Generation and characterization of a training population in Vitis vinifera for enhanced genomic selection

Abstract

Context and purpose of the study. Modern viticulture is facing significant challenges due to global climate changes, spanning from extreme heat spells and water scarcity to the acceleration of grapevine’s phenological development with important consequences from budbreak to harvest. These include increased spring frost vulnerability, as well as changes in grape quality and composition. Genomic selection (GS) has emerged as a powerful tool to address such challenges, being capable of accelerating breeding programs by supporting the prediction of phenotypic performance based on genomic data. A robust training population that captures the genetic diversity and trait variability of interest is crucial to implement GS.

Materials and methods. On these premises, a semidiallelic cross including 7 parental genotypes of interest was carried out in spring 2023 to generate a GS training population. The resulting seedlings from 21 biparental combinations were genotyped via ddRAD-seq. The implementation of hyperspectral imaging by infrared sensors is under way to capture hidden phenotypes of early developmental stages.

Results. The cross yielded a total population size of approximately 2100 individual plants, namely 100 from each of the 21 combinations. We successfully obtained sequencing data from an excess of 40,000 loci across the genome and detected point polymorphisms in an average of 90,000 sites. Collection of hyperspectral images is scheduled for Spring 2025.

This training population will serve as a foundational asset for the development of predictive models aimed at guiding the selection of superior Vitis vinifera cultivars. These efforts are expected to contribute substantially to the advancement of breeding programs, enabling the identification of traits associated with resilience and quality under the challenges posed by global climate change.