Assessment of river ecosystems by a multitemporal and multilayered landscape approach. The semiaquatic otter (Lutra Lutra L) as a paradigmatic species

Fluvial habitats represent key ecosystems for landscapes equilibrium: to preserve such a crucial role scientifically based strategies taking into account their multilayered structure are mandatory. The research aimed to identify and test a kit of sound indicators of fluvial basin criticalities able to synthesize the river multi-layered structure. As an indicator of the entire basin status we used the distribution of suitable habitats for a key species, the Eurasian otter (Lutra lutra L.). Suitable habitat distribution was provided by inferential Habitat Suitability (HS) models that are considered fundamental for providing a geographic perspective on supporting conservation strategies both for the chosen species and for its habitat. The study was carried out in Southern Italy (Basilicata Region), in a consistent part of the current Otter’s Italian core area. We used two different HS modelling: the ENFA model and the multi-modelling BIOMOD platform. ENFA model was applied at a single period (2006) and was first selected since it produces reliable results by using only-presence data. The links between the river and its upland areas were described by measuring a fine scale index of Fluvial Functionality (FF): the Italian standard protocol for FF. Results from both HS ENFA model and FF analyses were integrated in order to identify the main critical river sectors, showing different vulnerability levels. Each level synthesizes information coming both from the broad and the fine scale of the river system and provides a prioritization of the tracts needing restoration. The integration was tested in two sub-basins, by analysing the land-cover fragmentation inside each original and derived map, through the calculation of specific Landscape Metrics. Considering that river systems are highly dynamic, we implemented a multi-temporal HS model for two different periods (1985 - 2006) and we compared them in order to identify the environmental factors accompanying the natural key species expansion. Such an analysis was performed by using the BIOMOD ensemble approach to test previous modelling findings and to improve the final product. The results demonstrated that distribution models for a key aquatic species can be considered robust indicators of the river health status and correctly captured the system complexity, which is crucial for defining indicators aiming at monitoring purposes. The HS and FF integration effectively synthesized the relation between the water courses functions, the landscape structure and the key species distribution, representing a helpful tool for understanding the effect of anthropogenic pressures on river functionality. Furthermore, the approach resulted very effective for the identification of most vulnerable fluvial tracts, allowing for a correct prioritization of restoration interventions, which represents a crucial challenge. The multi-temporal analysis evidenced that the improvement of Otter population is coherent with the recovering of natural land covers confirming the essential role of riparian vegetation. Finally, considering that the correct management of endangered species is the most effective way for preserving a satisfactory biodiversity level, the approach represents a concrete answer for dealing with the conservation of river’s biodiversity and functionality.