About the SA Blog Network

Tetrapod Zoology

Tetrapod Zoology

Amphibians, reptiles, birds and mammals - living and extinct
Tetrapod Zoology Home

Life and times of the wild Axolotl

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

There haven’t been enough lissamphibians on Tet Zoo lately. So here’s a recycled section of text on axolotls, originally from a 2008 ver 2 article. I haven’t updated it properly, but I have added a new section of text at the end.

Captive Axolotl, image by LoKiLeCh, licensed under Creative Commons Attribution-Share Alike 3.0 Unported license.

Thanks to its perennial use in the pet and laboratory industries, the Axolotl Ambystoma mexicanum is one of the world’s most familiar lissamphibians. It’s certainly the most familiar member of its specific salamander group, the Ambystomatidae or mole salamanders [photo above by LoKiLeCh]. As everyone knows, the Axolotl is neotenous: it retains juvenile characters into sexual maturity, and even an adult is essentially an unmetamorphosed larva, exhibiting a perennibranchiate morphology (meaning that external gills are obvious). Unlike a metamorphosed salamander, a neotenous one retains external gills and a tail and body fin, lacks moveable eyelids, and of course is obligatorily aquatic.

Wild Axolotl, photo (c) Richard Griffiths, from EDGE Amphibians.

Little known is that the wild Axolotl population is restricted entirely to Mexico’s Lake Xochimilco and Lake Chalco, and here they are critically endangered thanks to the pollution, habitat degradation and water extraction and resulting desiccation that has occurred. Use of the animals for food and medicinal purposes, and predation from introduced fish have also contributed to Axolotl decline, and a 2004 survey failed to capture any in areas where they had formerly occurred. They aren’t thought to be wholly extinct, however, as specimens continue to show up in local markets [UPDATE: this was written in 2008 and may not be true as of 2012. Anybody know of more recent Axolotl captures or sightings?]. A. mexicanum is one of the larger Ambystoma species, with some individuals reaching 45 cm in total length (although this is exceptional and half this length is more normal). [Adjacent photo (c) Richard Griffiths, from the Axolotl page at EDGE Amphibians. Image below by ZeWrestler.]

Captive axolotls of the typical pale sort, photographed at Vancouver Aquarium by ZeWrestler; image licensed under Creative Commons Attribution 3.0 Unported license.

The word ‘axolotl’ (‘ajolote’ in Spanish*) is usually thought to mean ‘water dog’ (it is probably not coincidental that other, non-ambystomatid perennibranchiate salamanders are today termed ‘mudpuppies’ and ‘waterdogs’). It also has links to the concept of ugliness or monstrousness and is also often translated as ‘water monster’ (Smith 1969). All neotenous mole salamanders were originally termed ‘axolotls’, as were the larvae of metamorphosing species like the Tiger salamander A. tigrinum. Indeed, this is still the case among laypeople (some people even, apparently, use ‘axolotl’ as the name for any larval salamander). Herpetologists, however, have restricted the name to the perennibranchiate species A. mexicanum. Individuals of this species have been well known to scientists since 1864 when a shipment was sent to Paris. Lake Xochimilco animals were again sent to Paris in 1868, and these Parisian Axolotls were the ancestors of virtually all of the laboratory Axolotls bred forever afterwards.

* Which confuses things, because this name is today used for Bipes, the limbed amphisbaenians.

Ambystoma tigrinum, a mole salamander no longer regarded as especially close to A. mexicanum. Photo by Carla Isabel Ribeiro, licensed under Creative Commons Attribution-Share Alike 3.0 Unported license.

Early confusion as to whether perennibranchiate individuals should be given their own species or genus relative to metamorphosed individuals, and arguments as to whether A. mexicanum should be regarded as conspecific with the metamorphosing Tiger salamander, mean that the Axolotl has a complex nomenclatural history involving several appeals to, and actions of, the ICZN. Oh, and, remember that the ‘Tiger salamander’ of tradition isn’t one species, but a species complex, the members of which are not all close relatives (Shaffer 1993, Shaffer & McKnight 1996). Some of the animals involved may be “among the most recently derived vertebrate species known” (Shaffer & McKnight 1996, p. 430). Incidentally, the ‘Tiger salamanders’ most closely related to A. mexicanum are not part of the lineage associated with the name A. tigrinum: instead, the Plateau tiger salamander A. velasci is the Axolotl’s closest living relative.

The Axolotl’s distinctive morphology and highly endangered status mean that it’s among the top 100 amphibians on the EDGE list: EDGE species are Evolutionarily Distinct and Globally Endangered. The EDGE website on globally endangered amphibians can be found here, and here you can find a huge amount of information on endangered mole salamanders and other species. Nine mole salamanders are on the EDGE list, and several are poorly known, neotenous, Axolotl-like species.

