Plankton, constituted by phyto-, zoo- and bacterio-/virioplankton, occupies the lowest levels of the trophic web in many marine ecosystems. Phytoplankton, represented by photoautotroph organisms, is the main primary producer in the oceans. Zooplankton is the heterotrophic part of the plankton. As such, it has a key role in transfer from those primary producers to higher trophic levels. It also takes part of a side-way recycling of organic matter in the water column, initiated by the bacterioplankton (not forgetting archaea), which is called the microbial loop. Many plankton organisms play various roles simultaneously in the ecosystem: they are mixotrophs. Ecological plankton studies are thus complex, due to the diversified and intricate ecological role of the many marine planktonic organisms.
Zooplankton as well as phytoplankton are considered as excellent bioindicators of global environmental changes. Indeed, their response to environmental changes is very fast, both in term of abundance and/or species composition and in biomass. Moreover, phytoplankton frequently blooms (rapid exponential multiplication of one or several species). Some of these blooms are harmful: they can create anoxic conditions during the night, or they can poison other organisms in the ecosystem by the production of toxins. For all these reasons, the highly dynamic plankton communities are studied to get major information on the state and changes of oceanic environments.
Nowadays, planktonic organisms are still frequently enumerated and measured manually under a binocular or a microscope by specialized taxonomists. Complete analysis of one sample can last for 1 to 3 days, according to required taxonomic level. This limits the number of samples that can be processed in a reasonable amount of time. Such a manual approach is thus not very suitable to obtain a high spatial and temporal resolution required to study the patchy distribution of the plankton. A higher number of samples often need to be processed.
Since the 1980's, image analysis coupled with machine learning techniques is considered as a possible alternative to (at least partly) automatize and accelerate the process of plankton samples. Starting from digital images of the plankton, various parameters are measured and they are used as discriminant variables by machine learning algorithms to automatically determine the taxonomic group of each digitized particle. Recently, the constant raise of the calculation power of computers, the increase of the resolution of digital images and the development of more sophisticate machine learning algorithms allow to use this approach routinely in oceanographic campaigns. We participate to this tendency by developing and using specialized tools that bring speed and comfort to plankton analyses. Taxonomists should benefit from these tools as a help to automatically sort the bulk of particles and to get focus on the most important or most difficult ones to identify.
In our laboratory, we develop a software called Zoo/PhytoImage which targets the creation of plankton space-time ecological series by automating a part of the process of the samples. This open source software allows to analyze various kinds of digital plankton images (micro- or macrophotographies, scanned images, or pictures acquired using a FlowCAM), It also enumerates, measures and classify the various plankton organisms "numerically" preserved on these images. It then calculates derived variables that are of ecological importance, like abundances, biomasses, or size spectra per taxonomic group, or for the whole sample.
Zoo/PhytoImage was, partly, developed in collaboration with ESA (Ecologie des Systèmes Aquatiques – ULB – staff of the professor Ch. Lancelot) in the framework of the project AMORE III funded by the Belgian Federal Science Policy Office in order to detect blooms of Phaeocystis globosa as well as other changes in the phytoplankton communities in the Belgian Coastal Zone of the North Sea.
The development of this software continues now for its use by IFREMER in the framework of the REPHY (Phytoplankton Network) that surveys the changes of phytoplankton communities all along the French coasts (see also IFREMER environment). It is also used in many other studies of zoo- or phytoplankton all around the world.
Our scientific publications relating to this research topic are collected in the ULB-UMONS institutional database DI-Fusion .
PhD and Master theses and premaster placements in our laboratory in relationship with plankton studies:
Workshops and specific software developments around Zoo/PhytoImage: We organize, on request, workshops about Zoo/PhytoImage. We can also provide specific custom-made applications (bacteria counting, enumeration of other organisms than plankton, characterization of organisms in culture,, ...). Contact-us for more information.
The FlowCAM on the Belgica: Use of the FlowCAM aboard of the Belgica: we develop a software to analyze data from a FlowCAM in real-time aboard of our oceanographic ship, the "Belgica".
A couple of plankton images: from our campaigns with the FlowCAM (click on the picture).