What Is Soundscape Ecology?

Soundscape ecology examines the total acoustic environment—biological (biophony), geophysical (geophony), and human-made (anthrophony)—as an integrated indicator of ecosystem condition. Unlike species-specific bioacoustics, this field treats every sound as part of a dynamic signature that reflects biodiversity, habitat quality, and anthropogenic pressure. The concept was popularized by ecologist Bernie Krause in the 1970s, and technological advances over the past decade—cheap autonomous recording units (ARUs) like the AudioMoth and open-source software such as seewave and PAMGuide—have made landscape-scale deployments feasible. Researchers place ARUs in grids across rewilding sites, capture continuous audio for weeks or months, and compute acoustic indices that distill massive sound files into ecologically meaningful numbers. For a foundational treatment of the field, see “Acoustic indices for biodiversity assessment: moving from single-call to landscape-level monitoring” (Sueur et al., 2014).

Key Acoustic Indices Used in Urban Studies

  • Acoustic Complexity Index (ACI): Measures temporal variation in sound intensity across frequency bins. High ACI values often correlate with high species richness because vocalizations are irregular and diverse.
  • Acoustic Diversity Index (ADI): Applies Shannon entropy to the spectrogram’s frequency distribution. Higher ADI indicates that acoustic activity is spread evenly among frequency bands, typical of a balanced soundscape.
  • Normalized Difference Soundscape Index (NDSI): Computes the ratio of biophony to anthrophony. Values near +1 signify a pristine natural soundscape; negative values denote heavy urban noise dominance.
  • Bioacoustic Index (BI): The integrated area under the frequency spectrum curve, used as a proxy for bird abundance and activity, especially during dawn choruses.
  • Acoustic Evenness Index (AEI): The inverse of ADI; low AEI means sound energy is concentrated in a few frequency bands, often a sign of disturbance or low diversity.

These indices are not perfect in isolation, but when combined they offer a robust statistical description of the soundscape’s ecological quality. Researchers frequently validate indices against traditional survey data—point counts for birds, pitfall traps for insects—to confirm correlations.

Applying Soundscape Ecology to Urban Rewilding

Urban rewilding projects operate in challenging contexts: small fragmented patches, high human disturbance, and infrastructure constraints. Traditional biodiversity metrics (e.g., species richness, abundance) are often impractical to collect at the necessary temporal resolution. Soundscape ecology fills this gap by providing continuous, non-invasive, and scalable monitoring. For example, the rewilding of a 2-hectare urban wetland in Philadelphia recorded a 52% increase in the Bioacoustic Index within three years of planting native shrubs and removing invasive vines. The soundscape data also showed a shift from dawn-only bird activity to extended morning and late-afternoon singing periods, indicating improved foraging habitat and lower perceived predation risk.

A rigorous study on the topic is “Assessing urban rewilding success using soundscape metrics in New York City parks” (Frontiers in Ecology and Evolution, 2021), which demonstrated that acoustic diversity was significantly higher in rewilded parks with complex understory vegetation compared to manicured lawns.

Methodologies for Urban Soundscape Monitoring

  1. Site selection and stratification: Pair rewilded sites with matched control sites (e.g., similar size, adjacent land use, and background noise levels). Include multiple replicates to account for microclimates and edge effects.
  2. Deployment protocol: Place ARUs (e.g., AudioMoth, Swift One) at 1.5 m height on trees or poles, spaced at least 50 m apart to avoid spatial autocorrelation. Use a schedule of 1 minute recording every 10–15 minutes for 24/7 coverage, a common compromise between data volume and battery life.
  3. Data processing and index calculation: Use R packages seewave, soundecology, or PAMGuide to compute ACI, ADI, NDSI, BI, and AEI. Apply band-pass filters (e.g., 2–8 kHz) to reduce low-frequency traffic noise when analyzing bird song.
  4. Machine learning for species identification: Tools like BirdNET (for birds) and ARBIMON (for amphibians and insects) can classify vocalizations to species level, though accuracy depends on training data completeness for local species.
  5. Human perception component: Conduct listening walks or crowd-sourced acoustic preference surveys alongside recorder deployment. Data on perceived naturalness, annoyance, and aesthetic value complement acoustic indices and inform management decisions.

Benefits of Using Soundscape Ecology

Soundscape monitoring outperforms traditional methods in several dimensions. It is non-invasive—no traps, nets, or disturbance to sensitive species. It captures temporal patterns that human observers cannot, including nocturnal activity (owls, bats, katydids) and responses to transient events (storms, holiday traffic). The data are inherently digital and reproducible, supporting long-term comparisons and meta-analyses. For example, the U.S. National Park Service’s Natural Sounds and Night Skies Division uses ARUs across urban parks to benchmark soundscape quality. Their data show that even tiny rewilded plots—0.3 hectares—can exhibit higher biophony than adjacent untended lots, providing evidence that small-scale interventions matter.

Public engagement is another major benefit. Live-streaming a rewilded soundscape in a park kiosk or via a website invites residents to hear the return of crickets and warblers. This sensory connection fosters stewardship and reduces conflict over management decisions. A study in Ottawa found that park visitors rated the acoustic environment of rewilded areas as significantly more “restorative” than traditional mowed lawns, directly linking ecological health to human well-being.

