The current enzymatic assay approach (AACC International Approved Method 32-23) for the measurement of mixed-linkage β-glucan in small grains was modified to a cost-efficient and high-throughput format without compromising the accuracy of the results. Ten barley (Hordeum vulgare L.) genotypes used in the study represented a wide range of β-glucan content levels. A reduced reaction volume is used in the new protocol to adapt to a 96-well plate format. The volume of key components lichenase and β-glucosidase were reduced to 25% of the volume required in the original protocol and the cost per sample was reduced to 22% of that in the original protocol. Labor cost was also decreased to 25% of the original protocol as a result of format changes. The accuracy of the measurement from the modified protocol was comparable to the current standard enzymatic procedure. β-Glucan measurement accuracy of the modified and original protocols were also compared using 21 oat (Avena sativa L.) samples. The results indicated that the new protocol consistently produced accurate measurements in both barley and oat.
The Perten Single Kernel Characterization system is the current reference method for determination of single wheat kernel texture. However, the SKCS 4100 calibration method is based on bulk samples. The objective of this research was to develop a single-kernel hardness reference based on single-kernel particle-size distributions (PSD). A total of 473 kernels, drawn from eight different classes, was studied. Material from single kernels that had been crushed on the SKCS 4100 system was collected, milled, then the PSD of each ground single kernel was measured. Wheat kernels from soft and hard classes with similar SKCS hardness indices (HI 40-60) typically had a PSD that was expected from their genetic class. That is, soft kernels tended to have more particles at <21 μm than hard kernels after milling. As such, a combination of HI and PSD gives better discrimination between genetically hard and soft classes than either parameter measured independently. Additionally, the use of SKCS-predicted PSD, combined with other low level SKCS parameters, appears to reduce classification errors into genetic hardness classes by ≈50% over what is currently accomplished with HI alone.
Wheat flour and water mixtures at 90% absorption (dry flour basis) prepared at various mixing times were examined using Fourier transform infrared (FT-IR) reflectance spectroscopy. Spectra were obtained using a horizontal attenuated total reflection (ATR) trough plate. The apparent amount of protein and starch on the surface of the dough varied with mixing time but this was likely due to the polyphasic nature of the substrate and the changing particle distributions as the batter matrix was developed. Deconvolution of the Amide I band revealed contributions from alpha helical, beta-turn, beta-strand, beta-sheet, and random conformations. The ratio of beta-sheet to nonsheet conformations reached its greatest value about the same time that the mixture was most effectively separated by a laboratory-scale, cold-ethanol-based method but before the peak consistency measured by a microfarinograph.
Abrasion techniques were used to remove the hull and pericarp layers of barley kernels to obtain a smaller kernel enriched in endosperm. The objective of this study was to evaluate the fractions produced by two alternative abrading systems on four barley cultivars for potential use in fuel ethanol processes that feature an upstream (of the fermentation) dry fractionation system. Four barley cultivars, two hulled (Thoroughbred and Nomini) and two hulless (Doyce and Merlin), were scarified and whitened at 22 scarification times and three milling degrees (settings 2, 4, and 6), respectively. Three different abrasive surfaces (36, 40, and 50 grit) were used in the scarifier to determine the material removal ratio for each barley cultivar. Material balances and color analyses were conducted for all of the fractions produced. Three fractions were produced with the whitener at each milling degree: broken kernels, fine fractions >323 μm, and fine fractions <323 μm. Setting #2 seems to be the milling level that releases most of the hull in the hulled barley with the whitener. After 50 sec of scarification, rougher surfaces produced more fine material (<1,410 μm diameter) and consequently less coarse material (>1,410 μm diameter). A lower grit (36 grit) abrasive surface induced faster hull removal in hulled barley. Color parameters L* and b* were good indicators of the fine and coarse fractions produced by abrasive methods because they indicate the kernel layer removed and were modeled as a function of the fraction of the material produced. The information obtained in this study has application in designing processes capable of removing and recovering hull and pericarp layers of barley kernels and thereby producing smaller kernels or kernel pieces containing mainly endosperm tissue.
