On the spatial and resilience-related features of natural soundscapes
File(s)
Author(s)
Le Penru, Neel P.
Type
Thesis
Abstract
Ecosystems worldwide are under unprecedented, intensifying pressure from human activity and climate change, with profound implications for nature and society. To understand and mitigate the ecological impacts of these global crises, we need accurate, scalable, responsive tools to monitor ecosystem health and biodiversity. Ecoacoustic monitoring is a powerful and versatile method to this end, owing to the information-rich nature of soundscapes (comprising sounds from animals, humans, other organisms and environmental noise) and cheaper, more capable autonomous recording hardware and analysis software. In this thesis, I develop and apply novel technologies for acoustically monitoring some of the world’s most biodiverse and vulnerable ecosystems, with a focus on two twin strands: leveraging the spatial distribution of sounds for enhanced biodiversity assessment with multichannel recording and analysis, and exploring whether acoustic monitoring can detect early warning signals of declining ecosystem resilience (capacity to recover from perturbation) towards tipping points (nonlinear shifts to alternative, often undesirable states triggered by small changes or perturbations). I first investigate ambisonic soundscape reconstruction as a lifelike laboratory test environment for spatial ecoacoustic technologies. I evaluate the effectiveness of this method by using it to test the Multichannel Acoustic Autonomous Recording Unit (MAARU), a custom, six-microphone recorder. I then refine MAARU’s design, co-supervise its deployment in Brazil, and deploy it in the UK, USA and Malaysia to investigate its potential to improve avian call classification and distinguish neighbouring habitats via beamforming, which amplifies sounds from target directions. Finally, I explore satellite and acoustic data from Malaysia and Brazil for early warning signals of rainforest to savannah tipping points. I close by considering how these technologies advance our ability to understand and intervene in rapid ecological change, and relate the overall work to the vital need for ecosystem, individual and societal resilience – Earth's future depends upon it.
Version
Open Access
Date Issued
2025-10-04
Date Awarded
2026-06-01
Copyright Statement
Attribution-NonCommercial 4.0 International Licence (CC BY-NC)
License URL
Advisor
Picinali, Lorenzo
Ewers, Robert M.
Sethi, Sarab S.
Sponsor
Natural Environment Research Council
Grant Number
NE/S007415/1
Publisher Department
Dyson School of Design Engineering
Department of Life Sciences
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)