Data-Driven Identification of Environmental Variables Influencing Phenotypic Plasticity to Facilitate Breeding for Future Climates: A Case Study Involving Grain Yield of Hybrid Maize

Abstract

Phenotypic plasticity describes the ability of a genotype to produce different phenotypes in response to different environments. A key component for the quantification of phenotypic plasticity is the set of environmental variables that influence a particular phenotype. These variables are typically selected using domain-specific knowledge or, when the set of variables is suitably small, exhaustive search. Two factors complicate these strategies. First, environments are shifting and becoming more variable due to global climate change which may introduce novel stresses that are not yet captured by domain-specific knowledge. Second, environments are inherently infinite-dimensional not only in terms of the variables that can be measured and their temporal resolution but also on the timescales at which organisms perceive different environmental variables throughout development. This size makes exhaustive search unfeasible without potentially erroneous simplifying assumptions, especially when assessing the simultaneous influence of multiple environmental variables on a phenotype. To address these challenges, we propose the use of a genetic algorithm to efficiently identify informative sets of environmental variables for the quantification of phenotypic plasticity. We apply this procedure to a hybrid maize dataset and demonstrate its utility for characterizing phenotypic plasticity and identifying directions for future research into the biology of plastic responses.

Keywords: genotype-environment interactions, phenotypic plasticity, maize, genetic algorithm, environmental covariates, Climate Change, breeding for future climates