‘Nam Dok Mai’ mango is a luxury commercial fruit in Thailand, but post-harvest diseases infecting the ripe fruit is a major problem affecting marketability. The objective of the present study was to map the supply chain of ‘Nam Dok Mai’ mangoes exported to Japan and analyze the critical points of post-harvest disease infection caused mainly by Colletotrichum gloeosporioides. Risk points of the post-harvest diseases were found by examining the material and information flows from processes ranging from field production to post-harvest handling, and these were obtained from mango growers and an exporter. The findings of interviews with mango growers and observations of the mangoes in field production were that the first point of risk was cultivar selection, while branch pruning and fruit bagging were further important processes causing post-harvest fruit decay. On the other hand, it was found that post-harvest handling was significant in decreasing anthracnose disease infection; this was seen at the step of dipping the fruit in 50 °C hot water for 3 min at the processing line.

Shredded cabbage is widely used in much ready-to-eat food. Therefore, rapid methods for detecting and monitoring the contamination of foodborne microbes is essential. Short wavelength near infrared (SW-NIR) spectroscopy was applied on two types of solutions, a drained solution from the outer surface of the shredded cabbage (SC) and a ground solution of shredded cabbage (GC) which were inoculated with a mixture of two bacterial suspensions, Escherichia coli and Salmonella typhimurium. NIR spectra of around 700 to 1100 nm were collected from the samples after 0, 4, and 8 h at 37 °C incubation, along with the growth of total bacteria, E. coli and S. typhimurium. The raw spectra were obtained from both sample types, clearly separated with the increase of incubation time. The first derivative, a Savitzky–Golay pretreatment, was applied on the GC spectra, while the second derivative was applied on the SC spectra before developing the calibration equation, using partial least squares regression (PLS). The obtained correlation (r) of the SC spectra was higher than the GC spectra, while the standard error of cross-validation (SECV) was lower. The ratio of prediction of deviation (RPD) of the SC spectra was higher than the GC spectra, especially in total bacteria, quite normal for the E. coli but relatively low for the S. typhimurium. The prediction results of microbial spoilage were more reliable on the SC than on the GC spectra. Total bacterial detection was best for quantitative measurement, as E. coli contamination could only be distinguished between high and low values. Conversely, S. typhimurium predictions were not optimal for either sample type. The SW-NIR shows the feasibility for detecting the existence of microbes in the solution obtained from SC, but for a more specific application for discrimination or quantitation is needed, proving further research in still required.