Pseudosciaena crocea producing area traceability discrimination method based on element and FTIR fusion fingerprint spectrum

Abstract

The invention discloses a large yellow croaker origin traceability discrimination method based on an element and FTIR fusion fingerprint spectrum. The method comprises the following steps: S1, sample treatment; s2, data acquisition: S2.1, mineral element data acquisition, and S2.2, FTIR spectrum acquisition; and S3, data analysis. Precise classification of production areas can be realized through principal component analysis (PCA), clustering analysis (CA) and linear discriminant analysis (LDA), the overall discriminant accuracy of back substitution inspection of the discriminant model reaches 100%, and the overall discriminant accuracy of cross inspection reaches 97.3%. The element and infrared spectrum fusion fingerprint technology is applied to the source tracing of the large yellow croaker producing area, and the application prospect is wide.

Description

Technical Field

[0001] The present invention relates to a method for discriminating the origin of products, in particular to a method for discriminating the origin of large yellow croaker based on the fusion fingerprint spectra of elements and FTIR. Background Art

[0002] The large yellow croaker (Larimichthys crocea) belongs to the family Sciaenidae. It has a streamlined body shape, a golden-yellow body surface, delicious meat rich in high-quality protein and trace elements, and is deeply loved by consumers, with extremely high economic value. In recent years, with the breakthrough of artificial breeding technology and cage culture technology of large yellow croaker, the culture of large yellow croaker has developed rapidly. However, researchers have found that due to certain differences in the culture environments (such as family lines, hydrological conditions, bait sources and geological backgrounds) in different production areas, there are differences in the nutritional components and taste of large yellow croaker from different origins. At present, the identification of the geographical indication of large yellow croaker mainly relies on empirical means such as distinguishing the appearance and smelling the odor to judge its origin. Therefore, developing efficient and objective origin identification technology is of great significance for ensuring the sustainable development of the large yellow croaker industry and strengthening the protection of geographical indications.

[0003] In the field of food origin traceability, the "fingerprint recognition" technology based on chemical composition analysis has become a research hotspot. Among them, elemental fingerprint analysis can effectively correlate organisms with the geochemical background of their growth environment by detecting the enrichment characteristics of elements in organisms. Stable isotope tracing technology has been widely used in the origin traceability of fruits, tea, vegetable oils, aquatic products, etc. Mineral element tracing technology has been widely used in the origin traceability and discrimination of agricultural products due to its good tracing properties. On the other hand, infrared spectroscopy technology is a rapidly emerging rapid, green and environmentally friendly detection technology in recent years, which comprehensively reflects the compositional differences between samples based on the characteristic absorption of infrared spectra. Infrared spectroscopy technology has been applied to the origin traceability of products such as tea, rice, fruits and vegetables, livestock and poultry meat, and aquatic products. In the early stage, single technology was mostly used in origin traceability research. However, single fingerprint analysis technology cannot characterize the complex chemical information in food, and is easily interfered by individual metabolic differences or environmental pollution, with certain limitations and unsatisfactory discrimination accuracy. In recent years, the trend of origin traceability research has gradually developed towards the combination of multiple technologies. The data fusion strategy is to merge information from two or more sources, and through exploring the complementary and synergistic effects between data from different sources, to more comprehensively characterize the sample information, which can further improve the accuracy of traceability. Based on the fusion of stable isotope and element technologies for the origin traceability of various agricultural products such as hairy crabs, milk, fruits, dried tangerine peels, wines, beef and mutton, the origin discrimination accuracy is significantly higher than that of single technology. Mei Guangming et al. used the differences in elements and stable isotope ratios to distinguish large yellow croakers from Zhejiang and Fujian. Yu Hanbing et al. traced strawberries using the combination of stable isotopes and elements and were able to accurately distinguish the origins. Wu Zhongyu achieved the origin identification of Bingtang oranges through the data fusion strategy of mineral elements, infrared spectroscopy and chromatographic data. The above research results show that the data fusion discrimination results are better than single information, and rich chemical information is conducive to accurate sample characterization. Compared with single instrument information, data fusion can more comprehensively characterize the information of samples and is more suitable for studying foods with complex and diverse chemical compositions.

