Structural Characteristics Investigation Of Forest Landscape Patches Coupling Network Based On Adjacency Relations

Forest patches exchange complex ecological flow between each other, and various patches coupling bodies maintain a sustainable form, which leads to the formation of patches coupling network. In the process, the forest patches serve as spatial functional unit in order to extract nodes, the common boundary among different patches as a border, and the forest patches and the patches coupling network as the framework of the system, which contributes to structural properties of the forest landscape patches coupling network. By analyzing the structural and landscape characteristics of the patches coupling network, the results are as follows.(1) For the purpose of forest health management, practicality as the premise of dividing different kinds of landscape, based on the typical landscape features in Taoyuan County, a forest landscape classification index system is constructed based on the four factors including land type, dominant tree species, age group, and tree species composition.(2) In order to elaborate the complexities of the network,75 different sizes of forest landscape patches coupling bodies are taken as samples from Taoyuan County in Changde so as to draw out the forest landscape patches coupling network. This paper intends to study the structural properties of forest landscape patches coupling network in terms of Small-world properties, scale-free property, and community characteristic. Small-world properties of the forest landscape patches coupling is universal. The scale-free characteristics of forest landscape patches in the coupling network is related to the size of forest landscape patches. The larger size of forest landscape patches, the more significant scale-free properties present. On the whole, the forest landscape patches coupling network is more stable in the linear fitting correlation coefficient with logarithmic coordinates, and the distribution of the network is in line with the power-law distribution. The community characteristics of the coupling network is universal, and with the increase of forest landscape patches coupling body size, the community characteristics show more significant.(3) An independent correlation analysis is applied to analyze the forestry landscape pattern index of patches and landscape level respectively. The results showed that the correlation between forestry landscape pattern index and patches level can be divided into three categories:First, the number is greater than 0.75, which means the relation is very stable, free from the influence of landscape scale; Second, the significance of correlation has something to do with the landscape index; Third, the correlation index is less than 0.75, regardless of the size of the forest landscape. For the forestry landscape pattern index, the average value of the index, a weighted average and variance value correlation is not significant. However, the correlation among the weighted area index, the weighted related circumcircle index and other landscape index shows the strongest.(4) The writer screened more appropriate landscape size from five landscape particle size of 4m, 11m,13m,15m,18m. The results show that there is no significant difference in forest landscape pattern index under the condition of 11m,13m,15m.In this study, landscape particle size of 13m is selected.(5) From the perspective of a single patch of forest landscape coupled networks, clustering coefficient, patches of forest landscape and the landscape characteristics of the coupling network as a whole of the 75 forest landscape patches coupled network nodes are analyzed individually. The results show:node clustering coefficient, patches of forest landscape and the landscape characteristics of the coupling network is not obvious; From the perspective of the landscape characteristics of the coupling network as a whole, by analyzing the node clustering coefficient, and the same degree of clustering coefficient, the correlation analysis shows that the indicators of absolute value is greater than 0.75, which shows a negative correlation, indicating that the greater the node clustering coefficient is, the smaller plaque size will be, and with the same area, the more complex of the landscape patch. From the landscape level analysis, the higher of the clustering coefficient of the network described, the smaller forest landscape patches in the shape of the rules or landscape patches average area; If the averages relations is greater than 0.75, and the correlation between landscape indices is also higher than 0.75, which indicates that the higher the degree of the nodes, the larger the area corresponding to the forest landscape patches. With the same area, the more complex the shape of patches, the higher network average degree is, showing that the larger of the forest landscape patches on average or more irregular shape of the patches.