Automated Color Classification of Single Wheat Kernels Using Visible and Near-Infrared Reflectance

Modification of an existing single kernel wheat characterization system allowed collection of visible and near-infrared (NIR) reflectance spectra (450–1,688 nm) at a rate of 1 kernel/4 sec. The spectral information was used to classify red and white wheats in an attempt to remove subjectivity from class determinations. Calibration, validation, and prediction results showed that calibrations using partial least squares regression and derived from the full wavelength profile correctly classed more kernels than either the visible region (450–700 nm) or the NIR region (700–1,688 nm). Most results showed >99% correct classification for single kernels when using the visible and NIR regions. Averaging of single kernel classifications resulted in 100% correct classification of bulk samples.     

Single Wheat Kernel Color Classification by Using Near-Infrared Reflectance Spectra

An optical radiation measurement system, which measured reflectance spectra, log (1/R), from 400 to 2,000 nm, was used to quantify single wheat kernel color. Six classes of wheat samples were used for this study, including red wheat that appears white and white wheat that appears red. Partial least squares regression and multiple linear regression were used to develop classification models with three wavelength regions, 500–750, 500–1,700, and 750–1,900 nm, and three data pretreatments, log (1/R), first derivative, and second derivative. For partial least squares models, the highest classification accuracy was 98.5% with the wavelength region of 500–1,700 nm. The log (1/R) and the first derivative yielded higher classification accuracy than the second derivative. For multiple linear regression models, the highest classification accuracy was 98.1% obtained from log (1/R) spectra from the visible and near-infrared wavelength regions.     

Evaluation of a Near‐Infrared Reflectance Spectrometer as a Granulation Sensor for First‐Break Ground Wheat: Studies with Hard Red Winter Wheats

A single wheat class or blended wheats from two wheat classes are usually milled in a flour mill. A near‐infrared (NIR) reflectance spectrometer, previously evaluated as granulation sensor for first‐break ground wheat from six wheat classes, was evaluated for a single wheat class, hard red winter (HRW) wheat, using offline methods. The HRW wheats represented seven cultivars ground by an experimental roller mill at five roll gap settings (0.38, 0.51, 0.63, 0.75, and 0.88 mm) which yielded 35 ground wheat samples each for the calibration and validation sets. Granulation models based on partial least squares regression were developed with cumulative mass of size fractions as a reference value. Combinations of four data pretreatments (log 1/R, baseline correction, unit area normalization, and derivatives) and subregions of the 400–1,700 nm wavelength range were evaluated. Models that used pathlength correction (unit area normalization) predicted well each of the four size fractions of first‐break ground wheat. The best model, unit area normalization and first derivative, predicted all the validation spectra with standard errors of performance of 3.80, 1.29, 0.43, and 0.68 for the >1041, >375, >240, and >136 μm size fractions, respectively. Ground HRW wheats have narrower particle size distribution and better sieving properties than ground wheat from six wheat classes. Thus, HRW wheat granulation models performed better than the previously reported models for six wheat classes.     

Effect of Wheat Moisture Content on Meal Apparent Particle Size and Hardness Scores Determined by Near-Infrared Reflectance Spectroscopy

The AACC Approved Method for near-infrared reflectance (NIR) spectroscopy to produce a wheat hardness score for wheat market classification can be corrected for variation in wheat moisture content. The cause of the variation in NIR spectra resulting from variation in wheat moisture was investigated. Ten samples each of soft red winter, soft white winter, hard red winter, and hard red spring wheats were stored at 20, 40, 60, and 80 equilibrium relative humidity. Wheats were then ground on a cyclone grinder as required by the standard method. Variation in unground wheat kernel moisture content resulted in variation in NIR data. NIR log 1/reflectance values increased at all wavelengths as wheat moisture content increased. Spectral changes were related to changes in the apparent particle size of ground wheat meal as it was influenced by moisture content. Higher moisture contents produced slightly higher apparent particle size in meal, suggesting larger particles of pericarp that became more pliable at higher moisture (temper) levels. The apparent particle size of meal of high moisture wheats resulted in greater NIR radiation scattering and decreased reflectance. Meal moisture content itself had no effect on the two NIR wavelengths used to evaluate wheat hardness.     

