Most ecological surveys work with what can be seen: counts of animals observed, trees measured, or scat samples collected. A growing complementary tradition works with what can be heard. The discipline of acoustic ecology treats the sound of a landscape as a measurable characteristic, recording and analysing the composition of sound across time and space to track the presence, abundance, and behaviour of species that might be difficult to observe directly. In regions where traditional surveys are expensive or intrusive, and for species that are active at night or in dense vegetation, acoustic monitoring has become increasingly important.

A. The equipment has become remarkably affordable. Autonomous recording units, weatherproof devices the size of a small book that can be left in a forest for weeks at a time, now cost a small fraction of the specialised equipment available a decade ago. A single unit can produce gigabytes of audio per day, and a dozen units deployed across a reserve provide a continuous record of the soundscape over many weeks. The bottleneck has shifted from collection to analysis: a week of recordings from a single unit would take a human listener months to process manually, so automated methods are essential for any but the smallest studies.

B. Pattern recognition software can identify many bird, bat, and frog calls automatically, and the accuracy has improved sharply in recent years with the development of machine-learning classifiers trained on labelled reference recordings. The Spanish bioacoustician Carmen Espinosa has built a reference library of more than four hundred European bird calls, and her team's software can identify individual species with accuracy exceeding ninety percent under most field conditions. The remaining ten percent of errors, which typically involve species with very similar calls or calls at the edge of the recording range, are important enough that expert human review is still required for publication-quality results.

C. Beyond species identification, the composition of the soundscape itself carries information. A healthy tropical forest produces a characteristic pattern of overlapping bird, frog, and insect calls that has been called the dawn chorus. A degraded forest, by contrast, often shows a sparse soundscape with quieter calls and longer silences. Espinosa's group has developed quantitative indices of soundscape diversity and richness that correlate with more conventional biodiversity measures, and has argued that these indices can be used as a rapid proxy for ecosystem health in regions where species-by-species inventory is impractical.

D. The acoustic approach is particularly valuable for nocturnal and secretive species. Owls, nightjars, and many bats are difficult to observe directly but produce distinctive calls. Bats use ultrasonic echolocation calls that are inaudible to humans but can be recorded with specialised microphones and analysed to identify the species. A single night's recording at a well-chosen site can document several species of bat that might not be detected by a month of visual observation. The technique has proved useful in assessing the impacts of wind-farm developments on local bat populations, where the direct risks to bats need to be weighed against the indirect benefits of renewable energy.

E. Cetaceans - whales and dolphins - are another group for which acoustic monitoring has been transformative. Most cetacean species produce calls that can be recorded at long ranges underwater, and systematic hydrophone networks can track populations across entire ocean basins. The analysis of songs from a single population of humpback whales over many years has revealed that the song evolves continuously, with new elements being introduced and old ones retired in a pattern that resembles cultural fashion. The Australian bioacoustician Nathan Fraser has described the humpback song as among the most complex non-human cultural traditions for which long-term records exist.

F. Acoustic ecology has also begun to reveal the effects of human noise on wildlife. In marine environments, shipping noise interferes with the communication ranges of several cetacean species, and the recovery of baseline ocean sound during the reduction in shipping caused by the COVID-19 pandemic provided an accidental experiment whose effects on whale behaviour have been extensively studied. On land, the noise of motor traffic has been shown to change the timing and pitch of bird song in urban populations, as discussed in the dialect literature. Espinosa's work has documented similar, though smaller, effects on rural populations affected by heavy agricultural machinery.

G. The field continues to develop. Better microphones, cheaper storage, more powerful machine-learning classifiers, and the accumulation of reference libraries all make it easier to extract information from recorded sound. Fraser has argued that the next decade will see acoustic monitoring become a routine part of biodiversity assessment, comparable in importance to the visual surveys that have been the standard for more than a century. The role of acoustic data in conservation will probably expand, particularly in remote regions where sending expert human observers is expensive or impractical. The challenge of this expansion, as in other areas of ecology, is the integration of methods rather than their replacement: a well-designed study uses both traditional and acoustic methods, each addressing questions the other cannot answer.