Visual Versus Marker‐Based Selection of Hybrid Progeny in Fertile × Fertile Beet and Carrot Crosses

Abstract

ABSTRACTCytoplasmic‐genic male sterility systems (CMS) help control pollination in carrot and table beet breeding programs and hybrid seed production. A challenge common to these breeding programs is identification of F1 hybrid progeny in crosses where both parents carry normal (N) cytoplasm. Such fertile × fertile crosses often produce both F1 and selfed progeny, making identification of F1 hybrids difficult. Such identification has traditionally been accomplished in carrot and table beet by visually scoring roots using heterosis for size. However, if heterosis is lacking, misidentification can result in systematic errors in the breeding process. We sought to determine how effective visual selection of hybrid progeny is in the context of table beet and carrot breeding programs. We used amplified fragment length polymorphisms (AFLP) markers to genotype parents and progeny in a variety of fertile × fertile crosses. Progeny were also grown in the field and evaluated for hybridity based on visual selection using the phenotype of heterosis for size. Two of five fertile × fertile table beet crosses and three of five fertile × fertile carrot crosses showed highly significant differences between visual selection for hybridity and AFLP marker analysis (p < 0.01). Misidentification of roots as F1 hybrids when they were actually self‐pollinated occurred in some crosses as much as 75% of the time. Both type I, a hybrid root scored as an inbred, and type II, an inbred root scored as a hybrid, errors were committed. Across all matings, type I errors occurred on average 40.9% of the time and type II errors occurred on average 8.5% of the time. Type II errors have important consequences because both parents are not represented and recombined in the progeny resulting in further inbreeding. The results of this study indicate that using heterosis for size as a hybrid selection method can be inaccurate for identification of hybrids in certain crosses and may introduce error into breeding programs.