Bumble bees and honey bees on islands harbour reduced viral species richness, yet honey bee populations are dominated by a deformed wing virus recombinant
Data files
Sep 03, 2025 version files 49.22 MB
-
01_CoverM_analysis.R
21.78 KB
-
02_diversity_heatmap.R
10.77 KB
-
03_NMDS_permanova.R
9.90 KB
-
04_DWV_cov_plots.R
14.71 KB
-
blastx_results_unfiltered.zip
5.17 MB
-
CoverM_read_depth.zip
43.63 MB
-
read_coverage.csv
341.12 KB
-
README.md
5.88 KB
-
reference_names_final.csv
4.19 KB
-
sample_info.csv
2.88 KB
-
virus_wide_summary.csv
10.38 KB
Abstract
Pollinators harbour diverse RNA viromes that play a vital role in their health. Yet, factors that shape viral communities are often unclear. The European honey bee (Apis mellifera) is experiencing a viral epidemic since the emergence of the parasitic mite Varroa destructor (varroa) introduced vector-borne transmission, which has also been linked to increased viral spillover into wild pollinator communities. Varroa-free island populations provide natural laboratories to study the effect of varroa, while also allowing us to ask how islands affect viral communities. Barriers that restrict the dispersal of wild pollinators and their pathogens to islands may be overcome by human-mediated transport in managed honey bees. Here we used islands with and without varroa and matched mainland populations of honey bees (A. mellifera) and bumble bees (Bombus terrestris) from 2015 and 2021 to explore how varroa presence and island location affect the virome of managed and wild bees. We find lower viral richness on islands in both species. Bumble bees harbour a distinct viral community that was not affected by varroa but geographically structured. In honey bees, however, varroa-present populations contained more viral reads driven by a high abundance of deformed wing virus (DWV). Within the six years between the sampling events, DWV underwent a shift from mostly DWV-B towards a mix of DWV-B and recombinant strains. Surprisingly, these shifts appeared independent of varroa. Viewing pollinator virome composition within an ecological framework provides valuable insights into the barriers to virus spread and could help to predict drivers of disease emergence.
Dataset DOI: 10.5061/dryad.jsxksn0ns
Description of the data and file structure
This dataset contains the raw and processed data, BLASTx results, R scripts, and supporting files used in the analyses described in "Bumble bees and honey bees on islands harbour reduced viral species richness, yet honey bee populations are dominated by a deformed wing virus recombinant" by Jana Dobelmann and Lena Wilfert. All files are provided to facilitate reproducibility.
Files and variables
File: 02_diversity_heatmap.R
Description: This R script is used to analyse the virome diversity and create plots for viral richness, simpson's diversity and a heatmap. It requires "read_coverage.csv.csv" as input.
File: 01_CoverM_analysis.R
Description: This R script is used to apply detection threshold for read coverage files and gives basic statisical analyses of viral read coverage and abundance. It is also used to plot a map with sampling locations and pie charts with viral abundance. the virome diversity and create plots for viral richness, simpson's diversity and a heatmap. It requires "read_coverage.csv" as input.
File: 03_NMDS_permanova.R
Description: This R script is used to analyse beta diversity in virome data, including a permanova and creates NMDS plots. It requires "virus*wide*summary.csv" as input.
File: read_coverage.csv
Description: Reads mapping viral references using CoverM.
Variables
- X: Running number
- reference: Name of reference in FASTA files used during mapping
- sample: RNA-seq library (numbered)
- Site: sampling site
- Virus: Virus name
- Island: yes - islands, no - mainland
- Year: collection year, 2015 or 2021
- Species: host species, Apis mellifera or Bombus terrestris
- Varroa: varroa presence at site, yes or no
- Latitude: Latitude sampling site
- Longitude: Longitude sampling site
- sum_depth: Sum of mapped reads
- length: Reference length or genome size in bases
- average_depth: average read depth
- coverage: bases covered by reads
- perc_coverage: percent genome covered by reads
- perc_RdRp_coverage: percent RdRp covered by reads
- N: number of individuals in sample pool
- n_reads: total reads in library
File: virus_wide_summary.csv
Description:
Variables
- sample: RNA-seq library (numbered)
- Site: sampling site
- Island: yes - islands, no - mainland
- Year: collection year, 2015 or 2021
- Species: host species, Apis mellifera or Bombus terrestris
- Varroa: varroa-presence at sampling site
- Latitude: Latitude
- Longitude: Longitude
- N:number of individuals in sample pool
- n_reads: total reads in library
- BQCV Cherb.: normalised average read coverage over genome
- BQCV B-I: normalised average read coverage over genome
- Bt Iflavirus 2: normalised average read coverage over genome
- Bt Iflavirus 1: normalised average read coverage over genome
- SBPV roth: normalised average read coverage over genome
- SBPV harp: normalised average read coverage over genome
- LSV: normalised average read coverage over genome
- LSV1: normalised average read coverage over genome
- LSV TO: normalised average read coverage over genome
- LSV Lpool: normalised average read coverage over genome
- LSV France: normalised average read coverage over genome
- LSV3 Scilly: normalised average read coverage over genome
- AGV: normalised average read coverage over genome
- AMV1: normalised average read coverage over genome
- AMV3: normalised average read coverage over genome
- MfV1: normalised average read coverage over genome
- MLV: normalised average read coverage over genome
- ARV1: normalised average read coverage over genome
- LSV4: normalised average read coverage over genome
- LSV3 Liaoning: normalised average read coverage over genome
- MfV2: normalised average read coverage over genome
- BMlV: normalised average read coverage over genome
- BCD: normalised average read coverage over genome
- DWV-A: normalised average read coverage over genome
- DWV-B: normalised average read coverage over genome
- SBV: normalised average read coverage over genome
- SBV-like: normalised average read coverage over genome
- Bt Iflavirus 3: normalised average read coverage over genome
File: sample_info.csv
Description:
Variables
- sample: RNA-seq library (numbered)
- Site: collection site
- Year: sampling year, 2015 or 2021
- Species: host species
- Varroa: varroa presence at site, yes or no
- Island: yes - islands, no - mainland
- Latitude: Longitude
- Longitude: Latitude
File: CoverM_read_depth.zip
Description: Archive containing 44 files with the read depth. Each file has three columns which show the reference name, genome position, and number of reads that mapped at the genome position.
File: 04_DWV_cov_plots.R
Description: R script uder to generate plots that show the coverage of deformed wing virus reads over the genome. Requires "CoverM_read_depth.zip", "sample_info.csv" and "reference_names_final.csv".
File: reference_names_final.csv
Description:
Variables
- Name: Reference name
- Description: Virus name
- Virus: Virus acronym
- Family: Virus family (if known)
- Genome: Virus genome organization
- Detected: if virus was detected with >33% genome coverage and >75% RdRp coverage in at least one library
- Length: Reference length in bases
- RdRp_start: start of the RdRp in bases
- RdRp_end: end of the RdRp in bases
File: blastx_results_unfiltered.zip
Description: Archive containing results from blastx searched of de novo assembled contigs against a large protein reference database for all 44 samples.
Code/software
R
