The Gibbs Paradox & Ecology

["Tim Gwinn" <***>]

I was walking in a local wildlife refuge the other day, and interspersed
along the trails there were occasionally informational signs - most of which
were pictures and descriptions about species one might see. Two of the signs
were quite unusual: they explained why the ponds were occasionally drained.
One explained that "non-native" water chestnuts were too prolific and
therefore draining was necessary in order to stem the water chestnut's
growth. The other sign described how "non-native" carp were causing some
kind of problem by their presence and therefore, the ponds needed to be
drained so that all the carp were killed. (The sign also noted that after
the carp died, staff members had to walk thru the muck, picking up the dead
fish to prevent outbreaks of avian botulism. Ick.)

I found these signs remarkable. The idea that the US Fish & Wildlife Service
was engaged in such active "remedial" intervention surprised me. (Perhaps
this is common knowledge to others.) It was also surprising that they took
such a specific view of that ecosystem: the purpose of the refuge was as a
place for migratory birds to stop, and therefore species that interfered
with this perceived purpose were eliminated. It also appeared to be
important that the offending species were identified as "non-native" so as
to lend extra justification to their destruction.

=== Just to be clear: I do not argue herein either for or against such
actions. ===

It occurred to me that the discrimination processes involved had alot of
similarity (or analogy) to "Gibbs paradox" in thermodynamics. In ultra-short
form, Gibbs paradox is a scenario where seemingly contradictory entropy
results appear to challenge the validity of the second law of
thermodynamics. Imagine a closed vessel with a divider separating the volume
into two halves. In one side is a volume of a gas A1 at a temperature T and
pressure P. On the other side, is a gas, A2, also at temp T and pressure P.
When the divider is removed, does the entropy increase? Well, if A1 is the
same gas as A2, then there will be no entropy change, but if A1 <> A2, then
the entropy will increase as the two distinct gases intermingle.  However,
our ability to distinguish A1 from A2 is central to deciding if entropy
increases or not, and if so, by how much. Entropy, it turns out, is not the
"real" physical quantity that it appeared to be.

The seeming 'paradox' rests on informational issues. Rosen remarked on Gibbs
paradox: "I argued a long time ago that determining whether two gases were
"identical" or not, on the basis of discriminators, involved solving a word
problem (i.e., it was algorithmically unsolvable), and hence in particular
that entropic computations about them were unreliable." [EL p. 124] (I
believe Rosen was referring to this paper (which I do not have): "The Gibbs
Paradox and the Distinguishability of Physical Systems." Philosophy of
Science 31: 232-236. 1964)

E.T. Jaynes also gives a lucid portrayal of the informational basis of this
(supposed) paradox: http://bayes.wustl.edu/etj/articles/gibbs.paradox.pdf
For an information theory perspective:
http://arxiv.org/abs/cond-mat/0109324

My thought was that this issue of discrimination - or, solving a word
problem - was relevant to ecology, and to the choices of remedial actions.
What is it, for example, that allows us to label a given thriving species in
an ecosystem as "non-native"? It seems to rest on our historical knowledge
of a species relation to that ecosystem, and especially to knowledge of
whether a human was, shall we say, "causally efficacious" in historically
introducing the species to the ecosystem.

On the other hand, it has nothing whatsoever to do with the actual observed
behavior of the species in that ecosystem. It may well be that an ecosystem
has many "non-native" species in it; however, perhaps most of these species
fail to be spectacular in any way that draws attention to them, such that we
have never discriminated their "non-native" status.

One might claim that it is an issue of actual objective facts with
incomplete knowledge. But, consider: do "non-native" species ever become
"native" to an ecosystem? After 10 years? 100 years? 1,000 years? 1,000,000
years? One would be hard-pressed to claim that after 1,000,000 years that
carp will still be considered "non-native" to the refuge (assuming both
still existed, of course). If a new species of fish were introduced to the
wildlife refuge by, say, birds transporting them rather than humans, is that
fish species non-native or not? How could it be non-native if it arrived by
such "natural" means? Particularly, when we consider that this refuge as an
ecosystem is but one portion of a larger ecosystem in which this
transport-by-bird took place and where both bird and fish are "natives"?

