Spatial variability of soluble solids or dry-matter content within individual fruits, bulbs, or tubers : Implications for the development and use of NIR spectrometric techniques

Additional index words. nondestructive quality evaluation, constituent variability, NIR calibration equations, Malus ×domestica, Cucumis melo, Citrus paradisi, Mangifera indica, Citrus sinensis, Prunus persica, Ananas comosus, Lycopersicon esculentum, Allium cepa, Solanum tuberosum Abstract. Spatial variation in soluble solids content (SSC) of fruits of apple ( Malus ×domestica Borkh. cv. Red Delicious), cantaloupe (Cucumis melo L. Cantaloupensis group), grapefruit (Citrus paradisi Macf. cv. Indian River Ruby Red), honeydew melon (Cucumis melo L. Inodorus group), mango (Mangifera indica L. cv. Hayden), orange (Citrus sinensis L. Osbeck. cv. Valencia), peach (Prunus persica L. Batsch. cv. Windblow), pineapple (Ananas comosus L. Merr. cv. Kew) and tomato (Lycopersicon esculentum Mill.), and of bulbs of onion ( Allium cepa L. Cepa group) and in dry-matter content (DMC) of potato (Solanum tuberosum L. cv. Russet Burbank) tubers was measured along three directional orientations (i.e., proximal to distal, circumferentially midway along the proximal to distal axis, and radially from the center of the interior to the outer surface). The pattern and magnitude of constituent variation depended on the type of product and the direction of measurement. Radial and proximal to distal variation was greater than circumferential variation in all the products tested. Honeydew had the highest radial variation with a SSC difference of 6.0 % and a CV of 22.8%, while tomato displayed lower radial variation with a CV of 1.0%. Pineapple had a proximal to distal SSC difference of 4.6% with a CV of 13.8%, while the difference in tomato was 0.6% with a CV of 5.1%. Circumferential variation of SSC in all products tested was <2% with CV ranging from 1.1% to 3.8%. The results confirm that considerable constituent variabil- ity exists within individual fruit and vegetable organs. This variability may affect the accuracy of calibration equations and their prediction capability. Therefore, within- unit constituent variability should be meticulously assessed when an NIR spectrometric method is being developed for the nondestructive quality evaluation and sorting of a product.