What does it mean to be stable?
Communities (in the ecological sense of the term) are
assemblages of populations of organisms, occupying specific habitat conditions.
Sometimes communities persist over long periods without exhibiting any
substantial change in their appearance or composition, and when such
“periods-without-change” prevail, we often refer to the communities as
“stable”. Other communities seem to be more or less constantly in a state of
flux, gaining and losing components, and hardly looking the same way twice;
such communities may be referred to as “unstable”. We should ask questions
about these terms and their application: are they appropriate terms? What
assumptions are concealed behind their use? Is there a definite boundary we can
draw between “stable” and “unstable”? Are the descriptions consistent with
other ecological principles? The present document cannot answer these questions
completely, but should help you as you think through the principles of
community ecology.
Like a lot of terminology in ecology, “stable” has a
curious history. It’s a term that we think we understand, but which may
not always end up defined the same way by all people. One thing stable cannot
mean, of course, is the same thing as “static” – a stable community is not
one in which nothing is changing! Organisms are being born, they are maturing,
breeding, and dying, and all the while the climatic conditions are varying,
small disturbances are occurring, and so forth. There may be a hive of
activity, yet over time roughly the same patterns can be observed at the site.
Whatever is lost (say, to death) tends to be regained, and most new things
(say, chance arrival of immigrant species) fail to become permanent new parts
of the system. These facts do not, however, tell us why such a state may
be maintained. In a stable population, we know that deaths can be
replaced equivalently by births, but how will lost species be “replaced” in
communities? The question is subtly but significantly different from the
population-level concept of stability.
Theoretically – maybe philosophically would
be a better way of saying it – “stable” carries with it a connotation of
“appropriate” or even “right”. Whether motivated by the principles of evolution
by natural selection, in which we expect the optimum or “best-case”
outcome to emerge by differential success of variants, or by the
principles of creationism, in which we might expect a “designer” to conjure up
a sensible system from scratch, the notion of stable states is intuitively
appealing. People studying complex systems want to believe that they are
orderly, even if only to have some confidence in the interpretations they make.
Unless you start with the expectation that at least some order exists,
how can you examine any system rationally? It’s a short step from this position
to the position in which one requires order – sometimes, for example in
physics and mathematics, students are encouraged to pursue the most elegant
solution to a problem, on the grounds that it is more likely to be the correct
one, even in the absence of data, though as long as data is eventually acquired
this is not a problem. The more complex the system under study, the more
necessary it may be to assume rather than demonstrate order, and the harder it
is to find solid justifications for the assumptions.
Operationally, that is for practical purposes, we
can define community stability in several different ways. Most broadly, we
could define it on the community level as “the presence of the same set of
species over long periods of time”, for example longer than the lifespan of the
longest-lived species present. We could also make narrower, more limited
definitions: stability might require not only that the same set of
species be present, but also that they continue to interact in the same
way with each other over time. More narrowly still, we might add a
requirement to the previous definitions: the relative abundances (or biomasses,
depending on the type of organism) of all species must also remain the same.
The narrower the definition, the more subtle a deviation from it we would call
a “loss of stability”.
This leads us to a new kind of problem: if stability
(of whatever type!) is “lost”, and the result is a new state, what are we to
make of this state? An unstable system won’t be truly chaotic or random
– it can’t be. Even if a system is unstable, it will still
contain resources, certain atmospheric and substrate conditions, and organisms
experiencing those factors and interacting with each other. You can’t call such
a system random, even if it does tend to change in unpredictable ways, because
it still features the operation of basic ecological principles… just not in such
a well-ordered way, or at least not in the way it did before. Now we come to
the crux of the whole stability argument: the search for a borderline between
stable and unstable states.
Is stability, or instability, likely
to be the “default” in nature? Evolutionary theory implies the former, but only
if the correct assumptions are in place. A stable state is likely to exist in a
system if there has been ample time for it to become established, and
if there has been not interference imposed from outside the system in
the process. An unstable state would then exist where outside forces
continually upset a system, especially if the impact these forces have is
severe. So where in the biosphere would we expect to see systems which we would
be forced to call unstable? Most habitat types, and most sites in those habitat
types, experience conditions which are not severe enough to enforce truly
unstable states.
All places on the Earth experience disruption (disturbance)
from time to time. Disruptions may be severe, but infrequent, or they
may be less severe, but more frequent. Some environments feature lots of small
disruptions punctuated by the occasional big one, while other environments have
the occasional big ones only. There really aren’t any livable
environments in which there are severe and frequent disruptions… this
would be like the lava-filled crater of an active volcano, or the middle of a
permanent hurricane. Thus when we speak of “unstable” states we are really
referring to deviations from the stable state typical for that habitat,
and indeed in most case the implication is that these unstable states will
eventually be replaced by more stable ones (see
below).
