Density as a mechanism linking habitat disturbance to increased disease prevalence: evidence from a natural experiment
Data files
Oct 10, 2025 version files 354.08 KB
-
Abundance-Estimates_Aboreas_2015-2019.txt
1.79 KB
-
Capture-Surveys_Aboreas_2015-2019.txt
150.03 KB
-
Density-Estimates_Aboreas_2015-2019.txt
184 KB
-
Habitat-Disease_Aboreas_2015-2019.txt
756 B
-
Pond-Area_2015-2019.txt
2.25 KB
-
README.md
15.25 KB
Abstract
Sudden habitat loss associated with environmental disturbance can trigger animals to move from affected to undisturbed areas, where increases in local density may occur. Such increases in density can affect a number of ecological processes. Although pathogen transmission is strongly related to local density, how crowding after habitat loss affects infection dynamics in wild populations remains unclear. Here we conceptualize the Disturbance-Density-Disease (DDD) hypothesis, which posits that disturbance-induced habitat loss results in increased pathogen prevalence via increases in local density at adjacent, undisturbed patches. We then used empirical data from before, during, and after an extreme flooding event to test the DDD hypothesis in boreal toads Anaxyrus boreas boreas co-occurring with the pathogenic fungus Batrachochytrium dendrobatidis (Bd). We collected Bd samples from each captured individual during a 5-year (2015–2019) mark-recapture study of boreal toads (n = 1,295) that breed in beaver ponds in western Wyoming, USA. During spring of 2017, an extreme flooding event destroyed several beaver dams, resulting in the loss of breeding habitat. We compared host density and pathogen prevalence pre- and post-disturbance at sites affected versus unaffected by flooding. At affected sites, both population density and Bd prevalence increased at adjacent, undisturbed ponds following the sudden loss of habitat. Moreover, neither host density nor Bd prevalence increased at control sites that were in areas unaffected by flooding. Taken together, our results provide evidence that supports hypothesized links between disturbance, adjacent increases in density, and subsequent increases in pathogen prevalence. We thus demonstrate an important consequence of disturbance beyond proximate habitat loss, and introduce a clear conceptual approach (the DDD hypothesis) to understanding how pathogen transmission can be affected by disturbance via alterations to local density.
DATASET 1 of 5
#############################################################################
GENERAL INFORMATION
Title of Dataset: Capture-Surveys_Aboreas_2015-2019
Author Information
A. Principal Investigator Contact Information
Name: Gabriel Barrile
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: gbarrile@uwyo.edu
B. Associate or Co-investigator Contact Information
Name: Jerod Merkle
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: jmerkle@uwyo.edu
C. Alternate Contact Information
Name: Gabriel Barrile
Email: gbarrile15@gmail.com
Date of data collection:
May 2015 - June 2019
Geographic location of data collection:
Bridger-Teton National Forest, western Wyoming, USA
Funding source that supported the collection of the data:
Wyoming Game and Fish Department for funding (grant number 1003570-13403), Clear Creek Foundation, University of Wyoming School of Computing, and the Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health (2P20GM103432).
SHARING/ACCESS INFORMATION
Recommended citation for this dataset:
Barrile, G. M., Chalfoun, A. D., Walters, A. W., & Merkle, J. A. (2025). Data from: Density as a mechanism linking habitat disturbance to increased pathogen prevalence: evidence from a natural experiment. Dryad Digital Repository.
METHODOLOGICAL INFORMATION
Description of methods used for collection/generation of data:
Capture-mark-recapture methods of adult boreal toads
Methods for processing the data:
Data are in the raw format collected in the field.
Instrument- or software-specific information needed to interpret the data:
We used Program R to format and analyze data. All R packages and functions needed to interpret the data are cited in the manuscript.
DATA-SPECIFIC INFORMATION FOR: Capture-Surveys_Aboreas_2015-2019
Number of variables: 5
Number of cases/rows: 3084 (including the top row of column names)
Variable List and Information:
-
Survey_Date = date of each capture-recapture survey in year-month-day format
-
Survey_Number = the survey number at that given site (used to create capture histories)
-
Site = stream reach where the individual was captured (i.e., Buck, Chall, Lower Gypsum, and Upper Gypsum creeks)
-
Mountain_Range = mountain range where the individual was captured (i.e., "WYO" = Wyoming Range and "WIND" = Wind River Range)
-
Pit_Tag = the unique identifier for the individual that was captured. Numbers in this column were derived from passive integrated transponder tags, which were inserted subcutaneously in each adult boreal toad upon capture. In other words, each unique number in this column represents a unique individual.
*There should be no "NA" in the dataset
######################################################################################
DATASET 2 of 5
#####################################################################################
GENERAL INFORMATION
Title of Dataset: Pond-Area_2015-2019
Author Information
A. Principal Investigator Contact Information
Name: Gabriel Barrile
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: gbarrile@uwyo.edu
B. Associate or Co-investigator Contact Information
Name: Jerod Merkle
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: jmerkle@uwyo.edu
C. Alternate Contact Information
Name: Gabriel Barrile
Email: gbarrile15@gmail.com
Date of data collection:
May 2015 - June 2019
Geographic location of data collection:
Bridger-Teton National Forest, western Wyoming, USA
Funding source that supported the collection of the data:
Wyoming Game and Fish Department for funding (grant number 1003570-13403), Clear Creek Foundation, University of Wyoming School of Computing, and the Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health (2P20GM103432).
SHARING/ACCESS INFORMATION
Recommended citation for this dataset:
Barrile, G. M., Chalfoun, A. D., Walters, A. W., & Merkle, J. A. (2025). Data from: Density as a mechanism linking habitat disturbance to increased pathogen prevalence: evidence from a natural experiment. Dryad Digital Repository.
METHODOLOGICAL INFORMATION
Description of methods used for collection/generation of data:
We recorded the surface area of each pond using a handheld GPS to trace the pond perimeter (and subsequently calculated area from perimeter measurements).
Methods for processing the data:
Data are in the raw format collected in the field.
Instrument- or software-specific information needed to interpret the data:
We used Program R to format and analyze data. All R packages and functions needed to interpret the data are cited in the manuscript.
DATA-SPECIFIC INFORMATION FOR: Pond-Area_2015-2019
Number of variables: 5
Number of cases/rows: 66 (including the top row of column names)
Variable List and Information:
-
Site = study stream (i.e., Buck, Chall, Lower Gypsum, and Upper Gypsum creeks)
-
Year = year of the survey
-
Pond = the specific pond name/number at each site
-
Pond_Area = the area of each pond in sq. m.
-
Disturbance = a categorical variable denoting whether the survey was conducted before, during, or after the flooding event
*There should be no "NA" in the dataset
######################################################################################
DATASET 3 of 5
#####################################################################################
GENERAL INFORMATION
Title of Dataset: Abundance-Estimates_Aboreas_2015-2019
Author Information
A. Principal Investigator Contact Information
Name: Gabriel Barrile
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: gbarrile@uwyo.edu
B. Associate or Co-investigator Contact Information
Name: Jerod Merkle
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: jmerkle@uwyo.edu
C. Alternate Contact Information
Name: Gabriel Barrile
Email: gbarrile15@gmail.com
Date of data collection:
May 2015 - June 2019
Geographic location of data collection:
Bridger-Teton National Forest, western Wyoming, USA
Funding source that supported the collection of the data:
Wyoming Game and Fish Department for funding (grant number 1003570-13403), Clear Creek Foundation, University of Wyoming School of Computing, and the Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health (2P20GM103432).
SHARING/ACCESS INFORMATION
Recommended citation for this dataset:
Barrile, G. M., Chalfoun, A. D., Walters, A. W., & Merkle, J. A. (2025). Data from: Density as a mechanism linking habitat disturbance to increased pathogen prevalence: evidence from a natural experiment. Dryad Digital Repository.
METHODOLOGICAL INFORMATION
Description of methods used for collection/generation of data:
Mean abundance estimates and error around such estimates were obtained from a robust-design mark-recapture model.
Methods for processing the data:
Data are in the raw format obtained from the model output.
Instrument- or software-specific information needed to interpret the data:
We used Program R to format and analyze data. All R packages and functions needed to interpret the data are cited in the manuscript.
DATA-SPECIFIC INFORMATION FOR: Abundance-Estimates_Aboreas_2015-2019
Number of variables: 7
Number of cases/rows: 21 (including the top row of column names)
Variable List and Information:
-
estimate = mean abundance estimate for each site and year, obtained from the mark-recapture model. The units for abundance are the 'number of toads' (i.e., the number of adult male boreal toads).
-
se = standard error around the mean abundance estimate
-
lcl = lower 95% confidence limit around the mean abundance estimate
-
ucl = upper 95% confidence limit around the mean abundance estimate
-
Year = year of the study
-
Site = study stream ("LG" = Lower Gypsum Creek and "UG" = Upper Gypsum Creek)
-
Year2 = a duplicate column of year, used in data formatting at a later step in the workflow
*There should be no "NA" in the dataset
######################################################################################
DATASET 4 of 5
#####################################################################################
GENERAL INFORMATION
Title of Dataset: Density-Estimates_Aboreas_2015-2019
Author Information
A. Principal Investigator Contact Information
Name: Gabriel Barrile
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: gbarrile@uwyo.edu
B. Associate or Co-investigator Contact Information
Name: Jerod Merkle
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: jmerkle@uwyo.edu
C. Alternate Contact Information
Name: Gabriel Barrile
Email: gbarrile15@gmail.com
Date of data collection:
May 2015 - June 2019
Geographic location of data collection:
Bridger-Teton National Forest, western Wyoming, USA
Funding source that supported the collection of the data:
Wyoming Game and Fish Department for funding (grant number 1003570-13403), Clear Creek Foundation, University of Wyoming School of Computing, and the Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health (2P20GM103432).
SHARING/ACCESS INFORMATION
Recommended citation for this dataset:
Barrile, G. M., Chalfoun, A. D., Walters, A. W., & Merkle, J. A. (2025). Data from: Density as a mechanism linking habitat disturbance to increased pathogen prevalence: evidence from a natural experiment. Dryad Digital Repository.
METHODOLOGICAL INFORMATION
Description of methods used for collection/generation of data:
Mean predicted abundance estimates and error around such estimates were derived from a robust design capture-mark-recapture model. Density was calculated via abundance divided by pond area at each study stream. Density values were derived from random draws from a log-normal distribution, using mean abundance estimates and 95% confidence limits in each site-year. We drew 100 values in each site-year to generate distributions of host density values that reflected the skewed uncertainty in our abundance estimates (i.e., confidence intervals for abundance were asymmetrical, with a slight right-skew).
Methods for processing the data:
Data are in the format used for analysis.
Instrument- or software-specific information needed to interpret the data:
We used Program R to format and analyze data. All R packages and functions needed to interpret the data are cited in the manuscript.
DATA-SPECIFIC INFORMATION FOR: Density-Estimates_Aboreas_2015-2019
Number of variables: 8
Number of cases/rows: 2001 (including the top row of column names)
Variable List and Information:
-
Site = study stream (Buck, Chall, Lower Gypsum, and Upper Gypsum Creeks)
-
Year = year of the study
-
SiteYear = a combination of the two variables described above (e.g., Buck-2015)
-
abundance = abundance estimate for each site and year. The units for abundance are the 'number of toads' (i.e., the number of adult male boreal toads).
-
Density = density estimates for each site and year. The unit for density is simply toad abundance divided by the total pond area at each study stream (i.e., the number of individuals per square meter of available breeding habitat).
-
Pond_Area = the total area of ponds at each site (in sq. m.)
-
Prevalence = disease prevalence, or the number of Bd positive individuals out of the total number of individuals tested
-
Disturbance = a categorical variable denoting whether the survey was conducted before, during, or after the flooding event
*There should be no "NA" in the dataset
######################################################################################
DATASET 5 of 5
#####################################################################################
GENERAL INFORMATION
Title of Dataset: Habitat-Disease_Aboreas_2015-2019
Author Information
A. Principal Investigator Contact Information
Name: Gabriel Barrile
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: gbarrile@uwyo.edu
B. Associate or Co-investigator Contact Information
Name: Jerod Merkle
Institution: University of Wyoming
Address: 1000 E University Ave, Laramie, WY, 82071
Email: jmerkle@uwyo.edu
C. Alternate Contact Information
Name: Gabriel Barrile
Email: gbarrile15@gmail.com
Date of data collection:
May 2015 - June 2019
Geographic location of data collection:
Bridger-Teton National Forest, western Wyoming, USA
Funding source that supported the collection of the data:
Wyoming Game and Fish Department for funding (grant number 1003570-13403), Clear Creek Foundation, University of Wyoming School of Computing, and the Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health (2P20GM103432).
SHARING/ACCESS INFORMATION
Recommended citation for this dataset:
Barrile, G. M., Chalfoun, A. D., Walters, A. W., & Merkle, J. A. (2025). Data from: Density as a mechanism linking habitat disturbance to increased pathogen prevalence: evidence from a natural experiment. Dryad Digital Repository.
METHODOLOGICAL INFORMATION
Description of methods used for collection/generation of data:
We recorded the surface area of each pond using a handheld GPS to trace the pond perimeter (and subsequently calculated area from perimeter measurements). We also collected disease samples from every individual encountered in a given breeding season using standardized protocols to swab ventral skin surfaces of host amphibians for Bd. Disease samples were analyzed by Pisces Molecular® (Boulder, Colorado, USA) via a quantitative polymerase chain reaction to detect Bd.
Methods for processing the data:
Data are in their raw format.
Instrument- or software-specific information needed to interpret the data:
We used Program R to format and analyze data. All R packages and functions needed to interpret the data are cited in the manuscript.
DATA-SPECIFIC INFORMATION FOR: Habitat-Disease_Aboreas_2015-2019
Number of variables: 6
Number of cases/rows: 21 (including the top row of column names)
Variable List and Information:
-
Site = study stream (Buck, Chall, Lower Gypsum, and Upper Gypsum Creeks)
-
Year = year of the study
-
Pond_Area = the total area of ponds at each site (in sq. m.)
-
Bd_Pos = the number of toads that tested positive for Bd for a given site-year
-
Bd_Total = the total number of toads that were tested for Bd for a given site-year
-
Disturbance = a categorical variable denoting whether the survey was conducted before, during, or after the flooding event
*There should be no "NA" in the dataset
