How Lyrebirds Mimic: The Strange Vocal Engineering of a One-Bird Soundscape Archive

The superb lyrebird (Menura novaehollandiae) lives in the wet forests of southeastern Australia and is one of the largest passerine birds in the world, about the size of a chicken. Male lyrebirds spend the breeding season standing on display mounds and producing songs that last 15-30 minutes. Roughly 80 percent of the song content is direct mimicry of other species: the calls of more than 20 native bird species in a typical male's repertoire, plus mechanical sounds (camera shutters, chainsaws, dog barks, car alarms, human voices) that have entered the population's vocabulary within the last century. The mimicry is sufficiently accurate that ornithologists with field recording equipment routinely mistake lyrebird recordings for the species being mimicked, and the BBC's 2007 Life of Birds segment showing a lyrebird producing camera-shutter sounds prompted independent verification because the original recording was initially suspected of being a hoax.

The vocal anatomy

The lyrebird's vocal capability rests on an unusually elaborate syrinx, the vocal organ of birds. The syrinx is located at the base of the trachea where it splits into the two bronchi, and most songbirds have a syrinx with two pairs of vibrating membranes (one per bronchial side), each independently controlled by separate sets of muscles. Lyrebirds extend this basic plan with three pairs of intrinsic syringeal muscles (versus the typical four to nine pairs in other songbirds), but with unusual configurations that allow finer modulation of each vibrating element. The two sides of the syrinx can produce different sounds simultaneously, and lyrebirds exploit this to reproduce sounds that involve multiple simultaneous frequencies (the chainsaw mimicry, for example, requires producing two distinct frequencies at the same time, one from each side of the syrinx).

The frequency range of lyrebird vocalizations spans roughly 50 Hz to 6000 Hz, which is wider than most songbirds and overlaps usefully with most of the sounds in their environment. The mechanical-sound mimicry takes advantage of the low-frequency end, which is unusual for a passerine the lyrebird's size. The bird-call mimicry uses the entire range. The accuracy of the mimicry is sufficient that spectrograms of lyrebird-produced calls and species-produced calls are essentially indistinguishable for the better-rehearsed mimicry items.

The neural architecture

The neural systems supporting song learning in songbirds are well-studied in zebra finches and canaries, where the song system consists of a small set of forebrain nuclei (HVC, RA, Area X, LMAN, and others) that handle song template storage, motor production, and learning. Lyrebirds have not been studied at the level of detail the laboratory species have (they do not breed well in captivity, are large and difficult to handle, and have a long generation time), but comparative neuroanatomy shows that the basic song-system nuclei are present in lyrebirds with significant size enlargement relative to body size. The HVC and RA in particular appear to be larger than the body-size scaling for non-mimicking songbirds would predict, consistent with the larger song repertoire and more elaborate motor control.

The 2007 Zann and Dunstan paper in Animal Behaviour documented juvenile lyrebird vocal development across multiple sites in Sherbrooke Forest in Victoria and showed that mimicry items are learned across the first few years of life from adult males the juveniles can hear, with strong population-specific variation in which species get mimicked. A juvenile lyrebird in one valley will learn a different repertoire than a juvenile in the next valley, depending on which sounds the local adult males are producing. The mimicry items show stable population-level traditions over multi-generational time scales, which makes lyrebird mimicry a documented case of cultural transmission of acoustic patterns in a non-human species.

The function puzzle

Why lyrebirds mimic is less well-understood than how they mimic. The straightforward answer is sexual selection: females choose males with larger and more accurate repertoires, so any male with the capacity for accurate mimicry has a reproductive advantage. The evidence for this is largely correlational: males with more elaborate mimicry sequences hold larger display territories and attract more visiting females, but the causation is hard to establish because confounding traits (age, body size, territory quality) covary with mimicry skill.

The more interesting question is why mimicry specifically, rather than just an elaborate species-typical song. Lyrebirds do have a species-typical song that is part of the territorial display, but it makes up only 20-30 percent of the total song output. The remainder is mimicry. One hypothesis is that mimicry signals general cognitive capacity in a way that species-typical song cannot, because mimicry requires both auditory memory and fine motor control across an unusually broad acoustic range. A male that can reproduce 20 species accurately has demonstrated capabilities beyond what a species-typical song could demonstrate.

An alternative hypothesis is that mimicry confuses or deters predators. The mimicked calls include alarm calls of other species, which could prompt predator-avoidance behavior in nearby animals. The evidence for this is weak: lyrebirds mostly mimic during sustained display sequences in safe locations, not during predator encounters. The cognitive-signal hypothesis is currently the best supported, though the field is small enough that strong empirical resolution has not been achieved.

The mechanical-sound expansion

The most famous aspect of lyrebird mimicry is the inclusion of mechanical sounds, which expanded dramatically during the 20th century as European settlement brought new acoustic content into Australian forests. Camera shutters, motor-drive winds, and human voices appear in some populations starting from at least the 1960s. Chainsaw sounds appear in populations near logging operations starting from at least the 1980s. Car alarms, mobile phone ringtones, and other electronic sounds have been documented in populations near urban edges starting from the 2000s.

