Gradient diffusion entropy corrected ALNS optimization for vegetation topology interaction networks
Vegetation interaction network structures can reveal ecological relationships and interactions between ecological communities such as Grasslands, Croplands and Forests, and are essential for fitting and optimizing ecological numerical systems as well as ecological restoration evolution. In this stud...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Ecological Indicators |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1470160X24015012 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Vegetation interaction network structures can reveal ecological relationships and interactions between ecological communities such as Grasslands, Croplands and Forests, and are essential for fitting and optimizing ecological numerical systems as well as ecological restoration evolution. In this study, two characteristics of the vegetation interaction network constructed through density analysis linking land cover and land class transfer space can realize the objectivity of ecosystem simulation: (1) Constructing relationships between vegetation types using spaces where transfer patterns occur as aggregated clusters. (2) Calculation of relationship strength based on density analysis. In addition the interaction topology network was optimized using the gradient diffusion entropy correction ALNS to further enhance the stability of the proposed ecological interaction network structure. The following findings were obtained by parsing the corresponding study area through the constructed interaction network structure. From 2000 to 2020. All types of ecological nodes showed an increasing trend. the number of ecological nodes in cropland, grassland and forest in 2020 was 27, 8 and 19 respectively. The interaction structure is dominated by competition for cropland and forest in the middle and lower reaches. The GDEO-ALNS optimized interaction structure slows down the ECV drop rate in topological attacks by 0.03 and 0.02, respectively, compared to the original interaction structure, which provides better immunity to interference compared to the optimized structure of ALNS. The optimized vegetation structure consists of three optimal subplot structures that form a sub-ecosystem within the spatial region with a strong anti-disturbance capability. |
|---|---|
| ISSN: | 1470-160X |