Skip to main content
NCBI home page
Ambio logoLink to Ambio
. 2019 Oct 12;49(6):1146–1160. doi: 10.1007/s13280-019-01268-w

Using social media images to assess ecosystem services in a remote protected area in the Argentinean Andes

Sebastian Dario Rossi 1,✉, Agustina Barros 2, Chelsey Walden-Schreiner 3, Catherine Pickering 4
PMCID: PMC7128015  PMID: 31606883

Abstract

Social media images are a novel source of data to assess how people view and value the environment. Access to these images is often free, the volume and spread of images is expanding rapidly and hence they are an increasingly valuable source of data complementing and expanding on other data. Recently, coding images has been used to assess sociocultural values relating to ecosystem services including those provided by national parks. To further explore the use of social media images, including for remote environments, we analysed the content of images posted to Flickr by people visiting a national park that contains the highest mountain in the southern hemisphere, Mt. Aconcagua, in Argentina, South America. The saliency of aesthetic landscapes, recreation, social relations and fresh-water provisioning was high across the 334 images posted to Flickr by 104 visitors to the Park, but location mattered. Images from visitors in easily accessible day-use areas were significantly more likely to include content that reflects biodiversity-existence, geology, culture and education services, while the content of images from remote areas was more likely to reflect social relations and fresh-water provision services. Comparisons of the content of images from Mt. Aconcagua with other studies in Europe, South America, Asia, Africa and Australia highlight similarities and differences in people’s views of the diversity of locations, but also the benefits and limitations of user-generated social media content when assessing environmental and management issues.

Keywords: Aconcagua Provincial Park, Content analysis, Cultural ecosystem services, Flickr, Social media

Introduction

Nature, including in national parks, provides a myriad of ecosystem services to humans (Costanza et al. 1997; Millennium Ecosystem Assessment 2005). These include provisioning (e.g. fresh water, food, raw material, genetic resources), regulating and maintaining (e.g. climate, water quality, erosion, pollination, soil formation, nutrient cycling, primary productivity, biodiversity) and cultural ecosystem services (e.g. recreation and tourism, aesthetics, education and research, spiritual and religious, and cultural identity) (Millennium Ecosystem Assessment 2005; Stolton and Dudley 2014; Haines-Young and Potschin 2017).

Mountain areas provide many of these services and account for 32% of all protected areas globally (Grêt-Regamey et al. 2012; World Commission on Protected Areas 2019). They are the source of the worlds’ major rivers and play a critical role in water cycles by capturing moisture from the air, storing it as snow or ice providing water for agriculture, communities and industries downstream (Hamilton and McMillan 2004; Grêt-Regamey et al. 2012). Mountains are also important centres of biological diversity, providing refuge for many rare and endemic plants and animals, and conserving rich assemblages of species and ecosystems (Hamilton and McMillan 2004). Many contain sites of great cultural, sacred or spiritual significance (Bernbaum 2006) and are often popular tourism and recreation destinations (Debarbieux et al. 2014). Despite this, research on mountain parks, including the ecosystem services they provide and how these are valued by visitors, is sparse, particularly for remote areas where research, monitoring and management is often limited both by resources and capacity (Hamilton and McMillan 2004; Grêt-Regamey et al. 2012).

An increasingly important challenge for the management of mountain parks is assessing how visitors view and value these environments including the ecosystem services they provide (Martinez Pastur et al. 2015; Richards and Friess 2015; Leung et al. 2018; Oteros-Rozas et al. 2018). Previously, data on the views and values of visitors to parks have been collected using surveys, choice experiments, community consultation and focus groups, among others (Newsome et al. 2012; Stolton and Dudley 2014). More recently, data from social media have been used to address important environmental and management questions (Ghermandi and Sinclair 2019), including identifying how visitors view and value natural environments and potential ecosystem services (Oteros-Rozas et al. 2018; Teles da Mota and Pickering 2018; Rosário et al. 2019).

Social media data have been used to assess the relative popularity of different parks (Wood et al. 2013; Tenkanen et al. 2017), as well as spatial and temporal patterns of visitor use (Norman and Pickering 2017; Walden-Schreiner et al. 2018; Norman et al. 2019). Recently, researchers have started using social media data to evaluate how people value natural environments (Calcagni et al. 2019; Ghermandi and Sinclair 2019). Much of this research leverages data from the image-sharing platform Flickr (Calcagni et al. 2019; Ghermandi and Sinclair 2019). For instance, data from Flickr were used to assess spatial patterns of ecosystem services in different landscapes based on where images were taken (Martinez Pastur et al. 2015; Van Zanten et al. 2016; Langemeyer et al. 2018). Studies have also started to analyse the content of image to gauge how people value natural environments (Richards and Friess 2015; Thiagarajah et al. 2015; Angradi et al. 2018; Hausmann et al. 2018), agricultural landscapes in Europe (Oteros-Rozas et al. 2018; Tieskens et al. 2018) and countries (Stephchenkova and Zhan 2013; Richards and Tunçer 2018).

Although analysis of user-generated content from social media is still a novel approach for addressing environmental and management issues (Sherren et al. 2017), it has already provided important insights, particularly in Europe and North America. Yet, research from South America is limited (Calcagni et al. 2019; Ghermandi and Sinclair 2019). The few studies from the region include Martinez Pastur et al. (2015) who conducted a spatial analysis of cultural ecosystem services in southern Patagonia using the text and location of images posted on the image-sharing platform Panoramio. They found that aesthetic, existence, local identity and recreational values were common in text, while locations with water bodies and tourism opportunities were important based on the geolocation of images. Martinez-Harms et al. (2018) used geolocated images from Flickr to assess visitor use (i.e. photo user days) of 65 parks in Chile and compared the results with visitor numbers for 32 of the parks. They assessed the effects of distances and social inequality on visitation rates using the Flickr users’ home-location, obtained from users’ public-profiles, and local-government socioeconomic data. Results indicated that people from lower socioeconomic areas tended to visit parks closer to them, while people from wealthier areas travelled further. Walden-Schreiner et al. (2018) used geolocated images for Aconcagua Provincial Park, in Argentina, to look at temporal and spatial patterns of use. This paucity of research using social media data in South America also reflects the broader issue of a dearth of research on monitoring and managing park-visitors in the region compared to Europe and North America (Barros et al. 2015a; Pickering et al. 2018a).

