Ecosystem Services 60 (2023) 101518 Contents lists available at ScienceDirect Ecosystem Services journal homepage: www.elsevier.com/locate/ecoser The future of crowd-sourced cultural ecosystem services assessments A R T I C L E I N F O Keywords Cultural ecosystem services Crowd-sourced data Social media Automated image content analysis Semantic analysis 1. Introduction incentivising and improving global environmental stewardship and na­ ture conservation (Spotswood et al., 2021). Cultural Ecosystem Services (CES) represent one of the primary in­ This Special Issue includes 16 individual contributions that showcase teractions between humans and ecological systems, with implications cutting edge research in the use of crowd-sourced social media ap­ for environmental stewardship and awareness of environmental issues. proaches, including: 1) novel assessment opportunities for studying CES; 2) CES include the physical, experiential, intellectual and representative methodological advances through semantic analyses, automated image interactions we have with nature, our social relations within it, as well as content analyses, and validations; and 3) conceptual advances to better spiritual and symbolic attachments (Andersson et al., 2015; Chan et al., understand the representation of CES and creation of relational values 2012; Huynh et al., 2022). Consequently, the relationships that result within virtual social-media communities. This Special Issue overview from CES is a key motivation for caring for the environment (Dutcher provides critical reflection and future direction on the potential of social et al., 2007). Yet, while material ecosystem services (provisioning and media analyses for CES assessments, including specific limitations, key regulating services) have been mainstreamed into sustainability policies assumptions and biases, and remaining methodological, theoretical, and and landscape planning arenas in recent years, CES are missing from interpretive challenges. In the following sections we highlight insights many ecosystem service assessments (Nowak-Olejnik et al., 2022; from the assembled papers and present ideas on the future of crowd- Kosanic and Petzold, 2020). Challenges related to the recognition and sourced cultural ecosystem services research. quantification of CES have raised concerns about a lack of consideration of CES in decision making at local to global scales (IPBES, 2018). 2. Novel assessment opportunities Recently, however, there has been renewed interest in CES, with a focus on relational values (Chan et al., 2016; Chan et al., 2018; IPBES, When combined with geolocated social media data, advances in the 2022). Relational values of CES create additional incentives to care for quality, resolution, and availability of remotely sensed data can enhance the environment (e.g. Langemeyer et al., 2018) in ways that complement understanding of the distribution and value of CES. In Karasov et al. utilitarian motivations (Himes and Muraca, 2018). There are also (2022), the authors leverage high-resolution land cover models and important methodological advances that support the wider consider­ digital elevation models from state-of-the-art Earth observation data to ation of CES in ecosystem services assessments. New advances in social- investigate patterns in the distribution and content of geolocated photo environmental research based on passively crowd-sourced digital data, data in Estonia. The authors also demonstrate how researchers are notably from social media platforms, has led to novel empirical ap­ increasingly relying on platforms such as Google Earth Engine and proaches to assess CES (Ghermandi and Sinclair, 2019; van Zanten et al., advanced statistical models to analyse datasets with high levels of 2016) that overcome some of the limitations of time- and resource- spatial heterogeneity and spatial autocorrelation. As the global avail­ intensive methodologies, including surveys, interviews and workshops ability of remotely sensed data and geolocated social media data ad­ (Kosanic and Petzold, 2020). vances, the Special Issue papers demonstrate the potential for high With more than half of the world’s population active on social resolution monitoring of the status and trends in CES at different scales. networking sites (Statista, 2022), new opportunities are opening up for Two papers in this Special Issue explore the application of social the study of