Method Developments to Identify Loci and Selection Patterns Associated with Genotype by Environment Interactions in Soybean

Authors

Mary M. Happ, University of Nebraska - LincolnFollow

First Advisor

David L. Hyten

Date of this Version

Summer 7-28-2023

Document Type

Dissertation

Citation

Happ, M. (2023). METHOD DEVELOPMENTS TO IDENTIFY LOCI AND SELECTION PATTERNS ASSOCIATED WITH GENOTYPE BY ENVIRONMENT INTERACTIONS IN SOYBEAN.

Comments

A DISSERTATION Presented to the Faculty of The Graduate College at the University of Nebraska In Partial Fulfillment of Requirements For the Degree of Doctor of Philosophy, Major: Agronomy & Horticulture (Plant Breeding & Genetics), Under the Supervision of Professor David L. Hyten. Lincoln, Nebraska: July 20, 2023

Copyright © 2023 Mary M. Happ

Abstract

For many complex traits such as grain yield, genotype by environment (GxE) interactions are a prevalent source of phenotypic variation. Exploring the capacity of different methodologies to help describe and quantify the GxE interaction landscape for grain yield is an important step in informing plant breeders what the most viable strategies for management and exploitation of GxE may be. In this endeavor, we compared the results from multiple genome wide association studies (GWAS) that used either stability estimators as a phenotype to capture GxE variance, or directly mapped GxE in a mixed model for yield. Leading into this study, a method was developed to enable the cost-effective ascertainment of genotypic information via skim sequencing supplemented with imputation, where it was discovered that imputation accuracy could be maintained at ~98% down to a 0.3X genome coverage. This imputed genotype information was then used in the GWAS analysis that leveraged data from 213 elite local breeding lines tested over the course of three years at multiple sites in eastern Nebraska. Results from the GWAS showed minimal overlap in quantitative trait loci (QTL) discovered between the modeling methods, and that the majority of QTL discovered displayed a crossover effect. These results prompted our final investigation, where high depth sequencing data was obtained for our study population and used to investigate the effect of artificial selection on genomic windows contributing to GxE interactions. As part of this exploration, several improvements were introduced in the modeling procedure to avoid the inherent biases associated with comparing variance estimates to selection statistics. It was determined through this combination of novel methods that GxE experiences less directional selection pressure than main genetic effects. Interestingly, in contrast to the GWAS results, this also revealed a rich landscape of small, conditionally neutral loci drove the majority of GxE interactions and appeared to be under more directional selection that other GxE effect types.

Advisor: David L. Hyten