To further evaluate how social media data can be used to assess how people view and value natural environments, we analysed user-generated content created by visitors to the highest mountain in the southern hemisphere, Mt. Aconcagua in Argentina. The content of images posted to Flickr by visitors to Aconcagua Provincial Park was used: (1) to assess the relative popularity of different aspects of the Park, and how they relate to ecosystem services, and (2) how the content of the images differs between the easily accessible day-use areas and the more remote areas of the Park that are only accessible to adventure-based tourists. Research has shown that the activities, values and attitudes of day-use visitors to parks can differ from those of adventure-based tourists who access more remote areas, with important implications for management (Buckley 2006; Abbe and Manning 2007; Pierce and Manning 2015). We also compared the results from Aconcagua Provincial Park with those from 12 other locations where the content of images from Flickr has also been used to assess visitor views and values and associated ecosystem services.

Materials and methods

Study area

This study assessed images from Aconcagua Provincial Park (~ 700 km2), an IUCN Category II protected area in Argentina containing Mt. Aconcagua (6962 m a.s.l.). The Park has high aesthetic, conservation, biodiversity and cultural values and is the major water catchment in the region providing drinking water and irrigation to over one million people. However, monitoring and management of the Park is severely limited by resources and research capacity (Barros et al. 2015a).

As the Park is easily accessible from the international highway linking Mendoza in Argentina, with Santiago in Chile, it is relatively popular with approximately 41 000 visitors during summer (Barros et al. 2015a). Most visitors spend less than a day in the Park (35 000 people), going on short walks or sightseeing in the Horcones Valley. This valley is the main entrance to the Park and has a range of infrastructure including a short public road, carparks, visitor centre, self-guided 2-km circuit trail, paleontological and archaeological sites and a park ranger station with helicopter pad (Barros et al. 2015a; Walden-Schreiner et al. 2018) (Table 1; Fig. 1). In addition, about 6000 people in summer access more remote areas of the Park on multi-day trips, including attempting to summit Mt. Aconcagua (15–20 day trip) (Barros et al. 2015b). During autumn, winter and spring, remote areas of the Park are closed and visitation is limited to the day-use areas of the Horcones Valley, including the visitor centre and short walks. Previous research using geolocated images from Flickr found that most images were taken during summer on trails, at campsites, or at buildings such as the visitor centre, with very few images taken in winter, and they were almost exclusively in the Horcones Valley (Walden-Schreiner et al. 2018).

Table 1.

Important natural, cultural, tourism and recreational features within the day-use and remote areas of Aconcagua Provincial Park, Argentina

Feature Day-use areas Remote areas
Visitor centre Permits, toilets and interpretive material –
Interpretative trail 2 km trail exploring natural and historic features –
Historic buildings 3 (Puente del Inca Hotel Ruin, Crucified-Christ in Park entry, Hanging Wooden Bridge) 5 (Incas Ruin, Ibañez Refuge, Colombia Refuge Ruin, Hotel Plaza de Mulas, Berlin Refuge)
Lakes Two lakes (Horcones and Espejo) Two glacier lakes near Plaza de Mulas Campsite
River Two rivers (Horcones and Vacas) Four rivers (Horcones, Horcones Superior, Horcones Inferior, Vacas) with trails next to and crossing the rivers, also glacial feed streams and springs
Glaciers 1 Visible glacier of 6 km2 from day-use area (Horcones Inferior Glacier) 242 Glaciers, total surface area of 82 km2
Alpine meadows 10 Meadows. Total surface area of 0.10 km2 40 Meadows. Total surface area of 0.43 km2
Distinctive rock formation Erratic-blocks, coloured-sedimentary formation, fossil shells incrusted in big rock Coloured-sedimentary formation of Aconcagua west face, landslides, iconic formations that climbers clearly identify and use as meeting/resting points
Campsites and park rangers Basic-campsite next to car-park and the park rangers station including helicopter and rescue facilities Seven campsites and three park ranger stations on Horcones route and five campsites and three park ranger stations on Vacas route to summit
Recreation activities and services Includes trekking, walking, birdwatching, picnic, sightseeing and helicopter use to move rangers and medical emergencies Trekking and climbing, medical service, rescue patrols, art gallery at 4300 m a.s.l. and helicopters used to move rangers and for medical emergencies
Traditional cultural activities Muleteers (mule/horse riders)
Native flora 120 Species
Native fauna 60 Bird species and 11 mammals including puma, condor (Vultur gryphus), zorro colorado (Lycalopex culpaeus) and lamas (Lama guanicoe)
Open in a new tab

Fig. 1.

Fig. 1

Open in a new tab

Aconcagua Provincial Park in Argentina showing location of images from Flickr taken in the day-use area and remote areas of the Park. More images were taken within day-use areas than in remote areas (a), but images from remote areas were more popular online (b)

Content analysis of images

Data were obtained from the popular social media platform Flickr, which has over 65 million users, and over 10 billion publicly available images on the platform (Smith 2018). Flickr’s Application Programming Interface (API) and an R script were used to retrieve metadata for publicly available geotagged images taken in the Park between November 2010 and July 2018. Metadata leveraged in this analysis included when the image was taken, when it was uploaded to Flickr, owner ID, number of views, image URL, camera type and geolocation where the image was taken (i.e. longitude and latitude). Data were written to a comma-separated values (csv) file and imported into Excel for analysis.

Before analysing the content of images, the total number of images per visitor were capped. Specifically, the 10 most-viewed images per user were selected. This was done to ensure that results did not over-represent the views and values of prolific posters of images from the Park. Then images were divided into those taken within day-use areas of the Park or the remote areas using ArcGIS (version 10.5) combined with management zonation GIS layers for the Park. Images were also reviewed to ensure they were taken in the Park, that they were still publicly available on Flickr and that they were not duplicates posted by the same visitor. The final 334 geolocated images (Fig. 1) were then viewed online using the image URL from the metadata and their content classified by one person using variables reflecting features such as landscapes and geology, number of people, activities, infrastructure, flora and fauna, using a quantitative metonymic approach, where the features or attributes in the image were classified based on their apparent content (Stephchenkova and Zhan 2013; Sherren et al. 2017). Using this approach, it was then possible to statistically compare the content of images taken in the day-use area—Horcones Valley—with those taken in more remote areas of the Park. Each image could represent multiple categories, so an image taken in summer, that included the main summit, glaciers and four hikers on the side of a trail looking at a camera represented nine categories including: summer, aesthetic, mountain range, glacier, people present, number of people, trails, looking at camera and activity (e.g. hiking and resting). Data for specific variables including landscape, snow, storm, sunny, mammals, birds, native flora, mountains, boulder, lake, river, creek, glacier, hotel, railway, mules, walking, horse riding, birdwatching, among others were then combined to calculate the number of images that had content relating to different types of ecosystem services in the Park (Table 2). This was done using criteria/categories from previous studies with similar aims (Martinez Pastur et al. 2015; Richards and Friess 2015; Oteros-Rozas et al. 2018), which were then slightly adjusted to reflect specific aspects of the natural environment of the Park, and hence the types of ecosystem services that it may provide.

