Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Comparison of FT-NIR and NIR for evaluation of phyisco-chemical properties of stored wheat grains

Comparison of FT-NIR and NIR for evaluation of phyisco-chemical properties of stored wheat grains Objectives: Storage studies were carried out in wheat grains with different moisture contents, level of infestation, and storage days. Material and Methods: Wheat grain samples were infested with Rhyzopertha dominica and stored for up to 90 days under ambient conditions. Every 45 days, samples of wheat were collected and evaluated for protein, fat, ash, 1000 kernel weight, and hardness. Results: The physicochemical parameters, namely, protein, 1000 kernel weight, and hardness decreased while fat and ash content increased with the storage. Methodology for identification of infested samples was developed in Fourier transform near infrared (FT-NIR) and near infrared (NIR) using infested wheat and control samples. The linear regression plots for different quality parameters gave an R value of 82.04% and 97.15% via FT-NIR and 81.61% and 98.07% via NIR. The RMSEP values by NIR were in the range of 0.03 to 0.7, whereas the RMSECV values of FT-NIR were in the range of 0.03 to 1.2. Conclusions: Both the models performed well for the cross validation studies; hence, they can be used in future for the rapid assessment of storage quality of wheat grains. Key words: wheat grain; storage; FT-NIR; NIR; infested. was shown in terms of the rheological properties as dough stability, Introduction water absorption, and time to break down. Storage is an important part of a food grain processing chain. Storage of Consumption of cereals and legumes by pests such as insects dur- raw and processed wheat has been used by humans since the beginning ing storage and microbial spoilage or contamination may make these of history as a prerequisite for ensuring food security due to off time totally inedible (Basavaraja et al., 2007). In past few years, many dif- availability, and for withholding seed grain for long periods. Storability ferent methods of analysis had been employed to determine quality can be defined as the ability of grain to maintain their sensory charac- parameters in wheat. On farms, manual samples, traps, and probes teristics, nutritional value, and physico-chemical properties (Iconomou have been used to determine the presence of insects. Manual inspec- et  al., 2006; Hruskova and Machova, 2002). The stored grains are tion, sieving, cracking-floatation, and Berlese funnels are being used at subjected to infestation which severely affects the quality of the stored present to detect insects in grain-handling facilities. These methods are grains and making it unfit for human consumption (Bulla et al., 1978). not efficient and are time consuming. Acoustic detection, carbon diox- Crude fat and total sugars increased in infested grains while ide measurement, uric acid measurement, near-infrared spectroscopy, ash content and crude fibre decreased (Boyacioǧlu et  al., 1995) in and soft x-ray method have the potential for use at the industry level contaminated wheat grains. Edde et  al., (2005) studied the effect to detect insects in grain samples as their usefulness has been demon- of stored pests on the quality of commercial wheat. Pests reduced strated in the research laboratories (Dong et al., 2000; Karunakarana the quality of wheat by reducing water content and hectoliter mass. et al., 2004; Koljonen et al., 2008; Neerthirajan et al., 2005). The lower quality of flour obtained by milling of infected wheat © The Author(s) 2018. Published by Oxford University Press on behalf of Zhejiang University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 166 P. Pandey et al., 2018, Vol. 2, No. 3 Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 Table 1. Storage effects on physical and chemical properties of wheat grain Variables Parameters M.C (%) Days Level of Infestation*(%) Moisture*(%) Fat*(%) Protein*(%) Ash*(%) 1000 kernel wt** (g) Hardness*** (kg) 10 0 0 10.12 ± 0.14 1.28 ± 0.08 9.89 ± 0.03 1.63 ± 0.11 41.85 ± 0.12 15.21 ± 0.09 45 0 9.01 ± 0.11 1.34 ± 0.09 9.745 ± 0.04 1.57 ± 0.05 41.52 ± 0.07 15.46 ± 0.12 5 9.08 ± 0.16 1.34 ± 0.14 9.449 ± 0.02 1.64 ± 0.16 40.14 ± 0.11 12.77 ± 0.13 10 9.55 ± 0.22 1.41 ± 0.09 9.433 ± 0.04 1.65 ± 0.18 39.21 ± 0.08 10.36 ± 0.15 15 8.75 ± 0.31 1.43 ± 0.08 9.404 ± 0.03 1.63 ± 0.12 38.63 ± 0.1 7.62 ± 0.19 90 0 9.42 ± 0.2 1.23 ± 0.25 9.631 ± 0.02 1.57 ± 0.23 41.45 ± 0.21 13.81 ± 0.22 5 9.99 ± 0.12 1.25 ± 0.23 9.448 ± 0.06 1.58 ± 0.12 40.53 ± 0.14 9.02 ± 0.16 10 9.37 ± 0.14 1.31 ± 0.12 9.396 ± 0.01 1.58 ± 0.14 39.35 ± 0.13 7.43 ± 0.13 15 9.04 ± 0.14 1.23 ± 0.07 9.372 ± 0.01 1.6 ± 0.14 38.89 ± 0.12 6.47 ± 0.1 12 0 0 12.23 ± 0.05 1.29 ± 0.04 10.631 ± 0.14 1.52 ± 0.19 42.68 ± 0.11 14.35 ± 0.04 45 0 8.9 ± 0.13 1.22 ± 0.14 9.892 ± 0.06 1.55 ± 0.09 42.66 ± 0.11 13.85 ± 0.13 5 9.38 ± 0.05 1.253 ± 0.02 9.438 ± 0.02 1.6 ± 0.24 41.4 ± 0.13 10.7 ± 0.03 10 9.42 ± 0.11 1.31 ± 0.13 9.436 ± 0.02 1.56 ± 0.11 40.4 ± 0.11 7.91 ± 0.12 15 8.83 ± 0.14 1.295 ± 