October 11th 2017: Molecular tools and databases

08:30 Workshop Introduction

Session 1: 09:00

Keynote: Chris Bowler

Tara Oceans: Eco-Systems Biology at Planetary Scale

The ocean is the largest ecosystem on Earth and yet we know very little about it. This is particularly true for the plankton that drift within. Although these organisms are at least as important for the Earth system as the forests on land, most of them are invisible to the naked eye and thus are largely uncharacterized, even though they form the base of marine food webs and are key players in Earth’s biogeochemical cycles. To increase our understanding of this underexplored world, a multidisciplinary consortium, Tara Oceans, was formed around the 110-ft research schooner Tara, which sampled plankton at more than 210 sites and multiple depth layers in all the major oceanic regions during expeditions from 2009-2013 (Karsenti et al. Plos Biol., 2011). The seminar will summarize the first foundational resources from the project (see Science special issue May 22, 2015 and Nature 28 April, 2016) and their initial analyses, illustrating several aspects of the Tara Oceans’ eco-systems biology approach to address microbial contributions to macro-ecological processes. The project provides unique resources for several scientific disciplines, capturing biodiversity of a wide range of organisms that are rarely studied together, exploring interactions between them and integrating them with environmental conditions to further our understanding of life in the ocean and beyond in the context of evolution, adaptation and ongoing climate changes.

Rowena Stern

Can molecular methods be integrated into long-term plankton observations?

Rowena Stern*, Alex Kraberg, Katja Metfies

'Sir Alister Hardy Foundation for Ocean Science, rost@sahfos.ac.uk

Over the last fifteen years, molecular genetics has rapidly advanced the rate of plankton species discovery in the oceans and the advent of high-throughput DNA sequencing has revealed a vast level of diversity previously unknown. As the use of genetics is more accepted, should we consider its routine use for reporting plankton species to national government and European bodies?  Most long-term ecological research sites (LTERS) that report plankton still use microscopy-based species identification that favours the identification of larger organisms over smaller ones. This is done for continuity even though there is awareness that microscopy methods are insufficient to monitor all plankton diversity.  Several LTERs are now additionally using molecular method to augment the species they find as there is a need for this information from a variety of stakeholders, most notably for four of the descriptors of the new MSFD legislation. However there are several obstacles in integrating molecular methods into established plankton time-series. We conducted a literature survey and a questionnaire from marine molecular ecologists, many who also carry out microscopic surveys, to find out the benefits and challenges of integrating molecular identification methods into routine long-term observations.

Keywords: Plankton, Long-term ecological research sites, molecular, microscopy, survey

Sir Alister Hardy Foundation for Ocean Science, rost@sahfos.ac.uk

Tom Richards

Getting a grip on uncultured heterotrophic eukaryotes and their genomes

Jeremy G Wideman1, Raquel Rodriguez1, David Milner1, Adam Monier1, Guy Leonard1, Raquel Rodriguez1, David Milner1, Camille Poirier2, Patrick Keeling3, Alyson Santoro4, Alexandra Worden2, Thomas A Richards1


1 Biosciences, University of Exeter, Exeter, UK

2 Monterey Bay Aquarium Research Institute, Moss Landing, USA

3 University of British Columbia, Vancouver, Canada

4 University of California Santa Barbara, Santa Barbara, USA

 Environmental DNA sequencing has offered a glimpse at natural microbial diversity in both the sea and on land, revealing that we actually know very little about the microbes that occupying natural environments. This is a tantalising perspective, but poses many problems for understanding the tree of life and how microbial communities function. Here I discuss our progress using single cell genomics approaches to target heterotrophic eukaryotes, a group traditionally thought of as hard to get a grip on because they depend on trophic interactions which are hard to propagate in culture. Single-cell genomics is an under-developed molecular tool with a history of limited application to marine microbes. We set out to develop a single cell genomics approach to target small heterotrophic flagellates by sorting for cells of 3-10 um lacking a chlorophyll fluorescent signature but possessing fluorescently stained proteins that predominately function in flagella. This approach allowed us to recover numerous genomic samples of uncultured eukaryotes from across the tree of life. Single-cell genomics is of course noisy, allowing only limited and fragmented recovery of the nuclear genome; however, our approach consistently recovered useful mitochondrial genome assemblies, allowing us to re-investigate the evolution of mitochondrial genome complexity of ancestral eukaryotes. Interestingly, this work also uncovered several unique lineage-specific mitochondrial features, including a novel mitochondrial genetic code and the first eukaryotic-encoded restriction-mediated selfish genetic element. This eukaryote-encoded restriction-mediated selfish element represents a recent horizontal gene transfer event. This study has provided the foundation for future single cell genomics investigations into the diversity and evolution of organelle genomes, an approach that will allow us to get a grip on phylogenomic datasets that can be used to resolve population and biogeographical hypotheses relating to heterotrophic aquatic protists.

Adriana Zingone

Integrating metabarcoding results into the current knowledge on protists

Zingone A

Piredda R., K. Pargana, M.P.Tomasino, M. Montresor, W.H.C.F. Kooistra, D. Sarno  & A. Zingone

Stazione Zoologica Anton Dorhn , Villa Comunale, 80121 Napoli, Italy


DNA-Metabarcoding studies deliver huge amounts of data, enabling a revolution of our knowledge on microbial diversity. However, methodological choices concerning sample treatment, marker tags, primers, templates and analytical procedures affect the outcomes of these studies. Comparison of results from different exercises with the knowledge accumulated with traditional methods can help identify potential pitfalls and avoid inconsistent interpretations. We will use results from a time series in the Gulf of Naples (LTER-MC), as well as examples from other datasets such as Ocean Sampling Day, BioMarKs and Tara Oceans, to highlight some concordant or contrasting patterns that emerge from our metabarcode analyses. At LTER MC, results of protist metabarcode analysis conducted on eight samples taken along one year indicated clear seasonal patterning (spring-early summer, late summer and autumn-winter), with a peak of diversity in winter, and this patterning was confirmed by results obtained from a dataset of 48 samples gathered along the seasonal cycles over three years. The use of V4 or V9 produced similar seasonal and diversity patterns, although different annotation performances were often obtained with either markers, also due to the lower richness of the V9 reference database.  At the species level, both marker tags allowed identification of multiple taxa within individual species of the diatom genus Leptocylindrus, with different temporal occurrence and temperature preferences, which highlights the possibility of metabarcoding data to unveil cryptic or even intraspecific diversity. The same approach was tested successfully to address diversity of diatom resting stages in sediment samples, and to describe harmful species distribution across different coastal areas. Despite long term studies on HABs in the Gulf of Naples, a number of harmful species were detected for the first time, pointing at the relevance of the metabarcoding approach in monitoring aimed at the protection of human health. 

Session 2: 11:00

Katja Metfies

Molecular time series in the Arctic with special reference to the Hausgarten LTER station

Metfies, K. , Bienhold, C. , Boetius, A. , Buttigieg, P. , Fadeev, E. , Frickenhaus, S. , Hardge, K. , Jacob, M. , Neuhaus, S. , Noethig, E. M. , Peeken, I. , Rapp, J. , Salter, I. , Wenzhöfer, F. and Wolf, C.


