DATA AND MODEL-BASED RESOURCES TO SUPPORT ITALIAN RICE BREEDING

The central challenge that humanity is facing is the need to meet the nutritional needs of a growing population. After the tremendous progress achieved during the green revolution, the yields of the primary cereal crops are now stagnating and the undergoing climatic changes represent a further threat. Among the technologies available to allow a further increase in yield, genetic improvement is the most promising. Plant breeding, though, is an expensive, time consuming and labour- intensive activity which relies on a thorough knowledge of the available germplasm for its efficient exploitation requiring the integration of the phenotypic expression with molecular data. The analysis of the interactions between genetic makeup, pedo-climatic conditions and management practices is thus essential to guide breeding programs aimed at improving the agronomic traits of the main herbaceous crops. Crop simulation modelling can be used to support such activities, via a cost- and time-efficient analysis of the performances of a wide range of phenotypes in different weather, soil and management conditions. The requirement is the minimum deviation between the phenotypic expression and its model representation, which should consider the known physiological limits and compensatory effects among traits. The lack of an extensive characterisation of available germplasm often impedes the availability of exhaustive data to support breeding programs via crop modelling. This applies to Italian rice agriculture, being characterized by a long history of cultivation with a vast varietal landscape. Crop model-based studies and services have already been developed in the area to support rice growers and local stakeholders, thus outlining a proficient case study for their implementation in breeding programs. This doctoral project aimed at analysing the morpho-physiological characteristics of the Italian rice germplasm mostly contributing to the yield increase in the 20th century, highlighting the evolutionary trends, and the associations with published molecular data. The released information enlarges previous findings and can be used to guide genetic improvement programs aimed at further improve current rice varieties. The field experimental activity produced ready-to-use quantitative data to further refine crop modelling capabilities in the area. Their integration in a crop model study allowed correlating yield component traits and model parameters, fostering the design of synthetic cultivars to facilitate and prioritize new breeding efforts.