Understanding Picoides articus (Black-backed Woodpecker) occupancy in fire-suppressed forests of the Northern Blue Mountains, USA
Data files
Jan 05, 2026 version files 69.34 MB
-
bbwo_detection_data_090425.csv
36.41 KB
-
BBWO_Occupancy_Model_Script_102125.R
36 KB
-
BBWO_occupancy_probability_102925.tif
69.22 MB
-
bbwo_site_covariates.csv
30.75 KB
-
data_code_for_submission.Rproj
205 B
-
README.md
9.59 KB
Abstract
Fire is a significant disturbance agent, impacting forest structure and composition. In fire-adapted forests, fire can foster biodiversity by resetting successional processes and increasing habitat heterogeneity. However, fire suppression practices have altered historical fire regimes and left fire-adapted ecosystems largely unburned and the use of unburned forests by post-fire habitat specialists is understudied. Picoides arcticus (Black-backed Woodpecker) is a post-fire specialist that thrives in recently burned coniferous forests by exploiting heterogeneous patches of live and dead trees. Despite being a regional species of conservation concern, its occurrence and habitat use remains poorly understood in the managed forests of the northern Blue Mountains of Washington and Oregon, USA. Here, we combine passive acoustic monitoring and Bayesian occupancy modeling to understand Picoides arcticus occurrence across two National Forests during the breeding season. We examined how the prevalence of prominent tree species, attributes of forest structure, and fire history relate to occurrence. We found the odds of Picoides arcticus occupancy was 3.67 times larger for every 2.88 m2/ha increase in Pinus contorta basal area and 2.16 times higher with every 5.04 m2/ha increase in Pinus ponderosa basal area. The small number of survey stations classified as burned limited our ability to identify an effect of fire history. Our findings provide insight into how a fire-adapted species is using largely unburned forest. We also fill important knowledge gaps for a priority species of conservation concern by creating a species distribution map. Collectively, this information can inform active forest management and conservation decisions.
Dataset DOI: 10.5061/dryad.k98sf7mjw
Description of the data and file structure
Audio recordings were collected using automated recordings units randomly dispersed across the Umatilla and Wallowa-Whitman National Forests in 2022 and 2023. Audio recordings were processed using BirdNET v2.4. Detection/Non-detection data were collected for each week of recording by manually reviewing Black-backed Woodpecker detections using a top-down listening approach wherein detections were sorted in descending order by BirdNET confidence scores before manual review. Habitat covariates were generated using remotely-sensed datasets (LEMMA and LANDFIRE) and averaging or classifying values for each site using a 200m radius, circular buffer centered on each sampling site. Bayesian occupancy models and slab/spike priors were used to estimate and identify covariates relevant to Black-backed Woodpecker occurrence. See Method's section of manuscript for more details.
Files and variables
File: bbwo_site_covariates.csv
Description:
Variables
- station: location recordings were collected, first five numerical characters denote hexagon
- fire_one_year: binary variable denoting if the site burned in the last 1 year
- fire_two_five_year: binary variable denoting if the site burned in the last 2 to 5 years
- fire_six_ten_year: binary variable denoting if the site burned in the last 6 to 10 years
- med_snag_biomass: biomass of snags with a DBH of 25-50 cm, measured in kg/ha
- lrg_snag_biomass: biomass of snags between 50-75 cm DBH, measured in kg/ha
- med_downed_wood: biomass of downed wood with a DBH of 50-75 cm DBH, measured in kg/ha
- lrg_downed_wood: number of pixels within the 200m radius, circular buffer with a non-zero biomass of downed wood with a DBH > 100 cm
- grand_fir_BA: basal area of grand fir, measured in m2/ha
- can_cov: average canopy cover within the 200m radius, circular buffer, measured in percantage
- larch_BA: basal area of larch, measured in m2/ha
- lodgepole_BA: basal area of lodgepole pine, measured in m2/ha
- spruce_BA: basal area of spruce, measured in m2/ha
- ponderosa_BA: basal area of ponderosa pine, measured in m2/ha
- douglas_fir_BA: basal area of douglas fir, measured in m2/ha
- elevation: average elevation of each station, measured in meters
File: bbwo_detection_data_090425.csv
Description: Black-backed Woodpecker (BBWO) occupancy data and monitoring meta data used in occupancy analysis
Variables
- year: year station was monitored (2022 or 2023)
- station: identifier of location samples were collected, first 5 numerical characters refer to the hexagon ID and the leter denotes the station ID within a given hexagon.
- week: week of sampling within the survey period of a given station. First week of sampling denoted by 1, second week of sampling denoted by 2, etc.
- total_monitoring_time: number of hours of audio recording collected during a given sampling period
- min_date: date that the corresponding survey period began on
- bbwo_state: binary variable denoting if a BBWO was detected or undetected within a given survey period
- jday: day of the year a station's survey period began, denoted by the julian day value of min_date
File: BBWO_Occupancy_Model_Script_102125.R
Description: R script containing code used to fit Bayesian occupancy models, conduct the indicator analysis using spike and slab priors, and plot results.
File: BBWO_data_code_for_submission.Rproj
Description: .Rproj file used to increase the reproducibility of the analysis. This file should be opened before using the provided R script.
File: BBWO_occupancy_probability_102925.tif
Description: This raster depicts spatially continuous estimates of Black-backed Woodpecker (Picoides arcticus) occupancy probability across the study region. Each pixel value represents the posterior mean probability of occupancy predicted from a Bayesian occupancy model.
Format: GeoTIFF (single-band raster)
Coordinate reference system: NAD83 / Conus Albers (EPSG:5070)
Spatial resolution: 30 × 30 meters
Number of Layers: 1
Value type: Continuous
Value range: 0.196–0.999
Interpretation: Pixel values represent the estimated probability that a site is occupied by Black-backed Woodpeckers. Higher values indicate greater predicted occupancy probability.
Code/software
Occupancy Analysis
R version 4.2.0 (2022-04-22)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Monterey 12.3.1
Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.2/Resources/lib/libRlapack.dylib
locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] cowplot_1.1.3 tidyr_1.3.1 ggplot2_3.5.1 dplyr_1.1.4 reshape2_1.4.4 jagsUI_1.6.2 R2jags_0.8-9
[8] rjags_4-16 coda_0.19-4.1
loaded via a namespace (and not attached):
[1] Rcpp_1.0.12 pillar_1.9.0 compiler_4.2.0 plyr_1.8.9 tools_4.2.0 boot_1.3-30
[7] lifecycle_1.0.4 tibble_3.2.1 gtable_0.3.5 lattice_0.22-6 pkgconfig_2.0.3 rlang_1.1.3
[13] cli_3.6.2 rstudioapi_0.16.0 parallel_4.2.0 withr_3.0.0 stringr_1.5.1 generics_0.1.3
[19] vctrs_0.6.5 grid_4.2.0 tidyselect_1.2.1 glue_1.7.0 R6_2.5.1 fansi_1.0.6
[25] purrr_1.0.2 magrittr_2.0.3 scales_1.3.0 R2WinBUGS_2.1-22.1 abind_1.4-5 colorspace_2.1-0
[31] utf8_1.2.4 stringi_1.8.4 munsell_0.5.1
Maps and Species Distribution Map
R version 4.2.0 (2022-04-22)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Monterey 12.3.1
Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.2/Resources/lib/libRlapack.dylib
locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
attached base packages:
[1] tools stats graphics grDevices utils datasets methods base
other attached packages:
[1] viridis_0.6.5 viridisLite_0.4.2 purrr_1.0.2 dplyr_1.1.4 terra_1.7-46 sf_1.0-12
loaded via a namespace (and not attached):
[1] Rcpp_1.0.12 rstudioapi_0.16.0 magrittr_2.0.3 units_0.8-5 munsell_0.5.1 tidyselect_1.2.1
[7] colorspace_2.1-0 R6_2.5.1 rlang_1.1.3 fansi_1.0.6 grid_4.2.0 gtable_0.3.5
[13] KernSmooth_2.23-22 utf8_1.2.4 cli_3.6.2 e1071_1.7-14 DBI_1.2.3 class_7.3-22
[19] tibble_3.2.1 lifecycle_1.0.4 gridExtra_2.3 ggplot2_3.5.1 codetools_0.2-20 vctrs_0.6.5
[25] glue_1.7.0 proxy_0.4-27 compiler_4.2.0 pillar_1.9.0 scales_1.3.0 generics_0.1.3
[31] classInt_0.4-10 pkgconfig_2.0.3
| QGIS version | 3.32.3-Lima | QGIS code revision | 67d46100b5b |
|---|---|---|---|
| Qt version | 5.15.2 | ||
| Python version | 3.9.5 | ||
| GDAL/OGR version | 3.3.2 | ||
| PROJ version | 8.1.1 | ||
| EPSG Registry database version | v10.028 (2021-07-07) | ||
| GEOS version | 3.9.1-CAPI-1.14.2 | ||
| SQLite version | 3.35.2 | ||
| PDAL version | 2.3.0 | ||
| PostgreSQL client version | unknown | ||
| SpatiaLite version | 5.0.1 | ||
| QWT version | 6.1.6 | ||
| QScintilla2 version | 2.11.5 | ||
| OS version | macOS 12.3 |
Access information
Other publicly accessible locations of the data:
- LANDFIRE and LEMMA were used or remotely sensed habitat covariates
Data was derived from the following sources:
