Great tits sang with higher minimum song frequencies in the city than in the nearby forest in all 10 of 10 European city-forest comparisons (Slabbekoorn and den Boer-Visser, Current Biology 2006). That result, together with a 2003 Leiden study before it, anchored two decades of research on how birds adjust to human noise, and the adjustments turned out to be faster and more costly than anyone expected. When San Francisco traffic fell during the spring 2020 shutdown, white-crowned sparrows sang about 35% more softly, while the Lombard effect predicted a drop of only about 3% (Derryberry et al., Science 2020, as reported by Eos). Zebra finches exposed to traffic noise as embryos and nestlings went on to produce 59% fewer offspring as adults (Meillère et al., Science 2024), and at least 29% of Europe’s Natura 2000 protected areas are affected by transport noise at levels that could put terrestrial wildlife at risk (European Environment Agency, Environmental noise in Europe 2025). We aggregated data from Nature, Science, PNAS, the Proceedings of the Royal Society B, Behavioral Ecology, Biology Letters, Environmental Evidence, the European Environment Agency and recent 2025 to 2026 field experiments. For the human side of the same soundscapes, see our urban green space soundscape statistics.
TL;DR
- Great tits sang with higher minimum frequencies in 10 of 10 European city-forest comparisons (Slabbekoorn and den Boer-Visser, Current Biology 2006).
- San Francisco white-crowned sparrows raised their minimum frequency from 2780 Hz to 3077 Hz between 1969/1970 and the 1990s (Luther and Baptista, Proceedings of the Royal Society B 2010).
- Across 102 studies of 101 animal species, noise had a significant overall effect, SMDH 0.57 (95% CI 0.27 to 0.88) (Kunc and Schmidt, Biology Letters 2019).
- Birds are the subject of 64% of studies in the 2025 evidence map of airborne noise impacts on wildlife (Terray et al., Environmental Evidence 2025).
- Zebra finches sang 20.1 dB louder in noise, 95% CI 17.30 to 22.85 (Zollinger, Goller and Brumm, PLoS ONE 2011).
- During the 2020 shutdown, Bay Area urban noise fell by roughly 7 dB and sparrow songs could be heard about twice as far (Derryberry et al., Science 2020, via Eos).
- A phantom road (speakers playing traffic noise in a roadless area) drove away 31% of the bird community (Ware et al., PNAS 2015).
- Phantom wind turbine noise cut Sardinian warbler abundance by 43% (Lehnardt et al., Environmental Science and Pollution Research 2025).
- Noise-exposed zebra finches had telomeres 38 to 46% shorter (Meillère et al., Science 2024, via Science News).
- Female great tits responded less to low-frequency male songs played with noise, P < 0.001 (Halfwerk et al., PNAS 2011).
1. Pitch Shifts: Minimum Song Frequency in Noisy Cities
Urban noise is loudest at low frequencies, so the lowest notes of a song are the first to be masked (drowned out). The most common finding is that urban birds move the floor of their song upward, and in Leiden the average minimum frequency of 32 male great tits ranged from 2.82 to 3.77 kHz, lower in quieter places and higher near noise (Slabbekoorn and Peet, Nature 2003, as reported by Science News). The later 10-city replication showed this was not a local quirk of one Dutch town.
The white-crowned sparrow record from San Francisco adds a time dimension. Over roughly 30 years, the dialect with the higher minimum frequency spread while lower-pitched dialects shrank or disappeared: at Lake Merced the local dialect fell from 93% to 32% of singers while the San Francisco dialect rose from 2% to 34% (Luther and Baptista, Proceedings of the Royal Society B 2010). Pitch shifting, in other words, can operate as cultural evolution, not just as an individual reflex.