Information on current diversity and biogeography of Arctic marine microbes (bacteria, archaea and single cell eukaryotes) with adequate temporal, spatial and taxonomic resolution is urgently needed to better understand natural dynamics of ecosystem states in space and time, and consequences of environmental change by anthropogenic factors. Here, we introduce a standardized molecular-based observation strategy for high resolution assessment of marine microbes in space and time, even in remote areas such as the Arctic Ocean. The observation strategy involves molecular analyses such as Next Generation Sequencing (NGS) and quantitative polymerase chain reaction (qPCR) of diverse environmental samples, collected from sea ice, water column and seafloor with a complementary set of automated and ship-based sampling approaches. This includes newly developed automated under-way sampling, moored sediment traps and year-round water samplers, as well as CTD-casts, multi-corers, bottom landers and in the future seafloor crawlers. An integrated standardized dataset including linked, searchable information on synchronous environmental variables provides comprehensive information on the diversity, abundance and biogeography of Arctic marine microbes, covering all three domains of life. The development of the observation strategy involves a set of coordinated pilot studies testing questions of temporal and spatial resolution, i.e. to assess the impact of sea-ice on Arctic marine single-cell eukaroyte community composition, or of ocean warming in Eastern Fram Strait since the year 2000. In the future, the observation strategy for Arctic marine microbes will be implemented as a distributed Molecular Microbial Observatory in the framework of the Arctic observatory FRAM (Frontiers in Arctic Monitoring) and contributes to the ATLANTOS strategy for an integrated Atlantic observatory including genomic information.

Anne Vader

High-throughput methods in a high-Arctic marine time series; does “Atlantification” matter?

Anna Vader, Ragnheid Skogseth, and Tove M. Gabrielsen

University Centre in Svalbard, Longyearbyen, Norway

Arctic plankton communities show strong seasonal changes in abundance, community composition and timing of life cycle events. These changes are strongly driven by light, nutrients, prey-predator interactions, temperature and by the history of the water masses they inhabit. West Spitsbergen fjords have recently experienced increased heat content from inflowing warm and saline Atlantic Water originating from the West Spitsbergen Current. This has hindered sea ice formation and led to a range expansion of boreal species (e.g. mackerel and herring) into Svalbard Waters. To study seasonal and interannual variability as well as the potential effect of “Atlantification” on Arctic plankton communities we established a high-Arctic time series station in Isfjorden, West Spitsbergen (the IsA time series station) which has been sampled regularly since 2011. The data obtained include vertical salinity, temperature, light and fluorescence profiles, “classical” net samples for determination of different size groups of plankton as well as high throughput flow cytometry and microbial metabarcoding and metatranscriptomic data. An in-depth analysis of the first three years of high-resolution (weekly to monthly) biodiversity data showed high interannual variability in hydrography and also in the timing and magnitude of the spring bloom. The community composition of microbial eukaryotes (0.45-10 μm size) also displayed strong seasonality with large shifts during spring and summer, when also interannual differences were most apparent. In contrast, highly similar and stable communities were observed during the three sampled winters, suggesting a high degree of resilience in the system. Thus, although “Atlantification” seems to matter, the long-term influence of southern immigrants may be limited by the extreme winter light conditions. Our data show that high-resolution time-series are essential to disentangle the effects of natural variability from climate change impacts on these high-Arctic systems.

Connie Lovejoy

Perspectives for the Arctic.

Connie Lovejoy

Département de biologie, Québec Océan, Institut de biologie intégrative et des systèmes (IBIS) and Takuvik Joint International Laboratory (UMI 3376), Université Laval (Canada) - CNRS (France).

Université Laval, Québec QC G1V 0A6, Canada 

Marine microbial communities, phytoplankton and heterotrophic protists, referred to as microbial eukaryotes, Bacteria and Archaea are the base of oceanographic food chains and mediate many of the steps in global biogeochemical cycles. However, despite the ecological importance, apparent abundance and wide distribution, most aspects of ecology, diversity and oceanography of microbes are poorly understood. 

The microbial communities of the Arctic Ocean are taxonomically distinct from other oceans, suggesting vulnerability due recent climate related changes. In the Arctic, the environments that select different microbial communities are influenced by physical oceanography at multiple scales and oceanographic conditions follow regional differences in summer ice extent and freshwater input into the Arctic. Changes in the Arctic will affect phytoplankton and other microbial communities in a number of ways, such as altered nutrient supply, lower mixed layer salinities, and increased variability in surface temperatures. For example, in the Canada Basin smaller phytoplankton species are becoming more prevalent, which has implications on the feeding ecology of calenoid zooplankton by limiting the range and size of prey items available. Smaller average phytoplankton size also has an effect on the net carbon flux in the Arctic Ocean and the carbon cycle generally and a taxonomic comparison of microbial communities before and after the first drastic sea ice minimum in 2007 indicated significant changes in all three domains of life. Such changes signal the development of a more complex microbial foodweb where unicellular microzooplankton and bacteria become relatively more central in the transfer of energy and carbon to higher food webs compared to classical diatom, copepod based food chains.  Progress has been made characterizing the taxonomic composition of arctic microbial communities using targeted high throughput sequencing (HTS) approaches, and data from multiple Arctic expeditions have provided a platform to test spatial and temporal variability of these microorganisms. However, knowledge of the prevalence of putative ecological functions and microbial metabolic activities are needed.  Arctic microbes, including phytoplankton produce and degrade marine DOM and some of these microbes will be specialists, metabolizing different components of DOM. For this reason, I would advocate implementing high throughput metagenomics and metaproteomics to study microbial metabolic diversity and activity along with developing methods for high resolution characterization of organic molecules produced and processed by Arctic microbial communities from different water masses.  As change continues, knowledge of the taxonomic diversity and functional capacity of microbial life will become critical for predicting consequences of a fresher, likely more stratified Arctic Ocean.

Linda Amaral Zettler

A Multiphasic Approach to Studying Biodiversity in the "Plastisphere" Microbiome


Plastic Marine Debris (PMD) is a major source of marine pollution and potential source of invasive alien species (including pathogens and harmful algal blooming species), two important ocean health index criteria. While macroplastic is the most conspicuous and iconic debris in the environment, micro (< 5 mm) and nano-sized (<50 µm) plastic particles are now recognized as a growing concern. Broadening interest in the topic has extended to plastic debris in rivers and other freshwater environments. Plastics are almost instantaneously colonized upon contact with water of any kind by a thin film of microorganisms, what we refer to as the "Plastisphere" microbiome. My lab has been studying microbial interactions with PMD using a multiphasic approach including next-generation amplicon and metagenomics sequencing, culturing, Scanning Electron Microscopy, and most recently Combinatorial Labelling and Spectral Imaging – Fluorescence In Situ Hybridization (CLASI-FISH). These techniques all lend themselves to time-series sampling of developing biofilms on "virgin" substrates that can be sampled and preserved at desired time-points thereafter. Our early field investigations revealed that Plastisphere communities are quite distinct from the surrounding environment, but time series experiments provide a time-stamp on the succession and community assembly and seasonality in Plastisphere communities that is difficult to achieve in naturally collected samples. My talk will review what is known about diversity in the "Plastisphere" to date and discuss the advantages and disadvantages of different technologies in addressing some of the most urgent questions regarding this newest of marine habitats.  

