Abstract
The kelp species Saccharina latissima is gaining attention due to the growing interest in kelp cultivation in Europe and North America. The current common practice of propagation is to generate seedlings through mixed hybridization of individuals collected from the wild. However, virtually nothing is known about rates of self-fertilization, apomixis, and unintended selection in this process. Moreover, efforts to develop genetic material suitable for cultivation remain limited. We conducted a selection experiment to test how mixed hybridization and selection could be used for initial breeding for large-size S. latissima sporophytes. We mixed spores from 159 wild sporophytes to produce an F1 generation. Approximately 800 offspring were phenotyped and genotyped using “genotyping by sequencing” (GBS). Then, 25 large-sized individuals were selected as parents for an F2selected population, while 25 random individuals served as parents of an F2control population. Approximately 800 individuals from F2selected and 200 individuals from F2control were phenotyped and genotyped. Parentage assignment revealed considerable variation in transmission between parents. Selfing rates exceeded expectations under random mating in F2selected (0.10), and particularly in F2control (0.73), where it led to high inbreeding levels (FIS = 0.16) and contributed to reducing the expected heterozygosity from 0.20 to 0.18. In F1, 4 % of the sporophytes were apomictic, while in F2selected and F2control the fraction of apomictic sporophytes was 0 % and 1 %, respectively. Additionally, spontaneous triploid offspring were identified on the basis of GBS data at a rate of 0.3 %–0.6 %. We discuss implications of our results for practical propagation and breeding of S. latissima.