Method for identifying tea leaves of different seasons

Abstract

The present application relates to a method for identifying different seasons of tea, and belongs to the technical field of tea quality detection. The existing tea season identification relies on sensory or single component analysis, which is highly subjective and lacks accuracy. The present application collects tea samples, detects the δ 15 N stable isotope ratio and the contents of EGC, EGCG, ECG and CAF organic components in the tea samples, and substitutes these data into the linear discriminant models for spring, summer and autumn respectively to calculate the values of Y 春季 , Y 夏季 and Y 秋季 . The season corresponding to the largest value is the season to which the tea sample belongs. This method uses multiple indexes in combination with linear discriminant models to objectively identify the season of tea, and can be used in tea quality control, market supervision and tea traceability, etc., providing a scientific basis for accurate judgment of the season of tea.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tea identification. More particularly, the present application relates to a method for identifying tea leaves of different seasons. BACKGROUND

[0002] Camellia sinensis is a perennial economic crop. The climate factors such as light, precipitation and temperature change in different seasons. The changes in climate factors in different seasons will cause differences in the physiological needs, absorption and enrichment of tea plant tissues to nutrients, and further affect the composition and content of chemical components in tea leaves, so that the tea quality is closely related to the season, and the price difference between spring tea and summer tea and autumn tea is large. In spring, the temperature is suitable and the rainfall is sufficient, and the tea plant has been recuperating for a long time in winter, so the nutrient components in the tea plant are rich, the spring buds are plump, and the contents of effective components such as amino acids and various vitamins related to improving tea quality are rich. The spring tea has a fresh and refreshing taste, strong aroma and obvious health care effect. Moreover, there is no disease and pest damage during the spring tea period, and no pesticide is needed, so the tea is pollution-free, and the green tea quality is the best in a year. In summer, the tea shoots grow rapidly but are prone to aging, the contents of amino acids and vitamins in tea leaves are significantly reduced, and the contents of anthocyanins, caffeine and tea polyphenols are significantly increased, so the taste is bitter and astringent. In autumn, the climate is between spring and summer, and the climate is mild in the later period of autumn tea, but the rainfall is often insufficient, so the tea leaves are dry and old. In addition, the tea plant has been deficient in nutrients after the spring tea and summer tea harvesting, so the autumn tea is poor in internal substances, and the taste is light and the aroma is not high. Therefore, it is urgent to establish a rapid, simple and accurate technical method for identifying tea leaves of different seasons.

[0003] Stable isotope refers to a kind of isotope that does not decay or decays very little, most of which is derived from nature itself, a small part of which is derived from stable products after radioactive isotope decay, and which is less affected by human or environment, and has a certain stability. It is an important index in tea origin tracing, and its basic principle and basis is the isotope fractionation effect. This effect is mainly affected by external environment (climate, soil, topography) and its own metabolism. The content of stable isotope of different organisms will be enriched or depleted under the action of fractionation mechanism, resulting in differences in natural abundance of stable isotope of organisms under the influence of different seasons. Due to the fractionation effect of isotope of elements in tea plant body affected by factors such as precipitation and sunlight, the abundance difference is caused, and the difference in isotope ratio affected by the interaction between organism and natural environment becomes the natural imprint of tea, which can indirectly reflect the environmental information of tea growth. The stable isotope ratio (δ 13 C, δ 15N) are different. At the same time, organic components also play a crucial role in tea metabolism, and the contents of free amino acids, tea polyphenols, caffeine and soluble sugars in tea leaves are different in different seasons, thereby affecting the sensory characteristics of tea leaves. Therefore, the season of tea leaves can be inferred by analyzing the stable isotope ratio and the content of organic components in the tea leaves. SUMMARY

[0004] Another object of the present application is to provide a method for identifying tea leaves of different seasons, which can infer the season of tea leaves by analyzing the stable isotope ratio and the content of organic components in the tea leaves.

[0005] In order to achieve these objects and other advantages according to the present application, a method for identifying tea leaves of different seasons is provided, characterized in that it comprises the following steps:

[0006] Step one, collecting a tea leaf sample, detecting the stable isotope ratio and the content of organic components in the tea leaf sample, wherein the stable isotope is δ 15 N, and the organic components are EGC, EGCG, ECG and CAF;

[0007] Step two, substituting the detected δ 15 N ratio and the contents of EGC, EGCG, ECG and CAF into the following linear discriminant model, respectively, to calculate the values of Y 春季 , Y 夏季 and Y 秋季 :

[0008] Y 春季 = -79.166δ¹³C + 4.269δ 15 N - 0.590EGC - 0.044EGCG - 2.083ECG +9.610CAF - 1221.849;

[0009] Y 夏季 = -75.594δ¹³C + 4.351δ 15 N - 0.664EGC + 0.208EGCG - 1.793ECG +9.124CAF - 1144.902;

[0010] Y 秋季 = -78.922δ¹³C + 4.214δ 15 N - 0.802EGC - 0.011EGCG - 2.107ECG +10.037CAF - 1226.854;

[0011] Step 3: Based on Y 春季 Y 夏季 Y 秋季 The numerical value is used to determine the season to which the tea sample belongs.

[0012] Preferably, the pretreatment of the tea sample in step one employs a simultaneous extraction method: the pulverized tea sample is extracted with a mixture of formaldehyde and aluminum hydroxide at a volume ratio of 7:3.

[0013] The alcohol-water solution was ultrasonically extracted for 30 minutes. After filtration, the supernatant was divided into two portions. One portion was used for HPLC detection of EGC, EGCG, ECG, and CAF. The other portion was freeze-dried and used to stabilize the isotopes δ¹³C and δ¹³C. 15 EA-IRMS detection of N.

[0014] Preferably, before pre-processing the tea, the origin type of the tea is first determined by near-infrared spectroscopy, either mountain tea or lowland tea, and an origin correction coefficient K is introduced into the linear discriminant model in step two: when it is mountain tea, Yspring, Ysummer, and Yautumn are multiplied by 1.05, 1.02, and 1.01, respectively; when it is lowland tea, they are multiplied by 0.95, 0.98, and 0.99, respectively.

[0015] Preferably, in step one, the tea saponin (TS) content of the tea sample also needs to be detected, and a TS correction term is added to the linear discriminant model in step two: Y 春季修正值 = Y 春季 -0.15TS, Y 夏季修正值 = Y 夏季 +0.18TS, Y 秋季修正值 = Y 秋季 -0.07TS.

[0016] Preferably, in step one, the detection of stable isotope ratios and organic component content is validated using parallel samples. Each tea sample is tested at least three times in parallel, and the average value is used in the linear discriminant model. If the relative standard deviation of the parallel samples exceeds 5%, the test is repeated.

[0017] Preferably, according to Y 春季 Y 夏季 Y 秋季 The method for determining the season of a tea sample based on numerical values ​​is to first compare Y... 春季 Y 夏季 Y 秋季 The numerical value is then compared, and a multi-level judgment mechanism is executed:

[0018] First level: Calculate the relative range R=(Y max -Y min ) / Y maxX 100%, if R ≥ 4%, then determine Y max The corresponding season is the season to which the tea sample belongs, wherein Y max is the maximum value obtained by substituting the linear discriminant model, Y min is the second largest value obtained by substituting the linear discriminant model.