Table 2.

The types of ecosystem services relating to the specific types of content of images on Flickr taken in Aconcagua Provincial Park, Argentina

Ecosystem service Definition Examples in Aconcagua
Aesthetic Sites of particular beauty Natural landscapes
Existence/biodiversity The intrinsic value of nature and biodiversity. Belief that all species are worth protecting regardless of their utility Fauna and flora, alpine meadows
Provision: geological Distinctive/iconic rock formations
Provision: fresh water Snow on ground, rivers, lakes, alpine meadows, glaciers
Culture and heritage Sites relevant to local history and culture Historic buildings, historic railways, folk-muleteers, mules
Recreation

Outdoor recreational activities

Tourism and recreational infrastructure

Activities: trekking, walking, birdwatching, sightseeing, camping, climbing

Infrastructure: visitor centre, trails, campsites, tents, refuges
Education Sites that widen the knowledge about plants and animal species Information signs, visitor centre, interpretative trails
Social relations Sites serving as meeting points with people/friends Hotel Plaza de Mulas, Campsites, visitor centre, tents, refuge and selfies and group photos
Open in a new tab

To assess relationships among variables, exploratory Categorical Principal Component Analyses (CATPCA) were performed in the statistical software SPSS (version 25). The CATPCA analysis is analogous to Principal Component Analysis (PCA), except that it is suitable for categorical variables (i.e. nominal or ordinal) and non-linear relationships. In CATPCA, variable categories are transformed into numerical values and analysed as conventional linear PCA (Linting, et al. 2007). Differences in the image content from the day-use and remote-use areas of the Park were also statistically tested using χ2 and ANOVA tests at 95% confidence intervals. Results from Aconcagua Provincial Park were then compared with results from 12 other locations where the content of Flickr images has also been analysed to assess potential ecosystem services using similar methodology.

Results

There were 902 publicly shared geolocated images taken in the Park and posted to Flickr by 111 visitors (Table 3). Of these, most visitors (77%) uploaded less than five images, with half of the visitors uploading only one image. Six visitors uploaded more than 30 images, including 1 person who uploaded 151 images from the Park. After capping the number of images per visitor at 10 and confirming that the content of the images was within the Park, 334 images posted by 104 visitors remained for further coding. On average, these images were viewed 287 times each (SD ± 830) with a total 95 899 views (Table 3). Most of the images were taken in summer (64%), or autumn (21%), with few images in winter (4%) or spring (11%). Those posting the images to Flickr about the Park tended to use dedicated digital cameras (78%), although some (22%) used mobile phones with cameras. The average number of images of the Park per year was 37, with a minimum of 5 taken in 2010 and a maximum of 72 in 2012.

Table 3.

Number of geotagged images taken in Aconcagua Provincial Park, Argentina and uploaded to Flickr between November 2010 and July 2018. To avoid bias, images were capped at a maximum of 10 per visitor

Total Analysed Day-use Remote Summer Autumn Winter Spring
Number of images 902 334 214 (64%) 120 (36%) 214 (64%) 69 (21%) 14 (4%) 37 (11%)
Number of visitorsa 111 104 79 (76%) 33 (32%) 60 (58%) 28 (27%) 10 (10%) 16 (15%)
Total views 216 066 95 899 33 535 62 364 79 458 7670 2201 6570
Average views per image (min–max) 240 (0–6399) 287 (0–6399) 157 520 371 111 157 178
Average images per visitor 8 3 3 4 4 2 1 2
Open in a new tab

aPercentages and number of visitors when divided into categories can add up to over 100% and sum up to over 104 as some visitors posted images from both day-use and remote areas of the Park and took images in different seasons

Natural landscapes appeared in many images (62%), including mountains (39%) and glaciers (43%) (Table 4). In contrast, there were few images of specific plants, animals or buildings and cultural features such as the visitor centre, park entrances, or ranger facilities. Ecosystem services associated with the content of the images included aesthetic (87%), fresh-water provisioning (65%) and recreation and tourism activities (64%). Although the Park is known for high biodiversity and significant cultural and geological values, images with content related to biodiversity (20%), cultural heritage (11%) and iconic geological features (11%) were not common (Table 4).

Table 4.

Results of the content analysis including potential ecosystem services represented in images taken by visitors to Aconcagua Provincial Park, Argentina. Clear = χ2 test could not be applied, due to small numbers in one of the cells, but differences between day-use and remote-use areas were clear

Content of images Total n = 334 Day-use area n = 214 Remote area n = 120 χ2 Tests
Ecosystem services related to the content of images
 Aesthetic 289 (87%) 188 (88%) 101 (84%) 0.895, p = 0.344
 Biodiversity-existence 68 (20%) 53 (25%) 15 (13%) 7.135, p = 0.005
 Provision geology (rock formations) 37 (11%) 33 (15%) 4 (3%) Clear
 Provision fresh water 217 (65%) 137 (64%) 80 (67%) 0.237, p = 0.626
 Culture and heritage 37 (11%) 23 (11%) 14 (12%) 0.066, p = 0.797
 Recreation 215 (64%) 137 (64%) 78 (65%) 0.032, p = 0.857
 Education 102 (31%) 97 (45%) 5 (4%) Clear
 Social relations 148 (44%) 74 (35%) 74 (62%) 22.861, p < 0.001
Detailed features in images
 Landscape 280 181 99 0.245, p = 0.364
 Mountain (singular) 131 97 (45%) 34 (28%) 9.314, p = 0.002
 Mountain range 145 83 (39%) 62 (52%) 5.193, p = 0.015
 Fauna 26 16 10
  Mules and hares 9 5 4
  Birds 17 11 6
 Flora (only natives shown) 6 6 0
 Sources of fresh water 196 128 68 0.314, p = 0.328
 River 15 9 6
 Creek 3 3 0
 Lake/lagoon 48 45 3
 Glacier 142 80 (37%) 62 (61%) 6.418, p = 0.008
 Meadow 42 37 5
 Snow on ground 45 16 29 18.372, p < 0.001
Infrastructure and facilities
 Hotel Base Camp Mulas 2 2
 Railway 13 13
 Historic building 15 13 2
 Visitor centre 5 5
 Carpark 3 3
 Park entry 6 6
 Trails 101 91 10 42.605, p < 0.001
 Durazno Bridge 1 1
 Aconcagua view point 22 22
 Summit-from-summit 0
 Campsites 18 0 18 Clear
 Tents 32 0 32 Clear
 Mountain refuges 7 0 7
 Helicopter 2 1 1
 Park rangers 2 1 1
Recreational activities
 Trekking 52 47 5 Clear
 Walking/hiking 32 32 0 Clear
 Horse riding 3 2 1
 Photography 7 5 2
 Birdwatching 3 3 0
 Sightseeing 18 6 12 Clear
 Yoga/stretching 3 1 2
 Camping 23 0 23 Clear
 Posing 59 30 29 5.444, p = 0.015
People
 Selfie 6 3 3
 Group looking at camera 18 8 10 3.184, p = 0.065
 Group walking off camera 50 28 22 1.664, p = 0.130
 Portrait looking at camera 35 19 16 1.627, p = 0.138
 Portrait looking away 12 7 5 0.178, p = 0.444
Number people in focus
 #1 61 33 28 ANOVA F = 0.718, p = 0.399
 #2 20 13 7
 #3 6 2 4
 #4 7 4 3
 #5 2 1 1
Number in background
 #1 50 22 28 ANOVA F = 1.284, p = 0.187
 #2 26 15 11
 #3 10 6 4
 #4 7 4 3
Open in a new tab