Session 3: 13:30

Wiebe Kooistra

Kooistra, Wiebe H. C. F., Gaonkar, C. C., Zingone, A., Piredda, R., Sarno, D., Montresor, M., Stazione Zoologica Anton Dohrn, Villa Comunale, Naples, Italy


Diversity assessment in the planktonic diatom family Chaetocerotaceae by means of strain cultivation and morphological and genetic characterization uncovered ca. 80 species, 44 of which were observed in our study area, the Gulf of Naples (GoN). Subsequently, we assessed the diversity in the family by means of HTS meta-barcoding of 48 plankton samples collected at the LTER MareChiara (GoN) between 2011 and 2013. As meta-barcode we used the V4 hypervariable region in the 18S rDNA. The 13.6M obtained eukaryotic contigs grouped into 615,142 haplotypes, of which 18.625 were assigned to Chaetocerotaceae (EPA and BLAST using chaetocerotacean reference barcodes). Following elimination of haplotypes with <3 contigs, the retained haplotypes were aligned with the ca 80 Chaetocerotacean reference barcodes to build an ML tree. This tree resolved 66 terminal clades assigned putatively to Chaetoceros species and 10 to Bacteriastrum. Of the 76 clades: 39 were assigned to species present in the GoN (either already described or newly discovered) and 7 to a species known from elsewhere; not from the GoN. The remaining 30 lacked any reference barcode.. Whilst gathering rDNA sequences of the chaetocerotacean strains, group I introns and spliceosomal introns were encountered all over the place in the 18S rDNA alignment, but fortunately, the V4-region, showed introns in only two species and the V4 universal primer positions were fatally corrupted in only one species. Numerous haplotypes resolved on long branches inside terminal clades or as solitary lineages deeper down in the tree. Most of these were recombinants of pairs of common chaetocerotacean species. Such intra-generic recombinants are difficult to spot automatically, but it is relevant to identify them as such because otherwise they inflate “rare biodiversity.” Thus, HTS-meta-barcoding works well in detecting – and discriminating among the members of - this common and abundant family, all the way down to the species-level. Moreover, HTS metabarcoding results show that even in the taxonomically well-explored GoN a considerable percentage of the chaetocerotacean species is still to be descr

Fabrice Not

Bernhard Fuchs

A High-Throughput Automated Cell Enumeration System for the Quantification of Microbial Populations


Fuchs B. 1*, Bennke C.  1, Ellrott A.  1, Reintjes G.  1, Schattenhofer M.  1, Wulf J., Zeder M.  1

1 Max Planck Institute for Marine Microbiology, Bremen


* Corresponding author: bfuchs@mpi-bre­men.de



In the age of ever-increasing “-omics” studies, the accurate and statistically robust determination of microbial cell numbers within often-complex samples remains a key task in microbial ecology. Microscopic quantification is still the only method to enumerate specific subgroups of microbial clades within complex communities by, for example, fluorescence in situ hybridization (FISH). In the past this quantification was often done manually at the microscope and was rather time-consuming and relied on the experience of the individual person counting. Moreover only a small part of the sample was analyzed with usually 12 to 20 fields of view (FOV) per preparation. Over the years an automated microscopic image acquisition and cell detection and enumeration system was developed at the Max Planck Institute for Marine Microbiology, which made the quantification of many more FOVs in hundreds to thousands of samples feasible. It was first applied to several hundreds of FISH assays on samples taken along a transect across the Atlantic Ocean and subsequently used to quantify the abundance of microbial populations after spring diatom blooms at Helgoland in the North Sea comprising in several thousands of preparations. Lately we improved and adapted this automatic image acquisition and cell enumeration system for usage at high seas on board an oceanographic research ship. During a cruise through the Atlantic Ocean in 2012 it was robust enough to produce high-quality images even with ship heaves of up to 3 m and pitch and roll angles of up to 6.3°. On board the research ship, on average, 25% of the images acquired from plankton samples on filter membranes could be used for cell enumeration. The newly adapted microscope system allows insights into the abundance and distribution of specific microorganisms already on board no later than 36 hours after sampling.

Ann Bucklin

Young scientists' session:

Heidi Yeh

Comparative metagenetic analysis of the diet of the copepod Calanus finmarchicus in the Gulf of Maine (NW Atlantic) in years of low and high abundance.


Heidi Yeh, Jennifer Questel, and Ann Bucklin

Department of Marine Sciences, University of Connecticut, Groton, CT 06340 USA



The copepod Calanus finmarchicus occupies a pivotal position in the pelagic food web of the Gulf of Maine (Northwestern Atlantic), providing a key link between primary producers and predators. The abundance of this species has been recorded by Continuous Plankton Recorder (CPR) surveys since 1961 and during NOAA-NEFSC Ecosystem Monitoring (EcoMon) surveys since 1977. This study used metagenetic analysis (i.e., high throughput sequencing of orthologous gene regions for assessment of biodiversity) to analyze the gut contents of pooled C. finmarchicus specimens from samples collected during spring and summer EcoMon surveys in 2012 and 2013. These two years represent successive, exceptional conditions, with marked shifts in large-scale forcing (NAO index), water temperature, and abundance of the target copepod. The V4 hypervariable region of the small subunit (18S) ribosomal RNA (rRNA) gene was analyzed for DNA extracted from C. finmarchicus. Blocking PCR primers were used to reduce amplification of predator DNA and enhance detection of prey DNA. Operational Taxonomic Units (OTUs) were analyzed using the SILVA and an 18S rRNA reference database for potential prey items (phytoplankton, heterotrophic protists, and microbes) reported to occur in the Gulf of Maine. Our comparative assessment of variation in the diet of the copepod between an anomalously warm year, with high C. finmarchicus abundance (2012) and the year immediately following (2013) will allow examination of potential variation in food web dynamics driven by inter-annual environmental variation

Florian Mundt

A High-Throughput Automated Cell Enumeration System for the Quantification of Microbial Populations

Fuchs, BM., C. Bennke, A. Ellrott, G. Reintjes, M. Schattenhofer, J. Wulf, M. Zeder

Max Planck Institute for Marine Microbiology, Bremen

In the age of ever-increasing “-omics” studies, the accurate and statistically robust determination of microbial cell numbers within often-complex samples remains a key task in microbial ecology. Microscopic quantification is still the only method to enumerate specific subgroups of microbial clades within complex communities by, for example, fluorescence in situ hybridization (FISH). In the past this quantification was often done manually at the microscope and was rather time-consuming and relied on the experience of the individual person counting. Moreover only a small part of the sample was analyzed with usually 12 to 20 fields of view (FOV) per preparation. Over the years an automated microscopic image acquisition and cell detection and enumeration system was developed at the Max Planck Institute for Marine Microbiology, which made the quantification of many more FOVs in hundreds to thousands of samples feasible. It was first applied to several hundreds of FISH assays on samples taken along a transect across the Atlantic Ocean and subsequently used to quantify the abundance of microbial populations after spring diatom blooms at Helgoland in the North Sea comprising in several thousands of preparations. Lately we improved and adapted this automatic image acquisition and cell enumeration system for usage at high seas on board an oceanographic research ship. During a cruise through the Atlantic Ocean in 2012 it was robust enough to produce high-quality images even with ship heaves of up to 3 m and pitch and roll angles of up to 6.3°. On board the research ship, on average, 25% of the images acquired from plankton samples on filter membranes could be used for cell enumeration. The newly adapted microscope system allows insights into the abundance and distribution of specific microorganisms already on board no later than 36 hours after sampling.

