**3.2 Home range size**

314 Modern Telemetry

Over a three 3 year bear monitoring period (2007-2009) during the project's second phase, field work has focused on satellite telemetry combined to systematic collection of bear signs of presence and activity. An additional monitoring protocol was developed using thermo-

In total twenty two (22) adult and sub-adult brown bears have been fitted with GPS/GSM radiocollars: eighteen (15) males and seven (7) females. Bears were trapped from April to mid May and from September to mid-October in 2007 and 2008 within a buffer zone of 10 kilometres along the Egnatia Highway stretch 4.1 (routing from "Grevena to Metsovo"). We captured bears using Aldrich foot snares (Johnson and Pelton 1980) and immobilized bears with a zolazepam-tiletamine /medetomidine combination and reversed with atipamezole (Riegler et al. 2009). Body measurements were recorded and a premolar was extracted to

Photo 11., 12., 13. and 14. Bears equiped with radio-collars – radio-collars types:

sensitive, IR and conventional pre-programmed video and photo cameras.

**3. Materials and methods** 

determine bear age (Stoneberg and Jonkel 1966).

**3.1 Telemetry protocol** 

Simplex, Tellus GSM

We calculated home ranges with Arc View 3.2.a and the Home Range Extension (A. R. Rodgers and A. P. Carr, Ontario Ministry of Natural Resources) using the minimum convex polygon (100%MCP), 95%MCP, Fixed Kernel method and 50% contours of activity for core areas (areas of high intensity of use). The 100% MCP estimates were used to facilitate comparisons between studies and regions. Fixed Kernel Method range analysis was performed because, in addition to estimating range size, it reveals range use patterns, using a smoothing factor determined by least squares cross validation (LSCV) (Seaman and Powell 1996). We ignored autocorrelation within the data because the data continued to exhibit a high degree of dependence even when using extended fix intervals (24 hr; e.g., Reynolds and Laundre´ 1990, Rooney et al. 1998, De Solla et al. 1999). We tested collars GPS accuracy in the field and the mean error was 30m (Giannakopoulos et al. 2010).

Data gathered from the aforementioned methods were mainly used to identify bear presence as well as bear movements patterns and spatial behavior versus the highway alignment and especially in correlation to two main factors:

