Bulk populations in wheat breeding: comparison of single-seed descent and random bulk methods

Abstract

An accelerated-generation increase program for establishing near-homozygous lines was evaluated by both single-seed descent (SSD) and random bulk-population (BP) methods. After four generations, random SSD- and BP-derived lines from two hybrid populations were evaluated in the F4, F5, and F6 generations in field experiments at two locations. In separate populations, selection was applied for plant height (tall and short) and early maturity during the accelerated-generation cycle. Generation means showed increased plant height and yield from F3 to F6 in one hybrid but no change in the other hybrid with the BP method, whereas the means did not change with the SSD method except for plant height where recessive genes for dwarfness were segregating. Genetic variation among lines within each generation was greater for SSD than BP populations for heading time and grain yield of Hybrid I, but for Hybrid II the reverse was true for height and yield, with no difference for heading time. Response to selection during the accelerated growth cycle was significant for tallness and shortness, but was greater for shortness. Selection response for earliness was only slight.