Actually, I suggest that no species are native in any strict sense. Species
adapt to changing ecosystems, and themselves change ecosystems into
different ecosystems. Indeed, if within a given ecosystem a new species
arose, that species would not fit with the notion of 'native' insofar as: 1)
it is a new, previously absent, entry into the ecosystem, and 2) its
introduction may severely alter the current stability of the ecosystem. We
could say that since it arose inside the ecosystem that qualifies it as a
native. But then, if that were the criteria, I suspect that exceedingly few
species today would be native in that sense to their current ecosystem.
Species inevitably roam into new locales - probably what we see today in
nature is more of a snapshot in a ongoing history of transience rather than
a reflection of some original situation of points of arisings of species .

As a result, I find the Fish & Wildlife Service's use of "non-native"
classification as justification for the wholesale slaughter of a species in
the ecosystem to be specious. It may certainly provide justification from a
moral or stewardship point of view (e.g., "humans did a wrong thing by
introducing that species, we are merely righting that wrong with our
'remedial actions'"). However, this was not the way the sign was phrased -
it was implied that this was a biological justification. The F&WS was simply
abiding by a scientific fact.

I am unconvinced that - from a biological perspective -  there is: 1) any
objective basis for categorization of "native"/"non-native" species; and,
perhaps more importantly, 2) that there is no objective basis for asserting
that "the purpose or function of this ecosystem IS x" (in the case of the
refuge, it was that "the refuge IS a waystop for migratory birds").

The tacit assertion in the claim "the purpose of this ecosystem IS x" is
that it is eternally true: the statement is unqualified with reference to
time. If at one time, the apparent purpose of an ecosystem WAS x, then it is
STILL x, and it will ALWAYS be x. Such a lack of reference to time is
grossly over-simplistic. It would take immense efforts to maintain the
ecosystem in its specified "purpose" for any extended period of time,
against the inevitable alterations due to geologic, climactic, evolutionary,
and other influences, both internal and external. It may also be that, in
the refuge example, migratory birds at some point will no longer frequent
that region at all, in favor of more enticing habitats thereby nullifying
the "purpose" of the ecosystem.

But it would seem that the Fish & Wildlife's assertion that "the refuge IS a
waystop for migratory birds" is an attempt to invoke just such a tacit
eternality as additional justification for their actions. But, clearly
ecosystems are not, and cannot be, static in that way. In short, it would
seem that any assertion of fixedness of processes in an ecosystem across
time is essentially a claim that the ecosystem is a mechanism.

I suspect that "classification as native/non-native" and "purpose of
ecosystem" are but two instances of a large number of terms and concepts
related to ecology that have similar roots in both discrimination and
criteria, and that, like 'entropy', are not "real" physical quantities or
properties.

Finally, I just mention that divorcing these terms from their appearance as
physical quantities or properties does not make them useless. What it does
do is shift the responsibility for the basis and consequences of these terms
from appearing to be mostly objective data to instead resting heavily upon
the shoulders of the observers themselves. In my view, this may shift much
of what occurs under the heading of "ecology" from being classified as
science to being classified as technology: perhaps to a technology of
ecological stewardship, for example. The latter is not thereby made
unscientific by being a technology; but it does emphasize the willful human
criteria and value judgements that (seem to me to be) what actually leads to
decisions like those taken in the example of the wildlife refuge. In this
way, when we intervene intentionally, we do so with as clear a distinction
as possible between our scientific justifications and our technological
justifications. And it is the latter, not the former, that is properly
malleable to the moral, spiritual, social, and similar influences that, in
my opinion, are the ultimate basis for any notion of ecological
preservation, remediation, stewardship, or the like.

Regards,
Tim