Urban land use change and its effect on social metabolism: An empirical study in Shanghai

Habitat International 49 (2015) 251e259

Contents lists available at ScienceDirect

Habitat International journal homepage: www.elsevier.com/locate/habitatint

Urban land use change and its effect on social metabolism: An empirical study in Shanghai Xuezhu Cui*, Xuetong Wang School of Business Management, Guanghzou University, Guangzhou 510000, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 16 November 2014 Received in revised form 23 April 2015 Accepted 21 May 2015 Available online xxx

Numerous breakthroughs in multiple areas concerning land use change have been introduced in recent years. This phenomenon indicates that people have begun to realize the high environmental costs of their actions and are beginning to address this situation. From the perspective of environmental economics, this research focused on urban land use and explored the relationship between land use change and social metabolism flow through canonical correlation analysis (CCA). An empirical study was conducted in Shanghai to prove that urban land management could be seen as a means of balancing the social metabolism flow. First, for urban land use quantity, increasing the warehousing and traffic land areas could significantly affect the metabolism amount; meanwhile, the cultivated land and urban green land had crucial roles in controlling metabolism amount. Second, for urban land use efficiency, the intensity and efficiency of social metabolism could be influenced by land use efficiency, which indicated that a highly efficient industrial estate and storage land use could decrease the intensity of material consumption. These findings provide the government with new ideas and methods for urban planning and land management. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Shanghai Land use change Metabolism response Canonical correlation analysis Land use policy

1. Introduction Rapid urbanization changes the land use structure along with a continuous “colonization” to nature, which expands the urban scope and diminishes rural areas (Lambin et al., 2011). The unlimited demand for limited land areas has spurred serious urbanization-related problems, such as urban sprawl, food security, and ecological damage (Wang, Shen, Tang, & Martin, 2013). The increasing urban population and social economic development in developing countries have also increased the need for housing and infrastructure, which may be satisfied at the expense of losing agriculture land and consuming natural resources and energy (Giuseppina, 2012). Urbanization also produces a large amount of wastes that damage the ecological environment and threaten the ecological system (Stephan & Friedrich, 2000). The amount of these wastes all over the world has rapidly increased in recent years and will continue to increase in the coming decades, particularly for developing countries such as China (United Nations, 2008). Therefore, if this land conversion trend continues, its negative externalities will seriously affect the sustainable development of humans (Hall, Perez, & Leclerc, 2000).

* Corresponding author. http://dx.doi.org/10.1016/j.habitatint.2015.05.018 0197-3975/© 2015 Elsevier Ltd. All rights reserved.

The International GeosphereeBiosphere Program (IGBP) and International Human Dimensions Programme on Global Environmental Change (IHDP) showed that the extant studies on global environmental change were primarily focused on land use/land cover change and served as mediums through which humans could address global change (IGBP & IHDP, 1999). Land provides people with the essential materials, such as biomass, thus increasing the dependence of humans on land area. Human activities, such as material and energy transportation, storage, conversion, consumption, and waste treatment, are all linked with land use (Giuseppina, 2012). Previous studies have identified the negative effects of land use change on the world (Foley et al., 2005), environment (Alberti et al., 2007; Mark et al., 2012), ecosystem (Polasky, Nelson, Pennington & Johnson, 2011), climate (Chen, Zhao, Li, & Yin, 2006; Seto & Shepherd, 2009), and human health (Patz & Olson, 2008; Xu, Rita, Jing, & Xu, 2008). These studies indicate that most of the negative effects of land use change on the ecosystem and the environment can be addressed through ecological remediation or environmental governance, but the scarcity of non-renewable resources poses a more serious problem (Polasky, Nelson, Pennington & Johnson, 2011). In the current research, social metabolism refers to the exchange and consumption of materials and energy between nature and

252

X. Cui, X. Wang / Habitat International 49 (2015) 251e259

human society, and urban land use is considered the core problem from the social metabolism perspective. Such perspective aims to track the material flows among different social sectors and analyze their state and efficiency to understand how urban land use change influences social metabolism. In this manner, how the land use activities of humans affect the natural resources and the environment can be reflected. A theoretical analysis and empirical study were performed using Shanghai data to construct a method that could connect land use change with social metabolism and to propose several suggestions that could mitigate the negative effects of land use on the environment and natural resources. 2. Literature review After global industrialization and urbanization have accelerated economic development during the 20th century and have induced environmental pollution and ecosystem risks all over the world, scholars have begun to investigate the effects of urbanization on the environment and natural resources (Yan et al., 2003). Most of these researchers have adopted the social metabolism perspective in their studies. Social metabolism has evolved from the concept of metabolism in life science. This concept is currently used to reflect the material and energy exchange between social and natural systems and to examine the exploitation, manufacturing, and application of natural resources as well as the emission of waste materials (Han, 2002). The social metabolism perspective considers the human society as a system that interacts with nature through the exchange of materials and energy as well as links the diminishing natural resources to the land use behaviors of humans (Heinz & Niels, 2002). Domestic scholars indicate that land use change is a consequence of human activities that can affect landscape conversion and metabolism change (Xie, 2008). Researchers from Austria perceive land use as a colonization process that continuously intervenes with the natural ecosystem (Helmut, Simon, & Emilio, 2001). Previous studies identify two features of social metabolism, namely, a continuous exchange between human society and nature and the storing or discharge of some materials and energy into the social system (Huang & Hsu, 2003). The changes in urban land use during industrialization have altered the throughput of energy, and the interrelations between social metabolism and land use change can be detected through material flow accounting (Krausmann & Haberl, 2002), which is a useful tool for tracking material and energy flows as well as for exploring the effect of land use change (Helmut, Marina, Fridolin, Helga, & Verena, 2004). A biomass flow model has been established to evaluate the relationship between land use and social metabolism, which depends on the land use type and its corresponding demands of biomass change (Krausmann et al., 2003). Another study demonstrates that the changes in land use and cover can influence the exchange of materials and energy between nature and humans, which can be evaluated using a metabolism frame (Fischer & Rotmans, 2009). Land use primarily changes the regional environment and influences the changes in the bio-geographical processes that are driven by the evolution of landscape structure (Stephanie, Paul, & Tisha, 2012). Urban development and planning must be connected with social metabolism in the cities to increase the ecological capacity and reduce the dependence of humans on the environment (Boyen, Millar, Newcombe, & O'Neill, 1981). Studies on the metabolism response of land use change have also identified different forms of land use and their corresponding influence. A study from Austria reveals that land use/cover is closely related with social metabolism during the transition of lands from agricultural lands to industrial lands (Krausmann, 2001). Heinz and Niels (2002) investigated the relationship

