Predicting Scab, Vomitoxin, and Ergosterol in Single Wheat Kernels Using Near-Infrared Spectroscopy

Cereal Chem. 76(4):573–576 Near-infrared spectroscopy (NIRS) was used to detect scab damage and estimate deoxynivalenol (DON) and ergosterol levels in single wheat kernels. Results showed that all scab-damaged kernels identified by official inspectors were correctly identified by NIRS. In addition, this system identified more kernels with DON than did a visual inspection. DON and ergosterol were predicted with standard errors of ≈40 and 100 ppm, respectively. All samples with visible scab had single kernels with DON levels >120 ppm, and some kernels contained >700 ppm of DON. This technology may provide a means of rapidly screening samples for potential food safety and quality problems related to scab damage. Scab damage in wheat (Triticum aestivum L.) can occur when damp and cool conditions during the maturing and harvesting seasons create a favorable environment for the growth of the mold Fusarium graminearum. The mold can cause kernels to appear dull, lifeless, or chalky and may produce the toxin deoxynivalenol (DON). Scab does not uniformly affect all heads in a field or all kernels in a head. Often, only a portion of all heads or kernels is infected (Klenda 1995), with infection rates of heads reaching 50% (Liu 1985). The presence of scab adversely affects flour ash, flour color, glutenin levels, dough properties, and loaf volume (Dexter et al 1996). In addition, the toxin can cause digestive disorders, diarrhea, refusal to eat, and death in animals. DON is also a suspected human carcinogen (Luo et al 1990). Besides adversely affecting grain quality and food safety, scab can reduce yields by 50%. Crop losses in the United States in some years have exceeded $1 billion (Liu 1985, Meronuck 1997). Also, in years when DON levels are excessive, individual millers may spend more than $1 million as they attempt to blend wheat to meet Food and Drug Administration (FDA) guidelines (Anonymous 1998). FDA guidelines are 1 ppm for finished wheat products for human consumption, 10 ppm for cattle or poultry feed, and 5 ppm for other animals (Herrman et al 1995). Cleaning, milling, and baking can reduce DON levels. However, concentrations in finished products may not be significantly less than levels measured before milling (Scott et al 1984, Abbas et al 1985, Seitz et al 1986, Nowicki et al 1988, Trigo-Stockli et al 1996). Several methods exist to detect scab-damaged kernels, DON, or ergosterol levels. Ergosterol indicates the presence of fungal invasion (Seitz et al 1977). The Grain Inspection, Packers and Stockyards Administration (GIPSA) includes visual scab detection as part of their routine grain inspection procedures (USDA 1991). GIPSA inspection procedures consider kernels as scab-damaged if they have a significant amount of discoloration attributable to the fungus. However, kernels with little or no visible scab can have significant levels of ergosterol and DON (Seitz and Bechtel 1985) but not meet GIPSA criteria for damage. Chemical tests such as thin-layer chromatography, gas chromatography, or HPLC can measure DON or ergosterol levels (Miller et al 1983, Seitz and Bechtel 1985, Nowicki et al 1988, Trigo-Stockli et al 1996). However, these tests can take several hours to complete and typically measure DON or ergosterol in bulk samples. Bulk-sample results do not indicate whether the DON or ergosterol average values resulted from a few highly infected kernels or from many kernels infected at a lower level. Information about the distribution of DON in single kernels may assist millers in reducing DON to acceptable levels. Thus, a rapid, objective means of detecting scab-damaged kernels and indicating levels of DON and ergosterol is needed. Ruan et al (1998) showed that machine vision could estimate scab damage more accurately than a human expert panel and can be fast. They did not attempt to use machine vision to estimate DON or ergosterol levels. Near-infrared spectroscopy (NIRS) can estimate wheat quality characteristics such as internal defects, color class, protein, and hardness (Delwiche and Norris 1993, Delwiche and Massie 1996, Dowell 1998, Dowell et al 1998). Thus, applying NIRS to the detection of scab-damaged kernels and the estimation of DON and ergosterol in wheat may be possible. The objective of this research was to assess the feasibility of single-kernel spectral analysis as an objective method for measuring scab-damaged wheat kernels and for measuring DON and ergosterol levels. MATERIALS AND METHODS Ten samples of hard red spring wheat were collected from lots with high levels of scab damage. Lots originated from commercial sources and from scab nurseries. The GIPSA Board of Appeal and Review separated samples into scab-damaged and sound kernels. Kernels determined as sound kernels by GIPSA criteria were further separated into kernels that appeared healthy and kernels with any visible scab damage. Table I shows the number of kernels used for subsequent NIR, DON, and ergosterol measurements. About 45 kernels per sample were used for subsequent tests. Spectra Collection and Data Analysis A diode-array, near-infrared spectrometer integrated with a singlekernel characterization system (Perten Instruments, Reno, NV) was used to collect spectra from single wheat kernels. The spectrometer measures absorbance at 400–1,700 nm using an array of silicon and indium-gallium-arsenide sensors and collects data at 30 spectra/sec. The system can automatically deliver single kernels in the spectrometer viewing area at a rate of two kernels/sec. However, kernels were hand-placed in this research to minimize placement errors. Six spectra were collected from each kernel and averaged to reduce noise. Data were recorded in 5-nm increments. The spectra were stored on a hard disk for subsequent analysis using GRAMS/32 software (Galactic Industries Corp., Salem, NH). Spectra were analyzed by using partial least squares (PLS) regression, a spectral decomposition technique similar to principal component regression (Martens and Naes 1989). The PLS regression uses 1 Agricultural Engineer and Research Chemist, respectively, USDA ARS Grain Marketing and Production Research Center, Manhattan, KS 66502. Names are necessary to report factually on available data; however, the USDA neither guarantees nor warrants the standard of the product, and the use of the name by the USDA implies no approval of the product to the exclusion of others that may also be suitable. 2 Corresponding author. E-mail: [email protected] 3 Research associate, Department of Grain Science and Industry, Kansas State University, Manhattan, KS 66506. Publication no. C-1999-0526-06R. This article is in the public domain and not copyrightable. It may be freely reprinted with customary crediting of the source. American Association of Cereal Chemists, Inc., 1999.