3.1. Land Use Change from 1999 to 2014
Farmland, forestland and orchard were the main land use types in the Beibu Gulf area (
Table 5), and their areas accounted for 82.24% of the total area. Farmland was mainly represented by paddy fields, forestland was mainly evergreen broad-leaved forests, while orchards were mainly the dwarfed economic forests such as litchi, longan, and passion fruit. In 1999, farmland was the dominant land use type, accounting for 41.61% of the study area. The following was orchard, which accounted for 28.88%. However, orchard was the dominant land use type in 2014, accounting for 30.76% of the study area, followed by farmland with 29.82%. Forestland was consistently the third landscape during 1999–2014, accounting for about 20% of the study area. Between 1999 and 2014, farmland decreased significantly from 3677.51 km
2 in 1999 to 2635.85 km
2 in 2014. Mud flat and bare land were also decreasing with a reduction of 67.28 km
2 and 59.16 km
2, respectively. In contrast, other land use types were increasing in area. The largest growth was observed in construction land, which increased from 1.38% in 1999 to 6.74% in 2014. The second was forestland, which increased from 17.25% in 1999 to 21.66% in 2014.
The land use change in the Beibu Gulf presented a significant spatial heterogeneity (
Figure 2). In 1999, farmland was primarily distributed in the eastern of Beibu Gulf area, including Haicheng District, Yinhai District, Tieshangang District, Hepu County, and Qinnan District, and it accounted for about 90% of the total farmland area. The Fangcheng District and Dongxing District have about 90% of the total forestland area. Orchard was mostly clustered in Qinan District and the northern of Hepu County. The construction land was gathered within each district and county by region, mainly concentrated in Haicheng District, Qinnan District, and Hepu County. Moreover, water bodies were largely located in Hepu County and Qinan District, while aquaculture land was aggregated in coastal areas, such as Hepu County and Yinhai District. Although the area occupied by bare land and mud flat in the study area were small, but they were scattered in each district and county. However, the farmland in Qinnan District, Hepu County, and Gangkou District was largely reduced and converted to forestland, orchard and construction land during 1999–2014. Therefore, farmland was mainly distributed in Yinhai District, Hepu County, and Tieshangang District in 2014, while forestland was mainly located in Fangcheng District, Dongxing District, Qinnan District, and the eastern of Hepu County. Meanwhile, the orchard area had grown significantly in the Gangkou District. In addition to Haicheng District, Qinnan District, and Hepu County, construction land was also sporadically distributed in other districts. Other land use types such as mud flat and aquaculture land only had increased or decreased in area, their distribution in Beibu Gulf area remained consistent during 1999–2014.
Based on the kernel density analysis from 1999 to 2014, the three increasing land use types have more obvious spatial clustering characteristics in Beibu Gulf area (
Figure 3). According to the kernel density of each interval, they could be divided into high-density hotspot areas (the interval of >400), medium-density hotspot areas (the interval of 100–400), and low-density hotspot areas (the interval of 0–100). The high-density hotspot areas of increased forestland were mainly concentrated in the southwestern of Qinnan District and the northeastern of Hepu County. The medium-density hotspot areas of increased forestland were mainly distribution in Fangcheng District and Dongxing District. Increased orchard was the most highly clustered in the eastern and southern of Qinan District, the northwestern of Hepu County and the Gangkou District. The medium-density hotspot areas of it was concentrated in the eastern of Hepu County and Fangcheng District, the northern of Yinhai District and Tieshangang District, and Qinnan District. In addition, the high clusters of increased construction land were mainly concentrated in the Haicheng District, the western of Yinhai District and Qinnan District, the southeastern of Tieshangang District and the southern of Dongxing District. Meanwhile, the medium-density hotspot areas of increased construction land were mainly located in the southern of Fangcheng District and Gangkou District, as well as the southern and eastern of Hepu County.
3.2. Land Use Scenario Simulation in 2030
According to the land use area for three scenarios, farmland, forestland, and orchard were the main land use types during 2014–2030 (
Table 6). In the three scenarios, farmland area was the largest in the ecotourism development scenario with 2539.31 km
2 and the smallest in the regular growth scenario with 2479.59 km
2. Forestland area was the largest with 2297.50 km
2 in the ecological protection scenario, followed by the ecotourism development scenario with 1915.09 km
2. The orchard area was the largest in the regular growth scenario with 2801.93 km
2 and the smallest in the ecological protection scenario with 2687.46 km
2. In addition, the areas of aquaculture land, bare land and construction land were the largest in the regular growth scenario with 312.51 km
2, 282.18 km
2, and 896.65 km
2, respectively. The smallest area was observed in the ecological protection scenario with 212.41 km
2, 53.75 km
2, and 595.54 km
2, respectively. In contrast, mud flat area was the largest with 202.14 km
2 in the ecological protection scenario, followed by the ecotourism development scenario with 150.46 km
2.
