Enhancing Alfalfa Breeding through Genomic Prediction with Exotic Germplasm Resources

Authors: CHEN, SHUFEN - Breeding Insight; LIN, MENG - Breeding Insight; Tilhou, Neal; ZHAO, DONGYAN - Breeding Insight; BEIL, CRAIG - Breeding Insight; Riday, Heathcliffe; SHEEHAN, MOIRA - Breeding Insight

Submitted to: Plant and Animal Genome VX Conference Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 11/14/2024
Publication Date: N/A
Citation: N/A

Interpretive Summary: Alfalfa (Medicago sativa L) is the most important forage legume globally, valued for its high biomass yield and nutritional value in animal feeding. Increasing genetic diversity using exotic germplasm collections in alfalfa breeding can reduce the risk of crop vulnerability due to genetic bottlenecks and genetic drift. However, the lack of genetic relationship between the breeding germplasm and exotic collections can be a challenge to perform genomic prediction during the incorporation of the exotic germplasm. To establish effective genomic prediction workflows for alfalfa using exotic germplasm, a set of 778 alfalfa samples collected from different geographic origins (CASIA, EURO, OTTM and SYBR) and U.S. breeding materials (CHECK) were evaluated in a multipleyear field trial for plant height and vigor. Using genotyping results of the alfalfa 3K DArTag panel, these five groups of materials showed distinctive patterns of linkage disequilibrium patterns. In the first tested genomic prediction scheme, the predictive abilities were low (-0.065 ~ 0.351) for one of the five groups using the rest four groups in the training set. To address this challenge, in the second scheme, a training set was built to include four groups of samples and an addition of 50% of the fifth group samples, and the predictive abilities for the rest 50% of the fifth group samples increased by ~90%. To further enhance the predictive ability, in the third scheme, the training set was designed to include the same proportion of samples from each of the five groups but with an increasing sample size ranging from 10% to 90% of overall samples by adding 10% of the entire population each time. It was shown that the predictive abilities of the tested 10% samples were improved as the size of the training set increased. Overall, this study indicated that the incorporation of exotic germplasm would require the evaluation of the newly introduced base population and their inclusion in the training set for a successful implement of genomic selection in alfalfa.

Technical Abstract: Alfalfa (Medicago sativa L) is the most important forage legume globally, valued for its high biomass yield and nutritional value in animal feeding. Increasing genetic diversity using exotic germplasm collections in alfalfa breeding can reduce the risk of crop vulnerability due to genetic bottlenecks and genetic drift. However, the lack of genetic relationship between the breeding germplasm and exotic collections can be a challenge to perform genomic prediction during the incorporation of the exotic germplasm. To establish effective genomic prediction workflows for alfalfa using exotic germplasm, a set of 778 alfalfa samples collected from different geographic origins (CASIA, EURO, OTTM and SYBR) and U.S. breeding materials (CHECK) were evaluated in a multiple-year field trial for plant height and vigor. Using genotyping results of the alfalfa 3K DArTag panel, these five groups of materials showed distinctive patterns of linkage disequilibrium patterns. In the first tested genomic prediction scheme, the predictive abilities were low (-0.065 ~ 0.351) for one of the five groups using the rest four groups in the training set. To address this challenge, in the second scheme, a training set was built to include four groups of samples and an addition of 50% of the fifth group samples, and the predictive abilities for the rest 50% of the fifth group samples increased by ~90%. To further enhance the predictive ability, in the third scheme, the training set was designed to include the same proportion of samples from each of the five groups but with an increasing sample size ranging from 10% to 90% of overall samples by adding 10% of the entire population each time. It was shown that the predictive abilities of the tested 10% samples were improved as the size of the training set increased. Overall, this study indicated that the incorporation of exotic germplasm would require the evaluation of the newly introduced base population and their inclusion in the training set for a successful implement of genomic selection in alfalfa.