A big surprise from the world of unisexual mole salamanders

California tiger salamanders (A. californiense), Barred tiger salamanders (A. mavortium), and hybrids between the two. Neither species is relevant to the discussion here, but this is a nice picture.

While I’m here, it’s worth discussing something new on mole salamanders that’s been discovered since I wrote the text above. Several Ambystoma species are unisexual – that is, they consist only of parthenogenetic females. Unisexuals always seem to be of hybrid origin; in these cases, combining mtDNA of one species (A. barbouri) with nuclear material from four bisexual ancestors (A. laterale, A. jeffersonianum, A. texanum and A. tigrinum). Some of the unisexuals are diploid, others are triploid, others tetraploid, and others pentaploid!

Because unisexual species are lower in genetic diversity than bisexual ones, and because deleterious mutations accumulate in their genomes (these don’t get purged every time meiosis occurs), a fair and apparently correct assumption is that unisexuals work on a risky, ‘boom and bust’ evolutionary strategy of short-term success, thriving in habitat edges where bisexual species don’t have an ecological advantage.

Parthenogenesis in unisexual species is perhaps best known for certain whiptail lizards (Cnemidophorus sensu lato). This photo shows C. inornatus, C. neomexicanus and C. tigris: C. neomexicanus has arisen through hybridisation between the other two. Photo (c), by Alistair J. Cullum.

Based both on this logic, and on the genetic history of unisexual populations themselves, it has near-universally been thought that unisexual populations are short-lived and unable to persist as long as bisexual ones (Tet Zoo trivia: you may remember me using this argument when responding to the hilarious suggestion that the Loch Ness Monster might be parthenogenetic [it's here]). However, genetic data from unisexual mole salamander species has challenged this view: the amount of sequence divergence between unisexuals and their closest relatives indicates that the unisexuals have been in existence for 2, 3, or even as long as 5 million years or so (Bi & Bogart 2010). In other words, they’ve been around for a long time and cannot be considered short-lived entities.

How, then, are these animals able to avoid the problems otherwise associated with unisexual reproduction? The answer seemingly lies in the fact that they are not truly parthenogenetic – despite being unisexual, members of the salamander species concerned still mate with males of other species: note to fertilise their eggs, but to trigger the start of embryonic development. This is thus a form of parthenogenesis, but it’s a ‘leaky parthenogenesis’ (Lampert & Schartl 2010).

The biology and evolution of unisexual mole salamanders has become a contentious and much-debated subject in lissamphibian research and my coverage of it here is but fleeting and cursory – note that several key papers on the subject are open-access.

For previous Tet Zoo articles on salamanders, see…

Refs –

Bi, K. & Bogart, J. P. 2010. Time and time again: unisexual salamanders (genus Ambystoma) are the oldest unisexual vertebrates. BMC Evolutionary Biology 2010; 10: 238  doi: 10.1186/1471-2148-10-238

Lampert, K. P. & Schartl, M. 2010. A little bit is better than nothing: the incomplete parthenogenesis of salamanders, frogs and fish. BMC Evolutionary Biology 2010; 8: 78  doi: 10.1186/1741-7007-8-78

Shaffer, H. B. 1993. Phylogenetics of model organisms: the laboratory axolotl, Ambystoma mexicanum. Systematic Biology 42, 508-522.

- . & McKnight, M. L. 1996. The polytypic species revisited: genetic differentiation and molecular phylogenetics of the tiger salamander Ambystoma tigrinum (Amphibia: Caudata) complex. Evolution 50, 417-433.

Smith, H. M. 1969. The Mexican axolotl: some misconceptions and problems. BioScience 19, 593-597, 615.

Darren Naish About the Author: Darren Naish is a science writer, technical editor and palaeozoologist (affiliated with the University of Southampton, UK). He mostly works on Cretaceous dinosaurs and pterosaurs but has an avid interest in all things tetrapod. His publications can be downloaded at He has been blogging at Tetrapod Zoology since 2006. Check out the Tet Zoo podcast at! Follow on Twitter @TetZoo.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 21 Comments

Add Comment
  1. 1. vdinets 3:31 pm 12/30/2012

    One interesting thing about axolotls is that they can be easily induced to undergo metamorphosis. Here’s one account:

    Link to this
  2. 2. BrianL 5:26 pm 12/30/2012

    Something I’ve wondered: Surely, given how common axolotls (and tiger salamanders) are worldwide in captivity and that they are not overly difficult to keep and breed plus the general ease with which herps end up getting released, one would expect there to exist quite a few feral/naturalised populations of the big lugs. Is that indeed the case? I often get the impression that such non-native populations of lissamphibians tend to go under-reported and understudied unless they’re red-eared sliders, cane toads or American bullfrogs.

