Location: Hard Winter Wheat Genetics Research
Project Number: 3020-21000-012-034-S
Project Type: Non-Assistance Cooperative Agreement
Start Date: Sep 1, 2026
End Date: Aug 31, 2028
Objective:
The primary objective of the research supported by this agreement is to simultaneously improve both the efficiency and the power of multi-environment yield testing to identify superior breeding lines at all stages of line advancement. The research will optimize sparse testing designs for wheat variety development yield trials in order to accelerate advancement of high-performing wheat lines to cultivar release. This research will couple high throughput digital phenotyping, using unmanned aerial systems, and cost-effective genomic characterization in a pedigree-aware field testing structure. The secondary objective is to use artificial intelligence, machine learning, and other modern data analysis tools to predict the value of new lines as breeding parents at the earliest stages of testing, which will reduce breeding cycle time and accelerate genetic gain.
Approach:
The research will integrate the agronomic performance testing of inbred lines at all stages of testing. Lines in the first year of yield will be evaluated in multi-location trials that also include superior lines identified in prior years in a sparse testing structure. Each year, as superior lines advance, increasing replication across and within environments will be used to strengthen the inference of line performance within and across testing environments.
High throughput digital phenotyping data will be collected using camera(s) mounted on unmanned aerial systems at key wheat growth stages. Genomic data from a core set of ~ 4,000 variant sites will be collected on new lines entering yield testing, and this data will be coupled with genomic data on lines advanced from prior year testing. In combination, these data will support the prediction of performance of lines in trial environments in which they were not tested.
A key element of this project is to improve the trial designs to optimize the assignment of new lines, which only have sufficient seed for testing in one environment, to trial environments using the pedigree structure of lines and their genetic relatedness to advanced lines. AI and machine learning approaches will be used for this optimization.