Understanding effects of genotype × environment × sowing window interactions for durum wheat in the Mediterranean basin

International audience Durum wheat is one of the most important crops in the Mediterranean basin. The choice of the cultivar and the sowing time are key management practices that ensure high yield. Crop simulation models could be used to investigate the genotype × environment × sowing window (G × E×SW) interactions in order to optimize farmers’ actions. The aim of this study was to evaluate the performance of the wheat model SiriusQuality in simulating durum wheat yields in Mediterranean environments and its potential to explore the G × E×SW interactions. SiriusQuality was assessed in multiple growing seasons at seven sites located in Italy, Spain and Morocco, where locally adapted cultivars were grown. The model showed good ability in predicting anthesis and maturity date (Pearson r >0.8), as well as above ground biomass and grain yield (6 % < nRMSE < 18 %). The model was then used to find the optimal 30-day sowing window to maximize grain yields at four sites, two were located in Italy (Florence, Foggia), and the other two were in Spain (Santaella) and Morocco (Sidi El Aydi) respectively. Among the cultivars, on the average between all sowing window, Amilcar had the best performance in Foggia (+33 % compared to the traditional cultivar Simeto) and in Sidi El Aydi (+22 % compared to Karim), Karim in Florence (+19 % compared to Creso) and in Santaella (+6 % compared to Amilcar). Instead Creso and Simeto showed the lowest production at all locations. The results showed that an earlier sowing window compared to the traditional one would have a positive effect on wheat yields in all environments tested, because of increased maximum leaf area index, grain number and size, and grain filling duration. Moreover, with earlier sowing, grain filling coincides with higher soil water availability, reducing the water stress and increasing the accumulation of dry mass in grains. In cooler and wetter locations, cultivars characterized by higher leaf area index and radiation use efficiency had the higher number of grains, while in the hottest and driest locations, short-cycle cultivars with high grain dry matter potential (e.g. through enhanced “stay green” capacity) should be preferred.