Prediction of chemical and sensory properties in strawberries using Raman spectroscopy

Abstract

Non-destructive methods for assessing the chemical composition of fruit and berries are being developed for use in e.g., sorting, storage management or as part of decision making in autonomous harvesting systems. In this study, Raman spectroscopy (RS) was used to estimate a selection of chemical components using partial least squares regression (PLSR) from early (June) and late (September) in the Norwegian strawberry harvesting season. Satisfactory PLSR models were made for total soluble solids (TSS), fructose, glucose, sum of sugars (SS), citric acid and sum of acids (SA) with coefficient of determination (R2) ranging from 0.81 to 0.92 when evaluated by cross validation. PLSR models for total acid content, sucrose and malic acid did not perform as well, with R2 ranging from 0.42 to 0.68, when evaluated by cross validation. Strawberries harvested in September showed significant difference between samples in sweet, sour, and acidic taste (p < 0.001). Results from RS demonstrated that the method can be used to determine sensory properties, where e.g., the correlation between predicted values of TSS and SS/SA with sensory sweet taste were 0.80 and 0.87, respectively. In conclusion, RS performed very well for characterization of both chemical and sensory properties in fresh strawberries.