Genome-wide estimation of heritability and its functional components for flowering, defense, ionomics, and developmental traits in a geographically diverse population of Arabidopsis thaliana.

Narrow-sense heritability (portion of the total phenotypic variation attributable to additive genetic effect, h) is a critical parameter in plant breeding and genetics, but its estimation is difficult for populations with unknown pedigree information. This study applied a marker-based linear mixed model (LMM) analysis to estimate narrow-sense heritability and its seven functional components corresponding to SNPs in coding and noncoding regions for each of 107 flowering, defense, ionomics, and developmental traits in an Arabidopsis (Arabidopsis thaliana) population of 199 inbred lines with unknown genetic relatedness. Genetic relationship matrix (GRM) based on 214 051 SNPs and component GRMs based on seven subsets of SNPs were computed for LMM estimation of h and functional components contributing to h, respectively. The h estimates for flowering traits were higher than those for defense, ionomics, and developmental traits, supporting a general view that the fitness-related traits have lower heritabilities than other traits. The function component owing to SNPs in coding (exon) regions was the least contributor to h. Our LMM analysis provides an opportunity to gain a comprehensive view on heritability and its functional components for populations with unknown structure but with genome-wide DNA markers.