Bigg’s bangs: the sound of marine mammal death heard live?

In the December darkness of a recent Friday night (~18:00 on 12/7/18) Orcasound app listeners witnessed a rare and unprecedented live acoustic performance by Bigg’s killer whales. It sounded like these mammal-eating orcas were killing their dinner by beating it to death. Listen to this 45-second clip (especially the big “bang” at 0:37!) —

45-second clip of Bigg’s killer whale percussive or cavitation sounds
Transients in Haro Strait on 4/20/2009, including T40 & four females. (Credit: beamreach.org)

While bioacousticians normally think of Bigg’s killer whales (aka “transients”) as stealth predators who vocalize only after a kill, in this case it sounds like they were vocalizing and hunting simultaneously during about a half-hour period. Below are links to the amazing sounds we recorded continuously at the Orcasound Lab location (just south of Roche Harbor, WA, in Haro Strait) along with some preliminary bioacoustic analysis.

Recordings

Thanks to the AWS S3 archive of the live-streamed audio data, we were able to retrieve the ~hour-long recording that contains the acoustic event that occurred late in the last twilight of December 8, 2018. Most of the vocal activity occurred over 27 minutes between 17:32 and 17:59, Pacific time.

Audacity spectrogram and annotations for ~1-hour Bigg’s recording bracketed by vessel noise. These labels are available along with the ~700Mb Audacity project files in the raw data directory.

Analysis

The entire recording was unusual in that the Bigg’s calls were abundant early on and only afterwards did we hear intense impulsive sounds,  possibly caused by percussive behavior or cavitation, and likely associated with hunting. Usually we hear the opposite: a kill, and then a lot of vocalization. Listening more carefully, there are some fainter impulsive sounds early in the recording.

Overall, it sounds like a group of Bigg’s whales moved through Haro Strait in the darkness (sunset was about an hour earlier, at ~16:17), possibly coming quite close to the Orcasound Lab (~5 km north of Lime Kiln State Park) when the most intense impulsive sounds were recorded. Interestingly, I don’t hear a lot of echolocation clicks which is consistent with the suggestion by Baird (1994, Table 4.1) that Bigg’s killer whales don’t echolocate while foraging, but conflicts with Morton (1990) that observed clicks whenever calls were heard.

Here’s a quick chronology of the ~1/2 hour-long event (times represent mm:ss into the annotated raw data), starting at 17:32:

  • [17:32:16 PST] 0:00-8:00 minutes — faint calls and whistles; around 4:05.75 there are faint percussive sounds
  • [17:40:16 PST] 08:00-17:30 — ~10 minutes of more-intense calls interspersed with low-frequency sounds (from another marine mammal?)
  • [17:49:46 PST] 17:30-24:15 — about 7 minutes of calls & whistles with intermittent boat noise
  • [17:56:31 PST] 24:15-26:30 — about 2 minutes of intermittent intense impulsive sounds!
  • [17:58:46 PST] 26:30-34:00 — over about 8 minutes, the calls slowly fade out as a ship approaches

The calls and whistles are exciting by themselves, mainly because there are so many and the background noise levels were low, but the most remarkable sounds in the recordings are the occasional incredibly-intense impulsive sounds. I’m going to call them “Bigg’s bangs” from here on out. Here is an impressive series of four bangs, the last of which preceded the most intense bang in the whole recording by ~24 seconds (which you can hear in the first clip above at 0:37).

A series of four impulsive sounds, presumably made by Bigg’s killer whales.

Each impulsive wave form has two (sometimes three) peaks in its amplitude envelope, with the 2nd peak arriving about 40-50 milliseconds after the first. This suggests that the 2nd peak is a surface bounce after the direct arrival causes the 1st peak. Since the speed of sound in salt water is about 1500 m/s, a 50 millisecond delay is means the sound wave which arrived 2nd traveled an additional 7.5 meters. This suggests the sound source was within a meter or three of the surface (or the bottom, or a reflective vertical surface).

A 2nd of the 4 Bigg’s bangs with 45 milliseconds between the 1st & 2nd amplitude peaks. The onset is certainly sudden, but not instantaneous… you can hear a little rumble leading up to the main bang when the recording is played back at 10% normal speed (e.g. in Audacity).
3rd of the 4 Biggs bangs with 44 milliseconds between 1st and 2nd amplitude peaks. This one has a much more sudden onset.
4th of the 4 Bigg’s bangs with 39 milliseconds between 1st and 2nd amplitude peaks. This one has a very sudden onset, but is preceded by a smaller amplitude peak (about 30 milliseconds earlier). Is this from a separate source, or a pre-cursor to some types of Bigg’s bangs?
Spectrogram of the four Bigg’s bangs showing power peaks near 2 kHz and — for most bangs — excursions from background levels all the way through 16 kHz, the highest frequency resolved by the Orcasound live stream (HLS segments with lossy AAC compression).

