Download - Predicting Tomato Spotted Wilt Virus on tobacco in North Carolina A.L. Mila Dept. of Plant Pathology
Predicting Tomato Spotted Wilt Predicting Tomato Spotted Wilt Virus on tobacco in North Virus on tobacco in North
CarolinaCarolina
A. L. Mila
Dept. of Plant Pathology
TSWV symptoms on tobacco
Thrips: the culprit
Frankliniella fusca
TSWV epidemiology on tobacco
• Wide range of hosts
• (>500 spp.,>50 families)
TSWV incidence in NC
2.12.82.9
3.53.9
1.240.48
6.30.70.71
10.08
0 2 4 6 8
199819992000200120022003200420052006200720082009
Question:
What is next season’s expected TSWV risk
TSWV in NC (1993-2007)
Severe TSWV epidemics (8 counties)
TSWV epidemic in 2002 and then occasionally (12 counties)
TSWV in “trace” (38 counties)
What can a grower do?
Imidacloprin (such as, Admire Pro: 0.8 oz/1000 pl)AND Actigard (0.5 oz – 1 oz/50,000 pl drench OR 10 – 25
ppm float water)
Chemical control in the greenhouse
Untreated control
Actigard 25 ppm [float water]
Actigard: 4 weeks after transplant
Wake co. 2006
Actigard: 6-7 weeks after transplant
Untreated controlActigard 25 ppm [pretra]
Wake co. 2006
Question:
Which fields will have > 25% TSWV?
What can a grower do?
• Imidacloprin (such as, Admire Pro: 0.8 oz/1000 pl)
AND (if TSW incidence > 25%) • Actigard (0.5 oz – 1 oz/50,000 pl drench OR 10 –
25 ppm float water)
Two questions to answer:
• Question - pre-season risk
What is next season’s expected TSWV risk?
• Question - in-season riskWhich fields will have > 25% TSWV?
TSWV in NC (1993-2007)
Severe TSWV epidemics (8 counties)
TSWV epidemic in 2002 and then occasionally (12 counties)
TSWV in “trace” (38 counties)
A total of 489 cases 58 counties 14 years (1993 – 2007)
Objectives
• Determine if any weather factors are predictors of TSWV risk at the county level
Time: September before transplant to August (harvest)
Weather factors: Total precipitation, days with rain, degree days, and average air temperature
Pre-season risk - Results
Precipitation and temperature of Dec – Feb most significant explanatory weather factors
Parameter Estimate
Total Precipitation (December) -0.007
Total Precipitation (January) -0.014
Total Precipitation (February) -0.008
Sum of monthly average temperature (December, January & February)
0.14
* Magnitude of effect different for counties
Pre-season risk – “county effect”
►Correction between counties?
(Structured heterogeneity)
► Other missing factors?
(Unstructured heterogeneity)
Investigate the “county” effect: Bayesian disease mapping
yi ~ Bin (ni, pi) (likelihood)
pi = f (T, Pr) + Ui + Vi
Structured (autocorrelation)
“Auto-regression function”
Unstructured
N (0, σ2)
• Structured heterogeneity? NO
• Unstructured heterogeneity? YES
Pre-season risk – “county effect”
Sampson co.
TSWV incidence
Pro
bab
ilit
y o
f T
SW
V i
nci
den
ce
Validation (2008 – present)
Predicted Pre-season riskNorth Carolina - 2008
County Station Predicted
Craven Havelock 14 – 22Duplin Mount Olive 4 – 9 Sampson Clinton 13 – 26Lenoir Kinston 4 – 11 Pitt Ayden 1 – 4 Johnston Clayton 0.6 – 3.5Wilson Sims 0 – 1
Reported
14
8
2
10
2.3
4
1
Predicted Pre-season riskNorth Carolina - 2010
County Station Predicted
Craven Havelock 1 – 14 Duplin Mount Olive 1 – 5 Sampson Clinton 0.5 – 10 Lenoir Kinston 1 – 12 Pitt Ayden 0 – 3 Johnston Clayton 0 – 0.5 Wilson Sims 0 – 1
Conclusions
• Winter weather good predictor of TSWV risk for upcoming season
* Importance of spring weather was not discussed…
Pre-season risk on the web
Expand effort - 2009
• Pre-season risk
(Georgia, North Carolina, South Carolina and Virginia)
• In-season risk
(Georgia, North Carolina, South Carolina)
• K. Cherry
• J. Radcliff & M. Browne
• NC Tobacco Agents
• NC Tobacco Research Commission
• Tobacco Education & Research Council
Acknowledgements