Predicting Bird Community Recovery with Acoustic Indices - data and code

Predicting Bird Community Recovery with Acoustic Indices This repository contains the data and scripts for the manuscript: "Acoustic indices predict taxonomic, functional, and phylogenetic recovery of bird communities in tropical forest restoration." Manuscript Summary Quantifying ecosystem restoration success is a priority of the UN Decade on Ecosystem Restoration. In this study, we tested whether acoustic indices can effectively predict bird community recovery in abandoned agricultural areas within a biodiversity hotspot in Ecuador. Using extensive sound recordings from lowland tropical forests, we: Identified 334 bird species through expert annotation and AI-based recognition. Calculated standard acoustic indices from the recordings. Analyzed community composition using Hill numbers and accounting for incomplete sampling. Evaluated predictions of taxonomic, functional and phylogenetic diversity based on acoustic indices. Key Findings - Acoustic indices predicted validated species data with R² values between 0.59 and 0.76.- Taxonomic recovery was best predicted for common and dominant species.- Functional and phylogenetic recovery was best predicted for both rare and common species.- A small set of validated acoustic indices can serve as an efficient tool to monitor large-scale tropical restoration, including recovery of functionally rare bird species.