Biology Reference
In-Depth Information
Further, according to Connell, a direct measurement of competitive
effects would be transplantation of individuals from an allopatric popula-
tion into a sympatric location, and observation of the effects where a
competitor is present and where it has been removed. Individuals that are
not transplanted but handled and observed in the same way as the
transplanted ones could serve as controls. Competitive effects would be
shown by compression of the niche in the presence of competitors.
Finally, a genetic basis for the differences would have to be demonstrated.
These criteria are very strict and very difficult to apply. Therefore, many
authors have used plausibility arguments. Colwell (
1984
) and Cooper
(
1993
) emphasized the usefulness of plausibility arguments based on
background information. However, it is important not to forget that
such arguments may easily mislead if not applied very critically. If not
used critically they may, for instance, lead to a confusion of interspecific
competition with other factors such as reinforcement of reproductive
barriers responsible for species segregation.
Nonlinear dynamics (chaos) in populations and the outcome
of competition
A considerable amount of work has been done on the non-linear
dynamics of populations in one- and multi-species systems. Here, we
discuss some selected models in order to show that the outcome of
competition between species may be largely unpredictable.
Hawkins (
1993
) used a simple model incorporating local dynamics and
dispersal for studying interactions between dispersal, dynamics, and chaos.
His aims were to show that the scale at which density dependence can be
detected, depends strongly on the intensity of dispersal between patches,
that dispersal plays a stabilizing role, and that there may be extreme,
sensitive dependence of long-term behavior on initial conditions.
Hawkins refers to Park's experiments on Tribolium, which had been
interpreted as showing that the outcome of competition was due to initial
conditions, or - in cases where it was not - the result of stochastic
processes. However, the model established by Hawkins showed that a
simple deterministic process, and not stochasticity, may be the explanation
for such cases. This is apparent even when numbers within each patch are
very small. Hawkins concludes that ''This extreme dependence of quali-
tative behavior on historical effects calls into question many of the para-
digms underlying ecological explanations.'' Hawkins' results further show
that the appropriate spatial scale must be selected to demonstrate density
