Investigating the genetic basis of complex traits is critical for crop improvement. The focus of my project is to study genotype by environment interaction (G Ã Â E) of flowering time and heterosis of plant height in sorghum. These two traits, flowering time and plant height, were measured on a recombinant inbred line (RIL) population in 7 environments (Guayanilla Puerto Rico and Manhattan Kansas in 2011 and 2012, Ames IA in 2013 and 2014, and Santa Isabel Puerto Rico in summer 2014). As a follow up of the flowering time QTL mapping study, we used RNA-sequencing (RNA-seq) to study gene expression patterns underlying flowering time. Leaf tissue RNA was extracted from the two parents (Tx430 and P898012) and four RILs with different flowering time. During the 2013 growing season, four time points at Iowa and three time points at Puerto Rico were selected to do RNA extraction. To further elucidate the allele effects of plant height QTL, representative RILs with different allele combinations were selected for crossings during summer 2014 at Iowa. The F1 hybrids were planted together with the parental lines in Puerto Rico nursery during winter 2014. Plant height components were measured as the same for the RIL population.

Significant G Ã Â E was observed for flowering time but not for plant height. I found that effect direction change and conditional neutrality are the main causes of the observed G Ã Â E. Two major QTL identified for flowering time showed epistatic interaction. The combined effect of day length and temperature is the major factor underlying the observed phenotypic plasticity and QTL effect changes. Genomic prediction using the joint regression analysis approach provided a framework to link environmental factors with genomic information to predict crop performance in diverse environments. The RNA-Seq experiment identified genes that show G Ã Â E expression patterns, providing evidence for the involvement of the genes in flowering time regulation.

Two QTL identified for plant height are located 29 cM apart on chromosome 7 and are in repulsion linkage. By analyzing different plant height components, the QTL next to Dw3 on chromosome 7, qHT7.1, shows effect on both the upper and lower part of the stem, while Dw3 only affects the lower part of the stem. Genome-wide association study (GWAS) in a diverse sorghum population confirmed this QTL. When two inbred parents have repulsion linkage with opposite effects at the two QTL, the hybrid can show heterosis in plant height. Computer simulation showed that repulsion linkage could influence the separating of two closely linked QTL and lead to overestimated dominant effect and underestimated additive effect. This study provides an example of heterosis caused by pseudo-overdominance and a tool box for sorghum improvement for biomass or grain production.