Abstract
Near-infrared spectra (NIRS) from plant tissues can be used to predict traits owing to their relationship to internal biochemical states, shaped by both environmental and genetic components. Here, we tested the use of NIRS as predictors of budbreak the following year. We measured NIRS on leaf and bud tissue, collected at several dates during the growing season, of 240 dessert apple cultivars in 2021 and 2022. NIRS collected in 2021 and budbreak of 2022 were used to train partial least squares (PLSR) models, then tested using NIRS of 2022 to predict budbreak in 2023. A GWAS using these predictions identified a QTL, previously associated to budbreak in apple, indicating a significant genetic component was maintained in the predictions. Our results demonstrate the potential of NIRS to predict future developmental stages, such as budbreak, by detecting the metabolic states that precede them and could aid in genetic studies of difficult-tomeasure traits.