In the last few months, two things happened that got me to thinking about salamanders. First, a friend adopted an arboreal salamander she found ascending a set of concrete steps in her treeless apartment complex. It’s now ensconced in a terrarium, with all the earthworms and crickets it can eat. I’m told it zaps them with an impressively extensible tongue, though never when I visit.
Then I bought the third edition of Robert L. Stebbins’ Field Guide to Western Reptiles and Amphibians. I was struck by one change in particular. My 1966 Stebbins describes three species of slender salamanders, small amphibians of retiring habits, often found under boards or rocks in urban backyards. The new edition has 20 species and hints there might be even more. Clearly the taxonomists had been busy.
Some of the new ones have been discovered in out-of-the-way places, like springs and seeps in the Inyo mountains. Most, though, are “splits,” formerly included in two widespread species, the California and Pacific slender salamanders. Genetic research has proved them to be members of distinctive lineages, separated from their nearest relatives by chasms of geologic time.
My curiosity led me to a 1986 article by Dean Hendrickson in the journal Cladistics, with the daunting title of “Congruence of Bolitoglossine Biogeography and Phylogeny with Geologic History: Paleotransport on Displaced Suspect Terranes?” It also led me to David B. Wake at UC Berkeley’s Museum of Vertebrate Zoology, who has spent most of his career untangling the evolutionary history of these obscure creatures. The story he and his co-workers have pieced together is a remarkable one.
Most of North America’s salamanders belong to the family Plethodontidae, the lungless salamanders. Under circumstances that are still controversial — maybe they lived in rushing mountain streams, where buoyancy would increase the risk of being swept away — their ancestors dispensed with lungs. Modern plethodontids take in oxygen through their skins or gulp air into their throats. Judging from the number and variety of species there, lungless salamanders may have evolved in the Appalachian region.
My friend’s arboreal is the standard-model plethodontid. It has its quirks — lacking vocal cords, an arboreal salamander can produce an irritated squeak by retracting its eyeballs into their sockets and forcing air through its mouth — but it’s conventional compared to the bolitoglossines.
Bolitoglossines (“mushroom-tongued”) are a subgroup of lungless salamanders, the only ones to have established a foothold in the New World tropics, with a puzzling distribution. There are none in the Appalachians, the putative ancestral home of the whole family. There’s a raft of species in Central and South America. The West Coast has at least 23. All those slender salamanders, 19 in California and one in Oregon, are in the genus Batrachoseps. California also has three species of web-toed salamanders, genus Hydromantes. And their only other relatives are in southern Europe: southeastern France, Sardinia, and northern Italy. (No fossils, either; salamanders have not left much of a fossil record, so their evolutionary history has been inferred from morphological and genetic studies.) What could have accounted for such a scattered distribution?
There are other odd things about these salamanders. The eight species of Hydromantes are wall-crawlers, bracing themselves with their tails as they make their way over granite and limestone cliff faces. Sensitive binocular vision helps them spot prey in the dimmest light. When threatened, one California Hydromantes, the Mount Lyell salamander, coils itself into a ball and rolls away from danger.
Then there are the tongues. The Sardinian cave salamander has evolved a ballistic mechanism that fires its tongue from its mouth to hit insects a couple of inches away. At full stretch, tongue extension is close to 80 percent of the amphibian’s body length. I don’t know if comparable studies have been done on the California species, but they have similar equipment.
The slender salamanders seem less specialized. All 20 Batrachoseps species are small (less than three inches long) and, well, slender. They’re often mistaken for earthworms. They coil up tightly when alarmed, and when picked up may suddenly uncoil, slip from their captor’s hand, and bounce away. Unlike other lineages where species have differentiated in shape and behavior to exploit different niches, the slender salamanders haven’t diverged far from a common ancestral pattern. They all look pretty much alike and have much the same behavior, living under things and feeding on snails and small insects.
Yet something strange lies within the genes and the nervous systems of both the slender and web-toed salamanders. The Italian cave salamander may have the largest genome of any terrestrial animal. You wouldn’t think building a salamander would require such a thick instruction manual, and indeed there’s a lot of non-coding “junk” DNA. Along with the big genome, and possibly related to it, goes a simplified brain and stripped-down neural wiring — some salamander brains contain fewer neurons than a honeybee’s. We think of evolution as a process of increasing complexity, in both structure and behavior. But bolitoglossine salamanders seem to have evolved away from complexity.
Back to the geography, though. Dean Hendrickson tried to explain the peculiar distribution of these salamanders in terms of movements of the earth’s crust. If you’ve read John MacPhee’s Assembling California, you know about suspect terranes. Like the drifting continents, these smaller bits of crust, also called microplates, are propelled by the planet’s inner fires. Some fetch up against continental margins, building them outward. Much of California is a mosaic of exotic terranes.
Hendrickson believes microplate movement is a reasonable way of getting bolitoglossine salamanders from their presumed Appalachian point of origin to where they’re found today; he can even correlate tectonic movements with branching points in salamander evolution. It’s tempting to use microplate movement as the key to the California-plus-Mediterranean pattern of Hydromantes. If the European species were separated from their North American kin by the widening North Atlantic, and the ancestral North American species rode south on the Maya terrane (now a chunk of the Yucatan), and fragments of that took their descendants to the western edge of North America.
But it’s a beautiful theory undone by ugly little facts. When Wake and his colleagues looked at genetic variation among Hydromantes, they found that the European and Californian lineages must have diverged about 28 million years ago — much too recently for the terrane-riding scenario. The implausible-sounding alternative is that these sedentary creatures — many wander no more than a few yards from their natal spots — made their way from wherever in North America they first evolved, across the Bering bridge into Asia and on down into southern Europe, finding suitable habitat all along the way.
The slender salamanders do appear to have been crustal hitchhikers. Wake and former student Elizabeth Jockusch have concluded that several species now inhabiting the central Coast Ranges evolved somewhere to the south and were transported when crustal fragments moved north along the San Andreas Fault. Something like this may also have happened in the southern Sierra Nevada, home to another complex of salamander species, although the geologic history of this region is less well understood. (Tectonic movements may also explain the presence of arboreal salamanders on the virtually treeless Farallon Islands.)
Wake and his students are looking at the mitochondrial DNA of Hydromantes, Batrachoseps, and their tropical relatives. “One thing is clear,” Wake says: “California has been the staging ground for a great deal of radiative evolution in Batrachoseps, and this has been going on for a very long time.” Something to think about the next time you find one of these unassuming creatures under a flowerpot.