Environmental Effects on Developing Wheat as Sensed by Near‐Infrared Reflectance of Mature Grains

For 30 years, near‐infrared (NIR) spectroscopy has routinely been applied to the cereal grains for the purpose of rapidly measuring concentrations of constituents such as protein and moisture. The research described herein examined the ability of NIR reflectance spectroscopy on harvested wheat to determine weather‐related, quality‐determining properties that occurred during plant development. Twenty commercial cultivars or advanced breeding lines of hard red winter and hard white wheat (Triticum aestivum L.) were grown in 10 geographical locations under prevailing natural conditions of the U.S. Great Plains. Diffuse reflectance spectra (1,100–2,498 nm) of ground wheat from these samples were modeled by partial least squares one (PLS1) and multiple linear regression algorithms for the following properties: SDS sedimentation volume, amount of time during grain fill in which the temperature or relative humidity exceeded or was less than a threshold level (i.e., >30, >32, >35, <24°C; >80%, <40% rh). Rainfall values associated with four pre‐ and post‐planting stages also were examined heuristically by PLS2 analysis. Partial correlation analysis was used to statistically remove the contribution of protein content from the quantitative NIR models. PLS1 models of 9–11 factors on scatter‐corrected and (second order) derivatized spectra produced models whose dimensionless error (RPD, ratio of standard deviation of the property in a test set to the model standard error for that property) ranged from 2.0 to 3.3. Multiple linear regression models, involving the sum of four second‐derivative terms with coefficients, produced models of slightly higher error compared with PLS models. For both modeling approaches, partial correlation analysis demonstrated that model success extends beyond an intercorrelation between property and protein content, a constituent that is well‐modeled by NIR spectroscopy. With refinement, these types of NIR models may have the potential to provide grain handlers, millers, and bakers a tool for identifying the cultural environment under which the purchased grain was produced.     

A Comprehensive Genotype and Environment Assessment of Wheat Grain Ash Content in Oregon and Washington: Analysis of Variation

A comprehensive analysis of the variation in wheat grain ash content has not been previously conducted. This study assessed the relative contribution of genotype and environment to variation in ash content, with a particular aim of ascertaining the potential for manipulating the trait using contemporary adapted germplasm. A total of 2,240 samples were drawn from four years of multilocation field plots grown in the wheat production areas of Oregon and Washington states. Genotypes included commercial cultivars and advanced breeding lines of soft and hard winter, and soft and hard spring wheats with red and white kernel color, several soft white club wheats, and one soft white spring waxy wheat cultivar. In addition to ash, protein content, test weight, and Single Kernel Characterization System kernel hardness, weight and size were also measured. In total, 20 separate fully balanced ANOVA results were conducted. Whole model R² values were highly significant, 0.62–0.91. Nineteen of the 20 ANOVA results indicated significant genotype effects, but the effects were not large. In contrast, environment effects were always highly significant with F values often one to two orders of magnitude larger than the genotype F values. The grand mean for all samples was 1.368% ash. For individual data sets, genotype means across environments varied ≈0.1–0.3% ash. The genotypes judged noteworthy because they had the highest least squares mean ash content were OR9900553 and ClearFirst soft white winter, NuHills hard red winter, Waxy‐Pen and Cataldo soft white spring, and WA8010 and Lochsa hard spring wheats. Genotypes with lowest least squares mean ash were Edwin (club) soft white winter, OR2040073H hard red winter, WA7952 soft white spring, and WA8038 hard spring wheats. In conclusion, wheat grain ash is more greatly influenced by crop year and location than by genotype. However, sufficient genotype variation is present to plausibly manipulate this grain trait through traditional plant breeding.     

Ultrastructure of Developing Hard and Soft Red Winter Wheats After Air- and Freeze-Drying and Its Relationship to Endosperm Texture

