A forest ecological restoration method based on community integrity and functional traits

Abstract

The present invention belongs to the field of ecological restoration and discloses a forest ecological restoration method based on community integrity and functional traits, including S1, obtaining the community integrity index (CCI) of the study area; S2, determining whether the CCI is greater than 0. If so, the species in the study area are protected; if not, entering S3; S3, obtaining the dark diversity probability of the species in the study area; S4, selecting species with a dark diversity probability less than a set threshold as candidate species; S5, constructing an underdetermined equation system based on the expected values ​​of the functional traits of the candidate species to solve the relative abundance of the candidate species; S6, performing species restoration based on the relative abundance of the candidate species. The present invention can quantitatively assess missing species in the community, predict the abundance ratio of restoration species, and optimize functional restoration combinations, solving the problem that traditional species selection based on reliance on existing community structure alone cannot accurately identify potential restoration species and achieve function-oriented restoration configuration.

Description

Technical Field

[0001] The present invention relates to the field of ecological restoration, and in particular to a forest ecological restoration method based on community integrity and functional traits. Background Art

[0002] Forest ecosystems currently suffer from degradation of community structure and function, as well as declines in biodiversity, following natural disturbances or human damage. Traditional restoration measures often rely on existing community characteristics or empirical judgment to select restoration species. This often overlooks the importance of "dark diversity" species—potential but not yet present in regional species pools—and their functional traits in ecosystem recovery, resulting in limited restoration effectiveness.

[0003] Therefore, there is an urgent need to develop a forest ecological restoration technology method that systematically integrates dark diversity, species functional traits and environmental factors and has quantitative prediction capabilities, so as to provide a more scientific and accurate restoration path for degraded forest ecosystems. Summary of the Invention

[0004] The purpose of the present invention is to disclose a forest ecological restoration method based on community integrity and functional traits to solve the technical problems raised in the background technology.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] The present invention provides a forest ecological restoration method based on community integrity and functional traits, comprising

[0007] S1, obtain the community integrity index CCI of the study area;

[0008] S2, determine whether CCI is greater than 0. If so, protect the species in the study area; if not, enter S3;

[0009] S3, obtain the dark diversity probability PDD of species in the study area;

[0010] S4, species with dark diversity probability PDD less than the set threshold are regarded as candidate species;

[0011] S5, based on the expected values ​​of the functional traits of the candidate species, construct an underdetermined system of equations to solve the relative abundance of the candidate species;

[0012] S6. Based on the relative abundance of candidate species, species restoration treatment was performed on the study area.

[0013] Preferably, S1 includes:

[0014] The community integrity index (CCI) of the study area was calculated using the following formula:

[0015]

[0016] OD is the measured diversity index of the study area; It is the dark diversity indicator of the study area.

[0017] Preferably, the calculation formula for the measured diversity index of the study area is:

[0018] OD=

[0019] N is the total number of species present in the study area as recorded in the species bank; Used to indicate whether the zth species is observed. If it is observed, then is 1; if it is not observed, is 0.

[0020] Preferably, the process of obtaining the dark diversity index of the study area includes:

[0021] Step 1: training the prediction model;

[0022] Step 2: Input the environmental data of each sample plot in the study area into the prediction model to obtain the existence probability of each species in each sample plot;

[0023] Step 3, calculate the probability of species presence in the study area based on the presence probability of all quadrats;

[0024] Step 4: Obtain the dark diversity index of the study area based on the probability of species existing in the study area.

[0025] Preferably, training the prediction model includes:

[0026] The first step is to prepare training data:

[0027] The training data includes species matrix and environment matrix;

[0028] The second step is to select the prediction model:

[0029] The prediction model is a Bayesian hierarchical model or a Markov random field model;

[0030] The third step is to train the prediction model based on the training data.

[0031] Preferably, the probability of a species existing in the study area is calculated based on the probability of existence of all quadrats, including:

[0032] For species i, the probability of species i existing in quadrat k is expressed as ;

[0033] The probability that species i does not exist in quadrat k is 1- ;

[0034] The probability of species i not existing in the study area is , KU is the set of quadrats;

[0035] The probability of species i existing in the study area is .

[0036] Preferably, the dark diversity index of the study area is obtained based on the probability of species existing in the study area, including:

[0037] The following calculation is performed for each species i:

[0038] Determine whether the probability of species i existing in the study area is greater than the set probability threshold. If so, store it in the set U1;

[0039] Store the species in U1 that are not observed in any quadrat into set U2;

[0040] The total number of species in U2 was used as the dark diversity indicator of the study area.

