Effect of Kernel Size and Mill Type on Protein,Milling Yield,and Baking Quality of Hard Red Spring Wheat

Optimization of flour yield and quality is important in the milling industry. The objective of this study was to determine the effect of kernel size and mill type on flour yield and end‐use quality. A hard red spring wheat composite sample was segregated, based on kernel size, into large, medium, and small kernels, as well as unsorted kernels. The four fractions were milled in three roller mills: Brabender Quadrumat Jr., Quadrumat Sr., and Bühler MLU‐202 laboratory mills. Large kernels had consistently higher flour yield than small kernels across mills, with the Quadrumat Jr. mill showing the lowest flour yield. Mill type and kernel size significantly affected variation in flour protein molecular weight distribution. When compared with larger kernels, flour milled from the small‐kernel fraction contained a higher gliadin fraction and SDS‐unextractable high‐molecular‐weight polymeric proteins, which had positive correlations with bread loaf volume (r = 0.61, P < 0.05) and mixograph peak time (r = 0.84, P < 0.001). Overall, small kernels could contribute to enhancing flour breadmaking quality while having a detrimental effect on milling yield.     

Milling of Regular and Waxy Starch Hull-less Barleys for the Production of Bran and Flour

Five registered cultivars of hull-less barley (HB) with regular or waxy starch were milled in a Quadrumat Jr. mill to obtain whole grain flour; pearled in a Satake mill (cultivar Condor only), and the pearled fractions examined by microscopy to determine true HB bran. The samples were milled after tempering and drying in a Buhler mill to obtain bran and flour yields. Flour color and composition of HB were unaltered on milling in the Quadrumat Jr. mill. Microscopic evidence showed that a 70% pearl yield was devoid of the grain's outer coverings, including the aleurone and subaleurone layers. Therefore, the balance of 30% constitutes true bran in HB. Dry milling (as-is grain moisture) of regular starch HB in the Buhler mill gave 59% total flour and 41% bran (bran + shorts) yields, the comparative values for the waxy starch HB were 42 and 58%. On tempering HB from 9 to 16% grain moisture, the total flour yield decreased in both types of HB but to a lesser extent in the waxy starch HB due to decreases in reduction flour. On drying HB to 5 or 7% moisture, total flour yields increased due to contamination with bran and shorts. The milling study led to the conclusion that HB, at best, be dry-milled and a bran finisher be used to obtain commercial flour extraction rates. Lower total flour yields in the waxy starch HB than in the regular starch HB milled at the same grain moisture levels seemed due to higher β-glucan rather than grain hardness. Waxy starch HB flour had higher mixograph water absorption and water-holding capacity than regular starch HB or soft white wheat flour milled under identical conditions. Roller-milled HB products offer the best potential for entry into the food market.     

Effect of Soft Kernel Texture on the Milling Properties of Soft Durum Wheat

Worldwide, nearly 20 times more common wheat (Triticum aestivum) is produced than durum wheat (T. turgidum subsp. durum). Durum wheat is predominately milled into coarse semolina owing to the extreme hardness of the kernels. Semolina, lacking the versatility of traditional flour, is used primarily in the production of pasta. The puroindoline genes, responsible for kernel softness in wheat, have been introduced into durum via homoeologous recombination. The objective of this study was to determine what impact the introgression of the puroindoline genes, and subsequent expression of the soft kernel phenotype, had on the milling properties and flour characteristics of durum wheat. Three grain lots of Soft Svevo and one of Soft Alzada, two soft‐kernel back‐cross derived durum varieties, were milled into flour on the modified Quadrumat Senior laboratory mill at 13, 14, and 16% temper levels. Samples of Svevo (a durum wheat and recurrent parent of Soft Svevo), Xerpha (a soft white winter wheat), and Expresso (a hard red spring wheat) were included as comparisons. Soft Svevo and Soft Alzada exhibited dramatically lower single‐kernel characterization system kernel hardness than the other samples. Soft Svevo and Soft Alzada had high break flour yields, similar to the common wheat samples, especially the soft hexaploid wheat, and markedly greater than the durum samples. Overall, Soft Svevo and Soft Alzada exhibited milling properties and flour quality comparable, if not superior, to those of common wheat.     

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.     

