Effect of interaction of inoculum dose, cultivar, and geographic location on the development of foliar symptoms of bacterial ring rot of potato

Proportional hazards models were used to estimate hazard and survival functions for both the onset and maximum incidence of foliar symptoms of bacterial ring rot in three potato cultivars (Norchip, Norland, and Russet Burbank) grown at seven locations across the United States (Colorado, Maine, New York, North Dakota, Oregon, Washington, and Wisconsin) over the period 1988-1990. The models predicted a minimum of 50 and 82 days after planting for survival probabilities of P < 0.95 and P < 0.05 for symptom onset and maximum disease incidence, respectively. The time frame predicted for these events was affected by cultivar, location, and cultivar by location interaction. There was no proportional increase in the relative hazards for symptom onset and maximum disease incidence due to increased inoculum dose. An increased probability of survival, however, was associated with an increased incidence of right-censoring of the data in plants given the lower inoculum doses. The relationship between cultivar maturity and the relative hazard for maximum disease incidence was not consistent, an observation attributable to the fact that Russet Burbank displayed stunting (dwarf rosette) as an early season symptom of bacterial ring rot. Symptom onset and maximum disease incidence were influenced most by location and location by cultivar interaction, with certain locations displaying a trend toward an increased hazard for these events. Differences in the relative hazard for onset and maximum disease incidence among locations, in some cases, were associated with higher rates of right-censoring of the data. However, environmental conditions also appeared to influence the time when onset and maximum disease incidence occurred, because significant differences in the hazards for these events were observed in locations in which no right-censored observations were recorded.