4 DISCUSSION

4.1 Causes of deforestation in Zhejiang Province

Zhejiang Province has experienced massive deforestation since 2000, representing a significant land use change in the region. During the study period, remote sensing data showed a decrease in forest area in Zhejiang Province between 2000 and 2020.
According to satellite-based land use and land cover products, we detected a substantial decrease in forest area in the early 2000s (Figure 2b). While data on forest decline are not uniform, inconsistent definition of forest may be a plausible reason for explaining the different rates of forest loss among various land cover products (H. Chen et al., 2020; Zeng et al., 2018). The Hansen’s forest dataset show that extensive deforestation occurred in Zhejiang Province in the early 2000s (Xiong et al., 2020). However, the GlobeLand30 (30 m) imagery, which has a higher resolution, showed less deforestation compared to the Hansen’s dataset. The GlobaLand30 dataset was produced with a globally trained machine learning algorithm. This algorithm, less suitable for detecting irregularly shaped mountain crop fields, may account for its underestimation of forest loss. The resolution of the ESA-CCI (300m) product requires a classification scheme to use mosaic classes, which might result in mixing cropland with other land cover types. These mosaic categories may also underestimate the area of agricultural land (Ozdogan & Woodcock, 2006), and thus may also have underestimated forest loss. Yet, despite the inconsistent changes in forest loss area, we believe that remotely sensed land data, not relying on government statistics, are more likely to show objective patterns with relatively high reliability (Zeng et al., 2018).
Forest transition theory assumes that forest in a country or a region will shift from a net loss to a net gain along with socio-economic development (Meyfroidt, Rudel, & Lambin, 2010). However, it has also been shown that if a country or region is engaged in cross-region trade flows (e.g., South America soybean), then deforestation may positively correlate with the growth of urban population and agricultural exports (DeFries et al., 2010). Some studies also show that forestland tends to be lost as China’s GDP per capita rises (Vina, McConnell, Yang, Xu, & Liu, 2016). The accelerated urbanization in Zhejiang Province from 2000 to 2020 has resulted in conversion from agricultural land to urban land (Figure 4). Besides, we also observed a large amount of forest loss (i.e., land converted to arable land) (Figure 4) and continued forest fragmentation in Zhejiang Province, which might arise from reclamation of agricultural land (Xiong et al., 2020). Such large-scale and rapid deforestation should be of concern to governments and the scientific community.

4.2 Policy evolution on the Balance of Arable Land System (BALS)

The establishment and implementation of the BALS policy, established to play a dual role of guaranteeing economic development and protecting the environment, might have achieved this goal (Long et al., 2012; Xin & Li, 2018). However, this policy also attracted great debate and was changed several times.
China’s BALS policy can be roughly divided into three stages (Figure 8). The initial stage was for systematic policy construction. China first proposed the arable land acquisition and compensation system in 1997. The BALS policy was formally written into the new Land Management law in 1998. To protect arable land, China amended the Land Management Law to maintain the amount of arable land with a new arable land replenishment mechanism, giving rise to the “occupy one ha and replenish one ha” policy in 1999. In 2001, all provinces (including autonomous regions) basically achieved a balance in the amount of arable land, ensuring that the total area of arable land was not in decline. The focus of land consolidation during this period was to increase the amount of arable land, provide space for urbanization and industrialization, ensure food security, and increase farmers’ income (Wei Song & Pijanowski, 2014).
Figure 8. The Balance of Arable Land System (BALS) policy evolution
The second stage featured systematic policy improvement. From 2004 onwards, Chinese government has been focusing on the quality of newly created arable land, issuing a series of policies. In 2004, the government introduced measures to control the quality of land during land conversions, assuring that the quantity and quality of newly created arable land should meet national standards. The year of 2005 saw a shift of focus to administrative accountability, starting to implement an accountability target assessment and chief executive responsibility system. In 2006, several specific institutional assessment mechanisms were introduced, which issued several pilot projects demanding that decreases in rural construction land be linked to increases in urban construction land (Long & Li, 2012). In 2008, China proposed to designate permanent basic cropland to ensure that the quantity of arable land would not be reduced and its quality would be improved. In 2009, arable land was required to be fully replenished before land occupation, and subsequent operational rules were issued to strengthen the quality and management of replenished arable land. At this stage, the quantity and quality of arable land was protected, and the productivity of arable land was relatively ensured (Long, 2020).
The third stage was characterized by a “Trinity” of quantitative, qualitative, and ecological policy. The Chinese government authorities have further improved their management approach by emphasizing a policy of balancing land occupation and compensation between provinces. To ensure the quality of replenished arable land, a decree was issued in 2012 to strengthen the management of arable land and maintain land quality. In 2014, the Chinese government proposed that the occupation of high-quality land and paddy land should be compensated; in 2016, a combination of quality improvement and remediation was proposed. In 2018, China formed the Ministry of Natural Resources to strengthen the unified integration and management of natural resources. In 2018, the Ministry of Natural Resources issued a national coordinated management decree to establish the inter-provincial replenishment of arable land along with a management scheme for construction land transfer between rural and urban areas in different provinces. This provides a clear direction for the unified management of arable land resources, focusing more on the ecological functions of land use (Long, 2020; Long et al., 2018).
Zhejiang Province, the first province to implement the BALS policy in China, needs to evaluate its consequences and effectiveness. It is easy to focus on assessing the quantitative balance and difficult to assess the qualitative and ecological balance (Lin et al., 2017). As cropland varies greatly over spatial and temporal dimensions, the policy needs to be further adjusted in the future to effectively protect high-quality cropland resources and support national food security and ecological safety. In Zhejiang, for example, the conversion of forestland to agricultural land has occurred in parallel with the BALS policy implementation, inevitably leading to loss of forestland (Xiong et al., 2020).