Rapid Assessment and Prediction of Wheat and Gluten Baking Quality With the 2‐g Direct Drive Mixograph Using Multivariate Statistical Analysis

A computerized 2‐g direct drive mixograph was used to study the mixing characteristics of flours milled from a range of breadmaking cultivars obtained from five separate locations around the UK, providing 54 flour samples. Fifteen parameters were extracted from each mixograph trace using the Mixsmart software program and correlated with baking volume using partial least squares multiple regression statistical analysis to give a prediction of baking volume. Location had a considerable influence on the prediction of baking volume. Excellent predictions of baking volume were obtained from flours from individual locations (R² = 0.805–0.995), but predictions based on all cultivars without discriminating locations were poor. When mixograph and baking volume data for each cultivar were averaged over all five locations, a very high correlation was obtained (R² = 0.999). Preparation of flour samples using rapid, small‐scale milling procedures (Brabender Quadrumat Jr. mill and Perten 3100 hammer mill) did not have any adverse effect on prediction of baking volume. Mixograph parameters obtained from six commercial glutens of varying quality gave good correlations with test baking volumes, based on 6% gluten addition to a control flour.     

Application of the Single-Kernel Characterization System to Wheat Receiving Testing and Quality Prediction

Single-kernel characterization system (SKCS) 4100 measurements on wheat were reproducible and stable and gave good correlations with relevant reference data, e.g., kernel weight vs. 1,000 kernel weight, kernel hardness vs. particle size index, and kernel moisture vs. oven moisture. Under field conditions at a receiving station in Coleambally (NSW, Australia), the SKCS 4100 operated faultlessly and the reproducibility of the results was as good as in the laboratory. The measurements were completed within the time taken for the normal testing sequence, and the histograms were shown to provide valuable information about the samples that would not otherwise be available. For example, the distribution of moisture contents of individual kernels provides additional information about the samples' potential storage stability. Data on the uniformity of hardness could be interpreted in terms of the potential of the wheat to provide a consistent milling performance. An imprecise (r² = 0.44) but potentially useful calibration was obtained for the prediction of flour yield under test milling conditions using SKCS 4100 measurements on wheat. A much stronger correlation (r² = 0.83) was obtained between SKCS data on wheat and the starch damage contents of flours produced on a pilot mill. Thus, the SKCS 4100 has the potential for early generation screening of wheat lines for flour yield and starch damage.     

Application of the Single‐Kernel Characterization System to Durum Wheat Testing and Quality Prediction

The Single Kernel Characterization System (SKCS 4100) measures single kernel weight, width, moisture content, and hardness in wheat grain with greater speed than existing methods and can be calibrated to predict flour starch damage and milling yield. The SKCS 4100 is potentially useful for testing applications in a durum improvement program. The mean SKCS 4100 kernel weight and moisture values from the analysis of 300 individual kernels gave good correlations with 1,000 kernel weight (r² = 0.956) and oven moisture (r² = 0.987), respectively. Although significant correlations were obtained between semolina mill yield and SKCS 4100 weight, diameter, and peak force, they were all very low and would be of little use for prediction purposes. Similarly, although there were significant correlations between some SKCS 4100 parameters and test weight and farinograph parameters, they too were small. The SKCS 4100 has been calibrated using either the single kernel hardness index or crush force profile to objectively measure the percentage vitreous grains in a sample with reasonable accuracy, and it correlates well with visual determination. The speed and accuracy of the test would be of interest to grain traders. An imprecise but potentially useful calibration was obtained for the prediction of semolina mill yield using the SKCS 4100 measurements on durum wheat. The SKCS 4100 is useful for some traits such as hardness, grain size and moisture for early‐generation (F3) selection in a durum improvement program.     

Influence of Soft Kernel Texture on the Flour,Water Absorption,Rheology, and Baking Quality of Durum Wheat

