Recommended Lists disease ratings

In the AHDB Recommended Lists for cereals and oilseeds (RL), variety disease ratings are presented using a simple scale – from 1 (least resistant) to 9 (most resistant). Find out how we calculate them.

Produced for several key diseases of cereals and oilseed rape, ratings are calculated using disease information from three types of replicated RL trials (not treated with fungicides):

• Inoculated trials (inoculated)
• Untreated (naturally infected)
• Disease observation plots (naturally infected, single replicate only)

Inoculated trials are particularly important, as they promote high disease pressure to bring out the differences between varieties.

As the inoculations use pathogens with known characteristics, effects can be ascribed to a particular disease with greater confidence.

For rusts, the UK Cereal Pathogen Virulence Survey (UKCPVS) provides isolates that are broadly representative of UK populations.

RL trial managers use visual disease assessment keys to allocate scores consistently: from 0% to 100% infection (Table 1).

In wheat, the top four leaves are assessed at various intervals. Where a top leaf has been fully expanded for less than 14 days, the second leaf is referred to as the top 'top leaf'.

Naturally senescent leaf tissue is not assessed. All chlorosis and necrosis attributable to disease is recorded. Where disease is patchy, disease is scored as an average over the whole plot.

Table 1. Example visual disease assessment keys for two wheat diseases

In July, AHDB receives disease data for validation.

Only data from trials in which disease reaches 5% are included in the final data set (although all data may be used for wider research purposes).

An over-years mean of disease severity is calculated for each disease in each variety.

The varieties with the highest and lowest disease severity are usually used to set high and low disease fixed points from which to compare other varieties (Figure 1).

Due to the influence of the fixed points, only varieties that have been on the RL for several years, with well-established ratings or established disease control varieties, are used.

In addition, a variety with very high levels of disease may not be selected, if its inclusion would skew the ratings for other varieties.

For most diseases, a low disease fixed point of 9 = 0% disease is used to ensure that an RL rating of 9 reflects resistance.

For diseases where both natural and inoculated data are available, disease ratings are calculated from each data set.

These are then combined, with a greater emphasis placed on the inoculated data, to give an overall disease rating.

Careful checks are made to compare new disease ratings with those from the previous year to ensure that there is a good relationship for established varieties, unless there is evidence for a change in disease races.

Figure 1. How fixed points are used to determine RL disease ratings

This graph shows how a susceptible variety and a resistant variety are selected to create 'fixed points' at each extreme.

A straight line between these fixed points is used to calculate disease ratings for all other varieties.

In this example, variety 'A' has a disease level of 10% on the horizontal axis, which results in a disease rating of 4 on the vertical axis.

The RL 2021/22 introduced major changes to the disease-rating-calculation approach for winter wheat yellow rust and brown rust.

Typically, cereal disease ratings are based on the average level of disease in trials, observed over a three-year period.

Over the last decade, rust ratings have failed to reflect in-season field observations for some varieties at some locations.

This is partly due to the diverse and dynamic nature of the UK’s rust populations and the rating-calculation method, but not the quality of disease data.

In response, two changes were made to the way the yellow rust ratings are calculated. One of these changes was also applied to the brown rust ratings.

1. Weighted ratings: Yellow rust and brown rust

Ratings were historically based on three-year average disease ratings, with each year of data contributing an equal amount to the rating.

Where pathogen populations are relatively stable, this method provides reliable and stable ratings.

As wheat yellow and brown rust populations are dynamic and diverse, a different approach to the rating-calculation was necessary.

From the RL 2021/22 edition, winter wheat rust ratings are ‘weighted’, so that the most recent year of data has the largest (and the oldest year of data the smallest) influence on the rating.

This approach makes the rating more sensitive to changes in rust population structure in the most recent year, while still using the valuable three-year data set.

However, when a major change to the pathogen population occurs, the data set may be restricted to the most recent trial year. 

For example, this occured for wheat yellow rust in the harvest 2025 trials, with ratings in the following RL edition (2026/27) based on a one-year data set.

Table 2. The previous rating calculation

Table 3. The current rating calculation

2. Reset ratings slope: Yellow rust

Over time, the position of the susceptible fixed point moved as a result of yellow rust population changes.

This has had an undesired effect – susceptible varieties were achieving a higher rating at the end of the last decade, than the start of it.

For example, a variety with 10% disease would have received a rating of 4 in 2012, but 5 in 2019.

The RL 2021/22 saw the scale reset to the 2012 slope, with a lower susceptible fixed point. This means, compared to the previous edition (RL 2020/21), the same amount of disease resulted in a lower rating.

RL 2021/21 also saw the introduction of new disease ratings for winter rye (brown rust) and winter triticale (yellow rust), in addition to changes to the way spring oat mildew ratings are calculated.