4.3 Telecoupling and the no net loss of cropland policy

The telecoupling framework can explain the long-range land change mechanisms (Jianguo Liu, 2017; Sun Jing et al., 2020), which may apply to this particular deforestation phenomenon in Zhejiang Province. The expansion of arable land, although able to satisfy market demands and conform with policies, may lead to ecological problems such as deforestation (Henders, Ostwald, Verendel, & Ibisch, 2018). Our study shows that the largest source of urban construction land in Zhejiang Province is arable land, and the largest source of arable land is through converting forest land (Figure 4). In the context of the BALS policy, the balance of cropland quantity needs to be guaranteed, which inevitably leads to converting forest land to arable land in Zhejiang Province. In China, large municipalities like Hangzhou are often blamed for the massive loss of high-quality arable land (Hu et al., 2020). The largest land change in Hangzhou is the conversion of arable land to urban land; while in Lishui, the major changes feature mutual conversions between arable land and forest land with the area of forestland loss larger than that in Hangzhou (Figure 5). The less-urbanized Lishui municipality also has the most deforested area in Zhejiang Province (Xiong et al., 2020). A stable or depopulated landscape may face a higher deforestation pressure largely because distant urbanization processes would demand increases in arable land in surrounding places. In contrast, the BALS policy allows for the deployment of land resources within the province. In this context, Lishui municipality becomes the place that supplies land resources for other cities and towns because of its lower level of urbanization. As a result, Lishui will likely experience land degradation and ecosystem loss.
Figure 9. Mutual relationships between productive, living, and ecological land
The simultaneous demand of land for urban development, ecological conservation, and rural revitalization is a challenge for China’s policy and planning for land and other related resources (R. Chen, Ye, Cai, Xing, & Chen, 2014). A rational land planning and management system needs to reconcile the balance among productive land, living land, and ecological land (Figure 9). However, many land use policies implemented in China for more than 20 years are irrational in some aspects as they do not coordinate well with each other. These policies have either failed to ensure the productivity of arable land, or resulted in ecological land degradation, which clearly contradict China’s strategy for sustainable development (Bryan et al., 2018). Moreover, appropriate land use policies must be in line with socioeconomic development (Xin & Li, 2018). With urbanization, the continuing migration of rural population to cities will inevitably lead to the emergence of hollow villages and the abandonment of large amounts of arable land in rural areas (Zhang, Song, & Chen, 2018), which will further challenge the effectiveness and efficiency of BALS (S. Li et al., 2018).