516 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124. No. 3. September 2012 
TABLE 1. Analysis (Generalized Linear Mixed Model) of factors possibly affecting nest 
Shrikes (n = 30). 
predation of Bull-headed 
Explanatory terms 
df 
P 
Estimate ± SE 
Nest height 
0.34 
1 
0.56 
Number of branches 
11.36 
I 
0.0006 
0.38 ± 0.13 * 
Number of thorny branches 
0.10 
1 
0.75 
Significant values arc highlighted in 
bold. Research years (0.00 
— 0.00) were included as random terms. 
birds or reptiles were the main predators based on 
physical evidence at depredated nests. I identified 
Carrion Crows as egg predators based on two 
video recordings and the main predators in this 
study area are most likely birds. Predators that use 
visual cues to detect nests do not easily discover 
well-concealed nests (Lima 2009). Thus, a high 
density of branches around a nest may ensure that 
nests ol Bull-headed Shrikes are not easily 
discovered and depredated. 
Nest concealment affects the probability of nest 
predation for many passerines (Martin and Roper 
1988, Flaspohler et al. 2000). Nest success 
increased with nest concealment of Red-backed 
Shrikes {Lanins collurio) (Muller et al. 2005). 
Nest concealment on the horizontal and/or vertical 
planes has been used as an index for nest cover in 
previous studies (Holway 1991. Burhans and 
Thompson 2001). High-density branches support¬ 
ing a nest provided underside and lateral cover for 
nests in mv study. This branch cover likely made 
it difficult for ground predators to find nests 
which suggest branch cover may reduce nest 
predation by mammals. My results suggest that 
concealment from below the nest should receive 
more attention as an important nest-site charac¬ 
teristic. 
Previous studies have suggested that thorns 
could reduce nest depredation (Lack and Lac 
1958, Col lias and Collias 1964). Loggerhea 
Shrikes (Lanins hulovicianns) in South Carol in; 
USA nested more frequently in red ceda 
(Junipems virginiana) trees, which may hav 
more needles than any other species of tree 
shrikes in north-central South Carolina that nestei 
in red cedar trees fledged one additional youn; 
per nest than shrikes nesting in other types of tree 
(Gawlik and Bildstein 1990). Two recent fieh 
studies have suggested that presence of thorns die 
not correlate with the probability of nest predatiot 
(Barnard and Markus 1990. Mezquida ant 
Maione 2002). Similarly. I found no effect ol 
thorny branches on nest depredation. Mezquida 
and Marone (2002) suggested that thorns do not 
function as anti-predator devices for some bird 
species, because their main predators also m.i> 
nest in thorny plants. Thus, the effect of thorns, it 
any, may depend on the type of nest predators. 
Nest height did not correlate with nest preda¬ 
tion in this population of Bull-headed Shrikes A 
study of Bull-headed Shrikes in Hokkaido by 
Takagi and Abe (1996) reported that effect of nest 
height on nest predation varied depending upon 
when a female laid the first egg. Nest height did 
not affect nest predation of shrikes which started 
nesting earlier in the breeding season, but had a 
positive effect for later nesting birds. This 
difference between the two sites suggests that 
effect of nest-site characteristics on nest predation 
varies across different environments. Variable 
effects may be due to differences in the types of 
predators present, which may vary in their 
susceptibility to different nest defenses, or 
because shrub species used as nest sites may 
differ geographically as for Loggerhead Shrikes in 
North America (Gawlik and Bildstein 1990). 
Forty-six percent of nest sites selected by Bull- 
headed Shrikes in Hokkaido were in dwarf 
bamboo (Sasa kurilensis ), and 25% were in 
deciduous shrubs. However, deciduous shrubs 
accounted for 60% of nest sites in Karuizawa 
(S. Endo. unpubl. data). It is likely that accessi¬ 
bility for predators differs between dwarf bamboo 
and deciduous shrubs because dwarf bamboo have 
smooth trunks with no strong branches, whereas 
deciduous shrubs have rough trunks and many 
branches. The type of tree in which a bird nests 
may affect its reproductive success, as has been 
found for the American Robin (Tardus mi grown - 
its) (Schmidt and Whelan 1999). However, the 
probability of nest success did not differ among 
nest shrub types used by Loggerhead Shrikes in 
southwest Idaho, USA (Woods and Cade 1996i. 
Bull-headed Shrikes that attached their nest' to 
more branches had a higher probability of breeding 
success in my study. However, despite the 
