Studying phytoplankton parasites in a Baltic Sea spring bloom using imaging flow cytometry

Abstract

Phytoplankton, important primary producers in aquatic environments, are known to be infected by several kinds of parasites. Chytridiomycota, or chytrids, are among the most common fungal parasites found to infect phytoplankton. They have been shown to have significant impacts on phytoplankton communities and species composition. However, even though phytoplankton parasites have been recognized for decades, many questions are still unresolved about their role and impact, especially in marine environments and in the Baltic Sea. With the help of imaging flow cytometry, such as Imaging FlowCytobot (IFCB), it is possible to continuously sample and analyse phytoplankton blooms. The IFCB takes water samples every ~20 minutes that are run through the machine, while taking images of all phytoplankton in the samples. The images are automatically classified using machine learning techniques, sorting them into categories based on morphology. Chytrid infections are possible to detect in these images, thus, IFCB can also be utilised to detect fungal parasites on phytoplankton. In this study, phytoplankton parasites are studied using IFCB. The purpose of this thesis is to study the presence and impact of chytrid parasites on phytoplankton in the Baltic Sea, to increase the knowledge of phytoplankton parasites in the area. The material used in this study, provided by the Finnish Environmental Institute, was sampled in the Archipelago Sea with IFCB during spring bloom in 2021. The images were automatically classified in taxonomic groups, provided in daily or hourly folders. Four of these groups were chosen for this study to be analysed further: Centrales, Chaetoceros spp. (provided in folders with chains and single cells separately), Pauliella taeniata and Skeletonema marinoi. The bloom period for each group was analysed for putative infections, with a set of predetermined selection criteria. Putative infections were found on all groups, but the infection rates were lower than expected, especially on the dominating diatom S. marinoi. The most infected group was P. taeniata. Considering both infection rates and quality criteria, only Centrales and P. taeniata indicate actual chytrid infections, whereas infections on S. marinoi and Chaetoceros spp. are more uncertain. Further, there was no indication that the infection rates influence the growth rate of phytoplankton host blooms negatively. While temperature could explain infection abundances for most groups (except single-celled Chaetoceros spp.), nutrient changes did not show similar patterns. It seems, therefore, that infection abundances are more likely connected to changes in host bloom abundances thanenvironmental changes. Future studies should include evaluation of the putative infections observed from images taken with IFCB, for example, by comparing chytrid infections detected on IFCB images with infections detected with microscopy. However, the potential of this methodology is huge, as automated phytoplankton and anomaly identification would enable more extensive research on phytoplankton parasites.