GrainGenes

Estimation of wheat flour lipids and lipid fractions by near-infrared spectroscopy.

Free lipids were extracted from 12.5 g (db) flours milled from 12 hard winter wheat varieties grown at six regions in Kansas. The extracted FL were dissolved in 5 ml hexane, and the mixtures were used for NIR measurement and further fractionated into nonpolar, glyco-, and phospho-lipids using a solid phase extraction system. Digalactosyldiglycerides (DGDG) were determined from glycolipids (GL) on HPLC. The absorbance (log 1/transmission) values of the FL-hexane mixture in 2 mm liquid cuvette were collected at 8-nm intervals between 400-2,500 nm using an NIRSystem 6500. SELECT (Version 3.0, Infrasoft Int., NIRSystems Inc.) software was used to choose 51 samples as the calibration set based on an H distance of 0.3 among 72 samples. The remaining 21 samples were used as a validation set. The best model for estimating FL contents was obtained by stepwise multiple regression (SMR) using the spectra that were transformed by weighted multiplicative scattering correction and first derivatization. The coefficients of determination (R²) were 0.94 for the calibration set and 0.88 for the validation set. GL content could be predicted by a model developed by SMR with scatter correction by detrend and standard normal variate transformation and second derivatization, which gave R² values of 0.87 for the calibration set and 0.89 for the validation set, respectively. For DGDG, the model obtained by SMR after multiplicative scattering correction and second derivatization of spectra showed R² values of 0.89 and 0.86 for calibration and validation sets, respectively.

Wheat and flour properties in relation to bread crumb grain.

Among breadmaking quality parameters, crumb grain is considered to be one of the most important. Yet, it is not clear what causes differences in crumb grain. We selected 12 hard winter wheat flours with protein contents of 11.8-13.6 % (14 % mb) that gave pup loaf volumes of 910-1,035 cc and different crumb grain scores ranging from 1 (questionable to unsatisfactory) to 4 (satisfactory). There were no significant correlations of crumb grain with grain hardness scores, protein and ash contents of wheat and flour, flour swelling power, starch damage, pasting properties by Rapid Visco-analyzer (RVA), mixograph parameters, or other baking parameters including loaf volume. However, the granulation of flour measured by both the Alpine Air Jet Sieve and NIR methods showed significant correlations with bread crumb grain. The weight % of particles less than 38 [micro]m in size and representing 9.6-19.3 % of the flour showed a positive and high correlation (r = 0.78, P < 0.01) with crumb grain, whereas flour particles larger than 125 [micro]m had an inverse relationship (r = -0.60, P < 0.05) with crumb grain.

Fortifying bread with each of three antioxidants.

White bread was fortified individually with fat-coated L-ascorbic acid (AsA), cold-water-dispersible (CWD) beta-carotene, and CWD all-rac-a-tocopheryl acetate (ToAc) at levels of 64 mg, 5 mg, and 100 mg, respectively, of active ingredient per 100 g flour (14 % mb). The freshly baked pup loaves retained 76 %, 67 %, and 96 % of the added antioxidants, respectively. To extract ToAc quantitatively from bread or dough, dimethyl sulfoxide gave better results than hexane, which indicated that ToAc was in a partially bound state. Protein-encased (PE) beta-carotene did not impart a yellow color to bread crumb and had one-fourth higher retention in fresh bread compared to CWD beta-carotene. In loaves stored at 25 C for 1-7 days, AsA disappeared rapidly and PE beta-carotene disappeared slowly; CWD beta-carotene and ToAc were stable. In spite of storage loss, bread fortified with PE beta-carotene retained significantly more beta-carotene compared to bread fortified with CWD beta-carotene. One serving (one slice, 28 g) of 3-day-old bread fortified with one of the three antioxidants was calculated to provide 6 %, 103-128 %, and 13-16 %, respectively, of the daily adult recommended dietary allowance (RDA) of vitamins C, E, and A. When bread was fortified with both fat-coated AsA and CWD beta-carotene and stored for 5 days, no sparing effect on the retention of the antioxidants was found.

Effects of genotype and environment on hard winter wheat flour pasting characteristics and their relationships with baking characteristics.