between land use and social metabolism in England from the 1850s to the present, provided a metabolism account of land use change, and analyzed different aggregates of inputs, such as domestic material extraction, foreign trade of materials, and energy input. Krausmann, Haberl, Erb, and Wackernagel (2004) analyzed the relationship among economic growth, socio-economic metabolism, and land use using four scenarios of land use pattern. They revealed a complex feedback mechanism between energy and land use policies as well as emphasized the usefulness of the socioeconomic metabolism approach in examining land use. Given that the human appropriation of net primary production (HANPP) could be influenced by land use change, Wrbka et al. (2004) studied the relationship between social metabolism and land use intensity by calculating HANPP. They reported that strong monotonous correlations were also found between HANPP and urbanity, and land form could influence metabolism patterns, but not entirely. Lee, Huang, and Chan, 2009 used a spatial system modeling method to develop a socio-economic metabolism and land use change model that could simulate the spatialetemporal dynamics of socioeconomic metabolism and land use change. Chun asserted that material inflows could stimulate the accumulation of urban assets, and that some urban assets were out flowed to surrounding areas upon reaching the upper limit, thereby triggering land use change. Joan, Joan, Enric, and María (2010) used the Barcelona Metropolitan Region as an example to explore the synergies between sociometabolic energy use and various landscape patterns. They concluded that the simultaneous loss of energy and land use efficiencies from the mid-19th century to present could be tracked by the changes in the functional landscape structure, thus revealing the importance of traditional rural landscapes in maintaining the ecological quality of non-built-up land. Erb (2012) reviewed the socio-ecological metabolism method for examining the changes in land use intensity and noted that material flow analysis can be used to study the vital aspects of land intensification by collecting, deriving, or modeling information on the stocks and flows of materials, energy, or substances between socioeconomic and natural systems. Land use change has been recently investigated by Chinese scholars, and some regions in China have been used as research cases. By taking Jiangsu province as a case, Huang, Yu, and Ma (2006) studied the effect of land use change on social metabolism and reported that land use style, intensity, and pattern could directly affect the changes in the metabolism input and output of the socioeconomic system. A 10% increase in land use intention is accompanied by a 5.62% increase in social metabolism flux, whereas every 10% increase in land use intensity can induce a 5.03% increase in social metabolism efficiency. Through material flow analysis, Ma and Huang (2008) examined the responses of metabolism during the process of land use change in Jiangsu province from 1996 to 2005 and found a close relationship between these two variables. The changes in agricultural land did not increase the scale of output, whereas the changes in construction land increased the scale of input. Taiwanese scholars have analyzed the relationship between land use change and social metabolism using material flow accounting. By applying Geographic Information System (GIS) and its visualization in analyzing their data, these scholars argue that the development of Taipei was highly dependent on the flow of non-renewable energy; land use in Taiwan also demonstrated a significant layer structure that concerned the consumption of non-renewable energy (Chun et al. 2009). Wu, Yan, and Xu (2009) reviewed the sustainability aspect of energy-based urban metabolism and identified four points that should be further analyzed in the future. These points include (1) an integrated model for the ecological mechanism between urban metabolism and land use change, (2) government policies that are based on the

X. Cui, X. Wang / Habitat International 49 (2015) 251e259

metabolic course, (3) influence of global environmental change on land use and urban metabolism, and (4) interrelationships between urban metabolism and sustainable development. Wu and Yan (2011) investigated the interaction between land use change and urban metabolism by evaluating the energy-based urban metabolism in Guangzhou over the past 17 years. They reported that land use change significantly affected the material and energy flow, particularly in construction land. From the social metabolism perspective, this study focuses on urban land use transition and non-agricultural changes as well as evaluates the effect of land use change on the amount and efficiency of material flow. This paper is organized as follows. First, the existing studies are reviewed and the general research framework is constructed. Second, the indicators are selected and an appropriate methodology is defined. Third, Shanghai is used as a case study area and the data are collected. Fourth, the model is calculated using a computer software, and the results are obtained. Fifth, the results are analyzed and several suggestions for promoting the sustainability of urban land use are proposed. The flow of this research is summarized in Fig. 1. 3. Data collection and study method 3.1. Study area and data collection 3.1.1. Study area Shanghai is used as a case for this study. As one of the municipalities of China, Shanghai is located in the middle of the northesouth coastline and is considered the economic, financial, trade, and shipping center of mainland China (see Fig. 2). It has witnessed significant changes in its land use along with its rapid economic development these years. Miao, Cui, Luan, and He (2011) investigated the spatial land use changes in Shanghai and revealed that the multi-satellite towns of Shanghai were developed along with its central area, including the eastern part along the river, where many croplands have been turned into construction lands during this process. Shi, Wang, Yao, Niu, and Yu (2012) studied the spatial and temporal variations of land use in Shanghai between 1994 and 2006 and indicated that 74,825.86 h m2 of farmland were sacrificed for urban construction, resulting in a 67.12% reduction in the number of farmlands in the city. Wang, Huang, and Wang (2010) used the TM satellite images from 1987, 1997, and 2004 to prove that land use change could significantly affect the total cropland net primary productivity (NPP) in Shanghai. The decrease in land use has contributed 78% to the change of total NPP from the 1980s to

Fig. 1. Flow chart of the study process.

