Potential for the improvement of turf quality in crested wheatgrass for low-maintenance conditions

With the exception of the undesirable characteristic of summer dormancy and the accompanying low aesthetic value, crested wheatgrass has many desirable characteristics in semiarid environments, making it a promising candidate for lower water use turf. Using a population of 27 half-sib families, this study characterized the underlying genetics of turf quality (based on a 1–9 rating scale) of crested wheatgrass and compared the performance of crested wheatgrass turf with traditional control cultivars (‘Cody’ buffalograss, ‘Gazelle’ tall fescue, ‘Manhattan 3’ perennial ryegrass, and ‘Midnight’ Kentucky bluegrass) over 2 years under space-planted conditions. Heritability estimates were generally high (h = 0.44 to 0.84) and suggested a strong additive genetic component for crested wheatgrass turf quality throughout the summer months. Genotypic correlations among the monthly turf quality scores were very high (greater than 0.90) indicating a strong commonality for the genetics underlying turf quality during any point in the growing season. Thus, a breeding program aimed at improving turf quality in this population of crested wheatgrass would stand a good chance for success. However, primarily as a result of summer dormancy, the crested wheatgrass turf performed poorly compared with the control cultivars during late spring and early summer. Turf quality scores in early July were ’3 for the crested wheatgrass half-sib families compared with scores between 5 and 6 for the traditional turf species. Thus, crested wheatgrass, for the near future, will likely be a viable turf candidate only in situations in which turf aesthetics are secondary to a desire for low-input requiring species. As water resources in many areas, but particularly the western United States, become more limiting and are sought by competing interests, there is a need for turfgrass species that require less irrigation for low-maintenance situations (Feldhake et al., 1983). The need to identify new lower waterrequiring turfgrass species has resulted in the characterization of many perennial grass species for turfgrass potential (Diesburg et al., 1997), including crested wheatgrass [Agropyron cristatum (L.) Gaertn.]. Crested wheatgrass is a perennial Triticeae grass species, is well suited to harsh semiarid conditions, and is a key species for revegetation and forage production on rangelands of the Great Plains and intermountain regions of the United States (Asay and Jensen, 1996). Several studies have investigated the potential (Bushman et al., 2007; Diesburg et al., 1997; Hanks et al., 2006; Robins et al., 2006) and breeding (Asay et al., 1999; Hanks et al., 2005) of crested wheatgrass for turf. ‘RoadCrest’ was the first crested wheatgrass cultivar specifically developed for turf use (Asay et al., 1999), although other cultivars such as ‘Ephraim’ (Stevens et al., 1983) and ‘Fairway’ (Kirk, 1932) are commonly used in roadside stabilization and other lowmaintenance turf situations. Breeding efforts are ongoing for further crested wheatgrass turf improvements. A potential roadblock to higher public acceptance of crested wheatgrass turf is its poorer performance for turf characteristics such as color, quality, and increased tendency for summer dormancy when compared with traditional turf species (Bushman et al., 2007; Robins et al., 2006). Turf quality is a particularly important trait and consists of the combination of color intensity, leaf texture, and tiller density (Gibeault et al., 1989; Skogley and Sawyer, 1992) and can be affected by pest resistance. Hanks et al. (2005) identified the rapid spring growth and reduced summer turf quality as the most limiting characteristics of crested wheatgrass for turfgrass. However, the study also identified high levels of broad-sense heritability for turf quality and suggested that breeding efforts aimed at increasing crested wheatgrass turf quality would be successful (Hanks et al., 2005). Although high broad-sense heritabilities suggest the importance of genetic as compared with environmental factors (Holland et al., 2003; Nyquist, 1991), they are not indicative of the potential that can be made in a breeding program. The objective of this study was the estimation of genetic variation and narrow-sense heritability for turf quality in a population of half-sib families. These families represent breeding materials used in ongoing efforts at improving