The World Ecoacoustics Congress (WEC) is a global platform that brings together scientists, researchers and many other stakeholders who are using ecological sound to understand the dynamics of landscapes. This multidisciplinary congress presents attendees with an opportunity to participate in presentations and debates about the advances and challenges in the world of acoustic monitoring, which includes the latest developments in machine learning and artificial intelligence combined with acoustics and conservation biology. 

As a PhD student merging the goals of the Expanded Freshwater and Terrestrial Environmental Observation Network (EFTEON) with new approaches in passive acoustic monitoring, I had the privilege to attend the biannual WEC in Spain this past July.

Objectives

EFTEON is working towards implementing long-term acoustic surveillance into its environmental monitoring platforms, so this congress was important to create a platform for future collaboration with international long-term monitoring infrastructures, and to understand the advances in this rapidly growing field of study.

I presented a poster showcasing the deployment plan and objectives of EFTEON’s acoustic monitoring project. The key objective was to receive feedback from leading experts to better implement the long-term goals of our research infrastructure. As a result of this congress, I am confident that our plan is on the right track, and that we are working at the leading edge of long-term environmental monitoring.

Deployment feedback  

Most of the researchers in this field use distinct deployment techniques for different landscapes, and there is still a lack of standard methods for specific geographical areas and landscapes.

Oliver Metcalf, the author of “Guidelines to Acoustic Monitoring in the UK”, said he was comfortable with EFTEON modifying his guidelines to meet our needs and answer our research questions.

To use acoustic indices, which are known as proxies for species richness, the advice from delegates was to use a one-minute recording to generate representative indices of the landscape. However, there was a debate regarding the practicality of using acoustic indices in acoustic monitoring. This reflects the novelty of this approach, which is not as well-established as other indices linked to environmental or habitat quality.

New analysis tools

One of the greatest perks of being at this meeting of minds was that I learned first-hand about the pros and cons of a variety of software applications used in the analysis of large acoustic datasets, such as Kaleidoscope and Queensland University Technology (QUT) software. Some of these known applications have their limitations, which are mostly attributed to their confusing interface and constant need to troubleshoot.

Cornell lab researchers have developed BirdNet, which is user-friendly and can be easily trained to identify different soniferous species in various geographic locations. Dr Llarisa Sugai conducted a workshop where she introduced us to BirdNet and various ways that can be used to annotate and train machine-learning models for automated species in BirdNet.

This tool seems to be popular in Europe and has a high confidence level, but more classifier training is required for birds of other continents. This is where researchers and practitioners in Africa can contribute significantly to the growth of an internationally recognised tool and to the global movement of making science more accessible and affordable.

Collaborations

There were many opportunities for collaboration with other research infrastructures such as Birds of Norway, the Okinawa Acoustic Monitoring Program and the Australian Acoustic Observation. I had a chat with most of the people from these infrastructures and they were keen to collaborate.

Professor Lin Schwarzkopf from Australia was mostly interested in comparing and improving her methods using our proof of concept. Sethi Sarab is responsible for Birds of Norway, which is monitoring birds of Norway at a larger scale with real-time automated identification of species. He was willing to chip in for future work, once we are ready to expand to other SAEON landscapes.

We have an existing collaboration with Nicholas Friedman, who is responsible for the acoustic monitoring of Okinawa birds, and he was quite excited for our upcoming deployment in the Drakensberg. Dr Friedman is my PhD co-supervisor, bringing many international collaborators and sound ecology experience to the team.

One of the leading experts in acoustic indices, Tom Bradfer-Lawrence, is interested in collaborating with us to help characterise our landscape using indices.

Overall, this congress yielded many potential collaborators and integrated EFTEON into the world acoustic network.