AMIME: Artificial intelligence-assisted plankton Monitoring with Imaging flow cytometry and MEtabarcoding

Single-celled eukaryotic plankton (protists) form the productive base of marine ecosystems and are key drivers of global biogeochemical cycles of carbon and nutrients.

Consequently, plankton have fundamental impacts on both fish stocks and the climate and understanding the factors that control the abundance and distribution of different plankton species is of great concern. Monitoring of eukaryotic plankton has traditionally been conducted by manual microscopic detection.

Recently, alternative approaches have emerged such as high-throughput imaging and DNA metabarcoding. While promising, these methods have their challenges, not least in how to translate between these disparate datasets.

In this project we will utilize state-of-the-art image analysis and deep learning approaches to maximise the information gained from these types of data and to translate between them.

We will leverage on existing imaging data from the new Imaging FlowCytobot (IFCB) instrument mounted on the research vessel Svea as well as generate new parallel IFCB and DNA metabarcoding datasets for 500 water samples spanning the Baltic Sea, Kattegat and Skagerrak.

The methodology developed in the project will advance plankton research and ecology in general and plankton monitoring in Sweden in particular.

It will bridge the gap between imaging and DNA-based diversity data and increase the information output from both approaches.