Slimy Signals Save Salamanders

Everyone has experienced it: you hear a scream or shout of alarm coming from someone nearby, and instantly you feel your own blood pressure rise. Your metabolism revs up, you scan the area around you, and you might even physically jolt.

All of this can happen before you've even observed the potential threat--you're reacting entirely to a cue from another person. We are tightly tuned to react to the alarm signals of others, because a threat to one member of our species (a fire, an oncoming truck, an axe-murderer, and so on) may be a threat to other humans in the vicinity as well. The blinking lights attached to many fire alarms exemplify what a handicap it is

to be unable to hear commonly used alarm sounds.

Humans are not the only animals that react to each other's alarm calls, of course. Back in 2007, I had the chance to work on a fascinating alarm call study, which demonstrated that descriptive information is encoded in the alarm calls of Gunnison’s prairie dogs (Cynomys gunnisoni). I have been extremely interested in studies on alarm calls ever since.

Gunnison's prairie dog (Cynomys gunnisoni) emitting an alarm call. Photo by Anne-Marie Hodge.


It should not be at all surprising that conspecifics react to each other’s calls; this seems to be a natural part of communication. Furthermore, an individual can help preserve some of its own genes if its alarm calls help protect relatives from predators or other deadly situations. The phenomenon of alarm sharing doesn’t end at species boundaries, however: researchers are discovering more and more cases of cross-species alarm sharing. Some of the most common examples are of birds that travel in mixed-species flocks (Goodale et al. 2005; Lea et al. 2008; Satischandra et al. 2010) and fish that forage and travel in mixed-species shoals (Brown et al. 2001; Júnior et al. 2010).

Alarm sharing is common within taxa (ie, between different birds or different fish) but less so between different classes, orders, or phyla (This is a well-done video about the taxonomic hierarchy for animals, if you want to refresh). Still, cross-species alarms do occur: on the plains of Africa, impala (a species of ungulate) will pay attention to alarm calls emitted by baboons (Kitchen et al. 2010), and dwarf mongoose are thought to respond to the alarm calls of hornbills (Rasa 1983).

How do two species develop the ability to cue into each other's alarm calls? This is clearly not a signaling system that just pops up overnight. A shared habitat is a must, but other factors can make alarm-sharing more likely. It is thought that one primary determinant is the presence of a shared predator (Mathis & Smith 1993), especially one that seems to specialize specifically on the alarm-sharing prey species.

Ringneck snake (Diadophus punctatus). Photo by Brian Gratwicke.

For example, the ringneck snake (Diadophis punctatus), common to the southeastern United States, is thought to specialize on both earthworms and salamanders. Earthworms communicate largely using chemical cues secreted within their mucus. Likewise, salamanders have specialized chemosensory organs--the vomeronasal cirri—for picking up chemical signals from predators, prey, potential mates, and a variety of other environmental signals. The fact that the ringneck snake's two primary prey species both have highly attuned chemical communication systems seems to create a perfect opportunity to look for shared alarm calling behavior, with the “call” in this case being chemical rather than auditory. A team of researchers from Missouri State University decided to investigate the situation, and recently reported their results in the journal Ethology Ecology & Evolution (Crane et al. 2013).

Ozark zigzag salamander (Plethodon angusticlavis). Image via Wikimedia Commons

Earthworms often exhibit high densities in areas frequented by Plethodontid salamanders (also known as the "lungless salamanders"), and the salamanders sometimes actually co-opt burrows originally created by earthworms (Caceres-Charneco & Ransom 2010). Crane and colleagues collected wild earthworms and Ozark zigzag salamanders (one of the coolest common names I’ve seen for a salamander; Plethodon angusticlavis) from Missouri for their experiment. They hypothesized that chemical cues from stressed earthworms would induce the salamanders to exhibit antipredator behavior and increase their oxygen consumption. These are signs that they’re revving up their metabolisms, as if going into “fight or flight” mode . . . although in reality there is no predator, just stress-tainted worm mucus. Yum. The researchers tested this hypothesis by exposing adult salamanders to three different treatments: water containing chemical stimuli from stressed earthworms, water containing chemical stimuli from nonstressed earthworms, and a blank water control.

Earthworm. Image via Wikimedia Commons

Now this might make you wonder: how do you actually collect chemical stimuli from a stressed earthworm? The answer: you stress it out, methodically. Each worm was taken from its refrigerated holding cell, “repeatedly captured with forceps,” and swirled around in the water with the forceps for one minute. Sounds plenty stressful to me…the water they were swirled in then contained their mucosal alarm secretions (visible in the water after the worm-swirling), and was used as the “stress” stimulus. For the “non-stressed” stimulus, the worms were simply picked up gently with the forceps and placed in the water (unswirled) for one minute, then removed.

Next, the salamanders were placed in small chambers and exposed to one of the three water treatments. In the first experiment, salamanders were observed for five minutes after the stimulus was added, during which time the researchers recorded the amount of time the salamander spent with its head or body pressed against the edge of the chamber (an escape/hiding behavior), the number of times it tapped its nose on the bottom of the chamber (an exploratory chemosensory behavior), and the number of steps it took (overall activity level, used as a proxy for stress). In the second experiment, the salamanders were exposed to the stimuli within special metabolic chambers that measured each animal’s oxygen consumption during the trials. This was done to determine whether the salamanders’ metabolic rates increased when exposed to the stressed worm stimulus, relative to nonstressed worm and control stimuli.