| Metric | Value | Source |
|---|---|---|
| Great tit city-forest pairs where city birds had a higher minimum song frequency | 10 of 10 European pairs | Slabbekoorn and den Boer-Visser, Current Biology 2006 |
| Range of average minimum song frequency, 32 male great tits recorded across Leiden | 2.82 to 3.77 kHz | Slabbekoorn and Peet, Nature 2003 (via Science News) |
| White-crowned sparrow minimum frequency, San Francisco dialect, 1969/1970 vs 1990s | 2780 ± 106 Hz vs 3077 ± 48 Hz | Luther and Baptista, Proc. R. Soc. B 2010 |
| White-crowned sparrow minimum frequency, Lake Merced dialect, 1969/1970 vs 1990 | 2469 ± 41 Hz vs 2640 ± 70 Hz | Luther and Baptista, Proc. R. Soc. B 2010 |
| Share of Presidio sparrows singing the local Presidio dialect, 1969/1970 vs 1998 | 64% vs 0% | Luther and Baptista, Proc. R. Soc. B 2010 |
| Species whose minimum song frequency rose with noise (6 species, 28 urban-to-rural sites) | 2 of 6 (5 of 6 trending in the predicted direction) | Dowling, Luther and Marra, Behavioral Ecology 2012 |
| Most-used song element frequency of blackbirds, Vienna Woods forest vs inner-city Vienna | 1.8 to 1.9 kHz vs 2.2 to 2.3 kHz | Nemeth et al., Proc. R. Soc. B 2013 |
| Mean background noise in blackbird territories, inner-city Vienna vs Vienna Woods | LAeq 54 dB vs 45 dB | Nemeth et al., Proc. R. Soc. B 2013 |
Most recent available data: Nemeth et al., Proceedings of the Royal Society B 2013. The landmark pitch-shift field studies date from 2003 to 2013; newer work (section 2) mostly re-analyses or extends them rather than replacing them.
2. What the Meta-Analyses and Evidence Maps Say
Individual studies can be noisy in more than one sense, so the pooled numbers matter. Birds dominate the evidence base: 445 articles, 64% of studies, in the 2025 systematic map of airborne noise impacts on wildlife, against 22% for mammals (Terray et al., Environmental Evidence 2025). That map counts 191 studies pairing birds with communication as the outcome, and the USA alone produced 226 articles (34% of the record) across 59 countries in total.
The pooled effect sizes tell a consistent story: noise changes animal behaviour and physiology across taxa, and birds are not a special case. In the Kunc and Schmidt meta-analysis, phylogeny accounted for only 3.37% of the heterogeneity (variation between effect sizes), meaning the response is spread across the tree of life rather than clustered in a few sensitive families. The earlier Roca et al. meta-analysis of 36 studies narrowed the question to frequency: smaller bird species showed broader shifts than larger ones, and the effect was strongest in the low-frequency part of the song.
| Metric | Value | Source |
|---|---|---|
| Studies, experiments and species in the meta-analysis of noise-related frequency shifts in birds and anurans (frogs and toads) | 36 studies, more than 160 experiments, about 60 species | Roca et al., Behavioral Ecology 2016 |
| Overall effect of anthropogenic noise on animals (SMDH, a standardized mean difference) | 0.57 (95% CI 0.27 to 0.88), p < 0.001; 102 studies, 101 species | Kunc and Schmidt, Biology Letters 2019 |
| Effect of anthropogenic noise on birds only (SMDH) | 0.53 (95% CI 0.37 to 0.68), p < 0.001; 36 studies, 38 species, 149 effect sizes | Kunc and Schmidt, Biology Letters 2019 |
| Share of heterogeneity explained by phylogeny (all animals) | 3.37% | Kunc and Schmidt, Biology Letters 2019 |
| Articles in the 2025 evidence map of airborne noise impacts on wildlife | 863 (665 primary studies, 196 reviews and meta-analyses, 2 modelling papers) | Terray et al., Environmental Evidence 2025 |
| Citations screened in the 2020 to 2023 search update | 13,698 citations, 397 relevant articles | Terray et al., Environmental Evidence 2025 |
| Share of studies on birds vs mammals | 64% vs 22% | Terray et al., Environmental Evidence 2025 |
| Share of studies measuring behaviour vs vocal communication as the outcome | 27% vs 25% | Terray et al., Environmental Evidence 2025 |
Context note: the Kunc and Schmidt introduction states 108 studies and 109 species, while the main model uses 102 studies and 101 species because six studies lacked phylogenetic information. The same evidence-mapping logic applies under water, where ship noise is the dominant source (see our marine vessel propeller cavitation and underwater noise statistics).
