Artificial intelligence techniques combined with data from the EU Copernicus Marine Service have been applied to the study of cetacean habitats in the northern Ionian Sea for the first time. This will help expand the understanding of their bioecology and improve conservation and protection efforts. The study, published in Scientific Reports, involved CNR-STIIMA, Jonian Dolphin Conservation, CMCC Foundation, the University of Bari and the University of Basilicata.

To better understand cetacean habitats and uses, the Institute of Industrial Systems and Technologies for Advanced Manufacturing of the National Research Council (CNR-STIIMA) in Bari conducted a study using artificial intelligence technologies for the first time, in collaboration with the Jonian Dolphin Conservation, the CMCC Foundation – Euro-Mediterranean Centre on Climate Change, the Department of Biosciences, Biotechnology and Environment and the Department of Computer Science of the University of Bari and the University of Basilicata’s School of Engineering. The study is published in Scientific Reports.

“Cetaceans are exposed to multiple anthropogenic stresses and climate change. Assessing the conservation status of these species is strategic for setting up effective plans for the sustainable management of the sea and, at the same time, for the conservation of sites that are crucial for marine fauna of Community interest,” explains Rosalia Maglietta, an artificial intelligence expert at the CNR-STIIMA. “We have used environmental descriptors provided by the CMCC Foundation, which were obtained through the use of spatial remote sensing techniques and numerical modeling products from the European Marine Core Service. These descriptors provide a wide range of information regarding the data collected on the environment in which dolphins live, for the first time investigated and presented in a scientific study. By exploiting the information contained in the environmental descriptors provided by the Copernicus Service, artificial intelligence made it possible to predict the abundance of three of the most common odontocetes in the northern Ionian Sea: the striped dolphin, the common bottlenose dolphin and the Risso’s dolphin”.

“Artificial intelligence has shown that the concentration of inorganic nutrients, such as nitrates, phytoplankton, temperature and salinity, are the environmental variables that most influence the distribution and abundance of the cetaceans under study. Indeed, as shown by the analyses conducted by the ecologists of the Department of Biosciences, Biotechnology and Environment of the University of Bari, these specific environmental parameters of the marine habitat are directly linked to primary production and, therefore, to the abundant occurrence of prey,” explains Roberto Carlucci of the University of Bari.

“The sighting data collected in the study area by the Jonian Dolphin Conservation Association have been invaluable for habitat examination. They were collected following a strict scientific protocol for over 10 years, between the summer of 2009 and the summer of 2022.

The Citizen Science activities developed by the association, with the involvement of citizens, students, and tourists, were also strategic for the acquisition of this data over time,” concludes Maglietta. “The study, with its multidisciplinary character, brings progress and new knowledge on how these odontocetes use their habitat. Moreover, the strategy of analysis and study developed could also be applied effectively in other geographical areas and on different cetacean species”.

The study: Maglietta, R., Saccotelli, L., Fanizza, C. et al. Environmental variables and machine learning models to predict cetacean abundance in the Central-eastern Mediterranean Sea. Sci Rep 13, 2600 (2023). https://doi.org/10.1038/s41598-023-29681-y