TL;DR
This paper introduces GOLUM-CNP, a novel data-driven model that integrates carbon, nitrogen, and phosphorus cycles in terrestrial ecosystems. By combining existing carbon cycle models with new global datasets for nutrient inputs and losses, it provides a more consistent global picture of these crucial biogeochemical cycles and their interactions.
What Problem Does This Paper Address?
Understanding how terrestrial ecosystems cycle carbon (C), nitrogen (N), and phosphorus (P) is critical for predicting their response to climate change and rising atmospheric CO2. While C-cycle models are relatively advanced, current models struggle to accurately represent N and P cycles due to a lack of comprehensive global data for validation. This leads to significant uncertainties in estimates of nutrient pool sizes, fluxes, and turnover rates. The authors aim to bridge this gap by developing a data-driven framework that can provide a more robust and globally consistent diagnostic of these coupled cycles.
Methods
The core of this work is the GOLUM-CNP (Global Observation-based Land-ecosystems Utilization Model of Carbon, Nitrogen and Phosphorus) model. It builds upon the existing CARbon DAta MOdel fraMework (CARDAMOM) for the C cycle. The key innovation is the integration of spatially explicit, data-driven estimates for N and P inputs (deposition, fixation, weathering) and outputs (denitrification, leaching, fire). The model operates under a steady-state assumption, meaning it represents an equilibrium state of the biogeochemical cycles under current climate conditions. This assumption simplifies the model and allows for direct comparison with other models that also simulate steady states. The model uses observed stoichiometric ratios (C:N:P) to link the C, N, and P cycles.
Datasets
The study leverages a variety of global, spatially explicit datasets:
- C-cycle data: Derived from the CARDAMOM model, which itself is constrained by observations of MODIS LAI, soil C, burned area, and tropical biomass.
- N-cycle data:
- N deposition: Spatially explicit estimates from Wang et al. (2017).
- N fixation: Spatially explicit estimates from CABLE model simulations (Peng et al., 2018).
- N:C ratios: Biome-scale synthesis from Zechmeister-Boltenstern et al. (2015).
- Leaching: Derived from soil moisture and runoff data (ECMWF Interim Reanalysis, GRDC).
- Denitrification fraction: Based on global soil 15N data sets (Goll et al., 2017b).
- P-cycle data:
- P deposition: Spatially explicit estimates from Wang et al. (2017).
- P weathering: Derived from river discharge and mineral composition data (Hartmann et al., 2014).
- P:C ratios: Biome-scale synthesis from Zechmeister-Boltenstern et al. (2015).
- Labile soil P fraction: Based on Hedley soil P fractions (Yang and Post, 2011).
- Biome classification: A global land-cover map from ESA CCI (LC_CCI) was used to aggregate data into seven major terrestrial biomes.
Key Insights and Ideas
- Internal P recycling is crucial: The study highlights that new P inputs (deposition, weathering) contribute a much smaller fraction to plant uptake than new N inputs. This underscores the critical importance of internal P recycling within ecosystems to support plant growth.
- Tropical forests show unique nutrient efficiency patterns: Tropical forests exhibit the lowest N-use efficiency (NUE) and the highest P-use efficiency (PUE) among forest biomes, suggesting an adaptation to the typically low P availability in these regions.
- NPP allocation is a major source of uncertainty: Sensitivity analyses reveal that NPP allocation fractions (how much carbon goes to leaves, roots, wood) are a dominant driver of uncertainty in nutrient-use efficiencies, particularly for forest biomes. This points to the need for more precise data on how plants partition carbon.
- Openness of nutrient cycles varies significantly: The study quantifies the "openness" of N and P cycles (proportion of nutrient uptake from external inputs). While N inputs are more significant, P recycling plays a more dominant role in supporting plant growth, especially in P-limited ecosystems.
Key Results
- GOLUM-CNP successfully calculates steady-state N- and P-pool sizes and fluxes globally for major biomes.
- New N inputs supply over 20% of total plant uptake in most forest ecosystems, but less in boreal forests and grasslands.
- The model provides estimates for nutrient-use efficiencies, openness, and residence times across biomes, offering a consistent global picture.
- The study identified significant differences in nutrient cycling between biomes, with tropical forests showing distinct NUE and PUE patterns.
- Sensitivity analysis indicated that NPP-allocation fractions and stoichiometric ratios are key drivers of uncertainty in the model's nutrient-use efficiency estimates.
Limitations and Future Work
- Steady-state assumption: The model assumes a steady state, which may not accurately reflect current terrestrial ecosystems that are actively responding to climate change and elevated CO2. A transient simulation is needed for future projections.
- Data uncertainties: While the model integrates many datasets, uncertainties remain, particularly in N fixation and P deposition estimates, which can lead to differences in calculated openness. The authors emphasize the need for future data syntheses that include systematic uncertainty evaluations.
- Process simplification: Some processes, like organic N uptake and N inputs from rock weathering, were not explicitly included due to data limitations or model structure.
- Biome-scale vs. finer-scale variability: The study uses biome-averaged stoichiometric ratios, but acknowledges that significant variability exists within biomes, suggesting future work could incorporate trait-based approaches.
- Future work: The authors suggest extending the model to a transient state, incorporating more detailed processes, and improving the representation of stoichiometric variability. The GOLUM-CNP framework is presented as a traceable tool for comparing and evaluating other complex dynamic global vegetation models (DGVMs).
Summary updated