0.14 9.505 ± 0.1 1.57 ± 0.05 40.21 ± 0.09 7.28 ± 0.14 90 0 8.88 ± 0.14 1.42 ± 0.13 9.745 ± 0.03 1.63 ± 0.14 40.83 ± 0.13 14.1 ± 0.13 5 9.79 ± 0.12 1.417 ± 0.08 9.583 ± 0.03 1.63 ± 0.12 38.75 ± 0.11 8.03 ± 0.1 10 9.06 ± 0.25 1.4 ± 0.12 9.445 ± 0.02 1.58 ± 0.25 38.41 ± 0.21 7.03 ± 0.18 15 10.18 ± 0.01 1.427 ± 0.09 9.427 ± 0.01 1.58 ± 0.17 36.6 ± 0.11 6.58 ± 0.05 1 4 0 0 13.9 ± 0.19 1.20 ± 0.33 10.024 ± 0.2 1.47 ± 0.08 42.85 ± 0.17 13.48 ± 0.26 45 0 8.1 ± 0.21 1.24 ± 0.16 9.745 ± 0.07 1.55 ± 0.07 41.64 ± 0.12 13.19 ± 0.18 5 8.29 ± 0.38 1.227 ± 0.12 9.464 ± 0.02 1.58 ± 0.02 40.23 ± 0.13 11.25 ± 0.25 10 9.54 ± 0.18 1.26 ± 0.2 9.374 ± 0.01 1.63 ± 0.2 39.52 ± 0.19 10.56 ± 0.19 15 9.33 ± 0.22 1.28 ± 0.14 9.087 ± 0.01 1.63 ± 0.14 39.24 ± 0.16 9.67 ± 0.18 90 0 8.18 ± 0.1 1.24 ± 0.08 9.431 ± 0.02 1.65 ± 0.08 42.52 ± 0.08 14.25 ± 0.09 5 8.16 ± 0.15 1.23 ± 0.1 9.429 ± 0.03 1.6 ± 0.1 40.28 ± 0.11 10.83 ± 0.12 10 7.99 ± 0.13 1.273 ± 0.19 9.398 ± 0.01 1.65 ± 0.11 39.6 ± 0.13 10.25 ± 0.16 15 8.4 ± 0.19 1.273 ± 0.3 9.263 ± 0.01 1.63 ± 0.09 38.26 ± 0.16 9.39 ± 0.24 *Mean values ± SD of three replications. **Mean values ± SD of five replications. ***Mean values ± SD of ten replications. Evaluation of stored wheat grains, 2018, Vol. 2, No. 3 167 Keeping the above points in mind, the techniques employed Clean wheat samples infested with R. dominica in triplicate; each in our study are FT-NIR and NIR spectroscopy, as rapid analysis infestation level of 0, 5, 10, and 15 insects per sample at 10%, 12%, methods for qualitative and quantitative measurements of wheat and 14% moisture content were stored in plastic jars with a hole quality. provided on top of jars, covered with net. The samples were stored under ambient conditions (28 ± 2°C and 65 ± 5% RH). Infested and control samples were analysed after 0, 45, and 90 days of storage. Materials and Methods Methodology for identification of infested samples Sample preparation FT-NIR for identification of infested samples Experimental culture, lesser grain borer Rhyzopertha dominica TM FT-NIR MPA spectometer make Bruker Optics (Germany) (F.) were obtained from the Entomology Department, College of combined with OPUS IDENT software was used to identify Agriculture, G.  B. Pant University of Agriculture and Technology, substances by their IR spectra. An unknown spectrum was directly Pantnagar. Insect rearing was performed in jars with screen lids at compared with reference spectra of a library. IDENT identifies those RH 65%–70% and temp 25ºC. (Tilley et al., 2014). Figure  1. Effect of level of infestation and number of storage days on (a) protein, (b) hardness, (c) 1000 kernel wt, (d) ash, (e) fat, and (f) moisture in wheat samples. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 168 P. Pandey et al., 2018, Vol. 2, No. 3 reference spectra which are closest equivalent to the test spectrum Storage studies and determines the deviations between these spectra and the test The storage studies were performed for 27 wheat samples for 0, 45, spectrum. This allows IDENT to identify unknown substances, e.g. and 90 days under storage conditions (65% RH and 25°C) in plastic polymers, and to evaluate the conformity degree of a substance with containers with 0%, 5%, 10%, and 15% level of infestation, respect- a reference standard. ively. Changes in wheat quality were analysed and shown in Table 1. The protein, 1000 kernel weight, and hardness decrease as the level of infestation and number of storage days increased while fat and NIR for identification of infested samples ash content increased (Manley et al., 2002; Pearson et al., 2003). In Zeutec Spectraanalyser (double beam NIR) (Germany) spectometer Changes in protein, 1000 kernel weight, hardness, fat, ash, and mois- (1200–2600 nm), a reference library was prepared using Application TM ture are shown in Figure 1. worx software with multiple linear regression (MLR) model, and then unknown samples were fed in rotating cup drawer and analysed. The scanning was done everytime with all the NIR-filters included. In Quality evaluation of unknown wheat sample a calibration, a correlation between the NIR absorbance values at the Twenty-seven known wheat spectra were analysed using FT-NIR different wavelengths and the sample composition is established, and and NIR. Spectra of the same are shown in Figures  2 and 3. In calibration constants were calculated for each wavelength. When the FT-NIR, ident method is developed using absorption spectra of desired products and properties have been set up and the calibration stored wheat samples which is used as reference spectra for deter- coefficients have been entered, samples can be analysed with a predic- mination of unknown wheat samples. The analysis compares the tion calculation of the property results. test spectrum with all the reference spectra. Spectra of stored wheat grains were assigned in group according to the level of infestation Methodology for calculation of physical and and number of storage days, and then compared with spectra of chemical properties of wheat grains fresh wheat samples