54 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124. No. I. March 2012 
TABLE 2. Characteristics ot environmental parameters of farmland and high altitude meadows used by Red-backed 
Shrikes for breeding in central Italy. 
Characteristics 
Farmland (n = 132) 
Meadows (n = 18) 
u 
p 
Building, % 
Cultivated. % 
Vineyards, % 
Forest. % 
Grasslands, % 
Uncultivated and shrubs, % 
Water. % 
Roads, % 
Landuse categories, number 
Near road, m 
Near building, m 
3.4 ± 6.8 
57.4 ± 26.1 
0.9 ± 3.6 
6.7 ± 13.2 
14.2 ±21.5 
11.0 ± 19.9 
0.3 ± 1.4 
6.0 ± 3.2 
3.7 ± 1.2 
30.6 ± 43.7 
186.4 ± 136.7 
0.8 ± 2.6 
2.4 ± 3.9 
0.0 ± 0.0 
27.6 ± 28.4 
65.2 ± 27.9 
0.6 ± 2.1 
2.2 ± 8.4 
1.2 ± 2.6 
2.9 ± LI 
145.6 ± 123.6 
781.6 ± 475.2 
862.0 
64.0 
1.080.0 
490.0 
189.0 
818.5 
1.052.5 
362.0 
692.5 
369.5 
355.5 
0.028 
0.000* 
0.184 
0.000* 
0.000* 
0.016 
0.070 
0.000* 
0.003* 
0.000* 
0.000* 
* Significant at P < 0.01. 
density values differed between the two areas used 
for breeding ( Z 2 = 47.3, P < 0.05). The percentage 
of open space in high altitude meadows (67.0 ± 
28.2%) was higher than in farmland (31.3 ± 
26.5%). In particular, the main types of land-use 
cover functioning as open space in farmland were 
roads, and uncultivated land, while in meadows the 
main type was grassland. The edge density value in 
farmland (93.4 ± 45.8) was higher than in meadows 
(50.9 ± 54.3). 
There were significant structural differences 
between the two habitats in terms of distance of 
nests from the nearest road and building (the 
shortest distance was in farmland) (Table 2). The 
model produced using stepwise logistic regression 
analysis was based on 35 nest sites selected at 
random from the farmland breeding habitat and 35 
other sues where the species had not been present 
Oeal absence) within the same studied farmland 
Only five variables had a substantial effect on 
probability of the occurrence of the species in 
farmland (Table 3). These were: edge density 
open space, uncultivated land, and shrubs forests’ 
and grassland. The first three had a positive effect 
on the probability of the occurrence of Red- 
backed Shrikes, whereas the other two had a 
marginally negative effect. 
DISCUSSION 
Red-backed Shrikes seem to be much more cc 
mon ,n low-altitude farmland than in mount 
environments in the Marche Region of central It 
(FoKon, ,007. Morelli and Pandojfi 2009. Mon 
-01 la), as is also the case in other parts oft 
country. However, the results of my study indie- 
•bat Red-backed Shrikes can us/very7fte 
environments and select, if available, high meadow 
areas for nesting. 
1 he two populations studied were in very dif¬ 
ferent altitude zones with farmland birds breeding 
at a mean elevation of 350 m. while the mea¬ 
dow- population was at an elevation >800mask 
However, population densities of shrikes were 
similar in both areas, as were numbers of fledged 
young per pair. 
1 hese results highlight the ecological plasticity 
of Red-backed Shrikes in selecting breeding 
habitats, as this species can use different types 
ol landscape in central Italy for breeding. This 
suggests these two habitat types meet the eco¬ 
logical requirements of Red-backed Shrikes. 
Farmlands where Red-backed Shrikes were 
detected were characterized by higher landscape 
heterogeneity (edge-density values, roads and 
buildings in the vicinity), reducing risks from 
predators or providing natural hunting perches 
(Lefranc 1993, Yosef and Grubb 1994, Lefranc 
and Worfolk 1997, Bechet et al. 1998, Tryjanow- 
ski et al. 2000. Roos 2002. Roos and Part 20041 
Variables that appear to have made a major con¬ 
tribution to the suitability of farmland as breeding 
TABLE 3. Logistic regression of presence and absence 
ol breeding Red-backed Shrikes in farmlands (only 
significantly associated 
variables are 
shown). 
Variable 
B 
SE 
P 
Edge density 
0.05 
0.012 
<0.001 
Open space 
45.12 
17.67 
0.011 
Uncultivated and shrub 
1.13 
0.525 
<0.001 
Forest 
-1.12 
0.436 
0.012 
Grassland 
-21.42 
9.21 
0.05 
