Seed production studies in Ruzi grass (Brachiaria ruziziensis Germain and Everard) : a thesis presented in partial fulfilment of the requirement for the degree of Master of Agricultural Science in Seed Technology, at Massey University, Palmerston North, New Zealand

The effect of photoperiod on seed production in Ruzi grass (Brachiaria ruziziensis) was investigated in order to gain a better understanding of reproductive behaviour. The experiment was conducted in a controlled temperature glasshouse which provided day and night temperatures of 25°C±5. Plants were raised from seed obtained from Thailand. During the early stage of growth (10 days after germination) Ruzi grass plants were exposed to controlled photoperiods of 14, 13, 12 or 11 hours for a period of 30 days followed by exposure to the decreasing natural daylength occurring in the New Zealand autumn. The results of this experiment showed there was no "trigger" daylength requirement for reproductive initiation, since Ruzi grass was able to produce flowers in all daylength treatments (14, 13, 12 and 11 h). However, the data did show that the shorter the daylength, the greater the seed yield (i.e. potential seed yield was greatest at 11 h). Accordingly, the results confirm the conclusion of Dirven et al., (1979), that Ruzi grass is a quantitative short-day plant. The results of this study appear to contradict the reproductive behaviour of Ruzi grass in Thailand, where a critical daylength of approximately 12½ h to trigger reproductive development seems necessary. These results suggest that it is theoretically possible for Ruzi grass to produce seed over the entire range of daylengths which occur in Thailand, as recorded in this experiment. The fact that this does not occur is possibly due to a juvenility problem, as the plants are only 4-6 weeks old after the onset of the rainy season in late April/early May when the daylength is approximately 13 h, and due to drought conditions which occur in December under the shortest daylength of approximately 11½ h. However, this needs to be confirmed, possibly by conducting trials in Thailand under an 11½ h daylength with irrigation to overcome lack of water. Daylength strongly affected inflorescence numbers and inflorescence components. As daylength declined the number of racemes/inflorescence arising from basal tillers tended to decrease. This was accompanied by a corresponding increase in aerial tiller numbers. Floret number/raceme was a more important factor influencing seed yield than raceme number/inflorescence. Basal reproductive and vegetative tiller numbers were not significantly affected by daylength, although aerial reproductive tillers did increase as daylength declined. Total tiller numbers were low, even although they continued to show a steady increase through to harvest, when approximately 30% were reproductive and 70% vegetative. Ruzi grass produced more inflorescences from aerial tillers than from basal tillers. The morphological changes occurring during the changeover from vegetative to reproductive development were divided into five stages. The time required from early raceme initiation in the "double ridge' stage to inflorescence exsertion was 22 days. It appears that Ruzi grass does not have a characteristically prolonged anthesis within an individual inflorescence, as all anthers were exserted within 1-2 weeks. Despite this, approximately 80% of florets within an inflorescence completed anthesis in 7 days. Within individual inflorescences, anthesis began in the middle region of the uppermost raceme and subsequently extended to the upper and lower raceme(s). Ruzi grass exhibited a prolonged head emergence period of about 3 months which was highly variable both within and between individual plants. Seed development studies suggested that although some variation in the extent and timing of seed shedding may occur between plants, harvesting should not be carried out before 20-25 days after anthesis (maximum viability), and should not be delayed longer than 30 days after anthesis (maximum dry weight)