Submitted to: Journal of the American Society for Horticultural Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: August 21, 1997
Publication Date: N/A
Interpretive Summary: The traditional method for making inbred broccoli varieties is to self pollinate plants for several generations. A newer way to make inbreds is to use anther or pollen culture, which when successful can result in inbreds in a single generation. A problem with anther culture is that some of the plants that develop are abnormal and sterile. These abnormal plants scannot be readily distinguished from normal ones. If broccoli breeders us anther culture, they must be able to identify the fertile plants. ARS research has shown that a new technology, flow cytometry, can be used to identify normal plants from culture. With flow cytometry, cells from leaves of anther-derived plants are compared to cells from normal check plants to check for normality. If normal, plants are pollinated and seed is collected. Time is not wasted with abnormal, sterile plants. With these procedures, a breeder can make inbreds very efficiently and save substantial time. These studies also show that normal, anther culture-derived plants produce equal amounts of seed as inbreds produced in the traditional way. This research provides public and private broccoli breeders useful information to help them breed improved varieties. These varieties will enhance production and provide a stable supply of quality broccoli for consumers.
Technical Abstract: The use of anther culture to generate doubled-haploid (DH) lines is a common practice in breeding broccoli. Regenerants from anther culture can be haploids, diploids, triploids, tetraploids, octoploids or aneuploids. The objectives of this research were to determine the influence of F1 source on frequency of different ploidy levels among regenerants and compare seed set in broccoli inbreds developed in a traditional selfing program compared to seed set in DH broccoli derived from anther culture. In two cycles of culture, anther-derived regenerants were developed using four F1 hybrids as sources of anthers. In 1994, 'Everest', 'HiSierra', and 'Futura' yielded populations not significantly different from one another with 2 to 7% haploids, 53 to 56% diploids, 32 to 38% tetraploids, and 5 to 6% other types. 'Marathon'-derived regenerants were significantly different from all other populations with 5% haploid, 78% diploid, 15% tetraploid, and 2% other. These results were repeated in 1995, confirming that genotype of the anther source affects frequency of particular ploidy levels among regenerants. In self pollinations on 1994 regenerants, only diploids and rare tetraploids set seed. Seed production following manual self pollinations of 1995 DHs was not significantly different from that of traditional inbreds derived from the same F1 sources.