An Integrated Microstructural, Sensory and Instrumental Approach to Describe Potato Texture

A new method was developed to objectively measure a range of microstructural features in thin sections of raw and cooked potatoes using light microscopy and semi-automatic digital image analysis. Scanning electron microscopy was employed for qualitative visualization of the texture showing the degree of cell adhesion and cell wall softening. Sensory texture analysis and uniaxial compression tests were used to determine the textural properties of cooked potato samples. Six conventionally and two organically grown potato cultivars were used in the study. Eight image features grouped the potato cultivars with respect to three different structural qualities by principal component analysis. The most important structural differences between potatoes is characterized as a contrast between potatoes having a low volume fraction of cell walls (large cells), a small number of cells and starch grains, and large numbers of elongated cells. Another type of variation in the structural data grouped the potato cultivars with respect to volume fraction of starch grains, volume fraction of gelatinized starch, and formfactor of cells. The volume fraction of raw and cooked starch and the formfactor of cells were significantly lowered during storage. Image features were correlated with sensory and textural properties using multivariate data analysis (Partial Least Square Regression). Volume fraction of raw starch, volume fraction of gelatinized starch, and dry matter content were found to be highly correlated with sensory geometrical qualities. Mechanical sensory properties were found to be more correlated with a uniaxial compression measure (stress at fracture) than was the image data.