Monday, February 02, 2009

What the definition of "Species" is...

Therese, who knows about and studies such things, left a long comment on the last post that I think deserves to be up front.

"I’ve got some insight into these questions. I apologize in advance for both these long rants, I’m currently working on some species redescriptions and its not progressing as fast as it needs to be.

"So subspecies are a relic of 1950's species concepts. I know they still teach in biology (even university bio) that for something to be a species it has to be reproductively isolated from other similar groups. Conversely then if two groups are able to reproduce they were the same species. The concept of a subspecies was a way for a taxonomist to be like these things mate just fine, but they look different, have different ranges, etc. Since biological species concept has gone the way of the dinosaurs, in the real world we now operate under about 20 different species concepts and subspecies as such are no longer necessary. Infact what most people think of as subspecies are no more than races. May sound like a wording difference but the two are not even remotely the same. Subspecies, as a taxonomic unit, implies a monophyletic grouping while a race does not. There have been some interesting papers looking at subspecies in birds and in some papers up to 97% of the recognized subspecies looked at are not monophyletic. Most of us these days ask a series of questions before describing a new species or revising old ones. The two I ask as a follower of the phylogenetic species concept are 1) is the group diagnosably different and 2) is it following its own evolutionary path. Question 1 means I need to be able to look at a specimen and say this belongs to morphgroup x. Not based on range, not based on behavior, based on actual morphological/molecular evidence. Question 2 determines if the group is monophyletic or not. If the group is in question meets the 2 criteria it’s a valid species. The fact that its able to interbreed with morphgroup y (say in captivity or in hybrid zones) doesn’t mean anything. Since those incidents are rare they don’t affect the overall evolutionary path of the two species. However if these events become common- say 2 groups of birds that look a bit different but have massive range overlap and commonly interbreed the evolutionary path of the two groups is intertwined and they are not monophyletic in respect to each other. In this instance there is a single valid species. If you like splitting things you can have a few races- group 1 and group 2- but because the two groups arnt monophyletic they aren’t an actual taxonomic unit.

"As a side note a number of the new “insert large mammal/bird” here discoveries announced are elevations of old subspecies as the groups are being revised using modern phylogenetic theory. The Bornean Clouded Leopard (described in 1823 as a subspecies of Clouded leopard) comes to mind- read the Yahoo news story (or any other paper) and they conveniently neglect to mention that detail, after all the headline: Bornean Clouded Leopard now a species 200 years after discovery doesn’t quite have the same ring to it does it?

"As far as reintroductions go I tend to like the peregrine approach used by the p-fund in parts of the US. Use birds from across the entire range of the species and let the best genes win. Overtime the animal will either resemble the original animal (if that morphology is still the best for the environment) or it will create a new morphology to reflect what is now the best. In general reintroductions/restorations are sort of silly. People hate to hear it but landscapes aren’t static, they aren’t naturally so why are we trying to make them static now. I have no problem with reintroductions where there is a single or small group of known causes for the extinction event (peregrine and eggshell thinning comes to mind) but when the reality of the situation is the bird was on its way out anyway, it’s right on the edge of its range to start with, throwing a few million dollars at the problem isn’t going to fix it and that money could have been better spent on habitat protection for other things. Extinction is a natural event and is not uniformly bad. Take prairie chickens for example. I’ve been involved with various aspects of the Attwater’s prairie chicken for about 6 years now- raising birds for release,ooking at insect population at APCNWR, and that sort of thing. APCs are the southern most grouse species period. Grouse are an northern bird, the fact that a group living literally at the edge of where a grouse can survive is basically extinct isn’t surprising. But it makes people feel bad because they look cool and habitat loss is partially to blame so they throw money at it. Large sums of money at it. Hey its paid my salary for a while so I’m a hypocrite but whatever. Interestingly enough the northern populations are also doing pretty badly in some areas mainly due to habitat loss and there are some projects trying to keep numbers up. Only one problem- if you’d been in those areas 250 years ago there were no prairie chickens- they moved in as European settlers cleared forests for farming. So what point in history are we picking as “ideal”? "

I believe Moro said something like the last line in her comment.