3. The Lombard Effect: Singing Louder, Not Just Higher
Raising pitch is only one tool. The Lombard effect, the reflex of raising voice level when background noise rises, was first documented in humans more than a century ago and in a wild bird only in 2004, when Berlin nightingales in noisier territories were found to sing at higher sound levels (Brumm, Journal of Animal Ecology 2004). People do exactly the same thing in a crowded dining room, as our restaurant noise and Lombard effect statistics show.
Zebra finches sang 20.1 dB louder in noise than in quiet (95% CI 17.30 to 22.85), and the respirometry part of the same study did not find a consistent rise in oxygen consumption, which challenges the assumption that loud singing is always metabolically expensive (Zollinger, Goller and Brumm, PLoS ONE 2011). The Vienna blackbird data show why louder and higher go together: higher notes can be produced louder, so the two adjustments stack (Nemeth et al., Proceedings of the Royal Society B 2013).
| Metric | Value | Source |
|---|---|---|
| Song amplitude increase, zebra finches singing in noise vs quiet (birds without a respirometry helmet) | 20.1 dB (95% CI 17.30 to 22.85) | Zollinger, Goller and Brumm, PLoS ONE 2011 |
| Mean song amplitude of one zebra finch, quiet vs loud singing (bird G16) | 79.9 vs 106.1 dB(A) | Zollinger, Goller and Brumm, PLoS ONE 2011 |
| Fastest Lombard response measured in canaries after noise onset | as little as 300 ms | Hardman et al., Journal of Experimental Biology 2017 |
| Nightingales that raised their mean song level in white noise | all subjects (n = 4) | Brumm, Bioacoustics 2002 |
| Blackbird song amplitude, city birds vs forest birds | 2.5 ± 0.82 dB louder in the city | Nemeth et al., Proc. R. Soc. B 2013 |
| Communication distance gain in city noise, frequency shift alone vs frequency shift plus amplitude gain | about 13% vs about 43% | Nemeth et al., Proc. R. Soc. B 2013 |
| Distance blackbird songs carry in city noise, forest-type songs vs city-type songs | 30 ± 2.1 m vs 43 ± 2.8 m | Nemeth et al., Proc. R. Soc. B 2013 |
| Vermilion flycatcher males that stopped singing at noise onset, with traffic noise filtered below 1 kHz vs below 2 kHz (noise that does not overlap the song) | 4 of 21 vs 8 of 21 | Muñoz-Santos et al., Behavioural Processes 2025 |
Outlier note: the 2025 vermilion flycatcher experiment is a reminder that not every response is about being heard. Noise that did not even overlap the song’s frequencies still silenced some males, so masking release is not the only reason birds go quiet.
4. The Anthropause: What Happened When Traffic Stopped
The COVID-19 shutdowns of spring 2020 were an accidental experiment that no ethics board could approve: the noise source was switched off city-wide, then switched back on. Golden Gate Bridge crossings fell from more than 100,000 vehicles a day in recent decades to fewer than 35,000, a level not seen since the 1950s, and urban background noise in San Francisco and Richmond dropped by roughly 7 dB (Eos, reporting Derryberry et al., Science 2020). The sparrows responded within the same spring, singing more softly and with higher-performance songs.