human-nature relationships through an unprecedented media data to non-market, economic valuation of ecosystem services. In amount of user-generated social media data (Ghermandi et al., 2023). As an investigation of recreational activities in the entire network of na­ such, the analysis of geo-located social media content has become a tional parks in Germany, Sinclair et al. (2020) demonstrate how crowd- promising method for assessing and mapping CES at scale (Ilieva and sourced travel cost models result in consumer surplus estimates that are McPhearson, 2018; Ghermandi and Sinclair, 2019) and to gain insights highly correlated with those generated by more resource-intensive on- into the relationship between humans and nature, crucial for site surveys. Kolstoe et al. (2022) perform a discrete choice experiment https://doi.org/10.1016/j.ecoser.2023.101518 Received 31 January 2023; Accepted 10 February 2023 Available online 20 February 2023 2212-0416/© 2023 Elsevier B.V. All rights reserved. J. Langemeyer et al. Ecosystem Services 60 (2023) 101518 concerning recreational access to the Waikamoi Preserve on the island of models, trained on larger but not location — or CES — specific classified Maui, Hawai‘i. Comparison of the results obtained from a convenience pictures (e.g., ImageNet, Places365). sample of Facebook, Instagram and Twitter users to those of a represen­ Fox et al. (2021), Lee et al (2022) and Kaiser et al. (2021) use tative sample highlights systematic differences in the distributions of automated text annotation based on supplied image tags (using the Willingness To Pay (WTP), and provides technical guidance on how to Google Cloud Vision label detection, and Clarifai general image recognition correct for sample selection bias. tools). Fox et al. (2021) apply this approach to the social-media platform The Special Issue also highlights novel opportunities for CES as­ Reddit, a social news aggregation and discussion orientated social media sessments emerging from the use of crowd-sourced social media data, platform. Comparing their results with data from Flickr, the authors find and with these advances, uncovers numerous methodological chal­ the automated, machine-based annotation approach is promising for lenges. For instance, Zapata et al. (2022) apply a crowd-sourced photo- identifying recreational activities. While Reddit presents a novel data interpretation approach to assess CES in the city of Cali in Colombia. source for CES assessment, there are difficulties in geo-locating these Although hampered by the limited use of Flickr in the Latin American posts when compared to other social media data sources. context, the study clearly outlines the potential for social-media based Similarly, Lee et al (2022) applied Clarifai to Flickr photographs assessments in data-sparse urban contexts. Richards and Lavorel (2022), taken in the Mulde basin in Germany to identify CES related to land­ use a machine learning approach, supported by automated object scape aesthetics (i.e. scenic landscapes) and existence values. While the identification by Google Cloud Vision, to analyse landscape appreciation authors similarly acknowledge the promise of machine learning to in New Zealand. The novelty in their approach is to estimate CES under quickly and accurately analyse image content, they agree that this may future reforestation scenarios. The authors project the impact to CES not indicate the true motivation for the photograph. The need to provisioning given planned changes (i.e., spatial configuration) based “translate” identifiable features and objects to CES is common in auto­ on a spatial analysis of CES occurrence in relation to different social and mated approaches, suggesting the need to develop models more spe­ environmental variables. cifically tailored to CES assessment. Kaiser et al. (2021) apply automated labelling of Instagram, Flickr and 3. Methodological advances VKontakte via Google Cloud Vision and cluster analysis to investigate the CES provided by a prominent river restoration project in Israel. The 3.1. Semantic analyses authors highlight the potential of this novel data source to reveal pre­ viously unrecognised instrumental and relational values associated by While early social-media based CES assessments have broadly relied visitors to the ecosystem restoration project, but also acknowledge that on the interpretation of an image’s content (Ghermandi and Sinclair, photo