RESULTS

3.1 Deforestation in Zhejiang Province

Figure 2. Temporal changes and spatial patterns of deforestation in Zhejiang Province from 2000 to 2020. (a) Cumulative forest loss (line) and annual loss (bar) from 2001 to 2019. (b) Scale of forest change. The bar charts are estimates based on satellite-based land-cover change products and statistical yearbook data (GlobeLand30, 2000–2020; ESA-CCI, 2000–2018; Statistical yearbook, 2000-2018). (c) Spatial distribution of net forest loss converted to other land classifications from 2000 to 2020. (d) Spatial distribution of net forest gain derives from other land classifications from 2000 to 2020
Our estimation based on the Hansen’s datasets shows that the total net loss of forest cover in Zhejiang Province from 2001-2019 is 279,501 ha (Figure 2a), a 4.7% decline in forest from the 2000 baseline. Annually, the rate of forest loss is about 12,549 ha per year during 2001-2008, representing an six-fold increase from about 4,245 ha in 2001 to nearly 26,993 ha in 2008. After 2008, the forest loss appeared to be stable during 2009-2019, but at a high rate of 16,282 ha each year. The cumulative forest loss monotonically increased during the entire study period.
In addition to the Hansen’s dataset, other data products including Globeland30 and ESA-CCI also show forest loss in Zhejiang Province after 2020 (Figure 2b). The Globeland30 dataset shows that forest cover decreased by 22,823 ha in Zhejiang during 2000-2020 and ESA-CCI 186,014 ha during 2000-2018, but Zhejiang’s statistical yearbook data show an increase in forest at the magnitude of 80,400 ha in Zhejiang from 2000-2018 (Figure 2b). We believe that remote sensing based land change results do not rely on government statistics and have higher spatial resolutions, are thus more likely to be less biased.
At the same time, we analyzed the spatial distribution of forest loss and gain in Zhejiang Province using GlobeLand30 high-resolution satellite images (Figure 2c and d). Over the space of Zhejiang Province, forest loss was found to be a prevalent phenomenon in many parts across the whole province. The landscape in terms of forest loss was shown to be scattered and rarely interconnected in general, but forest loss mainly dominated the mountainous regions within the province (Figure 2c). Overall, forest loss happened mainly in western and southern parts of Zhejiang, while the lost area was relatively small in the northeastern part. Forest loss exhibited heterogenous patterns at the municipality level (Figure 2c). Among all the municipalities in Zhejiang Province, the greatest loss in forest cover was observed in Lishui municipality in South Zhejiang, which is followed by Hangzhou municipality in West Zhejiang (the most economically developed city and the capital city of the province). Jiaxing municipality in proximity to Shanghai (the economic center in eastern China) possessed the least loss of forest. Forest loss was mainly due to conversion to cropland, with the largest amount occurring in Lishui.
Forest gain also occurred in almost all areas of Zhejiang Province. But the increase in forestland in most of those regions was less than the loss of forests. Among all the municipalities, Hangzhou has the largest increase and Jiaxing the smallest increase. The main source of forest gain is through converting cultivated land to forest, which is observed most in Wenzhou, followed by Lishui and Hangzhou (Figure 2d).

3.2 Urban expansion in Zhejiang Province

Figure 3. Temporal changes and spatial patterns of urban land in Zhejiang Province. Data of temporal change of land use are from (a) Statistical Yearbook between 2000-2018 and (b) GlobeLand30 between 2000-2020. (c) Spatial distribution of net forest loss converted to other land classifications from 2000-2020. (d) Spatial distribution of net forest gain derives from other land classifications over the same time span
According to a statistical yearbook (2001-2018) (Figure 3a), the proportion of urban area in Zhejiang Province increased steadily between 2001 and 2018. During the same period, the proportions of urban land area and forest land area increased by 70.29% and 1.45%, respectively, while the arable land area decreased by 5.26%. The rest of land types had a gradual decrease (Figure 3a). Using the Globeland30 data, we found that during 2000-2020, the area of urban land use increased by 169.45% while the area of forest and arable land decreased by 0.40% and 19.20%, respectively (Figure 3b).
Meanwhile, we used GlobeLand30 to analyze the spatial distribution of urban land changes in Zhejiang Province (Figure 3c and d). Spatially, urban area loss and gain occurred in almost all over the province, but the lost patches were dispersed and much smaller compared to the gained areas although the amount of net loss is only around 6% of the net gain. Of these, Ningbo and Hangzhou had the largest urban area gains, while Lishui and Zhoushan had the smallest gains (Figure 3c and d).
Figure 4. Sankey Map of Land use Transformation in Zhejiang Province from 2000 to 2020
The largest land use changes from 2000 to 2020 took place in the form of land use conversions between urban construction land, arable land, and forest in Zhejiang Province (Figure 4). The conversion from arable land to urban area is the largest (562,399 ha), followed by conversions from arable land to forest land (260,322 ha) and from forest land to arable land (210,474 ha). The changes in other land uses are relatively small. It seems that urban land expansion primarily comes from losses of arable land, while arable land increases as a result of deforestation.

3.3 Linkages of deforestation and urbanization: Lishui and Hangzhou

Figure 5. Land use change in Hangzhou and Lishui from 2000 to 2020
By comparing the land use change patterns of Hangzhou and Lishui, we found that the largest conversion from cultivated land to urban area in Hangzhou was 77,288 ha, followed by 37,036 ha from cultivated land to forest land and 26,551 ha from forest land to cultivated land. In Lishui, the maximum conversion was 37,198 ha of cultivated land to forest land, 36,787 ha of forest land to cropland, and only 16,910 ha of cultivated land to urban area. These results show that the biggest land change in Hangzhou is from cultivated land to urban land. The change in Lishui municipality is mainly due to conversion between cropland and forest land, and its area of forestland loss is larger than that of Hangzhou municipality (Figure 5).
Figure 6. Socioeconomic change in Hangzhou and Lishui municipality of Zhejiang Province from 2000 to 2019
Hangzhou’s GDP continued to increase from 138.256 to 1537.305 billion yuan during 2000-2019, representing an 11-fold increase. The resident population of Hangzhou municipality increased by 148% from 7.02 to 10.36 million. Lishui’s GDP, on the other hand, only increased from 13.676 to 147.661 billion yuan during 2000-2019 with a slow resident population growthfrom 2.16 to 2.21 million (Figure 6).
Figure 7. Telecoupling of deforestation and urbanization in Hangzhou and Lishui
Considering the representative land use change in Hangzhou and Lishui municipalities, they illustrate an ideal case for demonstrating the telecoupling concept (Jianguo Liu, 2017; Sun Jing et al., 2020). In Zhejiang Province, Hangzhou is the most urbanized municipality while Lishui is comparatively less urbanized. As Hangzhou’s urban population increased, its limited land resources can no longer meet the needs of the region’s urbanization. Therefore, urban expansion inevitably took arable land resources from surrounding areas, leading to a reduction in arable land area. The governments could enact policies to find back-up land resources through the role of an intermediate agent (such as a higher-level government). The municipality of Lishui, however, has sufficient back-up land resources due to its lower level of economic development, massive out-migration of rural population, and slower urban development. Through the role of an agent, a local government can supply its reserved land resources to meet the need of other areas in need of land supply. Land and money may be exchanged between two such regions in the form of flows to get what each need. For instance, Hangzhou may buy back-up land resources from Lishui for urbanization, whereas Lishui could receive a corresponding compensation. Meanwhile, Lishui may also suffer from land degradation and loss of ecosystems in the region as a result (Figure 7).