253

the 1990s, and such contribution increased to 92% from the 1990s to the 21st century. Recent studies have indicated that the urbanization in Shanghai has also induced an inefficient and imbalanced land use. It must be further regulated to improve the intensive and sustainable use of land resources (Li, Li, Li, Xu, & Qin, 2008; Liu, Ji, & Duan, 2009). 3.1.2. Indicators Land use change and its ecological effect have been studied for several years from different perspectives (Bai & Imura, 2001; Carlson & Traci Arthur, 2000; Walter & Stützel, 2009), including social metabolism changes (Haberl & Schandl, 1999). The current study considers social metabolism as a response to land use change, and several indicators must be selected to ensure that the effect of land use change can be properly reflected. Therefore the indictor frame consists of two parts, namely, the indictors of land use change and of social metabolism. 3.1.2.1. Indicators of land use change. Previous studies on land use and land cover change indicate that the indicators for reflecting land use change can be categorized into land use structure and land use efficiency (Chen, Jia, & Lau, 2008; Qian, 2008; Thinh, Arlt, Heber, Hennersdorf, & Lehmann, 2002). Land use structure refers to different types of land use, and they are classified into urban development and non-development lands in China (Ministry of Housing and UrbaneRural Construction of the People's Republic of China, GBJ137-90). Urban development lands include residential, industrial, storage, public facility, green, and road and traffic lands, all of which reflect the urban land use structure in China. The following dynamic index is used to define the changes in the urban land use structure:

K¼

Ub  Ua 1 $  100%; T Ua

(1)

where K refers to the degree of dynamic land use change, which can reflect the changes in the scope and pace for any type of land use within a certain period. Ua and Ub represent the land use area at the start and end periods, respectively, and T denotes the study period. This index is typically applied in regional difference analysis and tendency prediction to understand the changing characteristics of land use structure and ratio (Zhang, Ma, Wang, & Ji, 2012). Land use efficiency normally concern the economic density of an industrial space (Qin, 2011), which is calculated by dividing the industrial comparative added value by the land area. Therefore, land use must be categorized based on the GDP structure, which includes the three major industries that comprise the GDP of China. The land use for the first, second, and third industries is represented by the area of cultivated, industrial, and commercial lands, respectively. Such division of economic density can reflect and differentiate the land use efficiency of the three major industries (Luo, Wu, & Feng, 2010). Cultivated land is a subcategory of non-development land according to the official land use classification of China. It is among the largest segments of land use in China, comprising 8% of all land areas in the country. Most of the urban development lands in China have been transferred from cultivated lands during the land expropriation process, and the efficiency of cultivated lands has a crucial effect on social metabolism efficiency and intensity (Zhao, Xu, & Mei, 2005). Industrial land is contained in the urban development land as specified above, whereas commercial land is a subcategory of public facility land (Ministry of Housing and UrbaneRural Construction of the People's Republic of China, GBJ137-90) as shown in the official statistical yearbooks of China. Land use efficiency is computed as follows:

254

X. Cui, X. Wang / Habitat International 49 (2015) 251e259

Fig. 2. Case study area.

Ki ¼

Gi g

Si

;

(2)

where Ki is the economic density of the cultivated, industrial, and commercial lands, Gi denotes the added GDP on each land area, g is the price index, and Si denotes the land area. 3.1.2.2. Indicators of social metabolism. The materials must be classified before selecting the indicators of social metabolism. The economy-wide material flow accounts of Euro stat and the derived indicator frame are used as the methodological guides for this study (EUROSTAT, 2001), and these flow accounts have been modified in this study to fit the Chinese context. The metabolic process includes seven major parts, namely, (1) domestic extraction of fossil energy, metal ores, and non-metallic minerals; (2) import and export of agricultural and industrial products; (3) net addition of socioeconomic stocks for housing, construction, and manufacturing of durable goods; (4) environmental pollution; (5) deliberate discharge of chemical fertilizers, agricultural chemicals, and plastic film for farming; (6) balancing the input of oxygen and the output of carbon dioxide and other gases; and (7) hidden flows or untapped parts of materials during the mining of raw materials. The data are subsequently divided into three aspects, namely, material input, material output, and material intensity and efficiency (Ma & Huang, 2008) as shown in Table 1. The input indexes include direct material input (DMI) and total material requirement (TMR), whereas the output indexes include direct material output (DMO) and total material output (TMO). The intensity and efficiency indexes include material consumption intensity (MCI) and material productivity (MP). Hidden flow is obtained from the material mass and the hidden flow ratio. Given that the national hidden flow is seldom recorded, the hidden flow ratio in this paper is derived from the world average ratio that is provided by the Wuppertal Institute and the Sustainable Development Center of Taiwan. 3.1.3. Data collection The land use area and added GDP data are directly collected from the statistical yearbooks of Shanghai for the years 2001e2013.

In order to make the data for material consumption, storage, and discharge be collected completely, the flow process of social metabolism (input and output) in different sections must be tracked and understand. As shown in Fig. 3, the human society is divided into five parts, namely, agriculture, industry, consumer, construction, and transportation sections, to track the material flow (Bao, 2010; Erb, 2012). The original data are collected based on the classification and flow of materials in each section. For the agriculture section, biomass and industrial product input data can be collected from the statistical yearbooks of Shanghai. For the industrial section, industrial products data can be collected from the same yearbooks, whereas raw material input and industrial product output data can be collected from the national inputeoutput tables for 1992 (National Bureau of Statistics, National Economy Accounting Department, 1994). For the consumer section, the biomass and industrial product input data are estimated based on the product consumption quantity per person in a household and of the entire city population, whereas the durable goods input data are calculated based on the indexes that are presented in the statistical yearbook of Shanghai, including the indexes of ownership per hundreds of households. For the construction section, data on the wood, steel, aluminum, cement, and glass inputs can all be found in the statistical yearbook of Shanghai. Given that some data from 2000 to 2003 are unavailable, these data are calculated based on the material unit consumption in 2004 and the annual construction area. For the transportation section, the energy consumption of all of the transportation modes is calculated based on their number of passengers, unit consumption of fuel, and conversion coefficients (Li & Xie, 1999). Its data on the number of vehicles are obtained from the statistical yearbook of Shanghai and are converted into mass. 3.2. Study method The relationship between land use change and social metabolism is examined through canonical correlation analysis (CCA), which is often used to examine the causal relations between two sets of variables according to statistical data.

