Wildlife responses to recreational trail density are mediated by landscape configuration rather than short-term increases in recreation intensity

Wildlife responses to recreational trail density are mediated by landscape configuration rather than short-term increases in recreation intensity | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Wildlife responses to recreational trail density are mediated by landscape configuration rather than short-term increases in recreation intensity Wim Kuypers, Jim Casaer, Nicolas Dendoncker, Natalie Beenaerts This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7926808/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Protected areas play a key role in conserving biodiversity, yet they increasingly face the challenge of balancing wildlife conservation with rising recreational use. Understanding how wildlife responds to human activity across space and time is therefore crucial to support coexistence. We used camera trap and visitor data to examine how recreational trail density, weekend visitation peaks, and landscape configuration influence spatiotemporal behaviour of roe deer, wild boar, and red fox in the densely visited Hoge Kempen National Park, Belgium. Despite a doubling of on-trail visitation during weekends, activity patterns of wildlife remained unchanged. Moreover, trail density negatively affected wildlife detection rates, but did not differ between weekdays and weekends, indicating that short-term increases in visitor numbers did not alter species’ responses to recreational infrastructure. Instead, wildlife responses to trail density were mediated by landscape configuration: while roe deer and wild boar were more frequent in forest interiors and red fox in open areas, the negative effect of trail density was strongest in open landscapes and declined progressively with distance into forest interiors, becoming negligible deep in forested areas. These findings demonstrate that in densely visited protected areas, landscape configuration, rather than short-term visitor peaks, mediates wildlife responses to recreational infrastructure. Biological sciences/Ecology Earth and environmental sciences/Ecology Biological sciences/Zoology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Introduction Protected areas (PAs) play a vital role in global biodiversity conservation while also serving as places for public recreation and enjoyment [1,2]. This dual mandate creates inherent management challenges, particularly as human activities continue to modify natural landscapes [3.4]. At the same time, recreational use of PAs is rising worldwide [5], leading to increasing levels of disturbance for wildlife. Hence, wildlife survival in the Anthropocene relies on the ability of species to cope with landscape modifications and intensifying human presence [6]. Wildlife responses to human activity are highly variable and shaped by multiple factors. Some species perceive humans as threats and adjust their behaviour by avoiding areas of high human activity or shifting their activity to periods with lower human presence [6–10]. Conversely, other species may exploit human-modified environments, gaining access to resources or protection from predators [11–13]. These divergent responses can be shaped not only by species-specific traits such as body size, ecological niche, and life-history characteristics [14], but also by extrinsic factors like the type and intensity of recreation [15, 16]. Abrupt changes in human activity, such as those caused by the COVID-19 pandemic, provided a unique opportunity to investigate how wildlife responds to substantial, sustained reductions or increases in human disturbance over extended periods. These studies revealed that animals can rapidly adjust spatiotemporal behaviour in response to marked changes in human presence [6,10,17]. In contrast, the global work-week generates recurring, short-term fluctuations in human activity, with visitor numbers in protected areas typically peaking during weekends or national holidays. Despite this widespread pattern, empirical studies examining how such short-term, periodic increases in recreational use simultaneously affect the spatial and temporal behaviour of wildlife remain scarce [18,19]. To cope with these fluctuations, wildlife may rely on temporal and spatial refuges. Temporal refuges, such as nocturnal or crepuscular activity, allow animals to avoid peak periods of human presence, while spatial refuges, such as forested interiors or areas distant from trails, provide shelter from disturbance and help maintain essential behaviours [7–9]. However, the effectiveness of these refuges is strongly influenced by the configuration and management of the surrounding landscape. For instance, habitat suitability and broader landscape context can either amplify or buffer the impacts of recreational activities, shaping whether temporal or spatial refuges are sufficient for species persistence [20–22]. In this way, landscape configuration emerges as another extrinsic factor determining how wildlife responds to human disturbance and how effectively refuges function. Taken together, these responses determine how wildlife can coexist with humans and persist in protected areas with high recreational use [6]. Despite these insights, the interplay between species traits, changes in recreation intensity, and landscape configuration remains poorly understood. This knowledge gap limits the ability of managers to design interventions that balance human use and wildlife conservation. By integrating both spatial (trail density) and temporal (week-weekend) recreational covariates and considering interactions with landscape features such as distance to the forest edge, our study aims to address this gap. We specifically investigate how weekend and holiday visitor peaks, recreational trail density, and distance to forest edges influence activity and camera trap detection rates of roe deer, wild boar, and red fox in the densely visited Hoge Kempen National Park (Belgium). 2. Materials and methods 2.1 Study area The study area (longitudes: 5.552°W − 5.703°W; latitudes: 50.899°N − 51.016°N) is situated in the core region of Hoge Kempen National Park (NPHK), eastern Belgium (Fig. 1 ). It has a total surface area of ∼60km2, consisting of large areas of planted pine forests (41% Pinus sylvestris and Pinus nigra ). These afforested regions are undergoing a systematic transition toward a more natural deciduous forest ecosystem, characterised by the presence of Quercus spp. and Betula spp. (9%). The park features valuable dry (mainly Calluna vulgaris ) and wet (featuring Erica sp. and Myrica sp.) heathlands (11%), along with shrub vegetation (7%), predominantly dominated by Molinia sp . NPHK has altitudes ranging from 50 to 100 m above sea level. The study area has a cool, temperate, and moist climate, with a mean annual temperature of 10.9°C and 816.4 mm rainfall [23]. The study area is situated within a densely populated urban matrix, with an average of 443 inhabitants per square kilometre in the surrounding municipalities [24]. Hence, the region is extensively utilised for recreational purposes, including walking, cycling, MTB, horse riding, and hunting. The study area features an exceptionally dense network of recreational trails, with approximately 100 km of advertised hiking trails (1.657 km/km2), 50 km of cycling paths (0.814 km/km2), and 60 km of designated MTB routes (1.008 km/km2). The estimated annual number of visitors within our study area exceeds 500,000 (Visitor counters NPHK (2021–2025)). Furthermore, there are five official entrance gates at the borders of the study area, providing visitors with parking opportunities and direct access to these various recreational trails. NPHK faces typical challenges, including increasing human recreational pressures, limited data on human usage and wildlife behaviour, and a lack of comprehensive understanding of how these factors influence management strategies. Additionally, the park is becoming increasingly isolated from other prime wildlife habitats due to surrounding road networks, fencing and urban expansion. 2.2 Camera trapping network A stratified random sampling design of 30 motion-sensing camera traps (CTs) (Reconyx HC600 Hyperfire) was applied in NPHK [25]. Sixty sampling locations from a systematic random sampling design (designed by Wevers et al., 2020; described and visualised in detail by Kuypers et al., 2025) were selected to reflect the proportional representation of the main habitat types (heathland, coniferous forest, deciduous forest, mixed forest, and scrubland) (Fig. 1 ). These locations were divided into two subsets of 30, which were alternately sampled every two months. Camera setup and image processing as described by Kuypers et al., 2025. For this article, we focused on images obtained from January 2021 to January 2025. In total, after accounting for camera malfunction or stolen devices, 60 unique sites were sampled with an average trapping effort of 587 days per site. Within these periods, we identified and documented observations of 13 wildlife species (excluding birds, domestic species and humans, Table A1). Among these, roe deer ( Capreolus capreolus ), wild boar ( Sus scrofa ), and red fox ( Vulpes vulpes ) provided sufficient data to model the effect of covariates on space use. 2.3 Recreational pressure To quantify recreational pressure, we used trail density from officially designated, marked hiking, cycling and MTB trails as a proxy for recreational use. All officially designated recreational trails were mapped using data from the NPHK official website. For each 300 × 300 m grid cell [16], trail density was calculated by summing the total length (in meters) of trails within the cell boundaries. Trail length variables were standardised (z-transformed) using the scale function in R to ensure comparability across covariates. These trails are actively promoted by NPHK management and are well known to visitors through multiple information sources, including on-site signposting, the park’s official website, and popular recreational applications such as Komoot and Fietsnet [26–28]. Visitor intensity was further quantified using data from fourteen automated visitor counters installed along officially designated trails (Fig. 1 ). These counters recorded daily visitor numbers throughout the study period. From these data, we calculated mean daily visitor counts separately for weekdays and weekends (including public holidays) from 2021 to 2025 (Fig. 2 ). 2.4 Habitat and landscape metrics To account for habitat-related variation, we integrated landscape composition data derived from Belgium’s Corine Land Cover map [29]. For each grid cell, we quantified the proportion of the main habitat categories: coniferous forest, deciduous forest, mixed forest, heathland, transitional woodland–shrub, wetland, and urban area. To assess whether finer-scale local habitat information could improve model performance, we also included habitat metrics collected within a 20 m buffer around each camera trap. These included dominant habitat type and the proportion of understory vegetation in three height classes: 120 cm. In addition, we measured the distance (in meters) from each camera trap to the nearest forest edge using Google Maps. Locations situated deeper within the forest were assigned positive values, whereas locations extending into open landscapes were assigned negative values. All numeric habitat covariates, expressed as proportions between 0 and 1, were standardised (z-transformed) using the scale function in R to allow for direct comparison among variables. These metrics ensured that the effects of recreational pressure on wildlife behaviour could be assessed while controlling for differences in habitat composition and structure. 2.5 Activity patterns We analysed activity patterns of humans, roe deer, wild boar, and red fox. Human activity was derived from hourly visitor counter data, whereas wildlife activity was obtained from continuous camera trap detections. Activity analyses were conducted in R using the activity package [30]. To account for seasonal variation in day length, we applied the double anchoring method [31]. Corrected activity patterns were estimated using kernel density functions. We compared activity patterns between weekdays and weekends (including national holidays) for each species separately using the Watson-Wheeler test. Differences in activity levels between weekdays and weekends were assessed using a Wald test, based on the relative proportion of active time estimated from the fitted kernel density functions. Additionally, we calculated the percentage of overlap between weekday and weekend activity patterns [30]. 2.6 Space use We analysed spatial patterns in wildlife detections using Poisson regression models to evaluate how recreational pressure and habitat characteristics influenced detection rates of roe deer, wild boar and red fox. To represent recreational pressure, we included (i) trail density within each grid cell as a proxy for spatial recreational intensity and (ii) a categorical time period variable distinguishing between weekdays and weekends/national holidays as a proxy for overall temporal recreational intensity. To account for potential variation in trail use intensity between weekdays and weekends, we included an interaction term between trail density and time period. We further added an interaction between trail density and distance to forest edge to examine whether the effects of recreation differed depending on proximity to forest boundaries. To ensure that the effects of recreation were not over- or underestimated due to omission of key environmental factors, we incorporated landscape covariates representing habitat composition and structure across grid cells, as described in Section 2.4 . Before fitting models, we screened all covariates for collinearity using the Spearman's rank correlation with a threshold of ρ = |0.7| [32] A total of 24 candidate models were constructed, containing different combinations of recreational and habitat covariates, including interaction terms (Table A2). After model fitting, model selection was conducted using Akaike’s Information Criterion (AIC) to assess the relative quality of the models [33].Top-ranked models with ΔAIC < 2 and greater weight than the null model were deemed competitive. We evaluated the significance of covariate effects of the top-performing model based on p-values, where a p-value < 0.05 (95% confidence level) was considered statistically significant. Furthermore, we used the predict function in R to visualise relevant covariate interactions based on the top-performing model. Analyses were conducted in R [34]. 3. Results Table 1: Comparison of weekday and weekend activity for humans, roe deer, wild boar, and red fox. The Wald test evaluates differences in activity levels, while the Watson-Wheeler test assesses differences in activity patterns. Statistically significant p-values (p < 0.05) are shown in bold. The overlap coefficient indicates the percentage of overlap between activity curves. Wald test (p-value) Whatson-Wheeler test (p-value) % overlap Human week vs weekend 0.2890976 < 0.0001 0.9679006 Roe deer week vs weekend 0.2770225 0.9755 0.971504 Wild boar week vs weekend 0.4977604 0.5115 0.9593221 Red fox week vs weekend 0.7811122 0.2149 0.9075656 3.1.1 Humans Human activity exhibited similar overall diel patterns during weekdays and weekends, with a clear concentration of activity between sunrise and sunset (Fig. 3 ). The Watson-Wheeler test indicated a statistically significant difference between weekday and weekend activity patterns (p < 0.0001; Table 1), whereas the Wald test showed no significant difference in overall activity levels (p = 0.289; Table 1). The activity curves showed that during weekdays, activity remained elevated for a longer period into the evening, particularly around and after the twilight period. In contrast, during weekends, activity was slightly higher from shortly after sunrise until the afternoon (Fig. 3 ). 3.1.2 Roe deer Roe deer exhibited a clear crepuscular activity pattern, with peaks occurring around sunrise and sunset (Fig. 4 ). The Watson-Wheeler test indicated no significant difference in activity patterns between weekdays and weekends (p = 0.975; Table 1), and the Wald test similarly showed no significant difference in overall activity levels (p = 0.277; Table 1). 3.1.3 Wild boar Wild boar exhibited a predominantly nocturnal activity pattern, with activity peaking during nighttime and remaining elevated until the end of the sunrise twilight (Fig. 5 ). The Watson-Wheeler test indicated no significant difference in activity patterns between weekdays and weekends (p = 0.512; Table 1), and the Wald test showed no significant difference in overall activity levels (p = 0.498; Table 1). 3.1.4 Red Fox Red fox displayed a predominantly nocturnal activity pattern, with distinct peaks shortly after sunset twilight and before sunrise twilight (Fig. 6 ). The Watson-Wheeler test indicated no significant difference in activity patterns between weekdays and weekends (p = 0.215; Table 1), and the Wald test similarly revealed no significant difference in activity levels (p = 0.781; Table 1). Furthermore, we observed a 91% overlap between red fox’ week and weekend activity. 3.2. Space use In the following results, we use the terms “negative effect” and “positive effect” solely to describe the direction of statistical relationships between covariates and the detection rates of roe deer, wild boar, and red fox in NPHK. These terms do not imply beneficial or detrimental impacts on the species. A negative effect indicates that an increase in a covariate is associated with a decrease in detection rates, whereas a positive effect indicates that higher covariate values correspond to increased detection rates. 3.2.1 Roe deer For roe deer, the top-performing model (mod4.3; Table A3) included habitat at the camera site (coniferous forest, deciduous forest, heathland, woodland-shrub), undergrowth vegetation (70–130 cm height), an interaction between trail density and distance to forest edge, and an interaction between trail density and the week-weekend period. Model estimates showed no significant interaction between the week-weekend period and trail density. Therefore, we removed this interaction term from the model to obtain a more parsimonious structure. The simplified model, which retained only the main effect of the week-weekend period, resulted in an AIC difference of less than 2 compared to the previous top model, indicating comparable model performance (Table A3). Following the principle of parsimony, we therefore selected the simpler model without the interaction term between week-weekend period and trail density as the final model. In this final model, the week-weekend period had a significant positive effect on roe deer detection rates (Table A4), indicating overall higher detections during weekends and holidays compared to weekdays. Distance to the forest edge had a strong positive effect, with higher detection rates occurring deeper within forested areas. A significant interaction between trail density and distance to forest edge (Fig. 7 ) further indicated that the negative influence of trail density on detection rates was most pronounced in open landscapes away from the forest edge, showing a slight inverse yield curve when near the forest edge or further away into open landscape, while this negative effect and the corresponding inverse yield-like curve diminished deeper inside the forest. 3.2.2 Wild boar For wild boar, the top-performing model (mod4.5; Table A3) included habitat at the camera site (coniferous forest, deciduous forest, heathland, woodland-shrub), undergrowth vegetation (0–130 cm), an interaction between trail density and distance to forest edge, and an interaction between trail density and week-weekend period. Model estimates indicated that the interaction between trail density and the week-weekend period was not significant. To achieve a more parsimonious model structure, this interaction term was removed. The simplified model, which retained the week-weekend period as a main effect, yielded an AIC difference of less than 2 compared to the more complex model, suggesting equivalent model support (Table A3). Following the principle of parsimony, we retained this simpler model for further analysis. In the final model, the week-weekend period had a significant positive effect on wild boar detection rates, with detections being higher during weekends and holidays than on weekdays (Table A5). Trail density showed a significant negative relationship with detection rates, while distance to the forest edge had a positive effect, indicating that wild boar were more frequently detected deeper within forested areas. A significant interaction between trail density and distance to forest edge (Fig. 7 ) further showed that the negative effect of trail density was strongest in open landscapes further away from the forest edge, where the slope of the relationship was steep and curved downward in an inverse yield shape, more pronounced than in roe deer. Deeper inside the forest, this relationship flattened considerably, indicating that the effect of trail density on detection rates diminished with increasing distance from the forest edge, deeper into the forest. 3.2.3 Red Fox For red fox, the top-performing model (mod4; Table A3) included grid-based habitat proportions (coniferous forest, mixed forest, woodland-shrub, heathland, wetland, and urban), an interaction between trail density and distance to forest edge, and an interaction between trail density and the week-weekend period. Model estimates indicated that the interaction between trail density and the week-weekend period was not significant. Similar as for roe deer and wild boar, we removed the interaction term to obtain a more parsimonious structure. In the simplified model, the week-weekend period also showed no significant effect on detection rates. Consequently, we removed the week-weekend period covariate as well. The final model, which excluded both the interaction and the week-weekend period, performed best (ΔAIC < 2) (Table A3). Trail density had a significant negative effect on detection rates, indicating reduced red fox activity with increasing trail density (Table A6). A significant interaction between trail density and distance to forest edge (Fig. 7 ) revealed a pattern consistent with that observed for roe deer and wild boar, whereby the negative effect of trail density on detection rates was strongest further away from the forest edge into open landscape and gradually weakened further inside the forest. However, contrary to roe deer and wild boar, the deeper into the forest, the lower the red fox detection rates. Finally, the proportion of urban area showed a significant positive effect, indicating higher detection rates in areas with more urbanised environments. 4. Discussion Using four years of continuous camera-trap data, we examined (i) whether increased visitor pressure during weekend and national holiday periods affected wildlife activity patterns, (ii) how spatial variation in recreational trail density influenced detection rates during weekdays and weekends, and (iii) how landscape features, particularly distance to forest edge, modulated recreation effects. By jointly analysing spatial and temporal dimensions of recreation, our study offers new insights into how wildlife responds to varying intensities of human presence in recreation-dominated protected areas. This approach contributes to a broader understanding of wildlife-recreation dynamics in densely visited protected areas, a setting that is becoming increasingly representative of protected areas across Western Europe and beyond. 4.1 No weekend effect on wildlife activity patterns or detection rates Roe deer, wild boar, and red fox did not exhibit any differences in activity patterns between weekdays and weekends, suggesting that a doubling of relative daily visitor numbers during weekends and holidays (Fig. 2 ) does not markedly alter their diel activity (Fig. 4 – 6 ). This pattern is likely explained by the fact that the activity peaks of all three species minimally overlap with human activity (Fig. 3 – 6 ), and that human week-weekend activity patterns, although statistically different, did not visibly shift towards a more nocturnal orientation. Hence, the predominantly crepuscular to nocturnal activity patterns observed for roe deer, wild boar and red fox may already represent a long-term adjustment to human disturbance. This encompasses an established temporal avoidance strategy that enables these species to minimise encounters with humans [6–10,15,35]. However, despite being most active during periods of low human activity, all three species consistently avoided areas with dense recreational trail networks. This pattern indicates that, although temporal avoidance can reduce direct encounters, it does not prevent spatial displacement from heavily used recreational zones [16]. Consequently, temporal avoidance alone may be insufficient to secure coexistence in high-use protected areas. Hence, maintaining or creating spatial refuges remains critical for the persistence of wildlife [9,36] Furthermore, for all study species, the non-significance of the interaction term between week-weekend period and trail density, demonstrates that, although visitor numbers roughly double on weekends and holidays, this increase does not amplify the general negative effect of trail density on wildlife detection rates. Taken together, we argue that recreational pressure along designated trails in the study area is already sufficiently high, such that a doubling in visitor numbers during weekends or holidays does not produce detectable changes in the effects of recreational trails on wildlife in NPHK. Nevertheless, the main effect of the weekend period itself was significantly positive for roe deer and wild boar, indicating overall higher detection rates during weekends and holidays, whereas red fox detection rates were unaffected. This apparent increase in camera trap detections during periods of higher human presence could result from greater alertness and movement of animals in response to disturbance, which raises the probability of detection [37]. Unlike for roe deer and wild boar, the top-performing model for red fox included corine land cover-based habitat proportions rather than local camera-site habitat variables. Among these, the proportion of urban area had a strong positive effect on detection rates, consistent with previous findings from the same area showing higher red fox space use intensity in more urbanised zones during certain seasons [16]. This pattern aligns with the species’ status as an urban adapter, capable of thriving in human-modified environments through flexible diet and denning behaviour [13,38,39]. Consequently, the superior performance of models including corine habitat proportions likely reflects the importance of urbanisation, which was not captured by the self-assessed camera site-specific habitat covariates. Nevertheless, red fox avoided trail dense areas, which is in contrast with other studies finding no, or positive associations between red fox space use and recreational trails [9,10,16,40]. The reasons underlying this discrepancy remain unclear. 4.2 Recreation effects are landscape context dependent Even within spatial refuges, factors such as habitat suitability and the broader management context and configuration of the landscape can strongly shape wildlife-recreation dynamics [20–22]. Moreover, the effects of recreation can depend on the type of recreation, seasonality, species-specific traits, and life-history characteristics [6,14,16], further underscoring the complexity of understanding and managing human-wildlife interactions in multi-species systems. Understanding these nuances is therefore essential for developing effective management strategies. Across all studied species, distance to the forest edge emerged as a key predictor of detection rates, substantially improving model performance and reducing AIC values, highlighting its strong explanatory value. For roe deer and wild boar, detection rates increased with greater (positive) distance from the forest edge, indicating a preference for interior forest habitats (Table A4-A5, Fig. 7 ). In contrast, red fox exhibited the opposite pattern, with higher detection rates closer to forest edges and more into open landscapes (Table A6, Fig. 7 ). Most notably, we found that the negative effect of recreational trail density on detection rates of roe deer, wild boar, and red fox diminished with increasing distance into the forest (Fig. 7 ). Our results suggest that within forest interiors, the impact of trail density weakens and can even level off entirely. Conversely, in open landscapes far from the forest edge, the relationship followed an inverse yield-like pattern, with detection rates declining sharply as trail density increased, before reaching an apparent threshold beyond which additional trail density had little further effect. These findings highlight the importance of maintaining or restoring forested spatial refuges as buffers against recreational disturbance. Moreover, we argue that in densely visited PAs, landscape configuration plays a crucial role in mediating wildlife responses to recreation, as it determines whether animals can find sufficient refuge from human disturbance. This underscores that managing spatial structure is as important as managing visitor numbers when planning or maintaining recreational infrastructure. Although our findings suggest potential temporal avoidance of humans, evidenced by limited overlap in activity patterns, as well as spatial avoidance of trail-dense areas and a buffering effect of forested habitats on recreational impact, it remains challenging to fully capture how these spatiotemporal mechanisms interact. This is particularly true during daytime, when low wildlife activity results in few detections, constraining the ability of camera-trap studies to disentangle diurnal from nocturnal responses. Future research would benefit greatly from large-scale, multi-species tracking efforts using GPS or bio-logger data, combined with high-resolution spatiotemporal information on human activity. Such an integrated approach would allow for a more continuous and mechanistic understanding of human-wildlife interactions in densely visited protected areas. Declarations Funding This work makes use of data and/or infrastructure provided by INBO and UHasselt and funded by Research Foundation Flanders (FWO) as part of the Belgian contribution to LifeWatch. Wim Kuypers is a joint PhD fellow funded by a BOF mandate at Hasselt University and the University of Namur. Additional information The authors declare no competing interests. Author Contribution Wim Kuypers: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Conceptualization. Jim Casaer: Writing – review & editing, Validation, Supervision, Project administration, Methodology, Funding acquisition. Nicolas Dendoncker: Writing – review & editing, Validation, Supervision, Project administration, Methodology, Funding acquisition. Natalie Beenaerts: Writing – review & editing, Validation, Supervision, Project administration, Methodology, Funding acquisition. Acknowledgement We are grateful to ANB, Regionaal Landschap Kempen & Maasland, the municipalities of As, Dilsen-Stokkem, Lanaken, Maasmechelen and Zutendaal, the tourist offices of NPHK, hunters and residents for allowing us to place camera traps on their property. Further, we thank all colleagues, students and volunteers who aided in the field or processed and annotated pictures. Data Availability Data will be made available on request. References IUCN. Guidelines for Protected Area Management Categories. (1994). IPBES. The Global Assessment Report on BIODIVERSITY AND ECOSYSTEM SERVICES. (2019). Kareiva, P., Watts, S., McDonald, R. & Boucher, T. Domesticated nature: shaping landscapes and ecosystems for human welfare. Science (1979) 316, 1866–1869 (2007). Venter, Z. S., Gundersen, V., Scott, S. L. & Barton, D. N. Bias and precision of crowdsourced recreational activity data from Strava. Landsc Urban Plan 232, 104686 (2023). Ferguson, M. D. et al. The nature of the pandemic: Exploring the negative impacts of the COVID-19 pandemic upon recreation visitor behaviors and experiences in parks and protected areas. Journal of Outdoor Recreation and Tourism 41, 100498 (2023). Burton, A. C. et al. Mammal responses to global changes in human activity vary by trophic group and landscape. Nat Ecol Evol (2024). Gaynor, K. M., Hojnowski, C. E., Carter, N. H. & Brashares, J. S. The influence of human disturbance on wildlife nocturnality. Science (1979) 360, 1232–1235 (2018). Nickel, B. A., Suraci, J. P., Allen, M. L. & Wilmers, C. C. Human presence and human footprint have non-equivalent effects on wildlife spatiotemporal habitat use. Biol Conserv 241, 108383 (2020). Lewis, J. S. et al. Human activity influences wildlife populations and activity patterns: implications for spatial and temporal refuges. Ecosphere 12, e03487 (2021). Anderson, A. K., Waller, J. S. & Thornton, D. H. Partial COVID-19 closure of a national park reveals negative influence of low-impact recreation on wildlife spatiotemporal ecology. Sci Rep 13, 687 (2023). Berger, J. Fear, human shields and the redistribution of prey and predators in protected areas. Biol Lett 3, 620–623 (2007). Bubnicki, J. W., Churski, M., Schmidt, K., Diserens, T. A. & Kuijper, D. P. J. Linking spatial patterns of terrestrial herbivore community structure to trophic interactions. Elife 8, e44937 (2019). Handler, A. M., Lonsdorf, E. V & Ardia, D. R. Evidence for red fox (Vulpes vulpes) exploitation of anthropogenic food sources along an urbanization gradient using stable isotope analysis. Can J Zool 98, 79–87 (2020). Suraci, J. P. et al. Disturbance type and species life history predict mammal responses to humans. Glob Chang Biol 27, 3718–3731 (2021). Procko, M. et al. Quantifying impacts of recreation on elk (Cervus canadensis) using novel modeling approaches. Ecosphere 15, e4873 (2024). Kuypers, W., Bollen, M., Casaer, J. & Beenaerts, N. Trapped by trails: How different types of recreational trails influence seasonal space use of wildlife in a densely visited national park. Science of the total environment 996, 180091 (2025). Boone, H. M. et al. Recreational trail use alters mammal diel and space use during and after COVID-19 restrictions in a US national park. Glob Ecol Conserv 57, e03363 (2025). Nix, J. H., Howell, R. G., Hall, L. K. & McMillan, B. R. The influence of periodic increases of human activity on crepuscular and nocturnal mammals: Testing the weekend effect. Behavioural Processes 146, 16–21 (2018). Green, A. M. et al. Variation in human diel activity patterns mediates periodic increases in recreational activity on mammal behavioural response: investigating the presence of a temporal ‘weekend effect’. Anim Behav 198, 117–129 (2023). Coppes, J., Burghardt, F., Hagen, R., Suchant, R. & Braunisch, V. Human recreation affects spatio-temporal habitat use patterns in red deer (Cervus elaphus). PLoS One 12, e0175134 (2017). Coppes, J. et al. Habitat suitability modulates the response of wildlife to human recreation. Biol Conserv 227, 56–64 (2018). Marion, S. et al. Mammal responses to human recreation depend on landscape context. PLoS One 19, e0300870 (2024). Klimaatstatistieken van de Belgische Gemeenten Maasmechelen (Nis 73107). www.meteo.be. Statbel. https://statbel.fgov.be/nl/themas/bevolking/structuur-van-de-bevolking/bevolkingsdichtheid#figure Indestege, S., Bollen, M., Casaer, J. & Beenaerts, N. Evaluating Ecological Inferences from Camera Trap Data: A Comparative Analysis of Two Sampling Designs. (2025). Fietsnet. https://www.fietsnet.be/routeplanner/default.aspx Hoge Kempen National Park. https://www.nationaalparkhogekempen.be/en Komoot. https://www.komoot.com/ EEA. Corine Land Cover . (2020). Rowcliffe, J. M., Kays, R., Kranstauber, B., Carbone, C. & Jansen, P. A. Quantifying levels of animal activity using camera trap data. Methods Ecol Evol 5, 1170–1179 (2014). Nouvellet, P., Rasmussen, G. S. A., Macdonald, D. W. & Courchamp, F. Noisy clocks and silent sunrises: measurement methods of daily activity pattern. J Zool 286, 179–184 (2012). Dormann, C. F. et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36, 27–46 (2013). Akaike, H. A new look at the statistical model identification. IEEE Trans Automat Contr 19, 716–723 (1974). R Core Team. (2023). R: A Language and Environment for Statistical Computing (Version 4.3.1) [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/. Fennell, M. J. E., Ford, A. T., Martin, T. G. & Burton, A. C. Assessing the impacts of recreation on the spatial and temporal activity of mammals in an isolated alpine protected area. Ecol Evol 13, e10733 (2023). Larson, C. L., Reed, S. E., Merenlender, A. M. & Crooks, K. R. A meta-analysis of recreation effects on vertebrate species richness and abundance. Conserv Sci Pract 1, e93 (2019). Olejarz, A. et al. Worse sleep and increased energy expenditure yet no movement changes in sub-urban wild boar experiencing an influx of human visitors (anthropulse) during the COVID-19 pandemic. Science of The Total Environment 879, 163106 (2023). DeCandia, A. L. et al. Urban colonization through multiple genetic lenses: The city-fox phenomenon revisited. Ecol Evol 9, 2046–2060 (2019). Gil-Fernandez, M., Harcourt, R., Newsome, T., Towerton, A. & Carthey, A. Adaptations of the red fox (Vulpes vulpes) to urban environments in Sydney, Australia. Journal of urban ecology 6, juaa009 (2020). Naidoo, R. & Burton, A. C. Relative effects of recreational activities on a temperate terrestrial wildlife assemblage. Conserv Sci Pract 2, e271 (2020). Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterials.pdf Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7926808","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":537021279,"identity":"7f5bc42d-eb78-427a-bba1-65e16c8368ab","order_by":0,"name":"Wim Kuypers","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+UlEQVRIiWNgGAWjYDACZgYGxgYIk/EBhMXcABcipIXZAMJibGzAq4UBoYVNgigtBse5Exhnth3O45c+/qyad4edvXx7Y/sDxh02uLUc5t3AuLHtcLFkX47Zbd4zyYkbzhwE2nImDacWM5CWh22HgSp52G7ztjEnGEgkArW0HSasZf8Z9mfFvG319vLzH4K0/MevZSPIFh4GM2betsOMDTdA3m87gFOLPVDLwRnn0oslzvAYS85tOw50YWLjjMQzyTi1SPaf3fiwp8w6j7+H/eGHt23VwBA7fODDxx12OLWAANANzQmoQgmYqtBBHRFqRsEoGAWjYMQCAISbXpvjOAyoAAAAAElFTkSuQmCC","orcid":"","institution":"Hasselt University","correspondingAuthor":true,"prefix":"","firstName":"Wim","middleName":"","lastName":"Kuypers","suffix":""},{"id":537021281,"identity":"6008c318-77a0-4ffd-b222-a9af0e88246c","order_by":1,"name":"Jim Casaer","email":"","orcid":"","institution":"Research Institute for Nature and Forest","correspondingAuthor":false,"prefix":"","firstName":"Jim","middleName":"","lastName":"Casaer","suffix":""},{"id":537021283,"identity":"226bc739-7ef1-49f3-8d05-10fd39c193bf","order_by":2,"name":"Nicolas Dendoncker","email":"","orcid":"","institution":"University of Namur","correspondingAuthor":false,"prefix":"","firstName":"Nicolas","middleName":"","lastName":"Dendoncker","suffix":""},{"id":537021285,"identity":"0b45cf35-7462-411e-aba5-3ca67ee5f29a","order_by":3,"name":"Natalie Beenaerts","email":"","orcid":"","institution":"Hasselt University","correspondingAuthor":false,"prefix":"","firstName":"Natalie","middleName":"","lastName":"Beenaerts","suffix":""}],"badges":[],"createdAt":"2025-10-22 22:08:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7926808/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7926808/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97856380,"identity":"df4b7c31-b4e9-403b-a0f6-19fb01ef9686","added_by":"auto","created_at":"2025-12-10 07:56:05","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3511250,"visible":true,"origin":"","legend":"","description":"","filename":"ManuscriptWimKuypers.docx","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/ff85719220999beb6fa16436.docx"},{"id":97856356,"identity":"436b3110-8393-44bb-a318-795169a48834","added_by":"auto","created_at":"2025-12-10 07:56:00","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6423,"visible":true,"origin":"","legend":"","description":"","filename":"0dcae98e3bd14713ac594757bdc6bc9a.json","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/77b860d8a1df6ac55dfb140b.json"},{"id":97856373,"identity":"5bf8bac8-e86f-458a-a6b3-775e13d1c9da","added_by":"auto","created_at":"2025-12-10 07:56:04","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":251761,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterials.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/c6a73f5dbb73fb7d4ba66402.pdf"},{"id":97856382,"identity":"c8d1ec07-eafa-4d68-9f9d-47e1f7aaa703","added_by":"auto","created_at":"2025-12-10 07:56:05","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":94256,"visible":true,"origin":"","legend":"","description":"","filename":"0dcae98e3bd14713ac594757bdc6bc9a1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/ab484102cbda31392e9a87a5.xml"},{"id":97856372,"identity":"8c3129ec-8f1c-4a53-84e5-a5f0b781b1c2","added_by":"auto","created_at":"2025-12-10 07:56:04","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2657615,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/aaecddee538dd033c44e89ef.png"},{"id":97856363,"identity":"605f9253-a97e-481f-ba36-094229407d2e","added_by":"auto","created_at":"2025-12-10 07:56:01","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":89898,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/145a156917948b526799fb76.jpeg"},{"id":97856381,"identity":"3710b700-4796-4008-8b9c-5f1d26e8729a","added_by":"auto","created_at":"2025-12-10 07:56:05","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":310708,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/9f4361bb53a23cf5044233dc.jpeg"},{"id":97856357,"identity":"08333b24-40d8-41b2-b525-88d0e5c55af8","added_by":"auto","created_at":"2025-12-10 07:56:00","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":135014,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/36af47d9e7e75fcca25c81a3.png"},{"id":97856362,"identity":"d5e73109-408d-492d-a511-7d0b448aa770","added_by":"auto","created_at":"2025-12-10 07:56:01","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":121993,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/3e3efc49d28f2efa44a8a51d.png"},{"id":97856376,"identity":"ccc615a5-5ed1-4a97-b775-6037c29a9e78","added_by":"auto","created_at":"2025-12-10 07:56:05","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":121035,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/98e8558b0cfb4385407842ad.png"},{"id":97856388,"identity":"ce4b1647-3daf-4c5d-8251-72ae7991f5c4","added_by":"auto","created_at":"2025-12-10 07:56:07","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":135398,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/2df10030dd02c65317fa18cb.png"},{"id":97856399,"identity":"16a4768a-76d7-4040-a810-d0e1d94b4500","added_by":"auto","created_at":"2025-12-10 07:56:10","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":204315,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/087c59ce12fad435cecbd02e.png"},{"id":97856385,"identity":"1cf6a1b0-811a-4a22-b9d7-d1de2dfff732","added_by":"auto","created_at":"2025-12-10 07:56:06","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":364751,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/279ef5be847dc3095797766c.png"},{"id":97856358,"identity":"063b3cbe-4b21-4f48-8d70-079328a27eda","added_by":"auto","created_at":"2025-12-10 07:56:01","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8390,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/661089b25427e17dfbb4c8aa.png"},{"id":97856369,"identity":"9305c973-1230-4ede-aca9-577c93364c48","added_by":"auto","created_at":"2025-12-10 07:56:03","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":44064,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/562cfc5eb797d6a04f741c72.png"},{"id":97900167,"identity":"da3208ab-bcfd-458a-a809-f59c2e1f23ea","added_by":"auto","created_at":"2025-12-10 15:45:16","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35463,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/c9f56f1040822cf23e73c78f.png"},{"id":97856359,"identity":"1c06f3de-994b-442b-909a-3ede06cdcd55","added_by":"auto","created_at":"2025-12-10 07:56:01","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":31849,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/fb690a3d615798e7702f3b19.png"},{"id":97856386,"identity":"6068dc09-c673-43d2-9196-c4ec66b463ec","added_by":"auto","created_at":"2025-12-10 07:56:06","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":31315,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/faca60842094234ae2bd297e.png"},{"id":97856366,"identity":"e0b0f868-1d6c-472b-ba84-5d6b76119a22","added_by":"auto","created_at":"2025-12-10 07:56:02","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35772,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/c6734be06a5c7787218e5948.png"},{"id":97856365,"identity":"69b92cfe-ab7f-49d0-986e-2e46398ed286","added_by":"auto","created_at":"2025-12-10 07:56:01","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39002,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/1e684ad7b64a8a75110dfb84.png"},{"id":97856379,"identity":"503922f1-fb40-4dfc-9139-91e3dce44eb3","added_by":"auto","created_at":"2025-12-10 07:56:05","extension":"xml","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":86787,"visible":true,"origin":"","legend":"","description":"","filename":"0dcae98e3bd14713ac594757bdc6bc9a1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/992d05cf939e830170e22dd0.xml"},{"id":97856394,"identity":"1974708d-bb92-461a-b30d-3fc728483c1a","added_by":"auto","created_at":"2025-12-10 07:56:08","extension":"html","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":105101,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/d5a5fab8e9d9d5e760a03f61.html"},{"id":97856390,"identity":"36a3c539-7ee3-4080-8aad-b22e456eb30e","added_by":"auto","created_at":"2025-12-10 07:56:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":999171,"visible":true,"origin":"","legend":"\u003cp\u003eMap of the study area in Hoge Kempen National Park (Belgium). Two subsets of 30 camera trap locations (white and yellow coloured), and officially designated recreational trails are illustrated on the map. The inset map (lower left) shows the study area within Belgium.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/50160e24b03714ba04db22af.png"},{"id":97856367,"identity":"d0daebf4-5107-4ceb-bc4e-4644db7f94cc","added_by":"auto","created_at":"2025-12-10 07:56:03","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":323047,"visible":true,"origin":"","legend":"\u003cp\u003eHistogram showing the difference in average week (Blue) and weekend/holiday (Orange) day counts, including standard errors, based on data from 14 visitor counters across the study area between 2021 and 2025.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/9960a9562d7d0aff5678423b.png"},{"id":97856387,"identity":"45416af8-3172-4b7e-b8df-a2be8eacf7ae","added_by":"auto","created_at":"2025-12-10 07:56:06","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":216999,"visible":true,"origin":"","legend":"\u003cp\u003eWeek-weekend activity patterns of humans in NPHK based on solar times. The dark grey zone indicates night-time, light grey zones indicate twilight periods, and white zones indicate daytime. Blue line indicates the weekday activity pattern, and the orange line indicates the weekend/holiday activity pattern. Solid lines show the mean activity trend, while the dashed lines represent the 95% CI.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/fc5cbc7f1514a2bd499bbaa7.png"},{"id":97856383,"identity":"ec21f249-7fa8-4d21-9d98-9fb5d4cd3374","added_by":"auto","created_at":"2025-12-10 07:56:05","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":130947,"visible":true,"origin":"","legend":"\u003cp\u003eWeek-weekend activity patterns of roe deer in NPHK based on solar times. The dark grey zone indicates night-time, light grey zones indicate twilight periods, and white zones indicate daytime. Blue line indicates the weekday activity pattern, and the orange line indicates the weekend/holiday activity pattern. Solid lines show the mean activity trend, while the dashed lines represent the 95% CI.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/d20751e92687b99efabd070e.png"},{"id":97856384,"identity":"ccd3e06c-49fb-493d-b0b4-ddef87faad60","added_by":"auto","created_at":"2025-12-10 07:56:06","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":212605,"visible":true,"origin":"","legend":"\u003cp\u003eWeek-weekend activity patterns of wild boar in NPHK based on solar times. The dark grey zone indicates night-time, light grey zones indicate twilight periods, and white zones indicate daytime. Blue line indicates the weekday activity pattern, and the orange line indicates the weekend/holiday activity pattern. Solid lines show the mean activity trend, while the dashed lines represent the 95% CI.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/731bd9fd06b4cc6c696f2cb5.png"},{"id":97856370,"identity":"3f1cd854-8e47-4b26-9b25-18263e2cf974","added_by":"auto","created_at":"2025-12-10 07:56:04","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":228857,"visible":true,"origin":"","legend":"\u003cp\u003eWeek-weekend activity patterns of red fox in NPHK based on solar times. The dark grey zone indicates night-time, light grey zones indicate twilight periods, and white zones indicate daytime. Blue line indicates the weekday activity pattern, and the orange line indicates the weekend/holiday activity pattern. Solid lines show the mean activity trend, while the dashed lines represent the 95% CI.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/764959acc4a609224a329991.png"},{"id":97856374,"identity":"55c08ad9-b219-42aa-962e-33e512b8a154","added_by":"auto","created_at":"2025-12-10 07:56:04","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":271185,"visible":true,"origin":"","legend":"\u003cp\u003eInteraction effect between trail density and distance to forest edge in relation to detection rates of roe deer, wild boar and red fox. The Y-axis shows 24-hour detection rates, while the X-axis indicates trail density in 300x300m grids around camera traps. Coloured lines indicate the effects of recreational trail density over a variety of distances to forest edges. The colours range from dark green to dark red, indicating the distance from the camera trap to the forest edge. Dark green being 800 meters from the\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/a2d5cf9d17cce09abfbef6dd.png"},{"id":104217838,"identity":"9942eab5-81cd-4e5c-92fe-b61e6f23651e","added_by":"auto","created_at":"2026-03-09 09:27:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2971942,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/9aeab891-ed7f-4e4b-9227-f0e54e97db76.pdf"},{"id":97856368,"identity":"c4a78a82-eaf1-4772-9747-57d31116199b","added_by":"auto","created_at":"2025-12-10 07:56:03","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":251761,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterials.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7926808/v1/a8676ae5e59f16dadba779e3.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Wildlife responses to recreational trail density are mediated by landscape configuration rather than short-term increases in recreation intensity","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eProtected areas (PAs) play a vital role in global biodiversity conservation while also serving as places for public recreation and enjoyment [1,2]. This dual mandate creates inherent management challenges, particularly as human activities continue to modify natural landscapes [3.4]. At the same time, recreational use of PAs is rising worldwide [5], leading to increasing levels of disturbance for wildlife. Hence, wildlife survival in the Anthropocene relies on the ability of species to cope with landscape modifications and intensifying human presence [6].\u003c/p\u003e\u003cp\u003eWildlife responses to human activity are highly variable and shaped by multiple factors. Some species perceive humans as threats and adjust their behaviour by avoiding areas of high human activity or shifting their activity to periods with lower human presence [6\u0026ndash;10]. Conversely, other species may exploit human-modified environments, gaining access to resources or protection from predators [11\u0026ndash;13]. These divergent responses can be shaped not only by species-specific traits such as body size, ecological niche, and life-history characteristics [14], but also by extrinsic factors like the type and intensity of recreation [15, 16]. Abrupt changes in human activity, such as those caused by the COVID-19 pandemic, provided a unique opportunity to investigate how wildlife responds to substantial, sustained reductions or increases in human disturbance over extended periods. These studies revealed that animals can rapidly adjust spatiotemporal behaviour in response to marked changes in human presence [6,10,17]. In contrast, the global work-week generates recurring, short-term fluctuations in human activity, with visitor numbers in protected areas typically peaking during weekends or national holidays. Despite this widespread pattern, empirical studies examining how such short-term, periodic increases in recreational use simultaneously affect the spatial and temporal behaviour of wildlife remain scarce [18,19].\u003c/p\u003e\u003cp\u003eTo cope with these fluctuations, wildlife may rely on temporal and spatial refuges. Temporal refuges, such as nocturnal or crepuscular activity, allow animals to avoid peak periods of human presence, while spatial refuges, such as forested interiors or areas distant from trails, provide shelter from disturbance and help maintain essential behaviours [7\u0026ndash;9]. However, the effectiveness of these refuges is strongly influenced by the configuration and management of the surrounding landscape. For instance, habitat suitability and broader landscape context can either amplify or buffer the impacts of recreational activities, shaping whether temporal or spatial refuges are sufficient for species persistence [20\u0026ndash;22].\u003c/p\u003e\u003cp\u003eIn this way, landscape configuration emerges as another extrinsic factor determining how wildlife responds to human disturbance and how effectively refuges function. Taken together, these responses determine how wildlife can coexist with humans and persist in protected areas with high recreational use [6]. Despite these insights, the interplay between species traits, changes in recreation intensity, and landscape configuration remains poorly understood. This knowledge gap limits the ability of managers to design interventions that balance human use and wildlife conservation.\u003c/p\u003e\u003cp\u003eBy integrating both spatial (trail density) and temporal (week-weekend) recreational covariates and considering interactions with landscape features such as distance to the forest edge, our study aims to address this gap. We specifically investigate how weekend and holiday visitor peaks, recreational trail density, and distance to forest edges influence activity and camera trap detection rates of roe deer, wild boar, and red fox in the densely visited Hoge Kempen National Park (Belgium).\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Study area\u003c/h2\u003e\u003cp\u003eThe study area (longitudes: 5.552\u0026deg;W \u0026minus;\u0026thinsp;5.703\u0026deg;W; latitudes: 50.899\u0026deg;N \u0026minus;\u0026thinsp;51.016\u0026deg;N) is situated in the core region of Hoge Kempen National Park (NPHK), eastern Belgium (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). It has a total surface area of \u0026sim;60km2, consisting of large areas of planted pine forests (41% \u003cem\u003ePinus sylvestris\u003c/em\u003e and \u003cem\u003ePinus nigra\u003c/em\u003e). These afforested regions are undergoing a systematic transition toward a more natural deciduous forest ecosystem, characterised by the presence of \u003cem\u003eQuercus\u003c/em\u003e spp. and \u003cem\u003eBetula\u003c/em\u003e spp. (9%). The park features valuable dry (mainly \u003cem\u003eCalluna vulgaris\u003c/em\u003e) and wet (featuring \u003cem\u003eErica\u003c/em\u003e sp. and \u003cem\u003eMyrica\u003c/em\u003e sp.) heathlands (11%), along with shrub vegetation (7%), predominantly dominated by \u003cem\u003eMolinia sp\u003c/em\u003e. NPHK has altitudes ranging from 50 to 100 m above sea level. The study area has a cool, temperate, and moist climate, with a mean annual temperature of 10.9\u0026deg;C and 816.4 mm rainfall [23]. The study area is situated within a densely populated urban matrix, with an average of 443 inhabitants per square kilometre in the surrounding municipalities [24]. Hence, the region is extensively utilised for recreational purposes, including walking, cycling, MTB, horse riding, and hunting. The study area features an exceptionally dense network of recreational trails, with approximately 100 km of advertised hiking trails (1.657 km/km2), 50 km of cycling paths (0.814 km/km2), and 60 km of designated MTB routes (1.008 km/km2). The estimated annual number of visitors within our study area exceeds 500,000 (Visitor counters NPHK (2021\u0026ndash;2025)). Furthermore, there are five official entrance gates at the borders of the study area, providing visitors with parking opportunities and direct access to these various recreational trails. NPHK faces typical challenges, including increasing human recreational pressures, limited data on human usage and wildlife behaviour, and a lack of comprehensive understanding of how these factors influence management strategies. Additionally, the park is becoming increasingly isolated from other prime wildlife habitats due to surrounding road networks, fencing and urban expansion.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Camera trapping network\u003c/h2\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eA stratified random sampling design of 30 motion-sensing camera traps (CTs) (Reconyx HC600 Hyperfire) was applied in NPHK [25]. Sixty sampling locations from a systematic random sampling design (designed by Wevers et al., 2020; described and visualised in detail by Kuypers et al., 2025) were selected to reflect the proportional representation of the main habitat types (heathland, coniferous forest, deciduous forest, mixed forest, and scrubland) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). These locations were divided into two subsets of 30, which were alternately sampled every two months. Camera setup and image processing as described by Kuypers et al., 2025. For this article, we focused on images obtained from January 2021 to January 2025. In total, after accounting for camera malfunction or stolen devices, 60 unique sites were sampled with an average trapping effort of 587 days per site. Within these periods, we identified and documented observations of 13 wildlife species (excluding birds, domestic species and humans, Table A1). Among these, roe deer (\u003cem\u003eCapreolus capreolus\u003c/em\u003e), wild boar (\u003cem\u003eSus scrofa\u003c/em\u003e), and red fox (\u003cem\u003eVulpes vulpes\u003c/em\u003e) provided sufficient data to model the effect of covariates on space use.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Recreational pressure\u003c/h2\u003e\u003cp\u003eTo quantify recreational pressure, we used trail density from officially designated, marked hiking, cycling and MTB trails as a proxy for recreational use. All officially designated recreational trails were mapped using data from the NPHK official website. For each 300 \u0026times; 300 m grid cell [16], trail density was calculated by summing the total length (in meters) of trails within the cell boundaries. Trail length variables were standardised (z-transformed) using the scale function in R to ensure comparability across covariates. These trails are actively promoted by NPHK management and are well known to visitors through multiple information sources, including on-site signposting, the park\u0026rsquo;s official website, and popular recreational applications such as Komoot and Fietsnet [26\u0026ndash;28].\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eVisitor intensity was further quantified using data from fourteen automated visitor counters installed along officially designated trails (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). These counters recorded daily visitor numbers throughout the study period. From these data, we calculated mean daily visitor counts separately for weekdays and weekends (including public holidays) from 2021 to 2025 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.4 Habitat and landscape metrics\u003c/h2\u003e\u003cp\u003eTo account for habitat-related variation, we integrated landscape composition data derived from Belgium\u0026rsquo;s Corine Land Cover map [29]. For each grid cell, we quantified the proportion of the main habitat categories: coniferous forest, deciduous forest, mixed forest, heathland, transitional woodland\u0026ndash;shrub, wetland, and urban area. To assess whether finer-scale local habitat information could improve model performance, we also included habitat metrics collected within a 20 m buffer around each camera trap. These included dominant habitat type and the proportion of understory vegetation in three height classes: \u0026lt;70 cm, 70\u0026ndash;120 cm, and \u0026gt;\u0026thinsp;120 cm. In addition, we measured the distance (in meters) from each camera trap to the nearest forest edge using Google Maps. Locations situated deeper within the forest were assigned positive values, whereas locations extending into open landscapes were assigned negative values.\u003c/p\u003e\u003cp\u003eAll numeric habitat covariates, expressed as proportions between 0 and 1, were standardised (z-transformed) using the scale function in R to allow for direct comparison among variables. These metrics ensured that the effects of recreational pressure on wildlife behaviour could be assessed while controlling for differences in habitat composition and structure.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e2.5 Activity patterns\u003c/h2\u003e\u003cp\u003eWe analysed activity patterns of humans, roe deer, wild boar, and red fox. Human activity was derived from hourly visitor counter data, whereas wildlife activity was obtained from continuous camera trap detections. Activity analyses were conducted in R using the \u003cem\u003eactivity\u003c/em\u003e package [30]. To account for seasonal variation in day length, we applied the double anchoring method [31]. Corrected activity patterns were estimated using kernel density functions. We compared activity patterns between weekdays and weekends (including national holidays) for each species separately using the Watson-Wheeler test. Differences in activity levels between weekdays and weekends were assessed using a Wald test, based on the relative proportion of active time estimated from the fitted kernel density functions. Additionally, we calculated the percentage of overlap between weekday and weekend activity patterns [30].\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e2.6 Space use\u003c/h2\u003e\u003cp\u003eWe analysed spatial patterns in wildlife detections using Poisson regression models to evaluate how recreational pressure and habitat characteristics influenced detection rates of roe deer, wild boar and red fox. To represent recreational pressure, we included (i) trail density within each grid cell as a proxy for spatial recreational intensity and (ii) a categorical time period variable distinguishing between weekdays and weekends/national holidays as a proxy for overall temporal recreational intensity. To account for potential variation in trail use intensity between weekdays and weekends, we included an interaction term between trail density and time period. We further added an interaction between trail density and distance to forest edge to examine whether the effects of recreation differed depending on proximity to forest boundaries. To ensure that the effects of recreation were not over- or underestimated due to omission of key environmental factors, we incorporated landscape covariates representing habitat composition and structure across grid cells, as described in Section \u003cspan refid=\"Sec6\" class=\"InternalRef\"\u003e2.4\u003c/span\u003e. Before fitting models, we screened all covariates for collinearity using the Spearman's rank correlation with a threshold of ρ = |0.7| [32]\u003c/p\u003e\u003cp\u003eA total of 24 candidate models were constructed, containing different combinations of recreational and habitat covariates, including interaction terms (Table A2). After model fitting, model selection was conducted using Akaike\u0026rsquo;s Information Criterion (AIC) to assess the relative quality of the models [33].Top-ranked models with ΔAIC\u0026thinsp;\u0026lt;\u0026thinsp;2 and greater weight than the null model were deemed competitive. We evaluated the significance of covariate effects of the top-performing model based on p-values, where a p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 (95% confidence level) was considered statistically significant. Furthermore, we used the \u003cem\u003epredict\u003c/em\u003e function in R to visualise relevant covariate interactions based on the top-performing model. Analyses were conducted in R [34].\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003e\u003cstrong\u003eTable 1: \u003c/strong\u003eComparison of weekday and weekend activity for humans, roe deer, wild boar, and red fox. The Wald test evaluates differences in activity levels, while the Watson-Wheeler test assesses differences in activity patterns. Statistically significant p-values (p \u0026lt; 0.05) are shown in bold. \u0026nbsp;The overlap coefficient indicates the percentage of overlap between activity curves.\u0026nbsp;\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWald test (p-value)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eWhatson-Wheeler test (p-value)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e% overlap\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHuman week vs weekend\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.2890976\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.0001\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.9679006\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRoe deer week vs weekend\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.2770225\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.9755\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.971504\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWild boar week vs weekend\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.4977604\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.5115\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.9593221\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRed fox week vs weekend\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.7811122\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.2149\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.9075656\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cdiv id=\"Sec10\" class=\"Section3\"\u003e\u003cdiv class=\"Heading\"\u003e3.1.1 Humans\u003c/div\u003e\u003cp\u003eHuman activity exhibited similar overall diel patterns during weekdays and weekends, with a clear concentration of activity between sunrise and sunset (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The Watson-Wheeler test indicated a statistically significant difference between weekday and weekend activity patterns (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; Table\u0026nbsp;1), whereas the Wald test showed no significant difference in overall activity levels (p\u0026thinsp;=\u0026thinsp;0.289; Table\u0026nbsp;1). The activity curves showed that during weekdays, activity remained elevated for a longer period into the evening, particularly around and after the twilight period. In contrast, during weekends, activity was slightly higher from shortly after sunrise until the afternoon (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\u003cdiv class=\"Heading\"\u003e3.1.2 Roe deer\u003c/div\u003e\u003cp\u003eRoe deer exhibited a clear crepuscular activity pattern, with peaks occurring around sunrise and sunset (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The Watson-Wheeler test indicated no significant difference in activity patterns between weekdays and weekends (p\u0026thinsp;=\u0026thinsp;0.975; Table\u0026nbsp;1), and the Wald test similarly showed no significant difference in overall activity levels (p\u0026thinsp;=\u0026thinsp;0.277; Table\u0026nbsp;1).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\u003cdiv class=\"Heading\"\u003e3.1.3 Wild boar\u003c/div\u003e\u003cp\u003eWild boar exhibited a predominantly nocturnal activity pattern, with activity peaking during nighttime and remaining elevated until the end of the sunrise twilight (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The Watson-Wheeler test indicated no significant difference in activity patterns between weekdays and weekends (p\u0026thinsp;=\u0026thinsp;0.512; Table\u0026nbsp;1), and the Wald test showed no significant difference in overall activity levels (p\u0026thinsp;=\u0026thinsp;0.498; Table\u0026nbsp;1).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section3\"\u003e\u003cdiv class=\"Heading\"\u003e3.1.4 Red Fox\u003c/div\u003e\u003cp\u003eRed fox displayed a predominantly nocturnal activity pattern, with distinct peaks shortly after sunset twilight and before sunrise twilight (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). The Watson-Wheeler test indicated no significant difference in activity patterns between weekdays and weekends (p\u0026thinsp;=\u0026thinsp;0.215; Table\u0026nbsp;1), and the Wald test similarly revealed no significant difference in activity levels (p\u0026thinsp;=\u0026thinsp;0.781; Table\u0026nbsp;1). Furthermore, we observed a 91% overlap between red fox\u0026rsquo; week and weekend activity.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e3.2. Space use\u003c/h2\u003e\u003cp\u003eIn the following results, we use the terms \u0026ldquo;negative effect\u0026rdquo; and \u0026ldquo;positive effect\u0026rdquo; solely to describe the direction of statistical relationships between covariates and the detection rates of roe deer, wild boar, and red fox in NPHK. These terms do not imply beneficial or detrimental impacts on the species. A negative effect indicates that an increase in a covariate is associated with a decrease in detection rates, whereas a positive effect indicates that higher covariate values correspond to increased detection rates.\u003c/p\u003e\u003cdiv id=\"Sec15\" class=\"Section3\"\u003e\u003ch2\u003e3.2.1 Roe deer\u003c/h2\u003e\u003cp\u003eFor roe deer, the top-performing model (mod4.3; Table A3) included habitat at the camera site (coniferous forest, deciduous forest, heathland, woodland-shrub), undergrowth vegetation (70\u0026ndash;130 cm height), an interaction between trail density and distance to forest edge, and an interaction between trail density and the week-weekend period. Model estimates showed no significant interaction between the week-weekend period and trail density. Therefore, we removed this interaction term from the model to obtain a more parsimonious structure. The simplified model, which retained only the main effect of the week-weekend period, resulted in an AIC difference of less than 2 compared to the previous top model, indicating comparable model performance (Table A3). Following the principle of parsimony, we therefore selected the simpler model without the interaction term between week-weekend period and trail density as the final model.\u003c/p\u003e\u003cp\u003eIn this final model, the week-weekend period had a significant positive effect on roe deer detection rates (Table A4), indicating overall higher detections during weekends and holidays compared to weekdays. Distance to the forest edge had a strong positive effect, with higher detection rates occurring deeper within forested areas. A significant interaction between trail density and distance to forest edge (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) further indicated that the negative influence of trail density on detection rates was most pronounced in open landscapes away from the forest edge, showing a slight inverse yield curve when near the forest edge or further away into open landscape, while this negative effect and the corresponding inverse yield-like curve diminished deeper inside the forest.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section3\"\u003e\u003ch2\u003e3.2.2 Wild boar\u003c/h2\u003e\u003cp\u003eFor wild boar, the top-performing model (mod4.5; Table A3) included habitat at the camera site (coniferous forest, deciduous forest, heathland, woodland-shrub), undergrowth vegetation (0\u0026ndash;130 cm), an interaction between trail density and distance to forest edge, and an interaction between trail density and week-weekend period. Model estimates indicated that the interaction between trail density and the week-weekend period was not significant. To achieve a more parsimonious model structure, this interaction term was removed. The simplified model, which retained the week-weekend period as a main effect, yielded an AIC difference of less than 2 compared to the more complex model, suggesting equivalent model support (Table A3). Following the principle of parsimony, we retained this simpler model for further analysis.\u003c/p\u003e\u003cp\u003eIn the final model, the week-weekend period had a significant positive effect on wild boar detection rates, with detections being higher during weekends and holidays than on weekdays (Table A5). Trail density showed a significant negative relationship with detection rates, while distance to the forest edge had a positive effect, indicating that wild boar were more frequently detected deeper within forested areas. A significant interaction between trail density and distance to forest edge (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) further showed that the negative effect of trail density was strongest in open landscapes further away from the forest edge, where the slope of the relationship was steep and curved downward in an inverse yield shape, more pronounced than in roe deer. Deeper inside the forest, this relationship flattened considerably, indicating that the effect of trail density on detection rates diminished with increasing distance from the forest edge, deeper into the forest.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec17\" class=\"Section3\"\u003e\u003ch2\u003e3.2.3 Red Fox\u003c/h2\u003e\u003cp\u003eFor red fox, the top-performing model (mod4; Table A3) included grid-based habitat proportions (coniferous forest, mixed forest, woodland-shrub, heathland, wetland, and urban), an interaction between trail density and distance to forest edge, and an interaction between trail density and the week-weekend period. Model estimates indicated that the interaction between trail density and the week-weekend period was not significant. Similar as for roe deer and wild boar, we removed the interaction term to obtain a more parsimonious structure. In the simplified model, the week-weekend period also showed no significant effect on detection rates. Consequently, we removed the week-weekend period covariate as well. The final model, which excluded both the interaction and the week-weekend period, performed best (ΔAIC\u0026thinsp;\u0026lt;\u0026thinsp;2) (Table A3). Trail density had a significant negative effect on detection rates, indicating reduced red fox activity with increasing trail density (Table A6). A significant interaction between trail density and distance to forest edge (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) revealed a pattern consistent with that observed for roe deer and wild boar, whereby the negative effect of trail density on detection rates was strongest further away from the forest edge into open landscape and gradually weakened further inside the forest. However, contrary to roe deer and wild boar, the deeper into the forest, the lower the red fox detection rates. Finally, the proportion of urban area showed a significant positive effect, indicating higher detection rates in areas with more urbanised environments.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eUsing four years of continuous camera-trap data, we examined (i) whether increased visitor pressure during weekend and national holiday periods affected wildlife activity patterns, (ii) how spatial variation in recreational trail density influenced detection rates during weekdays and weekends, and (iii) how landscape features, particularly distance to forest edge, modulated recreation effects. By jointly analysing spatial and temporal dimensions of recreation, our study offers new insights into how wildlife responds to varying intensities of human presence in recreation-dominated protected areas. This approach contributes to a broader understanding of wildlife-recreation dynamics in densely visited protected areas, a setting that is becoming increasingly representative of protected areas across Western Europe and beyond.\u003c/p\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003e4.1 No weekend effect on wildlife activity patterns or detection rates\u003c/h2\u003e\u003cp\u003eRoe deer, wild boar, and red fox did not exhibit any differences in activity patterns between weekdays and weekends, suggesting that a doubling of relative daily visitor numbers during weekends and holidays (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) does not markedly alter their diel activity (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). This pattern is likely explained by the fact that the activity peaks of all three species minimally overlap with human activity (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e), and that human week-weekend activity patterns, although statistically different, did not visibly shift towards a more nocturnal orientation. Hence, the predominantly crepuscular to nocturnal activity patterns observed for roe deer, wild boar and red fox may already represent a long-term adjustment to human disturbance. This encompasses an established temporal avoidance strategy that enables these species to minimise encounters with humans [6\u0026ndash;10,15,35]. However, despite being most active during periods of low human activity, all three species consistently avoided areas with dense recreational trail networks. This pattern indicates that, although temporal avoidance can reduce direct encounters, it does not prevent spatial displacement from heavily used recreational zones [16]. Consequently, temporal avoidance alone may be insufficient to secure coexistence in high-use protected areas. Hence, maintaining or creating spatial refuges remains critical for the persistence of wildlife [9,36]\u003c/p\u003e\u003cp\u003eFurthermore, for all study species, the non-significance of the interaction term between week-weekend period and trail density, demonstrates that, although visitor numbers roughly double on weekends and holidays, this increase does not amplify the general negative effect of trail density on wildlife detection rates. Taken together, we argue that recreational pressure along designated trails in the study area is already sufficiently high, such that a doubling in visitor numbers during weekends or holidays does not produce detectable changes in the effects of recreational trails on wildlife in NPHK. Nevertheless, the main effect of the weekend period itself was significantly positive for roe deer and wild boar, indicating overall higher detection rates during weekends and holidays, whereas red fox detection rates were unaffected. This apparent increase in camera trap detections during periods of higher human presence could result from greater alertness and movement of animals in response to disturbance, which raises the probability of detection [37].\u003c/p\u003e\u003cp\u003e Unlike for roe deer and wild boar, the top-performing model for red fox included corine land cover-based habitat proportions rather than local camera-site habitat variables. Among these, the proportion of urban area had a strong positive effect on detection rates, consistent with previous findings from the same area showing higher red fox space use intensity in more urbanised zones during certain seasons [16]. This pattern aligns with the species\u0026rsquo; status as an urban adapter, capable of thriving in human-modified environments through flexible diet and denning behaviour [13,38,39]. Consequently, the superior performance of models including corine habitat proportions likely reflects the importance of urbanisation, which was not captured by the self-assessed camera site-specific habitat covariates. Nevertheless, red fox avoided trail dense areas, which is in contrast with other studies finding no, or positive associations between red fox space use and recreational trails [9,10,16,40]. The reasons underlying this discrepancy remain unclear.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Recreation effects are landscape context dependent\u003c/h2\u003e\u003cp\u003eEven within spatial refuges, factors such as habitat suitability and the broader management context and configuration of the landscape can strongly shape wildlife-recreation dynamics [20\u0026ndash;22]. Moreover, the effects of recreation can depend on the type of recreation, seasonality, species-specific traits, and life-history characteristics [6,14,16], further underscoring the complexity of understanding and managing human-wildlife interactions in multi-species systems. Understanding these nuances is therefore essential for developing effective management strategies.\u003c/p\u003e\u003cp\u003eAcross all studied species, distance to the forest edge emerged as a key predictor of detection rates, substantially improving model performance and reducing AIC values, highlighting its strong explanatory value. For roe deer and wild boar, detection rates increased with greater (positive) distance from the forest edge, indicating a preference for interior forest habitats (Table A4-A5, Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). In contrast, red fox exhibited the opposite pattern, with higher detection rates closer to forest edges and more into open landscapes (Table A6, Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Most notably, we found that the negative effect of recreational trail density on detection rates of roe deer, wild boar, and red fox diminished with increasing distance into the forest (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Our results suggest that within forest interiors, the impact of trail density weakens and can even level off entirely. Conversely, in open landscapes far from the forest edge, the relationship followed an inverse yield-like pattern, with detection rates declining sharply as trail density increased, before reaching an apparent threshold beyond which additional trail density had little further effect. These findings highlight the importance of maintaining or restoring forested spatial refuges as buffers against recreational disturbance. Moreover, we argue that in densely visited PAs, landscape configuration plays a crucial role in mediating wildlife responses to recreation, as it determines whether animals can find sufficient refuge from human disturbance. This underscores that managing spatial structure is as important as managing visitor numbers when planning or maintaining recreational infrastructure.\u003c/p\u003e\u003cp\u003eAlthough our findings suggest potential temporal avoidance of humans, evidenced by limited overlap in activity patterns, as well as spatial avoidance of trail-dense areas and a buffering effect of forested habitats on recreational impact, it remains challenging to fully capture how these spatiotemporal mechanisms interact. This is particularly true during daytime, when low wildlife activity results in few detections, constraining the ability of camera-trap studies to disentangle diurnal from nocturnal responses. Future research would benefit greatly from large-scale, multi-species tracking efforts using GPS or bio-logger data, combined with high-resolution spatiotemporal information on human activity. Such an integrated approach would allow for a more continuous and mechanistic understanding of human-wildlife interactions in densely visited protected areas.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eThis work makes use of data and/or infrastructure provided by INBO and UHasselt and funded by Research Foundation Flanders (FWO) as part of the Belgian contribution to LifeWatch. Wim Kuypers is a joint PhD fellow funded by a BOF mandate at Hasselt University and the University of Namur.\u003c/p\u003e\n\u003ch2\u003eAdditional information\u003c/h2\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eWim Kuypers: Writing \u0026ndash; review \u0026amp;amp; editing, Writing \u0026ndash; original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Conceptualization. Jim Casaer: Writing \u0026ndash; review \u0026amp;amp; editing, Validation, Supervision, Project administration, Methodology, Funding acquisition. Nicolas Dendoncker: Writing \u0026ndash; review \u0026amp;amp; editing, Validation, Supervision, Project administration, Methodology, Funding acquisition. Natalie Beenaerts: Writing \u0026ndash; review \u0026amp;amp; editing, Validation, Supervision, Project administration, Methodology, Funding acquisition.\u003c/p\u003e\n\u003ch2\u003eAcknowledgement\u003c/h2\u003e\n\u003cp\u003eWe are grateful to ANB, Regionaal Landschap Kempen \u0026amp; Maasland, the municipalities of As, Dilsen-Stokkem, Lanaken, Maasmechelen and Zutendaal, the tourist offices of NPHK, hunters and residents for allowing us to place camera traps on their property. Further, we thank all colleagues, students and volunteers who aided in the field or processed and annotated pictures.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eData will be made available on request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eIUCN. Guidelines for Protected Area Management Categories. (1994). \u003c/li\u003e\n\u003cli\u003eIPBES. The Global Assessment Report on BIODIVERSITY AND ECOSYSTEM SERVICES. (2019). \u003c/li\u003e\n\u003cli\u003eKareiva, P., Watts, S., McDonald, R. \u0026amp; Boucher, T. Domesticated nature: shaping landscapes and ecosystems for human welfare. Science (1979) 316, 1866\u0026ndash;1869 (2007). \u003c/li\u003e\n\u003cli\u003eVenter, Z. S., Gundersen, V., Scott, S. L. \u0026amp; Barton, D. N. Bias and precision of crowdsourced recreational activity data from Strava. Landsc Urban Plan 232, 104686 (2023). \u003c/li\u003e\n\u003cli\u003eFerguson, M. D. et al. The nature of the pandemic: Exploring the negative impacts of the COVID-19 pandemic upon recreation visitor behaviors and experiences in parks and protected areas. Journal of Outdoor Recreation and Tourism 41, 100498 (2023). \u003c/li\u003e\n\u003cli\u003eBurton, A. C. et al. Mammal responses to global changes in human activity vary by trophic group and landscape. Nat Ecol Evol (2024). \u003c/li\u003e\n\u003cli\u003eGaynor, K. M., Hojnowski, C. E., Carter, N. H. \u0026amp; Brashares, J. S. The influence of human disturbance on wildlife nocturnality. Science (1979) 360, 1232\u0026ndash;1235 (2018). \u003c/li\u003e\n\u003cli\u003eNickel, B. A., Suraci, J. P., Allen, M. L. \u0026amp; Wilmers, C. C. Human presence and human footprint have non-equivalent effects on wildlife spatiotemporal habitat use. Biol Conserv 241, 108383 (2020). \u003c/li\u003e\n\u003cli\u003eLewis, J. S. et al. Human activity influences wildlife populations and activity patterns: implications for spatial and temporal refuges. Ecosphere 12, e03487 (2021). \u003c/li\u003e\n\u003cli\u003eAnderson, A. K., Waller, J. S. \u0026amp; Thornton, D. H. Partial COVID-19 closure of a national park reveals negative influence of low-impact recreation on wildlife spatiotemporal ecology. Sci Rep 13, 687 (2023). \u003c/li\u003e\n\u003cli\u003eBerger, J. Fear, human shields and the redistribution of prey and predators in protected areas. Biol Lett 3, 620\u0026ndash;623 (2007). \u003c/li\u003e\n\u003cli\u003eBubnicki, J. W., Churski, M., Schmidt, K., Diserens, T. A. \u0026amp; Kuijper, D. P. J. Linking spatial patterns of terrestrial herbivore community structure to trophic interactions. Elife 8, e44937 (2019). \u003c/li\u003e\n\u003cli\u003eHandler, A. M., Lonsdorf, E. V \u0026amp; Ardia, D. R. Evidence for red fox (Vulpes vulpes) exploitation of anthropogenic food sources along an urbanization gradient using stable isotope analysis. Can J Zool 98, 79\u0026ndash;87 (2020). \u003c/li\u003e\n\u003cli\u003eSuraci, J. P. et al. Disturbance type and species life history predict mammal responses to humans. Glob Chang Biol 27, 3718\u0026ndash;3731 (2021). \u003c/li\u003e\n\u003cli\u003eProcko, M. et al. Quantifying impacts of recreation on elk (Cervus canadensis) using novel modeling approaches. Ecosphere 15, e4873 (2024). \u003c/li\u003e\n\u003cli\u003eKuypers, W., Bollen, M., Casaer, J. \u0026amp; Beenaerts, N. Trapped by trails: How different types of recreational trails influence seasonal space use of wildlife in a densely visited national park. Science of the total environment 996, 180091 (2025). \u003c/li\u003e\n\u003cli\u003eBoone, H. M. et al. Recreational trail use alters mammal diel and space use during and after COVID-19 restrictions in a US national park. Glob Ecol Conserv 57, e03363 (2025). \u003c/li\u003e\n\u003cli\u003eNix, J. H., Howell, R. G., Hall, L. K. \u0026amp; McMillan, B. R. The influence of periodic increases of human activity on crepuscular and nocturnal mammals: Testing the weekend effect. Behavioural Processes 146, 16\u0026ndash;21 (2018). \u003c/li\u003e\n\u003cli\u003eGreen, A. M. et al. Variation in human diel activity patterns mediates periodic increases in recreational activity on mammal behavioural response: investigating the presence of a temporal \u0026lsquo;weekend effect\u0026rsquo;. Anim Behav 198, 117\u0026ndash;129 (2023). \u003c/li\u003e\n\u003cli\u003eCoppes, J., Burghardt, F., Hagen, R., Suchant, R. \u0026amp; Braunisch, V. Human recreation affects spatio-temporal habitat use patterns in red deer (Cervus elaphus). PLoS One 12, e0175134 (2017). \u003c/li\u003e\n\u003cli\u003eCoppes, J. et al. Habitat suitability modulates the response of wildlife to human recreation. Biol Conserv 227, 56\u0026ndash;64 (2018). \u003c/li\u003e\n\u003cli\u003eMarion, S. et al. Mammal responses to human recreation depend on landscape context. PLoS One 19, e0300870 (2024). \u003c/li\u003e\n\u003cli\u003eKlimaatstatistieken van de Belgische Gemeenten Maasmechelen (Nis 73107). www.meteo.be. \u003c/li\u003e\n\u003cli\u003eStatbel. https://statbel.fgov.be/nl/themas/bevolking/structuur-van-de-bevolking/bevolkingsdichtheid#figure\u003c/li\u003e\n\u003cli\u003eIndestege, S., Bollen, M., Casaer, J. \u0026amp; Beenaerts, N. Evaluating Ecological Inferences from Camera Trap Data: A Comparative Analysis of Two Sampling Designs. (2025). \u003c/li\u003e\n\u003cli\u003eFietsnet. https://www.fietsnet.be/routeplanner/default.aspx\u003c/li\u003e\n\u003cli\u003eHoge Kempen National Park. https://www.nationaalparkhogekempen.be/en\u003c/li\u003e\n\u003cli\u003eKomoot. https://www.komoot.com/\u003c/li\u003e\n\u003cli\u003eEEA. Corine Land Cover . (2020). \u003c/li\u003e\n\u003cli\u003eRowcliffe, J. M., Kays, R., Kranstauber, B., Carbone, C. \u0026amp; Jansen, P. A. Quantifying levels of animal activity using camera trap data. Methods Ecol Evol 5, 1170\u0026ndash;1179 (2014). \u003c/li\u003e\n\u003cli\u003eNouvellet, P., Rasmussen, G. S. A., Macdonald, D. W. \u0026amp; Courchamp, F. Noisy clocks and silent sunrises: measurement methods of daily activity pattern. J Zool 286, 179\u0026ndash;184 (2012). \u003c/li\u003e\n\u003cli\u003eDormann, C. F. et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36, 27\u0026ndash;46 (2013). \u003c/li\u003e\n\u003cli\u003eAkaike, H. A new look at the statistical model identification. IEEE Trans Automat Contr 19, 716\u0026ndash;723 (1974). \u003c/li\u003e\n\u003cli\u003eR Core Team. (2023). R: A Language and Environment for Statistical Computing (Version 4.3.1) [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/. \u003c/li\u003e\n\u003cli\u003eFennell, M. J. E., Ford, A. T., Martin, T. G. \u0026amp; Burton, A. C. Assessing the impacts of recreation on the spatial and temporal activity of mammals in an isolated alpine protected area. Ecol Evol 13, e10733 (2023). \u003c/li\u003e\n\u003cli\u003eLarson, C. L., Reed, S. E., Merenlender, A. M. \u0026amp; Crooks, K. R. A meta-analysis of recreation effects on vertebrate species richness and abundance. Conserv Sci Pract 1, e93 (2019). \u003c/li\u003e\n\u003cli\u003eOlejarz, A. et al. Worse sleep and increased energy expenditure yet no movement changes in sub-urban wild boar experiencing an influx of human visitors (anthropulse) during the COVID-19 pandemic. Science of The Total Environment 879, 163106 (2023). \u003c/li\u003e\n\u003cli\u003eDeCandia, A. L. et al. Urban colonization through multiple genetic lenses: The city-fox phenomenon revisited. Ecol Evol 9, 2046\u0026ndash;2060 (2019). \u003c/li\u003e\n\u003cli\u003eGil-Fernandez, M., Harcourt, R., Newsome, T., Towerton, A. \u0026amp; Carthey, A. Adaptations of the red fox (Vulpes vulpes) to urban environments in Sydney, Australia. Journal of urban ecology 6, juaa009 (2020). \u003c/li\u003e\n\u003cli\u003eNaidoo, R. \u0026amp; Burton, A. C. Relative effects of recreational activities on a temperate terrestrial wildlife assemblage. Conserv Sci Pract 2, e271 (2020). \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7926808/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7926808/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eProtected areas play a key role in conserving biodiversity, yet they increasingly face the challenge of balancing wildlife conservation with rising recreational use. Understanding how wildlife responds to human activity across space and time is therefore crucial to support coexistence. We used camera trap and visitor data to examine how recreational trail density, weekend visitation peaks, and landscape configuration influence spatiotemporal behaviour of roe deer, wild boar, and red fox in the densely visited Hoge Kempen National Park, Belgium. Despite a doubling of on-trail visitation during weekends, activity patterns of wildlife remained unchanged. Moreover, trail density negatively affected wildlife detection rates, but did not differ between weekdays and weekends, indicating that short-term increases in visitor numbers did not alter species\u0026rsquo; responses to recreational infrastructure. Instead, wildlife responses to trail density were mediated by landscape configuration: while roe deer and wild boar were more frequent in forest interiors and red fox in open areas, the negative effect of trail density was strongest in open landscapes and declined progressively with distance into forest interiors, becoming negligible deep in forested areas. These findings demonstrate that in densely visited protected areas, landscape configuration, rather than short-term visitor peaks, mediates wildlife responses to recreational infrastructure.\u003c/p\u003e","manuscriptTitle":"Wildlife responses to recreational trail density are mediated by landscape configuration rather than short-term increases in recreation intensity","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-10 07:54:27","doi":"10.21203/rs.3.rs-7926808/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"dd032c9e-19da-474d-93f2-ec883d6b78e3","owner":[],"postedDate":"December 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":57107996,"name":"Biological sciences/Ecology"},{"id":57107997,"name":"Earth and environmental sciences/Ecology"},{"id":57107998,"name":"Biological sciences/Zoology"}],"tags":[],"updatedAt":"2026-03-09T09:26:51+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-10 07:54:27","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7926808","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7926808","identity":"rs-7926808","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}