A rapid visco-analyzer was used to measure pasting characteristics for flours milled from 12 hard winter wheat varieties harvested at six regions in Kansas. The measured pasting characteristics included peak viscosity (PV), trough viscosity (TV), final viscosity (FV), peak time (PT), pasting temperature (PsT), trough temperature (TT), break down (BD), and set back (SB). These pasting characteristics showed significant genotypic (G), environmental (E), and `G x E' interaction effects by the analysis of variance (P < 0.05). However, E or `G x E' interaction effects were dominant over G effect by variance component analysis, indicating unstable responses of varieties to E. Flour protein content (PC) was correlated with TV, FV, PT, and SB, with linear correlation coefficients (r) of 0.380 (P < 0.01), 0.267 (P < 0.05), and 0.268 (P < 0.05), respectively. PsT had significant negative linear relationships with bread loaf volume (LV) and its potential, with r values of -0.348 (P < 0.01) and -0.333 (P < 0.01), respectively. TAM 200, TAM 107, and 2163 had significantly lower varietal averages of PsT than other varieties. LV was also linearly correlated with SB (r = 0.322, P < 0.01) and TV (r = 0.263, P < 0.05). Bake water absorption (WA) was linearly related with SB (r = -0.470, P < 0.01), TV (r = 0.448, P < 0.01); PT (r = 0.394, P < 0.01); FV (r = 0.315, P < 0.01); and PV (r = 0.314, P < 0.01). PsT was significantly (P < 0.01) included together with PC in a model developed by stepwise multiple regression (SMR) for prediction of LV with the coefficient of determination (R²) of 0.657. Bake WA was also modelled by SMR with R² of 0.814 using mixograph WA (P < 0.01), PC (P < 0.05), PT (P < 0.05), FV (P < 0.01), and SB (P < 0.05) as independent variables.

Effect of genotype and environment on gluten characteristics and their relationships with baking characteristics of hard winter wheats.

Gluten characteristics were measured by a glutomatic system using flours milled from 12 hard winter wheat varieties harvested at six regions in Kansas. Growing environment strongly influenced the dry gluten content (DGC), wet gluten content (WGC), and water amount in wet gluten (WAWG), even though there was also a significant varietal effect shown by analysis of variance. However, varietal effect was dominant for qualitative characteristics of gluten including gluten index (GI), % water of wet gluten (PWWG), and ratio of wet gluten to dry gluten (WG/DG), but the interaction effect of genotype and environment was insignificant (P < 0.05). Bread loaf volume (LV) had significant linear relationships with WGC, DGC, and WAWG as shown by linear correlation coefficients (r) of 0.648, 0.607, and 0.578, respectively (P < 0.001). Bake water absorption (WA) also had significant positive relationship with DGC (r = 0.834, P < 0.001); WGC (r = 0.825, P < 0.001); and WAWG (r = 0.811, P < 0.001). Bake mixing time (MT) was negatively correlated with qualitative characteristics including WG/DG (r = -0.444, P < 0.001) and PWWG (r = -0.428, P < 0.001), but positively with GI (r = 0.308, P < 0.01). Prediction equations were developed for bake WA and MT by stepwise multiple regression (SMR). The prediction model for bake MT included mixograph MT, flour protein content, and DGC as independent variables with a coefficient of determination (R²) of 0.775. For a model for bake WA, mixograph WA and DGC were used as independent variables with an R² of 0.772.

Rapid estimation of quality parameters of U.S. hard winter wheat breeding lines.

The USDA-ARS Hard Winter Wheat Quality Laboratory at the Grain Marketing and Production Research Center is the federal regional lab evaluating quality parameters of hard winter wheat progenies from federal, state, and private nurseries. For calibration development, we used an NIRSystems near-infrared reflectance (NIR) scanning spectrometer to construct a spectra database (400-2,498 nm) of over 8,000 wheats and meals each, and over 5,000 wheat flours from numerous nurseries. Equations developed from the flour spectra were better than those from wheat or meal spectra. Equations developed from a specific nursery gave much higher predictability than universal calibrations using all nurseries. For a federal nursery (Southern Regional Performance), the R² values for predicted values using calibration (based on flour spectra) versus lab data were > 0.95 for flour color and protein content; > 0.80 for flour ash content and mixograph absorption; > 0.75 for loaf volume; > 0.71 for bake absorption; and > 0.55 for bake mix time. NIR technology shows a high potential for rapid estimation of flour quality of early generation breeding lines.

Properties of commercial noodle flours.

