Please note that this is an on-going project.
Project Description
There is a growing need for non-invasive, scalable methods to monitor freshwater ecosystems, especially as traditional approaches struggle to detect subtle or unknown biological signals in complex river environments. This study sought to overcome these limitations and develop a new method for exploring freshwater soundscapes to better monitor river health, biodiversity, and ecological change. Researchers deployed waterproof microphones across rivers in South-East Queensland to record underwater soundscapes. They developed a semi-automated tool using R software which combines automated data processing with user-guided steps to streamline analysing the large audio dataset. The semi-automated tool identifies the constant background sound level—such as flowing water—and then isolates all other sound events that rise above it, regardless of their frequency or density within a recording. By calculating acoustic dissimilarity and applying nested unsupervised clustering, it groups similar sound events and distinguishes biological or human-made noises from ambient flow. The tool was tested on 22 streams and correctly identified nearly 90% of distinct sound types, dramatically reducing manual analysis time.
The research enabled scalable, non-invasive monitoring of freshwater ecosystems. It helps detect subtle biological signals, track changes over time, and potentially discover new species. By making the tool free and publicly available, the researchers hope to democratize ecoacoustic exploration and foster broader engagement with river conservation.
Project Personnel and Beneficiaries
The study benefits a wide range of stakeholders committed to environmental monitoring, conservation, and sustainable development. It empowers scientists and researchers by providing a scalable, semi-automated tool to detect and classify underwater sounds, making it easier to monitor biodiversity and ecological change in freshwater systems remotely and at a low cost. This type of remote monitoring can be deployed in areas that have limited cash flow to monitor because it does not require field workers to stay out and collect data manually. Conservation organisations and government agencies gain a non-invasive method to assess river health and inform policy, while local communities and scientists can engage with river ecosystems through accessible, open-source technology.
More broadly, the study supports ecosystem resilience and biodiversity protection, contributing to healthier waterways that benefit agriculture, fisheries, tourism, and cultural heritage.
Outcomes to Date
The semi-automated analysis tool being publicly available enables broader use by conservation groups, the community and environmental agencies. The research has already raised public awareness through media coverage and interactive soundscape tools, helping communities connect with their rivers.
Project Significance
Freshwater ecosystems are under-monitored and vulnerable. Rivers and streams face increasing threats from pollution, climate change, and land-use pressures, yet they’re often overlooked in biodiversity monitoring compared to marine or terrestrial systems. The study introduces a semi-automated tool that can detect and classify underwater sounds which can enable researchers to monitor ecosystem health without disturbing habitats. By identifying sound types that may be missed by traditional methods, the tool helps detect subtle biological signals and track changes over time, even in remote environments.
The study contributes to sustainable development by enhancing freshwater monitoring (SDG 6), protecting aquatic biodiversity (SDGs 14 and 15), and fostering open collaboration (SDG 17). Its publicly available tool empowers researchers, including those working in terrestrial environments, conservationists, and communities to engage with river health in new ways, supporting inclusive, data-driven stewardship of freshwater systems.