[0019] The second level: if 3% ≤ R < 4%, the synergistic verification method of δ 13C value and seasonal characteristic component is used: the spring season is tea amino acid Thea, the summer season is EGCG, and the autumn season is ECG. The correlation index K of the δ 13C value and the content of the seasonal characteristic component is calculated, K = (δ 13C measured value / δ 13C seasonal average) x (seasonal characteristic component measured value / seasonal characteristic component average). When 0.9 ≤ K ≤ 1.1, then determine Y max The corresponding season is the season to which the tea sample belongs.

[0020] Preferably, when K < 0.9 or K > 1.1 in the second level judgment, the tea sample enters the secondary verification. The method of the secondary verification is: calculating Y max The corresponding season and Y min The ratio of the content of the characteristic component between the corresponding seasons, the ratio of Thea content / EGCG content for spring and summer, the ratio of EGCG content / ECG content for summer and autumn, and the ratio of ECG content / Thea content for autumn and spring. If the ratio is in the transition interval of the corresponding season, the tea sample is determined to be a transition season sample. If the ratio exceeds the transition interval of the corresponding season, Y min The corresponding season is determined to be the season to which the tea sample belongs.

[0021] Preferably, the δ 13C seasonal average is: the δ 13C spring average is -27.1‰, the δ 13C summer average is -26.38‰, and the δ 13C autumn average is -26.35‰.

[0022] The seasonal characteristic component average: the EGCG characteristic component average in summer is 142.57, the tea amino acid characteristic component average in spring is 28, and the ECG characteristic component average in autumn is 25.13.

[0023] The seasonal transition interval is: the transition interval of spring and summer is 1.2-1.8, the transition interval of summer and autumn is 0.8-1.2, and the transition interval of autumn and spring is 2.0-3.0.

[0024] Preferably, a third level is further included, if R < 3%, the following verification is performed:

[0025] The ratio of cellulose to lignin content CLR is determined. The CLR of spring tea is ≤1.2, the CLR of summer tea is 1.3-1.8, and the CLR of autumn tea is ≥1.9.

[0026] The total amount of free amino acids FAA is determined by the ninhydrin colorimetric method, the FAA of spring tea is greater than or equal to 4.5%, the FAA of summer tea is 2.0-4.4%, and the FAA of autumn tea is 3.0-5.0%;

[0027] The potassium / calcium content ratio is detected, the K / Ca of summer tea is greater than or equal to 1.5, and the K / Ca of spring and autumn tea is less than or equal to 1.4;

[0028] The total flavonoid content TF is determined, the TF of summer tea is greater than or equal to 3.2%, the TF of spring tea is less than or equal to 2.5%, and the TF of autumn tea is 2.6-3.1%;

[0029] If the measured value of the index falls into the season corresponding to the maximum value or the season corresponding to the second maximum value, the corresponding season is scored 5 points, and if it does not fall into the season, it is scored 0 points, the full score is 20 points, and the higher the score is, the more the tea is determined to be the season;

[0030] If at least three of the four indexes in one season fall into the corresponding interval, the tea is determined to be the season;

[0031] If the total score of two seasons is equal, the content of cis-3-hexenol glycoside is detected, the content of cis-3-hexenol glycoside in spring tea is greater than or equal to 12 mu / g, the content of cis-3-hexenol glycoside in summer tea is 5-11 mu / g, and the content of cis-3-hexenol glycoside in autumn tea is less than or equal to 4 mu / g, and the final determination of the belonging season is made according to the interval to which the measured value belongs.

[0032] The present application at least includes the following beneficial effects:

[0033] Firstly, the present application can determine the belonging season of tea according to objective data by detecting the stable isotopes and organic components related to seasons in tea and combining linear discriminant model for calculation and comparison, thereby reducing the influence of subjective factors in traditional sensory identification, providing a quantifiable and repeatable method for tea season identification, and being suitable for confirming the season source of tea in tea quality control, market supervision and other scenes.

[0034] Secondly, the present application adopts this multi-level judgment mechanism, which can take different verification strategies according to the difference degree of values. For samples with obvious difference, the first level can be used for rapid determination to improve the detection efficiency; for samples with small difference, the second level is used for cooperative verification, and the double information of stable isotopes and characteristic components is combined to reduce the probability of misjudgment. In practical application, this mechanism can improve the accuracy of judgment, and is especially suitable for identifying tea samples picked in the season transition period.

[0035] Thirdly, the secondary verification mechanism can effectively deal with the edge samples with abnormal correlation index, identify the tea leaves in the transition season through the proportional relationship of characteristic components, and make the season determination more accurate. In the actual measurement of tea areas in the south of the Yangtze River, the method can effectively reduce the misjudgment rate of the transition season samples, and clearly determine the attribution of the non-transition season samples, thereby providing a more accurate basis for tracing the tea picking time. For tea processing enterprises, the processing technology can be adjusted according to the determination result of the transition season, so as to avoid the quality fluctuation caused by the ambiguous season attribute of raw materials.

[0036] Fourthly, by establishing a dynamic fault tolerance interval based on historical data, the method can further judge by combining the statistical distribution characteristics of the residual when part of the measured values of the to-be-measured sample exceeds the predicted interval, avoid the data misjudgment caused by accidental error or slight fluctuation, make the data effectiveness determination more flexible and adaptive, and thus provide a more reliable data basis for the subsequent steps of tea season identification.

[0037] Fifthly, by determining the establishment steps of the dynamic fault tolerance interval, including the time span of sample collection, sample quantity requirement, residual calculation method and quantile determination method, the establishment process of the dynamic fault tolerance interval is more standardized and operable, specific and unified standards are provided for the data effectiveness determination, and the consistency and reliability of the data verification in the tea season identification are improved.

[0038] Other advantages, objects and features of the present application will be apparent from the following description, and will be understood by those skilled in the art through the study and practice of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 Figure 1 is a scatter plot of the main component scores of green tea in different seasons according to one of the technical solutions of the present application;

[0040] Figure 2 Figure 3 is a discriminant analysis of stable isotopes and organic components of green tea in different seasons according to one of the technical solutions of the present application. DETAILED DESCRIPTION

[0041] The present application will be further described in detail below with reference to the embodiments, so that those skilled in the art can implement the present application according to the description.

[0042] According to one of the embodiments of the present application, the identification method of tea leaves in different seasons comprises the following steps:

[0043] Step 1, collecting tea samples, detecting the stable isotope ratio and organic component content in the tea samples, wherein the stable isotope is δ¹³C, δ 15N, the organic components are epigallocatechin EGC, epigallocatechin gallate EGCG, epicatechin gallate ECG, caffeine CAF;

[0044] Step two, the detected δ¹³C, δ 15 N ratio (as shown in Table 3) and the content of EGC, EGCG, ECG, CAF are respectively substituted into the following linear discriminant model, and the values of Y 春季 , Y 夏季 , Y 秋季 are calculated:

[0045] Y 春季 = -79.166δ¹³C + 4.269δ 15 N - 0.590EGC - 0.044EGCG - 2.083ECG +9.610CAF - 1221.849;

[0046] Y 夏季 = -75.594δ¹³C + 4.351δ 15 N - 0.664EGC + 0.208EGCG - 1.793ECG +9.124CAF - 1144.902;

[0047] Y 秋季 = -78.922δ¹³C + 4.214δ 15 N - 0.802EGC - 0.011EGCG - 2.107ECG +10.037CAF - 1226.854;

[0048] Step three, according to the values of Y 春季 , Y 夏季 , Y 秋季 to determine the season to which the tea sample belongs.