using ident test method with vector normal- The fresh and infested wheat samples were subjected to analysis of ization pre-processing method. All groups were identified uniquely. various physical and properties of grains such moisture, protein, ash, Figure  4 shows FT-NIR shows cross-validation results of stored fat, 1000 kernel wt, and hardness with changes in number of days wheat grains. In FT-NIR, the PLS regression showed a perfect fit of along with perectage level of insect infestation with the standard AOAC the model by higher values of R and lower RMSECV and RMSEE methods (AOAC, 2005; Mishra et al., 2018; Neethirajan et al., 2005). values (Table 2). In Zeutec NIR, the unknown wheat samples were compared with reference library and analysed with a prediction cal- Results and Discussion culation of the property results. Figure 5 shows NIR cross-validation Forty fresh samples and 60 infested samples of different wheat vari- results using Zeutec NIR. The best number of factor was determined ties were analysed with FT-NIR and NIR. by the lowest RMSEP value and highest value of R . The RMSEP Figure 2. FT-NIR absorption spectra of stored wheat samples used for identification. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 Evaluation of stored wheat grains, 2018, Vol. 2, No. 3 169 Figure 3. Zeutec-NIR absorption spectra of wheat samples used for identification. Table 2. Analytical features of the different regions and preprocessing methods for calibration and validation models in FT-NIR Parameters Wave number region (cm-1) Preprocessing method PLS factors Validation Calibration 2 2 RMSECV R RMSEE R (%) (%) (%) (%) Moisture 12489.4–7498.3and 5454-4242.8 Min Max normalization 10 0.731 95.26 0.601 97.15 Protein 12489.4 – 5446.3 and 4605.4-4242.8 First derivate 7 0.242 91.79 0.179 95.84 Ash 12489.4–7498.3, 6102- 5446.3 and First derivate + MSC 4 0.0392 83.18 0.0341 87.86 4605.4 – 4242.8 Fat 12489.4–7498.3, and 6102 – 5168.6 Straight line subtraction 8 0.0366 82.75 0.0287 90.45 1000 Kernel wt. 12489.4–7498.3 and 4605.4 – 4242.8 First derivate + MSC 6 1.00 90.30 0.748 95.01 Hardness 12489.4 – 4242.8 First derivate + straight line 2 1.23 80.78 1.21 82.02 subtraction 2 2 and R values for moisture, protein, fat, ash, 1000 kernal weight, 1.0, and 1.3 and maximum correlation coefficient (R ) 0.95, 0.91, and hardness determination with different spectral pre-processing 0.83,0.82, 0.90, and 0.80, respectively, were obtained (Sanchez- methods are presented in Table 3. During the storage studies of con- Marinez et al., 1997; Singh et al., 2008; Srivastava et al., 2018). trol and infested wheat samples, protein, hardness, and 1000 kernel weights decreased with an increase in days of storage and level of Conclusion infestation while the fat and ash increased. Moisture content is important for shelf-life and storage. Very high Infested and fresh wheat samples can be identified efficiently by moisture content (greater than 14.5%) attracts mold, bacteria, and the method developed in both FT-NIR and NIR (Armstrong et al., insects, all of which can result in storage issues or baking quality 2006; Amir et al., 2013; Salgo et al., 2012). In NIR, the reference deterioration. These parameters are commonly quantified using library for moisture, protein, ash, fat, 1000 kernel weight, and hard- either primary methods or FT-NIR or NIR spectroscopy. Replacing ness with lowest RMSEC values 0.62, 0.11, 0.025, 0.019, 0.81, and the primary methods with FT-NIR or NIR will provide faster results 0.56 and maximum correlation coefficient (R ) 0.86, 0.91,0.81, and accurate quantification, guaranteeing flour meets specifications. 0.96, 0.85, and 0.98, respectively, was obtained. In FT-NIR, the val- FT-NIR and NIR are an excellent means to visualize the chem- ues of validation for moisture, protein, ash, fat, 1000 kernel weight, ical composition of different wheat varieties with being very quick, and hardness with lowest RMSECV values 0.7, 0.2, 0.03, 0.03, Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 170 P. Pandey et al., 2018, Vol. 2, No. 3 Figure 4. Linear regression plot of measured vs. predicted values of (a) moisture, (b) protein, (c) ash, (d) fat, (e) hardness, and (f) 1000 kernel weight for cross- validation data set using FT-NIR. reliable, and cheaper analytical technique which can effectively be collecting a scan every second has minimum sample preparation. used for estimation of different wheat quality parameters. The values Both FT-NIR and NIR can be used for the rapid detection of quality determined by this technique are very close to the values determined of wheat grains. NIR is best suited if one has to deal with primary by the standard procedures and preferred over dispersive chemical bonds and simpler molecules/compounds but if the structure is a means due to following reasons: non-destructive technique provides complex one then FT-NIR suits better than NIR as it focuses on the precise measurement with no external calibration, increased speed, overtones and stretching frequency. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 Evaluation of stored wheat grains, 2018, Vol. 2, No. 3 171 Figure 5. Linear regression plot of (a) moisture, (b) protein, (c) ash, (d) fat, (e) hardness, and (f) 1000 kernel weight of stored wheat samples for cross-validated data set using Zeutec NIR. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 172 P. Pandey et al., 2018, Vol. 2, No. 3 Table  3. MLR validation methods developed for the parameters Dong, J., van de Voort, F. R., Ismail, A., Akochi-Koble, E., Pinchuk, D. (2000). using Zeutec NIR Rapid determination of the carboxylic acid contribution to the total acid number of lubricants by Fourier transform infrared spectroscopy. Lubr Parameters MLR factors RMSEC (%) RMSEP (%) R Eng, 56: 12–17. (%) Edde, P. A., Phillips, T. W., Toews, M. D. (2005). Responses of Rhyzopertha dominica (Coleoptera: Bostrichidae) to its aggregation pheromones Moisture 2 0.62 0.70 86.41 as influenced by trap design, trap height, and habitat. Environmental Protein 5 0.11 0.14 91.97 Entomology, 34: 1549–1557. Ash 5 0.025 0.031 81.61 Hruskova, M., Machova, D. (2002). Changes of wheat flour properties during Fat 7 0.0197 0.030 96.09 short term storage. Czech Journal of Food Sciences, 20: 125–130. 1000 Kernel wt 1 0.81 0.87 85.36 Iconomou, D., Athanasopoulos, P., Arapoglou, D., Varzakas, T., Christopoulou, Hardness 5 0.567 0.776 98.07 N. (2006). Cereal quality characteristics as affected by controlled atmospheric storage conditions. American Journal of Food Technology, 1: 149–157. Karunakarana, C., Jayasa, D. S., White, N. D. (2004). Detection of internal Funding wheat seed infestation by Rhyzopertha dominica using X-ray imaging. No funding was received for the study. Journal of Stored Products Research, 40: 507–516. Koljonen, J., Nordling, T., Alander, J. (2008). A review of genetic algorithms in near infrared spectroscopy and chemometrics: past and future. Journal of Acknowledgements Infrared spectroscopy, 16: 189–197. Manley, M., Zyl, L. V., Osborne, B. G. (2002). Using Fourier transform near infra- Conflict of interest statement. All the authors declare that there is no conflict red spectroscopy in determining kernel hardness, protein, and moisture con- of interest. tent of whole wheat flour. Journal of Near Infrared Spectroscopy, 10: 71–76. Mishra, G., Srivastava, S., Panda, B. K., Mishra, H. N. (2018). Rapid assessment References of quality change and insect infestation in stored wheat grain using FT-NIR spectroscopy and chemometrics. Food Analytical Methods, 11: 1189–1198. Amir, R. M., Anjum, F. M., Khan, M. I., Khan, M. R., Pasha, I., Nadeem, Neethirajan, S., Karunakaran, C., Jayas, D. S., White, N. D.  G. (2005). M. (2013). Application of Fourier transform infrared (FTIR) spectros- Detection techniques for stored-product insects in grain. Food Control, copy for the identification of wheat varieties. Journal of Food Science and 18: 157–162. Technology, 50: 1018–1023. Pearson, T. C., Brabec, D. L., Schwartz, C. R. (2003). Automated detection of AOAC. (2005). Officials Methods of Analysis. 18th ed. Association of Officials internal insects infestations in whole wheat kernels using a Perten SKCS Analytical Chemists, Washington, DC. 4100. Applied Engineering in Agriculture, 19: 727–733. Armstrong, P. R., Maghirang, E. B., Xie, F., Dowell E. (2006). Comparison Salgo, A., Gergely, S. (2012). Analysis of wheat grain development using NIR of dispersive and Fourier-transform NIR instruments for measuring grain spectroscopy. Journal of Cereal Science, 56: 31–38. and flour attributes. American Society of Agricultural and Biological Sanchez-Marinez, R. I., Cortez-Rocha, M. O., Ortega-Dorame, F., Silveira, M. Engineers, 22: 453–457. ISSN 0883−8542. I. (1997). End-use quality of flour from Rhyzopertha dominica infested Basavaraja, H., Mahajanashetti, S. B., Udagatti, N. C. (2007). Economic wheat. Cereal Chemistry, 74: 481–483. analysis of post –harvest losses in food grains in India: a case study of Singh, C. B., Jayas, D. S., Paliwal, J., White, N. D.  G. (2008). Detection of Karnataka. Agricultural Economics Research Review, 20: 117–126. insect-damaged wheat kernels using near-infrared hyperspectral imaging. Bulla Jr, L. A., Kramer, K. J., Speirs, R. D. (1978). Insects and microorgan- Journal of Stored Products Research, 45: 151–158. isms in stored grain and their control. Advances in cereal science and Srivastava, S., Mishra, G., Mishra, H. N. (2018). FTNIR-A Robust Diagnostic technology. Tool for the Rapid Detection of Rhyzopertha dominica and Sitophilus Boyacioǧlu, D., Hettiarachchy, N. S. (1995). Changes in some biochemical oryzae Infestation and Quality Changes in Stored Rice Grains. Food and components of wheat grain that was infected with Fusarium gramine- Bioprocess Technology, 1–12. arum. Journal of Cereal Science, 21: 57–62. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Food Quality and Safety Oxford University Press

Comparison of FT-NIR and NIR for evaluation of phyisco-chemical properties of stored wheat grains

Food Quality and Safety , Volume 2 (3) – Sep 1, 2018

Loading next page...
