This is no Nutcracker. After sundown, the pocket mice of Saguaro National Park are working it like a strip joint. Check it out (second one has my narration):
A brief summary also appear’s on Felicity Muth’s blog (Not Bad Science) over at http://www.ScientificAmerican.com. What started as bored amusement at this energetically costly behavior turned to curiosity as I realized how often I was seeing it. Idle curiosity turned into fascination as I realized how few mammologists had even seen the dance before.
But beyond the novelty and the entertainment factor, does this behavior matter at all to the ecology of the species? Does it affect their population dynamics?
Maybe.
1. What if this is anti-snake flagging behavior? (Cool video.) I have already posted about the ways in which an energetically costly anti-predator behavior can affect the energy flows in an food chain, and the densities of both the predators and their prey. So if this is tail flagging aimed at a snake, it could decrease the population growth rate of the mice.
2. What if mice are marking territories? I actually had more trouble than I expected finding general theoretical conclusions on the effects of territoriality on population dynamics, but please feel free to correct me if you know of some (or have research of your own I overlooked!). A trivial answer is that dividing space into territories is a form of intraspecific density dependence – that means the species is limiting its own population growth. But does that limit growth differently than pure resource competition? What about in a spatially variable environment? Maybe I can contribute something through my own research here eventually.
3. And of course, what if this is sexual behavior? Male mice have been reported to do a butt wiggle before mounting a female (Bret Pasch, personal communication), so these could be males smelling a potential mate and getting excited. If they are wasting energy dancing too early in anticipation of mating before even locating the female, that could decrease their eventual longevity or mating success, I suppose. That might not even matter to population dynamics if there are plenty of males to get out there and mate, though. Especially if dancing is a minimal cost relative to the total effort of running around through the desert searching for females.
Finally, how can I start testing these hypotheses? Proving the presence of an off-camera snake that is triggering flagging behavior might be hard, but a good first step would be to put a mouse and a snake together and video the encounter to compare with what I have. To detect urine spray for territoriality, I have tried placing filter paper around the camera to inspect with a black light in search of urine droplets. No dancing over my filter paper, yet, sadly. Additionally, if I want to find out whether only males do the dance, I could capture and mark mice in a study area before re-deploying the cameras.
Help me out by posting your own ideas about why the mice are shaking their booties, how to test the ideas, and what the important implications are!
[Update 1: I did not speed up or slow down any of the videos. All I did was select the parts of clips where they are “dancing.” Nothing else is changed from raw footage.]
[Update 2: Noelle Bitner, a doctoral student (candidate?) in Dr. Michael Nachman’s lab at Berkeley, reports she has two female pocket mice who “dance.” So much for the male-specific sexuality hypotheses? Also, Vicky Chan, a graduate student in Optical Sciences, reports her captive pocket mouse (unknown gender) often dances after caching seeds. Support for scent marking?]
Biologist. Explorer. Storyteller. 
My guess: scent marking. Chaetodipus intermedius are thought to be territorial. Would be interesting to know if your “booty-shaking” mice are males, females or both.
You may also consider posting this on MAMMAL-L@si-listserv.si.edu to gather more opinions from professional mammalogists.
Thanks! That’s a great idea.
I’ve had some chaetodipus eremicus in captivity. They do a lot of waggling. I assumed it was territorial, but could never see any liquid being left behind. Filter paper sounds like a good idea. I may try that out. The eremicus I’ve kept would all dance in response to my presence, some more reliably than others. The length of the dance in the captive animals seemed to depend on their personality. They would stop dancing once they became too scared by my approach or became comfortable with my presence.
Very cool! I’d love to talk with you more about your observations. From what I’ve learned since posting this, I suspect it is more of a tail-flagging behavior akin to that of squirrels. The only mention in the scientific literature I have found is in a 1963 monograph on heteromyid behavior that categorizes it as a conspecific aggression signal. Since the mice are almost always looking at the camera when they dance, I suspect it is a more generalized warning behavior. It would be interesting to see if more waggling correlates with other behaviors of dominance and aggression, or if it correlates more with fearfulness and lack of curiosity about new objects.