Individual Physiological Adaptations Enable Selected Bacterial Taxa To Prevail during Long-Term Incubations

Enclosure experiments are frequently used to investigate the impact of changing environmental conditions on microbial assemblages. Yet, how the incuba- tion itself challenges complex bacterial communities is thus far unknown. In this study, metaproteomic profiling, 16S rRNA gene analyses, and cell counts were com- bined to evaluate bacterial communities derived from marine, mesohaline, and oli- gohaline conditions after long-term batch incubations. Early in the experiment, the three bacterial communities were highly diverse and differed significantly in their compositions. Manipulation of the enclosures with terrigenous dissolved organic car- bon resulted in notable differences compared to the control enclosures at this early phase of the experiment. However, after 55 days, bacterial communities in the ma- nipulated and the control enclosures under marine and mesohaline conditions were all dominated by gammaproteobacterium Spongiibacter. In the oligohaline enclo- sures, actinobacterial cluster I of the hgc group (hgc-I) remained abundant in the late phase of the incubation. Metaproteome analyses suggested that the ability to use outer membrane-based internal energy stores, in addition to the previously de- scribed grazing resistance, may enable the gammaproteobacterium Spongiibacter to prevail in long-time incubations. Under oligohaline conditions, the utilization of ex- ternal recalcitrant carbon appeared to be more important (hgc-I). Enclosure experi- ments with complex natural microbial communities are important tools to investi- gate the effects of manipulations. However, species-specific properties, such as individual carbon storage strategies, can cause manipulation-independent effects and need to be considered when interpreting results from enclosures. This study was financially supported by the SAW-funded ATKiM project, which provided funds to D. P. R. Herlemann, C. Meeske, K. Jürgens, S. Markert, and T. Schweder. D. P. R. Herlemann was also supported by the European Regional Develop- ment Fund/Estonian Research Council funded Mobilitas Plus Top Researcher grant MOBTT24. We thank the crew and captain of the RV Meteor (M86, M87) for support during the research cruise. The computations were performed on resources provided by the Swedish National Infrastructure for Computing (SNIC) at the PDC Centre for High Performance Computing (PDC-HPC) and Uppsala Multidisciplinary Center for Advanced Computational Science (UPPMAX). We thank Jana Matulla for excellent technical assis- tance and Stephan Fuchs for his help and advice in MS database construction. We also thank Stefan E. Heiden for valuable help with the CDD BLAST analyses. This study was financially supported by the SAW-funded ATKiM project, which provided funds to D. P. R. Herlemann, C. Meeske, K. Jürgens, S. Markert, and T. Schweder. D. P. R. Herlemann was also supported by the European Regional Develop- ment Fund/Estonian Research Council funded Mobilitas Plus Top Researcher grant MOBTT24. We thank the crew and captain of the RV Meteor (M86, M87) for support during the research cruise. The computations were performed on resources provided by the Swedish National Infrastructure for Computing (SNIC) at the PDC Centre for High Performance Computing (PDC-HPC) and Uppsala Multidisciplinary Center for Advanced Computational Science (UPPMAX). We thank Jana Matulla for excellent technical assis- tance and Stephan Fuchs for his help and advice in MS database construction. We also thank Stefan E. Heiden for valuable help with the CDD BLAST analyses.