Multiple trait selection in a recurrent selection population in oat (Avena sativa L.)

Single trait selection is often utilized to maximize genetic gain in recurrent selection systems. However, traits negatively correlated to the primary trait can deteriorate with single trait selection. Multiple trait selection can be used to prevent or correct correlated trait deficiencies. Our objective was to compare the effectiveness of several multiple trait selection schemes with selection for grain yield alone. Five multiple-trait selection schemes were evaluated for their effect on total genotypic worth and individual trait responses in an 'opened' oat (Arena saliva L.) recurrent selection population. Grain yield, days-to-heading, plant height, and reaction to harley yellow dwarf virus (BYD) were studied in several environments. The phenotypic correlations with yield were 0.44 and 0.60 for heading date and plant height, respectively. There was no association between yield and BYD. A restricted index, a Smith index, a modified base index, a multiplicative index, and independent culling were evaluated and compared with selection for yield alone. The restricted and Smith selection indices, when compared with selection for yield alone, selected genotypes with improved heading date, plant height, and BYD with only negligible decreases in yield. Greater correction for heading date and plant height was achieved with the modified base index, the multiplicative index and independent culling, but yield decreased accordingly. Realized gains in total worth for the restricted index were 15% above, and the Smith index 14% above, the gain realized by selection for yield alone. Restricted and Smith index selection were effective at preventing secondary traits from deteriorating through selection for grain yield.