*Article* **Ecosystem Services Changes on Farmland in Response to Urbanization in the Guangdong–Hong Kong–Macao Greater Bay Area of China**

**Xuege Wang 1,2,3,4, Fengqin Yan 2,3,5, Yinwei Zeng <sup>6</sup> , Ming Chen <sup>2</sup> , Bin He <sup>7</sup> , Lu Kang <sup>2</sup> and Fenzhen Su 1,2,3,7,\***


**Abstract:** Extensive urbanization around the world has caused a great loss of farmland, which significantly impacts the ecosystem services provided by farmland. This study investigated the farmland loss due to urbanization in the Guangdong–Hong Kong–Macao Greater Bay Area (GBA) of China from 1980 to 2018 based on multiperiod datasets from the Land Use and Land Cover of China databases. Then, we calculated ecosystem service values (ESVs) of farmland using valuation methods to estimate the ecosystem service variations caused by urbanization in the study area. The results showed that 3711.3 km<sup>2</sup> of farmland disappeared because of urbanization, and paddy fields suffered much higher losses than dry farmland. Most of the farmland was converted to urban residential land from 1980 to 2018. In the past 38 years, the ESV of farmland decreased by 5036.7 million yuan due to urbanization, with the highest loss of 2177.5 million yuan from 2000–2010. The hydrological regulation, food production and gas regulation of farmland decreased the most due to urbanization. The top five cities that had the largest total ESV loss of farmland caused by urbanization were Guangzhou, Dongguan, Foshan, Shenzhen and Huizhou. This study revealed that urbanization has increasingly become the dominant reason for farmland loss in the GBA. Our study suggests that governments should increase the construction of ecological cities and attractive countryside to protect farmland and improve the regional ESV.

**Keywords:** ecosystem service value; farmland loss; construction land expansion; remote sensing

#### **1. Introduction**

Unprecedented urbanization has been occurring worldwide, which has caused the urban population to currently account for more than half of the world0 s population; this proportion is expected to increase to nearly three-quarters by 2050 [1]. Considerable urban expansion has mostly taken place in developed countries in the past, but it will mainly occur in developing countries in the coming decades. It is estimated that the urban area in developing countries will increase to 1.2 million km<sup>2</sup> by 2050, which is four times the urban area in 2000 [2]. As the largest developing country in the world, China has been experiencing unprecedented urban expansion after the implementation of the reform and opening up policy in 1978, with the increase of the urban population from 17.92% in 1978 to 58.5% in 2017, and a projected rise to 70% by 2030 [3,4]. Urban agglomerations in China

**Citation:** Wang, X.; Yan, F.; Zeng, Y.; Chen, M.; He, B.; Kang, L.; Su, F. Ecosystem Services Changes on Farmland in Response to Urbanization in the Guangdong–Hong Kong–Macao Greater Bay Area of China. *Land* **2021**, *10*, 501. https://doi.org/10.3390/ land10050501

Academic Editors: Alessio Russo and Giuseppe T. Cirella

Received: 22 March 2021 Accepted: 4 May 2021 Published: 8 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

are hotspots of urban expansion because of intensive anthropogenic activities and rapid economic growth, such as the Jing-Jin-Ji urban agglomerations, the Yangtze River Delta urban agglomerations, and the Pearl River Delta urban agglomerations [3,5,6].

Extensive urbanization has substantially encroached on farmland, wetlands and forestland, which has further led to decreases in the food supply, the loss of natural habitat, and the dramatic degradation of ecosystem services (ESs) [7–11]. Previous studies proved that urban expansion mainly originates from the conversion of farmland, and has also become the main source of farmland loss around the world [4,12–14]. In China, 74% of the lost farmland was converted into 85% of the newly increased urban land in the past decades [15]. Farmland preservation and urban expansion has become a hot topic. Urbanization has not only encroached substantially on farmland, but also impacted farmland quality and the intensity of agricultural activities [10,15,16]. Continued urban expansion will accelerate agricultural land expansion, putting high pressure on the natural ecosystems [10]. Moreover, urbanization has negative impacts on the ESs supported by farmland; e.g., food production ES, cultural heritage, wildlife habitat and open space [17–20]. Narduccia et al. showed that urban areas have more negative impacts on ESs than farmland [12]. The massive farmland loss due to urbanization is increasingly threatening farmland preservation and relative ESs [10,21–23]. Previous studies put more attention on the spatial and temporal relationship between urbanization and farmland, but they rarely explored the additional influences of urbanization on ESs and ecosystem functions supported by farmland. Therefore, it is necessary to explore the impacts of urbanization on farmland and associated ESs.

Estimate ESs are significant and alert decision makers and the public to the importance of these services. In 1997, Costanza et al. introduced the principle, methods and results of estimating the global ecosystem service value (ESV) by synthesizing previous studies based on a wide variety of methods, and they found that the global gross domestic product is much lower than the ESV [24]. In 2007, Xie et al. improved the valuation methods of Costanza et al. and developed a unit value adapted to China, which can be applied to assess ESs in specific land use areas [25]. In 2015, Xie et al. modified their original approach, developing a method for evaluating the value equivalent factor in the per unit area, and proposed an integrated method for the dynamic evaluation of the Chinese terrestrial ESV [26]. Xie et al.'s equivalent value per unit area method for the Chinese ESV has been widely adopted, but if it is directly applied to the regional scope, then the results will be biased. Xu et al. proposed a regional revision method, which entails revising the average ES equivalent values of the entire country to that of the study area based on the data of food products per unit area [27]. Xu et al.'s method has been widely implemented in different regions and has increased the accuracy of change detection [8,9,28].

The Guangdong–Hong Kong–Macao Greater Bay Area (GBA) is one of the most urbanized, industrialized and populous areas in China. However, the GBA used to be a predominantly agricultural region before 1980, was dominated by farms and rural villages and was highly suitable for agricultural development [29,30]. Extensive urbanization in the GBA has led to a large area of farmland loss, but few studies have examined the farmland loss in the GBA due to urbanization over recent decades. In previous studies, urbanization mostly refers to the expansion of urban land or urban areas, but it also includes the expansion of residential, road and industrial land [11,12,31,32]. To achieve a comprehensive understanding of urbanization, this study defines urbanization as construction land sprawl; namely, the expansion of residential areas, industrial areas, mining areas, transportation land and other construction land. Therefore, the changes in farmland and its ESV because of urbanization were examined in the GBA from 1980 to 2018 in this study. The objectives include the following: (1) to analyze the spatial and temporal changes in construction land in 1980, 1990, 2000, 2010 and 2018; (2) to examine the characteristics of farmland changes due to urbanization during the past 38 years; and (3) to evaluate the ESV of farmland loss under the effects of urbanization over the past four decades.

#### **2. Materials and Methods 2. Materials and Methods**

*Land* **2021**, *10*, x FOR PEER REVIEW 3 of 18

#### *2.1. Study Area 2.1. Study Area*

The GBA is located in the south of China, covering a total area of approximately 55,000 km<sup>2</sup> , with a population of approximately 45 million people in 2018 (Figure 1). The study area consists of 11 cities; specifically, Guangzhou, Shenzhen, Zhuhai, Foshan, Dongguan, Zhongshan, Jiangmen, Huizhou, Zhaoqing, Hong Kong and Macao. It has several typical geomorphic types, including mountains, hills, terraces and plains. The study area is located in a subtropical climate zone with an average annual temperature of 22.3 ◦C and an average annual rainfall of 1832 mm, and 83% of the rainfall occurs in the rainy season [33]. The study area has transformed from a predominantly agricultural region into one of the most developed urban agglomerations in the world since China's reform and opening-up policy [30,34]. As one of the regions in China with the highest degree of openness and a notable economic vitality, the GBA occupies an important strategic position in the overall national development situation. The GBA is located in the south of China, covering a total area of approximately 55,000 km<sup>2</sup> , with a population of approximately 45 million people in 2018 (Figure 1). The study area consists of 11 cities; specifically, Guangzhou, Shenzhen, Zhuhai, Foshan, Dongguan, Zhongshan, Jiangmen, Huizhou, Zhaoqing, Hong Kong and Macao. It has several typical geomorphic types, including mountains, hills, terraces and plains. The study area is located in a subtropical climate zone with an average annual temperature of 22.3 °C and an average annual rainfall of 1832 mm, and 83% of the rainfall occurs in the rainy season [33]. The study area has transformed from a predominantly agricultural region into one of the most developed urban agglomerations in the world since China's reform and opening-up policy [30,34]. As one of the regions in China with the highest degree of openness and a notable economic vitality, the GBA occupies an important strategic position in the overall national development situation.

farmland changes due to urbanization during the past 38 years; and (3) to evaluate the

ESV of farmland loss under the effects of urbanization over the past four decades.

