Using multivariate techniques to evaluate the effect of cluster thinning in the wine quality: a comprehensive analysis on Tempranillo variety

Optimizing vineyard yield is essential for maintaining wine quality, yet determining the ideal cluster thinning strategy remains complex. This study investigates the effect of different yield levels on wine composition using a novel electrochemical biosensor array. The aim is to evaluate the potential of these biosensors as an advanced analytical tool for viticulture, providing a more precise and accessible alternative to conventional techniques. Tempranillo vines were subjected to three levels of cluster thinning—High (HY), Medium (MY), and Low (LY) yields—across three vintages (2020, 2021 and 2022). A bioelectronic tongue based on enzyme-functionalized potentiometric sensors and gold nanoparticles (AuNPs) was used to selectively detect phenolic compounds and sugars, key markers of wine quality. The biosensor system effectively differentiated wines according to yield level, showing greater sensitivity and specificity than Fourier Transform Infrared (FTIR) spectroscopy. Principal Component Analysis (PCA) revealed clear clustering patterns, with MY and LY wines exhibiting greater phenolic and sugar content than HY wines. Support Vector Machine Regression (SVM) confirmed strong correlations (R 2 > 0.9) between biosensor responses and chemical parameters, outperforming traditional analytical methods (R 2 ∼ 0.7). The integration of multivariate data analysis with this advanced biosensor highlights the potential of electrochemical sensing technologies for real-time applications in viticulture. This work advances analytical chemistry by demonstrating the efficacy of enzyme-modified potentiometric sensors and supports precision agriculture practices to optimize grape yield and wine quality. The biosensor system is currently at a pre-commercial development stage and has been implemented in a portable format, making it suitable for potential integration into decision-support tools in wineries and viticulture operations • Electrochemical biosensors assess wine quality from different cluster thinning levels. • Sensor array integrates enzymes and gold nanoparticles for selective compound detection. • Wines from medium and low yields showed higher phenols and sugars than high yield. • PCA and SVMR reveal strong discrimination and correlation with chemical parameters. • Biosensor system outperforms FTIR in sensitivity, specificity, and in-field applicability.