Ernest Small, Eastern Cereal and Oilseed Research Centre,
Research Branch, Agriculture and Agri-Food Canada,
|Central Experimental Farm, Ottawa, ON, Canada K1A 0C6
(e-mail: smalle@ em.agr.ca ; website: http://res.agr.ca/ecorc/)
and David Marcus, Natural Hemphasis,
a division of Natural Emphasis Ltd.
Toronto, ON Canada
(e-mail: dave@hemphasis.com ; website: www.hemphasis.com)

Presented at the 3rd International Symposium BIORESOURCE HEMP Wolfsburg, Germany, Sept. 13-16, 2000


Canada has an extensive history of germplasm research on Cannabis sativa, including taxonomic studies that established the 0.3% THC level widely used to define low-THC cultivars that can be grown under license. In 1999 we carried out the most extensive comparative germplasm trial of hemp conducted to date in North America. In a licensed research location in southern Ontario, we grew 62 different accessions, comprising cultivars, seed bank accessions, breeding lines, and wild plants. These were statistically evaluated for a number of agronomic characteristics indicative of the suitability of the plants for Canadian production, particularly as an oilseed. Some of the accessions proved unfit as breeding stock for Canada, mostly because the maturation time was too late for the Canadian climate, but in some cases because the acceptable Canadian content limit of 0.3% THC was exceeded. There was a very wide range of development of useful traits, indicating that there is a huge reservoir of genetic variation available for improving hemp. Plants that have been growing wild in Canada for hundreds of years are also potentially useful as germplasm. For the most part, all of the plants cultivated were very resistant to insects and diseases, but except for a few resistant forms, most of the accessions were infested by the European corn borer, which caused many plants to develop numerous branches at and below the point of damage. The result was to delay maturation of these plants, but remarkably the overall biomass and seed productivity of the damaged plants were very significantly increased by comparison with the undamaged plants.



In 1938 the cultivation of Cannabis became illegal in Canada under the Opium and Narcotics Act, although there was some renewed production during the war years, 1939-1945. In 1961 the Canadian Narcotics Control Act provided for cultivation for research purposes, but only at the discretion of the Minister of Health. In May of 1997 there was an amalgamation of the Narcotic Control Act which governed illicit aspects of cannabis drugs and the Food and Drugs Act, to produce the Controlled Drugs and Substances Act. From 1994-1998 experimental cultivation was allowed, but only for research purposes (i.e. not for commercial profit). In March 1998, new regulations (under the Controlled Drugs and Substances Act) were provided to allow the commercial development of a hemp industry, and several hundred licenses were issued to cultivate the crop, with about 5,000 ha grown. Information on the commercial potential of hemp in Canada is in Marcus 1998, Pinfold Consulting 1998, and Blade 1998; and at http://www.hemphasis.com. Until very recently, it has not been possible to develop a germplasm collection in Canada. Our mandate is to establish and characterize a germplasm collection for agronomic purposes. This report deals primarily with 62 accessions we grew and studied in1999; this year (2000) we are studying an additional 120 accessions. Gehl (1995) reviewed fibre hemp development in Canada in the early 20th century, and concluded that the prospects for resuming a traditional fibre industry were poor. There is some development of fibre hemp in Canada, but mostly dual-purpose cultivars are grown at present. Canada’s large cash crop economy is based on cereals and oilseeds, and we believe that of the diverse usages of hemp, its largest potential for our country is as an oilseed, and accordingly we have been particularly concerned with this aspect of our program.


