THE RELATIONSHIP BETWEEN TEMPORAL DYNAMICS OF LANDSCAPE PATTERN AND PLANT BIODIVERSITY IN COASTAL DUNES OF PROTECTED AREA “CASTELPORZIANO PRESIDENTIAL ESTATE”

Coastal dunes are dynamic ecosystems located along sandy shorelines following strong sea-inland environmental gradients. Conservation status of coastal plant communities is crucial for habitat preservation, biodiversity protection and to provide essential ecosystem services [1]. Worryingly, these communities are facing significant threats from human activities leading to strong landscape alterations [2]. Improvements on monitoring with innovative cost-effective tools are a challenge for coastal plant communities. The availability of open access, remotely sensed data with increasing temporal and spatial resolution is promising in this context. Our aim is to better understand the long-term relation between landscape pattern and plant biodiversity of coastal plant communities in the protected area “Castelporziano Presidential Estate” over time 35 years by combining aerial orthophotos and on field resurveyed data. Inside the protected area, 58 plant community plots were sampled between 1988 and 2000 and resurveyed in 2023. At the same time, we mapped the protected area using free aerial orthophoto of 1988, 1998 and 2023 coeval to plot surveys. According with literature [3], we identified the following landcover classes: Beach Pioneer Vegetation, Herbaceous Dune Vegetation, Woody Dune Vegetation, Forests, Afforestation, Artificial, River, Sea, Semi Natural Vegetation. We calculated landscape metrics (e.g. Pland, Mean Patch Area, Patch density and Edge density) on land cover map buffers of increasing diameter (from 25 to 150 m) around the resurvey sites for analysing the temporal landscape pattern dynamics. In each past and present plot, we calculated diversity using Shannon index (S) both at taxonomical and landscape level on buffers. Thereafter we fitted Linear Mixed Models (LMM) for analysing diversity changes of habitat classes over the years. Our preliminary results show a decreasing trend in artificial cover, related to natural encroachment and the creation of new natural patches. Overall, our models show that variability is well explained (R2> 0.85), with higher values for larger buffers. Specifically, we mostly assessed significant differences between 1998 and 2023 showing a general trend of decrease in landscape diversity; we can interpret this result as a consequence of natural habitats encroachment which drives to larger homogeneous patches over the years. In fact, we observed a combination between increasing in Patch Area and a concomitant decreasing trend in Patch Density at the landscape level across the habitat classes. At taxonomical level, our results demonstrated a general habitat homogenization associated with forest canopy closure. Further analyses are needed to better understand the relationship between landscape and plant species diversity.