How do boa constrictors know when to stop squeezing?

January 20, 2012 • 6:36 am

The question above is one Andy Rooney never asked, but biologists are a curious lot.  Boas (yes, their Latin name is Boa constrictor) squeeze their prey, killing them by suffocation, after first biting them to get a grip.

I’ve put below a video of a boa constrictor killing a large rat this way, but here’s a WARNING: don’t watch it if you’re squeamish! Predation is never pretty, but of course natural selection doesn’t make predators kind.

Squeezing prey in this way is costly for the snake. It takes a lot of energy for a cold-blooded (ectothermic) reptile to squeeze that hard (its metabolism goes up sevenfold when squeezing), and during the act the snake itself could be subject to predation.You don’t want to squeeze longer than you have to, but how do you know your prey is dead?


Many of you might have said, “When the breathing stops,” but that’s apparently not correct.

The answer, which might seem obvious but is still cool, is in a new paper in Biology Letters by Scott Boback et al. (access is free), is that the snake monitor’s the prey’s heartbeat, and stops squeezing when the heart stops beating.

Well, the experiment is gruesome, and personally I wouldn’t kill rats to answer this question, but Boback et al. did the carnage. They used 16 wild-caught boas, and offered them killed but still-warm rats, kept warm by a sort of heating pad. The rats were also implanted with water-filled bulbs to measure the pressure of the snake’s squeezing, and then—the key—an artificial heart, formed by another water-filled bulb that was set to beat in three treatments: 1. no heartbeat, 2. heartbeat throughout a 20-minute period (the average time a snake squeezes a rat) and 3. heartbeat stops after 10 minutes.

The results:

  • snakes periodically tightened their coils in the constant-heartbeat treatment, but didn’t do that when there was no heartbeat from the outset
  • rats with the constant heartbeat were squeezed nearly twice as long as those with no heartbeat, and the pressure for the former was about 20% higher.
  • in the treatment for which artificial heartbeat was discontinued after 10 minutes, the snakes stopped their periodic increase in pressure at about 10 minutes, and pressures and squeezing times were intermediate between the full-heartbeat and no-heartbeat treatment. :’

    Snake pressure (vertical axis) versus time of squeezing prey (horizontal axis). Top: full 20-minute heartbeat; Middle: no heartbeat from outset; Bottom: heartbeat discontinued after 10 minutes.

Is this learned or innate? Experiments with naive boas, who had never squeezed a live prey, suggest there’s at least a strong innate component, i.e., an evolved, genetically-based ability to stop squeezing when you detect a stopped heart. But snakes also apparently learn to squeeze less when the heart stops, too. Regardless, this is the first demonstration that snakes can detect and monitor the heartbeat of prey that they’re squeezing.

The authors raise one caveat: warm-blooded (“endothermic”) prey like birds and mammals are killed much more quickly than 20 minutes, and snakes could use the cessation of movement as a cue, saving them lots of squeezing time.  Why the extra time and heartbeat monitoring? The authors theorize that the heartbeat monitoring originally evolved when the snakes preyed on cold-blooded (“ectothermic”) animals like lizards, which can live a lot longer than a few minutes without respiration and movement, and so cessation of heartbeat was a more reliable cue to death.

This explanation is interesting, but if boas’ current prey are mostly endothermic, there should have been selection to eliminate the heartbeat cue.  I have no idea what kinds of prey predominate in a modern boa’s diet.


Boback, S. M., A. E. Hall, K. J. McCann, A. W. Hayes, J. S. Forrester and C. F. Zwemer. 2012. Snake modulates construction in response to prey’s heartbeat.  Biology Letters online, doi: 10.1098/rsbl.2011.1105

25 thoughts on “How do boa constrictors know when to stop squeezing?

  1. But are the boas not fairly ‘primitive’ (I do not like to use the term as it seems to suggest inferior), I mean retain more ancestral features, and maybe they are generalists and have been hunting whatever they can get for a very long time so they have retained the ability? I would have also thought that if they started to swallow sooner the prey would be quickly suffocated anyway, but perhaps that increases the injury risk?

  2. “This explanation is interesting, but if boas’ current prey are mostly endothermic, there should have been selection to eliminate the heartbeat cue.”
    Why would there be if the process basically works as it is?

    1. Because of the costs. Even a working system is subject to selection if there are ways to reduce the costs or increase the benefits.

      Costs described here include risk of predation while busy dealing with prey, and the metabolic costs of contracting all those muscles for that length of time.

      How strong selection is in this case depends on many factors, including the predation risk for a Boa, and the proportion of endothermic vs. ectothermic prey in its diet.

        1. That’s irrelevant (peacocks). I don’t think there’s any evidence that the male boas invite females to come watch how hard and long they can squeeze their prey.

          1. You’ve completely missed the point…Everything that an animal does has costs – but after a certain level, there’s no little guy deciding that this particular set a genes is more fuel efficient than another. I think that what we’re looking at may be a maximum or an average of what it takes to ensure that the boa’s prey is dead (and that’s just a small sample of snakes); there simply may be not pressure enough (or enough cost) selecting out a time that’s measured in minutes.

            1. You couldn’t be more wrong. A measure of minutes is plenty of time for a predator to happen along and take out a snake that constricted longer than was necessary.

              I know it’s difficult to understand how small differences can be amplified over generations and time, but this is not a particularly small difference we’re speculating about.

              1. You couldn’t be more inaccurate-it’s a great deal more complex than just one measurement; you’ve over-simplified it to seem as if the only thing that matters is the time the snake is exposed. There are many other parameters involved. My point stands- apparently, with what little info we have on this -there just doesn’t seem to be enough selection pressure to make the span in constriction times really matter.

