Investigation of the internal quality of ‘Palmer’ and ‘Tommy Atkins’ mangoes by near-infrared spectroscopy

Abstract

This study focuses on optimizing mango harvesting and minimizing waste by using a handheld near-infrared (NIR) spectrophotometer to develop predictive models for assessing the quality of ‘Palmer’ and ‘Tommy Atkins’ mangoes. It aims to enhance mechanical resistance and reduce post-harvest losses. The study created prediction models for key quality attributes: soluble solids (SS), titratable acidity (TA), and dry matter (DM). For ‘Palmer’ mangoes, the first derivative of Savitzky–Golay (SG1) yielded the best SS predictions (the coefficient of determination for prediction [R²P] = 0.69, the square root of the mean error of prediction [RMSEP] = 1.56%, and the standard deviation ratio of prediction [SDRP] = 1.80). For ‘Tommy Atkins’ mangoes, the second derivative of Savitzky–Golay (SG2) was more effective (R²P = 0.72, RMSEP = 2.43%, and SDRP = 1.85). TA prediction models showed that SG2 was more effective for ‘Palmer’ (R²P = 0.56, RMSEP = 0.14%, and SDRP = 1.48), while multiplicative signal correction pre-treatment worked better for ‘Tommy Atkins’ (R²P = 0.59, RMSEP = 0.13%, and SDRP = 1.55). For DM predictions, SG1 was optimal for ‘Palmer’ (R²P = 0.83, RMSEP = 0.95, and SDRP = 2.44) and SG2 for ‘Tommy Atkins’ (R²P = 0.79, RMSEP = 1.36, and SDRP = 2.00). In conclusion, the handheld NIR spectrophotometer shows promise for accurate quality assessments in the mango production chain, enabling better decision-making on harvest timing and reducing post-harvest losses.