Sensory interpretation of instrumental measurements 2: sweet and acid taste of apple fruit

The relationship between objective and sensory measurements of apple taste and flavour was investigated. The aim was to determine the objective parameters that were best correlated with sensory evaluation, and then to identify the minimum objective difference that was required before a trained sensory panellist could detect a difference in apple taste and/or flavour. Objective measures included titratable acidity, degreesBrix (soluble solids content), levels of individual volatiles, sugars and acids, as well as calculations of degreesBrix/titratable acid ratio. Sensory panellists were trained to assess sweet taste, acid taste, apple flavour, and overall flavour. Titratable acidity was the best predictor of acid taste (correlation of 0.86 for the median panellist), and differences between apples of 800 mg kg(-1) (0.08% titratable acid) were required before the average trained panellist could detect a difference in acid taste (P = 0.90). This value represented about a tenth of the range of titratable acidity values of the treatments presented to the trained panel. Sweet taste was difficult to predict using any of the objective methods. Indeed, the best objective predictor of sweetness was degreesBrix (correlation of 0.41 for the median panellist), which could predict a difference in taste when apples differed by more than 1 degreesBrix (P = 0.90). This value represented about a third of the range of degreesBrix levels presented to the trained panel. Thus, while acid taste may be predicted on the basis of titratable acidity, we recommend that evaluation of sweet taste and flavour attributes continue to require assessment by trained sensory panels. (C) 2002 Elsevier Science B.V. All rights reserved.