Durum wheat (Triticum turgidum subsp. durum ) production worldwide is substantially less than that of common wheat (T. aestivum ). Durum kernels are extremely hard; thus, most durum wheat is milled into semolina, which has limited utilization. Soft kernel durum wheat was created by introgression of the puroindoline genes via homoeologous recombination. The objective of this study was to determine the effects of the puroindoline genes and soft kernel texture on flour, water absorption, rheology, and baking quality of durum wheat. Soft Svevo and Soft Alzada, back‐cross derivatives of the durum varieties Svevo and Alzada, were compared with Svevo, a hard durum wheat, Xerpha, a soft white winter wheat, and Expresso, a hard red spring wheat. Soft Svevo and Soft Alzada exhibited soft kernel texture; low water, sodium carbonate, and sucrose solvent retention capacities (SRCs); and reduced dough water absorptions similar to soft wheat. These results indicate a pronounced effect of the puroindolines. Conversely, SDS flour sedimentation volume and lactic acid SRC of the soft durum samples were more similar to the Svevo hard durum and Expresso samples, indicating much less effect of kernel softness on protein strength measurements. Alveograph results were influenced by the inherent differences in water absorption properties of the different flours and their genetic background (e.g., W and P were markedly reduced in the Soft Svevo samples compared with Svevo, whereas the puroindolines appeared to have little effect on L ). However, Soft Svevo and Soft Alzada differed markedly for W and L . Soft durum samples produced bread loaf volumes between the soft and hard common wheat samples but larger sugar‐snap cookie diameters than all comparison samples. The soft durum varieties exhibited new and unique flour and baking attributes as well as retaining the color and protein characteristics of their durum parents.     

Predicting Wheat Quality Characteristics and Functionality Using Near‐Infrared Spectroscopy

The accuracy of using near‐infrared spectroscopy (NIRS) for predicting 186 grain, milling, flour, dough, and breadmaking quality parameters of 100 hard red winter (HRW) and 98 hard red spring (HRS) wheat and flour samples was evaluated. NIRS shows the potential for predicting protein content, moisture content, and flour color b* values with accuracies suitable for process control (R² > 0.97). Many other parameters were predicted with accuracies suitable for rough screening including test weight, average single kernel diameter and moisture content, SDS sedimentation volume, color a* values, total gluten content, mixograph, farinograph, and alveograph parameters, loaf volume, specific loaf volume, baking water absorption and mix time, gliadin and glutenin content, flour particle size, and the percentage of dark hard and vitreous kernels. Similar results were seen when analyzing data from either HRW or HRS wheat, and when predicting quality using spectra from either grain or flour. However, many attributes were correlated to protein content and this relationship influenced classification accuracies. When the influence of protein content was removed from the analyses, the only factors that could be predicted by NIRS with R² > 0.70 were moisture content, test weight, flour color, free lipids, flour particle size, and the percentage of dark hard and vitreous kernels. Thus, NIRS can be used to predict many grain quality and functionality traits, but mainly because of the high correlations of these traits to protein content.     

Characterization of a Unique “Super Soft” Kernel Trait in Wheat

Kernel texture in wheat (Triticum sp.) is central to end‐use quality and utilization. Here we report the discovery of a novel soft kernel trait in soft white winter wheat (T. aestivum L.). Two heritable kernel phenotypes were selected among F₃‐derived sibs, hereafter designated “normal soft” (wild‐type) and “super soft.” Normal soft lines exhibited single kernel characterization system (SKCS) hardness index (HI) values typical of soft wheat (HI ≈ 20), whereas the super soft lines were unusually soft (HI ≈ 5). Under some environments, individual super soft lines exhibited HI values as low as HI = –4. The super soft trait was manifested in reduced SKCS kernel texture and higher break flour yields, with some increase in sodium carbonate SRC (solvent retention capacity) values and sponge cake volumes. Straight‐grade flour yield, flour ash, milling score, and cookie diameter were largely unaffected. With the possible exception of the sodium carbonate SRC values, we observed no indication that the super soft trait conferred any negative aspects to commercial soft wheat quality. As such, the super soft trait may provide wheat breeders with new opportunities to modify the end‐use quality of wheat.     

Determination of Bran Contamination in Wheat Flours Using Ash Content,Color, and Bran Speck Counts

A fast new method based on image analysis, ScanPro Speck Expert (SPX), to determine the bran contamination in wheat flour was studied and compared with existing methods (air-oven, ash, and color measurements) using an Agtron color meter and a Minolta chromameter. Twenty-one hard red winter wheat flour samples with ash contents of 0.30–0.58% were collected from the Kansas State University pilot mill and used for this study. Intrinsic variability in the flour sample because of randomness of bran speck orientation and distribution in the sample holder could result in variation in the speck count. Simple and multiple linear regression analyses showed that estimation of flour ash content from the SPX results (R² = 0.91) would be more accurate than the results from color measurements (R² = 0.66 [Agtron color meter] and 0.74 [L*]). The added capability of SPX image analyzer to not only count the number of bran specks but also to measure their areas probably increases the accuracy of determining the bran contamination in wheat flour by image processing.     

