In an effort to get through all the blog posts I’ve started but have yet to finish, I thought I may as well start with this one on, of course, plethodontid salamanders (aka lungless salamanders). It started life as part of the same article as the ver 1 post here: this was essentially an introduction to plethodontid diversity and phylogeny. In particular I waxed lyrical about the huge number of new species that have been named since 1985 (many of which come from well-studied regions of North America), and I frothed at the mouth with excitement over the recently described Korean crevice salamander Karsenia koreana Min et al., 2005, the very first plethodontid to be reported from Asia. Here, we look at two particularly interesting recent discoveries about plethodontids. Whether you hold salamanders in any special esteem or not, you are guaranteed to find the following interesting…

[adjacent photo, showing the plethodontids Pseudoeurycea leprosa (on the left) and Lineatriton lineolus is by Gabriela Parra-Olea/UC Berkeley, and is borrowed from here]

All salamanders are carnivorous. Or are they? During study of the cave-adapted plethodontid Eurycea spelaea (until recently given its own genus, Typhlotriton), Fenolio et al. (2006) repeatedly observed salamander larvae ingesting bat guano. Yes, bat shit. Adult cave salamanders were observed foraging on guano piles, though weren’t seen to eat the stuff. Because salamanders are notorious for accidentally ingesting bits of plant material, silt and other detritus accidentally, Fenolio et al. (2006) used both isotope analysis and nutritional study to work out whether ingestion was useful to the salamanders, and by inference deliberate. It turned out that the C and N isotopes present in both the guano and salamander tissues were highly similar, and that the salamanders were therefore assimilating the dung they were eating.

Furthermore, bat poo is – comparatively – quite nutritious and similar in protein content and calorific value to the McDonald’s Corporation Big Mac sandwich Fenolio et al. (2006) compared it with. Insert joke about junk food here. The short gut carrying time of the bats that produced the guano (Myotis grisescens, the Grey bat) means that their dung is quite rich in calories and nutrients. Because the bats and their resulting fresh guano are not a permanent feature of the relevant caves (they use them as summer maternity roosts), it is assumed that the salamanders exploit guano as a temporary resource in an otherwise nutrient-poor system. Coprophagy is known elsewhere in lissamphibians – tadpoles do it of course to inoculate themselves – but Fenolio et al.’s (2006) study is the first to document this behaviour in a salamander. They speculated that coprophagy might prove widespread in subterranean vertebrates. And of course this brings us back to olms [covered on ver 1 here]. You can download a free pdf of Fenolio et al. (2006) from Fenolio’s site here [adjacent image shows Eurycea spelaea].

The evolution of fossoriality and extreme morphological and ecological homoplasy

Many tropical plethodontids are fossorial, and it’s been argued that the evolution of fossoriality has been one of the key events that allowed this group to invade the lowlands of the American tropics (a region difficult for salamanders to colonise). In general, fossorial plethodontids are characterised by reduced limbs and an elongate body where there are 18-22 dorsal vertebrae as opposed to the ordinary 14 seen in other plethodontids. However, one elongate-bodied fossorial plethodontid, the bizarrely attenuate, miniaturised Mexican slender salamander Lineatriton lineolus, is special in that, while possessing the primitive count of 14 dorsal vertebrae, it’s the individual vertebrae that have become tremendously narrow and elongate. Parra-Olea & Wake (2001) termed the evolutionary mechanism behind this morphology the ‘giraffe-neck’ developmental program. It shows that plethodontids have used two different evolutionary strategies to evolve the long-bodied fossorial morphotype. But wait, there’s more.

Imagine for a moment that only the morphology of Lineatriton was studied. Because of its highly autapomorphic morphology, I don’t think we’d ever question the monophyly of this taxon. This is where the results of genetic analyses can be truly amazing and astounding. For, after including Lineatriton within a new molecular analysis (incorporating mtDNA from three genes), Parra-Olea & Wake (2001) found that Lineatriton is not monophyletic. Different populations fell in different places within Pseudoeurycea, with those from SE Veracruz (Mexico) forming the sister-taxon to a P. werleri + P. mystax clade, and those from central western Veracruz forming the sister-taxon to the clade that includes P. leprosa. If you’re sceptical of this and are wondering if there might be other explanations behind this discovery [e.g., mitochondrial polymorphism or hybridization and later introgression], check the paper, as Parra-Olea & Wake show why these can be discounted (to explain this in detail would essentially mean reproducing their paper in this article, and I ain’t doing that).

