Data from: Individual movements drive community-level space and resource use across Yellowstone ungulates
Data files
Apr 29, 2026 version files 130.54 MB
-
community_union_hypervolume_data_seasyr.rds
618 B
-
Hypervolumes_Step1_CalculateHypervolumes.Rmd
9.90 KB
-
Hypervolumes_Step2_UnionHypervolumes.Rmd
5.96 KB
-
Hypervolumes_Step3_CalculateMetrics.Rmd
9.71 KB
-
Hypervolumes_Step4_MetricBayesianModels.Rmd
55.41 KB
-
Hypervolumes_Step5_OStatistics.Rmd
7.72 KB
-
indiv_hypervolume_data.rds
33.97 KB
-
population_union_hypervolume_data_seasyr.rds
1.80 KB
-
README.md
2.93 KB
-
YNP_Ungulate_Hypervolume_Input_Data.csv
130.41 MB
Abstract
This repository contains data and code to evaluate how individual movements influence resource use across a community of sympatric, migratory ungulates. We use remotely-sensed environmental data and GPS locations of 5 ungulates (bison, bighorn sheep, elk, pronghorn, and mule deer) in Yellowstone National Park, USA (2017-2021). We generated hypervolumes (following methods of Bolker et al. 2018) of derived variables from GPS locations and remotely-sensed environmental data of Yellowstone ungulates. The repository datasets are the derived hypervolumes used in the associated code. We do not include any raw GPS locations. The GPS data & derived metrics in this repository are owned by NPS, to use contact Chris Geremia chris_geremia@nps.gov.
https://doi.org/10.5061/dryad.d2547d8cm
Description of the data and file structure
This data corresponds to the manuscript "Individual movements drive community-level space and resource use across Yellowstone ungulates" by Molly R. Caldwell, Chris Geremia, Daniel R. Stahler, Daniel R. MacNulty, Douglas W. Smith, and Jerod A. Merkle. The dataset contains the following variables with metrics derived from GPS collar data from 5 ungulate species (2017-2021, Yellowstone National Park, USA). Note that the data is randomly sampled to 300 locations or 600 locations (for winter) per individual per season and year. All metrics have been centered and scaled, except for latitude and longitude locations that were only centered. Note that the latitude and longitude no longer have any geographic relevance and do not represent real locations of animals. Full methodology descriptions for each variable are in the associated manuscript.
Files and variables
File: YNP_Ungulate_Hypervolume_Input_Data.csv
Description: Data derived from GPS collar locations of Yellowstone ungulates. Used to build individual hypervolumes following the methodology of the associated manuscript.
Variables
All variables are scaled (no relevant units)
- animal_id: unique identifyer for each collared animal
- season: season of data location based on phenology of plants and snow conditions
- yr: year of data location
- speed: speed between consecutive locations (euclidean distance / time between locations)
- displ_24hr: greatest euclidean distance between any 2 locations during prior 24 hours
- long: transformed longitude of location, has no geographic relevance
- lat: transformed latitude of location, has no geographic relevance
- elev: elevation at location
- slope: slope at location
- perc_treecov: percent tree cover at location
- NDVIVals: scaled NDVI values at location
- IRGVals: scaled IRG values at location
- forbgrass_biomass: sum of annual and perennial forbs and grasses biomass at location
- snowdepth: snow depth at location
- species: species of animal
Code/software
Code that uses the data to build hypervolume and extract environmental niche metrics following the associated manuscript methodology are the following .Rmd files (all code is in the form of R Markdown files, and requires the program R):
-
Hypervolumes_Step1_CalculateHypervolumes.Rmd
-
Hypervolumes_Step2_UnionHypervolumes.Rmd
-
Hypervolumes_Step3_CalculateMetrics.Rmd
-
Hypervolumes_Step4_MetricBayesianModels.Rmd
-
Hypervolumes_Step5_OStatistics.Rmd
The following files are generated hypervolumes used within the codes:
- community_union_hypervolume_data_seasyr.rds
- indiv_hypervolume_data.rds
- population_union_hypervolume_data_seasyr.rds
The hypervolume input dataset [YNP_Ungulate_Hypervolume_Input_Data.csv] contains metrics derived from GPS collar data from 5 ungulate species (2016-2021, Yellowstone National Park, USA). This dataset is randomly sampled to 300 locations or 600 locations (for winter) per animal per season and year. All metrics have been centered and scaled, except for latitude and longitude locations that were only centered. Note that the latitude and longitude no longer have any geographic relevance and do not represent real locations of animals.