Habitat complexity and prey composition shape an apex predator’s habitat use across contrasting landscapes
Data files
Mar 06, 2026 version files 475.75 KB
-
Fitting_HMSC_code.R
3.18 KB
-
Prediction_maps_code.R
11.44 KB
-
README.md
3.52 KB
-
ResultsPlots_code.R
14.16 KB
-
studyDesign.csv
82.73 KB
-
Triangles_coordinates.csv
34.57 KB
-
XData.xlsx
264.36 KB
-
YData.csv
61.79 KB
Abstract
This dataset contains snow-track observations of Eurasian lynx (Lynx lynx) and its main prey species collected across central and southern Finland between 2016 and 2020 within the Wildlife Triangle Scheme. The data include spatial records of species presence along standardized transects, combined with environmental and landscape variables describing forest structure, terrain complexity, and land cover. In addition, the dataset integrates information on prey species, including mountain hare (Lepus timidus), roe deer (Capreolus capreolus), and white-tailed deer (Odocoileus virginianus), enabling the analysis of predator–prey spatial associations.
The dataset was used to support joint species distribution modelling (HMSC) to evaluate relationships between lynx habitat use, environmental conditions, and prey distribution at a large spatial scale. Analyses based on these data indicate that lynx habitat use is consistently associated with structurally complex forests and heterogeneous terrain across regions. However, regional differences in prey availability are reflected in contrasting spatial associations between lynx and prey species.
These results indicate that lynx adjust their space use in response to prey availability while maintaining a consistent preference for structurally complex habitats typical of stalk-and-ambush predators. The analysis demonstrates how variation in environmental conditions and prey community composition is reflected in spatial patterns of habitat use, and how these patterns may contribute to population-level resilience.
Dataset Overview
This dataset contains snow-track observations of Eurasian lynx (Lynx lynx) and its main prey species collected across central and southern Finland between 2016 and 2020 within the Finnish Wildlife Triangle Scheme. The dataset integrates species data, environmental covariates, and spatial information to support joint species distribution modelling (HMSC) and analyses of predator–prey spatial associations.
FILES INCLUDED
1. YData.csv Snow-track count data per species, triangle, and year.
All values represent the number of snow-track crossings recorded during standardized winter surveys.
All counts are integer values. Zero indicates no tracks detected.
2. XData.xlsx Environmental and spatial covariates associated with each wildlife triangle and year.
Variables:
triangle_id: Unique numeric identifier of wildlife triangle.
year: Sampling year (2016–2020).
log_effort: Log-transformed sampling effort (continuous).
snow_depth: Mean winter snow depth (standardized; original unit: cm).
residential_distance: Distance to nearest residential area (standardized; original unit: meters).
roads_distance: Distance to nearest road (standardized; original unit: meters).
streams_distance: Distance to nearest stream (standardized; original unit: meters).
peatbogs: Proportion of peatbog habitat within triangle buffer (standardized; original unit: proportion or %).
streams_length: Total length of streams within triangle buffer (standardized; original unit: meters or km).
dead_wood_index: Index of dead wood availability (standardized; unitless index).
terrain_ruggedness: Terrain ruggedness index (standardized; unitless index).
forest_cover: Proportion of forest cover within triangle buffer (standardized; original unit: proportion or %).
triangle_x: X-coordinate of triangle centroid (projected coordinate system; meters).
triangle_y: Y-coordinate of triangle centroid (projected coordinate system; meters).
region: Categorical variable indicating geographic region.
Unless otherwise stated, environmental covariates were standardized (mean = 0, SD = 1) prior to modelling.
3. Triangles_coordinates.csv Spatial coordinates of wildlife triangle centroids.
Variables:
x: X-coordinate of triangle centroid (projected coordinate system; meters).
y: Y-coordinate of triangle centroid (projected coordinate system; meters).
Note: Coordinates were randomly rotated to protect sensitive species locations.
4. studyDesign.csv Study design matrix describing the hierarchical sampling structure used in HMSC models.
Variables:
Triangles: Unique identifier of wildlife triangle.
Year: Sampling year (2016–2020).
Region: Geographic region (categorical).
CODE FILES
Fitting_HMSC_code.R R script used for fitting joint species distribution models using the HMSC framework.
Analyses were performed on the Puhti supercomputer provided by CSC – IT Center for Science, Finland.
ResultsPlots_code.R R script used to produce figures and result visualizations.
Prediction_maps_code.R R script used to generate spatial prediction maps presented in the supplementary material.
DATA LINKAGE
Data files are linked via triangle identifiers and year variables. The variables ‘triangle_id’ (XData.xlsx), ‘Triangles’ (studyDesign.csv), and the spatial coordinates correspond to the same wildlife triangle units.