[Calibration Transfer of Near Infrared Spectrometric Models for Crude Protein of Protein Feed Materials].

The near infrared spectrometric quantitative model of protein feed and its sharing in different instruments can greatly improve the utilization efficiency of the model and meet the needs of rapid development of feed industry. Considering the issue of applicability of near infrared spectrometric models for crude protein of protein feed materials, calibration transfer was explored among three types of instruments using spectral subtraction correction, direct standardization and piecewise directs standardization methods for the first time. Four kinds of protein feed raw materials were involved in the present study, corn protein powder, rapeseed meal, fish meal and distillers dried grains with soluble. The experimental instruments included MATRIX-I Fourier transform near infrared instrument (master instrument), Spectrum 400 Fourier transform near infrared instrument (slave 1 instrument), and SupNIR-2750 grating near infrared instrument (slave 2 instrument). Results showed that the spectral data difference for all the samples between the master and slave 2 instrument was relatively small, and the difference between the master and slave 1 instrument, and slave 1 and slave 2 instrument were relatively large. All the root mean square error of prediction and bias values after calibration transfer were lower than the values before calibration transfer, except that no improvement was found for the prediction of corn protein powder of slave 2 instrument corrected by piecewise direct standardization method. The relative prediction deviation (RPD) of corn protein powder, rapeseed meal and distillers dried grains with soluble transferred by all three methods were higher than 3, which indicated good predictions, while the RPD of fish meal were all higher than 2.5, which indicated relative good predictions. All three techniques used in the study were effective in the correction of the difference between different instruments for protein feed materials. This study is of important practical significance for the application of near infrared spectrometric models for crude protein of protein feed materials.

[Statistical comparison of independent validation results for near infrared spectroscopy models predicting calorific value of straw].

Two hundred and twenty-two straw samples, consisting of 170 rice straw samples and 50 wheat straw samples, were collected from 24 provinces of China. Near infrared spectroscopy (NIRS)was applied to build quantitative models for calorific value of straw combining the use of principal component regression (PCR), partial least square regression (PLS)and modified partial least square regression (MPLS). Different scatter correction methods and derivative treatments were adopted to help improve the accuracy of NIRS models. A total of 54 NIRS models were obtained and independent validations were conducted using the same validation set of samples. A statistical comparison of independent validation results was then introduced to evaluate whether the models perform significantly. Bias and bias corrected standard error of prediction (SEP(C)), which are the mean and the standard deviation of the prediction residuals respectively, were compared by the proposed statistical procedures. It was concluded that near infrared spectroscopy was able to predict the calorific value of straw samples rapidly and accurately, with resulting SEP(C)s between 134 and 178 J x g(-1); statistical comparison of biases and SEP(C)s was a reasonable and efficient way to compare spectral pre-processing methods, and select NIRS models predicting calorific value of straw.

[Proximate analysis of straw by near infrared spectroscopy (NIRS)].

Proximate analysis is one of the routine analysis procedures in utilization of straw for biomass energy use. The present paper studied the applicability of rapid proximate analysis of straw by near infrared spectroscopy (NIRS) technology, in which the authors constructed the first NIRS models to predict volatile matter and fixed carbon contents of straw. NIRS models were developed using Foss 6500 spectrometer with spectra in the range of 1,108-2,492 nm to predict the contents of moisture, ash, volatile matter and fixed carbon in the directly cut straw samples; to predict ash, volatile matter and fixed carbon in the dried milled straw samples. For the models based on directly cut straw samples, the determination coefficient of independent validation (R2v) and standard error of prediction (SEP) were 0.92% and 0.76% for moisture, 0.94% and 0.84% for ash, 0.88% and 0.82% for volatile matter, and 0.75% and 0.65% for fixed carbon, respectively. For the models based on dried milled straw samples, the determination coefficient of independent validation (R2v) and standard error of prediction (SEP) were 0.98% and 0.54% for ash, 0.95% and 0.57% for volatile matter, and 0.78% and 0.61% for fixed carbon, respectively. It was concluded that NIRS models can predict accurately as an alternative analysis method, therefore rapid and simultaneous analysis of multicomponents can be achieved by NIRS technology, decreasing the cost of proximate analysis for straw.

