Acoustic Soundscapes as Early Warning Indicators for Biodiversity Collapse in Restored Savanna Ecosystems

Abstract

The rapid biodiversity loss in savanna habitats endangers ecosystem stability and long-term ecological
restoration success, but detecting ecosystem collapse early is difficult because traditional biodiversity
surveys are expensive and time-consuming. The research measures the biodiversity collapse of
restored savanna as an indicator of acoustic soundscapes. Passive acoustic observations were
performed at 18 restored and reference savanna sites over 24 months, revealing diel and seasonal
variation in biotic sound. Quantitative measurements of acoustic indices, such as the Acoustic
Complexity Index (ACI), Bioacoustic Index (BI), and Acoustic Entropy (H), were compared with
vegetation structure, insect abundance, and vertebrate species richness data collected independently.
Findings show that there was a consistent precedence of deterioration in soundscapes, with a resultant
measurable reduction in biodiversity. The ecological degradation sites recorded a 2238% decrease in
ACI and a 31% decrease in BI up to 69 months before the reporting of substantial losses in species
richness. There were strong positive correlations between the acoustic indices and field-based
biodiversity indices (e.g., ACI vs. species richness: r = 0.74, p < 0.001). Classification models trained
with machine-learning methods on acoustic features accurately identified stable and collapsing sites
with 87 % accuracy. Notably, the restored sites with a greater level of early acoustic diversity had a
higher level of long-term resilience indicating that soundscapes are indicative of both the state of
biodiversity and the functionality of an ecosystem. The paper has shown that acoustic soundscape
analysis is a tool which can be scaled, non-invasive, and cost-effective in identifying the early warning
of biodiversity breakage in restored savanna ecosystems. Acoustic monitoring can be effectively
incorporated into the restoration assessment systems to help achieve better adaptive management and
better conservation outcomes with the growing environmental stresses.