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



pattern of simulated microbial plus root respiration (Figure 9-12). There 

 are at least two reasons for this early season disparity between simulated 

 and measured values. The soil temperature employed in the model is that 

 at 5 cm depth. Early in the season field temperatures near the surface 

 permit respiration while simulated temperatures at 5 cm depth are too 

 low to allow significant respiration. Because the lower threshold for root 

 respiration is higher than that for microbes, the disparity between simu- 

 lated and measured values is more obvious when roots are considered. 

 The release of carbon dioxide trapped during freeze-up is not simulated 

 but will appear in field measures (Benoit, pers. comm.; Coyne, pers. 

 comm.), and this also contributes to the early spring disparity between 

 field and simulated values. 



Over the 85-day sample period the accumulated totals of measured 

 respiration and simulated whole system respiration are 159 and 165 g C 

 m"^ respectively. The difference between measured and simulated values 

 over this period is thus 6 g C m"^ or 3.797o of the measured values. A dis- 

 parity of less than 5% is well within the sample error associated with data 

 for root biomass. Thus the simulated dynamics of respiration of mi- 

 crobe, root and other contributing compartments must be assumed real- 

 istic within the accuracy of available data. The proportion of simulated 

 total soil respiration that originates with the roots on any given day 

 varies between 33 and 70%, and lies at the lower end of the range of re- 

 ported values of 50 to 93% (Billings et al. 1978). Errors in different pro- 

 cesses might compensate to produce an invalid sense of accuracy. How- 

 ever, the generally realistic behavior of the model for other tundra areas, 

 including Devon Island, Moor House and Abisko (Bunnell and ScouUar 

 1981), suggests that microbial respiration for specific substrates in the 

 upper 10 cm of soil follows the relationship expressed by the function 

 GRESP. The dynamics of microbial respiration at depth are much less 

 clear and are confused by anaerobic conditions and poorly understood 

 changes in substrate quality with advancing age. 



In summary, the concepts of decomposition discussed above appear 

 sufficiently comprehensive to allow laboratory measures to be related ef- 

 fectively to field measures through the vehicle of simulation models. The 

 manner in which microbial respiration responds to temperature, mois- 

 ture and broad chemical groups predicts not only weight loss but chemi- 

 cal composition of aboveground substrates. Upper limits on decay rates 

 are established by chemical composition, but are modified by abiotic var- 

 iables. Respiration is most sensitive to temperature and the microflora 

 has responded by extending its capabilities to grow, respire and utilize 

 substrates at low temperatures. Respiration declines with both increasing 

 and decreasing moisture levels. At low moisture levels degradation and 

 loss of chemicals from standing dead vegetation is temporarily sus- 

 pended; at high moisture levels respiration becomes the province of bac- 



