Stacy A. Bonos, Ph.D. 
Professor and Director of the Cool-Season Turfgrass Breeding Program at Rutgers University 

Currently, more than a dozen cool-season turfgrass species are cultivated for use on golf courses, home lawns, parks, and sports fields around the world. The most popular species belong to the genera Lolium, Poa, Festuca, and Agrostis. Some minor genera, such as Koeleria and Deschampsia, may also be used in specific situations. 

Perennial grasses like turfgrasses present unique characteristics compared to other annual grass crops. First, turfgrasses have very small floral parts, which makes artificial hybridization tedious and manual pollination control a very complex task. Therefore, these practices are not commonly used in the genetic improvement of cool-season turfgrasses. Second, many turfgrass species exhibit strong self-incompatibility, which maintains genetic diversity and reduces inbreeding depression, but also results in highly heterozygous genotypes. This heterozygosity becomes a challenge when attempting to use and develop genomic tools such as whole-genome sequencing and genetic linkage maps. 

In addition, many turfgrass species are polyploid, adding complexity to both the development of genomic tools and the selection of genotypes with improved agronomic traits. Another unique aspect of turfgrasses is that parental germplasm can be vegetatively propagated. This can be advantageous, as it allows breeders to preserve parent lines for maintaining Syn 0 generations (as in the case of Penncross bentgrass [Hein et al., 1958]) or Syn 1 breeder seed, and to recombine parents based on progeny performance. 

Lastly, the primary use of turfgrasses is as mown turf/forage, not just seed production — which is typically the main selection trait in most annual monocot crops. To further complicate matters, plant evaluations in row plots for seed production do not necessarily reflect performance under mowed turf conditions. As a result, there is a need for parallel selection — one for seed yield and one for turf performance — or for modifying spaced plant selection methods to simulate turf conditions (Heineck et al., 2021). Moreover, historically, there has been a negative correlation between seed yield and turf quality (Johnson et al., 2003). However, recent studies have shown no correlation between turf quality and seed yield, suggesting that cultivars can be developed with both high turf quality and improved seed yield (https://turf.umn.edu/news/potential-tradeoffs-between-turfquality-and-seed-yield). 

With the exception of Poa pratensis, most cool-season turfgrasses are cross-pollinated species. Conventional breeding methods for cross-pollinated plants — including ecotype selection, genotypic and phenotypic recurrent selection, development of synthetic and composite cultivars, and modified backcrossing — continue to be used in turfgrass breeding programs, with specific adaptations for turfgrasses (Funk, 1981; Funk et al., 1983; Bonos and Huff, 2013). Turfgrass breeders take advantage of positive unique traits such as the ability to vegetatively propagate superior genotypes, while managing challenges like self-incompatibility limitations.