Oilseed rape tool predicts sclerotinia infection risk

AHDB’s sclerotinia infection risk alerts service for oilseed rape provides a strong indication of risk and can help focus disease monitoring.

The alerts use forecast and observed weather data to flag potential infection risk periods during the main oilseed rape flowering period in the UK.

Based on an analysis of over-season performance – forecast alerts compared to actual alerts – the service has consistently delivered a high rate of accuracy (greater than 90%).

The tool uses current and historic weather data – temperature and relative humidity (RH) – to provide an area forecast of how favourable the environmental conditions are for infection over the next 48-hour period.

The map-based view uses a traffic-light system to highlight sites at ‘low risk’, ‘near-miss’ or ‘high risk’ of infection.

Catherine Harries, who manages disease research at AHDB, said: “After several seasons, it is reassuring that post-season analyses show the infection-risk forecast gets it right most of the time.

“For example, the forecast alerts were highly accurate (92%) in 2021, when compared to actual alerts during the period 15 March to 23 June 2021. However, accuracy was down on the 2020 results, mainly due to the wet May.”

In 2021, the service accurately predicted that conditions were unconducive for sclerotinia infection for around 80% of the time, with conditions conducive for around 10% of the time.

For any disease forecasting tool, it is important that the number of ‘false negatives’ reported is low. This is when conditions are predicted to be unconducive for disease when they are conducive. This can lead to missed fungicide applications when crops require protection.

Compared to temperature, relative humidity (RH) is challenging to model accurately. May 2021 was wet – in fact, the UK experienced the fourth wettest May on record. This contributed to higher RH levels than forecast. Despite this, the tool only generated around 5% false negatives in 2021.

Catherine said: “Although forecasts are unlikely to ever be 100% accurate, they can guide activity on the ground. If relative humidity is high, then use the service with more caution. RH spikes are more common near the coast or following localised showers.”

AHDB-funded research found that decisions based on weather-based infection risk and sclerotinia inoculum levels resulted in 26% fewer crops needing treatment.

The AHDB sclerotinia infection risk alerts web page also features inoculum-pressure data, from a small network of spore traps, and provides information on crop growth stages.

When crops are in flower (growth stages 60–69), spores are being detected – in the spore trap network or via local petal tests – and the forecast is showing high-risk infection conditions over the next 48 hours, then a protectant spray should be considered.

Typically, the optimum time for a single spray is just before mid-flowering on the main raceme and, since fungicides have protectant activity, should be applied prior to an infection risk alert.

Persistence of full-dose fungicides is approximately three weeks. If a spray is made earlier, or if the flowering period is extended, a second spray may be required under conducive conditions.

For the first time, the AHDB light leaf spot forecast reacted to winter rainfall as it fell during the winter. Prior to this, only two forecasts were produced – a preliminary and final forecast, at the beginning and end of winter respectively. The light leaf spot forecast web page now includes a month-by-month assessment of risk over the winter period. Although January was relatively dry, wet periods in February meant that risk-levels finished broadly in line with an ‘average’ winter.

With January–February 2022 air temperatures above the 30-year average, aphids are anticipated to take flight around three weeks earlier in Scotland and two weeks earlier over most areas of England compared to historic averages. The annually updated aphid forecasts can be accessed from the Rothamsted Research Insect Survey website.