Quantifying microhabitat selection of snowshoe hares using forest metrics from UAS-based LiDAR
Data files
Oct 09, 2025 version files 507.50 KB
-
pellets.csv
488.04 KB
-
README.md
3.83 KB
-
WLB-2025-01529.R2.Rmd
15.64 KB
Abstract
This dataset and code was used to evaluate the influence of forest structure (mapped using high-resolution light detection and ranging [LiDAR] data, collected on unoccupied aerial systems [UAS]) on snowshoe hare (Lepus americanus) microhabitat use within the temperate-boreal forests of northern New England. We mapped 14 forested stands (each 20 ha) using UAS-LiDAR (>300 returns/m2) and conducted fecal pellet surveys to estimate snowshoe hare microhabitat selection during the leaf-off and leaf-on seasons of 2022. We used the LiDAR point cloud data to generate traditional forest metrics (canopy closure, sapling prevalence, and shrub prevalence) that are often used in snowshoe hare habitat studies and three-dimensional metrics that we hypothesized to represent predator exposure (viewshed and lacunarity) and food availability (lacunarity). The data and code herein was used to evaluate which of these most influenced snowshoe hare microhabitat use during the leaf-off and leaf-on seasons.
Dataset DOI: 10.5061/dryad.zcrjdfns6
Description of the data and file structure
For a complete description of the experimental efforts for which the data was collected, see: 10.1002/wlb3.01529.
Files and variables
The code WLB-2025-01529.R2.Rmd does not contain external data (see below).
The empirical snowshoe hare data pellets.csv file is a data.frame that contains 1,389 observations of the pellet count response data and covariates to model snowshoe hare (Lepus americanus) microhabitat use. Below is a description of these data:
Response, spatial (study, x_scal, y_scal), temporal (season), offset, and random variables*
- count: Counts of snowshoe hare pellets from 1 m2 circular plots
- study: fixed effect factor variable with two levels ("NB", "WMNF") to account for density differences between study regions
- x_scal: scaled easting (UTM) values
- y_scal: scaled northing (UTM) values
- yr_season: 2022spring is the leaf-off season and 2022fall is the leaf-on season
- days: Number of days that elapsed since the last plot counts were conducted. This was used as an offset in models to account for differences in accumulation that would vary by differing count dates
- stand_id: Identification name of the forested stand where pellets were counted. stand_id was used as a random effect to account for potential autocorrelation of counts of plots that were nested within stands.
Traditional forest metric predictors*
- shrubs_8.5: prevalence of shrubs within 8.5-m radius plot
- shrubs_17: prevalence of shrubs within 17-m radius plot
- saplings_8.5: prevalence of saplings within 8.5-m radius plot
- saplings_17: prevalence of saplings within 17-m radius plot
- closure_8.5: canopy closure of 8.5-m radius plot
- closure_17: canopy closure of 17-m radius plot
Three-dimensional viewshed metric predictors*
- mean_count_8.5: average number of filled voxels between the center of a plot and the plot extent along observation angles within a 8.5-m radius plot
- mean_count_17: average number of filled voxels between the center of a plot and the plot extent along observation angles within a 17-m radius plot
- mean_distance_8.5: average unoccluded distance from plot center for all observation angles within a 8.5-m radius plot
- mean_distance_17: average unoccluded distance from plot center for all observation angles within a 17-m radius plot
- mean_proportion_8.5: proportion of unoccluded observation angles in a hemispherical viewshed within a 8.5-m radius plot
- mean_proportion_17: proportion of unoccluded observation angles in a hemispherical viewshed within a 17-m radius plot
Three-dimensional lacunarity metric predictors*
- L1: distribution of window size 3x3x3; 0.42 m3
- L3: distribution of window size 7x7x7; 5.36 m3
- L5: distribution of window size 11x11x11; 20.80 m3
- L7: distribution of window size 15x15x15; 52.73 m3
- L11: distribution of window size 23x23x23; 190.11 m3
- L15: distribution of window size 31x31x31; 465.48 m3
- L17: distribution of window size 35x35x35; 669.92 m3
- L27: distribution of window size 55x55x55; 2599.61 m3
- L51: distribution of window size 103x103x103; 17073.86 m3
Code/software
The file WLB-2025-01529.R2.Rmd contains code to analyze data using the pellets.csv data described above.
Access information
Data and R code used in this study are available at Dryad (Sirén et al. 2025): [10.5061/dryad.zcrjdfns6](10.5061/dryad.zcrjdfns6). We politely request to be contacted by parties interested in data reuse from the empirical hare study to discuss collaboration.
See materials and methods in 10.1002/wlb3.01529.