Reptile & Amphibian Forums

NEW SCIENTIST (London, UK) 25 April 08 Salamanders formed new species despite interbreeding (Bob Holmes)
Evolutionary biologists, from Charles Darwin onwards, have believed that isolation plays a key role in the origin of species.
Now a study of cave-dwelling salamanders that have evolved into separate species from their surface-dwelling kin despite regularly interbreeding suggests that isolation is not necessary for speciation.
Cave-dwelling species were thought to be classic examples of how isolation is necessary for speciation. Cave populations isolated underground gradually evolve to lose newly useless structures such as eyes and pigmentation.
But when Matthew Niemiller, an evolutionary biologist at the University of Tennessee in Knoxville, and his colleagues saw hints of hybridisation between Tennessee cave salamanders (Gyrinophilus palleucus) and the surface-dwelling spring salamander (G. porphyriticus), they decided to take a closer look.
Regular mixing
The researchers sequenced DNA samples from 109 cave and spring salamanders from 43 locations throughout Tennessee.
They then plugged the data into a sophisticated computer model that compared possible evolutionary histories for the salamanders and calculated which scenario provided the most likely explanation for the genetic patterns observed.
The results suggested that the cave salamanders could not have evolved in isolation from the surface species. In fact, the most likely history was one in which spring salamanders regularly interbred with cave salamanders even as the two species were diverging about 2 million years ago.
Natural selection in the cave salamander for traits helpful for life in caves – such as sharper non-visual senses and a permanently aquatic lifestyle – must have been strong enough to override this gene flow between the surface and cave animals, says Niemiller.
Applying the same analysis to other cases where species were thought to have evolved in isolation – both within caves and without – is likely to yield more examples, he speculates.
"It's very difficult to show that two divergent forms have experienced gene flow," he says. "It might be more prevalent than we currently recognise."
Common occurrence?
Andrew Hendry, an evolutionary biologist at McGill University's Redpath Museum in Montreal, Canada, agrees that speciation occurs in the face of gene flow much more often than many biologists are willing to admit. Niemiller's study provides further proof, particularly because of the timing of the gene flow.
"What is interesting is that the migration events cluster fairly early on," he says. "This suggests gene flow was initially fairly high, and then as they adapt to the new environment, you get the buildup of barriers due to adaptation, and then gene flow goes down."
However, says Hendry, the computer model at the heart of Niemiller's analysis is still relatively new, he says, and it takes so much computer time to run it that theoreticians have not yet learned all of its limitations.
As a result, Niemiller's study may do little to convince confirmed isolationists, he says.
Journal reference: Molecular Ecology (DOI: 10.1111/j.1365-294X.2008.03750.x)
Salamanders formed new species despite interbreeding

>>NEW SCIENTIST (London, UK) 25 April 08 Salamanders formed new species despite interbreeding (Bob Holmes)
>>Evolutionary biologists, from Charles Darwin onwards, have believed that isolation plays a key role in the origin of species.
>>Now a study of cave-dwelling salamanders that have evolved into separate species from their surface-dwelling kin despite regularly interbreeding suggests that isolation is not necessary for speciation.>>

Of course [geographic] isolation is not necessary for speciation. The cichlid fishes of the African lakes have taught us that a long time ago. That is why the so-called "evolutionary species concept" that is currently in vogue makes no sense. This concept assumes that two geographically isolated populations will eventually become two species, so we might as well recognize them as distinct species now. The problem is that geographical isolation does not necessarily lead to speciation, nor is it necessary for speciation to occur.

I don't think they were referring to geographic isolation, but rather reproductive isolation.

I've been following Niemiller and Miller's work on Gyrinophilus for a while,and it just keeps getting more interesting.

>>I don't think they were referring to geographic isolation, but rather reproductive isolation.>>

It is actually quite obvious they mean geographic isolation.

"Cave populations isolated underground gradually evolve to lose newly useless structures such as eyes and pigmentation."

In that statement, people originally thought that cave salamanders evolved because they were isolated geographically underground. Of course geographic isolation also precludes gene flow. Anyhow, the African cichlids have shown that reproductive isolation can evolve even without geographic isolation, and of course geographic isolation does not necessarily lead to reproductive isolation, contra the assumptions of many. The Eastern tiger salamander (A. t. tigrinum) has the exact same reproductive behavior as the California tiger salamander (A. t. californiense) despite millions of years of geographic isolation between them. Since mating behavior is a very important pre-mating isolation mechanism, the identical mating behaviors insure that there will be no barrier to interbreeding if the gametes are fertile. I read that some people were puzzled why introduced tiger salamanders would interbreed freely with the California tiger salamanders when they are supposed to be different species and therefore reproductively isolated. That is because these tiger salamanders do not even know that they are supposed to be different species. One way they can tell is if their mating behaviors are different. If their behaviors are not different, then they would treat each other as conspecific. So, no matter what the DNA says about their divergence times, they can still be conspecific if neither has changed.