 
/lp/oxford-university-press/comparison-of-ft-nir-and-nir-for-evaluation-of-phyisco-chemical-i4UV8VNFnm
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of Zhejiang University Press.
ISSN
2399-1399
eISSN
2399-1402
DOI
10.1093/fqsafe/fyy015
Publisher site
See Article on Publisher Site

Abstract

Objectives: Storage studies were carried out in wheat grains with different moisture contents, level of infestation, and storage days. Material and Methods: Wheat grain samples were infested with Rhyzopertha dominica and stored for up to 90 days under ambient conditions. Every 45 days, samples of wheat were collected and evaluated for protein, fat, ash, 1000 kernel weight, and hardness. Results: The physicochemical parameters, namely, protein, 1000 kernel weight, and hardness decreased while fat and ash content increased with the storage. Methodology for identification of infested samples was developed in Fourier transform near infrared (FT-NIR) and near infrared (NIR) using infested wheat and control samples. The linear regression plots for different quality parameters gave an R value of 82.04% and 97.15% via FT-NIR and 81.61% and 98.07% via NIR. The RMSEP values by NIR were in the range of 0.03 to 0.7, whereas the RMSECV values of FT-NIR were in the range of 0.03 to 1.2. Conclusions: Both the models performed well for the cross validation studies; hence, they can be used in future for the rapid assessment of storage quality of wheat grains. Key words: wheat grain; storage; FT-NIR; NIR; infested. was shown in terms of the rheological properties as dough stability, Introduction water absorption, and time to break down. Storage is an important part of a food grain processing chain. Storage of Consumption of cereals and legumes by pests such as insects dur- raw and processed wheat has been used by humans since the beginning ing storage and microbial spoilage or contamination may make these of history as a prerequisite for ensuring food security due to off time totally inedible (Basavaraja et al., 2007). In past few years, many dif- availability, and for withholding seed grain for long periods. Storability ferent methods of analysis had been employed to determine quality can be defined as the ability of grain to maintain their sensory charac- parameters in wheat. On farms, manual samples, traps, and probes teristics, nutritional value, and physico-chemical properties (Iconomou have been used to determine the presence of insects. Manual inspec- et  al., 2006; Hruskova and Machova, 2002). The stored grains are tion, sieving, cracking-floatation, and Berlese funnels are being used at subjected to infestation which severely affects the quality of the stored present to detect insects in grain-handling facilities. These methods are grains and making it unfit for human consumption (Bulla et al., 1978). not efficient and are time consuming. Acoustic detection, carbon diox- Crude fat and total sugars increased in infested grains while ide measurement, uric acid measurement, near-infrared spectroscopy, ash content and crude fibre decreased (Boyacioǧlu et  al., 1995) in and soft x-ray method have the potential for use at the industry level contaminated wheat grains. Edde et  al., (2005) studied the effect to detect insects in grain samples as their usefulness has been demon- of stored pests on the quality of commercial wheat. Pests reduced strated in the research laboratories (Dong et al., 2000; Karunakarana the quality of wheat by reducing water content and hectoliter mass. et al., 2004; Koljonen et al., 2008; Neerthirajan et al., 2005). The lower quality of flour obtained by milling of infected wheat © The Author(s) 2018. Published by Oxford University Press on behalf of Zhejiang University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 166 P. Pandey et al., 2018, Vol. 2, No. 3 Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 Table 1. Storage effects on physical and chemical properties of wheat grain Variables Parameters M.C (%) Days Level of Infestation*(%) Moisture*(%) Fat*(%) Protein*(%) Ash*(%) 1000 kernel wt** (g) Hardness*** (kg) 10 0 0 10.12 ± 0.14 1.28 ± 0.08 9.89 ± 0.03 1.63 ± 0.11 41.85 ± 0.12 15.21 ± 0.09 45 0 9.01 ± 0.11 1.34 ± 0.09 9.745 ± 0.04 1.57 ± 0.05 41.52 ± 0.07 15.46 ± 0.12 5 9.08 ± 0.16 1.34 ± 0.14 9.449 ± 0.02 1.64 ± 0.16 40.14 ± 0.11 12.77 ± 0.13 10 9.55 ± 0.22 1.41 ± 0.09 9.433 ± 0.04 1.65 ± 0.18 39.21 ± 0.08 10.36 ± 0.15 15 8.75 ± 0.31 1.43 ± 0.08 9.404 ± 0.03 1.63 ± 0.12 38.63 ± 0.1 7.62 ± 0.19 90 0 9.42 ± 0.2 1.23 ± 0.25 9.631 ± 0.02 1.57 ± 0.23 41.45 ± 0.21 13.81 ± 0.22 5 9.99 ± 0.12 1.25 ± 0.23 9.448 ± 0.06 1.58 ± 0.12 40.53 ± 0.14 9.02 ± 0.16 10 9.37 ± 0.14 1.31 ± 0.12 9.396 ± 0.01 1.58 ± 0.14 39.35 ± 0.13 7.43 ± 0.13 15 9.04 ± 0.14 1.23 ± 0.07 9.372 ± 0.01 1.6 ± 0.14 38.89 ± 0.12 6.47 ± 0.1 12 0 0 12.23 ± 0.05 1.29 ± 0.04 10.631 ± 0.14 1.52 ± 0.19 42.68 ± 0.11 14.35 ± 0.04 45 0 8.9 ± 0.13 1.22 ± 0.14 9.892 ± 0.06 1.55 ± 0.09 42.66 ± 0.11 13.85 ± 0.13 5 9.38 ± 0.05 1.253 ± 0.02 9.438 ± 0.02 1.6 ± 0.24 41.4 ± 0.13 10.7 ± 0.03 10 9.42 ± 0.11 1.31 ± 0.13 9.436 ± 0.02 1.56 ± 