|NYANKANGA, R - University Of Nairobi|
|OJIAMBO, P - North Carolina State University|
|WIEN, H - Cornell University - New York|
|KIRK, W - Michigan State University|
Submitted to: Crop Protection
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/22/2010
Publication Date: 3/8/2011
Citation: Nyankanga, R.O., Olanya, O.M., Ojiambo, P.S., Wien, H.C., Honeycutt, C.W., Kirk, W.W. 2011. Validation of a tuber blight (Phytophthora infestans) prediction model. Crop Protection Journal. 30:547-553.
Interpretive Summary: Late blight is a devastating disease of potato worldwide. Models that accurately predict late blight on potato tubers could be very useful for controlling this disease. We evaluated a tuber blight prediction model developed in New York using weather data and disease observations in Michigan. The model correctly predicted late blight on tubers in 7 out of nine years. This model has the potential to improve our control of late blight disease on potato.
Technical Abstract: Potato tuber blight caused by Phytophthora infestans accounts for significant losses in storage. There is limited published quantitative data on predicting tuber blight. We validated a tuber blight prediction model developed in New York with cultivars Allegany, NY 101, and Katahdin using independent foliar and tuber blight incidence data from Michigan on the cultivar Snowden. In both New York and Michigan, disease was initiated by artificial inoculation with pathogen isolates (US 8 genotype, A2). Late blight severity ranged from 0-94% at New York, and 0-93% at Michigan field sites. Similarly, mean tuber blight incidence ranged widely, from 0.7-40% in New York and 0-15% in Michigan. The tuber blight prediction model was validated by fitting a regression equation to weather and foliar blight data from the field site in Michigan, and comparing predicted vs. observed tuber blight incidence at the same location. The model correctly predicted tuber blight incidence in 7 out of 9 years (2000 to 2009), with low CV (9%) and standard errors. Although inoculum availability is assumed in the model, additional improvements in prediction accuracy might be accomplished by incorporating P. infestans inoculum density and propagule survival.