PNNL

Abstract

Chemodiversity of dissolved organic matter (DOM) has been proposed as an ecosystem property controlling the microbial metabolism and, thus, the fate of carbon (C) in soils. New work suggests that the chemodiversity, in particular its energetic favorability, of dissolved organic matter can improve the accuracy of process-based C cycling models; however, this approach has never been validated at large scales and remains sparsely implemented. In this contribution, we assessed how the chemodiversity of DOM affects soil respiration and evaluated whether incorporating chemodiversity in kinetic models improves respiration prediction. We used paired high-resolution Fourier Transform Ion Cyclotron Resonance mass spectrometry (FTICR-MS) descriptions of DOM chemistry and soil respiration rate measurements from 63 topsoils provided by the Molecular Observation Network across the USA. First, we evaluated the relationship of DOM with soil respiration by constructing a statistical model incorporating DOM chemodiversity and other common biogeochemical variables. Regression analysis revealed that DOM alpha diversity (defined as the number of detected organic compounds) interacted nonlinearly with dissolved organic C (DOC) and water-extractable total nitrogen (WETN) concentrations in relationship to soil respiration. For soils with high DOC and WETN concentrations, increasing alpha diversity decreased soil respiration, while for soils with low DOC and WETN concentrations, increasing alpha diversity increased respiration rates. Next, we then compared three kinetic model formulations for predicting soil respiration: (1) as a function of DOC concentration only, (2) with model parameters informed by average DOM chemodiversity and by chemodiversity calculated within chemical classes of DOM. Models including DOM chemodiversity predicted respiration rates similar to those of models using only DOC concentration. The disconnect between statistical and kinetic models of soil respiration suggests that while DOM chemodiversity has the potential to increase the accuracy of soil C cycling models, current kinetic formulations do not adequately represent these dynamics. Therefore, we encourage future experimental work and decomposition model development to systematically explore mechanistic relationships between soil organic matter chemodiversity and microbial functions.