Colors of noodle doughs made from hard white winter wheat flours from Oregon were measured at optimum noodle water absorptions (NWA). Partial correlations, removing effect of protein concentration, indicated that NWA had negative relationships with 0 hr L* and 24 hr b*, and positive relationships with 0 and 24 hr a*. Kernel hardness index had positive simple and partial correlations with NWA without any significant (P < 0.05) correlation with color parameters. High molecular weight glutentin subunits (HMW-GS) significantly (P < 0.05) affected all measured noodle parameters except for 0 hr L*. Covariance analysis, using protein concentration as a covariate, indicated that HMW-GS significantly affected NWA and a* (P < 0.01). Wheat cultivars with HMW-GS 17+18 showed significantly higher mean NWA and a* values than those with alternative Glu-B1 subunits. Protein molecular weight distributions affected noodles, as shown by significant correlations with absorbance areas and % areas of protein size exclusion (SE) HPLC chromatograms. Protein fractions that had positive correlations with redness had negative correlations with yellowness. Applying multivariate analyses to SE-HPLC data to derive calibration models to predict fresh noodle dough a* and b* values had R2 > 0.91 and cross validations values of R2 > 0.75.
Cast zein films are brittle at room conditions, so plasticizers are added to make them more flexible. The tensile properties of these films are known to be affected by the relative humidity (RH) of the ambient air. However, little is known about how the plasticizers are affected by RH. Cast zein films were plasticized with either glycerol (GLY), triethylene glycol (TEG), dibutyl tartrate (DBT), levulinic acid (LA), polyethylene glycol 300 (PEG), or oleic acid (OA). Mechanical properties and moisture content (MC) of the films were measured after one week of storage at 3, 20, 50, 70, 81, and 93% RH. The relative humidity of the films' storage had a great effect on the films' tensile properties. All the films' tensile strength and Young's modulus values decreased as RH increased. Films containing DBT, TEG, LA, or PEG showed an increase in the percent elongation with increasing RH. Films containing GLY, OA, or no plasticizer did not show any increase in percent elongation as RH increased. The changes seen in tensile properties with increasing RH are because of zein's hygroscopic nature. The absorbed water will further plasticize the zein. The type of plasticizer used determined the extent of the changes seen in the tensile properties of films stored at different RH values. Depending on the plasticizers used in the film, there were large differences in the amount of water absorbed. Films increasingly absorbed water depending on the plasticizer they contained in the order GLY > TEG > LA > PEG > NONE > DBT > OA. Films containing hygroscopic plasticizers like TEG absorbed too much water at high RH and became weak, but they absorbed enough water at lower RH values to not be brittle. While films containing the more hydrophobic plasticizer DBT were brittle at intermediate RH values, they had good mechanical properties at high RH values.
Whole kernel corn was wet ashed with nitric and perchloric acids, and the resultant salts were dissolved in dilute acid. Concentration levels in corn samples were determined for zinc, manganese, copper, lead, cadmium, and chromium by flame atomic absorption. Because this technique is not sensitive enough to measure the small quantities of mercury present, an oxidation procedure was developed that decomposed organic material in corn at a temperature of 70°C. The mercury was retained effectively in an aqua regia solution until the elemental mercury was de emanated for measurement of the vapor by a nonflame atomic absorption technique. The aqua regia solution ensured that mercury was completely solubilized. In addition, mercury standards prepared in a 10% aqua regia medium have proved to be stable at low concentrations for extended periods. Mean concentration values for the seven metals studied in 11 different corn samples ranged from a high of 23 γ/g for zinc to a low of approximately 0.0024 γ/g for mercury.
Arabinoxylans and glutenins are two critical network-forming polymeric components of wheat flours, functionally related to mixing and baking performance and baked goods quality. For soft wheat flour applications, solvent-accessible arabinoxylans (pentosans) play a dominant role in both sweet and savory products. Solvent retention capacity (SRC) testing (Approved Method 5611, AACC International 2000) is a solvation assay for flours based on the swelling behavior of polymer networks in biochemically selective diagnostic solvents (Slade and Levine 1994; Kweon et al 2009). SRC is used to predict the functional contribution of each flour component capable of forming transient entanglement networks or permanent gel networks. Water SRC is related to overall water-holding capacity by all network-forming flour components. Diagnostic solvent that exaggerates swelling of arabinoxylan networks is 50% w/w sucrose solution in water compared to water alone. The diagnostic solvent that exaggerates swelling of glutenin networks is 5% w/w lactic acid solution in water compared to water alone. Therefore, SRC measures the magnitude of exaggerated swelling to predict the extent of func
Phenolic compounds function in natural resistance of plants to insects, pathogens, and abiotic stresses. We investigated the flavonoid and phenolic acid content in developing maize kernels from 8 to 28 days after pollination and found a complex and developmentally dynamic mixture of compounds. LC-MS analysis showed no detectable levels of the flavonoid maysin in any of the kernel tissue tested. Accumulation of the three most abundant analytes was developmentally and spatially regulated and independent of the regulatory transcription factor pericarp color 1 (P1). Major analytes have UV characteristics similar to hydroxycinnamic acids. LC-MS indicates molecular weights of 410, 440, and 862 mass units, representing coumaroylferuloylputrescine (CFP), diferuloylputrescine (DFP), and a novel compound, respectively. There are large differences in the abundance of these compounds between genetic backgrounds, and accumulation increases with age. We dissected and analyzed various tissue types and demonstrated that the majority of these compounds are present in pedicels, with lesser amounts in pericarp and placento-chalazal regions. These data indicate that complex phenolic acids and amides rather than flavonoids are the major phenolic constituents in developing kernels, and accumulation of these compounds is developmentally and spatially regulated, with the greatest abundance in maternal tissues surrounding the kernel.
