatory encumbrances", cited to Andersen (2010). Her article has no such statement. It says the reverse for seed: patents and breeders' rights "prohibit farmers from exchanging seed", and marketing rules may bar the exchange or sale of traditional varieties. The sentence is no longer in the article.
Chouvy (2024, p. 6) writes that the UPOV conditions apply to "new varieties" and not to landraces, "for they are inherited and clearly not new". He also calls an appellation of origin "the only protection that can benefit a landrace". Intellectual property covers the rights themselves.
Andersen's piece is a signed article in D+C Development and Cooperation on crops in general. It does not discuss cannabis.

Gene banks and stored seed

Ramirez-Villegas and colleagues (2020) write that ex situ crop genebanks "have become essential", and that methods for judging representation have limited application to cultivated plants. They find both common bean genepools relatively well conserved. An earlier sentence had existing gene-bank cataloguing falling short for landraces, which goes beyond the paper.
The same paper mentions wheat landraces in the Fertile Crescent in passing, citing Gepts (2006) and Harlan (1975), as a crop with only a fraction of its diversity left in fields. Cannabis figures from Clarke and Merlin (2016, p. 318) now stand in that place.
Clarke and Merlin (2016, p. 318) give the Vavilov Institute's holding as 563 seed accessions, citing Grigorev (2015). Their statement on its losses cites Hillig (2004b).