X. Cui, X. Wang / Habitat International 49 (2015) 251e259

255

Table 1 Social metabolism indexes. Material category

Index name

Implication

Composition

Input

Direct material input (DMI)

Domestic obtaining þ import

Output

Total material requirement (TMR) Direct material output (DMO)

The amount of materials that directly take part in social metabolism All of the materials that are involved in social metabolism

Total material output (TMO) Intensity and efficiency

Material consumption intensity (MCI) Material productivity (MP)

The direct material outputs of the economy All of the output materials from social metabolism that are discharged into the economy The annual consumption per person in the economy The economic benefits of the per material consumption in the economy

CCA was introduced by H. Hotelling in 1936 and became a relatively complete theory in the 1970s after years of development. CCA is often considered a dimension reduction technique in multivariate analysis (Ma, 2002). The optimum linear combination of two groups of variables is identified to maximize the relationship between these groups. The basic idea of CCA is to identify vectors or the linear combination of variables in each group according to the maximum correlation principle, which determines several pairs of canonical variables. The vectors that maximize the same correlation subject are then identified amid the constraint that these vectors may not be correlated with the canonical variables. Such procedure may be continued until the entire relationship is extracted (Chinese Academy of Agricultural ScienceeInstitute of Agricultural Economics and Development, 2008) and the correlation between two sets of variables can be understood. Given that CCA can help describe the correlation between two sets of variables, this technique has the preconditions that the two groups of variables are continuous and that the data must follow a multivariate normal distribution (Hardoon, Szedmak, & ShaweTaylor, 2004). And the correlation between the two groups must also be linear to allow the CCA to accurately explore their relationship (Hotelling, 1936). As shown in Table 2, all of the variables for land use and social metabolism are continuous, and the linear correlation between these two groups of variables can be tested by

DMI þ input hidden flows Pollution þ deliberate discharge þ balancing items þ output DMO þ output hidden flows TMR/population GDP/TMR

describing the tendency of social metabolism and urban land use before the establishment of the CCA model. The general form of CCA maximizes the correlation between two groups of variables by exploring their optimum linear combination as shown in the following equation:

c ¼ a0 $X ¼ a11 X1 þ a12 X2 þ … þ a1m Xm h ¼ b0 $Y ¼ b11 Y1 þ b12 Y2 þ … þ b1n Yn

(3)

The variance and covariance between two sets of random variables can be shown as follows:

DðcÞ ¼ Dða0 $XÞ ¼ a0 CovðX; XÞa ¼ a0 S11 a DðhÞ ¼ Dðb0 $YÞ ¼ b0 CovðY; YÞb ¼ b0 S22 b Covðc; hÞ ¼ a0 CovðX; YÞb ¼ a0 S12 b

(4) 0

CovðX; YÞ a S12 b pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Corrðc; hÞ ¼ pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffi ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi DðcÞ$ DðhÞ a0 S11 a$ b0 S22 b CCA aims to obtain several canonical equations that maximize the correlation between two datasets by estimating the weighting coefficient in Equation (3). CCA normalizes the mean of the variables to zero to quantify the

Fig. 3. Flow of materials during the metabolism.

256

X. Cui, X. Wang / Habitat International 49 (2015) 251e259

Table 2 Indicators in the equation frame. Equation

Category

Indicators

Unit

Eq. 1

Urban land use structure

X1- dynamic degree of residential land X2- dynamic degree of industrial land X3- dynamic degree of storage land X4- dynamic degree of public facility land X5- dynamic degree of green land X6- dynamic degree of road and traffic land X7- dynamic degree of cultivated land Y1- direct material input Y2- direct material output X8- economic density of cultivated land X9- economic density of industrial land X10- economic density of commercial land Y3- material consumption intensity Y4- material productivity

% % % % % % % t t Yuan/m2 Yuan/m2 Yuan/m2 t/person Yuan/t

Social metabolism amount Eq. 2

Urban land use efficiency

Social metabolism efficiency

The indicators, including TMR, TMO, and center city area, are obtained and calculated based on the collected data. The changing tendency of these indicators is illustrated in Fig. 4. The social metabolism indicators (TMR and TMO) have a linear correlation with the center city area, which indicates that the relationship between land use and social metabolism fits the request of CCA. Table 1 shows that TMR contains all of the resources (e.g., biomass, metal, non-metallic minerals, other materials that are inputted into the human society, and hidden flows) that are involved in social metabolism, except for water and air. Given that TMR comprises all of the primary resources that are used in the production side of an economy (e.g., trade and service activities),

this indicator can reflect the overall pressure that is generated by human activities on the environment (Environmental Signals, 2000). Huge changes are observed in the center city area between 2000 and 2006 (Fig. 4), and a slight increase has been observed after 2006. Similar to center city area, TMR shows an increasing trend between 2000 and 2012, during which the indicator increases rapidly and then slightly fluctuates after 2006. This consistent tendency implies that the urbanization in Shanghai from 2000 to 2012 is accompanied by increasing material requirements. TMO, which is calculated based on DMI and hidden flows, comprises all of the output materials that have come from social metabolism and are discharged to the environment. It depicts the ecological pressures that are generated by the urbanization of Shanghai between 2000 and 2012. Fig. 4 shows that the curve of TMO demonstrates an increasing trend similar to that of the center city area, with TMO having a higher rate of increase. The tendencies of TMR and TMO are highly consistent with those of the center city area because of the rapid urbanization in Shanghai. Such urbanization has resulted in the construction of numerous infrastructures that are largely composed of nonmetallic minerals (Fig. 5). The urban expansion of Shanghai has exhibited an outward wave-like pattern before 2005 (Li, Li, Zhu, Song, & Wu, 2013), during which the consumption of nonmetallic minerals has peaked in 2004 and has remained stable between 15,000 and 20,000 tons. Given that urbanization is always typically characterized by an increase in per capita energy consumption (Liu, 2009), the consumption of fossil energy in Shanghai has demonstrated an increasing trend between 2000 and 2012 as well. A gradual increase in the consumption of energy, mineral resources, and land source may be observed with further urbanization, and the influencing mechanism of land use change on material requirement, output, and utilization must be understood to achieve a more sustainable development.