[0049] In this technical solution, the detection of stable isotope ratio can use an isotope mass spectrometer, the detection of organic component content can use a high-performance liquid chromatograph (HPLC), sample crushing can use a high-speed pulverizer, and filtration can use a microporous filter membrane;

[0050] The working process is as follows: first, collect representative tea samples, remove impurities, and crush to 80 mesh sieve; take the crushed sample, add 7:3 methanol-water solution, ultrasonic extraction for 30 minutes, filter through a microporous filter membrane after extraction is completed to obtain supernatant; divide the supernatant into two parts, one part is injected into a high-performance liquid chromatograph to detect the content of EGC, EGCG, ECG, CAF, and the other part is sent into an isotope mass spectrometer after freeze-drying to detect δ¹³C, δ15 N ratio; the obtained data is respectively substituted into the linear discriminant model of Y spring, Y summer and Y autumn, three Y values are calculated; the size of the three Y values is compared, the corresponding season with the largest value can be determined as the season of the tea sample, or the size of the three values is compared, and the relative difference between the maximum value and the second maximum value is determined;

[0051] By adopting the technical scheme, the application can determine the season of tea according to objective data by detecting stable isotopes and organic components related to seasons in tea, calculating and comparing by using a linear discriminant model, thereby reducing the influence of subjective factors in traditional sensory identification, providing a quantifiable and repeatable method for tea season identification, and being suitable for confirming the season of tea in tea quality control, market supervision and other scenes.

[0052] According to another embodiment of the application, the pretreatment of the tea sample in step one adopts a simultaneous extraction method: after the tea sample is crushed, the tea sample is extracted with a methanol-water solution with a volume ratio of 7:3 for 30 minutes (the ultrasonic power during extraction can be selected as 300W, the extraction temperature is controlled at 25 DEG C, and the pore size of the microporous filter membrane used for filtration can be 0.45 mu m), after filtration, the supernatant is divided into two parts, one part is used for HPLC detection of EGC, EGCG, ECG and CAF, and the other part is used for stable isotope δ 15 N detection by EA-IRMS. An appropriate amount of tea sample is taken, impurities are removed, and the sample is crushed to 80 mesh fineness by a high-speed universal crusher; a certain amount of the crushed sample is weighed and placed in a conical flask, methanol-water solution is added at a volume ratio of 7:3, the conical flask is placed in a numerical control ultrasonic cleaner, the power is set to 300W, the temperature is set to 25 DEG C, and ultrasonic extraction is performed for 30 minutes; after extraction is completed, the extraction liquid is vacuum filtered through a 0.45 mu m organic phase microporous filter membrane to obtain the supernatant; the supernatant is evenly divided into two parts, one part is directly transferred to a sample bottle for high-performance liquid chromatograph detection of the contents of EGC, EGCG, ECG and CAF; the other part is placed in a petri dish and placed in a vacuum freeze dryer for freeze-drying to a constant weight, and the dried sample is loaded into a tin boat and placed in an elemental analyzer-isotope mass spectrometer for detection of δ 15 N ratio.

[0053] By adopting the technical scheme, the simultaneous extraction method of the application can provide qualified sample solutions for the detection of organic components and stable isotopes at the same time, reduce the operation steps of sample pretreatment, avoid errors that may be caused by multiple extractions, ensure the consistency of samples required for the two detections at the same time, improve the detection efficiency and the reliability of data, and provide accurate basic data for subsequent tea season identification.

[0054] According to another embodiment of the application, before the tea leaf pretreatment, the tea leaf origin type mountain tea or plain tea is pre-judged by near-infrared spectroscopy, and an origin correction coefficient K is introduced into the linear discriminant model in step two: when it is mountain tea, Y spring, Y summer and Y autumn are multiplied by 1.05, 1.02 and 1.01 respectively; when it is plain tea, they are multiplied by 0.95, 0.98 and 0.99 respectively. The existing method for distinguishing plain tea and mountain tea by near-infrared spectroscopy is mainly based on the principle that the internal chemical components (such as moisture, amino acids, polysaccharides, etc.) of tea leaves from different origins have different contents due to differences in growth environment, and show specific absorption characteristics in the near-infrared spectroscopy region. Usually, a large number of tea leaf sample spectrum data from known origins (clearly mountain or plain) are collected first, such as using a Fourier transform near-infrared spectrometer, setting the wavelength range to 780-2500 nm, the resolution to 8 cm-1, and the scanning number to 32 times, taking air as the reference, crushing the tea leaves through a 40-mesh sieve, taking an appropriate amount and loading it into a quartz sample cell to scan and obtain the spectrum. Then, chemometrics methods such as successive projections algorithm (SPA) are used to screen characteristic wavelengths related to origin, such as 1450 nm (related to moisture), 1940 nm (related to amino acids), 2200 nm (related to polysaccharides), etc. By analyzing the differences in absorbance of mountain tea and plain tea at these characteristic wavelengths, a discriminant model such as partial least squares discriminant analysis (PLS-DA) is established. When discriminating tea leaves from unknown origins, the near-infrared spectrum is collected, the characteristic wavelength absorbance is extracted and substituted into the model, and according to the output "mountain tea probability" and "plain tea probability", when "mountain tea probability>80%", it is determined as mountain tea, otherwise it is determined as plain tea, thereby realizing the distinction between plain tea and mountain tea.

[0055] By using the technical scheme, the application can quickly pre-judge the origin type by near-infrared spectroscopy, and introduce a targeted correction coefficient, which can reduce the influence of mountain and plain environmental differences (such as altitude, temperature, soil composition) on the detection index, make the calculation result of the linear discriminant model more consistent with the actual growing season of the tea leaves, and further improve the accuracy of the season identification of tea leaves from different origins.

[0056] According to another embodiment of the application, the content of tea saponin TS in the tea leaf sample is also detected in step one (the content of tea saponin TS can be detected by using a UV-visible spectrophotometer, and the unit of TS content is mg / g, and the numerical part is used in the calculation of the linear discriminant model), and a TS correction term is added to the linear discriminant model in step two: Y 春季修正值 = Y 春季 -0.15TS, Y 夏季修正值 = Y 夏季 +0.18TS, Y 秋季修正值 = Y 秋季-0.07TS. In step one, while collecting the tea sample and performing pretreatment, another pulverized tea sample is taken, and a 80% by volume ethanol solution is used to extract the sample in an ultrasonic extraction instrument for 40 minutes, and after filtration, a tea saponin extract is obtained; an appropriate amount of the extract is taken, vanillin-ice acetic acid color reagent is added, and the mixture is reacted in a 60 DEG C water bath for 15 minutes; after cooling, the absorbance is measured at a wavelength of 548nm by using a UV-visible spectrophotometer, and the TS content is calculated according to a standard curve. In step two, after the initial values of Y spring, Y summer and Y autumn are calculated, the TS content is substituted into the corresponding correction formula, for example, when the TS content is 1.2%, the corrected value of Y spring is Y spring-0.15*1.2, the corrected value of Y summer is Y summer+0.18*1.2, and the corrected value of Y autumn is Y autumn-0.07*1.2, and then the sizes of the corrected Y values are compared to determine the season to which the tea belongs.

