Bumble bee nest density is lower in drought years
Data files
Aug 25, 2025 version files 49.88 KB
-
all_years_capture_history.csv
4.52 KB
-
drought.figures.R
3.58 KB
-
nestRmarkPalmerDrought.R
22.74 KB
-
palmer_index_coastal_MA_monthly.csv
1.79 KB
-
pmdi_monthly_coastal_MA_1960.2024.csv
10.60 KB
-
README.md
6.65 KB
Abstract
Climate change is driving more frequent and intense droughts in many parts of the world. While we hold a broad understanding of how drought impacts plant populations, predicting drought’s demographic effects on animal populations remains a challenge, particularly for invertebrates. Here, we use seven years of bumble bee nest surveys spanning a period with two historic droughts to estimate the impacts of drought on bumble bee nest abundance and apparent survival. Nest abundance significantly declined with drought (average of 19.3 vs. 11.3 nests in non-drought vs. drought years). Relative nest abundances of Bombus impatiens and Bombus griseocollis, two common species in this system, did not change with drought despite differences in life history strategies and apparent survival. Bombus impatiens abundance in forests increased during droughts (~60% of nests in forests, 20%-40% in drought versus non-drought years). Surprisingly, nests of rarer species were absent from our survey area after the first of the two droughts, but the two common species recovered quickly in the year after each drought. This study demonstrates that well-known effects of drought on floral resources likely translate to population-level impacts in bumble bees. It also highlights the importance of long-term monitoring for detecting impacts of intermittent environmental disturbances.
Dataset DOI: 10.5061/dryad.6t1g1jxbn
Description of the data and file structure
These data come from the current manuscript, which was a study of bumblebee nest density in relation to drought.
Files and variables
File: drought.figures.R
Description: This file contains code for Figure 1 and Figure 2 (in the main manuscript) using two datasets of Palmer Drought Indices.
File: nestRmarkPalmerDrought.R
Description: This file performs mark-recapture analyses of bumble bee nest detection and creates Figures 3A and 3B (in the main manuscript) from the output of the mark-recapture analysis. It also includes post-hoc analyses using linear regression and creates Figures 4A, 4B, and 4C (in the main manuscript) from the output of these post-hoc analyses. The supplementary tables S1, S2, and S3 were produced from the mark-recapture analysis model selection output.
File: palmer_index_coastal_MA_monthly.csv
Description: This file includes Palmer Drought Indices (PDI) averaged for each month from 2005-2024 in coastal Massachusetts. The data are sourced from https://www.ncei.noaa.gov/access/monitoring/weekly-palmers/time-series/1903. This file was used to create Figure 1 in the main manuscript. These PDI values were also used in mark-recapture analysis to interpret the effects of drought on bumblebee nest detection. The columns have the following meanings:
Variables
- year: year in which values were recorded (2005-2024)
- january: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during January
- february: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during February
- march: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during March
- april: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during April
- may: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during May
- june: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during June
- july: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during July
- august: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during August
- september: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during September
- october: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during October
- november: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during November
- december: Values are average PDI values (index values can range from minimum -6 to maximum 6) calculated from weekly PDI values during December
File: all_years_capture_history.csv
Description: This file contains capture history data for all bumblebee nests detected in our study site over seven years. Each row represents one nest and its associated capture history. This file was used to perform mark-recapture analysis and produce Figures 3 (A&B) and Figures 4 (A&B&C) in the main manuscript. The columns have the following meanings:
Variables
- year: Year in which data was collected (2018-2024)
- Species: Species identification of the bumblebee nest
- Habitat Type: Habitat in which nest was detected (Meadow, Forest, or Hayfield)
- S1: First survey event during which a nest was or was not detected. “1” = nest was detected during the survey. “0” = nest was not detected during the survey.
- S2: Second survey event during which a nest was or was not detected. “1” = nest was detected during the survey. “0” = nest was not detected during the survey.
- S3: Third survey event during which a nest was or was not detected. “1” = nest was detected during the survey. “0” = nest was not detected during the survey.
- S4: Fourth survey event during which a nest was or was not detected. “1” = nest was detected during the survey. “0” = nest was not detected during the survey.
- S5: Fifth survey event during which a nest was or was not detected. “1” = nest was detected during the survey. “0” = nest was not detected during the survey.
- ch: Capture history for each nest, created by collating the detection values across five surveys and representing a record of whether an individual nest was detected or not detected over the course of five surveys.
File: pmdi_monthly_coastal_MA_1960.2024.csv
Description: This file includes monthly Palmer Modified Drought Index (PMDI) values from 1960-2024 in coastal Massachusetts. The data are sourced from https://www.ncei.noaa.gov/access/monitoring/climate-at-a-glance/divisional/time-series. This file was used to create Figure 2 in the main manuscript. The columns have the following meanings:
Variables
- year: year in which values were recorded (1960-2024)
- month: month in which values were recorded, presented as numerical representations of each month (i.e., “1” = ”January”)
- pmdi: Palmer Modified Drought Index value for specified month (index values can range from minimum -6 to maximum 8). PMDI is the operational Palmer Drought Severity Index.
Code/software
No software is needed to view the data. The code is for doing data analysis in the open-source software R.
Access information
Other publicly accessible locations of the data:
- These data are also available on GitHub
Data was derived from the following sources:
- Monthly PMDI values from 1960 to 2024 are sourced from https://www.ncei.noaa.gov/access/monitoring/climate-at-a-glance/divisional/time-series
- Monthly PDI values from 2005 to 2024 are sourced from https://www.ncei.noaa.gov/access/monitoring/weekly-palmers/time-series/1903
