During our stopover in Villefranche-sur-Mer, the sailors left the boat and were invited to visit the Observatoire Océanologique. The research station was created in 1884 in a former prison for galley slaves of the kings of Sardinia – a charming place organized around a lovely courtyard. The buildings of the prison and the old forge are a maze of laboratories, offices, aquariums, animal tanks, and a library. Part of the space serves as a storehouse for Niskin bottles (well known to Taranautes), sea kayaks and diving equipment, opening onto the jetty just a few meters beyond. During the visit, by chance the crew encountered several scientists who had voyaged aboard Tara. Among them, Jean-Baptiste Romagnan, an engineer specialized in studying and analyzing plankton using imaging tools. He is still working on the analysis of data from Tara Oceans (2009-2012) – a major scientific adventure he participated in twice. He talked to us about the data collected during the mission.
When did you join the Tara Oceans expedition, and for what mission?
The first time was for my thesis, in October 2009, between Naples and Malta. The second time was in the fall of 2011, between Ascension Island and Rio. During my first voyage, we were still finalizing the protocols. I was in charge of collecting zooplankton in nets, and worked on the deck alongside the engineer Sarah Searson. In the fall of 2011, I again participated in sampling with nets and deployment of instruments with the deck engineer.
During that mission, a great quantity of data were collected. How are they being treated?
We collected thousands of samples – tubes containing plankton – during the Tara Oceans expedition and also Tara Oceans Polar Circle (2013), and we will continue collecting in the coming months in the Mediterranean. When these tubes are brought into the lab, we extract information in several ways: some scientists do the genetics, others analyze the samples using imaging tools. This is what I’ve been doing with the samples from Tara Oceans, with the help of many students, for almost 5 years using the Zooscan, a plankton scanner.* We have processed approximately 75% of the data collected. The procedure is always the same: we remove the formalin, take a portion of the sample and place it on the Zooscan to obtain images – small thumbnails of each object – from the big scanned picture. These images are analyzed, and from measurements on the thumbnails we do “automatic learning” – in other words, we ask the computer to identify the plankton, before confirming the identifications ourselves. This used to be done using a dissecting microscope, but it took time and required a lot of expertise. At present, we have developed tools that allow us to go faster and analyze a large number of samples.
A group of samples represents how much archived data?
Billions! The Zooscan is a tool that was developed to meet several needs. The first is to generate data from oceanographic campaigns soon after collection, because in the past it would take several years to analyze plankton data like ours. The second is a need for storage: samples in tubes are not eternal, they can be damaged or subject to accidents. Digital archiving allows us to store our data in several places for optimal conservation. The third need responds to scientific issues such as measuring biomass, biovolume, size and size spectrum. In fact with the pictures, we can automatically measure each organism and obtain accurate and consistent measurements. From these measurements, we derive information on the functioning of ecosystems. Plankton can be seen through the “lens of biodiversity,” or through the “lens of structure size” to answer different questions: How many of them are there, why, where are they, etc.
What would you say about the global scale of the Tara Oceans expedition?
This was an exceptional expedition, as was Tara Oceans Polar Circle. They were expeditions of a new type that had not been done for decades, even centuries. They can be classed among the great naturalist expeditions like those of Darwin or the Challenger. The original idea was to sample all the living plankton, only plankton, but all the organisms – from viruses and bacteria to the largest species of gelatinous plankton. The goal was to develop a sampling of all the biodiversity and complexity of plankton, in order to establish a “photographic” inventory of plankton biodiversity on a global scale. Expeditions like these generate decades of research work. Analyses are currently underway here in Villefranche, and also at the Station Biologique in Roscoff, as well as in other partner laboratories.
How was your experience aboard Tara?
My times on board were awesome! Tara is a relatively small boat compared to usual oceanographic research ships. The implementation of such a complex and comprehensive sampling system aboard the schooner is a remarkable feat. It’s a different approach to oceanography, and very intense. Finally, we work as a group, interacting with Tara Oceans consortium partners. We meet several times a year and it’s really interesting to do the science together. This community is very endearing. And personally, I grew up as a scientist with Tara. It’s been a formative project for me, and I’ll continue to work on the data.
*Invented in 2003 by Gaby Gorsky, Philippe Grosjean and Marc Picheral at the Observatoire Océanologique of Villefranche-sur-Mer.