Dengetal. • CONSPECIFIC BROOD PARASITISM IN MANDARIN DUCKS 
481 
14 
FIG. I. Distribution of Mandarin Duck first egg dates from 40 nests. April (III) last 10 days of April: May (I) first 
10 days of May, May (II) middle 10 days of May. May (III) last 10 days of May; June (I) first 10 days of June, and June (II) 
middle 10 days of June. 
that contained a full lining of down were 
incubated, and presence of warm eggs and 
repeated flushing of a bird from the box 
confirmed this assumption. We removed nests 
that were deserted during the laying period 
(incomplete clutches) in calculating clutch size. 
We converted nest initiation date to Julian day 
before analysis. 
The independent sample Mest was performed 
to compare clutch initiation dales and nest 
characteristic variables between un-parasitized 
and parasitized nests. We used linear regression 
to examine whether clutch size was related to 
dutch initiation date. We analyzed the relation¬ 
ship between frequency of CBP and nest densities 
using Spearman correlation coefficients. SPSS for 
Windows. Version 14.0 (SPSS Science. Chicago. 
IL, USA) was used for statistical analyses. Means 
± SD are presented. 
RESULTS 
Conspecific Brood Parasitism. —Forty-two of 
410 nest-box years were occupied by Mandarin 
Ducks with an average of 7.0 ± 5.4 (n = 6 yrs) 
boxes used per year (range = 1 to 15 nests 
annually). CBP occurred commonly in Mandarin 
Ducks. Eighteen of 39 (46.1%) nests with a 
complete clutch were parasitized. Frequency of 
CBP by Mandarin Ducks was positively correlated 
with nest densities (Spearman’s correlation: r = 
0.880, n = 6, P = 0.021). We estimated that 2.5 ± 
0.7 in - 15) females laid in each parasitized nest as 
indicated by eggs appearing per day. 
Nest Chronology .—Clutch initiation dates 
ranged from 22 April to 11 June (n = 41) with 
two laying peaks (Fig. 1). Clutch initiation dates 
for un-parasitized (n = 21) and parasitized nests 
(n = 18) differed </ = 2.594, P = 0.014). CBP 
occurred more frequently during the early portion 
of the breeding season. The median First-egg date 
lor parasitized nests was 5 May (n = 18), and was 
17 May for un-parasitized nests (n = 21). Eggs of 
a clutch in individual nests were laid on 
consecutive days or at longer intervals from 1 to 
6 days. The laying period differed between un- 
parasitized (n = 21) and parasitized (n = 18) 
nests (t = -4.126, P < 0.001), and was 11.9 ± 
2.4 days and 15,3 ± 2.9 days, respectively. The 
mean incubation period for successful nests was 
32.7 ± 1.4 days (n = 23, range = 31-36 days): no 
significant difference was detected between un- 
parasitized (n ~ 16) and parasitized (n = 7) nests 
(t = 0.031, P = 0.98). 
Clutch Size. —Three nests were deserted during 
egg laying and incomplete clutches ranged from 
one to six eggs. Clutch size of complete un- 
parasitized (n = 21) and parasitized nests (n = 
18) differed (/ = -10.787. P < 0.001) and was 
9.7 ± 1.6 eggs (range = 7-12), and 18.8 ± 3.2 
eggs (range = 15-25). respectively (Table 1). The 
combined clutch size was 14.0 ± 5.1 (n = 39). 
Clutch size for un-parasitized nests decreased as 
the nesting season progressed (p = —0.730. P 12 o 
= 21.625. P < 0.001). but not for parasitized 
nests ( P = -0.255, P U7 = 1.112. P = 0.307). 
Nesting Success .—Thirty-nine of 42 nest at¬ 
tempts reached a complete clutch size during the 
breeding seasons of 2004-2009 (Table 1), and 23 
nests successfully produced young that left the 
nest box. Failure of 19 unsuccessful nests 
primarily occurred during incubation (84.2%) 
followed by the egg laying period (15.8%); no 
