In this study a marine microbial community was collected from 70 m depth in the Trondheim fjord in Norway and used to inoculate twelve bioreactors (volume 250 mL). The experimental had a 2x2 factorial crossover design with disturbance (yes/no) and carrying capacity (low/high) as the variables. Each experimental condition was replicated tree times, and the microbial communities were cultivated in seawater spiked with f/2- and M65-media for 50 days. The disturbances were introduced as 1:50 dilutions of the reactor volume with sterile media. Undisturbed bioreactors were operated as chemostats with a dilution rate of 1 day^-1. We switched the disturbance regime between day 28 and 29. The amount of M65 was 5x higher in “high” compared to “low” carrying capacity groups. Samples for characterization of the microbial communities were taken at day 0, 1, 2, 4, 6, 8, 12, 16, 20, 24, 28, 29, 30, 32, 36, 40, 44 and 50. Amplicons representing the variable regions tree and four of the bacterial 16S rRNA gene were sequenced on two Illumina MiSeq run.