Role of wind-driven rain, aerosols, and contaminated budwood in incidence and spatial pattern of fire blight in an apple nursery.
The role of wind-driven rain, rain-generated aerosols and Erwinia amylovora contaminated budwood in the epidemiology of fire blight was investigated in a Michigan apple nursery in 1992 and 1993, and in a simulated nursery planting at Michigan State University (MSU) in 1993. A degree-day model, MARYBLYT, was used to assess whether fire blight outbreaks were associated with storms containing wind-driven rain. Most outbreaks, especially those occurring early in the season, were associated with previous storms. Spatial lag autocorrelation analysis indicated that when fire blight was initially detected in the nursery each season, significant autocorrelations among spatial lags generally were within-row or formed tight clusters. Following storms containing wind-driven rain, autocorrelation matrices showed significant values across rows, often at high-order, noncontiguous spatial lags. Ordinary runs analysis indicated strong within-row aggregation of fire blight in the MSU nursery throughout the season, whereas significant across-row aggregations were apparent only later in the season following storms containing wind-driven rain. The possibility that E. amylovora becomes suspended in aerosols during rain was investigated by collecting air samples using a 6-stage microbial impaction air sampler. All air samples collected during rain contained E. amylovora, whereas samples collected during dry periods contained very few or no c.f.u. of the pathogen. Using traditional plating techniques, E. amylovora was not detected in 115 10-bud samples collected in 1992 from a budwood orchard. In 1993, E. amylovora was detected in 6 and 16% of washings from budsticks and leaves by plating and a polymerase chain reaction method, respectively. It is concluded that wind-driven rain was the most important factor involved in the spread of E. amylovora in the nursery.