Towards a strategy for phenotyping architectural traits in mature F1 hybrids of an apple progeny

UMR DAP équipe AFEFInternational audienceOver the last decade several genetic studies investigated the genetic determinisms of architectural traits. However, most of these works were carried out in juvenile progenies, and the genetic determinisms of adult tree architecture remained poorly understood. Within this context, the present study is aimed at defining a phenotyping strategy at the beginning of mature stage that should provide both quantitative traits and qualitative criteria for genetic studies and selection schemes respectively. Accurate measurements and visual observations were performed on four-year-old apple hybrids focusing on the main architectural processes: primary and secondary growth, branching and form. Among the set of quantitative traits and qualitative criteria analyzed, a selection of relevant descriptors was carried out: (i) for quantitative traits on the basis of their genetic parameters (i.e. broad sense heritability and genetic correlations); (ii) for qualitative criteria on the basis of the observation robustness, their correlation with quantitative traits, and their genotypic effect. Considering quantitative variables, a set of 8 relevant traits that were related to an architectural process or an axis type was selected to represent the complexity of tree architecture at the beginning of mature stage. Clustering the genotypes from these 8 traits revealed many recombinations. The analysis of qualitative criteria showed that visual observations were efficient for traits that characterized global processes such as tree growth habit or vigor. However, discrepancies were found between quantitative and qualitative clustering, probably because some important traits that recombined in the quantitative clustering were either poorly correlated to qualitative criteria (e.g., percentage of long shoots), or not accounted in qualitative observations (e.g., annual shoot morphology). Further investigations are thus required in particular to define new criteria corresponding to the traits that were found to be relevant in the quantitative stud