There were differences in the content of images between the day-use and remote areas of the Park (Table 4; Figs. 2, 3), reflecting variation in the locations and how they are valued by visitors (Table 1). For instance, images in the day-use areas were more likely to include content relating to biodiversity-existence, geology and education ecosystem services (Fig. 3). They were also more likely to focus on features such as the single peak of Mt. Aconcagua, historical structures, recreational trails and activities including walking and birdwatching. In contrast, images from remote areas were more likely to contain content relating to social relations and fresh-water provisioning ecosystem services (Fig. 3). For example, images were more likely to show mountain ranges, glaciers, rivers, campsites, tents and people looking directly at the camera (Table 4).

Fig. 2.

Fig. 2

Open in a new tab

Distribution of potential ecosystem services relating to the content of the 334 images taken by visitors in the day-use and remote areas of Aconcagua Provincial Park, Argentina based on the results of Categorical Principal Component Analysis. Variable projections represent the relationship among them with those close together positively related and variables at 90° angle not related to each other. Cronbach’s α = 0.752 and total variance explained 37%

Fig. 3.

Fig. 3

Open in a new tab

Features seen among 334 images taken in day-use and remote areas of Aconcagua Provincial Park based on the results of Categorical Principal Component Analysis. Variables’ projections represent the relationship among them with those close together positively related and variables at 90° angle not related to each other. Cronbach’s α = 0.836 and total variance explained 34%

Discussion

Images from Aconcagua Provincial Park posted on Flickr represent many of the places, objects and experiences valued by visitors and reflected what some visitors share publicly. This study adds to the emerging literature about how images and texts from social media expand the types of data available to assess visitors’ views and values of natural environments, including in parks. This includes how image content can relate to potential ecosystem services (Calcagni et al. 2019; Ghermandi and Sinclair 2019), especially for regions where other data-sources are limited (Barros et al. 2015a; Pickering et al. 2018a).

Differences in visitors’ views and content relating to ecosystem services in Aconcagua Provincial Park

The differences between the images taken in Aconcagua Provincial Park within day-use and remote areas are likely to reflect differences in what can be seen in these areas, what can be visualized/interpreted from the content of the images and potential differences in what visitors to the two areas value. Other studies have also found spatial differences in potential ecosystem services based on image content analysis (Stephchenkova and Zhan 2013; Richards and Friess 2015; Oteros-Rozas et al. 2018; Richards and Tunçer 2018; Clemente et al. 2019).

Although services, including recreation, were equally important in both day-use and remote areas of Aconcagua, the types of recreation activities differed. Walking was more common in day-use areas and camping was only captured in images from remote areas. Interestingly, some other activities were rarely shown in images including climbing, despite many visitors to remote areas attempting to summit Mt. Aconcagua. Some features were location specific. For example, education services were represented by images of historic buildings, the visitor centre, interpretive trails, the main summit view point, and other interpretive signs, and these features were nearly entirely restricted to the day-use areas.

There were few images from Aconcagua Provincial Park of animals in either the day-use or remote areas. The few animal images were of introduced mammals, specifically mules and hares. The absence of native mammals in images may be because guanaco (Lama guanicoe), zorro colorado (Lycalopex culpaeus), puma and other species avoid areas used by visitors (Barros et al. 2015b). Furthermore, although the Park has over 90 species of birds, there were very few images of the iconic Andean condor or other large birds, probably due to the difficulties in photographing them. Where there were images of birds, they were mostly small birds that nest on the ground, close to campsites.

There was a greater emphasis on images showing social relations in remote areas, despite far fewer people accessing these areas. It is possible that adventure tourists may focus on people in their images as they want to capture a shared experience in remote locations, where adventure tourists are often motivated by connectedness in extreme conditions as part of transformative experiences (Buckley 2006).

Comparison among studies using social media to assess potential ecosystem services

The content of Flickr images has been used to evaluate how people view and value nature in Africa, Europe, Asia, Australia and North and South America (Calcagni et al. 2019; Ghermandi and Sinclair 2019). This includes studies at a range of spatial scales from as small as 1.4 km2 for areas of mangroves in Singapore to the whole country of Peru, and covers a range of landforms and land uses such as mountains, coastal areas, savannahs, agricultural and urban areas (Table 5).

Table 5.

Summary of the results of research across 13 locations where the content of Flickr images was used to assess the views and values of visitors and how image content relates to potential cultural ecosystem services and other features of the locations. NR = not recorded or not relevant to the location. NA = not available, as the categories used in the research were defined in such a way that the value could not be extracted from the paper. Where a paper used images from several platforms (Stephchenkova and Zhan 2013; Hausmann et al. 2018; Oteros-Rozas et al. 2018), only the data from Flickr images were included here. Agri = agriculture. Approximate “~” as in the paper no exact values are given but a bar-figure represents them

