Mein Forschungsschwerpunkt ist das Wechselspiel zwischen Planktongemeinschaften und biogeochemischen Stoffkreisläufen, z.B. Interaktionen von Zooplankton mit dem Fluss organischen Materials von der Meeresoberfläche in die Tiefsee ("biologische Pumpe"). Um die Möglichkeiten ökologischer Beobachtungen zu verbessern arbeite ich zudem an der Entwicklung verschiedener Technologien wie beispielsweise Unterwasserkamerasysteme um das Plankton in seiner natürlichen Umgebung zu untersuchen. Solche neuartigen Methoden können uns helfen, kleinskalige räumliche und zeitliche Verbreitungsmuster pelagischer Lebewesen zu erfassen und die Reaktion marine Ökosysteme auf den Klimawandel zu beleuchten.
MicroZooImager: An integrated optical system for microzooplankton analysis
Microzooplankton, a group of heterotrophic organisms in the size range of 20-200 μm, are major consumers of marine primary production in the world oceans. Data on their taxonomy and biomass is traditionally obtained by microscopy, which is very labor- and time-intensive. This impedes a high spatial and temporal resolution of data, which would be needed to improve our mechanistic understanding of the role of microzooplankton in marine ecosystems and carbon cycling. To solve this problem, we propose to develop a novel imaging system for rapid data acquisition on microzooplankton abundance, biomass, body size and taxonomy. The instrumental design will be based on a high-resolution line-scan camera with telecentric optics and illumination and provide high-quality images of the multitude of microzooplankton specimen contained in natural plankton samples. Acquired images will be analyzed with existing software for image processing and automated classification, thus allowing for a swift workflow and acquisition of final data. Our interdisciplinary project unifies marine ecology, optical engineering, and computer sciences with the ultimate goal of establishing a rapid and efficient method for studying microzooplankton communities.
Combining deep learning, in situ imaging and citizen science to resolve the distribution of giant protists in major upwelling regions
Recently, a first global in situ imaging survey comprising ~ 1.8 million semiautomatically annotated images obtained with an Underwater Vision Profiler 5 (UVP5) was finalized. This survey revealed that giant protists belonging to the super-group rhizaria contribute significantly to biomass in the mesozooplankton size class in tropical and subtropical regions of the oceans. These giant protists were also found to be especially abundant in the Californian upwelling system. We will analyze further UVP5 image data from the Peruvian, Mauretanian and Benguela upwelling systems to resolve the rhizarian environmental niche and their role in biogeochemical cycles in these regions. To enhance our abilities for high-throughput image annotation needed for an efficient sensor-to-information pathway, we will combine state-of-the-art deep learning image recognition algorithms and a citizen scientist approach to annotate the > 850000 images for this project. Involvement of citizen scientists will lead to a direct dissemination of results to the general public and will contribute to the outreach activities of the Future Ocean during the German Year of the Oceans.
