Ecological Indicators (Feb 2023)
Constructing an indices system for evaluating the ecological integrity of forests in western Sichuan, China based on structural equation modeling
Abstract
With the concern for biodiversity conservation and ecological sustainability, the traditional method of assessing conservation management based on single elements such as timber production, stand structure, and forest area is no longer applicable. Especially for China, where the primary goal of national park conservation management is to “protect and maintain the authenticity and integrity of natural ecosystems”, it is necessary to construct a set of indicators that can comprehensively evaluate the effectiveness of natural ecosystem conservation management. Therefore, in this study, we take Jiuzhaigou and Giant Panda National Park, in Sichuan, China, as examples, and construct a set of ecological integrity evaluation indices system from the perspective of attributes of forest ecosystems. Initially, we screened the candidate indicators based on the coefficient of variation, distribution range, and redundancy analysis and identified 16 indicators as the evaluation indices of ecological integrity. Next, the model fit and the interrelationship of the components of the ecological integrity model were verified using a combination of structural equation modeling and actual data. The model validation results showed that: (1) CMIN/DF = 1.66, RMSEA = 0.058, GFI = 0.954, and other model fit metrics have reached the best reference standard, indicating that the ecological integrity model overall fit is high. (2) In the parameter estimation results of the model observed and latent variables, the ecological integrity model explained 96%, 85% and 89% of the variance in species composition, stand structure and ecological processes, respectively, indicating that the evaluation indices selected can effectively predict the ecological integrity of forests. In addition, we used the factor loadings in the structural equation model to calculate the weights of the indices and used the weighted sum to calculate the ecological integrity. A one-way ANOVA was conducted to compare the differences among the ecological integrity of forests of different origins, forest types and life forms. The relative importance and linear regression were used to elucidate the main factors affecting forest ecological integrity. The results showed that (1) except for the scrub community, the mean value of forest ecological integrity was >0.45, which was in the middle-to-upper level. (2) the ecological integrity of forests of different origins and different forest types showed a changing pattern: natural forest > plantation forest, mixed forest > pure forest. (3) Stand age and elevation strongly influenced the ecological integrity of forests, with relative contribution rates of 64.54% and 11.88%, respectively.
