Zhou et al • NEST-SITE SELECTION AND NESTING SUCCESS OF THRUSHES 499 
TABLE 5. Model selection for variation of daily survival rate of Grey-backed Thrushes incorporating nest-location and 
nest-patch characteristics in Dagang Forestry Farm. China, in 2008. Models were ranked according to AIC c . -21og(/) 
referred to -21og-Iikelihood. All models used n = 64 nests. _ 
ID 
Predictors. 
-2log(/) 
K 
AAIC, U 
w. 
1 
Distance from nest to main stem + horizontal exposure 
+ density of shrubs 
201.24 
4 
0.16 
0.186 
2 
Null model (onlv intercept) 
209.95 
1 
2.83 
0.049 
3 
Distance from nest to main stem + horizontal exposure 
203.10 
3 
0.00 
0.202 
4 
Density of shrubs 
205.30 
2 
0.19 
0.184 
5 
Distance from nest to main stem + density ot shrubs 
203.65 
3 
0.55 
0.154 
6 
Horizontal exposure + density of shrubs 
203.65 
3 
1.03 
0.120 
7 
Distance from nest to main stem 
206.43 
2 
1.32 
0.105 
* The lowest AIC, value for this analysis was 209.13. 
Grey-backed Thrushes preferred species of 
shrub such as Manchurian Schneider buckthorn 
and Chinese hawthorn, whose umbrella-like 
crowns increased coverage over nests and aided 
in concealment. High density of small trees may 
have been important for shading and concealing 
nests, preventing potential visually oriented aerial 
predators, such as Eurasian Jay (Garrutus glan- 
durius), owls, and woodpeckers from discovering 
the nests (Weidinger 2(X)9). 
Eiserer (1980) found that American Robins ( T. 
migratorius) preferred to forage in short grass 
because of a lower predator attack rate, greater 
ease of locomotion, and hunting efficiency. Tall 
grass may make foraging difficult when Grey- 
backed Thrushes forage in the leaf litter layer 
around the nest. Thus, they tended to nest in areas 
with less short ground cover although shorter 
grass may increase their exposure to predators. 
They preferred areas with shorter grass, and chose 
areas with dense shrubs to compensate for the 
high exposure from shorter ground cover. High- 
density shrubs might inhibit growth of herbs on 
the ground based on their negative correlations 
(Table 2). We believe Grey-backed Thrushes 
probably make trade-offs between height of 
ground cover and density of shrubs. 
Successful Nests Versus Depredated Nests .— 
Nearly 60% of nests of Grey-backed Thrushes 
were depredated and nest predation is a major 
cause of nest failure for forest passerines (Ricklefs 
1969). Successful nests were constructed rela¬ 
tively close to the main stems of trees and shrubs, 
and also tended to be more concealed horizon¬ 
tally. 
We agree with other researchers that nesting 
close to the main stem provided strong support to 
the nest structure (Delannoy and Tossas 2002, 
Sargent et al. 2003). We speculate nests near the 
main stem may also decrease the upper exposure 
to predators based on preference of thrushes for 
nest trees with umbrella-like crowns. 
Predation rates were lower at nests with more 
dense foliage around the nest and less exposure in 
29 of 36 studies, including both grassland/marsh 
and shrub/woodland habitats and across a diversity 
of species (Marlin 1992). The impacts of horizontal 
exposure were embodied in the best model of DSR 
for Grey-backed Thrushes that incorporated habitat 
characteristics. We believe high visual exposure on 
a single side (upper, ground-level, or horizontal) 
resulted in high exposure and predation risk to the 
nest, regardless of the amount of concealment 
offered on the remaining sides. 
We preliminarily identified the seasonal pattern 
of nesting success of Grey-backed Thrushes, i.e., 
nest attempts late in the breeding season are more 
likely to be depredated. High levels of predation 
occurred during the incubation and brooding 
periods for Grey-backed Thrushes, which differs 
from Wood Thrushes. Anders et al. (1997) 
calculated DSR of Wood Thrushes in southern 
Missouri in 1994-1995. and found the DSR of the 
incubation period was highest among the tour 
nestling periods. However, both Wood and Grey- 
backed thrushes had the lowest DSR in the brood¬ 
ing period. Begging calls and increasing visits by 
adults possibly exposed the nests to predators 
(Briskie et al. 1999). The specific factors 
influencing DSR dependent on year, breeding 
season, and nest periods needs further studies. 
Gates and Gvsel (1978) suggested forest edge is 
an 'ecological trap’ that concentrates nests, and 
may increase nest predation. Flaspohler et al. 
(2001) reported Mayfield nest success was greater 
in the interior of large, closed-canopy hardwood 