The low upper frequency limitation of the Orcasound live stream prevents us from discerning whether the bangs come from cavitation (which would have a very broad-band signature and an initial negative pressure excursion) versus physical contact between the fluke and a marine mammal prey. For example, Simon et al. (2005) provide measures of the peak frequency, 97% bandwidth, and tail slap duration. A recording that resolves higher frequencies (perhaps from our collaborators at Lime Kiln on San Juan Island (SMRU/TWM) or East Point on Saturna (SIMRES) might also let us determine if harbor or other porpoises were involved by detecting their very high frequency signals. (Do all pinnipeds haul out at night?)

Generally, these bangs are separated from each other by seconds, but sometimes two successive bangs are only 0.5 seconds apart. They also seem to come in groups which rarely have interspersed calls or whistles. If these bangs are related to foraging, it seems that there is little (audible?) communication during the kill. Perhaps when they’re killing, they’re all business?

The literature suggests otherwise.  At the north end of Vancouver Island, in the Broughton Archipelago: 

“Transients made calls only during play or after a kill…. Calls were detected in 15.6% of the transient encounters, but in 55.6% of resident encounters. Whenever transients were observed making sounds, both calls and echolocation clicks were produced together.”

Morton (1990)

At the south end of Vancouver Island:

Felleman (1991) reported that percussive behavior in transient killer whales is only regularly exhibited during predation. Transients in this study engaged in social/play behaviors, not associated with prey captures, for 3.78% of their time, and this typically involved percussive behavior.

Baird, 1994

What acoustic evidence is there of marine mammal prey? To me the low-frequency grumbles and groans that are faintly audible in the recording suggest that a pinniped that is in distress. Of course, these sounds could also be coming from the Bigg’s killer whales themselves, and/or a larger whale like a humpback.

Discussion

Since I’ve not yet heard of any visual/surface observations of Bigg’s in Haro Strait on 12/7/18 or 12/8/18, we’re left to hypothesize about the source of these bangs, and to compare their acoustic nature with related findings in the literature. Here are four sources of the bangs that I find most plausible (add your own in the comments!):

  1. percussive sounds (made when a pectoral fin, fluke, or breaching body slaps the sea surface powerfully)
  2. underwater tail slaps (like those used by Norwegian KWs hunting herring)
  3. fluke cavitation (in a form more powerful than that observed in Norway)
  4. underwater impact between predator and prey

Let’s consider these possibilities one at a time…

1) Percussives

Percussive sounds are fairly familiar to Orcasound app listeners, and to bioacousticians who study southern resident killer whales (SRKWs). These endangered salmon-eaters often slap the surface of the water. Some think they do it to herd fish; others think it is a way to communicate group activities, like changing the direction of the pod. And of course who knows why they breach, but it definitely makes a sound when a whale falls in the ocean. 

Here’s an example from 2008 at Lime Kiln when SRKWs were slapping the sea surface with moderate to low intensity. (I’m not sure if these were tail or pectoral fin slaps…) You can hear that the initial slaps sound similar to the Bigg’s bangs, but with much less intensity and reverberation, and less of a low-frequency power peak. The latter group of slaps is gentle enough that some sound like a paddle being pushed through the water (making lots of gurgling noise).

2) Underwater tail slaps (like Atlantic herring-hunters)

Killer whales in the North Atlantic use the tails to stun herring, which makes them easy to snack upon. Here’s what this acoustic foraging technique looks and sounds like for Norwegian orcas (Simon et al., 2005):

Underwater video of Norwegian orcas — compelling evidence of what underwater fluke cavitation looks and sounds like. Note that there don’t seem to be any/many herring near the first cavitation event, suggesting that this really is cavitation — as opposed to the sound of the herrings’ bodies getting smacked by the fluke. (source: movie2.mpg at http://jeb.biologists.org/content/suppl/2005/05/31/208.12.2459.DC1)
Underwater video of Norwegian killer whales. In this case there are many herring in the path of the fluke, the tail slap is less rapid, and it sounds like 2-3 events, each with multiple impacts (except maybe the 2nd hardest-to-see one which sounds more like the cavitation-only 1st strike in the preceding movie). This seems more like the sound of underwater collisions than cavitation. (source: movie3.mpg at http://jeb.biologists.org/content/suppl/2005/05/31/208.12.2459.DC1)

(Why don’t orcas do that in the Pacific Northwest, where herring are often much more abundant than the Chinook salmon that the endangered southern residents strongly prefer to eat?)

