What role does distribution play in the species concept?

[Jools]

This has been troubling me for a while now and a post in the cory forum brought it up again so I thought I would ask the question here. From the original post...

Its down to location of the individuals, although two fish may look identical, because they are separated geographically buy hundreds if not thousands of Killometers, C. julii and C. trilineatus are a tipical examples. There is no way that they could have evolved as the same species.

This is why great importance is given to type locality, without knowing where a fish has come from it is very difficult to identify it with any accuracy.

Ian

The same argument exists elsewhere in the catfish world, Royal Plecos are another good example. The "real" (perhaps original would be a better word) comes from Venezeula and, given Ian's reasoning above, has to be a different species from L90 etc. There are other small differences too.

The problem I seek to discuss in this : Given the rivers these fish live in were connected in the past (albiet millions of years ago) but now form distinct basins or systems that do not interconnect even in times of high water. How do we know if the two populations have "evolved enough" to be classed as two species. Can we tell they had a common ancestor? Or is the fact that there is a physical barrier between them enough to say, OK, this is a different species becuase it is, for example, on this side of the Guyana shield from this other one.

What I'm asking is (Hopefully without slipping into the old "what is a species" debate), what are the rules or general practices when looking at two very similar fish seperated by land and also how do we know if the fish has evolved away from what it was when the rivers were connected.

Jools

[Silurus]

This is a difficult question to answer, but here are some points to consider:

1. Familiarity with the fish. This would be necessary for understanding intraspecific variation, bearing in mind what is construed as INTRAspecific variation by one may be the INTERspecific variation of another.

2. Speciation is an ongoing process. Trying to understand define the limits of a species is sometimes akin to trying to deduce the plot of a movie by looking at a single frame from the show. Bearing in mind the dynamic nature of the species, it's sometimes difficult to say for sure the number of species involved. After all, it is very likely that the two populations are in the process of speciating (and are thus incipient species), but enough potential introgression exists for the two populations to be treated as one. Again, it's all a matter of whether you lump or split.

3. Geography is sometimes a good indicator of what a species is, but it should not be used as a key factor, certainly not when the species involved occur in broad sympatry (and sometimes, syntopy).

So what is the gist of all this? Try as we might, there is still enough subjectivity in taxonomy that one man's species is another man's population (a point that I seem to harp on here).

[Jools]

HH,

Quick question, what does introgression mean in this context.

Thanks,

Jools

[Silurus]

Introgression is just a fancy way of saying that they can exchange genes, i.e. interbreed.

One point I omitted to mention is that speciation can be extremely rapid in catfishes (yes, it's not just a cichlid thing). The river drainages on mainland and Sundaic Southeast Asia have only been completely separated within the last 17,000 years, yet quite a number of populations have speciated in this time.

[Dinyar]

So if two freshwater fish specimens were morphologically identical but separated by thousands of miles, would you consider them distinct species based on the geographical evidence?

Dinyar

[Silurus]

If I can find no evidence that they are different other than geography, then yes, I would consider them the same species.

[pturley]

This is a very slippery slope to step out on, but I'll toss in a few pennies and try to keep my footing...

The most common metric in speciation (at least in a "splitters" view) is reproductive isolation. That is, are the two populations divergent enough so that they no longer represent a single gene pool. Of course this is an oversimplification, but a useful one none the less. If the species represents a single reproductive population over a broad geographical area, then it is one species. If the species has diverged in response to geographic isolation (spawning in different seasons, in response to different cues, physical differentiation (IE: This NO longer FITS!?")) then the populations likely represent different species.

A "lumper" would likely say that behavior alone is insufficient to differentiate the separate populations. The sum of statistical variation of morphological features would be the key to differentiation between the populations... If statistical significance is demonstrated (in addiition to other isolating mechanisms) then the two populations "could" be considered two separate species. If there is allot of statistical overlap in morphometrics, then the populations represent geographical variants of the same species.

