Time series barley germination is predictable and associated with known seed dormancy loci

Abstract

Preharvest sprouting (PHS) resistance is essential in malting barley (Hordeum vulgare L.) production to prevent damage caused by late season moisture but may be negatively correlated with malting quality traits. The germination percentage and germination rate (GI) in malting barley are measures of both PHS resistance and suitability for malting. Functional principal component analysis (FPCA) and logistic regression on time series spring malting barley GE and GI were performed to identify genetic regions associated with germination trait changes from 0 to 22 wk after maturity. Both analysis methods identified the SD1 and SD2 regions known to affect germination traits, as well as one novel locus on chromosome 6H at 473 Mbp. Several other associations were detected with a single analysis method. Genomic prediction (GP) was used in conjunction with the time series models to test GE and GI predictability over time. Germination traits of unobserved lines at unobserved times were predicted with high accuracy using ridge regression GP coupled with FPCA or logistic regression. The results demonstrate applicability of time series data analysis in determining genetics of germination trait changes in malting barley and that time series germination is predictable. Threshold traits calculated with these time series models, such as the after-ripening time needed until a line achieves a GE of 95%, can be used as an estimate for dormancy duration, which will be useful in breeding for combined PHS resistance and malting quality.