2.1. Study area
Xinggang town is located in the south of Guangxi Province of south China, covering an area of 123.42 km2. This town acts as a critical zone for the development in the Beibu Gulf of Guangxi Province for its favorable conditions (e.g., the natural harbor (31 km coastline), climate and topography) (Fig. 1). The altitude of this town ranges between 0 and 18 m with a flat terrain. It is recognized as a subtropical monsoon climate area with an average annual temperature of 22.6℃ and an average annual rainfall of 1,663.7 mm. As a key agricultural production region, rice and vegetables are broadly cultivated as agricultural crops. Furthermore, there are abundant shallow beach resources and diverse aquatic products. Over the past few years, Xinggang Town has witnessed the rapid economic development and the rapid population growth, and the total population reached up to 65,000 by 2019.
2.2 Data source and processing
In the present study, the remote sensing images (1 m spatial resolution) of Pleiades satellite in 2005 and 2020 presented by Europe Astrium company acted as the data source. With the support of ERDAS software, the following steps of geometric correction, image registration, image mosaic were processed based on 1: 10000 topographic map and field survey. Given the characteristics of the study area, the landscape types fell to eight types (i.e., farmland, aquaculture land, forest, built-up land, road, water body, mudflat and unused land) by using the method of manual visual interpretation. After the field correction, the Kappa coefficient of landscape type data exceeded 0.9, which revealed that it could comply with the data accuracy requirements of the study. By complying with the scope of the study area, Xinggang town fell to gradient Ⅰ (0-2 km), gradient Ⅱ (2-4 km), gradient Ⅲ (4-6 km) and gradient Ⅳ (>6 km) with an interval of 2 km from the coastline to the inland (Fig. 2).
2.3 Methods
2.3.1 Habitat quality assessment method
Landscape type data were adopted to assess regional habitat quality based on the impact intensity, distance and sensitivity of various threat factors in the habitat quality module of InVEST model (Aneseyee et al., 2020; Yohannes et al., 2021). The operation of the module required the data of the habitat type, the intensity and distance of the effect of the threat factors on each habitat, and the sensitivity of the respective habitat to threat factors. The farmland, aquaculture land, built-up land and road obtained from remote sensing interpretation were considered the landscape types significantly disturbed by human, and they were identified as the threat factors. The sensitivity of the threat factors to habitat, the distance and weight of the effect of the threat factors and other parameters were determined based on the expert score, the recommended value of the model user’s guide (Sharp et al., 2014) and related literatures (Han and Dong, 2017; Xu et al., 2019). The parameter settings of habitat quality module are listed in Tables 1 and 2.
Where Qxj denotes the habitat quality; Hj expresses the habitat suitability index; Dxj is the habitat degradation degree; Z equals 2.5, k equals half of the grid cell resolution; Wr represents the weight of the threat factors; ry is the number of ecological threat factors; irxy is the effect of threat factor r in grid y on grid x; βx is the degree of protection (unconsidered in this study); Sjr expresses the sensitivity of the threat factors to different habitats. The assessment results ranged from 0 to 1. The closer to one, the higher the habitat quality would be, while the closer to zero, the lower the habitat quality would be.
2.3.2 Ecological risk assessment methods
From the perspective of landscape structure of regional ecosystem, ecological risk at the landscape level was assessed with the area proportion of the respective landscape component, landscape structure index and landscape fragility index (Wang et al., 2020; Xie et al., 2021).
Where ERIk denotes the ecological risk index of the k fishnet; Aki is all landscape areas of fishnet; Ak is the area of the k fishnet; m expresses the number of landscape types in fishnet; LLi is landscape ecological loss index; Ui represents landscape disturbance index; Ci denotes landscape fragmentation degree; Fi is landscape separation degree; Di is landscape dominance degree; a, b and c respectively denote the weight of fragmentation degree, separation degree and dominance degree, and 0.3, 0.2 and 0.5 are assigned to the values of a, b and c, respectively; Si is the vulnerability index of landscape. Given the ability of each landscape type to resist external influences, the vulnerability index of the respective landscape was assigned as water body =7, mudflat =6, aquaculture land =5, farmland =4, unused land =3, forest =2, built-up land and road =1. The calculation of ecological risk complied with 400 m×400 m fishnets in the study area, and the fishnet tool of ArcGIS software was adopted to generate 847 fishnets in the study area.
2.3.3 Landscape metrics selection
Landscape metrics act as the indicators to quantify landscape pattern, which are capable of indicating the effect of human activities on landscape pattern (Mayer et al., 2016). Six landscape metrics at a landscape level, i.e., number of patches (NP), patch density (PD), largest patch index (LPI), landscape division index (DIVISION), shannon's diversity index (SHDI) and shannon's evenness index (SHEI), were selected to express the landscape pattern in this study (Table 3).
2.3.4 Correlation analysis of landscape metrics with habitat quality and ecological risk
Six landscape metrics (i.e., NP, PD, LPI, DIVISION, SHDI and SHEI) were selected, and the pre-processed landscape type data in 2005 and 2020 were inputted into the Fragstats 4.2 software as the data sources to generate landscape metrics in the respective fishnet. Subsequently, the landscape metrics distribution maps in 2005 and 2020 were drawn with the ArcGIS software. Lastly, the spatial analyst tools of ArcGIS software were employed to generate the spatial map of landscape metrics change in the respective fishnet from 2005 to 2020.
Based on the value of landscape metrics, habitat quality and ecological risk in the respective fishnet, the conversion tool of ArcGIS software was adopted to convert raster data into ASCII files. Next, the correlations of landscape metrics with habitat quality and ecological risk were analyzed by conducting the Pearson correlation analysis in SPSS software.
The Bivariate Local Moran's I tool of GeoDA software was applied for analyzing the spatial correlation of landscape metrics with habitat quality and ecological risk in the respective fishnet, and LISA cluster maps between landscape metrics and habitat quality and between landscape metrics and ecological risk were obtained (Anselin and Sergio, 2014).