Abstract
Technological advances have promoted the identification of microbial community compositions, but the underlying assembly mechanisms and rules governing successions remain poorly studied and understood. Previous studies tend to focus on bottom-up processes, but the focus on bottom-up selection pressures leaves much of the community variance unaccounted for. The inclusion of other processes, like top-down controls (e.g. predation), and identifying the relative influence of bottom-up and top-down mechanisms, may help close the gap and provide improved insight into previously unexplained variance. Leading to increased understanding of the rules governing community assembly and better succession outcome predictions. This study utilised a marine biofilm as an in situ model system to test and compare the effects of bottom-up (substrate type) and top-down (predation) selective pressures on community assembly over time. Different substrates were chosen based on surface properties (i.e. more degradable [wood] vs inert [plastic, tile, glass] at the times scales measured). Community development was monitored with 16S and 18S rRNA gene amplicon sequencing. Biofilm development occurred in two stages based on compositional differences: an unstable early, and stable late stage. Substrate differences shaped the establishment and early succession of the microbiome, while the influence of predation decreased substrate dependent differences and lead to distinct late successional compositions. Autotrophs : heterotroph : mixotroph ratios reflected the selective pressures influence best. Early wood communities contained 9-50% more mixotrophic Proteobacteria and Euglenozoa compared to inert substrates, which instead showed twice the abundance of autotrophic Cyanobacteria and Ochrophyta. Late non-enclosed (predated) biofilms contained 50% more autotrophs, with corresponding mixotroph and heterotroph decreases, compared to their enclosed counterparts. Additionally, enclosure status clustered with a correlation of > 18.9% within both early and established communities. The integration of a top-down selective pressure decreased unexplained variance by 18-52.6% and lead to the development of an assembly model. Early, unstable, communities are driven primarily by nutritional availability, but these differences decrease over time as community biomass increases. As the community ages higher trophic levels are recruited, until selective predation determines late community composition. The use of a bottom-up and top- down selection mechanism decreased the amount of unexplainable community variance, while enabling a deeper understanding of the underlying microbial ecological community assembly framework.