The effect of time of harvest on seed production of three red clover cultivars

Small plots of red clover cv. Sabtoron, S123 (diploids) and Hungaropoly (tetraploid) were harvested for seed production at two-week intervals from 19 August to 17 October inclusive in 1981, inflorescence appearance rate, bee density and components of yield having been monitored throughout the summer. Inflorescence appearance rate reached a maximum at the end of July/early August for Hungaropoly and Sabtoron and during mid-August for S123. Bee density followed a similar pattern. Florets and seeds per inflorescence and 1000-seed weight decreased as flowering progressed except during the first three weeks in July. Potential seed yield was calculated from the components of yield for the harvests taken on 3 and 18 September. Losses in seed yield (difference between actual and potential) up to and during harvesting, threshing and cleaning were lower in the tetraploid cultivar (27–39%) than the diploid cultivars (35–l55%). Each cultivar had an optimum harvest time before which yield was affected by immature seeds and beyond which it was adversely affected by shedding of inflorescences and seeds and sprouting of seeds on the inflorescences. The optimum time to harvest Sabtoron was early September, Hungaropoly early to mid-September and the late flowering cultivar S123 mid-September when less than 4% of the inflorescences were still unripe. Seed yield and inflorescences per unit area were lower in the tetraploid cv, Hungaropoly (maximum 542 kg ha^?1) than diploid cv. Sabtoron and S123 (864 and 897 kg ha^?1 respectively) although the tetraptoid had heavier seeds. It is concluded that the optimum time to harvest red clover for seed production is about three or four weeks after the end of the period of rapid inflorescence production and that this coincides with the time when only a small proportion of unripe inflorescences remain.