PS 103-230
Home-range scale's attributes limiting the habitat selection pattern on magellanic woodpecker

Friday, August 14, 2015
Exhibit Hall, Baltimore Convention Center
Gerardo E. Soto, Ingenieria Geografica, Universidad de Santiago de Chile, Santiago, Chile
Pablo M. Vergara, Gestión Agraria, Universidad de Santiago de Chile, Santiago, Chile
Ingo Hahn, Geo-ecology Research Group, Munich University
Christian Perez-Hernandez, Departamento de Gestión Agraria, Universidad de Santiago de Chile

We study how habitat disturbances at home range scale influences the tree selection pattern of Magellanic woodpeckers (Campephilus magellanicus), under the assumption that foraging quality for magellanic woodpeckers can be addressed by measuring tree's senescence, and based on hierarchical habitat selection theory.

                We used VHF telemetry data from two years and from six different family groups (n=516), from summer 2012 to fall 2013 in the Navarino island, Cape Horn, Chile (55°S). Using a high resolution WorldView imagery we mapped each tree by using the multiresolution segmentation algorithm in eCognition software and we classified them by Plant Senescence Reflectance Index (PSRI), age and species.  Using Kernel density estimators, we estimated the home range limits and area. We calculated the mean PSRI (mean quality), proportion of Old-growth forest and species composition at this scale. We carried out Synoptic habitat selection models to assess the selection patterns of family groups for all the measured attributes at tree scale. Last, we used the results of Generalized Linear Models (GLM) to test our hypotheses. 


We found a strong negative relation between the size of home range and mean PSRI (β=-103.4, p=0.063). Mean quality at home range scale was explained by both Old-growth forest and Lenga (Nothofagus pumilio) proportion (β=0.892, p=0.043 and β=0.806, p= 0.062 respectively). The only significant covariate at tree scale explaining the mean quality at home range scale was the PSRI of the selected trees (β=-6.034, p=0.072), showing a steep negative effect.

                These results corroborates our hypotheses based on hierarchical habitat selection theory and show that future efforts on forest management must be directed to address the processes at home range scale, since these processes rules the dynamics at lower scales.

Acknowledgments: Funding was provided by FONDECYT N°1131133 grant