Fig. 4. Tendencies of the TMR, TMO, and center city area from 2000 to 2012.

Fig. 5. Components of regional material consumption.

influence of each variable. This technique is more effective than calculating the weight coefficients of these variables while identifying their optimum linear combination. Given that several canonical equations can be extracted by CCA, the number of valid equations must be selected. According to Bartlett (1941), the number of these equations is selected based on the significance of the equations. The canonical loading coefficients can be calculated using the weighing coefficients of the selected canonical equations. The effect of land use change on social metabolism change can be interpreted through these loading coefficients. In Equation (3), c1 represents the linear combination of the land use variables, which include X1 … Xp, and h1 denotes the linear combination of the social metabolism variables, which include Y1 … Yq. To explore their relationship, two sets of equations are established; the corresponding indicators in the CCA model are shown in Table 2. 4. Analysis of results 4.1. Tendency of social metabolism and urban land use

X. Cui, X. Wang / Habitat International 49 (2015) 251e259

4.2. Effect on social metabolism amount

257

Table 4 Canonical loading coefficients of land use structure.

In this section, the relationship between social metabolism quantity and land use structure is revealed through CAA. The degree of dynamic residential, industrial, storage, public facility, green, road and traffic, and cultivated land use change are denoted as X1, X2, … X7 in the established equation. For the social metabolism aspect, the direct material inputs and outputs are denoted as Y1 and Y2, respectively (Table 2). As shown in Table 3, the estimated results from SPSS 13.0© reveal that the relationship between social metabolism and land use structure is statistically significant for the first canonical equation. The sig-value (0.052) demonstrates that the first pair of canonical variates can reflect the relationship between land use structure and total metabolism amount, with the correlation coefficient (0.997) representing the bivariate correlation between the two sets of indicators. The correlation coefficient also shows that urban land use change can gradually increase the social metabolism quantity. Such positive influence explains the similar change tendencies of TMR, TMO, and center city area as shown in Fig. 2. However, as illustrated in Table 4, these indicators have different effects on social metabolism quantity for each type of land use. The canonical loading in Table 5 shows the simple linear correlation between the independent variables and the corresponding canonical variate, whereas the cross canonical loading shows the correlation of the observed dependent/independent variables with the corresponding canonical variate (Joseph, 2011). Both of these loadings reflect the relationship between the independent/dependent variables and the corresponding canonical variate. Table 4 shows that the storage and transportation lands have large loadings on canonical function (canonical loadings of 0.652 and 0.668, respectively; cross loadings of 0.690 and 0.599, respectively), which indicates that social metabolism quantity is closely related with increasing storage and transportation lands. Such close relationship may be attributed to the rapid urbanization of Shanghai and its large urban population, which has expanded urban boundaries and traffic roads as well as increased the number of infrastructures and demand for building materials. All of these factors have induced serious environmental stress (Jeremy, 2003) and largely influenced the amount of social metabolism. According to Capello and Camagni (2000), the need for transportation and infrastructure is constantly accompanied by increasing material inputs and outputs, which not only include the resources to be used, but also the wastes to be discharged. Additionally, as one of the largest transportation hubs in China, Shanghai has a large flow of goods that requires additional land for storage use. Therefore, the warehouse layout must be centralized and arranged, and vertical storage facilities must be established to meet the requirements of urban goods and material flow. Residential, industrial, storage, public facility, and traffic lands have positive effects on material inputs and outputs. The increasing need for urban land among urban residents has increased to support their living, working, transportation, and storage needs. Such increasing demand has also enlarged the urban scale and its material demands as well as extended travel distances (Yeh & Li, 2000). By contrast, green and cultivated lands negatively affect the social metabolism amount according to the loading coefficients in Table 4 (green land with a coefficient of 0.337 and cultivated land with a coefficient of 0.077). Therefore, urban green lands Table 3 Correlation coefficients of land use structure and metabolism amount. Canonical equation

Correlation coefficient

Chi-square

df

Sig

1 2

0.997 0.835

22.660 6.323

14 6

0.052 0.391

Variable

Canonical loading

Cross canonical loading

X1 X2 X3 X4 X5 X6 X7

0.490 0.115 0.652 0.373 0.337 0.690 0.077

0.472 0.104 0.668 0.355 0.351 0.599 0.073

Table 5 Correlation coefficients of land use productivity and metabolism efficiency. Canonical equation

Correlation coefficient

Chi-square

Df

Sig

1 2

0.989 0.900

28.2433 9.3110

6 2

0.000 0.018

must be expanded to relieve the metabolism pressure and balance the amount of material inputs and outputs, whereas cultivated lands must be strictly protected from the transition to construction land during an urban sprawl. These proposals must be adopted by the government to guide its urban development and land use policies (Zhu, Zhang, Li, & Zhu, 2014). From this perspective, the planning, utilization, and consolidation of lands have become critically important. 4.3. Effect on social metabolism efficiency The metabolism efficiency responses that are caused by land use productivity change are explored in this section. As mentioned in Table 2, the land use efficiency indexes are denoted as X8, X9, and X10, which represent the economic densities of cultivated, industrial, and commercial lands, respectively. The social metabolism efficiency indexes are represented by the material productivity and material consumption intensity, which are denoted as Y3 and Y4, respectively. The estimated results from CCA show that the two canonical equations are both statistically significant (see Table 5). The relationship between social metabolism efficiency and land use productivity can be understood from their correlation coefficients (0.989 and 0.900, respectively), which implies a close relationship. Table 6 shows the canonical loadings of the two sets of equations. For the first equation, a significant relationship between X8 and Y3 is revealed by the loadings of 0.978 and 0.869, which indicates that the economic productivity of cultivated land largely affects material productivity. This positive effect on material productivity is crucial for the economic assessment of cultivated land use, and the enhancement of economic density can help increase material productivity (e.g., GDP per unit of material consumption). This finding is consistent with that of Erb (2012), who reports that the changes in land use intensity will not change land cover, but