[0057] By adopting the technical scheme, the present application can compensate for the influence of the difference in tea saponin content in tea leaves in different seasons on the linear discriminant model by introducing a correction term of the tea saponin TS content, so that the result of season identification is more in line with the actual growth of tea leaves, and the accuracy of tea season identification is further improved, and the present application is especially suitable for tea varieties with obvious fluctuations in tea saponin content.

[0058] According to still another embodiment of the present application, the detection of stable isotope ratios and the content of organic components in step one is verified by parallel samples, and each tea sample is detected at least 3 times, and the average value is substituted into the linear discriminant model. When the relative standard deviation of the parallel samples exceeds 5%, the detection is performed again.

[0059] By adopting the technical scheme, the present application can effectively reduce random errors that may exist in a single detection process by verifying the detection of stable isotope ratios and the content of organic components by parallel samples. The detection of each tea sample is performed at least 3 times, and the average value is substituted into the model, so that the detection data is closer to the true value, and the influence of accidental factors on the result is reduced. When the relative standard deviation of the parallel samples exceeds 5%, the detection is performed again, so that systematic deviation or operation errors that may exist in the detection process can be found in time, and the data substituted into the linear discriminant model has high precision and reliability. Thus, the method provides data quality guarantee for the accuracy of subsequent season identification, reduces misjudgment caused by fluctuations in data, and makes the final result of season identification more persuasive and repeatable, and is suitable for tea quality analysis and related research scenarios with high requirements for detection accuracy.

[0060] According to still another embodiment of the present application, Y 春季 , Y 夏季 , Y 秋季The method for determining the season to which the tea sample belongs by the numerical value of Y 春季 , Y 夏季 , Y 秋季 is as follows: first, the numerical values of Y

[0061] First level: calculate the relative range R=(Y max -Y min ) / Y max ×100%, if R≥4%, then determine that the season corresponding to Y max is the season to which the tea sample belongs, wherein Y max is the maximum value obtained by substituting the linear discriminant model, and Y min is the second largest value obtained by substituting the linear discriminant model;

[0062] Second level: if 3%≤R<4%, then use the synergistic verification method of δ¹³C value and seasonal characteristic component: the seasonal characteristic component is tea amino acid Thea in spring, EGCG in summer, and ECG in autumn, calculate the correlation index K of δ¹³C value and seasonal characteristic component content, K=(δ¹³C measured value / δ¹³C seasonal average)×(seasonal characteristic component measured value / seasonal characteristic component average), when 0.9≤K≤1.1, then determine that Y maxThe corresponding season is the season to which the tea sample belongs. The relative range is the difference between the maximum value and the second maximum value in the three seasonal indexes, divided by the maximum value and then multiplied by 100%, used to reflect the dispersion degree between the values. The seasonal characteristic component refers to the representative component in tea leaves in different seasons, wherein the spring is theanine, the summer is epigallocatechin gallate, and the autumn is epicatechin gallate. The correlation index is obtained by multiplying the ratio of the measured stable carbon isotope ratio to the average value of the corresponding season by the ratio of the measured content of the seasonal characteristic component to the average content of the component in the season, and is used to verify the reliability of the seasonal determination result. The relative range is calculated by first subtracting the second maximum value from the maximum value in the three values, then dividing the result by the maximum value, and finally multiplying by 100% to obtain the percentage of the relative range. If the percentage is greater than or equal to 4%, the season corresponding to the maximum value is directly determined as the season of the tea; if the relative range is between 3% and 4%, the cooperative verification link is entered: the stable carbon isotope ratio and the content of the corresponding seasonal characteristic component of the tea sample are measured, and the ratios of the two to the average value of the respective season are calculated, and the two ratios are multiplied to obtain the correlation index. When the correlation index is between 0.9 and 1.1, the season corresponding to the maximum value is determined as the season of the tea. By using this technical scheme, the present application adopts this multi-level judgment mechanism, which can adopt different verification strategies according to the difference between the values. For samples with obvious differences, the first level can be used for rapid determination to improve the detection efficiency; for samples with smaller differences, the second level of cooperative verification is used, and the double information of stable isotope and characteristic component is combined to reduce the probability of misjudgment. In practical application, this mechanism can improve the accuracy of judgment, and is especially suitable for identifying tea samples picked in the transition period between seasons.

[0063] According to another embodiment of the present application, in the second level judgment, when K < 0.9 or K > 1.1, the tea sample enters the secondary verification, and the method of the secondary verification is as follows: calculating Y max The corresponding season and Y min The ratio of the content of the characteristic component between the corresponding seasons, wherein the ratio of the content of theaflavins to the content of EGCG in spring and summer, the ratio of the content of EGCG to the content of ECG in summer and autumn, and the ratio of the content of ECG to the content of theaflavins in autumn and spring. If the ratio is in the transition interval of the corresponding season, the tea sample is determined as a transition season sample; if the ratio exceeds the transition interval of the corresponding season, Y minThe corresponding season determination is the season to which the tea sample belongs. When the correlation index of the second level determination is less than 0.9 or greater than 1.1, secondary verification is started, first, the season corresponding to the maximum value and the season corresponding to the second maximum value are determined, for example, the maximum value is the spring index and the second maximum value is the summer index, the ratio of the content of theanine to the content of epigallocatechin gallate of the tea sample is calculated. Then the ratio is compared with the corresponding transition interval. By adopting the technical scheme, the secondary verification mechanism of the present application can effectively process the edge sample with abnormal correlation index, identify the tea in the transition season through the proportional relationship of the characteristic components, and make the season determination more accurate. In the actual measurement of the tea area in the south of the Yangtze River, the method can effectively reduce the misjudgment rate of the sample in the transition season, and clearly determine the attribution of the sample in the non-transition season, thereby providing a more accurate basis for the tracing of the tea picking time. For the tea processing enterprise, the processing technology can be adjusted according to the determination result of the transition season, so as to avoid the quality fluctuation caused by the ambiguous season attribute of the raw material.

[0064] According to another embodiment of the present application, the average δ 13C seasonal values are as follows: the average δ 13C spring value is -27.1‰, the average δ 13C summer value is -26.38‰, and the average δ 13C autumn value is -26.35‰.

[0065] Seasonal characteristic component average: the summer EGCG characteristic component average is 142.57 (unit: mg / g), the spring theanine characteristic component average is 28 (unit: mg / g), and the autumn ECG characteristic component average is 25.13 (unit: mg / g) (obtained based on tea samples of known seasons);

[0066] The seasonal transition interval is: the transition interval of spring and summer is 1.2-1.8, the transition interval of summer and autumn is 0.8-1.2, and the transition interval of autumn and spring is 2.0-3.0. The data are calculated by collecting historical samples. By adopting the technical scheme, the accuracy of the determination can be effectively improved.

