Data from: Frequency masking drives species-specific temporal avoidance strategies in boreal songbirds
Data files
Sep 24, 2025 version files 1.01 MB
-
dataset.csv
1.01 MB
-
README.md
3.24 KB
Dec 03, 2025 version files 1.01 MB
-
dataset.csv
1.01 MB
-
README.md
3.78 KB
Abstract
Interference in acoustic signal transmission can impair communication, and many vocalizing species use various strategies to avoid signal masking. Past studies have focused primarily on the effect of anthropogenic noise and overlap in sound frequency range, leaving a gap in our understanding of how an individual’s vocal signal structure affects other animals in its acoustic community. Using playback experiments we tested whether five species of European boreal songbirds adjust their singing behavior to avoid overlap with three novel acoustic intruders from sub-Saharan Africa, each with a different song structure customized to each study species: (i) continuous songs of narrow, partially-masking frequency based on the study species’ peak frequency, (ii) non-continuous songs of broad, fully-masking frequency and (iii) continuous songs of broad, fully-masking frequency. All species showed evidence of temporal song avoidance only with competitors whose songs were broad-spectrum and completely overlapped in frequency range. The species differed in their behavioral response, either modifying the song rate, song duration, or the timing of their songs. These differences in behavior may be related to limitations imposed by the species-specific information encoded in different song parameters, and allow for predicting the consequences of changes in acoustic community structure.
Dataset DOI: 10.5061/dryad.rn8pk0pqs
Corresponding author: Agata Staniewicz, email: agata.staniewicz@gmail.com
Description of the data and file structure
Data were collected between 22 May and 7 June 2023 in Ekopark Käringberget, northern Sweden. We focused on five forest songbird species: European pied flycatcher, common chaffinch, common chiffchaff, willow warbler, and goldcrest, locating 60 individual males (12 per species) in their singing territories. Each bird was exposed to controlled playback experiment consisting of three non-native bird songs selected for frequency overlap: scaly-breasted illadopsis (Illadopsis albipectus), green-backed camaroptera (Camaroptera brachyura), and northern double-collared sunbird (Cinnyris reichenowi), and a silence control. Playbacks were broadcast in randomized 10-minute sessions, each composed of 5-second alternating intervals of song and silence, from two speakers placed near the bird’s singing location. We observed the birds to ensure they were alone in their territories and singing actively before playback began.
Acoustic responses were recorded using shotgun and handheld microphones, along with autonomous recorders. All recordings were synchronized into three-channel .wav files and manually scanned using Raven Pro software. We used the selection table tool to mark every bird song, noted wheter it began during a playback song, and measured the song duration, noting whether it overlapped (>50%) with a playback song. This threshold was chosen to identify cases where most of the bird's song—and the information it carries—would be acoustically masked. We analyzed a total of 9,361 bird songs.
Files and variables
File: dataset.csv
Description: The file contains data on study species' song timing, duration and overlap during the experimental treatments.
Variables
- Species: experimental study species; there are five: chaffich = Fringilla coelebs, chiffchaff = Phylloscopus collybita, goldcrest = Regulus regulus, flycather = Ficedula hypoleuca, willow_warbler = Phylloscopus trochilus
- Observers: observer initials (authors)
- Filename: unique experiment recording file name
- Playback: species-specific playback combination ID
- Date: date of the experiment (DD.MM.YYYY)
- Time: time at the start of the experiment (HH:MM:SS)
- GPS: unique assigned location ID point of the experiment
- BeginTime: start time (in seconds) of song, measured as number of seconds passed from the start of the experiment
- EndTime: end time (in seconds) of the song, measured as number of seconds passed from the start of the experiment
- Begin: binomial values of whether the song began during the playback song (1) or silence (0)
- Overlap: binomial values of whether the song overlapped by >50% with the playback (1) or <50% overlap (0)
- treatmentOrder: order of treatment within the experiment
- treatmentType: type of acoustic intruder treatment; the three main treatments are I = illadopsis (continuous narrow frequency song, partially masking), C = camaroptera (non-continuous broad frequency song, partially masking), R = sunbird (continuous broad frequency sog, fully masking); additionally K = control (silence), pre = songs recorder before the start of the experiment
- gpsCoordinates: latitude and longitude coordinates (DMS) of the experiment
Data analyzed using R v4.2.1 with glmmTMB and emmeans v1.8.7 packages.
Changelog:
December 2, 2025: Minor changes to README to add clarity on variable units, R packages, and corresponding author information.