Camera traps and acoustic approaches are needed to monitor Malagasy seed dispersers

Despite extensive conservation efforts worldwide, biodiversity loss continues, particularly in tropical regions. It is, therefore, essential to develop biodiversity monitoring methods that can be effectively scaled up in both time and space, enabling wildlife managers to quickly assess and respond to changes in populations or communities. We explored the effectiveness of combining camera traps with acoustic methods to survey seed dispersing birds and mammals. We deployed arboreal and terrestrial camera traps alongside AudioMoth sound recorders with varying recording schedules to survey 220 sampling locations across five distinct sites in Madagascar’s eastern humid forests, though acoustic analyses were limited to one site. Based on all three methods (arboreal, terrestrial camera trap and AudioMoths), we successfully identified 78 % of the seed disperser species assumed to be present at our sites. The species accumulation curves estimated higher species richness for arboreal cameras than for terrestrial cameras, and for mature protected forests compared to other habitats. Species composition differed between the two types of cameras, while bird species composition in Ranomafana differed between camera traps and AudioMoths. Arboreal camera traps detected more frugivorous primate, bird, and bat species while terrestrial camera traps detected more ground-dwelling vertebrate species. Where both cameras and AudioMoths were deployed species composition differed significantly, with AudioMoths detecting more vocal bird species. These results highlight the need for a combined approach to fully survey the seed-dispersal community, along with further development of acoustic analysis techniques for future studies.