[0041] Preferably, obtaining the dark diversity probability PDD of species in the study area includes:

[0042] The dark diversity probability of species b in the study area is calculated using the following formula :

[0043]

[0044] It represents the number of times species b belongs to set U2 in the most recent NP observations.

[0045] Preferably, the threshold is set to 0.4.

[0046] Preferably, the underdetermined system of equations is as follows:

[0047]

[0048] in represents the value of species j on functional trait q, is the relative abundance of species j, is the expected value of the functional trait q in the community of the study area, q∈bU, bU is the set of functional traits.

[0049] This invention discloses a forest ecological restoration method based on community integrity and functional traits. By integrating regional species pool potential with ecological functional trait information, the method develops a new method for quantitatively predicting optimal restoration species combinations. This method can be used to quantitatively assess missing species in a community, predict the abundance ratio of restoration species, and optimize functional restoration combinations. This method addresses the technical issues of traditional species selection based solely on existing community structure, which cannot accurately identify potential restoration species and implement function-oriented restoration configurations. It is fast, effective, convenient, and practical. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0051] Figure 1 This is a schematic diagram of a forest ecological restoration method based on community integrity and functional traits of the present invention.

[0052] Figure 2 Schematic diagram of the relative abundance of the present invention. DETAILED DESCRIPTION

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0054] Basic forest data can be collected using existing conventional technology or drones. There is no restriction on the specific brand of drone. All instruments, software, and tools involved are commercially available. I will not elaborate on these details.

[0055] Glossary: ​​The terms used in the examples are generally understood in the art, for example:

[0056] Measured diversity: refers to the number of species that actually exist and have been recorded in a specific area, which can also be understood as species richness;

[0057] Community integrity refers to the intact and stable state of a forest ecosystem in terms of species composition, ecological processes, structure, and function. It encompasses the maintenance of biodiversity, the smooth operation of material cycles, and the normal flow of energy. Community integrity is assessed by measuring the ratio of observed diversity to hidden diversity, reflecting the degree to which the species pool within a local community is realized. This concept quantifies the relationship between realized and potential diversity within a community, providing a more intuitive understanding of the community's ecological status and its recovery process.

[0058] refer to Figure 1 The present invention provides a forest ecological restoration method based on community integrity and functional traits, including

[0059] S1, obtain the community integrity index CCI of the study area;

[0060] S2, determine whether CCI is greater than 0. If so, protect the species in the study area; if not, enter S3;

[0061] S3, obtain the dark diversity probability PDD of species in the study area;

[0062] S4, species whose dark diversity probability PDD is less than the set threshold are regarded as candidate species;

[0063] S5, based on the expected values ​​of the functional traits of the candidate species, construct an underdetermined system of equations to solve the relative abundance of the candidate species;

[0064] S6. Based on the relative abundance of candidate species, species restoration treatment was performed on the study area.

[0065] Preferably, S1 includes:

[0066] The community integrity index (CCI) of the study area was calculated using the following formula:

[0067]

[0068] OD is the measured diversity index of the study area; It is the dark diversity indicator of the study area.

[0069] Preferably, the calculation formula for the measured diversity index of the study area is:

[0070] OD=

[0071] N is the total number of species present in the study area as recorded in the species database; Used to indicate whether the zth species is observed. If it is observed, then is 1; if it is not observed, is 0.

[0072] Preferably, the process of obtaining the dark diversity index of the study area includes:

[0073] Step 1: training the prediction model;

[0074] Step 2: Input the environmental data of each sample plot in the study area into the prediction model to obtain the existence probability of each species in each sample plot;

[0075] Step 3, calculate the probability of species presence in the study area based on the presence probability of all quadrats;

[0076] Step 4: Obtain the dark diversity index of the study area based on the probability of species existing in the study area.

[0077] This method makes full use of environmental information and species co-occurrence patterns between sample plots, quantitatively identifying species that should theoretically exist in the community under current environmental conditions but are actually missing, providing a more accurate basis for assessing community integrity and conservation management.

[0078] Preferably, training the prediction model includes:

[0079] The first step is to prepare training data:

[0080] The training data includes species matrix and environment matrix;

[0081] The second step is to select the prediction model:

[0082] The prediction model is a Bayesian hierarchical model or a Markov random field model;

[0083] The third step is to train the prediction model based on the training data.

[0084] Preferably, the species matrix is ​​NK , NK is the total number of quadrats. In the species matrix, if the species exists in the quadrats, the value of the element corresponding to the quadrats is 1, otherwise it is 0.

