European variety testing improves when countries link their data

In Europe, new varieties are assessed for their Value for Cultivation and Use (VCU). This is mainly done through national field trial networks. These trials are important for breeders, growers and policymakers, as they help determine which varieties are allowed onto the market.

Until now, this assessment has mainly been organised nationally, which means that information from other countries is used only to a limited extent. “Such information can be valuable,” says lead author Jip Ramakers of WUR’s Biometris chair group. “Data from other countries show how varieties perform under different growing conditions, such as differences in climate, soil and cultivation area. Often, the trials also cover different years. That extra variation helps researchers better understand how stable a variety’s performance is.”

Adding more trial sites may seem a logical route. The more often a variety is tested, the more accurately its performance can be estimated. But beyond a certain point, extra trials add very little. Many of the trial networks studied are already close to that efficient level.

Additional field trials are most valuable when they are located in other countries or under clearly different conditions. Expanding within such a network often mainly confirms the picture that is already there.

The greatest gain therefore lies in connecting existing networks more intelligently. In the two studies, the researchers show how such a joint analysis works. “You cannot simply compare a yield measurement from France with one from Germany,” Ramakers explains. “A variety may perform well somewhere because of its own genetic traits, but also because of favourable weather, more suitable soil or the local cultivation protocol, such as crop protection and fertilisation. By analysing datasets jointly, we can better estimate which part of the performance truly belongs to the variety and which part is the result of the conditions.”

Linking data works particularly well when varieties show a comparable performance pattern in different countries. In statistical terms, this means that the genetic correlation between the national trial networks is high enough. Information from one country can then help compare varieties in another country more effectively, even when not all of those varieties have been tested there.

In the study on wheat and maize, the precision of variety comparison increased by up to 46 percent for wheat and 27 percent for maize. This gain was mainly found in comparisons where varieties had not been assessed together in the same network.

The soybean study shows that Austrian and French VCU networks can also benefit from joint data analysis. The researchers looked not only at seed yield, but also at protein yield and protein content. This is relevant because soybean is an important protein crop. In both countries, seed yield and protein yield in soybean varieties have clearly improved. Protein content did not increase significantly.

In addition to linking existing trial data, the researchers point to a next step: making better use of environmental information, such as weather data, soil characteristics and location data. Such information can give researchers a clearer understanding of why varieties perform well or less well under specific conditions. It can also improve predictions of variety performance in similar growing environments. This allows more value to be gained from existing VCU networks, without immediately requiring many additional trial fields.

• Evaluation and optimization of wheat and maize national variety evaluation systems in Europe
• Evaluation and improvement of two European soybean VCU networks