Assessment and implementation of new breeding methods in the Kansas State winter wheat breeding program

Abstract

Growing populations and shifting climatic conditions are placing constraints on global food security that have not previously been experienced. Novel technologies are being developed to combat this challenge at many levels including crop germplasm improvement. It is unlikely that all of these technologies will provide a significant benefit to crop breeding programs whilst still being economically viable. The use of high-throughput phenotyping (HTP) in crop breeding programs is becoming more commonplace due in part to the relatively low establishment costs for several of the technologies. Genomic prediction models are also becoming more common in crop breeding programs due to their success in animal breeding schemes and decreases in sequencing costs for genotyping. Here we examine the utility of several HTP technologies and genomic prediction in wheat breeding in Kansas. An uncrewed aerial system (UAS) measuring reflectance values at different bandwidths was used to formulate vegetation indices (VIs) which are known to correlate with economically valuable phenotypes. The genetic architecture of these VIs was examined in an association mapping population of winter wheat (Triticum aestivum) and their use for genomic prediction was examined in the Kansas State University (KSU) winter wheat breeding program. Several economic and population parameters were determined under which genomic prediction would be favored in the breeding program. Based on simulated and empirical observations of model accuracy the KSU breeding program could potentially make larger genetic gains using genomic prediction than traditional phenotypic selection methods.