Considerable research occurred in Canada in the 1970's on a collection of about 400 Cannabis accessions, and while it was not possible to preserve this collection, it did serve to establish a system of classification that is useful for categorizing germplasm. Delimitation of groups, as shown in Fig. 1, is based on the two most important selective forces: 1) selection for fiber and oilseed characteristics vs. selection for narcotic properties, and 2) selection for any of these domesticated purposes vs. selection for wild existence. The first set of divergent selective forces has resulted in a) predominantly low-THC fiber and oilseed cultivars (var. sativa in Fig. 2), as well as wild plants (var. spontanea in Fig. 2); and b) predominantly high-THC narcotic cultivars (var. indica in Fig. 2), as well as wild plants (var. kafiristanica in Fig. 2). Collectively, the low-THC plants are placed in subsp. sativa; these are mostly found in north-temperate climates, and are photoperiodically adjusted to mature by the fall season in such locations. Collectively, the high-THC plants are placed in subsp. indica; they are (or at least used to be) mostly found in semi-tropical climates, and are photoperiodically adjusted to a relatively long season, and when grown in north-temperate countries tend to mature very late or not at all. This classification was established by Small and Cronquist in 1976, and the level of 0.3% that was used to delimit the two basic groupings in Cannabis has since been adopted in much of Europe and North America as a dividing line between cultivars that can be legally cultivated under licence and forms that are considered to have too high a narcotic potential to be permitted to be cultivated.

We recognize a wild and a cultivated variety within both the high-THC and low-THC subspecies. In contrast to domesticated seeds, wild seeds are smaller (generally less than 3.8 mm long), disarticulate more readily (facilitated by an attenuated base), are covered by a tightly adhering camouflagic mottled layer (the persistent perianth), have relatively thick walls, are relatively long-lived, and germinate irregularly (and not in the fall). By examining these features, one can often evaluate quite easily whether a cannabis plant growing outside of cultivation is spontaneous (i.e. a cultivated form that has recently escaped from cultivation) or ruderal (i.e. derived from escapes which have become efficient weeds by virtue of having re-acquired adaptations to the wild).

Cannabis is usually regarded as having just one species, although the issues was complicated by a forensic debate over the existence of "legal species" of marijuana (Small 1979). When more than one species is recognized, the name C. indica Lamarck is usually applied to the drug phase (both the domesticated and related wild forms), and sometimes the name C. ruderalis Janischevsky (= C. sativa var. spontanea Vavilov) is applied to wild north-European Old-World forms, while the name C. sativa L. is essentially restricted to the domesticated fibre form (cf. Schultes et al. 1974). An identification scheme based on the one-species concept is given below.

Key to subspecies and varieties of Cannabis sativa L.1

C. sativa subsp. sativa

C. sativa subsp. sativa var. sativa

C. sativa subsp. sativa var. spontanea Vavilov

C. sativa subsp. indica (Lam.) E. Small & Cronq.

C. sativa subsp. indica var. indica (Lam.) Wehmer

C. sativa subsp. indica var. kafiristanica (Vavilov) E. Small & Cronquist

1 Only substantially mature seeds (achenes) exhibit the diagnostic characteristic clearly. Cannabinoid concentration is known to vary somewhat with environment, but for most practical purposes the level of 0.3% THC suffices. Some plants or populations will be found to be intermediate and not easily assigned to one of the groups, but this is a well known aspect of classifying groups within a species. The authority names (e.g. (Vavilov) E. Small & Cronquist) are not necessary for most purposes. Also, it isn’t necessary when using the name of a variety to repeat the name of the subspecies (e.g., C. sativa var. kafiristanica is sufficient, and it isn’t necessary to specify C. sativa subsp. indica var. kafiristanica). The taxonomic category var. (variety) is not equivalent to horticultural varieties (i.e. cultivars).


Planting occurred at the beginning of June 1999, in a field in the southern, Ontario region, licensed as required under current regulations governing hemp cultivation. Sixty-two accessions (Table 1) were grown in a 4-replicate randomized complete block layout, each plot established with 15 plants in a 4.68 m (15 ft) row, 1.25 m (4 ft.) between rows Male (staminate) plants were not studied since they are generally not nearly as useful as the females. Individual plants with female flowers (both from monoecious and dioecious accessions) were scored for height, above-ground fresh weight, and other characteristics. Plants were evaluated when ready for seed harvest, which ranged, depending on accession, from mid summer to mid October. Only accession 30, a form with narcotic tendencies that was unintentionally grown, failed to produce seeds, consistent with the observation that most narcotic forms are not photoperiodically adapted to mature in southern Ontario. Some plants, although possessing female appearance, were sterile, bearing mostly male flowers, and were not scored (the possession of at least some female flowers appeared to prevent the early death of the plant that always occurs in exclusively male plants). This is a preliminary report, and more detailed information will be published elsewhere.