[0067] According to another embodiment of the present application, a third level is further included, and when R < 3%, the following verification is performed:

[0068] The cellulose to lignin content ratio CLR is determined, the spring tea CLR is ≤1.2, the summer tea CLR is 1.3-1.8, and the autumn tea CLR is ≥1.9;

[0069] The total amount of free amino acids FAA is determined by the ninhydrin colorimetric method, the spring tea FAA is ≥4.5%, the summer tea FAA is 2.0-4.4%, and the autumn tea FAA is 3.0-5.0%;

[0070] The potassium / calcium content ratio is detected, the summer tea K / Ca is ≥1.5, and the spring and autumn tea K / Ca is ≤1.4;

[0071] The total flavone content TF is measured, and the TF of the summer tea is greater than or equal to 3.2%, the TF of the spring tea is less than or equal to 2.5%, and the TF of the autumn tea is 2.6-3.1%;

[0072] If the measured value of the index falls into the season corresponding to the maximum value or the season corresponding to the second maximum value, 5 points are given to the corresponding season, otherwise 0 points are given, the full score is 20 points, and the higher the score is, the more the tea is determined to be the season;

[0073] If at least three of the four indexes in one season fall into the corresponding interval, the tea is determined to be the season;

[0074] If the total scores of two seasons are equal, the cis-3-hexenol glycoside content is detected, the cis-3-hexenol glycoside content of the spring tea is greater than or equal to 12 μg / g, the cis-3-hexenol glycoside content of the summer tea is 5-11 μg / g, and the cis-3-hexenol glycoside content of the autumn tea is less than or equal to 4 μg / g, and the final determination of the season is made according to the interval to which the measured value belongs. The cellulose and lignin detection kit and the ultraviolet-visible spectrophotometer can be used for CLR determination; the spectrophotometer is used for FAA determination by using the indene trinitrile colorimetric method; the atomic absorption spectrophotometer is used for K / Ca detection; the aluminum salt colorimetric method and the spectrophotometer are used for TF determination; and the gas chromatography-mass spectrometry (GC-MS) is used for cis-3-hexenol glycoside detection.

[0075] When R is less than 3%, the tea sample to be detected is crushed, and each index is detected according to the corresponding method: the sample is weighed, the cellulose and lignin are extracted by using the kit, the absorbance is measured to calculate CLR; the extract is taken, the indene trinitrile reagent is added for color development, and FAA is measured; the atomic absorption spectrophotometer is used to measure the potassium and calcium concentrations to calculate K / Ca; TF is measured by aluminum salt color development; the sample is extracted by using an organic solvent, and the cis-3-hexenol glycoside content is measured by using GC-MS. The measured values of CLR, FAA, K / Ca and TF are compared with the seasonal intervals, each falling into the corresponding seasonal interval is given 5 points, the full score of the four indexes is 20 points, and the season with a higher score is the determination result; if at least three of the four indexes in a season fall into the interval, the season is directly determined; if the total scores of two seasons are equal, the interval to which the measured value of cis-3-hexenol glycoside belongs is compared, and the season corresponding to the interval is finally determined.

[0076] By using the technical scheme, the four auxiliary indexes with high correlation with seasons and cis-3-hexenol glycoside are introduced as supplements, the problem that it is difficult to accurately determine the season when the Y value is close to each other is solved, the tea season identification can still obtain reliable results according to multi-dimensional characteristics in the case of fuzzy data, and the adaptability and determination accuracy of the method to complex samples are improved.

[0077] According to another embodiment of the present application, δ¹³C, δ 15After the N ratio and the content data of EGC, EGCG, ECG and CAF are obtained and before being substituted into the linear discriminant model, the following steps are further included:

[0078] Based on the tea sample data of known seasons, a quantitative relationship model between the δ 15 N ratio and the content of CAF is established, and the first deviation interval of the quantitative relationship model corresponding to the δ 15 N ratio and the content of CAF is determined. 15 The second deviation interval of the quantitative relationship model corresponding to the δ

[0079] For the sample to be tested:

[0080] The δ

[0081] The δ 15 N value is substituted into the δ 15 N-CAF quantitative relationship model to obtain the predicted value of CAF, and the second prediction interval is generated based on the second deviation interval.

[0082] If the measured value of EGCG and the measured value of CAF of the sample to be tested are both located in the corresponding prediction interval, it is judged that the data is valid, and the model is continued to be substituted and calculated, otherwise, it is judged that the data is abnormal, and the sample needs to be re-detected before being discriminated.

[0083] By establishing the quantitative relationship model and the deviation interval of the known sample, the effectiveness of the detection data can be verified before the season discrimination, the wrong discrimination caused by the detection error or the abnormal sample is reduced, the reliability of the data in the tea season discrimination process is improved, the season judgment result based on the linear discriminant model is more persuasive, and the method system of the tea season discrimination is further improved.

[0084] According to another embodiment of the present application, when the quantitative relationship model between the δ 15 N ratio and the content of CAF is established, the sample data of tea of known seasons is used to calculate and determine the quantitative relationship model of each season, specifically:

[0085] Based on the tea sample data of known spring, summer and autumn, the quantitative relationship model between the δ 15 N ratio and the content of CAF is calculated and determined, and the first deviation interval of the quantitative relationship model corresponding to the δ 15The second deviation interval of the quantitative relationship model corresponding to the N ratio and the CAF content;

[0086] For the sample to be detected:

[0087] Substitute the δ¹³C value of the sample into the δ¹³C-EGCG quantitative relationship model of all seasons to generate the corresponding EGCG predicted value of each season and generate the corresponding first prediction interval; substitute the δ 15 N value into the δ 15 N-CAF quantitative relationship model to generate the corresponding CAF predicted value of each season and generate the corresponding second prediction interval;

[0088] Compare the measured values of EGCG and CAF of the sample to be detected with the first prediction interval and the second prediction interval corresponding to the same season, if the measured values of EGCG and CAF of the sample to be detected can match at least one season, so that the measured value of EGCG of the sample to be detected falls within the first prediction interval of the season and the measured value of CAF falls within the second prediction interval of the season, then the data is determined to be valid.

[0089] First, collect tea samples of known spring, summer and autumn, detect the δ¹³C, δ 15 N ratio and the contents of EGCG and CAF of each sample, input these data into a computer, and use statistical analysis software to respectively establish the quantitative relationship model (such as a linear regression model) of δ¹³C and EGCG, δ 15 N and CAF, and calculate the standard deviation of the model, and determine the range of ±2 times the standard deviation as the first deviation interval and the second deviation interval; for the tea sample to be detected, similarly detect its δ¹³C, δ 15 N ratio and the contents of EGCG and CAF; substitute the δ¹³C value of the sample to be detected into the δ¹³C-EGCG quantitative relationship model to obtain the predicted value of EGCG, and generate the first prediction interval in combination with the first deviation interval, substitute the δ 15 N value into the δ 15 N-CAF quantitative relationship model to obtain the predicted value of CAF, and generate the second prediction interval in combination with the second deviation interval; if the measured value of EGCG of the sample to be detected is within the first prediction interval and the measured value of CAF is within the second prediction interval (of the same season), then the data is determined to be valid, which can be substituted into the linear discriminant model in step two for season identification; if any measured value is not within the corresponding prediction interval, then the data is determined to be abnormal and the sample needs to be re-detected;