Table 6 Canonical loading coefficients of the two equations. 1

2

Loadings of equation on land use efficiency X8 0.978 X9 0.113 X10 0.277 1

0.120 0.579 0.893 2

Loadings of equation on social metabolism efficiency Y3 0.869 Y4 0.592

0.433 0.729

258

X. Cui, X. Wang / Habitat International 49 (2015) 251e259

induce ecological changes (material productivity in this study) within the same land cover type. For the second equation, X9 (the economic density of the industrial land) and X10 (the economic density of the commercial land) both significantly affect Y4 (the material consumption intensity) as revealed by their coefficients 0.579, 0.893, and 0.729, respectively. This negative influence indicates that a higher economic efficiency and land use intensity will result in lower material consumption intensity. Therefore, the ecological and environmental pressures can be reduced by increasing the intensity of the industry and the commercial land to some extent. Given that the enhancement of land use efficiency may reduce the land area demand as well as prevent a considerable amount of carbon emissions and deforestation, such enhancement is often considered essential for ensuring sustainability (Burney, Davis, & Lobell, 2010). The CCA results specify the sustainable aspects of material productivity and material consumption intensity, which is denoted as metabolism efficiency in the present study. Therefore, several points concerning urban land use in Shanghai should be considered to enhance the efficiency of social metabolism and achieve a more sustainable development. First, for the agricultural land, certain cultivated land areas should be strictly protected, modern agriculture including scaled production should be developed, and farmland use should be intensified. Natural goods supply by agricultural land not merely views single ecosystems in isolation, but also understands the interplay in a specific context (Fürst, Helming, Lorz, Müller, & Verburg, 2013). Improvement in farmland use efficiency is useful for land conservation on the one hand, and is beneficial for urban material productivity in terms of freeing up non-renewable resources on the other hand. Second, increasing the economic productivity of urban commercial and industrial lands is the key approach for reducing the intensity of material consumption. For the center city, heavy industrial land use should be transferred to technology industrial use to effectively enhance land use efficiency and reduce material consumption. Other aspects, such as total land use control, function replacement, and function combination, are also useful for saving the limited land resource and improving social metabolism efficiency. In the suburbs, the scattered industrial enterprises that are located in towns and villages require further planning to achieve integration and induce a cluster effect. The optimization and adjustment of industrial land appear to be the major target for the suburbs. The government may have to undertake industrial zone planning to avoid land waste, which is also beneficial for conserving materials and protecting the environment. 5. Conclusion Land use is a major driver for habitat encroachment and biodiversity loss (Sala et al., 2000). The current study provided a method for measuring the relationship between changes in urban land use and social metabolism. The results indicated that changes in land use and the subsequent changes in land cover played an important role in material production and consumption, and ultimately altered the metabolism quantity and efficiency. Data in Shanghai illustrated a consistency trend with urban land use scale and metabolism quantity during the study period from 2000 to 2012. The total social metabolism was influenced by the structure of urban land use, in which the material inputs and outputs were largely affected by storage land and transportation land. It recognized the detrimental effects of residential, industrial, and transportation lands on metabolism throughput, however, green and cultivated lands were found to play a reverse role. For the efficiency aspect, material productivity was positively influenced by

the efficiency of cultivated land, whereas material consumption intensity was negatively affected by the efficiency of industrial and commercial lands. Regional land planning and its management policy were proposed as one of the best means of efficiently and equitably distributing land resource. Several suggestions concerning urban land use can be adopted to alleviate the immense pressure of social metabolism. This study cited cultivated land as a crucial factor that affects social metabolism; hence, its conservation is expected to benefit both food production and material productivity. For other urban land-use types, controlling land for storage and traffic use and protecting the open space and green land are deemed to be the effective methods for reducing the total material inputs and outputs during their flow process. The efficiency of industrial and commercial lands should be enhanced to alleviate the pressure on the ecological environment, and such an approach is associated with the intensity of social metabolism. These findings may contribute to a more comprehensive understanding of the metabolism response of urban land use change. They may also be useful for land use policy planning in urban settlements. Understanding the internal relationship between urban land use and social metabolism response should be regarded as the initial step in achieving ecosystem balance and controlling urban sprawl. For future research, we should explore the spatial shifts in urban land use change and its corresponding dynamic material input and output changes. Funding information This paper is funded by the Project of philosophy and social sciences in Guangdong Province (GD14CGL02); the Project of science and technology for local universities in Guangzhou city (1201420951); the Project of philosophy and social sciences in Guangzhou University for young Ph.D (201403QNBS). References Alberti, M., Booth, D., Hill, K., Coburn, B., Avolio, C., Coe, S., et al. (2007). The impact of urban patterns on aquatic ecosystems: an empirical analysis in puget lowland sub-basins. Landscape and Urban Planning, 80, 345e361. Bai, X. M., & Imura, H. (2001). Towards sustainable urban water resource management: a case study in Tianjin, China. Sustainable Development, 9, 24e35. Bao, Z. M. (2010). Material flow analysis (MFA) of the environmentaleeconomic system of Dalian. Dalian University of Technology (in Chinese). Bartlett, M. S. (1941). The statistical significance of canonical correlations. Biometrika, 32, 29e37. Boyden, S., Millar, S., Newcombe, K., & O'Neill, B. (1981). The ecology of a city and its people: The case of Hong Kong. Canberra: Australian National University Press. Burney, J. A., Davis, S. J., & Lobell, D. B. (2010). Greenhouse gas mitigation by agricultural intensification. Proceedings of the National Academy of Sciences, 107, 12052e12057. Capello, R., & Camagni, R. (2000). Beyond optimal city size: an evaluation of alternative urban growth patterns. Urban Studies, 37, 1479e1496. Carlson, T. N., & Traci Arthur, S. (2000). The impact of land use-land cover changes due to urbanization on surface microclimate and hydrology: a satellite perspective. Global Planetary Change, 25, 49e65. Chen, H. Y., Jia, B. S., & Lau, S. S. Y. (2008). Sustainable urban form for chinese compact cities: challenges of a rapid urbanized economy. Habitat International, 32, 28e40. Chen, X. L., Zhao, H. M., Li, P. X., & Yin, Z. Y. (2006). Remote sensing image-based analysis of the relation between urban heat island and land use/cover changes. Remote Sensing of Environment, 104, 133e146 (in Chinese). Chinese Academy of Agricultural ScienceeInstitute of Agricultural Economics and Development. (2008). National agricultural policy analysis platform and decision support system: Agricultural econometric model and its application. Beijing: Agriculture Press (in Chinese). Erb, K. H. (2012). How a socio-ecological metabolism approach can help to advance our understanding of changes in land-use intensity. Ecological Economics, 76, 8e14. EUROSTAT. (2001). Economy-wide material flow accounts and derived indicators: A methodological guide (pp. 7e89). Luxembourg: Office for Official Publications of the European Communities. Fischer, K. M., & Rotmans, J. (2009). Conceptualizing, observing and influencing

