Thank you, I was mixing up terms. I suppose I was thinking of phonemes, but I see they're also not purely the sound... Though (I didn't actually read the article yet!) I wondered if that is what they think they found: units of sound that can vary in exact audio/phonetic expression but 'mean' the same sound to the whales. (And from which longer audible communication structures are built.)

Okay, side thought, since I'm also tired and don't feel like looking things up properly:

In simple communication, such as one might assume whale-baleen* to be, perhaps a one to one mapping of phonemes to morphemes is likely.

*I think the baleen is that krill-filtering thing you were after?