VG06047 Celery – Septoria Predictive Model

Research has evaluated modifications to a computer model that can reduce the number of sprays for control of late blight in celery.

Late blight is a fungal disease that attacks the leaves and stalks of celery crops as they mature. Celery is usually sprayed weekly to control late blight, which can result in up to 16 sprays being applied per crop.

The model showed that savings on sprays could be made in the early stages of crop production, before the plant canopy closed in. Most savings were made on winter crops. In these crops the model predicted a saving of up to 8 sprays.

Although there are additional hardware and monitoring costs, the reduced spray program under the model, was as economical as weekly fungicide applications.

At present for winter celery crops the model calculates the need to spray when temperatures exceed 13 °C. Our research recommends a systemic fungicide 10 weeks after planting or at canopy closure and then the use of the model at a lower temperature of 8 °C to calculate the need for further spraying.

This following report details the outcomes of a 2 year study of late blight of celery which investigated efficacy and economics of the TomCast disease forecasting model for timing fungicide sprays to control late blight without reducing quality or yield.


Elizabeth Minchinton Victor Galea
Desmond Auer Dean Harapas
Fiona Thomson Simone Vassiliadis
Lindsay Trapnell Slobodan Vujovic

Validation of a disease forecasting model to manage late blight (Septoria) in celery - 2008
Download 228kb

Findings :

The TomCast disease-predictive model for late blight in celery which estimates disease activity commences at 13 °C, requires modification, as our growth chamber studies demonstrated spore release was substantial at 10 °C, measurable at 8 °C, but sparse at 5 °C.

  • The TomCast model is very effective as a decision support tool in the early stages of crop growth prior to canopy closure, where it can save 6–8 sprays with spray thresholds of 10 or 15 DSVs for winter grown crops.

    At 10 weeks, first lesions or canopy closure (whichever comes first), application of a registered systemic fungicide followed by weekly applications of chlorothalonil will produce an economic yield equal to weekly sprays, with yields based either on grower estimates or incidence data.

  • This is the first report of the TomCast model being deployed until harvest, by reducing the start temperature to 8 °C at either 10 weeks, or first lesions or canopy closure (which ever comes first). This protocol reduced the number of sprays by 5 to control the disease and produced an economical yield, based on incidence data.
  • The Disease Doctor™ computer program designed to deliver the TomCast model was validated and produced similar or better control of Septoria late blight than the Excel equivalent.
  • Desk-top simulations of the Septoria predictor and Cercospora model, which have been touted as alternatives to TomCast for Septoria late blight control, overestimated or underestimated the number of sprays required, respectively, and are consequently inferior to TomCast.
  • Gibberellic acid may have the potential to enhance late blight control as two applications in glasshouse trials considerably reduced lesion size and the number of pycnidia on lesions. • Vapour Pressure Deficit (VPD) cannot be use to replace leaf wetness sensors under field conditions due to air movement.
  • A fuzzy logic model which estimates leaf wetness based on measurement of temperature, relative humidity and wind speed predicted periods of leaf wetness under field conditions with an accuracy of only 75%.

Acknowledgments :

The researchers gratefully acknowledge the financial support for this project from Horticulture ustralia Limited (HAL), AUSVEG, the Federal Government, the Department of Primary Industries Victoria and the Biosciences Research Division.

The authors thank the members of the advisory groups: Debra Corrigan, Silvio Favero, Glenn Favero, Paul Gazzola, Russell Lamattina, Mark Milligan, Tom Schreurs, Harry Velisha, Karl Riedel, Matt Newland, Ian Willert, Tim Harslett, Denise and Alex Harslett for their valuable contribution to this project.

South Pacific Seed Pty. Ltd. and Boomaroo Nurseries Pty. Ltd. are thanked for supplying seedlings.

The authors thank Tom Schreurs for supply field sites and seedlings and Dr Robert Holmes for his useful comments on the final report.

^ Back to top