trations hourly and leachate contents weekly for 1 75 days. After 90 days they treated 

 all but two systems with cadmium chloride, a known soil poison. Prior to treatment 

 leachate nutrient loss had stabilized. The addition of cadmium greatly stimulated loss 

 of calcium and other nutrients, which then gradually returned to normal (pre- 

 treatment) values. 



Using standard ecological procedures, they examined plant species diversity and 

 richness and analyzed residual nutrients in each microcosm. Through computer 

 time-series analysis of the carbon dioxide data, they found patterns of peak 

 frequencies of carbon dioxide flux inversely correlated with the extent and duration 

 of excessive nutrient loss. The microcosms with the greatest diversity, as revealed by 

 the carbon dioxide analysis, were most resistant and resilient; those with the lowest 

 diversity were most vulnerable and least resilient. Although predicted by ecological 

 theory, this relationship of ecosystem complexity to functional stability has seldom 

 been quantified so clearly. 



This study also demonstrates the superior role that microcosms can play in 

 probing the effects of pollutants under precisely controlled conditions. This 

 experiment could only have been performed in a laboratory. Even though technically 

 sophisticated, the approach is actually fairly simple. The process can be measured 

 with inexpensive, widely available equipment and does not require highly skilled 

 personnel. The technique could be applied to any given chemical or site that might be 

 impacted. 



A host of excised aquatic systems have been studied for each of the several major 

 types of ecosystems: lake, stream, estuary, and marine environment. Some systems 

 include the sediments; others concentrate on the water column. The simplest is the 

 estuarine Eco-core microcosm developed by the scientists at the EPA laboratory at 

 Gulf Breeze, Florida. It functions much like the smaller soil core systems, both in 

 terms of results of chemical metabolism and in the role it can play in evaluation. 

 Larger systems have been employed by the Narragansett and Corvallis EPA labs 

 independently to examine community structure and impact of pollutants from 

 dredging and ocean dumping of sludge, etc. The largest microcosm is the 1 1,000-gal. 

 Marine Ecological Research Laboratory (MERL) system in Rhode Island, which has 

 been employed for studies of both scale and complexity regarding microcosm 

 structure and the impact of petroleum on marine organisms. 



Several significant freshwater systems of diverse scale are also in use or have been 

 employed extensively over the past decade. The laboratory stream of Charles 

 Warren'-* and colleagues is a simple double channel connected at each end by a 

 paddlewheel to provide movement and aeration of water (Figure 5). This type of 

 system has been useful in examining problems as diverse as effects of logging on 

 forest streams, impact of pesticides on intra-species and inter-species community 

 species structure, and toxicity of pulpmill wastes to anadromous fish. Other stream 

 systems of not caret he multiple channels at the Savannah River Ecology Laboratory 

 in Aiken. Georgia. The Monticello, Minnesota channel is being used by scientists 

 from Michigan State University and two EPA labs to validate the Exposure 

 Assessment Model System (EXAMS)'" developed at the Athens, Georgia EPA lab 

 over the past several years. The EXAMS model can be used to predict chemical 

 concentration of a pollutant based on the loading or input of the chemical into a 

 stream, pond or lake. Such predictions are of great utility in comparing single-species 

 responses from laboratory studies to concentrations the test species might encounter. 

 EXAMS was derived in part from studies in the pioneering Artificial Ecosystem 

 Simulator (AECOS) model stream at Athens. 



Important studies on the criteria which might be employed in evaluation of 

 microcosms have been carried out at the Oak Ridge National Laboratory, Tennessee, 

 and Lawrence-Berkeley, California, Radiation Laboratory (LRL). These studies 

 have addressed very serious questions and problems that have plagued researchers 

 since they first began developing laboratory systems for evaluation of pollutant 



160 



