6.1.3 Microcosms
Phosphorus has a major role in determining productivity of plankton communities. The pioneer experiments on the role of phosphorus in aquatic communities were carried out by introducing radiophosphorus (32P), which was then followed as a community-level tracer.
Such experiments can be done in the laboratory, using either a plankton-and-water sample from a natural aquatic community or the small-scale community that develops through some weeks in an aquarium provided with water and nutrients and an initial seeding with pond organisms. Such systems are called microcosms.
For example, Into a 200-liter aquarium microcosm of the latter sort 100 microc of 32P-labeled phosphate (as phosphoric acid) were introduced.  After the tracer introduction the movements pattern showed the following sequence of four phases

1. There was initial, very rapid movement from water into plankton organisms and turnover from these back to water. In the experiment illustrated one half of the 32P had moved into the plankton (predominantly single-celled green algae) within two hours, and by twelve hours the distribution of 32P was in a steady-state balance between plankton and water. Other experiments, using millipore filters for fuller separation of small plankton cells and particles with bacteria from the water, have shown even faster uptake (up to half the amount in the water in three minutes) and balance (93 per cent in the plankton and particles after twenty minutes). Turnover rates (in fractions of the 32P in the plankton returning to the water per unit time) were 0.27/hr and 0.013/min for the two experiments. These rapid movements of 32P are affected by adsorption onto the surfaces of cells as well as absorption through those surfaces. The mass of the plankton is very small in relation to the water in which it is suspended. With the greater part of the phosphate in this small mass, the concentration of the 32P is many thousand times as high in the plankton as in the water. Concentration ratios (expressed as 32P content per unit dry mass of organisms divided by 32P content of an equal mass of water) are in some cases, with low phosphate content in the water, of the order of one to two million- fold.

2 Somewhat more slowly through the first few hours, the 32P moved into the filamentous algae growing on the sides and on the mud-containing bottom trays of the aquarium. Maximum 32P content per unit mass and approximate equilibrium with the water were reached for attached algae within twenty-four hours. (In general, the larger the organism, the larger the mass or pool of phosphate it contains, the lower the turnover rate for this pool, the slower the uptake per unit mass, the later the equilibrium is reached, and the slower the decline from that equilibrium if the experiment is continued long enough for such decline to occur.) As a substantial part of the 32P moved into the attached algae, the 32P in the water-and-plankton together declined. Even though equilibrium of 32P content per unit mass of algae and water was reached during the first day, the total content of 32P in attached algae continued to increase, because of the growth in mass of the algae, until the sixth day. Thereafter, because of the turnover of 32P between algae and water, there was net movement of 32P out of the algae as 32P levels in the water declined.

3. The tracer moved into animals (water fleas grazing the plankton algae, snails grazing the sidewall algae) more slowly than into their food, at rates that varied with size, food habits, and other characteristics of the animals. From the grazing animals the 32P reached the carnivorous animals (fish feeding on water fleas). Although the rate of uptake decreases along a food chain, the concentration ratios in animals may be very high.

4. Through the latter part of the experiment an increasing fraction of the tracer moved into the bottom mud, the sediment (of recently settled plankton and animal feces), and the film of microorganisms on the aquarium walls. Some turnover of 32P between these three parts of the ecosystem and the water continued. There was, however, net movement of 32P "downward"—out of the water and active circulation between water and organisms, and into the less active or bound forms of the sediment, mud, and surface films. By gradual accumulation three quarters of the 32P had moved into these pools of less active turnover by the end of the experiment, forty-five days after tracer introduction.

The tracer approach can be applied with isotopes other than 32P, but such work is limited as yet. The 32P transfer patterns are representative in broad features of the ways other substances circulate, but details for other isotopes and compounds necessarily vary. It is of interest to see whether the patterns of 32P movement in lakes are similar to those in aquaria; in general they are. 32P introduced into the surface waters of a lake is rapidly taken up by plankton, particles, and bacteria, while only a small amount remains in the water. There is a slower, longer-term movement of 32P into rooted plants along the shores of the lake, into animals, and downward into the sediments. The plants are analogous to the side-wall algae of the aquarium, but phosphorus movement through these higher plants appears to involve two rates: a faster uptake and release by the diatoms and other microorganisms forming the surface film on the plants, and a slower turnover through the tissues of the plants themselves.