Morphometrics of Southern Ocean diatoms using high throughput imaging and semi-automated image analysis


Contact
Bank.Beszteri [ at ] awi.de

Abstract

Since the ADIAC project, which ended more than 15 years ago, not much progress in automating morphometric analysis of diatoms from slide-mounted material has been published, and no ready-touse system has become available. This thesis work is the first to implement such a system completely, covering all aspects of the underlying imaging and image processing pipeline, by combining a commercially available slide scanning microscope with my diatom morphometry software SHERPA. I was able to show the applicability as well as the potential of this approach by executing a series of smaller and two large-scale morphometry projects. The extensive sampling sizes, which were made possible only by the new workflow, enabled the first observations of life cycle related size distribution changes of F. kerguelensis in its natural habitat, leading to hypotheses on influences of reproduction, grazing and environmental changes in one of the most important diatom species of the Southern Ocean. In a second large-scale investigation, SHERPA’s precise morphometric measurements revealed a second F. kerguelensis morphotype, which has not been recognized before, even though the species, as well as the very material I analyzed, have been investigated intensely before by experienced diatomists; a result not disqualifying their work, but rather underlining that explicit and precise quantification of morphological information has a strong potential to generate novel scientific insights. This new morphotype has implications on the utilization of paleo-proxies which are based on geometrical valve features of F. kerguelensis. Differentiating both morphotypes might improve established methods and possibly provides a new proxy for summer sea surface temperature. As image processing and classification nowadays develop faster than ever before, further development will allow application of my workflow in automatic diatom identification, water quality monitoring, paleoceanography, paleolimnology and many other fields of diatom research.



Item Type
Thesis (PhD)
Authors
Divisions
Primary Division
Programs
Primary Topic
Peer revision
Not peer-reviewed
Publication Status
Published
Eprint ID
48754
Cite as
Kloster, M. (2018): Morphometrics of Southern Ocean diatoms using high throughput imaging and semi-automated image analysis , PhD thesis, Universität Bremen.


Download
[img]
Preview
PDF
KlosterM_PhD-Thesis.pdf

Download (8MB) | Preview

Share

Research Platforms

Campaigns


Actions
Edit Item Edit Item