[Methods of analyzing soybean meal adulteration in fish meal based on visible and near infrared reflectance spectroscopy].

The present study investigated the feasibility of visible and near infrared reflectance spectroscopy (NIRS) method for the detection of fish meal adulteration with vegetable meal. Here the authors collected fish meal and soybean meal (representative vegetable meal) which were common used in our country. Fish meal was adulterated with different proportion of soybean meal and then the doping test samples were prepared. Qualitative discriminant analysis and quantitative analysis were studied with representative fish meal adulterated with soybean meal. Two hundred and six calibration samples and 103 validation samples were used in the qualitative discriminant analysis. The effects of different spectrum pre-treatment methods and spectrum regions were considered when the qualitative discriminant analysis model was established. Based on the smallest standard error of cross validation (SECV) and the correct rate, the spectrum region of visible and NIR was chosen as the best region. The eventually established pre-treatment methods were the standard multi-scatter correction (Std MSC) combined with the second derivative (2, 4, 4, 1). Then the independent external validation set was used to test the model, and there was no false positive samples and false negative samples. The correct discriminant rate was 96.12%. In quantitative analysis, 130 fish meal samples adulterated with soybean meal were used as calibration set. The calibration model was established by partial least squares (PLS). Furthermore, the effect of different spectrum pre-treatment methods and the spectrum region were considered. The results showed that the best pre-treatment method was the standard normalized variate (SNV) combined with the second derivative (2, 4, 4, 1). The coefficient of determination (R2) and the standard errors of calibration (SEC) were 0.989 0 and 1.539 0 respectively between the predictive value and the actual value. Sixty five fish meal samples adulterated with soybean meal were used as independent validation set. The coefficient of determination (R2) and the standard errors of prediction (SEP) were 0.988 8 and 1.786 0 respectively, and the ratio of standard deviation of reference data in prediction sample set to the standard errors of prediction (RPD) was 8.61. The results showed that the NIRS could be used as a method to detect the existence and the content of soybean meal in fish meal.

[NIRS method for determination of meat and bone meal content in ruminant concentrates].

Feed contaminated with MBM is commonly accepted as the main transmission carrier of bovine spongiform encephalopathy (BSE). To prevent BSE many countries have banned MBM as a feed ingredient. In the People's Republic of China, the ban was first applied to ruminant feed. In order to investigate the feasibility of near infrared diffuse reflectance spectroscopy method for rapidly quantitative determination of meat and bone meal content in ruminant concentrates, 225 representatively commercial ruminant concentrates samples and 75 meat and bone meal (including cattle, sheep, pig and poultry meat and bone meal) samples were collected in the People's Republic of China. Two hundred twenty five ruminant concentrates samples of adulterated meat and bone meal (0.5%-35%) were prepared including 135 calibration samples and 90 independent validation samples. For the calibration set samples, 3 samples were prepared at each concentration. For validation set samples, 2 samples were prepared at each concentration. Any one commercial ruminant concentrates was used once only. The spectra were scanned by raster near infrared diffuse reflectance spectroscopy instrument, and the effect of spectrum pretreatment methods (mathematic pretreatments and scatter correction) and spectrum region (visible and NIR) on the calibration results was considered. The calibration equation was established by modified partial least squares method. The result showed that the calibration gave r2 of 0.979, a standard error of calibration (SEC) of 1.522% and a standard error of cross validation (SECV) of 1.582%. The 90 independent validation samples were used to validate the quantitative equation. The r2, a standard error of prediction (SEP) and ratio of performance to standard deviation (RPD) were 0.972, 1.764% and 5.99 respectively. The results of this study indicated that near infrared diffuse reflectance spectroscopy method could provide rapidly quantitative prediction for meat and bone meal percent in ruminant concentrates. This method was significant in practice for enriching the rapidly quantitative methods of determining animal feed materials.

