Data and script from: Recreational activities and human hunting strongly influence red deer space use to trails
Data files
Jun 01, 2026 version files 60.36 MB
-
analysis_and_figs.R
17.48 KB
-
data_for_model.rds
60.34 MB
-
movement_kernel.rds
971 B
-
README.md
3.46 KB
Abstract
Assessing how animals adjust their movements to two widespread human activities - hunting and recreation- can provide critical insights into how these activities reshape red deer spatial behavior. In this study, we used GPS-telemetry data from red deer in the Bavarian Forest National Park (Germany) to analyze their habitat selection and displacement in relation to hiking trails with varying visitor numbers. Using integrated step selection analyses, we looked at different human disturbances for four scenarios (non-hunting season, hunting season, day, and night). Despite research on the combined effects of hunting and recreation on ungulates, how responses vary between day and night and across areas with varying levels of human disturbance remains poorly understood. By incorporating visitor numbers, this study provides a novel perspective on how red deer adjust their behavior to recreational intensity. We hypothesized that red deer individuals would avoid trails, avoid crossing trails, increase their displacement, and select habitats with dense understory vegetation during periods of greater human disturbance (i.e., during the day and hunting season). We expected this effect to be stronger within areas with high visitor numbers during the day and within the hunting zone during the hunting season. We found that red deer avoided trails during the day, with a stronger avoidance of trails with high visitor numbers, and favored denser vegetation for concealment during daytime. Trails were increasingly avoided with distance from trails; areas further away from trails were selected when visitor numbers were high, and step length was minimally higher close to trails. The strong aversion to crossing trails suggest that trails act as barriers, influencing displacement and habitat use. Our results demonstrate that trails associated with varying levels of human presence are important features shaping red deer displacement and habitat selection. These behavioural modifications might cause cascading ecological effects, i.e., affect vegetation and nutrient dynamics, underlining the importance of visitor management such as spatial zoning, temporal access limitations, and modifications of trail networks to mitigate impacts.
Dataset DOI: 10.5061/dryad.b8gtht7tf
Data
Dataset: data_for_model.rds
Format: Random steps (amt package), data frame
Rows: 2,041,891
Columns: 18.
| Variable | Type | Description |
|---|---|---|
case_ |
logical | Whether the step is the observed (TRUE) or a random alternative (FALSE) |
step_id_ |
numeric | Step identifier, links observed and random steps within a stratum |
id |
numeric | Individual animal ID |
t1_ |
POSIXct | Timestamp at the start of the step (UTC) |
nohunting |
numeric | Whether the step occurred in a no-hunting zone (1 = yes) |
trail.cross |
logical | Whether the step crosses a trail |
overstory_start |
numeric | Overstory canopy cover at the start location (proportion) |
overstory_end |
numeric | Overstory canopy cover at the end location (proportion) |
understory_start |
numeric | Understory cover at the start location (proportion) |
understory_end |
numeric | Understory cover at the end location (proportion) |
dem |
numeric | Elevation at the end location (metres) |
TRI |
numeric | Terrain Ruggedness Index at the end location |
dist_start |
numeric | Distance to trail at the start location (metres) |
dist_end |
numeric | Distance to trail at the end location (metres) |
n.visitors_start |
numeric | Visitor index at the start location |
n.visitors_end |
numeric | Visitor index at the end location |
sl_ |
numeric | Step length (metres) |
tod_end_ |
factor | Time of day at step end (day / night) |
Dataset: movement_kernel.rds
This data set contains the tentative step length distributions needed for creating
Figure 1. Code included in Rscript: "Analysis_and_figs.R"
Script
All statistical analyses were conducted in R (version 4.5.3).
Rscript: analysis_and_figs.R
This R script fits step selection functions and creates figures for the manuscript.
| Loaded packages | Version |
|---|---|
glmmTMB |
1.1.14 |
tidyverse |
2.0.0 |
broom.mixed |
0.2.9.7 |
amt |
0.3.1.0 |
legendry |
0.3.0 |
| kableExtra | 1.4.0 |
Access information
All data is extracted from red deer GPS-position data. Raw data is not publicly available but can be accessed via the Eurodeer database (https://euromammals.org/eurodeer/) or upon request by contacting the corresponding author: suzanne.van-beeck@uni-ulm.de