In Canada, every acquisition of hemp grown at a particular place and time must be tested for THC content by an independent laboratory, although we were licensed to collect samples and did so. Our samples were collected at the early flowering stage, and somewhat higher levels than those reported here probably would have developed had we sampled at early fruiting. As shown in Fig. 3, most of the accessions (87%) were below the permitted level of 0.3%. Those at or above this level included (see Table 1) accessions: 11, 14, 17, 21, 29, 30, 38, and 44. Given their relatively high THC content, these latter accessions would not be suitable for incorporation into a breeding program, unless they possessed exceptionally valuable characteristics.


As shown in Fig. 3, maturation times (judged by 50% seed maturation) varied considerably for the different accessions. The cultivar FIN314 matured at only 11 weeks following planting. On the other hand, accessions 14, 29 and 30 were not mature by mid-October, just before frost occurred (their maturation times have been estimated in Fig. 4). These latter three accessions all had THC levels higher than 0.3%, and reflect the tendency of high-THC Cannabis sativa to mature very late when grown in North-temperate conditions. The following accessions collected at 19 weeks are too late-maturing to be suitable for much of Canada: 17, 18, 22, 26, 27, 44, 46, 48, 49, 50, 60. Thus 14/62 (23%) of the accessions tested were marginally or not at all suitable for producing seed in much of Canada, although most of these could be useful exclusively for fibre production.


As illustrated in Fig.5, about half of the accessions were dioecious, a third were predominantly monoecious, seven were all-female, and one was strictly monoecious. Cannabis sativa is naturally dioecious, and divergence from this state was found in many of the cultivars and experimental breeding lines examined.


In view of the large potential of hemp as an oilseed in North America, we conducted a number of studies intended to evaluate the oilseed potential of the germplasm collection. These are summarized below. Oil analysis was conducted on mature seed only, since it is known that immature seed may be qualitatively different from mature seed.


Mean values for the principal fatty acids are given in Fig. 6, based on open-pollinated seeds produced in 1999 by almost all of the accessions. As can be seen in Table 2, the data fall within the ranges reported in the literature. Percentages of the essential fatty acids (linoleic, linolenic) are shown in Figs. 7 and 8, and percentages of GLA are given in Fig. 9. The highest percentages of GLA (>4%) were in accessions 6, 8, 35, 45, 61, and 64.


Seed weight (using mature seeds) was determined on seeds produced by open-pollination in the experimental field in 1999. As shown in Fig. 10, mean seed weight among the accessions varied from about 17 to 32 grams. For some applications, heavier seeds may be desirable. Accessions with the heaviest seeds included 2, 5, 24, 25, 31, and 51.


For some applications, hemp seed is de-hulled, and the hull is of little or no value. Accordingly, we determined the comparative proportion the hull makes up of seeds of the accessions we studied, again on seeds produced by open-pollination in 1999. As can be seen in Fig. 11, the hull varied among accessions from about 30 to 42% of the weight of the seed. Accessions with the least proportion of hull in the seeds included 4, 20, 22, 31, 45 and 55.


Germination tests were performed on seeds that had been produced by open pollination in the experimental field. These trials were conducted about 3 months after harvest, at room temperature, on seeds that had been stored at room temperature. The seeds of all accessions showed some germination after only 24 hours, except for those of accession 16, the most obviously wild form (C. sativa var. spontanea) in the collection. Accession 16 showed delayed germination, but after 3 days 70% of the seeds had germinated. Most of the accessions showed at least 95% germination after 3 days. Delayed germination is well known in C. sativa var. spontanea.


