Controls on Forest Soil Carbon

Printer-friendly version

Changes in tree species composition and forest productivity can alter the sources and quality of soil organic matter (SOM) inputs to forest soils, yet the relative importance of these sources remains weakly understood and quantified.  We examined controls on SOM quantity and quality using the Detrital Input and Removal Treatment Experiment, established at the Harvard Forest LTER (MA), HJ Andrews LTER (OR),  Allegheny College Bousson Experimental Forest (PA), University of Wisconsin Arboretum, and the Sikfokut International LTER site (Hungary).  Detrital inputs were altered by excluding roots, leaves, or both, or by doubling annual leaf litter inputs.  Despite 20 years of doubled leaf litter inputs at Harvard Forest and Bousson, mineral soil C remained unchanged.  In contrast, doubling litter inputs increased surface mineral soil C approximately 30% after eight years at Sikfokut and 40% after 50 years at the Wisconsin site. Greater mineral soil C at the Sikfokut site is likely an effect of low soil moisture and hence reduced decomposition.  At the Wisconsin site, five decades of additional litter inputs finally resulted in increased soil C.

Halting litter inputs resulted in a range of responses.  Reduction of leaf or root litter inputs did not alter mineral soil C at the Harvard Forest.  In contrast, Bousson mineral soil C was reduced upon cessation of leaf litter inputs but not root litter inputs; conversely, mineral soil C at the Sikfokut site was reduced by halted root inputs, but not halted leaf inputs. The variable responses to halted litter inputs are likely due to differences in litter quality and input quantity.  For example, at Bousson, leaf litter is dominated by rapidly decomposing sugar maple and black cherry, however total leaf litter inputs are greater than root litter inputs, thus resulting in relatively equal net contributions from each source to total SOM.  Suberin and cutin biomarkers indicate that Bousson SOM is comprised equally of leaf- and root-derived litter sources.  At Harvard Forest and Sikfokut, litter inputs are dominated by oak leaves, which, with slow rates of decay, allow greater C retention of this C input into the SOM pool. 

 Differential scanning calorimetry at Bousson indicated no difference in the degradation quality of SOM among the treatments.  Conventional paradigms of SOM decomposition suggest that alterations in C input rates would affect labile SOM pools more than recalcitrant pools, but these predicted shifts in SOM quality were not borne out by the thermal analysis data. Interestingly, there were no differences (after 12 yr of treatment) in the mean residence time of light or heavy soil C fractions at Bousson, also indicating that SOM of different quality was subject equally to decomposition.

 This work shows that increased forest productivity in current forests does not necessarily lead to increased soil C, at least not in the short term.  Changes in forest productivity or composition that modify sources of SOM are likely to alter SOM pools.