Long-term tracing of whole catchment 15N additions in a mountain spruce forest: measurements and simulations with the TRACE model
- First Online:
- Cite this article as:
- Krause, K., Providoli, I., Currie, W.S. et al. Trees (2012) 26: 1683. doi:10.1007/s00468-012-0737-0
Despite numerous studies on nitrogen (N) cycling in forest ecosystems, many uncertainties remain, particularly regarding long-term N accumulation in the soil. Models validated against tracer isotopic data from field labeling experiments provide a potential tool to better understand and simulate C and N interactions over multiple decades. In this study, we describe the adaptation of the dynamic process-based model TRACE to a new site, Alptal, where long-term N-addition and 15N-tracer experiments provide unique datasets for testing the model. We describe model parameterization for this spruce forest, and then test the model with 9- and 14-year time series of 15N-tracer recovery from control and N-amended catchments, respectively. Finally, we use the model to project the fate of ecosystem N accumulation over the next 70 years. Field 15N recovery data show that the major sink for N deposition is the soil. On the control plot, tracer recovery in the soil increased from 32 % in the second year to 60 % in the ninth year following tracer addition, whereas on the N-saturated plot, soil recovery stayed almost constant from 63 % in the third year to 61 % in the twelfth year. Recovery in tree biomass increased over the decadal time scale in both treatments, to ca. 10 % over 9 years on the control plot and ca. 13 % over 14 years on the N-amended plot. We then used these time series to validate TRACE, showing that the adaptation and calibration procedure for the Alptal site was successful. Model-data comparison identified that the spreading method of 15N tracers needs to be considered when interpreting recovery results from labeling studies. Furthermore, the ground vegetation layer was recognized to play an important role in controlling the rate at which deposited N enters soil pools. Our 70-year model simulation into the future underpinned by a Monte-Carlo sensitivity analysis, suggests that the soil is able to immobilize a constant fraction of 70 and 77 % of deposited N for the treated and the control plot, respectively. Further, the model showed that the simulated increased N deposition resulted in a relatively small elevated C sequestration in aggrading wood with an N use efficiency of approximately 7 kg C per kg N added.