Print
Recreation effects on wildlife: a review of potential quantitative thresholds
expand article infoJeremy S. Dertien, Courtney L. Larson§, Sarah E. Reed
‡ Wildlife Conservation Society, Fort Collins, United States of America
§ Colorado State University, Fort Collins, United States of America
Open Access

Abstract

Outdoor recreation is increasingly recognised for its deleterious effects on wildlife individuals and populations. However, planners and natural resource managers lack robust scientific recommendations for the design of recreation infrastructure and management of recreation activities. We reviewed 38 years of research on the effect of non-consumptive recreation on wildlife to attempt to identify effect thresholds or the point at which recreation begins to exhibit behavioural or physiological change to wildlife. We found that 53 of 330 articles identified a quantitative threshold. The majority of threshold articles focused on bird or mammal species and measured the distance to people or to a trail. Threshold distances varied substantially within and amongst taxonomic groups. Threshold distances for wading and passerine birds were generally less than 100 m, whereas they were greater than 400 m for hawks and eagles. Mammal threshold distances varied widely from 50 m for small rodents to 1,000 m for large ungulates. We did not find a significant difference between threshold distances of different recreation activity groups, likely based in part on low sample size. There were large gaps in scientific literature regarding several recreation variables and taxonomic groups including amphibians, invertebrates and reptiles. Our findings exhibit the need for studies to measure continuous variables of recreation extent and magnitude, not only to detect effects of recreation on wildlife, but also to identify effect thresholds when and where recreation begins or ceases to affect wildlife. Such considerations in studies of recreation ecology could provide robust scientific recommendations for planners and natural resource managers for the design of recreation infrastructure and management of recreation activities.

Keywords

Distance to people, human disturbance, park management, protected areas, recreation impacts, wildlife conservation

Introduction

Human disturbance is widely recognised for its deleterious effects on the physiology, behaviour and demographics of individuals and populations of wild animals (Steven and Castley 2013; Coetzee and Chown 2016). Sources of disturbance are extremely diverse and include mortality from hunting and roadkill (Scillitani et al. 2010) to non-consumptive sources, such as hiking, boating and wildlife watching (Cowling et al. 2015; Tarjuelo et al. 2015). Whereas the population- or community-level effects of human disturbance via take are more apparent, effects of non-consumptive human disturbance on wildlife physiology and behaviour are less easily identified or separated from other confounding environmental factors. A growing body of research has focused on the effects of non-consumptive human disturbance with a specific focus on outdoor recreation (Larson et al. 2016).

Outdoor recreation is growing rapidly around the world and has been identified as one of the greatest threats to protected areas (Balmford et al. 2015; Schulze et al. 2018). In the United States, visitation to developed recreation sites is projected to increase by 23% by 2030 (White et al. 2014). Human disturbance on wildlife from non-consumptive recreation can result in altered spatiotemporal habitat use (Kangas et al. 2010; Rösner et al. 2014), decreased survival and reproduction (Iverson et al. 2006; Baudains and Lloyd 2007) and, ultimately, decreased population abundance (Miller et al. 1998; Bejder et al. 2006) or extirpation from otherwise suitable habitat (Steven and Castley 2013). To reduce or eliminate negative effects of recreation on wildlife, land managers require explicit recommendations for how to design trails, manage visitors and otherwise balance the multi-use objectives of many protected areas.

Identifying the effect threshold or the point at which wildlife begins to be disturbed by such recreation activities is key to providing informed recommendations to land managers and planners attempting to make decisions regarding infrastructure construction and visitor management (Braunisch et al. 2011; Rösner et al. 2014; Monz et al. 2016). Data on effect thresholds give protected area planners and managers a better understanding of, for example, the overall effect area for each trail (Lenth et al. 2008), buffer zones around birds of prey nests (Swarthout and Steidl 2001; Keeley and Bechard 2011) and evidence to defend limits on visitation numbers or seasonal closures (Schummer and Eddleman 2003; Malo et al. 2011). Researchers who study the effects of recreational activities on wildlife often attempt to estimate quantitative effects thresholds as effect distances from people or infrastructure (Pittfield and Burger 2017; Bötsch et al. 2018), density of trails and other infrastructure (Braunisch et al. 2011; Harris et al. 2014) or visitation rates (Kerbiriou et al. 2009; Malo et al. 2011).

