EXERCISE 23
Experimental Manipulation
of Model Ecosystems
Ecosystems have been described as the "functional units of the landscape" (Odum, 1971),
where organisms interact with their physicalchemical environment and with other organisms.
For limnologists, lakes, ponds, streams, and
rivers and their associated drainage basins represent the ecosystems of interest. However, these
ecosystems are large and complex, and thus it is
difficult to decipher the various interactions that
occur within them.
Natural ecosystems are exposed continuously
to changing environmental conditions, of both
internal and external types. These conditions can
be changed either by natural (for example,
seasonally) or by unnatural means (e.g., through
human intervention). Whatever the cause, the
effects of any single change may be difficult to
separate from those resulting from all of the other
changes that are occurring simultaneously. A
changing natural ecosystem can be likened to an
experiment in which the investigator is trying to
control and, at the same time, understand many
known and unknown variables. For the creative
scientist, this situation can present a real challenge. Nevertheless, the understanding of causation in the ecosystem rarely can be delineated
without the benefit of some carefully designed
and rigorously controlled experimental work.
The cost in both money and time of doing
experimental work on whole ecosystems often is
prohibitive. Moreover, while such studies have
been done [see, for example, Hasler and Johnson
(1954), Likens et al. (1970) and Schindler (1974)],
we usually do not have the privilege of being able
to alter or to disturb seriously an entire ecosystem for experimental purposes. A common
solution is to recreate, in the laboratory, microecosystems or microcosms [see, for example,
Warington (1851) and Beyers (1963)]. An ecosystem brought into the laboratory can mimic the
natural ecosystem in some respects but will differ
in others: a microcosm is a simplified ecosystem
with discrete boundaries. The scales of events in
both time and space of events are abbreviated.
Succession to a new steady state takes place in
weeks, rather than in years. Microcosms generally have fewer species than do natural ecosystems and have, in consequence, simpler communities of organisms. Some characteristics of
microcosms make them valuable objects of
study. Microcosms are expendable, and the experimenter has control over the environmental
boundary conditions to a degree impossible to
achieve in the field. Also, it generally is assumed
that the investigator can establish reproducible
or replicable units, thereby allowing statistical
evaluation of the data obtained from experimental treatments and controls for each
manipulation.
In this exercise, two different approaches, the
chemostat and the microcosm, will be described
for the study of ecosystems in the laboratory.
Both of these approaches require several weeks
for stabilization, manipulation, and evaluation.
301
Experimental Manipulation
of Model Ecosystems
Ecosystems have been described as the "functional units of the landscape" (Odum, 1971),
where organisms interact with their physicalchemical environment and with other organisms.
For limnologists, lakes, ponds, streams, and
rivers and their associated drainage basins represent the ecosystems of interest. However, these
ecosystems are large and complex, and thus it is
difficult to decipher the various interactions that
occur within them.
Natural ecosystems are exposed continuously
to changing environmental conditions, of both
internal and external types. These conditions can
be changed either by natural (for example,
seasonally) or by unnatural means (e.g., through
human intervention). Whatever the cause, the
effects of any single change may be difficult to
separate from those resulting from all of the other
changes that are occurring simultaneously. A
changing natural ecosystem can be likened to an
experiment in which the investigator is trying to
control and, at the same time, understand many
known and unknown variables. For the creative
scientist, this situation can present a real challenge. Nevertheless, the understanding of causation in the ecosystem rarely can be delineated
without the benefit of some carefully designed
and rigorously controlled experimental work.
The cost in both money and time of doing
experimental work on whole ecosystems often is
prohibitive. Moreover, while such studies have
been done [see, for example, Hasler and Johnson
(1954), Likens et al. (1970) and Schindler (1974)],
we usually do not have the privilege of being able
to alter or to disturb seriously an entire ecosystem for experimental purposes. A common
solution is to recreate, in the laboratory, microecosystems or microcosms [see, for example,
Warington (1851) and Beyers (1963)]. An ecosystem brought into the laboratory can mimic the
natural ecosystem in some respects but will differ
in others: a microcosm is a simplified ecosystem
with discrete boundaries. The scales of events in
both time and space of events are abbreviated.
Succession to a new steady state takes place in
weeks, rather than in years. Microcosms generally have fewer species than do natural ecosystems and have, in consequence, simpler communities of organisms. Some characteristics of
microcosms make them valuable objects of
study. Microcosms are expendable, and the experimenter has control over the environmental
boundary conditions to a degree impossible to
achieve in the field. Also, it generally is assumed
that the investigator can establish reproducible
or replicable units, thereby allowing statistical
evaluation of the data obtained from experimental treatments and controls for each
manipulation.
In this exercise, two different approaches, the
chemostat and the microcosm, will be described
for the study of ecosystems in the laboratory.
Both of these approaches require several weeks
for stabilization, manipulation, and evaluation.
301