This study Stephchenkova and Zhan (2013) Pickering et al. (2018b) Pickering et al. (in press) Richards and Friess (2015) Richards and Tunçer (2018) Hausmann et al. (2018) Clemente et al. (2019) Oteros-Rozas et al. (2018)
Location Aconcagua Provincial Park Peru The Spit Gold Coast Kosciuszko Alpine Area Four mangrove sites Singapore Kruger National Park Sudoeste Alentejano and Costa Vicentina Natural Park Lake Peipsi Lesvos Madrid Uppland Obersimmental
Landform/land use Mountain Variable Coastal Mountain Coastal Mostly urban Savannah Coastal Agri. Agri. Agri. Agri. and urban Alpine pasture
Size of area (km2) 700  1.3 million 2 200 1.4 722 19 485 605 162 86 183 47 334
Country Argentina Peru Australia Australia Singapore Singapore South Africa Portugal Estonia Greece Spain Sweden Switzerland
Data collected Geodata Text Text Geodata Geodata Geodata Geodata Geodata Geodata for all five sites
Number of images 334 500 493 656 762 23048 4597 1378 22 28 200 15 200
Number of categories 79 20 70 69 7 7 57 6 25
More than one category per image? Yes Yes Yes Yes No No Yes No Yes for all five sites
Landscape (aesthetic) (%) 87 NR NA 50.4 20.5 14.4 ~ 60 NA NA NA NA NA
Rock formations (%) 11 NR NR 30 NR NR NR ~ 25 18.2 24.1 40 4 56
Water bodies (%) 65 NR 82.7 11.4 NR NR NR ~ 28 59.1 82.8 20 22.6 24
Plant and/or animal (biodiversity-existence) (%) 20 18.2 18.6 13.9 35.3 23.6 84.5 ~ 19 NA NA NA NA NA
People (%) 36 30.6 45 29 NA 8.5 NA NR 9.1 3.5 13.3 21.1 18.5
Culture/heritage/history (%) 11 47.6 NA 1.2 0.3 NA 0.3 ~ 5 18.2 17.2 20 24.1 4
Recreation, tourism and sports activities (%) 27 12.8 58 32 13.6 3.2 5 ~ 25 40.9 41.4 26.7 44.7 32.5
Recreation and tourism facilities (%) 56 4.2 NA 42 9.4 NA 1.6 NR
Education/research (%) 31 10.8 NR ~ 1
Social (%) 44 NA NA 50 17.2 20 30.2 6
Spiritual (%) NR NR NR NR NR NR ~ 2 13.6 3.5 6.7 20.1 2
Open in a new tab

Among the studies, there was variation in how data were obtained and coded. For instance, some studies retrieved image-metadata based on keywords within the image tags while others retrieved images located within defined polygons based on the geolocation of the images (Table 5). In some studies, images were only assigned to single categories, while in others single images could represent multiple categories. Most studies coded images manually, except Richards and Tunçer (2018) who used machine learning. In some cases, it was not possible to directly equate the categories used in the different studies due to differences in definitions, coding and locations. However, preliminary comparisons could be made relating to the frequency and types of values associated with ecosystems services represented in social media images from different landscapes.

Landscapes were common both in the content of images among locations and in coding categories within studies (Table 5). In several cases, landscapes were treated as equivalent to aesthetic values including in this study, Clemente et al. (2019), Richards and Tunçer (2018) and Pickering et al. (in press) (Table 5). Landscape features, such as mountains, lakes or oceans, were popular in images despite often accounting for a small proportion of the total area of the location. For instance, in Aconcagua Provincial Park, glaciers and lakes were popular in the images (42% and 14%, respectively) even though they account for only 15% of the Park’s land area. This is also the case for the peak of Mt. Aconcagua, which was common in images (39%) although it represents a relatively small area and it can only be seen from a few places in the Park.

The presence of people in images reflects differences in visitors’ values and perceptions of the places they visit (Stephchenkova and Zhan 2013). In more remote protected areas where few people live, most images show fellow tourists, and hence the presence of people in images is equivalent to tourism activities (e.g. Aconcagua Provincial Park, Kosciuszko National Park, Kruger National Park). The dominance of images of tourists in these parks varies and can be quite low even though visitors may be sharing places with hundreds of others such as in Kosciuszko National Park (Pickering et al. in press) (Table 5). In urban areas, images of people may still predominantly represent recreation (e.g. Spit, Gold Coast), but they can also show locals engaged in day-to-day activities (e.g. Peru). In some cases, images of locals may represent cultural values reflecting apparent difference in ideas, customs or social behaviour with those of tourists (Stephchenkova and Zhan 2013). In some studies, images were coded to indicate if people were looking at the camera and hence could be recognized (e.g. Aconcagua, Spit, Kosciusko, Kruger). This may reflect people’s personal history, where seeing a person in a place is important. Although authors included ‘selfies’ in most image classifications, there were few if any selfies in most studies (e.g. Aconcagua, Spit, Kosciusko). Such images may be more common on other social media platforms such as Facebook, Instagram and Snapchat where images are more likely to be directly shared within social groups or used to market popular identities.

For recreation and tourism activities there are obvious differences among locations. Interestingly, the frequency with which an activity is shown in images does not always match reality, with some types of activities more common in images than in reality, or missing entirely from images (Pickering et al. 2018b). For instance, there were few if any images within the visitors centre at Aconcagua Provincial Park despite its popularity with tourists, while for the Spit, Gold Coast, of the 35 recreation activities known to occur on the Spit, many were not seen in any images (Pickering et al. 2018b).

Some studies separate recreation into activities and facilities, such as in Aconcagua, Spit, Mt. Kosciuszko and Peru, while others may have recorded activities and facilities separately to start with, but then combined them when calculating overall human activities/tourism (Hausmann et al. 2018). A wide range of different types of facilities were recorded across the different studies including accommodation (i.e. hotels, huts, tents), transport (i.e. trolley cars, tour boats, trails, boardwalk) and others (i.e. viewpoints, visitor centres, information boards, surf lifesaving structures, pools, toilets) (this study, Stephchenkova and Zhan 2013; Richards and Friess 2015; Pickering et al. 2018b; and Pickering et al. in press). Culture, heritage and history services were coded in several studies, but they represent different contents including different types of historical buildings, ruins, monuments, monasteries, fortresses and/or burial sites. Sometimes, historical services was combined with cultural services when there were images of people engaged in cultural and other activities reflecting their way of life.

Most studies assessed if images included animals and plants (Table 5). This was considered to reflect existence values, or nature appreciation in some studies. The frequency with which images presented specific plants and animals varied among locations ranging from 14% for the Kosciuszko Alpine Area in Australia to 84.5% in images from Kruger National Park in South Africa. A diversity of animals was shown within and among locations, but most often they were larger mammals or birds, while images of small mammals, reptiles, amphibians or arthropods were rare (Table 5).

