94 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 1, March 2011 
first nests, whereas those found with < three eggs, 
or nests found late in the season were categorized 
as re-nesting attempts (Yang et al. 2009; KZ, 
unpubl. data). We recorded the following variables 
for each nest: breeding group composition (breed¬ 
ing pairs and helpers), nesting attempt, nest type 
(either ground or tree nest), nest characteristics 
(structure and material), nest contents, nest tree 
(scrub) species, and habitat characteristics (habitat 
type, canopy cover, and scrub cover). Canopy cover 
and scrub cover were measured within a 10-m 
diameter plot around each nest. 
Statistical Analyses .—Analyses were conducted 
using SPSS Version 13.0 (SPSS Institute 2004). 
We included only the first nesting attempt by each 
female in the subsequent analyses to avoid 
pseudo-replication, and analyzed the few re¬ 
nesting attempts separately. Chi-square test of 
independence was used to compare the proportion 
of nest types between nesting attempts. Data 
corresponding to re-nesting attempts were exclud¬ 
ed from the analysis. We also used Chi-square 
Goodness of Fit tests to examine tree nest site 
preferences among habitat types. Mann-Whitney 
U-tests were used to detect differences between 
ground and tree nests. Statistical tests were two- 
tailed and P < 0.05 was interpreted as being 
statistically significant. Mean ± SD values are 
presented. We combined all egg measurement 
data from ground and tree nests as they did not 
differ (Z= -1.824, P = 0.068; Z = -1.178, P = 
0.24, for egg length and width, respectively). 
RESULTS 
Nest Surveys. Sixty-eight breeding groups 
were detected over the four consecutive breeding 
seasons (18 in 2006, 16 in 2007, 17 in 2008, and 17 
in 2009). Szechenyi’s Monal-Partridge in the 
Pamuling Mountains treeline habitats constructed 
nests either on the ground or in trees. Fifteen 
ground nests (all active; 68% of total active nests) 
and 63 tree nests (56 used and 7 active; 32% of 
total active nests) were found during the study 
period. Thirteen ground nests were first nests and 
two were re-nesting attempts. Three active tree 
nests were first nests while the remaining four were 
classified as re-nesting attempts. No third nesting 
attempts were identified. Tree nests were more 
common in re-nesting attempts than in first nests 
(50 and 17%, respectively), but the difference was 
not significant (Fisher’s exact test, P = 0.40). 
Nest Characteristics and Nest-site Selection .— 
Ground nests (n = 15) were scrapes in the soil and 
lined with leaves, sticks, and bark, usually 
positioned at the base of a tree or scrub (Fig. 1A). 
They occurred in all habitats except Sichuan 
kobresia meadow, and had lower canopy vegeta¬ 
tion cover (13 ± 14%, range = 0-40%, n = 13) 
when compared with tree nest sites (30 ± 17%, 
range = 20-50%, n = 3); this difference was not 
significant (Z = — 1.835, P = 0.066). Tree nests (// 
= 63) were cup shaped (Fig. IB) and constructed 
of moss, lichen, and feathers although some nests 
also contained fragments of man-made cloth. They 
were positioned 1.9-12.0 m above ground level 
either at the base of the tree branches and adjacent 
to the main trunk in coniferous forest and 
rhododendron scrub, or at the top of trees in oak 
forest. Tree species used included Abies squamata 
(61%), Quercus aquifolioides (31%), and Larix 
potaninii (8%). No tree nests were found within 
either the scrub holly leaf-like oak habitat or 
Sichuan kobresia meadow; they were distributed 
proportionally among the three remaining habitat 
types (x 2 = 0.194, df = 2, P = 0.91, n = 56). 
Eggs and Nesting Success .—Eggs (n = 46) 
were pinkish brown with fine, well-dispersed 
burgundy-colored spots, and measured 53.8 ± 
2.6 mm (range = 47-59 mm) X 37.4 ± 1.9 mm 
(range = 33-42 mm). Only 54% (n = 7) of 
ground nests survived until hatching, compared 
with 33% (n = 1) of tree nests. The principal 
cause of ground nest failure was predation which 
accounted tor 50% of all failed nests, while severe 
weather conditions (33%), and human disturbance 
(17%) accounted for the remaining nest failures. 
Hatching success was 97% (n = 31 eggs from 8 
nests). Twelve chicks (n = 21 from 7 ground 
nests) survived to fledge, while three chicks from 
the one tree nest failed to survive. 
Changes in Nest Location and Nest Type.—We 
documented re-nesting attempts on six occasions 
during the 4-year period. There were two 
occasions where only re-nests were detected: 
one was a ground nest in 2008, and the other 
was a tree nest in 2009. We documented the nest 
location and nest type for both nesting attempts on 
the other four occasions. On two occasions (once 
each in 2008 and 2009), females re-nested in the 
same nest types. One female reused the tree nest 
after the chicks from first nest failed to survive, 
the other female changed ground nest location in a 
i e-nesting attempt following nest predation. 
Females switched nest locations and nest types 
from ground to tree nests on the other two 
occasions (once each in 2007 and 2009) following 