A transmission electron microscopic study was conducted on air- and freeze-dried developing wheat to determine the effects of drying on the structure of the starchy endosperm. Field-grown hard red winter wheat (Karl) and soft red winter wheat (Clark) were harvested at 15, 18, 21, 23, 25, 28, and 35 days after flowering (DAF). Wheat was dried by either air-drying in the spike at 28°C or freeze-drying following freezing in liquid nitrogen. Dried wheat was prepared for microscopy. Fresh samples of Karl and Clark were also harvested on the same days and prepared immediately for microscopy. The method of drying greatly affected cellular ultrastructure. The most pronounced change upon air-drying of developing samples was disappearance of individual protein bodies and conversion of the cytoplasm into a matrix-like material similar in appearance to storage protein matrix found in mature wheat endosperm. Freeze-dried wheats maintained nearly natural ultrastructure but exhibited various amounts of freeze damage. Conversion of protein bodies to a matrix was not observed in freeze-dried samples. The results suggest that hardness develops as a result of endosperm senescence rather than accumulation of particular grain components. Senescence may cause changes in the starch granule surface such that surrounding components bind tightly in hard wheats, whereas the binding is weaker in soft wheats. Therefore, the surface of starch granules might be more important than components the starch granules bind to in determining hardness.     

Development of Standard Procedures for a Simple,Rapid Test to Determine Wheat Color Class

Growing conditions, kernel characteristics, and genetics affect wheat kernel color. As a result, red and white wheats sometimes cannot be differentiated by visual examination. Soaking wheat kernels in a sodium hydroxide solution enhances the difference in color; red wheat turns a darker red, and white wheat turns straw‐yellow. Previously, when NaOH was used for wheat determination of color class, only a visual assessment was made under arbitrary conditions, many times not suitable for field work. In the present work, visible reflectance spectroscopy and visual assessments were used to optimize NaOH (2 mL/g of wheat) soak time (10 min), concentration (5M or 20%), and temperature (60°C). The optimal procedure will provide users who are not laboratory trained with inexpensive, safe procedures to definitively assign wheat color class in the shortest time in field locations. Calibration and prediction of several wheat cultivars using partial least square regression were used to validate the optimal test procedure. The test differentiated even rain‐bleached wheat and cultivars that were difficult to classify visually. No distinct correlation occurred between predicted color value and the number of red genes.     

Origin of French virgin olive oil registered designation of origins predicted by chemometric analysis of synchronous excitation-emission fluorescence spectra

The authentication of virgin olive oil samples requires usually the use of sophisticated and very expensive analytical techniques, so there is a need for fast and inexpensive analytical techniques for use in a quality control methodology. Virgin olive oils present an intense fluorescence spectra. Synchronous excitation-emission fluorescence spectroscopy (SEEFS) was assessed for origin determination of virgin olive oil samples from five French registered designation of origins (RDOs) (Nyons, Vallée des Baux, Aix-en-Provence, Haute-Provence, and Nice). The spectra present bands between 600 and 700 nm in emission due to chlorophylls a and b and pheophytins a and b. The bands between 275 and 400 nm in emission were attributed to alpha-, beta-, and gamma-tocopherols and to phenolic compounds, which characterize the virgin olive oils compared to other edible oils. The chemometric treatment (PLS1) of synchronous excitation-emission fluorescence spectra allows one to determine the origin of the oils from five French RDOs (Baux, Aix, Haute-Provence, Nice, and Nyons). Results were quite satisfactory, despite the similarity between two denominations of origin (Baux and Aix) that are composed by some common cultivars (Aglandau and Salonenque). The interpretation of the regression coefficients shows that RDOs are correlated to chlorophylls, pheophytins, tocopherols, and phenols compounds, which are different for each origin. SEEFS is part of a global analytic methodology that associates spectroscopic and chromatographic techniques. This approach can be used for traceability and vindicates the RDOs.     

Automated Single‐Kernel Sorting to Select for Quality Traits in Wheat Breeding Lines

An automated single kernel near‐infrared system was used to select kernels to enhance the end‐use quality of hard red wheat breeder samples. Twenty breeding populations and advanced lines were sorted for hardness index, protein content, and kernel color. To determine whether the phenotypic sorting was based upon genetic or environmental differences, the progeny of the unsorted control and sorted samples were planted at two locations two years later to determine whether differences in the sorted samples were transmitted to the progeny (e.g., based on genetic differences). The average hardness index of the harvested wheat samples for segregating populations improved significantly by seven hardness units. For the advanced lines, hardness index was not affected by sorting, indicating little genetic variation within these lines. When sorting by protein content, a significant increase from 12.1 to 12.6% was observed at one location. Purity of the red samples was improved from ≈78% (unsorted control) to ≈92% (sorted samples), while the purity of the white samples improved from 22% (control) to ≈62% (sorted samples). Similar positive results were found for sorting red and blue kernel samples. Sorting for kernel hardness, color, and protein content is effective and based upon genetic variation.     

