From spectral measurement to softening trajectory prediction: Integrating spatially resolved spectroscopy with kinetic modeling in Kiwifruit

Accurate prediction of postharvest quality evolution and remaining shelf-life is essential for effective management of kiwifruit. However, most spectroscopic studies focus on quality assessment at a single time point and provide limited information on subsequent quality changes. In this study, visible and near-infrared spatially resolved spectroscopy (Vis–NIR SRS) was applied to monitor quality evolution of gold-fleshed kiwifruit at the individual fruit level during shelf-life storage. Compared with conventional NIR spectroscopy, SRS collects reflectance signals at multiple source–detector distances, enabling depth-dependent characterization of both scattering-related structural properties and absorption-related chemical composition. Multi-distance SRS measurements were used to develop partial least squares regression (PLSR) models for predicting key quality attributes, including firmness, soluble solids content, dry matter content, pH, and weight loss. Based on the predicted firmness, a zero-order kinetic model was applied to characterize softening behavior, and the softening rate constant ( k ) was estimated for individual fruits. To examine whether kinetic parameters are sensitive to mechanical stress during postharvest handling, controlled mechanical compression was introduced as a perturbation to the ripening process. Compression treatments resulted in increased k values and shortened shelf-life, indicating accelerated and more heterogeneous softening behavior. Both k and remaining shelf-life were accurately predicted from SRS data, with coefficients of determination exceeding 0.94 and residual predictive deviation values above 4.2. These results demonstrate that integrating SRS with kinetic analysis enables non-destructive and dynamic prediction of kiwifruit quality evolution and shelf-life at the individual fruit level, highlighting its potential for practical postharvest quality management.