hey - if it's any consolation, I live and look like a tramp - I'm told I smell alright though
I reckon for people who can't face reading the whole paper, it's worth reading the conclusion
Small and Cronquist (1976) assumed that the sativa and indica taxa diverged primarily as a result of human selection for fiber and/or seed production on the one hand and for high THC production on the other. However, the present investigation does not support this hypothesis. The high frequency of BT in the feral biotype of C. indica suggests that this allele may have been present at high frequency in the C. indica gene pool prior to domestication. Human selection of plants carrying two copies of the BT allele appears only to be of appreciable significance in the domestication of the NLD biotype. Human selection may have resulted in an increase in the quantitative levels of cannabinoids produced by the WLD biotype, but the average amount of CBD + THC produced by the NLD biotype did not significantly differ from the hemp and feral biotypes of C. indica. In fact, the average amount of THC + CBD produced by the NLD accessions was not significantly greater than the average amount of these two cannabinoids produced by the hemp accessions of C. sativa. Small and Beckstead (1973b) also reported comparable levels of total cannabinoids in their "psychotomimetic" and "non-psychotomimetic" strains. Plants with enhanced levels of THCV were uncommon in most drug accessions, except those from southern Africa and an accession from Afghanistan. This suggests that humans may have selected against this trait in cultivated drug strains.
In contrast with the NLD biotype, the WLD biotype did not have a significantly higher frequency of BT than the hemp and feral biotypes of C. indica. This may be explained by the different products obtained from the two drug biotypes of C. indica. NLD strains are usually cultivated for the production of marijuana (pistillate inflorescences), and it is the product of a single plant that is utilized. WLD strains are traditionally cultivated for the production of hashish (detached glandular trichomes), which is the combined product of many plants. A marijuana plant with two CD alleles (chemotype III) would be ineffectual for its intended use, whereas hashish made from a population of chemotype I, II, and III individuals typically contains more THC than CBD and depending on the proportions of the different chemotypes is more or less psychoactive (Ek et al., 1972 ; Clarke, 1998 ). Thus, human selection is expected to favor chemotype I plants as seed sources for marijuana cultivation, whereas the chemotypes of individual plants cultivated for hashish production are usually unknown to the cultivator (Clarke and Watson, 2002 ).
The presence of allele BT in the C. sativa gene pool suggests that introgression from C. indica might have played a role in the evolution of C. sativa. Wind-blown pollen may have contributed to allele migration between the two gene pools (Cabezudo et al., 1997 ). Relatively high BT frequencies (range 0.38–0.55) were detected in seven hemp accessions from Turkey, Spain, Italy, former Yugoslavia, and southern Russia, which are assignable to the southern eco-geographical group of C. sativa (Davidyan, 1972 ). Additional allozyme markers and morphological traits typical of C. indica were also observed in the southern group of C. sativa (Hillig, 2004 , in press). Cannabis indica may have been introduced into Asia Minor for the purpose of hashish production and hybridized with C. sativa. Davidian (1972) cited evidence that Cannabis was introduced into Europe by both a northern and a southern route. Introgressed stock may have spread into new areas through trade or human migration (Heiser, 1973 ).
The patterns of cannabinoid variation provide evidence of progenitor-derivative relationships. The low frequency of BT and the low levels of propyl cannabinoids in accessions assigned to C. ruderalis suggest that this putative taxon could be the progenitor of C. sativa, but not of C. indica. The feral biotype of C. sativa may be comprised of "escaped" populations of cultivated C. sativa that have merged with naturalized populations of C. ruderalis (Vavilov, 1926 ). The wide range of cannabinoid variation within feral accessions of C. indica suggests that this biotype could be the progenitor of the cultivated biotypes of C. indica. The high incidence of plants in this taxon with enhanced levels of propyl cannabinoids suggests that it is not the progenitor of C. sativa, in which plants with enhanced levels of propyl cannabinoids are much less common. It is unlikely that the feral biotype of C. indica represents an escape of NLD strains from cultivation because the NLD biotype is lacking in cannabinoid variation. The high frequency of BT in the hemp biotype of C. indica suggests the possibility that one or both drug biotypes could have been secondarily derived from this taxon's gene pool. More feral and cultivated populations of C. indica and C. sativa will have to be studied to further resolve these issues.
This study of cannabinoid variation supports a two-species concept for Cannabis. A taxonomic revision that applies valid scientific names to the biotypes of C. indica recognized herein is supported by the chemotaxonomic data.