First steps towards DNA barcoding and gene expression in the Zygnematophyceae


Mundt, F., Department Aquatic Ecophysiology and Phycology, University of Hamburg, Germany,


Harms, L., Scientific Computing, Alfred Wegener Institute, Bremerhaven, Germany

Heinrich, S., Department Molecular Plant Genetics, University of Hamburg, Germany,

Hanelt, D., Department Aquatic Ecophysiology and Phycology, University of Hamburg, Germany,



The Zygnematophyceae form a diverse group of freshwater algae with more than 3500 species. Recent

studies on evolution of land plants indicate that Zygnematophyceae are sister group to the

Embryophyta, which has important implications for early land plant evolution. Despite the importance in

terms of ecology and evolutionary processes, molecular studies on this group are still scarce.

Due to elasticity in phenotypic traits and a lack of molecular data, a determination on the species level is

only possible for taxonomists. Hence, the first part of this project focuses on the development of a

barcoding system. DNA of more than 550 culture strains of the Microalgae and Zygnematophyceae

Collection Hamburg (MZCH) was extracted. The rbcLa gene region is used as a pre-barcode, resulting in

clades distinguishing on the genera level. First analyses indicate a lack of gene references in the group of

Zygnematophyceae to further investigate the genetic diversity.

The second part of the project focuses on molecular acclimation of a polar isolate of Cosmarium

crenatum, a cosmopolitan species. A de novo reference transcriptome by Illumina HiSeq sequencing in

300 bp paired-end modus was established. The resulting reads were assembled to 141,711 transcripts

and annotated to 58,826 gene functions via blastx. The reference transcriptome covers all major

pathways i.e. lipid, carbohydrate, amino acid and nucleotide metabolism. Subsequently, the first

expression changes in response to a one week darkness exposure were investigated, resulting in more

than 7,900 differentially expressed genes. The dataset provides new insights into molecular acclimation

of Cosmarium crenatum. It is the first comprehensive reference transcriptome in the developmental

lineage of the Desmidiaceae.

Laura Käse

Unveiling new phytoplankton community connections after comparison of detailed counts and next generation sequencing

Laura Käse1, Katja Metfies2, Karen H. Wiltshire1, Maarten Boersma1, Bernhard Fuchs3, Alexandra Kraberg1

1 Biologische Anstalt Helgoland, Alfred-Wegener-Institute Helmholtz Centre for Polar and Marine Research, 27498 Helgoland, Germany

2 Alfred-Wegener-Institute Helmholtz Centre for Polar and Marine Research, 27570 Bremerhaven, Germany

3 Department of Molecular Ecology, Max Planck Institute for Marine Microbiology, 28359 Bremen, Germany

Plankton time-series are an important component in the study of long-term changes in marine biodiversity. However, most conventional time-series like the Helgoland Roads time-series still use traditional microscopy techniques. Using the traditional Utermöhl method causes underreporting of small sized protists and therefore leaves a considerable research gap. The resolution of the time-series should be improved by incorporating different monitoring methods to include so far underreported pico- and nanoplankton. However, one obstacle in using these new methods is to generate comparable datasets to the existing long-term data. Therefore it needs to be tested to what extent counting data using the traditional Utermöhl method can be complemented with the new information on the diversity of the samples. Next generation sequencing is a presented as a good option to obtain additional data particularly on these smaller taxa e.g. different small flagellates. Phytoplankton samples were taken two times per week and cells were counted optically via microscopy. The same samples were used for next-generation sequencing. Microscopic Utermöhl counting resulted in high abundances of different small-sized flagellates and cryptophytes, which were categorized in size groups and are not counted in such detail in the regular time series. Smaller organisms included in the picoplankton group could not be distinguished further. Bigger organisms like the diatom Pseudonitzschia sp. are also hardly distinguishable at species level using this method. Therefore, comparison of these counts and the regular Helgoland Roads counts with sequencing data is necessary. Additionally the sequencing data can give lots of information about the whole community structure and it might be possible to distinguish different cryptophytes on genus or species level that were not further distinguished via microscope. Taking into account the ongoing changes in the plankton community due to climate change, new insights in the food web and occurring parasite-hosts interactions at Helgoland are gained.

Sarah Cosgrove

The quantitative analysis of Alexandrium catenella in Monterey Bay, California using imaging flow cytometry


Harmful algal blooms pose a threat to the health of both humans and surrounding marine life and require consistent and reliable methods of monitoring. Current methods of monitoring including light microscopy and fluorescent in situ hybridisation (FISH) can be time-consuming. The FlowCAM, an imaging flow cytometer, provides a rapid, automated and economical method of analyzing phytoplankton communities within natural samples. We investigated the ability of the FlowCAM to identify and enumerate Alexandrium catenella cells within natural samples collected weekly from the Santa Cruz Municipal Wharf. Cell concentrations recorded using both the FlowCAM and more traditional methods of enumeration were compared over an eight week period. No significant difference was found between the quantitative analysis of A. catenella using FISH and the FlowCAM (p=0.16). To increase the accuracy of species specific detection, the FlowCAM was coupled with the oligonucleotide probe (NA1) to target A. catenella cells within mixed natural samples. The reliability of this technique is discussed. Imaging flow cytometry is an important technology with the ability to provide precise and simultaneous measurements of phytoplankton populations to facilitate their continuous monitoring and create reliable time-series data sets.

Alison Chase

Using Automated Imagery to Investigate Phytoplankton Size Structure and Community Composition in the North Atlantic


Alison Chase, Lee Karp-Boss, Emmanuel Boss, Nils Haëntjens


School of Marine Sciences, University of Maine, Orono, ME 04469 USA


Automated imagery techniques are becoming more prevalent in studies of phytoplankton community composition and diversity. Using an Imaging FlowCytobot (IFCB, McLane Research Laboratories, Inc.) deployed at sea on a continuous flow-through system, we have collected millions of phytoplankton images during several recent cruises in the North Atlantic. Currently, we are working to manage the data and sort the images taxonomically using the EcoTaxa software (http://ecotaxa.obs-vlfr.fr/). Initial results from these datasets will be presented and examine both the size distribution of living particles and phytoplankton community composition. In addition, we show a comparison between imagery-based results and High Performance Liquid Chromatography (HPLC)-based phytoplankton size structure, in an effort to define the uncertainties in the pigment-based analysis of phytoplankton community. Building the expertise for collection, management, and application of automated phytoplankton imagery data will allow application of this type of instrumentation to answer questions on time and space scales that have been previously unanswerable.

Deo Florence Onda

Multiyear study revealed changes in microzooplankton assemblages towards ice-related events in the changing Arctic Ocean

Deo Florence L. Onda1,2 and Connie Lovejoy1

1 Département de Biologie and Québec-Océan, Université Laval, G1V 0A6, Canada Takuvik Joint International Laboratory, UMI 3376, Université Laval (Canada) - CNRS (France), Institut de Biologie Intégrative et des Systèmes (IBIS), Université Laval, Québec, QC, Canada

2 Green Talents – German Federal Ministry on Research and Education (BMBF) / Hemholtz Center for Polar and Marine Research, Alfred Wegener Institute, Bremerhaven, Germany 

Multiyear studies on phytoplankton showed decreasing trends in the abundance of large-celled taxa due to strengthened stratification, nutrient limitation, and surface freshening in the Arctic Ocean, associated with the changing climate. However, little is still known about the responses of the other components of the microbial food webs, such as the ciliates and dinoflagellates (microzooplankton), which have multiple trophic roles. In this study, we identified potential annual species occurrence patterns from Spring and Summer 2003 to 2010, and linked them with environmental drivers. High-throughput DNA amplicon sequencing coupled with a curated Arctic 18S rRNA gene database was used to identify sequences to the lowest rank possible.  We found that microzooplankton community composition change was associated with the record ice minimum in the summer of 2007. Specifically, reads from smaller predatory species like Laboea, Monodinium and Strombidium and several unclassified ciliates increased in the summer after 2007, while the other usually summer-dominant dinoflagellate taxa decreased. The ability to exploit smaller prey, which are predicted to dominate the future Arctic, could be an advantage for these smaller ciliates in the wake of the changing climate. Network analysis further revealed increased co-occurrence and potential interactions between small phytoplankton and the microzooplankton groups, suggesting changing food web complexity with the changing environment.

