Figure 1

Location of Jinghong County and its urban land expansion during three periods"

Table 1

The socio-economic development of Jinghong County from 1970 to 2010"

Figure 2

Schematic framework of the study (a) and the framework of the PANDORA model (b) (Pelorosso et al., 2017). BESV represents the ES for biodiversity conservation. The full line represents the steps of the study, the dashed line represents the indexes and methods used in this study, and the dotted line represents the original data and the implementation of PANDORA model."

Figure 4

The uncertainty analysis of the PANDORA model. a, b, c, and d represent the Scenario 1 (the original sample), Scenario 2 (reduces roads based on the original sample), Scenario 3 (adds roads based on the original sample), and Scenario 4 (increases the local fragmentation based on the original sample), respectively. e, f, g, and h represent the BESV of Scenario 1, the BESV of Scenario 2, the BESV of Scenario 3, and the BESV of Scenario 4, respectively."

Table 2

Indexes of analyses of landscape changes and the BESV variation"

Figure 3

Land use map of the study area in 2010 (a), and the comparison of the performance of the PANDORA model (b), a unit value-based approach (c), and the InVEST model to evaluate the BESV (d). UL, IU, OP, SC, AV, AL, PL, FO, WA, and WE separately represent urban fabric, industrial units, open spaces, scrub, green space, arable land, plantation, forests, inland water, and inland wetland. b, c, and d represent the BESVs evaluated by PANDORA model ($·m-2 year-1), by the unit value-based approach (Costanza et al., 1997a; Ng et al., 2013) ($·m-2 year-1), and the carbon storage evaluated by the InVEST model (mg ha-1) in 2010, respectively. The classification of low, medium and high level applies the equal interval method."

Figure 5

Annual variation of both urban land and the region’s BESV. ACV and ACI represent annual change velocity and annual change intensity, respectively."

Figure 6

Distributions of urban land and its expansion from the 1970s to 2010 (a. The distribution of urban land; b. The distribution of the newly grown urban land)"

Figure 7

The distance gradients for the difference between urban land areas (?Area) and the difference between BESVs (?BESV) in different periods. The left longitudinal axis shows the standardized ?Area, and its value increases from bottom to top along the axis. The right longitudinal axis shows the standardized ?BESV, and its value decreases from bottom to top. The 0 in the X-axis represents the city center. Because of the normalization, the plus-minus state in the Figure is not the same as those in the raw data while the fluctuated trends of these data were consistent with those of the raw data."

Table 3

Pearson correlation coefficients between the difference (?) of BESVs and the difference (?) of landscape metrics"

Table 4

The stepwise regression analysis for the difference (?) of BESVs and the difference (?) of landscape metrics"

Figure 8

Statistics of BESV changes (?BESV) of projected urban land in 2030. Different colors represent the different range of variations in BESV. The small figure with different colors above represents the spatial distribution that the color represents below."