SUMMARY
Phytophthora colocasiae was successfully isolated by baiting with detergent‐treated taro leaf discs 8 cm diameter placed on water slurries of soil, on suspensions of macerated infected leaf lesions or on the washings from petioles of harvested plants. Taro root tips, detached or left on corms, were not susceptible to zoospores of P. colocasiae nor were detached root tips of Lupinus angustifolius. Cubes of taro corm used as baits, and agar selective for Phycomycetes which was inoculated directly with soil, both became too heavily overrun by Phythium splendens to allow detection of P. colocasiae. Investigations indicated that inoculum on lesions of detached leaves and in soil remains viable for only a few days. Petiole bases which comprise the bulk of the ‘tops’ used for vegetative propagation, lost detectable natural inoculum rapidly (2 days) if stored dry, but less rapidly (14 days) if planted immediately in the field. Artifically augmenting surface inoculum with naturally produced sporangia considerably extended the periods of detectability, probably by increasing the chances that a few propagules would survive, especially during dry storage. Incubation of inoculated tops in high humidity led to active infection and sporulation on petioles, especially on cut ends, a situation that might be paralleled under suitable moisture conditions in the field. Of several aroid species tested by artificial inoculation only Alocasia macrorrhiza was susceptible. Natural infection of this plant has not been seen, making it an unlikely alternate host of P. colocasiae under field conditions. Thus perennation between taro crops is effected by shortlived surface propagules and possibly also by mycelium within lesions on petioles. Reduction of the former and prevention of the latter might be achieved by dry storage of tops for 2 to 3 wk.