turf-quality traits in crested wheatgrass for low-maintenance turfgrass conditions. Materials and Methods Plant materials. Plant materials consisted of 27 half-sib crested wheatgrass families and six control cultivars representing various turfgrass species. The half-sib families came from the polycross nursery of 27 crested wheatgrass genotypes selected from ‘RoadCrest’ crested wheatgrass (Asay et al., 1999). Selection criteria were spreading ability, short growth stature, and fine leafiness. The six control cultivars were ‘Cody’ buffalograss [Buchloë dactyloides (Nutt.) Engelm.] (Severmutlu et al., 2005), ‘Fults’ weeping alkaligrass [Puccinellia distans (L.) Parl.] (identified by S.E. Metsker in 1979), ‘Gazelle’ tall fescue (Festuca arundinacea Schreb.) (Rutgers University and Pure Seed Testing, Inc.), ‘Manhattan 3’ perennial ryegrass (Lollium perenne L.) (Rose-Fricker et al., 2002), ‘Midnight’ Kentucky bluegrass (Poa pratensis L.) (Meyer et al., 1984), and ‘RoadCrest’ crested wheatgrass. Experimental design. The study location was the Utah Agricultural Experiment Station Evans Farm in Millville, UT (41 45#N, 111 8#W; 1350 m above sea level; Nibley silty clay loam [fine, mixed, mesic Aquic Argiustolls]). The experimental design was a randomized complete block design with four complete blocks. Transplanting of the control cultivars and half-sib families occurred in Apr. 1999. Individual plots consisted of 10 plants. Spacings were 1 m between rows and 0.5 m between plants within a row. Although sward plots would have been preferable to space-planted plots for the evaluation of turf quality, the polycross used to derive the halfsib families did not result in sufficient seed production for a sward plot study. Thus, seed limitations required the utilization of spaceplanted plots. Space-planted plots have been used previously to characterize turf quality (Hanks et al., 2005). Throughout the study, irrigation occurred weekly from April to October at 50% evapotranspiration (ET0) replacement. Approximate ET0 values for turfgrass at Millville, UT, were: April, 48 mm; May, 87 mm; June, 109 mm; July, 121 mm; August, 107 mm; September, 68 mm; and October, 29 mm Received for publication 30 Apr. 2007. Accepted for publication 1 July 2007. Joint contribution of the USDA-ARS and the Utah Agr. Exp. Stat. Paper No. 7906. Mention of a trademark, proprietary product, or vendor does not constitute a guarantee or warranty of the product by the USDA or Utah State University. To whom reprint requests should be addressed; e-mail [email protected]. 1526 HORTSCIENCE VOL. 42(7) DECEMBER 2007 (Hill and Kopp, 2002). Throughout the evaluation, plots were mowed at a height of 7.62 cm (3.0 in) with a rotary mower at an interval that removed 33% of growth at each mowing. The clippings were left on the ground and 49 kg ha of nitrogen (1 lb 1000 ft) was applied in early June and again in September. Phenotypic data collection and analysis. Data collection took place from April through October in both 2000 and 2001. However, data were not collected during May 2001. Turf quality was measured monthly and twice in July during these time periods. The turf quality assessment combined visual ratings of color intensity, leaf texture, and tiller density (Gibeault et al., 1989; Skogley and Sawyer, 1992). Ratings followed a 1 to 9 scale. A score of 9 was given to the plot with the highest turf quality in the study and indicated a plot with solid green color, fine leaves, and dense sod with solid groundcover. A score of 5 was given to plots with acceptable turf quality and indicated a plot with uniform, mostly green color, acceptable leaf fineness, and a uniform sod with little openground area. A score of 1 was given to plots with unacceptable turf quality and indicated a plot with brown color, coarse leaves, and an open canopy. Other values were given to plots with characteristics intermediate to these descriptions. Using the MIXED (Littell et al., 1996) and IML procedures of SAS (SAS Institute, 2006), variance components and heritabilities with their standard errors were computed with control genotypes removed from the analysis (Holland et al., 2003) based on a split-plot-in-time modification of the randomized complete block design with four complete blocks. The model used was appropriate for the analysis of genetic variation and heritability in a set of half-sib families (Nguyen and Sleper, 1983) based on entry means and was as follows: h = sHSF sHSF + s2 HSFY y + s2 HSFR r + se yr