110 LEFFLER 



All parameters were combined, and rankings of each treatment 

 level were assessed by Friedman nonparametric analyses of variance 

 and by sign tests (P < 0.200) (Siegel, 1956). These results are shown 

 in Table 2. Energy subsidies were not significantly related to any 

 type of stability. Constancy stability and to a lesser extent resilience 



TABLE 2 



EFFECTS ON ECOSYSTEM STABILITY OF 



INCREASING NUTRIENT AND ENERGY SUBSIDIES 



IN AN AQUATIC MICROCOSM EXPERIMENT* 



*Based on summation of all measured parameters. Friedman 

 analyses of variance and the sign test were used to detect significance 

 at the P < 0.200 level. Abbreviations H, M, and L are high, medium, 

 and low, respectively; NS is not significant. The predicted rankings 

 from most to least stable are H-M-L or H-L. Parentheses indicate ties 

 in ranking. 



stability increased with increasing nutrient subsidies, but response 

 times decreased. Increased levels of nutrient— energy combinations 

 led to decreases in resistance stability and response time. Table 3 

 summarizes the behavior predicted by the hypotheses for large 

 abiotic nutrient reserves and the empirically determined results of 

 the microcosm experiments. The observed increase in constancy 

 stability with increasing nutrients verified the predictions of both 

 Pomeroy (1975) and Webster, Waide and Patten (1975). This 

 suggests that nutrient availability may play an important role in 

 ecosystem dynamics. It is obvious, however, that the mechanisms by 

 which nutrient reserves and energy subsidies affect ecosystem 

 stability are poorly understood, as evidenced by contradictions 

 between observations and hypotheses for the four types of ecosys- 

 tem stress response. 



