Process-based simulation of nutrient (N, P, Si, Ca) effects on permafrost carbon cycling under present and future climate conditions

Abstract

Many aspects regarding biogeochemical cycles in carbon-rich permafrost ecosystems remain poorly constrained to date, resulting in a major source of uncertainty for prognostic simulations of the global greenhouse gas budget, and the associated design of effective future climate policies. Only very few studies have investigated the role of nutrients on carbon cycle processes in Northern ecosystems, and existing data is particular limited for elements beyond nitrogen (N) and phosphorus (P). Consideration of their impacts may be particularly relevant for simulating a warmer future Arctic with substantially increased thaw depths, and associated input of nutrients from currently deep-frozen permafrost pools.For the presented study, we enhanced a high-latitude version of the terrestrial ecosystem model QUINCY, which fully couples carbon (C), nitrogen (N) and phosphorus (P) cycles in vegetation and soil, with an additional first-order factor derived from soil incubation experiments that accounts for stabilization and mobilization of soil organic matter through Calcium (Ca) and Silicon (Si). In a preparation step, based on thaw depths of CMIP6 models we first computed the pan-Arctic scale magnitude of Si and Ca susceptible to release to the active layer under different climate warming scenarios. Subsequently, considering changes in the active layer depth computed by QUINCY, we calculated historical and future changes in active layer Ca and Si contents at selected sites. Element availability and associated effects on carbon cycle processes were simulated for three Siberian permafrost observatories, Chersky, Spasskaya Pad and Chokurdakh.For a historical time period, testing the Ca/Si relationship at the Chersky site for the carbon cycle resulted in a slightly improved agreement between model results and eddy covariance flux data, mainly linked to an increase in organic matter stabilization induced by higher Ca content in the soil. To illustrate the potential future implications of Ca and Si on the permafrost carbon cycle, we then compared a historical (2000 – 2020) against a future (2060 – 2080) period, with the simulations for the latter based on RCP 4.5 emissions. The substantial increase in active layer depth (0.5 – 0.8m) between these periods led to various changes in Ca/Si availability across our three study sites, including neutral to positive trends for Si and both increases and decreases for Ca. Since Ca dominated the net effect on carbon cycling, accordingly we observed both increases and decreases in GPP and ecosystem respiration linked to the consideration of Ca/Si effects, with mean changes in component fluxes reaching up to ±24 g m-2 yr-1. This implies that considering stabilization factors induced by Ca/Si, and potentially other soil minerals, could be important for a process-based reproduction of present-day permafrost carbon cycles, and projections of future scenarios.