Abstract:
Low seed set and hollow seededness are amongst the main problems of the sunflower oilseed production industry in South Africa. This study of eighteen sunflower cultivars on seventeen different localities was conducted to address these and other problems. It was undertaken in collaboration with the national sunflower trials. The study has revealed that a lack of pollinators had the least significant influence on seed set. Neither could the percentage hollow seeds at each locality be attributed to a low number of insect pollinators. Commercial sunflowers in cages with honeybees gave a seed yield of I,859 kg ha-I (72% seed set) compared to I,22I kg ha-I (45% seed set) in cages with no insect pollinators. The percentage unfilled seeds differed significantly between localities and cultivars. However, the incidence of both hollow seeds and unfilled seeds differed highly significantly with the increase in plant density from I 5,000 plants per ha to 90,000 plants per ha. Yields also decreased as plant density increased. Although the boron levels differed between soil, leaf and floret samples for the variables of locality, cultivar and plant density, no relationship was found between the boron levels and differences observed in the percentage hollow seeds and unfilled seeds between the eighteen sunflower cultivars. Slow pollen movement was identified as the main cause responsible for hollow seededness. A lack of nutrients resulted in the high levels of unfilled seeds found. Recommendations are given regarding these problems. Viability and vitality of the sunflower pollen was good. It was, however, not possible to germinate the trinucleate sunflower pollen in vitro using conventional palynological methods and techniques. It is hypothesized that an unidentified "pollen growth factor", occurring on the stigmas, regulates germination of sunflower pollen.
