Search

Hainan Island has undergone rapid landscape transformation leading to habitat fragmentation.Thus, exploring the spatiotemporal change from its past habitat quality provides insights into its future ecological protection. This study investigated the spatiotemporal changes of the habitat quality of Hainan Island from 1980 to 2020 using the InVEST and CA-Markov models and predicted the habitat quality for the time frame of 2020–2040. We found that the habitat quality index first increased from 0.73 in 1980 to 0.74 in 1990 and then will decrease to 0.70 in 2040 with an accelerating rate. The habitat quality inside the national park was significantly better and more stable than that outside the national park. The simulation and prediction results showed that the habitat quality inside the national park would continue to improve from 2020 to 2040, while that outside the national park would decline. The spatial distribution of the changes in the habitat quality of Hainan Island was significantly autocorrelated and clustered, enhanced over time. Our results suggest that the overall decline in the habitat quality of Hainan Island mainly result from the habitat quality change outside the national park, and the habitat quality change inside the national park will play a key role in stabilizing and supporting the habitat quality of Hainan Island, which highlights the importance of national park conservation.

National parks promote the protection of the ecological environment and global biodiversity conservation. However, the changes in land use obviously damages the natural habitat, leading to habitat fragmentation and degradation. Land use changes affect ecosystem service value (ESV) and its spatial distribution. In this study, a framework for the assessment of regional spatial patterns in land use/cover changes (LUCC) and ESV was established; the framework relies on the application of a classical spatial autocorrelation model. Remote sensing data for Hainan Tropical Rain Forest National Park (HNTRFNP) were used as a test case for the autocorrelation model approach to analyzing spatial patterns in ESV. Across HNTRFNP, ESV averaged 6201.44 million USD, with high ESV seen in aquatic and forested areas. Furthermore, land use types in HNTRFNP showed spatial autocorrelation (P