Data from: Canopy cover has divergent effects on movement of closely related bumble bees in managed conifer forest landscapes
Data files
Jun 26, 2026 version files 1.01 MB
-
colony_abundance.csv
9.74 KB
-
fit_stand_age_and_roads.R
4.88 KB
-
fit_stand_age_and_roads.sh
343 B
-
fit_stand_age.R
4.20 KB
-
fit_stand_age.sh
182 B
-
floral_cover_midnorth.txt
807 B
-
floral_cover_north.txt
719 B
-
floral_cover_south.txt
732 B
-
GEE_script.txt
9.15 KB
-
genotyped_captures_Bcal_midnorth.csv
4.74 KB
-
genotyped_captures_Bcal_north.csv
5.47 KB
-
genotyped_captures_Bcal_south.csv
3.22 KB
-
genotyped_captures_Bvos_midnorth.csv
16.09 KB
-
genotyped_captures_Bvos_north.csv
20.42 KB
-
genotyped_captures_Bvos_south.csv
8.72 KB
-
midnorth_agemap_90m.tif
14.91 KB
-
midnorth_canopy_cover_map_90m.tif
21.43 KB
-
midnorth_early_seral_binary_map.tif
14.91 KB
-
midnorth_latlongs.csv
1.36 KB
-
midnorth_roads_binary.tif
18.94 KB
-
midnorth_roads_proportion_90m.tif
6.64 KB
-
midnorth_trap_coords.txt
1.37 KB
-
midnorth.bomcal.90m.RData
39.66 KB
-
midnorth.bomvos.90m.RData
40.07 KB
-
midnorth.roads.R
3.17 KB
-
midnorth.singletons.R
1.82 KB
-
midnorth.stand_age.R
2.54 KB
-
north_agemap_90m.tif
29.63 KB
-
north_canopy_cover_map_90m.tif
41.61 KB
-
north_early_seral_binary_map.tif
39.18 KB
-
north_latlongs.csv
1.21 KB
-
north_roads_binary.tif
45.35 KB
-
north_roads_proportion_90m.tif
14.24 KB
-
north_trap_coords.txt
1.30 KB
-
north.bomcal.90m.RData
71.93 KB
-
north.bomvos.90m.RData
72.03 KB
-
north.stand_age.R
2.52 KB
-
ocr_age_canopycover.csv
8.52 KB
-
package_data.R
2.25 KB
-
README.md
8.67 KB
-
south_agemap_90m.tif
35.51 KB
-
south_canopy_cover_map_90m.tif
50.18 KB
-
south_early_seral_binary_map.tif
39.60 KB
-
south_latlongs.csv
1.22 KB
-
south_roads_binary.tif
55.31 KB
-
south_roads_proportion_90m.tif
17.34 KB
-
south_trap_coords.txt
1.30 KB
-
south.bomcal.90m.RData
87.99 KB
-
south.bomvos.90m.RData
88.16 KB
-
south.stand_age.R
2.49 KB
-
structural_connectivity.R
18.72 KB
-
structural_connectivity.RData
5.74 KB
-
tmp2.R
2.89 KB
-
total_captures_Bcal_midnorth.csv
798 B
-
total_captures_Bcal_north.csv
837 B
-
total_captures_Bcal_south.csv
721 B
-
total_captures_Bvos_midnorth.csv
857 B
-
total_captures_Bvos_north.csv
780 B
-
total_captures_Bvos_south.csv
760 B
Abstract
Context: Functional connectivity-the degree to which landscape features facilitate or impede movement among resource patches— affects animal survival, health, and ability to cope with environmental changes. This is particularly important in heterogeneous landscapes prone to rapid change, such as intensively managed forests.
Objectives: We aimed to quantify the effects of forest canopy cover and roads on foraging movement of two closely related bumble bee species (Bombus vosnesenskii and Bombus caliginosus) in conifer forest landscapes intensively managed for timber. We also assessed whether early seral forest configuration predicted site-level colony abundance for each species.
Methods: We sampled bumble bees and surveyed floral resources in 75 sites in three study landscapes in the central Oregon Coast Range. We modeled the effects of forest canopy cover and secondary road cover on foraging range using a genetic capture-recapture approach and modeled colony abundance as a function of site-level structural connectivity of early seral forest.
Results: We found evidence that canopy cover impeded foraging movement of B. vosnesenskii but not of B. caliginosus. Roads had neutral or inhibitory effects on B. vosnesenskii movement depending on the landscape but only inhibitory effects on B. caliginosus movement. Colony abundance was not related to site-level connectivity for either species.
Conclusions: Our study demonstrates that landscape features can have divergent effects on space use of even closely related taxa and that site-level response variables may not always reflect functional connectivity. Management of early seral forest habitats should consider not only patch quality but also configuration and species-specific permeability of matrix features.
https://doi.org/10.5061/dryad.zw3r228g6
This repository contains data and scripts associated with the manuscript: "Canopy cover has divergent effects on movement of closely related bumble bees in managed conifer forest landscapes" prepared to submit to Landscape Ecology. The model demonstrated here can be used to assess effects of landscape features on movement of central place-foraging eusocial insects via a genetic capture-recapture approach.
In the example dataset, workers of two bumble bee species (Bombus vosnesenskii and Bombus caliginosus, abbreviated in data as "Bvos" or "bomvos" and "Bcal" or "bomcal") were sampled and genotyped at 75 sites (1-25) in 3 replicate intensively managed forest landscapes ("south", "midnorth", and "north") in western Oregon to examine effects of forest cover on space use and site-level colony abundance.
Description of the data and file structure
raw data file descriptions:
- colony_abundance.csv includes standardized colony abundance estimates (Nns) along with intermediary calculations for each species x site
- Site = site name
- Species = bumble bee species epithet
- N = total number of captured individuals (all castes)
- Ni = total number of captured workers
- Ng = number of successfully genotyped individuals
- Nnr = number of unique colonies detected
- Nns = standardized colony abundance estimate
- Landscape = study landscape
- flower_density = flower density at the time of sampling
- genotyped_captures_species_landscapename.csv (6 files: genotyped_captures_Bcal_midnorth.csv, genotyped_captures_Bcal_north.csv, genotyped_captures_Bcal_south.csv, genotyped_captures_Bvos_midnorth.csv, genotyped_captures_Bvos_north.csv, genotyped_captures_Bvos_south.csv) is a matrix of workers of a single species genotyped at each site in a landscape. Rows represent colonies and columns (named) represent sites. Abbreviated species name is specified by "species"; landscape name is specified by "landscapename".
- Site = site name
- Species = bumble bee species epithet
- n_F = number of females
- n_M = number of males
- total = total number of individuals
- total_captures_species_landscapename.csv (6 files: total_captures_Bcal_midnorth.csv, total_captures_Bcal_north.csv, total_captures_Bcal_south.csv, total_captures_Bvos_midnorth.csv, total_captures_Bvos_north.csv, total_captures_Bvos_south.csv) is a table containing counts of individual bees of a single species captured at each site in a landscape. Counts are given for each sex separately (n_M/n_F) and together (total).
- n_M = number of males
- n_F = number of females
- total = total number of individuals of both sexes (n_M + n_F)
- floral_cover_landscapename.txt (3 files: floral_cover_midnorth.txt, floral_cover_north.txt, floral_cover_south.txt) is a vector of floral density values for each site in a landscape.
- landscapename_latlongs.csv (3 files: midnorth_latlongs.csv, north_latlongs.csv, south_latlongs.csv) is a table of lat/long coordinates (WSG84) for each site in a landscape.
- Site_Name = site name (Landscape + Trap_num)
- Long = longitude
- Lat = latitude
- Elev_m = meters above sea level
- Landscape = study landscape
- Trap_num = trap ID number (ordinal categorical variable); numbers range from 1-25 and repeat in each landscape.
- landscapename_trap_coords.txt (3 files: midnorth_trap_coords.txt, north_trap_coords.txt, south_trap_coords.txt) is a table of coordinates (Albers Equal Area Conic Projection) for each site in a landscape.
- ocr_age_canopycover.csv is a table of stand age and average canopy cover collected in 60 intensively managed Douglas-fir stands in the central part of the Oregon Coast Range during 3 sampling rounds in the late spring-summer of 2018 and 2019 (data from Zitomer et al. 2023, Dryad repository).
- Stand = site name
- Year = data collection year
- Round = sampling round for each year
- Age = stand age (years since harvest at the time of data collection)
- AvgCanopy = average canopy cover estimated from 4 points within sample plot using a convex densiometer
- landscapename_agemap_90m.tif (3 files: midnorth_agemap_90m.tif, north_agemap_90m.tif, south_agemap_90m.tif) is a raster (90 m resolution) of forest age---years since stand-replacing disturbance in summer 2021---estimated using the disturbance detection algorithm LandTrendr for the specified landscape.
- landscapename_canopy_cover_map_90m.tif (3 files: midnorth_canopy_cover_map_90m.tif, north_canopy_cover_map_90m.tif, south_canopy_cover_map_90m.tif) is a raster (90 m resolution) of canopy cover projected from forest age rasters based on stand age and canopy cover data collected in a previous study (see ocr_age_canopycover.csv).
- landscapename_early_seral_binary_map.tif (3 files: midnorth_early_seral_binary_map.tif, north_early_seral_binary_map.tif, south_early_seral_binary_map.tif) is a binary raster (90 m resolution) of early seral forest derived from the previously described canopy cover raster. Early seral forest pixels are defined as those with less than 0.75 canopy cover.
- landscapename_roads_binary.tif (3 files: midnorth_roads_binary.tif, north_roads_binary.tif, south_roads_binary.tif) is a binary raster (30 m resolution) of roads in a landscape.
- landscapename_roads_proportion_90m.tif (3 files: midnorth_roads_proportion_90m.tif, north_roads_proportion_90m.tif, south_roads_proportion_90m.tif) is a raster (90 m resolution) of roads in a landscape aggregated from 30-m resolution binary maps. Cell values represent proportion of road cover.
data file descriptions:
- landscapename.species.90m.RData (6 files: midnorth.bomcal.90m.RData, north.bomcal.90m.RData, south.bomcal.90m.RData, midnorth.bomvos.90m.RData, north.bomvos.90m.RData, south.bomvos.90m.RData) is packaged data ready for import in foraging model analyses.
- structural_connectivity.RData is packaged data ready for import in structural connectivity analyses
Sharing/Access information
Links to other publicly accessible locations of the data:
Code/Software
Unless otherwise specified, scripts are to be used in R (v. 4.1.1 in manuscript). Necessary packages are loaded within scripts and include:
- Spatial packages
- raster
- sf
- rgeos
- Tidying and visualization packages
- optparse
- tidyverse
- cowplot
- grid
- gridExtra
- patchwork
- GGally
- smoothr
- viridis
- linear modeling packages
- glmmTMB
- DHARMa
- ggeffects
- emmeans
- autocorrelation packages
- gstat
- ncf
- data package loaded from github:
- sibships
script file descriptions:
- package_data.R packages raw_data into .RData files used in model fitting
- fit_stand_age.R and fit_stand_age_and_roads.R establish foraging model fitting and bootstrapping with command line parsing (supported by shell files fit_stand_age.sh and fit_stand_age_and_roads.sh) for entering model specs--study landscape, resolution, species, lower/upper bounds and grid search size for theta parameters, number of bootstrap simulations and block size--used in fitting for a foraging model containing 1 (stand age) or 2 (stand age and road cover) theta parameters of interest.
- GEE_script.txt is code used in Google Earth Engine to obtain stand age and NLCD rasters
- structural_connectivity.R is code used in structural analysis, including a function to calculate a modified Hanski connectivity index
- tmp2.R is a script demonstrating model fitting and bootstrapping for B. vosnesenskii in the midnorth landscape and produces data objects: bomvos.midnorth.90m.stand_age_and_roads.fitted.RData, bomvos.midnorth.90m.stand_age_and_roads.bootstrap.RData
- landscapename.stand_age.R (3 files: midnorth.stand_age.R, north.stand_age.R, south.stand_age.R) demonstrate resistance and foraging model fitting for effects of stand age on foraging range of B. vosnesenskii in the each study landscape using 90 m resolution data and produces figures: midnorth.sampling_dist.png, midnorth.sim_at_mle.png, midnorth.sim_at_null.png
- midnorth.roads.R demonstrates resistance model and foraging model fitting for effects of road cover on foraging range of B. vosnesenskii in the midnorth study landscape using 90 m resolution data
- midnorth.singletons.R demonstrates foraging model fitting with only singleton colonies for B. vosnesenskii in the midnorth landscape.
