Fine quantification method for mangrove plant source contribution in mangrove forest wetland sediment carbon sink

Abstract

The invention provides a fine quantification method for mangrove plant source contribution in mangrove forest wetland sediment carbon sink. The method comprises the following steps: acquiring mangrove forest wetland sediment sample data; constructing a multi-source observation data coupling model; total organic carbon constraint is adopted to calibrate isotope mass balance; estimating prior distribution of a source contribution proportion by using Dirichlet; the posterior distribution of the source contribution ratio is estimated through MCMC sampling. According to the invention, the monomer delta 13C, the concentration and the total organic carbon delta 13C are fused, so that the limitation of a traditional single data source is solved; and the technical problem of mangrove forest carbon source signal overlapping is solved. Connecting isotope signals with different scales by using a fractionation constraint mechanism, and eliminating independent estimation deviation of end member parameters through a total organic carbon constraint mechanism; meanwhile, end member variability, analysis errors and system errors are considered, and the quantification accuracy is further improved; according to the method, the sampling deviation is reduced through probabilistic contribution proportion estimation and end member parameter posteriori estimation.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological environment protection, and in particular to a fine quantification method for the contribution of mangrove plant sources to carbon sinks in mangrove wetland sediments. Background Art

[0002] Mangrove wetlands have a significant carbon sink function and are a crucial link in the global carbon cycle. Accurately identifying and quantifying the carbon source composition of mangrove wetland carbon pools is key to assessing ecosystem services and carbon sequestration potential.

[0003] The existing technologies mainly use stable isotope analysis and total organic carbon δ 13 C TOC However, the current traditional stable isotope analysis is limited by the overall isotope signal and cannot accurately distinguish the specific carbon source; and the traditional total organic carbon δ 13 C TOC This method cannot distinguish between isomeric carbon inputs. Compound-specific stable isotope analysis is limited to the analysis of individual compounds, making it difficult to effectively probabilistically quantify the uncertainty of source contributions. Pure isotope analysis ignores the systematic constraints of total organic carbon and does not consider the complex structure of analytical and systematic errors. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides a method for finely quantifying the contribution of mangrove plant sources to the sediment carbon sink in mangrove wetlands. The present invention organically couples the high-resolution traceability capability of compound-specific stable isotope analysis technology with the probabilistic quantification of the Bayesian mixture model to finely quantify the contribution ratio of mangrove plant-derived carbon to the sediment carbon sink and its uncertainty range.

[0005] The technical solution of the present invention is: a method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments, comprising the following steps:

[0006] S1), obtaining mangrove wetland sediment sample data;

[0007] S2), building a multi-source observation data coupling model;

[0008] S3), calibrating the isotope mass balance using total organic carbon constraints;

[0009] S4), using Dirichlet to estimate the prior distribution of source contribution ratio;

[0010] S5) Estimate the posterior distribution of source contribution ratio through MCMC sampling.

[0011] Preferably, in step S1), the mangrove wetland sediment sample data includes: monomer isotope δ 13C CSIA , total organic carbon isotope δ 13 C TOC , and the biomarker concentrations C of each source conc .

[0012] Preferably, in step S1), the biomarkers include lignin phenols and long-chain alkanes.

[0013] Preferably, in step S2), the expression of the multi-source observation data coupling model is:

[0014]

[0015] Where, δ 13 C CSIA represents monomeric isotopes; C conc Biomarker concentrations from various sources; δ 13 C TOC Total organic carbon isotope; K represents the number of sources; f k represents the contribution ratio of source k, ∑f k =1;δ 13 C CSIA,k , δ 13 C TOC,k 、C conc,k are all end member parameters, among which δ 13 C CSIA,k Represents the monomeric isotope δ of source k 13 C value; δ 13 C TOC,k represents the total organic carbon isotope value of source k; C conc,k represents the biomarker concentration of source k; ε δ,CSIA , ε C , ε δ,TOC Represents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC Observation error.

[0016] Preferably, in step S2), under the Bayesian framework, the observation error is assumed to consist of analysis error and system error, thus obtaining:

[0017]

[0018] Where, N(0,σ 2 ) indicates that the parameter has a mean of 0 and a variance of σ 2 Normal distribution; σ δ,CSIA , σ C , σ δ,TOCRepresents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC The systematic error to be estimated; σ δ,CSIA,anal , σ C,anal , σ δ,TOC,anal Represents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC known analytical errors.