An improved means of isolating zein is needed to develop new uses for corn zein. We have measured the yield of zein and evaluated the ability of acetic acid to remove zein from corn gluten meal, distillers dried grains, and ground corn using acetic acid as solvent. Acetic acid removed zein more quickly, at lower temperatures, and in higher yields when compared with alcoholic solvents. After 60 min at 25°C, approximately 50% of the zein in corn gluten meal was removed. A step change in yield from 43 to 50% occurs as the extraction temperature is increased from 40 to 55°C after mixing for 30 min at 25% solids. The protein composition of the zein removed from corn gluten meal using acetic acid is very similar to that of commercial zein by SDS-PAGE. The zein obtained from corn gluten meal using acetic acid had higher amounts of fatty acids and esters according to IR analysis, leading to slightly lower protein content. Films made from zein extracted from corn gluten meal using acetic acid had lower tensile strength (approximately 60% lower) than films produced from commercial zein. Fibers with very small diameter (0.4-1.6 μm) can be produced by electrospinning using the AcOH solution obtained after corn gluten meal extraction.
Two corn hybrids (3394 and 33R87) were steeped with three sulfite salts and five acids to test the effect of sulfur dioxide (SO2) source and acid sources on wet-milling yields and starch properties. Milling yields from each treatment were compared with a control sample that was steeped with 2,000 ppm of SO2 (using sodium metabisulfite) and 0.55% lactic acid. Sulfur dioxide sources were potassium sulfite, sodium sulfite, and ammonium sulfite; acids were acetic, hydrochloric, oxalic, phosphoric, and sulfuric. Starch yields were affected by the SO2 source and steep acids but the effects were hybrid-dependent. Different steep acids gave different starch yields when wet milled at the same pH. Among the acids tested, weak acids (lactic and acetic) tended to give higher starch yields compared with strong acids (hydrochloric, sulfuric, phosphoric, and oxalic). Some differences were observed with different sulfite salts and acids on starch pasting properties; however, there were no clear trends.
High-amylose corn starch was cooked in an excess-steam jet cooker in the presence of 5% oleic or palmitic acid, based on amylose. The cooked product was rapidly cooled in an ice bath and then freeze-dried or drum-dried. Amylose was removed from solution by forming helical inclusion complexes with the fatty acid, and the inclusion complexes formed submicron spherical particles upon cooling. The dried material was reconstituted to form a paste that exhibited gel-like properties upon standing, but that flowed readily when shear was applied. The rheological properties of these pastes were measured to determine the effects on the flow properties of 1) the solids concentration in the reconstituted paste, 2) the method of sample drying and reconstitution, and 3) the fatty acid used. The materials were very spreadable, and at the highest concentrations their flow properties were similar to a commercial shortening. The pasting properties of the dried solids were also examined.
We report efficient sample extraction and assay methods allowing quantitative determinations of proteinase activities from barley malt. The improved methods are used to assay > 2,200 developmental lines of malting barley for two subsets of proteinase activity. The distributions of the resulting activities suggest differences in population structures between the two types of protemases. Comparison of the activities of the green malt proteinases with standard malting quality measurements show highly significant correlations that differ between the proteinase subsets. The pH 4.5 hydrolysis of the artificial substrate Z-Phe-Arg-AMC correlates well with the traditional malting quality measurements, supporting the role of cysteine-class proteinases in mobilization of grain reserves during malting and mashing. Results from assays of gelatin hydrolysis at pH 6.0 suggest that these proteolytic activities may be involved in other aspects of seed C and N dynamics also linked to malting quality measurements. The differences between the pH 4.5 and 6.0 activities assayed here and their association with malting quality measurements suggest different physiological roles for the two proteinase activities in several aspects of seed germination. Either assay could be useful for population surveys, depending on the particular facet of seed metabolism under study.
