| dc.description.abstract | This study was conducted in two different Mediterranean areas of central Greece and aimed at (i) comparing the breeding density and frequency of occurrence of shrike species between these two areas, (ii) identifying the key characteristics of habitats for each species, and (iii) evaluating possible underlying environmental factors determining the shrike's presence. To assess which factors mostly affect density, estimated with the point count method, we used the non-parametric test Kruskal - Wallis, Spearman's correlation coefficient, Canonical Analysis and Canonical Correspondence Analysis (CCA). Environmental variables tested were microclimate (air temperature and humidity), vegetation (herbaceous cover, shrub cover and shrub height), topography (altitude, slope, road presence, distance from nearest road, settlement and river) and habitat type (broad leaf forest, agricultural land, rangeland and abandoned olive groves). Between the two areas types, only Lanius senator mean breeding density was significantly different. The CCA model was significant (p < 0.05). Our results highlighted that L. minor did not occupy plots of agricultural land (cereals) and was negatively related to the distance of the nearest road. L. senator did not occupy plots of broad leaf forest and was positively related to the existence of the nearest road. Furthermore, the species exhibited the habit of perching along roadsides. On the contrary, L. collurio was characterized as habitat generalist in landscapes of central Greece. | en |
