TERRESTRIAL DETRITUS AND THE CARBON CYCLE 



311 



tundra, 47 but it is difficult to know if this is generally true. The value of 95% 

 may be too high, but, since the worldwide production by these types is so low, 

 the error makes little difference in the total estimate. The 95% estimate for 

 desert scrub is based on the Chew and Chew 48 estimate of 2% loss to small 

 mammals in Larrea desert and on my assumption that insects and large mammals 

 graze a further 3%. 



The estimate of 50% loss to detritus for agricultural land may be too high. A 

 more rigorous estimate could be developed by a careful review of distribution of 

 production in crop plants and the degrees to which roots and stems, as well as 

 fruits, are utilized. 



The Second Approximation 



The second approximation results from Whittaker and Likens' estimates of 

 biomass 40 in major ecosystem types together with averages of Rodin and 

 Bazilevich's 49 estimates of total detritus ("litter") as percentages of biomass 

 (Table 3). The number of examples on which these averages were calculated 



TABLE 3 



A SECOND APPROXIMATION OF CARBON TURNOVER IN 

 TERRESTRIAL DETRITUS* 



*Based on estimates of world net biomass 40 and on estimates of detritus input as 

 percentages of biomass in ecosystem types by Rodin and Bazilevich. 49 Carbon has been 

 assumed to equal one-half of dry weight. 



tAverage of Rodin and Bazilevich's grass tugai (70%) and a Thuja swamp (3%). 44 



$ Average for tropical forests, Table 53 of Ref. 49. 



§ Average for numbers 7 to 14, Table 9 of Ref. 49. 



1 Author's estimate. 



