Sporadic P limitation constrains microbial growth and facilitates SOM accumulation in the stoichiometrically coupled, acclimating microbe-plant-soil model

Abstract

Requirements for biomass carbon (C), nitrogen (N), and phosphorus (P) constrain organism growth and are important agents for structuring ecosystems. Arctic tundra habitats are strongly nutrient limited as decomposition and recycling of nutrients are slowed by low temperature. Modeling interactions among these elemental cycles affords an opportunity to explore how disturbances such as climate change might differentially affect these nutrient cycles. Here we introduce a C–N–P-coupled version of the Stoichiometrically Coupled Acclimating Microbe-Plant-Soil (SCAMPS) model, “SCAMPS-CNP”, and a corresponding modified CN-only model, “SCAMPS-CN”. We compared how SCAMPS-CNP and the modified SCAMPS-CN models project a moderate (RCP 6.0) air warming scenario will impact tussock tundra nutrient availability and ecosystem C stocks. SCAMPS-CNP was characterized by larger SOM and smaller organism C stocks compared to SCAMPS-CN, and a greater reduction in ecosystem C stocks under warming. This difference can largely be attributed to a smaller microbial biomass in the CNP model, which, instead of being driven by direct costs of P acquisition, was driven by variable resource limitation due to asynchronous C, N, and P availability and demand. Warming facilitated a greater relative increase in plant and microbial biomass in SCAMPS-CNP, however, facilitated by increased extracellular enzyme pools and activity, which more than offset the metabolic costs associated with their production. Although the microbial community was able to flexibly adapt its stoichiometry and become more bacteria-like (N-rich) in both models, its stoichiometry deviated further from its target value in the CNP model because of the need to balance cellular NP ratio. Our results indicate that seasonality and asynchrony in resources affect predicted changes in ecosystem C storage under warming in these models, and therefore build on a growing body of literature indicating stoichiometry should be considered in carbon cycling projections.