Increased drought drives avian community declines in the warm deserts of the United States
Data files
Jul 24, 2025 version files 2.64 MB
-
birddata_bigger.csv
2.58 MB
-
Final_Script_BBS.R
58.52 KB
-
README.md
2.46 KB
Abstract
The frequencies, intensities, and durations of extreme weather are increasing under climate change, furthering biodiversity loss. In the Southwestern USA, rising drought frequencies and intensities are anticipated to increasingly affect wildlife and their habitat. Particularly in highly arid regions, where species live closer to the limit of physiological tolerances, increasing drought emerges as a major threat to biodiversity. Using North American Breeding Bird Survey data (1998–2022) in combination with the Standardized Precipitation-Evaporation Index (SPEI), we estimated the effects of short-term (3-month), annual (12-month), and prolonged (36-month) drought on avian abundance and species richness in the warm deserts of the USA. We assessed community-level drought effects using six Bayesian relative abundance and species richness models. Our results indicate declines in avian community abundance and richness under multiple drought durations, with the strongest effects associated with annual droughts. Our models estimated that in years where annual drought met the threshold of ‘severe’, 9 out of 38 species would have 18–34% declines in relative abundance compared to a year with long-term average conditions, and species richness would decline by 10%. Availability of open water slightly mitigated negative drought effects on the avian community. Our results suggest drought will increasingly contribute to the collapse of aridland avian communities through declines in habitat generalists and obligate aridland species. Beyond mitigating future climate change, developing local adaptation strategies in arid regions, such as the conservation of water sources, appears necessary for biodiversity conservation.
Dataset DOI: 10.5061/dryad.866t1g229
Description of the data and file structure
Final_Script_BBS.R: All code required to create the models and results visualization, as well as links to the open-source data used for the analysis.
birddata_bigger.csv: CSV file with breeding bird survey data used for the analysis.
Files and variables
File: Final_Script_BBS.R
Description: All code required to create the models and results visualization, as well as links to the open-source data used for the analysis.
File: birddata_bigger.csv
Description: CSV file with breeding bird survey data used for the analysis.
Variables
- Year: Year of survey
- English_Common_Name: Common name of bird
- Sroute: Route number attached to a state number
- ObsN: Unique observation number
- TotalSpp: Total species observed during the particular survey
- count: count of the species in question
- Route_year: Combination of the route number and survey year
- Julian_day: Julian day at which the survey took place
- water: Percentage of open water during this year at this location
- wetland: Percentage of wetland during this year at this location
- shrub: Percentage of shrubland during this year at this location
- barren: Percentage of barren land during this year at this location
- crop: Percentage of cropland during this year at this location
- developed: Percentage of developed land during this year at this location
- spei_03: SPEI-3 values during this year at this location
- spei_12: SPEI-12 values during this year at this location
- spei_36: SPEI-36 values during this year at this location
Code/software
We used R version 4.4.3 for analysis. All necessary software packages are listed within the .R file.
Access information
Data was derived from the following publicly available sources:
- SPEI-gamma data were downloaded from the nClimGrid monthly: https://www.ncei.noaa.gov/pub/data/nidis/indices/nclimgrid-monthly/
- Annual NLCD data were obtained from: https://www.mrlc.gov/data
- North American Breeding Bird Survey data were obtained from: https://www.sciencebase.gov/catalog/item/52b1dfa8e4b0d9b325230cd9