Summary of the 2009 Blue Mold Epidemic in North America
Thomas Keever, Lead Forecaster
North American Plant Disease Forecast Center
NCSU, Department of Plant Pathology
We have reached the conclusion of the 14th year of forecasting for Tobacco Blue Mold epidemics in North America. Forecasts were produced three times per week in 2009 from mid-March through early September. The forecasting website provided visitors with up-to-date information about the status of the epidemic and forecasts of potential epidemic spread to other regions from the known sources. We hope the forecasts assisted the growing community via improved crop management and assessment of the risk of developing tobacco blue mold. Two hundred twenty-three forecasts were produced on 70 days covering the six-month period.
As has become common in recent years, the 2009 epidemic was comparatively weak when viewed from a historical perspective. Only 33 counties reported blue mold in six U.S. states. Disease incidence was focused in four main areas: east-central and northeast KY / southern Ohio, southeastern PA, west-central MA, and isolated outbreaks in a loose grouping from northeast TN eastward along southern VA and northern NC. Most major growing regions reported little or no disease. The most active disease presence was found in KY / OH and southeast PA. Impact from the 2009 epidemic was probably very light; however, some fields or areas in the aforementioned KY / OH and southeast PA regions likely experienced more significant losses.
For the fifth time in the past six years, there were no reports from the northern FL / southern GA region. Though possible, this was somewhat surprising, as there were more than a half-dozen transport episodes between mid-February and mid-April which might have deposited live spores in that area. These transport events originated in Cuba, where disease was active this year according to communications with other scientists. Most of these events involved conditions that were unfavorable for survivable transport but that coincided with weather suitable for deposition and infection on the U.S. mainland. Given the unfavorable transport conditions, the majority of these transport events would have produced Low Risk for U.S. tobacco plants.
The first report in the U.S. did not arrive until June 5. This was much later than average. Further, this report was from southeast PA! While blue mold is reported from this area in most years, it has never before been the location of the first report from the United States.
Origins of the outbreak in southeast PA (Chester County) remain unknown. At the time of discovery, there were no other reports of blue mold in the United States. The closest known sources in Cuba were over 1000 miles away, and no transport events from Cuba drew near southeast PA. Later outbreaks from VA, NC, and TN could not possibly have been in existence long enough to produce this outbreak, as well.
The first report from NC arrived on June 28 from Forsyth County, NC. A field of burley was approximately 60% infected. Symptoms were estimated to have appeared around June 12. Extensive investigation was performed to deduce the origin of the outbreak. There were no known sources to the south except for Cuba; however, disease activity there had ended. Focus shifted to the possibility of introduction from southeast PA. This introduction would have had to occur during or before the first week of June. The only plausible scenario involved transport episodes on June 4 or June 5 in which airborne spores from PA were deposited on June 6/7, via dry deposition, on fields / plants that were soaked from recent rains in west-central NC on June 3-5. While the evidence here is mixed and quite weak, no other possibilities exist that fit the field observations or an appropriate biological timeline. The events from PA would have occurred just prior to the report from that region.
The first report from TN arrived on July 2 from the northeast mountains in Greene County. First symptoms on these burley plants were estimated to have appeared between June 22 and June 25. Confidence is high regarding the origins of this outbreak. It was due to a transport event from Forsyth County, NC on June 16. This was after the appearance of symptoms in NC (June 12), but prior to the discovery and report from NC (June 28). Conditions were likely mixed to favorable for this event. Both the rainfall data (showers in the mountains) and the timing (event on June 16/17, first appearance of symptoms in TN approx. 1 week later) coincide well with this transport event. Except for the June 15 event from Forsyth County, which was plausible but not as good a candidate as the event the next day, there were no other episodes within a week to 10 days of June 16 that could have reasonably led to this outbreak. The event is illustrated in Figure 1 below.