Compressive Strength of Wheat Endosperm: Analysis of Endosperm Bricks

The material properties of wheat grain endosperm are central to its processing and end‐use quality. The preparation of geometrically‐defined endosperm specimens free of bran, germ, and pigment strand can facilitate the objective study of endosperm material properties. This study was conducted to characterize the material properties of wheat endosperm from two soft, two hard, and one durum wheat varietal samples. Additionally, each varietal sample was sorted according to vitreous or mealy kernel type. Endosperm ‘bricks’ approximately 0.76 × 2.08 × 1.06 mm were prepared using an abrading (Kernel Sanders, KS) device. Bricks were tested in compression using a texture analyzer (TA.XTPlus). Stress‐strain curves were used to calculate failure strain, failure stress, failure energy, and Young's modulus. Additionally, the effect of brick aging up to one month, and changes in moisture content (freeze drying, oven drying, and equilibration to ≈10.5–11% mc) were studied. Intrakernel variation was assessed by preparing two sibling bricks (one from each cheek) from individual kernels. Failure strain, stress, and energy all had relatively high model R² values (0.68, 0.79, and 0.75, respectively). The ANOVA model R² for Young's modulus was 0.46. All models indicated variety as a highly significant source of variation in brick material properties. The effect of vitreous versus mealy kernel type was not consistent across varietal samples. Brick age and moisture content did not significantly affect brick material properties. Analysis of sibling bricks indicated that the magnitude of intrakernel variation was similar to that observed for individual varietal lots of uniform vitreous or mealy kernel type. Overall, failure strain provided a ranking and mean separation most consistent with kernel texture market class. The results obtained in the present study, although similar to other published reports do not closely agree with them on the material properties of wheat endosperm. Similarly, published results of material properties often differ considerably. The source of these discrepancies are at present unknown, but in some circumstances they may relate to specimen orientation relative to the source kernel, as there was evidence for anisotropic behavior. A companion study compares the variation in kernel texture obtained with the single kernel characterization system (SKCS) with that obtained here using bricks.     

Investigation of Wheat Milling Yield Based on Grain Hardness Parameters

The objectives of this study were to investigate the relationship between milling yield and grain hardness. A preliminary study was carried out with 20 samples (both hard and soft wheats) using the Brabender hardness tester (BHT) with two grind settings: one‐step grind (0‐10) and two‐step grind (2‐12: coarse; 0‐8: fine). The two‐step grind was correlated with particle size index, single‐kernel characterization system (SKCS) hardness, break yield, and reduction yield (P < 0.05), whereas there was no correlation with the one‐step grind method. An additional 64 samples were ground with the two‐step grind setting to further validate this method. In terms of the BHT crush profile, no discernible differences were observed between varieties for the coarse grind, whereas for the fine grind, hard wheat gave a higher BHT maximum peak height and shorter grinding time compared with soft wheat. The break and reduction yields were significantly correlated with both BHT and SKCS hardness (P < 0.05). The findings indicated that the BHT method could be used to differentiate for milling yield among the different varieties. Based on the results, two milling yield models were developed, and both gave highly significant correlations between the predicted and Buhler mill break (R² = 0.791, P < 0.05) and reduction yield (R² = 0.896, P < 0.05).     

Effect of Different Wheat Classes and Their Flour Milling Streams on Textural Properties of Flour Tortillas

Wheat flour tortillas were made from flour streams of three wheat cultivars: Jagger hard red winter wheat, 4AT-9900 hard white winter wheat, and Ernie soft red winter wheat. Wheat samples were milled on a Miag experimental mill. Twelve flour streams and one straight-grade flour were obtained. Tortillas were made from each flour stream and the straightgrade flour by the hot-press method. Tortilla stretchability and foldability were evaluated by a texture analyzer and six panelists, respectively. Flour protein and water absorption affected tortilla texture. The foldability evaluated by panelists was positively correlated with flour protein content, farinograph water absorption, and damaged starch (P < 0.05). The 2BK and 3BK streams of hard wheat produced tortillas with strong stretchability and good foldability. Middling streams of hard wheat yielded tortillas with lighter color and less stretchability. Under the conditions tested in this study, soft wheat flours were not good for producing flour tortillas.     

