Envirotyping to control genotype x environment interactions for efficient soybean breeding

In the context of the European protein deficit and the need for climate change mitigation by agriculture, soybean (Glycine max L. Merr) is of major interest to farmers and breeders. Expanding the crop to new cultivation areas requires understanding and control of the Genotype by Environment Interactions (GEI) that impede the genetic gain. New envirotyping methods, including the Target Population of Environments (TPE) characterisation, are key to the development of efficient breeding programs for specific or broad adaptations. The objectives were to (i) determine the environment types describing the European early soybean TPE, (ii) characterise the distribution and repeatability of the environment types across the TPE to identify the best suited adaptation strategy and (iii) demonstrate the importance of assessing the alignment between multi-environmental trials and the TPE for breeding decisions. In this study, 602 environments from France to Russia were clustered into five environment types using twelve eco-climatic factors, i.e. environmental variables that were calculated over specific phenological periods, such as the number of days below 15 degrees C between flower induction and first flower stages or the solar radiation quantity between first pod and first seed stages. These factors were previously identified as main drivers of GEI for yield in early maturity soybean (maturity groups '000 ' and '00 '). The environmental clustering explained 88% of the GEI effect on soybean yield. The five environment types that composed the TPE, mainly contrasted in the intensity and timing of stresses related to temperature (cold stress during the vegetative growth and heat stress during the reproductive growth) and water availability (precipitation amount, evapotranspiration and drought throughout the crop cycle). Interestingly, we observed geographical and temporal variations in the environment types distributions across the TPE as well as in their repeatability. These variations attested to the TPE heterogeneity and thus suggested that selection strategies based on either specific or broad adaptations should be combined. For example, specific adaptations to the third and fourth environment types were best suited in Eastern Europe while the broad adaptation to all environment types could be recommended in Western Europe. When broad adaptation was required, we demonstrated the need to assess the alignments between the environment types frequencies in the TPE and those observed when multi-environmental trials were conducted. This work will contribute to improving the existing soybean germplasm by considering the risks linked with weather variations and unpredictability so as to design elite soybean according to environment type.