[Taxacom] Diagnosing species

[Richard Pyle]

>     As for Richard's fishes, I think he may well be moving 
> toward regarding some of these so-called separate species as 
> actually being subspecies (or just populations in a more 
> continual spectrum of populations) within a larger species. 

Perhaps.  But as you might imagine, I regard them neither as "species" nor
"subspecies".  Rather, they are what they are (indivudal organisms;
collective sets of which have various patterns of similarity, which we may
or may not infer as representing patterns of homology), and I would advocate
labelling with either binomials or trinomials -- whichever a competent
taxonomist or community of taxonomists decides (ultimately by consensus via
usage, perhaps biased by a desire for nomenclatural stability) is most
appropriate for effective communication.
 
> Speciation simply hasn't been achieved and sampling is 
> finally showing that reproductive isolation has not taken 
> place.  

This makes it sound like "reproductive isolation" is a singular event that
we could mark on a historical timeline. In most cases, I suspect, it's
probably a long gradual transition; perhaps often marked by cross-branches
of hybridization/introgression/reticulation.

> For some organisms, 
> documenting gene flow may be so labor intensive as to be 
> largely futile at the present time, but a generation from now 
> (with whole genome barcoders and the like), it will become 
> easier, cheaper, and less labor intensive.  It may also lead 
> to a general consensus how low gene flow needs to become to 
> indicate speciation has occurred (although it would be 
> different for different groups of organisms).

I'm a firm believer in the promise of future cheap/ubiquitous whole-genome
sequencing.  And I am confident we will also eventually learn how to milk
every last drop of phylogenetically informative information out of those
whole genomes.  And I seriously doubt that, at such time, we will be any
closer to an objective definition of species than we are now.  The patterns
represented in those genomes came about by processes operating at the
resolution of individual reporductive events -- which in the vast majority
of cases occur across at least two, and usually three or more orders of
magnitude finer timescales than those associated with divergence of what we
traditionally think of as distinct species.  It would be nice if we
discovered something other than a smooth transition between these two points
of divergence (two sibling individuals from the same reproductive event at
one end of the spectrum, and sibling species populations at that other end)
-- such as a consistent inflection in the pattern of genetic divergence --
that would mark a speciation event (analagous to a pair of adjacent annual
growth rings on a tree trunk cross-section).  But I'd be willing to bet that
no such consistent demarcation will be found.  

In which case, we'll be then where we are now, and where we've been for the
past 250 years: recognizing "species" and other taxa as convenient units of
communication, defined as whatever a competent taxonomist or community of
taxonomists say they are.

Aloha,
Rich