Integrating genomic selection and speed breeding to increase genetic gain in spring wheat (Triticum aestivum) breeding

Abstract

There is a need to develop novel breeding approaches to increase the rate of genetic gain for wheat yield due to rising demand and climate volatility. Genomic selection (GS) can increase genetic gain in quantitative traits, such as yield, principally through a shorter selection cycle. 'Speed breeding' (SB) could further accelerate the breeding cycle by using extended photoperiod in the glasshouse, enabling rapid generation advance. An integrated breeding approach could include fast inbred line development, phenotyping of yield secondary traits for multivariate GS and indirect phenotypic selection for important traits prior to field trials in a target environment. The objective of this thesis was to explore a comprehensive breeding strategy that combines the aforementioned tools, aimed at increasing the rate of genetic gain in wheat. Firstly, to investigate the potential for SB application across major crop species, a sample of wheat, barley, canola and chickpea cultivars were evaluated under either 22-hour light (SB) or natural, diurnal lighting in a temperature-controlled glasshouse to quantify and compare the effects on development. Wheat and barley were also grown at high densities to investigate potential single seed descent (SSD) programs. Time to anthesis was significantly reduced for all species relative to the natural photoperiod conditions. Coupled with harvesting seed two weeks post-anthesis, SB could enable up to six generations of wheat per year. Grain and spike number were similar between lighting treatments and germination percentage was higher in grain of SB-grown plants at two weeks after flowering. These findings indicated that SB had potential to be used at multiple stages of a breeding program. In order to identify traits that could be measured under SB (SB traits) that were correlated to field-based yield and could be used in multivariate GS, an explorative study was carried out on a 135-line, bi-parental (SeriM82 x Hartog, SxH) bread wheat population. Days to anthesis (SB-DTA), plant height (SB-height), spike length and flag leaf length were measured under SB and narrow-sense heritability and genetic correlation with field-based yield was determined from previously performed field trials in three environments that varied with respect to extent and timing of water deficit. Field-based traits were also measured, including DTA, plot height and two senescence indices. The heritability of most SB and field-based traits were higher than yield heritability and SB-DTA and SB-height showed strong genetic correlations with yield (up to-0.80 and-0.66 respectively) in the irrigated environment. Further investigation with a larger, more diverse population was carried out using a previously developed 256-line, multi-reference bread wheat population, and data from two previous rain-fed yield trials. Multivariate models iii incorporating the field-based senescence indices significantly increased yield prediction ability although the addition of SB traits did not improve prediction above that of univariate mo...

Benefit of Introgression Depends on Level of Genetic Trait Variation in Cereal Breeding Programmes

“…Two traits, disease resistance and grain yield were simulated, with narrow-sense heritabilities of 0.5 ( Arabi, 2005 ; Xi et al, 2019 ) and 0.25 ( Watson, 2018 ; Guo et al, 2020 ; Tsai et al, 2020 ), respectively, at two environments. The genetic correlation between the two environments was 0.8 for disease resistance (low level of genotype-by-environment interaction (G × E)) and 0.4 for grain yield (high level of G × E; Li et al, 2017 ).…”