Two New Strategies for Detecting and Understanding QTL × Environment Interactions

• Published in: Crop science Vol. 51; no. 1; pp. 96 - 113
• Main Authors: Gauch, Hugh G., Rodrigues, Paulo C., Munkvold, Jesse D., Heffner, Elliot L., Sorrells, Mark
• Format: Journal Article
• Language: English
• Published: Madison, WI The Crop Science Society of America, Inc 01.01.2011; Crop Science Society of America; American Society of Agronomy
• Subjects: Accuracy; Agronomy. Soil science and plant productions; Biological and medical sciences; Factorial experiments; Fundamental and applied biological sciences. Psychology; Genetics; Genetics and breeding of economic plants; Genotype & phenotype; Grants; Life sciences; Noise; Principal components analysis; Statistical methods; Studies; Varietal selection. Specialized plant breeding, plant breeding aims; Quantitative trait loci; Genotype environment interaction; Detection
• ISSN: 0011-183X; 1435-0653
• DOI: 10.2135/cropsci2010.04.0206

Two new strategies are proposed to improve the detection and understanding of quantitative trait loci (QTL), especially those exhibiting QTL × environment interactions (QEI), in the context of experiments conducted in multiple environments. First, a parsimonious Additive Main effects and Multiplicative Interaction (AMMI) model is applied to the phenotypic data to gain accuracy and thereby to increase the logarithm of odds (LOD) scores for QTL detections. Second, the environments are ordered by AMMI parameters that summarize genotype × environment interaction information to reveal consistent patterns and systematic trends that often have an evident ecological or biological interpretation. The combination of greater accuracy for the phenotypic data and systematic trends for the environments provides for more consistent and understandable QTL results. These new strategies are illustrated with two examples: preharvest sprouting scores of a biparental wheat (Triticum aestivum L.) population from 14 environments spread over 5 yr, and yield for a doubled‐haploid barley (Hordeum vulgare L.) population tested in 16 environments. AMMI parameters can also provide successful predictions of entire QTL scans for new environments. The statistical methods developed here are of great generality, applicable across microbial and plant populations grown in multiple environments, and may be adapted to animal and human genetic studies.