content does not capture the full range of recreational activities 2019), an important frontier in current research is the additional that are known to take place at the site. consideration of social media text. Schirpke et al. (2021) use a text analysis of tags linked to photographs from Flickr in order to access CES 3.3. Validation from mountain lakes across the European Alps. The study provides a means to compare CES across large and remote region region while As a nascent field of research, many approaches to assess CES from identifying the importance of small lakes for aesthetic experiences. social media data are still in an exploratory stage that requires evalua­ Following a similar approach, applied to a peri-urban park in Barcelona, tion of their efficacy. Validation of results is therefore a current focus in Spain, Calcagni et al. (2022) combined manual content analysis of Flickr social-media based CES research. The work by Dai (2022) refines our pictures with the coding of titles, tags and text into nine CES categories. understanding of potentials and limitations in automated semantic The studies by Schirpke et al. (2021) and Calcagni et al. (2022) highlight analysis by comparing two social media data streams, i.e., reviews from the capacity to identify a wider spectrum of CES, including cognitive and Dianping and location-based text posts from Weibo, with face-to-face spiritual CES, when combining picture content with semantic (text) surveys in urban parks in Shanghai, China. While the results based on analyses. Dianping showed low reliability, strong discrepancies were also observed The broader range of CES that can be assessed through semantic between revealed (Weibo) and stated CES (survey). These discrepancies analyses is also explored by Kong and Sarmiento (2022), who differen­ were particularly related to spirituality, inspiration, and knowledge, and tiate eight CES by means of natural language processing (NLP) and text- indicate the need to further understand reporting biases on social media mining techniques to examine reviews derived from TripAdvisor and platforms. For example, social media might reveal different CES than Google Reviews. This study capitalizes on web-based travel apps that are surveys, but particular platform do match specific CES. promising social media data sources but still rarely used in CES research. Depietri et al. (2021) address the case study of green and blue areas The automated approach based on a phrasal lexicon (in contrast to in the city of Haifa (Israel). As in the above mentioned study by Dai single word lexica) was applied to the El Cajas National Park, Ecuador, (2022), the validation is conducted through the comparison of the re­ and further underlines the potential for future applicability and refine­ sults obtained with social media data streams to those from more ment of semantic analyses for CES. established and more broadly tested methodologies. The authors compare the distribution of seven different CES obtained from the 3.2. Automated image content analysis analysis of photographs from three social networking sites (Flickr, Pan­ oramio and VKontakte) to those from a public participation mapping Another critical research frontier in social-media based CES research (PPGIS) exercise. The results of this study also point to complementarity is the automated image content assessment based on artificial intelli­ in the insights that can be obtained with the two methods. The combi­ gence. Cardoso et al. (2022) explore the potential of deep learning for nation of research approaches and ancillary data sources requires automated analysis of crowd-sourced social media data through the further research as it helps to understand and overcome particular biases training of neural networks. Applying this deep learning approach to the related to social media-based assessments that might allow for their comparative case studies of Peneda-Gerês, Portugal, and Sierra Nevada, systematic accounting in the future. Spain, the authors highlight that the automated identification of CES is not straightforward. While the study notes challenges distinguishing 4. Conceptual advances specific CES categories, the ability to identify the biophysical nature attributes that underpin the generation of CES is real. The authors also The infancy of using social media for the study of CES means that found that their own models, trained on purposely developed and there are still few conceptual frameworks with which to orient and place manually pre-coded data-sets, perform less well than established the innovative work being conducted (Havinga et al., 2020). 