Microcosms were sampled 4 days after the start of the selection experiment, and on the day of the common garden before inoculation of the common garden (day 78), each time by taking a sample of 500 µl. The common garden experiment was sampled 4 days after inoculation, extracting two samples of 250 µl each on day 82. Video recording and automated analysis were used to collect data on population densities (number of cells) and traits (bio-area, aspect ratio, and gross speed)67 (link),68 (link). We recorded 20 s videos (25 frames per second) of an effective sampled volume of 34.4 µl using a Leica M205C stereomicroscope at a 16-fold magnification and an Orca Flash 4 camera (Hamamatsu). Videos were analyzed using a customized version of the Bemovi package in R (available on https://github.com/efronhofer/bemovi 67 (link)).
Data output of the videos was visually inspected, and upon analysis, data entries reflecting P. aurelia or S. teres particles with a bio-area smaller than 1000 µm2 were removed from the data as these reflected non-living particles that were accidentally tracked. After species identification, videos were again visually inspected to ensure the identification was performed correctly. Trait values of those videos that contained only one of the species were manually assigned to the correct species identity when needed.
Data output of the videos was visually inspected, and upon analysis, data entries reflecting P. aurelia or S. teres particles with a bio-area smaller than 1000 µm2 were removed from the data as these reflected non-living particles that were accidentally tracked. After species identification, videos were again visually inspected to ensure the identification was performed correctly. Trait values of those videos that contained only one of the species were manually assigned to the correct species identity when needed.
Free full text: Click here
