Item

Abstract

Three different approaches were used to calculate heterotrophic soil respiration (Rh) and soil carbon dynamics in an old-growth deciduous forest in central Germany. A root and mycorrhiza exclosure experiment in the field separated auto- and heterotrophic soil respiration. It was compared to modeled heterotrophic respiration resulting from two different approaches: a modular component model of soil respiration calculated autotrophic and heterotrophic soil respiration with litter, climate and canopy photosynthesis as input variables. It was calibrated by independent soil respiration measurements in the field. A second model was calibrated by incubation of soil samples from different soil layers in the laboratory. In this case, the annual sum of Rh was calculated by an empirical model including response curves to temperature and a soil moisture. The three approaches showed good accordance during spring and summer and when the annual sums of Rh calculated by the two models were compared. Average Rh for the years 2002-2006 were 436 g C m(-2) year(-1) (field model) and 417 g C m(-2) year(-1) (lab-model), respectively. Differences between the approaches revealed specific limitations of each method. The average carbon balance of the Hainich forest soil was estimated to be between 1 and 35 g C m(-2) year(-1) depending on the model used and the averaging period. A comparison with nighttime data from eddy covariance (EC) showed that EC data were lower than modelled soil respiration in many situations. We conclude that better filter methods for EC nighttime data have to be developed.