Using the soluble solids accumulation in tomatoes from fruit setting until harvest for the construction of a predictive model by hand-held NIR Spectroscopy

Tomato of the cultivar 'Endeavour' grafted on rootstock of the cultivar 'Kaiser' were grown on substrate in a greenhouse. The accumulation of fruit soluble solids content was followed from fruit setting to harvest in two different seasons: a first set during spring and a second one during summer. The aim was to get spectral acquisition for these periods using a hand-held NIR device in order to develop a calibration model for soluble solids content prediction. Fruits developed from fruit setting to harvest over 8 weeks in spring and 6 weeks in summer. The dynamics of soluble solids accumulation were also influenced by the season. PLS models for predicting SSC were determined on the base of two criteria: (1) the growing season (spring and summer) and (2) fruit maturity (before maturation and ripening period). The results showed that the performance of models could be influenced by the growing season and the maturity stage. During spring, the best model was achieved with fruit picked before ripening period while during summer the best model was achieved with fruits picked during ripening. Global SSC models based on fruits from the two seasons was possible for fruits harvested during ripening but not with fruits picked before ripening.