Diversity in cannabis

In an effort to reduce the overproduction of a too limited number of arable food crops, several research programs focusing on industrial crops have recently been initiated in the Netherlands. The 'Hemp research program' investigated from 1990 to 1994, the feasibility of hemp as an arable crop and as a raw material for paper pulp. In this context, breeding, agronomy, plant pathology, mechanization, processing and economics were the subject of a comprehensive study.This thesis reports on a part of the breeding research. A collection of Cannabis germplasm, covering variation within the genus, was established and evaluated at the DLO-Centre for Plant Breeding and Reproduction Research (CPRO-DLO, Wageningen). Stem production, stem quality, psychoactive potency and resistance to soil pathogens were considered relevant criteria for the introduction of hemp in crop rotations in the Netherlands and its utilization as a raw material for paper pulp. Prospects for the breeding of improved cultivars were based on the diversity for these traits, their stability, and mutual relations.In its final state the established germplasm collection comprised ca. 200 accessions (entries, populations), including fibre strains, drug strains, an ornamental cultivar and truly wild or naturalized populations (Chapter 2). According to consulted references, considerable mutual relatedness seemed to exist among fibre cultivars. Italian and German strains, especially, have directly been the basis of, or have been used as breeding parents for many of the present cultivars. Also, the central European naturalized populations and the fibre landraces seemed quite closely related to modern fibre cultivars. The drug strains were expected to be more distinct. The collection was considered to be sufficiently representative for investigating diversity in Cannabis.Traits related to stem production, stem quality and psychoactive potency were observed for ca. 160 accessions in field trials, or determined afterwards with field grown materials. Trials were carried out at CPRO (52° N latitude). Accessions were grown in small plots in two replicates, in high plant density to obtain results that are readily interpretable for fibre hemp cultivation. Host reactions to the root-knot nematode Meloidogyne hapla Chitwood were assessed for ca. 150 accessions in a seedling test, and verified in a field trial for a subset of accessions.Variation in phenological development and in stem elongation was studied in relation to stem production (Chapter 3). Large variation among accessions was found for the day of anthesis (ranging from approximately 10 May to 30 September) and consequently for the day of seed maturity. A higher latitude of origin was associated with earlier anthesis and seed maturity (r=-0.75). The phenological pattern was stable over years. Stem elongation was characterized by the parameters of a sigmoid curve fitted to periodical measurements of stem length. Significant differences among accessions were found for these parameters. The final stem length ranged from 60 to 360 cm. Stem elongation was less stable over years than the phenological pattern. Day of anthesis and final stem length were strongly and positively related to stem dry matter production (r>0.7). Some very lateflowering fibre landraces, with low bark fibre contents, produced up to 2200 g of stem dry matter per m 2and exceeded the fibre cultivars (producing 800 to 1750 g/m 2) in stem yield. Late-flowering drug strains were less persistent in a dense crop than late fibre strains. It was concluded that in an efficient crop growth system, seed reproduction and stem production should occur in separate geographic areas. Seed reproduction (adaptation) should be organized at lower latitude to obtain late-flowering, high yielding cultivars for stem production at higher latitude. At a given phenological pattern, breeding can contribute to yield potential by increasing both the persistency of cultivars and the efficiency of stem dry matter accumulation.The large differences between bark tissue and woody core in anatomical and chemical properties, make a quantification of these two fractions important for stem quality assessment. The bark is the most valuable component for pulp production. Accuracy and power of discernment of current procedures for the analysis of the composition of hemp stems were evaluated (Chapter 4). A method using stem segments, intended to reduce the amount of material to be handled, gave somewhat less accurate, but still usable information, than analysis of entire stems. The relation was studied between the simple estimation of the fractions of bark fibre and woody core, commonly applied in fibre hemp breeding, and the assessment of potential pulp recovery according to a more laborious standard method of the pulp and paper industry. The recovery of bark in the pulping process (pulp yield) is accurately predicted by the fraction of bark fibre. The woody core recovered as pulp is a fixed fraction (ca. 69%) of the intact woody core.Variation among accessions was assessed for the woody core fraction in the stem