Trait mediated effects of anthropogenic noise on bird behavior and fitness
Data files
Dec 22, 2025 version files 429.38 KB
-
bird-acoustic-meta-analysis_DATA_2025-12-22.csv
351.74 KB
-
Bird-acoustic-meta-analysis-code_2025-11-03.R
73.69 KB
-
README.md
3.94 KB
Abstract
Birds are considered especially vulnerable to anthropogenic noise because of their reliance on acoustic information. Single-species research shows that noise can impact different aspects of bird behavior and consequently reduce their fitness. However, we have a limited understanding of how ecological and life-history traits mediate responses to anthropogenic noise across species. We performed a meta-analysis to quantify noise impacts on bird behaviors (communication, cognition, aggression, risk, foraging, and habitat use) and fitness-related responses (growth, physiology, and reproduction), and how bird traits, such as nesting and habitat type, mediated those responses. Using 944 effect sizes from 160 bird species across 6 continents, we found that anthropogenic noise significantly affected various behaviors as well as physiology and has strong negative effects on reproductive responses. We also found that anthropogenic noise had stronger negative effects on bird reproduction for species that nest nearer to the ground, while growth and physiological responses were stronger for species that nested in open rather than cavity nests and those living in deciduous forests, respectively. Our results highlight the characteristics of those birds most vulnerable to noise pollution and inform how conservation actions can best reduce the impacts of human-made noise in those species’ habitats.
Dataset DOI: 10.5061/dryad.rfj6q57q6
Description of the data and file structure
Data was collected through multiple literature searches. One done in 2017 and two co-occurring searches done in 2025. The 2025 searches were done through Web of Science and using Elicit AI.
Files and variables
File: Bird-acoustic-meta-analysis-code_2025-11-03.R
Description:
File: bird-acoustic-meta-analysis_DATA_2025-12-22.csv
Description:
Variables
- studyid: study specific ID
- effectid: individual ID given to each individual response
- species: Latin name of species
- common.x: English common name of species
- citation: lead author of paper from which data was extracted and year it was published
- response_cat: response category determined by authors based on response measured in original study (aggression, cognition, communication, foraging, habitat use, risk, growth, physiology, reproduction)
- response: response measured in original study
- explanatory: source of noise as stated in original study
- sound_type: category of source of noise (urban, military, industrial, other)
- amplitude: amplitude measured in study at which noise was either measured or projected in experimental conditions
- experiment_type: observational vs. experimental
- yi_og: original Hedges’ g that was calculated from our analysis
- vi: variation associated with Hedges’ g
- yi: altered Hedges’ g used for further analysis. The sign of Hedges’ g values were altered for reproduction and growth responses to reflect impact to fitness (i.e. a positive sign represents a positive impact on fitness)
- family: taxonomic family of the given species
- body.size: average body mass of adult individuals (g)
- diet: the dominant diet type (invertebrate, omnivore, plant/seed, fruit/nectar, vertebrate/fish/scavenger)
- diet.breadth: the number of diet categories a species falls into
- clutch.size: average clutch size of species
- foraging.h: dominant height at which species feeds. If species forages at multiple heights equally, both heights are listed (ground, ground/understory, midhigh, midhigh/canopy, canopy)
- nest.type: cavity vs. open nest
- nesting.h: average height of nest for given species (m)
- peak.freq: average frequency of the highest decibel of song for given species (dB)
- song.l: average song length for given species (seconds)
- range: average frequency range of song for given species (dB)
- migratory: whether a species migrates or not.
- habitat: dominant habitat type. (coastal, coniferous forest, deciduous forest, generalist, grassland, mixed forest)
- yi_abs: absolute value of yi_og. Used for analysis of behavioral responses (aggression, cognition, communication, foraging, habitat use, risk) and physiology
- logsize: logarithm of body size
- logsong: logarithm of song length
- JetzEnglish: English common name from Jetz et al. 2012
- JetzSpecies: taxonomic genus and species from Jetz et al. 2012
- log_yi_abs: absolute value of the log of yi_og
Data description:
NA: not available. Any data we were not able to obtain either from the original research from which the effect size was collected or trait data for the species is denoted with NA. Blank cells in column log_yi_abs indicate values whose original effect size was zero and we could therefore not calculate the logarithm for.
Code/software
All analyses were done in R using the following packages: MuMin, lme4, lmerTest, jtools, tidyverse, metafor, ggsignif, ggpubr, magrittr, dplyr, ape, geiger, phytools, corrplt, gridExtra, and grid.
Access information
Other publicly accessible locations of the data:
Data was derived from the following sources:
- n/a
