Methods for marker assisted breeding in Octoploid Strawberry (Fragaria × ananassa)

He, J. (2021) Methods for marker assisted breeding in Octoploid Strawberry (Fragaria × ananassa). PhD thesis, University of Reading. doi: 10.48683/1926.00108101

Abstract/Summary

Strawberry is an economically important crop with global and UK production on an upwards trend. Strawberry breeding efforts attempt to generate novel varieties that have increased yields, resistance to pathogens, good eating quality and high nutritional content. Genomic prediction (GP) is an advanced marker assisted prediction (MAP) technique that makes predictions about agronomically important traits in crops. Three areas of improvement were identified to assist commercial deployment of GP for strawberry. Breeding efforts currently rely primarily on visual and mechanical measurement of plant phenotypes, which is slow, imprecise and liable to human biases. A strawberry phenotyping platform was developed that captured images from 360° around the straw�berry fruit to generate 3D representations. Seven fruit quality traits were calculated from the representations, which showed good concordance with manual measurements. Deployment of the system could lower phenotyping costs, increase throughput, increase precision and thus improve GP accuracy. Current genotyping approaches for dense marker panels in strawberry are too expen�sive for commercial deployment. A rational design process was implemented to generate amplicon sets that would genotype a panel of markers to maximise information for GP, with scalability to accommodate resources available to different breeding programmes. The design process failed to generate marker information due to unexpected interaction in the multiplexed PCR reaction. The relative effectiveness of phenotypic prediction and MAP in strawberries is unclear. Moreover, existing models of GP in strawberry do not represent all the traits of interest to breeders. Between years predictions of 15 fruit quality traits were implemented using phenotype only, traditional MAP (tMAP) and GP models. GP had similar selection accuracy compared to phenotype only prediction, but tMAP performed significantly worse than the other models. It was concluded that GP would likely yield benefits to strawberry breeding in the context of speed breeding.

Item Type	Thesis (PhD)
URI	https://centaur.reading.ac.uk/id/eprint/108101
Identification Number/DOI	10.48683/1926.00108101
Divisions	Life Sciences > School of Agriculture, Policy and Development
Download/View statistics	View download statistics for this item