Killer whale call detection rates vary among subspecies and populations in the North Pacific

Killer whales (Orcinus orca) in the North Pacific have diverged into subspecies with distinct diets, social structures, and vocal behaviors. Understanding the rate at which killer whales of different subspecies vocalize provides insight into their communication with conspecifics as well as the environmental factors shaping vocalization patterns. Estimating a population's calling rate can also enable the number of animals present to be estimated acoustically. We used killer whale dialects, long-term photo-identification records, and year-round autonomous acoustic data across multiple locations to estimate call detection rates during vocal bouts for three killer whale populations, one resident (fish-eating) and two transient (mammal-eating). We focused on call detection rates, or the number of calls detected rather than the true number produced, in order to capitalize on this large autonomous dataset to investigate the influence of spatiotemporal and social factors across many observations. Generalized additive mixed models showed that the call detection rate for southern Alaska resident killer whales was consistent across locations, seasons, which pod was calling, and the presence of other pods. Gulf of Alaska transient killer whale call detection rates differed across locations. We detected fewer, more tightly spaced discrete calls from AT1 transients than from Gulf of Alaska transients, and the mean AT1 call detection rate was stable across spatiotemporal factors. Although transients call less often than residents, we found that once vocalizing, they did so at a higher rate. These call detection rates provide novel insight into killer whale vocal behavior and can support passive acoustic group size estimation for these populations.