On 26th April 2017, at 11:00, Gianpaolo Coro from the Istituto di Scienza e Tecnologie dell’Informazione (ISTI) gave a presentation entitled "Analysis of invasive species using species distribution models:  the silver-cheeked toadfish (Lagocephalus sceleratus) use case" in Rome, Italy, at the Food and Agriculture Organisation (FAO) of the United Nations headquarters.

The presentation summarized a method to estimate the possibile extension and impact of the silver-cheeked toadfish (an aggressive predatory causing devastating effect to native species and habitats) in the Mediterranean Sea based on species distribution modeling techniques. The approach consists on the following steps:

  1. Generation of species distribution models (SDM) showing the potential distribution of the species using different algorithms: Aquamaps Suitable Model, Maximun Entropy (MaxEnt), Artificial Neural Networks (ANN) and Support Vector Machines (SVM).
     
  2. An overall projection for the Mediterranean Sea was produced by merging together all the models distributions.
     
  3. Obtaining of an "observations-distance weighted habitat map" showing the reachable locations from known observations on the Mediterranean Sea.
     
  4. Intersection of the “observations-distance weighted habitat map” with common use Mediterranean sub-divisions, to obtain distributions for FAO Areas, Exclusive Economic Zones (EEZ) and General Fisheries Commission for the Mediterranean (GSA) subdivisions.
     
  5. Analysis of the distribution maps for 2050 to evaluate the future potential impact of the species.
     
  6. Analysis of the agreement between different models.
     
  7. Evaluation of the spread of the species from the Red Sea using a drifting simulation considering habitat suitability maps. 

 

The presentation is available to view here and can be dowloaded from here

 

Researchers

Gianpaolo Coro is a Physicist with a Ph.D. in Computer Science. His research focuses on Artificial Intelligence, Data Mining and e-Infrastructures. He has been working for more than fifteen years on Machine Learning and Signal Processing with applications to Computational Biology, Brain Computer Interfaces, Language Technologies and Cognitive Sciences. The aim of his research is the study and experimentation of models and methodologies to process biological data and to apply the results to fields in Ecological Modelling, Vessel Monitoring Systems and Ecological Niche Modelling with an approach oriented to Science 2.0. His approach relies on distributed e-Infrastructures and uses parallel and distributed computing via Grid- and Cloud-based technologies. He is one of the authors of the CMSY model for stock assessment in data-limited scenarios.