    Link to this
  3. 3. vdinets 5:57 pm 12/30/2012

    BrianL: there are introduced populations of tiger salamanders from eastern US all over the West; that’s a major threat to California tiger salamander.

    Link to this
  4. 4. Jerzy v. 3.0. 6:34 pm 12/30/2012

    A bit shame that huge captive population of axolotls is not maintained with conservation in mind. Captive animals are inbred, most are albinos etc…

    Link to this
  5. 5. jbrougham 7:00 pm 12/30/2012

    Wow, fascinating article. I did not know that they were tiger salamanders. As for their name referring to “dog”, that reminded me that dogs and axolotls were both prized for food by the Aztecs. Maybe the Nahuatl word is the equivalent of “hot dog”. One last bit of trivia: When Trotsky came to Mexico to visit Diego Rivera they, along with Andre Breton, apparently went fishing for axolotls. Now that would be a surreal afternoon. You can find photos of the fishing party on the internets.

    Link to this
  6. 6. souhjiro 10:00 pm 12/30/2012

    About the possible feral populations of axolotls, if they can resist other introduced species, such red crayfish, trout or tilapia, they can exist as “microcryptids” on many places.How can these populations be detected?

    Link to this
  7. 7. BrianL 10:11 am 12/31/2012

    What do you mean by ‘microcryptids’ and does the ‘micro’-part have anything to do with the presence of the fish species you mentioned?

    I was mostly referring to naturalised populations outside the USA. Surely there must exist/have existed some in Europe?

    Link to this
  8. 8. jwmeritt 10:52 am 12/31/2012

    Three legs is interesting, but do all of them show bilateral symmetry>

    Link to this
  9. 9. naishd 12:07 pm 12/31/2012

    Thanks for the great comments. On feral axolotls, usual port of call for this sort of thing is Christopher Lever’s Naturalized Reptiles and Amphibians of the World. I don’t own a copy, but part of it is available online at google books. Therein we find the rumoured presence of naturalised colonies in the Waitako region of New Zealand. As of 1996, it was intimated that some of the animals might be breeding.

    I haven’t heard of any other feral colonies.


    Link to this
  10. 10. Acoyauh2 12:11 pm 12/31/2012

    As of 2011, few sightings have been recorded in the wild in Xochimilco and none that I know of in Chalco. The species has been introduced in other bodies (Chapala, Cuitzeo, Madín, and Tequesquitengo) although no monitoring seems to be maintained for those.

    Regarding the Nahuatl name, the meaning is actually ‘water monster’. The reference to ‘water dogs’ comes somewhat misgudedly from the xoloitzcuintle dog breed, but this name actually derives from Xolotl, the monster god that takes the dead to the underworld – xoloitzcuintles were believed to be Xolotl’s helpers.

    @ 4. Jerzy v. 3.0.: ‘Albinos’ or pale-colored are the most common ocurrence of the wild axolotl – they are not caused by the inbreeding in captivity. However, your concern does make sense – we don’t know the genetic variey (or concern for) of the lab axolotls going around. One of the many unmonitored things regarding the species that should be addressed given their conservation status in the wild.

    Link to this
  11. 11. Heteromeles 12:54 pm 12/31/2012

    The whole build-up of deleterious alleles thing does seem suspect to me. If they are that deleterious, how does the individual expressing them pass these genes on?

    Of course I’m biased. I worked on the glomeromycota years ago. It’s still unclear how (or even if) they have sex, and they’re among the most common fungi on the planet. This tends to make me look for holes in any argument that says syngamy is always best, period.

    As my old plant ecology teacher put it, when you look at a 10,000 year old aspen clone, or a 15,000 year-old creosote clone, it’s really hard to say these particular organisms are unsuccessful.

    There’s a whole problem here with frame shift in arguments about fitness. Organisms are successful in a whole variety of ways. Aspen clones, for example, are forests, a few of which are over a kilometer wide. Are they unsuccessful? If one organism has 10000 offspring, is it more successful than an organism that has two? What if, of those 10,000 offspring, only one survives to adulthood, while the organism that has two offspring raises both to adulthood? The problem here is that success if framed in three different ways: ecological dominance, number of offspring produced, number of offspring surviving to adulthood. These aren’t comparable, but too often, we try to compare across frames anyway, or (worse) deliberately shift the frame-of-reference to advance an argument in the face of the evidence.

    Good thing we still have axolotls to study.