SDS-Protein Gel Test for Prediction of Bread Loaf Volume

Flour dispersed in aqueous solutions of sodium dodecyl sulfate (SDS) forms a proteinaceous gel when centrifuged at high speed. The conventional methodology for SDS gel testing was modified to develop a small-scale (<1 g of flour or wheat meal) screening test for evaluation of the protein quality of wheat for breadmaking. The principal modification involved centrifugation with a swinging-bucket rotor to facilitate direct measurement of gel height, which is the primary test parameter. The effects of suspension temperature and time, centrifugation speed, sample size, and sieving of ground wheat or flour on the efficacy of the test were examined. Gel height, wet weight, and protein content were assessed as test parameters. In the standard test procedure that was developed, 0.67 g of flour or ground whole wheat was dispersed in 13.5 mL of 1.5% SDS solution for 15 min at 20°C, followed by centrifugation at 80,000 × g for 30 min. The test was evaluated using seven Canadian commercial wheat flours with diverse breadmaking quality. For the samples, gel height was strongly related to loaf volume (R² = 0.89 and 0.95 for flour and ground wheat, respectively). Sieving flour through a 75-μm sieve slightly increased the predictive power of the test (R² = 0.94). SDS gel height gave better discrimination of samples for prediction of loaf volume than did the traditional SDS sedimentation test. The performance of the sedimentation test improved when sieved ground wheat was used. The relationship between gel height or protein content and flour protein content was comparatively poor (R² = 0.25). The SDS gel test appears to primarily measure the effects of flour protein quality.     

Relationship of Bread Quality to Kernel,Flour, and Dough Properties

This study measured the relationship between bread quality and 49 hard red spring (HRS) or 48 hard red winter (HRW) grain, flour, and dough quality characteristics. The estimated bread quality attributes included loaf volume, bake mix time, bake water absorption, and crumb grain score. The best‐fit models for loaf volume, bake mix time, and water absorption had R² values of 0.78–0.93 with five to eight variables. Crumb grain score was not well estimated, and had R² values ≈0.60. For loaf volume models, grain or flour protein content was the most important parameter included. Bake water absorption was best estimated when using mixograph water absorption, and flour or grain protein content. Bake water absorption models could generally be improved by including farinograph, mixograph, or alveograph measurements. Bake mix time was estimated best when using mixograph mix time, and models could be improved by including glutenin data. When the data set was divided into calibration and prediction sets, the loaf volume and bake mix time models still looked promising for screening samples. When including only variables that could be rapidly measured (protein content, test weight, single kernel moisture content, single kernel diameter, single kernel hardness, bulk moisture content, and dark hard and vitreous kernels), only loaf volume could be predicted with accuracies adequate for screening samples.     

Wheat Grain Hardness Among Chromosome 5D Homozygous Recombinant Substitution Lines Using Different Methods of Measurement

The level of grain hardness of wheat (Triticum aestivum) cultivars profoundly affects milling properties and end-use. We examined grain hardness among a genetically defined set of 83 chromosome 5D homozygous recombinant substitution lines derived from soft wheat cv. Chinese Spring and hard wheat cv. Cheyenne and compared four common methods of measuring wheat grain hardness. Measures of grain hardness included a modified particle size index, Brabender Quadrumat flour milling, near-infrared reflectance (NIR) spectroscopy, and the single-kernel characterization system (SKCS). Duncan's multiple range test was used to group recombinant lines according to parental classes. Quadrumat milling fractions, percent bran and middlings, were well correlated to NIR and SKCS grain hardness, whereas break flour, a traditional measure of grain hardness, was poorly correlated to other hardness measures. NIR and SKCS grain hardness measures provided the greatest and similar mean separations. Both methods identified recombinant lines as being significantly outside either parental class and significantly different from and in between the two parental classes. Between two divergent environments, correlations (r) for Quadrumat bran and middlings percents and NIR and SKCS hardness ranged from 0.83 to 0.94. Analysis of variance indicated that lines differed substantially for hardness, and hardness was highly influenced by environment, albeit consistently, as indicated by low line-location model interaction terms. The results confirmed the presence of major allelic differences assignable to chromosome 5D and suggested the action of minor gene(s). Break flour, in particular, showed strong indications of transgressive segregation independent of the Hardness (Ha) locus. The Perten 4100 SKCS provided the best (most discriminating) measure of the material properties of the wheat endosperm manifested by the action of the Ha locus.     

