Susan E. Cosens
Canada Dept. of Fisheries and Oceans, 501 University Cres., Winnipeg, MN R3T 2N6, Canada
Offshore industrial noise sources ensonify marine mammal habitat and may mask detection of social signals and important environmental sounds. A probabilistic model of sound detection was developed for assessing the impact of underwater icebreaker noise on signal detection by beluga whales and narwhals. The probability of detecting sample signals, in the absence of vessel noise, was compared to that of detecting the same signals in ship noise, using the MV arctic, as a sample noise source. Analysis of signal detection probabilities showed that loud signals centered on the 5-kHz critical band were more severely masked by ship noise than were quiet 5-kHz signals or loud 2-kHz signals. Thus long-range calls seem to be more susceptible to masking by ship noise than are short-range calls. The model could also be applied to other critical bands of interest. Changes in the probability of detecting vessel noise, as estimated by the model, were also correlated with changes in beluga and narwhal behavior, observed in response to the vessel in operation.