Posted by Ines Geraldine Moran

Birds’ melodious songs, bats’ echolocations, insects’ crackling lisps and shuffles are sounds heard in nature that have fascinated humans for many centuries. Bioacoustics, the science of natural sounds produced by living organisms, is a relatively new field of science that has become central to the study of linguistics, animal behaviour, animal ecology and animal conservation. 

Prior to any technological tools in the field of bioacoustics, scientists described animal sounds using various medium such as music notes, intricate words, or onomatopoeia with letter combinations that attempted to reproduce particular sounds. In order to accurately identify sounds in nature, scientists needed detailed behavioural notes associated to phonetic references. One may imagine how difficult it would have been to walk in a forest and try to detect an animal sound described as Grea-deal for example. For the curious minds, Grea–deal was a phonetic sound that referred to Alder Flycatchers from Massachusetts.

Like with many advances in science, new technologies often play an essential role in making new discoveries. In the mid 20s century, a technological revolution changed how scientists studied animal sounds. In 1950s with the invention of recorders and sound visualization tools, a new era in the field of bioacoustics began. Thanks to these devices, scientists could record and visualize sounds of wild species. A new window in the inner lives of animals opened up to scientists. For the first time, scientists could record and measure complex vocalizations and repertoires, vocal differences between individuals, sound variation throughout seasons or even vocalizations produced during specific breeding stages in wild animals. With these technologies, new horizons opened up in linguistics, animal behaviour, animal ecology and conservation. For example, new sound libraries, like the Macaulay Library, have built up impressive collections of animal sounds from the wild. Playback experiments, in which animal sounds are played back to live animals, became a common technique for wildlife biologists and allowed researchers to answer new questions about animal behaviour. Later, automated recorders, devices left in nature for long periods of time, allowed researchers to record the sounds of habitats known as soundscapes, which in return provided important information about the health of ecosystems. 

Recently, the Cain lab – at the University of Auckland where I am conducting a PhD in bioacoustics- started to use some of the latest technologies available in the field of bioacoustics, to advance our knowledge on the evolution of vocal learning in birds. Research in the Cain Lab investigates the vocal learning abilities of rifleman (a small passerine) in a remote reserve, Boundary Stream Mainland Island, New Zealand. Researchers at the Cain Lab use relatively novel bioacoustics technologies, such as automated recording devices, computer programming and machine learning, to record and analyse bird vocalizations.

The development of new technology in the field of wildlife bioacoustics has changed the way we study the vocal world of wild animals. New technologies in bioacoustics are rapidly advancing, and with them new questions are emerging. Animal vocalizations has fascinated humans for many centuries and will keep doing so for many more centuries. As frogs would say: ribbit ribbit!

R packages:

Recommended resources for the detection and analysis of animal sounds.

Several R packages, in particular warbleR, SeeWave, bioacoustics, and monitor, and software are available to analyse, detect and classify sound. Here are few examples of great R packages and software:​

warbleR : warbleR is R package that combines analytic tools used to measure and detect acoustic signals. Authors: Marcelo Araya-Salas & Grace Smith-Vidaurre (araya-salas@cornell.ed)

Seewave Seewave offers a wide array of tools to analyze animal sounds with R signals. Acoustic template detection and monitoring database interface. Authors: Jerome Sueur et al. (sueur@mnhn.fr)

monitoR monitoR uses acoustic template to detect sounds. Authors: Sasha D. Hafner (sdh11@cornell.edu)

bioacousticsbioacoustics contains tools to transform, detect and classify animal sounds. Authors: Jean Marchal et al. (jean.marchal@wavx.ca) 

Sound autodetection software

Kaleidoscope Kaleidoscope uses sound recognizers to detect animal sounds. This software saves a lot of time when processing numerous and long audio files.

Interactive sound analysis software

Raven– Cornell Lab of Ornithology Raven is a user-friendly platform that allows visualizing of sounds and annotation of animal vocalizations.