Daniela Gigante

The Annex I of the 92/43/EEC “Habitat” Directive includes semi-natural dry grasslands, with the code 6210, among the habitats of European relevance, holding a priority importance (indicated with an asterisk) when hosting a remarkable presence of orchids. Semi-natural dry grasslands are among the most species-rich plant communities in the world (the richest in the Temperate climates), making their sampling particularly complex, and reserved for highly skilled operators. Periodic monitoring of their state of conservation is a duty for each EU27 country where they are present, implying a huge effort in terms of time, human power and money. In this frame, robotic monitoring might represent a valuable support, provided that robust proxies are identified, appropriate in terms of both efficiency in predicting the state of an ecosystem (in terms of structure and functions), and recognizability by the AI operators. Semi-natural dry grasslands present a huge variability in terms of substrata, soil, slopes, cover, size, structure, physiognomy, color, only to mention the most relevant ones. In the frame of “NI” Project ("Natural Intelligence for Robotic Monitoring of Habitats”, call H2020-ICT-2020-2, ICT-47-2020), a set of synthetic indicators will be selected, concerning physical environment, vegetation structure, and key species occurrence, in order to use the robot's intelligence to help assessing the habitat's conservation status, by detecting occurrence and distribution of the implemented indicators. In particular, “typical” species (sensu Habitat Directive) useful for status assessment will be pointed out, and their features taught to robots, through images at both detailed and landscape scale. At the same time, plants indicative of grasslands transformation, such as edge species (e.g. Asphodelus macrocarpus, Pteridium aquilinum), shrubs species (e.g. Prunus spinosa, Cytisophyllum sessilifolius, Rosa sp. pl.), young trees (e.g. Acer opalus subsp. obtusatum, Quercus cerris, Sorbus aria), alien invasive taxa (e.g. Senecio inaequidens), will be chosen as early warning species when suitable to be easily detected by the monitoring robots. The final aim is to equip the quadruped robots under development by the “NI” Project with tools for indicators detection, in order to be able to successfully operate in different grassland environments, implementing monitoring protocols, supporting the human investigation for those repetitive and time-consuming activities and, by that, contributing to biodiversity conservation.