Ecologists have new ways to keep their ears to the ground. Rather than crouch in the underbrush to listen for an elusive species, they increasingly rely on recording equipment that they can leave in the field. But relaying thousands of hours of audio and sorting the signal from the noise brings new headaches. A team of ecologists and computer scientists at the University of Puerto Rico, Rio Piedras, in San Juan hope to change that with their Automated Remote Biodiversity Monitoring Network (ARBIMON),  described today in the journal PeerJ. The system captures 1-minute audio segments at regular intervals using an iPhone in a solar-powered field station and immediately sends them to a master server. Relevant calls, like the subtle grunt of the critically endangered coquí llanero (Eleutherodactylus juanariveroi, pictured), are singled out using a computer algorithm, which scientists can train to identify any species they choose. The team suggests that even ecologists without computer programming savvy can now set up more sophisticated surveillance for species threatened by climate change and habitat loss.