Fig 1.
Detail of study area in Scoresby Sound, East Greenland, as shown in (A) with the turquoise rectangle.
The four sub-areas used in the statistical analyses are delimited by the thick red lines: Øfjord (Ø) to the northwest; Fønfjord (F), which includes the tagging area and the fjord along the southwestern shore of Milne Land; Gåsefjord (G), the large fjord south of Gåseland; and Outer Gåsefjord (OG), where Gåsefjord broadens and merges into Hall Bredning, at the widest part of Scoresby Sound. (B) Acousonde on female Freya, 8 Aug. 2013 (picture by MP Heide-Jørgensen). The map was generated with ArcGIS using bathymetric data from the International Bathymetric Chart of the Arctic Ocean (IBCAO) and coastline data from the Geological Survey of Denmark and Greenland (GEUS) [19].
Table 1.
Morphometrics and record information for instrumented whales.
Body mass is estimated from standard length [20]. Based on his length, Balder is considered immature. The females are considered adult, but Thora only marginally so.
Fig 2.
Satellite tracks of the six subjects over the duration of their Acousonde records.
All tracks begin near the field camp (Hjørnedal), shown with a yellow dot, and end at the star symbol for each whale. The map was generated with ArcGIS using coastline data from the Geological Survey of Denmark and Greenland (GEUS) [19].
Table 2.
Various parameters related to clicking behavior (echolocation).
The percentage of time spent clicking excludes the post-release silent period. Clicking bouts of less than 1 min in duration are excluded from the five rightmost columns of data.
Fig 3.
Dive depth as a function of time (gray lines) for all instrumented whales in 2013–2016.
Periods of echolocation by the tagged whale are shown in blue and each buzz is indicated by a black dot. The red-edged yellow squares show calls made either by the whale carrying the tag or other whales nearby (see text for more information).
Fig 4.
Histograms of the length of clicking and non-clicking periods, for all records but Mára’s.
(A) Percentage distribution of clicking bouts. Each bin on the x-axis is 2 min wide except for the first one (<1 min). (B) Number of click-free bouts of various durations. The y-axis above 1 is logarithmic because of the range of values. Bins on the x-axis have a width of 5 min from 0–60 min, then a width of 1 hour from 1–37 h. Bins above 10 h are non-continuous and are only shown if they contained data. The whales’ names are placed above the first click-free bout of each record, following release. Two bouts [(1) and (2)] are shown for Balder, since he used echolocation for 23 s about 12 h after release, following which he was silent for another 24 h. The vertical arrow shows the length of the non-clicking period following Mára’s release (her data were not otherwise included in these plots).
Table 3.
Various parameters related to buzzes and calls.
Buzzes only include those made by the tagged whale, whereas calls include those made by the tagged whale and other nearby whales. The calculation of mean buzzing and calling rates omitted the post-release silent period. A few calls in Thora’s and Eistla’s records, which took place before the tagged narwhal started clicking, were excluded (see Fig 3). IQR = inter-quartile range, NA = not applicable. All depths are in meters. Buzz durations are based on 35 analyzed buzzes per whale, and 18 for Mára.
Fig 5.
Preferential depths for buzz and call production, for 17 50-m depth bins encompassing the range of depths visited by the whales (all but Mára).
(A) Depth distribution of buzzes produced by the tag-bearer. (B) Depth distribution of calls detected in the acoustic records. All calls detected were included, i.e., calls made by the tag-bearer as well as other nearby whales. (C) and (D) show ratios of probabilities for (C) buzzes and (D) calls, for the combined records. (See text for more information.) Ratios >1, shown in black, indicate a higher-than-expected rate of buzzing or calling, while ratios <1, shown in gray, indicate a lower-than-expected rate. The star symbols indicate depth bins that did not include any calls, so for which ratios of probabilities could not be calculated.
Table 4.
Odds ratios for buzzes and calls for the four whales that entered more than one area: Thora, Frida, Eistla, and Balder.
Fønfjord (area F) is the reference for Thora, Eistla, and Balder, and Gåsefjord (area G) is the reference for Frida. Values indicating odds ratios > 1 are shown in bold on a yellow background, values indicating odds ratios < 1 are shown on a light blue background, and 95% confidence intervals on the odds ratios are shown in italics. NA = not applicable.
Fig 6.
Buzzing and calling rates as a function of location for females Thora and Eistla.
Locations devoid of buzzes or calls are shown as white circles. Locations with associated sound production are shown with increasingly darker and larger symbols, in a red color scheme for buzzes [panels (A) and (B)] and a blue color scheme for calls [panels (C) and (D)]. The three sound production categories were limited by the 33rd and 66th percentiles of the buzzing or calling rate for each female, with Low < 33rd percentile, Medium > 33rd and ≤ 66th percentile, and High > 66th percentile. These percentiles were as follows—Thora (left panels): 33rd = 0.39 buzzes / min and 0.22 calls / min, 66th = 0.96 buzzes / min and 0.64 calls / min; Eistla (right panels): 33rd = 0.69 buzzes / min and 0.10 calls / min, 66th = 1.01 buzzes / min and 0.33 calls / min.
Fig 7.
These express the relationship between depth [(A) and (B)] or time-of-day [(C) and (D)] and the occurrence of buzzes [(A) and (C)] or calls [(B) and (D)].