Salta al contenuto principale
Passa alla visualizzazione normale.

MAURIZIO SARA'

Spatial analysis of lanner falcon habitat preferences: Implications for agro-ecosystems management at landscape scale and raptor conservation

Abstract

Sicily hosts the largest European population of the endangered lanner falcon, a poorly known species which needs conservation planning based on habitat preferences. A distribution model on 10 10 km cells of Sicily was described using Generalized Linear Models and variation partitioning methods. This modelling approach extracted explanatory factors, pure and joint effects of greatest influence from subsets of variables controlled for multi-collinearity and spatial autocorrelation. Analytical cartography used the environmental favourability function to assess habitat preferences, and the insecurity index estimated the degree to which lanner falcon occupancy is represented in the Natura2000 networks of Sicily. The lanner population is not randomly distributed across the geographical space as the significant latitudinal effect revealed. The most parsimonious explanatory model suggested traditional agro-ecosystems (i.e. arable lowlands and grasslands, with rugged terrains and cliffs) as the best predictors of lanner occupancy, and gave strong support to the negative effects of land abandonment and intensification (i.e. increase of heterogeneous areas and shrubby vegetation). The variation partitioning method suggested how an alteration of traditional agro-ecosystems might enhance interference competition with the peregrine falcon and limit lanner falcon occupancy. Most of the lanner falcon favourable habitats fall outside the Natura2000 networks, thus the main challenge for its conservation is represented by agri-environmental measures to be taken within the Common Agricultural Policy of the European Union. Conservation of traditional agro-ecosystems devoted to top-predators, like the lanner falcon, requires single environmental management agreements for multiple farm units, deployed at landscape scale on a network of favourable areas.