Downy Mildew (Peronospora viciae) in Spring Beans: Season 2 (2006)

Evaluation of a forecast scheme to warn of infection by airborne inoculum

The PGRO funded project "Spring beans - effect of downy mildew control on yield", which ran from 2000 to 2002, developed a qualitative warning scheme based on rainfall and air temperature. The aim of the CropMonitor bean downy mildew monitoring is to undertake validation testing before the warning scheme can be recommended for use by industry. After an encouraging start in 2005, this report summarises the results of the second season of validation, during which, a more specific numerical indication of risk was produced.

During April and May air temperature and rainfall monitoring at eight locations were started in early April, continuing into June as crops across the country approached flowering. A three of these locations, a local crop identified was monitored for development of bean downy mildew. The sites were located at:

• Bourne (Cambridgeshire)
• Huntingdon (Hertfordshire)
• Louth (Lincolnshire)
• Morley (Norfolk)
• Rosemaund (Hereford)
• Drayton (Warwickshire)
• Andover (Hampshire)
• York (North Yorks.)

An aim of the second season of evaluation was to extend the forecasting scheme to produce a daily numerical figure related to downy mildew infection risk, rather than a generalised commentary approach.

This risk figure was based on tracking how short period changes in rainfall (a six day total) and temperature (three day average) indicated when conditions turned particularly favourable for the disease. The figure produced is higher as infection risk rises, which enables easier comparison of the risk status at different locations as the season progresses.

How well did the forecasts work?

• All eight locations exhibited similar rainfall patterns during Spring 2006.
• Most locations experienced drier than average spells in early April (compared to the 30 years average), but all experienced a wetter than expected May and a return to drier conditions in early June. The conditions during May undoubtedly contributed to what many observers thought was a severe season for this disease.
• At the monitoring locations, all instances of downy mildew appearance were preceded by 8 to 19 days of conditions flagged as high risk by the forecast scheme - particularly with the rainfall criteria; cooler air temperatures played a lesser role.
• The average number of days warning of a likely infection was 15 days. Although this is dependent on where the 'significant' value of the warning is set, there was good degree of robustness in the timeliness of the warning. Raising (or reducing) the threshold value by 100% shortens (or lengthens) the warning by between seven and 11 days.
• Although the weather at all the sites was broadly similar the forecast scheme identified the Morley, Louth and York sites as those at greatest risk during the monitoring period. Seasonal risk is however related to flowering date and the most favourable weather for the disease may not coincide with the critical period as crops approach flowering.

Previous work funded by PGRO (which lead to the development of downy mildew forecasting) has shown that crops kept free of the disease during the two weeks leading up to flowering benefit the most from fungicide applications. The difficulty for growers remains that critically favourable conditions for downy mildew may be problematic for spraying. Further validation and monitoring will be undertaken in 2007.