Welcome to nature. com!Your current browser version has limited support for CSS. For the best experience, we recommend using a more updated browser or disabling compatibility mode in Internet Explorer. In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript. Researchers have utilized artificial intelligence (AI) to track the movements of two endangered species of dolphins in the Amazon River. By training a neural network to recognize the unique clicks and whistles of the dolphins, the researchers were able to build an accurate picture of their movements across the rainforest, which becomes submerged after the rainy season. This method, using sound, is less intrusive compared to conventional tracking techniques like GPS tags, boats, or drones. Co-author Michel André, a bioacoustician at the Technical University of Catalonia in Barcelona, Spain, highlights that sound is a universal sense that is shared by all living beings on Earth. André and his colleagues focused on studying the activity of the boto (Inia geoffrensis) and the tucuxi (Sotalia fluviatilis) across the floodplains of the Mamirauá reserve in northern Brazil. Underwater microphones were placed at various locations to monitor the dolphins' whereabouts. To distinguish the dolphin sounds from the ambient noise of the Amazon region, the researchers employed AI. They used a deep-learning neural network capable of categorizing sounds in real-time, mimicking the human brain's ability.

Federico Mosquera-Guerra, a researcher studying Amazonian dolphins at the National University of Colombia, notes that this technology allows analysis of vast amounts of information that would otherwise be nearly impossible. The neural network was trained to identify three types of sounds: dolphin sounds, rainfall, and boat engines. The dolphins rely heavily on echolocation clicks to navigate their environment and communicate with each other through whistles. By detecting these clicks and whistles, the researchers were able to map the dolphins' movements. The distinct whistles of botos and tucuxis allowed the neural network to differentiate between the species. This study is a collaboration between the Technical University of Catalonia and the Mamirauá Institute of Sustainable Development, aiming to utilize this technology for monitoring biodiversity and threats in the Amazon. Both dolphin species are currently endangered, with the boto population declining by 50% every ten years and the tucuxi population declining every nine years. Understanding their movements will enable researchers to protect their populations and develop measures to facilitate the coexistence of dolphins with indigenous communities. Dolphins can have a disruptive impact on fisheries by competing for fish or becoming entangled in nets. Consequently, collecting such information is crucial in informing conservation decisions across the Amazon region. In the future, the team plans to expand the system's capabilities by training the neural network to detect other aquatic species and deploying it over a larger area. This approach could also be applied to the ocean. André's previous work using this system has shed light on the effects of human-made noise pollution on sperm whales and has led to the development of a warning system for ships to mitigate harm to the whales.