The frequency of images showing animals partly depends on how easy it is to capture an image of the animal and this is affected by the size of the animal and its behaviour (Castley et al. 2013). For example, bigger herd mammals active during the day are easier to photograph than smaller, solitary and/or nocturnal species (Castley et al. 2013). For Flickr and other social media platforms, the frequency of different types of animals in images is affected by the preferences of those posting and sharing images and reflect the animals ‘value’ in a specific context (Willemen et al. 2015; Hausmann et al. 2018). For instance, images from protected areas can often include iconic species such as elephants and lions in South Africa (Hausmann et al. 2018) or llama, alpaca and condors in Peru (Stephchenkova and Zhan 2013). They may also show animals that are integral to tourism such as mules in Aconcagua. In agricultural and urban areas, images may include domesticated as well as native animals (Stephchenkova and Zhan 2013; Oteros-Rozas et al. 2018). This could include recreation images such as dogs being taken on walks (Spit, Gold Coast), or reflecting the nature of landscapes such as images of livestock (Oteros-Rozas et al. 2018). Individual plants tend to be uncommon in images in contrast to broad vegetation types (e.g. forests, mangroves, lawns, pastures). As with animals they can have different meaning, such as native plants in protected areas (Aconcagua, Kosciuszko), compared to agricultural landscapes which may represent traditional land uses (Oteros-Rozas et al. 2018).

Many ecosystem services would be hard to capture in images and/or are difficult to code and categorize. This remains an important limitation of the method. This limitation applies particularly to supporting ecosystem services such as soil formation, nutrient cycling, primary productivity, genetic resources, pollination, wellbeing and cultural services such as peace. It is also possible that visitors do not know about these types of values, or do not value them even if they know about them. There may also be issues due to a disconnect between the ‘gaze’ of the person taking and posting the image and that of the coder (Stephchenkova and Zhan 2013).

An obvious example is spiritual values that are hard to code when it is not associated with specific content such as objects (e.g. churches, temples, monuments) or activities (e.g. religious ceremonies, people in prayer). Many locations embody spiritual values beyond those including sacred landscapes, mountains and rivers, caves and individual trees or where people have a sense of attachment to specific places (Oteros-Rozas et al. 2018; Wartmann and Purves 2018). In some cases, there can also be overlaps between spiritual, cultural and historical services wherein the same object, such as a tomb or temple, which may reflect culture to some people and spiritual values to others. In coding the content of images, we and other researchers have used a metonymic interpretation of images that reflect specific ‘norms’ of interpretation, while Flickr-users taking and sharing images may ‘see’ the images from a metaphoric perspective ascribing meaning not apparent to the coder (Stephchenkova and Zhan 2013).

Management implications

Social media as a mean of communication is increasingly popular with people sharing news and experiences, as well as using social media to gain information about places and plan their trips (Miller et al. 2019). This provides managers with opportunities to reach a broader audience, communicating their purpose, functions and opportunities to local residents and tourists. Protected areas managers and tourism agencies can use the knowledge gained through the analysis of user-generated content, including images and text, to better understand Cultural Ecosystem Services demands and preferences (Wolff et al. 2015). At a finer scale, managers can also better understand where and what visitors look for and share within parks helping to tailor education programs and their own social media posts. For instance, in Aconcagua Provincial Park, managers could use information from the Flickr images to encourage more responsible environmental behaviours and safety measures to current, virtual and future visitors (Miller et al. 2019).

Benefits and limitations

Content analysis of social media images enables the rapid assessment of some visitors’ perceptions of landscapes. Data from social media can be collected remotely, quickly and cheaply, and provides both spatial and temporal data about visitor use, views and values (Calcagni et al. 2019; Ghermandi and Sinclair 2019). This information is hard to obtain even in easily accessed and well-funded protected areas and is extremely limited for more remote parks and locations. The types of values and preferences reflected in the content of images can match those reported in surveys of visitors (Hausmann et al. 2018).

There are, however, limitations that need to be considered. Only a very small proportion of people visiting parks post images to Flickr (Hausmann et al. 2018), with Flickr and other social media sites dominated by younger, wealthier and more highly educated people (Smith and Anderson 2018). Hence, images on Flickr only represent the ‘views’ and/or ‘interests’ of some visitors to parks and the content can be dominated by frequent users of the platforms. They may not show some types of popular activities (this study and Pickering et al. 2018b) but over-represent novel experiences and activities (Calcagni et al. 2019; Ghermandi and Sinclair 2019). There are also important social factors shaping what people post on social media (Oeldorf-Hirsch and Sundar 2016), with variation in content among platforms, users and places (Angradi et al. 2018; Hausmann et al. 2018). As highlighted above, there can also be differences in what is ‘seen’ by researchers coding images versus their meaning to those posting images (Stephchenkova and Zhan 2013).

Conclusions

Researchers are starting to evaluate benefits and limitations of social media for managers, practitioners and natural resource professionals. This includes how social media such as Flickr images can be used to better understand the ways people relate to natural environments including the services they provide. But more research is required including comparing platforms, locations and between social media and more traditional methods. What is clear is that user-generated content is increasing in volume and variety with more people posting it from a diversity of places to different platforms where this content itself is shaping and influencing not only visitation to parks, but the ways people value the natural environment.

Acknowledgements

We thank Ruben Massarelli and Dirección de Recursos Naturales Renovables, Mendoza for providing the visitor statistic data and GIS layers for Aconcagua Provincial Park and we thank Dr. Charles Lawson for his valuable inputs made to this manuscript. We also thank CONICET-Argentina for their support through the funded Short-Internship-Program (Pasantias Breves en el Exterior) and Griffith University for the Visiting-Fellow Position that made this collaborative-article possible. We also thank the funding provided by The Scientific and Technological Research Fund (FONCYT, PICT 2015-1455 and PICT 2017-1869).

Biographies

Sebastian Dario Rossi

is a Researcher at the Instituto Argentino de Investigaciones de Zonas Áridas (IADIZA) in CONICET, Argentina. His research interests include protected areas planning and management, outdoor recreation and nature–society interactions in natural areas including environmental psychology and place making.

Agustina Barros

is a Researcher in Ecology at the Instituto Argentino de Nivología y Glaciología y Ciencias Ambientales (IANIGLA), Centro Científico Tecnológico (CCT) CONICET, Mendoza, Argentina. Her research interests include recreation and alpine ecology, tourism and protected areas management.

Chelsey Walden-Schreiner

is an Independent Researcher based on USA. Her research interests include recreation ecology, visitor use and impact monitoring, and applications of cloud and geospatial technologies for conservation and resource management.

Catherine Pickering

is a Professor, Researcher and Head of Discipline in Ecology and Evolution in the School of Environment and Science, Griffith University, Australia. Her research interests include ecology, tourism and protected area management.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Sebastian Dario Rossi, Email: srossi@mendoza-conicet.gob.ar.

Agustina Barros, Email: anaagustinabarros@gmail.com, Email: abarros@mendoza-conicet.gob.ar.

Chelsey Walden-Schreiner, Email: waldenschreiner@gmail.com.

Catherine Pickering, Email: c.pickering@griffith.edu.au.