The speed matters more than the direction. Birds that had spent decades singing over traffic dropped their amplitude far below what the Lombard effect alone predicts, which suggests the urban song is an ongoing, reversible adjustment rather than a fixed trait. Elsewhere the picture was mixed: an analysis of 6,369 escape responses from 147 species in five countries found urban birds’ tolerance of people largely unaffected by shutdown-related changes in human presence (Mikula et al., Communications Biology 2024).
| Metric | Value | Source |
|---|---|---|
| Drop in urban background noise, San Francisco and Richmond, spring 2020 vs 2015-2016 recordings | roughly 7 dB | Derryberry et al., Science 2020 (via Eos) |
| How much more softly sparrows sang vs the drop predicted by the Lombard effect | about 35% vs about 3% | Derryberry et al., Science 2020 (via Eos) |
| Distance at which a sparrow could be heard by another bird | roughly doubled | Derryberry et al., Science 2020 (via Eos) |
| Golden Gate Bridge vehicle crossings per day, recent decades vs April 2020 | more than 100,000 vs fewer than 35,000 | Eos 2020 |
| Pre-shutdown loudness of San Francisco sparrow territories vs rural western Marin County | nearly 3 times louder | George Mason University 2020 |
| Bird detectability at sunrise, lockdown vs historical urban checklists, northeastern Spain | on average 27% higher | Gordo et al., Proc. R. Soc. B 2021 |
| Urban bird species with no significant change in occurrence during lockdown | 12 of 16 | Gordo et al., Proc. R. Soc. B 2021 |
| Increase in species recorded on eBird in India’s 20 most populous cities during lockdown | 16% | Madhok and Gulati, Biological Conservation 2022 |
Context note: the Spanish study pooled 126,315 citizen-science records and found detectability peaking at dawn during lockdown, about 2 to 3 hours earlier than in historical urban surveys (Gordo et al. 2021). Higher detectability is partly a human-listener effect: quieter streets make birds easier to hear, even when the birds themselves have not changed. Most recent available data: Mikula et al., Communications Biology 2024.
5. Abundance and Body Condition: The Phantom Road and Phantom Turbines
Correlational studies cannot separate noise from everything else a road brings, such as collisions, edges and lighting. The phantom road experiment solved that by building a road out of sound only: 15 pairs of speakers in a roadless migratory stopover site, switched on and off in 4-day blocks. Thirty-one percent of the bird community avoided the phantom road, and the birds that stayed paid for it in body condition (Ware et al., PNAS 2015). Lab trials with white-crowned sparrows pointed to the mechanism: more time looking up for predators, less time eating.
The design has now been copied for renewable energy. A 2025 phantom wind turbine experiment found that the drop in warbler calls came from fewer birds being present, not from each bird calling less, so noise reshaped the community rather than the song (Lehnardt et al., Environmental Science and Pollution Research 2025).
| Metric | Value | Source |
|---|---|---|
| Share of the bird community that avoided the phantom road | 31% | Ware et al., PNAS 2015 |
| Body condition of birds that stayed during noise playback | fell by a full standard deviation vs the community mean in quiet conditions | Ware et al., PNAS 2015 |
| Noise level at the phantom road, playback on vs off (1-hour Leq) | 48 dB(A) vs 37 dB(A) | Ware et al., PNAS 2015 |
| Lab white-crowned sparrows at 61 dB(A): change in foraging, vigilance and feeding duration | about -8%, +21% and -30% | Ware et al., PNAS 2015 |
| Sardinian warbler abundance during phantom wind turbine noise vs before | -43% | Lehnardt et al., ESPR 2025 |
| Mean number of warbler calls during wind turbine noise vs before | -25% | Lehnardt et al., ESPR 2025 |
| Probability of warbler vocal presence, centre of the noisy area vs its edges | 63% lower | Lehnardt et al., ESPR 2025 |
Data caveat: the phantom road paper reports 5 of 21 well-sampled species with a significant body condition drop in the main text, but other passages of the same paper give 9 of 21 and 7 species. We report only the community-wide result as a headline (see Data watch).
6. Breeding, Development and Fitness
Song shifts are not cost-free if mates prefer the original version. In a Dutch nest-box population, female great tits responded less to low-frequency male songs when noise was added (P < 0.001), but not to high-frequency songs (P = 0.39), and males that were not cuckolded sang lower on average than males that were (Halfwerk et al., PNAS 2011). Noise forces a trade-off between being heard and being attractive.
