Mixotrophic Phytoflagellate Bacterivory Field Measurements Strongly Biased by Standard Approaches: A Case Study.

Bacterivory among small (≤20 μm) phytoflagellates (SP) is increasingly recognized as a globally relevant phenomenon, impacting a wide range of aspects from primary production levels to marine fisheries. However, to correctly parametrize mixotrophic SP in biogeochemical and food web models, a better understanding of the magnitude and regulation of in situ SP feeding is urgently needed. Current methods to determine SP bacterivory in the field may introduce biases by treating these organisms as equivalent to heterotrophic nanoflagellates (HNF). In the present case study we experimentally tested two generally employed assumptions of such studies: (A) bacterivory rates of the whole SP community and of distinct SP groups remain constant over 'short' time scales (hours to a day) and (B) SP community ingestion rates approximate the average ingestion rate of all feeding individuals. Food vacuole markers (acidotropic probes), were applied along the diel cycle at three stations in December 2015, and May and June 2016. In December and June, surrogate prey (fluorescently labeled bacteria) were used in parallel at one sampling station. Sampling at different times of day produced an up to fourfold difference in estimates of SP daily bacterivorous impact. In contrast, daily bacterivory estimates for HNF remained constant in almost all cases. The perceived principal SP bacterivorous groups also shifted strongly. As an example, picoeukaryotes dominated total SP bacterivory in daylight hours but completely ceased to feed at night. Finally, a large fraction of the SP community was not feeding at all time points tested. This lead to significant errors in estimated ingestion rates determined using the whole SP community, being up to 16 times lower than those determined solely for actively feeding mixotrophic SP. Overall, this case study indicates that applying the two commonly used premises outlined above can introduce significant biases and considerably alter our perception of mixotrophy in a given system.