Seven Cycles of Reciprocal Recurrent Selection in BSSS and BSCB1 Maize Populations1

Abstract

Reciprocal recurrent selection was proposed as an efficient method of improving the population cross of two broad‐genetic base populations. Seven cycles of reciprocal recurrent selection for grain yield have been completed in ‘Iowa Stiff Stalk Synthetic’ (BSSS) and ‘Iowa Corn Borer Synthetic no. 1’ (BSCB1) maize (Zea mays L.) populations. To determine the effectiveness of reciprocal recurrent selection for grain yield improvement, we evaluated the C0, C1, C3, C5, and C7 of BSSS(R) and BSCB1(R), and the C0 ✕ C0, C5 ✕ C5, and C7 ✕ C7 population crosses in six Iowa environments. Data were collected to determine the direct response to selection for grain yield and the correlated responses for eight other agronomic traits. Observed response for grain yield was compared with computer simulated reciprocal recurrent selection as it was conducted with BSSS and BSCB1. Ten cycles of reciprocal recurrent selection were simulated for starting conditions with three levels of dominance and two initial gene frequencies.Grain yield of the BSSS(R) ✕ BSCB1(R) crosses increased 1.75 ± 0.37 q/ha per cycle, but yield of the parental populations did not change significantly. Observed midparent heterosis increased from 14.9% for the C0 population cross to 41.7% for the C7 population cross. Mean yield of the population crosses increased from 58.5 q/ha for the C0 population cross to 70.7 q/ha for the C7 populations cross. Estimates of genetic components of variance determined from the yield selection trials for each cycle showed no evidence that genetic variability in BSSS(R) and BSCB1(R) was declining because of selection. The greatest correlation between observed and simulated responses was obtained for the starting condition of complete dominance and equal initial allele (p=q=0.5) frequencies in the simulated populations.