The ability to characterise cannabinoid chemical phenotype (chemotype) accurately is important for the development of Cannabis sativa L. cultivars specific for pharmacological, hemp fibre, or seed end use. Although a number of chemotyping and genotyping methods have previously been developed to predict and characterise cannabinoid composition, only a subset of the gene pool has been examined. A representative survey from a wide range of geographically and genetically diverse C. sativa accessions using liquid chromatography-mass spectrometry (LC-MS) cannabinoid profiling together with dominant and co-dominant DNA marker assays was performed. Overall variability of chemotype across the gene pool was found to be three-fold greater within heterozygote genotypes than previously reported. Interestingly, an individual plant of East Asian origin was found to exhibit a rare propyl alkyl cannabinoid homologue and a chemotype inconsistent with the predicted genotype. We propose that in order to carry out comprehensive screening of genetic resource collections and to identify chemotypic variants specific for end-use pharmacological applications, a strategy which adopts both cannabinoid profiling and the co-dominant DNA marker assay is required. Further research with consideration of propyl-alkyl-cannabinoid homologues should explore the relationship between chemotype and genotype in greater detail.

Cannabis is a predominantly diecious phenotypically diverse domesticated genus with few if any extant natural populations. International narcotics conventions and associated legislation have constrained the establishment, characterization, and use of Cannabis genetic resource collections. This has resulted in the underutilization of genepool variability in cultivar development and has limited the inclusion of secondary genepools associated with genetic improvement strategies of the Green Revolution. The structured screening of ex situ germplasm and the exploitation of locally-adapted intraspecific traits is expected to facilitate the genetic improvement of Cannabis. However, limited attempts have been made to establish the full extent of genetic resources available for pre-breeding. We present a thorough critical review of Cannabis ex situ genetic resources, and discuss recommendations for conservation, pre-breeding characterization, and genetic analysis that will underpin future cultivar development. We consider East Asian germplasm to be a priority for conservation based on the prolonged historical cultivation of Cannabis in this region over a range of latitudes, along with the apparent high levels of genetic diversity and relatively low representation in published genetic resource collections. Seed cryopreservation could improve conservation by reducing hybridization and genetic drift that may occur during Cannabis germplasm regeneration. Given the unique legal status of Cannabis, we propose the establishment of a global virtual core collection based on the collation of consistent and comprehensive provenance meta-data and the adoption of high-throughput DNA sequencing technologies. This would enable representative core collections to be used for systematic phenotyping, and so underpin breeding strategies for the genetic improvement of Cannabis.