Josephine Rapp

Diversity of hydrolytic enzymes among Arctic deep-sea sediment bacteria

Josephine Z. Rapp1,2, Christina Bienhold1,2, Halina E. Tegetmeyer2,3, Claudia Pala2, Pierre

Offre2 and Antje Boetius1,2

1Alfred Wegener Institute Helmholtz Center for Polar and Marine Research, HGF-MPG

Group for Deep Sea Ecology and Technology, Bremerhaven, Germany

2Max Planck Institute for Marine Microbiology, HGF-MPG Group for Deep Sea Ecology and

Technology, Bremen, Germany

3Bielefeld University, Center for Biotechnology - CeBiTec, Bielefeld, Germany

The vast majority of deep-sea ecosystems are sustained by exported organic material from

the productive, sunlit surface ocean. Bacteria dominate benthic communities both in

biomass and abundance, and have been recognized as the key players in the

remineralization of organic material. Since most sediment bacteria remain however

uncultivated and represent unknown taxa, we have very limited knowledge of their

metabolic capabilities and enzymatic machinery. Here we studied deep-sea surface

sediments along a seafloor depth gradient from 1200 m to 5500 m at the Arctic long-term

ecological research station HAUSGARTEN. We applied Illumina 16S rRNA gene surveys

based on DNA and cDNA, as well as metagenomic and -transcriptomic sequencing to

elucidate total and active bacterial community composition and gain insight into the

carbohydrate processing and uptake capabilities of deep-sea benthic bacteria. We identified

specific taxa of interest and quantified their cellular abundance using CAtalyzed Reporter

Deposition–Fluorescence In Situ Hybridization.

Results from the different molecular approaches were in good agreement and suggested

similar community structures with the same dominant members. Interestingly, typically

predominant sediment taxa, i.e. the JTB255 marine group, the Sh765B.TzT29 group or the

OM1 clade, were underrepresented in the active part of the community, while other usually

low-abundant taxa, i.e. Flavobacteriia and the SAR202 clade, were overrepresented. At low

taxonomic resolution, communities along the slope were similar, yet showed high turnover

at species level. Although, the repertoire of carbohydrate-active enzymes (e.g.

polysaccharide hydrolases) appeared unchanging along the depth gradient, the relative

contribution of distinct enzyme-coding genes varied. Specific glycoside hydrolases involved

in polysaccharide degradation of algae material (e.g. for laminarin; xylan) had higher counts

at shallow depth, while others responsible for the breakdown of bacterial cell walls (e.g. for

components of peptidoglycan) were more strongly represented at deep stations. Our

findings indicate an adaptation of the communities to differences in organic matter quality.

Andrea Garvetto

Phylogenetic characterisation of oomycetes infecting

toxic species of the marine diatom genus Pseudo-nitzschia


Garvetto A., Nézan E., Badis Y., Bilien G., Arce P., Bresnan E.,  Gachon C.M.M, Siano R.

In aquatic ecosystems parasitic interactions in the plankton have been shown to shape host species dynamics. At sea, sampling problems, the limited number of cultivable species, and the dearth of morphological information of protist parasites hampered the assessment of their diversity, phylogenetic position and ecological importance. According to morphological observations, a parasitoid of the marine toxigenic diatom Pseudo-nitzschia pungens has been tentatively affiliated to the oomycete genus Ectrogella, reported to infect both marine and freshwater diatoms. No phylogenetic affiliation was provided, opening the debate about its systematic position. By single-cell genetic analyses, we obtained 18s-rDNA sequences of seven distinct intracellular eukaryotic parasitoids, infecting four toxic Pseudo-nitzschia and one Melosira species in the North Atlantic coastal waters. Robust phylogenetic analyses demonstrate that our sequences cluster into two separate clades within the phylum Oomycota, being related to the seaweed parasite Anisolpidium and Olpidiopsis. Morphological features of our specimens were not sufficient to unambiguously attribute these parasites to any Ectrogella species. Therefore, we named our two Oomycota clades OOM-1 and OOM-2, awaiting additional morphological and genetic information. A screening of global databases of the regions V4 and V9 of the 18s-rDNA, demonstrated the presence of our parasites beyond the North Atlantic coastal regions. In a biweekly metabarcoding survey of the diatom communities in the Concarneau Bay (France, Brittany), high abundances of the parasite OOM-1-13-374 coincided once with the decline of Pseudo-nitzschia spp. and then with that of Cerataulina pelagica. This finding, together with the genetic identification of closely related oomycetes infecting both Pseudo-nitzschia australis and Melosira sp. supports the hypothesis of a broad host range for the studied parasites. Our finding highlights a complex and still unexplored genetic diversity within oomycete parasitoids of diatoms and calls for new biological evidence to unveil the evolutionary history and ecological role of these marine protists.   

October 12th 2017: Imaging techniques

Session 1: 09:00

Keynote: Heidi Sosik

Long-Duration High-Resolution Time Series of Plankton from Automated Flow Cytometry and Imaging


Heidi M. Sosik, Woods Hole Oceanographic Institution, Woods Hole, MA USA 02543


Many aspects of how and why marine plankton communities change through time remain poorly understood, in large part because traditional organism-level sampling strategies are not amenable to high-frequency, long-duration application. The combination of ocean observatories and automated sensors is now addressing this gap and accelerating the pace of discovery. Submersible flow cytometers and imaging-in-flow cytometers are capable of rapid, unattended analysis of individual cells and colonies. Over a decade of high-resolution observations in US coastal waters have provided measurements of 100s of millions of cells, which in many cases can be classified to genus or species with automated analysis. These taxon-specific, high-resolution records are revealing extraordinary detail about the biology and dynamics of plankton ecosystems, including occurrence patterns of harmful species, effects of life cycle transitions during blooms, influences of taxon-specific parasites on mortality, and climate-related impacts on phenology. In this talk, I will highlight examples of new insights and provide some perspective on the challenges in scaling these observational approaches to a more extensive network of aquatic locations. 

Nicole Poulton

Veronique Creach

New technology: beyond the challenges for a better use of phytoplankton data.


Véronique Créach (1), Felipe Artigas (2), Soumaya Lahbib (3), Melilotus Thyssen (3), Lennert Tyberghein (4).

