COMBINING UAV-BASED HIGH-THROUGHPUT PHENOTYPING AND GWAS FOR TRAIT DISCOVERY IN A TOMATO MAGIC POPULATION

Tomato is one of the most economically important vegetable crops worldwide. Breeding objectives vary according to end use, and must address the increasing demands of consumers and producers while coping with emerging environmental challenges exacerbated by climate change. Achieving these goals requires access to broad genetic diversity and the identification of reliable molecular markers associated with key agronomic traits. Multiparental Advanced Generation Intercross (MAGIC) populations represent a powerful resource for capturing genetic diversity and improving the resolution of genetic studies. SABER (Solanum lycopersicum Allele Biodiversity Enriched Resource) is a tomato MAGIC population developed from eight founder genotypes, including seven cultivated tomato lines and one wild Solanum cheesmaniae accession (LA1407). The founders were crossed through a stepwise scheme followed by successive generations of selfing to generate 445 recombinant sister lines, currently at generation 14 (G14). Genotypic characterization was performed on 431 G11 individuals using ddRADseq, generating a high-density SNP dataset distributed across the genome. In parallel, an extensive phenotypic characterization was conducted over multiple years, focusing on agronomic and fruit traits. This comprehensive dataset provides a framework for genome-wide association studies (GWAS) aimed at dissecting the genetic basis of traits relevant to tomato breeding. To complement conventional field evaluations, high-throughput phenotyping was performed through multitemporal UAV-based remote sensing, using a DJI Mavic 3 Multispectral equipped with RGB and multispectral (green, red, red-edge, and near-infrared) cameras. Images were acquired at successive phenological stages, allowing crop growth, and vigor to be tracked over time. Vegetation indices, such as the Normalized Difference Vegetation Index (NDVI), the Green NDVI (GNDVI), and the Visible Atmospherically Resistant Index (VARI), were used as proxies for biomass, vigor, and chlorophyll concentration. Beyond these spectral indices, canopy structure was characterized through image texture metrics, computed for every index at each acquisition date to capture spatial patterns within the canopy. Preliminary GWAS integrating genotypic and phenotypic datasets identified 60 significant SNPs associated with remote sensing-derived traits, with particularly promising associations detected for VARI. In contrast, field-based phenotypes revealed two significant SNPs associated with average fruit weight and average juice weight (measured on five fruits). These findings highlight the value of the SABER population as a reservoir of genetic and phenotypic diversity for tomato improvement. The integration of high-density genotyping, conventional phenotyping, and UAVbased high-throughput phenotyping provides a powerful platform for the genetic dissection of complex traits and the development of molecular markers to enhance breeding efficiency.