Molecular versus microscopic analyses tell different stories of little brown bat (Myotis lucifugus) diet in Alaska
Data files
Jan 15, 2026 version files 3.44 MB
-
AlaskaMyLu_ASV_Table.csv
413.25 KB
-
ASVfasta.txt
269.07 KB
-
DietComparisonAnalysis.html
1.08 MB
-
DietTrial.html
1.67 MB
-
MyotisASVData.csv
2.24 KB
-
MyotisMetadata.csv
1.29 KB
-
README.md
4.98 KB
Abstract
Populations of wide-ranging species on the edge of their geographic range may experience environmental conditions unlike those of conspecifics from the core of the range. We conducted a study on the diet of little brown bats (Myotis lucifugus) near the northern edge of their range in Alaska, USA, using molecular fecal dietary techniques and compared these results to a previous diet study using microscopic analysis of the same samples. Molecular dietary analysis suggested that M. lucifugus in this population rely heavily on flies (Diptera), moths (Lepidoptera), and caddisflies (Trichoptera). Our results are consistent with other studies in the insect orders eaten but differ in the relative importance of those orders compared with molecular analyses of diet from M. lucifugus populations nearer the core species range. Our results also contrast with microscopic assessments of diet from these Alaskan samples, which identified spiders (Araneae) as the primary prey item for this population. Importantly, our study demonstrates how different types of analyses can yield different characterizations of diet, even from the same samples.
Dataset DOI: 10.5061/dryad.1c59zw49t
Description of the data and file structure
The DietTrial.html RMarkDown file documents the assembly and filtering of fastq files from DNA metabarcoding of bat feces on the MiSeq platform. Assembly was conducted using dada2.
ASVfasta.txt contains the assembled and filtered fasta file output of dada2, and AlaskaMyLu_ASV_Table.xlsx contains the taxonomic IDs for each sequence from BOLD.
I created the DietComparisonAnalysis.html RMarkDown file to document my analysis of diet metabarcode data. I have characterizations of diet from the same samples analyzed (A) with molecular metabarcoding and (B) with microscopic analyses. I want to know whether differences between diet methods are statistically larger than inter-individual variances for the same method. I want to (1) visualize clustering by method using NMDS or PCoA, (2) evaluate significance of clustering by method using PERMANOVA or AnoSim, and (3) use SimPer to determine which taxa are driving differences by method. This work is adapted from Dill-McFarland & Cox (2019) workshop at [https://rpubs.com/maddieSC/R_SOP_BRC_Oct_2019].
The csv datafiles include both the percent occurrence of arthropod orders in each sample (MyotisASVData.csv), and metadata (MyotisMetadata.csv) for each sample. These are input files for the diet comparison analysis.
Files and variables
File: DietComparisonAnalysis.html
Description: This is an RMarkDown file documenting the comparative analysis of diet as assessed by DNA metabarcoding and by microscopic analyses.
File: MyotisASVData.csv
Description: This is the percent occurrence of arthropod orders in each sample.
Variables
- Taxon: Sample IDs. The first 22 rows of data are from DNA barcoding on a MiSeq platform. The last 22 rows of data are from visual/microscopic analyses of arthropod parts in feces.
- Araneae: Spiders
- Trombidiformes: Mites
- Coleoptera: Beetles
- Diptera: Flies
- Hemiptera: True bugs
- Hymenoptera: Wasps
- Lepidoptera: Moths
- Neuroptera: Lacewings
- Trichoptera: Caddisflies
- Other_Insecta: Pooled classification of sequences or prey items that could not be reliably identified below the Class level
File: MyotisMetadata.csv
Description: This is a file of metadata for all samples.
Variables
- Sample: Sample IDs. The first 22 rows of data are from DNA barcoding on a MiSeq platform. The last 22 rows of data are from visual/microscopic analyses of arthropod parts in feces.
- Analysis: Diet was assessed either by DNA metabarcoding ("Molecular") or by visual/microscopic identification of parts of prey items ("Microscopic").
- Date: The date on which the fecal sample was collected in the field.
- Sex: The sex of the bat that contributed the fecal sample.
File: ASVfasta.txt
Description: A fasta file containing 1608 partial COX1 sequences from DNA metabarcoding of bat fecal samples on a MiSeq platform
Variables
- Reads are numbered sequentially (ASV1-ASV1608) and correspond to the "Sample" column in the AlaskaMyLu_ASV_Table.xlsx file.
File: AlaskaMyLu_ASV_Table.csv
Description: This is the prey identification output for the DNA metabarcoding assessment of diet.
Variables
- Sample: Sequence read codes. These correspond to the codes in the ASVfasta.txt fasta file.
- 13-10-S68 to WRST1-S20: Each column is a different fecal sample. Counts in each column indicate the number of reads present for a given prey item.
- NEG-S80: Negative control in the MiSeq run.
- Total hits: Number of reads present for a given prey item across all samples.
- Phylum to Species: Taxonomic ID for a given read. Identifications are from BOLD. Taxonomic levels for which confident identifications could not be made are indicated with "NA".
- Similarity: Percent similarity to BOLD reference sequence.
- Status: Reference sequence status in BOLD.
- Flags: Taxonomic ID warnings from BOLDigger. ASVs with no taxonomic ID warnings are indicated with "NA".
- Notes: Details on the taxonomic ID warnings from BOLDigger. ASVs with no taxonomic ID warnings are indicated with "NA".
File: DietTrial.html
Description: This is an RMarkDown file documenting the assembly and filtering of DNA metabarcoding data using dada2.
Code/software
All analyses were conducted in RStudio v.2025.05.1 using packages 'dada2' v.1.16.0 and ‘vegan’ v.2.7-1.
Access information
Data was derived from the following sources:
- Boyles, J. G., L. P. McGuire, E. Boyles, J. P. Reimer, C. A. C. Brooks, R. W. Rutherford, T. A. Rutherford, J. O. Whitaker Jr., and G. F. McCracken. 2016. Physiological and behavioral adaptations in bats living at high latitudes. Physiology & Behavior 165: 322–327.