Elucidating an effect threshold can be difficult because a threshold may not exist, the study sample was not large enough or inferring an effect threshold was not of interest during the study design. Therefore, often the mean distance, mean disturbance intensity or an index of disturbance is reported rather than an effect threshold (Bennett et al. 2013; Costello et al. 2013). The mean effect level is important and valuable information for conservation, but likely does not capture the point at which all or, at least, a large portion of wildlife individuals are affected. Estimating the complete extent of potential recreation impacts provides a more complete robust assessment for conservation planning.

Our objective was to identify quantitative thresholds of non-consumptive recreation in order to provide clearer data to nature professionals about the potential extents and limits of recreation impacts on wildlife. We conducted a systematic review of the published scientific literature of non-consumptive human recreation effects on wildlife in terrestrial environments. We analysed articles to determine if the authors detected a quantitative threshold where recreation began to impact wildlife at the individual, population or community level or cause habitat degradation. We summarise the findings descriptively, reviewing the species and ecosystems that have been studied and identifying gaps in the available literature. We identify quantitative thresholds across a wide array of recreation activity types, wildlife species and response measurements which only allow summation of our findings across broad categories. In addition, we investigated whether the threshold effect depends on body size, predicting a positive relationship between body size and quantitative thresholds (i.e. larger birds and mammals would respond to disturbance at further distances) (Blumstein et al. 2005; Piratelli et al. 2015; Battisti et al. 2019). Finally, we discuss the limitations of these findings and how future research should consider study designs that explore the quantitative thresholds of systems as a means of providing the best recommendations for natural resource professionals.

Methods

We used a database of primary literature compiled for a systematic review of the effects of recreation on wildlife (Larson et al. 2016), supplemented with additional articles published through December 2018 that matched the criteria of Larson et al. (2016) for a total of 38 years of publications. Their criteria were limited to journals (n = 166) in the Web of Science database (Thompson Reuters, New York, NY, USA) in the categories: biodiversity conservation, ecology, zoology and behavioural sciences. The criteria included articles that focused on non-consumptive human recreation activities (i.e. did not include hunting or fishing), studied one or more animal species, assessed recreation effects using statistical tests and were published in English. For the purpose of our review of quantitative thresholds, we included only studies of terrestrial species or interactions with aquatic animals while they were on land. This resulted in 330 articles remaining in our database.

We sought to determine which papers identified a minimum effect threshold, which we defined as the point at which ≥ 90% of sampled wildlife individuals already showed a behavioural or physiological response (e.g. flushing, increased heart rate) to a recreation disturbance or the point at which recreation disturbance begins to reduce the presence, abundance or survival probability of a population or degrade the habitat. For example, Thomas et al. (2003) found that 96% of sanderlings (Calidris alba Pallas, 1764) were disturbed at a distance of ≤ 30 m and Malo et al. (2011) found that detections of guanaco (Lama guanicoe Müller, 1776) began to reduce at > 250 visitors/day. We chose this definition because of the preponderance of studies that identified the 90th or 95th percentile of threshold distance (Swarthout and Steidl 2001; Muposhi et al. 2016). A threshold of habitat degradation was highly study-specific and, therefore, was generally the point at which a specific paper’s metric of habitat alteration began to exhibit a negative change correlated with recreation (Bennett et al. 2013). We did not include papers that reported only the mean level of disturbance (e.g. mean flush distance, mean recreation group size), as this value does not represent the full distribution of disrupted animals. We did include papers that presented graphical representations that allowed for estimation of a threshold effect, even if that threshold was not explicitly stated in the article text.

We recorded the details of each quantitative threshold, including the measure of wildlife or indirect response (behavioural, occurrence, physiological, relative abundance, reproduction and habitat degradation), the measure of recreation disturbance (e.g. number of visitors, distance to people) and the value at which the disturbance threshold was observed (e.g. > 14 visitors/day, < 100 m from people). Some articles recorded multiple threshold effects per species that varied by season or recreation type; therefore, several articles had multiple database inputs. To avoid pseudo-replication, we took the largest threshold response if there were multiple values for one species across seasons or for the same recreation activity. We did record all values across different recreation types for the same species since recreation types can be viewed as different treatments. We classified each article into nine different ecosystem classifications alpine/tundra, coast/shoreline, desert, forest, grassland, polar, savannah, scrub/shrub and wetland. Studies were classified into all the ecosystems that the authors identify in the paper. In addition, we extracted details on study type (e.g. observational or experimental), species of interest and publication information.