Developmental effects reach further. Zebra finch embryos and nestlings exposed to traffic noise, with parents kept out of the treatment, grew less and had shorter telomeres (chromosome end caps linked to ageing), and early noise reduced lifetime reproduction (Meillère et al., Science 2024). A 2025 follow-up found that zebra finches reared with highway noise recordings (45 to 55 dB(A) or 55 to 75 dB(A), depending on treatment) tolerated noise as subadults but avoided it again when retested about 2 months later, so tolerance was only transient (Liu et al., Biology Letters 2025).
| Metric | Value | Source |
|---|---|---|
| Female great tit emergence in response to low-frequency songs, with vs without added noise | reduced (P < 0.001); no significant change for high-frequency songs (P = 0.39) | Halfwerk et al., PNAS 2011 |
| Average song frequency, non-cuckolded vs cuckolded great tit males | non-cuckolded males sang lower, F(1,21) = 6.84, P = 0.018 | Halfwerk et al., PNAS 2011 |
| Nest boxes in the gas-well noise study, New Mexico, 2011-2014 | 240 boxes at 12 paired control and noise sites | Kleist et al., PNAS 2018 |
| Noise level up to which nestling feather growth increased before declining sharply | about 70 dB(F) (flat, unweighted level) | Kleist et al., PNAS 2018 |
| Body mass of zebra finches exposed to traffic noise, 12 days after hatching | 14.5% lighter on average | Meillère et al., Science 2024 (via Science News) |
| Telomere length in noise-exposed zebra finches | 38 to 46% shorter on average | Meillère et al., Science 2024 (via Science News) |
| Offspring produced in adulthood by noise-exposed zebra finches | 59% fewer (roughly four fewer birds) | Meillère et al., Science 2024 (via Science News) |
| Nest records used to model light and noise effects on US bird reproduction (NestWatch, 2000-2014) | 58,506 nests, 142 species | Senzaki et al., Nature 2020 |
Context note: in the New Mexico study, chronic noise lowered baseline stress hormone (corticosterone) in adult females and nestlings, and western bluebirds were the only one of three species with a significant negative noise effect on hatching success (Kleist et al., PNAS 2018). The continental NestWatch analysis is described by Hokkaido University, one of the partner institutions.
7. Exposure Context: How Much Habitat Is Noisy
The ecological numbers above only matter at scale if the noise is widespread, and the European data say it is. About 112 million Europeans, over 20% of the population, are exposed to long-term transport noise above EU reporting thresholds, rising to about 150 million (over 30%) against stricter WHO recommendations (European Environment Agency, Environmental noise in Europe 2025). The same report finds at least 29% of Natura 2000 protected areas affected by transport noise at levels that could pose risks to terrestrial wildlife, and notes that EU law covers underwater noise but not noise impacts on terrestrial species.
Progress is slow. Source measures exist (see our porous asphalt quiet pavement statistics for what road surfaces can achieve), but the number of people highly annoyed by transport noise barely moved between 2017 and 2022. Meanwhile, the bird chorus itself is thinning: soundscape analysis at more than 200,000 sites across North America and Europe found a loss of acoustic diversity and intensity over the past 25 years, driven by bird population declines and species turnover (Morrison et al., Nature Communications 2021).
| Metric | Value | Source |
|---|---|---|
| Europeans exposed to long-term transport noise above EU Environmental Noise Directive thresholds | about 112 million (over 20%) | EEA, Environmental noise in Europe 2025 |
| Europeans exposed above stricter WHO recommended levels | about 150 million (over 30%) | EEA, Environmental noise in Europe 2025 |
| People exposed to road traffic noise above thresholds, day-evening-night vs night | 92 million vs 58 million | EEA, Environmental noise in Europe 2025 |
| Natura 2000 protected natural areas affected by transport noise levels that could pose risks to terrestrial wildlife | at least 29% | EEA, Environmental noise in Europe 2025 |
| Change in the number of people highly annoyed by transport noise, 2017 to 2022 | down only about 3% | EEA, Environmental noise in Europe 2025 |
| Share of wildlife-noise studies on transportation noise vs urban noise | 43% vs 24% | Terray et al., Environmental Evidence 2025 |
| Distance from a Brazilian highway within which noise most reduced understory bird responses to playback | about 200 m | da Silva et al., Research Square preprint 2026 |
| Sites analysed for changes in spring bird soundscapes, North America and Europe | more than 200,000 sites, past 25 years | Morrison et al., Nature Communications 2021 |
Context note: the Brazilian Atlantic Forest result is a 2026 preprint that has not yet been peer-reviewed; treat it as an early signal. The EEA exposure figures refer to the 2022 Environmental Noise Directive reporting round.