Overall these herring “slaps” and/or fluke cavitation events don’t sound similar to the Bigg’s bangs at all (IMHO). Simon et al. (2005) note that the peak power for the sounds they think are related to cavitation have center frequencies of 46+/-22 kHz, but a 97% energy bandwidth of 130 kHz, and peak power levels “below 10 kHz.” So, the spectra may be similar for cavitation versus impact (why didn’t Simon et al. tease apart the different types of sounds in the “multi-pulsed structure” of the waveforms?)…

Fig. 4B from Simon et al. (2005). How is this pulse, presumably from cavitation during a tail slap different from the spectrum of an underwater impact. Or is this pulse actually due to a herring getting hit by an orca tail?

3) Fluke cavitation (in Pacific Bigg’s KWs)

Thanks to experienced marine bioacousticians like John Ford, we know another possible source for these impulsive sounds is cavitation from Bigg’s whale flukes. While the Norwegian whales cause cavitation mostly when they strike herring with their flukes, in Bigg’s KWs cavitation could happen during bursts of speed, and/or could be a way of stunning prey.

If you ever seen these marine-mammal hunters in action, you know that they are very powerful animals. They move fast and hit hard in comparison with the slow-swimming, almost lackadaisical Norwegian herring hunters. For example, Bigg’s killer whales (aka West Coast transients) are famous for   with the sound wave that is emitted when the low pressure void (or cavity in the water) suddenly collapses on itself.

Here’s an example of what Ford says (via an ONC moderator in the YouTube video comments) fluke cavitation sounds like in a recording from the archives of Ocean Networks Canada:

Bigg’s calls interspersed with impulsive sounds, and a possible baleen whale vocalization. (Recording by Ocean Networks Canada via their Neptune Canada YouTube channel, suggesting this recording was made on the outer, west coast of Vancouver Island or at the western end of the Strait of Juan de Fuca.)

The very first impulsive signal sounds most similar to the Bigg’s bangs from Haro Strait. The others are similar, but a couple have double peaks, and the most intense one sounds more powerful and extended (a breach?). Overall, the power in these pulses peaks below 2 kHz (see vertical axis in the ONC video above, where parenthetical numbers are in Hz) which matches the spectral peaks in the Bigg’s bangs from Haro Strait.

Another thing that jumps out about the ONC recording is that in the first part the Bigg’s calls and the impulsive sounds are interspersed, in contrast to our recording where they are more clearly separated into periods of calling or banging.

4) Underwater collisions of marine mammals

While the intensity is indeed impressive from extreme cavitation or other sudden collapses of voids of vacuums — like imploding light bulbs that are often used for impulsive artificial sources in marine acoustics — I’d like to propose that there’s an even more plausible explanation for Bigg’s bangs, especially the most intense, explosive ones (like the one at 0:37 in the first clip; or the one at 0:30 in the ONC recording). I think the most likely cause of most of the sound is the physical impact of a fast moving killer whale hitting a relatively motionless pinniped or porpoise. Three basic observations motivate this hypothesis.

A Bigg’s killer whale attacks a Stellar sea lion (credit: Jeanne Hyde)

First, I know from trying to communicate while SCUBA diving or just playing in a pool that the best way to make a sharp, intense noise underwater is to slam the thumb-side of your fist into your palm. Done correctly, this action makes a snap that can be heard 10s of meters away without much effort expended (i.e. it doesn’t take all my strength to make an impressive sound). I’m imagining that a MUCH bigger mammal hitting a much bigger surface a lot faster and more powerfully could conceivably make a tremendously intense impulsive sound.

Second, the spectra from the Bigg’s bangs recorded in Haro Strait and by ONC seem to have most of their power below 2 kHz. By contrast, I think of cavitation as producing a much broader-band and flatter spectrum.

Finally, though I haven’t had the privilege of watching Bigg’s killer whales make a kill, I can tell from online videos (see below) that the most-common method of dispatching their marine mammal prey is to hit it so hard that it flies out of the water. In the best videos, it appears that the hardest hits are delivered with the head or upper back, often with the orca flying out of the water along with the battered sea lion or porpoise.

Watch a few examples and imagine what it must sound like underwater: either an occasional bang from an isolated attack, or a series of bangs due to coordinated blows delivered by a group.

