Wild yak habitat suitability evaluation method and device and storage medium

Abstract

The invention relates to the technical field of animal ecological protection, in particular to a wild yak habitat suitability evaluation method and device and a storage medium. According to the method, environment variables influencing the habitat of the wild yak are specifically screened according to life habits of the wild yak, geographical factors of the habitat of the wild yak and the like, so that the accuracy of an evaluation result of the suitability of the MAXENT model to the habitat of the wild yak can be remarkably improved; the precision of the MAXENT model is evaluated through the ROC curve, so that the MAXENT model with higher precision can be selected to evaluate the habitat suitability of the wild yaks, and the evaluation accuracy is further improved; the evaluation method is used for evaluating the wild yak habitat suitability, has the advantages of being high in evaluation accuracy and good in universality, is suitable for effective evaluation of the wild yak habitat suitability, and plays a positive role in ecological protection of the wild yak habitat.

Description

Technical Field

[0001] The invention relates to the technical field of animal ecological protection, in particular to the ecological protection of wild yaks, and in particular to a method and device for evaluating the suitability of wild yak habitats and a storage medium. Background Art

[0002] Wild yaks live in high altitude areas of 3,000-6,000 meters, living in a specific high-altitude alpine climate zone, and adapting to the unique ecological and geographical environment, such as alpine tundra, alpine meadows, alpine deserts, river valley forests, etc. They have a strong adaptability to the high-altitude oxygen-deficient environment and can endure the extreme cold of -38 degrees Celsius.

[0003] Wild yaks are gentle, gregarious, reflexive, and resistant to disease, and are known as the "boats of the plateau". Wild yaks can also roll over, a habit not found in other cattle species. Wild yaks have good working performance, carrying 50-100 kg of load in alpine grassland areas and traveling 20-30 kilometers a day. In addition, they can be used for riding, plowing, pulling carts, etc., and their service life can reach 15 years.

[0004] With the increase of human activities in the plateau, the habitat of wild yaks is under serious threat. The evaluation of the suitability of wild yak habitat plays an important role in assessing the impact of human activities on wild yak populations. Through the evaluation of the suitability of wild yak habitat, it can better provide basic information and basis for the protection of wild yak populations.

[0005] At present, the habitat suitability evaluation of wild species mainly adopts MaxEnt model, niche model, PCA and AHP, entropy weight method, generalized additive model, multi-source remote sensing satellite data fusion method, etc. However, the accuracy of habitat suitability evaluation results is not only affected by the type of evaluation model selected, but also directly affected by factors such as the type, number and weight division of the selected influencing factors. Therefore, the evaluation of the suitability of different wildlife habitats requires the targeted selection of corresponding evaluation models and influencing factors to obtain the most accurate evaluation results. Summary of the invention

[0006] The purpose of the present invention is to propose a method, device and storage medium for evaluating the suitability of wild yak habitats in view of the problem that the prior art has not been able to achieve accurate evaluation results on the suitability of wild yak habitats.

[0007] The present invention provides a method for evaluating the suitability of wild yak habitat, comprising the following steps: S1. Targeted selection of environmental variables that affect the suitability of wild yak habitats; S2, substituting the environmental variables into the MAXENT model, and obtaining the logical relationship between the environmental variables and the probability of wild yaks appearing when the entropy of the MAXENT model is maximum; S3. Randomly obtain a number of points where wild yaks are detected, and construct a set based on the environmental variable data of the points; randomly divide 75% of the environmental variable data in the set into a training set, and 25% of the environmental variable data into a test set, perform several iterations on the logical relationship between environmental variables and the probability of wild yaks appearing, and use the jackknife method to determine the importance weight of each environmental variable in the MAXENT model; S4. Substitute the importance weight of each environmental variable of the wild yak habitat into the MAXENT model to obtain the wild yak habitat suitability evaluation model, and make a receiver operating characteristic curve (ROC curve) to evaluate the accuracy of the model prediction results; S5. When the accuracy of the model prediction results meets the requirements, the wild yak habitat evaluation model is used to obtain the logistic probability map of the wild yak habitat (the probability curve map in the logistic regression model), and the suitability level of the wild yak habitat is evaluated according to the logistic probability map; when the accuracy of the model prediction results does not meet the requirements, the number of points and / or the number of iterative operations are increased, and steps S3, S4 and S5 are repeated until the accuracy of the model prediction results meets the requirements.

[0008] The invention discloses a wild yak habitat suitability evaluation method. The method not only performs targeted screening on environmental variables affecting the wild yak habitat according to the living habits of the wild yaks and the geographical factors of the wild yak habitat, so that the accuracy of the suitability evaluation result of the wild yak habitat by the MAXENT model can be significantly improved; the accuracy of the MAXENT model is also evaluated by the ROC curve, so that the MAXENT model with higher accuracy can be selected to evaluate the suitability of the wild yak habitat, and the accuracy of the evaluation is further improved; the evaluation method evaluates the suitability of the wild yak habitat, has the advantages of high evaluation accuracy and good universality, is suitable for the effective evaluation of the suitability of the wild yak habitat, and has a positive effect on the ecological protection of the wild yak habitat.

[0009] Among them, in step S1, the environmental variables include land use, vegetation coverage, altitude, slope, slope aspect, minimum distance between habitat and water source, minimum distance between habitat and settlement, annual average temperature, annual maximum temperature, annual minimum temperature, annual precipitation, annual maximum monthly precipitation, and annual minimum monthly precipitation.

[0010] Preferably, the water source is at least one of a river, a lake, a reservoir, and a swamp.

[0011] Among them, in step S2, the method for obtaining the logical relationship is: Find the MAXENT model:

[0012] The maximum entropy characteristic root is the optimal solution under the constraints, that is: ; The constraints are: ; in, It represents the conditional entropy of random variable y under the condition of given environmental variable set X; is the probability distribution of the environmental variable X, is given The conditional probability of y when Represents the probability distribution given X Next, find the conditional probability The conditional entropy maximum; is a constraint condition, which ensures that is a valid probability distribution representation, that is, the sum of the probabilities of all possible y values ​​is 1; Among them, X is the set of environment variables, that is, ; y is the probability of wild yaks appearing, which is a random discrete variable; is the probability distribution of the environmental variable X, .

[0013] Among them, in step S3, preferably, the number of points is greater than 150; the larger the number of points, the higher the accuracy of the trained MAXENT model, and the higher the accuracy of the obtained wild yak habitat suitability index results.

[0014] Among them, in step S3, preferably, the number of iterative operations is more than 15 times; the greater the number of iterations, the more accurate the importance weight obtained, and the higher the accuracy of the wild yak habitat suitability index result obtained.

[0015] Among them, in step S4, preferably, the method for evaluating the accuracy of the model prediction results includes: evaluating the accuracy of the model prediction results by the AUC value (area under curve, AUC, represented by "Q" in the present invention) of the ROC curve diagram; the AUC value is the area enclosed by the receiver operating characteristic curve and the horizontal axis in the ROC curve diagram.

[0016] Preferably, the accuracy evaluation standard of the model prediction result is: when the AUC value (Q) Q < 0.6, it means that the model prediction effect is not ideal; when the AUC value is 0.6≤Q<0.7, it means that the model prediction result is poor; when the AUC value is 0.7≤Q<0.8, it means that the model prediction result is general; when the AUC value is 0.8≤Q<0.9, it means that the model prediction result is good; when the AUC value Q≥0.9, it means that the model prediction result is excellent.

[0017] Among them, in step S5, preferably, the accuracy requirement of the model prediction result is AUC value Q≥0.8; more preferably, the accuracy requirement of the model prediction result is AUC value Q≥0.9.

[0018] Preferably, the method for evaluating the suitability of wild yak habitats based on the logistic probability map includes: obtaining the probability value (denoted by "P" in the present invention) of the occurrence of wild yaks at random points in the wild yak habitat based on the logistic probability map, and judging the suitability level of the wild yak habitat by the size of the probability value.

[0019] Preferably, the judgment criteria for the suitability level of wild yak habitat are: when more than 40% of the random points in the wild yak habitat meet P≥0.9, the suitability level of the wild yak habitat is suitable; when more than 40% of the random points in the wild yak habitat meet 0.8≤P<0.9, the suitability level of the wild yak habitat is relatively suitable; when more than 40% of the random points in the wild yak habitat meet 0.7≤P<0.8, the suitability level of the wild yak habitat is generally suitable; when more than 40% of the random points in the wild yak habitat meet 0.6≤P<0.7, the suitability level of the wild yak habitat is unsuitable; when more than 40% of the random points in the wild yak habitat meet P<0.6, the suitability level of the wild yak habitat is extremely unsuitable.

[0020] The present invention also provides a device for evaluating the suitability of wild yak habitat, comprising at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned method for evaluating the suitability of wild yak habitat.

[0021] The present invention also provides a computer-readable medium on which instructions executable by a processor are stored. When the instructions are executed by the processor, the processor executes the above-mentioned method for evaluating the suitability of wild yak habitat.

[0022] Compared with the prior art, the present invention has the following beneficial effects: The invention discloses a wild yak habitat suitability evaluation method. According to the living habits of wild yaks and the geographical factors of wild yak habitats, environmental variables affecting wild yak habitats are screened in a targeted manner, thereby significantly improving the accuracy of suitability evaluation results of wild yak habitats by a MAXENT model.

[0023] The wild yak habitat suitability evaluation method of the present invention evaluates the accuracy of the MAXENT model through the ROC curve, so that the MAXENT model with higher accuracy can be selected to evaluate the suitability of the wild yak habitat, further improving the accuracy of the evaluation.

[0024] The wild yak habitat suitability evaluation method of the present invention has the advantages of high evaluation accuracy and good universality, is suitable for the effective evaluation of the suitability of the wild yak habitat, and has a positive effect on the ecological protection of the wild yak habitat. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Schematic diagram of the process of the method for evaluating the suitability of wild yak habitat in Example 1 of the present invention; Figure 2 It is a knife-cut diagram of environmental variables predicted by the MAXENT model in Example 1 of the present invention; Figure 3 is a response curve diagram of the environmental variable (land use) in the MAXENT model in Example 1 of the present invention; Figure 4 is a response curve diagram of the environmental variable (vegetation coverage) in the MAXENT model in Example 1 of the present invention; Figure 5 is a response curve diagram of the environmental variable (altitude) in the MAXENT model in Example 1 of the present invention; Figure 6 is a response curve diagram of the environmental variable (slope) in the MAXENT model in Example 1 of the present invention; Figure 7 is a response curve diagram of the environmental variable (slope aspect) in the MAXENT model in Example 1 of the present invention; Figure 8 is a response curve diagram of the environmental variable (minimum distance from the water source) in the MAXENT model in Example 1 of the present invention; Fig. 9 is a response curve diagram of the environmental variable (minimum distance from the settlement) in the MAXENT model in Example 1 of the present invention; Fig.10 is a response curve diagram of the environmental variable (annual precipitation) in the MAXENT model in Example 1 of the present invention; Fig.11 It is a response curve diagram of the environmental variable (annual maximum monthly precipitation) in the MAXENT model in Example 1 of the present invention; Fig.12 It is a response curve diagram of the environmental variable (annual minimum monthly precipitation) in the MAXENT model in Example 1 of the present invention; Fig.13It is a response curve diagram of the environmental variable (annual average temperature) in the MAXENT model in Example 1 of the present invention; Fig.14 is a response curve diagram of the environmental variable (annual minimum temperature) in the MAXENT model in Example 1 of the present invention; Fig.15 is a response curve diagram of the environmental variable (annual maximum temperature) in the MAXENT model in Example 1 of the present invention; Fig.16 It is a ROC curve diagram of the prediction results of the MAXENT model in Example 1 of the present invention; Fig.17 This is a distribution map of the suitability assessment results of wild yak habitats along a certain highway in Example 1 of the present invention. DETAILED DESCRIPTION

[0026] The present invention is further described in detail below in conjunction with specific embodiments. However, this should not be understood as the scope of the above subject matter of the present invention being limited to the following embodiments, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0027] Example 1 The method of the present invention was used to evaluate the habitat suitability of wild yaks within a 20 km radius along a highway in 2020 (see flowchart for details). Figure 1 ), the specific steps are as follows: S1. Targeted selection of environmental variables that affect the suitability of wild yak habitats, including land use, vegetation coverage, altitude, slope, slope aspect, minimum distance between habitat and water source, minimum distance between habitat and residential area, annual average temperature, annual maximum temperature, annual minimum temperature, annual precipitation, annual maximum monthly precipitation, and annual minimum monthly precipitation; S2. Substitute the environmental variables into the MAXENT model to obtain the logical relationship between the environmental variables and the probability of wild yaks appearing when the entropy of the MAXENT model is maximum; find the MAXENT model:

[0028] The maximum entropy characteristic root is the optimal solution under the constraints, that is: ; The constraints are ; Among them, X is the set of environment variables, that is, ; y is the probability of wild yaks appearing, which is a random discrete variable; The probability distribution of , ; S3. Randomly obtain 100 monitoring points where wild yaks are detected, and construct a set based on the environmental variable data of the points; randomly divide 75% of the environmental variable data in the set into a training set, and 25% of the environmental variable data into a test set, perform several iterations on the logical relationship between environmental variables and the probability of wild yaks appearing, and use the jackknife method to determine the importance weight of each environmental variable in the MAXENT model ( Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure 7 , Figure 8 , Fig. 9 , Fig.10 , Fig.11 , Fig.12 , Fig.13 , Fig.14 , Fig.15 ); S4. Substitute the importance weight of each environmental variable of the wild yak habitat into the MAXENT model to obtain the wild yak habitat suitability evaluation model, and make a receiver operating characteristic curve (ROC curve) diagram ( Fig.16 ) The accuracy of the model prediction results was evaluated (the AUC value was 0.982, and the prediction result was excellent); The accuracy evaluation method includes: evaluating the accuracy of the model prediction results through the AUC value of the ROC curve; the AUC value is the area enclosed by the receiver operating characteristic curve and the horizontal axis in the ROC curve; The accuracy evaluation criteria of the model prediction results are as follows: when the AUC value (Q) Q < 0.6, it means that the model prediction effect is not ideal; when the AUC value is 0.6≤Q < 0.7, it means that the model prediction result is poor; when the AUC value is 0.7≤Q < 0.8, it means that the model prediction result is average; when the AUC value is 0.8≤Q < 0.9, it means that the model prediction result is good; when the AUC value Q ≥ 0.9, it means that the model prediction result is excellent; S5. When the accuracy of the model prediction results meets the requirements (AUC value Q ≥ 0.8), the wild yak habitat evaluation model is used to obtain the logistic probability map of the wild yak habitat (the probability curve map in the logistic regression model), and the suitability level of the wild yak habitat is evaluated according to the logistic probability map to obtain the evaluation results (Table 1, Fig.17 ); When the accuracy of the model prediction result does not meet the requirements, the number of points and / or the number of iterative operations are increased, and steps S3, S4 and S5 are repeated until the accuracy of the model prediction result meets the requirements; The suitability evaluation method includes: obtaining the probability value of wild yaks appearing at random points in the wild yak habitat according to the logistic probability map, and judging the suitability level of the wild yak habitat by the size of the probability value.

[0029] The judgment criteria for the suitability level of wild yak habitat are as follows: when more than 40% of the random points in the wild yak habitat meet P≥0.9, the suitability level of the wild yak habitat is suitable; when more than 40% of the random points in the wild yak habitat meet 0.8≤P<0.9, the suitability level of the wild yak habitat is relatively suitable; when more than 40% of the random points in the wild yak habitat meet 0.7≤P<0.8, the suitability level of the wild yak habitat is generally suitable; when more than 40% of the random points in the wild yak habitat meet 0.6≤P<0.7, the suitability level of the wild yak habitat is unsuitable; when more than 40% of the random points in the wild yak habitat meet P<0.6, the suitability level of the wild yak habitat is extremely unsuitable.

[0030] Table 1 Evaluation results of wild yak habitat suitability .

[0031] Results: The habitat distribution of wild yaks decreases with the increase of altitude, and the suitable range is 3000m-3500m. The suitable range of slope is 0-10°, and it gradually decreases and tends to be stable when it exceeds 5°. The habitat selection of wild yaks is affected by human disturbance factors. The probability of wild yaks appearing in areas close to water sources is higher. As the distance from residential areas increases, the probability of wild yaks appearing shows a trend of gradually decreasing. As the annual average temperature increases, the probability of wild yaks appearing gradually increases. The appearance of wild yaks shows a stable-decreasing trend with the change of annual precipitation. When choosing habitats, wild yaks tend to choose vegetation types that are more targeted. According to field survey results, wild yaks are distributed along the K0-K175 section, where the main vegetation is subalpine meadows and shrubs. There are many plateau pastures with a large number of wild yaks. Some wild yaks are also found mixed together, and they almost all move within 1,500 meters from water sources. The K175-249 section of a certain highway is mainly composed of cloud, fir and mixed forests, and the terrain is mostly alpine woodland, which is not suitable for the activities of ungulates such as wild yaks.

[0032] The evaluation results show that the area of ​​suitable habitat for wild yaks along a highway is about 280.67 km², and the area of ​​relatively suitable habitat is about 1046.19 km², accounting for 13.69% of the total area of ​​the study area. Most wild yaks are distributed in the central, western and central areas along a highway, concentrated in areas close to rivers, and are extremely unsuitable to occupy most areas along the highway. Therefore, the proportion of suitable habitats available to wild yaks is low.

[0033] Example 2 A device for evaluating the suitability of wild yak habitats includes a processor and a memory connected to the processor in communication; the memory stores instructions that can be executed by the processor, and the instructions are executed by the processor so that the processor can execute a method for evaluating the suitability of wild yak habitats in Example 1.

[0034] Example 3 A computer-readable medium stores instructions executable by a processor, and when the instructions are executed by the processor, the processor executes a method for evaluating the suitability of a wild yak habitat in Example 1.

[0035] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating the suitability of wild yak habitat, characterized in that: The following steps are involved: S1. Targeted selection of environmental variables that affect the suitability of wild yak habitats; The environmental variables include land use, vegetation coverage, altitude, slope, slope aspect, minimum distance between habitat and water source, minimum distance between habitat and settlement, annual average temperature, annual maximum temperature, annual minimum temperature, annual precipitation, annual maximum monthly precipitation, and annual minimum monthly precipitation; S2, substituting the environmental variables into the MAXENT model, and obtaining the logical relationship between the environmental variables and the probability of wild yaks appearing when the entropy of the MAXENT model is maximum; S3, randomly obtain a number of points where wild yaks are monitored, and construct a set based on the environmental variable data of the points; 75% of the environmental variable data in the collection were randomly divided into training sets, and 25% of the environmental variable data were divided into test sets. The logical relationship between environmental variables and the probability of wild yaks appearing was iterated several times, and the importance weight of each environmental variable in the MAXENT model was determined by the jackknife method. S4. Substitute the importance weight of each environmental variable of the wild yak habitat into the MAXENT model to obtain the wild yak habitat suitability evaluation model, and make a receiver operating characteristic curve to evaluate the accuracy of the model prediction results; S5. When the accuracy of the model prediction results meets the requirements, the wild yak habitat evaluation model is used to obtain the logistic probability map of the wild yak habitat, and the suitability level of the wild yak habitat is evaluated based on the logistic probability map; when the accuracy of the model prediction results does not meet the requirements, the number of points and / or the number of iterative operations are increased, and steps S3, S4 and S5 are repeated until the accuracy of the model prediction results meets the requirements.

2. The method for evaluating the suitability of wild yak habitat according to claim 1, characterized in that: In step S1, the water source is at least one of a river, a lake, a reservoir, and a swamp.

3. The method for evaluating the suitability of wild yak habitat according to claim 1, characterized in that: In step S2, the method for obtaining the logical relationship is to obtain the MAXENT model: The maximum entropy characteristic root is the optimal solution under the constraints, that is: ； The constraints are: ; Among them, X is the set of environment variables, that is, ; y is the probability of wild yaks appearing, which is a random discrete variable; is the probability distribution of the environmental variable X, ; It represents the conditional entropy of random variable y under the condition of given environmental variable set X; is given The conditional probability of y when .

4. The method for evaluating the suitability of wild yak habitat according to claim 1, characterized in that: In step S3, the number of points is greater than 150.

5. The method for evaluating the suitability of wild yak habitat according to claim 1, characterized in that: In step S3, the number of iterative calculations is 15 or more.

6. The method for evaluating the suitability of wild yak habitat according to any one of claims 1 to 5, characterized in that: In step S4, the method for evaluating the accuracy of the model prediction results includes: evaluating the accuracy of the model prediction results by the AUC value of the ROC curve; the accuracy evaluation standard of the model prediction results is: When Q < 0.6, it means that the model prediction effect is not ideal; When 0.6≤Q<0.7, it means that the model prediction results are poor; When 0.7≤Q<0.8, it means that the model prediction results are average; When 0.8≤Q<0.9, it means the model prediction results are good; When Q ≥ 0.9, it means that the model prediction results are excellent.

7. The method for evaluating the suitability of wild yak habitat according to claim 6, characterized in that: In step S5, the accuracy requirement of the model prediction result is Q ≥ 0.

8.

8. The method for evaluating the suitability of wild yak habitat according to claim 7, characterized in that: When more than 40% of the random points in the wild yak habitat meet P≥0.9, the suitability level of the wild yak habitat is suitable; when more than 40% of the random points in the wild yak habitat meet 0.8≤P<0.9, the suitability level of the wild yak habitat is relatively suitable; when more than 40% of the random points in the wild yak habitat meet 0.7≤P<0.8, the suitability level of the wild yak habitat is generally suitable; when more than 40% of the random points in the wild yak habitat meet 0.6≤P<0.7, the suitability level of the wild yak habitat is unsuitable; when more than 40% of the random points in the wild yak habitat meet P<0.6, the suitability level of the wild yak habitat is extremely unsuitable.

9. A wild yak habitat suitability assessment device, characterized in that: It includes at least one processor and a memory that is communicatively connected to the at least one processor; the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a wild yak habitat suitability assessment method according to any one of claims 1 to 8.

10. A computer-readable medium, characterized in that Instructions executable by a processor are stored thereon, and when the instructions are executed by the processor, the processor executes a method for evaluating the suitability of a wild yak habitat as described in any one of claims 1 to 8.

Images