PBurns said...

A very good post.

I mull over some of the same questions in a post entitled:
"Thinking About Species Loss", structuring the post around 5 questions:

1. Do the animals exist at all?
2. Is the animal being described really a species?
3. Was the species ever very common?
4. What is the population of the species now?
5. What has really caused the decline (or increase) of this species?

In another post on dogs, I note that "Ironically, it turns out that maintaining a [dog] breed and keeping it more-or-less heterogeneous is neither a contradiction nor a difficulty. The trick is simply to follow Mother Nature and to occasionally do true outcrosses to animals that are entirely outside of the gene pool being crossed into. In the case of cattle and chickens, this is commonly achieved by crossing in an animal of similar size and traits, but with a very different genetic history.

It surprises people to find out that Mother Nature does much the same thing. Most people assume a Mallard duck is a Mallard duck. Aren't all Mallards simply clones of each other?

Well, No. You see, ducks hybridize all the time. What appears to be a Mallard may, in fact, have a little Gadwall crossed into it, or a little Black Duck, or even a bit of Greenwinged duck tucked into its double-helix.

In the duck world, where success is defined in Darwinian terms, there are no closed registries. While animals within a species tend to mate with others of the species in the same area, new blood flies, walks or swims in all the time. In the case of ducks, it may even come from across the ocean -- or from an entirely different duck species....

What is true for ducks is true for a lot of animals. Not only will individual animals often travel great distances to find unoccupied territories, they may also cross the species barrier as they do so. A wolf will mate with both a dog AND a coyote, while finches leap across the species barrier at the drop of a hat. A spotted owl will freely mate with a barred owl, while most amazon parrots freely cross-breed. A lion can mate with a tiger and produce fertile offspring, and an African elephant can cross breed with an Asian elephant. A muskellunge will cross with a northern pike, and a sunfish will cross with a bluegill. Trout and salmon species readily hybridize. Many species of hawks and falcons will also cross the species line, while a buffalo will cross with a cow. Just last week a hunter in Alaska shot an animal that turned out to be a cross between a polar bear and a grizzly.

The point here is not that trans-species outcrosses are common, but that even between distinct species Mother Nature often runs her train "loose on the tracks," and a considerable amount of genetic wobble is allowed.

Mother Nature allows outcrosses because she values heterogeneous genes, while she punishes homogeneous genes by "culling" animals through a process of dwindling survivorship (neonatal mortality), shortened lifespans, and infecundity.


Ken Chiacchia said...

It gets worse -- in bacteria, the concept of species begins to dematerialize. Bacteria of different species -- recognizably, indisputably (by the old standards) different bugs -- have been regularly exchanging DNA for billions of years, in a kind of sexual reproduction called F-plasmid exchange.

And of course, "race" is becoming bioligically fishy as well. We've known for a while that the genes that determine "race" as we commonly think of it don't obediently follow the geneological and geographic lines we expected. Many folks studying the human genome have begun to think of it as a completely artificial construct with no biolgical meaning.

There was just an article in the Jan. 22 Nature that took it one step further -- there has been so much interbreeding going on at the edges of "racial" and "ethnic" groups that they have no real genetic coherency. And soon, when individual DNA sequences become common, we're likely to find that we have significant shared "ancestry" with people on the far side of the globe who look nothing like us -- because the genes kept intermixing long after the groups that mean so much to so many people came into being.

therese said...

Ken: at this point I too use race as a term that holds no real genetic conotation- just a way to group things that look similar together, similar to morph or phase.

You're exactly right on the bacteria. Between them and some of the paraistoid wasps (bacteria induced asexual reproduction which can be "cured" with antibiotics) its funny to think people thought biological species concept would work. I enjoy going to PEET meeting where there are bacteria people, they have quite interesting perspectives on things.