Using Bioacoustics to Unlock the Ecological Information in Soundscapes

If you have ever gone bird watching under a dense canopy, you've probably strained your ears more than your eyes to find a specific bird. Several species are cryptic and rarely seen or caught on camera, and the best way to find them is by their calls or songs. But, relying on calls and songs in a natural environment is more than just assessing the presence of a species. The most melodious songs and noisy cackles of birds, the stridulations of the insects, and even the swoosh of wind hold valuable ecological information. This valuable ecological information can be accessed by lending an ear to entire ecosystems.

Bioacoustics, or the sounds of nature, in their entirety, make up a ‘soundscape'. There are several thousands of species which vocalise for different reasons, such as to look for a mate or to warn other competitors vying for resources. All of these species, calling and singing to different tunes, together, contribute to the soundscape. An emerging field with the aim of unlocking this ecological information in soundscapes, soundscape ecology is the study of a group of vocalising species to understand the relationship between the species and to their habitat. So, how does one tune into nature to understand the relation between species and their habitat?

Two types of recorders I use to collect my data. The first and larger one is a SWIFT recorder by the Cornell Lab of Bioacoustics and the second, smaller recorder is by AudioMoth.
Photo Credit: Sarika Khanwilkar

As part of my doctoral research, I study the impact of forest restoration on birds and insects in and around Kanha National Park by studying their vocalisations. I compare certain ecological indicators of vocalising fauna across a range of study sites. These sites include species diverse forests, forests that are currently being resorted to unrestored forests and vast patches of Lantana camara, an invasive but ubiquitous shrub in India associated with forest degradation. I study not only the differences in the community of species across these sites, but also the possible difference in their vocalising behaviour as a group.

This research is timely for several reasons. First, large charismatic carnivores, especially big cats, receive the bulk of the attention and conservation efforts. For example, in recent times, insects have experienced sharp declines around the world but with fairly insufficient attention from the research and conservation community. Second, forest degradation is cryptic unlike deforestation, and understanding what a slow change in a forest does to species is necessary to design and enhance conservation programs. Lastly, the use of remote sensing technology for research in today's world is critical. Bioacoustics provides a means to broadly and rapidly study an ecosystem. The technology often is a first pass at monitoring an ecosystem, to be followed by more human- intensive surveys and research efforts where necessary.

In order to capture these vocalisations to study them, I climb up trees, and sometimes creep up to the ends of branches to attach my acoustic recorders. These sturdy recorders, braving wind and rain, and the often brutal attacks by curious palm squirrels and garden lizards, capture all vocalising species that call and sing between 0 to 24,000 hertz. As always, using technology in harsh conditions outdoors with limited resources requires one to be creative to protect and maintain equipment. I use a wooden covering for the recorder, an idea that a collaborator's brother, Arjun Ramesh, came up with in the early days of my research. While the AudioMoth recorders are more delicate than the SWIFT recorders as they lack an outer metal or plastic case, they have been easier to work with given their weight and compact size. Moreover, a zip-lock bag suffices in protecting the AudioMoth recorders from harsh weather conditions.

Tying a recorder to a tree and using a protective wooden covering.
Photo Credit: Sarika Khanwilkar

Remote sensing technologies such as the acoustic recorders I use allow for scientists to passively collect data with minimal human disturbance in a habitat. It is akin to having a peep hole through which one can observe, or in my case listen to, species without letting our presence interfere with their behaviour. Leaving the recorders out for seven days at a time, letting it record continuously through that period, a surprise always awaits when I finally retrieve the recorders. I once chanced upon the alarm calls of a group of hanuman langurs and barking deers because of the drama unfolding around them. At some sites, I hear a village at a distance waking up while the birds, like clockwork, begin their dawn chorus. The acoustic data is a source of joy as much as it is a scientific endeavour. I often find myself imagining and trying to piece together the events taking place at the time the acoustic data was collected.

Spectrograms of a restored (L) and unrestored (R) forest at the edge of Kanha National Park.
Photo Credit: Pooja Choksi

The data I collect is, ironically, analysed using a visual aid — a spectrogram. The spectrogram visualises every species vocalising within the acoustic radius of the recorder. A limited range of frequencies is audible to humans. So, a spectrogram allows us to visualise the vocalisations of all species, even those at extremely high and low frequencies, often inaudible to us. The spectrogram also acts as an image to be analysed using Artificial Intelligence (AI).

The biggest advantage of using remote sensing technologies for research is the ability to gather vast amounts of data. However, to sift through terabytes of acoustic data I collect, AI is proving to be the most efficient way forward. After manually analysing a small proportion of my data, I am able to train AI to look for the presence of a particular species or a particular type of call in the data. Such AI abilities and methods provide the tools to undertake large-scale research necessary to understand patterns, if any, in the broader impact of humans on nature and wildlife. As novel technologies continue to be developed, the future of research and conservation looks exciting and full of potential to uncover another dimension of an ecosystem and address pressing challenges in the field.

Pooja Choksi is currently a Ph.D. candidate at Columbia University in New York and co-founder of Project Dhvani, a long term acoustic monitoring project in India. She studies the impact of forest regeneration and restoration on fauna in central India.

This series is an initiative by the Nature Conservation Foundation, under their programme Nature Communication to encourage nature content in all Indian languages. If you're interested in writing on nature and birds, please fill up this form.

How to find the best deals during online sales? We discussed this on Orbital, our weekly technology podcast, which you can subscribe to via Apple Podcasts, Google Podcasts, or RSS, download the episode, or just hit the play button below.