Mayr on "What is a Species"

[pierre deleporte]

A 09:03 26/04/2004 -1000, Richard Pyle wrote:
>But if the "biological system" depends on some sort of cohesive
>inter-connectedness among "individuals" (sensu Deleporte), and you do not
>include the dead individuals as part of this system, then what, exactly, is
>it that establishes the interconnectedness?  Wouldn't a "population", then,
>be restricted to simultaneously extant direct kin sets, or perhaps a flock
>of birds or school of fish, or other set of individuals that have direct
>"here-and-now" interconnectedness?  When I think of the word "population", I
>infer a set of individuals joined by genetic heritage.  And as such, I find
>it hard to think of a population without also including the dead ones.

Effectively, what we call "a given population" is possibly, and I would say 
frequently, a very loosely biologically connected system between living 
individuals. This is why I introduced "social group", and you introduce 
"flock" here.
But as Ron states himself clearly, the individual (just a bird in the hand) 
is far more a cohesive biological system than any population, or part of.

All members of a poipulation, including dead ones if we include them in the 
definition, can be said to share a property of effective or potential 
interbreeding, as well as members of species. The notion of "effective or 
potential" seems important to me in this discussion. The individuals 
potentially reproducing together are not really reproducing with one 
another. They share a property (reproducing compatibility), which we use as 
a classificatory criterion. Such a species or population is a taxon, not a 
real network of effectively interbreeding  individuals. Just think of 
completely separate populations of a "same species", and of  course you can 
have completely separate strains in a "same population".

If we quit just one second the anthropocentric view of considering sexual 
species, and consider strictly clonal species, what is a population? In 
this case, a population, and a species, become typologically defined taxa, 
possibly with an additional geographic criterion for "population". They 
cannot be defined as potentially interbreeding, but only as sharing a 
common descent and looking like a given "type". A mutant judged 
"significantly different" by the taxonomist is immediately the first member 
of a new species.

This is, I think, the problem Kirk Fitzhugh unfortunately avoids dealing 
with in his essay about species. I suggest that Kirk, and all of us, begin 
with thinking of a species in non-sexual critters. Typology will impose 
itself as the taxonomic (classificatory) criterion.
 From this solid ground, extension to the more complexe figure of sexual 
lineages will be possible, by trying to deal with the cross-breeding mess 
in addition to (not instead of) the typological criterion. [Back to my pink 
elephants...]

But if a species (lineage) is an "individual", then all life from the 
origins up to now is an "individual"... who really buys this?
[Not to mention that, in classificatory logics, an individual cannot have 
individuals as parts, while classes may have subclases as components, with 
individuals composing only in the terminal classes... the notion of 
"individuals as part of individuals" seems flawed to me - as well as to 
Mahner and Bunge - anyway you take it, logically and ontologically... but 
will I again be taxed of semantic games?...].

>I think I understand where you are coming from on this; taking a population
>to be some sort of "snap-shot" of a cohesive set of organisms that somehow
>comprise a biological system.

I mean "some" biological cohesion here and now. But you're right in your 
comments.

>But the VAST majority of individuals within
>most sets of organisms that I would consider to constitute a "population"
>would never, in any way, interact with each other during the course of their
>lives.  They are only joined, it seems to me, by their shared evolutionary
>history.

Agreed 100%

>I would agree is you were defining a single coral head as a "population" of
>individual polyps.

It's effectively a much more cohesive "population", because it involves 
anatomo-physiological links, which define an individual organism. Nature 
decidedly makes nothing to facilitate our urge for universal classifications!

>Geographically-defined populations are, in a way, potentially
>much less genetically cohesive than units we refer to as "species",

I can't follow you here, if populations are subclasses of a species they 
cannot be more loosely defined given we use the same set of criteria. You 
add at least spatial proximity for populations, and then in much cases 
improved chances of local cross-breeding. But of course you can find or 
imagine exceptions to the rule (nature is pitiless for classifiers).

>It seems we differ only in how we scope a "population".  If I understand
>your definition/scope, I would tend to use words like "school", "herd",
>"flock", "colony", etc. to circumscribe a set of directly-interacting
>individuals.

You understand quite well. But I learnt with my professor of population 
ecology that "a population is what you decide to define as a population 
given the needs of your ongoing research". Classification again, allbeit 
based on biologically relevant criteria. Populations are frequently defined 
as loose systems, the loosiest being "mere geographic proximity".

>   The word "population", to me, transcends the direct
>interactions among simultaneously extant individuals, and neccessarily
>includes the dead ones (and yet-to-be-born ones).

Under this definition, it's clearly a class ("taxonomic", Ron would say), 
not a real biologically cohesive whole.

Best, Pierre


Pierre Deleporte
CNRS UMR 6552 - Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 66
Télécopie : 02 99 61 81 88