Description of a plankton filtration bias in sequencing-based community analysis and of an Arduino microcontroller-based flowmeter device that can help to resolve it

Sequencing-based diversity studies of aquatic picoplankton (bacterioplankton) communities using size-class filtration, DNA extraction, PCR and sequencing of phylogenetic markers, require a robust methodological pipeline. Biases have been demonstrated essentially at all levels, DNA extraction, primer choice and PCR. Even different filtration volumes of the same water sample and, thus, different biomass loading of the filters, can distort the sequencing results. We attempted to normalize biomass loading during filtrations, independent of plankton density in water samples, and designed a microcontroller-based flowmeter device for monitoring the decrease of initial flowrate as a proxy for increasing biomass loading and clogging of the filters. Test filtrations using freshwater plankton (Lake Constance) were done, DNA was extracted and an amplicon of the 16S rDNA was sequenced. The impact of filter loading was examined both at the community and at the ASV level.