Soluble protein extracts of germinating maize seedlings exhibited a limited ability to hydrolyze purified xylans, and specific assays were unable to confirm the presence of endo-beta-1,4-xylanase activity. However, extracts contained a variety of aryl-glycosidase activities, including beta-glucosidase, beta-xylosidase, and alpha-L-arabinofuranosidase. These activities peaked in three- to four-day seedlings and were particularly concentrated in shoot and root tissues. Maximal levels of beta-glucosidase were two orders of magnitude greater than those of beta-xylosidase or alpha-L-arabinofuranosidase. Isoelectric focusing gels revealed multiple forms of these enzymes. The principal beta-glucosidase and alpha-L-arabinofuranosidase protein species were clustered at pI 4.8-4.9 and pI 5.8-6.0, respectively. beta-Xylosidase activity appeared to be associated with both of these enzymes, and no evidence was obtained for a distinct beta-xylosidase.
Hybrids with high grain yield and higher starch, protein, or oil content are available to corn growers; however, they result from crossing adapted Corn Belt inbred lines that rarely include exotic germplasm. This study was conducted to determine whether Corn Belt lines introgressed with exotic materials from Argentina, Chile, Uruguay, Cuba, and Florida have appropriate wet-milling characteristics in their hybrids. Ten lines from the Germplasm Enhancement of Maize (GEM) project with different starch contents were crossed to three adapted inbred lines used as testers. The B73×Mo17 hybrid was used as a control. The F1 generation of these 30 experimental hybrids was analyzed using both near-infrared transmittance (NIT) technology and a 100-g modified wet-milling procedure, and measuring test and 1,000-kernel weight. There was great variation among physical, compositional, and wet-milling characteristics of the experimental hybrids, suggesting that exotic germplasm can be used to improve wet-milling characteristics of Corn Belt hybrids.
One of the fastest growing industries in the United States is the fuel ethanol industry. In terms of ethanol production capability, the industry has grown by more than 600% since the year 2000. The major coproducts from corn-based ethanol include distillers dried grains with solubles (DDGS) and carbon dioxide. DDGS is used as a livestock feed because it contains high quantities of protein, fiber, amino acids, and other nutrients. The goal of this study was to quantify various chemical and physical properties of DDGS, distillers wet grains (DWG), and distillers dried grain (DDG) from several plants in South Dakota. Chemical properties of the DDGS included crude ash (5.0-21.93%), neutral detergent fiber (NDF) (26.32-43.50%), acid detergent fiber (ADF) (10.82-20.05%), crude fiber (CF) (8.14-12.82%), crude protein (27.4-31.7%), crude fat (7.4-11.6%), and total starch (9.19-14.04%). Physical properties of the DDGS included moisture content (3.54-8.21%), A(w) (0.42-0.53), bulk density (467.7-509.38 kg/m3), thermal conductivity (0.05-0.07 W/m·°C), thermal diffusivity (0.1-0.17 mm2/sec), color L* (36.56-50.17), a* (5.2-10.79), b* (12.53-23.36), and angle of repose (25.7-47.04°). These properties were also determined for DWG and DDG. We also conducted image analysis and size determination of the DDGS particles. Carbon group characterization in the DDGS and DDG samples were determined using NMR spectroscopy; O-alkyl comprised >50% of all DDGS samples. Results from this study showed several possibilities for using DDGS in applications other than animal feed. Possibilities include harvesting residual sugars, producing additional ethanol, producing value-added compounds, using as food-grade additives, or even using as inert fillers for biocomposites.
Distillers dried grains with solubles (DDGS) is one of the coproducts obtained from dry-grind ethanol manufacturing. As the ethanol industry is Growing exponentially, the production of DDGS has been significantly increasing as well. To optimize the use of DDGS, it has to be economically transported from one part of the country to other parts, and stored efficiently. But DDGS has some flow issues, which often makes storage and transportation very problematic. So the objective of this study was to investigate the dynamic water adsorption characteristics of DDGS with four soluble levels at four temperatures and four relative humidities. Three mathematical models were then used to fit the adsorption data (Peleg, Pilosof, and Singh-Kulshrestha). As there was no model available for describing the water adsorption characteristics of DDGS with varying soluble levels at various temperature and relative humidity conditions, a new comprehensive model was developed. The new model, Ganesan-Rosentrater-Muthu (GRM) model, encompassed soluble level, temperature, and relative humidity effects, along with time and moisture content. The GRM model (R-2 = 0.94; F =16503.90) provided a good description of DDGS water adsorption behavior and can be used to predict the dynamic adsorption of water in DDGS for a broad range of storage conditions.