>>I've been following Niemiller and Miller's work on Gyrinophilus for a while,and it just keeps getting more interesting.>>

There are a lot of assumptions made about speciation and gene flow between species and so forth, and these assumptions are getting shattered by new data. It has been known for a long time, for example, that coyotes interbreed with wolves and yet both maintain their distinctness as species. It is now known that the "red wolf" is in fact a hybrid species between the coyote and the Gray Wolf. Even more interesting is the revelation that Eumeces gilberti is not a true species in the classic sense, but an ecological race of Eumeces skiltonianus. Different E. gilberti populations are derived from different E. skiltonianus populations, and yet E. gilberti does not interbreed with E. skiltonianus but different populations of E. gilberti will interbreed among themselves despite their separate origin.

In all these cases, and in the case of the cave salamanders, it shows that natural selection plays a bigger role than gene flow in maintaining species distinctness. If gene flow results in hybrids with phenotypes that are less adaptive to the environment, then these will be weeded out. If not, then a species can and do maintain its species distinctness even if gene flow occurs. My guess is that the selective forces that shaped the cave salamanders would eliminate any hybrids which are phenotypically maladaptive. So, even though gene flow may occur, it may not change an animal's phenotype because of natural selection.

It is actually quite obvious they mean geographic isolation.

"Cave populations isolated underground gradually evolve to lose newly useless structures such as eyes and pigmentation."

Technically this is not geographic isolation, it is habitat partitioning. The animals are well within one another's geographic ranges (they are sympatric or parapatric in many areas) and dispersal areas. Besides, Gyrinophilus porphyriticus (the "surface species" ) is well known to use caves throughout its range.

In that statement, people originally thought that cave salamanders evolved because they were isolated geographically underground. Of course geographic isolation also precludes gene flow. Anyhow, the African cichlids have shown that reproductive isolation can evolve even without geographic isolation, and of course geographic isolation does not necessarily lead to reproductive isolation, contra the assumptions of many. The Eastern tiger salamander (A. t. tigrinum) has the exact same reproductive behavior as the California tiger salamander (A. t. californiense) despite millions of years of geographic isolation between them. Since mating behavior is a very important pre-mating isolation mechanism, the identical mating behaviors insure that there will be no barrier to interbreeding if the gametes are fertile. I read that some people were puzzled why introduced tiger salamanders would interbreed freely with the California tiger salamanders when they are supposed to be different species and therefore reproductively isolated. That is because these tiger salamanders do not even know that they are supposed to be different species. One way they can tell is if their mating behaviors are different. If their behaviors are not different, then they would treat each other as conspecific. So, no matter what the DNA says about their divergence times, they can still be conspecific if neither has changed.

You seem hung up on the idea that the ability to interbreed is proof of conspecificity. This is not a species definition supported by any working biologists that I am aware of, not even the late great Ernst Mayr who first laid out the biological species concept. The fact that gene flow could occur if two populations (let's use your eastern and California tiger salamanders as an example) were sympatric does not reflect the fact that gene flow is not occurring. The two populations are as isolated from another as they are from pigeons or hedgehogs. Of course this reasoning can be taken too far (does each of the many isolated populations of four-toed salamanders deserve species status?), but I think you are in an indefensible position by claiming interbreeding=conspecificity.

Reproductive incompatibility, whether behavioral, physical, or genetic, is often evolved in closely related species that are sympatric to guard against hybridization; there is no reason for allopatrically distributed species to evolve such safeguards. This does not make them conspecific.

>>I've been following Niemiller and Miller's work on Gyrinophilus for a while,and it just keeps getting more interesting.>>

There are a lot of assumptions made about speciation and gene flow between species and so forth, and these assumptions are getting shattered by new data. It has been known for a long time, for example, that coyotes interbreed with wolves and yet both maintain their distinctness as species. It is now known that the "red wolf" is in fact a hybrid species between the coyote and the Gray Wolf. Even more interesting is the revelation that Eumeces gilberti is not a true species in the classic sense, but an ecological race of Eumeces skiltonianus. Different E. gilberti populations are derived from different E. skiltonianus populations, and yet E. gilberti does not interbreed with E. skiltonianus but different populations of E. gilberti will interbreed among themselves despite their separate origin.