Nineteen wheat flours were obtained from different sources. Three Australian standard white (ASW) flours segregated for Japanese salt noodles (Udon) were from Agriculture Western Australia. Two sets (7 and 3 each) of commercial Udon noodle flours, and two other sets (3 and 3 each) of commercial Chinese noodle flours were from two Asian mills. Higher protein contents were observed in Chinese noodle flours (11.3-11.6 %) than Udon noodle flours (8.5-10.4 %) and ASW flours (9.1-9.2 %). Fine particles (< 38 [micro]m) ranged from 38 to 48 % in Udon noodle flours, which was similar to ASW flours (35-43 %) but higher than in Chinese noodle flours (26-36 %). The swelling power (SP) of flours at 92.5 C, and their peak viscosities (Rapid Visco-Analyser) were measured. The average of flour SP was 19.4 (g/g) for ASW flours, 18.0 (g/g) for Udon noodle flours, and 15.8 (g/g) for Chinese noodle flours. The average peak viscosity (14.0 g solids/100 ml) was highest for ASW flours (204 rvu), followed by Udon noodle flours (185 rvu), and least for Chinese noodle flours (166 rvu). The average gluten index of Udon noodle flours was 92 %, which was similar to that of ASW flours (92 %) but higher than that of Chinese noodle sdfflours (88 %). SDS sedimentation volume ranged from 29 to 35sdddd dddddml for Chinese noodle flours, from 32 to 40 ml for Udon noodle flours, and from 43 to 45 mL for ASW flours.

Storage insects and grain odor.

This particular experiment was conducted using sorghum. However, similar effects would be observed with wheat. Lesser grain borer (Rhyzopertha dominica), red flour beetle (Tribolium castaneum), rice weevil (Sitophilus oryzae), saw-toothed grain beetle (Oryzaephilus surinamensis), and rusty grain beetle (Cryptolestes ferrugineus) were placed in sorghum (1 kg, 14 % moisture content) for 5, 7, and 10 weeks at 27 C. Infested samples were analyzed for insect numbers, frass, odor, and volatiles. Volatiles from whole grain at 60 C were collected on Tenax absorbent and thermally desorbed and analyzed by gas chromatography using infrared and mass detectors for component identification. Odor was assessed by sensory panels at the Federal Grain Inspection Service and in our own laboratory. Lesser grain borer caused severe off-odor, and red flour beetle caused some off-odor. The other three insects caused little or no objectionable odor, even though infestation was heavy. High concentrations of 2-pentanol and the known aggregation pheromones, dominicalure 1 and 2, were consistently present in samples infested with lesser grain borer. These compounds were caused only part of the odor from lesser grain borer. Several metabolites from lesser grain borer not previously reported were tentatively identified.

Electronic nose technology for identifying off-odors in grains.

Devices using arrays of coated organic polymer sensors were found to be capable of distinguishing pairs of samples with different odors as perceived by human panelists. The conductivity of each sensor changes in response to different volatiles, depending on molecular size, configuration, or polarity. Sensor response patterns were clearly different when viewed by subtraction or by Sammon mapping. Temperature and moisture of the samples had a major effect on the sensors, so it was necessary to calibrate the instrument using reference-air humidity and temperature similar to those of the samples. Groups of samples with similar odors could be distinguished from other groups by using artificial neural networks to classify sensor response patterns. Electronic nose devices have the potential to be more objective and consistent than human grain inspectors and should be more easily adapted to routine field use than conventional analytical chemical methods for identifying volatiles.

Digital imaging for bread crumb grain evaluation.

Computer vision techniques were applied to evaluate bread crumb in several studies at GMPRC to develop an objective automated method. Images of two commercial bread slices were acquired, and image texture features were extracted. A computer model comprised of digital image texture features was tested for ability to distinguish crumb grain differences between the two commercial bread brands. In a second bread study on commercially baked bread, image texture features and shape and size of the slices were used to distinguish different bread formulations. In the third study, a crumb grain evaluation algorithm was tested on bread baked at GMPRC from various wheat varieties. The computer vision method was applied to scanner images of pup loaf slices (100 g flour loaves). Correct recognition was computed for three bread categories based on expert scores (40 %), loaf features (96 %), and slice features (91 %). Preliminary data analysis showed potential for objective evaluation of bread crumb by image features or image descriptors.

Application of machine vision to pup loaf bread evaluation.

