GENETIC VARIANCES OF AGRONOMIC AND MORPHOLOGICAL TRAITS OF DIPLOID CRESTED WHEATGRASS

Crested wheatgrass breeding efforts in the USA have focused on tetrapl oid germplasm, resulting in little genetic information available for d iploid germplasm. This study was conducted to estimate progeny-mean he ritability (h(2)) of eight traits of diploid crested wheatgrass (Agrop yron cristatum L.) grown under seeded competitive conditions in the no rthern Great Plains. Thirty half-sib families were randomly selected f rom a genetically broad-based population, and evaluated at one locatio n over 2 yr, in single-row 1.5-m plots, spaced 38 cm apart, and seeded at 90 seed m(-1). Heritabilities based on the progeny means of 30 fam ilies were determined for forage dry matter yield (h(2) = 52%), tiller height (h(2) = 63%), first-cut vigor (h(2) = 15%), regrowth vigor (h( 2) = 68%), proline content (h(2) = 49%), spikelets per spike (h(2) = 5 9%), anthocyanin pigmentation of stem nodes (h(2) = 36%), and flag lea f pubescence (h(2) = 70%). Heritability of forage yield, tiller height , regrowth vigor, number of spikelets per spike, and nag leaf pubescen ce were estimated with the greatest precision. First-cut vigor, prolin e content, and node color were estimated with less precision. Proline accumulation was net associated with either dry matter yield or regrow th vigor under water stress. Regrowth vigor was not associated with fi rst-cut yield (P greater than or equal to 0.24), indicating that impro vement in first-cut field would have little effect on forage regrowth potential. A moderately high heritability estimate for regrowth vigor indicated that selection for this trait could extend the grazing seaso n of crested wheatgrass. Genetic variance components for all traits ex cept first-cut vigor exceeded their standard errors. Significant famil y x year effects were detected only for node color. The results indica te that evaluation over multiple years would not be needed for most tr aits studied.