Solvent Retention Capacity Values in Relation to Hard Winter Wheat and Flour Properties and Straight‐Dough Breadmaking Quality

Solvent retention capacity (SRC) was investigated in assessing the end use quality of hard winter wheat (HWW). The four SRC values of 116 HWW flours were determined using 5% lactic acid, 50% sucrose, 5% sodium carbonate, and distilled water. The SRC values were greatly affected by wheat and flour protein contents, and showed significant linear correlations with 1,000‐kernel weight and single kernel weight, size, and hardness. The 5% lactic acid SRC value showed the highest correlation (r = 0.83, P < 0.0001) with straight‐dough bread volume, followed by 50% sucrose, and least by distilled water. We found that the 5% lactic acid SRC value differentiated the quality of protein relating to loaf volume. When we selected a set of flours that had a narrow range of protein content of 12–13% (n = 37) from the 116 flours, flour protein content was not significantly correlated with loaf volume. The 5% lactic acid SRC value, however, showed a significant correlation (r = 0.84, P < 0.0001) with loaf volume. The 5% lactic acid SRC value was significantly correlated with SDS‐sedimentation volume (r = 0.83, P < 0.0001). The SDS‐sedimentation test showed a similar capability to 5% lactic acid SRC, correlating significantly with loaf volume for flours with similar protein content (r = 0.72, P < 0.0001). Prediction models for loaf volume were derived from a series of wheat and flour quality parameters. The inclusion of 5% lactic acid SRC values in the prediction model improved R² = 0.778 and root mean square error (RMSE) of 57.2 from R² = 0.609 and RMSE = 75.6, respectively, from the prediction model developed with the single kernel characterization system (SKCS) and near‐infrared reflectance (NIR) spectroscopy data. The prediction models were tested with three validation sets with different protein ranges and confirmed that the 5% lactic acid SRC test is valuable in predicting the loaf volume of bread from a HWW flour, especially for flours with similar protein contents.     

Determining Weight and Moisture Properties of Sound and Fusarium‐Damaged Single Wheat Kernels by Near‐Infrared Spectroscopy

Single kernel moisture content (MC) is important in the measurement of other quality traits in single kernels because many traits are expressed on a dry weight basis. MC also affects viability, storage quality, and price. Also, if near‐infrared (NIR) spectroscopy is used to measure grain traits, the influence of water must be accounted for because water is a strong absorber throughout the NIR region. The feasibility of measurement of MC, fresh weight, dry weight, and water mass of single wheat kernels with or without Fusarium damage was investigated using two wheat cultivars with three visually selected classes of kernels with Fusarium damage and a range of MC. Calibration models were developed either from all kernel classes or from only undamaged kernels of one cultivar that were then validated using all spectra of the other cultivar. A calibration model developed for MC when using all kernels from the wheat cultivar Jagalene had a coefficient of determination (R²) of 0.77 and standard error of cross validation (SECV) of 1.03%. This model predicted the MC of the wheat cultivar 2137 with R² of 0.81 and a standard error of prediction (SEP) of 1.02% and RPD of 2.2. Calibration models developed using all kernels from both cultivars predicted MC, fresh weight, dry weight, or water mass in kernels better than models that used only undamaged kernels from both cultivars. Single kernel water mass was more accurately estimated using the actual fresh weight of kernels and MC predicted by calibrations that used all kernels or undamaged kernels. The necessity for evaluating and expressing constituent levels in single kernels on a mass/kernel basis rather than a percentage basis was elaborated. The need to overcome the effects of kernel size and water mass on single kernel spectra before using in calibration model development was also highlighted.     

基于SWAP-IES的旱区春小麦长势和产量模拟

基于观测数据和作物模型相同化的田块尺度作物生长监测,对于农田精准管理具有重要意义。为构建能准确模拟旱区春小麦长势和产量的同化模拟模型,该研究利用SWAP(soil-water-atmosphere-plant)模型和迭代集合平滑器算法(iterative ensemble smoother,IES),构建了适合旱区春小麦的SWAP-IES同化模拟系统,并利用2019—2020年田间观测试验数据,评估了同化叶面积指数(leaf area index,LAI)、土壤水分(soil water content,SW)及其组合在旱区春小麦生长模拟和估产中的作用。结果表明,相较于无同化情景,在吸收6次土壤水分观测数据后,模型对土壤水分模拟的R²从0.48提升到0.87。同化LAI时,各水分胁迫处理下LAI的模拟精度均最高,R²从无同化的0.35～0.62提升到0.76～0.96。同化LAI+SW时,各处理对生物量模拟的精度均最高,R²从无同化的0.40～0.67提升到0.73～0.96。轻度水分胁迫处理(T4～T5)下,仅同化LA...