The intrinsic end-use quality of hard winter wheat breeding lines is routinely evaluated at the USDA-ARS GMPRC, Hard Winter Wheat Quality Laboratory. The ultimate test for evaluating hard wheat quality is the experimental baking of pup loaves. Computer vision was used to develop several objective methods for bread crumb evaluation for samples of the 1994 and 1995 hard winter wheat breeding lines. Computer-extracted features for bread crumb grain were studied, using subimages (`32 x 32' pixel) and features computed for the slices with different threshold settings. A subsampling (image) grid was located with respect to the axis of symmetry of a slice to provide identical topological sub-image information. Several ranking algorithms and data visualization techniques were employed to create a sensitive scale for porosity patterns of bread crumb. Significant linear correlations were found between the machine vision-extracted features and bread-making parameters. Crumb grain scores by human experts were correlated more highly with some image features than with other bread-making parameters.

Image texture analysis for discrimination of mill fractions of hard and soft wheat.

Digital image texture analysis was utilized to identify mill fractions from different mill streams and to assess wheat hardness differences. The study was conducted using a SRWW (Terra SR-87) and a HRWW (Thunderbird). Black and white images were acquired in a `256 x 256' pixel format to examine samples of coarse and fine mill fractions. Sixteen `64 x 64' pixel subimages per image were evaluated using texture analysis. Software was developed to calculate the image textural features used to develop the mill stream and hardness classification models. Several models based on image textural features were computed for different sets of subimages belonging to wheat of different hardness or mill stream. Recognition of `hard wheat vs. soft wheat' was achieved with 100 % correct recognition rate for each mill fraction when a three-feature model was used for pairwise analysis. Different mill fractions of the same wheat, `coarse vs. fine', were similarly discriminated with 100 % accuracy for each pairwise comparison. All four mill fractions were successfully recognized, with 100 % correct recognition rate when a three-feature model was used for four class analysis. Discrimination of wheat classes and mill fractions was achieved with less than 3 g (equivalent to 0.2 g per subimage) of material.

Wheat classification using image analysis and crush-force parameters.

A study was conducted to develop methodology to identify wheat class by combining image analysis techniques, wheat hardness, and other physical measures of single wheat kernels. Wheat kernel morphometrical parameters were extracted from digitized images, and hardness parameters and kernel weight were obtained from a single kernel wheat characterization system. Pattern recognition methods were applied to the data collected for six classes and 17 samples of soft and hard wheat. Class recognition rates based on combined image and physical (hardness and kernel weight) parameters were higher than those based on either the image or physical measures alone. Single kernel hard and soft recognition rates of 94 % were achieved when image and physical measures were combined. A version of the algorithm developed was implemented on a PC and tested with the same data to confirm practicality (speed and transferability) of PC implementation.

Computer vision for grain quality assessment.

The ultimate goal of computer-vision studies conducted at GMPRC was to train a machine and develop systems to automatically grade and classify grain and grain products. Recognition of soft and hard wheats was the objective of several studies. Recent studies on identification of wheat of different hardness showed a high degree of correct classification using a model based on kernel shape and size and kernel hardness scores from single kernel crushing evaluations. A study using wheat starch globule shape and size to identify hard and soft wheats also produced high recognition rates. A high degree of recognition was achieved for identification of foreign material in wheat, and the system has potential to identify different foreign material components.

Project summary on mixograph instrumentation and mixogram sensitivities.

An experimental 10-gram mixograph was constructed with the capabilities of variable mixing speed, forward or reverse planetary mixing pin rotation, acquisition of digital data at known planetary pin positions, and bowl platform moving or fixed at clockwise and counterclockwise 0, 7, and 14 degree positions. Moving and fixed bowl systems were compared and the effects of different fixed bowl positions evaluated. Mixograms obtained with the bowl fixed at a position clockwise from chart paper center were compared to those with the bowl fixed at a corresponding counterclockwise position and the reversed planetary pin rotation. Repeatability was studied as were the effects of variable absorption (44, 48, 54, 60, and 64 % for soft wheat flour, and 52, 56, 62, 68, and 72 % for hard wheat flour); temperature (67, 77, and 87 F); pin to bowl clearance (0, 1 and 2 mm); planetary input shaft speed (44, 88, 132, 176, and 250 rpm); and flour (5, 7.5, 10, 12.5, and 15 g) or dough load on mixogram characteristics. Recording mixograms in terms of torque rather than chart paper position facilitated calibration and standardization of the instrument and definition of the resulting data. Data analysis and reporting is underway.