Now on the other hand, if we see a broad overlap
between the set of species that occupies one “community” (e.g. a forest)
and the set occupying a neighbouring community (e.g. a grassland), would
we be justified in saying that either is a stable state in itself? If we were
using the species-composition definition, we would have to say no, but if we
were using the set-of-interactions definition we might be able to say yes: the
species common to both systems would engage in unique sets of interactions in
each community! You can see from these arguments that often there is really no
hard and fast line to be drawn.
We can define, for convenience, two broad categories
of community-stability, based on physiological and ecological principles
applying to the organisms the communities contain.
Systems filled mainly with short-lived,
fast-reproducing organisms can become established quickly, but such systems are
extremely prone to disruption by even mild outside (extrinsic)
forces, so they may not ever meet an abstract criterion of stability, but even these
systems are stable in a way: they are easily disrupted, but just as
easily return to their original state. We call this short-term and changeable
kind of stability resilience.
It may not seem stable when you first think about it, but consider: as long as
you don’t look at the system right at the moment it is being disrupted,
it is likely to appear consistent over time. If you know that it is easily
disrupted, but also know that it returns to the same consistent state,
that is still stable.
Systems filled mainly with long-lived, large-bodied
organisms take a long time (usually involving slow growth of well-constructed
bodies) to become established, and the process of establishment may be repeatedly
set back by inopportune disruptions, but once such systems are fully
established, their stability is maintained largely by intrinsic forces like
competition. Small extrinsic disruptions are more or less shrugged off by such
systems, since the organisms present are physically strong enough to stand up
to environmental forces (e.g. trees with very thick bark can “ignore”
small to moderate fires). This kind of stability is called resistance (def. 2),
and no special pleading is likely necessary to convince you that it’s stable! A
disruption occurs, but has little or no effect on community structure, though some
individual organisms are bound to be injured or killed.
Of course there are some disruptions so severe that neither
type of stability given above will prevail. In the event of an asteroid impact,
for instance, nearly every type of organism present before the disturbance will
be unavailable to recolonize the area afterward, and contribute to the
re-forming of pre-existing communities, no matter how much time might elapse...
because they have gone extinct. Such truly global disruptions are like
discontinuities in the history of community stability, over which smaller-scale
descriptions like resilience or resistance fail to register.
If a community is disturbed, and it was a “stable” community by one definition or another, it should eventually recover back to that state – but how, exactly? Here we encounter another partly philosophical issue: is there some inherent “driving force” that makes a community return to the stable state? In the early days of community ecology, about 100 years ago, some people seemed to think so. Communities were viewed by Frederic Clements and his followers as integrated units, more or less “superorganisms”, whose constituent populations were analogous to the cells of a body. In the same way that a damaged body might heal itself, so might a community, to return to “health”. Clements viewed the sequence of species-sets involved in the recovery of a system as deterministic, in both their order and their composition. Other researchers, led by Henry Gleason, disagreed, and claimed that communities were not nearly so integrated, and in fact that in some cases it was “every population for itself”, each type of organism simply selecting its preferred conditions, and the resulting coincidental mixtures being misinterpreted as integrated “units”.
Today most post-disturbance ecology – usually called
succession theory – leans towards an individualist, Gleasonian world view, one
in which each type of organism accepts a slot if it is available and thus finds
itself now with one set of neighbours, now with another. Species do well, then
fail as they die out through competition or predation or simply the changes in
shade or moisture as a system matures. This is not to say that determinism, or
integrated species-sets, don’t sometimes exist, just that they cannot stand
alone as an explanation. No matter the view you might take of the reason for
replacements, each intermediate stage is unstable in the presence of
later-stage colonizers, so replacements do occur – whether of one species or of
a set of species at a time isn’t too important in the model – until a set
establishes which cannot be replaced by another. Thus the process is apparently
driven not by a goal but by measurable step-by-step interactions along the way,
a much more satisfactory model in science even if it isn’t as elegant as a
predetermined endpoint!
At any stage, further disturbances can act to “re-set”
the series to an earlier stage, and there is still considerable dispute as to
whether an actual endpoint – a so-called climax community – is ever
truly reached. In many systems it can be demonstrated that even the “latest”
observable stage of a succession is itself unstable, suggesting that if only
the environment were unchanging there might be later, still-more-stable stages
to enter… but of course this cannot be tested in the real world. A century of
theory and experiment has not yet produced a complete understanding of
post-disturbance ecology, but it has strengthened the resolve of ecologists to
keep testing, since much has already been added to the earlier level of
understanding; no doubt more can be discovered.