We further binned each paper based on recreation activities into either hiking-only, multi-use non-motorised or motorised categories. This was done in order to compare threshold effects across general recreation types. The multi-use non-motorised included both papers that had hiking as one of multiple activities and the motorised category included papers that were motorised-only and which had multiple motorised and non-motorised recreation activities. We used a single-factor analysis of variance to test if there was a significant difference in the threshold effects amongst these recreation categories.

Finally, we researched body masses for all bird and mammal species (Dunning 2007; Williamson et al. 2013) and used linear regression to analyse the relationship between mass and effect distance for birds and mammals separately, with body mass as an explanatory variable. We excluded two studies on flightless birds given the mass disparity to flighted birds and two studies on mammal populations that were habituated to close human presence. We log-transformed bird (n = 50) and mammal (n = 21) body mass and effect distance to conform to assumptions of normality. Significance of all tests was set at 0.05 and analyses were performed in programme R (R Core Development Team 2020).

Results

We reviewed 330 journal articles, of which 53 articles identified one or more quantitative threshold effects. The vast majority of the 53 articles focused on bird or mammal species, with little representation of invertebrates, amphibians or reptiles. Studies of birds focused primarily on species in the orders Charadriiformes (e.g. wading birds and gulls), Accipitriformes (e.g. hawks, eagles and vultures) and Passeriformes (i.e. perching birds) (Fig. 1A). Studies of Strigiformes (i.e. owls) and Galliformes (i.e. upland birds) were notably under-represented. Mammal studies primarily focused on species in the orders Artiodactyla (i.e. even-toed ungulates) and Carnivora (i.e. bears and cats) (Fig. 1B).

Figure 1.

Recreation effect threshold articles by bird and mammal orders (a) bird and (b) mammal orders studied in papers that identified an effect threshold. Several articles contained more than one order, thus, the total number of articles sums to more than all the threshold effects papers.

Studies that identified threshold effects were conducted predominately in forest or coastal/shoreline ecosystems with limited representation in the other ecosystems (Fig. 2A). Hiking was by far the most studied recreational activity, followed by wildlife viewing on land, beach use and dog-walking (Fig. 2B). Most studies examined only non-motorised activities (71.7%), while fewer studies examined only motorised activities (15.1%) or both (13.2%). Nearly half (39.6%) of the articles examined two or more recreation activities, two-thirds of which included hiking as one of the activities.

Figure 2.

Descriptive statistics of recreation threshold articles. Summary of a ecosystem types b recreation activities and c disturbance variables of articles that identified an effect threshold. Several papers studied more than one ecosystem, recreation activity or disturbance variable, therefore, percentages in one plot sum to greater than 100%. Aquatic recreation only included those water-based activities that effected wildlife on land. Disturbance variable distance to trail included all forms of recreation (e.g. motorised, non-motorised and dogs allowed and not allowed) and infrastructure referred to density of human built strucutres.

Quantitative thresholds were identified for a variety of recreation disturbance variables, but can be generally grouped into distance effects, visitation rates and infrastructure density effects (Fig. 2C). Distance effects included distance to people, trails and vehicles. Studies that focused on the distance effects to people included observational studies in coastal ecosystems where trails are less well defined and quasi-experimental studies, in which researchers approached individual animals to measure alert and flight initiation distances. Quantitative thresholds for distance to trail were identified in studies of birds, mammals and invertebrates. Several studies were precluded from the possibility of finding a threshold effect because the researchers only focused on categorical differences between trail types.

Articles examining thresholds of visitation rates or the number of people or vehicles per unit time, were comparatively less well represented (Fig. 2C). Those measuring threshold numbers of people focused on human visitation effects on primate group behaviour, decreasing detections correlated with increasing magnitude of visitation and behavioural disturbance to animals from tourist group visits to wildlife concentrations. Visitor numbers, as low as one person or off-road vehicle per day, were shown to negatively affect the habitat use of studied species in some cases. Very few articles focused on or found recreation infrastructure density effect thresholds (Fig. 2C).

The vast majority of threshold studies focused on the behavioural response of wildlife to a human disturbance, followed by measurements of occurrence and relative abundance (Table 1). Of the behavioural response measurements, over half were measured as a flight initiation distance (i.e. the distance at which wildlife began to move due to a human disturbance). Other behaviour measurements included the number of wildlife individuals feeding or standing, vigilance behaviour and changes in activity budget; however, each of these was measured in less than 4% of papers. Occurrence measurements were a derivation of presence or detection and abundance measurements included counts of individuals or faecal pellet densities. Physiological, reproductive or habitat degradation response thresholds were represented in less than 2% of papers (Table 1).

