PT - JOURNAL ARTICLE AU - Leigh M. Schmidtke AU - Lachlan J. Schwarz AU - Claudia Schueuermann AU - Christopher C. Steel TI - Discrimination of <em>Aspergillus</em> spp., <em>Botrytis cinerea</em>, and <em>Penicillium expansum</em> in Grape Berries by ATR-FTIR Spectroscopy AID - 10.5344/ajev.2018.18048 DP - 2019 Jan 01 TA - American Journal of Enology and Viticulture PG - 68--76 VI - 70 IP - 1 4099 - http://www.ajevonline.org/content/70/1/68.short 4100 - http://www.ajevonline.org/content/70/1/68.full SO - Am J Enol Vitic.2019 Jan 01; 70 AB - Attenuated total reflectance Fourier-transform infrared (ATR-FTIR) spectroscopy, in conjunction with chemometric modeling and machine-learning algorithms, was successfully applied to objectively differentiate fungal mycelia of Aspergillus carbonarius, Aspergillus niger, Botrytis cinerea, or Penicillium expansum and mature winegrape berries (Vitis vinifera, cultivar Chardonnay) infected with any of these bunch-rot pathogens. The differentiation of B. cinerea-infected grape berries from those infected with either Aspergillus or Penicillium species shows promise as a tool for the rapid detection of the pathogen when grapes are received at the winery for processing. Support vector modeling provided class prediction for pathogen and control samples that was superior to other modeling techniques, whereas random forest models were successful in classifying samples infected with Aspergillus spp., illustrating the potential for these techniques to be applied to the assessment of bunch-rot pathogens. The use of ATR-FTIR shows potential for assessing the phytosanitary aspects of grapes.