TERRESTRIAL DETRITUS AND THE CARBON CYCLE 305 



output from this compartment which might have an influence on global carbon 

 balances. The objectives of this paper are to review methods for estimating 

 carbon turnover in the detritus pool and to estimate turnover of terrestrial 

 detritus. 



METHODS OF MEASURING TURNOVER OF TERRESTRIAL 

 DETRITUS POOLS 



A wide variety of methods has been used to gain estimates of decomposition 

 rates. These methods range from microbiological incubation techniques to 

 measurement of weight loss by tethered leaves in the field. All these approaches 

 are of some value in gaining an understanding of cause and effect relations, but 

 most provide only relative measures of decomposition or are restricted to such 

 limited portions of detritus pools as litter layers or humus. Methods discussed in 

 this section include only those designed to estimate turnover of entire detritus 

 pools under field conditions. These fall into three classes, which may be 

 designated as: (1) the input method, (2) the C0 2 -evolution-measurement 

 method, and (3) the 14 C-isotope methods. 



Input Method 



This method is based on the principle that, for detritus pools in steady 

 states, outputs must equal inputs. Therefore a measure of carbon input is 

 equivalent to carbon output, and the ratio of input to average pool size expresses 

 the fractional turnover rate for the pool. The inverse of this ratio expresses the 

 average residence time for carbon. 



The input method is dependent on the assumption of a steady-state detritus 

 pool. It would obviously be inappropriate for young secondary successional 

 ecosystems, for even quite old primary successional systems, or for systems 

 undergoing some kind of disturbance such as fertilization or irrigation. On the 

 other hand, the error is likely to be small for detritus pools that are 

 asymptotically approaching steady states since the degree of change in pool size 

 may be less than errors of estimate for carbon inputs. 



This method also requires measurement of both aerial and subterranean 

 inputs. Aerial contributions are easy to measure in forests in particular, and a 

 significant volume of data on forest litterfall exists, much of it reviewed up to 

 1964 by Bray and Gorham. 12 Aerial-litterfall collection is methodologically 

 simple, but it is laborious and requires at least one annual collection cycle. 

 Furthermore, there can be wide differences in year-to-year variation. 12-14 Only 

 a few measurements on the contribution of foliar throughfall and stemflow to 

 detritus have been made, 1 5 but, while this fraction may be qualitatively 

 important, quantitatively it is probably very minor. 



Carbon contribution by underground organs is much more difficult to 

 measure, usually involving indirect methods and broad assumptions. Many of 