**Figure 1.** Location of the Greater Bay Area (GBA): from top to bottom, from left to right, cities are Hezhou (HZ), Wuzhou (WZ), Yangjiang (YJ), Zhaoqing (ZQ), Qingyuan (QY), Guangzhou (GZ), Foshan (FS), Jiangmen (JM), Zhongshan (ZS), Zhuhai (ZH), Macao (MC), Dongguan (DG), Shenzhen (SZ), Hong Kong (HK), Huizhou (HZ), Heyuan (HY), Shanwei (SW). **Figure 1.** Location of the Greater Bay Area (GBA): from top to bottom, from left to right, cities are Hezhou (HZ), Wuzhou (WZ), Yangjiang (YJ), Zhaoqing (ZQ), Qingyuan (QY), Guangzhou (GZ), Foshan (FS), Jiangmen (JM), Zhongshan (ZS), Zhuhai (ZH), Macao (MC), Dongguan (DG), Shenzhen (SZ), Hong Kong (HK), Huizhou (HZ), Heyuan (HY), Shanwei (SW).

#### *2.2. Data Source 2.2. Data Source*

The land use datasets of the farmland and construction land in this study related to 1980, 1990, 2000, 2010 and 2018 were extracted from the Land Use and Land Cover of China (CNLUCC) database [35]. The CNLUCC includes the land use and land cover (LULC) datasets for 1980, 1990, 1995, 2000, 2005, 2010, 2015 and 2018. These datasets were mainly produced by visual interpretation on the basis of a series of multitemporal Landsat images, including Thematic Mapper (TM), Enhanced Thematic Mapper (ETM) and Operational Land Imager (OLI) images. As one of the most accurate remote sensing monitoring The land use datasets of the farmland and construction land in this study related to 1980, 1990, 2000, 2010 and 2018 were extracted from the Land Use and Land Cover of China (CNLUCC) database [35]. The CNLUCC includes the land use and land cover (LULC) datasets for 1980, 1990, 1995, 2000, 2005, 2010, 2015 and 2018. These datasets were mainly produced by visual interpretation on the basis of a series of multitemporal Landsat images, including Thematic Mapper (TM), Enhanced Thematic Mapper (ETM) and Operational Land Imager (OLI) images. As one of the most accurate remote sensing monitoring data products in China, CNLUCC has played an important role in the investigation and research of national land resources, hydrology and ecology. In this study, the classification system included two primary classes and five subclasses, and the detailed information is listed in

Table 1. The two primary classes are farmland and construction land, and both were divided into several subclasses. Farmland is classified into two subclasses, namely paddy fields and dry farmland. Construction land includes three subclasses: urban residential land, rural residential land and other construction land. In addition, specific social statistical data, including the yield and area of the main grains in Guangdong Province and China, were obtained from the Statistical Yearbook of Guangdong Province and China of 1980, 1990, 2000, 2010 and 2018 and were downloaded from the China National Knowledge Infrastructure (CNKI) database (http://data.cnki.net/Yearbook/Navi?type=type&code=A (accessed on 20 May 2020)).


**Table 1.** Classification of farmland and construction land in this study.

#### *2.3. Methods*

2.3.1. Analysis of the Changes in Land Use Types

To analyze the degree of changes in land use types, the annual change area (ACA, km<sup>2</sup> /year) was calculated as follows:

$$\text{ACA} = \frac{A\_2 - A\_1}{t} \tag{1}$$

where *A*<sup>1</sup> and *A*<sup>2</sup> are the area of farmland or construction land at the start and end dates, respectively, and *t* is the time span of the study period.

#### 2.3.2. Assessment of the Ecosystem Service Value of Farmland

The ESV was divided into four classes and eleven subclasses based on previous research [24,26,36]. The four classes were provision services, regulation services, support services and cultural services. Provision services include food production, primary production and water supply. Regulation services comprise gas regulation, climate regulation, environmental purification and hydrological regulation. Support services consist of soil conservation, nutrient cycling and biodiversity conservation. Cultural services include recreational and aesthetic values. Based on the research of Costanza et al. and

Xie et al. [24,26,37], the ESV of farmland in the GBA from 1980 to 2018 was quantitatively estimated with the ecosystem service value coefficient method as follows:

$$\text{ESV} = \sum (A\_k \times V\_{ck}) \tag{2}$$

where ESV denotes the total annual value of the ESs, and *A<sup>k</sup>* and *Vck* are the area and value coefficient, respectively, for land use type *k*. In this study, the ESV of paddy fields and dry farmland were calculated first, then they were summed up to get the ESV of farmland.

To better reflect the ESV changes in the different study periods, the main grain yield in the GBA compared to that in all of China was used to revise the ecosystem services equivalent value [27]. The correction coefficient *β<sup>t</sup>* was calculated as follows:

$$
\beta\_t = y\_t / \Upsilon\_t \tag{3}
$$

where *y<sup>t</sup>* and *Y<sup>t</sup>* are the grain yields in Guangdong Province and China at time *t*, respectively. Because the data of grain yield in the GBA in the early study periods was not easy to acquire, the grain yield in the GBA was replaced by that of Guangdong Province in this study. Thus, *β*1980, *β*1990, *β*2000, *β*<sup>2010</sup> and *β*<sup>2018</sup> were 1.34, 1.24, 1.38, 1.05 and 0.99, respectively.

The economic value of the ecosystem services equivalent value per unit area (*Ea*) was calculated as follows [8,28]:

$$E\_a = 1/7 \times (\text{P} \times \text{Y})/A \tag{4}$$

where *Y* and *A* are the total yield and area of the main grains in Guangdong Province in 2018, respectively; *P* is the average price of the main grains. In order to eliminate the impacts of changing values of the currency (yuan), P was given the value of 3.77 yuan/kg, which is the average price of the main grains in 2018, obtained from the Grain Net of South China (https://gdgrain.com/#!/list?params=%7B%22type%22:8%7D (accessed on 15 May 2020)).

#### **3. Results**

#### *3.1. Urbanization in the GBA from 1980 to 2018*

From 1980 to 2018, the area of construction land in the GBA showed a discernible growth trend, increasing from 2607.4 km<sup>2</sup> to 8243.5 km<sup>2</sup> , a more than twofold increase (Figure 2a). A similar trend was observed in urban residential land and other construction land, which increased nearly sixfold (from 694.7 km<sup>2</sup> to 4778.5 km<sup>2</sup> and from 285.7 km<sup>2</sup> to 1991.2 km<sup>2</sup> , respectively) (Figure 2b). The area of rural residential land peaked in 2000 with 1855.9 km<sup>2</sup> and then declined. By 2018, urban residential land occupied a dominant position among construction land (58.0%), followed by other construction land (24.2%) and rural residential land (17.9%). Regionally, construction land expanded rapidly in the 11 cities of the GBA in the past 40 years (Figure 2c). Among the 11 cities, only one had an area of construction land over 500 km<sup>2</sup> in 1980 (Guangzhou), but there were 8 cities by 2018. Among the four different periods, the highest ACA of construction land in the GBA was 279.7 km2/year from 2000–2010, followed by 149.1 km2/year from 1990–2000 and 108.7 km2/year from 2010–2018 (Figure 2d). A similar trend also appeared in urban residential land and other construction land (Figure 2e). The ACA of rural residential land was negative in the final two study periods. The highest ACA of most cities was observed from 2000–2010 (Figure 2f). Lastly, although the urbanization and development degree of Hong Kong and Macao were much higher than that of the other cities in the GBA, their construction land expansion area and rate seemed much smaller than those of the other cities because their own administrative area was relatively small.

**Figure 2.** Construction land evolution of the GBA from 1980 to 2018: (**a**–**c**) are the growth of construction land for the total GBA, different types of construction land and 11 cities, respectively; (**d**–**f**) present the annual change area (ACA) of the construction land of the total GBA, different types and 11 cities, respectively. URL: urban residential land; RRL: rural residential land; OCL: other construction land. **Figure 2.** Construction land evolution of the GBA from 1980 to 2018: (**a**–**c**) are the growth of construction land for the total GBA, different types of construction land and 11 cities, respectively; (**d**–**f**) present the annual change area (ACA) of the construction land of the total GBA, different types and 11 cities, respectively. URL: urban residential land; RRL: rural residential land; OCL: other construction land.