As illustrated in Fig. 12, Cannabis sativa has been grown at low densities for narcotics and as an oilseed, and at higher densities as a fibre crop, oilseed, or for a combination of fibre and oilseed. Low density cultivation, as is well known, encourages branching and associated flower and seed production, but it must be remembered that strains of hemp differ greatly in their genetic capacity to branch. Cultivation at low densities is one strategy to increase oilseed production, particularly with cultivars having a natural capacity to branch. However, low-density cultivation of hemp makes for a much greater weed elimination challenge during seedling establishment, while high-density cultivation suppresses weeds very well. Also, highly branched plants are less desirable from the point of view of straw, hurd, and fibre utilization of the remains of the plant, and may also make harvest more difficult. An alternate architectural strategy is the fairly dense cultivation of unbranched plants, which may be short, like FIN314, and intended predominantly as an early-maturing oilseed, or tall, and intended for dual harvest of seeds and stalks, like many European cultivars. Plants with limited (or at least short) branching are naturally superior than irregularly branching plants for the purpose of fully and uniformly occupying a field, and maximally utilizing solar irradiation. This is obviously desirable for optimizing production.


Ideally, seed productivity should be measured by actually harvesting seeds from representative plots. Since we are evaluating large numbers of accessions (for the year 2000 we have cultivated well over a hundred) we needed a quicker, more easily obtained measure, and we adopted a "seed productivity index" that simply required visual inspection. For an accession to score well on our index, individual plants were required to bear many seeds. In Fig. 13, the accessions we studied are scored on our index, which varied from zero (minimum) to 10 (maximum). The highest ranking accessions were 2, 11, 15, 17, 18, 25, and 29. While no accession had a mean index higher than 8, many individual plants were rated higher. Many of the accessions were noticeably heterogeneous (with both excellent and poor plants according to our index), especially the dioecious accessions.


The harvest index is the fraction of economically valuable plant material of the total plant dry matter produced. During the so-called Green Revolution of the 1960s and 1970s, tremendous increases in crop yields occurred, based in part on the selection of new varieties of grains that had an increased harvest index, measured as the ratio of grain weight to above-ground weight. In historical times, straw production was a priority, so that a high harvest index was not necessarily desirable (Sinclair 1998). When hemp is grown simultaneously for seed and for stem products, a single index may not be very useful. However for grain crops knowing how much of the plant weight is made up of seeds is important, particularly when growing exclusively for seed. For breeding purposes, it may be desirable to consider selecting specifically for a high harvest index. Seeds were collected from most of the accessions, and in Fig. 14 we present the distribution of mean percentage that the seeds made up of the fresh weight of the plant (harvest index has not been calculated, since this should be based on dry weight, which we did not measure). Birds consumed more than 50% of the seeds, so that our data underestimate considerably the seed-bearing potential of hemp. As can be seen in Fig. 14, the percentage of the weight of the plant made up of seeds varied generally between 5 and 20%. The accessions with the highest percentages of seeds were 18, 20, 22, 24, and 27.


Although a relatively minor usage, there is interest in using hemp as a biomass crop. We measured the fresh above-ground weight of every female plant we grew, and the mean values for the accessions studied are shown in Fig. 15. We did not include FIN314 (accession 45) in this analysis, since the plants matured far earlier than all other accessions, when quite small, and can’t be fairly compared for biomass production with plants that grew for a much longer period. As indicated in Fig. 15, mean plant weight varied from less than 50 to almost 400 g. The heaviest plants (with period of growth in parenthesis) were produced by accessions 15 (16 weeks), 17 (19 weeks), 29 (19 weeks), 32 (17 weeks), and 48 (19 weeks). It should be remembered that our experimental plants were grown at very low density, and the data do not necessarily reflect biomass potential at higher densities for particular accessions.


Our hemp plants seemed quite resistant to most insects, but the European corn borer (Ostrinia nubilalis Hübner) infested the vertical stem leader of at least some plants of 55 of the 62 accessions we grew (15.5% of the infested lines were visibly damaged). At the site of infestation the main stem was destroyed, and the plant became strongly branched (Fig. 16). Seven of the 62 accessions (3, 16, 27, 30, 35, 43, and 45) completely escaped obvious infestation by the ECB. These included the three most unusual accessions studied: 16, a wild, highly-branched form with very woody stems; 30, the only accession that failed to mature more than a few seeds, a highly-branched narcotic form with very woody stems; and 45, the cultivar FIN 314, the shortest and earliest-maturing of all the plants grown, with very slim stems. The resistance of these three accessions, with the slimmest and/or woodiest stems, suggests a preference by the insect for larger and softer stems.

