Real Species

[pierre deleporte]

Starting back from the original question:

At 22:11 13/04/2004 -1000, Richard Pyle wrote :
>The more specific question that I'm interested in is whether
>boundaries between species actually exist in nature (independent of our
>ability to observe them) and it is our job to discover them; or such
>boundaries are (artificially) defined by us as a means to allow more
>effective communication.

Individuals are rather well circumscribed biological systems in themselves.
Classic species are classes of individuals. Species themselves are not 
"individuals". They are not even loosely connected biological systems (the 
dead individuals don't interact with the living ones, to the difference 
with spatially restricted populations composed of living, interacting 
individuals). Classic subspecies also are such kinds of classes and don't 
form a coherent interacting system (the dead members of a subspecies don't 
interact with the living ones, contrary to living members of an existing 
population). (possible reading on this: Mahner and Bunge 1997).

According to Popper (which cannot be always wrong ! ;-), all 
classifications are conventional. A species (or subspecies) is a terminal 
taxon = a "class" of some kind. Its delineation depends on your convention 
rules for delineating (sub)species.
Just define (sub)species a useful way, according to your problem at stake 
(evolutionary, or other...).
Seems all trivial... What is required is at least to expose clearly your 
criteria (which species concept you're using in a given context)... seems 
it's not traditionally done (see other posts).

This said, it is not because you have a clear criterion (or unambiguously 
combinable series of criteria) that species limits will not sometimes 
appear fuzzy in the real world (see other posts). This is "real" too: the 
easiness with which natural features will fit (or not well fit) into your 
conventional classification grid. Part of the explanation is the ongoing 
evolutionary process, and possibly "speciation" process, which is 
progressive in populations and biotas (see many other posts).

Thus, OK with Richard's hint that species "boundaries are (artificially) 
defined by us as a means to allow more effective communication". And 
"effective communication" (hence relevant criteria) may change according to 
different questions at stake. Classification implements useful conventions 
(at best!).

Some will call "real species" the classes of individuals fitting 
unambiguously their classification criteria; they rather should be called 
"well discriminated species" according to these particular criteria.

Example: suppose you have a population of elephants. You may class it into 
a species, and possibly a subspecies (...your criteria, for your specific 
purposes). Two pink mutants occur in this population.
- if your problem is regional conservation biology, you can simply go on 
calling this a population: all the local living elephants, including the 
pink ones (i.e. you don't propose to shoot them down in order to preserve 
the "original species" unspoiled, neither to breed them separately in order 
to try to create a "new species"). Newborns join the population (even if 
pink), dead ones quit it (even if grey).
- now if your problem is evolutionary biology and speciation processes, 
it's interesting to distinguish the pink individuals as the first members 
of a possible new (sub)species. If you mix up the pink ones with their 
colleagues because they belong to the same population, you can't easily 
engage into the analysis of the possible speciation process (i.e. follow 
the particular fate of the pink elephant's descent). Evolutionary studies 
must account for the apparition and fate of novelty. Hence some 
'typological' classification is, paradoxically enough, useful for 
evolutionary investigations: how do new "types" appear, and succeed or are 
wiped out, or coexist in the long run with ancient types. If you have a 
definition of "species" (or whatever way you decide to name your terminal 
classes) which means "lineage" or "sympatry" (or both) whatever the changes 
in "types", you cannot easily study what is classically called "speciation" 
(= a process of divergent evolution).
- other question: an extreme proposition for phylogenetic analysis is to 
use individuals as terminal taxa. Seems it should work. Certainly for 
clonally reproducing organisms. And likely for other ones too (you simply 
expect fuzziness in the "tokogenetic" apical zone of the tree... and some 
extra computing time!).

Hence classifications and their corresponding classes are not "real or not 
real" (they exist from the moment we forge them), they are "useful or not 
useful" in a specified context, and classes or real objects appear more or 
less well discriminated, this possible fuzziness being itself a possible 
source of further investigation. Different problems, different optimal 
classifications (criteria and delineation rules...), and this vails for, 
let's say... terminal taxa of all imaginable kinds and names, defined for 
optimal communication in specific contexts.

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