The long-short day requirement for flowering in Stylosanthes guianensis

Five ‘tardio’ selections of Stylosanthes guianensis ssp. guiunensis var. pauciflora and S. guiunensis var. guianensis cv. Cook were grown in a glasshouse where natural daylength was shortencd, or extended by incandescent lamps. They did not flower or flowered weakly in 152 cycles of 10 h short days (SD), but flowered more rapidly and prolifically if exposed to 30, 60 or 90 16 h long days (LD) followed by 30 10 h cycles than under natural daylength conditions at 27' 30' S. Increasing the previous exposure to LD reduced the number of 10 h cycles to floral initiation. Similar behaviour was exhibited by cv. Bandeirante which did not flower in 131 cycles of 11 h, but which flowered when SD induction followed 45 to 65 cycles of 14 h. Spike density was positively related to number of LD, which favoured first flower appearance on the terminal apices of lateral rather than of main shoots. CIAT 1283 and cv. Cook grown in controlled environment cabinets at 30±/23± (day/night) did not flower in 140 cycles of 10 h SD, but flowered if LD were interposed before SD induction. Cook had a greater LD requirement of 50 cycles of 14 or 15.5 h for floral initiation, whilst CIAT 1283 had a lesser LD requirement and flowered after 20 cycles of 14 or 15.5 h or after 50 cycles of 12.5 h. These data indicate a qualitative long-short day flowering response. This has implications which favour the higher latitudes for seed production and the early summer sowing of seed crops when plants would receive maximum LD exposure before SD induction occurs