用于污染黄曲霉毒素花生分选的荧光信号研究

为在加工前将黄曲霉毒素超限的带衣花生米从原料中剔除,参照已有的色选系统,提出一种依据黄曲霉毒素含量超限带衣花生米的专属荧光信号进行逐粒分选的技术构想。采用Cary Eclipse荧光分光光度计测定100粒外观具有代表性的带衣花生米表面的紫外-荧光规律,通过与免疫亲和层析净化荧光光度法(GB/T18979-2003)检测结果对比,判定了黄曲霉毒素超限带衣花生米的荧光光谱特征;通过绘制450/490、460/490荧光强度比值的箱线图,评估了表面荧光法判断黄曲霉毒素超限带衣花生米的准确率;在搭建的荧光成像系统上,对黄曲霉毒素超限带衣花生米进行了荧光成像。检测发现,在365 nm波长激发下,黄曲霉毒素超限带衣花生米在420~460 nm处有荧光峰;以450/490荧光强度比值为依据剔除超限值带衣花生米的判断准确率为81%;a.u.40的带衣花生米可在图像中呈现亮蓝荧光光斑。表明表面荧光信号可作为带衣花生米在线、无损、逐粒分选的专属光学信号,用于黄曲霉毒素超限带衣花生米的剔除。     

Assessment of Heat-Damaged Wheat Kernels Using Near-Infrared Spectroscopy

Heat damage is a serious problem frequently associated with wet harvests because of improper storage of damp grain or artificial drying of moist grain at high temperatures. Heat damage causes protein denaturation and reduces processing quality. The current visual method for assessing heat damage is subjective and based on color change. Denatured protein related to heat damage does not always cause a color change in kernels. The objective of this research was to evaluate the use of nearinfrared (NIR) reflectance spectroscopy to identify heat-damaged wheat kernels. A diode-array NIR spectrometer, which measured reflectance spectra (log (1/R)) from 400 to 1,700 nm, was used to differentiate single kernels of heat-damaged and undamaged wheats. Results showed that light scattering was the major contributor to the spectral characteristics of heat-damaged kernels. For partial least squares (PLS) models, the NIR wavelength region of 750–1,700 nm provided the highest classification accuracy (100%) for both cross-validation of the calibration sample set and prediction of the test sample set. The visible wavelength region (400–750 nm) gave the lowest classification accuracy. For two-wavelength models, the average of correct classification for the classification sample set was >97%. The average of correct classification for the test sample set was generally >96% using two-wavelength models. Although the classification accuracies of two-wavelength models were lower than those of the PLS models, they may meet the requirements for industry and grain inspection applications.     

Puroindoline Genotype of the U.S. National Institute of Standards & Technology Reference Material 8441, Wheat Hardness

Grain hardness (kernel texture) is of central importance in the quality and utilization of wheat (Triticum aestivum L.) grain. Two major classes, soft and hard, are delineated in commerce and in the Official U.S. Standards for Grain. However, measures of grain hardness are empirical and require reference materials for instrument standardization. For AACC Approved Methods employing near‐infrared reflectance (NIR) and the Single Kernel Characterization System (39‐70A and 55‐31, respectively), such reference materials were prepared by the U.S. Dept. of Agriculture Federal Grain Inspection Service. The material was comprised of genetically pure commercial grain lots of five soft and five hard wheat cultivars and was made available through the National Institute of Standards and Technology (SRM 8441, Wheat Hardness). However, since their establishment, the molecular‐genetic basis of wheat grain hardness has been shown to result from puroindoline a and b. Consequently, we sought to define the puroindoline genotype of these 10 wheat cultivars and more fully characterize their kernel texture through Particle Size Index (PSI, Method 55‐30) and Quadrumat flour milling. NIR, SKCS, and Quadrumat break flour yield grouped the hard and soft cultivars into discrete texture classes; PSI did not separate completely the two classes. Although all four of these methods of texture measurement were highly intercorrelated, each was variably influenced by some minor, secondary factors. Among the hard wheats, the two hard red spring wheat cultivars that possess the Pina‐D1b (a‐null) hardness allele were harder than the hard red winter wheat cultivars that possess the Pinb‐D1b allele based on NIR, PSI, and break flour yield. Among the soft wheat samples, SKCS grouped the Eastern soft red winter cultivars separate from the Western soft white. A more complete understanding of texture‐related properties of these and future wheat samples is vital to the use and calibration of kernel texture‐measuring instruments.     