Two very interesting texts on the Subject:
"The Diversity of Life", by E.O. Wilson (a self proclaimed "Splitter") and "What Evolution is" by Ernst Mayr (generally considered a "Lumper").

I am currently 80%+ of the way through E.O. Wilson's book and will follow with Ernst Mayr's. Interesting note, E.O. Wilson wrote a good deal in his book in response to parts of Mayr's work which was written a few years prior. Should be interesting to read the debate in these two texts. Given the text I am currently reading, I am somewhat more versed in the "Splitters" viewpoint at the moment. There are several interesting examples I can think of that give credence to a "Lumpers" viewpoint as well though.

Later,
Paul E. Turley

EDITS 9-23-03: Ernst Mayr (not Mayer),& E.O. Wilson is a self-proclaimed splitter

[Yann]

Hi!

I have seen so far many population of Crencichla regani, they look very similar but only differ from colour pattern but are considered to be the same species because of same morphological evidences. Even if they are the same species, I really don't think as a good idea to mix up geographical population, and promote breeding between these population.
I am also believe that this should not be done with loricariids and other catfishes.
These geographical population still may differ only by color so far, but are probably evoluating into seperated species??!!!
This is why I am not fond of having L-number that are supposed to be the same species but coming form different rivers, such as Hypancistrus L04-5 L28 and L73...

I think that Claude Weber is currently working on the population of Hypostomus watwata from the Maroni and Oyapok rivers to see how close or how distinct these population are...

just my 2cents on the subject..
Cheers
Yann

I

[pturley]

I wholeheartedly agree not to mix geographical populations. I prefer my aquarium subjects to represent the full width and breadth of evolutionary diversity. Cross breeding potentially erases distinct feature between these populations. To me, this diminishes the value of the aquarium stocks as a whole.
Just look at the state of the Malawian Cichlid fishes. For decades they were sold in strores as "Mixed Africans". It's only now, 40 years later, after extensive research, publicatioons on the subject and direct F0 imports that we as Aquarists come to appreciate the full level of diversity in this lake.
And how interesting would Tropheus moorii be today if aquarists had mixed each regional variant into one "Mutt" Tropheus representative?


Later,
Paul E. Turley

[Dinyar]

I expect that all responsible aquarists would agree that geographical variants of the same species should not be interbred. But I think that is a separate issue from the one of when a geographical variant becomes a separate species.

As Silurus said in an earlier post in this thread, "species" are constantly evolving and are as such dynamic, while our concept of "species" is a static snapshot in time. Though we may put a tight boundary around a "species", in fact that boundary is blurred by the fact of ongoing speciation. Hence the inevitable and irresolvable debate between "lumpers" and "splitters".

Seems to me (and of course, I'm just a rank amateur sliding blithely down slippery slopes) that one relatively objective yardstick that could be explicitly used to adjudicate the tolerance with which we define "species" is the rate of speciation. In taxa where this occurs rapidly, it would seem to make more sense to lean towards a splittist view, anticipating further divergence. Thus, hypothetically given two pairs of closely related populations with an identical degree of morphological variation, the pair that could be shown to be speciating more rapidly would be a better candidate for splitting into separate "species". (Of course, this is just an unrealistic thought experiment .)

Dinyar

[Silurus]

one relatively objective yardstick that could be explicitly used to adjudicate the tolerance with which we define "species" is the rate of speciation.

This isn't any more objective than what I've said so far.
How does one determine a "rapid rate of speciation"? What is going to be the yardstick? Number of nucleotide changes per unit time? Even this is not constant depending on where the nucleotides reside (i.e. in mitochondria or in the nucleus). Number of morphological character changes per unit time? This is even more tricky, given the fact that many character states are usually continuous variables, especially in incipiently diverging lineages.

[Dinyar]

Quantification in most sciences does not take place in the absence of judgment.

[Silurus]

Which is why what you propose does not make things any more objective.