1.     Cefas, Pakefield Road, NR33 0HT Lowestoft, UK


2.     UMR CNRS 8187 LOG, 32 av. Foch, 62930 Wimereux, France

3.     UMR 110 MIO, 163 Avenue de Luminy, 13288 Marseille, France,

4.     Flanders Marine Institute, Wandelaarkaai 78400 Ostend, Belgium



A common problem in environmental data management systems is the incoherence between the sequential steps of data collection, analysis and distribution of data to potential users. Beside the heterogeneity of the sample collection steps, the differences in subsequent techniques of processing, quality control, and transfer of data into information all require standardisation. Marine observing technologies add today a new challenge by increasing the frequency of measurements and by consequence the data to process and store. Prioritised by policy makers but also due to a longer experience in in-situ technology, physical and chemical data have increased their reliability and their visibility developing standardisation networks and delivering information in real-time to European portals. Today, the improvement and the availability of more flexible, and more reliable instruments for measuring the phytoplankton community could make a huge difference in the way of handling and using the biological data. During the last 10 years, scientists have developed collaboration around the North Sea, Channel, Baltic and Mediterranean Sea, sharing experience and knowledge on online flow cytometry measurement at high frequency. This consortium of scientists aims to establish now terms of reference, metadata, and quality control processes needed to create a framework of a new database for storing flow cytometry data for mapping the diversity of the phytoplankton in the areas of interest. This new access to data will ensure their optimal use and joint development for future services. This initiative has been funded by three European projects (DYMAPHY, JERICO-NEXT and SeaDataCloud) aiming to bring together key national and regional research infrastructures and European researchers, from both academia and private sector. An overview of the activities of the projects will be presented, highlighting the different steps and challenges encountered by the consortium related to data handling.

Lisa Campbell

High temporal resolution phytoplankton time series reveals bloom dynamics in the Gulf of Mexico

Lisa Campbell and Darren W. Henrichs

Department of Oceanography, Texas A&M University, College Station, TX 77843 USA

The Imaging FlowCytobot (IFCB) combines flow cytometry and video technology to capture images of individual cells which, together with machine-learning technology, enables near real-time reporting of individual phytoplankton species abundance and community composition. Continuous automated operation of the IFCB in the Gulf of Mexico has produced a 10-year phytoplankton time series at the Texas Observatory for Algal Succession Time series (TOAST). From these high temporal resolution observations (hourly to daily) at TOAST, the IFCB has provided successful early warning for eight harmful algal bloom (HAB) events since 2007. Initial bloom stages were detected with sufficient time to close shellfish harvesting and prevent human illness. Data have also been used to develop models to determine the origin of HAB populations and to provide new insights into dynamics of phytoplankton community structure, including the response to environmental forcing. In addition, near real-time data permits targeted sampling. Examples of targeted metatranscriptomics to assess community responses, demonstrates the benefit of combing IFCB and molecular approaches to assess the variability and diversity of the phytoplankton community.  

Session 2: 11:00

Klas Möller

A novel plankton observatory for integrative ecosystem monitoring within the framework of COSYNA: Resolving coastal plankton and particle dynamics in high-resolution using imaging and acoustics


Klas Ove Möller1, Boris Cisewski2, Philipp Fischer3, Christian Möllmann4, Holger Brix1, Markus Brand3, Gisbert Breitbach1, Wilhelm Petersen1, Maarten Boersma5, Burkard Baschek1


1Helmholtz-Zentrum Geesthacht, Institute of Coastal Research / Operational Systems, Germany

2Thuenen Institute, Institute of Sea Fisheries, Germany

3Alfred Wegener Institute - Helmholtz Centre for Polar- and Marine Research, Research group Fish Ecology, Germany

4University of Hamburg, Institute for Hydrobiology and Fisheries Science, Germany

5Alfred Wegener Institute - Helmholtz Centre for Polar- and Marine Research, Research group Shelf Sea System Ecology, Germany



Increasing human activities and climate change have been shown as major stressors especially in coastal marine ecosystems world wide. Although many methods have been developed to evaluate the status of single components of these ecosystems, there is still a crucial lack in assessing multiple ecosystem components in a holistic way at appropriate scales. This is particularly true for the zooplankton community being a sensitive indicator to environmental changes. We here present first results from a novel zooplankton underwater observatory which has been recently deployed in the German Bight (Southern North Sea). The cabled underwater observatory combines a remote-controlled underwater imaging system (Video Plankton Recorder), Acoustic Doppler Current Profiler and a CTD-probe, allowing continuous and automatic small-scale observations in near real-time of zooplankton species abundance and behavior (e.g. vertical migration and trophic interactions) and the associated hydrography covering temporal scales from hours to several months. This observatory is part of the Coastal Observing System for Northern and Arctic Seas (COSYNA) providing a unique dataset in combination with a suite of other sensor platforms including e.g. FerryBoxes, Research vessels, Gliders, HF-Radar, remote sensing as well as modeling. Furthermore, the zooplankton observatory is located in close vicinity to the Helgoland Road time series station allowing comparisons, ground truthing and combination of modern optical and traditional plankton sampling methods. Our integrative monitoring approach is bridging the gap between primary production and higher trophic levels, helps to identify and track rapidly occurring environmental changes and thereby provides a potential tool for integrated ecosystem assessment and management within the marine strategy framework directive.


Contact Author: Klas Ove Möller, Helmholtz-Zentrum Geesthacht, Institute of Coastal Research / Operational Systems, Max-Planck-Strasse 1, 21502 Geesthacht, Germany

e-mail: klas.moeller@hzg.de



Felipe Artigas

Automated in vivo approaches for currently monitoring phytoplankton in coastal marine waters: advantages and challenges

Artigas L. F.1*, Bonato S. 1, Claquin P.2, Créach V3, de Blok R.4, Deneudt K.5, Dugenne M.6, Grégori G6, Grosjean, P.7, Hamad D.8, Hébert P.-A.8, Houliez E 1,9, Karlson B.10,  Kromkamp J.11, Lahbib S.6, Lefebvre A.12, Lizon F.1, Louchart A.1, Ove Möller, K. 13, Petersen W.13, Poisson-Caillault E.8, Revilla M.14, Rijkeboer M.15, Rutten T.16, Tyberghein L.5, Thyssen M.6, Seppälä J.9, Stemmann L.17, Veen A. 15, Wacquet G.1, Wollschläger J.13

* Corresponding author: felipe.artigas@univ-littoral.fr

1 Laboratoire d’Océanologie et Géosciences (CNRS UMR 8187 LOG – Univ. Littoral Côte d’Opale – Univ. Lille), Wimereux, FR

2 CNRS UMR 7208, Biologie des Organismes et Ecosystèmes Aquatiques (BOREA- CNRS, MNHN, UPMC, IRD 207, UCN, UA) - Université de Caen, Caen, FR

3 Center for Environment, Fisheries and Aquaculture Science (CEFAS), Lowestoft, UK

4 Protistology and Aquatic Ecology, Ghent University, Ghent, BE

5 Vlaams Instituut voor de Zee (VLIZ), Ostende, BE

6 CNRS UMR 7294 Institut Méditerranéen d’Océanologie (MIO – CNRS, IRD, UM AMU 110), Aix-Marseille Université, Marseille, FR

7 Laboratoire d’Écologie Numérique des Milieux Aquatiques, Université de Mons, Mons, BE

8 Laboratoire d’Informatique Signal et Image de la Côte d’Opale - EA 4491, Université du Littoral Côte d’Opale, Maison de la Recherche Blaise Pascal, Calais, FR

