Study areas
We collected survey and salivary hormone data from people at the Camel’s Back—Hulls Gulch Reserve of the Ridge to Rivers Management Area in Boise, Idaho, USA (Fig. 1S). The Lower Hulls Gulch area is largely characterized by sagebrush steppe habitat as well as riparian areas that act as habitat to a variety of avian, mammal, and herpetological species. Common wildlife spotted by recreationists include birds, amphibians, reptiles, and small mammals. We collected data between March 24—May 18, 2017 on both the weekend and weekdays. Weather data, consisting of average wind speed and average air temperature were taken from the Crestline Trail Idaho (Boise, ID).
Participants
We recruited participants at a prominent trail head associated with the recreational area, and the purpose and protocols of the study were explained. Upon agreeing to participate, participants were asked to read an additional written statement of the project and sign a informed consent form. Each recruited individual was minimally required to participate in either the saliva collection or survey, although participation in all parts of the study was highly encouraged. Any recreationists who had been recreating for more than 10 min were not recruited. We defined hikers as recreationists who were intent on walking in the recreational area, and excluded users who were performing exertive activities such as running or biking. All participants were given an anonymous identifier which was used throughout the study. This research project was reviewed and approved by the IRB at Boise State University, Idaho, USA under #006-SB17-061.
Salivary cortisol and testosterone collection
We restricted all sampling until late morning (often 10:00am) to control for diel patterns in hormone concentrations after awakening25,26,27. To further account for any variations due to time of day, the time of collection was recorded for each saliva sample and factored into each analysis. We found that the time of day each cortisol sample was taken did not account for variation, and that the diurnal cycle of cortisol did not explain the change in cortisol concentration that we observed in hikers.
Saliva samples were collected from recreationists using a pre-post paired design. Participants were asked to give at least 0.25 mL of saliva using a saliva collection aid (2 mL cyrovial, Salimetrics PA, USA) via the passive drool method25,28. The passive drool method allows for the collection of large samples for multiple assays, reduces the risk of contamination by collection substances, and allows samples to be frozen without interfering with assay protocols28. Saliva was immediately stored in a portable cooler with ice until frozen at -10 °C. Cortisol and testosterone concentrations were assessed using a Salimetrics Cortisol Enzyme Immunoassay following the manufacturer’s protocols and design (Salimetrics, PA, USA). All assay plates were read using the Gen5 software and Biotek EL800 Plate Reader. Hormone concentrations were then calculated from the optical densities using a standard curve and the online elisaanalysis interface.
Survey collection
After recreating, all participants were asked to take a survey. Survey questions included the following: variables affecting psychological stress levels, observation of wildlife, motivations for recreation (ranging from social, personal challenge, wildlife, and solitude), perceived ecological impact that outdoor recreational activity has on wildlife and habitats, and basic demographics including age and gender. In addition, participants were asked to rate how psychologically stressed they felt after their recreational activity. Participants also had the opportunity to note any negative experiences they had while recreating.
GPS track collection
Participants were encouraged to carry a handheld GPS device to track their route while hiking. We used GPS tracks to connect hiker routes with land cover types such as % vegetation cover, urban cover, water, and riparian cover. Using a 100 m buffer as a standard measurement for visibility, we calculated the proportion of land cover types each hiker traveled through during their trip. We also used GPS tracks to compare line density calculations to estimate areas and trails with higher recreational traffic.
Photograph collection
We used volunteer employed photography and asked participants to take photographs of landscapes they found beautiful. At the end of the trip, participants were asked to select one photograph that captured an area of high aesthetic value to them. Photographs were spatially linked to the landscape using participant GPS tracks, and analyzed using kernel density estimates (KDE) to compare what areas and land cover types were photographed more frequently. In addition, the subject and photographic elements of each photograph were tallied and collected to compare what components of the landscape participants were more likely to take pictures of.
Statistical analyses
To assess the change in salivary cortisol to landscape aesthetics, we used a backward stepwise approach to create a linear model with the following predictor variables: perceived aesthetics (using principal component PC1 scores), land cover metrics, start time, duration, total wildlife observance, plant identification skills (as a measure of ecological familiarity), and all aesthetic interaction terms. . We checked for collinearity among variables and considered pairwise r >|0.70| to indicate a correlation. Due to high correlation between start time and duration, duration was removed from all models. A total of 6 saliva samples (all hikers) were not used due to unusually high (> 2 SD) cortisol concentrations.
Non-significant interaction variables were removed first based on lowest SS values, followed by main effect variables with the lowest SS value until only statistically significant variables remained. Cortisol was evaluated as the change in concentration from recreational activity (Post-collection—Pre-collection). Negative changes correspond to a decrease and positive changes to an increase in hormone concentration after recreating. To meet normality assumptions, the change in both cortisol was transformed using a square root function taken at absolute value. After data transformation, all original signs were returned to maintain the negative–positive spectrum of hormone change. The removal of any outliers did not change the interpretation of the results and were kept across analyses.
We used a backward stepwise approach to explain aesthetic perception with the following predictor variables: weekend/weekday (as a metric of visitor use), total wildlife observance, plant identification skills (as a metric of ecological familiarity), gender, age, duration, and all interaction terms. We used a backward stepwise approach and a generalized linear model to evaluate whether land cover, recreational use, and ecological familiarity explained the number of wildlife observations.No variables were correlated with each other, and all interaction terms were included. To investigate the relationship between perceived ecological impact metrics and recreational motivation, we used spearman correlations.
Source: Ecology - nature.com
