1 Section of Chemical Engineering, The Faculty of Engineering and Science, Aalborg University, VBN2 Department of Chemistry and Bioscience, The Faculty of Engineering and Science, Aalborg University, VBN3 The Faculty of Engineering and Science (ENG), Aalborg University, VBN4 Department of Food Science, Faculty of Science and Technology, University of Aarhus
Regression models for predicting preharvest dry matter (DM) and soluble solids content (SSC), based on two spectral ranges (680-1000 nm and 1100-2350 nm), were compared. Models based on longer NIR spectra were more successful for both parameters (DM/SSC: R2 = 0.78-0.84; RMECV = 0.78/0.44; LVs = 6/7). SSC prediction was better than expected considering the presence of starch in fruit. Generally poor SSC prediction in the presence of starch could be related to the inability of models to distinguish between forms of carbohydrate. Variable selection and regression coefficients highlighted the contribution certain wavelengths made to DM and SSC models, especially for SSC prediction between ∼1500-1800 nm. Pear physiology may influence the accuracy of DM and SSC prediction due to the presence of stone cells just below the skin. These cells contain lignin and are a source of soluble solids in ripening fruit. Further research is needed to qualify and build on the results presented here.
Lebensmittel - Wissenschaft Und Technologie, 2014, Vol 59, Issue 2, p. 1107-1113