The cultural-transmission dynamic of this expansion is the load-bearing detail. A single male lyrebird that learns a new sound from its environment will pass the sound to juveniles in its territory, who then carry the sound forward into subsequent generations. The result is geographically-clustered acoustic traditions where one valley has chainsaw sounds and another has camera shutters, depending on what the founding males happened to learn. The traditions can persist for decades after the original sound source has disappeared from the environment, because new juveniles learn from older males who learned from older males who learned from the original sound source.

The 2009 Hollien and Wilson paper documented that BBC-recording-prompted skepticism about lyrebird chainsaw mimicry was substantively unfounded: the recordings are genuine and the lyrebirds were producing the chainsaw-like sounds without obvious nearby chainsaw activity at the time of recording, which is consistent with the cultural-transmission interpretation. Some of the mechanical sounds in the lyrebird vocabulary may persist for generations after the human sources stop producing them.

What lyrebirds do not mimic

The selectivity of lyrebird mimicry is itself informative. They mimic other birds (especially birds with distinctive calls in their natural acoustic environment), they mimic mechanical sounds (especially those with patterns similar to bird calls or with strong rhythmic structure), and they mimic some human voice fragments. They mostly do not mimic continuous broadband sounds like wind or running water (which have no patterned structure), do not mimic the calls of non-vocal animals (frogs, insects), and do not mimic acoustic events that are too brief (single impact sounds without sustained content).

The selectivity is consistent with the mimicry mechanism being tuned to patterned vocal-range sounds, which is what the auditory system and vocal apparatus are optimized for. It is also consistent with the cognitive-signal hypothesis: mimicry items that are easy (broadband noise, single impacts) would not demonstrate the cognitive capabilities that the female is supposedly selecting for, so the selectivity is in part about what counts as a costly signal.

Comparison with other vocal mimics

Lyrebirds are not the only vocal mimics. The mockingbirds (Mimidae) include several species that mimic other birds across a broad range. African gray parrots (Psittacus erithacus) mimic mechanical and human sounds extensively, with the Alex studies from the 1980s through Pepperberg's career establishing that the mimicry is connected to genuine cognitive engagement rather than pure motor reproduction. Bowerbirds (Ptilonorhynchidae) include some mimics, especially the spotted bowerbird. The starling family (Sturnidae) includes several mimics, and the European starling commonly imitates car alarms and phone rings in urban environments.

What distinguishes lyrebirds from this broader set is the combination of accuracy, repertoire size, and the prominent role of mimicry in the species-typical mating display. Mockingbirds mimic accurately but mostly other bird species and mostly in lower volume. Parrots mimic mechanical sounds well but mostly in captivity or in close association with humans. Bowerbirds mimic accurately but with much smaller repertoires. Starlings mimic but mostly as a small fraction of total vocal output. The lyrebird is the only species where the elaborate-and-accurate mimicry of a broad range of sounds is the central feature of the male display, and the only species where the repertoire has expanded substantially to include mechanical sounds during the historical period.

Conservation status

The superb lyrebird is currently listed as Least Concern by the IUCN, though local populations have declined where forest habitat has been fragmented or burned. The 2019-2020 Australian bushfires destroyed substantial habitat in the species' range, and the post-fire recovery has been incomplete. Population estimates are difficult because the birds are cryptic and the forest habitat is hard to survey, but order-of-magnitude estimates are in the hundreds of thousands rather than millions. The Albert's lyrebird (Menura alberti), the second species in the genus, has a much smaller range in northern New South Wales and southern Queensland and is listed as Near Threatened.

The conservation question that is harder to address than the species-population question is the conservation of the population-specific mimicry traditions. The traditions are culturally transmitted across generations and depend on continuous adult-juvenile contact in stable forest territories. Fragmentation of habitat that reduces adult population density also reduces the transmission rate, and traditions can be lost when local populations drop below the threshold for stable transmission. The lyrebird mimicry repertoire is part of what makes the species itself distinctive, and the cultural component is at least as fragile as the genetic component.

Three observations and a deeper one

First, vocal mimicry as a male display has evolved independently in multiple bird lineages but reaches its most elaborate development in a single species in the southeastern Australian forests. The convergent evolution suggests the underlying selection pressure (some form of cognitive-or-motor-display signaling) is broadly applicable, but the lyrebird's full elaboration depends on specific ecological conditions that are not common.

Second, the cultural transmission of mechanical-sound mimicry is one of the most clearly-documented cases of acoustic-cultural-tradition expansion in a non-human species. The geographic clustering of which mechanical sounds appear in which populations, and the multi-decade persistence of sounds after the original source disappeared, are evidence of a cultural rather than purely genetic transmission system.

Third, the standard understanding of bird vocalization as either fixed species-typical calls or as song-learning within species was broadened by the lyrebird tradition long before laboratory work on cultural transmission in songbirds established the same point more rigorously. The lyrebird is an unusually visible case of capabilities that turned out to be more widely distributed in songbirds than the initial narrow framing suggested.

The deeper observation is that the inventory of biological capabilities is consistently larger and more varied than the canonical model-organism-centered framing of biology suggests. Lyrebirds were known to indigenous Australians for tens of thousands of years and to European settlers for over two centuries, but the systematic neurobiological and behavioral characterization of their mimicry has happened mostly within the last 50 years, and many basic questions about mechanism and function remain open. The pattern repeats across the catalog of species capable of unusual sensory or cognitive feats: the capabilities have been there all along, and the conceptual framework for noticing them has typically lagged the observable behavior by decades or longer.