Diverse prokaryotic communities consume and transform a broad suite of molecules in the dissolved organic matter (DOM) pool, which controls major biogeochemical cycles. Despite methodological advancements that provide increasingly more detailed information on the diversity of both prokaryotic communities and DOM components, understanding how these two component parts are structured to influence ecosystem functioning remains a major challenge in microbial ecology. Using empirical data collected along a gradient of productivity in the Labrador Sea, we characterized relationships among DOM compounds, metabolic processing, and prokaryotic diversity by structuring prokaryotic communities using spatial abundance distribution (SpAD) modeling. We identified strong associations of different SpAD taxonomic groups with specific organic substrates as well as with metabolic rates. Amplicon sequence variants (ASVs) with more cosmopolitan distributions (i.e. normal‐like) such as Bacteroidia were related to fresher DOM substrates such as free and combined amino acids whereas rare ASVs (i.e. logistic) like δ‐proteobacteria were associated with complex forms of organic matter. In terms of ecosystem function, rates of respiration and production were most strongly predicted by the abundance of certain SpAD taxonomic groups. Given the importance and complexity of linking environmental conditions, prokaryotic community structure, and ecosystem function, we propose a framework to bridge the gap between prokaryotic diversity, microbial ecology, and biogeochemistry among methods and across scales. Our work suggests that SpAD modeling can be used as an intermediate step to link prokaryotic community structure to both finer DOM details and larger ecosystem scale processes.