Compared to the land use area in 2014, only orchard area and construction land area increased in 2030 under the regular growth scenario. The construction land area increased the most at 301.11 km2, while the orchard area only increased by 82.64 km2. Meanwhile, forestland area and farmland area had the largest decrease with the decrement of 163.35 km2 and 156.26 km2, respectively. The bare land area remained constant from 2014 to 2030. In the ecological protection scenario, there was a significant increase in forestland area and mud flat area, with an increase of 382.88 km2 and 189.08 km2, respectively. The construction land area remained unchanged from 2014 to 2030. However, other land use areas showed a decline trend. Bare land, farmland and aquaculture land decreased the most, by 228.43 km2, 122.63 km2, and 101.68 km2, respectively. In the ecotourism development scenario, the mud flat area had the largest increment with 137.39 km2. The areas of construction land, orchard, and forestland also increased by 44.67 km2, 29.03 km2, and 0.47 km2, respectively. In contrast, other land use types were in decline from 2014 to 2030. Farmland area decreased the most with a reduction of 96.54 km2.
In the regular growth scenario, farmland was mainly distributed in Yinhai District, Tieshangang District and the southern of Hepu County (
Figure 4a). Forestland was mainly located in the Fangcheng District, Dongxing District and the western of Qinnan District, while orchard was mainly distributed in the Qinnan District, Gangkou District and the northwestern and eastern of Hepu County. Construction land was distributed within the eight districts and counties in a zoned aggregation, and aquaculture land was mainly located in coastal areas such as Hepu Country and Yinhai District. The water bodies were mainly located in the southern of the Beibu Gulf area and had a small area. There were also two large lakes in Hepu County. The mud flat was mainly located on the seashore, while the bare land was sporadically distributed within each district and county.
Compared with the regular growth scenario, the other two scenarios only had a small number of land use types distributed in a different way, while other land use types remained basically the same. In the ecological protection scenario (
Figure 4b), the construction land was mainly located in the Haicheng District, the western of Yinhai District, Gangkou District, and the Qinnan District, and the southern of Dongxin District. Other districts and counties also had a distribution of construction land, but the scale was much smaller than that of the regular growth scenario. Aquaculture land was mainly distributed in the southwest of Hepu County and the part of coastal areas in Yinhai District and Tieshangang District. The mud flat was mainly located in the southern of Yinhai District and the southeastern of Dongxing District. Besides, mud flat was also found in the coastal areas of Qinnan District and Hepu County. The distribution of other land use types was consistent with the regular growth scenario.
In the ecotourism development scenario (
Figure 4c), construction land had a similar distribution to the ecological protection scenario, but its scale was larger than that of the ecological protection scenario. Mud flat was mainly located in the coastal area of each district and county, and its spatial extent was the largest among the three scenarios. The distribution of aquaculture land was consistent with the ecological protection scenario, while that of other land use types—including farmland, forestland, orchard, bare land, and water—were consistent with the regular growth scenario.
Based on the kernel density analysis (
Figure 5a), construction land was mainly concentrated on the Haicheng District and the western of Yinhai District and Qinnan District. Compared to the distribution of construction land in 2014, construction land under the regular growth scenario increased significantly, aggregated within the eight districts and counties by subdivision. The distribution of construction land in the ecological protection scenario was consistent with that of 2014, which was almost unchanged. In the ecotourism development scenario, the clustering of construction land in the western of Gangkou District, the southeastern of Fangcheng District and the southwestern of Qinnan District had increased, while the clustering in other districts and counties remained the same as in 2014.
In 2014, forestland was mainly concentrated in Fangcheng District and Dongxing District (
Figure 5b). In contrast, the distribution of forestland under the three scenarios was consistent with that of 2014, but some districts or counties have significant changes in the clustering of it. The clustering of forestland within Dongxing District and Hepu County in the regular growth scenario was significantly weakened, while that of forestland within each district and county in the ecological protection scenario was significantly enhanced. The clustering of forestland in the ecotourism development scenario was consistent with 2014.