[Discriminant analysis of near infrared diffuse reflectance spectra to detect adulteration of non-ruminant meat and bone meal].

In order to study the feasibility of using near infrared (NIR) diffuse reflectance spectroscopy to discriminate adultera tion of non-ruminant meat and bone meal (MBM) with ruminant MBM, a total of 39 MBM samples made up of 15 from pig, 15 from poultry, 5 from cattle and 4 from sheep produced in different areas in China were chosen. The MBM samples were ground with 0. 5 mm sieve. 252 specimens were prepared by non-ruminant MBM deliberately adulterated with different proportion of ruminant MBM. The specimens were scanned by FOSS NIRSystem 6500. A calibration set of 180 specimens and an independent validation set of 72 specimens were randomly selected by the WINISI software. Discriminant analysis model was developed by partial least squares (PLS) on the calibration set and validated with independent validation set. The best discriminant model was obtained using standard normal variate and detrend (SNVD) and second derivative for spectrum pretreatment; this model had a coefficient of determination (R2(CV)) of 0.83 and a standard error of cross-validation (SECV) of 0. 147 1. For the independent validation set, the correct classification rate is 90%. There were a false negative specimen (0.5%) and two uncertain specimens (1%, 1.5%) in validation set. Results showed that it is feasible to use NIR diffuse reflectance spectroscopy to discriminate adulteration of non-ruminant MBM with ruminant MBM, but for specimens adulterated with ruminant MBM at less than 2%, the accuracy of calibration model needs to be improved. NIR was a rapid and non-destructive approach to discriminating adulteration of non-ruminant MBM with ruminant MIBM.

[Rapidly determination of compositions of animal manure compost using near infrared reflectance spectroscopy].

Composting is a process of aerobic thermophilic microbial degradation or an exothermic biological oxidation of various wastes by many populations of the indigenous microorganisms, which lead to a stabilized, mature, deodorized and hygienic product, free of pathogens and plant seeds, rich in humic substances, less volume, easy to store and marketable as organic amendment or fertilizer. Compared to the conventional wet chemical method, near-infrared reflectance spectroscopy (NIRS), a rapid, nondestructive, cost-effective technique, has been extensively used for qualitative and quantitative analysis in the field of agriculture. This study was to explore the capability of NIRS to analyze the compositions of Chinese animal manure compost. A representative population of 120 animal manure compost samples from 22 provinces in China was selected as research object, and this study explored the feasibility of analyzing animal manure compost compositions, which included moisture (Moist), volatile solid (VS), total organic carbon (TOC), total nitrogen (TN), C : N, pH and Electronic conductivity (EC) using NIRS. Original samples were scanned with a SPECTRUM ONE NTS (Perkin Elmer, New Jersey, USA) from 10 000 to 4 000 cm(-1). NIRS calibrations of a series of chemical parameters were developed by means of partial least-squares (PLS) regression. Results showed that the determination coefficient of calibration (r2) and the standard error of estimate (SEE) were Moist (0.981 6, 21.98), VS (0.936 5, 37.29), TOC (0.961 0, 16.46), TN (0.987 4, 1.61), C : N (0.741 0, 2.29), pH (0.788 0, 0.48) and EC (0.870 4, 1.74), respectively. The determination coefficient of validation (r2(V)) and the standard error of prediction (SEP) were Moist (0.983 2, 20.99), VS (0.938 1, 35.07), TOC (0.912 8, 26.34), TN (0.973 5, 3.96), C : N (0.830 8, 2.01), pH (0. 615 8, 0.60) and EC (0.895 3, 1.87), respectively. The value of RPD (SD/SEP) for Moist, VS, TOC, TN and EC were all greater than 3.0, 2.39 for C : N and 1.63 for pH. Together, results showed the feasibility and efficiency of NIRS to determinate compositions of animal manure compost.