Fig. 4
From: Rodent-avoidance, topography and forest structure shape territory selection of a forest bird
Habitat variables discriminating breeding territories and control areas. Shown are plots of the five variables whose model-averaged coefficients did not include 0 (cf. Table 3). The solid lines are fitted values based on model-averaged coefficients of the seven top-ranked GLMMs of the across-hypotheses analysis (Table 2), the dotted lines show 95 % confidence limits. To calculate the fitted values, the variable of interest (x-axis) was varied within the observed range while the others were fixed on their average values. For each variable, inset box plots show median (bold line), 25–75 % range (grey box), range of data within 1.5 times the interquartile range from the corresponding quartile (whiskers) and observations beyond this range (dots) for occupancy probability equaling 0 (control areas) or 1 (breeding territories). Nterritories = 73, ncontrol areas = 73