Developing sweet onions by recurrent selection in a short-day onion breeding program

Sweet or low-pungency onions have become popular in fresh market onion production and are commonly produced using low sulfur soil and/or innate low-pungency cultivars in specific regions. However, there have been very limited breeding efforts to create true low-pungency cultivars. This study reports the process and progress of a recurrent selective breeding program to produce low-pungency lines in short-day type onions. Two parent populations, 'T81079' (n = 754) and 'T81082' (n = 347), with mean pungencies of 4.1 and 3.5 mM pyruvic acid, respectively, were tested for pungency and bulbs were divided into five groups according to their pungency levels. Twenty bulbs were selected from each group and grown for seeds by mass-pollination. The pungency in bulbs produced from the seeds of each group showed genetic shifts towards the selection differential. The realized heritability (h(2)) values in the two populations varied between 0.48 and 2.21 or -2.09 and 6.16, respectively, and showed no consistency or reliability. However, a regression analysis between the selection differentials and the response to selection showed coefficients of 0.65 and 0.54 with high significance (r(2) = 0.94**. After three recurrent selections of 20 bulbs from the lowest pungency progenies, we obtained two populations with mean pungencies of 2.6 n = 1209 and 2.5 n = 1781 mM pyruvic acid, showing a reduction in pungency by 1.0-1.5 mM, as compared to the initial parent populations. The 4th selection of 20 bulbs from one of these populations yielded 11 families of onions with a mean pungency of 1.5 mM pyruvic acid. Our study demonstrated that the recurrent selection of low pungency actually shifted the pungency distribution toward lower pungency levels, and therefore developing low-pungency (1.5 mM pyruvic acid) lines will be feasible in about 10 years in short-day onions.