基于特征荧光信号的去囊衣带芯橘瓣分选

为实现去囊衣带芯橘瓣的机器视觉分选,给去囊衣带芯橘瓣的机器识别提供直接、准确的判别信号,采用荧光分光光度计对50颗去囊衣带芯橘瓣的橘芯和砂囊分别进行三维荧光光谱扫描,通过对橘芯、砂囊平均三维荧光光谱分析,确定了橘芯相对砂囊的特征荧光信号,据此对带芯橘瓣荧光图像识别的可行性进行了验证。检测发现,在370~390 nm紫外光激发下,橘芯和砂囊在440 nm处的荧光强度差异较大;橘芯与砂囊在370 nm紫外光激发下,440 nm处荧光强度分布的箱线图表明两者荧光强度分布存在明显差异,且以橘芯在440 nm处荧光强度的下四分位数(Q1)与砂囊在440 nm处荧光强度的上四分位数(Q3)的平均值作为分类标准,去囊衣带芯橘瓣检出准确率可达85%。对(370±2)nm激发下得到的单色(440±5)nm荧光图像进行二值化及形态学处理后,可在以砂囊为背景的橘瓣图形中显现橘芯亮斑。利用橘芯与砂囊的荧光特性差异进行机器视觉成像分析,可作为识别去囊衣带芯橘瓣的一种有效方法。     

Fluorescence spectroscopy of aqueous pine litter extracts: effects of humification and aluminium complexation

Fluorescence excitation, emission, and synchronous-scan spectra were obtained for aqueous extracts of needle and litter layer (O-horizon) samples from ponderosa pine collected at a plantation site. The spectral lineshapes differed markedly between the needle and litter samples, and showed an increasing overall intensity with increasing extent of humification (increasing depth in the litter layer). At a dissolved organic C (DOC) concentration of 100 g m^?3, these effects were accompanied by a general shift in spectral density from lower to higher wavelength such that, in the excitation spectra, there was increasing prominence of a peak at 390 nm. When the DOC concentration was decreased from 100 to 25 g m^?3, the overall spectral intensity decreased and the peak at 390 nm in the excitation spectra of the litter samples gave way to a rising peak at 340 nm. Changes in pH from 4 to 5, characteristic of the undiluted litter extracts, produced little effect on the spectra. The addition of A1 at 40 mmol m^?3 generally produced enhancement of the fluorescence intensity in all three kinds of spectra for the needle and upper litter-layer extracts at 100 g C m^?3 and pH 4. Spectral density below 320 nm in the excitation and synchronous-scan spectra of the needle solution (possibly attributable to gentisic acid on the basis of model experiments) was, however, diminished by Al addition. For the lower litter-layer extracts, A1 addition decreased the fluorescence intensity in excitation and emission spectra, whereas it increased the intensity in synchronous-scan spectra. These trends indicated that the water-soluble fluorophores in the litter layer differed significantly from those in the fresh needles, and changed with the extent of humification. The 390-nm peak in the excitation spectrum, particularly its behaviour in the presence of added Al, may be useful as a spectral signature of products formed by litter humification processes.     

Interactions between bovine beta-lactoglobulin A and various bioactive peptides as studied by front-face fluorescence spectroscopy

Front-face fluorescence spectroscopy was used for the first time to study the interactions between bovine beta-lactoglobulin variant A (beta-Lg A) and various beta-Lg-derived bioactive peptides. Fluorescence spectra were recorded for beta-Lg A-peptide mixtures at 25 degrees C and pH 6.8 with an excitation wavelength of 290 nm to characterize the molecular environment of tryptophan (Trp) residues present in the protein but absent in the peptides. Spectra remained unchanged following addition of peptides beta-Lg f92-100 and beta-Lg f125-135, while Phe-Phe interaction between beta-Lg f69-83 molecules interfered with analysis. Addition of beta-Lg f102-105 produced a blue shift (3 nm) and a significant increase in fluorescence intensity, while addition of beta-Lg f142-148 also caused a significant increase in fluorescence intensity but accompanied by a red shift (3 nm). These results indicate that the polarity of the Trp environment in the beta-Lg A structure may be modified differently depending on the peptide added.     

