Benthonic animal optimal selection method based on substrate feature matching and inhabiting structure device of benthonic animal optimal selection method

Abstract

The invention relates to the technical field of ecological restoration and environmental engineering, and solves the technical problems that in existing engineering, species screening lacks a quantitative basis for substrate adaptation, an ecological window is not brought into an optimal decision, the function of an inhabiting device is single, and an evaluation result is difficult to translate into executable layout parameters. The method comprises the following steps: collecting substrate factor data of a target water body, unifying dimensions, comparing the data with a species response database, calculating a substrate matching score, and combining ecological windows such as water temperature, flow velocity, disturbance, transparency and the like to generate an optimal selection index; and outputting the target species, the suitable putting time period and the device layout parameters. According to the method, an integrated decision of seed selection, timing and layout is achieved, the colonization success rate and community stability are remarkably improved, the method is suitable for various water bodies such as lakes, rivers and wetlands, and the long-term restoration effect and operation and maintenance economical efficiency can be improved based on monitoring feedback iteration optimization parameters.

Description

Technical Field

[0001] This invention relates to the field of ecological restoration and environmental engineering technology, and in particular to a method for selecting benthic animals based on substrate characteristic matching and its habitat structure device, which belongs to the cross-application technology of ecosystem restoration, water environment management and artificial habitat construction. Background Technology

[0002] Benthic animals are an important biological group in aquatic ecosystems, widely distributed in the surface or shallow media of sediments in rivers, lakes, and wetlands. They play a crucial role in organic matter decomposition, nutrient recycling, and biotransformation of pollutants. Their community structure and diversity are often used as important indicators of the ecological health of aquatic bodies and are priority targets for restoration in many aquatic ecological restoration projects. In existing aquatic ecological restoration practices, the reconstruction of benthic animals mainly follows two paths: natural restoration, which involves inducing the natural migration and recovery of target species by improving environmental conditions; and artificial assisted restoration, which involves actively introducing target species and providing certain forms of habitat carriers or platforms. However, the following problems are commonly encountered in the engineering applications of artificial assisted restoration: (1) Species selection lacks quantitative basis for substrate compatibility. In real-world projects, species selection is often based on regional lists, biological monitoring frequency, or expert experience through "analogical" methods, with less emphasis on incorporating the substrate characteristics of the target water body into the selection logic. Benthic animals are highly sensitive to substrate factors such as particle size structure (sand / silt / viscosity ratio), organic matter content, nutrient levels (e.g., TN, TP), pH and conductivity, porosity, or water content. Different species exhibit significant differences in their adaptation thresholds and tolerances to these factors. The lack of a quantitative matching and ranking mechanism based on substrate factor sets can easily lead to selected species failing to survive or form stable communities after release. Furthermore, existing multi-indicator evaluations suffer from non-standardized same-scale processing (direct superposition of indicators with different dimensions and distribution patterns or addition of empirical weights), resulting in systematic biases in the fusion results.

[0003] (2) The synergistic effects of ecological windows and hydrodynamic disturbances were not fully considered. The migration, settlement, and reproduction of benthic animals are significantly sensitive to time and space, and are affected by water temperature (T), current velocity (V), disturbance frequency (D), and transparency (e.g., the depth of the Seychelles disc). S d Environmental factors such as zoning and sedimentation have a combined impact. Current practices tend to focus on static habitat suitability assessments and lack a mechanism to systematically incorporate the environmental adaptation window into the selection process; even if species are compatible with the substrate, improper timing can significantly reduce the effectiveness of restoration.

[0004] (3) The habitat structures are functionally limited and cannot support the full range of ecological needs. Currently, the carriers used for benthic animal restoration are mostly made of stones, brick cages, ceramic pots, or simplified fish nests, emphasizing physical stability or shelter, but they do not adequately consider the comprehensive needs of nutrient supply, habitat microstructure, and attachment substrate. There is a lack of modular engineering devices that integrate resistance to currents, controllable nutrient release, multi-scale shelter / attachment space, and are easy to deploy and maintain quickly, making it difficult to continuously attract or maintain population establishment and expansion.

[0005] At the research level, some studies have attempted to use substrate suitability index, multi-factor evaluation, or niche models (such as Habitat Suitability Index, Ecological Niche Modeling, etc.) to help determine the habitat suitability of benthic animals, but most of them remain at the evaluation or prediction level and lack effective connection with the engineering implementation stage. Specifically, this manifests as follows: (1) Insufficient parameterization of the relationship between substrate factors and species response, lack of structured representation and comparable scale of "optimal value-tolerance-weight"; (2) Inconsistent quantitative scoring and ranking methods for multi-indicator fusion, lack of repeatable response functions and normalization constraints; (3) Lack of conversion rules between evaluation and deployment "from optimization results to device deployment parameters"; (4) Lack of a closed-loop mechanism of monitoring-feedback-model update after engineering implementation, making it difficult to iteratively optimize based on actual colonization effects.

[0006] In summary, existing technologies still have significant gaps in areas such as species selection based on substrate characteristics, inclusion and determination of ecological windows, and multifunctional habitat structures that provide stability, nutrition, and shelter. To improve the feasibility and success rate of benthic animal ecological reconstruction, it is necessary to provide a technical solution that uses a target water body substrate factor set as a basis, combines a species response parameterized database with environmental adaptation window modeling, forms a quantifiable and sortable selection process, outputs a parameter set that can be directly used for engineering deployment, and is equipped with an integrated habitat structure device that provides current resistance and stability, controlled-release nutrition, and multi-scale shelter, and has the ability to update the model based on monitoring feedback. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a method for selecting benthic animals based on substrate feature matching and a habitat structure device thereof. This solves the technical problems in existing engineering projects, such as the lack of quantitative basis for substrate compatibility in species screening, the failure to incorporate ecological windows into the selection decision, the limited functionality of habitat devices, and the difficulty in translating evaluation results into executable deployment parameters.

