Fluorescence fingerprint characterization of O’Neal blueberries from different maturity levels

Fluorescence spectroscopy was evaluated as a rapid approach for assessing maturity in ‘O’Neal’ blueberries. The aim of this study was to characterize EEM fluorescence fingerprints of blueberry peel and flesh across four maturity stages and to relate key fluorescence regions to reference quality parameters, including SSC, TA, pH, firmness, and pigment composition. Fluorescence-derived PARAFAC components showed strong predictive relationships with several physicochemical parameters of blueberry peel, with PLSR models demonstrating high predictive performance for most quality attributes. Chlorophyll exhibited the highest accuracy (Rcal2 = 0.96, Rcv2 = 0.95, RMSECV = 0.01 g kg−1), followed by carotenoids (Rcal2[jls-end-space/]= 0.92, Rcv2[jls-end-space/]= 0.98), indicating a strong association between fluorescence signals and pigment composition. The models also showed good predictive capability for pH (Rcv2[jls-end-space/]= 0.96), SSC (Rcv2[jls-end-space/]= 0.95), and firmness (Rcv2[jls-end-space/]= 0.89), reflecting the sensitivity of fluorescence signatures to biochemical and structural changes during fruit ripening. In contrast, anthocyanin prediction showed lower predictive performance (Rcv2[jls-end-space/]= 0.53), suggesting more complex fluorescence behavior and potential spectral overlap with other phenolic compounds. Overall, these results highlight the strong potential of EEM fluorescence combined with PARAFAC and PLSR modeling as an informative approach for evaluating blueberry quality.