2 J. Langemeyer et al. Ecosystem Services 60 (2023) 101518 Langemeyer and Calcagni (2022), aim to remedy this, conceptualising pluralistic portfolio of valuation methods that includes local, contextual, ‘digital relational values’ connecting the concept of relational values in and ethnographic data if we are to truly characterize the breadth and ES with these new digital spheres. Their framework distinguishes three depth of CES. phases. With ‘Seeding’ describing the sharing of CES within virtual The articles in this Special Issue provide evidence of the importance communities as the digital expression of relational nature-values; of crowd-sourced social media for understanding people’s appreciation ‘spreading’ considering the deliberation and multiplication of rela­ and relationship with nature across different spatial and temporal scales tional values within and across virtual communities; and ‘grounding’ as (Bubalo et al., 2019; Lu et al., 2021). Crowd-sourced assessments are the creation of environmental stewardship agency triggered by digital showing that CES are no longer ‘intangible’, as they have often been relational values. Establishing the hypothesis for virtual communities to described in the past and crowd-sourced social media assessments — at least partly — counteract the ‘extinction of experiences’ of nature constitute an important opportunity to further reverse the lack of visi­ (Miller, 2005), the authors call for the analysis of inherent dynamics bility and relevance of CES in research and practice. Thanks to the within virtual social media communities, in order to understand digital relative ease with which social media-based CES analyses can be scaled value shaping processes in relationship with the representation of CES up and updated in a timely fashion, there is a great potential for crowd- on social media and global environmental stewardship. sourced data to support, for instance, the ongoing efforts to extend the current systems of national accounts to adequately reflect changes in 5. Outlook natural capital and ecosystem services (e.g., UN SEEA). Yet, to unfold this potential a conceptual adaptation of the understanding of CES seems This Special Issue demonstrates how recent advances in analytical necessary (Havinga et al., 2020). Technical feasibility should not over­ techniques, including artificial intelligence, have allowed researchers to rule conceptual advances, where it is important to maintain a continued move beyond simple counts of media posts, tweets, or photos, which communication with the broader CES-research community, and a dominated early research in this field, to the analysis of the content of stronger engagement with parallel developments for instance within the images or text, enabling thus a differentiated understanding of CES IPBES. In particular, the emerging field of relational values constitutes a provided. Ultimately, however, more remains to be done before research window of opportunity to further foster the visibility of CES; it also in this field translates into actionable results for decision and policy- opens a new research frontier. Crowd-sourced assessments embed makers. Crowd-sourced CES assessments will require a stronger stan­ important potentials for the observation of digital value-shaping pro­ dardization, with harmonized protocols, established ethical standards, cesses (Calcagni et al. 2019; Ghermandi et al., 2023). Novel methodo­ and including widely agreed upon categories of CES, which allow for logical approaches developed for CES assessments, such as semantic comparison between different case studies — such standardization is analyses, will, in combination with other methods such as social net­ currently lacking. Furthermore, broadly tested methodological ap­ works analysis, likely underpin the empirical development of the novel proaches, including the systematic validation of assessment qualities field of research on digital relational values (Langemeyer and Calcagni, across data streams, including both social media and other empirical 2022). assessments, are necessary. The Special Issue includes several studies that are providing empirical insight into this representation bias — Declaration of Competing