What is the ICI distribution for SRKWs (& does it differ between coastal hydrophones and tags)?

[scottveirs]

What distribution of inter-click intervals should we expect in SRKW echolocation observed at the 24/7 monitoring of Orcasound hydrophone locations?

1. Should it be trimodal because they quantize their click rates: common slow clicks associated with navigation, rarer fast clicks with salmon searching, and very rare buzzes associated with prey capture (attempts or suceesses)?
2. Should it be Gaussian because they have a wide range of click rates that vary continuously?
3. Should it be bimodal because we almost never hear buzzes and navigation clicks sound "different" (slow, LF) than foraging clicks (fast, higher frequency)?

In other odontocetes, the sequence of echolocation clicks during a prey pursuit and capture event typically involves lower click rates during a search phase and then a buzz that has ~10x the click repetition rate. Here is an example from the Biosonar book (Surlykke et al., 2014) of sperm and beaked whales, as well as a bottle nosed dolphin and a harbor porpoise:

Another experimental observation to ponder is that in harbor porpoise the buzz doesn't necessarily stop upon prey capture, but can continue to prey ingestion (DeRuiter et al., 2009):

[scottveirs]

Leu et al. (2022) published a ICIgram (distributions of ICI within a 10-click sequence) and histogram of ICIs from click pairs for resident killer whales, but it isn't clear how many came from northern vs southern residents:

The HARPs that recorded these clicks were at a depth of 100-150m, so echolocation ranges were likely ~100m whereas Orcasound hydrophones in ~10m of water may sometimes sense clicks made at shorter echolocation ranges (order 10m).

[scottveirs]

In spring, 2024, after listening to a variety of recordings of SRKWs, Max Bolen decided upon the following nomenclature for echolocation click sequences (with ICI thresholds or ranges):

• <50ms = very fast (but what is threshold between "very fast" and buzz?)
• 50-100ms = fast
• 100-250ms medium
• 250-400ms slow
• > 400ms very slow

An alternative that we discussed was a slightly simpler nomenclature (with proposed annotation labels noted parenthetically):

• <50ms = buzz (buzz)
• 50-100ms = fast (clicks-fast)
• 100-400ms medium (clicks-medium)
• >400ms slow (clicks-slow)
• For any of these sequences, a pulse annotation could be added around each individual echolocation pulse.

[scottveirs]

Re-reading Tennessen+2024, I note this definition of slow vs fast click sequences, buzzes, and prey consumption sounds in SRKWs and NRKWs:

Tennessen, Jennifer B., Marla M. Holt, Brianna M. Wright, M. Bradley Hanson, Candice K. Emmons, Deborah A. Giles, Jeffrey T. Hogan, Sheila J. Thornton, and Volker B. Deecke. “Males Miss and Females Forgo: Auditory Masking from Vessel Noise Impairs Foraging Efficiency and Success in Killer Whales.” Global Change Biology 30, no. 9 (2024): e17490. https://doi.org/10.1111/gcb.17490.

If we adopted their ICI ranges, the nomenclature would look like this:

• <11ms = buzz (buzz)
• 11-100ms = fast (clicks-fast)
• >100ms slow (clicks-slow)

Why did they choose 11 rather than 10?