Case Study 38.2: Acoustic Ecology — Bernie Krause and the Physics of Natural Soundscapes

Forty-Five Years of Listening

Bernie Krause began recording natural soundscapes in 1968. He was already an accomplished musician — he had played guitar for the Weavers, had worked as a session musician and electronic music pioneer (he was among the first to use a Moog synthesizer commercially), and had developed a successful career in music production. But a chance assignment to record wildlife sounds for a film led him into a different kind of listening that consumed the next five decades of his life.

By 2022, Krause's Wild Sanctuary archive contained more than 5,000 hours of recordings from approximately 15,000 distinct locations around the world — a 45-year cross-section of Earth's acoustic environments. The archive is now among the most comprehensive records of what natural soundscapes sounded like in the late twentieth and early twenty-first centuries, and it has become an inadvertent document of their degradation.

Krause estimates that more than half of the locations where he recorded in the 1970s and 1980s have since become acoustically degraded — they produce recordings that are "impaired" by either human noise, the loss of species that previously inhabited them, or both. He can play recordings from the same location taken thirty years apart and demonstrate the change: where once there was a dense, layered acoustic environment with dozens of species occupying distinct spectral niches, there is now a sparser, simpler soundscape with fewer voices.

Niche Differentiation: The Ecological Acoustics of Biodiversity

Krause's most significant theoretical contribution is the concept of bioacoustic niche differentiation — the hypothesis that the species in a healthy ecosystem have acoustically evolved to occupy distinct spectral niches, minimizing interference between their vocal communications.

The evidence for this hypothesis comes from spectrogram analysis of healthy ecosystem recordings. When Krause displays a spectrogram of a healthy temperate rainforest or coral reef recording on a screen, the image is striking: the frequency spectrum from lowest to highest audible frequencies is densely and almost uniformly occupied. Different species occupy different frequency bands: insects dominate the high-frequency range (3-10 kHz); birds occupy middle frequencies (1-8 kHz, with each species at a specific frequency niche); frogs call in the lower middle range (0.5-3 kHz); large mammals vocalize at low frequencies (50-500 Hz). Within each species' frequency band, the temporal patterns of calling (rhythms, timing, duty cycle) further partition the acoustic niche to minimize overlap.

The spectrogram of a healthy soundscape looks, in Krause's description, like a symphony orchestra score — each "instrument" (species) occupies its own range of the stave, and together they fill the available acoustic space efficiently and without crowding. This is not a metaphor: the acoustic organization is real and physically measurable.

Compare this to a degraded soundscape — one that has been affected by logging, agriculture, climate change, or noise pollution. The spectrogram shows gaps: frequency bands that were previously occupied are now empty, because the species that occupied them are gone or have moved. Other bands show crowding, because the remaining species have shifted their frequency or timing to partially compensate. The overall density of acoustic occupation is lower, and the temporal structure is less organized.

The Three Categories of Sound

Krause's framework divides soundscape sounds into three categories:

Geophony encompasses all sounds generated by the physical, non-biological environment: wind, water, rain, thunder, geological activity (including distant earthquakes and volcanic events). These sounds form the acoustic backdrop against which all biological communication occurs. In evolutionary terms, biological vocal systems evolved in the presence of geophonic sounds and have adapted to communicate effectively despite them — bird song, for instance, has evolved in the context of wind noise in the frequency range where wind produces the most energy (below approximately 1 kHz), which is partly why bird song tends to occupy mid-to-high frequencies (above 1 kHz).

Biophony encompasses all sounds generated by living organisms: insect stridulation, frog choruses, bird song, whale vocalization, mammal calls — and human voices, which are biophony by definition even if their acoustic content is speech rather than purely biological signal. In healthy ecosystems, the biophony is the most information-rich layer of the soundscape: it encodes the presence, location, health, sex, and reproductive status of organisms, and the quality of biophonic signal is a direct indicator of ecosystem health.

Anthrophony encompasses all sounds generated by human technology: road traffic, aircraft, industrial machinery, construction equipment, electronic devices, and the electromagnetic interference that various technologies produce. Anthrophony is, in the context of ecosystems, almost entirely maladaptive for non-human species: it adds broadband noise that masks biophonic signals, disrupts the acoustic niche structure that organisms have evolved over millions of years, and can directly damage hearing in species whose acoustic ecology brings them into proximity with high-intensity human noise sources (cetaceans near naval sonar; birds near construction sites).

Climate Change and the Acoustic Record

One of the most striking uses of Krause's archive has been as an acoustic record of ecological change over time. By comparing recordings from the same locations made decades apart, researchers can track changes in species composition, population density, and seasonal timing with an acoustic precision that is often easier to achieve than visual observation.

Krause documented a location in Sugarloaf Ridge State Park in California that he had been recording since 1988. In recordings from the late 1980s and 1990s, the meadow soundscape in spring was dense with the calls of frogs, birds, and insects in a well-organized spectral arrangement. Recordings from the 2000s and 2010s showed a progressive thinning: first the frogs (affected by chytrid fungus, drought, and invasive species), then certain bird species (responding to shifts in insect populations and temperature-driven habitat changes), and finally a simplification of the insect chorus as drought conditions reduced insect populations.

