Data from: Predator-prey space-use and landscape features influence movement behaviors in a large-mammal community
Data files
Sep 06, 2024 version files 609.66 MB
-
coug_dat_all_for_pub.RData
14.63 MB
-
crwOut_ALL_wCovs_for_pub.RData
26.51 MB
-
elk_dat_all_for_pub.RData
36.84 MB
-
md_dat_all_for_pub.RData
68.50 MB
-
NE_1km_grid_mask.tif
34.63 KB
-
NE_30m_grid_mask.tif
25.05 MB
-
NE_covariates_1km.RData
234 KB
-
NE_covariates_30m.RData
154.21 MB
-
OK_1km_grid_mask.tif
46.67 KB
-
OK_30m_grid_mask.tif
35.81 MB
-
OK_covariates_1km.RData
316.92 KB
-
OK_covariates_30m.RData
212.11 MB
-
README.md
10.53 KB
-
wolf_dat_all_for_pub.RData
3.96 MB
-
wtd_dat_all_for_pub.RData
31.39 MB
Abstract
Predator hunting strategies, such as stalking versus coursing behaviors, are hypothesized to influence antipredator behaviors of prey and can describe the movement behaviors of predators themselves. Predators and prey may alter their movement in relation to predator hunting modes, yet few studies have evaluated how these strategies influence movement behaviors of free-ranging animals in a multiple-predator, multiple-prey system. We fit hidden Markov models (HMM) with movement data derived from >400 GPS-collared ungulates and large predators in eastern Washington, USA. We used these models to test our hypotheses that stalking (cougars [Puma concolor]) and coursing (gray wolves [Canis lupus]) predators would exhibit different broad-scale movement behaviors consistent with their respective hunting strategies in areas that increased the likelihood of encountering or capturing ungulate prey (e.g., habitats selected by deer [Odocoileus spp.]). Similarly, we expected that broad-scale movement behaviors of prey would change in response to background levels of predation risk associated with each predator’s hunting strategy. We found that predators and ungulate prey adjusted their broad-scale movements in response to one another’s long-term patterns of habitat selection but not based on differences in predator hunting strategies. Predators changed their movement behaviors based on the type of prey, whereas ungulates generally reduced movement in areas associated with large predators, regardless of the predator’s hunting strategy. Both predator and prey movements varied in response to landscape features but not necessarily based on habitat that would facilitate specific hunting behaviors. Our results suggest that predators and prey adjust their movements at broad temporal scales in relation to long-term patterns of risk and resource distributions, potentially influencing their encounter rates with one another at finer spatiotemporal scales. Habitat features further influenced changes in movement, resulting in a complex combination of movement behaviors in multiple-predator, multiple-prey systems.
README: Data from: Predator-prey space-use and landscape features influence movement behaviors in a large-mammal community
https://doi.org/10.5061/dryad.kh1893292
Recommended citation for this dataset:
Bassing, S. B., L. Satterfield, T. R. Ganz, M. DeVivo, B. N. Kertson, T. Roussin, A. J. Wirsing, and B. Gardner. 2024. Data for: Predator-prey interactions and landscape features influence movement behaviors in a large-mammal community. Dryad, Dataset. https://doi.org/10.5061/dryad.kh1893292
Description of the data and file structure
This repository contains movement data from collected from GPS-collared elk, mule deer, white-tailed deer, cougars, and wolves in eastern Washington, USA, from 2017 - 2021. The provided data are formatted for conducting resource selection function (RSF) analyses and hidden Markov models (HMM) for movement analysis. Formatted data are provided for several stages of the analysis.
Date of data collection: 2017-01-01 to 2021-03-31
Geographic location of data collection: Northeast study area centered around Chewelah, Washington, USA (117.7193° W, 48.28302° N); Okanogan study area centered around Winthrop, Washington, USA (120.1096° W, 48.42966° N)
Information about funding sources that supported the collection of the data: Funding was provided by the Washington Department of Fish and Wildlife, the Washington State Legislature, Federal Aid in Wildlife Restoration Grant no. F16AF00910, WDFW Aquatic Lands Enhancement Accounts, Seattle City Light Wildlife Research Grant no. 2015-04, NASA FINESST grant 80NSSC19K1334, and National Geographic grant EC-51129R-19.
Data & File Overview
Use-Available Data: Contains species and season-specific "used" and "available" data included in resource selection function (RSF) analyses. GPS-collar relocation data from each collared animal were classified as "used" locations and used to define the region within a study area that was "available" to it (i.e., its seasonal home-range). A ratio of 1:20 used:available locations were randomly sampled within this spatial extent. "Used" locations were labeled (1) and "available" locations were labeled (0). Covariate data were extracted at each used and available location and included here with its used or available classification. The coordinates of each observation are excluded due to sensitivity of the information. Contact information for complete data provided below. Files are RData files that can be read into R with the load() function. Start of file names relate to each species (coug = cougar, elk = elk, md = mule deer, wtd = white-tailed deer, wolf = wolf).
Files included:
- coug_dat_all_for_pub.RData
- elk_dat_all_for_pub.RData
- md_dat_all_for_pub.RData
- wtd_dat_all_for_pub.RData
- wolf_dat_all_for_pub.RData
Data description:
- ID: Individual animal identification number
- Season: Indicates season and year when each observation was made (Summer18 = summer 2018; Winter1819 = winter 2018-2019; Summer19 = summer 2019; Winter1920 = winter 2019-2020, Summer20 = summer 2020; Winter2021 = winter 2020-2021)
- Year (Year1, Year2, Year3): Year of study each observation was made (Year1 = 2018-2019; Year2 = 2019-2020; Year3 = 2020-2021).
- Elev: Elevation (m) of observation location
- Slope: Slope (degrees) of terrain at observation location
- RoadDen: Total road length/1 km-sq at observation location
- Dist2Water: Distance (m) to nearest water
- HumanMod: Percentage of human modification to the landscape
- CanopyCover: Percentage of tree cover
- Dist2Edge: Distance (m) to nearest forested to non-forested habitat edge
- PercForestMix: Percentage of forested habitat within 250 m of observation location
- PercXGrass: Percentage of xeric grassland habitat within 250 m of observation location
- PercXShrub: Percentage of xeric shrubland habitat within 250 m of observation location
- Landcover: Numerical value representing landcover classification from Cascadia Partner Forum TerrAdapt:Cascadia tool (30m resolution)
- Landcover_type: Landcover classification label
- obs: Unique value for each observation
- Area (NE, OK): Indicates which study area (Northeast or Okanogan) each observation was associated with
- Used: Indicates whether observation was "used" (1) or "available" (0) to an individual
- w: The weight of each observation ("available" = 5000, "used" = 1)
Movement Data: Contains movement data used in hidden Markov model analyses, derived from telemetry relocations for each species. The coordinates of each observation are excluded due to sensitivity of the information. Contact information for complete data provided above. File is an RData file that can be read into R containing a list of 14 data frames, one per species, season, and study area. Start of each named list indicates the species (e.g., md = Mule Deer, wtd = White-tailed Deer); season is indicated by smr = Summer or wtr = Winter; study area is indicated by NE = Northeast or OK = Okanogan. Mule deer data were only collected in the OK, elk and white-tailed deer data were only collected in the NE study area.
Files included:
- crwOut_ALL_wCovs_for_pub.RData
Data description:
- ID: Unique animal ID and burst number
- step: Step length
- angle: Turning angle
- FullID: Unique animal ID and year of observation
- time: Date and time of telemetry relocation, time floored to nearest hour
- burst: Numerical indicator for which burst of sequential locations each observation belongs to.
- AnimalID: Unique animal ID
- speed: Movement speed calculated based on time step and step length
- obs: Observation number
- Date: Date of observation
- month: Month of observation
- Dist2Road: Distance (m) to nearest road based on Cascadia Partner Forum TerrAdapt:Cascadia tool roads layer
- PercOpen: Percentage of open habitat within 250m of observation
- SnowCover: Binary indicator of whether snow cover was present at the animal location on day of observation
- TRI: Measure of landscape variability (Terrain ruggedness index) 30m resolution, values have been centered and scaled
- [species code]_RSF: Predicted relative probability of selection by each species of interest [ELK = elk, MD = mule deer, WTD = white-tailed deer, COUG = cougar, WOLF = wolf] based on species, season, and year specific resource selection functions (RSF), values have been centered and scaled
- hour: Hour of each observation
- hour_fix: Hour of each observation
- hour3: Hour of each observation transformed to categorical variable representing the number of relocations in a day
- daytime: Binary variable indicating whether observation occurred during daylight (0) or nighttime (1)
- Sex: Sex of collared animal
- StudyArea (NE, OK): Indicates which study area (Northeast or Okanogan) each observation was associated with
- Season: Indicates season and year when each observation was made (Summer18 = summer 2018; Winter1819 = winter 2018-2019; Summer19 = summer 2019; Winter1920 = winter 2019-2020, Summer20 = summer 2020; Winter2021 = winter 2020-2021)
Covariate Data: CSV files containing covariate values extracted at every 30-sq-m or 1-sq-km pixel across the Northeast and Okanogan study areas and their corresponding TIF files, each containing a single raster that matches the spatial location and resolution of the covariate data. Raster pixels can be converted to points, representing the location of each pixel centroid, and paired with each row of the covariate data. The covariates and rasters were used to predict and map the relative probability of selection for each RSF analysis. CSV file columns and units of measurement for each variable are described below. Waterways were masked in the raster data so no predictions can be made to pixels that overlap waterbodies or rivers.
Files included:
- NE_covariates_30m.csv
- OK_covariates_30m.csv
- NE_covariates_1km.csv
- OK_covariates_1km.csv
- NE_30m_grid_mask.tif
- OK_30m_grid_mask.tif
- NE_1km_grid_mask.tif
- OK_1km_grid_mask.tif
Data description:
- ID: unique identifier for covariate data extracted at each location, corresponding with the raster's grid index number
- Elev: Elevation (m) of camera site
- Slope: Slope (degrees) of terrain at camera site
- RoadDen: Total road length/km-sq at observation location
- Dist2Water: Distance (m) to nearest water
- HumanMod: Percentage (%) of human modification to the landscape (1 km resolution)
- CanopyCover: Percentage (%) of tree cover per year of study
Dist2Edge: Distance (m) to nearest forested to non-forested habitat edge per year of study
Landcover_type: Landcover classification from Cascadia Partner Forum TerrAdapt:Cascadia tool (30m resolution) per year of study
Sharing/Access information
The coordinates of telemetry re-locations from GPS-collared animals analyzed during the current study are not publicly available due to sensitive location information but are available to qualified researchers from the Wildlife Chief Scientist of the Washington Department of Fish and Wildlife by contacting (306) 902-2515.
Query details: Data collected Jan 2017 – Mar 2021 in Game Management Units 117, 121, 203, 218, 224, 231, 233, and 239, including animal ID, date, time, and coordinates of telemetry relocations for all cougars, elk, mule deer, white-tailed deer, and wolves GPS-collared as part of the Washington Predator-Prey Project. Anonymized data are available here. Detailed data collection, processing, and formatting methods described in manuscript.
The code used to analyze the datasets during the current study are available in the GitHub repository, https://github.com/SarahBassing/Bassing_et_al_Predator-Prey_Movement and is permanently archived with Zenodo to ensure permanency and versioning, https://zenodo.org/records/13381998 (DOI: 10.5281/zenodo.13381998).
Author Information
- Name: Sarah B. Bassing
- ORCID: 0000-0001-6295-6372
- Institution: University of Washington
- Address: 107 Anderson Hall, University of Washington, Seattle WA
- Email: sarah.bassing@gmail.com
Author/Principal Investigator Information
- Name: Beth Gardner
- ORCID: 0000-0002-9624-2981
- Institution: University of Washington
- Address: 107 Anderson Hall, University of Washington, Seattle WA
- Email: bg43@uw.edu
Methods
Telemetry relocation data were collected using Global Positioning System (GPS) radio collars (make and model varied by species) affixed to >400 individual animals, including adult female elk, mule deer, and white-tailed deer, and adult female and male cougars and wolves in two study areas in eastern Washington, USA, 2017 - 2021. Location data were used to generate resoure selection functions (RSFs), which were then used to predict the relative probability of selection of each species across each study area. Location data were further used to estimate the effects of predation risk (represented by the RSFs) and landscape features associated with predator hunting mode on animal movement using hidden Markov models (HMMs). Additional capture and handling information, as well as descriptions of data cleaning and analyses, are described in detail in text (Bassing et al. in review).
Usage notes
All analyses were conducted in program R.