So what was thought to be one highly autapomorphic taxon is, in fact, a convergently similar assemblage of different Pseudoeurycea populations. Given that the Pseudoeurycea species closest to the ‘Lineatriton‘ populations are stout-bodied non-fossorial plethodontids, an identical morphological solution has been evolved on more than one occasion. The reason that the products of this convergence are so stunningly alike is, presumably, due to the fact that the ancestral bauplan was essentially identical. Exactly this conclusion was reached in another study of convergence in plethodontids: Wiens et al. (2003) found that convergence was most likely to go undocumented in clades where specialised species evolved independently from ancestors that were already highly similar [adjacent image shows Pseudoeurycea smithi. Image © 1975 David Wake].

Of course, this doesn’t mean that Lineatriton lineolus doesn’t exist, as one of the populations that’s always gone by this name is still a distinct taxonomic entity. But which population was that? This is tricky to resolve as Cope’s original 1865 description* was vague on the type locality. It seems most likely that the central western Veracruz population represents the ‘real’ Lineatriton, in which case those from elsewhere will now need new names. However, even this probable type population contains so much genetic divergence that it likely represents more than one taxon, meaning that one specific population within the Veracruz population will have to be defined as the type population for L. lineolus proper. A revised nomenclature is clearly going to be needed for the giraffe-necked Pseudoeurycea populations that are not conspecific with the ‘true’ L. lineolus population, and I regret that I haven’t checked lately to see if this has been sorted out. Two new Lineatriton species were named by Brodie et al. (2002), but these were (I think) Lineatriton proper (i.e., close kin of L. lineolus), and thus don’t help resolve the situation. Do let me know if you know otherwise.

* I think I’ve now said this several times but… for those who only know Cope as a describer of Mesozoic reptiles, you might like to know that he was also a prolific describer of extant fishes, lissamphibians and reptiles.

Finally, Parra-Olea & Wake (2001) note that the ‘Lineatriton phenomenon’ appears to have been local and limited, only occurring in a few ‘terminal twigs in the bolitoglossine radiation’ (p. 7891). In contrast, the evolution of an increased vertebral count was, so far as we know, a far more successful strategy, as Oedipina – the fossorial long-bodied taxon characterised by its 18-22 (as opposed to 14) trunk vertebrae – occurs from southern Mexico, throughout Central America, and into Colombia and Ecuador, inhabits habitat from sea level to over 2500 m, and includes over 20 species. Different evolutionary strategies, similar morphological results, radically different results [adjacent image shows Oedipina taylori. Image copyright © 2006 Vladlen Henríquez].

Oh – and finally finally, I can’t leave without mentioning the fifth conference on the Biology of Plethodontid Salamanders, happening on August 3rd-6th at San Cristobale de las Casas, Chiapas, Mexico. Previous meetings happened in 1972, 1982, 1992 and 1998. Go here for the conference homepage.

Refs – -

Brodie, E. D., Mendelson, J. R. & Campbell, J. A. 2002. Taxonomic revision of the Mexican plethodontid salamanders of the genus Lineatriton, with the description of two new species. Herpetologica 58, 194-204.

Fenolio, D. B., Graening, G. O., Collier, B. A. & Stout, J. F. 2006. Coprophagy in a cave-adapted salamander; the importance of bat guano examined through nutritional and stable isotope analyses. Proceedings of the Royal Society of London B 273, 439-443.

Parra-Olea, G. & Wake, D. B. 2001. Extreme morphological and ecological homoplasy in tropical salamanders. Proceedings of the National Academy of Sciences 98, 7888-7891.

Wiens, J. J., Chippindale, P. T. & Hillis, D. M. 2003. When are phylogenetic analyses misled by convergence? A case study in Texas cave salamanders. Systematic Biology 52, 501-514.