0.11 40.4 ± 0.11 7.91 ± 0.12 15 8.83 ± 0.14 1.295 ± 0.14 9.505 ± 0.1 1.57 ± 0.05 40.21 ± 0.09 7.28 ± 0.14 90 0 8.88 ± 0.14 1.42 ± 0.13 9.745 ± 0.03 1.63 ± 0.14 40.83 ± 0.13 14.1 ± 0.13 5 9.79 ± 0.12 1.417 ± 0.08 9.583 ± 0.03 1.63 ± 0.12 38.75 ± 0.11 8.03 ± 0.1 10 9.06 ± 0.25 1.4 ± 0.12 9.445 ± 0.02 1.58 ± 0.25 38.41 ± 0.21 7.03 ± 0.18 15 10.18 ± 0.01 1.427 ± 0.09 9.427 ± 0.01 1.58 ± 0.17 36.6 ± 0.11 6.58 ± 0.05 1 4 0 0 13.9 ± 0.19 1.20 ± 0.33 10.024 ± 0.2 1.47 ± 0.08 42.85 ± 0.17 13.48 ± 0.26 45 0 8.1 ± 0.21 1.24 ± 0.16 9.745 ± 0.07 1.55 ± 0.07 41.64 ± 0.12 13.19 ± 0.18 5 8.29 ± 0.38 1.227 ± 0.12 9.464 ± 0.02 1.58 ± 0.02 40.23 ± 0.13 11.25 ± 0.25 10 9.54 ± 0.18 1.26 ± 0.2 9.374 ± 0.01 1.63 ± 0.2 39.52 ± 0.19 10.56 ± 0.19 15 9.33 ± 0.22 1.28 ± 0.14 9.087 ± 0.01 1.63 ± 0.14 39.24 ± 0.16 9.67 ± 0.18 90 0 8.18 ± 0.1 1.24 ± 0.08 9.431 ± 0.02 1.65 ± 0.08 42.52 ± 0.08 14.25 ± 0.09 5 8.16 ± 0.15 1.23 ± 0.1 9.429 ± 0.03 1.6 ± 0.1 40.28 ± 0.11 10.83 ± 0.12 10 7.99 ± 0.13 1.273 ± 0.19 9.398 ± 0.01 1.65 ± 0.11 39.6 ± 0.13 10.25 ± 0.16 15 8.4 ± 0.19 1.273 ± 0.3 9.263 ± 0.01 1.63 ± 0.09 38.26 ± 0.16 9.39 ± 0.24 *Mean values ± SD of three replications. **Mean values ± SD of five replications. ***Mean values ± SD of ten replications. Evaluation of stored wheat grains, 2018, Vol. 2, No. 3 167 Keeping the above points in mind, the techniques employed Clean wheat samples infested with R. dominica in triplicate; each in our study are FT-NIR and NIR spectroscopy, as rapid analysis infestation level of 0, 5, 10, and 15 insects per sample at 10%, 12%, methods for qualitative and quantitative measurements of wheat and 14% moisture content were stored in plastic jars with a hole quality. provided on top of jars, covered with net. The samples were stored under ambient conditions (28 ± 2°C and 65 ± 5% RH). Infested and control samples were analysed after 0, 45, and 90 days of storage. Materials and Methods Methodology for identification of infested samples Sample preparation FT-NIR for identification of infested samples Experimental culture, lesser grain borer Rhyzopertha dominica TM FT-NIR MPA spectometer make Bruker Optics (Germany) (F.) were obtained from the Entomology Department, College of combined with OPUS IDENT software was used to identify Agriculture, G.  B. Pant University of Agriculture and Technology, substances by their IR spectra. An unknown spectrum was directly Pantnagar. Insect rearing was performed in jars with screen lids at compared with reference spectra of a library. IDENT identifies those RH 65%–70% and temp 25ºC. (Tilley et al., 2014). Figure  1. Effect of level of infestation and number of storage days on (a) protein, (b) hardness, (c) 1000 kernel wt, (d) ash, (e) fat, and (f) moisture in wheat samples. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 168 P. Pandey et al., 2018, Vol. 2, No. 3 reference spectra which are closest equivalent to the test spectrum Storage studies and determines the deviations between these spectra and the test The storage studies were performed for 27 wheat samples for 0, 45, spectrum. This allows IDENT to identify unknown substances, e.g. and 90 days under storage conditions (65% RH and 25°C) in plastic polymers, and to evaluate the conformity degree of a substance with containers with 0%, 5%, 10%, and 15% level of infestation, respect- a reference standard. ively. Changes in wheat quality were analysed and shown in Table 1. The protein, 1000 kernel weight, and hardness decrease as the level of infestation and number of storage days increased while fat and NIR for identification of infested samples ash content increased (Manley et al., 2002; Pearson et al., 2003). In Zeutec Spectraanalyser (double beam NIR) (Germany) spectometer Changes in protein, 1000 kernel weight, hardness, fat, ash, and mois- (1200–2600 nm), a reference library was prepared using Application TM ture are shown in Figure 1. worx software with multiple linear regression (MLR) model, and then unknown samples were fed in rotating cup drawer and analysed. The scanning was done everytime with all the NIR-filters included. In Quality evaluation of unknown wheat sample a calibration, a correlation between the NIR absorbance values at the Twenty-seven known wheat spectra were analysed using FT-NIR different wavelengths and the sample composition is established, and and NIR. Spectra of the same are shown in Figures  2 and 3. In calibration constants were calculated for each wavelength. When the FT-NIR, ident method is developed using absorption spectra of desired products and properties have been set up and the calibration stored wheat samples which is used as reference spectra for deter- coefficients have been entered, samples can be analysed with a predic- mination of unknown wheat samples. The analysis compares the tion calculation of the property results. test spectrum with all the reference spectra. Spectra of stored wheat grains were assigned in group according to the level of infestation Methodology for calculation of physical and and number of storage days, and then compared with spectra of chemical