Wetland and Wildlife Monitoring Using Technology and Acoustics (WatchOut)

Project duration: November 2022. – November 2025.

 

Coordinator: Faculty of Agrobiotechnical Sciences Osijek, Josip Juraj Strossmayer University of Osijek (https://www.fazos.unios.hr/)

Project partners are:

• Rainforest Connection, nonprofit organization (USA) (https://www.rfcx.org)
• Faculty of Electrical Engineering, Computer Science and Information Technology, Josip Juraj Strossmayer University of Osijek (https://www.ferit.unios.hr/)
• Cultural and development centre of Bilje municipality (https://krc.bilje.hr)

The project goal is to monitor biodiversity, to predict the stability of the ecosystem of the Kopački rit Nature Park and to provide the solution to facilitate informed decision-making in the preservation of natural heritage. Development of a convolutional neural network for selected and/or protected animal species in Kopački rit Nature Park. Collecting, processing and analysing of large data with the help of machine learning.

Purpose of the project research is usage of the high technology in biomonitoring of flora and fauna in Kopački rit Nature Park. Results of the project will contribute the protection of endangered and protected animal species, distribution of monitored animal and plant species, and climate changes influence on their biology and ecology will be determined.

The remote monitoring of biodiversity by collecting audio and video records that are analysed by using special digital tools, like algorithm and artificial intelligence, for recognizing and identifying species will be carried out along with standard field research. Using aerial recordings taken by drones and satellite images representation and spatial coverage of certain land, wetland and water habitat will be assessed, as well as impact of nature and climate changes of the habitats of endangered and protected animal species.

Four groups of habitats are selected for implementation of project activities:

• Flood forests of white willow, white poplar and black poplar (connection Salicon albae and Populion albae);
• Pedunculate oak forests out of flood reach (connection Carpinion betuli);
• Wetland habitats, reeds and rushes (connection Phragmition australis), tall sedges and sedge-bed marsh vegetation (connection Magnocaricion elatae) and
• Freshwater habitats: watercourses, canals, lakes, ponds and puddles, overgrown with submerged, floating or amphibious vegetation (connection Lemnion minoris, Hydrocharition, Magnopotamion and Nanocyperion).

The priority species for the application of machine learning in the project are white-tailed eagle, red deer, great spotted woodpecker, European tree frog and golden jackal.