AHDB releases a variety blend tool for winter wheat

Varietal mixtures increase a field’s genetic diversity and may help slow the spread of some diseases and reduce the risk of resistance breakdown. Fuelled by a passion for adding value to numbers, AHDB’s Bastiaan Brak discusses how he’s used Recommended Lists (RL) data to build a tool to guide variety-blend decisions.

As a research data analyst, it should come as no surprise that I am fascinated by numbers. They can tell us so much about how the world works. Yet, all too often, their full potential goes unexploited. Even within the relatively narrow field of variety trials, we only scrape the surface of the data story. So, when several UK farmers mentioned (independently) that they were experimenting with variety blends, I knew RL data could be crunched in a way to guide their decisions on which varieties to test.

The UK is not alone, either. Globally, there has been a spike in interest in the use of mixtures. For example, in France, it is estimated that the bread-making wheat area grown as mixtures has more than doubled recently – from around 5% in 2017 to around 12% in 2020 (source: FranceAgriMer).

The primary reason people consider mixes is to add genetic diversity to a field as part of efforts to spread risk – with disease topping the risk-management list. Although several scientific studies have shown that the technique has promise as a disease management tool, it is a complex area – involving numerous genetic and environmental interactions. As a result, it is best to test the approach on the farm before adopting the approach more widely. And this is where the variety blend tool comes in.

When it comes to the development of a tool, it is often best to focus on the simplest option and add in the bells and whistles later. Even in the most basic form, a tool can be surprisingly complex. As a result, the variety blend tool has simple mathematics at its heart. Luckily, for the RL component values considered by the tool, bigger numbers are better. For the selected components, the tool simply adds together the associated values and divides the total by the number of varieties in the mix – the bigger the average score, the potentially stronger the mix.

Of course, it is not that simple. Although many components are associated with 1–9 values (brown rust, yellow rust, septoria tritici and lodging) others have relatively large values (Hagberg Falling Number, specific weight and untreated yield). For these components, the tool converts their values to a 1–9 scale – where 1 and 9 represent the minimum and maximum values, respectively (see table), with other values (between these points) determined by a simple straight line. The tool also considers protein content values, as published in the RL.

As a mixture of at least three varieties is considered best, the tool considers either three-way or four-way mixes.

The values generated by the tool are based on the performance of single varieties in RL trials.

As such, the tool is not able to capture the complex interactions associated with varietal mixtures or predict relative performance. 

The tool needs to be used with caution and varietal mixes not considered as a recommendation. However, it can aid the identification of mixes for subsequent on-farm testing.

Based on NIAB winter wheat parentage data, the tool also assesses the potential influence of parents, grandparents, great grandparents and great-great grandparents in the mix. Once again, it is a simple calculation – this time based on the number of times a variety features in its lineage. A score of ‘1’ indicates that varieties in the blend share no ancestors, whereas a score of ‘0’ indicates that all varieties in the blend share the same parentage.

The tool visualises data in numerous ways. As with any tool, the best way to find out what it can do is to experiment with it – it is resilient and tough to ‘break’.

We plan to update the tool around the time the RL booklet edition is launched (typically, January). However, if you do plan to test mixtures on the farm, be sure to talk with grain buyers to assess any potential specification issues first.

This tool is just one of several data-driven tools I’ve helped develop recently – you can access them via AHDB’s ‘tools’ page.

• Winter wheat variety blend tool
• Variety selection tool
• BYDV management tool
• Light leaf spot forecast
• Phoma leaf spot forecast
• Sclerotinia infection risk tool
• Excess winter rainfall tool