9 SYKE Finnish Environmental Institute, Helsinki, FI

10 Swedish Meteorological and Hydrological Institute (SMHI), Norrköping, SE

11 Royal Netherlands Institute for Sea Research (NIOZ), Yerseke, NL

12 Laboratoire Environnement Ressources, Institut Français pour l’Exploitation de la Mer (IFREMER), Boulogne sur Mer, FR

13 Institute for coastal Research, Helmholtz-Zentrum Geesthacht (HZG), Hamburg, DE

14 AZTI, Marine Research Unit, Pasaia, SP

15 Centre for Water Management, Laboratory for hydrobiological analysis, Waterdienst, RWS, Lelystad, NL

16 Thomas Rutten Projects, Middelburg, NL

17 Laboratoire d'Océanographie de Villefranche (CNRS-UPMC), Université Pierre et Marie Curie, Villefranche sur Mer, FR



Phytoplankton micro-organisms, which are at the base of most food webs, mediate biogeochemical cycles and can be responsible for harmful events. Moreover, changes in their community’s growth rate, size structure, taxonomic and pigmentary composition, can occur at different time and spatial scales, evidencing rapid as well as long-term changes in environmental conditions. Currently, phytoplankton monitoring is based on discrete sampling and reference laboratory methods such as microscopic identification and counts, as well as pigment analysis. However, sampling frequency (fortnightly to monthly) and spatial cover (mainly in coastal single stations) is not sufficient to fully understand and evidence of phytoplankton dynamics in marine waters. Therefore, in order to better understand phytoplankton changes and to increase both the spatial and temporal resolution and automated in approaches are being deployed during the last decade. Whether being less precise in identifying different phytoplankton taxa or pigmentary groups than in vitro laboratory techniques (including molecular methods), these approaches provide new insights into phytoplankton dynamics and thus allow to gather useful complementary information for robust calculation of indicators, which are crucial for better defining the environmental state, trends and potential regime shifts within marine ecosystems. Moreover, when implemented in automated environmental monitoring platforms, as fixed stations, moorings, research vessels and/or ships of opportunity, these techniques can represent early-warning systems of phytoplankton changes, as the occurrence of blooms and, in particular, of harmful algal blooms (HAB), of special interest in areas of fishing, aquaculture and tourism. Therefore, there is an urgent need to improve the operability and discrimination of automated techniques addressing phytoplankton diversity (at taxonomical and/or functional levels) and productivity. Innovative optical sensors have recently been explored in previous studies, as in the DYMAPHY (2010-2014) cross-border European project. Within the Joint European Research Infrastructure network for Coastal Observatories – Novel European expertise for coastal observatories (JERICO-Next – H2020, 2015-2019), scientists are applying automated observation approaches for addressing phytoplankton dynamics, based on single cell/particle or bulk optical characteristics, in several European coastal and shelf seas, at high resolution, in (near) real-time. Three main techniques, image in flow acquisition and analysis, pulse shape-recording flow cytometry, as well as a combination of multi spectral photometry and fluorometry are being critically explored in order to better define their range of applicability in different case studies and to analyse the new information gathered on phytoplankton dynamics in coastal marine waters. We discuss about how these techniques would be relevant in a time series context and which challenges might need to be addressed for allowing their integration of these automated high throughput methods into marine time-series

Marc Picheral

Automated laboratory and field imaging systems for plankton are becoming more common to acquire monitoring time series or digitize preserved plankton archives


Picheral M. 1*

1 CNRS/UPMC, LOBEPM, La Darse, 06234 Villefranche sur mer cedex 4


* Corresponding author: Picheral@obs-vlfr.fr



Beyond their high acquisition rate and the fine sampling resolution they provide, a great improvement brought by these techniques for marine biodiversity time series is the possibility to verify the consistency of taxonomic classifications over time and build on previous work to refine them.

The bottleneck of imaging techniques remains the classification of the immense number of images they acquire. Even if modern algorithms predict classes with good accuracy, a percentage of images must still be checked or corrected by experts in order to provide classification confidence indexes, fully validated datasets, or a finer level of sorting. The second major issue is the variability and the multiplicity of ad-hoc taxonomy systems used in various labs, which hiders the inter-comparison of different datasets or time-series.

Based on our 10 years experience developing the classification tools for the Zooscan and the Underwater Vision Profiler (UVP) instruments, we recently developed a dedicated database and web application to replace them (EcoTaxa: ecotaxa.obs-vlfr.fr), in the framework of the Tara/Oceanomics project. Two years in, EcoTaxa hosts over 30 million images and metadata from diverse instruments (Zooscan, UVP, Zoocam, FlowCytobot, FlowCam, LOKI, microscopes, …). Its interface eases the automatic classification of images and the explicit validation by users from all over the planet; about 30% of the images are currently validated. The taxonomy in EcoTaxa is provided by the UniEuk project and based on genetic markers, which makes it consistent and potentially universal.

In Villefranche-sur-mer, we use this architecture daily to classify images from the Point B plankton time-series, started in 1966. We also build on it to provide quality check routines and daily updated displays of the validated data (e.g. www.obs-vlfr.fr/data/view/zoo/b/wp2/).


Bengt Karlson

Studies of plankton with a focus on harmful algae using imaging in flow and metabarcoding in the Baltic Sea and the Kattegat-Skagerrak


Karlson B.1*, Andersson A. 2, Brosnahan M.3, Cembella A.4, Hu Y.2, Krieger E.4


1Swedish Meteorological and Hydrological Institute, Research & Development, Oceanography, Gothenburg, Sweden

2Science for Life Laboratory, Division of Gene Technology, School of Biotechnology, KTH Royal Institute of Technology, Stockholm, Sweden

3Woods Hole Oceanographic Institution, Woods Hole, USA

4Alfred-Wegener-Institut, Bremerhaven, Germany


* Corresponding author: Bengt.Karlson@smhi.se




The Baltic Sea and the Kattegat-Skagerrak, adjacent to the Baltic Sea, covers a larger salinity range and thus a large variety in plankton communities. A combination of methods was used in studies of the plankton communities. Preliminary data will be presented from a study focussed on phycotoxin producing algae in the Skagerrak where the Imaging Flow Cytobot was used in situ to characterize the plankton community every 30 minutes. Weekly water sampling facilitated samples for microscopy and sequencing of 16S and 18S rDNA. In another study water samples were collected as part of the Swedish National Marine Monitoring Program during monthly cruises covering the Baltic Proper and the Kattegat Skagerrak. In a third study a merchant vessel with a Ferrybox system was used to collect samples in the same area. Metabarcoding revealed a much higher biodiversity than microscopy. Microscopy has advantages e.g. when estimating biomass of individual taxa. The results will be discussed in the context of incorporating the novel methods in long term monitoring and for short term forecasts for harmful algal bloom warnings useful e.g. to aquaculture.

In Villefranche-sur-mer, we use this architecture daily to classify images from the Point B plankton time-series, started in 1966. We also build on it to provide quality check routines and daily updated displays of the validated data (e.g. www.obs-vlfr.fr/data/view/zoo/b/wp2/).

Session 3: 13:30

Bank Beszteri

Bank Beszteri

Hustedt Collection, Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research

Moving from photographic illustration to high throughput imaging in the Hustedt diatom collection

Images have always been central to diatom taxonomy, and will probably continue to be in the future, in spite of the (unquestioned and still increasing) importance of molecular methods. Nevertheless, the way images are used is in continuous transformation also in this field. Our team in the Hustedt diatom collection has worked on bringing such a high throughput transformation of diatom imaging practices forward during the last years. By pulling together existing tools from medical virtual microscopy and academic image analysis, and complementing these with targeted own software development, we created a highly automated, but also highly manually controllable imaging and image analysis workflow applicable to diatom permanent slides (i.e., microscopic preparations on which silicate shells are embedded into a high refractive index mountant on a cover slip). In this talk, I will introduce this workflow, give examples of how we are attempting to use it both to address research questions and for a deeper mobilization of the diatom collection, and also address some of the main challenges we perceive on the road forward.

