There is a growing interest in using artificial intelligence for passive acoustic detection and monitoring of avian biodiversity due to its potential to provide cost-effective and non-invasive methods for collecting large amounts of data. Traditional survey methods provide essential data that informs the diversity of many songbird communities, however, they are often costly and cumbersome compared to passive acoustic monitoring methods. Our study characterizes the soundscape of ephemeral wetlands to examine whether automated methods are redundant or complementary to human detection methods. We compare community diversity and composition of songbirds using human vs. automated detection. The soundscape is characterized using solo system audio recorders and the BirdNET artificial intelligence algorithm to identify species occurrences. We sampled 15 isolated forest ephemeral wetlands for 3 days each between May 15 - June 15, 2022. We compare the automated species detections to traditional point-count surveys to inform future monitoring schemes.
Results/Conclusions
The preliminary results suggest that passive acoustic monitoring is complementary to the point-count method because both methods capture species not detected by the other. In particular, the median species occupancy detected between methods was significantly different for several species and the point-count method detected higher species richness. Our results highlight the advanced capabilities and touch on the drawbacks of using passive acoustic detection in place of point-count methods. We conclude that researchers can use automated methods in a complementary manner to traditional monitoring methods. Efforts towards making avian monitoring more efficient and accurate will significantly improve and advance research on birds and conservation efforts, but the differences in monitoring shown in this analysis should be considered when designing a study that utilizes both complementary monitoring methods.
