Do Acoustic Indices Correlate with BirdNET Call Diversity Indices for Biodiversity Assessments? An Application to the Sundarbans Mangrove Forest Ecosystem

<p data-end="777" data-start="449">This dataset contains BirdNET-generated detections from passive acoustic monitoring conducted in the Sundarbans mangrove forest, the largest of its kind in the world. Acoustic sensors were deployed for 288 days across three ecological zones characterized by differing salinity levels (hypersaline, mesosaline, and hyposaline).</p>

<p data-end="1030" data-start="782">The BirdNET output includes species identifications, detection time intervals (start/end in seconds), and confidence scores for each avian vocalization detected. Each detection record is associated with its corresponding audio file and sensor ID.</p>

<p data-end="1360" data-start="1035">These data were used to calculate vocal species richness and to derive Vocal Shannon and Simpson Diversity Indices for temporal and spatial biodiversity assessments. The dataset supports ongoing research on the integration of automated species detection tools for large-scale biodiversity monitoring in mangrove ecosystems.</p>

<p data-end="1501" data-start="1365">Data are suitable for reuse in ecoacoustics research, avian monitoring studies, or validation of AI-based biodiversity assessment tools.</p>