[Taxacom] barcode of life: PS

Stephen Thorpe stephen_thorpe at yahoo.co.nz
Sun Jul 4 19:11:52 CDT 2010

PS: Putting it another way, I think that the C. flavissima / C. vroliki example demonstrates direct evidence of high reproductive isolation (and therefore distinction at the species level) just as the C. ferrugata / C. shepardi example does! The only difference is that the repro. isol. is evidently 100% in the latter case, but somewhat less (but still high) in the former case (because both forms of both examples are maintaining their integrity, except just along a narrow border between the respective ranges of the two forms in the former example)

>So, by anyone's metric, there are effectively no biological reproductive barriers between these two species [C. flavissima / C. vroliki ]

If the world was such that the level of reproductive isolation was often mid-range, then this would be a world in which species didn't exist (and so taxonomy at the species level would be impossible). But the world doesn't seem to be like that. The level of repro. isolation is mostly either low enough or high enough for species to be a useful concept

Analogy: if there were an equal proportion of people for every possible height, then there would be no useful concept of "tall people" or "short people". Of course you could arbitrarily set a particular height such that anybody above it was a "tall person", but this would not be a "real" distinction. A person who is 7ft tall really is a tall person simply because there is a distinct gap in nature between that and the normal range of heights (which is NOT to suggest that there aren't a few people with each "in between" height). Note that a 7ft tall person would still be a tall person if lots of people were 7ft. tall, provided only that there were relatively few people of heights somewhat less and then more again below that, so it isn't just about the rarity of 7ft. tall people, but about the whole frequency distibution of heights.

It is all about finding (describing) patterns in nature - patterns which "really are out there" (and need not have been if things were different)


From: Richard Pyle <deepreef at bishopmuseum.org>
To: Richard Zander <Richard.Zander at mobot.org>; Stephen Thorpe <stephen_thorpe at yahoo.co.nz>; taxacom at mailman.nhm.ku.edu
Sent: Mon, 5 July, 2010 10:18:12 AM
Subject: RE: [Taxacom] barcode of life

Hi Stephen,

> It is an in principle 
> refutable hypothesis that they are distinct species. 

In some cases, I would agree.  Those are cases where the two forms are
sympatric, but show no evidence of hybridization; or cases where
occasionally strays from one population find their way into the other
population, and no evidence of hybridization occurs.

But because the BSC pivots on the ability to produce fertile hybrids *in
nature*, in the vast majority of cases (i.e., allopatric sister-species or
sister-subspecies or whatever you want to call them), it's not really a
refutable hypothesis until a natural event allows us to refute it.  We can
certainly test (in many cases) whether viable hybrids can be produced when
individuals from the two populations are artificially given the opportunity
to do so, but that's not really how the BSC is defined.

Also, even when they do have the opportunity to reproduce in nature, there
may be a spectrum of circumstances ranging from totally uninhibited
reproduction, to highly assortative/disproportionate mating, whereby viable
offspring from the two different forms/populations result from only a very
small percentage of the opportunities to do so.  And in the case of rare
vagrants from one population to another, how rare does it need to be to
cross the threshold from "gene flow between two populations of the same
species" to "rare hybridization event"?

Following the lead of Curtis, I'll illustrate with a couple of examples from
the genus I worked on for my PhD.

Centropyge ferrugata Randall & Burgess in Burgess & Axelrod 1972 is
restricted to a small region in the western Pacific, from southern Japan to
the northern Philippines, and a population in the Ogasawara Islands. It
looks like this:

Centropyge shepardi Randall & Yasuda 1979 is restricted in distribution to
the Mariana and Ogasawara Islands. It looks like this:

These two species have only been known for a few decades, yet their status
as distinct species has already been questioned. Several people have
informally suggested that they should be regarded as the same species, based
on color/morphology alone.  A genetic investigation currently in progress
found identical haplotypes between the two forms.  It would be extremely
tempting to treat them as different forms of the same species.  Except for
one thing:  they both occur in the Ogasawara Islands, and despite two field
trips there, and many dives searching specifically for hybrids, none were
found.  You could often find harems of both species within a meter of each
other, so *clearly* they have opportunity to hybridize. Yet, without
exception, the harems are composed of one form or the other, without any
apparent inter-mixing or hybridization.