The damaged plants appeared to have comparatively large seed productivity, although floral production and seed maturation were delayed. Mean weight of the damaged plants was significantly heavier, by more than a third. This phenomenon of damaged plants growing more robustly and productively than their undamaged counterparts, is known as "overcompensation," and its significance is very controversial (e.g. Belsky 1986). Our observation suggests that European corn borer damage may, at least for some purposes, be good for hemp. However, this is not really all that surprising.The capacity to respond to damage to the main leader is important for hemp. Moes (1998) observed that following severe hail damage to a hemp plot in Manitoba, axillary branches developed at nodes below damaged stems, produced inflorescences, and a substantial (albeit reduced) seed yield. Pate (1999) noted that when growing hemp for seed, the number of flowers per plant and the number of seeds produced can be increased by "topping" the plants when 30-50 cm high. While European corn borer damage to hemp stem leaders is probably not normally beneficial, at least under some circumstances the reduction in crop yield may be limited.


Wild North American hemp is derived mostly from escaped European cultivated hemp imported in past centuries. Hemp was introduced to North America in Port Royal, Acadia (Nova Scotia) in 1606. It was a popular crop in Eastern and Central Canada during the 18th and 19th centuries, but by the mid 1930s production had ceased, except for a brief revival during World War II. The distribution of wild Canadian C. sativa is shown in Fig. 17. North American wild plants have re-evolved adaptations to wild existence, including the wild achene characters described above. The wild plants also typically differ from their domesticated relatives in having smaller stature, a much more branching habit with shorter internodes and woodier stems, smaller leaves and narrower leaflets. Forms that have re-evolved wild adaptations are concentrated along the St. Lawrence and lower Great Lakes, where considerable cultivation occurred in the 1800s. In the US, wild hemp is best established in the American Midwest and Northeast, where hemp was grown historically in large amounts. Decades of eradication have exterminated many of the naturalized populations in North America. In the US, wild plants are rather contemptuously called "ditch weed" by law enforcement personnel. However, the attempts to destroy the wild populations are short-sighted, because they are a natural genetic reservoir. Wild North American plants have undergone many generations of natural adaptation to local conditions of climate, soil and pests, and accordingly it is safe to conclude that they harbour genes that are invaluable for the improvement of hemp cultivars. Wild North American hemp is very low in THC, and is usually distinguishable from narcotic strains that are illicitly cultivated in clandestine locations. In North America, numerous wild plants are being used to revegetate and maintain landscapes, especially to improve habitat and provide food for wild birds. Hemp seeds are nutritious and extremely attractive to birds, and were it not for its reputation, wild hemp could be planted for the benefit of wildlife. However, present policies in North America require the eradication of wild hemp wherever encountered.


We particularly thank the individuals and institutions mentioned in Table 1 for contributing seeds. Drs. G. Butler and A. McElroy of Agriculture & Agri-Food Canada have provided advice on statistical design and will coauthor rigorous analyses to be published later.

Table 1. Brief details of accessions (study code, supplier, catalogue number, description, source; additional details are available in catalogues of supplier institutions)

