Evolution: Such a thing as a "middle species"?

Hi there,

Just have a question about evolution.

Does evolution sometimes throw up a situation in which you have 3 different “species”… X, Y and Z, such that:

X and Y can breed to produce fertile offspring.

Y and Z can breed to produce fertile offspring.

X and Z cannot breed to produce fertile offspring.

In the example above, you can see that ‘Y’ is kind of a “middle species”.

Is this phenomenon a standard part of evolutionary theory?

(Please note, I’m not really talking about chihuahuas breeding with St. Bernards, but moreso asking about instances of two clearly separate species).

Yes, as a generalized matter. Consider ring species, as well as the possibilities of clines in general. This is part of why the idea of “share the same species” as a Yes/No equivalence relation isn’t really the most coherent one (and shows how defining it somewhat arbitrarily in terms of ability to create fertile offspring isn’t particularly helpful for patching up such incoherences).

(Though, I should say, “can produce fertile offspring” isn’t necessarily really fruitfully understood as a Yes/No relation either)

Excellent, you’ve given me some good resources to ponder. Thank you.

You may have a problem with your premise. A species is defined as a “a group of organisms capable of interbreeding and producing fertile offspring.”

Are you thinking of a “transitional” species? Or what is your long-term intention with this line of questioning?

The point is presumably the semi-incoherency of such a definition (and related questions concerning the species problem); that is, if we think of every animal population as having a particular species, then it cannot actually be the case that, for every pair of animal populations, they are capable of interbreeding and producing fertile offspring just in case they have the same species.

[I say “animal populations” above rather than simply “animals” to ignore the fact that, for example, no two men can successfully interbreed, despite presumably being of the same species. Also, outside of animals, note that the definition of species in general cannot be simply what you posted, as many species do not use sexual reproduction to begin with]

I think the problem is actually fuzzy boundaries - the question ‘can individuals of these two populations interbreed successfully?’ may be answered ‘yes’, ‘no’, ‘usually’, ‘occasionally’, ‘under certain conditions’, etc. But the question ‘are these two populations the same species’ expects a straight yes or no answer.

The notion of ‘species’, although useful and often quite applicable to the real world, is still an artificial system of pigeonholing, to which the natural world will not necessarily always deliver clearly categorisable cases. It’s an attempt to impose a discrete scheme of measurement upon a continuous system (or if not actually continuous, discrete at a much finer level).

I agree with you entirely (this is what I was trying to get at with discussion above of how these things are not best understood as simply Yes/No relations).

The problem is that in the situation you describe, the X, Y, and Z would be the same species by definition. That’s why ring species are still classified as one species. Even though X and Z can’t directly interbreed, you can still get gene flow between them through the Y intermediate.

They’re not classified as one species though, are they? At the non-joining ends of the Larus gull ring, Herring Gulls are Larus argentatus and Black-Backed Gulls are Larus fuscus - with a few additional species and subspecies in the middle of the chain to boot.

This is a common misconception. According to the Biological Species Concept (one of several possible concepts) species must not only be capable of producing fertile offspring, they also must do so regularly in nature where their ranges overlap. (If ranges do not overlap, it’s generally a judgment call based on whether or not subspecies/species in the same group with a similar degree of similarity hybridize.)

Many species groups, such as ducks of the genus Anas and all species in the genus Canis (wolves, dogs, coyotes, and jackals) are capable of producing fully fertile hybrids. They normally do not do so in nature due to behavioral differences. Locally, however, in unusual situations (as where one species is very rare) this isolation may break down and they may hybridize.

They are not classified as one species, but then, neither are they truly an example of ring species after all (.pdf doc; this is the Liebers paper mentioned in the Wikipedia article).

Perhaps so - in fact, they’re a pretty apt example of how nature just goes on doing whatever it’s doing, with little regard for our attempts to describe it.

Ring species were classified as one species a few years ago when I went through that class. But it’s worth remembering that “species” is a completely human, artificial concept that we impose on the natural world. Gulls or dogs or frogs or whatever don’t care about gene flow or ancestral lines - they just think, “Hey! I bet I can hump that!”

ETA: I don’t mean to say it’s an invalid concept - it’s a very useful one, in fact. It’s just that I think some people sometimes start to think of “species” as something much more fundamental and, I don’t know, real than it really is.

For a second there I thought you said ducks and dogs could mate.

Duckdog! The flying dog with webbed feet and claws!

Its bark doesn’t echo - and no one knows why.

Actually, the exact opposite is rather common: species that could interbreed, but due to small differences in appearance, vocalization etc. rarely do so.

Like, maybe, reproducing “after his kind”?

Heh. ‘Species’ is difficult to exhaustively define for reasons mentioned above. ‘Kind’ simply isn’t rigorously defined because the concept exists to obfuscate, rather than clarify.