[0004] There are differences in the mineral element and organic matter contents in large yellow croakers from different origins, which provides the feasibility for multi-dimensional data fusion origin discrimination analysis. At present, there is no literature report on the application of the fusion fingerprint technology of elements and infrared spectroscopy in the origin traceability of large yellow croakers. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for discriminating the origin of large yellow croakers based on the fusion fingerprint of elements and FTIR in view of the deficiencies of the prior art.

[0006] S1 Sample treatment

[0007] The collected chilled large yellow croaker samples are washed and scaled, and the back muscle tissue is taken and homogenized for standby.

[0008] Accurately weigh 0.2 g of the homogenized large yellow croaker sample (accurate to 0.0001 g), put it into a polytetrafluoroethylene digestion tank, add 1.7 mL of nitric acid and 1.3 mL of primary water, and use super microwave digestion. Digestion procedure: (1) Ramp temperature to 80 °C, pressure to 100 bar, ramp time 5 min; (2) Ramp temperature to 130 °C, pressure to 100 bar, ramp time 8 min; (3) Ramp temperature to 180 °C, pressure to 160 bar, ramp time 5 min; (4) Ramp temperature to 220 °C, pressure to 160 bar, ramp time 7 min; (5) Temperature 220 °C, pressure 160 bar, holding time 15 min; (6) Cool from 220 °C to room temperature within 20 min. The cooled digestion solution is made up to 25 ml with primary water for ICP-MS detection.

[0009] Place the homogenized large yellow croaker sample in the freezer at -30 °C. After freezing, put the sample into a vacuum freeze dryer and dry for 48 h. The dried sample is thoroughly crushed in an agate mortar and then passed through a 16-mesh sieve, and then put into a sealed bag for storage for FTIR detection.

[0010] S2 Data collection

[0011] S2.1 Mineral element data collection

[0012] During the ICP-MS element analysis, add a mixed internal standard solution of 1 μg / ml Sc, Ge, Rh, and In online. Prepare and measure the blank solution and the metal element analysis quality control sample in shellfish (GBW10024) in the same way, and perform quality control on the results. The quality control results are within the standard reference range, indicating that the experimental data accuracy is good. The measured data is expressed as the mean ± standard deviation (SD) of 3 repeated measurements.

[0013] S2.2 FTIR spectrum collection

[0014] Weigh (1.5 ± 0.2) mg of the sample and (100 ± 2) mg of potassium bromide, place them in an agate mortar and mix and grind them into a powder, then pour it into a mold and press it into a thin slice. Set the instrument resolution to 4 cm-1, the scanning range to 4000 - 400 cm-1, preheat and then measure the spectrum. Before scanning, use a blank slice to remove the interference of carbon dioxide and water in the background. The sample is measured 3 times repeatedly, and the average spectrum is taken as the sample measurement spectrum.

[0015] S3 Data analysis

[0016] For the mineral element content data, the element distribution characteristics were characterized by the mean ± standard deviation (mean±SD). The original data of Fourier transform infrared spectroscopy (FT-IR) were standardized: First, baseline correction was performed to eliminate the instrument background interference, and then the transmittance-absorbance conversion was completed. To further eliminate the spectral scattering effect, standard normal variate transformation (StandardNormalVariate, SNV) was used to preprocess the mineral element dataset and the ordinate of the infrared spectrum.

