Improving yield and essential oil content in annual caraway accompanied by utilization of molecular-genetic methods

Abstract

Caraway (Carum carvi) is a species of the Apiaceae family with a long history of usage and cultivation. Fruits or extracted components are used as spice, in pharmaceuticals and for sprout inhibition in potato. Caraway cultivation in Europe was based on biennial varieties until annual varieties were introduced in the 1990s. Annual cultivation promises a reduction of production costs and better integration into crop rotation. However, current annual varieties do not gain satisfying yields and essential oil contents. Moreover, caraway cultivation in Germany becomes riskier due to ongoing climatic changes.

Therefore, a breeding project was started to increase yield potential and essential oil content in annual caraway and to develop material adapted to changing climatic conditions. To achieve yield increases, breeding of synthetic varieties should be tested as potential method in caraway breeding. To some extent, this method allows the systematical exploitation of heterosis. Breeders expected to find strong heterosis due to the assumed outcrossing nature of caraway. However, neither heterosis nor predominant outcrossing in open pollination had been experimentally verified for caraway. Hence, gathering first insights into heterosis and into the outcrossing rate of caraway were major scientific goals of this thesis. Moreover, breeding of caraway is hampered by a small amount of phenotypic and genotypic data on available germplasm. To lay a sound foundation for new breeding programs, a broad phenotypic and genotypic evaluation of available germplasm should be conducted within the project. This thesis is focused on the genotypic evaluation.

For phenotypic and genotypic evaluation, a set of 137 accessions consisting of 67 annual and 70 biennial accessions was composed (chapter 1, von Maydell et al. 2021a). A GBS was chosen for genotyping and led to the generation of a set of 13,155 SNP markers. This approach was unique in its breadth and depth for caraway breeding research. Population structure was investigated by using a PCoA, a Bayesian clustering analysis, phylogenetic trees and a neighbor network. Across all analyses a separation of accessions into two subpopulations was supported, which was associated with the flowering type. AMOVA indicated a rather low genetic variation between the two subpopulations of 7.84%. In addition, flow cytometry was conducted for 35 accessions to determine the genome size of caraway. An average genome size of 4.282 pg/2C was estimated, which is about half the size indicated in literature. In general, it was corroborated that GBS can be a robust method to analyze population structure and genetic diversity in minor crops like caraway without a reference genome. With further downstream applications, GBS should turn out to be highly cost-efficient and sustainable. In a subsequent study, generated SNP markers were already successfully be converted to diagnostic markers.

In this study, selected SNP markers were used to estimate the outcrossing rate in caraway (chapter 2, von Maydell et al. 2020). Selected annual inbred lines were placed in a neighbor-balanced polycross design to produce F₁ populations by open pollination. The genotyping system PACE was implemented to genotype seven produced F₁ populations. Moreover, a high-throughput DNA extraction was adopted. In total, more than 1,300 individual plants were analyzed. The outcrossing rate ranged from 51.6% to 82% and was on average 66.5%. Therefore, predominate outcrossing was confirmed for caraway. However, a large proportion of self-fertilization was found as well, so that the mating system of caraway can be defined as a mixed mating system. The estimation of the outcrossing rate in more genotypes, years and environments is desirable. Yet, those first insights indicated that heterosis should be strong in caraway, but due to incomplete outcrossing maximum potential heterosis might not be exploited in F₁/Syn₁ generation.

Finally, heterosis and GCA were investigated based on the 18 F₁ populations generated in the polycross (chapter 3, von Maydell et al. 2021b). 18 inbred lines were compared with 18 corresponding F₁ populations and four standard cultivars in a randomized complete block design with two years of growing and four repetitions per year and genotype. As pivotal finding, better-parent heterosis was detected. Significantly, F₁ populations showed a higher yield, thousand-grain weight and height and an earlier beginning of flowering, end of flowering and maturity than corresponding inbred lines. However, F₁ populations did not considerably outperform standard population cultivars. It remains to be seen whether the combination of the best components to a synthetic variety finally leads to the desired yield increases. Additionally, strong negative correlations were found between developmental traits and yield so that selection of early developing genotypes seems highly recommendable.

To conclude, this thesis can be seen as a pioneering step towards implementing heterosis breeding and utilizing molecular-genetic methods in caraway breeding. Results indicate that yield increments can only be accomplished by a breeding method that enables the exploitation of heterosis. Hence, breeding of synthetic varieties could be the method of choice. Genotypic data could be a valuable aid in selection decisions in new breeding programs. Moreover, genotypic data and the implemented molecular-genetic methods could be the foundation for future genetic analyses in caraway.