| Abstract |
Bearded Vulture [Gypaetus barbatus (Linnaeus, 1758)] is a solitary cliff-nesting raptor inhabiting mountain regions of Eurasia and Africa. In Greece, Bearded Vulture is one of the most endangered vertebrates. The island of Crete hosts 7-8 breeding pairs which constitute the last breeding population of the species in southeastern Europe, except Turkey. The current thesis examines the patterns of movement and habitat use of young individuals during their post-fledging dispersal. The time period between the first flight from the nest and the onset of dispersal is one of the most critical stages in the life history of birds. All birds were captured during 2001-2015 and were fitted with VHF transmitters. A total of nine individuals were radio-equipped twenty days before fledging and were monitored for ca. a year. The home range (95% of occurrences) of the species and the core areas (50% of occurrences) were calculated with the Kernel Density Estimators (KDE) which is an appropriate method for detecting the most frequently visited areas and the minimum convex polygon (MCP) which is commonly used for comparison purposes with other studies. Overall, the species uses almost the entire island of Crete (KDE 95%= 5.316,6 ± 2.524 km2) and core area at 2.556,3 ± 2.268,8 km2. MCP 100% was estimated at 3.700,5 ± 2.054,1 km2 and no overall spatial seasonal asymmetry was detected in the species distribution. However, six individuals showed uniformity of distribution when examined for home range overlap by applying a Utilization Distribution Overlap Index (UDOI). Moreover, the factors that determine the settlement areas of the species were investigated using a Geographic Information Systems (GIS) tools and applying a generalized linear regression model (GLM). Last, prediction maps of the spatial distribution of the species were constructed via maximum entropy machine learning software (MaxEnt). The prediction maps were examined regarding their percent overlap with the NATURA 2000 sites and the Important Bird Areas (IBAs) in Crete. Fourteen environmental variables regarding climate, topography and human pressure were used. However, after multicollinearity examination three variables, i.e. annual mean temperature, temperature annual range and elevation were excluded from further analysis. The most important variables for habitat selection were found to be the annual rainfall, the slope, and the human pressure. GLM also included air velocity, distance from water and the vegetation index NDVI in June, while the maximum entropy model included two bioclimatic variables (rainfall seasonality, temperature seasonality). The positive selection of rainfall can be explained through its correlation with altitude. Also, the increase of probability of presence with increasing of slope is associated with the existence of thermal currents, the availability of areas where the species break bones and presence of rocky ledges that can be used as food storages. The positive response between average air velocity and habitat selection, in GLM regression model correlate with the morphological characteristics of young vultures. Young individuals use thermal and deflected upward air currents to increase flight efficiency at the lowest
energy cost. The variable of human pressure seems to be a limiting distribution factor, with young Gypaetus barbatus selecting mountainous areas of Crete with low human presence. AUC score was calculated 0,89 for GLM model and 0,846 ± 0,016 for maximum entropy model. True Skill Statistic (TSS) was also calculated for the average of 10 replicates at 0.55, which indicated that the final model has good ability to predict the species presence. The overlap between the species post-fledgling dispersal areas with the NATURA 2000 site and the Important Bird Areas (IBAs) networks was 72.3% and 79.5% respectively. Understanding the movement of this endangered species through telemetry techniques proved to be important for its management and a very useful tool for future conservation planning.
|
Search