properties of wheat grains fresh wheat samples using ident test method with vector normal- The fresh and infested wheat samples were subjected to analysis of ization pre-processing method. All groups were identified uniquely. various physical and properties of grains such moisture, protein, ash, Figure  4 shows FT-NIR shows cross-validation results of stored fat, 1000 kernel wt, and hardness with changes in number of days wheat grains. In FT-NIR, the PLS regression showed a perfect fit of along with perectage level of insect infestation with the standard AOAC the model by higher values of R and lower RMSECV and RMSEE methods (AOAC, 2005; Mishra et al., 2018; Neethirajan et al., 2005). values (Table 2). In Zeutec NIR, the unknown wheat samples were compared with reference library and analysed with a prediction cal- Results and Discussion culation of the property results. Figure 5 shows NIR cross-validation Forty fresh samples and 60 infested samples of different wheat vari- results using Zeutec NIR. The best number of factor was determined ties were analysed with FT-NIR and NIR. by the lowest RMSEP value and highest value of R . The RMSEP Figure 2. FT-NIR absorption spectra of stored wheat samples used for identification. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 Evaluation of stored wheat grains, 2018, Vol. 2, No. 3 169 Figure 3. Zeutec-NIR absorption spectra of wheat samples used for identification. Table 2. Analytical features of the different regions and preprocessing methods for calibration and validation models in FT-NIR Parameters Wave number region (cm-1) Preprocessing method PLS factors Validation Calibration 2 2 RMSECV R RMSEE R (%) (%) (%) (%) Moisture 12489.4–7498.3and 5454-4242.8 Min Max normalization 10 0.731 95.26 0.601 97.15 Protein 12489.4 – 5446.3 and 4605.4-4242.8 First derivate 7 0.242 91.79 0.179 95.84 Ash 12489.4–7498.3, 6102- 5446.3 and First derivate + MSC 4 0.0392 83.18 0.0341 87.86 4605.4 – 4242.8 Fat 12489.4–7498.3, and 6102 – 5168.6 Straight line subtraction 8 0.0366 82.75 0.0287 90.45 1000 Kernel wt. 12489.4–7498.3 and 4605.4 – 4242.8 First derivate + MSC 6 1.00 90.30 0.748 95.01 Hardness 12489.4 – 4242.8 First derivate + straight line 2 1.23 80.78 1.21 82.02 subtraction 2 2 and R values for moisture, protein, fat, ash, 1000 kernal weight, 1.0, and 1.3 and maximum correlation coefficient (R ) 0.95, 0.91, and hardness determination with different spectral pre-processing 0.83,0.82, 0.90, and 0.80, respectively, were obtained (Sanchez- methods are presented in Table 3. During the storage studies of con- Marinez et al., 1997; Singh et al., 2008; Srivastava et al., 2018). trol and infested wheat samples, protein, hardness, and 1000 kernel weights decreased with an increase in days of storage and level of Conclusion infestation while the fat and ash increased. Moisture content is important for shelf-life and storage. Very high Infested and fresh wheat samples can be identified efficiently by moisture content (greater than 14.5%) attracts mold, bacteria, and the method developed in both FT-NIR and NIR (Armstrong et al., insects, all of which can result in storage issues or baking quality 2006; Amir et al., 2013; Salgo et al., 2012). In NIR, the reference deterioration. These parameters are commonly quantified using library for moisture, protein, ash, fat, 1000 kernel weight, and hard- either primary methods or FT-NIR or NIR spectroscopy. Replacing ness with lowest RMSEC values 0.62, 0.11, 0.025, 0.019, 0.81, and the primary methods with FT-NIR or NIR will provide faster results 0.56 and maximum correlation coefficient (R ) 0.86, 0.91,0.81, and accurate quantification, guaranteeing flour meets specifications. 0.96, 0.85, and 0.98, respectively, was obtained. In FT-NIR, the val- FT-NIR and NIR are an excellent means to visualize the chem- ues of validation for moisture, protein, ash, fat, 1000 kernel weight, ical composition of different wheat varieties with being very quick, and hardness with lowest RMSECV values 0.7, 0.2, 0.03, 0.03, Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 170 P. Pandey et al., 2018, Vol. 2, No. 3 Figure 4. Linear regression plot of measured vs. predicted values of (a) moisture, (b) protein, (c) ash, (d) fat, (e) hardness, and (f) 1000 kernel weight for cross- validation data set using FT-NIR. reliable, and cheaper analytical technique which can effectively be collecting a scan every second has minimum sample preparation. used for estimation of different wheat quality parameters. The values Both FT-NIR and NIR can be used for the rapid detection of quality determined by this technique are very close to the values determined of wheat grains. NIR is best suited if one has to deal with primary by the standard procedures and preferred over dispersive chemical bonds and simpler molecules/compounds but if the structure is a means due to following reasons: non-destructive technique provides complex one then FT-NIR suits better than NIR as it focuses on the precise measurement with no external calibration, increased speed, overtones and stretching frequency. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 Evaluation of stored wheat grains, 2018, Vol. 2, No. 3 171 Figure 5. Linear regression plot of (a) moisture, (b) protein, (c) ash, (d) fat, (e) hardness, and (f) 1000 kernel weight of stored wheat samples for cross-validated data set using Zeutec NIR. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018 172 P. Pandey et al., 2018, Vol. 2, No. 3 Table  3. MLR validation methods developed for the parameters Dong, J., van de Voort, F. R., Ismail, A., Akochi-Koble, E., Pinchuk, D. (2000). using Zeutec NIR Rapid determination of the carboxylic acid contribution to the total acid number of lubricants by Fourier transform infrared spectroscopy. Lubr Parameters MLR factors RMSEC (%) RMSEP (%) R Eng, 56: 12–17. (%) Edde, P. A., Phillips, T. W., Toews, M. D. (2005). Responses of Rhyzopertha dominica (Coleoptera: Bostrichidae) to its aggregation pheromones Moisture 2 0.62 0.70 86.41 as influenced by trap design, trap height, and habitat. Environmental Protein 5 0.11 0.14 91.97 Entomology, 34: 1549–1557. Ash 5 0.025 0.031 81.61 Hruskova, M., Machova, D. (2002). Changes of wheat flour properties during Fat 7 0.0197 0.030 96.09 short term storage. Czech Journal of Food Sciences, 20: 125–130. 1000 Kernel wt 1 0.81 0.87 85.36 Iconomou, D., Athanasopoulos, P., Arapoglou, D., Varzakas, T., Christopoulou, Hardness 5 0.567 0.776 98.07 N. (2006). Cereal quality characteristics as affected by controlled atmospheric storage conditions. American Journal of Food Technology, 1: 149–157. Karunakarana, C., Jayasa, D. S., White, N. D. (2004). Detection of internal Funding wheat seed infestation by Rhyzopertha dominica using X-ray imaging. No funding was received for the study. Journal of Stored Products Research, 40: 507–516. Koljonen, J., Nordling, T., Alander, J. (2008). A review of genetic algorithms in near infrared spectroscopy and chemometrics: past and future. Journal of Acknowledgements Infrared spectroscopy, 16: 189–197. Manley, M., Zyl, L. V., Osborne, B. G. (2002). Using Fourier transform near infra- Conflict of interest statement. All the authors declare that there is no conflict red spectroscopy in determining kernel hardness, protein, and moisture con- of interest. tent of whole wheat flour. Journal of Near Infrared Spectroscopy, 10: 71–76. Mishra, G., Srivastava, S., Panda, B. K., Mishra, H. N. (2018). Rapid assessment References of quality change and insect infestation in stored wheat grain using FT-NIR spectroscopy and chemometrics. Food Analytical Methods, 11: 1189–1198. Amir, R. M., Anjum, F. M., Khan, M. I., Khan, M. R., Pasha, I., Nadeem, Neethirajan, S., Karunakaran, C., Jayas, D. S., White, N. D.  G. (2005). M. (2013). Application of Fourier transform infrared (FTIR) spectros- Detection techniques for stored-product insects in grain. Food Control, copy for the identification of wheat varieties. Journal of Food Science and 18: 157–162. Technology, 50: 1018–1023. Pearson, T. C., Brabec, D. L., Schwartz, C. R. (2003). Automated detection of AOAC. (2005). Officials Methods of Analysis. 18th ed. Association of Officials internal insects infestations in whole wheat kernels using a Perten SKCS Analytical Chemists, Washington, DC. 4100. Applied Engineering in Agriculture, 19: 727–733. Armstrong, P. R., Maghirang, E. B., Xie, F., Dowell E. (2006). Comparison Salgo, A., Gergely, S. (2012). Analysis of wheat grain development using NIR of dispersive and Fourier-transform NIR instruments for measuring grain spectroscopy. Journal of Cereal Science, 56: 31–38. and flour attributes. American Society of Agricultural and Biological Sanchez-Marinez, R. I., Cortez-Rocha, M. O., Ortega-Dorame, F., Silveira, M. Engineers, 22: 453–457. ISSN 0883−8542. I. (1997). End-use quality of flour from Rhyzopertha dominica infested Basavaraja, H., Mahajanashetti, S. B., Udagatti, N. C. (2007). Economic wheat. Cereal Chemistry, 74: 481–483. analysis of post –harvest losses in food grains in India: a case study of Singh, C. B., Jayas, D. S., Paliwal, J., White, N. D.  G. (2008). Detection of Karnataka. Agricultural Economics Research Review, 20: 117–126. insect-damaged wheat kernels using near-infrared hyperspectral imaging. Bulla Jr, L. A., Kramer, K. J., Speirs, R. D. (1978). Insects and microorgan- Journal of Stored Products Research, 45: 151–158. isms in stored grain and their control. Advances in cereal science and Srivastava, S., Mishra, G., Mishra, H. N. (2018). FTNIR-A Robust Diagnostic technology. Tool for the Rapid Detection of Rhyzopertha dominica and Sitophilus Boyacioǧlu, D., Hettiarachchy, N. S. (1995). Changes in some biochemical oryzae Infestation and Quality Changes in Stored Rice Grains. Food and components of wheat grain that was infected with Fusarium gramine- Bioprocess Technology, 1–12. arum. Journal of Cereal Science, 21: 57–62. Downloaded from https://academic.oup.com/fqs/article-abstract/2/3/165/5048124 by Ed 'DeepDyve' Gillespie user on 28 August 2018

Journal

Food Quality and SafetyOxford University Press

Published: Sep 1, 2018

References