Orchard was mainly concentrated in Gangkou District, Qinan District, and Hepu District in 2014 (
Figure 5c). Overall, orchard was similarly distributed within the eight districts and counties in the three scenarios as in 2014, only varying in the clustering degree. In the regular growth scenario, the clustering of orchard in Dongxing District, the central of Qinnan District and the eastern of Hepu County was enhanced, while that of orchard in the southern of Gangkou District was weakened. However, the scale of orchard in this scenario was larger than that of 2014. In the ecological protection scenario, the clustering of orchard in the western of Fangcheng District was significantly weakened. In the ecotourism development scenario, the clustering of orchard in the eastern of Hepu County was significantly enhanced.
Mud flat was mainly located in the Dongxin District and Hepu Country, and its scale was extremely small in 2014 (
Figure 5d). The distribution and scale of mud flat in the regular growth scenario was consistent with that of 2014. The distribution of mud flat in the ecological protection scenario was consistent with that in the ecotourism scenario, which was mainly distributed in the coastal areas in each district and county such as Yinhai District, Gangkou District, and Dongxin District. However, the clustering of mud flat in the ecotourism scenario was higher than that of mud flat in the ecological protection scenario.
3.3. Ecosystem Service Value Dynamic
According to the ecosystem service value (ESV) in 1999, regulating services had the highest ESV in the study area at
$4.54 × 10
9, followed by supporting services at
$1.18 × 10
9, and provisioning services had the lowest ESV at
$1.20 × 10
8 (
Table 7). The ESV ranking among four categories for 2014 was consistent with that of 1999, but with a change in value. Overall, the ESV of four categories all increased during 1999–2014. The increase of supporting service value was
$1.65 × 10
8 and was the highest, while the increase of regulating service value was
$4.33 × 10
6 and was the lowest.
In the provisioning services, food production service had the highest ESV of $3.05 × 108 in 1999, followed by raw material production service with only $1.33 × 108. However, the ESV of water supply service was in deficit, with a loss of $3.18 × 108 in 1999. Between 1999 and 2014, only the ESV of food production service was decreasing with a reduction of $6.11 × 107, while the value of other ecosystem services had increased. The water supply service increased by $1.32 × 108 during 1999–2014. In the regulating services, hydrological regulation service and climate regulation service had the highest ESV of $2.29 × 109 and $1.21 × 109 in 1999 respectively, while purify environment service had the lowest ESV of $4.51 × 108. The ESV of hydrological regulation service decreased by $1.92 × 108 from 1999 to 2014, while that of other regulating services were increasing. Climate regulation service had the largest increase in ESV with $1.45 × 108, followed by purify environment service with $4.42 × 107, and gas regulation service had the smallest increase with $6.57 × 106. The ESV of biodiversity conservation service was the highest in the supporting services at $6.28 × 108 in 1999, followed by that of soil retention service at $4.77 × 108, while the ESV of nutrient cycling service was the lowest with $7.00 × 107. Between 1999 and 2014, the ESV of biodiversity conservation service and soil retention service increased by $9.26 × 107 and $7.62 × 107, respectively, while that of nutrient cycling service decreased by $3.80 × 106. Besides, the ESV of cultural services was higher in 2014 than that of 1999, and the value was $3.59 × 108. Its ESV increased by $4.69 × 107 during 1999–2014. In general, the total ESV of 11 ecosystem services in the Beibu Gulf area was growing from $6.15 × 109 to $6.45 × 109 between 1999 and 2014.
Compared to the ESV of ecosystem services in 2014, the ESVs of the provisioning services in 2030 for the ecological protection scenario and ecotourism development scenario were higher than those in 2014, while that of the provisioning services in the regular growth scenario was lower than those in 2014 (
Table 8). The ESV of the former two scenarios increased by
$8.24 × 10
7 and
$5.66 × 10
7, respectively, but the latter decreased by
$9.37 × 10
6 from 2014 to 2030. In the three scenarios, ecological protection scenario had the highest ESV for provisioning services, including food production service, raw material production service, and water supply service, while regular growth scenario had the lowest ESV. In the regular growth scenario, the ESV of water supply service was increased by
$9.87 × 10
6, and that of food production service and raw material production service was decreased by
$1.35 × 10
7 and
$5.75 × 10
6. In the ecological protection scenario, only the ESV of food production service was decreased by
$3.38 × 10
5. Similarly, only the ESV of food production service in the ecotourism development scenario decreased by
$2.53 × 10
6, while that of raw material production service and water supply service increased by
$1.01 × 10
4 and
$5.92 × 10
7, respectively.