Bile acid adsorption by lignified dietary fiber in the human intestine is proposed as a mechanism for lowering blood cholesterol level and reducing colon cancer risk. In this study, we investigated how the concentration and composition of lignin in fiber influences the in vitro adsorption of primary bile acids (glycocholate, taurocholate, and glycochenodeoxycholate) and a secondary bile acid (deoxycholate). Adsorption studies were performed by incubating nonlignified and artificially lignified maize cell walls (dehydrogenation polymer-cell walls) with bile acids under conditions imitating the small intestine and distal colon. Artificially lignified cell walls had varying but defined lignin concentrations (4.8-19.0%) and compositions (varying from pure guaiacyl to pure syringyl lignins) but a uniform polysaccharide-protein matrix. Adsorption of bile acids by cell walls was in a range of 6-31% (4-26 nmol of bile acids/mg of cell walls), with glycochenodeoxycholate showing the highest adsorption rates. Neither lignin concentration nor lignin composition influenced bile acid adsorption, thus disproving a major role of lignin in bile acid adsorption.
Wheat bran-derived arabinoxylan-oligosaccharides (AXOS) recently have been shown to potentially exert prebiotic effects. In this study, 15 bran samples obtained by milling different wheat cultivars were treated with xylanases from Hypocrea jecorina (XHJ), Aspergillus aculeatus (XAA), and Pseudalteromonas haloplanktis (XPH) to assess the effect of bran source and xylanase properties on the AXOS yield and structure. The total arabinoxylan (AX) extraction yield was higher with XHJ (8.2- 10.7%) and XAA (8.2-10.8%) than with XPH (6.9-9.5%). Irrespective of the enzyme, a significant negative correlation was observed between extraction yield and arabinose to xylose (A/X) ratio of bran AX (r = -0.7), but not between yield and bran AX level. The A/X ratio of the extracted material was 0.27-0.34 for all bran samples and all enzymes, which combined with yield data and microscopic analysis, indicated primary hydrolysis of aleurone and nucellar epidermis AX. The average degree of polymerization (avDP) of the extracted AX was very low for all enzymes (2-3), owing to the release of high levels of monomeric arabinose and xylose. The release of these monosaccharides could be ascribed to I) the activity of wheat bran-associated enzymes (arabinofuranosidases and xylosidases); 2) the hydrolytic properties of the xylanases themselves; and 3) the presence of xylosidases as contaminations in enzyme preparation, in that order of importance. Heat treatment of bran before xylanase treatment significantly decreased the levels of monomeric arabinose and xylose in the extract, without affecting the extraction yield, resulting in a higher avDP of 3-7, thus yielding true AXOS. Overall, for AXOS production, wheat cultivars with a low bran A/X ratio of the AX are preferable as starting materials, and inactivation of bran-associated enzymes before incubation is desirable. The XHJ xylanase was the best enzyme for wheat bran-derived AXOS production.
Important rice grain quality characteristics such as percentage of chalky rice kernels are affected by both high and low night temperatures and by different day and day/night temperature combinations. High nighttime temperatures have also been suspected of reducing rice milling quality including head rice yields. Experiments to confirm or refute this have not been reported. A controlled climate experiment was conducted. Conditions in the chambers were identical except between 2400 and 0500 hours (midnight and 5 am). For those times, two temperature treatments were imposed: 1) 18 degrees C (low temperature treatment) and 2) 24 degrees C (high temperature treatment). Two cultivars were tested: LaGrue and Cypress. The high temperature treatment reduced head rice yields compared with the low temperature treatment. Grain widths were reduced for the high temperature treatment compared with the low temperature treatment. There was no effect of temperature on grain length or thickness. Amylopectin chain lengths 13-24 were increased by the high temperature treatment by approximately equal to 1%. Future research will focus on determining whether genetic variability exists among cultivars in their head rice yield response to high temperatures. After identifying a source of resistance to high temperature effects, this characteristic can be incorporated into rice cultivars. In addition, ways to reduce this effect, including biotechnological remedies, have the potential for increasing rice yield and quality.