Given the relatively low sample size of articles that identified thresholds, we were only able to make meaningful conclusions about distance thresholds for birds and mammals (Fig. 3). Distance thresholds from people and trails varied amongst orders and species. For example, wading birds and passerines were generally affected at distances less than 100 m, whereas larger-bodied species, such as hawks and eagles, had threshold effect distances greater than 400 m (Fig. 4). Smaller rodent species avoided areas within 50-100 m of trails or people, whereas some carnivores and ungulates had minimum effect distances anywhere from 40 to 1000 m from trails and people. The median effect threshold distance was 80.0 m for birds and 77.5 m for mammals and mean thresholds were 112.1 m and 151.1 m for birds and mammals, respectively (Fig. 4). We found evidence of a positive correlation between increasing body mass of flighted birds (βˆ = 0.233 SE = 0.052; p < 0.001) and effect distance threshold (Fig. 5). We did not find the same relationship between mammal body mass (βˆ = 0.138 SE = 0.102; p = 0.192) and effect distance threshold (Fig. 5).

Figure 3.

Distance of effect thresholds of birds and mammals. Effect distance thresholds across all mammal (n = 24) and bird (n = 53) species studied for the impacts of recreation on wildlife. Thresholds included observed distances of direct human disturbance to wildlife and disturbance from recreation infrastructure. Outliers for mammals are effect distances for larger ungulates. Outliers for birds are effect distances for raptors, including hawks and eagles. Boxplots indicate median and 25th and 75th percentiles. Whiskers extend to data 1.5 times the interquartile range.

Figure 4.

Distance of effect thresholds across bird orders. Threshold distances of birds by taxonomic group. Black dots indicate individual data points. The only owl threshold distance (x = 55 m) is not presented in this figure. Boxplots indicate median and 25th and 75th percentiles. Whiskers extend to data 1.5 times the interquartile range.

Figure 5.

Wildlife body mass as a predictor of effect threshold distance. Regression analysis of body mass as a predictor for a taxon’s effect distance threshold for (a) birds and (b) mammals. Black dots indicate individual observations and the shaded area represents the 95% confidence interval.

Motorised recreation had the highest median threshold distance for birds (111.5 m), whereas multi-use non-motorised had the highest median value for mammals (100 m) (Fig. 6). Hiking-only recreation had the lowest median threshold distance for both birds (45 m) and mammals (40 m). However, there was substantial overlap of the distribution of values amongst all recreation types and single-factor ANOVA found no significant difference amongst recreation types for birds (F = 0.066, p < 0.936) or mammals (F = 0.760, p < 0.480).

Figure 6.

Effect thresholds across groups of recreation activities and taxa. Black dots indicate individual data points. Boxplots indicate median and 25th and 75th percentiles. One outlier of 1000 m is not shown for mammal motorised. Whiskers extend to data 1.5 times the interquartile range.

Discussion

There are numerous gaps in the scientific literature regarding quantitative thresholds of recreation effects on wildlife. While the publication rate on the recreation effects on wildlife has been increasing (Larson et al. 2016), there is still a need for science-based recommendations for management of recreation that present thresholds of disturbance. Further, certain taxonomic groups and ecosystems are substantially under-represented in this body of research. In this review, invertebrates were included in two articles and, while there are papers that have focused on reptile and amphibian behaviour (Moore and Seigel 2006; Bowen and Janzen 2008; Selman et al. 2013), we only found one paper each that presented evidence for a threshold to human disturbance on these taxa (Rodríguez-Prieto and Fernández-Juricic 2005; Pittfield and Burger 2017). Threshold studies were primarily conducted in forests or coastal ecosystems, with little representation of other ecosystems, especially deserts and savannahs.

We did however find numerous examples of minimum effect thresholds from certain taxa, especially shorebirds and ungulates. Studies of plover species (genera Charadrius and Pluvialis) provided some of the clearest examples of minimum effect thresholds and were primarily identified between 50-100 m (Fig. 4) (Lafferty 2001; Jorgensen et al. 2016). Ungulate species were the best represented mammalian group and had a broad distribution of effect distance thresholds from 40 to 1000 m (Borkowski et al. 2006; Preisler et al. 2006).

