Acoustics of Killer Whale (Orcinus orca) Sounds Recorded In Icelandic Waters
Supervisor: Marianne Helene Rasmussen (University of Iceland)
Killer whales (Orcinus orca) produce three types of vocalizations: echolocation clicks, calls and whistles. While these vocalizations, ave very well documented for the killer whales of British Columbia, for example, only a very sparse literature is available concerning the Icelandic killer whales. Successfully describing the basic acoustics of Icelandic killer whales would be a milestone into a better tracking of those animals and into a better assessment of anthropogenic sound impacts forced on them, which could be necessary for an area like Icelandic waters where fishing boats and whale-watching boats have an important activity. The main purpose of this study is therefore to describe the technical characteristics of sounds emitted by Icelandic killer whales under normal conditions. In order to achieve this goal, sound analysis were processed on opportunistic audio records done on the 27th of July, 2013, in the Faxaflói Bay, near Keflavík, Iceland. 109 clicks, 414 calls, 177 whistles and 49 ultrasonic whistles were selected among the 1 hour and 32 minutes of audio samples. Different variables of duration and frequencies structures were measured for each of the recorded vocalizations thanks to the software BatSound. Once the measurements were done, statistical analysis were performed in order to assess if they are subject to variations within the same vocalization type. Variables were also compared between the different call types. Clicks were shown to last about 0.51 ms and to have an average peak frequency of 25.4 kHz, an average -3 dB bandwidth of 17.9 kHz and an average -10 dB bandwidth of 41.4 kHz. Calls were found to have a duration ranging from 0.18 to 6.32 s, a start and an end frequency between 0.1 end 4.2 kHz, a minimum frequency that ranges from 0.1 to 3.4 kHz, a maximum frequency ranging from 0.3 to 4.2 k Hz and a number of harmonics that ranges from 0 to 60 harmonics. Calls were divided into 22 call types with an important unbalance between the sample sizes. The observations of the calls’ enhanced harmonics led to the conclusion that they have a start frequency ranging from 0.7 to 8.1 kHz, an end frequency ranging between 0.6 and 8.2 kHz, a minimum frequency ranging from 0.4 to 7.2 kHz and maximum frequency ranging from 0.7 kHz to 8.2 kHz. It has been established that whistles have an average duration of 0.89 s, an average start frequency of 5.3 kHz, an average end frequency of 6.1 kHz, an average minimum frequency of 4.8 kHz, an average maximum frequency of 6.6 kHz, a number of inflexions ranging from 0 to 14 inflexions and a number of harmonics ranging from 0 to 14 harmonics. The measurements showed that whistles’ enhanced harmonics have an average frequency of 9.7 kHz, an average end frequency of 11.2 kHz, an average minimum frequency of 11.8 kHz and a number of inflexions ranging from 0 to 3 inflexions. And the results showed that ultrasonic whistles have an average duration of 0.48 s, an average start frequency of 24.1 kHz, an average end frequency of 29.5 kHz, an average minimum frequency of 22.5 kHz, an average maximum frequency of 30.6 kHz, a number of inflexions ranging from 0 to 11 inflexions and a number of harmonics ranging from 0 to 4 harmonics. Almost all of the variables were shown to assume differences (Kruskall- Wallis test and ANOVA test: p < 0.05) within the same vocalization. This is also the case when comparing the variables between different call types. This surprising fashion of variance attests either the variable nature of calls or whistles, either the lack of efficiency of the recordings (signals recorded off-axis, or too far from the animals). Here are the limits of an opportunistic sampling. Therefore, extended and more precise studies are needed to confirm or invalidate the results of this study.