TERRESTRIAL DETRITUS AND THE CARBON CYCLE 307 



discouraging to attempts to relate CO,; evolution to soil or air temperatures and 

 to attempts to estimate rates with short-term measurements. 



An even more serious question arises from the possibility that, in such deep 

 detritus pools as northern forest floors or grassland soils, these C0 2 measure- 

 ments are controlled by physical processes and are only indirectly related to 

 biological rates of decomposition. Carbon dioxide is produced chiefly by 

 biological respiration and enriches soil air in C0 2 at the expense of oxygen. 

 Although most respiration occurs near the surface, the C0 2 content of soil air 

 generally increases with depth. Baver 2 reviewed some of the older literature 

 showing C0 2 concentrations in soil air as high as 15%, with considerable 

 variation with depth and season. Soil air exchanges with surface air at rates 

 controlled by diffusion, soil-temperature changes, barometric variations, wind 

 action, and pore displacement and gas transport by percolating rainwater. 2 Of 

 these factors, Baver felt that diffusion was the most important and the others 

 were minor. 



If the high potential air volume of soil (average porosity = 50%) and C0 2 

 contents 24 of up to 15% are considered, it is clear that soil has an enormous 

 storage capacity for C0 2 . If diffusion is the principal agency controlling C0 2 

 flux upward and across the soil— air interface, then C0 2 diffusion will be 

 responsive to concentration, temperature, and pressure gradients, and possibly to 

 surface ventilation. Only when C0 2 diffusion is in balance with C0 2 



production by respiration will C0 2 evolution measure metabolic activity of the 

 soil. Considering the C0 2 storage volume of soils and the complex influences on 

 diffusion, a balance between C0 2 production and C0 2 diffusion out of the soil 

 may be transitory and fortuitous. At the very least, it is mandatory that methods 

 of measuring C0 2 evolution do not influence normal diffusion processes in soils. 

 It is doubtful whether most of the apparatus described in the literature meets 

 this requirement. 2 5 This fundamental problem is probably of greatest concern 

 when deep detritus pools are involved; such systems as tropical forests may have 

 extremely active, shallow forest floors in which diffusion is less important than 

 ventilation. It is also possible that surface-layer metabolism (a fast process) is so 

 much more productive of C0 2 than deeper layer diffusion (a slow process) that 

 measurements of C0 2 evolution are mainly controlled by actual C0 2 production 

 near the surface. 



Inasmuch as C0 2 evolution measures total carbon coming from the detritus 

 pools and soil, it must include contributions from root respiration and 

 metabolism of mycorrhizal fungi. Harley 28 reviewed estimates of mycorrhizal 

 respiration which led him to doubt that measurements of "soil respiration" 

 would be as useful in estimating decomposition processes as commonly thought. 

 Certainly root respiration alone presents a highly significant addition to soil C0 2 

 content, and compartmental separation of root respiration from decomposition 

 metabolism is a difficult but necessary task. A few workers are attempting such 

 compartmentalization, 22 ' 29 but such attempts require rather broad assumptions 



