Data from: Collaborative camera-trap data set for evaluating the status of a rangewide carnivore in Latin America
Data files
Jun 10, 2026 version files 517.14 MB
-
0a_prep_data_IDS.R
6 KB
-
0b_prep_data_collaborator.R
28.52 KB
-
0c_prep_data_misc.R
16.70 KB
-
1_combine_data.R
8.34 KB
-
2_summaries_plots.R
15.36 KB
-
puma_rangewide_effort_detection_data.csv
74.92 MB
-
puma_rangewide_effort_detection_data.dbf
160.23 MB
-
puma_rangewide_effort_detection_data.geojson
239.17 MB
-
puma_rangewide_effort_detection_data.prj
145 B
-
puma_rangewide_effort_detection_data.Rdata
638.50 KB
-
puma_rangewide_effort_detection_data.shp
32.75 MB
-
puma_rangewide_effort_detection_data.shx
9.36 MB
-
README.md
3.18 KB
Abstract
This dataset comprises a comprehensive collection of puma (Puma concolor) detections obtained from camera-trap surveys across Latin America, encompassing over 1,169,563 camera trap nights across 19 countries, spanning from northern Sonora, Mexico, to southern Patagonia, Chile. The compiled dataset includes 12,312 unique puma detections, together with associated sampling metadata and geographic coordinates.
Dataset DOI: 10.5061/dryad.k6djh9wj8
Description of the data and file structure
This shareable dataset is formatted so that each row corresponds to a camera station location (identified by its latitude and longitude) and a specific date on which sampling effort occurred. Each row therefore represents a single 24-hour sampling period at a camera station. Camera station–date combinations that do not appear in the dataset indicate that no sampling effort occurred on that date at that station.
The dataset includes columns identifying the country and whether a puma detection occurred during that 24-hour period (detection: 1 = detection, 0 = no detection). The data are provided at a fine spatial and temporal resolution, with a simplified set of variables focused on camera station location, date, country, and detection status.
Files and variables
Files:
- "puma_rangewide_effort_detection_data.Rdata": this is a data frame R object
- "puma_rangewide_effort_detection_data.csv": csv format
- "puma_rangewide_effort_detection_data.shp": shapefile (spatial object that can be read directly into a GIS program)
- "puma_rangewide_effort_detection_data.geojson": another spatial object that can be read directly into a GID program
Scripts
- "0a_prep_data_IDS.R" details the cleaning procedures applied to data exported from PantheraIDS and the verification of camera station coordinates.
- "0b_prep_data_collaborator.R" documents the cleaning, restructuring, and manual correction steps applied to contributed datasets that followed the provided template format.
- "0c_prep_data_misc.R" describes the individualized processing required for contributed datasets that did not follow any standard format.
- "1_combine_data.R" describes the process for combining the sub-datasets generated in scripts 0a, 0b, and 0c, and the final checks for valid XY coordinates and country assignment.
- "2_summaries_plots.R" generates all summaries and plots described above.
Dataset Columns Description:
1. country: The country in which the camera station is located.
2. longitude: The longitude coordinate* of the camera station location. The dataset is structured so that each row corresponds to this location and a specific date of sampling effort.
3. latitude: The latitude coordinate* of the camera station location. Together with longitude, this identifies the camera station associated with each date of sampling effort.
4. date: The date corresponding to a 24-hour sampling period during which the camera trap was active.
5. effort: Effort indictor showing effort occurred at the camera station during 24-hour period (all rows are 1 since only dates of actively deployed camera stations are included).
6. detection: Detection indicator for puma during that 24-hour period at the camera station (1 = detection recorded, 0 = no detection).
*Coordinates are provided in decimal degrees using the WGS84 geographic coordinate system.
Camera-trap detection data for pumas (Puma concolor) across Latin America were compiled from a network of over 200 contributors recruited through systematic literature review and snowball sampling. Potential contributors were identified via Web of Science, Google Scholar, and Scopus using keywords including "puma," "Puma concolor," and "camera trap," and contacted by email in September 2020.
Contributors submitted data in a standardized occupancy modeling format: each row represents a camera-trap station with latitude, longitude, and country, and each column represents a survey day, coded as 1 (puma detected), 0 (no detection), or NA (camera inactive). Data were also obtained from Panthera's internal repository (PantheraIDS). All data underwent coordinate validation and correction, date format standardization, and removal of records with missing or invalid locations. Two datasets unavailable for public sharing were excluded.
The final dataset is in long format, with one row per camera-station and date with confirmed sampling effort, and a binary puma detection variable (1/0). Files are provided in .csv, .Rdata, .shp and JSON formats.