Comparing hemp (Cannabis sativa L.) cultivars for dual-purpose production under contrasting environments

• The cultivar effect on stem and seed yield was determined mainly by flowering time.
• A large variation in bast fibre content of commercial cultivar was found. Harvesting at seed maturity increased stem yield only of monoecious varieties.
• Modelling can aid design of production strategies for dual-purpose hemp.

Interest in hemp as a multi-purpose crop is growing worldwide and for the first time in 2015 it was cultivated in Europe on more than 20.000 ha as a dual-purpose crop, for the seeds and for the fibre. In the present study, fibre and seed productivity of 14 commercial cultivars were tested in four contrasting European environments (Latvia, the Czech Republic, France, Italy). At full flowering, the stem yield ranged from 3.7 Mg ha−1 to 22.7 Mg ha−1, the bast fibre content ranged from 21% to 43%, and the bast fibre yield ranged from 1.3 Mg ha−1 to 7.4 Mg ha−1. When harvesting was postponed from full flowering until seed maturity, the stem yield of monoecious cultivars significantly increased but in dioecious cultivars it decreased at all tested sites, except for Italy. Only the early cultivars Fedora 17 and Markant produced seed in the most northern location Latvia. The seed yield ranged from 0.3 Mg ha−1 to 2.4 Mg ha−1 in Italy, France and the Czech Republic. The cultivar effect on stem and seed yield was mainly determined by the genetic variation in time of flowering. Stem yield at full flowering was strictly related to the duration of the vegetative phase while seed yield was lowest in the late flowering cultivar. The late cultivar CS is suitable for stem and fibre production as it had the highest stem yield at full flowering in all locations. Both Fedora 17 and Futura 75 are candidate cultivars for dual-purpose production in Italy, France and the Czech Republic, with Fedora 17 being more suitable for seed production and Futura 75 for fibre production.The application of modelling to design production strategies for dual-purpose hemp is promising. However, accurate parameterisation is needed based on large data sets and diverse genetic background.