Vavilov Institute of Plant Industry, St. Petersburg, Russia (courtesy of S.V. Grigoryev):1. Local, Altai, Russia. 2. N.-Zaimskaia, Russia. 3. Toguchinskaia, Russia. 4. Local, Sverdlovsk, Russia. 5. YUS - 22, Ukraine. 6. YUSO- 14, Ukraine. 7. YUSO- 19, Ukraine. 8. YUSO- 21, Ukraine. 9. YUSO- 31, Ukraine. 10. YUSO- 33, Ukraine. Institut fuer Pflanzengenetik und Kulturpflanzenforschung, Gatersleben, Germany (courtesty of A. Boerner): 11. Accession CAN 16, SVK (Slowakei). 12. Accession CAN 17, Hungary. 13. Accession CAN 19), Italy. 14. Accession CAN 20, Korea. 15. Accession CAN 21, Romania. 16. Accession CAN 22, "C. sativa L. subsp. spontanea Serebr. ex Serebr. et Sizov," Georgia. 17. Accession CAN 23, Korea. 18. Accession CAN 24, Italy. 20. Accession CAN 26, Turkey. 21. Accession CAN 27, unknown origin (donated by Bot. Gart. AdW Vacratot, Ungarn). 22. Accession CAN 29, Romania. 24. Accession CAN 31, "cunepa," Romania. 25. Accession CAN 32, "conopla (ukr.),", Romania. 26. Accession CAN 33, "cinepa," Romania. 27. Accession CAN 34, "cinepa," Romania. 28. Accession CAN 35, "cinepa," Romania. 29. Accession CAN 36, unknown origin (donated by Herr Schulze, Greifenberg). 30. Accession CAN 37, France. 31. Accession CAN 38, "cinepa," Romania. 32. Accession CAN 39, wild plants from Beijing, China. UK Praha, Farmaceutická fakulta zahrada le…ivýh rostlin Heyrovského, Hradec Králové, Czech Republic (courtesy of curator): 33. From plants grown in Carolinae University Botanical Garden. AGRITEC, Research, Breeding & Services, Ltd., Zemedelska, Czech Republic (courtesy of P. Smirous): 34. Cultivar Beniko. 35. Cultivar Bialobrzezski. 36. Cultivar Juso-11. Escola Superior d’Agricultura de Barcelona, Spain (courtesy of G. Gorchs): 37. Delta 405 (an old Spanish cultivar). J.D. Spanring, Strossmayerjeva, Ljubljana, Slovenia: 38. Rudnik A16 (derived selection A16 from Rudnik, an old Slovenian land race from gene bank). 39. Rudnik T17 (derived selection from Rudnik). 40. Rudnik A16 (derived selection from Rudnik). 41. Pesnica (local race from Štajerska, an old Slovenian land race from gene bank). 42. Gazvoda (local race from Dolenjska, an old Slovenian land race from gene bank). Agra Seeds Inc., Winnipeg, MB (courtesy G. Cloutier): 43. Cultivar Fasamo. Utrecht University Botanic Gardens, The Netherlands (courtesy of J. Tolsma, Curator): 44. Land race cultivated in Utrecht Botanical Garden (originally from Botanical Garden of Gent University, Belgium, from cultivated origin). Department of Pharmaceutical Chemistry, University of Kuopio, Finland (courtesy of J. Callaway) and Gen-X Research Inc. (courtesy of S. Przytyk): 45. Cultivar Fin 314. 46. Agricultural Research Institute, Kompolt (Heves), Hungary (courtesy of I. Bócsa): Cultivar Kompolti. 47. Cultivar Uniko-B. 48. Hybrid: Kompolti hybrid TC. 49. Cultivar Fibriko. 50. Hybrid: Lipko(Fibr.xUniko-B). 51. Hybrid: FxT,F1. 52. Hybrid: K.monoecious x K.unisex. 53. Hybrid: Fibrimon x (FxT). 54. Hybrid: Fibrimon x K.unisex. Kenex Hemp, Pain Court, Ont.: 55. Cultivar Ferimon 12. 56. Cultivar Fedora 19. 57. Cultivar Felina 34. 58. Cultivar Fedrina 74. 59. Cultivar Futura 77. 60. Cultivar Uniko B. Consolidated Growers and Processors, Winnipeg MB (courtesy of S. Dushenkov): 61. Cultivar Zolo 11 (super elite). 62. Cultivar Zolo 15 (elite). 64. Cultivar USO 14 (super elite). 65. Cultivar USO 31 (elite).

Table 2. Average fatty acid profile of accessions grown in Canadian survey, based on means

Fatty acid Canada Literature2
Palmitic acid 7.93% 6 - 9
Stearic acid 2.34% 2 - 3
Oleic acid 12.52% 10 - 16
Linoleic (omega-6) 55.83%1 50 - 70
Alpha-linolenic (omega 3) 16.63%1 15 - 25
Gamma-linolenic (GLA) 2.34% 1 - 6
Total 97.56%

1 Ratio linoleic/linolenic: 3.36
2 Reported in Pate (1999)


Belsky, A.J. 1986. Does herbivory benefit plants? A review of the evidence. Am. Nat. 127: 870-892.

