Home · News · Antonio Garcia Quintas Predicts the Risk of Coastal Marine Mercury Pollution Using Seabirds and AI

Antonio Garcia Quintas Predicts the Risk of Coastal Marine Mercury Pollution Using Seabirds and AI

Antonio Garcia Quintas has first worked as an ornithologist in a research centre in Cuba before becoming an international ISblue postdoc. He obtained his PhD in Sète with the Marbec laboratory and the University of Montpellier. Let’s learn a bit more about him!

Your project is about: Tracking down anthropogenic mercury in coastal marine areas: a deep learning approach using satellite data and seabirds as ecosystem sentinels. Can you tell us a bit more about it?

This project is pioneering an interdisciplinary approach that combines marine ecology, chemistry, oceanography, and artificial intelligence (AI) to predict mercury (Hg) contamination risks in coastal ecosystems. It aims to use innovative methods to address gaps in data availability, particularly in Cuba and the Caribbean basin.

The project integrates traditional mercury concentration analysis with modern techniques like GPS tracking and satellite observations, alongside AI methods such as Convolutional Neural Networks (CNNs). By doing so, it aims to create a comprehensive view of marine mercury pollution in the Caribbean basin.

For me, it is an opportunity to expand my expertise beyond seabird ecology. I am learning more about oceanography and biogeochemistry. I am also working with experts like Ronan Fablet from IMT Atlantique to train a model using deep learning techniques. Collaborating with specialists in trace metals and isotopes will enhance the possibility to assess pollution at sea.

Specifically, this project will bring valuable insights, especially for Cuba and the wider Caribbean region, where environmental research remains limited.

How is the project going to be organised?

I am going to work on sterns and gulls because they are the most representative birds in Cuba. The idea is to study the foraging patterns (to search the food) of the birds when breeding – as it is unknown for Cuba so far.

I will be able to study samples of chicks and adults individuals to calculate the mercury level in their blood and feathers.  The idea is to register data from several breeding sites in Cuba and maybe another island of the Caribbean… Maybe in Granada, Guadeloupe or Bermudas even if it is not in the Caribbean basin. 

We also want to take satellite images from different satellites (Lansad, another oceanographic satellite…and that is why Elodie Martinez from Lops laboratory is involved in this project.

Based on satellite image, to identify the places where we take the samples and then in collaboration with La Rochelle University, we will make the mercury analysis. We will have a measurement of the mercury concentration from the tissues of seabirds and the satellite images to characterise the oceanography, the ecosystem of this place in particular. 

Then we want to build a deep-learning model with Ronan Fablet to develop a CNN (conventional neural network). The aim is to process this information to obtain a model that can classify risk of pollution of mercury at sea based only on satellite image to make predictions in the Caribbean islands where there is not yet information about; and to train the model on the potential possibility that it can be suitable for another tropical ecosystem.

It is a mix of researchers from different disciplines and that makes the project very interesting. For me, what is new is the tracking of seabirds during foraging with GPS and the oceanography side. Previously, in my PHD, I have been working with the CNN model but not so much with oceanography. Learning something new is always interesting but it can also be hard.

How did you hear about the ISblue postdoctoral call? 

Sophie LANCO (IRD MARBEC) in Sète was my PhD supervisor and she had already worked with Anne Lorrain (Lemar). Anne told me about the ISblue call and encouraged me to apply as the project was interdisciplinary and focused on another region to explore (Caribbean).

Will you go to Cuba to collect the samples on sea birds then?

Yes!

How do you collect the samples exactly?

We catch chicks and we collect a few feathers – it is very easy. For adult birds, we put a GPS for tracking but then we have to recapture them to pick up the GPS and collect the data on foraging. In addition, we want to capture the mercury and the isotopes of nitrogen and carbon to have an idea of the main foraging locality and the characteristics of their diet.

What are the implications for the bird to have mercury in their blood?

It is a very serious metal. We have previous data from my PhD where we evaluated the mercury concentration and in some areas of Cuba, the birds had very high values so apparently, there is high mercury pollution at sea yet we do not really know the source of this pollution. We want to make a deep approach in this context. There are similar problems on other islands but we do not have data yet.

For example, in some places, it could be a natural source of mercury around upwelling (volcanic area) but this is not the case for Cuba. That is why with this project, we want to make a prediction model for the authorities to enquire about the mercury pollution sources.

So now, can you tell us what you think of Brest?

I arrived in Brest in February and “la pluie, la pluie…”.However I find that Brest is a very modern city in terms of architecture compared to the rest of France. It is exciting to discover a new region of France – very contrasting to the south of France. It is great for me to be here for the science and the discovery of Brittany.

What do you think about this postdoc so far?

I am sure it is going to be a nice experience! It is a challenge as it is a big and complex project.

How did you come to do your PhD in Sète? 

The educational system is very different in Cuba compared to Europe. So I have worked for 10 years before doing my PhD as an ornithologist. I was working at the Coastal Ecosystems Research Center of Cayo Coco in Cubaand I obtained PhD grant from this center and IRD.  The PhD was in Sète with Sophie Lanco from Marbec (University of Montpellier) as I had worked in Cuba with her previously.

Did you manage to go back to Cuba from time to time during your PhD?

My IRD grant for my PhD was supposed to be 6 months in Sète and 6 months in Cuba for fieldwork but the situation became complicated because of Covid. In the end, I could only collect all the data for my thesis in one season instead of three. Thankfully, my PhD was extended by 6 months as Covid had affected my work plan so I was then able to complete my thesis.

Do you have any plans for after your ISblue postdoc?

Maybe I will do a new postdoc to reinforce my research and to carry on collecting ecological data for Cuba as it has been understudied.

More informations :