[0017] In the chemometrics modeling stage, by constructing a spectral preprocessing process: Savitzky-Golay (SG) convolution smoothing, first derivative (1D) and second derivative (2D) processing were performed in sequence, where the derivative calculation window parameters were consistent with the smoothing process. This combined algorithm can effectively eliminate baseline drift and enhance the resolution of spectral characteristic peaks. Data statistics were carried out, including one-way analysis of variance (oneway ANOVA), Fisher discriminant analysis, principal component analysis (principal component analysis, PCA) and cluster analysis (cluster analysis, CA). Description of the Drawings

[0018] Figure 1 Scatter plot of PC1 and PC2 scores in the PCA of large yellow croaker from different origins;

[0019] Figure 2 Original infrared spectra of large yellow croaker from 3250 - 750 cm-1 at different origins;

[0020] Figure 3 Infrared fingerprint spectra of large yellow croaker samples from different origins;

[0021] Figure 4 Second derivative map of the infrared spectra of large yellow croaker from 5 origins;

[0022] Figure 5 PCA of the combined data of elements and infrared second derivatives;

[0023] Figure 6 Cluster analysis of large yellow croaker from different origins. Detailed Implementation Modes

[0024] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with embodiments. The content mentioned in the implementation modes does not limit the present invention.

[0025] Example 1

[0026] A method for discriminating the origin of large yellow croaker based on the fused fingerprint spectra of elements and FTIR.

[0027] S1 Sample Processing

[0028] Wash and scale the collected fresh chub mackerel samples, take the back muscle tissue, and homogenize it for later use.

[0029] Accurately weigh 0.2 g (accurate to 0.0001 g) of the homogenized chub mackerel sample, put it into a polytetrafluoroethylene digestion tank, add 1.7 mL of nitric acid and 1.3 mL of primary water, and use super microwave digestion. Digestion procedure: (1) Ramp temperature to 80 °C, pressure to 100 bar, ramp time 5 min; (2) Ramp temperature to 130 °C, pressure to 100 bar, ramp time 8 min; (3) Ramp temperature to 180 °C, pressure to 160 bar, ramp time 5 min; (4) Ramp temperature to 220 °C, pressure to 160 bar, ramp time 7 min; (5) Temperature 220 °C, pressure 160 bar, holding time 15 min; (6) Cool from 220 °C to room temperature within 20 min. Dilute the cooled digestion solution to 25 ml with primary water for ICP-MS detection.

[0030] Freeze the homogenized chub mackerel sample at -30 °C, put the frozen sample into a vacuum freeze dryer, and dry for 48 h. Thoroughly crush the dried sample in an agate mortar and pass it through a 16-mesh sieve, then put it into a sealed bag for storage for FTIR detection.

[0031] S2 Data collection

[0032] S2.1 Mineral element data collection

[0033] During the ICP-MS element analysis, add a mixed internal standard solution of 1 μg / ml Sc, Ge, Rh, and In online. Prepare and measure the blank solution and the metal element analysis quality control sample in shellfish (GBW10024) in the same way, and perform quality control on the results. If the quality control results are within the standard reference range, it represents that the experimental data accuracy is good. The measured data is expressed as the average value (mean) ± standard deviation (standard deviation, SD) of 3 repeated measurements.

[0034] S2.2 FTIR spectrum collection

[0035] Weigh (1.5 ± 0.2) mg of the sample and (100 ± 2) mg of potassium bromide, put them into an agate mortar and mix and grind them into a powder, then pour it into a mold and press it into a thin slice. Set the instrument resolution to 4 cm-1, the scanning range to 4000 - 400 cm-1, preheat and then measure the spectrum. Use a blank slice to remove the interference of carbon dioxide and water in the background before scanning. Repeat the measurement of the sample 3 times, and take the average spectrum as the sample measurement spectrum.

[0036] S3 Data analysis

[0037] For the mineral element content data, the element distribution characteristics were characterized by the mean ± standard deviation (mean±SD). The original data of Fourier transform infrared spectroscopy (FT-IR) were standardized: first, baseline correction was performed to eliminate the instrument background interference, and then the transmittance-absorbance conversion was completed. To further eliminate the spectral scattering effect, standard normal variate transformation (StandardNormalVariate, SNV) was used to preprocess the mineral element dataset and the ordinate of the infrared spectrum.

