Scaling down the microbial loop: data-driven modelling of growth interactions in a diatom-bacterium co-culture

An intricate set of interactions characterizes marine ecosystems. One of the most important is represented by the microbial loop, which includes the exchange of dissolved organic matter (DOM) from phototrophic organisms to heterotrophic bacteria. Here, it can be used as the major carbon and energy source. This interaction is one of the foundations of the entire ocean food-web. The carbon fixed by phytoplankton can be redirected to bacteria in two main ways; either (i) bacteria feed on dead phytoplankton cells or (ii) DOM is actively released by phytoplankton (a process resulting in up to 50% of the fixed carbon leaving the cell). Here, we have set up a co-culture of the diatom Phaeodactylum tricornutum and the chemoheterotrophic bacterium Pseudoalteromonas haloplanktis TAC125 and used this system to study the interactions between these two representatives of the microbial loop. We show that the bacterium can thrive on diatom-derived carbon and that this growth can be sustained by both diatom dead cells and diatom-released compounds. These observations were formalized in a network of putative interactions between P. tricornutum and P. haloplanktis and implemented in a model that reproduces the observed co-culture dynamics, revealing an overall accuracy of our hypotheses in explaining the experimental data.