Data and code from: Evaluating eDNA metabarcoding techniques for pollinator community assessment in a field and controlled experiment
Data files
Aug 27, 2025 version files 206.91 KB
-
ButterflyExhibitData_071425.xlsx
51.34 KB
-
Manuscript_code.R
51.25 KB
-
PollBiodiversityData_071425.xlsx
100.61 KB
-
README.md
3.70 KB
Abstract
Pollinators play a critical role in ensuring the stability of food systems, yet their populations are in decline. To better understand and promote pollinator biodiversity, this study explored the use of environmental DNA (eDNA) metabarcoding techniques to assess plant–pollinator interactions. We conducted two experiments to optimize eDNA metabarcoding strategies for detecting pollinators on flowering plants. In the first experiment, we compared visual observations and eDNA detection via Illumina sequencing to characterize pollinator visitation on two native plant species in public and private green spaces across the Richmond, Virginia metropolitan area. Our findings revealed notable differences between the two methods, with visual surveys more sensitive to Hymenoptera and eDNA more sensitive to plant pests and other organisms. We devised a second experiment in a controlled environment at the Lewis Ginter Botanical Garden butterfly exhibit. Here, we tested different sampling strategies, primer pairs, and DNA sequencing platform (using Oxford Nanopore Technology). Despite detecting two butterfly species present in the exhibit, the eDNA metabarcoding showed limited sensitivity to the expected Lepidoptera. Our results indicate that while eDNA metabarcoding can detect a broader range of eukaryotic organisms, it may not be as effective for monitoring specific pollinator taxa like Hymenoptera and Lepidoptera. Factors such as sample type, primer bias, sequencing platform, and bioinformatics pipeline may impact detection outcomes. This study underscores the need for combining traditional monitoring techniques with eDNA metabarcoding to gain a comprehensive understanding of plant–pollinator interactions and improve pollinator conservation efforts.
Dataset DOI: 10.5061/dryad.j0zpc86tc
Description of the data and file structure
This study explored the use of environmental DNA (eDNA) metabarcoding techniques to assess plant-pollinator interactions. We conducted two experiments: in the first experiment, we compared visual observations and eDNA detection via Illumina sequencing to characterize pollinator visitation on two native plant species in public and private green spaces across the Richmond, Virginia metropolitan area (called Pollinator Biodiversity Field Survey). For the second experiment, we sampled in a controlled environment at the Lewis Ginter Botanical Garden butterfly exhibit, tested two primers and used Oxford Nanopore Technology for DNA sequencing (called Butterfly Exhibit Survey).
Files and variables
File: ButterflyExhibitData_071425.xlsx
Description: The output of sequencing, clustering, and matching OTUs to taxa for the Butterfly Exhibit Survey. There are two tabs, one tab for each primer used (Hebert and Folmer, see text).
Variables
- sample is the name of each sample
- sample type is where the sample came from [fruit, flower (swab or whole flower), bark]
- flower is species of flower sampled
- abundance is read number
- cluster ID is cluster OTU
- pident is percent identity of cluster sequence to blastn taxon
- length is length of sequence matched
- mismatch is number of SNP mismatches
- gapopen is number of gaps in alignment
- bitscore is alignment score
- Acc is Accession number OTU matched to
- Kingdom, Phylum, Class, Order, Family, Genus, Species are identifiers of OTU
- Taxon is the final Taxa assigned after accounting for rules of assignment (length and percent identity, see text)
File: Manuscript_code.R
Description: The R code used for statistical analysis and plot creation.
File: PollBiodiversityData_071425.xlsx
Description: The output of sequencing, clustering, and matching OTUs to taxa for the Pollinator Biodiversity Field Survey. There are four tabs. The first 'VisualPA' is the presence/absence data for visual observations of pollinator interactions (includes site name, plant name and taxon to lowest id). The tab called 'eDNA_Liatris_all' include all OTUs that matched given the length and percent identity threshold noted in the text, filtered to Eukaryotes for the samples swabbed on plant Liatris spicata. The tab called 'eDNA_Pycnanthemum_all' include all OTUs that matched given the length and percent identity threshold noted in the text, filtered to Eukaryotes. The tab called 'eDNA_arthropoda' combines OTUs from both plant species, filtered to length, percent identity, and to Phylum Arthropoda.
Variables
- Sample is described by L/P for which plant, Site #, Flower #
- Site is what site sample was swabbed
- Flower is flower swabbed
- DNA_Sequence is the OTU cluster
- Abundance/Reads is the read number of that OTU cluster
- pident is percent identity of OTU matched to taxa
- length is read length
- mismatch is number of SNP mismatches
- gapopen is number of gaps in alignment
- evalue is match probability due to random chance
- bitscore is the alignment score
- Acc is Accession number OTU matched to
- Kingdom, Phylum, Class, Order, Family, Genus, Species are identifiers of OTU
- Taxon is the final Taxa assigned after accounting for rules of assignment (length and percent identity, see text)
Code/software
RStudio and R were used to write code for statistical analysis. See code, packages, etc. in file called 'Manuscript code'.