Browsing by Author "Smith, Anne"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Evaluating near infrared spectroscopy as a rapid screening tool for microbiological safety of rooibos (Aspalathus linearis)
    (Stellenbosch : Stellenbosch University, 2023-03) Smith, Anne; Williams, Paul James; Stellenbosch University. Faculty of Agrisciences. Dept. of Food Science.
    ENGLISH ABSTRACT: Rooibos, or Aspalathus linearis, is a popular South African beverage, known worldwide for its wide array of health benefits. While rooibos is marketed as a health product, very limited information is available on the microbiological safety of rooibos. Microbiological testing methods currently available in the industry is time- consuming, costly and potentially inaccurate. Given rooibos’ known association with microbiological contaminants such as Salmonella, E. coli and coliforms, this study aimed to assess near infrared (NIR) spectroscopy as a feasible alternative for rapid screening of the microbiological safety of rooibos. Rooibos samples (n=30) were collected from the final stages of processing, to represent the samples used in industry for microbiological screening. Both control (n=15) and contaminated (n=15) samples from various regions and harvesting seasons were chosen to increase the variation within the sample set. All samples had known moisture contents (in the range of 8.8 % to 9.8 %) and microbiological profiles. All unsafe samples had only coliforms present as no samples containing E. coli or Salmonella could be obtained at the time of the study. Through principal component analysis (PCA), a distinctive separation emerged when plotting PC 1 (82.28%) against PC 4 (0.45%). This separation, observed in the direction of PC 4, suggested a significant distinction, between safe and unsafe samples. To investigate potential correlations with moisture content, samples were colour-coded accordingly. However, no discernible patterns related to moisture content emerged in the direction of PC 4. As a result, the conclusion was drawn that the observed separation along PC 4 was indicative of the presence of coliforms, serving as a marker for microbiological safety. The study demonstrated that employing NIR spectroscopy and partial least squares discriminant analysis (PLS-DA) enabled the development of a calibration and prediction model with 100% classification accuracy, sensitivity and specificity. With no instances of misclassification, this model emerges as a feasible option for use in industry. The study not only established the potential of NIR spectroscopy for microbiological safety screening, it also employed multivariate data analysis techniques. These included the examination of PC loadings and PLS-DA regression vectors, identifying critical wavelengths (1720 nm, 1930 nm and 2310 nm). These wavelengths, associated with moisture content and CH stretching, played a crucial role in discriminating between safe and unsafe samples. The presence of coliforms were inferred from these wavelengths. While addressing the current gap in microbiological safety assessments of rooibos, this research shows the potential of using NIR spectroscopy as a rapid and accurate screening tool for rooibos classification. The findings pave the way for improved food safety control practices and highlight the need for future research in the application of spectroscopic techniques to ensure the safety of rooibos.