Assessment of anthocyanins in grape (Vitis vinifera L.) berries using a noninvasive chlorophyll fluorescence method

Anthocyanins (Anths) in grape (Vitis vinifera L.) berries harvested at véraison from Pinot Noir and Pinot Meunier cultivars were assessed nondestructively by measuring chlorophyll fluorescence (ChlF) excitation spectra. With increasing Anth content, less excitation light was transmitted to the deeper Chl layers, and thus the ChlF signal decreased proportionally. By applying Beer-Lambert's law, the logarithm of the ratio between the fluorescence excitation spectra (log FER) from a green and a red berry gave the in vivo absorption spectrum of Anths, which peaked at about 540 nm. Absolute quantitative nondestructive determination of Anths for each berry was obtained by the log FER calculated for two excitation wavelengths, 540 and 635 nm (absorbed and not-absorbed by Anths, respectively) of ChlF at 685 nm. Over a range of skin colors going from green to purple, the relationship between the log [ChlF(635)/ChlF(540)] and the Anth concentration of berry extracts was fairly well fitted (r 2 = 0.92) using a power function. Reflectance spectra on the same berry samples were also measured, and Anth reflectance indices, which were originally developed for apples and table grapes, were derived. The log FER Anth index was superior to the reflectance-ratio-based index, but was as good as the color index for red grapes (CIRG) calculated from the whole visible reflectance spectrum. The proposed log FER method, applied by means of suitable portable devices, may represent a new, rapid, and noninvasive tool for the assessment of grape phenolic maturity in vineyards.     

Detection of Scab‐Damaged Hard Red Spring Wheat Kernels by Near‐Infrared Reflectance

Scab (Fusarium head blight) is a fungal disease that has become increasingly prevalent in North American wheat during the past 15 years. It is of concern to growers, processors, and the consumers because of depressed yields, poor flour quality, and the potential for elevated concentrations of the mycotoxin, deoxynivalenol (DON). Both wheat breeder and wheat inspector must currently deal with the assessment of scab in harvested wheat by manual human inspection. The study described herein examined the accuracy of a semi‐automated wheat scab inspection system that is based on near‐infrared (NIR) reflectance (1,000–1,700 nm) of individual kernels. Using statistical classification techniques such as linear discriminant analysis and nonparametric (k‐nearest‐neighbor) classification, upper limits of accuracy for NIR‐based classification schemes of ≈88% (cross‐validation) and 97% (test) were determined. An exhaustive search of the most suitable wavelength pairs for the spectral difference, [log(1/R)_λ1 ‐ log(1/R)_λ2], revealed that the slope of the low‐wavelength side of a broad carbohydrate absorption band (centered at ≈1,200 nm) was very effective at discriminating between healthy and scab‐damaged kernels with test set accuracies of 95%. The achieved accuracy levels demonstrate the potential for the use of NIR spectroscopy in commercial sorting and inspection operations for wheat scab.     

Evidence that Minor Sprout Damage Can Lead to Significant Reductions in Gluten Strength of Winter Wheats

Seventeen winter wheat lines were grown in triplicate plots at Warsaw, Painter, and Blacksburg, VA, during the 1999–2000 growing season. Hagberg falling numbers, protein content, farinograms, sedimentation volumes, and total glutenin content were determined for flours derived from 153 harvested wheat samples. Over three inches (8.2 cm) of rain fell during the week before harvest in Blacksburg, resulting in falling numbers of 100 for Recital and 137 for Heyne, two severely sprouted Blacksburg wheat samples, and falling numbers <250 in three other Blacksburg wheats. There were no significant differences across locations in falling numbers of four wheat lines, and one line had significantly greater falling numbers in Blacksburg than in Warsaw or Painter. All 18 Blacksburg flours had a significantly higher mixing tolerance index (MTI) and shorter departure times (DT) than corresponding Warsaw flours. Weaker gluten strength of Blacksburg flours suggestss that all 18 Blacksburg wheats were sprout‐damaged and contained active proteases. However, according to falling number data, five of these Blacksburg wheats were sprout‐free with falling numbers >400. These data indicate that Hagberg falling number should not be used as the sole criteria for determining the degree of sprout damage in wheat because it does not quantify, nor always accurately reflect changes in protein composition and quality due to grain weathering.