The call repertoire of Icelandic killer whales and how it varies between locations
Our latest research on Icelandic killer whale calls shows that killer whales in the major herring spawning and overwintering grounds in the South and West of Iceland share most of their calls but also have a proportion of calls that are unique to these locations. Furthermore, call types and subtypes used appear to be stable over a long time period, with minor variation between years.
Killer whales in Vestmannaeyjar, south Iceland. Photo: Filipa Samarra
For this study we had an exciting dataset available that included recordings of killer whales from five locations around Iceland that were collected over a 32-year time period, from 1985 to 2016. Most of the data were collected by the Icelandic Orca Project in herring spawning and overwintering grounds in the South and West of Iceland. But we also reached out to colleagues that provided us with recordings from other locations around Iceland. This allowed us to make a comprehensive comparison between locations and over time.
Killer whales produce different types of sounds, including echolocation clicks, whistles and calls. Calls are the most commonly produced sounds of killer whales and have received most attention in research. They are very stereotyped, meaning they can easily be grouped into different categories. While the basic properties of these calls are similar for all killer whales, different populations or social groups differ in the exact call categories used. In the resident killer whales in the North Pacific for example, different matrilines share only part of their repertoire of calls and the closer related they are the more calls they will share. Therefore, specific groups can be recognised simply by listening!
In a previous study, we found that killer whales in Iceland share part of their repertoire with whales from Shetland but not with whales from Norway. However, we knew very little about how the repertoire varies between different locations around Iceland and whether it changed over time.
The data set we had included over 3,000 hours of recordings, some of which were from the 1980s and still on cassette tapes that needed to be digitised before analysis. We found a total of 8,321 high-quality calls in these recordings and classified them into a total of 91 different call categories. These call categories included call types and subtypes. Different subtypes of a call type are calls that are very similar but have a specific element that is different or added. Humans have been found to be very good at classifying whale and dolphin sounds into meaningful categories. Nevertheless, the process is always slightly subjective. This study also gave us the opportunity to confirm our classification with two different computer models. We were very happy to see that the models showed high agreement with our manual classification. This provided us with confidence that we are not comparing arbitrary categories in the following analysis.
A treasure box of tapes with killer whale sounds recorded in the 1980s in the East of Iceland. Photo: Filipa Samarra
Comparing the call categories recorded at different locations, we found that about half of the call categories (56%) were found at two or more locations but only one call category was recorded in all five locations. The remaining categories were unique to either the south, the west, or the east of Iceland.
Map of Iceland showing the locations where acoustic recordings of killer whales were collected, including recording years. Pie charts show the patterns of call category matching between areas (n = number of call categories per location). Call categories occurring in only one location are indicated in yellow (Eastfjords), orange (Breiðafjörður), and turquoise (Vestmannaeyjar), while call categories occurring in multiple areas are indicated by varying shades of blue and grey.
Three examples of call type I11.4, a very common call type and the only one that was recorded in all 5 locations.
Three examples of call type I45, the most common call type.
Three examples of call type I43, showing different subtypes of this call type.
In the south of Iceland (Vestmannaeyjar) our recordings spanned a 14-year time period (2002-2016), which allowed us to investigate changes in the repertoire over time. As in other killer whale populations, the repertoire seems to be stable over several years, although fluctuations may occur over shorter time periods. In the beginning of our study period, for example, new call categories were steadily added with each new encounter and recording. However, there was a clear jump in new call categories in 2014, which corresponded to an increase in individually identified whales. Therefore, it seems that whales that had not previously been present started to use this location and brought their own sounds with them.
One of the most striking findings was that a large part of the repertoire was shared between the herring spawning and overwintering grounds in the south and west of Iceland but that both locations also had call categories that were unique and not recorded elsewhere. Furthermore, call categories that were the same in both locations were not used at the same rates. We already knew from our photo-identification studies that about half of the killer whales in Iceland seem to follow the herring year-round, moving between the herring spawning and overwintering grounds, while the other half is only ever seen in one of the two locations. The results of this acoustic study closely match those findings and support the idea that individuals or groups have different acoustic repertoires or that they use their repertoires differently in different social or environmental contexts. This emphasises the importance of long-term studies and the need for future research on whether Icelandic killer whales have group-specific repertoires.
This study was possible thanks to a collaboration across many institutions and years. Many thanks to our co-authors for their support and for searching their archives for any snippet of killer whale recordings!
The paper is available here: https://onlinelibrary.wiley.com/doi/full/10.1111/mms.13039