Killer whales (Orcinus orca) are top marine predators occurring globally. In Antarctic waters, five ecotypes have been described, with Type C being the smallest form of killer whale known. Acoustic recordings of nine encounters of Type C killer whales were collected in 2012 and 2013 in McMurdo Sound, Ross Sea. In a combined 3.5 hours of recordings, 6386 killer whale vocalisations were detected and graded based on their signal-to-noise ratio. Spectrograms of the highest-quality calls were examined for characteristic patterns yielding a catalogue of 28 call types (comprising 1250 calls). Acoustic parameters of each call were measured and summarised by call type. Type C killer whales produced complex calls, consisting of multiple frequency-modulated, amplitude-modulated and pulsed components. Often, two components occurred simultaneously, forming a biphonation; although the biphonic components did not necessarily start and end together, with one component lasting over several others. Addition and deletion of components yielded call subtypes. Call complexity appears stable over time and may be related to feeding ecology. Characterization of the Type C acoustic repertoire is an important step for the development of passive acoustic monitoring of the diverse assemblage of killer whale ecotypes in Antarctica’s rapidly changing marine ecosystems.

Acoustic data were collected near the fast ice edge in McMurdo Sound, Ross Sea, Antarctica, between December 2012 and January 2013. The hydrophone was hand-deployed into the water at the ice edge in the immediate vicinity of killer whales (i.e., at < 100 m range) approximately 3-4 m below the surface. Acoustic recordings were obtained using an M-Audio Microtrack 24-96 recording unit with a Lab-Core LAB-40 hydrophone system (~5 Hz - 85 kHz bandwidth) and a 20 dB in-line amplifier. Sound was sampled at 96 kHz, 24 bit or 44.1 kHz, 24 bit, providing a bandwidth of 48 kHz and 22.05 kHz, respectively. Acoustic recordings were inspected both visually and aurally using acoustic software Raven Pro 1.5. Calls were visually graded based on their signal-to-noise ratio (SNR). Calls graded 2 and 3 were sorted into distinct categories to produce a call catalogue.