The effectiveness of acoustic indices for forest monitoring in Atlantic rainforest fragments

Abstract

Forest monitoring is essential to maintain the integrity of protected areas, which has become a more pressing issue due to anthropogenic pressures and global climate change. Birds are widely used for this purpose, and as bird diversity is correlated with acoustic indices, it has been suggested that sounds can be used as an ecological indicator of ecosystem complexity and environmental changes. However, previous studies yield controversial results. Thus, we evaluated the effectiveness of six acoustic indices compared with the results obtained from a traditional point-counts survey for forest monitoring in an Atlantic rainforest fragment in Southern Bahia, Brazil. We also determined differences between acoustic indices calculated in the presence and absence of the researcher in the field to quantify the bias caused by human presence during data collection. There was a moderate correlation between the number of bird species just with the acoustic evenness index, from the six acoustic indices evaluated. We also determined that the presence of the researcher in the field at the point-count survey caused bias in acoustic indices. Therefore, more research must be done with autonomous recording units and the acoustic indices before they can be widely adopted for monitoring purposes in complex and diverse habitats such as the Atlantic tropical rainforest.