#### *3.2. Changes in Farmland in the GBA from 1980 to 2018 3.2. Changes in Farmland in the GBA from 1980 to 2018*

Over the past 38 years, the total area of the farmland in the GBA varied from 16,640.1 km<sup>2</sup> to 12,417.4 km<sup>2</sup> , and more than a quarter of the total farmland area of 1980 had been lost (Figure 3a). An overall decreasing trend was also observed in the different categories of farmland and 11 cities (Figure 3b,c). The area of paddy fields was far larger than the area of dry farmland in the GBA; correspondingly, the paddy field loss (3086.3 km<sup>2</sup> ) was much greater than the dry farmland loss (1136.4 km<sup>2</sup> ) (Figure 3b). From 1980 to 2018, the top five farmland losses were in Guangzhou (−827.2 km<sup>2</sup> ), Dongguan (−744.7 km<sup>2</sup> ), Jiangmen (−452.9 km<sup>2</sup> ), Foshan (−428.2 km<sup>2</sup> ) and Shenzhen (−414.1 km<sup>2</sup> ) (Figure 3c). The highest ACA of farmland was −174.1 km<sup>2</sup> /year from 2000–2010, followed by −158.0 km<sup>2</sup> /year from 1990–2000, while the lowest ACA was −27.3 km<sup>2</sup> /year from 2010–2018 (Figure 3d). The ACA of paddy fields, dry farmland and the farmland in 11 cites was also higher from 1990–2000 and 2000–2010 (Figure 3e,f). Over the past 38 years, the total area of the farmland in the GBA varied from 16,640.1 km<sup>2</sup> to 12,417.4 km<sup>2</sup> , and more than a quarter of the total farmland area of 1980 had been lost (Figure 3a). An overall decreasing trend was also observed in the different categories of farmland and 11 cities (Figure 3b,c). The area of paddy fields was far larger than the area of dry farmland in the GBA; correspondingly, the paddy field loss (3086.3 km<sup>2</sup> ) was much greater than the dry farmland loss (1136.4 km<sup>2</sup> ) (Figure 3b). From 1980 to 2018, the top five farmland losses were in Guangzhou (−827.2 km<sup>2</sup> ), Dongguan (−744.7 km<sup>2</sup> ), Jiangmen (−452.9 km<sup>2</sup> ), Foshan (−428.2 km<sup>2</sup> ) and Shenzhen (−414.1 km<sup>2</sup> ) (Figure 3c). The highest ACA of farmland was <sup>−</sup>174.1 km2/year from 2000–2010, followed by <sup>−</sup>158.0 km2/year from 1990–2000, while the lowest ACA was <sup>−</sup>27.3 km2/year from 2010–2018 (Figure 3d). The ACA of paddy fields, dry farmland and the farmland in 11 cites was also higher from 1990–2000 and 2000–2010 (Figure 3e,f).

tively.

**Figure 3.** Farmland evolution of the GBA from 1980 to 2018: (**a**–**c**) are the change of farmland for the total GBA, different types and 11 cities, respectively; (**d**–**f)** present the ACA of farmland of the total GBA, different types and 11 cities, respec-**Figure 3.** Farmland evolution of the GBA from 1980 to 2018: (**a**–**c**) are the change of farmland for the total GBA, different types and 11 cities, respectively; (**d**–**f**) present the ACA of farmland of the total GBA, different types and 11 cities, respectively.

From 1980 to 2018, the conversion from farmland to construction land was mainly distributed at the center of the GBA; e.g., Guangzhou, Dongguan, Shenzhen, Foshan and Zhongshan (Figure 4). Over the past 38 years, 3711.3 km2 offarmland was converted into construction land. Among the four adjacent periods, the farmland loss due to urbanization increased drastically from the first study period (479.5 km<sup>2</sup> ) to the third study period (1772.4 km<sup>2</sup> ) but then dropped sharply in the last study period (439.2 km<sup>2</sup> ) (Figure 5a). However, the percentage of farmland loss due to urbanization in the total farmland loss exhibited a growth trend, increasing from 39.3% to 96.6% from the first study period to the last study period (Figure 5b), which demonstrates that urbanization increasingly became the dominant reason for farmland loss. From 1980 to 2018, the urbanization-triggered losses of paddy fields and dry farmland were 2599.4 km<sup>2</sup> and 1038.7 km<sup>2</sup> , respectively (Figure 5c). The paddy field area lost due to urban sprawl was much more considerable than that of dry farmland over the past four decades. From 2000–2010, the losses of paddy fields and dry farmland due to construction land expansion were the highest, with areas of 1183.1 km<sup>2</sup> and 589.3 km<sup>2</sup> , respectively. Over the past 38 years, 2086.7 km<sup>2</sup> , 708.0 km<sup>2</sup> and 916.5 km<sup>2</sup> of farmland disappeared because of the expansion of urban residential land, rural residential land and other construction land, respectively (Figure 5d). The largest farmland loss due to urban residential land expansion was 1000.5 km<sup>2</sup> from 2000–2010, From 1980 to 2018, the conversion from farmland to construction land was mainly distributed at the center of the GBA; e.g., Guangzhou, Dongguan, Shenzhen, Foshan and Zhongshan (Figure 4). Over the past 38 years, 3711.3 km<sup>2</sup> of farmland was converted into construction land. Among the four adjacent periods, the farmland loss due to urbanization increased drastically from the first study period (479.5 km<sup>2</sup> ) to the third study period (1772.4 km<sup>2</sup> ) but then dropped sharply in the last study period (439.2 km<sup>2</sup> ) (Figure 5a). However, the percentage of farmland loss due to urbanization in the total farmland loss exhibited a growth trend, increasing from 39.3% to 96.6% from the first study period to the last study period (Figure 5b), which demonstrates that urbanization increasingly became the dominant reason for farmland loss. From 1980 to 2018, the urbanization-triggered losses of paddy fields and dry farmland were 2599.4 km<sup>2</sup> and 1038.7 km<sup>2</sup> , respectively (Figure 5c). The paddy field area lost due to urban sprawl was much more considerable than that of dry farmland over the past four decades. From 2000–2010, the losses of paddy fields and dry farmland due to construction land expansion were the highest, with areas of 1183.1 km<sup>2</sup> and 589.3 km<sup>2</sup> , respectively. Over the past 38 years, 2086.7 km<sup>2</sup> , 708.0 km<sup>2</sup> and 916.5 km<sup>2</sup> of farmland disappeared because of the expansion of urban residential land, rural residential land and other construction land, respectively (Figure 5d). The largest farmland loss due to urban residential land expansion was 1000.5 km<sup>2</sup> from 2000–2010, followed by 603.8 km<sup>2</sup> from 1990–2000 and 305.6 km<sup>2</sup> from 1980–1990. Farmland being converted

to rural residential land mainly occurred from 1990–2000 with an area of 305.6 km<sup>2</sup> , then from 2000–2010 (249.7 km<sup>2</sup> ) and from 1980–1990 (132.6 km<sup>2</sup> ). The largest farmland loss caused by other construction land was 522.2 km<sup>2</sup> from 2000–2010, followed by 242.4 km<sup>2</sup> from 2010–2018. In the past four decades, the top five cities that experienced farmland loss because of urbanization were Guangzhou (824.42 km<sup>2</sup> ), Dongguan (653.34 km<sup>2</sup> ), Foshan (591.03 km<sup>2</sup> ), Shenzhen (371.35 km<sup>2</sup> ) and Huizhou (362.22 km<sup>2</sup> ) (Figure 5e). In most areas, this conversion was concentrated between 1990 and 2010. converted to rural residential land mainly occurred from 1990–2000 with an area of 305.6 km<sup>2</sup> , then from 2000–2010 (249.7 km<sup>2</sup> ) and from 1980–1990 (132.6 km<sup>2</sup> ). The largest farmland loss caused by other construction land was 522.2 km<sup>2</sup> from 2000–2010, followed by 242.4 km<sup>2</sup> from 2010–2018. In the past four decades, the top five cities that experienced farmland loss because of urbanization were Guangzhou (824.42 km<sup>2</sup> ), Dongguan (653.34 km<sup>2</sup> ), Foshan (591.03 km<sup>2</sup> ), Shenzhen (371.35 km<sup>2</sup> ) and Huizhou (362.22 km<sup>2</sup> ) (Figure 5e). In most areas, this conversion was concentrated between 1990 and 2010.

from 1980–1990. Farmland being

from 1990–2000 and 305.6 km<sup>2</sup>

*Land* **2021**, *10*, x FOR PEER REVIEW 8 of 18

followed by 603.8 km<sup>2</sup>

**Figure 4.** Spatial and temporal distribution of farmland loss due to urbanization in the GBA from 1980 to 2018. **Figure 4.** Spatial and temporal distribution of farmland loss due to urbanization in the GBA from 1980 to 2018.

