318 P. W. Flanagan and F. L. Bunnell 



response of microbial respiration in substrates from tundra and taiga is 

 most sensitive to temperature, then to substrate chemistry, and least sen- 

 sitive to amounts of moisture, particularly high levels of moisture. Des- 

 pite the variable response to moisture, elimination of moisture effects 

 from the model reduces its predictive ability by a minimum of 23 to 31% 

 (Bunnell et al. 1977a). Also, it is important to note that while respiration 

 rates are relatively insensitive to moisture over the range of respirometry 

 data collected, moisture levels in the field may become high enough to 

 reduce respiration significantly. Temperature, moisture and oxygen are 

 all important modifiers of the rate of respiration of tundra 

 microorganisms. 



The overall effect of the microbial population and micrometeoro- 

 logical factors can be seen in the relative decomposition rates of a uni- 

 form substrate, cellulose, placed in three different microhabitats (Figure 

 9-10). Over several years, the weight loss from the cellulose was greatest 

 in the litter, intermediate in the standing dead, and lowest in the soil. As 

 both soil and litter have more cellulolytic decomposers than standing 

 dead (Figure 9-1), the results support the hypothesis that moisture may 

 be limiting in the standing dead (Chapter 8). The considerable decrease in 

 decomposition rate noted for the soil suggests that conditions in the soil 

 are less favorable overall to decomposition than those above ground. 



Substrate Chemistry and Microbial Respiration 



Microbial respiration is assume J to be influenced by substrate chem- 

 istry as well as by temperature and moisture. This assumption is broadly 

 accommodated by coefficient aj in the function gresp. Coefficient O} 

 represents the respiration rate at 10°C when neither moisture nor oxygen 

 are limiting, and is employed to establish the upper level or amplitude of 

 the response surface for respiration. The "quality" of the substrate is 

 thus directly proportional to the magnitude of a,, which in turn is di- 

 rectly correlated with the percentage of ethanol-soluble compounds or 

 percent glucose (Bunnell et al. 1977a). The significance of a, to broad 

 patterns of respiration is estimated by the preceding sensitivity analyses. 



Earlier works (Henin et al. 1959, Minderman 1968) proposed that 

 observed rates of weight loss result from the summation of rates from 

 specific chemical components, but did not relate these rates to microbial 

 activities. Bunnell et al. (1977b) extended these earlier models of decom- 

 position to encompass not only the observed patterns of weight loss, but 

 the microbial activities producing these patterns. Ethanol-soluble com- 

 pounds disappear five to six times as fast as other constituents of natural 

 substrates (Figure 9-11). Combination of these two chemically defined 

 groups produces the common departure from a simple exponential 