  3. Heart beat is amore reliable indicator even in warm blooded prey as so many stop moving before death as, it appears, a means of tricking the predator. The snakes ability to detect the heart beat thus is still a boon as energy spent is not wasted by letting up too soon. Well, it’s my guess anyway.

  4. I came to the same conclusion (heartbeat) when I saw reports of a python which seems to have tried to eat a not-quite-dead alligator. The python seemed to have been tricked by the gator’s capability of lowering its heartbeat to the level of being unobservable. Once inside the python, it woke up. Didn’t like it, left the building (IIRC).

    (I suppose here pythons and boas use the same sensors, which might of course be untrue).

  5. I find this interesting because I am a big fan of king snakes. They eat a lot of small mammals and kill them my constriction, but they are most famous for eating other snakes. They really do need their adaptations for killing cold blooded prey.

    Does anyone here know how king snakes kill other snakes? Seems to me that snakes would be hard to kill by constriction because of their weird configuration of lungs and trachea. Do they perhaps bash the snake into submission as black racers do with all their prey?

    1. Videos I’ve seen of king snakes eating other snakes shows them wrapping around them but eating them alive. A lot of king and rat snakes eat lizards when young, then switch to mammals when they get older. They often don’t bother to constrict lizards, especially small ones. Garter and water snakes grab amphibians and fish and just swallow them, often while they are still struggling. Sometimes they’ll press the prey against the ground or a rock to stop it from thrashing around too much. It doesn’t seem a stretch that constriction started as simply a way to control reluctant dinner while you’re ingesting it and then refined to actually killing the prey.

      1. I actually have some experience caring for boas, king snakes, and ball pythons on a temporary basis. The boas and pythons were adapted to dead rats. but the tiny baby king snake would only eat snakes, anoles, mice, and dwarf hamsters, and they had to be live.

        I had an out of control group of dwarf hamsters dumped on me to sex and separate. My friend didn’t believe it was the same king snake when he got back. A king snake will grow plenty fast if you give it one dwarf hamster every four days.

        I totally got along with the pythons and boas. But the king snake bit the crap out of me every time I cleaned the habitat.

        BTW, I managed to control the hamster colony and adopted a=out fifteen hamsters to good homes. Sorry I had to feed the most aggressive up and coming “queen bee” females to the king snake.

  6. I feed my corn previously frozen mice and he still squeezes for quite a while (but less than ten minutes), even though (obviously) the mouse has been dead for weeks. What’s interesting to me is that when he thinks my hand is the food he’ll squeeze for a much longer time, waiting for my heart to stop I suppose. Eventually he either figures out that I’m not edible or he tires himself out (most likely) and lets go.

    I’m always amazed at how strongly he can squeeze (and this certainly comes at a high cost). After eating he pretty much lays in his shelter for a few days which evens out this cost in energy output. Also, he’s safe from predators while he’s hidden, so this balances out also. In the wild he would only be exposed while he is hunting/eating, which may only be a few hours during one day/week.

    1. Perhaps the squeezing is required to stimulate the start of the digestive process? Like the posts of Jerry’s fab foody feasts making one start to drool?

  7. I have a pet constrictor (kingsnake) and could have answered this question. The snake is excited by motion. To say that the snake “monitors the heartbeat” makes it sound sophisticated. The snake is simply excited by any movement it feels. When it stops feeling movement, it calms down and eats.

  8. As I read it, the paper doesn’t say that the snakes squeeze warm-blooded prey longer than they need to. That would imply that the prey’s heart keeps beating long after it’s dead. Rather, the authors merely wonder why the snake detects heartbeat at all, given that warm-blooded prey could be reliably killed simply by squeezing for a short, fixed interval. Their answer, as you say, is that not all prey is warm-blooded. But it doesn’t follow that they’re wasting energy by over-squeezing warm-blooded prey. They squeeze until the heart stops, which presumably happens much sooner in warm-blooded prey.

    In fact one can concoct a just-so story that starts from a primordial behavior of squeezing for a fixed interval, which then became modified by selection to restart the interval timer each time the prey kicks or struggles. Eventually this struggle detection became sensitive enough to detect individual heartbeats. So today the rule is: keep squeezing for N seconds after each heartbeat, for some fairly small value of N.

  9. I think it would be difficult to jumpstart the selection for using visual cues rather than the inefficient heartbeat cue in the predation on endotherms.

    Wouldn’t there have be some variability in heartbeat detection and constrictor behavior to select? Boas with poor predatory abilities (heartbeat detection and constriction)are already at a huge disadvantage.

    Wouldn’t there also have to be some simultaneous compensatory switch in the sensory modality used to determine death in prey (from somatosensory to visual)? That seems like a rather high hurdle to jump.

    The costs of squeezing a few extra minutes probably isn’t quite costly enough to jumpstart the process.

  10. Loved this article. I keep snakes — no boas, but womas (an Australian python) and kingsnakes — and the researchers’ conclusions make a lot of sense. I typically feed frozen/thawed prey but occasionally offer live food and definitely see a difference in the constriction time. I’ve also had a woma strike and constrict my hand and he never did voluntarily let go, even when it should have been pretty clear he wasn’t going to get a meal out of it. I’ll assume the pulse in my hand was enough to keep him constricting.

    1. Impressive, as this is usually how snake/kitteh encounters go:

      Gotta say this rings true as my kitteh used to bring me copperheads when she was young. She did not carry them into my apartment to eat them. I just found two by my front door.

      Birds got their skulls, brains, and livers eaten. Rats were choked out and allowed to revive for fun play time, and then got their livers eaten after many more choke submissions. Best prey of all was the katydid, which lived longer under kitteh abuse than any other prey item, and they can fly. Super super fun time.

Leave a Reply