Research that identified effect thresholds were heavily skewed towards studies that measured the distance from which there was a behavioural response from wildlife. Few studies in recreation ecology identified a physiological or reproductive response threshold or showed a threshold of visitation numbers or density of human infrastructure. Previous work has shown that even low human presence can impact wildlife habitat use (Cornelius et al. 2001; Spaul and Heath 2016; Patten and Burger 2018); however, isolating and interpreting the impacts of visitor numbers or infrastructure density is arguably more difficult than the physical distance to humans or trails, which could explain the sparse examples of density impacts in our findings. Further, short-term behavioural responses to human disturbance can be difficult to link directly to population consequences (Gill et al. 2001). With the increasing visitation pressure on the world’s protected areas (Schulze et al. 2018), there is a great opportunity and need to focus on identifying physiological or reproductive effect thresholds of recreation and to measure when visitor numbers begin to deleteriously impact wildlife.

We found that the median threshold distance for birds and mammals across different recreation activities ranged from 40 to 111.5 m, but that the values were not significantly different amongst groups of recreation activities (Fig. 5). Though not statistically significant, the hiking-only recreation group for both mammals and birds had median threshold distance approximately half the magnitude of the non-motorised multiple-use or motorised recreation. This points to the magnitude of influence that even non-motorised recreation can have on the disturbance of wildlife (Stalmaster and Kaiser 1998; Reimers et al. 2006). Large buffer zones around human activities should always be considered during the planning and maintenance of parks and protected areas (Miller et al. 1998; Keeley and Bechard 2011). Based upon on our findings, efficient trail systems with significant gaps of at minimum 250 m between any two trails provide some undisturbed areas for most wildlife species. The suppression and restoration of social trails (i.e. non-designated informal trails) maintain these buffer zones between trails, one of several conservation benefits of reducing these unplanned features. However, even intact buffers between trails do not ensure all species will have areas free of human disturbance.

We found a positive correlation between flighted bird body mass and effect distance threshold, but no relationship between mammal body mass and effect distance threshold. Flight initiation distance, the predominant response measure in our review (Table 1), is shown to be significantly correlated with bird body mass (Piratelli et al. 2015). Similarly, Blumstein et al. (2005) found a significant relationship between body mass and alert distance from a sample of 150 species, suggesting that bird body mass could be a good predictor for conservation decision-making. However, this suggestion could be tempered by Larson et al. (2019) who found that high recreation levels had greater negative effects on small bird abundance than on large bird abundance. This indicates the importance of taking multiple response measures into account and understanding their link to individual fitness or population growth when making conservation policies and guidelines.

Table 1.

Wildlife response measurements across threshold articles. Measurement variables varied amongst the articles that identified an effect threshold. Habitat degradation was a measure of habitat response to recreation, an indirect effect to wildlife.

General response Measurement % of Articles
Abundance Density per site 1.9
Number of birds observed 1.9
Number of herds sighted daily 1.9
Pellet density 7.5
Relative abundance 1.9
Track detections 1.9
Behavioral Changes in activity budget of group 1.9
Distance at which animal changed direction 1.9
First reaction 1.9
Flight initiation distance 37.7
Max alert distance 1.9
Number feeding or standing 1.9
Number of moves 1.9
Probability of active response 1.9
Probability of disturbance 1.9
Probability of flight 1.9
Proportion of birds disturbed 1.9
Time spent alert 1.9
Time spent feeding/day 1.9
Vigilance behavior 3.8
Habitat Habitat degradation 1.9
Occurrence Avoidance of human areas 1.9
Community assemblage 1.9
Habitat selection 1.9
Presence 11.3
Reproduction Monthly juvenile survival 1.9
Physiological Heart rate 1.9

The relationship between mammal body mass and human disturbance distance appears less clear than for birds. While there is evidence that smaller-sized mammals are more tolerant of human disturbance and the proximity to human settlements (Battisti et al. 2019; Lhoest et al. 2020), these studies incorporate human disturbances beyond non-consumptive recreation. Larson et al. (2019) did find a similar lack of relationship between mammal body size and recreation effects on abundance, rather than effect distance. In addition, what influence human habituation may play in altering this relationship could not be quantified, though some studies in our analysis did state the likelihood of wildlife individuals habituated to human presence (Lott and McCoy 1995; Klailova et al. 2010). Ultimately, threshold data was much sparser for mammals than birds, thus making it difficult to draw any strong inferences from these results.

