UC Berkeley Press Release
(Rachel Mueller/UC Berkeley)
Genetic analysis rewrites salamander family tree
BERKELEY – Biologists take for granted that the limbs and branches of the tree of life - painstakingly constructed since Linnaeus started classifying organisms 270 years ago - are basically correct. New genetic studies, the thinking goes, will only prune the twigs, perhaps shuffling around a few species here and there.
Hence the surprise when a new University of California, Berkeley, study of the largest family of salamanders produced a genetic family tree totally inconsistent with the accepted classification, which is based primarily on physical features.
Salamanders formerly classified together because of similar characteristics, such as a tail that breaks at only one spot as opposed to anywhere when stressed, now appear not to be close relatives at all. And salamanders that go through an aquatic larval stage are scattered about on different branches instead of grouped on one limb of the tree: Apparently some salamander lineages lost the larval stage and then reacquired it again.
"For 40 years, we have had a very clear understanding of the evolutionary history of the largest family of salamanders, Plethodontidae," said David Wake, professor of integrative biology at UC Berkeley and an expert on salamanders. "We thought they arose in Appalachian mountain streams and then diverged in a highly patterned way, sequentially abandoning larvae for direct development, gaining highly specialized, projectile tongues, et cetera."
"The results were stunningly different than what we anticipated," he said. "Only one of the currently recognized four major groups is supported."
The study, published this week in the online edition of the Proceedings of the National Academy of Sciences, was conducted by Wake's graduate student, Rachel Mueller, to understand the evolution of the Plethodontid salamanders, a family that comprises 360 species - two-thirds of the world's 522 known species of salamander. Known for being one of few landlubbing vertebrates without lungs - they breathe through their skin - Plethodontids were thought to have originated in the Appalachians because the southern portion of that chain has the greatest diversity of species.
The new family tree, constructed by comparing the mitochondrial genomes of 22 representative Plethodontid species and five others from different salamander families, offers no support for the out-of-Appalachia theory, Mueller said.
"We can infer only a North American origin," Mueller said. "Most likely, where these species are now doesn't relate to where they were ages ago, because the climate and geology have changed so much, and the species have moved around."
Though results from this one family of vertebrates can't necessarily be generalized to other families, Mueller said, "this does tell us that, when reconstructing evolutionary relationships, you have to be careful which morphological features you assume are conservative and haven't evolved much, and which you think are likely to have changed over time."
Three years ago, Mueller teamed up with the Joint Genome Institute (JGI) in Walnut Creek, Calif., to sequence the complete mitochondrial genomes of 24 salamanders, all but two of them Plethodontids. The mitochondrial genome is smaller than the nuclear genome, containing only 37 complete genes in most animals - just those needed to produce energy for the cell. They're inherited without recombination from the mother, and therefore represent an unbroken female chain going back perhaps 100-150 million years in plethodontid salamanders. By looking at mutations in the mitochondrial DNA, biologists can infer the pattern of evolution.
"What is happening now is that by focusing on mitochondrial genes and adding particular nuclear genes, we really are teasing apart issues related to our understanding of the tree of life," Wake said.
Mueller obtained DNA samples from frozen salamander tissue in the collection of UC Berkeley's Museum of Vertebrate Zoology, and over a two-year period worked with JGI researchers to sequence the brief genomes. Once Mueller annotated the genomes, that is, identified the genes, she compared the complete genome sequences with three other known mitochondrial genomes, for a total sample of 27 species. Using Bayesian statistical techniques, she was able to reconstruct a "robust" family tree of the Plethodontid salamanders.
One feature that has been used to classify salamanders is their larval stage, much like the tadpole stage of frogs. The ancestral Plethdontid salamanders, like all other salamanders, apparently went through a larval stage in fresh water, where the eggs are laid, before losing their gills and swimming tails and emerging onto land. Salamanders that evolved to live totally out of the water, however, lost the larval stage, emerging from the egg as a smaller version of an adult.
The new family tree shows, however, that some terrestrial salamanders regained their larval stage after moving back to the water. This may have happened in three separate lineages of Plethodontids, which is surprising for a seemingly complex feature biologists have assumed arose just once, very early in the history of salamanders.
Tongues also are distinctive in plethodontid salamanders, coming in three different types: ones attached to the jaw so they stick out only slightly; tongues attached by a stretchy muscle, which allows them to be thrown out an appreciable distance to snag food; and so-called "ballistic" tongues that have no muscular attachment at the tip and can be hurled out nearly the length of the salamander's body. Wake has shown in previous studies that these tongues are not a good way to classify salamanders, because the various types of tongues have evolved several times in different lineages. The new data confirm that, Mueller said.
Mueller's coauthors, in addition to Wake, include Jeffrey L. Boore, adjunct professor of integrative biology at UC Berkeley and director of evolutionary genomics at JGI; herpetologist and evolutionary biologist Robert Macey of JGI; and former visiting student Martin Jaekel. Mueller, Wake, Macey and Jaekel also are associated with UC Berkeley's Museum of Vertebrate Zoology.