Project summary on identifying quality characteristics of single wheat kernels using visible and NIR reflectance.

The existing single kernel wheat characterization system (SKWCS) was integrated with a fiber-optic near-infrared (NIR) diode-array system to measure color class, protein, and moisture of individual kernels. The integrated system collects spectra at a rate of one kernel per two seconds. The kernel is viewed as it rests in the weighing bucket of the SKWCS. When determining color class, kernel orientation can be a factor when working only in the visible spectrum (VIS), because the crease tends to be darker and dorsal side lighter for both red and white kernels. This effect is reduced in the NIR region. Incorrect classifications were reduced by about 50 % when using both VIS and NIR wavelengths in lieu of VIS wavelengths only. A classification accuracy of 98 % was achieved for red and white kernels when using the wavelength range of 450-1,688 nm. For protein predictions, an R² = 0.94 and SECV = 0.751 was achieved. Kernel weight had no correlation to protein (R² < 0.0005). When predicting single kernel moisture content, an R² of 0.96 and SECV of 0.305 was achieved. These results compare favorably to those achieved under more controlled conditions and manual kernel positioning.

Application of the single kernel wheat characterization technology to sorghum grain.

A single kernel wheat characterization system (SKWCS) was recently developed by the USDA-ARS Grain Marketing Research Laboratory and is currently being marketed by Perten Instruments North America, Inc. This device has been shown to accurately measure individual seed hardness, moisture, and size of wheat. The objective of this study was to determine if the SKWCS technology could be applied to the measurement of sorghum grain. Grains from 64 sorghum plots grown at Mead, NE in 1992 were characterized using a prototype SKWCS at the USDA-ARS Grain Marketing Research Laboratory. Problems encountered were primarily associated with the single kernel feeder mechanism. Occasionally, two sorghum seeds were fed to the crushing device instead of a single kernel. These double sampling events were easily detected by examination of the size data, and software limits could be set to exclude such double sampling events from the data set. Errors in hardness and size values also occurred if broken seeds were not removed prior to measurement of the grain. These errors could usually be detected by examination of the data, and eliminated by adjustment of software limits. Inspection and hand cleaning of samples is highly recommended prior to characterization. Based on our results, SKWCS technology can be successfully applied to sorghum seed.

Evaluation of a rotor-crescent design for sensing wheat kernel hardness.

Alternative designs of an experiential hardness tester were studied to evaluate its ability to classify single wheat kernels as hard or soft. Test variables included the effect of adding guide channels to control kernel orientation, five rotor speeds and five rotor tooth designs. The addition of guide channels to the crescent had the greatest benefit, increasing the CI from 1.2 to 1.6. Hence, we decided to keep the channels in place while testing rotor speed and tooth design. Data showed that the system was not sensitive to rotor speeds in the range between 15 and 480 rpm, with the classification index (CI) remaining about 1.6 for all but the highest speed. Because the system was designed to operate at 120 rpm, we decided to use only 120 rpm in the tooth design test. This study showed that decreasing the amount of material removed to fabricate a given tooth pattern increased the CI. Sawtooth and shallow patterns had a similar CI, close to 1.6. The sawtooth pattern was selected for two reasons; first, the hardness sensitivity was more uniform; second, the machining operation was the least complicated.

Effect of hopper angle on the flow rate of grain through orifices.

Three hopper bins with hopper angles of 40, 50, and 60 degrees from the horizontal and three different orifice sizes (17.8, 20.3, and 22.9 cm diameter) were used to evaluate the effect of hopper angle on the grain flow rate. Wheat, corn and sorghum at different moisture contents were tested for flow rates. The flow rates of wheat, corn and sorghum were significantly affected by hopper angle. An equation expressing grain flow rate as a function of hopper angle and orifice diameter was determined. Average flow rates from 40, 50, and 60 degree hoppers were 135, 140, and 150 m³/h, respectively, for wheat; 118, 125, and 134 m³/h, respectively, for corn; and 130, 134, and 144 m³/h, respectively, for sorghum. For the same size orifice and grain, the volume flow rate from a 60 degree hopper was 11-14 % higher than that from a 40 degree hopper. For the same size of opening and same hopper angle, the flow rate of corn was lower than those of wheat and sorghum.

Review of research on reducing grain dust emission.