[0090] For example: known tea samples of spring, summer and autumn in a certain region for two consecutive years, 150 samples for each season, first detect the δ¹³C, δ 15The N ratio and EGCG and CAF contents were used. The δ¹³C values ​​and corresponding EGCG contents of the spring samples were input into the computer, and linear regression was performed using statistical software to obtain a quantitative relationship model of δ¹³C-EGCG, for example, EGCG predicted value = 0.5 × δ¹³C + 20 (the δ¹³C ratio is used for calculation when substituting into the formula, as shown in Table 3). Then, the standard deviation between the measured and predicted EGCG values ​​of all spring samples under this model was calculated, and the first deviation interval was determined to be the predicted value ± 0.8%. Similarly, the δ¹³C values ​​of the spring samples were used to calculate the standard deviation between the measured and predicted EGCG values. 15 Establish δ based on N value and CAF content 15 The N-CAF quantitative relationship model, such as CAF predicted value = 0.3 × δ 15 N + 1.2 (δ) 15 The N ratio is calculated using the numerical part when substituted into the formula. After calculating the standard deviation, the second deviation interval is determined to be ±0.3% of the predicted value. The same steps are followed to establish the models and deviation intervals for summer and autumn. The tested tea sample had a δ¹³C of -26‰ and a δ... 15 N is 3‰, the measured value of EGCG is 10%, and the measured value of CAF is 3%. First, substitute δ¹³C = -26‰ into the δ¹³C-EGCG model to calculate the predicted value of EGCG as 0.5 × (-26) + 20 = 7. Combine this with the first deviation interval to generate the first prediction interval as 7 ± 0.8%, i.e., 6.2% - 7.8%. Then, δ... 15 Substituting N=3‰ into δ 15 The N-CAF model yields a CAF predicted value of 0.3 × 3 + 1.2 = 2.1. Combining this with the second deviation interval, a second prediction interval of 2.1 ± 0.3% (1.8% - 2.4%) is generated. Next, comparing the measured values ​​with the prediction intervals, the measured EGCG value of the tested sample (10%) is outside the 6.2% - 7.8% range, and the measured CAF value (3%) is also outside the 1.8% - 2.4% range. In this case, the data is considered abnormal, and the tea sample needs to be retested. If the measured EGCG value of another tested sample is 7.5%, falling within the first prediction interval, and the measured CAF value is 2.2%, falling within the second prediction interval, then the data is considered valid and can be substituted into the linear discriminant model for further seasonal identification.

[0091] By adopting this technical solution, the present invention establishes a quantitative relationship model and deviation range for different seasons, making the validity judgment of the test sample data more consistent with the component characteristics of tea in different seasons, reducing data misjudgment caused by seasonal differences, further improving the accuracy of data validity verification, providing more reliable input data for the subsequent calculation of the linear discriminant model, and helping to improve the overall accuracy of tea seasonal identification.

[0092] According to still another embodiment of the present application, for a sample to be tested, the measured values of EGCG and CAF of the sample to be tested are compared with the prediction interval corresponding to any season, when one of the measured values is located in the prediction interval of a season, and the other measured value exceeds the prediction interval of the season, and the measured values do not satisfy the prediction of any season, if the following conditions are satisfied, the data is determined to be valid:

[0093] Based on the actual fluctuation of the known historical tea sample data δ¹³C and EGCG, δ 15 N and CAF of spring, summer and autumn, the dynamic fault tolerance interval of different seasons is established by using statistical methods;

[0094] For a sample to be tested, the season satisfied by the measured value of EGCG or CAF is first determined, and then whether the residual of the measured value exceeding the corresponding prediction interval and the predicted value is located in the dynamic fault tolerance interval of the corresponding season is determined; if it is located in the dynamic fault tolerance interval, the data is determined to be valid; otherwise, the data is determined to be abnormal. First, collect 120 tea samples in each season in three consecutive years, and detect and extract the δ¹³C and EGCG data pairs, δ 15 N and CAF data pairs of each sample; for the samples of each season, the residual ΔEGCG of the measured value of EGCG and the predicted value, and the residual ΔCAF of the measured value of CAF and the predicted value are calculated; for the residual sequences of ΔEGCG and ΔCAF of each season, the 0.5% and 99.5% quantiles are calculated by using statistical analysis software to determine the dynamic fault tolerance interval of each season, such as the dynamic fault tolerance interval of spring ΔEGCG is -1.2% to 1.3%. For a sample to be tested, the measured values of EGCG and CAF are first detected, and then substituted into the prediction model of each season, if it is found that the measured value of EGCG satisfies the prediction interval of spring, and the measured value of CAF exceeds the prediction interval of spring, the residual of the measured value of CAF and the predicted value of spring is calculated; it is checked whether the residual is located in the dynamic fault tolerance interval of CAF of spring, if it is located in the interval, the data is determined to be valid, and can be used for season identification; if the residual exceeds the interval, the data is determined to be abnormal, and needs to be re-detected.

[0095] By establishing a dynamic fault tolerance interval based on historical data, the present application can further judge by combining the statistical distribution characteristics of the residual when part of the measured values of the sample to be tested exceed the prediction interval, avoid data misjudgment caused by accidental error or slight fluctuation, make the judgment of data validity more flexible and adaptive, and thus provide a more reliable data basis for the subsequent steps of tea season identification.

[0096] According to another embodiment of the present application, the establishment of the dynamic fault tolerance interval comprises the following steps:

[0097] A1, collect tea sample data of known seasons in at least 3 consecutive planting years, the sample quantity of each season is not less than 100, and δ 13 C and EGCG data pairs, δ 15 N and CAF data pairs;

[0098] A2, calculate the residual ΔEGCG of the measured value and the predicted value of EGCG of all samples in each season, and the residual ΔCAF of the measured value and the predicted value of CAF;

[0099] A3, take the 0.5% to 99.5% quantile of the residual sequence as the boundary value of the dynamic fault tolerance interval for the residuals ΔEGCG and ΔCAF of each season. First, collect tea samples of known spring, summer and autumn in 4 consecutive planting years, 120 samples in each season, remove impurities and process according to the foregoing method, detect the δ 15 N and CAF data pairs; for the samples of each season, calculate the difference value of the measured value and the predicted value of EGCG to obtain ΔEGCG, and the difference value of the measured value and the predicted value of CAF to obtain ΔCAF; arrange the ΔEGCG and ΔCAF of each season into residual sequences respectively, import into statistical analysis software, and calculate the 0.5% quantile and the 99.5% quantile of each residual sequence, for example, the 0.5% quantile of the spring ΔCAF residual sequence is -1.3%, and the 99.5% quantile is 1.5%, so the dynamic fault tolerance interval of spring CAF is -1.3% to 1.5%.

[0100] By adopting the technical scheme, the present application clearly defines the establishment steps of the dynamic fault tolerance interval, including the time span of sample collection, sample quantity requirement, residual calculation method and quantile determination method, so that the establishment process of the dynamic fault tolerance interval is more standardized and operable, specific and unified standards are provided for the judgment of data effectiveness, and the consistency and reliability of data verification in tea season identification are improved.