[0038] In the chemometric modeling stage, by constructing a spectral preprocessing process: Savitzky-Golay (SG) convolution smoothing (window points 5), first derivative (1D) and second derivative (2D) processing were performed in sequence, where the derivative calculation window parameters were consistent with the smoothing process. This combined algorithm can effectively eliminate baseline drift and enhance the resolution of spectral characteristic peaks. IBM SPSS 22 was used for data statistics, one-way ANOVA, Fisher discriminant analysis, principal component analysis (principal component analysis, PCA) and cluster analysis (cluster analysis, CA). Finally, the visualization of the experimental data was completed through Origin 2021 to ensure the scientific expression of spectral characteristics and statistical results.

[0039] Analysis of the difference in mass fraction of mineral elements

[0040] The mass concentrations (μg kg-1) of AL, Ti, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, As, Se, Sr, Cd, Sn, Sb, Ba, Hg, and Pb in large yellow croaker samples were analyzed (Table 1). A one-way ANOVA was performed on the mass concentrations of these mineral elements among the large yellow croaker samples from the five production areas. As shown in Table 2, significant differences (P < 0.05) were found for AL, Ti, Cr, Mn, Fe, Co, Ni, Ga, As, Se, Sr, Sn, Sb, Ba, Hg, and Pb. No significant differences (P > 0.05) were found for Cu, Zn, and Cd. Analysis of mineral elements with significant differences found that the mass fractions of AL, Se, Hg and Pb in the yellow croaker cultured in Taizhou, Zhejiang were the highest, which were 1059.16μg·kg-1, 287.66±27.38μg·kg-1, 49.84±6.08μg·kg-1 and 16.15±4.02μg·kg-1 respectively, and the mass fraction of As was the lowest, which was 315.37±120.94μg·kg-1; the mass fractions of Ni, Cu and Zn in the yellow croaker cultured in Zhoushan, Zhejiang were the highest, which were The concentrations of Ti, Cr, Mn, and Sr in large yellow croaker cultured in Wenzhou, Zhejiang, were significantly higher than those in large yellow croaker from other production areas, at 2213.83 μg kg-1, 140.41 μg kg-1, 403.31 μg kg-1, and 5900.07 μg kg-1, respectively. The lowest concentration of Se was 217.49 ± 11.02. Large yellow croaker cultured in Zhanjiang, Guangdong, had the highest As concentration of 1073.11 μg kg-1, significantly higher than samples from the other four production areas. Large yellow croaker cultured in Zhanjiang, Guangdong, also had relatively low concentrations of Al, Ti, Cr, Mn, Fe, Co, Ni, Ba, and Pb, indicating high identification. Analysis of the mass fractions of the same elements in yellow croaker samples from five production areas revealed significant differences in their contents, which may be due to the different water environments of yellow croaker farming.

[0041] Principal component analysis based on mineral elements

[0042] PCA is a commonly used linear dimensionality reduction method in data processing, which simplifies the analysis complexity and maximally retains the information content of the original variables. Through orthogonal transformation, the original variables are converted into uncorrelated principal components. Among them, the first principal component explains the largest amount of information, and the subsequent components decrease in turn.

[37] The KMO (Kaiser-Meyer-Olkin) test was performed on the mass fractions of mineral elements in the large yellow croaker samples from 5 producing areas. The test value was 0.687 (KMO > 0.500), indicating that the correlation between elements was significant and PCA was meaningful. To enhance the accuracy and reliability of the model, elements with less influence on the principal components were removed, reducing the complexity of the model and also the influence of noise and errors, thus making the results of principal component analysis more stable and reliable. When constructing a model or making predictions, more accurate results can be obtained based on these screened and optimized principal components.

[0043] AL, Ti, Cr, Mn, As, Se, Sr, Ba, Hg, Pb, Fe, Ni, Sn with significant differences in mass fractions among the mineral elements were selected as variables for principal component analysis (see Figure 1 ). The principal component PC1 accounted for 46.81%, and the principal component PC2 accounted for 19.41%. As Figure 1 known, the scatter plots of large yellow croaker from the 3 producing areas of Zhanjiang in Guangdong, Wenzhou in Zhejiang, and Ningde in Fujian do not overlap with each other. Through PCA, large yellow croaker from the producing areas of Zhanjiang in Guangdong, Ningde in Fujian, and Wenzhou in Zhejiang can be distinguished. The large yellow croaker from the producing areas of Zhoushan and Taizhou cannot be completely distinguished. The investigation found that most of the yellow croaker in these two places are cultured using deep-water cages, and the culture cycle and water temperature are also relatively close, affecting the identification of mineral elements.

