Line of action:

Status:

Execution date:

The overall objective of the iGENTAC project is the development of innovative tools, based on minimally invasive sampling techniques, for the implementation of coastal species monitoring programmes in a marine protected area.

OE1: Development of artificial intelligence (AI) techniques for the photoidentification of individuals and implementation of a capture-mark-recapture (CMR) system.
OE2: Development of molecular markers to identify individual specimens that allow the implementation of a CMR based on genetics.
SO3: Estimation of population parameters and spatial activity of the target species using CMR models applied to the data obtained in SO1 and SO2.
OE4: Comparison and evaluation of the results of OE3 using the acoustic telemetry monitoring network as a control method.
SO5: Provide advice on the design and implementation of monitoring plans for populations of marine organisms in an MPA.
SO6: Disseminate the role of MPAs in the management of coastal ecosystems and encourage user participation in monitoring plans.

A1 Mark-capture-recapture campaigns, photographs and sampling.
A2. Development of an individual recognition system based on artificial intelligence.
A3. Development of molecular markers to identify marine organisms at the individual level.
A4. Estimation of population parameters and spatial activity of the target species using capture-mark-recapture models, CMR.
A5. Establishment of acoustic monitoring network.
A6. Campaign to mark individuals with acoustic emitters.
A7. Activity analysis using acoustic telemetry and comparison with the results of CMR models.
A8.1 Initial information day and A8.2 Final project information day with the PNMTIAG managers.
A9.1 Dissemination of the project on websites and social networks; A9.2 Dissemination day and visit to school, A9.3 Citizen science workshops and A9.4 Press releases.
A10. Project Conclusions Seminar

The project worked on the monitoring of a coastal aggregation of a local mosaic ray population, Raja undulata, which takes place within the Maritime Terrestrial National Park of the Atlantic Islands of Galicia (PNMTIAG). This species is considered near threatened by the IUCN and is also an important resource for the artisanal fishing sector.

Deep learning algorithms were developed for the photo-identification of individual specimens of Raja undulata using, for the first time, a Siamese network architecture. Our model exceeded 90% accuracy over the validation set and achieved 70% accuracy over the test set, including recaptures of some individuals. This work was the subject of the master’s thesis of Nuria Gómez Vargas (University of Seville) and the work has been submitted to the high-impact international scientific journal, Methods in Ecology and Evolution.

Genomic tools have been successfully applied to generate a unique genetic fingerprint for each individual sampled from the local mosaic ray population. A database has then been generated with this genetic code, which allows us to identify each of the cells quickly and accurately from small tissue samples. This will allow for long-term monitoring. This work was part of the master’s thesis of Miguel López Aguilar (University of Alicante).

Acoustic telemetry has proven to be a fundamental tool for monitoring the behavior and movement of marine organisms. In this particular case, the acoustic data demonstrated a seasonal use of the waters of the PNMTIAG by this local mosaic ray population and support the results obtained in the censuses obtained during the acquisition of the images for the development of the deep learning algorithm. It has made it possible to define a series of ecotypes depending on the relationship of the specimens with the waters of the park: residents, seasonal visitors, itinerant visitors.

These results have been transferred to the PNMTIAG through continuous interaction with the park’s management body through seminars where the main results with implications for the implementation of long-term monitoring have been presented and discussed. The communication of the project has been worked on through workshops and citizen science actions, encouraging the participation of users in the monitoring plans (capture of images for photo-identification). Schools have been visited and the results have been disseminated on TV, press and regional and national radio.