Figure 1: Multi-level trajectories from Forsyth County, NC for June 16, 2009
The first report from VA (Dinwiddie County, near Petersburg) arrived on July 13. First symptoms were estimated to have appeared between June 28 and July 4. Here again, confidence is strong as to the origins of this outbreak. The transport event from Forsyth County of June 25 is the most likely candidate. Conditions, rainfall data, and timing support this thinking. While a similar event on June 19 is also plausible, the June 25 event is a better fit for the timing and field observations. This event is shown in Figure 2 below.
Figure 2: Multi-level trajectories from Forsyth County, NC for June 25, 2009
The first report from KY arrived on July 17. Fields near the Clark / Montgomery County line in east-central KY were found to have actively sporulating blue mold the day before. Observations indicated that symptoms appeared in a window to time centered around 10 days prior to the report.
The origins of the KY outbreak most likely lie with the transport event of July 4 from northeast TN. This episode correlates reasonably with the field observations, rainfall estimates, and disease cycle timing. The TN source had just been discovered / reported at the time of this event; this was the first forecast in 2009 from that source. Other transport events from the TN mountains or from west-central NC are far less likely to have given rise to the KY infections. The forecast from Greene County, TN for July 4 is shown below (Figure 3). Plants in eastern KY were placed at Weakly Moderate Risk for that event.
Figure 3: Trajectory of airborne spores from northeast KY for July 4, 2009
The first report from OH arrived on August 4. A field on the border of Brown and Adams Counties was found to have active lesions estimated to have first appeared around July 28. This outbreak originated from infected locations in east-central / northeast KY. Any of a number of events may have been responsible.
The southern New England region discovered and reported its first outbreak on July 30/31. Shade tobacco in Hampden County, MA was infected. Observations indicated that 2 generations of lesions were present, placing the appearance of first symptoms between July 16 and July 21. The forecast of July 11 from southeast PA illustrates the best candidate for this outbreak, as it is a good fit for the transport and deposition, the disease cycle timing, and field data. Other episodes are far less suitable in one or more of these respects. Tobacco plants in southern New England were at Moderate Risk for this transport event, shown in Figure 4.
Figure 4: Trajectory of airborne spores from southeast PA for July 11, 2009
After the burst of initial reports in late June through mid-July, the epidemic continued to spread during periods of favorable weather for the remainder of the summer, mostly near the known sources. Conditions were more favorable in the Ohio Valley; this region had the greatest geographic spread of the epidemic by season’s end with 21 counties affected. NC ended the season with 5 counties after August reports from Caswell, Nash, Davidson, and Surry County. Nine counties total reported blue mold from the northeast TN / southern VA / northern NC corridor. The last reports came from KY on August 18.
Overall, the 2009 epidemic was very light and likely had few significant impacts. Many major growing regions reported little to no disease activity. The epidemic started in very unusual fashion with a late initial report in the U.S. (early June) from southeast PA (a first-time occurrence). The epidemic then spread to the southwest to NC. From there, transport events in June and early July moved the epidemic to points west (TN, KY) and east (southern VA). Further epidemic spread to central and eastern NC, OH, and MA occurred during various episodes during the remainder of the season while local to short-range transport expanded the epidemic in KY. Tracking of the epidemic proved problematic in the early stages, as outbreaks proved mysterious (southeast PA, west-central NC) or arose from transport events from sources that were undiscovered at that time (northeast TN, southeast VA).
The complete set of 2009 forecasts can be reviewed on the Blue Mold Forecasting website: http://www.ces.ncsu.edu/depts/pp/bluemold.
The NAPDFC thanks all the reporting coordinators in the U.S., Canada, and elsewhere for their participation during 2009. If you wish to become part of the forecast system, or need further information, please contact Dr. Mina Mila or the NAPDFC directly.
Dr. Mina Mila
Asst. Professor, Tobacco Specialist
mina_mila@ncsu.edu
Thomas Keever
Lead Forecaster / Meteorologist
thomas_keever@ncsu.edu