Interest however, a stronger standardization and guidelines for validation are urgently needed to enhance the scientific robustness of crowd-sourced The authors declare that they have no known competing financial CES assessments. Cooperation between social-environmental and data interests or personal relationships that could have appeared to influence scientists could aid in developing new ethical and technical standards. the work reported in this paper. The exploration of novel social media data streams that are not yet widely tested for studying CES are crucial in this context. Most studies in Data availability this Special Issue as well as in the wider research field (Ghermandi et al., 2023), are still relying on data from Flickr and, to a lesser extent, Twitter The authors do not have permission to share data. data, due to the ease of data accessibility and the availability of spatial meta-data. With the waning popularity of Flickr and Twitter’s uncertain future, our ability to use social media for CES assessment, and other Acknowledgements sustainability research, is in jeopardy (Ghermandi et al., 2022). Studies in this special issue have highlighted potential alternative crowd- AG and JL are supported by the German-Israeli Foundation for Sci­ sourced data streams, such as Dianping, Reddit, TripAdvisor and Google entific Research & Development #I-1533-500.15/2021. JL is supported Reviews, while other data sources, especially video content (e.g. TikTok, by the EU-H2020 through grants #818002, and #869324. DVB is sup­ YouTube) have not yet received wide attention. However, the opening of ported by the US National Science Foundation (NSF) through grant novel data streams raises important challenges for CES assessments, #73133571. The Special Issue development was additionally supported including challenges with the lack of spatial meta-data and the difficulty through discussions and insights from dedicated sessions during of data integration and analysis in data mashups (Fleming et al., 2014). Ecosystem Service Partnership Conferences, including “Making the To maintain access to this valuable data for CES and sustainability intangible tangible: Using social media data to assess cultural ecosystem research will require high ethical standards on the part of researchers to services” (ESP-EU Conference, San Sebastian, 2018), “Mapping cultural ensure the public that their data is being handled ethically and for the ecosystem service: use of social media to assess cultural ecosystem ser­ common good (Ghermandi et al., 2023). We also acknowledge that in­ vices” (ESP-WORLD Conference, Hannover, 2019), and “Cultural sights derived from crowd-sourced social media data will omit values Ecosystem Services: Time to advance theory and practice” (ESP-EU and perspectives from communities without access to these technologies Conference, Tallinn, 2020). Additional support emerged from the ac­ or who chose not to participate in social media platforms. Much of our tivities of Thematic Group 19 on “Big data and digital communication” current insights into the nature and value of CES are derived from of the Ecosystem Services Partnership. ethnographic and qualitative scholarship, often through deep engage­ ment with local stakeholders, including indigenous communities (Sat­ terfield et al., 2013, Gould et al., 2015). A desire to scale integration of References CES into policy at local and global scales may be enhanced by use of Andersson, Erik, Maria Tengö, Timon McPhearson, and Peleg Kremer. 2015. Cultural mobile technologies and data from web-based applications. However, Ecosystem Services as a Gateway for Improving Urban Sustainability. Ecosyst. Serv. these novel methods need to be considered as part of a diverse and 12: 165–68. https://doi.org/10.1016/j.ecoser.2014.08.002. 3 J. Langemeyer et al. Ecosystem Services 60 (2023) 101518 Bubalo, M., van Zanten, B.T., Verburg, P.H., 2019. Crowdsourcing Geo-Information on media data analysis facilitates assessment of cultural ecosystem services. Ecosyst. Landscape Perceptions and Preferences: A Review. Landsc. Urban Plan. 184, Serv. 53, 101391. 