    Link to this
  12. 12. BrianL 1:19 pm 12/31/2012

    Thanks for the information. Apparently then, I was wrong in assuming that there must be feral *Ambystoma* in Europe (and elsewhere outside the USA).

    Another question now: While I know that olms, mudpuppies and sirens are permanently neotenic in the sense that they can’t be ‘forced’ to undergo full metamorphosis in any (known) way, is that because their version of neoteny is fundamentally different from that found in, say, axolotls or montane populations of Alpine Newt or not? In the same veign, should we expect them to have evolved from non-neotenic salamanders or could their neoteny actually be the plesiomorphic condition for caudatans sister to the hynobiid-cryptobranchid clade?

    Link to this
  13. 13. souhjiro 3:47 pm 12/31/2012

    As “microcryptids” i mean “small sized unnoticed animals”(not sure if is a correct use of the word), the other species mentioned could are preventing axolotls from forming viable feral populations.The cryptic nature of the axolotl will render any feral population very hidden even from standard surveys.

    Link to this
  14. 14. Heteromeles 4:19 pm 12/31/2012

    If someone wanted to survey for salamanders, surely it’s possible by now to either come up with a PCR-based water sample, or possibly even something using a fluorescent marker? Then all you would need to do is to collect water samples to determine the presence or absence of various species from water bodies.

    Link to this
  15. 15. David Marjanović 7:08 pm 12/31/2012

    There haven’t been enough lissamphibians on Tet Zoo lately.

    There also haven’t been enough *coughtemnospondylscough* on Tet Zoo lately. :-}

    Never heard of feral axolotls. What kind of temperature range do they need?

    could their neoteny actually be the plesiomorphic condition for caudatans sister to the hynobiid-cryptobranchid clade?

    Good question, actually. The ability to undergo neoteny definitely is plesiomorphic for Caudata (and not just Urodela, its crown-group).

    Three legs is interesting, but do all of them show bilateral symmetry>

    …What? ~:-|

    Link to this
  16. 16. souhjiro 10:09 pm 01/2/2013


    In my experience, axolotls grow and breed pretty well on water temperatures from 9 to 19⁰C, the average temperatures now on Valley of Mexico being 12 to 15⁰C on average, perhaps they lived along temps below freezing point on the last glacial period

    Link to this
  17. 17. Chelydra 3:14 pm 01/5/2013

    My understanding of unisexual Ambystoma is that they could be thought of as a single species, or don’t fit the vertebrate concept of species at all by essentially constantly changing species. They share unique mitochondrial DNA, closest to A. barbouri, but always possess at least one set of chromosomes from A. laterale instead. They are mostly triploids, with tetraploids being frequent. The diploids are rare (and sterile?) and sometimes male. Pentaploids I believe have only been lab-raised from wild eggs, and exhibit some abnormalities. The eggs are easy to ID in the wild, as they’re usually mostly inviable, but those that live hatch out bigger and are apparently good competitors. It’s typical for individuals in a population to be of similar genotype, with two (out of three) sets of chromosomes from the most common normal species in the same pond. The going theory is that the offspring frequently incorporate a new full set of chromosomes from a normal male, dumping one of the maternal sets (or not). This way, when dispersing to a new location, within a few generations they essentially become the same species as the most common normal Ambystoma at that site. However, individuals from some localities have had one set from each of four species! They overlap with A. maculatum across their entire range, but apparently never involve its DNA.

    Link to this
  18. 18. David Marjanović 10:23 am 01/7/2013

    That’s impressive!

    The eggs are easy to ID in the wild, as they’re usually mostly inviable, but those that live hatch out bigger and are apparently good competitors.

    Makes sense; more DNA = bigger cells.

    Link to this
  19. 19. farandfew 6:54 am 01/14/2013

    Heteromeles: “If someone wanted to survey for salamanders, surely it’s possible by now to either come up with a PCR-based water sample,”

    Indeed. see

    Thomsen, PF , Kielgast, J , Iversen, LL , Wiuf, C , Rasmussen, M , Gilbert, MTP , Orlando, LAA & Willerslev, E 2012, ‘ Monitoring endangered freshwater biodiversity using environmental DNA ‘ Molecular Ecology , vol 21, s. 2565-2573.

    I believe this approach is currently being considered for at least one endangered salamander species. Whether for axolotls, I do not know.

    Link to this
  20. 20. amkim 3:50 pm 01/16/2013

    “I used to love axolotl tamales,” he says, rubbing his stomach and laughing. — Xochimilco fisherman Roberto Altamira

    Link to this
  21. 21. souhjiro 6:44 pm 02/3/2013

    a link about current wild axolotl conservation efforts

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article