[0044] Infrared spectrum analysis of large yellow croaker

[0045] The original infrared absorption spectra of 75 large yellow croaker samples from 5 producing areas collected at 3250 - 750 cm-1 are as Figure 2 shown.

[0046] After baseline correction, noise elimination, background interference deduction and other factors for the large yellow croaker spectra from 5 producing areas, the Spectrum10 software was used to make an average spectrum of the infrared spectra of large yellow croaker from each producing area as the common pattern of its FTIR fingerprint ([[]] Figure 3 ). Then, the similarity analysis between the infrared spectrum data of cultured large yellow croaker from each place and the common pattern was carried out using the "correl" function in Excel2020. Taking the cultured large yellow croaker in Taizhou as an example, the similarity between 15 batches of cultured large yellow croaker in Taizhou and the common pattern was above 0.95, as shown in Table 2. The similarity between the fingerprint spectra of large yellow croaker from the other 4 producing areas and their respective common patterns was also above 0.95, indicating that the established common pattern map can be used as the fingerprint spectrum of the samples from this producing area.

[0047] Table 1 Similarity of Infrared Spectra of Taizhou Large Yellow Croaker Samples

[0048]

[0049] There are 11 main common peaks in the fingerprint spectra of large yellow croaker from 5 origins, which can be used as the common characteristic peaks of large yellow croaker. Large yellow croaker is rich in proteins, lipid substances, vitamins, etc., and also contains a certain amount of taurine and curcumin. Table 2 shows the chemical bond attribution of the 11 common peaks, indicating that the FTIR spectra of large yellow croaker can provide rich chemical composition information. Taking the 6th peak marked in Figure 3 as the reference peak, the relative peak heights of other common peaks are calculated as shown in Table 3. The results show that there are differences in the relative intensities of the spectral absorption peaks of large yellow croaker from 5 origins, indicating that the contents of characteristic substances in large yellow croaker from different origins are different, with certain fingerprint characteristics. Since the second derivative can reduce the influence of spectral overlap and baseline drift, in the analysis of some complex systems, the method based on the second derivative spectrum may be more accurate and reliable than the method based on the original spectrum. Select the infrared bands (3250 - 2700 cm-1 and 1800 - 1000 cm-1) with characteristic large yellow croaker from different origins for second derivative pretreatment. The data after second derivative pretreatment of the infrared average spectra of large yellow croaker from 5 origins are as shown in Figure 4 Figure. The results of second derivative spectral analysis show that there are significant spectral characteristic differences among large yellow croaker samples from different origins in the range of 3250 - 2700 cm-1 and 1800 - 1000 cm-1.

[0050] Table 2 Chemical Bond Attribution of Common Infrared Absorption Peaks

[0051]

[0052] Table 3 Common Infrared Absorption Peaks and Their Relative Peak Heights of Large Yellow Croaker from 5 Origins

[0053]

[0054]

[0055] 3.4 Discrimination Model of Large Yellow Croaker from Different Origins

[0056] To reduce the interference of useless spectral information and lower the data processing volume, only the spectral acquisition information under the characteristic bands of 3050 - 2800 cm-1 and 1800 - 1000 cm-1 was selected for preprocessing. The infrared data of large yellow croaker were preprocessed with algorithms such as Savitzky-Golay (SG) smoothing, standard normal variate (SNV), and second derivative (2D) for the original spectra. The mineral elements AL, Ti, Cr, Mn, As, Se, Sr, Ba, Hg, Pb, Fe, Ni, and Sn with significant differences in mass fraction among 35 batches of large yellow croaker were concatenated with the preprocessed FTIR spectral data. After normalization, data fusion was performed to obtain a new data matrix, and principal component analysis (PCA), cluster analysis (CA), and linear discriminant analysis (LDA) were respectively carried out on the new data matrix of the combination of mineral elements and FTIR.