There were few examples of recreation infrastructure thresholds, beyond those describing distance to trail. Despite the small sample size, the findings were consistent: infrastructure, even at low densities, can be a contributing factor to altering the habitat use of birds and mammals (Braunisch et al. 2011; Harris et al. 2014; Richard and Côté 2016). At a regional scale, recreation infrastructure can also further exacerbate underlying human-wildlife conflicts (Ménard et al. 2014) and fragment habitats (Whittington et al. 2005). Better understanding of how the density and effect distance of buildings and trails influences the behaviour and survival of wildlife species is paramount for the creation of informed regulatory guidelines.

The detection of threshold effects, if present, can be constrained by the spatiotemporal extent and overall design of a study. In addition, the effect threshold of human presence or infrastructure may be outside the boundaries of the study area or may be difficult to disentangle from correlated effects of other variables. Future researchers should consider how their experimental design could isolate recreation activities and species to support the detection of specific quantitative thresholds. Rodríguez-Prieto and Fernández-Juricic (2005) provide a valuable example demonstrating the quantitative threshold of the effect of recreation activity on the Iberian frog (Rana iberica Boulenger, 1879). Their study design incorporated systematic exposure of the species of interest to human disturbance, which provided direct and measurable flight initiation distances of individual animals from humans. Although this study system is likely easier to control and observe than studies of larger bodied species, it is an important example of implementing a study design to quantify a threshold effect of recreation disturbance and how to effectively represent these results.

There remains a need to understand when and where recreation activities are affecting species negatively or positively (Larson et al. 2016). However, to provide information for future designation and management of recreation use, researchers must go beyond simple hypothesis testing. Studies that focused on categorical variables (e.g. low versus high visitation rates, hikers versus mountain bikers) to examine the potential effects of a recreation treatment, rarely identified the threshold at which the recreation activity may begin or cease to affect an animal species. Asking when and to what extent a species is being disturbed and measuring beyond the spatial or temporal magnitude where the disturbance is expected to begin or end allows researchers to identify important thresholds of recreation disturbance. Researchers should not provide a quantitative recommendation that is not justified by their results, but, where possible, researchers should provide resource managers with clear guidance and conservative estimates to support science-based management decisions. Ultimately, these thresholds allow for more informed and effective management decisions and a higher probability of successful conservation of species.

Acknowledgements

We would like to thank Tony Nelson, the Sonoma Land Trust and the Gordon and Betty Moore Foundation for their support of this project. We would also like to thank Jessica Sushinsky, Stacey Lischka, Sasha Keyel and Miguel Jimenez for discussion and support. Two anonymous reviewers provided very useful feedback and greatly improved the article.