It is my impression that the gray wolf/coyote hybrid origin for red wolves is speculative only; if you have links to any research on the issue I would be glad to read it. But certainly it would not be a unique situation; the European edible frog (Rana edulis) is a similar case. The hybrid Ambystoma of the northeast are even more peculiar. They have genomic components from four species, but their MtDNA is unrelated to any of the four; apparently all the 'hybrids' are the descendants of an ancestral stock whose nuclear genome has disappeared!

The Eumeces (I guess it's now Plestiodon) gilberti stuff is very interesting; it must be a fairly simple genetic cause to allow multiple independent origins from the E. skiltonianus stock. The same has been shown in the tetraploid gray tree frog Hyla versicolor.

>>Technically this is not geographic isolation, it is habitat partitioning. The animals are well within one another's geographic ranges (they are sympatric or parapatric in many areas) and dispersal areas. Besides, Gyrinophilus porphyriticus (the "surface species" ) is well known to use caves throughout its range.>>

They may be able to reach each other's habitat now, but it was apparently thought by some that in the past the cave salamanders were geographically isolated and thus evolved their cave adaptations like lost of eyesight and pigment. Once they have evolved these adaptations, the passage to the outside world reopened. But because "speciation" (which means different things to different people) has occurred, the 2 sibling species can no longer interbreed. That is the classic scenario of how isolation leads to speciation and it was probably what many people assume happened prior to the discovery of the amount of gene flow. But of course we now know that geographic isolation is not necessary.

>>You seem hung up on the idea that the ability to interbreed is proof of conspecificity. This is not a species definition supported by any working biologists that I am aware of, not even the late great Ernst Mayr who first laid out the biological species concept.>>

You misunderstand. I do not consider the ability to interbreed as proof of conspecificity. Otherwise, the corn snake could easily be classified as the same species as the common kingsnake, given their ability to produce hybrids in captivity. Instead, I consider the inability to interbreed as evidence that two populations are not the same species. Ernst Mayr points out that it is very easy to tell if two sympatric populations are conspecific, because if they do not interbreed freely, then they are two distinct species. It is much more difficult when two populations are allopatric. In that case the ability to interbreed may or may not be proof of conspecificity.

>>The fact that gene flow could occur if two populations (let's use your eastern and California tiger salamanders as an example) were sympatric does not reflect the fact that gene flow is not occurring. The two populations are as isolated from another as they are from pigeons or hedgehogs. Of course this reasoning can be taken too far (does each of the many isolated populations of four-toed salamanders deserve species status?), but I think you are in an indefensible position by claiming interbreeding=conspecificity.>>

Quite the contrary. According to Mayr, a species can consist of a group of potentially interbreeding populations. A. t. californiense and other subspecies of A. tigrinum clearly fit that description. They both retain the potential to interbreed. OTOH, those who treat geographically isolated populations as distinct species are really in an indefensible position.

>>Reproductive incompatibility, whether behavioral, physical, or genetic, is often evolved in closely related species that are sympatric to guard against hybridization; there is no reason for allopatrically distributed species to evolve such safeguards. This does not make them conspecific.>>

It is true that pre- and post- mating isolation mechanisms often evolve when two closely related species are sympatric with each other. Once again, we can point to the African cichlid fishes. These diverse species, which evolved in sympatry from a common ancestor, also evolved reproductive isolation mechanisms to guard against hybridization because hybrids are less fit. Hence one can infer species status indirectly by evaluating whether hybrids are less fit or not. In some cases, it is easy. For example, Hyla versicolor and H. chrysoscelis crosses are obviously unfit. In other cases it is not as easy. In the case of tiger salamander hybrids, it appears that they are every bit as fit as the native species because the researchers who worked with them fear that these hybrids may become dominant at the expense of the natives

>>It is my impression that the gray wolf/coyote hybrid origin for red wolves is speculative only; if you have links to any research on the issue I would be glad to read it.>>

Here is one link, with references therein.

http://www.montana.edu/~wwwbi/staff/creel/bio480/The red wolf.pdf

>>The Eumeces (I guess it's now Plestiodon) gilberti stuff is very interesting; it must be a fairly simple genetic cause to allow multiple independent origins from the E. skiltonianus stock. The same has been shown in the tetraploid gray tree frog Hyla versicolor.>>

I don't know what the underlying genetic mechanism is, but there is evidence that this ability to give rise to different ecological types exist in the Eumeces species found in Asia.