[0057] To explore the effect of concatenating and fusing the mineral elements AL, Ti, Cr, Mn, As, Se, Sr, Ba, Hg, Pb, Fe, Ni, Sn with the preprocessed FTIR spectral data to obtain a new data matrix for discriminating the origin of large yellow croaker, the new data matrix of the combination of mineral elements and FTIR was selected as a variable for PCA analysis. The results are as Figure 5 shown that the principal component PC1 accounted for 59.59%, the principal component PC2 accounted for 15.11%, and the cumulative contribution rate of the first two principal components accounted for 74.7%. This indicates that the first two extracted principal component factors can well represent the comprehensive information of the element-FTIR combined data model. The samples of large yellow croaker from 5 origins did not overlap in the directions of PC1 and PC2, and the large yellow croaker from different origins were relatively distinct, indicating that it has high feasibility for tracing the origin of large yellow croaker based on the element-FTIR combined data model.

[0058] Figure 5 PCA of the combined data of elements and the second derivative of infrared. Note: a: Large yellow croaker from Wenzhou, Zhejiang; b: Large yellow croaker from Zhoushan, Zhejiang; c: Large yellow croaker from Ningde, Fujian; d: Large yellow croaker from Zhanjiang, Guangdong; e: Large yellow croaker from Taizhou, Zhejiang

[0059] Cluster analysis is an unsupervised learning method that groups data objects according to their similarity, making the objects within the same group have high similarity while the objects between different groups have large differences. To more intuitively understand the distribution of the origin of large yellow croaker, cluster analysis was carried out with the new data matrix obtained by concatenating and fusing the mineral elements AL, Ti, Cr, Mn, As, Se, Sr, Ba, Hg, Pb, Fe, Ni, Sn and the preprocessed FTIR spectral data as variables. The Ward's method and Euclidean distance method were selected for cluster analysis. Cutting the dendrogram at the cluster distance, the large yellow croaker samples can be divided into 5 categories ( Figure 6As shown, one color line corresponds to the origin of large yellow croaker: all of the first category are large yellow croakers cultured in Ningde, Fujian (1 - 10, n = 10), all of the second category are large yellow croakers cultured in Taizhou, Zhejiang (11 - 20, n = 10), the third category are large yellow croakers cultured in Wenzhou, Zhejiang (21 - 30, n = 10), the fourth category are large yellow croakers cultured in Zhanjiang, Guangdong (31 - 40, n = 10), and the fifth category are large yellow croakers cultured in Zhoushan, Zhejiang (41 - 50, n = 10). The results show that cluster analysis can be used for the discriminant analysis of large yellow croakers from 5 different origins.

[0060] Figure 6 Cluster Analysis of Large Yellow Croakers from Different Origins

[0061] Note: Cultured in Ningde, Fujian from 1 - 10; cultured in Taizhou, Zhejiang from 11 - 20; cultured in Wenzhou, Zhejiang from 21 - 30; cultured in Zhanjiang, Guangdong from 31 - 40; cultured in Zhoushan, Zhejiang from 41 - 50

[0062] Fisher discriminant analysis and principal component analysis (PCA) are two data dimensionality reduction methods based on different principles. PCA realizes dimensionality reduction by extracting the direction with the largest data variance, while Fisher discriminant analysis aims to maximize the between-class scatter and minimize the within-class scatter, constructs a discriminant model through known class samples, and then classifies unknown samples. At present, this method has been successfully applied to the origin traceability research of dairy products, rice, swimming crabs, etc., showing good discriminant performance. To deeply explore the accuracy of the fusion matrix of the contents of elements such as AL, Ti, Cr, Mn, As, Se, Sr, Ba, Hg, Pb, Fe, Ni, Sn, etc. and the preprocessed data of Fourier transform infrared spectroscopy extracted by PCA dimensionality reduction in the origin identification of large yellow croakers, this study selects large yellow croakers from 5 different origins for Fisher discriminant analysis. In the experimental design, 10 samples are randomly selected from 15 batches of large yellow croakers from each origin to form a training set for constructing a discriminant model, and the remaining 5 samples are used as an independent test set, only for verifying the prediction performance of the constructed discriminant model. The discriminant functions for large yellow croaker samples from 5 different origins are respectively:

