GENETIC IMPROVEMENT OF PEANUT FOR RESISTANCE TO CERCOSPORA LEAF SPOTS USING HIGH-THROUGHPUT PHENOTYPING AND ARTIFICIAL INTELLIGENCE TOOLS

• Grant number: 25/10382-5
• Support Opportunities: Scholarships in Brazil - Post-Doctoral
• Start date: December 01, 2025
• End date: November 30, 2027
• Field of knowledge: Agronomical Sciences - Agronomy - Crop Science
• Principal Investigator: Sandra Helena Unêda-Trevisoli
• Grantee: Jardel da Silva Souza
• Host Institution: Faculdade de Ciências Agrárias e Veterinárias (FCAV). Universidade Estadual Paulista (UNESP). Campus de Jaboticabal. Jaboticabal , SP, Brazil

Abstract

Peanut (Arachis hypogaea L.) can be classified within the group of pulse crops, with its high protein content contributing significantly to the diet of populations in developing countries. In Brazil, peanut production has been increasing in recent years, with the state of São Paulo standing out as the leading producer. This expansion is driven by external market demand and aligns with the export trends of agribusiness in São Paulo. However, peanut cultivation faces several challenges, with diseases being one of the main limiting factors of grain production. Among these, early leaf spot [Cercosporidium personatum (Berk. & Curtis) Deighton] and late leaf spot (Cercospora arachidicola Hori) are the most destructive. Together, they are known as "leaf spots" and are present in most peanut-growing areas in Brazil. In the state of São Paulo, early leaf spot has shown to be the most prevalent and severe foliar disease affecting peanuts.The objective of this proposal is to perform partial diallel crosses, obtain F1 seeds, and apply modern phenotyping tools and artificial intelligence to initiate a peanut breeding program focused on resistance to leaf spot diseases. Resistance to these pathogens is controlled by multiple genes, allowing genetic improvement through successive cycles of selection and recombination of superior genotypes. To achieve this, crosses will be carried out between two groups of parents one resistant to leaf spots and the other with desirable agronomic traits, using a partial diallel scheme. The F1 hybrids obtained will be evaluated in field experiments, along with the parental lines, in standardized experimental plots.The plots will be assessed for traits related to disease resistance, productivity, and plant morphology. The data collected will allow diallel analysis to estimate the General Combining Ability (GCA) of the parents and the Specific Combining Ability (SCA) of the hybrid combinations. The most promising populations for the breeding program will be those resulting from hybrid combinations with favorable SCA values and involving parents with high GCA. In addition, high-throughput phenotyping techniques will be applied using optical sensors to capture images during plant development in the field. These images will be processed with machine learning and deep learning algorithms to enable the automated identification and classification of resistant and susceptible plants with greater precision and efficiency. This modern approach aims to accelerate the selection process, reduce subjectivity in traditional evaluations, and enhance the accuracy in identifying superior genotypes. (AU)