#### *3.3. The Impact of Urbanization on the Ecosystem Service Value of Farmland*

An unprecedented expansion of construction land has led to a great loss of farmland, which has further impacted the ESs and functions of farmland in the GBA. In the past four decades, the ESV of farmland decreased by 5036.7 million yuan due to construction land encroachment, accounting for 43.1% of the total ESV loss of farmland (Table S1). The highest ESV loss of farmland because of urbanization was 2177.5 million yuan from 2000–2010, followed by 1655.1 million yuan from 1990–2000 (Figure 6a). For paddy fields and dry farmland, the ESV losses caused by urbanization were 3442.1 million yuan and 1594.7 million yuan, respectively (Figure 6b). The largest ESV losses of both paddy fields and dry farmland appeared from 2000–2010 (1438.8 million yuan and 738.7 million yuan, respectively), and the second-largest ESV losses were from 1990–2000 (1040.0 million yuan and 615.1 million yuan, respectively). Because of urbanization, the most affected ecosystem service function was hydrological regulation with a loss of 2514.2 million yuan, followed by food production and gas regulation with losses of 1541.4 million yuan and 1248.6 million yuan, respectively (Figure 6c). Specifically, the ESV of the water supply of farmland was positive under the effects of urbanization, because paddy fields need a large amount water for the growth of aquatic plants. To some degree, the loss of paddy fields is a benefit for the water supply. Regionally, the total ESV loss of farmland caused by urbanization in Guangzhou was the largest (1110.85 million yuan), followed by that in

Dongguan (921.8 million yuan), Foshan (791.6 million yuan), Shenzhen (518.1 million yuan) and Huizhou (460.4 million yuan). The largest ESV loss of most cites in most areas was also concentrated from 2000–2010 (Figure 6d). In addition, Figure 7 shows the temporal and spatial changes of ESV of farmland caused by urbanization in 11 cities. From 1980–1990 and 2010–2018, the ESV losses of farmland because of urbanization in 11 cities were mostly less than 100 million yuan (green and dark green). However, from 1990–2000 and 2000–2010, the ESV losses of farmland because of urbanization in Guangzhou, Foshan and Dongguan were larger, especially in Guanghzou from 2000–2010, showed with red color. Overall, the ESV losses of farmland due to urbanization in Zhaoqing, Zhuhai, Macao and Hong Kong were relatively less than in other areas. *Land* **2021**, *10*, x FOR PEER REVIEW 9 of 18

**Figure 5.** Farmland loss due to urbanization in the GBA during the four study periods: (**a**) is the area of farmland converted to construction land in the GBA during the four study periods; (**b**) presents the percentage of the area of farmland converted to construction land in the total area of farmland loss in the GBA during the four study periods; (**c**) shows the variation among the losses of different farmland categories due to urbanization in the GBA during the four study periods; (**d**) is the variations in the area of farmland converted into different categories of construction land in the GBA during the four study periods; and (**e**) shows the area of farmland converted to construction land in the 11 cities during the four study periods. URL: urban residential land; RRL: rural residential land; OCL: other construction land. *3.3. The Impact of Urbanization on the Ecosystem Service Value of Farmland* **Figure 5.** Farmland loss due to urbanization in the GBA during the four study periods: (**a**) is the area of farmland converted to construction land in the GBA during the four study periods; (**b**) presents the percentage of the area of farmland converted to construction land in the total area of farmland loss in the GBA during the four study periods; (**c**) shows the variation among the losses of different farmland categories due to urbanization in the GBA during the four study periods; (**d**) is the variations in the area of farmland converted into different categories of construction land in the GBA during the four study periods; and (**e**) shows the area of farmland converted to construction land in the 11 cities during the four study periods. URL: urban residential land; RRL: rural residential land; OCL: other construction land.

An unprecedented expansion of construction land has led to a great loss of farmland,

decades, the ESV of farmland decreased by 5036.7 million yuan due to construction land encroachment, accounting for 43.1% of the total ESV loss of farmland (Table S1). The highest ESV loss of farmland because of urbanization was 2177.5 million yuan from 2000–2010, followed by 1655.1 million yuan from 1990–2000 (Figure 6a). For paddy fields and dry

**Figure 6.** The ecosystem service value (ESV) loss of farmland encroached by urbanization in the GBA during the four study periods: (**a**) is the ESV loss of farmland converted to construction land in the GBA during the four study periods; (**b**) is the ESV loss of different farmland types converted to construction land in the GBA during the four study periods; (**c**) shows the variation of ecosystem service losses of farmland due to urbanization in the GBA during the four study periods; and (**d**) is the ESV loss of farmland converted to construction land in the 11 cities during the four study periods. **Figure 6.** The ecosystem service value (ESV) loss of farmland encroached by urbanization in the GBA during the four study periods: (**a**) is the ESV loss of farmland converted to construction land in the GBA during the four study periods; (**b**) is the ESV loss of different farmland types converted to construction land in the GBA during the four study periods; (**c**) shows the variation of ecosystem service losses of farmland due to urbanization in the GBA during the four study periods; and (**d**) is the ESV loss of farmland converted to construction land in the 11 cities during the four study periods.

*Land* **2021**, *10*, x FOR PEER REVIEW 12 of 18

**Figure 7.** Spatial and temporal distribution of ESV loss of farmland because of urbanization in 11 cities in the four study periods. **Figure 7.** Spatial and temporal distribution of ESV loss of farmland because of urbanization in 11 cities in the four study periods.

#### **4. Discussion 4. Discussion**

#### *4.1. Farmland Changes Caused by Urbanization 4.1. Farmland Changes Caused by Urbanization*

The results of this study demonstrate that construction land in the GBA experienced dramatic changes in the overall extent, different types and regional extent, especially after 1990. Farmland is the land use type that contributed most to the expansion of construction land over the past four decades (Figure S1). Because of urbanization, 3711.3 km<sup>2</sup> of farmland disappeared in the GBA from 1980 to 2018. The highest farmland loss caused by urbanization was observed from 2000–2010 (1772.4 km<sup>2</sup> ), followed by that in the period from 1990–2000 (1020.1 km<sup>2</sup> ), but it was small from 1980–1990 (479.5 km<sup>2</sup> ) and from 2010–2018 (439.2 km<sup>2</sup> ). Our results are different from others showing the global conversion of farmland to other land use types; globally, farmland was mainly converted to grasslands and woodlands, which accounted for 57% and 36%, respectively [38]. These changes were strongly influenced by the social development background in China and local governments' behavior [39,40]. From 1980–1990, the first ten years after the implementation of the reform and opening-up policy in 1978, social and economic development was in the initial stage, production technologies lagged behind and infrastructure was relatively imperfect. All of these reasons contributed to a slower process of urbanization, which resulted in less farmland being converted to construction land. However, after ten years of development, the economy expanded to a certain extent, and the process of urbanization and industrialization also gradually accelerated, which led to farmland being intensively encroached on by construction land. Subsequently, China entered the World Trade Organization in 2001, which indicated that China's reform and opening up had entered a new stage in history. Rapid economic and scientific technology development impelled extensive urbanization, which caused the area of farmland loss to construction land to reach a peak from 2000–2010. As construction land expansion produced a great loss of farmland, a series of environmental problems appeared which threatened farmland preservation, food security, production capacity and social stability [41,42]. In addition, The results of this study demonstrate that construction land in the GBA experienced dramatic changes in the overall extent, different types and regional extent, especially after 1990. Farmland is the land use type that contributed most to the expansion of construction land over the past four decades (Figure S1). Because of urbanization, 3711.3 km<sup>2</sup> of farmland disappeared in the GBA from 1980 to 2018. The highest farmland loss caused by urbanization was observed from 2000–2010 (1772.4 km<sup>2</sup> ), followed by that in the period from 1990–2000 (1020.1 km<sup>2</sup> ), but it was small from 1980–1990 (479.5 km<sup>2</sup> ) and from 2010– 2018 (439.2 km<sup>2</sup> ). Our results are different from others showing the global conversion of farmland to other land use types; globally, farmland was mainly converted to grasslands and woodlands, which accounted for 57% and 36%, respectively [38]. These changes were strongly influenced by the social development background in China and local governments' behavior [39,40]. From 1980–1990, the first ten years after the implementation of the reform and opening-up policy in 1978, social and economic development was in the initial stage, production technologies lagged behind and infrastructure was relatively imperfect. All of these reasons contributed to a slower process of urbanization, which resulted in less farmland being converted to construction land. However, after ten years of development, the economy expanded to a certain extent, and the process of urbanization and industrialization also gradually accelerated, which led to farmland being intensively encroached on by construction land. Subsequently, China entered the World Trade Organization in 2001, which indicated that China's reform and opening up had entered a new stage in history. Rapid economic and scientific technology development impelled extensive urbanization, which caused the area of farmland loss to construction land to reach a peak from 2000–2010. As construction land expansion produced a great loss of farmland, a series of environmental problems appeared which threatened farmland preservation, food security, production capacity and social stability [41,42]. In addition, farmland conversion to construction

land could also lead to intensification of farming, and abandonment and degradation of farmland [43,44]. Governments try their best to maintain a balance between urbanization development and farmland preservation by implementing much stricter policies for farmland conversion and ecological urbanization and construction [10,14]. As early as 2005, the central land policy for the preservation of 1.8 billion mu (1.2 million km<sup>2</sup> ) farmland was established, which is a key environmental policy to cope with China's land transformation crisis since the 1990s [45]. In addition, to improve the urbanization quality and to strengthen the protection of farmland and natural ecosystems, the National Plan on New Urbanization and the Opinions on Accelerating the Construction of Ecological Civilization were promulgated in 2014 and 2015, respectively [39]. Under these policies, farmland loss due to urbanization sharply declined after 2010. In addition, local governments' behavior has vital impacts on the conversion of farmland to construction land [14,46,47]. Driven by particular interests, local governments converted farmland to urban land at a low compensation and leased to developers at a much higher price [48], which significantly affected farmland change under rapid urbanization [14]. To preserve agricultural land and food security, a centralized fiscal reform in 1994 induced local governments' land financing behavior to significantly influence farmland conversion [14]. Although the central government limits the total construction land quotas that local governments can lease to developers, local governments still have sufficient autonomy to determine which parcel of land to lease out [14], and they resorted to land leasing to gain extra revenue to finance urban construction and balance fiscal expenditures [49]. This extra local revenue from land leasing was approximately 33.7% of the local revenues from 2007 to 2012, which mainly came from farmland conversion [50,51]. Lastly, the urbanization in Guangzhou, Dongguan, Foshan, Shenzhen and Huizhou was faster than in other cities over the past four decades (Figure 2); correspondingly, the area of farmland conversion to construction land was greater in these cities than in the other cities (Figure 5). The same is observed in other developing counties, like India, where farmland conversion to construction land predominantly occurred in the districts with high rates of economic growth and higher agricultural land suitability [22].