X. Cui, X. Wang / Habitat International 49 (2015) 251e259 socio-ecological transitions. Ecology and Society, 14(2), 1e18. Foley, J. A., DeFries, R., Asner, G. P., Barford, C., Bonan, G., Carpenter, S. R., et al. (2005). Global consequences of land use. Science, 309, 570e574. Fürst, C., Helming, K., Lorz, C., Müller, F., & Verburg, H. P. (2013). Integrated land use and regional resource management: a cross-disciplinary dialogue on future perspectives for a sustainable development of regional resources. Journal of Environmental Management, 127, 1e5. Giuseppina, S. (2012). Urbanization strategies, rural development and land use changes in China: a multiple-level integrated assessment. Land Use Policy, 1, 165e178. Hall, C. A. S., Perez, C. L., & Leclerc, G. (2000). Quantifying sustainable development. The future of tropical economies. San Diego: Academic Press. Han, L. X. (2002). Marx's conception of material metabolism and the thought of environmental protection. Philosophical Researches, 2, 6e13 (in Chinese). Hardoon, D. R., Szedmak, S., & Shawe-Taylor, J. (2004). Canonical correlation analysis: an overview with application to learning methods. Neural Computation, 16(12), 2639e2664. Heinz, S., & Niels, S. (2002). Changes in the United Kingdom's natural relations in terms of society's metabolism and land-use from 1850 to the present day. Ecological Economics, 41, 203e221. Helmut, H., Marina, F. K., Fridolin, K., Helga, W., & Verena, W. (2004). Progress towards sustainability? what the conceptual framework of material and energy flow accounting (MEFA) can offer. Land Use Policy, 21, 199e213. Helmut, H., Simon, B., & Emilio, M. (2001). Using and shaping the land: a long-term perspective. Land Use Policy, 18, 1e8. Hotelling, H. (1936). Relations between two sets of variates. Biometrika, 28, 321e377. Huang, S. L., & Hsu, W. L. (2003). Materials flow analysis and energy evaluation of Taipei's urban construction. Landscape and Urban Planning, 63, 61e74. Huang, X. J., Yu, S. T., & Ma, Q. F. (2006). Studies on material metabolism response of regional land use change. Journal of Natural Resources, 21(1), 1e8 (in Chinese). IGBP, & IHDP. (1999). Land-use and land-cover change (LUCC): Implementation strategy. Jeremy, R. (2003). The hydrogen economy: the creation of the worldwide energy web and the redistribution of power on earth. Refocus, 4(3), 12. Joan, M., Joan, P., Enric, T., & María, J. C. (2010). Social metabolism, landscape change and land-use planning in the barcelona metropolitan Region. Land use policy, 27, 497e510. Joseph, F. H. (2011). Canonical correlation, a supplement to multivariate data analysis. Pearson Prentice Hall Publishing. Krausmann, F. (2001). Land use and industrial modernization: an empirical analysis of human influence on the functioning of ecosystems in Austria 1830e1995. Land Use Policy, 18(1), 17e26. Krausmann, F., & Haberl, H. (2002). The process of industrialization from the perspective of energetic metabolism: socioeconomic energy flows in Austria 1830e1995. Ecological Economics, 41, 177e201. Krausmann, F., Haberl, H., Erb, K. H., & Wackernagel, M. (2004). Resource flows and land use in Austria 1950e2000: using the MEFA framework to monitor societyenature interaction for sustainability. Land Use Policy, 21, 215e230. Krausmann, F., Haberl, H., Schulz, N., Erb, K. H., Darge, E., & Gaube, V. (2003). LandUse change and socio-economic metabolism in AustriadPart I: driving forces of land-use change: 1950e1995. Land Use Policy, 20(1), 1e20. Lambin, E. F., Turner, B. L., Geist, H. J., Agbola, S. B., Angelsen, A., Bruce, J. W., et al. (2011). The causes of land-use and land-cover change: moving beyond the myths. Global Environmental Change, 11, 261e269. Lee, C. L., Hang, S. H., & Chan, S. L. (2009). Synthesis and spatial dynamics of socioeconomic metabolism and land use change of Taipei metropolitan Region. Ecological Modelling, 220, 2940e2959. Li, C., Li, J. X., Li, R., Xu, M. C., & Qin, H. (2008). Impacts of rail transit in Shanghai on its urban land use change. Chinese Journal of Applied Ecology, 19(7), 1537e1543 (in Chinese). Li, J. X., Li, C., Zhu, F. G., Song, C. H., & Wu, J. G. (2013). Spatiotemporal pattern of urbanization in Shanghai, China between 1989 and 2005. Landscape Ecology, 28(8), 1545e1565 (in Chinese). Liu, Y. (2009). Exploring the relationship between urbanization and energy consumption in China using ARDL and FDM. Energy, 34(11), 1846e1854. Liu, H., Ji, D. M., & Duan, H. (2009). Discuss the prospect of soil utilization in Shanghai. Integrated Transportation, 1(10), 23e27 (in Chinese). Li, X. M., & Xie, R. H. (1999). The relationship of the transportation and energy, on the perspective of energy consumption and conservation of transportation. Shanghai Urban Planning Review, 88(5), 59e62 (in Chinese). Luo, W. B., Wu, C. F., & Feng, K. (2010). Temporal-spatial features and influence mechanism of urban land economic density: an empirical study based on city panel data in hunan Province. Urban Studies, 17(6), 68e74 (in Chinese). Ma, Z. H. (2002). Modern math handbook: Probability statistic and stochastic progress. Beijing: Tsinghua Press (in Chinese). Ma, Q. F., & Huang, X. J. (2008). Responses of metabolism during the process of land use change. Transactions of the CSAE, 24(8), 6e11 (in Chinese).

