This project highlights one component of a broader portfolio of center-funded research initiatives focused on developing novel approaches to complex scientific challenges.

ABSTRACT

Objectives

Developing new turfgrass varieties is slow and expensive. A major bottleneck is the nursery screening phase, where breeders spend three years evaluating thousands of seedlings across multiple field sites—only to advance fewer than 10% to advanced testing. Genomic selection offers a way to bypass this bottleneck by predicting plant performance from DNA markers alone, potentially cutting nearly a quarter of the time needed to release new cultivars, and freeing up resources for more extensive testing of promising individuals. The goal of this project is to extend our previous proof-of-concept work into actual breeding populations in the NC State St. Augustinegrass program. We aim to: (1) establish an affordable, high-throughput genotyping protocol and genotype >700 individuals from early-stage breeding crosses; (2) train genomic prediction models (GBLUP) and evaluate their accuracy using cross-validation and between-year prediction; and (3) quantitatively estimate how much genomic prediction can improve breeding program efficiency—in terms of both genetic gain and cost.

Key Findings and Progress

In our previous Turfgrass Center–funded work, we showed that genomic prediction is effective in a diverse panel of 157 St. Augustinegrass accessions, achieving prediction accuracies of 0.28–0.74 across six traits. This was a promising result, because prediction in diverse, unstructured populations is typically harder than in the structured families used in breeding programs. During this project period, PhD student Jaswinder Kaur has made substantial progress on Objective 1. She identified a high-throughput, low-cost reduced-representation genotyping protocol and is working with a third-party vendor (LGC) to bring it to production. Specifically, she completed quality-control filtering on existing genetic data to identify markers suitable for high-throughput genotyping, and oversaw trial runs of high-throughput DNA extraction by LGC—a key first step in the pipeline. Leaf tissue from the >700 nursery individuals has already been collected and is in storage, ready for DNA extraction and genotyping once the protocol is finalized. In the coming FY, Jaswinder will continue to finalize the genotyping protocol and apply it to the >700 breeding program individuals, while curating phenotypic data in preparation for fitting genomic prediction models.