[0008] To address the aforementioned technical problems, this invention provides a method for selecting benthic animals based on substrate feature matching. This method includes substrate factor standardization and parameter space construction, parametric modeling of a species response database, substrate matching scoring and candidate screening, establishment and same-scale fusion of ecological window fitness functions, mapping of selection results to deployment parameters, and structural design and deployment of multifunctional habitat structures. It also provides a monitoring feedback and parameter update mechanism during operation. This achieves quantitative coupling of "substrate-species response-ecological window" within a unified metric space and directly maps the selection results to on-site deployment and operation parameters, thereby improving colonization success rate and community stability.

[0009] Furthermore, the benthic animal selection method uses the bottom sediment factor data of the target water body as input to form a bottom sediment factor vector with a dimension of no less than five. The substrate factor vector The standardized underlying factor vector is constructed by performing dimension-by-dimensional standardization to form a unified metric standardization space. Standardization preferably adopts one or a combination of the following: Z-score transformation is suitable for cases where the single-dimensional sample size is ≥30 and approximately normally distributed. Min-Max normalization is suitable for situations where the dimensions of multidimensional indicators have a large span but the dimensional distribution is balanced. Box-Cox transformation with translation constants: For dimensions with zero or negative values, a translation constant is first introduced. This makes it applicable to all samples Its transformation is defined as: when hour, ; when hour, ; in, The Box-Cox transformation factor. For the first s The sample at the th j The original measurements of each substrate factor, For the Box-Cox power parameter of this dimension (in the th dimension) j (The dimension is fixed for all samples). Let be the positive domain translation constant of this dimension. ln This is the natural logarithm. The data obtained after the above transformation are then subjected to linear scaling to form... P′ .

[0010] Furthermore, to mitigate the impact of extreme values, the substrate factor data can be first truncated or robustned before standardization.

[0011] Furthermore, a species response database is constructed within the standardized space. And calculate the species response database. Any species Substance matching score ; The species response database For any candidate species Store its effects on various substrate factors optimal value With response tolerance And configure weight vectors that reflect relative importance. Satisfying weight and To ensure cross-dimensional comparability; the optimal value Response tolerance With substrate factors They exist in the same standardized space.

[0012] Furthermore, the optimal value It can be determined based on the recorded peak kernel density or the abundance-weighted robust center statistic; The response tolerance Robust estimates can be made based on the interquartile range or the absolute deviation of the median, and truncation can be performed within the empirical interval. The weight It can be obtained and normalized by constructing a regularized logistic regression or information entropy method based on "appearance / absence".

[0013] Furthermore, the parameters can be derived from historical monitoring, semi-natural domestication experiments, and screened literature data to form a parameter library with regional representativeness.

[0014] This method defines a unimodal response function whose range is limited to [0,1] in the dimension-wise similarity calculation. f The single-peak response function f Satisfying It reaches a maximum value of 1 at time, and then... Monotonically decreasing. The unimodal response function. f Specifically, it can take any of the following forms: Gaussian function: ; Hyperbolic tangent function: ; Penalty exponential function: ; in, For the first i The species in the first j The sensitivity / decay rate parameter on each factor is used to control the rate of decline of the exponential penalty; Preferably, This is used to unify the semantics of "tolerance-sensitivity". To avoid confusion, Box-Cox power parameter in standardization The meanings and indexes are different.

[0015] The dimension-wise similarity is weighted and summed to obtain the substrate matching score, i.e.: The substrate matching score This is used to characterize the compatibility between species and the substrate of the target water body. The preferred candidate strategy is a threshold-quota composite strategy: first selecting those that meet the criteria... A collection of species, if the number is insufficient to meet the quota. Then, the score is based on the substrate quality. Fill in the blanks from highest to lowest. One, of which , .

[0016] Furthermore, the threshold The number of slots is determined by the Youden index or the inflection point of the receiver operating characteristic (ROC) curve on the historical validation set. The carrying capacity of the target water body is initially estimated or determined by the minimum number of cover species.

[0017] Furthermore, to incorporate the influence of temporal environment into the selection process, a model based on the historical distribution frequency of candidate species and their environmental response was constructed, using water temperature as a criterion. Water flow velocity Sediment disturbance frequency With transparency Ecological window adaptation function with Seychelles disk depth as the independent variable .

[0018] The ecological window adaptation function The preferred approach is to use fuzzy logic reasoning to establish the ecological window parameters. The mapping is based on membership degree, fuzzy reasoning is performed based on the rule base, and the ecological window adaptation function is obtained by defuzzifying using the centroid method. Specifically: First Define three or more membership functions (such as low / medium / high trigonometric or trapezoidal functions), construct a rule base based on experience and data for fuzzy inference, and use the centroid method to defuzzify and obtain continuous fitness.

[0019] Furthermore, the ecological window adaptation function Alternatively, a parametric regression model or a Bayesian response model can be used to establish the model, based on the historical distribution frequency of candidate species and ecological window parameters. The paired samples are used as training data, and the output is normalized to [0,1] to match the bottom sediment score. Same scale.

[0020] This invention uses a bottom sediment matching score Adaptation function with ecological window Same-scale fusion yields the optimal index The fusion preferably adopts a linear weighted form, that is: in, This refers to the scene adaptation coefficient.

[0021] Furthermore, the fusion can also take two other specific forms, namely, geometric fusion: Or, for the preferred index Or the output of the second layer network Softmax is used to generate ranking weights, targeting the candidate set Ω. cand definition: in, , is a temperature parameter used to adjust the discrimination index; For species i In the candidate set Ω cand The softmax sorting weights satisfy the following conditions: ); The input score for softmax is taken from either U(k) or the output of the second-layer network (either one, and consistent within the same candidate set). k For the candidate set Ω cand The index in.

[0022] Furthermore, the preferred index Or its softmax weights are used to determine the target species set Ω.

[0023] This invention generates a habitat structure layout parameter set Ψ by translating the preferred results into engineering executable parameters.

[0024] Furthermore, the habitat structure deployment parameter set Ψ includes at least the following elements: anchoring method and mud penetration depth, number of single-point devices and spacing between them, deployment depth and shell attitude angle, nutrient release cycle and feeding frequency. Specifically: Anchoring structure type (cone or spiral) and depth into the mud; number of single points, spacing and array shape (linear, grid or local cluster); deployment depth, angle with the mainstream direction and attitude stability parameters; feeding frequency and controlled release cycle of the nutrient release module; pore size and target porosity of the shading and induction module.

[0025] The anchoring method and the depth of penetration into the mud can be selected according to the risk classification of near-bottom shear or disturbance. In areas with high flow velocity or frequent disturbance, spiral anchoring and increased penetration depth are preferred. In medium and low risk areas, cone anchoring can be used. The number of single-point devices and the spacing between them can be estimated based on the effective coverage area of ​​the target and the radius of influence of the devices. In combination with the mainstream direction, linear arrays can be selected or grid / cluster arrays can be used in low-flow areas to increase habitat heterogeneity. The preferred deployment depth and shell attitude angle are such that the long axis of the shell is parallel to the main direction and maintains a small pitch angle. The preferred nutrient release cycle and feeding frequency are small-dose, equally spaced strategies, with a cycle of 10–30 days and a carbon-to-nitrogen ratio of 10–25, and can be adapted to an ecological window function. The intensity is finely adjusted to balance induction efficiency and water quality safety.

[0026] This invention can update parameters during operation based on monitoring feedback. This includes collecting environmental parameters after deployment, colonization rate and individual density feedback, minimizing validation error or cross-entropy loss, and adjusting weights. Response tolerance Scene adaptation coefficient α or temperature parameter β Iterative optimization and updating of the species response database. The target species set Ω and habitat structure layout parameter set Ψ for subsequent batches are calculated again.

[0027] Furthermore, the update uses successful colonization or high abundance as the target label, and adopts the criterion of minimizing validation error or cross-entropy loss to adjust the weights. Response tolerance Scene adaptation coefficient α or temperature parameter β Implement periodic or event-triggered iterative optimizations, which take effect in subsequent selection and deployment, thereby maintaining the accuracy of selection and the economy of deployment under ecological fluctuations and seasonal changes.

[0028] This invention also provides a benthic habitat structure device applied to the aforementioned benthic animal selection method. This device supports the selection results and promotes stable settlement of benthic animals. It is equipped with a water depth adjustment support and a posture-stabilizing hinge structure to adapt to angle fine-tuning and stable deployment under different slope or depth conditions. It consists of a stable base module, a nutrient release module, a sheltering and induction module, and a quick-assembly and quick-disassembly module.

[0029] The stabilizing base module preferably uses stainless steel or high-density polyethylene (HDPE) mesh cages filled with porous ceramsite, crushed stone or river pebbles, and has a spiral or cone-shaped anchoring structure at the bottom to adapt to different bottom materials and improve its resistance to current.

[0030] The nutrient release module is equipped with a replaceable cavity and filled with controlled-release microcapsule feed coated with chitosan or gelatin. The microcapsule particle size is preferably 1-3 mm, the release kinetics are approximately zero or near zero order, the carbon-nitrogen ratio is 10-25, and the controlled-release cycle is 10-30 days. The masking induction module uses a porous ceramic shell with a stone-like texture, preferably with a pore size distribution of 5–30 mm, and the surface roughness is... R a The size is controlled within the range of 10-50μm to promote the formation of attached algae and biofilm, while taking into account both the needs of avoiding enemies and feeding. The quick-installation and quick-disassembly module enables modular and rapid deployment and maintenance through sliding rail slots, plug-in interfaces, or magnetic structures.

[0031] Furthermore, the mesh cage material of the stabilizing base module can be selected from 304 or 316 stainless steel wire mesh, or HDPE wire mesh.

[0032] By employing the above technical solution, the present invention provides a method for selecting benthic animals based on substrate feature matching and a habitat structure device thereof, which has at least the following beneficial effects: 1. This invention achieves parameterization of substrate response, integration of ecological windows at the same scale, and translation of engineering parameters within a unified measurement space, thus forming a closed loop of "seed selection - timing - deployment - operation and maintenance"; through the stabilization, nutrient supply, and multi-scale shading design of the multifunctional device, it significantly improves the initial survival rate and colonization stability; through the feedback update mechanism, it enables parameters to adapt to regional differences and temporal fluctuations, thereby ensuring ecological effectiveness while taking into account engineering economics.

[0033] 2. This invention enables quantitative coupling of "substrate-species response-ecological window" within a unified metric space, and directly maps the optimized results to on-site deployment and operation and maintenance parameters to improve colonization success rate and community stability. This significantly enhances colonization success rate and community stability, is applicable to various water types such as lakes, rivers, and wetlands, and allows for iterative parameter optimization based on monitoring feedback to improve long-term restoration effects and operational economics.

[0034] 3. Based on the target water body sediment factor set, this invention combines a species response parameterized database and environmental adaptation window modeling to form a quantifiable and sortable optimization process, and outputs a parameter set that can be directly used for engineering deployment. At the same time, it is equipped with an integrated habitat structure device that provides anti-current stabilization, controlled-release nutrient supply and multi-scale shelter, and has the ability to update the model under monitoring feedback. Attached Figure Description

[0035] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of the method for selecting benthic animals in this invention; Figure 2 This is a schematic diagram of the benthic animal habitat structure device in this invention; Figure 3 This is a partial detail view of the benthic animal habitat structure device of the present invention, which includes a quick-assembly and quick-disassembly module; Figure 4 This is a partial detail view of the anchoring spikes set at the bottom of the stabilizing base module in this invention.

[0036] In the diagram: 1. Stabilizing base module; 2. Nutrient release module; 3. Shading and induction module; 4. Quick-install and quick-disassemble module. Detailed Implementation

[0037] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0038] This embodiment proposes a method for selecting benthic animals based on substrate feature matching and a habitat structure device thereof, such as... Figure 1 - Figure 4 As shown, the benthic animal selection method collects data on bottom sediment factors such as particle size, organic matter, nutrients, pH, conductivity, and porosity of the target water body. After unifying the dimensions, the data are compared with a species response database to calculate a bottom sediment matching score. An selection index is generated by combining ecological windows such as water temperature, flow velocity, disturbance, and transparency, outputting the target species, suitable release time, and device deployment parameters. The benthic animal habitat structure device consists of a stable base module 1, a controlled-release nutrient module 2, a porous sheltering and induction module 3, and a quick-assembly and quick-disassembly module 4. It has anchoring and stabilization functions, provides nutrients, and avoids predators, supporting quick assembly and disassembly and array deployment. Compared with existing technologies, this invention achieves integrated decision-making of "species selection-timing-deployment," significantly improving the colonization success rate and community stability. It is applicable to various water types such as lakes, rivers, and wetlands, and can iteratively optimize parameters based on monitoring feedback to improve long-term restoration effects and operational economy.

[0039] Example 1: To verify the feasibility and operability of the benthic animal selection method based on substrate feature matching, a restored wetland area on the western shore of Chaohu Lake in Anhui Province was selected as the experimental setting. Substrate measurements, species response matching, and comprehensive fit score calculations were conducted. The nearshore substrate types in this area exhibit a small-scale mosaic pattern, which is representative and suitable for verifying species selection driven by substrate differences.

[0040] Three representative sampling points were set up at the experimental site under identical meteorological and hydrological conditions, corresponding to silty bottom (CH-1), silty bottom (CH-2, originally recorded as "clay" in the field, but the D50 value better reflects the silty description), and sandy bottom (CH-3). Bottom sediment parameters were determined using a combined in-situ and laboratory method, including organic matter (OM, %), median particle size D50 (μm), pH, redox potential (ORP, mV), and water holding capacity (WH, %). In this embodiment, ORP was used instead of conductivity as one of the "at least five parameters," still meeting the lower limit requirement for the parameter items. The original observation values ​​are shown in Table 1.

[0041] Table 1. Original values ​​of sediment parameters for representative sample points

[0042] First, the substrate vector is standardized to a unified metric space to eliminate the influence of dimensions and span. Considering the small sample size in this embodiment and the fact that the distribution of each dimension may not be close to normal, Min-Max normalization is preferably used, and its dimension-wise form is as follows: in, For sample points j In terms of indicators k The original value on, , These are the extreme values ​​of the index at the three sample points. The normalized results are shown in Table 2 (rounded to 3 decimal places).

[0043] Table 2. Min-Max standardized vector P′ of sediment parameters

[0044] Candidate species were selected from five common resident groups in Chaohu Lake and its tributaries: S1- *Viburnum cyrtonema* (high organic content, fine-grained, slightly reducing, and relatively high water content), S2- *Shrimp molluscium* (widely adaptable, slightly oxygenated, and relatively coarse-grained), S3- *Bellamya affinis* (medium particle size and medium nutrition), S4- *Clam slug* (filter-feeding, low organic content, coarse sand, and oxygenated), and S5- *Bugula spp.* (extremely high organic content, extremely fine-grained, and strongly reducing environment). Species responses were parameterized in a standardized space as "optimum-tolerance-weight," where the weight vector adopted [w...]. OM ,w D50 ,w pH ,w ORP,w WH The response function is defined as [0.25, 0.20, 0.15, 0.25, 0.15], satisfying the constraint that the sum is 1. Preferably, the response function... f A Gaussian model is used to express the ecological implications of achieving maximum fitness at the optimal value and then rapidly decaying with deviation: Among them, ,and f ∈[0,1].

[0045] The corresponding substrate matching score is: The final score is represented as 100×M(i) for easy interpretation.

[0046] Table 3 shows the optimal values ​​and tolerance values ​​(all within the 0–1 normalized space) for the five species in five dimensions. The values ​​are provided as demonstration parameters based on common sense in species ecology. For field engineering applications, they can be replaced by values ​​determined by a combination of regional monitoring, semi-natural domestication, and literature data.

[0047] Table 3 Species response parameters (normalized space) m,s )

[0048] To ensure the reproducibility of the scoring process, a step-by-step calculation example is given below. Taking CH-1 (silty) and S1 (round snail) as examples, the standardized vector is (1.000, 0.000, 0.000, 0.000, 1.000), which is then substituted into Table 3. (m,s) Dimensional computation f The scores for OM (Optimal Mean Square) are 0.8825, D50 (Digital Mean Square) is 0.9231, pH (Potential Mean Square) is 0.2560, ORP (Optimal Mean Square) is 0.9231, and WH (Wholeness Mean Square) is 0.8825. Weighted summation yields M = 0.807, which multiplied by 100 gives 80.7 points. The remaining combinations are calculated using the same method, and the results are shown in Table 4.

[0049] Table 4. Matching scores of species at the three substrate sampling sites (out of 100)

[0050] Optimization is performed based on a combined threshold-quota strategy. If the threshold... i =0.6 and quota k=2, then under CH-1 conditions, tubifex worms and round snails are preferred; under CH-2 conditions, bell-shaped snails and round snails (or black-shelled shrimp, depending on management objectives and redundancy) are preferred; under CH-3 conditions, snails and clams and black-shelled shrimp are preferred. The ranking is consistent with ecological expectations: silty-high organic-low oxygen environments favor oligochaetes and low-oxygen-tolerant snails; sandy-oxygenated-relatively coarse-grained environments favor filter-feeding bivalves and more adaptable crustaceans; under moderate conditions, intermediate species (bell-shaped snails) are dominant.

[0051] To verify the engineering feasibility and predictive effectiveness of the method, a small-scale colonization experiment was conducted in the same experimental area, in which each recommended species was introduced for each substrate. n =50 individuals, repeat r The scheme with a ratio of 3 was followed up for 3 months, and the survival and colonization status were recorded and significance was determined using Fisher's exact test. α =0.05). Preferably, under CH-3 (sandy) conditions, the average survival rates of black shrimp and snail clams were approximately 92% and 86%, respectively (all three replicates were above 80%), while the survival rate of tubifex worms, as a control, was approximately 26% under the same substrate. Under CH-1 (silty) conditions, the survival rates of tubifex worms and round snails were significantly higher than those of black shrimp and snail clams. The above differences reached statistical significance under the conditions of this embodiment, indicating that the matching score can effectively indicate species-substrate compatibility and guide engineering release and deployment.

[0052] This embodiment illustrates the computational path and result interpretation of the method of the present invention. In practical engineering applications, the ecological window adaptation function can be introduced following the same process. (Using water temperature, flow velocity, disturbance frequency, and transparency as independent variables, normalized to [0,1] after fuzzy logic or regression / Bayes modeling), and matched with the bottom sediment score. according to α The weights ∈ [0.4, 0.8] are fused to obtain the optimal index U( i )= α M( i )+(1- α )W( i To adapt to seasonal and time-varying hydrodynamic conditions. Based on U( i The final optimized result can be further translated into a set of habitat structure deployment parameters (including anchoring method and depth, spacing and array, deployment posture and nutrient supply cycle), which are then implemented in conjunction with the habitat structure device of this invention to promote rapid colonization and community stability of the target species. The above values ​​and parameters are illustrative and do not constitute a limitation on the scope of protection. For regional applications, adjustments can be made based on local monitoring and testing. μ、σ、w、α Wait until recalibration is performed.

[0053] Example 2: To verify the application effect of the benthic animal selection method based on substrate feature matching and its habitat structure device in a practical engineering scenario, a field test area was set up on the northwest shore of Dongda Lake in Anqing City, Anhui Province. This area is affected by historical pollution and differences in recharge, resulting in significant substrate heterogeneity. Field surveys and in-situ tests revealed two distinct substrate types: one is a silty substrate with high organic matter content and fine particle size (labeled Substrate I, OM=12.5%, D50=45μm, pH=6.8), and the other is a sandy substrate with high sand content and high permeability (labeled Substrate II, OM=4.2%, D50=320μm, pH=7.3). The average water depth in the test area was approximately 1.2m, and the near-bottom flow velocity was less than 0.05m / s. No strong disturbances such as strong winds, waves, or gate operation occurred during the test period.

[0054] The optimization method first standardizes and maps the two types of substrates based on substrate factor vectors and generates a target species set Ω by calling a species response database. Under substrate I conditions, the optimization method tends to output groups that are tolerant of low oxygen, have finer grains, and are high in organic matter as candidates. Under substrate II conditions, the optimization method tends to output groups that have higher oxygen requirements, stronger filter feeding and sand-boring abilities as candidates. The optimization method further generates a set of deployment parameters Ψ based on water depth, flow velocity, and substrate carrying capacity, and customizes and deploys habitat structures accordingly. The device consists of a stable base module, a shelter-inducing module, and a nutrient release module. Preferably, the shelter-inducing module uses a porous ceramic shell with a stone-like texture, a pore size distribution of 5-30 mm, and a surface roughness of... R a The particle size is controlled at 10-50 μm to enhance the formation of attached algae and biofilm. The nutrient release module is filled with controlled-release microcapsules coated with chitosan or gelatin, with a carbon-to-nitrogen ratio of 10-25, a release cycle of 10-30 days, and a particle size of 1-3 mm, providing initially available organic matter in small doses at equal intervals. For the stabilizing base module, spiral anchoring is preferred in substrate I, with the insertion depth set at 0.6-1.0 times the shell height; for substrate II, conical anchoring is preferred, with the insertion depth set at 0.3-0.6 times the shell height. Furthermore, the device array is arranged along the main flow direction, with a spacing of 2-3 m between devices. The long axis of the shell is parallel to the main flow direction and maintains a small attitude angle of 0-15° to obtain a stable near-bottom flow field while accommodating both water flow and predator avoidance space.

[0055] The experiment included five sites for the device group and five sites for the control group in each type of substrate. Random starting points and fixed spacing were used to control spatial effects. The device groups were deployed according to parameter set Ψ, while the control groups remained in their original substrate state without any devices. The monitoring period was 60 days, with sampling and recording every 10 days, including benthic fauna composition, individual density, and dry biomass. A 0.1m sampler was used. 2The sampled material (either a frame mesh or a columnar mud sampler of equivalent area) was washed off with a 300μm sieve on-site, then refrigerated and brought back to the laboratory for microscopic examination, identification, drying, and weighing. Density was calculated to ind / m² based on the sampling area. 2 Biomass is converted to g / m³ by dry weight. 2 To avoid duplicate counting, stratified sampling and fixed-trajectory repeated location methods were employed. Statistical analysis compared the means of five replicates between the device group and the control group. Differences were assessed using the Mann-Whitney U test or Fisher's exact test, with a significance threshold set at [missing value]. α =0.05; diversity is recommended using the Shannon index. H′ Pielou uniformity J′ express.

[0056] Monitoring results showed that the device significantly increased the individual density and biomass in both types of substrate and altered the dominant species structure. To ensure data readability and reflect the contributions of non-target groups, Table 5 also presents snail density, shellfish density, total density and biomass, as well as changes in dominant species. Total density is not equal to the sum of snail and shellfish density; the difference represents the contribution of other groups (such as oligochaetes and chironomid larvae).

[0057] Table 5. Key monitoring data for each group (end of 60 days)

[0058] To facilitate comparison of effects, the increase ratio and growth rate were calculated. In substrate I, the total density increased from 89 to 264, a 2.97-fold increase (+196.6%), and the biomass increased from 34.5 to 87.2, a 2.53-fold increase (+152.8%). In substrate II, the total density increased from 77 to 215, a 2.79-fold increase (+179.2%), and the biomass increased from 29.8 to 73.6, a 2.47-fold increase (+147.0%). The average increase ratios for the two substrate types were approximately 2.9 times (total density) and 2.5 times (biomass), respectively. The promoting effect of the structure on shellfish settlement was particularly significant in substrate II, where shellfish density increased from 38 to 109, an increase of approximately 186% (2.87 times); snail density increased from 24 to 83, an increase of approximately 246% (3.46 times). This demonstrates that the stable substrate and multi-scale porosity provided by the device significantly promoted the adsorption and settlement of aerobic filter-feeding groups. The dominant species in the device group shifted from a single dominant species to "dual dominance" or "co-dominance," reflecting the regulatory effect of the integrated microenvironment of shading, attachment, and nutrient supply on community structure.

[0059] The results were consistent with the expected outcomes of the optimized method: under silty-high organic-low pH conditions, the device group promoted the recovery of fine-grained and anaerobic-facultative taxa and significantly increased overall biomass; under sandy-permeable-near-neutral conditions, the device group significantly promoted the settlement of filter-feeding bivalves and highly adaptable crustaceans, and the dominant species shifted from a tax-tolerant taxa group (shellfish C) to a group with ecological sensitivity and resource value (shellfish D). These differences were statistically significant (p<0.05) in all five replicates of this embodiment, indicating that the device and the set of deployment parameters Ψ have a clear positive effect on community reconstruction. To ensure operability, the device was supplemented with an equal amount of microcapsules every 14-21 days during the experimental period, with 20-50g added per cycle per device. No risk of localized eutrophication or anaerobic blackening and odor was observed.

[0060] To reduce confounding factors, this embodiment underwent equal-intensity disturbance reduction treatment (light raking once to eliminate historical bioattachment differences) on both the device and control areas before the experiment, and the spacing between the sampling points was maintained at ≥10m to reduce recharge interference. During monitoring, temperature, transparency, and near-bottom flow velocity were recorded to confirm the absence of abnormal events. It is recommended that in subsequent engineering applications, the ecological window adaptation function W(i) and weight parameters be introduced simultaneously using the described method. α The optimal results and deployment parameters are continuously updated under seasonal-hydrodynamic variations; and the Shannon index is introduced. H′ With uniformity J′ Incorporating it into monitoring as a routine indicator allows for the quantification of the device's long-term effects at the community structure level.

[0061] The above embodiments demonstrate that the optimization method based on substrate characteristic matching can provide a clear set of target species Ω and a set of deployment parameters Ψ for on-site observation. The habitat structure device, through the combined functions of "stable base + porous shelter + controlled-release nutrient supply," achieves quantifiable and verifiable increases in density and biomass on two typical substrate types, and realizes positive regulation of dominant population structure, verifying the engineering feasibility and promotional value of the invention under complex substrate conditions. It should be noted that the pore size, roughness, anchoring depth, feeding cycle, and dosage given in this embodiment are preferred parameters. On-site adjustments can be made within the stated range according to the near-bottom hydrodynamics and carrying capacity of the target water body, without deviating from the essence of the invention.

[0062] Example 3: To verify the effectiveness of the benthic animal selection method based on substrate feature matching and its habitat structure device in riverine water bodies, an experimental area was set up in the middle and lower reaches of the Nanfei River in Hefei City, Anhui Province (117.26°E, 31.82°N). Based on the regionalized species response database formed in Example 1, a candidate species set was selected, and supporting devices were deployed under different substrate regulation conditions to test the closed-loop path of "substrate-species-window-deployment".

[0063] The test river section was approximately 20m wide, with a water depth of 0.8–1.2m and a main flow velocity of approximately 0.2m / s. Prior to sampling, sediment samples and water parameters were obtained from representative locations in the three types of colonization areas. Preferred sediment factors included organic matter (OM), median particle size (D50), pH, (either conductivity or ORP), and water holding capacity (WH), forming a sediment factor vector. ( n ≥5). To achieve comparison at the same scale, the vector is standardized dimension by dimension. Preferably, in this embodiment, Min-Max normalization is used to obtain the standardized substrate factor vector. If some dimensions contain zero / negative values ​​or are significantly skewed, a translation constant c>0 is first introduced into that dimension before applying a Box-Cox transformation, in the form of: Within a standardized space, the species response database is invoked. S The optimal values ​​for each substrate factor dimension were applied to target species A (Asian dragonfly nymphs) and B (schizothorax spp.). m Tolerance s( 0) and weight vector W The weights satisfy w ≥0 and Σ w =1. Response function f Using the Gaussian form (hyperbolic tangent or penalized exponential form can also be selected), it is defined as: Based on this, the substrate matching score is calculated: Substitute into the three types of colonization areas respectively and m , s , w The substrate matching score M for species A and B was calculated, and the selected species set Ω was selected using a threshold-quota strategy; threshold i The number of slots can be determined by the Youden index of the historical verification set of this region. k Take 1-2. If MA ≥ MB and MA ≥ i If species A is introduced first, then species B is introduced first; otherwise, species B is introduced; if both are ≥ i They can coexist and live together.

[0064] To incorporate temporal environmental impacts, the ecological window parameter E=[T, V, D, S] was obtained. d ], where T is water temperature, V is flow velocity, D is bottom sediment disturbance frequency (or over-threshold event count / near-bottom shear proxy), and S is... d The depth of the Seychelles disk. Ecological window adaptation function. W ( iThe region ∈ [0,1] is established using fuzzy logic reasoning: T, V, D, S d Define three or more membership functions (low / medium / high) to build a rule base (e.g., "V=medium and S..."). d =High→W=High), using the centroid method to resolve fuzziness and obtain continuous output; alternatively, parametric regression or a Bayesian model can be used, and the output can be normalized to [0,1]. Finally, the optimal index is obtained through fusion: According to the preferred index The sequencing determines the target species set Ω for each zone and generates a habitat structure deployment parameter set Ψ. The habitat structure deployment parameter set Ψ includes at least: anchoring method and depth into the mud, number of device points and spacing between points, deployment depth and shell attitude angle, nutrient release cycle and feeding intensity. Preferably, in river sections with a significant mainstream direction, the long axis of the device is aligned with the mainstream direction, with an attitude angle of 0-15°; when the bottom bearing capacity is low or the disturbance risk is high, spiral anchoring and increased depth into the mud are preferred; the spacing between points is 2-3m to balance near-bottom water flow and shelter; nutrient release adopts a small-dose, equally spaced strategy, with microcapsules C / N=10-25, a cycle of 10-30 days, and a particle size of 1-3mm.

[0065] This embodiment sets up three types of colonization areas: Area A has an optimized substrate of "clay + organic matter + gravel" and is equipped with a three-layer structure (stabilization + shelter + nutrition). The target species are selected according to the selection index. The sorting was selected in area A; area B consisted of natural substrate and was equipped with a standard planar device, with the target species sorted according to the selection index. The sorting was selected in area B; area C was for optimizing the substrate and installing a three-dimensional device, according to the optimization index. The sorting is A and B coexisting. Three repeating units are deployed in each area, with a spacing ≥3m, for a period of 30 days. The shading induction module has an aperture of 5-30mm and a surface roughness of... R a =10-50μm, nutrient release module controlled release cycle 10-30 days, microcapsule particle size 1-3mm, C / N 10-25, base material 304 / 316 stainless steel or HDPE, equipped with water depth adjustment bracket and attitude stabilization hinge structure.

[0066] To evaluate the project's effectiveness and provide a basis for online updates, four monitoring indicators were set: biomass per unit area (g / m²). 2 The regression rate (%), sediment stability index, and sediment nitrogen and phosphorus release reduction rate (%) were obtained and normalized to the maximum value. Comprehensive evaluation indicators: The bottom stability index is defined as follows: It can be seen that at the end of day 0, the average biomass of representative units in zones A, B, and C were as follows: Zone A biomass 92.4 g / m³ 2 The regression rate was 87.5%, pH remained stable with a 12.5% ​​decrease in OM, and nitrogen and phosphorus release decreased by 28.6%; Zone B had a concentration of 58.1 g / m³. 2 The regression rate was 64.2%, pH fluctuated with no significant change in OM, and the nitrogen and phosphorus release reduction rate was 11.3%; the concentration in zone C was 97.6 g / m³. 2 The regression rate was 90.1%, pH remained stable with a 15.2% decrease in OM, and nitrogen and phosphorus release decreased by 33.1%. After normalization and substituting into the above formula, we obtain S≈0.929 for region A, S≈0.455 for region B, and S=1.000 for region C. Statistical tests (...) α =0.05, Mann-Whitney U or Fisher exact test) shows that regions A and C are significantly better than region B. According to the "model update" of this invention, the result of S is used as a trigger signal to adjust the database parameter weights. Response tolerance With scene adaptation factor α or temperature parameter β Perform iterative optimization and select the optimal index in the next round. It takes effect in the calculation and Ψ generation, realizing a closed loop of "optimization-deployment-feedback-optimization".

[0067] The core of this invention lies in: constructing a sediment factor vector P using measured sediment parameters of the target water body, and then... P Dimensional standardization to a unified metric space P′ Construct a species response database within the same standardized space for each candidate species. With substrate factors Give the optimal value Response tolerance With weight Define a unimodal response function with a range of [0,1]. f Calculate the substrate matching score Obtain ecological window parameters Construct an ecological window adaptation function ∈[0,1], and the optimal index obtained by fusion at the same scale Generate a target species set Ω; map the target species set Ω to a habitat structure layout parameter set Ψ for the project, and deploy modular habitat devices for implementation to promote colonization and community stability; during operation, adjust the weights based on monitoring feedback. Response tolerance With scene adaptation factor α or temperature parameter β Iterative updates are performed to form a closed loop of "optimization-deployment-feedback-optimization" to continuously improve the accuracy of optimization and the success rate of colonization.

[0068] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0069] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Since the above embodiments are substantially similar to the method embodiments, they are described relatively simply; relevant parts can be referred to the descriptions of the method embodiments.

[0070] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for selecting benthic animals based on substrate feature matching, characterized in that, The method includes the following steps: Multiple representative sediment samples were collected from the target water body, and sediment factor data including at least five parameters were measured to form a sediment factor vector. ; For the substrate factor vector Standardization is performed dimension by dimension to construct a standardized foundation factor vector of a unified metric standardization space. ; Constructing a species response database within the standardized space And calculate the species response database. Any species Substance matching score ; Base quality matching score Sort by high to low, using a threshold. - Quota Combined candidate strategies, selection species If the number is insufficient to meet the quota Then score according to the substrate quality. Fill in the blanks from highest to lowest. indivual; Ecological window parameters of the target water body's environmental window period are obtained during the candidate phase. Establishment for characterizing species Ecological window parameters Ecological window fitness function of fitness ; Base quality matching score Adaptation function with ecological window Same-scale fusion yields the optimal index And based on the preference index The ordering determines the target species set Ω, and the selection... All species or pre- One species; Based on the target species set Ω and the water depth, flow velocity and substrate carrying capacity of the target water body, a habitat structure layout parameter set Ψ is generated. Within the adaptation area, habitat structure devices are deployed according to the habitat structure deployment parameter set Ψ, and target benthic animals indicated by the target species set Ω are introduced within the ecological adaptation window to promote their colonization and population stability.

2. The method for selecting benthic animals according to claim 1, characterized in that, The sediment factor data include particle size distribution, organic matter mass fraction, total nitrogen and total phosphorus concentrations, pH value and electrical conductivity, and sediment porosity or water content.

3. The preferred method for benthic animals according to claim 1, characterized in that, The standardization process employs one or a combination of the following methods: Z-score transformation is suitable for cases where the single-dimensional sample size is ≥30 and approximately normally distributed. Min-Max normalization is suitable for situations where the dimensions of multidimensional indicators have a large span but the dimensional distribution is balanced. The Box-Cox transform with translation constants is used to first introduce translation constants for dimensions that have zero or negative values. This makes it applicable to all samples Its transformation is defined as: when hour, ; when hour, ; in, The Box-Cox transformation factor; for s The sample at the th j Observed values ​​of each substrate factor; This is the power parameter for that dimension; ln It is the natural logarithm.

4. The preferred method for benthic animals according to claim 1, characterized in that, The species response database For any candidate species Store its effects on various substrate factors optimal value With response tolerance And configure weight vectors that reflect relative importance. Satisfying weight and To ensure cross-dimensional comparability; The optimal value Response tolerance With substrate factors They exist in the same standardized space; The substrate matching score The sediment matching score is used to characterize the compatibility between a species and the substrate of a target water body. The calculation formula is: ； in, f It is a unimodal response function that satisfies the condition that It reaches a maximum value of 1 at time, and then... Monotonically decreasing, therefore ; This represents the total number of standardized substrate factor vectors.

5. The preferred method for benthic animals according to claim 1, characterized in that, The threshold Determined by the Youden index or the inflection point of the receiver operating characteristic curve on the historical validation set; The quota The carrying capacity of the target water body is initially estimated or determined by the minimum number of cover species.

6. The preferred method for benthic animals according to claim 1, characterized in that, The ecological window adaptation function Established using fuzzy logic reasoning, including the ecological window parameters The mapping is based on membership degree, fuzzy reasoning is performed based on the rule base, and the ecological window adaptation function is obtained by defuzzifying using the centroid method. ; Alternatively, a parametric regression model or a Bayesian response model can be used to establish the model, based on the historical distribution frequency of candidate species and ecological window parameters. The paired samples are used as training data, and the output is normalized to [0,1] to match the bottom sediment score. Same scale.

7. The method for selecting benthic animals according to claim 1, characterized in that, The same-scale fusion yields an optimal index. Using a linear weighted form, that is: ； in, For scene adaptation coefficients; Alternatively, a geometric fusion approach can be adopted: ； Or, for the preferred index Or the output of the second layer network The ranking weights are generated using softmax, i.e.: ； in, , is a temperature parameter used to adjust the discrimination index; For species i In the candidate set Ω cand The softmax sorting weights satisfy the following conditions: ); The input score for softmax is taken from the candidate index. U ( k (or taken from the output of the second-layer network;) k For the candidate set Ω cand The index in.

8. The preferred method for benthic animals according to claim 1, characterized in that, The set of habitat structure deployment parameters Ψ includes at least: anchoring method and anchoring depth, number of single-point devices and spacing between points, deployment depth and attitude angle, nutrient release cycle and feeding frequency; The anchoring method and the depth of penetration into the mud are selected based on the risk classification of near-bottom shear or disturbance. In areas with high flow velocity or frequent disturbance, spiral anchoring is selected and the depth of penetration into the mud is increased. In medium and low risk areas, cone anchoring is used. The number of single-point devices and the spacing between them are estimated based on the effective coverage area of ​​the target and the radius of influence of the devices, and linear arrays are selected in combination with the main flow direction, or grid / cluster arrays are used in low flow areas to increase habitat heterogeneity. The deployment depth and shell attitude angle selection ensure that the long axis of the shell is parallel to the main direction and maintains a small pitch angle. The nutrient release cycle and feeding frequency adopt a small-dose, equally spaced strategy, with a cycle of 10–30 days and a carbon-to-nitrogen ratio of 10–25, and are adjusted according to the ecological window adaptation function. The intensity is finely adjusted to balance induction efficiency and water quality safety.

9. A benthic habitat structure device applied to the preferred method for benthic animals according to any one of claims 1-8, characterized in that, The benthic animal habitat structure is equipped with a water depth adjustment support and a posture-stabilizing hinge structure to adapt to angle fine-tuning and stable deployment under different slope or depth conditions, and also includes: The stabilizing base module (1) is made of stainless steel or high-density polyethylene mesh cage and filled with porous ceramsite, crushed stone or natural river pebbles; the bottom of the stabilizing base module (1) is provided with insertable cone spikes or spiral anchoring structures to adapt to different bottom materials and achieve anti-flow stabilization. The nutrient release module (2) is set inside the stable base module (1) and is filled with microcapsules of controlled-release organic bait coated with chitosan or gelatin. The carbon-nitrogen ratio of the microcapsules is 10-25 and the controlled-release cycle is 10-30 days. The shielding and guidance module (3) is set on top of the stable base module (1). It is made of a ceramic porous shell with a stone-like surface texture. The shell pore size is distributed from 5 to 30 mm, providing multi-scale enemy avoidance and attachment space. The quick-install and quick-release module (4) provides a sliding rail slot, plug-in interface or magnetic structure between the stable base module (1), the nutrient release module (2) and the shielding guidance module (3) to realize modular quick deployment and maintenance.

10. The benthic animal habitat structure device according to claim 9, characterized in that, The surface roughness of the shading induction module (3) R a The size should be controlled between 10-50 μm to enhance the ability of attached algae and biofilm to form. The mesh material of the stable base module (1) is 304 or 316 stainless steel wire mesh, or HDPE wire mesh. The microcapsule size of the nutrient release module (2) is 1-3 mm, and the release kinetics are approximately zero or near zero order.

Images