259

Mark, A. D., Roger, F. A., Krista, A. K., Kristi, L. S., Janis, L. T., & Thomas, R. L. (2012). Land change variability and humaneenvironment dynamics in the United States great Plains. Land Use Policy, 7, 710e723. Miao, L. J., Cui, X. F., Luan, Y. B., & He, B. (2011). Similarities and differences of Beijing and Shanghai's land use changes induced by urbanization. Journal of the Meteorological Sciences, 31(4), 398e404 (in Chinese). Ministry of Housing and UrbaneRural Construction of the People’s Republic of China. City land classification and land use planning and construction standards. GBJ137e90 (in Chinese). National Bureau of Statistics, & The National Economy Accounting Department. (1994). National inputeoutput tables (for material) 1992. Beijing: National Statistics Press (in Chinese). Patz, J. A., & Olson, S. H. (2008). Land use/land change and health. International Encyclopedia of Public Health, 15e21. Polasky, S., Nelson, E., Pennington, D., & Johnson, K. A. (2011). The impact of landuse change on ecosystem services, biodiversity and returns to landowners: a case study in the state of Minnesota. Environmental & Resource Economics, 48(2), 219e242. Qian, Z. (2008). Empirical evidence from Hangzhou's urban land reform: evolution, structure, constraints, and prospects. Habitat International, 32(4), 494e511. Qin, B. (2011). The density gradient of industries' spatial distribution in Shanghai and its influence factors. Human Geography, 26(1), 39e43 (in Chinese). Sala, O. E., Stuart, C. F., Armesto, J. J., Berlow, E., Bloomfield, J., Dirzo, R., et al. (2000). Global biodiversity scenarios for the year 2100. Science, 287, 1770e1774. Seto, K. C., & Shepherd, J. M. (2009). Global urban land-use trends and climate impacts. Current Opinion in Environmental Sustainability, 1, 89e95. Shi, L. J., Wang, S. Y., Yao, X. J., Niu, J. J., & Yu, L. Z. (2012). Spatial and temporal variation characteristics of land use and its driving force in Shanghai city from 1994 to 2006. Resource and Environment in the Yangtze Basin, 21(12), 1468e1479 (in Chinese). Stephan, P., & Friedrich, D. (2000). Assessing the environmental performance of land cover types for urban planning. Landscape and Urban Planning, 52, 1e20. Stephanie, P., Paul, B., & Tisha, H. (2012). An expanded urban metabolism method: toward a systems approach for assessing urban energy processes and causes. Landscape and Urban Planning, 107, 193e202. Thinh, N. X., Arlt, G., Heber, B., Hennersdorf, J., & Lehmann, I. (2002). Evaluation of urban land-use structures with a view to sustainable development. Environmental Impact Assessment Review, 22(5), 475e492. United Nations. (2008). World urbanization prospects, the 2007 revision. Walter, C., & Stützel, H. (2009). A new method for assessing the sustainability of land-use systems (II): evaluating impact indicators. Ecological Economics, 68, 1288e1300. Wang, Y., Huang, M., & Wang, X. G. (2010). Impacts of land use and climate change on agricultural productivity in Shanghai. Acta Scientiae Circumstantiae, 30(3), 641e648 (in Chinese). Wang, H., Shen, Q. P., Tang, B., & Martin, S. (2013). An integrated approach to supporting land-use decisions in site redevelopment for urban renewal in Hong Kong. Habitat International, 38, 70e80. Wrbka, T., Erb, K.-H., Schulz, N. B., Peterseil, J., Hahn, C., & Haberl, H. (2004). Linking pattern and process in cultural landscape: an empirical study based on especially explicit indicators. Land Use Policy, 21, 289e306. Wu, Y. Q., & Yan, M. C. (2011). Analysis of the indicators between urban metabolism and land use change in Guangzhou. Geographical Research, 30(8), 1381e1390 (in Chinese). Wu, Y. Q., Yan, M. C., & Xu, L. F. (2009). Review on the research of energy-based urban metabolism. Ecology and Environmental Sciences, 18(3), 1139e1145 (in Chinese). Xie, H. L. (2008). Review and the outlook of land use ecological security pattern. Acta Ecologica Sinica, 28(12), 6305e6311 (in Chinese). Xu, J. C., Rita, S. M., Jing, F., & Xu, Y. F. (2008). Critical linkages between land-use transition and human health in the himalayan region. Environment International, 34, 239e247. Yan, W. H., Liu, Y. M., Huang, X., & Hu, Y. J. (2003). The change of urban metabolism and the effect of waste being created of Shenzhen. Urban problems, 1, 40e44. Yeh, A. G. O., & Li, X. (2000). The need for compact development in the fast growing areas of China: The Pearl River Delta. Compact cities: Sustainable urban forms for developing countries, 73e90. Zhang, J., Ma, C. H., Wang, Q. M., & Ji, W. W. (2012). Analysis on land use dynamic changes in hanzhong city. Research of Soil and Water Conservation, 19(1), 112e116 (in Chinese). Zhao, J., Xu, J. H., & Mei, A. X. (2005). Fractal study on land use structure and form change: the case of central areas in Shanghai. Journal of East China Normal University (Natural Science), 1, 78e84 (in Chinese). Zhu, F. K., Zhang, F. R., Li, C., & Zhu, T. F. (2014). Functional transition of the rural settlement: analysis of land-use differentiation in a transect of Beijing, China. Habitat International, 41, 262e271. report No. 6 Environmental assessment. (2000). Environmental signals. OPOCE.