PhD defense: Kedra Mohammed Ousmael

Kedra Mohammed Ousmael defends her thesis:

Implementation of high-throughput DNA markers for efficient breeding tactics of forest tree species

Zoom link

Supervisors:
Associate Professor Ole Kim Hansen, IGN
Senior Researcher Jon Kehlet Hansen, IGN

Assessment Committee:
Associate Professor Maria del Rosario Garcia Gil, SLU - Sweden
Researcher Laurent Bouffier, INRAE - France
Associate Professor Anders Ræbild (chair), IGN

Summary:
Traditional tree breeding programs are hindered by their retrospective nature, focusing on a limited number of economically important species, and requiring significant time and resources. Therefore, and in response to climate change and the growing population, there is a need for new and more flexible breeding strategies suitable for a wider range of species. The challenge lies in balancing the resource intensive long-term genetic improvement with the urgency of addressing evolving environmental pressures. This thesis explores the use of highthroughput DNA markers to enhance tree breeding efficiency, particularly focusing on trees of lower economic interest with limited resources for breeding programs. The thesis includes four papers/manuscripts, and a brief overview is presented as follows: 
First, the effectiveness of partial SNP genotyping in increasing breeding efficiency and managing genetic diversity was evaluated. For this purpose, we used breeding seedling orchards (BSOs) – the primary breeding effort − of Cordia africana located in Ethiopia. Partial genotyping using the reduced representation-based SNP genotyping technique named DArTseq was used to estimate realized genomic relatedness between individuals in the BSOs. Genotyped and non-genotyped individuals were co-evaluated using a single-step approach. Genotyping a subset of the individuals not only improved genetic parameter estimates but also enabled diversity management by revealing inbreeding and genetic differentiation among provenances.
Recognizing the challenge posed by the scarcity of genomic resources in many tree species, the thesis work proceeded with investigating cost-effective methods of identifying markers in trees with no genomic resources. Firstly, the issue of trees with mega genomes was addressed by utilizing a reference genome of a closely related species, an own de novo assembly and
reference free approaches. Secondly, the simultaneous identification of markers in multiple angiosperm species by leveraging conserved regions within their genomes was explored, aiming to streamline the process. For the first case, Nordmann fir was used, a conifer with a large and complex genome. A combination of the above-mentioned techniques identified a high density of SNPs (up to 1 SNP every 176 bases), and a subset of them was validated via identity and genomic relatedness analysis. For the second case, universal target capture probes were used to sequence specific regions across 11 different angiosperm species and assess the presence of sufficient within species variation inside those regions for kinship analysis. Sufficient SNPs were identified in all 11 species and validated through cluster analysis, genomic relatedness analysis, and comparison to genome-wide SNPs. Cluster analysis grouped individuals according to their families. The relatedness values from angiosperm353-based SNPs correlated highly (r = 0.95, p < 0.001) with the values obtained using thousands of genome-wide DArTseq SNPs in C. africana. Potential for scaling up was shown through successful design and in silico primer amplifications in Alnus glutinosa and Faidherbia albida, with 85.3% and 87% success rates, respectively. Additionally, 57% of A. glutinosa primers were transferable to Alnus rubra.
Lastly, fast and efficient methods of incorporating markers in breeding to address occasional disorders or newly emerging challenges were investigated, exemplified by the study of current season needle necrosis in Nordmann fir. A total of 1545 randomly selected reference trees from two production stands (Mariager and Tørring) established with seeds from the FP.266
Skibelund clonal seed orchard were utilized for ad hoc breeding to enable backward selection among the parents. Additionally, to complement the ad hoc breeding, which employs quantitative genetic analysis, 495 moderately to severely damaged trees (Sel-495) were specifically selected from the most affected area at Mariager to screen parents based on the proportion of their offspring in these affected trees. The majority (55.6%) of the offspring in Sel-495 came from 20% of the clonal parents in FP.266. The low narrow-sense heritability (h2 = 0.08 ± 0.04) and low predictive accuracy (average in Mariager = 0.32 (0.12); average in
Tørring = 0.35 (0.12)) indicate that neither the simple ranking based on the proportion of affected offspring nor ranking based on breeding values offer a perfect solution. Therefore, a combination of breeding value-based ranking and ranking based on over-frequency of offspring in the Sel-495 trees emerged as a more effective approach for selection against the disorder.
Overall, the thesis provides practical solutions for developing and utilizing high-throughput DNA markers to improve tree breeding efficiency and implement new breeding tactics, ultimately contributing to genetic diversity management and sustainable forestry practices.

A digital version of the PhD thesis can be obtained from the PhD secretary at phd@ign.ku.dk before the defence. After the defence the thesis will become available from the Royal Danish Libary at kb@kb.dk