Comparison of methods to identify donor inbreds with favourable alleles for the improvement of elite maize single cross hybrids

Abstract

The success of single cross hybrids in maize primarily rests on the economical seed production of its inbred parents. Inbred parental improvement thus holds the key in modern maize breeding programmes. Identification of inbred lines (lw) as sources of favourble alleles not already present in an elite hybrid (11 x 12) is one of the approaches towards line enhancement in single cross hybrids. In the present study, a minimally biased estimate of the number of favourable alleles present in an inbred but not present in an elite hybrid (1lG') was compared with minimum upper bound estimates (UBND), net improvement statistic (NI), predicted three-way performance (PTC), an unbiased estimator based on combining ability model (lla[B+G]), probability of net gain of favourable alleleS (PNGg), inbred performance per se and general combining ability (GCA). These estimators were applied to grain yield and related characters among 45 F1's obtained from crossing ten maize inbred lines. The relative efficiency of these estimates was analyzed. The estimates, 1lG' and UBND were found to be the best in ranking the donor inbreds in a select group of five single crosses. A scheme for possible utilization of these estimates for inbred line enhancement of the target crosses is proposed.