Statistical evaluation of a newly modified Robbins device using a bioluminescent pseudomonad to quantify adhesion to plastic

A modified Robbins device (MRD) has frequently been used as a model system to study adhesion and biofilm formation. This study investigates the reproducibility of attachment and whether a statistically significant gradient of adhesion exists along the 25 sampling ports of a MRD. A simple, quantitative, non‐destructive, bioluminescence assay was developed in order to measure attachment of bioluminescent P. veronii BL146bio cells to plastic discs of Thermanox? in newly modified Robbins devices (nMRD). No statistically significant difference in mean bioluminescence values occurred between pairs of nMRDs run in parallel, but there was a significant difference in bioluminescence values between different batches of bacteria (p < 0.05). Generalised Linear Modelling showed that the position of the sample disc influenced the numbers attaching. In 50% of devices a significant positive gradient of attachment occurred and bioluminescence values varied from disc 1 to disc 25 by 29.6–58.0%. In the other 50% of nMRDs there was a smaller, non‐significant gradient. A disc sampling regime was devised to take this gradient into account and used to prove a positive correlation between bioluminescence and numbers of viable P. veronii BL146bio cells during a 6h biofilm accumulation period.