Using litter chemistry controls on microbial processes to partition litter carbon fluxes with the Litter Decomposition and Leaching (LIDEL) model. (September 2016)

• Record Type: Journal Article
• Authors: Campbell, Eleanor E.; Parton, William J.; Soong, Jennifer L.; Paustian, Keith; Hobbs, N. Thompson; Cotrufo, M. Francesca
• Is Part Of: Soil biology and biochemistry. Volume 100(2016)
• Page Start: 160
• Page End: 174
• Publication Date: 2016-09
• Subjects: Litter decomposition model -- Leaching model -- Microbial carbon use efficiency -- Hierarchical Bayesian modeling -- Lignocellulose -- Nitrogen; Soil biochemistry -- Periodicals; Soil biology -- Periodicals; Sols -- Biochimie -- Périodiques; Sols -- Biologie -- Périodiques; Sols -- Microbiologie -- Périodiques; Bodembiologie; Biochemie; 631.46
• DOI: 10.1016/j.soilbio.2016.06.007
• Languages: English
• ISSNs: 0038-0717

Abstract

Abstract: New understanding of the connection between dynamic microbial carbon use efficiency (CUE), litter decomposition products, and pathways of soil organic carbon (SOC) formation have not been fully integrated into current generalizable litter decomposition models. We developed a new approach, the Litter Decomposition and Leaching (LIDEL) model, that: 1) includes leaching and formation of dissolved organic carbon (DOC), important components of vertical C movement and SOC inputs into deeper soil layers, and 2) uses widely available litter chemistry data to drive the simulation of microbial processes that partition litter during decomposition, through affecting rates of CO2 respiration versus formation of microbial biomass and microbial products. Two ecologically important but poorly understood processes explored in this analysis include 1) the relationship between litter nitrogen (N) availability and rates of microbial decay and assimilation, and 2) the efficiency of DOC generation from the decomposition and leaching of soluble- versus cellulose-dominated plant litter fractions. We tested multiple hypothesis-driven model formulations, and for each estimated initial conditions and parameters using hierarchical Bayesian approaches. We combined data from experimental results and literature review for five types of litter that vary by initial lignin and N content. Our analyses showed the LIDEL model formulations with a logistic N limitation curve gave better predictions than model formulations using a linear N limitation curve. Model formulations with higher DOC generation efficiency from the soluble litter pool yielded more variable predictions and parameter estimations (shown by consistently wider 95% Bayesian credible intervals), but may have better simulated large DOC leaching events in early decomposition. Our analyses highlight a need for targeted studies clarifying measures of soluble litter and the generation of DOC during early litter decomposition, as well as rates of microbial biomass turnover and the flux of soluble versus non-soluble microbial products. Overall, the LIDEL model provides a robust generalizable framework to express and test hypotheses connecting litter chemistry and dynamic microbial CUE with the generation of DOC and microbial products during litter decomposition. Highlights: Model of litter decay with dynamic microbial carbon use efficiency and leaching. Model integrated new hypotheses for microbial role in litter carbon partitioning. Hierarchical Bayesian methods integrated multiple sources of data and variability. A logistic curve best expressed N limitation on microbial processes. Additional data needed to resolve microbial- versus litter-derived soluble products.