I missed this when it came out — Virgina Morell, "Elephants Have an Alarm Call for Bees", Science Now 4/26/2011:
East Africa’s elephants face few threats in their savanna home, aside from humans and lions. But the behemoths are terrified of African bees, and with good reason. An angry swarm can sting elephants around their eyes and inside their trunks and pierce the skin of young calves. Now, a new study shows that the pachyderms utter a distinctive rumble in response to the sound of bees, the first time an alarm call has been identified in elephants.
… [T]he study suggests that this alarm call isn’t just a generalized vocalization but means specifically, “Bees!” says Lucy King, a postgraduate zoologist at the University of Oxford in the United Kingdom and the study’s lead author.
When they hear buzzing bees, the pachyderms turn and run away, shaking their heads while making a call that King terms the “bee rumble."
The associated scientific paper is Lucy King et al., "Bee Threat Elicits Alarm Call in African Elephants", PLoS ONE 5(3) 2010. Here's what the "bee rumble" sounds like:
Interestingly, there was not much media uptake for this story — aside from Science, only New Scientist seems to have responded to the press release.
At the other end of the acoustic spectrum, there are the ultrasonic vocalizations of rats, which alleged have a simple relationship to their neurochemistry. Thus S.M. Brudzynski, "Communication of adult rats by ultrasonic vocalization: biological, sociobiological, and neuroscience approaches", ILAR J 2009:
Rats have developed antipredator defensive adaptations to protect themselves from the large number of animals that prey on them. One such adaptation is the ability to communicate by ultrasonic vocalization, which decreases the likelihood of detection by a predator. [...] Adult rats emit two types of ultrasonic calls: alarm (22 kHz) calls, in aversive and dangerous situations, and appetitive (50 kHz) calls, in appetitive or "friendly" (i.e., nonaggressive) behavioral situations. These two types of calls differ in all acoustic parameters, and their initiation depends on activity in different ascending tegmental pathways to the forebrain: 22 kHz calls require activity in the cholinergic system, and 50 kHz calls in the mesolimbic dopaminergic system. The release of acetylcholine in the diencephalon and forebrain is associated with the emission of 22 kHz vocalizations, and the release of dopamine in the nucleus accumbens with 50 kHz calls. Thus the two calls are reliable predictors of increased cholinergic or dopaminergic activity in the brain. The calls serve as indices of the rat's state, including its affective state, induced by activity of one or the other neurochemical system.
Here's 22 kHz rat vocalization, identified as a "distress call", and slowed way down:
Some similarly slowed-down calls in the 40-50-kHz neighborhood:
And here's a 60+ kHz vocalization, also slowed down:
A suggestion that things might not be so simple can be found in Ronald J. Barfield et al., "Ultrasonic Vocalizations in Rat Sexual Behavior", American Zoologist 19(2) 1979:
Ultrasonic vocalizations are very conspicuous during rat mating activity. Two types of calls are produced by both sexes. The first, brief complex calls with the main frequency centered about 50 kHz, occur primarily in conjunction with solicitation and mounting activity. The second type of call is the long, 22 kHz whistle which is emitted mainly by the male during the postejaculatory refractory period, but also by both male and female at other times during the copulatory sequence. [...] Following ejaculation, the male characteristically emits 22 kHz vocalizations and exhibits a sleep-like EEG pattern. The function of the postejaculatory vocalization may be to enforce separation between the mating pair, while at the same time maintaining contact between the partners.
An experiment sure to make it big in the media: the effect of humanized FOXP2 on rodent sexual vocalizations. The 2009 study of mice with humanized FOXP2 was sort of a disappointment ("Mice with the "language gene" stay mum", 6/4/2009), because pretty much the only measured effect was to make the separation cries of pups somewhat lower and flatter in pitch. Nevertheless, the study got wide coverage in the mass media. Can you imagine the impact if they had studied sex vocalizations instead?