In the three scenarios, the ESV of hydrological regulation service was the highest, followed by the climate regulation service, and the ESV of purify environment service was the lowest. Overall, the ESV of regulating service was the highest in the ecological protection scenario with $5.50 × 109 and the lowest in the regular growth scenario with $4.33 × 109. Compared to the ESV in 2014, the ESVs of all regulating services in 2030 under the regular growth scenario was reduced. The ESV of hydrological regulation service decreased notably with a reduction of $1.24 × 108, followed by climate regulation service with a reduction of $5.02 × 107. In contrast, the ESVs in the ecological protection scenario were increased. The ESV of hydrological regulation service and climate regulation service increased by $7.84 × 108 and $9.91 × 107, respectively. In the ecotourism development scenario, only the ESV of climate regulation service decreased by $4.63 × 106, while other regulation services had increased. In particular, the value of hydrological regulation service increased by $5.92 × 108 during 2014–2030.
For supporting services, the ESV of biodiversity conservation service was the highest, while that of nutrient cycling service was the lowest. In general, the ecological protection scenario had the highest ESV of supporting services at $1.37 × 109, followed by the ecotourism development scenario at $1.33 × 109, and the lowest ESV in the regular growth scenario at $1.28 × 109. In the regular growth scenario, the ESV of biodiversity conservation service was reduced by $3.88 × 107 during 2014–2030. Meanwhile, the ESV of soil retention service was reduced by $2.14 × 107, and that of nutrient cycling service was reduced by $3.07 × 106. In the ecological protection scenario, only the ESV of biodiversity conservation service decreased by $7.95 × 106, while that of soil retention service and nutrient cycling service increased by $3.37 × 107 and $1.38 × 106, respectively. In contrast, only the ESV of soil retention service increased by $6.38 × 105 in the ecotourism scenario for the period 2014–2030.
For cultural services, the ESV of aesthetic landscape service decreased in all scenarios compared to 2014. The regular growth scenario had the largest decrease with a reduction of $2.10 × 107. Among the three scenarios, the highest ESV of aesthetic landscape service presented in the ecotourism development scenario, while the lowest presented in the regular growth scenario with the value of $3.38 × 108. Overall, the total ESVs of 11 ecosystem services of the Beibu Gulf area was highest in the ecological protection scenario with the value of $7.51 × 109, followed by the ecotourism development scenario with the value of $7.11 × 109, and lowest in the regular growth scenario with the value of $6.14 × 109. In addition, the total ESV decreased by $3.16 × 108 during 2014–2030 in the regular growth scenario, while the total ESVs under the ecological protection scenario and the ecotourism development scenario both increased by $1.06 × 109 and $6.63 × 108, respectively.
According to the ESVs of various ecosystems in the three scenarios, there were significant differences between districts and counties (
Table 9). By comparing regular growth scenario and ecological protection scenario, we found that the ESV of farmland ecosystems in Fangcheng District under the regular growth scenario was higher than the ecological protection scenario, while the ESV of other districts and counties was lower than it. The ESVs of both forest ecosystems and wetland ecosystems within the eight districts and counties under this scenario were lower than the ecological protection scenario, of which the ESV of forest ecosystems had the largest gap of
$2.60 × 10
8 in Fangcheng District and the ESV of wetland ecosystems had the largest gap of
$2.40 × 10
8 in Qinnan District. However, the ESVs of desert ecosystems within each district and county were higher than the ecological protection scenario. Similarly, by comparing the regular growth scenario with the ecotourism development scenario, it can be demonstrated that only the ESV of farmland ecosystems in Fangcheng District and the ESVs of desert ecosystems within each district and county were higher than the ecotourism development scenario. The ESV of farmland ecosystems in Hepu County was consistent under the two scenarios. The ESV of forest ecosystems between these two scenarios had the largest difference of
$6.00 × 10
7 in Fangcheng District, followed by
$2.50 × 10
7 in Hepu County, while the ESV of wetland ecosystems had the largest difference of
$2.31 × 10
8 in Dongxing District, followed by
$2.17 × 10
8 in Yinhai District. By comparing ecological protection scenario and ecotourism development scenario, it could be found that the ESV of farmland ecosystems in Fangcheng District and Tieshangang District under the ecological protection scenario was lower than that of the ecotourism development scenario. The ESVs of forest ecosystems were higher than the ecotourism development scenario in all districts and counties, with the largest differences in Fangcheng District and Qinnan District of
$2.00 × 108 and
$8.00 × 10
7, respectively. In addition, the ESVs of wetland ecosystems in Dongxing District, Yinhai District, and Fangcheng District were lower than the ecotourism scenario, while other districts and counties were higher than the ecotourism scenario. The ESVs of desert ecosystems were higher in the ecotourism scenario, and their differences were highest in Fangcheng District and Qinnan District with
$6.59 × 10
5 and
$6.51 × 10
5, respectively.