Deterministic processes drive thermodynamics of stream corridor metabolites

Streams and rivers receive substantial inputs of carbon from terrestrial sources (~1.9 Pg C yr^-1) and transport it to the ocean (~0.95 Pg C yr^-1). As this dissolved organic carbon (DOC) travels downstream it interacts with microbes and undergoes significant biochemical transformations. However, there are significant gaps in our understanding of the principles governing the spatiotemporal organization of DOC thermodynamic properties, which strongly influence biogeochemical function. To understand factors governing DOC thermodynamics we leverage the recently proposed synthesis of meta-community ecology and metabolomics, termed "meta-ecosystem metabolomics." This framework treats metabolite profiles that comprise DOC as assemblages of chemical species that are analogous to ecological communities comprised of biological species. More specifically, we adapted and applied ecological null models to time-series metabolite profiles sampled in stream and porewater from the HJ Andrews experimental forest. Our results demonstrate that despite very similar bulk properties (through space and time) tied to molecular and thermodynamic characteristics (i.e., CHONSP-content, Gibbs free energy, etc.), deterministic processes nonetheless resulted in clear differentiation through time and between stream and pore waters. Dynamics in DOC molecular properties appear to be driven by biochemical transformations that vary in space and time. Furthermore, our results show that thermodynamic properties of DOC are highly constrained, despite differences in biochemical transformations; we refer to this as 'thermodynamic redundancy' as an analog to the ecological concept of functional redundancy. We further identify metabolites that are always deterministically organized, those are always stochastically organized, and those that vary through space or time in terms of deterministic and stochastic influences. Metabolites that are always deterministic are likely to be the most relevant to biogeochemical function and should be the focus of genome-scale metabolic models tied to reactive transport codes.