ABSTRACT
This chapter extends the reaction norm model for genomic prediction, using canopy coverage image data. It provides a brief description of the SoyNAM phenotypic and marker data. The chapter describes how the canopy coverage image data were collected, and why they have the potential to increase prediction accuracy when included in the prediction models, compared with traditional genomic prediction models. It describes the statistical models and cross-validation (CV) schemes used for genomic-enabled prediction. The chapter compares nine prediction models, with three different CV schemes for yield and date to maturity, for the SoyNAM data set. It compares the effects on prediction models when considering canopy data captured in only the early stages of the growing season instead of the entire growing season and discusses the results and some future research avenues. Grain yield in soybean is influenced by genetic and environmental factors and their interactions.