Blade, S. (ed.). 1998. Alberta Hemp Symposia proceedings, Red Deer, Alberta, March 10, 1998, and Edmonton Alberta, April 8, 1998. Alberta Agriculture, Food and Rural Development, Edmonton, AB. 85 pp.

Gehl, D. 1995. A summary of hemp research in Canada conducted by the fibre division of Agriculture Canada, 1923-1942. Agriculture and Agri-Food Canada, Indian Head, SK. 10 pp.

Marcus, D. 1998. Commercial hemp cultivation in Canada: an economic justification. (Based originally on a 1996 thesis at the University of Western Ontario). [Available at http://www.hemphasis.com]

Moes, J. 1998. Hemp research in Manitoba - 1995-1997. Pages 43-48 in S. Blade, ed. Alberta hemp symposia proceedings, Red Deer, Alberta, March 10, 1998, and Edmonton Alberta, April 8, 1998. Alberta Agriculture, Food and Rural Development, Edmonton, AB.

Pate, D.W. 1999. Hemp seed: a valuable food source. Pages 243-255 in P. Ranalli, ed. Advances in hemp research. Food Products Press, New York, NY.

Pinfold Consulting 1998 (G. Pinfold Consulting Economists Ltd. and J. White, InfoResults Ltd.). A maritime industrial hemp product marketing study. Prepared for Nova Scotia Agriculture and Marketing (Marketing and Food Industry Development), and New Brunswick Agriculture & Rural Development (Marketing and Business Development). 31 pp. + appendices.

Schultes, R.E., Klein, W.M., Plowman, T. and Lockwood, T.E. 1974. Cannabis: an example of taxonomic neglect. Bot. Mus. Leafl. Harvard Univ. 23: 337-367.

Sinclair, T.R. 1998. Historical changes in harvest index and crop nitrogen accumulation. Crop Sci. 38: 638-643.

Small, E. 1979. The species problem in Cannabis, science and semantics. Corpus, Toronto, ON. 2 vol.

Small, E. and Cronquist, A. 1976. A practical and natural taxonomy for Cannabis. Taxon 25: 405-435.


Fig. 1. Disruptive selective forces producing variants of Cannabis. Above, divergent selection for drug and fibre content. Below, divergent selection for different kinds of "seeds" (achenes) between domesticated and wild plants.

Fig. 2. Classification of Cannabis sativa into subspecies on the basis of selection for intoxicant potential, and of the subspecies into varieties on the basis of selection for domesticated or wild characteristics.

Fig. 3. THC content of accessions.

Fig. 4. Maturation times (for seed harvest) of accessions. Killing frosts occurred in mid-October, and the later maturation times were estimated.

Fig. 5. Sexual types among accessions ( % = male (staminate) plants; & = female (pistillate) plants; = plants with male and female flowers). Note that in accessions with a mixture of types including hermaphrodites, most plants tended to be hermaphrodites.

Fig. 6. Fatty acid profile, based on means of accessions (cf. Table 2).

Fig. 7. Frequency histogram, percentage linoleic acid among accessions.

Fig. 8. Frequency histogram, percentage linolenic acid among accessions.

Fig. 9. Frequency histogram, percentage gamma-linolenic acid among accessions.

Fig. 10. Frequency histogram, mean seed weight among accessions.

Fig. 11. Frequency histogram, mean percentage hull of seeds of accessions.

Fig. 12. Appearance of plants according to a combination of environment (modes of cultivation) and genetic background.

Fig. 13. Frequency histogram for accessions, mean scores for quality as oilseed producers (see text for explanation).

Fig. 14. Frequency histogram for accessions, mean percentage the seeds make up of fresh plant weight.

Fig. 15. Frequency histogram, mean above-ground biomass of accessions.

Fig. 16. Profiles of normal plant (left) and plant infested by the European corn borer (right).

Fig. 17. Distribution of wild hemp in Canada.


spy phone mobile buy gold bar