Dataset DOI: 10.5061/dryad.5dv41nshz

Dataset overview

This dataset contains R scripts and processed data supporting the manuscript Diversity regulation in local communities: Colonization and extinction dynamics in North American birds. The files are derived from the North American Breeding Bird Survey, 2022 release (https://doi.org/10.5066/P97WAZE5, Ziolkowski et al. 2022), and are organized as ten R scripts and six .RData files. The data files include temporally and spatially subsampled Breeding Bird Survey records, route-level environmental and metadata variables, calculated community variables for each route and year, variables averaged across years, and datasets used for supplementary analyses based on three-year bins and core species. The R scripts document the full analytical workflow, including data loading and subsampling, variable calculation, correlations and generalized linear models, mapping, random forest analyses, structural equation models, and supplementary analyses.

This archive contains R scripts and processed data supporting the manuscript Diversity regulation in local communities: Colonization and extinction dynamics in North American birds.

Uploaded: 30/11/2025
Authors: Eliska Bohdalkova, David Storch

Analyses are based on the North American Breeding Bird Survey (BBS) 2022 release (Ziolkowski et al. 2022; https://doi.org/10.5066/P97WAZE5).

The submission includes:

1. 10 R scripts (*.R)
2. 6 R binary data files (*.RData)

File naming conventions

Script files use numeric prefixes to indicate execution order:

1. 1_...R to 6_...R: main workflow
2. 7_...R and 8_...R: supplementary workflow using 3-year bins
3. 9_...R and 10_...R: supplementary workflow using core species only

Data file naming:

1. BBS_*: processed BBS-derived data objects used in analyses
2. routes_environment.RData: route-level metadata and environmental covariates

Software requirements

Required software:

1. R (tested with R 4.5.2 (2025-10-31))

R packages used across scripts:

1. dplyr
2. tidyverse
3. tidyr
4. tibble
5. ggplot2
6. corrplot
7. sf
8. rnaturalearth
9. viridis
10. randomForest
11. lavaan
12. raster

Optional installation command:

install.packages(c(
  "dplyr", "tidyverse", "tidyr", "tibble", "ggplot2", "corrplot",
  "sf", "rnaturalearth", "viridis", "randomForest", "lavaan", "raster"
))

External data dependency

Script 1_loading_subsampling_data.R requires the original BBS 2022 release files downloaded separately from:

1. https://doi.org/10.5066/P97WAZE5

The BBS source directory is expected as 2022BBSrelease/ under the project working directory.

Execution order and workflow

Main analysis:

1. 1_loading_subsampling_data.R -> creates BBS_subsampled.RData
2. 2_variables_calculation.R -> creates BBS_variables_allyears.RData and BBS_variables_averagedyears.RData
3. 3_correlations.R -> figures/correlations/GLMs (no file written)
4. 4_maps.R -> map figures (no file written)
5. 5_random_forest.R -> random forest analyses (no file written)
6. 6_SEMs.R -> SEM analyses (no file written)

Supplementary analysis A (3-year bins):

1. 7_3yearsbins_variables.R -> creates BBS_for_3yearsbins_supplement_analysis.RData
2. 8_3yearsbins_analyses.R -> analyses and figures (no file written)

Supplementary analysis B (core species):

1. 9_corespecies_variables.R -> creates BBS_cores_final_averagedyears.RData
2. 10_corespecies_analyses.R -> analyses and figures (no file written)

File-by-file description

Scripts (*.R)

1. 1_loading_subsampling_data.R
   Loads raw BBS files, filters to quality-approved runs (RunType == 1), passerines only, applies temporal and spatial subsampling, and saves filtered survey table.
2. 2_variables_calculation.R
   Computes route-year biodiversity and turnover metrics, population variability, route-level means, environmental joins, species-pool metrics; saves main analysis tables.
3. 3_correlations.R
   Produces scatterplots, correlation tests/matrices, and GLMs for main and supplementary datasets.
4. 4_maps.R
   Produces geographic maps of extinction probability and species richness.
5. 5_random_forest.R
   Fits random forest models for extinction probability and summarizes variable importance.
6. 6_SEMs.R
   Fits structural equation models (SEM) for extinction probability variants.
7. 7_3yearsbins_variables.R
   Aggregates data into 3-year bins, computes analogous metrics for supplementary analysis, and saves summary table.
8. 8_3yearsbins_analyses.R
   Runs analyses/visualizations for the 3-year bin supplementary dataset.
9. 9_corespecies_variables.R
   Classifies route-level species occupancy into core/transient groups, retains core species, recomputes analysis metrics, and saves summary table.
10. 10_corespecies_analyses.R
    Runs analyses/visualizations for the core-species supplementary dataset.

Data files (*.RData)

1. BBS_subsampled.RData
   Object: survey3 (data.frame, 604,788 rows x 9 columns)
   Description: temporally and spatially filtered survey records used as the analysis starting point.
2. routes_environment.RData
   Object: routes_environment (data.frame, 5,784 rows x 9 columns)
   Description: route metadata and environmental covariates (ndvi, seasonality) keyed by SpecificID.
3. BBS_variables_allyears.RData
   Object: bbs (data.frame, 16,150 rows x 19 columns)
   Description: route-year table (1995-2019) with extinction/colonization and community structure metrics.
4. BBS_variables_averagedyears.RData
   Object: bbs_summed (tbl_df, 646 rows x 26 columns)
   Description: route-level means of main-analysis metrics with environmental and species-pool variables.
5. BBS_for_3yearsbins_supplement_analysis.RData
   Object: bbs_3ybins_summed (data.frame, 646 rows x 20 columns)
   Description: route-level summary metrics for supplementary analysis using 3-year temporal bins.
6. BBS_cores_final_averagedyears.RData
   Object: bbs_cores_summed (data.frame, 646 rows x 26 columns)
   Description: route-level summary metrics for supplementary analysis restricted to core species.

Data dictionary (variables and interpretation)

Variables below appear across one or more tabular objects in this archive.

Identifier and key variables

1. SpecificID: route identifier as CountryNum_StateNum_Route (character).
2. route_year_code: composite key SpecificID_year (character).
3. RouteDataID: BBS route-run identifier from source data (integer).
4. CountryNum: BBS country code (integer).
5. StateNum: BBS state/province code (integer).
6. Route: BBS route number (integer).
7. RouteName: route name/label (character).
8. bcr: Bird Conservation Region code (integer).

Time variable

1. year: calendar year of survey/observation (integer; 1995-2019 in filtered tables).

Spatial/environmental variables

1. lat: latitude in decimal degrees (WGS84; north positive).
2. lon: longitude in decimal degrees (WGS84; west negative).
3. ndvi: normalized difference vegetation index (unitless index).
4. seasonality: NDVI seasonality metric (unitless, provided in routes_environment).

Species identity and abundance variables

1. aou: American Ornithological Union species code (integer).
2. SpeciesTotal: total count of individuals of species aou on a route in a given year (count).
3. tot_abun: total abundance across species for route-year (count).
4. mean_abun: mean abundance per species for route-year (count per species).
5. mean_logabun: mean of natural-log transformed species abundances for route-year (ln(count), unitless on log scale).
6. sp_rich: species richness (number of species detected; count).
7. sp_rich_cummul: cumulative species richness across a 3-year bin (count of unique species).
8. spool: species pool size in the corresponding BCR (count of unique species).
9. SrelativeToPool: ratio of richness to species pool (sp_rich / spool), unitless proportion.

Colonization/extinction variables

1. ext_count: number of species present at time t and absent at t+1 (count).
2. ext_prob: extinction proportion, ext_count / species richness (unitless, 0-1).
3. ext_count_nosingl: extinction count excluding singleton species (count).
4. ext_prob_nosingl: extinction proportion excluding singleton species (unitless, 0-1).
5. col_count: number of species present at time t and absent at t-1 (count).
6. col_prob: colonization proportion, col_count / species richness (unitless, 0-1).
7. col_count_nosingl: colonization count excluding singleton species (count).
8. col_prob_nosingl: colonization proportion excluding singleton species (unitless, 0-1).

Community structure and dynamics variables

1. dominance_rel: relative dominance, max species abundance / total abundance (unitless, 0-1).
2. rarity_singl: proportion of singleton species (species with abundance = 1), unitless (0-1).
3. rarity_5: proportion of species with abundance below 5% of route total abundance, unitless (0-1).
4. variability: mean absolute log population growth rate across species, based on |ln(N_t / N_t-1)| (unitless).
5. residuals_S_NDVI: residuals from route-level model log(species richness) ~ log(ndvi) (unitless residual on log scale).

Route filtering support variable

1. pair_years: number of consecutive year pairs available per route during filtering (count; higher values indicate more complete temporal coverage).

Missing values and notes

1. ndvi, seasonality, and residuals_S_NDVI contain missing values for a small subset of routes (8 routes in route-level analysis tables).
2. Some route-year metrics in BBS_variables_allyears.RData are NA where required consecutive years are unavailable.
3. routes_environment.RData includes more routes (5,784) than the filtered analysis set (646 routes).

Minimal example to inspect included objects

load("BBS_variables_averagedyears.RData")
str(bbs_summed)
summary(bbs_summed$ext_prob)

Access information

Data derived from:

1. North American Breeding Bird Survey, 2022 release
   https://doi.org/10.5066/P97WAZE5

References

R Core Team (2025). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/.

Ziolkowski Jr., D.J., Lutmerding, M., Aponte, V.I., and Hudson, M-A.R., 2022, North American Breeding Bird Survey Dataset 1966-2021: U.S. Geological Survey data release, https://doi.org/10.5066/P97WAZE5.