Besides, the surface material itself and its chemical composition, surface roughness, and hydrophobicity have a crucial influence on the accumulation of microorganisms. In most previous studies on bacterial adhesion on titanium and ceramic surfaces, the quantity of bacterial adhesion showed a direct positive correlation with surface roughness [4, 10, 18, 24,25,26]. In case of interacting surface roughness and hydrophobicity, roughness seems to be dominant in in vitro settings [11, 25, 27]. This phenomenon is enhanced in vivo because of the sheltering effect of rough surfaces against the removal forces present in the oral cavity [10, 28,29,30]. These observations were confirmed by one of our own studies, in which in vivo and in vitro initial bacterial adhesion followed the circular surface irregularities, consisting of the grinding tracks generated by the machine manufacturing of the specimen with a lathe [25]. Nevertheless, two in vivo studies reported contradictory observations on the impact of surface roughness on bacterial adhesion. Gatewood et al. [31] and Wennerberg et al. [32] worked with volunteers who carried specimens in their periodontal pockets respectively modified implant abutments for a test period up to 4 weeks and could not find any different amounts of adhering oral biofilms, neither on rough nor on smooth titanium surfaces.

In most in vivo studies on this matter, specimens are mounted on individual splints and thus exposed to shear forces related to salivary flow, muscles, and chewing activity [4, 10, 25, 33]. With regard to the “real in situ situation,” no corresponding removal forces are present in the peri-implant region, which is protected from such forces by the adjacent peri-implant mucosa. The tight contact between the peri-implant soft tissues and the implant abutment surface protects implant surfaces from extensive shear forces. Therefore, shear forces and the influence of surface roughness may be overestimated in these specific settings. As a result, we choose a semi-static experimental setup, in which specimens were placed in an orbital shaker to simulate fluid movements in the peri-implant sulcus. This consideration was approved by the findings of Elter et al. who investigated supra- and subgingival biofilm formation on implant abutments with different roughness values. Biofilm accumulation in supragingival areas was shown to be significantly increased by higher R _a values, whereas this correlation was not found in subgingival areas [5].