ECOSYSTEM RESPONSES TO STRESS 107 



able. This approach is subject to the appropriateness of the 

 experimental treatments for simulating mechanisms postulated by 

 the hypothesis. Negation of a hypothesis by microcosm experiments 

 limits its generality and, at least, indicates that caution is advisable in 

 extrapolating the hypothesis to all types of ecosystems. Several 

 examples of this procedure will illustrate the microcosm approach. 



The Diversity-Stability Hypothesis 



The diversity— stability hypothesis states that increasing diversity 

 within an ecosystem increases the system's stability. The first step in 

 evaluating this proposal by microcosm experiments was to define 

 operationally and to clarify the statement to be tested. Diversity was 

 defined simply as the total number of species within each micro- 

 cosm. Each of the five stability measures in Fig. 1 were evaluated. 

 Thus the hypothesis predicted that microcosms with greater numbers 

 of species should have greater constancy stability and that under 

 stressed conditions they should have greater resistance, response 

 time, resilience, and total relative stability. 



Four types of microcosms, each containing a different number of 

 taxonomic groups, including algae, protozoans, metazoans, and 

 bacteria, were established in 500-ml cotton-stoppered Erlenmeyer 

 flasks. System A contained 10 species; system B, 17; system C, 21; 

 and system D, 25. Each system was developed from stock cultures 

 that were originally derived intact from natural sources but had been 

 grown in flasks in the laboratory for a minimum of 6 months. Thus 

 the organisms had coevolved histories and were adapted to the 

 laboratory environment. Sterile techniques ensured against invasion 

 by additiongJ taxa. 



Fifty -four replicates of each system were used in the study. All 

 microcosms were maintained under identical environmental condi- 

 tions. After inoculation the systems underwent a 12-week succession 

 to allow each to attain a steady-state condition. The normal 

 operating range and constancy stability were determined for each of 

 the four systems by sacrificing unperturbed control microcosms 

 during weeks 13 to 18. Other microcosms were stressed by raising 

 their temperatures from 24 to 45°C during weeks 14 to 16. These 

 systems were sacrificed at intervals from weeks 13 to 40 to monitor 

 their response to the stress. Each microcosm was characterized by its 

 net day production, night respiration, production-to-respiration 

 (P/R) ratio, microscopic counts of dominant organisms, and nutrient 

 agar plate counts of bacterial types. An element distribution index 

 (EDI) was used to characterize the cycling of phosphorus and iron. 

 This is an isotope technique that is sensitive to changes in the size of 