#### *4.2. Ecosystem Service Value Changes due to Urbanization*

Extensive urbanization has resulted in the degradation of farmland, which poses a great threat to food provision security, ecological environmental protection, ESs and regional sustainability [52–55]. In the past four decades, the ESV loss of farmland caused by urbanization was 5036.7 million yuan in the GBA, and the highest loss was 2177.5 million yuan from 2000–2010, followed by 1655.1 million yuan from 1990–2000. The change trend of the ESV loss of farmland is similar to the change trend of farmland lost because of urbanization, which is because the calculation of the ESV is based on area. In addition, our results indicate that ESs such as hydrological regulation, food production and gas regulation are particularly vulnerable to urbanization impacts (Figure 6c). As construction land expands, a large area of farmland is converted to impermeable surfaces, which causes rain and runoff to not penetrate the ground in time and participate in the natural water cycle, which further heavily affects hydrological regulation [56]. For food production, urbanization has multiple effects. Rapid urbanization in the GBA has led to an increase in the agricultural land use intensity, cropping frequency and chemical fertilizer use, a decline in the per capita availability of food grains, and soil and water pollution [10,22,57,58]. In this complex context, both food production and security are seriously influenced. The process of urbanization and industrialization inevitably causes environmental problems such as air pollution [59], which predominantly originates from industry, transport, power generation and construction [60]. Air pollution could lead to air quality deterioration and thus influence regional gas regulation [59,60].

#### *4.3. Farmland Conservation and Ecosystem Services Protection*

Governments have been striving to maintain a balance between urbanization development and farmland preservation [10,14], and the results of this study show that farmland loss due to urbanization sharply declined after 2010, but the results also indicate that construction land expansion has increasingly become the dominant reason for the loss of farmland in the GBA over the past 40 years. Therefore, it is still vital for governments to take more effective steps to regulate construction land expansion and to develop trade-offs and synergies among urban development, agricultural production and ecosystem preservation [14,39]. Based on the results of this study, we provide several recommendations for farmland preservation. First, to protect farmland and improve regional ESs, central and local governments should strictly control the area of new construction land and strengthen the construction of ecological cities and beautify countryside. In urban areas, the government could protect, plan and establish natural parks or green belts, including forests, grasses and wetlands. In rural areas, the government should promote reasonable planning for farmland use or exploitation. Governments can also 'green' abandoned industrial and mining land with forests, grasses or water bodies to improve ESs. Second, governments must strictly protect the red line of high-quality farmland and prohibit exploitation without rational reasons. Third, local governments should maintain the requisition–compensation balance of farmland, including the quantity, quality and ecological balance. For example, if the quality of compensatory farmland is lower than the quality of requisitioned farmland, then governments should invest more money to improve the quality of medium- and low-yield farmland or exploit new farmland rather than invest in urban construction [41]. Fourth, local governments should take measures to prevent environmental pollution, such as educating farmers to strictly control the use of agrochemicals and other soil additives to prevent soil pollution, and strictly prohibit the direct discharge of domestic sewage and industrial wastewater into rivers without treatment. Finally, the public should respond to government policies and protect the environment. For example, to protect the water and soil environment of farmland, enterprises and factories near farmland should strictly treat sewage according to regulations, and discharge the sewage only after it reaches the required standard. Farmers could reduce the use of pesticides and other chemicals to reduce the pollution of the farmland environment.

#### *4.4. Limitations and Future Works*

Estimates of ESV regionally and globally in monetary units play a critical role in heightening awareness and estimating the overall level of importance of ESs relative to and in combination with other contributors to sustainable human well-being [37]. Therefore, it is better for decision-makers and the public to consider the ESs as public goods or natural resources, and take these values into account when scenarios and policies are changed. However, this valuation method could not be used to examine the spatial changes in ESV. In addition, the ESV of construction land in this study was considered as zero, which is not appropriate in some degree and probably led to errors in the results. In our future work, we would like to combine the valuation methods and other evaluation methods to acquire more accurate results.

Farmland conservation and ecosystem services protection require cooperation in many aspects, including from governments, experts and the public. Therefore, it is necessary to build a system integrating geographic information system (GIS), spatial multicriteria evaluation (SME) and participatory GIS (PGIS) approaches, where decision-makers, experts and the public can participate and identify a range of ecosystem services [61–64]. Decision maker can easily collect and manage the results, which is useful for land use planning. That would be conducted in our future work.

#### **5. Conclusions**

This study explored the impacts of urbanization on farmland in the GBA from 1980 to 2018 based on multiple temporal land use datasets of the CNLUCC database. The ESV of farmland was also estimated by the valuation methods to study the ecosystem service changes caused by urbanization. Our results showed that the total area of farmland loss caused by urbanization was 3711.3 km<sup>2</sup> over the past 38 years, leading to a direct decline in total ESVs by 5036.7 million yuan. Paddy fields suffered much more losses than dry farmland because of urbanization. The expansion of construction land increasingly became the dominant reason for farmland loss in the GBA with the influence increasing from 39.3% to 96.6%. The value of hydrological regulation, food production and gas regulation showed much more decline. Guangzhou, Dongguan, Foshan, Shenzhen and Huizhou had the greatest total ESV loss of farmland caused by urbanization. The social development background in China and local governments' behavior played a vital role in the farmland conversion to construction land. To protect farmland and improve regional ESs, the central and local governments should strengthen the construction of ecological cities and beautify countryside by increasing capital investment, strengthening supervision, and raising public awareness of environmental protection, and the public should respond to land use policies and protect the environment.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/land10050501/s1, Figure S1: Conversion percentages of the different land use types into construction land in the GBA from 1980 to 2018, Table S1: ESV changes of the farmland in the GBA from 1980 to 2018 (million yuan).

**Author Contributions:** Conceptualization, X.W., F.Y. and F.S.; methodology, X.W. and F.Y.; formal analysis, X.W., M.C., B.H. and L.K.; writing—original draft preparation, X.W., Y.Z. and M.C.; writing review and editing, X.W., F.Y., Y.Z. and F.S.; funding acquisition, X.W. and F.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by the National Natural Science Foundation of China (41890854), the Program B for Outstanding Ph.D. Candidate of Nanjing University (No. 202002B087), and the China Scholarship Council (CSC) (No. 201906190120).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** We appreciate critical and constructive comments and suggestions from the reviewers that helped improve the quality of this manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Applying the Evaluation of Cultural Ecosystem Services in Landscape Architecture Design: Challenges and Opportunities**

**Xin Cheng 1,\* , Sylvie Van Damme <sup>2</sup> and Pieter Uyttenhove <sup>3</sup>**


**Abstract:** Landscape architects play a significant role in safeguarding urban landscapes and human well-being by means of design and they call for practical knowledge, skills, and methods to address increasing environmental pressure. Cultural ecosystem services (CES) are recognized as highly related to landscape architecture (LA) studies, and the outcomes of CES evaluations have the potential to support LA practice. However, few efforts have focused on systematically investigating CES in LA studies. Additionally, how CES evaluations are performed in LA studies is rarely researched. This study aims to identify the challenges and provide recommendations for applying CES evaluations to LA practice, focusing specifically on LA design. To conclude, three challenges are identified, namely a lack of consistent concepts (conceptual challenge); a lack of CES evaluation methods to inform designs (methodological challenge); and practical issues of transferring CES evaluations to LA design (practical challenge). Based on our findings, we highlight using CES as a common term to refer to socio-cultural values and encourage more CES evaluation methods to be developed and tested for LA design. In addition, we encourage more studies to explore the links of CES and landscape features and address other practical issues to better transfer CES evaluations onto LA designs.

