Integration of phenomics and genomics: biomass development in barley

High-throughput (HT) genotyping technologies used in plant genetic studies are state-of the art procedures while HT-phenotyping methods were unavailable for a long time. The time-consuming work required to phenotype plant populations is known as the phenotyping bottleneck. A complete new field of (semi-) automated phenotyping technologies is rapidly developing.

We applied automated non-invasive daily imaging in the visible light range to screen a diverse spring barley panel for vegetative biomass development in three consecutive experiments in 2012. In total, 100 two-rowed spring barley cultivars and landraces were evaluated originating from 29 countries with 5 plants per genotype in one experiment. Image data were used to estimate plant biomass development, by calculating ‘digital biomass’ - a pixel volume showing a high correlation to plant fresh weight. A logistic model was fitted to the growth curves of individual plants that reveal the maximum vegetative biomass, the intrinsic growth rate and the point of inflection, where growth is maximal.

Genome-wide association studies (GWAS) were performed using the 9K iSelect array for early to late biomass as well as for the three logistic model traits. A number of 4873 SNPs was informative for our collection and 4224 of them have known genome positions. Population structure analysis was performed with Structure but resulted no useful Q-grouping. Also PCA showed no clear structuring and for GWAS the kinship matrix calculated in Tassel 2.1 was further used for controlling population structure effects. Using a mixed-linear model including the kinship, altogether 48 SNPs were significantly (-Log(p)>3) associated with biomass development over time using Tassel 2.1. GWAS for early and late biomass development resulted in different loci. Validation how vegetative biomass development is related with grain yield characteristics is currently performed in a experiment lasting until maturity is reached with a selected subset of 26 genotypes.