Barbara Niehoff

Comparing multi-net sampling with optical measurements: how efficient is the plankton recorder LOKI (Light frame In situ Key species Investigations) in analysing zooplankton communities?


Niehoff B. and Hildebrandt N.

Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research


Optical measurements are increasingly important in zooplankton studies as they allow for covering wide spatial ranges and study the distribution of the dominant taxa in greater detail than classical net twos. The plankton recorder LOKI provides high-resolution pictures, continuously taken by a 6 Megapixel camera during vertical hauls from 1000 depth to the surface. The build-in computer recognizes objects, i.e. particles and plankton organisms, and stores the respective images for later analyses. Linked to each picture, hydrographical parameters are being recorded, i.e. salinity, temperature, oxygen concentration and fluorescence. This allows to exactly identifying distribution patterns in relation to environmental conditions, rather than sampling depth intervals of up to several hundred meters as is possible with multiple net samplers. In order to compare the community composition, abundance and depth distribution of the species between samples taken by LOKI and multi-net Midi (Hydrobios) hauls, we have conducted parallel sampling of zooplankton during a RV Polarstern cruise. Both sampling devices were equipped with nets of 150 µm mesh size and vertically towed with 0.5m sec-1 from 1000m depth to the surface. The samples from the multi-net casts allowed for a high taxonomical resolution. In general, however, the abundances of the dominating large taxa (Calanus, Metridia, ostracods) determined by LOKI mirrored their abundance in net samples. Abundances of small copepods, especially Oithona spp. Were lower when determined from LOKI images, possibly due to optical restrains. Also, the abundances of fast swimming predators (chaetognathes, amphipods) were lower, in this case likely due to the construction of the LOKI net. Our optical method cannot fully replace net sampling, however, it yields high-resolution distributions of key zooplankton taxa and, in some species, developmental stages, which are not obtainable with traditional net sampling.

Tim Natkemper

Title: Marine image informatics  - New computational tools for the analysis of marine image collections


Abstract:Combining state-of-the art digital imaging technology with different kinds of marine exploration techniques such as modern autonomous underwater vehicle (AUV), remote operating vehicle (ROV) or other monitoring platforms enables marine imaging on new spatial scale and/or temporal resolution. A comprehensive interpretation of such image collections requires the detection, classification and quantification of objects of interest (OOI) in the images usually performed by domain experts. However, the data volume and the rich content of the images makes the support by software tools inevitable. In 2007 the Biodata Mining Group, led by Tim W. Nattkemper, started to investigate the potential of modern IT technology and machine learning algorithms for the support of manual annotation and the automation of particular steps and semantic annotation to overcome the serious bottleneck in marine image interpretation. The tall will give a review on the past projects and results of the group.


Prof. Dr.-Ing. Tim W. Nattkemper

Biodata Mining Group

Faculty of Technology

Bielefeld University

PO Box 100131

D-33501 Bielefeld


Mirco Scharfe

Emerging time series at the Helgoland Roads LTER site

A. Kraberg, K. Metfies, J. Hessel, P. Sprong, L. Käse, M. Scharfe

The Helgoland Roads LTER site has been sampled for phytoplankton continuously since 1962 (a bacterial time series was also started in 1962 but has since been discontinued). In 1975 a zooplankton time series was also added and these core measurements are accompanied by measurements of inorganic nutrients, temperature and salinity and more recently chlorophyll. 

While the time series on phytoplankton and zooplankton are still mostly sampled and analysed by traditional manual methods. Different high throughput methods are also becoming more important. In 2005 a stationary ferrybox system was installed which facilitates measurements of a range of physico-chemical parameters at a frequency of minutes. These provide important background data for the interpretation of the biotic data and are also used for the calibration of  instruments or the validation of models. 

Since 2015 molecular methods for eukaryotic taxa have also been introduced and the options for sustainably incorporating these into the routine phytoplankton monitoring are now being explored and the installation of a molecular biosensor into the ferrybox set up is planned. The level of comparability of data is now being evaluated.



SHERPA is a software tool developed by Michael Kloster implementing a highly customizable image processing workflow for the identification and quantification of object outlines. SHERPA was mainly developed for the analysis of bright field microscopic images of diatom valves but might be useful for other applications as well.

SHERPA is provided for free but without any guarantees, please read the below disclaimer before downloading.


New version SHERPA 1.1c available:


Ecotaxa (http://ecotaxa.obs-vlfr.fr/) is a web based application which permits to efficiently classify images of organisms using a reference taxonomy (http://unieuk.org/). The images, associated metadata and descriptive variables are loaded for each organism and permit to predict the identification (automatic classification). The registered users can then validate (check and perform a finer sorting) on the web by visualizing the classified images using an efficient interface. All operations are recorded to allow close control of the tasks. Ecotaxa hosts today more than 3000000 images acquired by different instruments and about 30% have been validated by experts which makes it the largest database of sorted images of plankton.

October 13th 2017:09:00

Synthesis session

Keynote: Frank-Oliver Glöckner

Challenges in Marine ‘Omics: Lessons from Ocean Sampling Day

Frank Oliver Glöckner

The Ocean Sampling Day (OSD) was initiated by the EU “Ocean of Tomorrow” project Micro B3 (Marine Microbial Biodiversity, Bioinformatics, Biotechnology) to obtain a global snapshot of marine microbial biodiversity and function. OSDs are simultaneous, collaborative, global mega-sequencing campaigns to analyze marine microbial community composition and functional traits on a single day. On June 21st 2014, 2015 and 2016 scientists from around the world collected more than 350 ribosomal DNA (rDNA) amplicon datasets and metagenomes plus a rich set of environmental metadata. Standardized procedures, including a centralized hub for laboratory work and data processing, assured a high level of consistency and data interoperability. Since 2015 OSD was accompanied by the Citizen Science campaign MyOSD to enhance the marine microbial community snapshot resolution of the OSD and to increase environmental awareness of the general public.

OSD was an experiment, not only by its research tasks, but also by its innovative character in activating and mobilising marine researchers and citizens alike to form a virtual research community that combines many brains, questions and approaches. OSD has shown that the full potential of recent technological advances can only unfold by moving towards an immediate and free exchange of data, technology and expertise to engage many brains from the start.

The talk will provide an overview about OSD including its first scientific results. It will stress the importance of contextual data and proper research data management for integrative research following the FAIR (Findable, Accessible, Interoperable, Reusable) data principles. It will reflect on supportive activities like the German Federation for Biological Data (www.gfbio.org) and the Council for Scientific Information Infrastructures (www.rfii.de).


Kopf A et al. (2015) The ocean sampling day consortium. GigaScience 4:27

Stefan Neuhaus

Q-Zip, a rapid meta-barcoding pipeline

Neuhaus S., Frickenhaus, S.

Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research, Am Handelshafen 12, 27570 Bremerhaven

With an ever growing amount of sequencing data, meta-barcoding analyses became very compute-time consuming. Furthermore, reproducibility and transparency of these analyses were hard to guarantee due to the complexity of the required analysis steps. The Q-Zip pipeline is a tool to tackle these challenges, incorporating established and modern tools into easily configurable and (re-)executable workflows. The pipeline provides a very fast way to reproducibly and transparently analyze and reanalyze thousands of samples with hundreds of millions of sequences due to largely parallelized processing, an easy workflow control and detailed logging.

Adriana Zingone

Session 2 11:00: Breakout groups