**Keywords:** landscape architecture; cultural ecosystem services; design; evaluation

#### **1. Introduction**

Cities experience increasing environmental stress given the sharp rise of city populations, and the consequent loss and degradation of urban green spaces. Landscape architects therefore play a significant role in maintaining the human environment and well-being by planning, design, and management of urban landscapes more than ever [1]. The Council of Europe [2] defines landscape as an area that is perceived by people, and the character is the result of the action and interaction of natural or human factors. Landscape architecture (LA), as a profession, provides site planning, design, and management advice to improve the character, quality, and experience of the landscape [3]. Design is the core activity of the LA practice, considering many factors, such as the landscape itself, the intentions of the clients, the interaction of the users and the design setting, the materials, and the processes of creative expression. To do this effectively, landscape architects apply solid design processes and implementation methods, elaborated on and explained in an increasing amount of studies and publications on those topics. For example, Langley et al. [4] and Yue and Shao [5] summarize the core knowledge domains of LA, including the evolution of analytical methods for LA planning and design. Milburn and Brown [6] and Lenzholzer et al. [7] focus on incorporating research into the LA design process. Some scholars focus on new tools, such as Li and Milburn [8] and Gu et al. [9], who work on geodesign as a design tool.

LA primarily aims to organize the complexity of the landscape into comprehensible, productive, and beautiful places to improve the function, health, and experience of life. However, landscape architects increasingly face new challenges within today's rapid

**Citation:** Cheng, X.; Van Damme, S.; Uyttenhove, P. Applying the Evaluation of Cultural Ecosystem Services in Landscape Architecture Design: Challenges and Opportunities. *Land* **2021**, *10*, 665. https://doi.org/10.3390/land10070665

Academic Editors: Alessio Russo and Giuseppe T. Cirella

Received: 18 May 2021 Accepted: 23 June 2021 Published: 24 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

urbanization, such as designing resilient landscapes for adjusting to a changing climate and controlling natural disasters, creating a sense of place, and managing the growth of informal settlement [10,11]. Meanwhile, clients and the general public are no longer only interested in the aesthetic value of landscape but also increasingly concerned about ecological functions and environment conservation [12]. To achieve these multiple, oftencompeting objectives, landscape architects need a clear understanding of the relationships between humans and the environment and the interaction processes on how humans shape the landscape they rely on and experience. Hence, there is an urgent call for practical knowledge, skills, and methods to address these challenges within the LA profession [12].

The concept of ecosystem services (ES) refers to the benefits that people obtain from ecosystems [13]. They highlight the contributions that environments and landscapes provide for society and the economy and, more generally, for human well-being. They are divided into provisioning services, regulating services, supporting services, and cultural services [13]. In the last decade, several researchers conducted studies on integrating ES evaluation into spatial planning policies such as land use/cover to foster sustainable development. More recently, the relationship between ES and urban landscapes has also been increasingly investigated. The outcomes of ES evaluations are recognized to have the potential to support practical application and decision-making [14–19]. Among ES, cultural ecosystem services (CES), namely the non-material benefits that people get from ecosystems, are regarded as directly influencing human-wellbeing and having the potential to motivate people's willingness to protect urban greens [14,20]. Whether or not people are familiar with the term, the concept resonates with nearly every human being, such as the experience of recreational activities and scenery appreciations. Although different definitions have been developed so far, CES are generally defined as the non-material benefits people obtain from ecosystems [13], which was adopted by this study. CES evaluation is one of the biggest challenges for further applying in practice. Evaluating what people obtain is a concise way of disseminating the importance of ecosystems or landscapes [14]. Various evaluation methods have been developed and they can generally be classified into monetary methods and non-monetary methods, depending on whether the evaluation outcomes are expressed by money or not [14]. The classification can be further divided into revealed preference and stated preference methods. The revealed preference method means observing the actual markets or human behaviors related to the CES to assess CES. The stated preference method means directly asking about one's values to assess CES [14]. CES evaluation studies and LA studies share many commonalities. First, they focus on the "human dimension" of landscapes [21]. For example, LA has long been concerned with questions of the human perception of and involvement with nature. Meanwhile, CES evaluations heavily depend on people's perceptions and preferences. Second, both have a broad knowledge about the assessment of aesthetic values, place identity, and cultural heritage on-site and in specific contexts [21]. Landscape architects may not be familiar with ES or CES; however, they are familiar with concepts such as aesthetics or heritage [22]. Third, both highlight the local scale of different landscape types, especially while studying urban green spaces like urban parks, gardens, and forests, and, fourth, both are concerned with the benefits provided by landscape features that can enhance the quality of the landscapes and the entire system [23]. For example, parks are recognized as providing a range of CES, and, in many areas, these relate to scientifically-assessed landscape features, such as the proportion of vegetation [24]. Similarly, De Valck et al. [25] test the different degrees of high or low vegetation that indicate outdoor recreation.

CES are highly related to LA studies, and the outcomes of CES evaluations have the potential to guide LA practices. Although they share many commonalities, few efforts have focused on systematically investigating CES in LA studies. Additionally, how CES evaluations are performed in LA studies is rarely investigated. This paper aims to identify the challenges and opportunities of integrating CES evaluations into LA designs. Specifically, it discusses the problems and challenges of existing studies, the examined CES categories, the

implications and applications of evaluation methods, the links between CES and landscape features, and opportunities for future study.

With this in mind, we specifically ask which CES are studied, what evaluation methods are used, and how CES are connected to landscape features in LA studies. To achieve this, we focus on three areas:


Combining these three methods provided a comprehensive knowledge of CES evaluation in LA fields. In addition, this study focused on urban parks in China. As an important component of urban green spaces, urban parks are crucial for securing human well-being by providing diverse benefits, and they have been studied by ES and LA fields. Urban parks are important indicators used to estimate the quality of life in a city. Moreover, studying urban parks is vitally important for their design, planning, and management, especially in the case of China, which is experiencing increasing social and environmental pressure caused by the rapid population growth in cities. Hence, it is an urgent call for valuable tools to aid urban park design.

#### **2. Methods**

We combined three methods to investigate the challenges and opportunities of applying CES evaluations in LA studies. Table 1 shows the study process. Specifically, we first reviewed existing scientific articles to have a general overview of the research on urban parks and identify items related to CES and the links between these terms. Then, we conducted a systematic review of selected park proposals based on a guideline (Table A1 in Appendix A). Subsequently, we interviewed landscape architects directly to know their thoughts about CES evaluations and other practical issues (Table A2).


**Table 1.** Overview of the study process.

CNKI is a key national information and knowledge database that includes journals, doctoral dissertations, master's theses, proceedings, newspapers, yearbooks, statistical yearbooks, patents, standards, etc. in China.

#### *2.1. Review of Keywords and Abstract of Scientific Publications*

To review the existing publications about CES within LA studies, we used key terms and phrases based on reviewing the abstracts and keywords of scientific publications. This technique can be used to quantify a large sample [26,27]. We first searched "landscape

architecture design\*" and "urban park\*" in the China National Knowledge Infrastructure (CNKI) database, including journals, dissertations, and conference papers. We set the timespan up to November 2019, which resulted in 5273 items. Then, we screened the key terms using CiteSpace.5.5.R2 (parameter settings: Years Per Slice: 2; Node Types: Keyword). The frequency of the keywords could reflect hot topics and core content over a set period [28]. Subsequently, we counted the occurrence frequency of key terms in relation to CES. There are different definitions and classifications of CES, and different authors use different terms based on their study focus. Here, we recorded the terms that the authors reported. Finally, other terms were also presented and counted if they were linked to CES terms.

#### *2.2. Review of Design Proposals*

A systematic review was conducted to review the design proposals of urban parks in China. More specifically, we first collected park proposals from LA companies, government agents, and internet searches. A balance of park types and scales were also noticed, and, finally obtained, 83 items were obtained. Second, we asked a set of questions when we reviewed each park proposal and recorded the answers (Table A1), including the time, park scales, park types, CES types, evaluation methods, and the relationships with landscape features. More specifically, we set the classification of the answers of identified proposals as shown in Table 2. There are several different classification mechanisms of urban parks, and it is beyond the scope of this study to give a comprehensive account of the various classifications. In this study, we based on the Standard for Classification of Urban Green Space of China.


**Table 2.** Review of design proposals.

#### *2.3. Interviews of Landscape Architects*

To reveal more detailed information from designers, 12 semi-structured phone interviews were conducted between November, 2019 and January, 2020. The interviewees were landscape architects with rich experience in park design in China and consisted of three groups. The first group consisted of designers in the government's LA design institute. The second group was made up of professors working at the university who were also

doing LA projects, and the third group contained landscape architects from LA design companies. The interview started with a brief description of this study. Subsequently, a set of questions was asked based on guidelines that directed towards our research questions (Table A2). The interviewees were prompted with a talk-generating question, and the structure of the interview was adjusted to their statements. Each interview was conducted over 30 to 90 min, depending on the interviewee's interests and schedule. In addition, the results were qualitatively analyzed according to the research goals. For the presentation of our results, quotations were translated from Chinese to English. The interview texts were summarized when interviewees shared the same ideas. The individual ideas were recorded independently.

