Floral morphology of 'Anand' rapid cycling Brassica oleracea .

Abstract

Abstract The 'Anand' type of cytoplasmic male sterility (CMS) was introduced to Brassica oleracea through protoplast fusion (Cardi and Earle, 1997), providing a source of male sterility that is of value in producing F1 hybrid varieties of this species without the cold sensitivity of 'Ogura' cytoplasm. In this species, as in others in which it has been introduced, 'Anand' mitochondria are associated with flower abnormalities. The best lines have been used by commercial breeders. This paper describes 'Anand' flower abnormalities in further detail. Rapid cycling brassica with 'Anand' cytoplasm may have promise as a model system for research on floral patterning. Materials and Methods. Three lines of rapid cycling B. oleracea carrying the 'Anand' cytoplasm and one wild type were grown. The flower morphology of 'Anand' lines was classified as good in F25-F-1A, fair in F25-F-6B, and poor in F23-F-1B (Cardi and Earle, 1997). Seedlings were started in an indoor growth room, but were maintained in a naturally lit greenhouse at 20° night, 25° day during flowering. Five flowers on 5 plants of each line were examined. Results. Morphological abnormalities affected all floral parts (Table 1). There was substantial variation among flowers on individual plants and among lines but little within lines. Conclusions. "Good" flowers have large petals that push open the sepals before the carpel elongates. Female fertility could be reduced due to occlusion of the stigma or physical damage caused by the stigma pressing against the end of a sepal and kinking the style. Female fertility in some flowers was unlikely because the carpel suture failed to fuse. In some cases the edge of the split carpel was sepal-like, consistent with the absence of SEPALLATA expression. The stigma on a sepal represents the converse. Stamens were most strongly affected, with outer stamens most reduced. This whorl is the one most frequently suppressed in flowering plants (Masters, 1869). The number of stamens was reduced to varying degrees. First to be lost was one or both of the pair outside the petals. Then the outer pair inside the petals was affected. One or both of these stamens could either be lost or doubled to give 3 to 5 stamens. Again a positional cue for stamens appears to be weakened. Anther development was often limited, the expected and desired consequence of the Anand genotype. Some undehisced " good " anthers contained pollen. There is variation even among the lines initially described as " good ". Further selection among these may help identify those that are most promising for further use in hybrid production.