Characterization and genetic analysis of durum wheat root system architecture under nitrogen deficiency in aeroponics

Roots capture the water and nutrients required for plant growth and thus, their function is crucial for crop production. The root system architecture (RSA) determines the plant adaptability to a specific environment, and it is the key to describe nutrient-efficient crops. Ultimately, learning more about the genetics behind the RSA would allow, via genomic-assisted selection, breeding for plants adaptation to difficult conditions while maintaining their yield potential. During the time of this master thesis, 218 elite varieties of durum wheat were phenotyped in the RootPhAir aeroponic platform while under either normal feeding (N+) or nitrogen stress (N-). The protocol was adapted to durum wheat necessities and noteworthy improvements were made, attaining a rate of nearly 100% of germinated seedlings and more than 95% of living plants at harvest. Multiple root weight traits were collected throughout the 4 experiments conducted and characteristics such as the number of seminal roots or the roots angle were measured trough image analysis software’s. The effects influencing the root traits were analyzed by using statistical mixed models. The lack of nitrogen caused significant changes to both shoots and roots in every experiment. Genotype-treatment interaction was found significant for root dry weight (RDW) in the third experiment and was close to statistical significance for the other trials. However, changes in environmental conditions caused significant differences across experiments for almost all the traits, and sometimes hiding the desired (N-) treatment effect. A genome wide association study (GWAS) was conducted after these phenotyping experiments and revealed the presence of 2 major QTLs for RSA for: total dry wight (GDW) and shoot to root ratio (SDW/RDW). Moreover, QTLs with high significance were detected for the interaction with the treatment, including the GWD_(N-) interaction. A total of 97 QTL clusters were found and, in this master thesis, we suggest that one of them may be of interest for future breeding programs.