Filtration artefacts in bacterial community composition can affect the outcome of dissolved organic matter biolability assays

Abstract

Inland waters are large contributors to global carbon dioxide (CO2) emissions, in part due to the vulnerability of dissolved organic matter (DOM) to microbial decomposition and respiration to CO2 during transport through aquatic systems. To assess the degree of this vulnerability, aquatic DOM is often incubated in standardized biolability assays. These assays isolate the dissolved fraction of aquatic OM by size filtration prior to incubation. We test whether this size selection has an impact on the bacterial community composition and the consequent dynamics of DOM degradation using three different filtration strategies: 0.2 µm (filtered and inoculated), 0.7 µm (generally the most common DOM filter size) and 106 µm (unfiltered). We found that bacterial community composition, based on 16S rRNA amplicon sequencing, was significantly affected by the different filter sizes. At the same time, the filtration strategy also affected the DOM degradation dynamics, including the δ13C signature. However, the dynamics of these two responses were decoupled, suggesting that filtration primarily influences biolability assays through bacterial abundance and the presence of their associated predators. By the end of the 41-day incubations all treatments tended to converge on a common total DOM biolability level, with the 0.7 µm filtered incubations reaching this point the quickest. These results suggest that assays used to assess the total biolability of aquatic DOM should last long enough to remove filtration artefacts in the microbial population. Filtration strategy should also be taken into account when comparing results across biolability assays.