[0019] Preferably, in step S2), the end member parameters are expressed as:

[0020]

[0021] Where, δ 13 C CSIA,k Represents the monomeric isotope δ of source k 13 C value; δ 13 C TOC,k represents the total organic carbon isotope value of source k; C conc,k represents the biomarker concentration of source k; N(μ,τ 2 ) indicates that the parameters have a mean of μ and a variance of τ 2 Normal distribution of μ δ,CSIA,k 、μ δ,TOC,k 、μ C,k Represents δ 13 C CSIA,k , δ 13 C TOC,k 、C conc,k The prior mean of τ δ,CSIA,k , τ δ,TOC,k , τ C,k represent the monomer isotope δ of sample n 13 Analytical errors of C values, biomarker concentrations, and total organic carbon isotope values.

[0022] Preferably, in step S3), the isotope mass balance is constrained using the total organic carbon constraint, specifically:

[0023] δ 13 C TOC,k =δ 13 C CSIA,k +Δ k +β k ;

[0024] Where, Δ krepresents the intrinsic fractionation value between the biomarker and total organic carbon of source k; β k It represents the allowable range of biological variation and is a constraint item for total organic carbon.

[0025] As a preference, in step S3), the inherent fractionation value Δ k and the total organic carbon constraint term β k Respectively expressed as:

[0026] Δ k =δ 13 C TOC -δ 13 C CSIA ;

[0027]

[0028] Where, Represents β k The mean is 0 and the variance is σ β,k Normal distribution; σ β,k Represents the variance of the biological variation value of source k.

[0029] Preferably, in step S4), the prior distribution of the source contribution ratio is estimated using Dirichlet, which is expressed as:

[0030] f k ~Dirichlet(α k );

[0031] Among them, α k is a hyperparameter; if there is no information prior, α k =1.

[0032] Preferably, in step S5), the posterior distribution of the source contribution ratio is estimated by MCMC sampling, specifically:

[0033] S51), according to Bayes' theorem, the posterior distribution is obtained based on the likelihood function of the prior distribution and the observed data;

[0034] S52) Use Hamiltonian Monte Carlo algorithm to perform posterior distribution sampling and iteration. After T iterations, the source contribution ratio f is obtained. k The posterior distribution of , and output the source contribution ratio f k The posterior mean and 95% confidence interval of .

[0035] Preferably, in step S51), the posterior distribution is obtained based on the likelihood function and the prior distribution, that is:

[0036] P(θ|D)∝L(D|θ)·P(θ);

[0037] Where θ = f, δ 13 C CSIA,k ,C conc,k ,δ 13 C TOC,k ,σ δ,CSIA ,σ C ,σ δ,TOC are all parameters to be estimated; D = δ 13 C CSIA ,C conc ,δ 13 C TOC is the observed data; L(D|θ) is the likelihood function; P(θ) is the prior distribution; P(θ|D) is the posterior distribution.

[0038] Preferably, in step S51), the likelihood function L(D|θ) is expressed as:

[0039]

[0040] Where, N(μ,σ 2 ) indicates that the parameter has a mean of μ and a variance of σ 2 Normal distribution of δ 13 C CSIA,n 、C cone,n , δ 13 C TOC,n represent the monomer isotope δ of sample n 13 C value, biomarker concentration, total organic carbon isotope value; σ δ,CSIA,anal,n , σ C,anal,n , σ δ,TOC,anal,n represent the monomer isotope δ of sample n 13 Analytical errors of C values, biomarker concentrations, and total organic carbon isotope values.

[0041] The beneficial effects of the present invention are:

[0042] 1. The present invention converts monomer δ 13 C, concentration and total organic carbon δ 13 C is integrated to overcome the limitations of traditional single data sources and the technical difficulties of overlapping mangrove carbon source signals;

[0043] 2. The present invention utilizes a fractionation constraint mechanism to connect isotope signals of different scales and eliminates the independent estimation bias of end-member parameters through a total organic carbon constraint mechanism;

[0044] 3. The present invention simultaneously considers end-member variability, analytical error and systematic error, further improving the accuracy of quantification;

[0045] 4. The present invention provides a probabilistic contribution ratio estimation and reduces sampling bias through posterior estimation of end-member parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 Schematic diagram of the process of the present invention;

[0047] Figure 2 This is a source contribution ratio trace diagram in an embodiment of the present invention;

[0048] Figure 3 Schematic diagram of analysis results of multi-source observation data coupling in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings:

[0050] like Figure 1 As shown, this embodiment provides a method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments, comprising the following steps:

[0051] S1), obtaining mangrove wetland sediment sample data;

[0052] In this embodiment, mangrove wetland sediment samples and end-member samples were collected, freeze-dried, and then ground through a 100-mesh sieve; organic carbon in the samples was separated and extracted, and specific biomarkers were purified;

[0053] In this embodiment, the total organic carbon isotope δ 13 C TOC GC-IRMS was used to determine the isotope δ 13 C CSIA .

