Summary and Future Work Improving model estimates of trace gas emissions from forest soils through incorporation of 15N fractionation processes into a widely used ecosystem model Zaixing Zhou, Changsheng Li, Scott Ollinger Complex Systems Research Center, Institute for the Study of Earth, Oceans, and Space University of New Hampshire, Durham, NH, 03824 USA Introduction δ15N (a) (c) Nitrogen transformations in ecosystems result in minute differences in the ratio between 15N and 14N in various N pools (expressed in δ15N ), offering potential insight into nitrogen cycling at different temporal and spatial scales. Levels of isotopic enrichment depend dynamically on the internal and external fluxes of N cycling, controlled by specific abiotic and biotic conditions (Hogberg 1997). The complexity of the N cycle has limited the development and broad application of mechanistic N cycling models. In fact, detailed process-based models capable of explaining spatial and temporal patterns of natural abundances of 15N have not yet been developed. Here, we present a new effort designed to improve the ability to simulate C and N cycling in forest soils by incorporating isotopic 15N fractionation processes into a soil biogeochemistry model known as Forest-DNDC (Li et al. 2000). Preliminary application of the model focuses on the effect of elevated N inputs on northeastern U.S. forests. Part of the motivation for this work stems from unexpectedly high observations of N retention in soil at a nitrogen addition experiment at Harvard Forest in Massachusetts. The mechanisms responsible are still unclear, but are thought to involve a combination of N gas emission and abiotic immobilization (Aber et al. 1997; Nadelhoffer et al. 2004). We expect that the revised model incorporating 15N isotopic processes could help quantify the relative importance of these two mechanisms. (b) Methods (d) N isotope fractionation are presented by the following functions in Forest-DNDC: N isotope abundance δ15N in products N isotope mass balance Figure 4. Validation and application of the revised Forest-DNDC for the 15N-labeled control treatment in the hardwood stand of the Harvard Forest Chronic N Amendment Study. a) Simulated (solid line) and observed (open squares and triangles) organic carbon stored in the wood and forest floor; b) simulated δ15N in various pools, where the arrow indicate the starting of label tracing in1991; c) simulated and observed δ15N values in various pools and along soil depth before 15N tracer additions (year 1990) and 7 years after the end of 15N tracer additions (year 1999); d) simulated and observed N2O and NO emission in 2000 and 2001. δ15N (a) (c) Figure 1. Schematic diagram of nitrogen transformations and processes affecting δ15N values in the Forest-DNDC model. Nitrification and denitrification are thought to be the most import N isotopic fractionation processes in forest soils, at least in the northeast hardwood forests of the U.S. Figure 2. Interface of revised Forest-DNDC for N isotope fractionation. (b) Results and Discussion (d) (a) Figures 4 and 5 show a validation of the revised model with the new feature of 15N prediction. The model was run for the hardwood stand with two 15N-labeled N input levels, control subject to a N input of 8 kg ha-1 per year (Fig. 4) and low N addition of 50 kg ha-1 per year (Fig. 5). Model simulations for the control consistently reflect the observations of forest carbon budget (Fig 4 a), δ15N values in vegetation and soil pools (Figs 4 b and c), and N2O and NO emissions from soils (Fig. 4 d). After 15N tracer addition in 1991 and 1992, δ15N in foliage increased initially and reached its peak in 1994. Wood δ15N followed the pattern of foliage δ15N with a one year lag. Due to soil microbial immobilization and the leaf residues returned to the forest floor, δ15N value in the forest floor increased for a decade and then slowly declined. δ15N values in the mineral soil also showed a gradual increase, while the difference along the soil profile was depleted over time. Simulated NO emissions were consistent with measured values, while the model underestimated N2O emissions. The annual total of NO emitted was about 4% of the total N input, mostly contributed by nitrification. The denitrification rate was simulated as 5% of the total N input. Similar results are found for the low N treatment. However, δ15N values in plants were less than those for control treatment. The model overestimated δ15N values in plants. Simulated NO emissions and denitrification rates amounted to 2% and 3% of the total N input, respectively. The discrepancy between simulated and observed values may be due to the fact that the model needs a new mechanism to predict the effect of additional N input on the forest, e.g. that places more importance on denitrification and/or abiotic N immobilization, as the field study suggested. (b) Figure 5. Validation and application of the revised Forest-DNDC for the 15N-labeled low N treatment in the hardwood stand of the Harvard Forest Chronic N Amendment Study. The additional N input for low N treatment is 50 kg ha-1 per year starting from year 1988. a) Simulated (solid line) and observed (open squares and triangles) organic carbon stored in the wood and forest floor; b) simulated δ15N in various pools, where the arrow indicates the starting of label tracing in 1991 and 1992; c) simulated and observed δ15N values in various pools and along soil depth before 15N tracer additions (year 1990) and 7 years after the end of 15N tracer additions (year 1999); d) simulated and observed N2O and NO emissions in 2000 and 2001. δ15N Summary and Future Work 15N isotopic processes were incorporated into Forest-DNDC. The preliminary validation and application of the revised model shows that the new feature of predicting 15N could help to interpret 15N abundance in soil-plant systems and therefore N cycling. The model performed better under natural conditions than N addition experiments. As denitrification normally has a strong fractionation against 15N, the model is expected to help to investigate the importance of denitrification along with validated N uptake by plants, soil 15N abundance, and N2O and NO fluxes. In the future, further efforts will involve improvements to the model prediction of the effects of elevated N input on forests, especially plant biomass and nitrate leaching. With these improvements, the model may be better able to reliably quantify the fraction of N input lost through denitrification. Figure 3. Estimated long term organic carbon (a) and δ15N (b) in various pools at Harvard Forest. The model was run for 120 years with the initial δ15N values in the soil and vegetation pools as zero under the climate, soil and vegetation conditions in year 1987 (forest age 40). It takes 30 years for the hardwood stand to reach maturity, at about age 70. The soil organic carbons remain stable during the simulation. δ15N values in foliage and wood decrease gradually for 60 years to relatively stable levels of -3.1 and -2.0 at age 100, consistent with the observed values of -2.9 and -2.1 at age 42, respectively. δ15N values along the soil profile show a clear increasing trend from the forest floor, to surface mineral soil (5cm deep) to deep mineral soil (10-20 cm deep). They remain relatively stable after 100 years at age 140 with the values of -3.0 for forest floor, 0.36 for surface mineral soil and 6.7 for deep mineral soil. These values are somewhat consistent with the observed values of -2.3, 4.6 and 7.1 at age 42, respectively. References Aber J, McDowell W, Nadelhoffer K, Magill A, Berntson G, Kamakea M, McNulty S, Currie W, Rustad L, Fernandez I (1998). Nitrogen saturation in temperate forest ecosystems. BioScience 48:921-934. Hoegberg P (1997) 15N natural abundance in soil-plant systems. New Phytologist 137:179-203 Li CS, Aber J, Stange F, Butterbach-Bahl K, Papen H (2000). A process-oriented model of N2O and NO emissions from forest soils: 1. Model development. Journal of Geophysical Research-Atmospheres 105:4369-4384. Nadelhoffer KJ, Colman BP, Currie WS, Magill A, Aber JD (2004). Decadal-scale fates of 15N tracers added to oak and pine stands under ambient and elevated N inputs at the Harvard Forest (USA). Forest Ecology and Management 196:89-107.