[0085] Preferably, the environment matrix is where M represents the total number of types of environmental data; the types of environmental data include temperature, precipitation, soil pH, altitude, etc.

[0086] The processing of environmental data is as follows:

[0087] Normalize environment variables.

[0088] Preferably, when the prediction model is a Bayesian hierarchical model, training the prediction model based on the training data includes:

[0089] The first step is to define the model structure:

[0090] (1) Ecological process layer:

[0091] The actual existence status of the modeled species at the u-th location , which is usually a binary latent variable (0 = not present, 1 = present).

[0092] Probability of real existence Usually through the logut-lunk function and environmental covariates Related:

[0093] , , ,...are the regression coefficients that need to be estimated.

[0094] is a random effect that captures spatial autocorrelation, site-specific variation, or other structured random variation (e.g. or spatial Gaussian process GP(0,Σ)).

[0095] (2) Observation process layer:

[0096] Modeling the conditional probability of observing a species in the vth survey at the uth site , provided the species actually exists at the location ( = 1). If the species does not exist ( =0), then the conditional probability of observing the species is 0.

[0097] The observed probability is usually also combined with the covariate through the logut-lunk function These covariates may include survey time, weather, observer experience, etc.

[0098] β0,β1,β2,... are the observed regression coefficients that need to be estimated.

[0099] Data Model:

[0100] The observation data (whether the species was observed in the vth survey at the uth site, 0 if not, 1 otherwise) and and Get in touch: | , Bernoullu )

[0101] This means:

[0102] If the species exists ( =1), then By probability When it is observed, it is 1 (observed), with probability 1- If observed, it is 0 (not observed).

[0103] If the species does not exist ( =0), then Must be 0 (not observed).

[0104] The second step is to prepare the data:

[0105] Response variable: Observation record of the vth survey at location u ( = 0 or 1). Note that multiple surveys are usually required for a location to reliably estimate the probability of an observation.

[0106] Ecological covariates : Variables that describe the environmental characteristics of site u (e.g., elevation, slope, vegetation type, climate variables, distance to water, etc.). These should be associated with site u, not with a single survey v. Covariates used to predict new sites must be the same as those in the training data and must be processed in the same way (standardization, etc.).

[0107] Observed covariates : Variables describing the conditions of the vth survey at the uth site (e.g. survey date, time of day, duration, cloud cover, observer, etc.). These are used to explain variations in the probability of observation.

[0108] The third step is to set the prior distribution:

[0109] For all parameters that need to be estimated (α0, α1, ..., β0, β1, ..., , , ...) specify the prior distribution.

[0110] Step 4: Model training:

[0111] Based on the observed data y and the set prior distribution, the model parameters and latent variables are calculated ( ), where θ includes all parameters (α, β, σ, ...).

[0112] The above process can be implemented using Markov Chain Monte Carlo (MCMC).

[0113] Preferably, after the prediction model is trained based on the training data, the prediction model outputs species-environment relationship parameters, thereby enabling the prediction model to have prediction capabilities.

[0114] Preferably, the probability of a species existing in the study area is calculated based on the probability of existence of all quadrats, including:

[0115] For species i, the probability of species i existing in quadrat k is expressed as ;

[0116] The probability that species i does not exist in quadrat k is 1- ;

[0117] The probability of species i not existing in the study area is , KU is the set of quadrats;

[0118] The probability of species i existing in the study area is .

[0119] Preferably, the dark diversity index of the study area is obtained based on the probability of species existing in the study area, including:

[0120] The following calculation is performed for each species i:

[0121] Determine whether the probability of species i existing in the study area is greater than the set probability threshold (for example, 0.5). If so, store it in the set U1;

[0122] Store the species in U1 that are not observed in any quadrat into set U2;

[0123] The total number of species in U2 was used as the dark diversity indicator of the study area.

[0124] Preferably, obtaining the dark diversity probability PDD of species in the study area includes:

[0125] The dark diversity probability of species b in the study area is calculated using the following formula :

[0126]

[0127] It represents the number of times species b belongs to set U2 in the most recent NP observations. Specifically, the value of NP can be determined based on the actual number of observations.

[0128] Preferably, the threshold is set to 0.4.

[0129] Preferably, the underdetermined system of equations is as follows:

[0130]

[0131] in represents the value of species j on functional trait q, is the relative abundance of species j, is the expected value of the functional trait q in the community of the study area, q∈bU, bU is the set of functional traits.

