Abstract
An experiment was conducted to determine critical factors in the recovery of embryos from cultured anthers of broccoli (Brassica oleracea var. ‘italica’) and to unambiguously distinguish whether embryos were of gametophytic origin. Among factors tested, genotype, genotype x anther developmental stage, and method of anther culture had a distinct impact on embryo recovery, whereas length of anther exposure to the culture medium did not. However, extreme heterogeneity of embryo emergence within and among replications precluded statistical contrasts. Among 762 plants derived from embryos of four independent cultivars, only one was determined to be of sporophytic origin by use of heterozygous codominant isozyme markers. Two of the cultivars tested were heterozygous at two or more loci. While segregation among loci was consistent with previously published linkage data, segregation of alleles was consistently non-random. In all of seven separate cases involving four cultivars, a significant over-representation of the fast-migrating class was observed. It appears, therefore, that populations of plants derived from microspores within cultured anthers of broccoli do not necessarily represent a random gametic array, and that care must be exercised in breeding and genetic applications.