[0063] Y Cultured in Ningde, Fujian = 236.242AL - 189.327Ti - 62.929Cr - 32.959Mn + 35.836As + 127.673Se + 377.04Sr - 190.685Ba - 13.675Hg + 71.956Pb + 29.116Fe - 178.578Ni - 63.127Sn - 54.957 + X(2864) + 1635.844X(2865) - 1239.438X(2853) + 2468.012X(2854) - 205.283X(1475) - 729.621

[0064] Y Zhejiang Taizhou Aquaculture = 265.27AL - 166.534Ti - 152.647Cr - 22.199Mn + 60.661As + 158.259Se + 282.969Sr - 223.223Ba + 51.921Hg + 125.728Pb + 101.724Fe - 129.462Ni - 42.853Sn - 299.351X(2864) + 1939.457X(2865) - 1310.697X(2853) + 2581.811X(2854) - 199.682X(1475) - 899.845

[0065] Y Zhejiang Wenzhou Aquaculture = 288.635AL - 57.72Ti + 101.839Cr + 651.87Mn + 63.792As - 57.681Se + 1289.517Sr - 409.231Ba - 13.89Hg + 142.605Pb - 33.338Fe - 250.593Ni - 24.622Sn - 66.15X(2864) + 1712.708X(2865) - 1545.933X(2853) + 2559.963X(2854) - 589.274X(1475) - 1059.949

[0066] Y Guangdong Zhanjiang Aquaculture = 144.698AL - 160.787Ti - 89.23Cr - 93.921Mn + 74.2As + 163.716Se + 245.351Sr - 166.431Ba + 42.9Hg + 88.686Pb + 69.339Fe - 148.619Ni - 83.779Sn - 19.187X(2864) + 1587.892X(2865) - 1011.486X(2853) + 2262.723X(2854) - 162.38X(1475) - 766.819

[0067] Y Zhejiang Zhoushan Aquaculture = 196.407AL + 53.639Ti - 191.998Cr - 85.964Mn + 133.349As + 131.931Se - 132.099Sr - 89.149Ba - 2.559Hg + 108.704Pb + 115.951Fe + 32.926Ni + 55.818Sn - 439.308X(2864) + 1800.425X(2865) - 1145.663X(2853) + 2334.98X(2854) - 33X(1475) - 917.615

[0068] In the formula: the elemental symbols (AL, Ti, Cr, Mn, As, Se, Sr, Ba, Hg, Pb, Fe, Ni, Sn) represent the mass fractions of the corresponding elements in the large yellow croaker samples, and X(i) is the second derivative value of the corresponding wave number of the large yellow croaker FTIR.

[0069] Using the above discriminant to predict the samples in the test set, the prediction accuracy of large yellow croakers from each origin is 100%. To test the reliability of the discriminant, the "leave-one-out" cross-validation method was used for testing. The accurate discrimination rates of large yellow croaker samples from 5 origins (Zhoushan, Zhejiang; Taizhou, Zhejiang; Wenzhou, Zhejiang; Ningde, Fujian; Zhanjiang, Guangdong) are all 97.3%, indicating that the established discriminant can preferably identify the origins of large yellow croakers from different origins.

[0070] Those skilled in the art of the present technology can understand that, unless otherwise defined, all terms used herein (including technical terms and scientific terms) have the same meaning as the general understanding of those of ordinary skill in the art to which the present invention pertains. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted in an idealized or overly formal sense unless specifically defined as here.