#### **3. Results**

#### *3.1. CES in Scientific Publications*

This section presents the high-frequency keywords assigned by the CNKI database. A minimum of two occurrences of each keyword was included in the network, resulting in 208 keywords and 1266 links. Notably, we merged similar terms, because the original language was Chinese, and some terms share the same meaning, and the final results contained 30 keywords (Table 3). Specifically, the results revealed that regional/local culture (frequency = 212), humanization/user-friendly (85), and cultural characteristic (70) were mentioned far more than other keywords. Other high-frequency keywords included historical culture (16), urban culture (9), theme culture (8), leisure and recreation (8), and place spirit (5). All of these keywords served as a reference point for finding and understanding the possibilities of applying CES in the landscape design of urban parks.

**Table 3.** List of keywords in relation to CES.


In addition, the intensity of the link between two keywords was computed based on the number of times they were mentioned together. Sixty-three keywords were found linked to the CES keywords (Table A3). The main connecting keywords were landscape architecture design (23), landscape architecture (23), urban park (18), theme park (11), and ecology (8). The first three keywords were expected as they were our search terms. Other highly linked keywords included design (7), plant design (7), local culture (7), urban culture (6), historical culture (6), waterfront landscape (6), landscape renovation (5), humanization/user-friendly (5), urban wetland park (5), and comprehensive park (5). Other items, such as sculpture, environmental infrastructure, and plant community, were related to landscape features. There were surprisingly no terms directly about evaluation or evaluation methods.

#### *3.2. CES in Park Proposal*

The sizes of the 83 reviewed parks ranged from 2.5 ha to 1500 ha between 2003 and 2019. The parks included 31 comprehensive parks, 37 topic parks, 7 community parks, and 8 belt-shaped parks (Figure 1). Topic parks included the wetland park (6), the sports park (6), the expo park (5), the children's park (3), the water park (3), the amusement park (2), the botanical garden (2), the shopping park (2), the ecological park (2), the cultural

*Land* **2021**, *10*, x FOR PEER REVIEW 6 of 15

*3.2. CES in Park Proposal*

park(2), the lake park (1), the cultural relics park (1), the commemorative park (1), and the zoological garden (1). (2), the botanical garden (2), the shopping park (2), the ecological park (2), the cultural park(2), the lake park (1), the cultural relics park (1), the commemorative park (1), and the zoological garden (1).

The sizes of the 83 reviewed parks ranged from 2.5 ha to 1500 ha between 2003 and 2019. The parks included 31 comprehensive parks, 37 topic parks, 7 community parks, and 8 belt-shaped parks (Figure 1). Topic parks included the wetland park (6), the sports park (6), the expo park (5), the children's park (3), the water park (3), the amusement park

Leisure and recreation 8 Culture heritage 2 Place spirit 5 Social life 2 Ecological aesthetic 4 Tradition 2

**Figure 1.** The types of the park.

**Figure 1.** The types of the park.

The concept of CES was not found in park proposals, but 50 terms were identified in relation to CES (Table 4). The most mentioned types were recreation (73), education (26), and aesthetic (13), followed by traditional culture (12), sense of place (11), social interaction (10), history (8), Shan-Shui culture (8), sense of nature (7), art (7), tourism (6), and experience (6). Recreation included walking, running, boating, fishing, biking, bird The concept of CES was not found in park proposals, but 50 terms were identified in relation to CES (Table 4). The most mentioned types were recreation (73), education (26), and aesthetic (13), followed by traditional culture (12), sense of place (11), social interaction (10), history (8), Shan-Shui culture (8), sense of nature (7), art (7), tourism (6), and experience (6). Recreation included walking, running, boating, fishing, biking, bird watching, tea dinking, flower watching, farming, bird watching, fishing, skating, etc.

watching, tea dinking, flower watching, farming, bird watching, fishing, skating, etc. **Table 4.** CES categories in park proposals. **CES Category Number CES Category Number** Ethnic culture 1 Spiritual 4 The majority of proposals (74) mentioned that they used a document research approach through analyzing materials such as landform, existing information and data, and master plan. Among them, 56 proposals analyzed successful case studies in China and other countries. Twelve proposals stated that they conducted field surveys, and four used questionnaires to ask people how they perceived the existing parks and what they liked.

Education 26 knowledge 1 Recreation 73 Tourism 6 Spiritual and religious value 3 Shopping experience 1 Cultural diversity 3 Identity 3 Traditional culture 12 Respect of nature 1 Landscape features were wildly used to link CES to designs, and 26 features were found in proposals (Table A4). Plants (41) were the most mentioned feature in park proposals, followed by architecture (28), playground (26), sculpture (21), and sports-/fitness facilities (20). Other features, such as signs (19), furniture (12), recreation facilities (10), path (9), water (8), and landform (7), were also mentioned often.

#### History 8 Love 2 *3.3. CES in the View of Landscape Architects*

Three interviewees who were also involved in landscape education in a university stated that they knew the concept of CES, but they had not used it in their designs. The other interviewees stated this was the first time to hear about this concept. However, the landscape architects were familiar with related terms, such as socio-cultural value. They stated that it is essential that parks are designed for recreation, aesthetics, and education purposes. They agreed on the importance of socio-cultural values and non-material benefits in LA designs. However, they also stated that it is difficult to be aware of them and capture them because of their abstract and intangible characteristics, which often depend on designer's experience, values, and preferences.


**Table 4.** CES categories in park proposals.

On the one hand, landscape architects recognized the huge potential of CES evaluations to support their practices. They claimed that it may help them capture cultural benefits and better understand human and environmental processes and further integrate them into the ES framework to achieve multifunctional landscapes, not single-function landscapes. Designers stated that it is difficult to find an overarching value that would allow aesthetic, social, economic, and environmental values to be measured against one another [34]. Moreover, it should be a useful tool to communicate with clients and stakeholders by showing them potential benefits. One designer stated, "It's very important for communicating with clients who make the final choice. Therefore, you're actually selling your ideas to them and somehow persuade them to trust your designs. So more supporting proofs are needed to strengthen your design plans."

However, designers also stated that there are still many issues to be addressed. The biggest challenges are the methodological and practical issues. For example, the interviewees asked series of questions: "when do I need to evaluate CES and in which stages (before design, during design, or after design)?" "What methods do I need, and how?" "What kind of expertise is needed for this?" "Do I need CES experts to aid me?" "Who is involved in this evaluation, and how can it be facilitated?" "What is the difference when dealing with different parks of different scales? For example, the design of children parks is totally different from the zoological garden." "Does it influence the designer's creative expression or artistry?"

When asked how they transfer the CES to their designs, designers stated that recreation, aesthetics, or educational values are easier to express. For example, recreational values are often expressed by creating spaces for recreation and equipping them with recreational facilities. As for the other more abstract CES, designers stated that they are often inspired by history, a story, a culture, or a celebrity, and they often use physical landscape features to create the cultural atmosphere. For example, Shan-Shui culture was highlighted by interviewees; one stated, "Shan-Shui refers to mountains and rivers and broadly refers to nature in general, which is an important source of inspiration for creating

places." Another example is the use of plants: "Bamboo, for example, is often used in designs to emphasize the elegance of the environment because, in Chinese culture, it is a symbol representing the character of moral integrity, modesty, and loyalty." Sculptures are also widely used to commemorate a celebrity or historical event or person. Designers also stated that the process is complex and depends on sites and specific contexts.

#### **4. Discussion: Challenges and Opportunities**

This study focuses on which CES are studied, what evaluation methods are used, and how the concepts connected to physical landscape featured in LA studies. In this section, we highlight three key challenges based on our findings and further discuss opportunities for landscape architects to contribute to apply CES evaluations into LA studies in the future.

#### *4.1. Conceptual Challenges: A Lack of Consistent Concepts*

Although the term CES was not been used, various socio-cultural values were found in the LA publications and design proposals. Local culture and historical culture were widely highlighted, benefiting from the fruitful history and culture of China. Landscape architects did not use the term CES because the concept of CES is relatively new, proposed in recent decades. Chinese scholars seemed to be unaware of this progress until recent years, and the number of studies focusing particularly on CES is still very limited [35]. In addition, some interviewees stated that they are unaware of the latest research trends or methods because they do not read international research publications as their English is poor. Another landscape architect stated that both CES and other terms he often uses such as historical culture or Shan-Shui culture all refer to the social-cultural values, so it does not matter whether he uses the term CES or not. The landscape architects also highlighted that the key is how to evaluate them and use them to guide the LA design practice.

