Predicted efficiency of spaced-plant selection to indirectly improve tall fescue sward yield and quality

The validity of spaced-plant evaluation to determine sward performance of forage grasses has been questioned. This experiment studied the efficiency of spaced-plant evaluation to indirectly improve sward yield and nutritional quality in tall fescue (Festuca arundinacea Schreb.). Narrow-sense heritabilities, genetic and rank correlations, and indirect selection efficiencies were estimated for a tall fescue population grown in spaced plant and seeded sward environments. Heritability for yield was similar between spaced plants and swards (0.43 and 0.44, respectively), but genetic correlation between the two was low (0.37 +/- 0.38). Inconsistency (r = 0.30, P = 0.17) in family ranking further suggested that spaced plants were not predictive of sward yield. Heritability of crude protein from swards was low (0.27 +/- 0.25) compared with 0.77 +/- 0.08 from spaced plants, but there was no genetic relationship between the two (r = -0.13 +/- 0.30). Moderate to high heritabilities and genetic correlations were observed for most fiber traits, but indirect selection efficiencies and rank correlations of <1.0 suggested that evaluation in a sward environment would be best to select for improved nutritional quality. Spaced-plant evaluation appears to be less effective, or ineffective, at improving sward yield and nutritional quality in tall fescue. New techniques are needed that maximize genetic expression but simulate actual sward production of forage grasses.