References

  1. Abbe JD, Manning RE. Wilderness day use: Patterns, impacts and management. International Journal of Wilderness. 2007;13:21–25. [Google Scholar]
  2. Angradi TR, Launspach JJ, Debbout R. Determining preferences for ecosystem benefits in Great Lakes Areas of Concern from photographs posted to social media. Journal of Great Lakes Research. 2018;44:340–351. doi: 10.1016/j.jglr.2017.12.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barros A, Monz C, Pickering CM. Is tourism damaging ecosystems in the Andes? Current knowledge and an agenda for future research. Ambio. 2015;44:82–98. doi: 10.1007/s13280-014-0550-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Barros A, Pickering C, Gudes O. Desktop analysis of potential impacts of visitor use: A case study for the highest park in the Southern Hemisphere. Journal of Environmental Management. 2015;150:179–195. doi: 10.1016/j.jenvman.2014.11.004. [DOI] [PubMed] [Google Scholar]
  5. Bernbaum E. Sacred mountains: Themes and teachings. Mountain Research and Development. 2006;26:304–309. doi: 10.1659/0276-4741(2006)26[304:SMTAT]2.0.CO;2. [DOI] [Google Scholar]
  6. Buckley R. Adventure tourism. Wallingford, UK: CABI International; 2006. [Google Scholar]
  7. Calcagni F, Maia ATA, Connolly JJT, Langemeyer J. Digital co-construction of relational values: Understanding the role of social media for sustainability. Sustainability Science. 2019 doi: 10.1007/s11625-019-00672-1. [DOI] [Google Scholar]
  8. Castley JG, Bennett A, Pickering CM. Wildlife visual imagery: Do pictures used to promote destinations on-line match on-site species visibility at two geographic destinations? Geographical Research. 2013;51:59–70. doi: 10.1111/j.1745-5871.2012.00767.x. [DOI] [Google Scholar]
  9. Clemente P, Calvache M, Antunes P, Santos R, Cerdeira JO, Martins MJ. Combining social media photographs and species distribution models to map cultural ecosystem services: The case of a Natural Park in Portugal. Ecological Indicators. 2019;96:59–68. doi: 10.1016/j.ecolind.2018.08.043. [DOI] [Google Scholar]
  10. Costanza R, D’Arge R, De Groot R, Farber S, Grasso M, Hannon B, Limburg K, Naeem S, et al. The value of the world’s ecosystem services and natural capital. Nature. 1997;387:253–260. doi: 10.1038/387253a0. [DOI] [Google Scholar]
  11. Debarbieux B, Oiry Varacca M, Rudaz G, Maselli D, Kohler T, Jurek M, editors. Tourism in mountain regions: Hopes, fears and realities. Geneva: UNIGE, CDE, SDC; 2014. [Google Scholar]
  12. Ghermandi A, Sinclair M. Passive crowdsourcing of social media in environmental research: A systematic map. Global Environmental Change. 2019;55:36–47. doi: 10.1016/j.gloenvcha.2019.02.003. [DOI] [Google Scholar]
  13. Grêt-Regamey A, Brunner SH, Kienast F. Mountain ecosystem services: Who cares? Mountain Research and Development. 2012;32:S23–S34. doi: 10.1659/MRD-JOURNAL-D-10-00115.S1. [DOI] [Google Scholar]
  14. Haines-Young, R., and M.B. Potschin. 2017. Common International Classification of Ecosystem Services (CICES) V5.1 and guidance on the application of the revised structure. www.cices.eu. Accessed 22 July 2019.
  15. Hamilton L, McMillan L. Guidelines for planning and managing mountain protected areas. Gland: IUCN World Commission on Protected Areas; 2004. [Google Scholar]
  16. Hausmann A, Toivonen T, Slotow R, Tenkanen H, Moilanen A, Heikinheimo V, Di Minin E. Social media data can be used to understand tourists’ preferences for nature-based experiences in protected areas. Conservation Letters. 2018 doi: 10.1111/conl.12343. [DOI] [Google Scholar]
  17. Langemeyer J, Calcagni F, Baro F. Mapping the intangible: Using geolocated data to examine landscape aesthetics. Land Use and Policy. 2018;77:542–552. doi: 10.1016/j.landusepol.2018.05.049. [DOI] [Google Scholar]
  18. Leung, Y.-F., A. Spenceley, G. Hvenegaard, and R. Buckley. 2018. Tourism and visitor management in protected areas: Guidelines for sustainability. Best Practice Protected Area Guidelines Series No 27. Gland: IUCN.
  19. Linting M, Meulman JJ, Groenen PJ, Van der Kooij AJ. Nonlinear principal components analysis: Introduction and application. Psychological Methods. 2007;12:336–358. doi: 10.1037/1082-989X.12.3.336. [DOI] [PubMed] [Google Scholar]
  20. Martinez-Harms MJ, Bryan BA, Wood SA, Fisher DM, Law E, Rhodes JR, Dobbs C, Biggs D, et al. Inequality in access to cultural ecosystem services from protected areas in the Chilean biodiversity hotspot. Science of the Total Environment. 2018;636:1128–1138. doi: 10.1016/j.scitotenv.2018.04.353. [DOI] [PubMed] [Google Scholar]
  21. Martinez Pastur G, Peri PL, Lencinas MV, Garcia-Llorente M, Martin-Lopez B. Spatial patterns of cultural ecosystem services provision in Southern Patagonia. Landscape Ecology. 2015;31:383–399. doi: 10.1007/s10980-015-0254-9. [DOI] [Google Scholar]
  22. Miller ZD, Taff BD, Neman P, Lawhon B. A proposed research agenda on social media’s role in visitor use and experience in parks and protected areas. Journal of Park and Recreation Administration. 2019 doi: 10.18666/jpra-2019-9553. [DOI] [Google Scholar]
  23. Millennium Ecosystem Assessment . Ecosystems and human well-being: Synthesis. Washington, DC: Island Press; 2005. [Google Scholar]
  24. Newsome D, Moore SA, Dowling RK. Natural area tourism: Ecology, impacts and management. Bristol: Channel View Publications; 2012. [Google Scholar]
  25. Norman P, Pickering CM. Using volunteered geographic information to assess park visitation: Comparing three on-line platforms. Applied Geography. 2017;89:163–172. doi: 10.1016/j.apgeog.2017.11.001. [DOI] [Google Scholar]
  26. Norman P, Pickering CM, Castley JG. What can volunteered geographic information tell us about the different ways mountain bikers, runners and walkers use urban reserves? Landscape and Urban Planning. 2019;185:180–190. doi: 10.1016/j.landurbplan.2019.02.015. [DOI] [Google Scholar]
  27. Oeldorf-Hirsch A, Sundar SS. Social and technical motivations for online photo sharing. Journal of Broadcasting and Electronic Media. 2016;60:624–642. doi: 10.1080/08838151.2016.1234478. [DOI] [Google Scholar]
