Efficient and sustainable plant protection is a cornerstone of modern viticulture. Tested in 2024 Certis Belchim the VineForecast algorithm in field trials on a Müller-Thurgau vineyard in Spiesheim, Germany. The aim was to reduce the number of plant protection measures to a necessary minimum and at the same time ensure effective control of Downy mildew and Powdery mildew to ensure that
The result was promising: up to three fewer sprays with the same level of disease control. This was achieved through the data-based planning of the VineForecast algorithm, which optimises the applications based on AI-optimised disease and weather forecasts as well as the growth phases of the vine.
VineForecast's lead modeller Paul Petersik, on the 2024 field trial
In addition to the written test presented here, you will find a summary of the field test in the video below.
Mit dem Laden des Videos akzeptieren Sie die Datenschutzerklärung von YouTube.
Mehr erfahren
Experimental setup and realisation
The experiments compared two strategies:
- IP (integrated pest management): Standard spraying sequences based on local agronomic recommendations.
- VineForecast: Spraying sequence optimised by the VineForecast algorithm, which takes into account local AI-optimised weather and disease data as well as growth parameters.
The trials took place under ideal conditions for disease development: The alternating humid and warm weather conditions favoured Downy Mildew development before flowering, while dry and warm conditions after flowering posed a high risk for Powdery Mildew.
The trial was coordinated by Eurofins. In addition to the treated areas, untreated controls were used to assess the disease pressure and the effectiveness of the strategies. The differences in spraying intervals and measures saved were analysed by the Eurofins documented.
Downy Mildew: One spraying saved with the same protective effect
Downy Mildew was a major threat before flowering due to frequent rainfall and high humidity. While the IP strategy envisaged applications at 11-day intervals, VineForecast extended the intervals in cooler phases to up to 16 days due to the reduced growth of the vine.
Important differences:
Third application (Profiler): Applied on 21 May (IP) and 28 May (VineForecast). The longer interval was made possible because the slower vine growth reduced the need for protection.
Skipped application: Due to the previously extended spraying intervals, the fourth measure (Zorvec Zelavin) was skipped in VineForecast.
At the end of the experiment:
Efficiency: 98% with both strategies.
Savings: One application less with VineForecast.
Powdery Mildew: Three measures saved and slightly better results
After flowering, the risk of Powdery Mildew increased due to warmer and drier conditions. The IP spraying sequence maintained constant intervals of around 11 days, while VineForecast extended the intervals and skipped several treatments.
Important differences:
Fourth application (Luna Max): The interval has been extended from 11 days (IP) to 18 days (VineForecast) thanks to the systemic protection of Luna Max.
Skipped measures: Three treatments (e.g. Kumar and Vivando) were omitted from VineForecast.
At the end of the experiment:
Efficiency: VineForecast showed equivalent results (93% grapes, 89% leaves) compared to the IP spraying sequence (81% and 87%).
Savings: Three fewer applications.
Conclusion: Optimised scheduling enables savings without compromising on the protective effect
The VineForecast algorithm showed that modern plant protection requires fewer measures through precise planning without compromising disease control. This saves resources, protects the environment and optimises the workload.
You are welcome to send us your questions about the test results to [email protected] or test the software free of charge for 30 days. Click here to register:
