Grassland Pest Occurrence Area Survey Method Based on Multi-stage Sampling
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 花立民
- Filing Date
- 2022-04-22
- Publication Date
- 2026-06-30
Smart Images
Figure CN115311343B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of grassland pest control technology, and in particular to a method for investigating the occurrence area of grassland pests based on multi-level sampling. Background Technology
[0002] Grasslands play a vital role in soil and water conservation, windbreak and sand fixation, air purification, and biodiversity maintenance. In recent decades, under the combined influence of human activities and climate change, natural grasslands have experienced varying degrees of degradation. Degraded grasslands provide excellent habitats for some harmful organisms, such as rodents, leading to their population expansion. When the population density reaches a certain level, it results in pest outbreaks. Grassland pests, such as grassland rodents, damage vegetation and soil through burrowing, foraging, and bulldozing, causing weed growth, low vegetation cover, and declining productivity in the damaged grasslands. Simultaneously, their frequent digging and bulldozing activities cause grassland patchy formation, reduce soil nutrients, and exacerbate soil erosion, negatively impacting the ecological, landscape, and economic value of grasslands. Pest problems have become one of the major obstacles to the protection, construction, and utilization of natural grasslands.
[0003] For a long time, pest control has been characterized by an "outbreak-and-exterminate" emergency management approach. Numerous studies have shown that only when a pest's population density is excessively high and its distribution range is too wide will it disrupt the balance of the grassland ecosystem and produce negative effects. Therefore, from an ecological perspective, the ultimate goal of pest control is not "population-clearing" extermination, but rather controlling its numbers within a reasonable range through scientific methods. A crucial aspect of scientific control is conducting grassland pest hazard assessments. Grassland pest assessment factors include "hazard level (qualitative)" and "affected area (quantitative)." While there is considerable research on hazard level assessment, there has been little breakthrough in exploring methods for investigating affected area. The affected area is a vital basis for formulating pest management strategies, including determining the pest's distribution range, monitoring and early warning systems, and planning pest control funding.
[0004] The area affected by pests refers to the area where the degree of pest occurrence reaches the control threshold through sampling surveys of various representative plots. Taking grassland rodent infestation as an example, the area affected by grassland rodent infestation refers to the total area of bare land caused by grassland rodents through activities such as digging, bulldozing, and foraging, including new and old mounds, mound runoff areas, burrow entrances, abandoned and collapsed burrows, obvious runways, and bald patches caused by their activities.
[0005] Currently, pest outbreak area surveys mainly employ mapping and delineation methods. Mapping involves placing a surveying net at the outbreak site, filling in the damage area proportionally on a calculation sheet, moving the net grid to complete the map, and finally calculating the total damaged area. This method has a small sampling area, is labor-intensive and time-consuming, and is difficult to apply to large-scale production practice surveys. Delineation involves surveyors marking the pest outbreak locations on-site or visually estimating the affected area, then drawing patches on a working map to estimate the affected area. Delineation is limited by roads, terrain, etc., making it impossible to survey remote areas. Furthermore, the delineation accuracy is poor and highly subjective, leading to less scientific rigor. In addition, both mapping and delineation methods are conducted in areas where pests have already occurred, making random sampling impossible. When the variability of the surveyed population is large, these methods fail to meet statistical requirements.
[0006] Grassland pests exhibit strong habitat selectivity, meaning that at the grassland landscape level, pest occurrence sites possess significant spatial heterogeneity and landscape complexity, rendering traditional simple sampling statistical methods unsuitable for this type of survey. Conducting grassland pest occurrence area surveys must adhere to ecological and statistical principles. Summary of the Invention
[0007] To address the aforementioned technical problems, this invention provides a stratified sampling method. First, the survey area is divided into several layers based on different landscapes, such as grasslands, shrublands, and wetlands, according to the habitat characteristics of grassland pests. For example, grasslands can be categorized into mountainous and plain areas. Then, sample units are randomly selected from each layer. The average grassland pest occurrence area of multiple sample units is used to calculate the grassland pest occurrence area of the current layer. Finally, the total pest occurrence area of the survey area is obtained by statistically analyzing the data from each layer.
[0008] To achieve the above technical objectives, this invention provides a method for investigating the area of grassland pest occurrence based on multi-level sampling, comprising the following steps:
[0009] S1. Select suitable habitats for harmful organisms within the target area as primary survey sample areas;
[0010] S2. Select areas with consistent community appearance and similar composition from the primary survey sample area as secondary survey sample plots, and divide the secondary survey sample plots into several grids as tertiary survey quadrats;
[0011] S3. Select several of the three-level survey plots and conduct drone aerial photography to obtain survey images;
[0012] S4. Calculate the area of pest occurrence corresponding to the third-level survey plot based on the survey images, and calculate the area of pest occurrence of the second-level survey plot and the first-level survey area accordingly.
[0013] In some preferred embodiments, the method for selecting suitable areas for pest habitat as primary survey sample areas in step S1 includes:
[0014] The survey obtains data on the distribution points of pests in the target area, selects environmental factors that have a significant impact on the suitability of pest habitats, inputs the distribution point data and environmental factors into the species distribution model, and obtains multiple pest suitable areas with different suitability factors. The areas with a suitability factor greater than 0.5 are selected as the primary survey sample areas.
[0015] In some preferred embodiments, the method for selecting environmental factors that have a significant impact on the suitability of pest habitats is as follows:
[0016] Obtain climate, food, and geographical information closely related to the habitat suitability of the pests, establish a correlation model, calculate and select the information indicators with the highest correlation as the environmental factors.
[0017] In some preferred embodiments, step S3 involves selecting a number of the third-level survey plots for drone aerial photography. The specific method for determining the number of samples to be selected includes:
[0018] S31. Randomly select several sample plots from the primary survey area, calculate their mean and variance of vegetation cover, and calculate the theoretical minimum sample size n according to the following formula:
[0019]
[0020] Where d′ is the allowable error of the sampling result, t is the standard error multiple, and w h For vegetation coverage, Let N be the variance of vegetation cover, and N be the population size.
[0021] S32. Select an integer n that is not less than the theoretical minimum number of samples as the number of samples to be drawn.
[0022] In some preferred embodiments, the method for calculating the pest occurrence area of the secondary survey plots in step S4 includes:
[0023] S41. Calculate the average incidence of harmful organisms in the secondary survey plots: Where Di is the area of pest occurrence in the i-th tertiary quadrat; Ai is the area of the i-th tertiary quadrat; and n' is the number of tertiary quadrats.
[0024] S42. Calculate the total area of the secondary survey plots, multiply it by the average incidence rate of pests in the secondary survey plots, and obtain the pest occurrence area of the secondary survey plots.
[0025] Beneficial effects
[0026] 1. This invention employs a stratified sampling method, avoiding the drawbacks of simple random sampling that tends to concentrate on a particular region or characteristic, or miss certain characteristics; 2. By rationally setting the hierarchical structure and sampling ratio, the sampling results of the lower-level survey area can better reflect the characteristics of the upper-level survey area, thereby showcasing the pest occurrence area characteristics of the entire target area; 3. By combining UAV aerial surveying technology and image recognition technology with stratified sampling, accurate field surveys of pest occurrence areas can be achieved, obtaining a reliable data calculation basis. Attached Figure Description
[0027] Figure 1 This is a flowchart illustrating a preferred embodiment of the present invention;
[0028] Figure 2 This is a diagram showing the results of the primary survey sample area division in a preferred embodiment of the present invention;
[0029] Figure 3 This is a diagram showing the division results of secondary and tertiary survey plots in a preferred embodiment of the present invention;
[0030] Figure 4 This is a diagram showing the results of image interpretation in a preferred embodiment of the present invention;
[0031] Figure 5 This is a graph showing the error relationship between measured and predicted values of the investigation image interpretation results in a preferred embodiment of the present invention. Detailed Implementation
[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described below with reference to the accompanying drawings. In the description of this invention, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention.
[0033] like Figure 1 As shown, this invention provides a method for investigating the area of grassland pest occurrence based on multi-level sampling, including the following steps:
[0034] S1. Select suitable habitats for pests within the target area as primary survey sample areas; these primary survey sample areas are the overall sampling areas for investigating the area of pest occurrence. Pests refer to major pests (including invasive species) that harm grassland vegetation and its products, causing economic or ecological losses, including rodents, insects, plant pathogens, and poisonous weeds, with grassland rodents and rabbits being the most common. The area of pest occurrence refers to the area where the degree of pest occurrence reaches the control threshold through sampling surveys of various representative plots. Taking grassland rodent infestation as an example, the area of grassland rodent infestation refers to the total area of bare land caused by grassland rodents through digging, bulldozing, foraging, etc., including new and old mounds, mound runoff areas, burrow entrances, abandoned and collapsed burrows, obvious runways, and bald patches caused by their activities.
[0035] There are many methods for selecting suitable habitats for pests, commonly including historical data review and on-site investigation confirmation. However, these methods mostly rely on empirical data, meaning they can only be used in areas where pests have already occurred or are currently occurring. They are not well-suited for areas where pests have not occurred but there is a potential risk. Therefore, in some preferred embodiments, a more scientific method for selecting suitable habitats for pests is provided, including:
[0036] The survey obtains data on the distribution points of pests in the target area, selects environmental factors that have a significant impact on the suitability of pest habitats, inputs the distribution point data and environmental factors into the species distribution model, and obtains multiple pest suitable areas with different suitability factors. The areas with a suitability factor greater than 0.5 are selected as the primary survey sample areas.
[0037] The selection of survey sites involves setting up reconnaissance routes while considering traffic conditions. These routes should, as far as possible, traverse the main grasslands and terrain types within the survey area. During the survey, coordinates of the distribution areas of harmful organisms are marked and their latitude and longitude recorded. The environmental factors that significantly impact the suitability of harmful organism habitats specifically include indicators closely related to the surveyed harmful organisms, such as climate information (e.g., temperature, precipitation), food information (e.g., biomass, canopy cover), and habitat geographic information (e.g., topography, altitude). This information can be obtained through field surveys and sampling, or by searching online databases (e.g., meteorological websites, Chinese Academy of Sciences databases).
[0038] Species Distribution Models (SDMs) are mathematical models based on species presence or richness data and environmental factor data. These models estimate the niche requirements of species in a multidimensional ecological space composed of environmental factors, based on statistical information provided by sampling points, and then project them onto a selected spatiotemporal range to reflect the species’ preference for habitat in the form of probabilities. The simulation results are generally given as Boolean values (1 represents potential distribution, 0 represents potential non-distribution). The model results usually reflect the distribution of suitable habitats for species on a large scale.
[0039] S2. In the primary survey area, select areas with consistent community appearance and similar composition as secondary survey plots. Divide the secondary survey plots into several grids as tertiary survey quadrats. Consistent community appearance means that areas suitable for the same type of pest generally have similar characteristics, such as open plains and gentle slopes (slope <7 degrees) suitable for plateau pikas. Similar composition refers to grassland species; generally, grassland plant species are relatively homogeneous, without drastic changes or significant differences. It should be understood that when dividing the secondary survey plots into grids, the actual size of the target area, the type of drone to be used in subsequent steps, and the image interpretation software and hardware computing power resources need to be considered to determine the grid size. Specifically, when the actual area of the target area is large, the drone model is high-end, and the image interpretation software and hardware computing power is abundant, larger grids can be divided; conversely, smaller grids should be divided. The specific division method should be determined by those skilled in the art based on the actual situation, and this invention does not further limit it.
[0040] S3. Select several of the aforementioned tertiary survey plots for aerial photography using drones to obtain survey images. Those skilled in the art should understand that the accuracy of the final survey results relative to the overall true value is directly related to the sample size. Theoretically, all tertiary survey plots should be surveyed in the field. However, in reality, there are often insufficient manpower and financial resources to support such extensive field sampling. Therefore, those skilled in the art should use sampling methods to conduct field surveys of the tertiary survey plots. The specific number of samples can be determined by those skilled in the art based on the project needs and budget. Current drone low-altitude remote sensing technology can ensure sufficiently clear aerial images. Combined with existing image interpretation technology (image processing technology that separates the characterized pest occurrence area from normal grassland), it is possible to quickly identify ground features such as burrows, mounds, and bare patches dug by pests and calculate their quantity and area. This eliminates the need for extensive on-site statistical mapping, making it suitable for large-scale determination of grassland pest occurrence areas. It conforms to ecological and statistical principles and has advantages such as high efficiency, operability, and strong scientific validity.
[0041] In some preferred embodiments, to save more manpower and financial resources, a lower sampling number needs to be set. However, a sufficiently large sampling number is needed to meet high sampling accuracy and thus more accurately measure the true value of the whole. This embodiment provides a sampling number determination method that can take into account all aspects, including:
[0042] S31. Randomly select several sample plots from the primary survey area, calculate their mean and variance of vegetation cover, and calculate the theoretical minimum sample size n according to the following formula:
[0043]
[0044] Where d′ is the allowable error of the sampling result, t is the standard error multiple, and w h For vegetation coverage, Let N be the variance of vegetation cover, and N be the total population.
[0045] S32. Select an integer n that is not less than the theoretical minimum number of samples as the number of samples to be drawn.
[0046] S4. Calculate the area of pest occurrence corresponding to the third-level survey plot based on the survey images, and calculate the area of pest occurrence of the second-level survey plot and the first-level survey area accordingly.
[0047] Calculating the area affected by pests based on survey images can be done using existing pest monitoring software, such as the "Grassland Rodent Monitoring Software v1.0" (registration number 2021SR0546708). It should be understood that the focus of image interpretation will differ depending on the type of pest. For example, when investigating the area affected by plateau mole rats, the focus should be on identifying and calculating the area of new and old mounds and mound runoff caused by mole rat activity. When investigating the area affected by greater gerbils, due to the unique characteristics of desert steppes, the focus should be on the burrow entrances and burrow atolls caused by gerbil activity.
[0048] In some preferred embodiments, a preferred method for calculating the area of pest occurrence in secondary survey plots is provided, including:
[0049] S41. Calculate the average incidence of harmful organisms in the secondary survey plots: Where Di is the area of pest occurrence in the i-th tertiary quadrat; Ai is the area of the i-th tertiary quadrat; and n' is the number of tertiary quadrats.
[0050] S42. Calculate the total area of the secondary survey plots, multiply it by the average incidence rate of pests in the secondary survey plots, and obtain the pest occurrence area of the secondary survey plots.
[0051] Example
[0052] This embodiment takes a survey of the area affected by plateau pika infestation on the grasslands of Maqu County, Gansu Province as an example.
[0053] 1. Determination of primary survey sample areas
[0054] The main environmental impact factors of the surveyed area were obtained from the database. In this embodiment, temperature, precipitation, and food resources (replaced by the normalized difference vegetation index) were selected for modeling. The Maxent model was chosen as the species distribution model. The calculation results show that, Figure 2 As shown, under the current environmental climate, the most suitable habitats for plateau pikas (suitability index > 0.5) are mainly concentrated in areas X1, X2, X3, and X4. According to the design, areas with a suitability index > 0.5 are selected as the primary sample area for rodent infestation survey, with an area of 2.5 × 10⁻⁶. 5 hm 2 .
[0055] 2. Delineation of secondary survey plots
[0056] After determining the primary survey areas, secondary survey plots were delineated. Field investigations revealed that the plateau pika habitat in this area is mostly located in open plains and gentle slopes (slope <7 degrees), with a single grassland type. Therefore, this embodiment primarily focuses on plains or gentle slopes as the survey areas. A 5km × 5km grid was created using ArcGIS 10.2 Fishnet analysis tools as the tertiary survey plots, as follows: Figure 3 As shown, a total of 175 tertiary quadrats were divided into primary survey areas within the four townships. Before the formal survey, a preliminary experiment was conducted on 10 randomly selected tertiary quadrats to obtain the mean and variance of vegetation cover. The allowable error for sampling results in this embodiment is set at 8%, and the standard error multiple is 95%. The final results were then... Calculations show that with a sampling accuracy of 92%, the theoretical minimum number of samples, n = 31.43, requires at least 32 sample plots to be selected for investigation. Therefore, after constructing the grid, this embodiment selected a total of 38 sample plots, representing a sampling rate of 21.7%, for UAV low-altitude aerial photography.
[0057] 3. Calculation of the area affected by rodent infestation
[0058] The survey images were interpreted using a self-designed grassland rodent monitoring software v1.0 (registration number: 2021SR0546708), which quickly identified ground features such as rodent burrows, mounds, and bare patches, and calculated their numbers and areas. Figure 4 As shown, after obtaining the vegetation cover of each survey plot, the accuracy of the UAV image interpretation was verified. The verification revealed that, as... Figure 5As shown, the average relative error between the measured and predicted values is MRE = 0.07, and the correlation coefficient R = 0.81, indicating that the UAV image interpretation values and the manually measured values in the study area are in high agreement.
[0059] To assess whether the target area has reached the level requiring control measures, based on the pika control indicators in the grassland pest control standards, this embodiment will separately count the bald patch area and the number of effective burrows in aerial images. Meeting either indicator is considered as meeting the "control indicator." After interpreting the plateau pika infestation area in different tertiary survey plots, the formula is used... The incidence rate of plateau pika pests in the secondary survey plots was calculated, and then the area of plateau pika pest infestation in the primary survey plots was calculated, as shown in Table 1 below. The total area of plateau pika pest infestation in the survey area was obtained by adding the areas of plateau pika pest infestation in different primary survey plots: 1.8 × 10⁻⁶. 4 hm 2 It accounts for 7.2% of the total suitable habitat area.
[0060]
[0061] Table 1. Area affected by rodent infestation in the surveyed area
[0062] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.
Claims
1. A method for investigating the occurrence area of grassland harmful organisms based on multi-stage sampling, characterized by, Including the following steps: S1. Select suitable habitats for harmful organisms within the target area as primary survey sample areas; S2. Select areas with consistent community appearance and similar composition from the primary survey sample area as secondary survey sample plots, and divide the secondary survey sample plots into several grids as tertiary survey quadrats; S3. Select several of the three-level survey plots and conduct drone aerial photography to obtain survey images; S4. Calculate the area of pest occurrence corresponding to the third-level survey plot based on the survey images, and calculate the area of pest occurrence of the second-level survey plot and the first-level survey area accordingly; The method for calculating the area of pest occurrence in the secondary survey plots in step S4 includes: S41, calculating the average occurrence rate of pests in the secondary investigation sample plot: wherein Di is the pest occurrence area of the i-th tertiary investigation sample plot; and Ai is the area of the i-th tertiary investigation sample plot; is the number of tertiary investigation sample plots; and n is the theoretical minimum sampling number; S42. Calculate the total area of the secondary survey plots, multiply it by the average incidence rate of pests in the secondary survey plots, and obtain the pest occurrence area of the secondary survey plots.
2. The rangeland pest occurrence area survey method based on multi-stage sampling according to claim 1, characterized by, The method for selecting suitable areas for pest habitats as primary survey sample areas in step S1 includes: The survey obtains data on the distribution points of pests in the target area, selects environmental factors that have a significant impact on the suitability of pest habitats, inputs the distribution point data and environmental factors into the species distribution model, and obtains multiple pest suitable areas with different suitability factors. The areas with a suitability factor greater than 0.5 are selected as the primary survey sample areas.
3. The rangeland pest occurrence area survey method based on multi-stage sampling according to claim 2, characterized by, The method for selecting environmental factors that have a significant impact on the suitability of habitats for harmful organisms is as follows: Obtain climate, food, and geographical information closely related to the habitat suitability of the pests, establish a correlation model, calculate and select the information indicators with the highest correlation as the environmental factors.
4. The rangeland pest occurrence area survey method based on multi-stage sampling according to claim 1, wherein In step S3, several of the three-level survey plots are selected for drone aerial photography. The specific method for determining the number of samples includes: S31. Randomly select several sample plots from the primary survey area, calculate their mean and variance of vegetation cover, and calculate the theoretical minimum sample size n according to the following formula: in, The allowable error for the sampling results is given by t, where t is a multiple of the standard error. For vegetation coverage, Let N be the variance of vegetation cover, and N be the population size. S32. Select an integer n that is not less than the theoretical minimum number of samples as the number of samples to be drawn.