Tropical forests are facing threats that may affect the dynamics of seed dispersers which participate in the forest regeneration. To implement appropriate conservation programs, it appears necessary to monitor seed dispersers and to estimate their response to local changes. Here, we used non-invasive ecoacoustic methods to monitor the activity of a canopy bird, the White-throated toucan, Ramphastos tucanus, a major seed disperser and flagship species of the Amazonian forest. We deployed nine acoustic recorders over 29 days along a road that connects French Guiana to Brazil. We used template matching to automatically detect the vocalizations of R. tucanus. This method, which can easily be repeated with limited human expertise, detected 1748 recordings with vocalizations. A GLMM analysis was applied to test for a possible effect of habitat type and human activity, while accounting for time of the day and rainfall. The number of vocalizations varied according to time of the day with peaks at dawn and dusk. The number of vocalizations did not differ significantly among sites, they were not affected by habitat type, and they were only marginally influenced by human activity. These results indicate that the vocal activity of a key conservation species can be monitored automatically in a non-invasive way. The species targeted, R. tucanus, does not seem to be significantly impacted by the road and local human activity. This might be related to the mobility of the species, which can easily cross the road, as well as low local forestry pressure.