[0054] In this example, the lignin phenolic monomer Λ8 was selected as a biomarker. Λ8 is a core indicator for quantifying the total amount of lignin in sediments, representing the total concentration of eight key phenolic monomers. Λ8 degrades macromolecular lignin polymers in sediment samples into small phenolic monomers through alkaline oxidative hydrolysis. After derivatization, these monomers are quantitatively analyzed by gas chromatography-mass spectrometry (GC-MS). The total organic carbon content of the sediment samples was also measured. The names and groupings of the lignin phenolic monomers Λ8 are shown in Table 1:

[0055] Table 1 Names and groups of lignin phenolic monomers Λ8

[0056]

[0057]

[0058] Wherein, the calculation formula of lignin phenolic monomer Λ8 is:

[0059] Λ8=∑(V+S+P);

[0060] Wherein, V, S, and P represent different components of different lignin phenols in Table 1, respectively.

[0061] The observation data are shown in Table 2:

[0062] Table 2 Observation data of mangrove wetland sediment samples

[0063] sample <![CDATA[δ 13 C CSIA ]]> <![CDATA[C conc ]]> <![CDATA[δ 13 C TOC ]]> <![CDATA[σ δ,CSIA,anal ]]> <![CDATA[σ C,anal ]]> <![CDATA[σ δ,TOC,anal ]]> 1 -28.28 102.53 -26.43 0.12 1.8 0.18 2 -28.12 99.71 -25.85 0.12 1.8 0.18 3 -27.22 99.57 -26.15 0.12 1.8 0.18 4 -27.96 113.69 -26.21 0.12 1.8 0.18 5 -27.94 97.74 -26.57 0.12 1.8 0.18 6 -27.14 115.16 -26.03 0.12 1.8 0.18 7 -27.77 84.51 -26.47 0.12 1.8 0.18 8 -28.63 105.85 -27.12 0.12 1.8 0.18 9 -28.34 101.24 -26.23 0.12 1.8 0.18 10 -28.22 102.16 -25.45 0.12 1.8 0.18

[0064] S2), building a multi-source observation data coupling model;

[0065] In this embodiment, the expression of the multi-source observation data coupling model is:

[0066]

[0067] Where, δ 13 C CSIA represents monomeric isotopes; C conc Biomarker concentrations from various sources; δ 13 C TOC Total organic carbon isotope; K represents the number of sources; f k represents the contribution ratio of source k, ∑f k =1;δ 13 C CSIA,k , δ 13 C TOC,k 、C conc,k are all end member parameters, among which δ 13 C CSIA,k Represents the monomeric isotope δ of source k 13 C value; δ 13 C TOC,k represents the total organic carbon isotope value of source k; C conc,k represents the biomarker concentration of source k; ε δ,OSIA , ε C , ε δ,TOC Represents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC Observation error.

[0068] In the Bayesian framework, assuming that the observation error consists of analytical error and systematic error, the observation error can be expressed as:

[0069]

[0070] Where, N(0,σ2 ) indicates that the parameter has a mean of 0 and a variance of σ 2 Normal distribution; σ δ,CSIA , σ C , σ δ,TOC Represents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC The systematic error to be estimated; σ δ,CSIA,anal , σ C,anal , σ δ,TOC,anal Represents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC known analytical errors.

[0071] The system error meets the following requirements:

[0072]

[0073] The end member parameters are expressed as:

[0074]

[0075] Where, δ 13 C CSIA,k Represents the monomeric isotope δ of source k 13 C value; δ 13 C tOC,k represents the total organic carbon isotope value of source k; C conc,k represents the biomarker concentration of source k; N(μ,τ 2 ) indicates that the parameters have a mean of μ and a variance of τ 2 Normal distribution of μ δ,CSIA,k 、μ δ,TOC,k 、μ C,k Represents δ 13 C CSIA,k , δ 13 C TOC,k 、C conc,k The prior mean of τ δ,CSIA,k , τ δ,TOC,k , τ C,k represent the monomer isotope δ of sample n 13 Analytical errors of C values, biomarker concentrations, and total organic carbon isotope values.

