Submitted to: Crop Science
Publication Type: Peer reviewed journal
Publication Acceptance Date: 3/5/1997
Publication Date: N/A
Citation: Interpretive Summary: Orchardgrass is a highly nutritious grass grown for hay and in pastures, especially in mixtures with alfalfa used as forage for dairy cattle in the northeast and upper midwest regions of the USA. Late maturing cultivars of orchardgrass are desirable so they better match recommended harvest management of alfalfa. Selection of elite germplasm lines with high seed production is vital to the development of acceptable forage cultivars. Unless seed can be efficiently produced, cultivars possessing even the most important forage traits will not be successful. The Pacific Northwest climate is ideal for orchardgrass seed production and over 90% of the USA orchardgrass seed is produced in the region. Genetic shift in grass cultivars with seed produced outside their region of forage production has been a concern in the grass seed industry for many years. To solve the problem, strict seed increase rules for seed certification with limited generations from breeder seed in designated seed producing areas were developed. Studies have suggested that orchardgrass germplasm must be evaluated for seed production potential in the region where seed will be grown. Convergent-divergent (C/D) selection is a method of selecting for broad adaptation in a germplasm pool. In part I of this series of three papers, we showed that both C/D and local selection at four forage-production locations can be used to increase panicle seed weight and total seed yield, and to shift heading date toward later maturity in orchardgrass grown for seed in the regions where selected for seed yield. It is possible to achieve high seed yields concomitant with later maturity.
Technical Abstract: Simultaneous breeding for improved forage traits & increased seed yield in orchardgrass Dactylis glomerata L. has been a consistent problem. This study's objective was to conduct & evaluate 2 cycles of convergent- divergent (C/D) & local selection for panicle seed weight (PSW) & for forage traits. Selection was conducted at 4 locations (Ames, IA; Mount Vernon, MO; Rock Springs, PA; & Arlington, WI) & initiated in 4 base populations (I79DT, MO2, PLS4, & W011). Phenotypic selection on spaced plants was based on a 0.25 selection intensity for forage traits & a 0.25 selection intensity for PSW. All populations were evaluated as spaced plants at the IA & PA locations in 1990 & 1991. On average, both selection methods were equally effective at increasing PSW, with responses averaging 1.0 to 18.2% per cycle. Realized heritabilities for PSW were highly variable among the 4 populations (mean of 0.03 to 0.23) & were lower for cycle 2 than for cycle 1 (mean of 0.16 vs. 0.02). In C/D selction, the 4 locations did not contribute equally to genetic gains for PSW, with means of 25, 39, 94, & 33 mg/panicle for IA, MO, PA, & WI, respectively. Three of the 4 populations had significant genetic changes toward later heading, by 2 to 4 d despite no direct selection pressure for heading date. Increased apparent rust resistance appeared to be due to a general resistance mechanism when observed for both crown rust (Puccinia coronata Cda.) at PA & stem rust (Puccinia graminis Pers.) at IA. There were very few significant & biologically meaningful genotype x location interactions for any trait; therefore, neither selection method resulted in plants with differential adaptation when evaluated as spaced plants.