Datasets and Model for `Political legacies shape bear use of anthropogenic spaces`
Authors/Creators
Contributors
Data curators:
Description
This entry offers five datasets and one R model object as they were used for analyses in the manuscript titled `Political legacies shape bear use of anthropogenic spaces`and available at doi: https://doi.org/10.1101/2025.11.20.689425
Abstract
Romania hosts Europe’s largest brown bear population west of Russia, but rising human-bear encounters and related casualties raise concerns about their drivers and future population management. Using >10,000 encounter reports (2019–2024) in a hierarchical occupancy model accounting for detection processes, we show that bear presence in human-dominated spaces is linked to food conditioning, habituation, and shaped by historical management. Bears occur more often in high wildland-urban interface areas, especially those closer to communist-era elite-hunting grounds with feeding sites. These patterns reflect a long-term legacy of 1960s population management policies interacting with environmental and behavioral processes to shape recent bear behavior. We call for management addressing ecological and social sustainability by phasing out supplemental feeding, reducing habituation through a restored landscape of fear, improving public knowledge of bear ecology, and recognizing historical legacies to promote coexistence.
Related Data
historical_hunting_districts_ro.csv
geogrid_bear_presence_ro.gpkg
grid_bear_presence_years_ro.csv
grid_detection_drivers_subsample.csv
grid_occupancy_drivers_subsample.csv
Please see detailed data descriptors below
Notes
historical_hunting_districts_ro.csv
The file depicts the approximate centroids of 632 historical hunting districts in they years 1970s and 1980s. Geographic Projection: EPSG: 3844 Data fields:
X = x coordinate (longitude)
Y = y coordinate (latitude)
Type = type of hunting district (1=buffer zone, designated for species protection, 2= elite hunting districts). Note: this dataset does not include all hunting districts in Romania (total 2156).
geogrid_bear_presence_ro.gpkg
The file represents a regular grid covering the country of Romania with a cell size of 3 × 3 km grid (N = 27,197 cells). For each grid, we report weather a bear-alert call has been made within this cell. All environmental and social variable are summarized at this spatial level. . Geographic Projection: EPSG: 3844- Data fields:
Grid_id = unique grid identifier
Area_mp = total area of cell within the territory of Romania (in square meters)
Jud = NUTS3 administrative unit of Romania that the majority of the grid cell covers
Presence = information on whether a RO-ALERT bear encounter has been reported within this cell or not between 2019 and 2024. 1 = bear alert present, 0= bear alert absent
grid_bear_presence_years_ro.csv
Data frame summarizing the presence of bear encounters in a subsample of 5000 grid cells within the brown bear range in Romania. Columns represents the years 2019 to 2025 and rows labels correspond to to grid_id from geogrid_bear_presence_ro.gpkg. Values as follows 1 = bear alert present, 0= bear alert absent
grid_detection_drivers_subsample.csv
Data frame summarizing the variables used in the detectability part of the species occupancy model. Each variable is available for each year between 2019-2024. The column names follow the format `variable.Xyear`.The prefix `n` indicates that the variable has been normalized prior to use in model. E.g. npopX2019 represents the normalized population value for year 2019. While pop.X2019 is the raw population value for 2019. The dataframe include a subsample of 5000 gridcells used for modelling. Row labels correspond to grid_id from geogrid_bear_presence_ro.gpkg. Following variables have been considered.
|
Variable abbreviation |
Variable |
Expected relationship |
Mechanism |
|
tod |
Time of day (minutes) |
+ |
Bears are crepuscular, most active dawn and dusk |
|
doy |
Time of year (days) |
- |
Immediately after emergence from hibernation or in the fall when fattening up |
|
road |
Road density (road length in km) |
+ |
Measure of effort: higher likelihood of encounters due to higher human presence |
|
pop |
Population (number of inhabitants in cells) |
+ |
Measure of effort: higher likelihood of observers being present and making calls |
|
amn |
Tourism (number of touristic amenities like restaurants, bars, supermarkets) |
+ |
Measure of effort: higher likelihood of observer presence |
grid_occupancy_drivers_subsample.csv
Data frame summarizing the variables used in the occupancy part of the species occupancy model.The prefix `norm` indicates that the variable has been normalized prior to use in model. E.g. norm.p_forest represents the forest cover, while p_forest is the raw forest cover. The dataframe includes a subsample of 5000 gridcells used for modelling. Row labels correspond to grid_id from geogrid_bear_presence_ro.gpkg. Following variables have been considered.
|
Variable abbreviation |
Variable |
Expected relationship |
Mechanism |
|
bd.9.km |
+ |
Competition and predation refuge |
|
|
amn |
Anthropogenic food waste sources (number of total waste bins (public), or garbage dumps from supermarkets, restaurants and hotels. |
+ |
Food conditioning and habituation |
|
p_forest |
Forest cover (percent of grid cell) |
- |
Habitat & natural food availability |
|
pWUI |
Wildland-urban-interface (WUI) - proxy for orchards, yards and subsistence farming (percent of grid cell) |
+ |
Food availability from small farms, orchards, bee hives |
|
dist_v_r |
Distance to historical elite hunting and supplemental feeding areas |
- |
Management legacy, potentially related to habituation through continued supplemental feeding |
human_bear_ecounter_model.Rdata is a spatial occupancy model object that was used to investigate the drivers of brown bear presence in Romania. The model is fit in a Bayesian hierarchical framework for a single species and was implemented in the spOccupancy package in R. The model consists of a detectability component, and a spatial occupancy component. The data used for the modelling is described separately below.
Files
grid_bear_presence_years_ro.csv
Files
(2.7 GB)
| Name | Size | Download all |
|---|---|---|
|
md5:59aae14691c7a2191b8fabed7d52e3b5
|
8.2 MB | Download |
|
md5:967b951f2c47015871a7df250a3d84c8
|
105.1 kB | Preview Download |
|
md5:7adb640cb9d732ec10fcdd91d72ceacb
|
1.8 MB | Preview Download |
|
md5:1e768759f366972b1b1ccba3bbd05322
|
682.8 kB | Preview Download |
|
md5:ca467e88edcbda08924f8ef5e17e380e
|
15.8 kB | Preview Download |
|
md5:5e9f631e2a9f08716556c5e0eeaf1a94
|
2.7 GB | Download |
Additional details
Identifiers
Related works
- Is supplement to
- Dataset: 10.1101/2025.11.20.689425 (DOI)