How to accurately assess cultural ecosystem services by spatial value transfer? An answer based on the analysis of urban parks

Abstract

Cultural ecosystem services (CESs) intangibly influence many aspects of public daily life, and when evaluating them, it is difficult to obtain perception data. Spatial value transfer predicts the ecosystem service value of policy sites based on the spatial associations of study sites with services value. This approach has the potential to produce an evaluation without perception data. Previous studies have preliminarily researched the validity and effects of using spatial value transfer to evaluate CESs, but, there are doubts about the transfer performance of indicators, differences among various CES types, and the impact of the physical environment. This study examines the above key points of spatial value transfer in depth, with urban parks as the study area to assess three types of CES values that are closely related to human well-being: aesthetic, historical and recreational value. Two popular indicators, area under the curve statistics and transfer error rate, are used to evaluate transfer performance. The results reveal that historical values have the best transfer performance, and transfer error is lower in transfers from small locations to large locations with environmental variable combinations that include road network, water coverage and attraction distribution. The transfer error rate indicator could not only assess the overall transfer performance but also generate maps to reflect the distribution and value of transfer error. We suggest that more knowledge of minor differences in physical factors and developing specific transfer coefficients for them are necessary to enhance the accuracy of CES evaluation by spatial value transfer.

Introduction

Ecosystem services (ESs), referring to the benefits that people receive from ecosystems, have widely contributed to descriptions and assessments of the relationship between ecosystems and human well-being (Haines-Young et al., 2010). As economic development is increasingly challenging ecosystems, many policy-makers have integrated ESs into their decision-making to better manage natural capital on the basis of its holistic benefits (Chan et al., 2012; Dunford et al., 2018). The Common International Classification of Ecosystem Services (CICES) indicates that ESs include provisioning services (e.g., food and water), regulation and maintenance services (e.g., climate regulation), and cultural services (e.g., aesthetic, leisure and entertainment experiences) (Affek et al., 2020). Cultural ecosystem services (CESs), defined as the nonmaterial benefits that people obtain from ecosystems, have strong ties to human wellbeing and directly influence the physical, emotional, mental, and social aspects of people’s daily lives (MEA, 2005, Fagerholm et al., 2012, Fazey, 2011, Gee and Burkhard, 2010). In contrast to provisioning and regulating services, which can be substituted with socioeconomic progress, human dependence on cultural services is increasing and irreplaceable (Guo et al., 2010, Plieninger et al., 2013). As the importance of CESs has been generally recognized, an increasing number of studies have aimed to develop a methodological framework (Chan et al., 2012, Fish et al., 2016). However, CESs are reflected mainly in intellectual and spiritual interactions that are difficult to quantify (Cabana et al., 2020). Additionally, CESs overlap other types of ESs and are related to economics, sociology, and a wide range of other subjects, which means that CES research is challenging and insufficient, especially in assessment applications (Daniel et al., 2012, Milcu et al., 2013, Small et al., 2017).

Methods of CES evaluation are classified as monetary and nonmonetary methods (Braat et al., 2015, Hirons et al., 2016). Monetary methods assess CES values by calculating the costs of acquiring cultural services from the ecosystem (e.g., the travel cost [TC] method, contingent valuation [CV] method, and choice experiment method) (Berkel & Verburg, 2014; Gandarillas et al., 2016; Ungaro et al., 2016) or evaluating the financial value provided by the cultural services (e.g., market price method and hedonic pricing method) (Garcia et al., 2016, Sumarga et al., 2015). Monetary valuation has long been used to evaluate ecosystem services, but some types of CESs cannot be evaluated by monetary valuation, and the validity and adequacy of this approach have been questioned (Christie et al., 2012, Cheng et al., 2019, Hernández-Morcillo et al., 2013). Nonmonetary methods, which evaluate CESs by arbitrary scales other than monetary values, are regarded as better at assessing public perceptions of the ecosystem and are used often by researchers and policymakers (Hirons et al., 2016; Kaplin, 2005). The observation method (Unnikrishnan & Nagendra, 2015) and the social media-based method (Willemen et al., 2015) are associated with revealed preference, which calculates the CES value by observing behavior or analyzing texts, photographs, and other documents representing preference. More nonmonetary methods are stated preference for directly acquiring CES assessments from the public, including interviews (Schmidt et al., 2016), questionnaires (Bryce et al., 2016), focus groups (Ajwang’ Ondiek et al., 2016), expert-based methods (Nahuelhual et al., 2013), participatory mapping (Brown & Hausner, 2017) and visitor-employed photography (Sun et al., 2019). However, data acquisition for most nonmonetary valuation methods is usually limited by institutional or legal factors and requires considerable manpower and resources to gather preference data (Cheng et al., 2019). The social media-based method offers the possibility of freely obtaining large amounts of public preference data and is regarded as the most promising evaluation method, but the image identification still requires an investment of manpower and time (Tian et al., 2021). When evaluating multiple study locations, the investment of manpower and time can be great.