The results showed that the salamanders did indeed react to the worms’ stress chemicals by exhibiting vigilant, antipredator behavior. The salamanders spent significantly more time at the edges of their enclosures—a tenfold increase— in the stressed worm treatment than in the nonstressed worm and water treatments. Nose-tapping behavior (associated with exploration) significantly decreased when exposed to stressed worm stimuli, which is to be expected if the salamanders suddenly felt the need to buckle down and watch for predators when they sensed the worm's stress cues.

The metabolic response was also significant: when the salamanders were exposed to the stressed worm treatment, they used significantly more oxygen than when exposed to either the nonstressed worm or blank water treatment. This was shown not to be a response to worm presence alone, as the salamander’s oxygen consumption when exposed to nonstressed worm stimuli didn’t differ from their rate when exposed to mere water. The salamanders took significantly more steps when exposed to either stressed or nonstressed worm stimuli than when exposed to water, which could simply be an exploratory response to the “presence” of a novel animal that they perceive to be in close proximity.

These findings provide strong support for the hypothesis that Ozark zigzag salamanders are able to perceive alarm chemicals emitted by earthworms when the worms are faced with stressful situations. The results also raise several questions. Could it be that the salamanders aren’t necessarily “reading” the stressed worm treatment as an alarm cue, but are simply not fond of earthworms and are exhibiting an aversive response to worm secretions? Crane and colleagues address this: they report that previous studies have shown that these salamanders readily approach and consume earthworms, so it is unlikely that their avoidance and metabolic stress demonstrated in this study are a result of aversion to the worm itself (also evidenced by the difference in reaction to stressed vs nonstressed worms). They also explain that it’s unlikely that the increased activity and oxygen consumption are predatory responses to earthworm presence, as previous studies have shown that predatory movements would decrease “edge behavior.” This behavior was actually shown to dramatically increase when the salamanders were exposed to stressed worm cues.

There are many avenues of inquiry that could enhance our understanding of these interactions in the future. One potential issue that remains to be resolved is that the salamanders used in this experiment were collected from the wild as adults. Thus, at this point we can’t be sure of how much this cross-species alarm detection is innate versus learned from experience. Given the lethal consequences of missing a predator alarm cue, innate ability isn’t impossible, but further experiments will be needed to demonstrate how the ability develops.

This new study by Crane and colleagues provides a fascinating example of not just cross-species communication, but cross-phyla communication, with an amphibian responding to the alarm signals of an invertebrate in order to reduce its vulnerability to predators. What’s the payoff for the earthworm, though? If the ringneck snake eats both of them, isn’t it to the earthworm’s advantage to let the reptile fill up on salamanders? That issue isn’t addressed in the present study, but it will be fascinating to see if follow-up studies are able to unveil more of the underlying intricacies of these interactions.





Brown, G. E., LeBlanc, V. J., & Porter, L. E. 2001. Ontogenetic changes in the response of largemouth bass (Micropterus salmoides, Centrarchidae, Perciformes) to heterospecific alarm pheromones. Ethology 107, 401-414.

Caceres-Charneco, R.I. & Ransom, T. S. 2010. The influence of habitat provisioning: use of earthworm burrows by the terrestrial salamander, Plethodon cinereus. Population Ecology 52, 517-526.
Crane, A.L., Lampe, M.J., & Mathis, A. 2013. Detecting danger from prey-guild members: behavioural and metabolic responses of Ozark zigzag salamanders to alarm secretions from earthworms Ethology Ecology & Evolution DOI: 10.1080/03949370.2013.800162

Goodale, E., & Kotagama, S. W. 2005. Alarm calling in Sri Lankan mixed-species bird flocks. The Auk, 122, 108-120.

Júnior, A. B., Magalhães, E. J., Hoffmann, A., & Ide, L. M. 2010. Conspecific and heterospecific alarm substance induces behavioral responses in piau fish Leporinus piau. Acta Ethologica, 13(2), 119-126.

Kitchen, D. M., Bergman, T.J., Cheney, D. L., Nicholson, J. R. & Seyfarth, R. M. 2010. Comparing responses of four ungulate species to playback of baboon alarm calls. Animal Cognition 13: 861-870.

Lea, A. J., Barrera, J. P., Tom, L. M. & Blumstein, D. T. 2008. Heterospecific eavesdropping in a nonsocial species. Behavioral Ecology 19, 1041–1046, doi:10.1093/beheco/arn064

Mathis, A. & Smith, R. J. F. 1993. Intraspecific and cross-superorder responses to chemical alarm signals by brook stickleback. Ecology 74, 2395-2404.

Rasa, A. 1983. Dwarf mongoose and hornbill mutualism in Taru Desert, Kenya. Behavioral Ecology and Sociobiology, 12, 181-190.

Satischandra, S. H. K., Kodituwakku, P., Kotagama, S. W., & Goodale, E. 2010. Assessing “false” alarm calls by a drongo (Dicrurus paradiseus) in mixed-species bird flocks. Behavioral Ecology, 21, 396-403.

Leave a Reply

eight − = 1