Making wheat more resilient to climate change without compromising yields has become an urgent priority for the agricultural sector. Now, a study led by a research team from the University of Barcelona and the Agrotecnio research centre has identified an innovative way to address this challenge: combining advanced technology and artificial intelligence to select the best varieties of this crop.

The study, published in the journal Plant Phenomics, suggests a shift in perspective: it is necessary to focus not only on yield, but also on wheat’s ability to maintain consistent harvests despite changing weather conditions. The findings indicate that this combination of productivity and stability is key to ensuring safe harvests under variable environmental conditions.

The authors of the study are researchers Jara Jauregui, José Luis Araus and Shawn Carlisle Kefauver, from the Department of Evolutionary Biology, Ecology and Environmental Sciences at the UB’s Faculty of Biology and Agrotecnio; Nieves Aparicio and Sara Álvarez, from the Agro-technological Institute of Castilla y León (ITACyL), and María Teresa Nieto, from the National Institute for Agricultural and Food Research and Technology (INIA-CSIC).

Drones for monitoring wheat crops

The team analysed 64 varieties of durum wheat grown under two different Mediterranean conditions: irrigated and rain-fed. The aim was to identify which genotypes combine high yields with a stable performance across variable environments, with differences in temperature and water availability.

One of the most surprising findings is that the selected varieties are not those that retain their green leaves the longest until the end of the season, but rather those that grow vigorously at the start and mature slightly earlier.

In contrast, the rejected lines showed low initial vigour and retained their green leaves for longer, which does not guarantee a better yield.

As part of the project, the team used ground-sensors and drones equipped with RGB, multispectral and thermal cameras, enabling them to monitor crop development throughout the entire growing cycle. This technology provides key information about the wheat before harvesting, eliminating the need for harvesting and reducing both the costs and the time required for analysis.

Using all this data, the team trained artificial intelligence models capable of predicting both the yield and the stability of production for the different varieties with a high degree of accuracy.

This strategy could be a very useful tool for plant breeding programmes and could help develop wheat varieties that are equipped to meet the challenges of climate change.

Greener doesn’t always mean better

The researchers first analysed, separately, the yield and stability traits of durum wheat. They found that the genotypes with the highest yields are characterised by high initial vigour and sustained greenness during the rapid growth phases up to the end of the growing season. In contrast, the most stable genotypes exhibit lower initial vigour, slower growth and a shorter cycle, enabling them to make better use of the resources available for grain production. To identify a balance between these compensatory mechanisms, the experts developed a variety selection method that combines competitive yield with good stability.

The study concludes that vigorous early growth combined with early maturation is a key factor to achieving more consistent yields under variable environmental conditions, helping wheat cope better with drought and high temperatures.