Sampling
To assess the abundance, species identity, biomass, and traits of protists in each patch, we recorded videos of 0.2 ml samples from each patch every seven days, starting from day five. The only exception was on day zero, when we filmed the autotrophic and heterotrophic in the two source mixtures as a proxy for the individual autotrophic and heterotrophic patches, respectively. While the first two time points (days 0 and 5) took place before the first disturbance, all other observations were always recorded five days after the disturbance to allow communities to recover from disturbances. Each video was 5 s long and recorded using a Hamamatsu Orca Flash 4.0 (Herrsching am Ammersee, Germany) camera at 1.6x magnification.