[0071] It should be understood that the detailed description of the technical solutions of the present invention with the aid of the preferred embodiments above is illustrative rather than restrictive. Those of ordinary skill in the art can modify the technical solutions recorded in each embodiment based on reading the specification of the present invention, or perform equivalent substitution on some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A method for discriminating the origin of large yellow croaker based on the fusion fingerprint spectrum of elements and FTIR, characterized in that S1 Sample treatment; S2 Data collection: S2.1 Mineral element data collection, S2.2 FTIR spectrum collection; S3 Data analysis.

2. The method according to claim 1, characterized in that S1 sample treatment includes: The collected ice fresh large yellow croaker samples are washed and scaled, the back muscle tissues are taken, and homogenized for standby; Accurately weigh 0.2 g of the homogenized large yellow croaker sample (accurate to 0.0001 g), put it into a polytetrafluoroethylene digestion tank, add 1.7 mL of nitric acid and 1.3 mL of primary water, and use super microwave digestion. Digestion program: (1) Ramp temperature 80 °C, pressure 100 bar, ramp time 5 min; (2) Ramp temperature 130 °C, pressure 100 bar, ramp time 8 min; (3) Ramp temperature 180 °C, pressure 160 bar, ramp time 5 min; (4) Ramp temperature 220 °C, pressure 160 bar, ramp time 7 min; (5) Temperature 220 °C, pressure 160 bar, holding time 15 min; (6) Cool from 220 °C to room temperature within 20 min. The cooled digestion solution is made up to 25 ml with primary water for ICP-MS detection; The homogenized large yellow croaker samples are frozen at -30 °C, and the frozen samples are put into a vacuum freeze dryer and dried for 48 h. The dried samples are fully crushed in an agate mortar and passed through a 16-mesh sieve, and then put into a sealed bag for storage for FTIR detection.

3. According to the method described in claim 1, wherein, S2.1 Mineral element data collection includes: During the ICP-MS element analysis, a mixed internal standard solution of 1 μg / ml Sc, Ge, Rh and In is added online. The blank solution and the metal element analysis quality control sample in shellfish (GBW10024) are prepared and measured in the same way, and the results are quality controlled. The quality control results are within the standard reference range, indicating that the experimental data accuracy is good. The measured data is expressed as the average value (mean) ± standard deviation (standard deviation, SD) of 3 repeated measurements.

4. According to the method described in claim 1, wherein, S2.2 FTIR spectrum collection includes: Weigh (1.5 ± 0.2) mg of the sample and (100 ± 2) mg of potassium bromide, put them into an agate mortar and mix and grind them into powder, then pour them into a mold and press them into a thin slice. Set the instrument resolution to 4 cm-1, the scanning range to 4000~400 cm-1, preheat and then measure the spectrum. Before scanning, use a blank slice to remove the interference of carbon dioxide and water in the background. The sample is measured 3 times repeatedly, and the average spectrum is taken as the sample measurement spectrum.

5. According to the method described in claim 1, wherein, S3 data analysis includes: For the mineral element content data, the element distribution characteristics were characterized by the mean ± standard deviation (mean±SD); the original data of Fourier transform infrared spectroscopy (FT-IR) were standardized: first, baseline correction was performed to eliminate the instrument background interference, and then the transmittance-absorbance conversion was completed; to further eliminate the spectral scattering effect, standard normal variate transformation (StandardNormalVariate, SNV) was used to perform standardized preprocessing on the mineral element dataset and the ordinate of the infrared spectrum. In the chemometrics modeling stage, by constructing a spectral preprocessing process: sequentially performing Savitzky-Golay (SG) convolution smoothing, first derivative (1D) and second derivative (2D) processing, where the derivative calculation window parameter is consistent with the smoothing process; performing data statistics, one-way ANOVA, Fisher discriminant analysis, principal component analysis (principal component analysis, PCA) and cluster analysis (cluster analysis, CA).

6. The method according to claim 5, wherein Determine the overall discrimination accuracy of the back-substitution test of the discrimination model and the overall discrimination accuracy of the cross-validation.

Images