Summary: Urban Songbird Pitch Shift and Anthropogenic Noise by the Numbers
| Metric | Value | Source |
|---|---|---|
| Great tit city-forest pairs with higher minimum song frequency in the city | 10 of 10 | Slabbekoorn and den Boer-Visser, Current Biology 2006 |
| San Francisco white-crowned sparrow minimum frequency, 1969/1970 vs 1990s | 2780 Hz vs 3077 Hz | Luther and Baptista, Proc. R. Soc. B 2010 |
| Species whose minimum frequency rose with noise | 2 of 6 | Dowling, Luther and Marra, Behavioral Ecology 2012 |
| Blackbird most-used song element frequency, forest vs city | 1.8 to 1.9 kHz vs 2.2 to 2.3 kHz | Nemeth et al., Proc. R. Soc. B 2013 |
| Studies in the bird and anuran frequency-shift meta-analysis | 36 | Roca et al., Behavioral Ecology 2016 |
| Overall effect of noise on animals (SMDH) | 0.57 (95% CI 0.27 to 0.88) | Kunc and Schmidt, Biology Letters 2019 |
| Share of wildlife-noise studies on birds | 64% | Terray et al., Environmental Evidence 2025 |
| Zebra finch song amplitude increase in noise | 20.1 dB | Zollinger, Goller and Brumm, PLoS ONE 2011 |
| Fastest canary Lombard response | as little as 300 ms | Hardman et al., Journal of Experimental Biology 2017 |
| Sparrows singing more softly vs Lombard prediction | about 35% vs about 3% | Derryberry et al., Science 2020 (via Eos) |
| Bird detectability at sunrise during lockdown, northeastern Spain | 27% higher | Gordo et al., Proc. R. Soc. B 2021 |
| Bird community that avoided the phantom road | 31% | Ware et al., PNAS 2015 |
| Sardinian warbler abundance under wind turbine noise | -43% | Lehnardt et al., ESPR 2025 |
| Female great tit response to low songs in noise | reduced, P < 0.001 | Halfwerk et al., PNAS 2011 |
| Offspring of zebra finches exposed to early-life traffic noise | 59% fewer | Meillère et al., Science 2024 (via Science News) |
| Natura 2000 areas affected by potentially harmful transport noise | at least 29% | EEA, Environmental noise in Europe 2025 |
| Noise level up to which nestling feather growth rose before declining sharply | about 70 dB(F) | Kleist et al., PNAS 2018 |
| Noise tolerance of zebra finches reared with traffic noise, retested as adults | not retained (only transient) | Liu et al., Biology Letters 2025 |
Methodology and Sources
- Slabbekoorn H. and Peet M., Birds sing at a higher pitch in urban noise, Nature 2003: PubMed record and Science News coverage.
- Slabbekoorn H. and den Boer-Visser A., Cities change the songs of birds, Current Biology 2006: Leiden University release via ScienceDaily.
- Luther D. and Baptista L., Urban noise and the cultural evolution of bird songs, Proceedings of the Royal Society B 2010: PMC full text.
- Dowling J.L., Luther D.A. and Marra P.P., Comparative effects of urban development and anthropogenic noise on bird songs, Behavioral Ecology 2012 (online 2011).
- Nemeth E. et al., Bird song and anthropogenic noise: vocal constraints may explain why birds sing higher-frequency songs in cities, Proceedings of the Royal Society B 2013: PMC full text.
- Roca I.T. et al., Shifting song frequencies in response to anthropogenic noise: a meta-analysis on birds and anurans, Behavioral Ecology 2016: IDEAS record.