Although CES have a range of definitions based on diverging views, leading to alternative classification schemes, they are generally described as socio-cultural values or non-material benefits. There are several mainstream classifications of CES in ES studies, such as the Millennium Ecosystem Assessment classification, Economics of Ecosystems and Biodiversity, Common International Classification of Ecosystem Services, Final Ecosystem Goods and Services Classification System, and, more recently, the Intergovernmental Platform on Biodiversity and Ecosystem Services. Diverse classifications were developed to clarify CES, and they are still in progress [36,37]. However, a consistent concept is necessary as this gives an opportunity to organize a dialogue and cooperation between LA and CES studies [21]. CES have the potential to be a common term referring to sociocultural benefits. The results show that many CES are referenced and studied in LA research and design proposals but not by explicitly using the term CES. Hence, an opportunity is missed by not highlighting the link between the environment and humans that implicitly exists within the research and proposals. Given that this anthropocentric rationale can be effective at generating support for environmental conservation [38], the explicit inclusion of the CES concept may also lead to better ecosystem protection in general [39]. Moreover, practitioners who lack basic knowledge about CES concepts may be less aware of what could benefit from their designs. Hence, the inclusion of a common term, concept, or definition for CES in the LA studies and design proposals and other similar documents is needed to promote LA practice. Furthermore, the existing CES evaluation studies by many other countries and their various methods and indicators can inspire LA studies in China. However, a concern about the use of CES evaluations by designers is that the application of CES as a scientific concept might cause a loss of creativity and artistic thinking. It is a challenge for designers to balance the evaluation and creativity, which requires designers to have a full understanding of the knowledge of CES and LA design.

#### *4.2. Methodological Challenges: Lack of Evaluation Methods to Inform Designs*

Most design proposals use document research, followed by questionnaires which are based on people's preferences and perceptions. In reality, expert-based methods are the

most used, even when this is not mentioned in the proposals. Indeed, the designs highly depend on the experiences, skills, and values of the designers as experts, and evidencebased design is widely used for LA designs in China. An evidence-based design highlights the process of design by critically and appropriately integrating various aspects, such as credible data, practitioner design expertise, client needs, and resources, in order to achieve project objectives [40]. Although evidence-based design has been widely used, it has been simultaneously critiqued for rigidity and misapplication in China. Hence, landscape architects gravitate to the support of research by integrating "research-informed design" as a broader term with concepts, fields, tools, or methods to support their practice. In this study, CES is regarded as a research tool that has the potential to aid LA practices. It is clear that, although CES are difficult to evaluate, a systematic evaluation of CES could guide the designers to capture and maximize them.

The diversity of CES study methods provides a rich inventory for scholars to assess CES. Instruments for assessing CES, including quantification, valuing, mapping, and modeling, are increasingly studied in CES and ES research. Many evaluation methods provide an opportunity to evaluate CES in LA studies. The evaluation methods are generally divided into monetary and non-monetary methods or revealed preference and stated preference methods as introduced in the first section. Cheng et al. [14]; Hirons et al. [41]; and Spangenberg and Settele [42] summarize the evaluation methods. In urban green space studies, non-monetary methods were used more, especially questionnaires and interviews, which emphasize the preferences and perceptions of people. It highlights the importance of the role that humans play in interaction with landscapes. Participatory mapping can be used to evaluate the crucial locations and settings for an urban park. Such assessments can provide valuable information for designers to increase CES provisions in the urban landscape. A social media-based method was used to reveal people's preference of CES based on the social media data from various resources such as Flickr or Instagram. Monetary methods, such as willingness to pay (WTP), can be used to know people's preferences for specific park settings. Both monetary and non-monetary methods and their combinations are encouraged to be used in LA studies. To achieve this, combing economics with other disciplines such as social or behavioral sciences is significant to know how to shift human aspects to environment setting, which is encouraged in future studies.

#### *4.3. Practical Challenges: Practical Issues of Transferring CES to LA Design*

In addition to the conceptual and methodological challenges, there are also more practical issues. They are: the relationships between CES and landscape features, the evaluation scopes (i.e., park types and size), data collection, and operation.

#### 4.3.1. CES and Landscape Features

CES are highly related to landscape features. Sculpture, environmental infrastructure, and plant community were identified in the publications review. In addition, plants, architecture, playground, sculpture, and sports-/fitness facilities were the features used most to indicate CES in the reviewed design proposals. Landscape architects also highlighted the importance of the landscape features to express their ideas about CES. Linking the CES to physical landscape features is regarded as an efficient way to ground the abstract concept into the designs. The link between CES and landscape features provided an opportunity to help designers easily grasp CES and create essential links between ecosystem processes and management actions [43]. For example, some studies have tried to link CES to a location by means of participatory mapping, which allows people to mark the sites of CES (see Brown and Hausner [44]; Bryan et al. [45]; and Plieninger et al. [46]). Such studies provide designers with insight into the location of CES supplies and their correlations with specific features [30]. De Valck et al. [25] test the influence of different degrees of high or low vegetation on recreation and how such specific connections have the potential to support the vegetation design and landscape management. Benches are also regarded as an indicator of aesthetic values that can guide LA designs [47]. However, the number

of studies linking CES and landscape features are still limited, and future studies should focus more on the links between CES and landscape features to guide LA designs.

#### 4.3.2. Scopes, Data Collection, and Implementation

Many other factors influence the transfer of the concept of CES into LA design. The design process is complicated, and landscape architects need to coordinate a series of tasks to be performed by a number of people over a set period. They need to consider factors such as scale, types, layout, and features and organize the landscape to create a functional, comprehensible, and beautiful place. In addition, landscape architects need to consider the data collection and operation process when integrating CES evaluation into this process.

Specifically, the first issue is the evaluation scope including park scales and types. Parks with different types and scales have different expressions. For instance, sports parks emphasize recreation more than commemorative parks, which highlight spiritual values as stated by designers, and, in this case, landscape architects often have to trade off different CES. Meanwhile, different parks with different sizes add complexity. Hence, defining types and scales is crucial at the beginning to clarify the study scope and further select evaluation methods.

The second issue is how to collect the data. Data collection is significant to support practices, and depends on what methods are selected. The data can be divided into primary data or secondary data and quantitative data or qualitative data. For example, LA designs often depend on the secondary data derived from the document research, and primary data derived from the interviews. The primary data often engage stakeholders because CES evaluation is recognized to rely heavily on people's preferences and perceptions. LA is a user-inspired and user-useful discipline that requires designers to achieve the requirements of the clients and collaborate with them, making it more complex. CES evaluation could contribute to the increased congruence between different stakeholders. It is a useful tool for communicating with clients by showing what benefits they can have based on design proposals.

The third issue is the operation of the CES evaluation, such as who is involved in the evaluation and how it can be implemented. There is often a lack of local research capacity to undertake valuation research [48], which requires designers to have the capacity to act as enumerators or facilitators. As most designers do not know about CES, training on how to evaluate CES is essential for designers. Additionally, the development of evaluation methods, toolboxes, and practical guidelines is important to aid designers. Meanwhile, cooperation with ecologists or other experts is also highly encouraged at the beginning.

We take account of the complexity of practical issues and emphasize that there is not one simple and straightforward way to address practical issues. This study presented assumptions and discussions for tackling the complexity of involving CES evaluation in landscape design.

#### **5. Conclusions**

Landscape architects are facing increasing pressure due to rapid urbanization and call for practical knowledge, skills, and methods, etc. to support their designs. This study proposes that the concept of CES could have the potential to address their pressures by integrating CES evaluation in LA designs. This study identified three challenges, including consistent concepts, methods for evaluating CES, and practical issues of transferring CES to LA design. We further provided recommendations about how to deal with these challenges by highlighting opportunities for designers to contribute to LA and CES research in the future. (1) We highlighted developing consistent concepts and highlighted using CES as a common term to refer to socio-cultural values. (2) We encouraged using more evaluation methods to assess CES in LA studies, including monetary and non-monetary methods, such as WTP, participatory mapping, and the social media-based method. In addition, (3) we encouraged more studies addressing various practical issues to better guide LA designs. The first issue is to define park types and scales, which is crucial at the beginning

to clarify the study scope and further select evaluation methods. The second issue is to collect data, especially the primary data that are often ignored in LA studies. The third issue is to develop evaluation methods, toolboxes, and practical guidelines to aid designers. In addition, training on how to evaluate CES is also essential for designers.

**Author Contributions:** X.C.: Conceptualization, Methodology, Investigation, Formal analysis, Writing—Original draft preparation. S.V.D.: Methodology, Writing—Reviewing and Editing. P.U.: Writing—Reviewing and Editing. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by The Talent Introduction Program of Xihua University.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are contained within this article.

**Conflicts of Interest:** We declare that we have no conflicts of interest to this work. We have no financial and personal relationships with other people or organizations that can inappropriately influence our work.

#### **Appendix A. Supplementary Data**

This appendix contains four tables.

**Table A1.** Set of questions asked for every design proposal reviewed.




The interview begins with an informed consent about the recording and an explanation about the confidentiality of the interview. Following is a rough and easy to understand description of the aim of our study.



**Table A4.** Landscape feature categories in park proposals.


#### **References**