[0076] The sources of sediment organic carbon should include: mangrove plant sources, terrestrial sources (terrestrial plants), and marine sources (microorganisms and algae, etc.). The end member library is obtained as shown in Table 3.

[0077] Table 3 End member parameter data (a priori parameters)

[0078]

[0079] Note: In the table, μ δ , τ δ are the prior means and prior standard deviations of the endmember parameters, respectively.

[0080] S3) Calibrate the isotope mass balance using total organic carbon constraints; specifically:

[0081] δ 13 C TOC,k =δ 13 C CSIA,k +Δ k +β k ;

[0082] Where, Δ k represents the intrinsic fractionation value between the biomarker and total organic carbon of source k; β k represents the total organic carbon constraint.

[0083] Among them, the intrinsic fractionation value Δ k and the total organic carbon constraint term β k Respectively expressed as:

[0084] Δ k =δ 13 C TOC -δ 13 C CSIA ;

[0085]

[0086] Where, Represents β k The mean is 0 and the variance is σ β,k Normal distribution; σ β,k Represents the variance of the biological variation value of source k.

[0087] S4) Use Dirichlet to estimate the prior distribution of source contribution ratio; its expression is:

[0088] f k ~Dirichlet(α k );

[0089] Among them, α k is a hyperparameter; if there is no information prior, α k=1, if there is prior knowledge, α can be adjusted k value.

[0090] S5) Estimate the posterior distribution of the source contribution ratio through MCMC sampling, specifically:

[0091] S51) According to Bayes' theorem, the posterior distribution is obtained based on the likelihood function of the prior distribution and the observed data; that is:

[0092] P(θ|D)∝L(D|θ)·P(θ);

[0093] Where θ = f, δ 13 C CSIA,k ,C conc,k ,δ 13 C TOC,k ,σ δ,CSIA ,σ C ,σ δ,TOC are all parameters to be estimated; D = δ 13 C CSIA ,C conc ,δ 13 C TOC is the observed data; L(D|θ) is the likelihood function; P(θ) is the prior distribution;

[0094] P(θ|D) is the posterior distribution.

[0095] The likelihood function L(D|θ) is expressed as:

[0096]

[0097] Where, N(μ,σ 2 ) indicates that the parameter has a mean of μ and a variance of σ 2 Normal distribution of δ 13 C CSIA,n 、C cone,n , δ 13 C TOC,n represent the monomer isotope δ of sample n 13 C value, biomarker concentration, total organic carbon isotope value; σ δ,CSIA,anal,n , σ C,anal,n , σ δ,TOC,anal,n represent the monomer isotope δ of sample n 13 Analytical errors of C values, biomarker concentrations, and total organic carbon isotope values.

[0098] S53) Use Hamiltonian Monte Carlo algorithm to sample and iterate the posterior distribution. After T iterations, the source contribution ratio f is obtained. k The posterior distribution of Figure 2As shown, and output source contribution ratio f k The posterior mean and 95% confidence interval of are shown in Table 4, Table 5, and Figure 3 The system error posteriori is shown in Table 6.

[0099] Table 4 Posterior summary of carbon source contribution ratio

[0100]

[0101] Table 5δ 13 C CSIA Posterior summary of endmember values ​​(‰)

[0102]

[0103] Table 6 Summary of system error posteriori

[0104]

[0105]

[0106] The above embodiments and descriptions are only for explaining the principles and best embodiments of the present invention. Without departing from the spirit and scope of the present invention, the present invention may be subject to various changes and improvements, which shall fall within the scope of the invention to be protected.

Claims

1. A refined method for quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments, characterized in that: The steps include: S1), obtaining mangrove wetland sediment sample data; S2), building a multi-source observation data coupling model; S3), calibrating the isotope mass balance using total organic carbon constraints; S4), using Dirichlet to estimate the prior distribution of source contribution ratio; S5) Estimate the posterior distribution of source contribution ratio through MCMC sampling.

2. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 1, characterized in that: In step S2), the expression of the multi-source observation data coupling model is: Where, δ 13 C CSIA represents monomeric isotopes; C conc Biomarker concentrations from various sources; δ 13 C TOC Total organic carbon isotope; K represents the number of sources; f k represents the contribution ratio of source k, ∑f k =1;δ 13 C CSIA,k , δ 13 C TOC,k 、C conc,k are all end member parameters, among which δ 13 C CSIA,k represents the monomeric isotope δ13C value of source k; δ 13 C TOC,k represents the total organic carbon isotope value of source k; C conc,k represents the biomarker concentration of source k; ε δ,CSIA , ε C , ε δ,TOC Represents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC Observation error.

3. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 2, characterized in that: In step S2), under the Bayesian framework, the observation error is assumed to consist of analytical error and systematic error, thus obtaining: Where, N(0,σ 2 ) means that the mean is 0 and the variance is σ 2 Normal distribution; σ δ,CSIA , σ C , σ δ,TOC Represents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC The systematic error to be estimated; σ δ,CSIA,anal , σ C,anal , σ δ,TOC,anal Represents monomeric isotope δ 13 C CSIA , biomarker concentration C conc , total organic carbon isotope δ 13 C TOC known analytical errors.

4. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 3, characterized in that: In step S2), the end member parameters obey the following normal distribution: Where, δ 13 C CSIA,k represents the monomeric isotope δ13C value of source k; δ 13 C TOC,k represents the total organic carbon isotope value of source k; C conc,k represents the biomarker concentration of source k; N(μ,τ 2 ) indicates that the parameters have a mean of μ and a variance of τ 2 Normal distribution of μ δ,CSIA,k 、μ δ,TOC,k 、μ C,k Represents δ 13 C CSIA,k , δ 13 C TOC,k 、C conc,k The prior mean of τ δ,CSIA,k , τ δ,TOC,k , τ C,k Represents monomeric isotope δ 13 C CSIA , total organic carbon isotope δ 13 C TOC , biomarker concentration C conc The prior standard deviation of .

5. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 4, characterized in that: In step S3), the isotope mass balance is constrained using the total organic carbon constraint, specifically: d 13 C TOC,k =d 13 C CSIA,k +D k +b k ; Where, Δ k represents the intrinsic fractionation value of the biomarker and total organic carbon of source k; β k It represents the allowable range of biological variation and is a constraint item for total organic carbon.

6. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 5, characterized in that: In step S3), the inherent fractionation value Δ k and the total organic carbon constraint term β k Respectively expressed as: D k =d 13 C TOC -d 13 C CSIA ; Where, Represents β l The mean is 0 and the variance is σ β,k Normal distribution; σ β,k Represents the variance of the biological variation value of source k.

7. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 1, characterized in that: In step S4), the prior distribution of the source contribution ratio is estimated using Dirichlet, which is expressed as: f k ~Dirichlet(a k ); Among them, α k is a hyperparameter; if there is no information prior, α k =1.

8. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 1, characterized in that: In step S5), the posterior distribution of the source contribution ratio is estimated by MCMC sampling, specifically: S51), according to Bayes' theorem, the posterior distribution is obtained based on the likelihood function of the prior distribution and the observed data; S52) Use Hamiltonian Monte Carlo algorithm to perform posterior distribution sampling and iteration. After T iterations, the source contribution ratio f is obtained. k The posterior distribution of , and output the source contribution ratio f k The posterior mean and 95% confidence interval of .

9. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 8, characterized in that: In step S51), the posterior distribution is obtained based on the likelihood function and the prior distribution, namely: P(θ|D)∝L(D|θ)·P(θ); Where θ = f, δ 13 C CSIA,k ,C conc,k ,δ 13 C TOC,k ,σ δ,CSIA ,σ C ,σ δ,TOC For all parameters to be estimated; D = δ 13 C CSIA ,C conc ,δ 13 C TOC is the observed data; L(D|θ) is the likelihood function; P(θ) is the prior distribution; P(θ|D) is the posterior distribution.

10. The method for finely quantifying the contribution of mangrove plant sources to the carbon sink in mangrove wetland sediments according to claim 9, characterized in that: In step S51), the likelihood function L(D|θ) is expressed as: Where, N(μ,σ 2 ) indicates that the parameter has a mean of μ and a variance of σ 2 Normal distribution of δ 13 C CSIA,n 、C cone,n , δ 13 C TOC,n represent the monomer isotope δ13C value, biomarker concentration, and total organic carbon isotope value of sample n, respectively; σ δ,CSIA,anal,n , σ C,anal,n , σ δ,TOC,anal,n represent the analytical errors of the monomeric isotope δ13C value, biomarker concentration, and total organic carbon isotope value of sample n, respectively.

Images