- Kunc H.P. and Schmidt R., The effects of anthropogenic noise on animals: a meta-analysis, Biology Letters 2019: PMC full text.
- Terray L. et al., What is the evidence for the impacts of airborne anthropogenic noise on wildlife? A systematic map update, Environmental Evidence 2025: PMC full text.
- Brumm H., The Lombard effect in nightingales, Bioacoustics 2002; Brumm H., The impact of environmental noise on song amplitude in a territorial bird, Journal of Animal Ecology 2004.
- Zollinger S.A., Goller F. and Brumm H., Metabolic and respiratory costs of increasing song amplitude in zebra finches, PLoS ONE 2011: PMC full text.
- Hardman S.I. et al., Lombard effect onset times reveal the speed of vocal plasticity in a songbird, Journal of Experimental Biology 2017.
- Muñoz-Santos I. et al., Non-masking noise suppresses bird song, Behavioural Processes 2025.
- Derryberry E.P. et al., Singing in a silent spring, Science 2020: Eos report and George Mason University release.
- Gordo O. et al., Rapid behavioural response of urban birds to COVID-19 lockdown, Proceedings of the Royal Society B 2021: PMC full text.
- Madhok R. and Gulati S., Ruling the roost, Biological Conservation 2022; Mikula P. et al., Urban birds’ tolerance towards humans, Communications Biology 2024.
- Ware H.E. et al., A phantom road experiment reveals traffic noise is an invisible source of habitat degradation, PNAS 2015: PMC full text.
- Lehnardt Y. et al., Wind turbine noise pollution reduces songbird vocal presence through changes in abundance, Environmental Science and Pollution Research 2025.
- Halfwerk W. et al., Low-frequency songs lose their potency in noisy urban conditions, PNAS 2011: PMC full text.
- Kleist N.J. et al., Chronic anthropogenic noise disrupts glucocorticoid signaling and has multiple effects on fitness in an avian community, PNAS 2018: PMC full text.
- Meillère A. et al., Pre- and postnatal noise directly impairs avian development, with fitness consequences, Science 2024: Monash University research record.
- Liu Q. et al., Growing up with chronic traffic noise exposure leads to transient but not long-term noise tolerance in a songbird, Biology Letters 2025: PMC full text.
- Senzaki M. et al., Sensory pollutants alter bird phenology and fitness across a continent, Nature 2020: Hokkaido University summary.
- European Environment Agency, Environmental noise in Europe 2025 and its summary for policymakers: EEA report page.
- da Silva J.N. et al., Road traffic noise masks bird vocalizations and disrupts communication in Atlantic Forest understory birds, Research Square preprint 2026; Morrison C.A. et al., Bird population declines and species turnover are changing the acoustic properties of spring soundscapes, Nature Communications 2021.
- Data watch: most of the classic pitch-shift and Lombard studies are older than 3 years (2002 to 2022), so they are flagged as most recent available data; the newest work (2024 to 2026) adds experiments rather than updated population-wide figures. The Science paper by Derryberry et al. was not accessible in full, so its 7 dB, 35% and doubled-distance figures are taken from Eos reporting with the authors’ quotes, and the Meillère et al. body mass, telomere and offspring figures from Science News; Eos also frames the 7 dB drop as about 55%, which is a loose conversion of a logarithmic unit. The phantom road paper is internally inconsistent on how many of 21 species lost body condition (5, 9 or 7 depending on the passage), so only the community-level 31% and one-SD results are used as headlines. The 2025 evidence map’s percentages (for example 64% birds) do not equal simple division by its 665 primary studies, likely because multi-taxon articles are counted more than once. A March 2026 corrigendum to the EEA report revised several of its health-outcome estimates, so we cite only its exposure and Natura 2000 numbers. The Atlantic Forest highway study is an unreviewed preprint.
Last updated: October 24, 2026. We update this roundup quarterly, and the next refresh is expected when the EEA publishes results from the next Environmental Noise Directive reporting round.