Challenges and Limitations

Urban soundscape monitoring faces serious methodological hurdles. The most obvious is noise pollution: traffic, construction, and social activity frequently mask biophony, especially at low frequencies below 2 kHz. Bird songs with dominant frequencies in this range (e.g., mourning doves, some thrushes) become undetectable. High-pass filtering and focus on ultrasonic bands (e.g., for bats) can help, but signal-to-noise ratio remains a persistent issue.

Weather introduces further variability. Rain, wind, and leaf rustling add acoustical clutter that can mislead indices—ACI, for instance, spikes during rain because of increased temporal variation. Researchers must either exclude rainfall periods using automated weather data or include equal amounts of dry hours across comparisons. Similarly, phenological shifts (migration, leaf emergence) cause dramatic seasonal changes; short recording windows produce unreliable conclusions.

Analytical blind spots also exist. Low-amplitude species (small insects, soft-voiced birds) may not register above the noise floor. Overlapping calls from multiple species can saturate the recording and confuse both indices and machine learning classifiers. Current classifiers trained on clean recordings often fail when applied to urban field conditions with reverberation, traffic hum, and unusual species mixes. Inter-study comparability remains low because index values depend on recorder gain, microphone sensitivity, and software parameters—standardization is urgently needed.

Finally, soundscapes capture only one sensory channel. The smell of freshly turned soil, the sight of wildflower blooms, and the tactile experience of rough bark are essential parts of rewilding’s impact. A complete assessment must integrate soundscape metrics with soil health indicators, vegetation surveys, and camera trap data. For a discussion of these limitations, see “The promise and pitfalls of soundscape ecology as a tool for conservation and management” (Pijanowski et al., 2019).

Case Studies: Listening to Rewilded Cities

London’s Wild West End

Starting in 2019, the London Wildlife Trust and researchers from Imperial College deployed AudioMoths in seven pocket parks and green roofs converted from abandoned office buildings. After five years, the rewilded sites exhibited a 40% increase in the Bioacoustic Index during spring dawn choruses. Notably, the Common Swift—a species that declined 60% in London over the past two decades—was detected in three of the rewilded sites, whereas it was absent from all control plots. Playback experiments with local residents showed that perceived acoustic naturalness was 2.5 times higher in rewilded than in manicured parks, and that participants reported feeling more “restored” after listening to rewilded recordings.

Portland’s Pollinator Corridors

Portland, Oregon, installed strips of native wildflowers along street medians, bioswales, and parking lot edges as part of its “Pollinator Connectivity Plan.” Soundscape monitoring using ARUs placed every 100 m along the corridors revealed that acoustic complexity (ACI) was 25% higher at corridor nodes than in isolated habitat patches of similar area. Moreover, NDSI values along corridors showed a 0.15-point improvement over three years, indicating that native plantings gradually reduced the perceptual dominance of traffic noise. This quantitative evidence helped the city secure funding to expand the corridor network by 12 km in 2023.

Singapore’s Vertical Rewilding

Singapore’s “City in a Garden” initiative integrates green walls, sky gardens, and rooftop forests. Acoustic recorders affixed to building facades at three heights (ground, 3rd floor, 10th floor) documented the arrival of urban-adapted songbirds (e.g., Olive-backed Sunbird, Asian Glossy Starling) within two years of installation. The NDSI shifted from -0.25 (heavy construction noise) to +0.18 on the 3rd floor, while ground-level recordings remained mostly anthrophony-dominated. This pioneering work proves that soundscape ecology can effectively monitor three-dimensional rewilding in dense urban cores.

Future Directions

The next frontier is real-time, edge-based computing. New ARUs with onboard microcontrollers (e.g., Raspberry Pi Pico-based recorders) can run lightweight models like BirdNET Nano, classifying calls instantly and transmitting summaries via LoRaWAN. This eliminates the need for massive storage and enables adaptive monitoring—for instance, intensifying recording after a rain shower or when rare species are detected. National-scale initiatives such as the Swiss “Ecoacoustic Observatory” and Australia’s “Acoustic Observatory” are already building RAM-standardized datasets that will power continent-wide urban rewilding comparisons.

Citizen science platforms will also expand. The app BirdNET (developed by the Cornell Lab of Ornithology) already allows anyone to record and identify bird calls. Integrating such grassroots data with formal ARU networks can dramatically increase spatial coverage. Policies like the European Commission’s Biodiversity Strategy for 2030 now include “soundscape quality” as a candidate indicator for urban green infrastructure. As costs fall and methods become more standardized, soundscape ecology will likely join soil testing and vegetation surveys as a routine component of rewilding assessment.

Conclusion

Urban rewilding is not merely a cosmetic landscaping exercise—it aims to restore self-sustaining ecosystems where humans and wildlife coexist. Soundscape ecology offers a uniquely scalable, non-invasive, and publicly engaging way to track that restoration. By listening to the acoustic fabric of a rewilded site, we can measure increases in bird song, insect stridulations, and the relative quieting of anthropogenic noise. These acoustic signatures, when paired with ground-truth ecological data and human perception studies, provide compelling evidence that rewilding works. The hum of bees, the trill of a warbler, the soft rustle of native grasses—these are the sounds of a city healing, and soundscape ecology gives us the tools to hear them clearly.