Plant breeding programs exert considerable effort evaluating new breeding lines across many locations to identify superior performing entries for release as new varieties. For this evaluation, regional testing networks have developed to provide additional information to breeders on the adaptation and performance of their lines. Previously, the information collected in these trials, was only applicable on the line per se. With the development of inexpensive, high-density genetic markers, whole-genome profiles can be obtained for every experimental breeding line and can be used to create genomic selection prediction models that take into account alleles common between lines even if they were not jointly evaluated across time or space. The developed models can then be used to predict phenotypes and pre-select new lines based on their allelic combinations. This new breeding approach shifts the unit of importance from a line per se to individual alleles. With this new focus, historical datasets from regional trials can be used for developing new prediction models, taking advantage of considerable investment in previous years of trials.
The Hard Winter Wheat Regional Nursery Program is an 81-year-old nursery established by the USDA to characterize performance and quality of near-release wheat varieties from breeding programs in the central plains (KSU, UNL, OSU, CSU, TAMU, SDSU, and private programs). Entries are submitted annually and genetic gain is measured across years by including multiple long-term check cultivars for comparison. Phenotypic data collected from the nurseries includes grain yield, test weight, plant height, and lodging.
The Southern Regional Performance Nursery , which is my current focus due to seed availability and geographic location, is grown at more than 30 locations each year across the Midwest. I’m working with seed stocks that date back to 1992 and contain a majority of the experimental lines tested by participating wheat breeding programs across the region as well as the varieties released by those programs. While locations have varied in use from year to year, twenty-two locations have been included in the nursery for more than ten years, allowing analysis and characterization of environment in addition to genotype. The 907 lines are currently in the process of being genotyped. Further work will determine the most suitable training population structure, determine the available prediction power from the germplasm, and compare prediction models across the set. In conjunction, I’ll quantify the environment in terms of effect on yield and be able to derive a more accurate assessment on reported genetic gain of wheat across the Midwest since 1992. This research will refine GS models that have the ability to leverage preexisting regional testing networks for increased genetic gain.
I am also working to integrate technology into agriculture programs. The ubiquity of computers and relatively inexpensive upfront cost for consumer electronics makes it irrational to ignore the potential of these tools to save time and make data collection easier. In the past, learning curves, lack of infrastructure, and cost barriers have prevented software design, development, and use in agronomy. One of my goals during my time at Kansas State is to lower these barriers and promote the wide use of a standard set of applications among all researchers. This vision has been started with the release of several free and open source software tools. I'm also responsible for updating and maintaining the lab website.