101–111. Kolstoe, S., Naald, B.V., Cohan, A., 2022. A Tale of Two Samples: Understanding WTP Calcagni, F., Maia, A.T.A., Connolly, J.J.T., Langemeyer, J., 2019. Digital Co- Differences in the Age of Social Media. Ecosyst. Serv. 55 (June), 101420 https://doi. Construction of Relational Values: Understanding the Role of Social Media for Sus­ org/10.1016/J.ECOSER.2022.101420. tainability. Sustain. Sci. 14 (5), 1309–1321. https://doi.org/10.1007/S11625-019- Kong, I., Sarmiento, F.O., 2022. Utilizing a Crowdsourced Phrasal Lexicon to Identify 00672-1/FIGURES/5. Cultural Ecosystem Services in El Cajas National Park, Ecuador. Ecosyst. Serv. 56 Calcagni, F., Nogué Batallé, J., Baró, F., Langemeyer, J., 2022. A Tag Is Worth a Thou­ (May), 101441 https://doi.org/10.1016/j.ecoser.2022.101441. sand Pictures: A Framework for an Empirically Grounded Typology of Relational Kosanic, A. and Petzold, J., 2020. A systematic review of cultural ecosystem services and Values through Social Media. Ecosyst. Serv. 58, 101495. human wellbeing. Ecosystem Services, 45, p.101168. Cardoso, A.S., Renna, F., Moreno-Llorca, R., Alcaraz-Segura, D., Tabik, S., Ladle, R.J., Langemeyer, J., Calcagni, F., 2022. Virtual spill-over effects: What social media has to do Vaz, A.S., 2022. Classifying the Content of Social Media Images to Support Cultural with relational values and global environmental stewardship. Ecosyst. Serv. 53, Ecosystem Service Assessments Using Deep Learning Models. Ecosyst. Serv. 54 101400. (February), 101410 https://doi.org/10.1016/j.ecoser.2022.101410. Langemeyer, J., Camps-Calvet, M., Calvet-Mir, L., Barthel, S., Gómez-Baggethun, E., Chan, K.M.A., Satterfield, T., Goldstein, J., 2012. Rethinking Ecosystem Services to 2018. Stewardship of Urban Ecosystem Services: Understanding the Value(s) of Better Address and Navigate Cultural Values. Ecol. Econ. 74, 8–18. https://doi.org/ Urban Gardens in Barcelona. Landsc. Urban Plan. 170, 79–89. 10.1016/j.ecolecon.2011.11.011. Lee, H., Seo, B., Cord, A.F., Volk, M., Lautenbach, S., 2022. Using Crowdsourced Images Chan, K.M.A., Balvanera, P., Benessaiah, K., Chapman, M., Díaz, S., Gómez- to Study Selected Cultural Ecosystem Services and Their Relationships with Species Baggethun, E., Gould, R., Hannahs, N., Jax, K., Klain, S., Luck, G.W., Martín- Richness and Carbon Sequestration. Ecosyst. Serv. 54, 101411. López, B., Muraca, B., Norton, B., Ott, K., Pascual, U., Satterfield, T., Tadaki, M., Lu, Y.i., Zhao, J., Xueying, W.u., Lo, S.M., 2021. Escaping to Nature during a Pandemic: A Taggart, J., Turner, N., 2016. Why protect nature? Rethinking values and the Natural Experiment in Asian Cities during the COVID-19 Pandemic with Big Social environment. PNAS 113 (6), 1462–1465. Media Data. Sci. Total Environ. 777 (14609), 2. https://doi.org/10.1016/j. Chan, K.MA., Gould, R.K., Pascual, U., 2018. Editorial Overview: Relational Values: What scitotenv.2021.146092. Are They, and What’s the Fuss About? Curr. Opin. Environ. Sustain. 35 (December), Miller, J.R., 2005. Biodiversity conservation and the extinction of experience. Trends A1–A7. https://doi.org/10.1016/j.cosust.2018.11.003. Ecol. Evol. 20 (8), 430–434. Depietri, Y., Ghermandi, A., Campisi-Pinto, S., Orenstein, D.E., 2021. Public Participa­ Nowak-Olejnik, A., Schirpke, U. and Tappeiner, U., 2022. A systematic review on sub­ tion GIS versus Geolocated Social Media Data to Assess Urban Cultural Ecosystem jective well-being benefits associated with cultural ecosystem services. Ecosystem Services: Instances of Complementarity. Ecosyst. Serv. 50 (10127), 7. Services, 57, p.101467. Dutcher, D.D., Finley, J.C., Luloff, A.E., Johnson, J.B., 2007. Connectivity with nature as Richards, D.R., Lavorel, S., 2022. Integrating Social Media Data and Machine Learning to a measure of environmental values. Environ. Behav. 39 (4), 474–493. Analyse Scenarios of Landscape Appreciation. Ecosyst. Serv. 55 (June), 101422 Fleming, L., Haines, A., Golding, B., Kessel, A., Cichowska, A., Sabel, C., Depledge, M., https://doi.org/10.1016/J.ECOSER.2022.101422. Sarran, C., Osborne, N., Whitmore, C., Cocksedge, N., Bloomfield, D., 2014. Data Satterfield, T., Gregory, R., Klain, S., Roberts, M., Chan, K.M., 2013. Culture, intangibles mashups: potential contribution to decision support on climate change and health. and metrics in environmental management. J. Environ. Manage. 117, 103–114. Int. J. Environ. Res. Public Health 11 (2), 1725–1746. Sinclair, M., Mayer, M., Woltering, M., Ghermandi, A., 2020. Using Social Media to Fox, N., Graham, L.J., Eigenbrod, F., Bullock, J.M., Parks, K.E., 2021. Reddit: A Novel Estimate Visitor Provenance and Patterns of Recreation in Germany’s National Data Source for Cultural Ecosystem Service Studies. Ecosyst. Serv. 50 (July), 101331 Parks. J. Environ. Manage. 