A better definition of a species is "an isolated breeding population".
However, there is not one single definition that fits all situations. A
given definition might be much too restrictive in one situation, and much
too liberal in another. The important point of the above definition is that
it =does not= preclude the populations in question from coexisting, only
that they do not naturally interbreed.
The Mbuna cichlids from Lake Malawi have been previously offered as an
example. There are those with dramatically different coloration but which
freely interbreed in nature and in aquaria, and are simply different color
morphs of the same species. On the other hand, there are colors and color
patterns that are quite "popular" among Malawian cichlids, and are sported
not only by different species, but across generic lines as well. The forms
may look superficially very similar, and may sometimes even co-exist, but
they do not interbreed in nature. They are different species.
The barriers to interbreeding are not all physical isolation (as in
different sides of a mountain chain, different river systems, etc) or
genetic (producing inviable and/or infertile offspring). There are
behavioural, social and habitat barriers as well.
So the problem that arises with the Malawian cichlids is that taken out of
their natural environment and placed into our artificial and crowded tanks,
usually these different species =will= interbreed in captivity, and what's
more, usually produce perfectly viable and fertile fry. The speciation in
Lake Malawi is so historically recent that species of grossly differing
size, shape, coloration and natural habitat (eg. rock-dwelling vs
open-water) can and will sucessfully interbreed in captivity. Behavioral,
social and habitat barriers to interbreeding either no longer exist or
break down quickly in our tanks. By the definition you mentioned above, we
would have to consider the tremendous genetic diversity of the endemic
cichlids of Lake Malawi as mostly belonging to one species. It just ain't
so! :-)
We don't want happening (any more than it already has) to the rainbows (and
the killies and the Apistos, etc, etc, etc) what has happened with the
Malawian cichlids. The genetic purity of many strains has been permanently
compromised in the hobby. Worse, the problem in some instances can no
longer be fixed by fresh importations of new wild stock. Fishes which would
normally never come in contact with one another in nature have been mixed
together in nearer, more easily collectible sites by collectors, resulting
in genetic corruption of wild breeding populations.
While this last scenario is not likely to happen with rainbows (or is
it??), something has now been lost there.
Genetic drift is part of all breeding populations. Within a freely
interbreeding population it is evolution of the species. When a population
becomes separated into one or more populations that for whatever reason no
longer interbreed, they each drift in their own way, eventually possibly
becoming separate species. At what point the different populations should
be considered separate "races", subspecies, species, genera, etc. is
pretty subjective and often hotly contested and debated. Just because we
consider two fish to be simply different populations of the same species
today doesn't mean we won't "determine" they are different species (of
possibly even different genera) tomorrow.
One last clunker thrown in for thought. I understand there are fish (I
can't remember which) that have a very broad range, and with which local
populations all freely interbreed from one end of its range to the other.
The breeding population is continuous. Yet specimens from one extreme of
the range cannot successfully interbreed with specimens from the other
extreme, with a sliding scale of success in between.
Well, enought is enough. Sorry to go on and on and on. But I watched this
thread the first time as a sometimes frustrated lurker, and now its popping
up again, right on the tail of the previous one. I guess I just finally
exploded! :-)
Ron West