[0101] <Embodiment 1>

[0102] The main production areas selected in the test are Sichuan Province with geographical indication product identity, specifically, Ya'an City, Yibin City and Dujiangyan City, and in 2024, 5 tea gardens are randomly selected in each region in spring (March), summer (June) and autumn (October) for sample collection, about 250 g of immature "bud leaves" (including one bud and two leaves) are collected at each sampling point, and the tea stable isotopes (δ 15N) and organic components (EGC, EGCG, ECG, CAF).

[0103] 1. Differences in stable isotopes and organic components in green tea from different seasons

[0104] Table 1 shows the average and standard deviation of mineral element content in tea leaves from different seasons. It is worth noting that, except for nitrogen isotopes (δ¹⁸O₂),... 15 In addition to N), carbon isotopes (δ) 13 C) and four organic components showed significant differences across different seasons. p <0.05). Furthermore, tea leaves from different seasons exhibit different isotopic characteristics, δ... 13 C and δ 15 The nitrogen isotope ratios are highest in autumn; for organic components, EGC is highest in spring tea; EGCG and ECG contents reach their peak in summer tea, while CAF content is higher in autumn tea. Therefore, it is speculated that the significant differences in various indicators in different seasons may be due to the influence of factors such as sunlight, precipitation, and temperature.

[0105] Table 1. Stable isotope ratios and organic component content of green tea from different seasons

[0106] Element Spring Summer Autumn delta 13 C (‰) -27.1 ± 1.06 b ]] -26.38 ± 0.59 a ]] -26.35 ± 0.74 a ]] 15 N (‰) ​ -0.12 ± 1.44 ab ]] -0.36 ± 3.55 b ]] 1.19 ± 2.92 a ]] EGC 71.03 ± 19.98 a ]] 58.7 ± 16.29 b ]] 30.85 ± 15.33 c ]] EGCG 123.30 ± 6.68 b ]] 142.57 ± 12.05 a ]]> 121.32 ± 14.81 b ]] ECG 29.70 ± 7.13 b ]] 39.35 ± 11.06 a ]] 25.13 ± 5.78 c ]] CAF 43.48 ± 7.47 ab ]]> 40.85 ± 3.48 b ]] 46.18 ± 4.70 a ]]

[0107] Note: Different lowercase letters in the same line indicate significant differences between different places of origin. p <0.05), the units for EGCG content, ECG content, EGC content, and CAF content are mg / g.

[0108] 2. Principal Component Analysis

[0109] Principal component analysis (PCA) can fit multiple variables into a few principal components through dimensionality reduction, thereby preserving the information of the original variables to the greatest extent. To further explore the characteristic relationships between stable isotopes and organic components under the influence of different seasons, six indicators (δ¹³C, ... 15 Principal component analysis was performed on N, EGC, EGCG, ECG, and CAF. The results are as follows: Figure 1 As shown, the first principal component (PC1) explains 34.9% of the total variance, while the second principal component (PC2) and the third principal component (PC3) account for 23.4% and 14.6% respectively. The three principal components contribute 72.9% of the variance, indicating that these three principal components can express the information on the content of stable isotopes and organic components in green tea, and can be used for seasonal traceability of green tea.

[0110] 3. Seasonal determination of green tea based on linear discriminant analysis

[0111] To verify the use of stable isotopes (δ¹³C, δ¹³C, δ¹³C) 15N) and four organic components (EGC, EGCG, ECG, CAF) to distinguish the green tea from different seasons. Linear discriminant models were established based on the data of the six indexes, which effectively realized the discrimination of the season of origin of green tea. Figure 2 To evaluate the robustness of the model, cross-validation method was used to classify the tea samples from different seasons. The results are shown in Table 2. The correct classification rates of the initial discriminant of spring, summer and autumn were 100%, 85.2% and 96.3%, respectively. The cross-validation results showed that the classification accuracy of the model reached 100% (27 / 27) for spring tea samples, and 81.5% (22 / 27) and 96.3% (26 / 27) for summer and autumn samples, respectively. In summary, the overall classification accuracy of the model reached 92.6%.

[0112] Table 2 Linear discriminant analysis results of green tea from different seasons

[0113] Season Spring Summer Autumn Total Initial discrimination Count Spring 27 0 0 27 Summer 4 23 0 27 Autumn 1 0 26 27 % 100 85.2 96.3 93.8 Cross-discrimination Spring 27 0 0 27 Summer 5 22 0 27 Autumn 1 0 26 27 % 100 81.5 96.3 92.6

[0114] Based on the results of linear discriminant analysis, the discriminant model of different seasons was obtained as follows:

[0115] Y 春季 = -79.166 δ¹³C + 4.269 δ 15 N - 0.590 EGC -0.044 EGCG - 2.083 ECG +9.610 CAF - 1221.849 (1)

[0116] Y 夏季 = -75.594 δ¹³C + 4.351 δ 15 N - 0.664 EGC + 0.208 EGCG - 1.793 ECG +9.124 CAF -1144.902 (2)

[0117] Y 秋季 = -78.922 δ¹³C + 4.214 δ 15 N - 0.802 EGC - 0.011 EGCG - 2.107 ECG +10.037 CAF -1226.854 (3)

[0118] 4. Verification

[0119] In the second year, nine fresh tea samples from spring, summer and autumn were collected, and the stable isotopes (δ¹³C, δ 15N) and 4 organic components (EGC, EGCG, ECG, CAF) are composed, and the size values of Y 春季 , Y 夏季 and Y 秋季 are calculated by substituting the above established discriminant models (1), (2) and (3). The calculation results are shown in Table 3:

[0120] Table 3 is the verification result

[0121] Season δ13C 15 N]]> ​ EGC EGCG ECG CAF

[00014] Y 春季 ]]

[00014] Y 夏季 ]]

[00014] Y 秋季 ]] Spring-1 -28.01 -1.13 79.55 134.45 39.5 31.8 1161.24 1159.09 1157.81 Spring-2 -28.87 -1.39 83.2 116.6 41.15 27.4 1181.12 1163.91 1172.85 Spring-3 -28.11 -1.28 77.9 118.6 40.95 34.4 1192.15 1177.44 1187.75 Summer-1 -27.38 -1.88 61.2 163.5 58.45 40.2 1162.03 1172.16 1158.20 Summer-2 -27.50 -1.43 51.2 171.65 62.25 39.8 1176.67 1180.70 1172.19 Summer-3 -27.50 -1.48 49.15 169.7 48.9 38.2 1187.86 1190.79 1186.56 Autumn-1 -26.39 -2.24 29.65 127 23.95 38.8 1149.66 1155.69 1160.38 Autumn-2 -26.52 -2.19 29.65 125.45 24.1 39.3 1166.07 1162.58 1175.55 Autumn-3 -26.23 -2.29 31.25 127.6 24.15 43.4 1181.46 1176.12 1192.01

[0122] According to the calculation results, the Y 春季 values of the 3 spring samples are all greater than the Y 夏季 and Y 秋季 values, the Y 夏季 values of the 3 summer samples are all greater than the Y 春季 and Y 秋季 values, the Y 秋季 values of the 3 autumn samples are all greater than the Y 春季 and Y 夏季 values, indicating that all the samples can be accurately discriminated.