Control of grain dust emission during bulk handling is needed to improve air quality and reduce dust explosion hazards. Dust control during handling can be achieved by collecting emitted dust with air cleaning systems or by preventing dust from being emitted, thus containing it within the grain stream. Chang et al. (1985) developed a grain-flow regulator that maintains a mass flow or 'choke' flow state and, thus, reduces grain dust emission. The grain-flow regulator and an 18-cm orifice installed at the end of a 25-cm diameter vertical spout were tested using shelled corn to determine dust emission during loading of grain into a bin. The 25-cm diameter vertical spout with no flow control device installed was also tested. When only the flow regulator, orifice, and spout were used to fill the bin, the average measured dust concentrations were 0.43, 0.57, and 2.57 gram per cubic meter, respectively. Lai et al. (1981) treated wheat and corn with mineral oil, soybean oil and water. The treated grains were transferred through a small grain elevator right after treatment and after 1, 3, 6, and 12 months of storage to evaluate the effectiveness of these liquid additives on the control of dust emission. Treatment with mineral and soybean oils was effective in reducing dust emissions from grain and remained effective for treated grain stored for as long as 12 months. Lai et al. (1982 and 1986) developed procedures and methodology to evaluate the effectiveness of liquid additives on grain dust emission in commercial elevators. Commercial elevator tests confirmed that mineral and soybean oil were effective in suppressing the emission of fine dust from corn, wheat and soybeans.

a-Amylase inhibitors from wheat: amino acid sequences and patterns of inhibition of insect a-amylases.

Four a-amylase inhibitors (WRP24, WRP25, WRP26, and WRP27) were purified from wheat flour by preparative, reversed-phase high performance liquid chromatography. All have polypeptide molecular masses of about 14 kDa and are members of the cereal superfamily of protease and amylase inhibitors. Sedimentation velocity analysis indicated that WRP25 and WRP27 are monomeric proteins, whereas WRP24 is a dimer. WRP24 is identical in N-terminal amino acid sequence to the well characterized 0.19 dimeric inhibitor from wheat kernels. WRP25 and WRP26 differ in sequence from each other at only three positions and represent previously unseparated forms of the 0.28 wheat inhibitor. WRP27 is a previously uncharacterized inhibitor and is more similar in sequence to the 0.28 inhibitor than to the 0.19 inhibitor. WRP25 and WRP26 inhibited a-amylase from the rice weevil, red flour beetle, and the yellow meal worm but did not inhibit human salivary-amylase. WRP24 inhibited the human and the insect a-amylase but inhibited one of the two rice weevil a-amylase much more strongly than the other. WRP27 was notable in that it strongly inhibited only the rice weevil a-amylase. We observed that the growth rate of red flour beetle larvae was slowed when purified WRP24 was included in the diet at a level of 10 %. Addition of WRP24 to corn starch resulted in greater weight loss of red flour beetle adults than occurred on control diets. Our results support the hypothesis that these a-amylase inhibitors provide wheat seeds with a selective evolutionary advantage, because they can slow the growth of insect pests that attack cereal grains.

Biotechnological approaches for stored-product insect control.

Transgenic plants that contain novel proteins detrimental to critical life functions in insects are becoming more significant in the development of pest management strategies. The pace of development of these genetically modified plants is being hastened not only by technical and scientific advances in molecular biology and genetic engineering, but also by the growing urgency for environmentally safe alternatives to synthetic chemical pesticides. Chemical defense proteins (CDPs) are unique, bioactive proteins that, when engineered into plants, offer one avenue of pest management that can be highly selective for the pest and more friendly to nontarget organisms and the environment than conventional pesticides. With the development of tissue-specific promoter systems, the potential for these CDPs to impact population growth of stored-product insects that feed on cereal and legume seeds as well as on their products has dramatically improved. There is a variety of CDPs that interact with different physiological and morphological systems in insects including the endocrine, nervous, skeletal, and digestive systems. CDPs currently being evaluated that target these systems include juvenile hormone esterases, insect-specific neurotoxins, vitamin-binding proteins, lectins, chitinases, proteinase inhibitors, and a-amylase inhibitors. Evidence from several laboratories indicates that engineering plants with multiple CDP genes offers the best hope for developing significant levels of resistance against stored-product insects. Although a polygenic approach to insect management in cereal and legume seeds will slow the adaptation process of these insect species to CDPs, stored-product insects will probably adapt to these transgenic crops, and strategies to manage or slow the adaptation process will need to be formulated.
go to next document