263 (11041), 8. https://doi.org/10.1016/j.ecoser.2021.101331. Statista. (2022d, July 26). 2022. Most popular social networks worldwide as of ranked by Ghermandi, A., Langemeyer, J., Van Berkel, D., Calcagni, F., Depietri, Y., Egarter Vigl, L., number of monthly active users (in millions). Statista.Com. Fox, N., Havinga, I., Jaeger, H., Kaiser, N., Karaxov, O., McPhearson, T., van Zanten, B.T., Van Berkel, D.B., Meentemeyer, R.K., Smith, J.W., Tieskens, K.F., Podschun, S., Ruiz-Frau, A., Sinclair, M., Venohr, M., Wood, S.A., 2023. Broad and Verburg, P.H., 2016. Continental-scale quantification of landscape values using so­ transparent social media data access can boost environmental sustainability cial media data. PNAS 113 (46), 12974–12979. research. One Earth (pre-accepted). Zapata, E., Calcagni, F., Baró, F., Langemeyer, J., 2022. Using Crowdsourced Imagery to Ghermandi, A., Sinclair, M., 2019. Passive Crowdsourcing of Social Media in Environ­ Assess Cultural Ecosystem Services in Data-Scarce Urban Regions: The Case of the mental Research: A Systematic Map. Glob. Environ. Chang. 55 (March), 36–47. Metropolitan Area of Cali, Colombia. Ecosyst. Serv. 56 (10144), 5. https://doi.org/10.1016/j.gloenvcha.2019.02.003. Ghermandi, A., Van Berkel, D., Langemeyer, J., 2022. Twitter’s Totter Must Prompt Research Rethink. Nature 612 (7939), 211. https://doi.org/10.1038/D41586-022- Johannes Langemeyera,b,*, Andrea Ghermandic, Bonnie Keelerd, 04362-4. Derek van Berkele Gould, R.K., Klain, S.C., Ardoin, N.M., Satterfield, T., Woodside, U., Hannahs, N., a Institute of Environmental Science and Technology (ICTA), Universitat Daily, G.C., Chan, K.M., 2015. A protocol for eliciting nonmaterial values through a cultural ecosystem services frame. Conserv. Biol. 29, 575–586. Autònoma de Barcelona (UAB), Edifici Z (ICTA-ICP), Carrer de les Havinga, I., Bogaart, P. W., Hein, L., & Tuia, D., 2020. Defining and spatially modelling Columnes s/n, Campus de la UAB, 08193 Cerdanyola del Vallès, Spain b cultural ecosystem services using crowdsourced data. Ecosyst. Serv., 43, 101091. Department of Geography, Humboldt Universität zu Berlin, Germany Himes, A., Muraca, B., 2018. Relational Values: The Key to Pluralistic Valuation of c Ecosystem Services. Curr. Opin. Environ. Sustain. 35 (December), 1–7. https://doi. Department of Natural Resources and Environmental Management, org/10.1016/j.cosust.2018.09.005. University of Haifa, Israel d Huynh, L.T.M., Gasparatos, A., Su, J., Dam Lam, R., Grant, E.I., Fukushi, K., 2022. Humphrey School of Public Affairs, Center for Science, Technology, and Linking the nonmaterial dimensions of human-nature relations and human well- Environmental Policy, University of Minnesota, United States being through cultural ecosystem services. Sci. Adv. 8 (31), eabn8042. e Ilieva, R.T., McPhearson, T., 2018. Social-Media Data for Urban Sustainability. Nat. School for Environment and Sustainability, University of Michigan, United Sustainab. 1 (10), 553–565. https://doi.org/10.1038/s41893-018-0153-6. States IPBES, 2018. Summary for Policymakers of the Regional Assessment Report on Biodi­ versity and Ecosystem Services for Europe and Central Asia. Bonn, Germany. https:// * doi.org/10.13140/RG.2.2.32932.58248. Corresponding author at: Institute of Environmental Science and Methodological Assessment of the Diverse Values and Valuation of Nature of the Inter­ Technology (ICTA), Universitat Autònoma de Barcelona (UAB), Edifici governmental Science-Policy Platform on Biodiversity and Ecosystem Services. Edited by P. Balvanera, U. Pascual, M. Christie, B. Baptiste, and D. González- Z (ICTA-ICP), Carrer de les Columnes s/n, Campus de la UAB, 08193 Jiménez. Bonn, Germany: IPBES secretariat. https://doi.org/https://doi.org/ Cerdanyola del Vallès, Spain. 10.5281/zenodo.6522522. E-mail addresses: johannes.langemeyer@uab.cat (J. Langemeyer), Kaiser, N.N., Ghermandi, A., Feld, C.K., Hershkovitz, Y., Palt, M., Stoll, S., 2021. Societal aghermand@univ.haifa.ac.il (A. Ghermandi), keeler@umn.edu (B. Benefits of River Restoration – Implications from Social Media Analysis. Ecosyst. Serv. 50 (February), 101317 https://doi.org/10.1016/j.ecoser.2021.101317. Keeler), dbvanber@umich.edu (D. van Berkel). Karasov, O., Heremans, S., Külvik, M., Domnich, A., Burdun, I., Kull, A., Helm, A., Uuemaa, E., 2022. Beyond land cover: How integrated remote sensing and social 4