Understanding Environmental Variables Influencing Hybrid Maize Performance Across the United States Corn Belt

Ani Anna Elias, Purdue University

Abstract

Maize (Zea mays L., common name: corn) is the most widely cultivated cereal food crop in the United States and is globally ranked first in production. The agronomic performance of hybrid maize is influenced by characteristics of the environment where it grows. This poses a challenge to plant breeders, agronomists, and crop producers in selecting a genotype that performs well over diverse environments. To understand the extent of the influence of environment on newly developed genotypes, hybrids need to be tested in multi-environmental trials (MET) before commercial release into a particular target environment. It is also important to link environmental effects to the developmental phases of the plants to provide a better understanding of environment interactions. Modern plant breeding evaluation techniques use augmented field designs with unreplicated test genotypes in MET. In this study, nested random regression models (NRRMs) are employed for the purpose of estimating genotype by environment interactions (GEI) and to identify environmental indices responsible for GEI. NRRMs model genotypic performance using random regression on environmental indices from various developmental phases to pinpoint the key indices for each stage of development. NRRMs are investigated via computer simulation studies to assess the range and effect of both genotype and environmental variables. The simulation results revealed that the NRRMs successfully partitioned variation for large MET studies. As shown by simulation studies, MET studies with between 10 and 100 locations and between 30 and 800 genotypes can accurately be evaluated using NRRMs with over 250 environmental indices. Further, based on the information obtained from NRRMs, strategies are developed for efficiently clustering locations, to ensure homogenous performance during commercialization of test genotypes. In actual data modeling, genotypic performance from METs conducted in various zones across the United States Corn Belt was evaluated for the purpose of negating any possible influence due to relative maturity. Based on these results, METs from the northern part of the Corn Belt are shown to be influenced by environmental characteristics during the vegetative stages while those in the middle of the Corn Belt are influenced by environmental conditions from the vegetative through grain filling stages. Environment clustering strategies, where locations with similar environmental characteristics are clustered based on the environmental indices, proved to be superior to other strategies except in a few examples. Identification of environmental influence and utilization of these effects to cluster locations can provide better information in decision making during commercialization of chosen test genotypes.

Degree

Ph.D.

Advisors

Doerge, Purdue University.

Subject Area

Agronomy

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS