Data from: Correlated divergence of ecology, morphology, and fine-scale vocal motor performance among sparrow subspecies
Data files
Jul 30, 2025 version files 2.77 MB
-
alldata3.csv
2.65 MB
-
GapBWPerf_pub.R
1.66 KB
-
GapLenPref_pub.R
3.98 KB
-
LenBWPerf_pub.R
8.68 KB
-
multiphrasesongs.csv
30.42 KB
-
NoSyll.csv
7.42 KB
-
README.md
3.63 KB
-
SubspeciesAnalysis_pub.R
33.07 KB
-
SubspeciesDQR_pub.R
21.88 KB
Abstract
Animal structures often evolve for multiple functions, such as limbs used for both climbing and fighting. Adaptations that optimize one function may limit performance of others. Here, we compare songs of inland (freshwater) and coastal (saltmarsh) subspecies of swamp sparrows (Melospiza georgiana) for which adaptation to distinct habitats has driven divergence in beak size and shape. Prior studies have shown that these subspecies’ songs differ in a broad-scale metric of vocal performance—vocal deviation—in a direction consistent with the hypothesis that vocal performance is disproportionately constrained in larger-beaked birds. We hypothesize that beak divergence has also driven divergence in additional, more fine-scale phonological and temporal measures of song including note types and the acoustic structure of notes and inter-note gaps. Consistent with expectations, coastal birds’ songs show greater proportions of low-performance notes and note transitions, notes and inter-note gaps that span narrower bandwidths, and notes situated farther from presumed performance maxima. Yet in contrast to expectations, coastal birds’ songs show shorter average inter-note gaps, suggesting a compensatory strategy to maintain baseline levels of vocal performance. We also find that, for each subspecies, males maximized performance of the parameters they sang with the least consistency, a finding that highlights a need to consider vocal consistency not in isolation, but rather in relation to other performance metrics. This study enhances our understanding of a previously developed example of how ecologically driven changes in morphology may cause divergence in display performance for sexually selected traits, which in theory could drive further divergence among lineages.
Dataset DOI: 10.5061/dryad.qnk98sfvn
Description of the data and file structure
Datasets and associated analysis code for a comparison of fine-scale song structure between two subspecies of swamp sparrow (Melospiza georgiana).
Datasets and variables
multiphrasesongs.csv: Summary sheet indicating syllables and notes within multiphrased swamp sparrow songs.
Variables:
- BirdID
- SongID
- MultiSyll: Essentially a syllableID indicating unique syllable types from each bird. The first syllable in the song is the same ID as the songID, and the second syllable is the songID followed by a 2.
- Syllable.Number: Which syllable in sequence within a song the note is a part of.
- Element.Number: Which note in sequence within a song.
- ElementType: Note type as described by Marler and Pickert 1984.
- Pop: Indicates which subspecies.
NoSyll.csv: Summary sheet with the number of syllables in each song.
Variables:
- Pop: Indicates which subspecies.
- SongID
- NoSyll: Number of syllables in the song.
alldata3.csv: Main dataset containing acoustic metrics for notes from swamp sparrow songs.
Variables:
- MultiSyll: Essentially a syllable ID indicating unique syllable types from each bird.
- ElementID: Which note in sequence within a song.
- SongID
- Element.Number: Which note in sequence within a song.
- Pop: Indicates which subspecies.
- Syllable.Number: Which syllable in sequence within a song the note is a part of.
- Echo.Tail: Dereverberation in ms (a spectrogram setting).
- Start.time: Start time of the note in the original recording file in ms.
- Length: Note duration in ms.
- Gap.after: Duration of the silent gap following the note in ms.
- Peak.frequency.Mean: Mean frequency of maximum intensity across all 5 ms time bins (includes overlapping time frames) in Hz.
- Peak.frequency.Start: Frequency of maximum intensity of first 5 ms of a note in Hz.
- Peak.frequency.End: Frequency of maximum intensity of last 5 ms of a note in Hz.
- Fundamental.frequency.Mean: Mean lowest frequency (first harmonic) across all 5 ms time bins (includes overlapping time frames) in Hz.
- Fundamental.frequency.Start: Lowest frequency (first harmonic) of first 5 ms of a note in Hz.
- Fundamental.frequency.End: Lowest frequency (first harmonic) of last 5 ms of a note in Hz.
- Wiener.entropy.Mean: Mean Wiener entropy across all 5 ms time bins (includes overlapping time frames) in Hz.
- Wiener.entropy.Start: Wiener entropy of first 5 ms of a note in Hz.
- Wiener.entropy.End: Wiener entropy of last 5 ms of a note in Hz.
- ElementType: Note type as described by Marler and Pickert 1984.
- BirdID
Code/software
All analyses were completed in R. Required packages are indicated in the scripts below.
GapBWPerf_pub.R: Code used for quantile regression analysis of gap bandwidth versus gap length for all note types combined. alldata3.csv is necessary as input.
GapLenPref_pub.R: Code used for quantile regression analysis of note length versus gap length by note type. alldata3.csv is necessary as input.
LenBWPerf_pub.R: Code used for quantile regression analysis of note length versus note bandwidth by note type. alldata3.csv is necessary as input.
SubspeciesDQR_pub.R: Code for double quantile regression analysis of all tradeoffs.
SubspeciesAnalysis_pub.R: Code used to perform statistical analysis for comparisons of descriptive stats across subspecies. All csv files in this folder are necessary as input.