[0132] It is based on fitting the relationship between traits and environment through a generalized linear model, and then predicting the optimal trait value under the target environment. This target trait value is used as a constraint and input into the linear equation system to generate species abundance distribution.

[0133] Preferably, the functional traits include wood density, diameter at breast height, tree height, leaf nitrogen content, etc. of the species.

[0134] Preferably, solving for the relative abundance of candidate species includes:

[0135] Establish the objective function Y:

[0136]

[0137] is the importance weight of the functional trait q, ;

[0138] The constraints on Y are YS:

[0139]

[0140] A heuristic algorithm was used to minimize the objective function and obtain the relative abundance of each species.

[0141] Preferably, a heuristic algorithm is used to minimize and optimize the objective function to obtain the relative abundance of each species, including:

[0142] The first step is to obtain the heuristic algorithm set GU. The number of heuristic algorithms contained in GU is represented by nGU.

[0143] In the second step, each heuristic algorithm in GU is used to optimize the objective function FM (for example, 20 times) to obtain the relative abundance of each species. , represents the relative abundance of species h obtained in the cth optimization of the bth heuristic algorithm;

[0144] The third step is to calculate the weight of each optimization result of each heuristic algorithm:

[0145] Calculate the score of each optimization result:

[0146]

[0147] represents the score of the cth optimization result of the bth heuristic algorithm, and ms is a constant to avoid the denominator being 0, for example, it can be e to the power of -5; is the final fitness obtained by the cth optimization of the bth heuristic algorithm;

[0148] Calculate the total score:

[0149]

[0150] Calculate the weight of each optimization result:

[0151]

[0152] The fourth step is to obtain the final relative abundance of each species:

[0153]

[0154] represents the final relative abundance of species h.

[0155] The above optimization scheme can significantly improve robustness and stability, reduce the impact of randomness, and reduce the risk of falling into local optimality.

[0156] Results from multiple runs of a single algorithm can fluctuate significantly. Ensemble methods significantly smooth out random fluctuations by combining the results of numerous independent runs (across different algorithms), resulting in results that are more reproducible across different random seeds.

[0157] Different algorithms, and multiple runs of the same algorithm, may fall into different local optima. Ensemble methods don't rely on luck to find the global optimum, but instead combine multiple potentially "good" solutions. Even if no single solution is globally optimal, the weighted average often results in a closer approximation to the global optimum or a better compromise than a single run. This effectively explores a wider region of the solution space.

[0158] Different algorithms excel at handling different types of optimization landscapes (e.g., GA excels at global exploration, PSO excels at fast convergence, and SA excels at escaping local optima). Ensemble methods combine the strengths of different algorithms and, in theory, can find solutions that are better than the average of any single algorithm.

[0159] A large number of candidate solutions contain information about the shape of the objective function and the structure of the solution space. Weighted averaging can be viewed as a "consensus" solution guided by this information.

[0160] Preferably, the heuristic algorithms in GU include genetic algorithm, particle optimization algorithm, simulated annealing algorithm and differential evolution algorithm.

[0161] Example:

[0162] The study area is located in Forest Farm C in City B, Province A. This area has a temperate continental climate with four distinct seasons and an average annual temperature of 3.6°C. The total area of ​​the restoration area is 150 mu, and the restoration target is a broad-leaved Korean pine forest dominated by native tree species. Because the bare land is not suitable for directly creating the key ecological species Korean pine, in 2014, the first artificial regeneration was to use Pinus sylvestris var. mongolica, which has a fast growth rate and strong adaptability to full-light environments. By the end of 2024, the survival density of Pinus sylvestris var. mongolica was low, but it effectively improved the habitat, prevented forest degradation, and created conditions for further restoration measures. In 2025, the species configuration for the next step of restoration will be determined with reference to the local zonal vegetation, and afforestation will be carried out accordingly in the spring of that year.

[0163] Based on regional forest survey data, species composition information of the sample plots was obtained. In combination with public databases (such as WorldClim, SoilGrids, etc.) or field sampling data, key environmental variables such as the average annual temperature, annual precipitation, and soil physical and chemical properties corresponding to the sample plots were extracted. A sample data table is shown in Table 1.

[0164] Table 1 Demonstration data table

[0165]

[0166] If the community integrity index of a sample site is greater than 0, conservation measures will be implemented for the community, and existing species will be managed according to local conditions. For example, natural restoration methods will be adopted to reduce the negative impact of human intervention on the ecosystem. For areas with high vegetation cover, minimal degradation, and a certain degree of natural recovery capacity, closed areas will be established. These measures may include full, partial, or rotational closures. Full closure prohibits all human activities during the closure period; partial closure allows limited tending and collection activities within a certain time and area; and rotational closure involves rotating closures of separate areas. Fences, boundary stakes, warning signs, and other containment measures will be installed at the boundaries of the closed areas to clearly define the scope and requirements of the closure.

[0167] Table 2 shows the top eight commonly used afforestation species ranked by calculated dark diversity probability. Because the restoration area is low-lying and predominantly wetlands, following the principle of "suitable trees for suitable sites," in addition to the ecologically keystone Korean pine (Pinus koraiensis), the adaptable walnut (Juglans mandshurica) and Manchurian ash (Fraxinus mandshurica) species were selected as restoration species for the Manchurian ash forest.

[0168] Table 2 Dark diversity probability ranking table

[0169]

[0170] As shown in Table 2, the present invention calculated the trait mean of each species and used a generalized linear model to predict the expected value of the trait in the target community environment. The relative abundance of each tree species was solved by substituting the underdetermined equations to obtain the attached Figure 2 .

[0171] Based on the output of the above equations, it can be decided to select Juglans mandshurica, Fraxinus mandshurica and Pinus koraiensis as the restoration species combination, and their planting ratios fall into the following range: Figure 2 The areas shown (i.e., 22% for Fraxinus mandshurica, 60% for Pinus koraiensis, and 18% for Juglans mandshurica) are used to guide the implementation of restoration operations in the target group.

[0172] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A forest ecological restoration method based on community integrity and functional traits, characterized by: include: S1, obtain the community integrity index CCI of the study area: The community integrity index (CCI) of the study area was calculated using the following formula: OD is the measured diversity index of the study area; DD is the dark diversity index of the study area; The process of obtaining dark diversity indicators in the study area includes: Step 1: training the prediction model; Step 2: Input the environmental data of each sample plot in the study area into the prediction model to obtain the existence probability of each species in each sample plot; Step 3, calculate the probability of species presence in the study area based on the presence probability of all quadrats; Step 4: Obtain dark diversity indicators of the study area based on the probability of species existing in the study area, including: The following calculation is performed for each species i: Determine whether the probability of species i existing in the study area is greater than the set probability threshold. If so, store it in the set U1; Store the species in U1 that are not observed in any quadrat into set U2; The total number of species in U2 was used as the dark diversity indicator of the study area; S2, determine whether CCI is greater than 0. If so, protect the species in the study area; if not, enter S3; S3, obtain the dark diversity probability PDD of species in the study area, including: The dark diversity probability PDD of species b in the study area is calculated using the following formula b : ND b represents the number of times species b belongs to set U2 in the most recent NP observations; S4, species whose dark diversity probability PDD is less than the set threshold are regarded as candidate species; S5, based on the expected values ​​of the functional traits of the candidate species, construct an underdetermined system of equations to solve the relative abundance of the candidate species; S6. Based on the relative abundance of candidate species, species restoration treatment was performed on the study area.

2. A forest ecological restoration method based on community integrity and functional traits according to claim 1, characterized in that: The calculation formula for the measured diversity index in the study area is: N is the total number of species present in the study area as recorded in the species database; S z Used to indicate whether the zth species is observed. If it is observed, S z is 1; if it is not observed, S z is 0.

3. A forest ecological restoration method based on community integrity and functional traits according to claim 1, characterized in that: Training a predictive model, including: The first step is to prepare training data: The training data includes species matrix and environment matrix; The second step is to select the prediction model: The prediction model is a Bayesian hierarchical model or a Markov random field model; The third step is to train the prediction model based on the training data.

4. The forest ecological restoration method based on community integrity and functional traits according to claim 1, characterized in that: The probability of species presence in the study area was calculated based on the presence probabilities of all quadrats, including: For species i, the probability of species i existing in quadrat k is expressed as p i,k ; The probability that species i does not exist in quadrat k is 1-p i,k ; The probability of species i not existing in the study area is KU is the set of quadrats; The probability of species i existing in the study area is P yes,i =1-P no,i .

5. The forest ecological restoration method based on community integrity and functional traits according to claim 1, characterized in that: The threshold is set to 0.

4.

6. The forest ecological restoration method based on community integrity and functional traits according to claim 1, characterized in that: The underdetermined system of equations is as follows: where t jq represents the value of species j on functional trait q, p j is the relative abundance of species j, T q is the expected value of the functional trait q in the community of the study area, q∈bU, bU is the set of functional traits.

Images