[0123] Although the embodiments of the present application have been disclosed as above, it is not limited to the application listed in the specification and the embodiments, and it can be fully applied to various fields suitable for the present application, and other modifications can be easily realized by those skilled in the art, and therefore the present application is not limited to the specific details and the examples shown and described herein, without departing from the general concept defined by the claims and the equivalent scope.

Claims

1. A method for identifying tea leaves from different seasons, characterized in that, Includes the following steps: Step 1: Collect tea samples and test the stable isotope ratios and organic component content in the tea samples, wherein the stable isotopes are δ¹³C and δ¹³C. 15 N, the organic components are epigallocatechin gallate (EGC), epigallocatechin gallate (EGCG), epigallocatechin gallate (ECG), and caffeine (CAF); Step 2: Measure the obtained δ¹³C and δ... 15 Substituting the N ratio and the contents of EGC, EGCG, ECG, and CAF into the following linear discriminant model, Y is calculated. 春季 Y 夏季 Y 秋季 The value: Y 春季 = -79.166δ¹³C + 4.269δ 15 N - 0.590EGC - 0.044EGCG - 2.083ECG + 9.610CAF- 1221.849; Y 夏季 = -75.594δ¹³C + 4.351δ 15 N - 0.664EGC + 0.208EGCG - 1.793ECG + 9.124CAF- 1144.902; Y 秋季 = -78.922δ¹³C + 4.214δ 15 N - 0.802EGC - 0.011EGCG - 2.107ECG +10.037CAF - 1226.854; Step 3: Based on Y 春季 Y 夏季 Y 秋季 The numerical value determines the season of the tea sample; Before pre-processing the tea leaves, the tea's origin type is first determined using near-infrared spectroscopy: mountain tea or plain tea. Then, an origin correction coefficient K is introduced into the linear discriminant model in step two: when it is mountain tea, Y... 春季 Y 夏季 Y 秋季 Multiply by 1.05, 1.02, and 1.01 respectively; for flatland tea, multiply by 0.95, 0.98, and 0.99 respectively. In step one, the tea saponin (TS) content of the tea sample also needs to be detected, and a TS correction term needs to be added to the linear discriminant model in step two: Y 春季修正值 = Y 春季 -0.15TS, Y 夏季修正值 = Y 夏季 +0.18TS, Y 秋季修正值 = Y 秋季 -0.07TS; According to Y 春季 Y 夏季 Y 秋季 The method for determining the season of a tea sample based on numerical values ​​is to first compare Y... 春季 Y 夏季 Y 秋季 The numerical value is then compared, and a multi-level judgment mechanism is executed: First level: Calculate the relative range R=(Y max -Y min ) / Y max ×100%, if R≥4%, then determine Y. max The corresponding season is the season to which the tea sample belongs, where Y max To obtain the maximum value by substituting into the linear discriminant model, Y min This is the second largest value obtained by substituting it into the linear discriminant model; Second level: If 3% ≤ R < 4%, then the co-verification method of δ¹³C value and seasonal characteristic components is adopted: among the seasonal characteristic components, theanine (Thea) is in spring, EGCG is in summer, and ECG is in autumn. Calculate the correlation index K between δ¹³C value and the content of seasonal characteristic components, K = (measured value of δ¹³C / seasonal mean of δ¹³C) × (measured value of seasonal characteristic component / seasonal mean of seasonal characteristic component). When 0.9 ≤ K ≤ 1.1, then Y is determined. max The corresponding season is the season to which the tea sample belongs.

2. The method for identifying tea leaves from different seasons as described in claim 1, characterized in that, The pretreatment of the tea sample in step one uses a simultaneous extraction method: the pulverized tea sample is ultrasonically extracted with a methanol-water solution at a volume ratio of 7:3 for 30 minutes. After filtration, the supernatant is divided into two portions. One portion is used for HPLC detection of EGC, EGCG, ECG, and CAF, and the other portion is freeze-dried for stabilizing isotopes δ¹³C and δ¹³C. 15 EA-IRMS detection of N.

3. The method for identifying tea leaves from different seasons as described in claim 1, characterized in that, In step one, parallel sample verification is performed for the detection of stable isotope ratios and organic component content. Each tea sample is tested at least 3 times in parallel, and the average value is substituted into the linear discriminant model. When the relative standard deviation of the parallel sample exceeds 5%, the test is repeated.

4. The method for identifying tea leaves from different seasons as described in claim 1, characterized in that, In the second-level judgment, when K < 0.9 or K > 1.1, the tea sample enters the secondary verification. The method for secondary verification is: calculate Y. max The corresponding season and Y min The ratio of characteristic component content between corresponding seasons is calculated as follows: Thea content / EGCG content for spring and summer, EGCG content / ECG content for summer and autumn, and ECG content / Thea content for autumn and spring. If this ratio falls within the transition range of the corresponding season, the tea sample is classified as a transitional season sample. If the ratio exceeds the transition range of the corresponding season, it is classified as a Y-type sample. min The corresponding season is determined by the season to which the tea sample belongs.

5. The method for identifying tea leaves from different seasons as described in claim 4, characterized in that, The seasonal mean values ​​of δ¹³C are as follows: δ¹³C in spring is -27.1‰, δ¹³C in summer is -26.38‰, and δ¹³C in autumn is -26.35‰. Mean values ​​of seasonal characteristic components: 142.57 for EGCG characteristic components in summer, 28 for theanine characteristic components in spring, and 25.13 for ECG characteristic components in autumn; The seasonal transition ranges are as follows: 1.2 to 1.8 for spring and summer, 0.8 to 1.2 for summer and autumn, and 2.0 to 3.0 for autumn and spring.

6. The method for identifying tea leaves from different seasons as described in claim 5, characterized in that, It also includes a third level, where if R < 3%, the following verification is performed: The cellulose to lignin ratio (CLR) was determined. For spring tea, the CLR was ≤1.2; for summer tea, the CLR was 1.3-1.8; and for autumn tea, the CLR was ≥1.

9. The total free amino acid (FAA) was determined using the ninhydrin colorimetric method. Spring tea had an FAA content ≥ 4.5%, summer tea 2.0-4.4%, and autumn tea 3.0-5.0%. The potassium / calcium ratio was tested; for summer tea, K / Ca ≥ 1.5, and for spring and autumn tea, K / Ca ≤ 1.

4. The total flavonoid content (TF) was determined, with summer tea having a TF ≥ 3.2%, spring tea having a TF ≤ 2.5%, and autumn tea having a TF of 2.6-3.1%. If the measured value of the indicator falls into the season corresponding to the largest value or the season corresponding to the second largest value, then the corresponding season is added 5 points; if it does not fall into either, then it is 0 points. The full score is 20 points. The season with the highest score is tea. If at least three of the four indicators for a particular season fall within the corresponding range, then it is determined that the season is tea. If the total scores for two seasons are equal, the content of cis-3-hexenol glycosides will be tested. For spring tea, the content should be ≥12 μg / g, for summer tea, it should be 5-11 μg / g, and for autumn tea, it should be ≤4 μg / g. The season will be determined based on the range of the measured values.

Images