Authentication of Mango Varieties Using Near-Infrared Spectroscopy

Mango cultivars are presently identified phenotypically. In the present study, near-infrared spectroscopy in the wavelength range of 1,200–2,200 nm, in combination with chemometrics was evaluated for the purpose. Principal component analysis and partial least square (PLS) discriminate analysis were applied using characteristic variables as 0 and 1 for cv Alphonso and Banganapalli, respectively, in one group and for Dasheri and Malda, respectively, in the other group. PLS was carried out with and without pre-processing of spectral data. Wavelength range of 1,600–1,674 nm was found to be suitable for the classification of Alphonso and Banganapalli after detrending of data, while Dasheri and Malda could be classified in the wavelength range 1,200–2,200 nm after computing Norris-Gap derivative. Alphonso–Banganapalli could be differentiated from each other with 99 % accuracy, whereas, accuracy for Dasheri and Malda was about 94 at 5 % level of significance.