Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Month
FIGURE 4.29 Comparison of litterfall from the mangrove forests in Biosphere 2 and Southwest Florida. (Adapted from Finn, M. 1996. Comparison of Mangrove Forest Structure and Function in a Mesocosm and Florida. Ph.D. dissertation, Georgetown University, Washington, DC.)
tests our understanding to the limit." (Lawton, 1995), and "If ecologists can learn to construct mesocosms that replicate important characteristic of natural ecosystems, this will provide unmistakable evidence that they really understand how these ecosystems function" (Cairns, 1988). Thus, the ability to create microcosms that adequately model real ecosystems not only allows experiments to be conducted and extrapolated, but this ability also provides a practical measure of how well real ecosystems are understood. This is the notion that Nixon (2001) was referring to when he stated that "every mesocosm is a living hypothesis."
It also may be possible to learn from microcosms that fail or from those that do not match with natural analogs. Most researchers have not followed this line of thinking and examples are not easy to find. One example of a system that "failed" was embedded in the Everglades mesocosm of the Smithsonian Institution. The problem here can be seen by comparing maps of the system soon after construction (see Figure 7 in Adey and Loveland, 1991, p. 580) and after 7 years of self-organization (see Figure 7 in Adey and Loveland, 1998, p. 417). This mesocosm was intended to be an abstract model of the gradient of systems that make up a transect across the Southwest Florida Everglades from freshwater (tank 7) through estuarine (tanks 2 to 6) to marine (tank 1) habitats (Figure 4.15). If success is judged by the stability of the ecological components within the tanks, then the overall system was a success because there was little change between the initial map and the map after 7 years of change. In fact, it was a remarkable system in containing so much biodiversity characteristic of the Everglades in engineered gradients of tide and salinity. However, tank 5 did change. This tank was originally intended to model a saltmarsh system with dominance of grasses and succulent herbs, as can be seen in the map from the first edition of Adey and Loveland's book. After 7 years, tank 5 changed to a mangrove-dominated system with white and red mangroves and mangrove ferns, as can be seen in the map from the second edition of Adey and Loveland's book (Figure 4.15). The change was due to natural succession in which mangroves were able to outcompete saltmarsh (Kangas and Lugo, 1990). In essence then, the "failure" of tank 5 is actually a verification of a successional hypothesis. Saltmarsh exists in South Forida either temporarily or in locations that exclude easy mangrove establishment. However, in the close confines of the greenhouse meso-cosm environment, saltmarsh was not able to find a refuge from the competitively superior mangrove system and it was out-competed.
In other cases ecosystems have emerged in microcosm experiments that do not match natural analogs due to scaling problems. Oviatt et al. (1979) found super blooms of phytoplankton in a pelagic microcosm which were difficult to explain but were probably due to altered light climates in the lab vs. the field. The best examples of emerging new systems are probably those attached to walls of microcosms. These can sometimes become interesting in themselves even though they confound the intended systems. For example, Twinch and Breen (1978) found an attached system develop on their limnocorrals that included about 90 snails/m2 supported by the growth of algae on the walls. Margalef (1967) recognized the significance of these types of wall growths in plankton experiments, even though they are unintended:
The attached species progressively invade all the flasks as time advances. The concentration of organic matter on the walls and the absorption of light are new factors and the whole pattern becomes blurred. The elegant simplicity of the experiments with free-floating algae is lost. The brutal competition for dominance based on the rates of increase has given way to more subtle and interminable processes and the chemostat is prevented from attaining a stationary state. The situation is interesting as an example of development of more organization than the experimenter desires.
His last sentence in this quote is particularly relevant in suggesting how ecosystems, which surprise the experimenter by generating more organization than was intended, can emerge. Perhaps the most unusual microcosms that indicate emergence of new ecosystems are systems that are given control over their own inputs. The first example of this type was an unpublished experiment by Beyers (1974; Kania and Beyers, no date) that was diagrammed by H. T. Odum (1983) in Figure 4.30. Here Beyers' flask microcosms were interfaced through a pH meter to the timer that controlled the lights in the growth chamber where the experiment took place. Changes in pH occur diurnally with uptake and release of CO2 in photosynthesis and respiration, respectively. Thus, the interfaced microcosm could control the duration of lighting through its own metabolism. Figure 4.31 shows light-dark patterns of the three replicate systems. Periods of lights "on" and lights "off' are plotted. Two of the replicates had longer "on" periods than "off' periods and they both eventually evolved to turn the lights on continuously. The third system had longer "off' periods than "on" periods, and it never turned on the lights continuously. This is a remarkable experiment that never got published for one reason or another. This story becomes even more interesting because Petersen (1998) performed an
independent experiment as part of his dissertation research that was very similar to Beyers' original work, without knowledge of it. He interfaced pelagic microcosms with an oxygen electrode and set up a routine whereby the systems could control the length of the light and dark periods. In general, his results were similar to Beyers' results, with the microcosms generating alternating periods of light and dark through their metabolism (Petersen, 2001). Both of these experiments are remarkable because they create an ecosystem that never existed previously: one that can control its energy source. In fact, many kinds of emerging new ecosystems can be developed in microcosms through creative ecological engineering. Every microcosm is a new ecosystem that never existed previously, even if it is intended to model a natural analog. Some of the most interesting microcosms may be those with strange, artificial scaling because they show what is possible in ecosystem development. In this sense, the differences between microcosms and their natural analogs are opportunities to learn, perhaps about some fundamental property of a simple food chain or of a succession sequence, or perhaps some great new truth of ecology.
Was this article helpful?