dry weight, which ranged from 50 to 78% (Chapter 5). The fraction of bark, being the complementary tissue, ranged hence from 22 to 50%. The fraction of primary bark fibre in the stem varied from 8 to 27%, the secondary bark fibre from 0 to 14%, and the total bark fibre from 9 to 34%. The fibre cultivars had strongly increased fractions of bark tissue and primary and secondary bark fibre, and reduced fractions of woody core in comparison to wild populations, drug strains and fibre landraces. The levels of the assessed stem fractions were stable over years. Since woody core fibres of hemp are on average too short (550 μm according to references) to produce high quality paper pulp, the evaluation was directed at detecting variation for wood fibre dimensions as well. Accession means for wood fibre length ranged only from 433 to 613 μm. Also the variation within accessions was limited. Breeding for improved woody core quality is considered not very promising. The best strategy for genetic improvement of pulping quality seems a continued selection for increased bark fibre content in fibre cultivars, which implies a reduction of the woody core fraction.A simple possibility to increase bark fibre yield for pulp production is indicated in Chapter 9. It consists of the adaptation of current, well performing fibre cultivars to low latitude, meanwhile maintaining the high bark fibre content by a mild selection. The resulting delay in phenological development is expected to increase the potential bark fibre yield.Large variation among accessions was found for the contents of the major cannabinoids delta-9- tetrahydrocannabinol (THC; 0.01 to 1.77% in the female inflorescence dry matter) and cannabidiol (CBD; 0.05 to 2.19%) (Chapter 6). THC is the psychoactive compound in Cannabis. Within accessions, THC content was less variable than CBD content. Both contents were strongly affected by the year of cultivation, but THC content was more stable than CBD content. Accessions were classified into the phenotypes 'drug', 'intermediate' and 'non-drug' on the basis of cannabinoid profiles. There were no strict relationships between the cannabinoid profiles and nonchemical traits. A few accessions combined for example a considerable psychoactive potency and a high content of bark fibre. On the accession average level, relatively wide leaflets and slow phenological development were weakly associated with a stronger psychoactive potency. There seems little need for efforts aimed at a further reduction of THC content in the current fibre cultivars. The collection comprised a number of accessions, suitable for the selection of genotypes for the production of either THC or CBD.Inoculation of Cannabis seedlings with a larval suspension of the root-knot nematode Meloidogyne hapla, resulted in significant variation for the number of galls and egg masses on the seedling roots (Chapter 7). These parameters were considered as estimates for nematode infection and larval multiplication, respectively. A subset of the tested accessions was grown on a naturally infested arable field to verify the relation between the test results and host reactions in the field. The ranking order of accessions for the seedling test parameters agreed satisfactory with that for nematode infection and multiplication in the field. The collection comprised one relatively resistant fibre cultivar with good agronomic properties which seemed a useful source for further selection. Some highly resistant individual Plants retained from accessions with poor agronomic properties were expected to be useful for cross-breeding.The diversity for the observed agronomic and morphological traits considered together was studied in Chapter 8. Mutual associations among traits were weak, indicating that many combinations of character states within the observed ranges can be established through breeding. For the interpretation of diversity patterns, accessions were grouped a priori on the basis of the presumed purpose and status of domestication resulting in the recognition of four 'plant-use groups': fibre cultivars, fibre landraces, drug strains, and truly wild or naturalized populations. Plant-use groups could be discriminated quite well on the basis of experimental observations. Contents of bark fibre and cannabinoids were most discriminative.A genetic characterization of accessions was additionally attempted on the basis of electrophoretic patterns of seed proteins. Although reproducible variation in banding patterns was found among accessions, the results were quite unsatisfactory as they did not reflect expected common ancestry. Furthermore, the banding patterns were independent from any grouping based on origin or agronomic and morphological traits.Individual accessions suitable for cultivation and breeding are identified in Chapter 9. Issues requiring further evaluation in case of actual hemp breeding and cultivation in the Netherlands are, resistance to the root- lesion nematode Pratylenchus penetrans Cobb and the fungus Botrytis cinerea Pers. ex Fr. (grey mold). Also genetic relationships among accessions need further clarification.