Examining population structure across multiple collections of Cannabis

Abstract

Cannabis sativa L. (Cannabaceae) is an annual flowering herb of Eurasian origin that has been associated with humans for over 26,000 years. Multiple independent domestications occurred with different events leading to use as food, fiber, and medicine, with human intervention likely accelerating a division in the genus with varietals broadly known today as either hemp-type or drug-type. Using publicly available sequence data, we assessed genome wide diversity and population relationships across seven independently developed datasets. Phylogenetic analysis was conducted, with data sources providing a unique sampling of Cannabis varieties, with landrace and modern cultivars represented. Comparing nucleotide diversity over chromosomal length in landrace and domesticated modern drug-type varieties, genomic regions with decreased diversity where humans may have selected for specific traits were identified. Population structure was evident based on use type. Evidence of hybridization was extensive across the datasets. In a subset of landrace individuals where geographic origin was known, population separation was observed between varieties collected from Northern India in the Hindu Kush Mountains and Myanmar. The use of publicly available data provides an initial impression of the complexity within the Cannabis genus and adds to our understanding of the genetics underlying evolutionary history and population stratification, which will be critical for future crop improvement for any potential human use.