Open access publications in the SLU publication database

Abstract

Small and structured populations are problematic for breeding due to low levels of genetic variability and increased levels of linkage disequilibrium (LD). The increased LD causes problems with utilization of common genomic tools and may lead into increased accruement of deleterious alleles through the intense selection and bottlenecks characteristic of breeding. In forest trees, maintaining a large breeding population to alleviate these issues is difficult due to large physical size of most trees. The aim of this thesis was to study and resolve some of the LD dependent issues of small populations and establish small populations as an option in breeding. We established the accuracy of LD-based estimation methods of recombination using data from European aspen, allowing confident use of these methods later in the project. The small black cottonwood breeding population contained enough genetic diversity to facilitate future adaptive selection to novel Swedish climate and light conditions, suggesting that small populations consisting of offspring with diverse parentage are feasible options for breeding in species with high outcrossing and recombination rates. We identified candidate genes that can be targeted for selection on phenology and growth using a genome wide association study (GWAS). We also show that GWAS is useful for identifying large effect alleles even in small populations and that efficient growth under novel conditions likely require different allele combinations than in native habitats. Finally, we quantified deleterious load and identified effects of the deleterious load on growth. Accounting for deleterious load allowed for more effective genomic selection and increased breeding cycle gain in breeding programs based on small populations.