With increasing production of distillers dried grains with solubles (DDGS), both fuel ethanol and animal feed industries are demanding standardized protocols for characterizing quality. AOCS Approved Procedure (Am 5-04) was used for measuring crude oil content in milled corn and resulting DDGS. Selected factors, including sample type (milled corn, DDGS), sample origin (ethanol plant 1, 2, 3), sample particle size (original matrix, <0.71 mm, <0.50 mm mesh opening; the last two materials were obtained by grinding and sieving), solvent type (petroleum ether, hexane), extraction time (30, 60 min), and postextraction drying time (30, 60 min) were investigated by a complete factorial design. For milled corn, only sample origin and extraction time had significant effects (P < 0.05) on crude oil values measured, but for DDGS, besides those two factors, sample particle size, solvent type, and drying time also had significant effects. Among them, the particle size of DDGS had the most effect. On average, measured oil content in DDGS ranged from 11.11% (original matrix) to 12.12% (<0.71 mm) and to 12.55% (<0.50 mm). For measuring the crude oil content of DDGS, particle size reduction, 60 min of extraction, and 60 min of drying are recommended. Regardless of the underlining factors, the method was very repeatable (standard errors <0.05). The observed particle size effect on crude oil analysis of DDGS suggests the need for similar confirmations using other analytical methods.
Beneficial reduction in glycemic response has been observed after consumption of some high-amylose foods. This study examined the effect of varying the moisture content and the particle size of the starch in a test food. Twelve men and 12 women consumed corn chips or corn muffins made with starch or starch plus cornmeal from standard corn (30% amylose, 70% amylopectin) or high amylose (70%) corn. Half of the subjects were hyperinsulinemic based on a prestudy glucose challenge. No gender differences were observed. The hyperinsulinemic subjects had significantly higher insulin and glucose responses and area under the curve as compared with the normal responders. Average glucose, insulin and glucagon were usually lower after muffins compared with chips or foods containing cornmeal. Insulin and glucose responses, but not glucagon, were significantly lower after the consumption of foods made with high-amylose compared with standard corn starch and in control versus hyperinsulinemic subjects. Average plasma glucose and insulin area under the curves after high-amylose foods were approximately half those after standard corn starch. The presence of cornmeal (increasing the particle size of the starch) had less effect on the response of glucose, insulin or glucagon than the type of starch or food consumed.
Cereal Chem. 63(5):414-419 A low-molecular-weight protein (S protein) fraction with a high affinity and SDS-PAGE patterns were similar for both varieties, but RP-HPLC for flour polar lipid was isolated from flours of one hard and one soft wheat patterns were quite different. Defatting the flour did not affect the yield of S variety. The fraction was characterized by lipid content and composition, protein or the PAGE and SDS-PAGE patterns. The two major acidic polyacrylamide gel electrophoresis (PAGE), gradient gel sodium components of the S protein fraction were isolated by Sephadex G-50 dodecyl sulfate PAGE (SDS-PAGE), and reversed-phase high- gel-filtration chromatography and characterized by electrophoresis, amino performance liquid chromatography (RP-HPLC). The content of this acid analysis, and RP-HPLC. PAGE, SDS-PAGE, and amino acid protein in the two wheat varieties was approximately the same. Lipid analysis results did not show any significant intervarietal differences, but contents and compositions of the two preparations were similar. PAGE the S protein fractions appeared distinctly different upon RP-HPLC.
A high-speed dual-wavelength sorter was tested for removing corn contaminated in the field with aflatoxin and fumonisin. To achieve accurate sorting, single kernel reflectance spectra (500-1,700 nm) were analyzed to select the optimal pair of optical filters to detect mycotoxin-contaminated corn during high-speed sorting. A routine, based on discriminant analysis, was developed to select the two absorbance bands in the spectra that would give the greatest classification accuracy. In a laboratory setting, and with the kernels stationary, absorbances at 750 and 1,200 nm could correctly identify >99% of the kernels as aflatoxin-contaminated (>100 ppb) or uncontaminated. A high-speed sorter was tested using the selected filter pair for corn samples inoculated with Aspergillus flavus; naturally infested corn grown in central Illinois; and naturally infested, commercially grown and harvested corn from eastern Kansas (2002 harvest). For the Kansas corn, the sorter was able to reduce aflatoxin levels by 81% from an initial average of 53 ppb, while fumonisin levels in the same grain samples were reduced an average of 85% from an initial level of 17 ppm. Similar reductions in mycotoxin levels were observed after high-speed sorting of A. flavus inoculated and naturally mold-infested corn grown in Illinois.