Acoustic detector and BirdNET denounce invasive bullfrog
Successful management of invasive species such as bullfrog depends heavily on early detection. The use of new technologies such as camera traps or environmental DNA is already well established, but acoustic monitoring is also on the rise. INBO was involved in a trial investigating the efficiency of passive acoustic monitoring for the detection of the invasive American bullfrog.
The test was carried out in the Nete valley in Flanders and in Umbria (Italy), where established populations of bullfrogs are abundant. In addition, the method was also used in the Spanish Ebro delta (Catalonia), where bullfrogs were recently eradicated. To pick up bullfrog sounds, we tested the Song Meter Micro (Wildlife Acoustics). For signal recognition, we used the BirdNET algorithm, known from the Merlin app for bird sounds.
BirdNET proved extremely effective in detecting bullfrog, with a remarkable precision of 99.7% correct identifications and only two false positive labelled sound recordings. The permanent recordings also provided information on ecology. For instance, the vocal activity of bullfrog was found to be higher at night than at dusk. Fortunately, no more bullfrogs were detected in Catalonia.
The study shows that passive acoustic monitoring can be a valuable complement to other methods of bullfrog detection. It can also be used in the future for real-time detection of this invasive species in new locations.
Sander Devisscher, Tim Adriaens