The acoustic change was measurable with scientific precision: the total spectral entropy of the recording (a measure of how uniformly the energy is distributed across the frequency spectrum) declined progressively from the 1990s to the 2010s. A high-entropy, well-filled soundscape became a low-entropy, sparsely occupied one. The physics of information theory provides an objective measure of what "acoustic ecological health" means: a healthy soundscape has high entropy because many distinct species are occupying distinct spectral niches. A degraded soundscape has lower entropy because fewer species are present and the spectral space is less fully occupied.

What Music Could Learn From Nature's Acoustic Design

Krause has explicitly drawn connections between his acoustic ecology research and music composition. He argues that the acoustic design of healthy ecosystems represents millions of years of evolutionary optimization for how sounds can coexist in a shared acoustic space — and that human music composition has arrived at some of the same principles independently.

Spectral niche allocation in orchestration. The orchestrator who assigns high-register melodic material to flutes and violins, midrange harmonic material to violas and horns, and bass foundation to cellos and double basses is applying the same principle as the ecosystem's acoustic niche differentiation: different voices occupy different frequency ranges to maintain clarity and mutual intelligibility. The "rules" of orchestration that theorists codified over centuries are, in part, rules for managing spectral niche allocation in an ensemble.

Temporal niche allocation in counterpoint. Contrapuntal music — Bach's fugues, African polyrhythmic drumming, interlocking hocket patterns in Indonesian gamelan — involves multiple voices that are temporally offset so that each voice has acoustic space in the temporal domain even if they share frequency ranges. This is analogous to the temporal acoustic niche differentiation that some ecosystems exhibit, where species that share a frequency range differentiate their temporal patterns (calling at different times of day, or with different rhythmic duty cycles) to reduce interference.

Dynamic niche allocation in musical form. The quiet introductory passage that precedes a dramatic entrance; the orchestral reduction that gives a soloist acoustic space; the ritardando that carves out time for a melodic climax — these are all forms of dynamic niche allocation, creating acoustic space (in the dynamic dimension) for specific musical elements to be heard clearly against the texture.

Krause summarizes the connection: "Nature's soundscapes are the world's oldest music. They solved the problem of acoustic coexistence long before any human composer thought about it. We could learn a great deal from listening to them."

The Archive as Memorial

There is an elegiac dimension to Krause's work that no amount of scientific framing can fully contain. The Wild Sanctuary archive is not just a scientific dataset — it is a memorial. Many of the ecosystems Krause recorded no longer exist in the form he recorded them. Species that he can hear singing in his 1970s recordings are now locally extinct or endangered. The specific acoustic communities he captured — each the result of millions of years of evolutionary coexistence — are gone.

Krause has described listening to these archival recordings as listening to the voices of the dead. The frogs that sang in that California meadow in 1993 are gone. The birdsong he recorded in a forest in Papua New Guinea in 1979, before logging cleared the area, is now the only evidence that those specific birds inhabited that specific place in that specific configuration. The archive preserves sounds that have no physical manifestation anywhere else in the world.

This is a profound and unusual function for a scientific archive — it preserves not data but presence, the acoustic presence of living beings whose physical existence has ended. In this, the archive functions somewhat like a piece of music preserved in notation or recording: it allows something that no longer exists physically to be heard again. The physics of sound waves — which dissipated into the atmosphere the moment they were produced — are reconstituted by the archive into something listenable.

Cage argued that ambient sound, heard with attention, is music. Krause's archive suggests that the most important ambient sound to hear with attention may be the sound of a world that is losing its voice.

Discussion Questions

Krause's niche differentiation hypothesis proposes that species evolve to occupy distinct acoustic niches to minimize communication interference. Design a study to test whether this hypothesis is correct — what data would you collect, what analysis would you perform, and what result would confirm or disconfirm the hypothesis?
The spectral entropy of a soundscape (how uniformly energy is distributed across frequency) is proposed as a measure of acoustic ecological health. What are the strengths and weaknesses of this measure? Can you think of a scenario in which a high-entropy soundscape would be acoustically unhealthy, or a low-entropy soundscape would be acoustically healthy?
The chapter on acoustic virality (Chapter 37) argued that music's acoustic features are shaped by the distribution environment. Krause's work suggests that the acoustic design of natural soundscapes is shaped by evolutionary "optimization" over millions of years. Compare these two "selection environments" for sound: the streaming algorithm and the evolutionary ecosystem. In what ways are they similar? In what ways are they fundamentally different?
Krause describes listening to archival recordings of now-extinct or degraded soundscapes as "listening to the dead." Does this make the recordings a form of art, a form of history, a form of conservation, or all three? How does the context in which you hear these recordings (a scientific presentation, a concert, a museum installation) change what you hear and how you value it?