Improvement of tannia species by reproductive means

A fair amount of genetic diversity exists among, Xanthosoma cultivars. There is a possibility of increasing this diversity through floral induction in cultivars which do not flower naturally, followed by hybridisation among the cultivars. Several cultivars of tannia were induced to flower through hormone treatment and attempts were made to produce viable seed through selfing and hybridisation. The effect of time, method and number of applications of growth regulator gibberellic acid (GA3) on floral induction was assessed in pot and field experiments. GA3 was applied at rates of 250 and 100 ppm in three applications commencing at fortnightly intervals. All treatments were replicated six times. GA3 stimulated 100 percent inflorescence in both cultivars within 130 days on the initial application and there was little difference between the GA3 rates. Pollination experiments resulted in a high incidence of fruit set (approx.-40 percent) fruits developed to maturity and large numbers of viable seeds were obtained. GA3 proved extremely effective in promoting flowering in local cultivars of tannia but had a somewhat variable effect on introduced types - complete inability in Barbados nut eddoe and cv. Cubana to about 20 percent promotion in Florida Giant. The fact that GA3 failed to promote flowering in some cultivars of tannia although inducing the apparently related axillary bud release indicates that some factor other than gibberellin may limit reproductive development subsequent to axillary bud release. In Colocasia esculenta var esculenta some evidence indicates a seasonal requirement for flowering. Concludes that much is still to be understood on the control of flowering and seed production on tannia but the development in floral induction and viable seed production in breeding programmes can lead to the production of improved tannia species.