Multiple Regression Methods to express ‘Melting Texture’ of ‘La France’ Pear by Physical Properties

Abstract

The melting texture of ‘La France’ pear develops only after the fruit has been pre-chilled at 1°C for several weeks after harvest. This melting texture has not been sufficiently evaluated based on physical properties, such as elasticity or penetration hardness. We investigated the melting texture characteristics of ‘La Prance’ pear by three physical methods: non-destructive vibration, visco-elastic, and conventional compression methods. Sinusoidal vibration signals from 100 to 2000 Hz were applied to the fruit. Then, the second resonant frequency (f₂) was obtained to calculate the fruit elastic index (E₂). Six visco-elastic indices: bulk elasticity (K), bulk viscosity (ζ), shear elasticity (μ), shear viscosity (η), Poisson’s ratio (σ), and apparent Young’s modulus (E_app) as well as strain were obtained by a damped vibration method. Another apparent Young’s modulus (E_comp) was also obtained by a conventional compression method. E_comp was significantly correlated with E₂ but not E_app. After the above physical measurements, the pear flesh was scored by an organoleptic test for texture and melting evaluation. Texture and melting scores were analyzed by four multiple regression methods (variable reduction, variable increase, stepwise, and forced input). The melting scores could be explained with two viscosity indices (ζ and η), suggesting that the viscosity is closely involved in the melting texture.