Optimizing mineral nutrition for flower production in Heliconia 'Golden Torch' using response surface methodology

Containerized plants of Heliconia psittacorum L.f. X H. spathocircinata Ari steguieta 'Golden Torch' were grown in a greenhouse for 8 months from early summer to winter under selected combinations of N, P, and K. Fertilizer ra tes ranged from zero to rates that exceeded those reported in the literatur e by 50% to 100%. Biomass variables (vegetative and inflorescence dry weigh t, and leaf area) were predicted to be maximized at high N and high N to P, and N to K ratios corresponding to N-P-K application rates of 1.2, 0.5, an d 0.6 kg.m(-3), respectively (approximate to 2:1:1). However, the number of shoots and flowers produced per rhizome were maximal at lower N to K ratio s (1:1). Flower yield could therefore be optimized with appropriate fertili zation, provided attention was paid to the N to K ratio so that the size of plants and their flowers was not compromised by efforts to increase shoot and flower number. The heavier the rhizome planted, the shorter the time fo r shoot emergence and flowering to occur, and the greater the number of flo wers harvest ed. However, rhizome weight had no effect on number of shoots to emerge. The probability of shoots flowering declined markedly with order of shoot emergence, although this could be increased with appropriate mine ral nutrition. The maximum number of leaves subtending the inflorescence (s even) was obtained at high N and P rates. Flower production was probably li mited by declining solar radiation in autumn, and by within-plant competiti on for roofing space.