A Bigg’s killer whale slams a Stellar sea lion almost clear out of the water. It looks like a second hit occurs moments later, mostly underwater.
Here’s a similar set of strikes by transients in Speiden channel (just north of the Haro Strait location) on a Stellar sea lion. The most spectacular blows by the larger male transient flip the (possible ~1000 kg?) Stellar half-out of the water.
This failed attack by a group of Bigg’s whales on a large (400 kg?) sea lion is interesting for two reasons. First there don’t seem to be many direct hits. And at 0:48 there is an impact or cavitation event just below the surface that has the sort of explosive quality that seems similar to some of the underwater bangs.
With both underwater and in-air video, David Attenborough’s team really captures the nature of the blows dealt to a Stellar sea lion by Bigg’s whales in this clip from the BBC series Nature’s Great Events, The Great Feast.  Most blows seem like direct hits from the orca’s head, shoulders, or upper back. Tail slaps seem like they may land some blows, but may also be primarily for distraction, confusion, and washing waves over the prey as it struggles to breath.
This single hit of a harbor porpoise off Victoria, BC, by a Bigg’s KW appears to have begun just below the surface (depth <1m?), but ended with the porpoise knocked ~5 meters out of the water.
This slowed footage clearly shows an impact of the head or back of the Bigg’s KW with the lower half or 2/3 of a Stellar sea lion near Hornby Island, BC.
A ~2,000 kg transient launches a ~200 kg dolphin about 10 meters out of the water. In this case it looks like the initial impact was underwater. Could this sort of hit be what causes the most intense bang in the Haro Strait recording — the one that sounds like an explosion (at 0:37)?
A big male killer whale head butts a minke whale in Antarctica. Notice how the killer whale’s head deflects during the impact.
Or are we just hearing the sound of harbor seals belly-flopping after being launched by talented transients like this!?

Conclusions

At both ends of Vancouver Island, researchers like Robin Baird (1994) and Alexandra Morton (1990) have observed the foraging behavior of Bigg’s killer whales. In a nutshell, these orcas are stealth hunters that pummel other marine mammals to death and devour them. I think we had a rare chance to hear the acoustic manifestation of Bigg’s killer whales attacking their prey.

If I’m right, these orca sounds are among the most powerful biological signals I’ve heard in 15 years of listening to the live hydrophones (e.g. Yukusam the sperm whale). Whether we’re hearing only (a) the violent underwater impacts that fling huge creatures out of the sea and/or (b) the cavitation of tail flukes during the dramatic acceleration towards the high-velocity impacts we observe near the surface remains to be determined.

The good news is we should be able to distinguish between the impact and/or cavitation hypotheses once we obtain some higher-quality (lossless) recordings, ideally ones that resolve frequencies up to ~100 kHz and that are contextualized by visual observations at the surface or — better yet — underwater. (Maybe an ambitious researcher should get a GoPro on a foraging transient?!) Impacts should have a more negative power spectral slope, while the spectrum of cavitation should be more flat. Also, the initial pressure variation should be negative for cavitation, but positive for an impact. Some acoustic lab experiments with underwater models of fins and/or colliding bodies may also be in order!

I’ll leave you with a final plea for more videos like this one (below). If we can get more hydrophone on more boats, or more cameras monitoring surface behavior where we also have recording high-quality hydrophones, we’ll be well on our way to new discoveries. I couldn’t convince myself that this footage from Orca Lab had the underwater sound synched with the surface video, but it’s the right idea — and some of the faintly audible underwater sounds appear to be similar to Bigg’s bangs!

Bigg’s whales attack a sea lion in Johnstone Strait. Note the (possibly simultaneous?) underwater sound, likely played through one of Orca Lab’s local hydrophones…

References

Future work, other questions

Historically, this event would be a very rare occasion, but perhaps the Bigg’s whales are just becoming more chatty because the southern resident fish-eating orcas haven’t been around as much the last few years?

Compare SL of pile drive and echolocation click w Norwegian tail slap. Use LK and -18 log(R) TL to estimate SL for Bigg’s bangs. At what RL from pile driving, air guns, or dynamite fishing are fish stunned and or killed and or damaged? How about marine mammals like pinnipeds or harbor porpoises?

Are Bigg’s eating herring as pre-pinniped hordouvres? Or do the LF groans in the recording suggest the Bigg’s are hunting CA or Stellar sea lions — in part by generating cavitation sounds with their flukes?

Why are SCUBA diver claps so loud? And how broadband are they?

Leave a Reply

Your email address will not be published. Required fields are marked *