Organic molecules are deterministically assembled in variably inundated river sediments, but drivers remain unclear.

Dissolved organic matter (DOM) is vital to ecosystem functions, influencing nutrient cycles and water quality. Understanding the processes driving DOM chemistry variation remains a challenge. By examining these processes through a community ecology perspective, we aim to understand the balance between stochastic forces (e.g., random mixing of DOM) and deterministic forces (e.g., systematic loss of certain types of DOM molecules) shaping DOM chemistry. Previous research on stochastic and deterministic influences over DOM chemistry applied null models to aquatic environments and subsurface pore water. Our study extends this to variably inundated riverbed sediments, which are widespread globally. We studied 38 river reaches across biomes, finding that DOM chemistry within most sites was governed by deterministic processes that were highly localized and led to spatial divergence in DOM chemistry within each reach. The degree of determinism varied substantially across reaches and we hypothesized this was related to differences in sediment moisture. Our findings partially supported this, showing that the upper limit of determinism decreased with increasing sediment moisture. We integrated our results with previous studies to develop a post-hoc conceptual model proposing that DOM assemblages become more deterministic along the continuum from river water to saturated sediment pore spaces to drier sediments or soils. This conceptual model aligns with previous work linking DOM chemistry to the Damköhler number and hydrologic connectivity, suggesting generalizable patterns and processes that can be further revealed by quantifying the stochastic-deterministic balance through space, time, and across scales.