References

  • Balmford A, Green JMH, Anderson M, Beresford J, Huang C, Naidoo R, Walpole M, Manica A (2015) Walk on the wild side: Estimating the global magnitude of visits to protected areas. PLoS Biology 13(2): 1–7. https://doi.org/10.1371/journal.pbio.1002074
  • Battisti C, Fanelli G, Cerfolli F, Amori G, Luiselli L (2019) Body mass and trophic level variations in relation to habitat disturbance in a set of mammal species. Life and Environment 69: 147–152.
  • Bejder L, Samuels A, Whitehead H, Gales N, Mann J, Connor R, Heithaus M, Watson-Capps J, Flaherty C, Krützen M (2006) Decline in relative abundance of bottlenose dolphins exposed to long-term disturbance. Conservation Biology 20(6): 1791–1798. https://doi.org/10.1111/j.1523-1739.2006.00540.x
  • Bennett VJ, Quinn VS, Zollner PA (2013) Exploring the implications of recreational disturbance on an endangered butterfly using a novel modelling approach. Biodiversity and Conservation 22(8): 1783–1798. https://doi.org/10.1007/s10531-013-0512-6
  • Bötsch Y, Tablado Z, Scherl D, Kéry M, Graf RF, Jenni L (2018) Effect of recreational trails on forest birds: Human presence matters. Frontiers in Ecology and Evolution 6: 1–10. https://doi.org/10.3389/fevo.2018.00175
  • Bowen KA, Janzen FJ (2008) Human recreation and the nesting ecology of a freshwater turtle (Chrysemys picta). Chelonian Conservation and Biology 7(1): 95–100. https://doi.org/10.2744/CCB-0650.1
  • Braunisch V, Patthey P, Arlettaz R (2011) Spatially explicit modeling of conflict zones between wildlife and snow sports: Prioritizing areas for winter refuges. Ecological Applications 21(3): 955–967. https://doi.org/10.1890/09-2167.1
  • Coetzee BWT, Chown SL (2016) A meta-analysis of human disturbance impacts on Antarctic wildlife. Biological Reviews of the Cambridge Philosophical Society 91(3): 578–596. https://doi.org/10.1111/brv.12184
  • Costello CMCM, Cain SSISI, Nielson RMRM, Servheen C, Schwartz CC (2013) Response of American black bears to the non-motorized expansion of a road corridor in Grand Teton National Park. Ursus 24(1): 54–69. https://doi.org/10.2192/URSUS-D-11-00027.1
  • Cowling M, Kirkwood R, Boren L, Sutherland D, Scarpaci C (2015) The effects of vessel approaches on the New Zealand fur seal (Arctocephalus forsteri) in the Bay of Plenty, New Zealand. Marine Mammal Science 31(2): 501–519. https://doi.org/10.1111/mms.12171
  • Harris G, Nielson RM, Rinaldi T, Lohuis T (2014) Effects of winter recreation on northern ungulates with focus on moose (Alces alces) and snowmobiles. European Journal of Wildlife Research 60(1): 45–58. https://doi.org/10.1007/s10344-013-0749-0
  • Iverson JB, Converse SJ, Smith GR, Valiulis JM (2006) Long-term trends in the demography of the Allen Cays Rock Iguana (Cyclura cychlura inornata): Human disturbance and density-dependent effects. Biological Conservation 132(3): 300–310. https://doi.org/10.1016/j.biocon.2006.04.022
  • Jorgensen JG, Dinan LR, Brown MB (2016) Flight initiation distances of nesting Piping Plovers (Charadrius melodus) in response to human disturbance. Avian Conservation & Ecology 11(1): 1–5. https://doi.org/10.5751/ACE-00826-110105
  • Kangas K, Luoto M, Ihantola A, Tomppo E, Siikamäki P (2010) Recreation-induced changes in boreal bird communities in protected areas. Ecological Applications 20(6): 1775–1786. https://doi.org/10.1890/09-0399.1
  • Keeley WH, Bechard MJ (2011) Flushing distances of ferruginous hawks nesting in rural and exurban New Mexico. The Journal of Wildlife Management 75(5): 1034–1039. https://doi.org/10.1002/jwmg.140
  • Kerbiriou C, Le Viol I, Robert A, Porcher E, Gourmelon F, Julliard R (2009) Tourism in protected areas can threaten wild populations: From individual response to population viability of the chough Pyrrhocorax pyrrhocorax. Journal of Applied Ecology 46(3): 657–665. https://doi.org/10.1111/j.1365-2664.2009.01646.x
  • Klailova M, Hodgkinson C, Lee PC (2010) Behavioral responses of one western lowland gorilla (Gorilla gorilla gorilla) group at Bai Hokou, Central African Republic, to tourists, researchers and trackers. American Journal of Primatology 72(10): 897–906. https://doi.org/10.1002/ajp.20829
  • Larson CL, Reed SE, Merenlender AM, Crooks KR (2019) A meta-analysis of recreation effects on vertebrate species richness and abundance. Conservation Science and Practice 1(10): 1–9. https://doi.org/10.1111/csp2.93
  • Lhoest S, Fonteyn D, Daïnou K, Delbeke L, Doucet JL, Dufrêne M, Josso JF, Ligot G, Oszwald J, Rivault E, Verheggen F, Vermeulen C, Biwolé A, Fayolle A (2020) Conservation value of tropical forests: Distance to human settlements matters more than management in Central Africa. Biological Conservation 241: e108351. https://doi.org/10.1016/j.biocon.2019.108351
  • Malo JE, Acebes P, Traba J (2011) Measuring ungulate tolerance to human with flight distance: A reliable visitor management tool? Biodiversity and Conservation 20(14): 3477–3488. https://doi.org/10.1007/s10531-011-0136-7
  • Ménard N, Foulquier A, Vallet D, Qarro M, Le Gouar P, Pierre JS (2014) How tourism and pastoralism influence population demographic changes in a threatened large mammal species. Animal Conservation 17(2): 115–124. https://doi.org/10.1111/acv.12063
  • Monz CA, Pickering CM, Hadwen WL (2016) Recent advances in recreation ecology and the implications of different relationships between recreation use and ecological impacts. Frontiers in Ecology and the Environment 11(8): 441–446. https://doi.org/10.1890/120358
  • Moore MJC, Seigel RA (2006) No place to nest or bask: Effects of human disturbance on the nesting and basking habits of yellow-blotched map turtles (Graptemys flavimaculata). Biological Conservation 130(3): 386–393. https://doi.org/10.1016/j.biocon.2006.01.001
  • Muposhi VK, Gandiwa E, Makuza SM, Bartels P (2016) Trophy hunting and perceived risk in closed ecosystems: Flight behaviour of three gregarious African ungulates in a semi-arid tropical savanna. Austral Ecology 41(7): 809–818. https://doi.org/10.1111/aec.12367
  • Pittfield T, Burger J (2017) Basking habitat use and response of freshwater turtles to human presence in an urban canal of Central New Jersey. Urban Ecosystems 20(2): 449–461. https://doi.org/10.1007/s11252-016-0606-5
  • R Core Development Team (2020) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. https://www.R-project.org/
  • Richard JH, Côté SD (2016) Space use analyses suggest avoidance of a ski area by mountain goats. The Journal of Wildlife Management 80(3): 387–395. https://doi.org/10.1002/jwmg.1028
  • Rodríguez-Prieto I, Fernández-Juricic E (2005) Effects of direct human disturbance on the endemic Iberian frog Rana iberica at individual and population levels. Biological Conservation 123(1): 1–9. https://doi.org/10.1016/j.biocon.2004.10.003
  • Rösner S, Mussard-Forster E, Lorenc T, Müller J (2014) Recreation shapes a “landscape of fear” for a threatened forest bird species in Central Europe. Landscape Ecology 29(1): 55–66. https://doi.org/10.1007/s10980-013-9964-z
  • Schulze K, Knights K, Coad L, Geldmann J, Leverington F, Eassom A, Marr M, Butchart SHM, Hockings M, Burgess ND (2018) An assessment of threats to terrestrial protected areas. Conservation Letters 11(3): 1–10. https://doi.org/10.1111/conl.12435
  • Schummer ML, Eddleman WR (2003) Effects of disturbance on activity and energy budgets of migrating waterbirds in south-central Oklahoma. The Journal of Wildlife Management 67(4): 789–795. https://doi.org/10.2307/3802686
  • Scillitani L, Monaco A, Toso S (2010) Do intensive drive hunts affect wild boar (Sus scrofa) spatial behaviour in Italy? Some evidences and management implications. European Journal of Wildlife Research 56(3): 307–318. https://doi.org/10.1007/s10344-009-0314-z
  • Selman W, Qualls C, Owen JC (2013) Effects of human disturbance on the behavior and physiology of an imperiled freshwater turtle. The Journal of Wildlife Management 77(5): 877–885. https://doi.org/10.1002/jwmg.538
  • Spaul RJ, Heath JA (2016) Nonmotorized recreation and motorized recreation in shrub-steppe habitats affects behavior and reproduction of golden eagles (Aquila chrysaetos). Ecology and Evolution 6(22): 8037–8049. https://doi.org/10.1002/ece3.2540
  • Stalmaster MV, Kaiser JL (1998) Effects of recreational activity on wintering bald eagles. Wildlife Monographs 137: 3–46.
  • Steven R, Castley JG (2013) Tourism as a threat to critically endangered and endangered birds: Global patterns and trends in conservation hotspots. Biodiversity and Conservation 22(4): 1063–1082. https://doi.org/10.1007/s10531-013-0470-z
  • Swarthout ECH, Steidl RJ (2001) Flush responses of Mexican spotted owls to recreationists. The Journal of Wildlife Management 65(2): 312–317. https://doi.org/10.2307/3802910
  • Tarjuelo R, Barja I, Morales MB, Traba J, Benítez-López A, Casas F, Arroyo B, Delgado MP, Mougeot F (2015) Effects of human activity on physiological and behavioral responses of an endangered steppe bird. Behavioral Ecology 26(3): 828–838. https://doi.org/10.1093/beheco/arv016
  • White EM, Bowker JM, Askew AE, Langner LL, Arnold JR, English BK (2014) National Center for Natural Resources Economic Research Federal Outdoor Recreation Trends: Effects on Economic Opportunities.
  • Whittington J, St. Clair CC, Mercer G (2005) Spatial responses of wolves to roads and trails in mountain valleys. Ecological Applications 15: 543–553. https://doi.org/10.1890/03-5317
  • Williamson EA, Maisels FG, Groves CP, Fruth BI, Humle T, Morton FB, Richardson MC, Russon AE, Singleton I (2013) Family Hominidae. In: Mittermeier RA, Rylands AB, Wilson DE (Eds) Handbook of the Mammals of the World – Volume 3. Lynx Edicions, Barcelona.