Value transfer (also known as benefit transfer) is defined as predicting the value of policy sites for which primary data are unavailable by transferring the assessment results of study sites (Rosenberger & Loomis, 2000). Unit value transfer, which applies a single value representing the characteristics of the study sites to the policy sites, and function transfer, which calculates the value at the policy sites by function variables describing the attributes of the study sites, are two primary types of transfer (Boyle et al., 2010). Earlier research focused more attention on the economic theory and practice of value transfer, in which the TC, CV, willingness to pay (WTP) function, and meta-regression analysis functions according to the WTP were popular (Brouwer, 2000, Richardson et al., 2015). With the increased understanding of the influence of spatial heterogeneity in transferring value and the good practice of using nonmonetary methods in primary evaluation, more researchers have attempted spatial transfer based on nonmarket values to obtain better transfer performance and demonstrate the spatial distribution for the services value (Troy & Wilson, 2006). Published analyses of spatial transfer for CES values are limited. Most studies have used a participatory geographic information system (PPGIS) to conduct primary evaluations and have differed in terms of the transfer approach, spatial unit value transfer, and spatial function transfer. Brown et al. (2016) identified the spatial association of CES values with land cover data and used land cover as the unit value for transfer. Sherrouse & Semmens (2014) provided a method of spatial function transfer by generating models describing the relationships between CES values and environmental variables. Among these studies, only aesthetic and recreational value types were examined, and there was a consensus that similar physical and social contexts of the study areas and a primary evaluation based on a large number of unbiased samples can reduce the transfer error.

Assessing the reliability of the value predicted by value transfer and quantifying the transfer error relies on transfer performance indicators (Lawton et al., 2021). The present indicators, the transfer error rate and ±2 value index (VI) (Sherrouse & Semmens, 2014), product-moment correlation (Brown et al., 2016), and area under the curve (AUC) statistics (Semmens et al., 2019), are controversial in their validity. The transfer error rate is limited by the fact that the formula is invalid when the value of the previous evaluation is 0. The ±2 VI is a supplemental indicator for the transfer error rate, and the product-moment correlation cannot provide statistically rigorous measurements. Semmens et al. (2019) evaluated the transfer performance by using the AUC, which quantitatively measured the predicted evaluation models with a threshold value of 0.70. Although the AUC could well evaluate the overall transfer performance, it could not indicate the distribution of areas with transfer errors. Identifying the region where transfer errors are generated is important to improve transfer performance in future studies. The transfer error rate of these indicators has the most potential for both quantifying the overall transfer performance and showing the distribution of transfer errors if the limitation on its calculation is removed. In addition, Sherrouse & Semmens (2014) discussed the correlation of the AUC of primary evaluation models with the error rate mean and cells within ±2 VI difference, and their results did not match expectations. Researching the correlation of indicators helps to understand their transfer performance assessment ability, but this is rarely studied.

In this study, we primarily selected urban parks for an evaluation of CES and spatial value transfer. Berghöfer et al. (2011) created a manual on ESs for cities that built upon the Economics of Ecosystems and Biodiversity (TEEB, 2010) and indicted that urban parks have been fundamental components of urban ecosystems and played the crucial role in human benefits (Mäntymaa et al., 2021). In efforts to establish CES assessment methods, urban parks have gained extensive attention, and the same problem of evaluations requiring significant investment exists in the CES assessment of them. There is no doubt that value transfer provides the possibility of efficiently and economically evaluating the CES values of urban parks. Social media combined with PPGIS are generally used to conduct primary evaluations. Based on this, we performed a primary evaluation of the aesthetic, historical and recreational values and predicted these values by spatial function transfer. We selected some environmental variables with differences among study locations to generate predicted models and compared the transfer performance of these models with different combinations of these environmental variables. In addition, we addressed the issue of invalid values when using transfer error rates for transfer performance evaluation through data standardization. Then, the transfer error rate and AUC indicators were used to assess the transfer performance, and their correlations were studied. We aimed to 1) identify the characteristics of the indicators and the most appropriate indicator for evaluating transfer performance; 2) discuss the transfer performance of various types of CESs; 3) demonstrate the variable combinations of the physical environment that generate the best predicted models; and 4) analyze the value transfer in urban parks and the possibility of reducing transfer error.