Eavesdropping on underwater worlds: the potential of aquatic ecoacoustics
Could listening to the underwater sounds made by freshwater life help us better document and protect aquatic ecosystems? A new special issue of the Freshwater Biology provides intriguing evidence to suggest that it could.
Acoustic monitoring has emerged as a key tool for ecologists and conservationists in recent years. Bioacoustics (the study of sounds produced by or affecting living things) and ecoacoustics (the study of environmental sounds relating to ecosystem processes) continue to grow in popularity as approaches to ecological monitoring.
These approaches centre on the idea of passive acoustic monitoring, or PAM, where researchers place autonomous acoustic sensors (aka microphones) in study sites to capture sound recordings of the environment over time.
The recordings – whether transcribed by human researchers listening back, or by computer algorithms – can then be used to calculate biodiversity metrics such as species abundance, behaviour and phenology. Technological advances increasingly make PAM an affordable, long-term and non-invasive ecological sampling approach for researchers: a ‘listening ear’ on a changing world.
However, the use of such acoustic monitoring techniques has yet to be fully explored or adopted in freshwater systems. The new special issue, edited by Dr Simon Linke, Dr Camille Desjonqueres and Dr Toby Gifford, outlines the opportunities acoustic monitoring offers to freshwater researchers and conservationists, in an effort to raise awareness of its potential.
“Monitoring freshwater ecosystems is time consuming and costly. Using acoustics enables us to observe what is going on 24/7,” says Dr Desjonqueres. “We took over the editorial desk of Freshwater Biology for an issue,” Dr Linke continues. “We invited the biggest names in the field to help us tackle some of the key steps towards operationalising acoustics in the freshwater realm.”
The special issue contains nine studies that investigate underwater acoustics (including the work by Ben Gottesman and colleagues which we covered last year), and three studies on water birds and frogs. Tracing a lineage of describing underwater sound back to Aristotle, the editors identify six key challenges for the widespread uptake of freshwater ecoacoustic monitoring.
1. Characterising sounds and linking them to organisms and ecosystem processes
Four main groups of freshwater organisms are known to produce sounds: amphibians, crustaceans, fish and insects. However, it is rare that researchers can visually identify the source of different sounds in underwater environments. Lowering a hydrophone beneath the water’s surface can be a surprising and disconcerting experience: the listener becomes immersed in the invisible soundworlds created by aquatic life. How might such soundscapes be translated into useful ecological metrics?
The editors highlight the need for more comprehensive catalogues of the sounds of freshwater life, which could offer researchers ‘reference recordings’ to compare to their own studies. In this issue, two studies develop such ‘soundtype references’ in Costa Rica and Northern Australia.
2. Improving automatic sound detection and analysis techniques
In addition to better identifying and cataloguing freshwater sounds, the editors highlight the need to improve how recordings are processed and analysed. Autonomous sound recorders have the potential to generate a lot of data, particularly if multiple recorders are used over an extended period.
Manual transcription of these recordings – whether through listening, or the use of visual spectrograms – is thus time-consuming. As such, automatic sound recognition technologies – which can identify organisms based on their sonic signatures – are needed.
In this special issue, two papers develop the basis of what editor Dr Gifford calls a “Shazam for fish” by documenting the calls of different species of piranhas in Peru, and the spawning calls of ‘love-sick’ burbot in northern Canada. Another study develops an automated detection algorithm for the underwater vocalisations of the spadefoot toad, whilst another uses a deep learning algorithm to acoustically detect the highly-endangered white-bellied heron in Bhutan.
3. Archiving and sharing freshwater acoustic data
As researchers make advances in identifying aquatic life through sound, it is important that their data is archived and shared amongst the global scientific community, the editors state. They write that, “While the Cornell Lab of Ornithology’s Macaulay Library contains some fish sounds (982, which represents 0.25% of all calls), these are mainly marine and from the 60s and 70s.”
Initiatives such as the Freshwater Information Platform are driving forward open-access sharing of datasets, and perhaps there is scope to develop their sound libraries in the future.
4. Understanding acoustic patterns across landscapes
The way that ecosystems and biodiversity vary across landscapes is called spatial heterogeneity by ecologists. Traditional ecological surveys account for spatial heterogeneity in their design, often by replicating study methods in different areas of a landscape.
The editors suggest that ecoacoustic methods have yet to adopt similar approaches. They suggest that this is due to the volume of data generated by ecoacoustic methods and the demands it places on computer analysis systems. In this issue, one paper uses a regular spaced set of hydrophones to show that acoustic activity of aquatic insects (Hempitera sp.) is higher in open water than vegetated areas.
5. Understanding acoustic patterns over time
Ecoacoustic methods offer researchers the potential to monitor ecosystems over long timescales, offering an insight into the diurnal and seasonal patterns of life that occur in them.
Two studies in the issue (here and here) highlight nightly aquatic insect activity patterns. The editors suggest that studies which seek to identify aquatic animals by their calls should focus on such times of day when activity is highest.
6. Making links between sound and ecological health
The end goal of all ecological assessments is to understand the ecological health and condition of a landscape. Whilst there are limits to the scope of this in freshwater environments (e.g. only 20% of fish are soniferous), the editors highlight three useful approaches.
First, changes in aquatic sound can indicate changes to the wider ecological community. Second, ecoacoustics can help us understand the effects of noise on aquatic ecosystems. Third, ecoacoustics could help the automatic detection of invasive species – such as the round goby – reaching an ecosystem.
Clearly, ecoacoustic techniques offer new opportunities for freshwater scientists and conservationists seeking to understand and protect aquatic ecosystems, and the wide-ranging and innovative studies in this special issue highlight their rich potential.
+++
The Freshwater Blog 
Comments are closed.