Now, another example in the same genus:

Centropyge flavissima (Cuvier in Cuvier & Valenciennes 1831) is widespread
throughout the tropical central Pacific, and looks like this:

Centropyge vroliki (Bleeker 1853) is widespread throughout the tropical
western Pacific, and looks like this:

The two species are parapatric, and within the vast majority of their
respective ranges, their coloration is extremely consistent.

However, in a few localities, along the border between their respective
ranges, the freely form hybrids:

In fact, in these specific areas, they form hybrid swarms -- where there is
a complete spectrum of forms from one pure parent to the other. Indeed,
"pure" parents are very hard to find.

So, by anyone's metric, there are effectively no biological reproductive
barriers between these two species.  When given the opportunity to
cross-breed in nature, they do so seemingly without restraint.  But for the
vast, vast, vast majority of the individuals of each species, there *is* a
*geographic* barrier to reproduction.  That is, looking at both species in
full, only a tiny, tiny percentage of individuals of either species
participate in hybridization episodes.  Moreover, both species have been
known since the early/mid 19th century, yet nobody, ever, in that 150-year+
history of taxonomists, has ever suggested that the two species should be

Would it best serve the communicative needs of biologists to regard them as
the same species?

So here's my point:  When we have direct information that, when given the
opportunity to interbreed in nature, two closely-related forms fail to do
so; then we have good reason to regard them as distinct species, no matter
how similar they are morphologically or genetically.  This is the strength,
in my opinion, of the BSC as a guide to defining species boundaries.

However, the converse is not necessarily true (and, I suspect, this is where
we may disagree).  That is, there are some cases (e.g., C. flavissima and C.
vroliki) where, despite a demonstrated ability to produce viable offspring
in nature, there is still good reason to regard them as distinct species.
You might want to maintain that these *are* the same species, according to
the BSC.  And I would have a hard time refuting that.  But I can also tell
you that if I published a paper declaring C. flavissima and C. vroliki to be
the same species, *NOBODY* in the ichthyological community would follow my
lead. Why? Because the communicative needs of biologists are best served by
regarding the two forms as distinct species.  

Now, if it were the case that the two forms (C. flavissima and C. vroliki)
were broadly overlapping, with hybrid swarms representing more than a tiny
fraction of the total set of populations of both species; or if there was a
continuum of variation from one end of the spectrum to the other, then it
would make much more sense that the communicative needs of biologists would
be better met by regarding them as the same species, and perhaps as
different subspecies.

One final thing:  The point has been raised several times in the past (in
print, at least -- if not on Taxacom), that the question of whether or not
two populations represent different species is separate from the question of
whether or not those populations should be labeled with different species
epithets.  I *think* I understand where this sentiment is coming from, but I
don't think I buy it.  I'm not aware of any case where biologists have said
"these are clearly the same species, but we'll continue to call them by
separate species names"; or "these are clearly distinct species, but we
think that people should continue calling them by the same name".  Much more
often we see the softer: "our evidence clearly suggests that these should be
treated as distinct species, but we'll leave it to someone else to change
the name".  In other words, while one may be able to make a philosophical
argument that whether they *are* distinct species is separate from whether
they should be *named* as different species; in practical terms, I think
there is no meaningful distinction.

But in this context, perhaps one's perspective on how lines should be drawn
between species depends on whether one is looking at the issue through the
eyes of a nomenclaturist, or through the eyes of an evolutionary biologist.


P.S. Apologies for the length of this post, which will (no-doubt) reinforce
my reputation as a voluminous poster to this -- despite my best efforts in
recent years to establish a track record to the contrary.


More information about the Taxacom mailing list