Leveraging the diversity of Beta maritima populations for sugar beet breeding

Abstract

Sea beet (Beta maritima) populations are of critical importance for sugar beet breeding because they represent a rich reservoir of genetic diversity and adaptive traits that have been lost during domestication. Their ability to thrive in diverse and challenging environments makes them invaluable for improving the resilience of cultivated sugar beet. This thesis investigates the genetic diversity of sea beet populations and their practical relevance for sugar beet breeding, combining a broad literature review, empirical population studies, and simulation-based breeding design. The work addresses the urgent need to broaden the genetic base of cultivated sugar beet, which has been narrowed by intensive breeding, by leveraging the rich genetic resources found in sea beet populations. The first objective was to assess the current state of research on Beta maritima diversity. The literature review reveals extensive morphological and genetic variation across European and North African populations, shaped by geography and local environmental pressures. Phenotypic diversity often exceeds genetic differentiation, emphasizing the influence of environmental factors and the need for genetic analyses to complement morphological studies. The review also highlights significant gaps in the genetic characterization of Mediterranean populations. The second objective was to characterize the genetic diversity and population structure of three Northern Atlantic Beta maritima populations using high-density SNP markers. Within this study, a total of 1,363 genotypes across three populations from Denmark, France, and Ireland were analyzed using 16,201 SNP markers. The findings reveal genetic variation among the populations, with the Irish population exhibiting the highest genetic diversity and pronounced population structure. The Danish population showed low genetic diversity and minimal population structure, while the French population displayed intermediate levels of both. In the Irish population, a pronounced population structure was detected even within a very small geographic area, illustrating that genetic diversity and population structure are shaped by more than just geographical distance. Also, all populations exhibit unique polymorphisms. This highlights the importance of in situ conservation to preserve unique genetic variants, as well as the need for broad sampling and comprehensive testing to fully uncover useful genetic diversity regions. The third objective was to evaluate how population characteristics influence the power and accuracy of genome-wide association studies for detecting quantitative trait loci (QTL). The study found that the low minor allele frequencies within sea beet populations require larger sample sizes to ensure rare alleles are adequately represented and statistical power is maintained. Populations with greater genetic diversity, such as those from France and Ireland, offer more potential for trait mapping, but strong substructure poses analytical challenges that must be addressed to avoid confounding effects. Careful consideration of population structure is essential to avoid false positives and misleading conclusions. Nevertheless, these populations offer great potential when analysis is carried out correctly and population structure is accounted for appropriately. The final objective was to simulate and compare crossing strategies between wild and elite genotypes to optimize mapping populations based on sea beet populations for QTL discovery, especially for complex traits like yield and drought tolerance. Simulation-based studies showed that mapping populations containing a high proportion of sea beet genome (up to 50%) have the greatest power to detect rare alleles, even those present at frequencies below 1%. However, practical breeding requires at least 75% elite genome content for reliable phenotyping, as wild traits can hinder trait evaluation. Crossing designs based on elite × wild beet F1s, followed by testcrossing, strike the best balance between maintaining wild allele representation and enabling proper phenotyping. This approach offers a practical solution for integrating wild genetic diversity into sugar beet breeding programs and can be adapted for other outcrossing crop species. By combining empirical population studies with simulation-based breeding design, this thesis provides a robust framework for leveraging wild genetic resources in breeding. The insights gained in this thesis not only contribute to sugar beet breeding but also serve as a model for utilizing crop wild relatives in other breeding programs, emphasizing the importance of in situ conservation, broad sampling, and integrative approaches.