  28. Oteros-Rozas E, Martín-López B, Fagerholm N, Bieling C, Plieninger T. Using social media photos to explore the relation between cultural ecosystem services and landscape features across five European sites. Ecological Indicators. 2018;94:74–86. doi: 10.1016/j.ecolind.2017.02.009. [DOI] [Google Scholar]
  29. Pickering CM, Rossi SD, Hernando A, Barros A. Current knowledge and future research directions for the Monitoring and Management of Visitors in Recreational and Protected Areas. Journal of Outdoor Recreation and Tourism. 2018;21:10–18. doi: 10.1016/j.jort.2017.11.002. [DOI] [Google Scholar]
  30. Pickering, C., M. Chabau-Gibson, and J. Raneng. 2018b. Using Flickr images to assess how visitors’ value and use natural areas: Lessons from a popular natural area on the Gold Coast, Australia. In Abstracts of the 9th international conference on monitoring and management of visitors in recreational and Protected Areas, ed. J. Dehez, Bordeaux, France, August 2018, 68–69.
  31. Pickering, C., A. Barros, C. Walden-Schreiner, and S.D. Rossi. in press. Using social media images and text to examine how tourists view and value the highest mountain in Australia. Journal of Outdoor Tourism and Recreation.
  32. Pierce WV, Manning RE. Day and overnight visitors to the Olympic Wilderness. Journal of Outdoor Recreation and Tourism. 2015;12:14–24. doi: 10.1016/j.jort.2015.11.002. [DOI] [Google Scholar]
  33. Richards DR, Tunçer B. Using image recognition to automate assessment of cultural ecosystem services from social media photographs. Ecosystem Services. 2018;31:318–325. doi: 10.1016/j.ecoser.2017.09.004. [DOI] [Google Scholar]
  34. Richards DR, Friess DA. A rapid indicator of cultural ecosystem service usage at a fine spatial scale: Content analysis of social media photographs. Ecological Indicators. 2015;53:187–195. doi: 10.1016/j.ecolind.2015.01.034. [DOI] [Google Scholar]
  35. Rosário IT, Rebelo R, Cardoso P, Segurado P, Mendes RN, Santos-Reis M. Can geocaching be an indicator of cultural ecosystem services? The case of the Montado savannah-like landscape. Ecological Indicators. 2019;99:375–386. doi: 10.1016/j.ecolind.2018.12.003. [DOI] [Google Scholar]
  36. Sherren K, Smit M, Holmlund M, Parkins JR, Chen Y. Conservation culturomics should include images and a wider range of scholars. Frontiers in Ecology and the Environment. 2017;15:289–290. doi: 10.1002/fee.1507. [DOI] [Google Scholar]
  37. Smith, A., and M. Anderson. 2018. Social media use in 2018. Pew Research Centre. http://www.pewinternet.org/2018/03/01/social-media-use-in-2018/. Accessed Mar 2018.
  38. Smith, C. 2018. Interesting Flickr stats and facts (February 2018): By the numbers. https://expandedramblings.com/index.php/flickr-stats/. Accessed March 2018.
  39. Stephchenkova S, Zhan F. Visual destination images of Peru: Comparative content analysis of DMO and user-generated photography. Tourism Management. 2013;36:590–601. doi: 10.1016/j.tourman.2012.08.006. [DOI] [Google Scholar]
  40. Stolton, S., and N. Dudley (Principle authors). 2014. Chapter 6: Values and benefits of protected areas. In Protected areas governance and management, eds., G. Worboys, M. Lockwood, A. Kothari, S. Feary, and I. Pulsford, 145–169. Canberra: ANU Press.
  41. Teles da Mota, V. and C. Pickering. 2018. How can we use social media to know more about visitors to natural areas? In Abstracts of the 9th international conference on monitoring and management of visitors in recreation and Protected Areas (MMV9), ed. J. Dehez, Bordeaux, France, August 2018, 72–74.
  42. Tenkanen H, Di Minin E, Heikinheimo V, Hausmann A, Herbst M, Kajala L, Toivonen T. Instagram, Flickr, or Twitter: Assessing the usability of social media data for visitor monitoring in protected areas. Scientific Reports. 2017;7:17615. doi: 10.1038/s41598-017-18007-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Thiagarajah J, Wong SK, Richards DR, Friess DA. Historical and contemporary cultural ecosystem service values in the rapidly urbanizing city state of Singapore. Ambio. 2015;44:666–677. doi: 10.1007/s13280-015-0647-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Tieskens KF, Van Zanten BT, Schulp CJE, Verburg PH. Aesthetic appreciation of the cultural landscape through social media: An analysis of revealed preference in the Dutch river landscape. Landscape and Urban Planning. 2018;177:128–137. doi: 10.1016/j.landurbplan.2018.05.002. [DOI] [Google Scholar]
  45. Van Zanten BT, Van Berkel DB, Meentemeyer RK, Smith JW, Tieskens KF, Verburg PH. Continental-scale quantification of landscape values using social media data. Proceedings of the National Academy of Sciences of the United States of America. 2016;113:12974–12979. doi: 10.1073/pnas.1614158113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Walden-Schreiner C, Rossi SD, Barros A, Pickering C, Leung Y-F. Using crowd sourced photos to assess seasonal patterns of visitor use in mountain protected area. Ambio. 2018;47:781–793. doi: 10.1007/s13280-018-1020-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wartmann FM, Purves RS. Investigating sense of place as a cultural ecosystem service in different landscapes through the lens of language. Landscape and Urban Planning. 2018;175:169–183. doi: 10.1016/j.landurbplan.2018.03.021. [DOI] [Google Scholar]
  48. Willemen L, Cottam AJ, Drakou EG, Burgess ND. Using social media to measure the contribution of Red List species to the nature-based tourism potential of African protected areas. PLoS ONE. 2015 doi: 10.1371/journal.pone.0129785. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wolff S, Schulp CJE, Verburg PH. Mapping ecosystem services demand: A review of current research and future perspectives. Ecological Indicators. 2015;55:159–171. doi: 10.1016/j.ecolind.2015.03.016. [DOI] [Google Scholar]
  50. Wood SA, Guerry A, Silver J, Lacayo M. Using social media to quantify nature-based tourism and recreation. Scientific Reports. 2013;3:2976. doi: 10.1038/srep02976. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. World Commission on Protected Areas. 2019. Mountains. https://www.iucn.org/commissions/world-commission-protected-areas/our-work/mountains. Accessed March 2019.

Articles from Ambio are provided here courtesy of Springer

RESOURCES