Gao et al • MARSH GRASSBIRD AND REEDBED MANAGEMENT 
759 
TABLE 4. Characteristics measured (two-independent samples test of 11 variables from 2006, 2007, and 2009) in 
breeding areas of Marsh Grassbirds, Yalu River Estuary, China. 
Variables 
Mean i SD (n = 53) 
Munn-Whitney U 
Wilcoxon W 
z 
P (2-tailed) 
Cover 
76.38 ± 11.93 
146 
237 
-2.363 
0.018 
Water depth 
3.51 ± 2.55 
7 
98 
-5.243 
<0.001 
Nest height 
17.97 ± 8.91 
116 
207 
-2.98 
0.003 
Number of live grasses 
374.83 ± 117.67 
133 
953 
-2.626 
0.009 
Height of live grasses 
99.87 ± 10.86 
48 
139 
-4.396 
<0.001 
Number of dry grasses 
148.83 ± 64.97 
15.5 
106.5 
-5.055 
<0.001 
Height of dry grasses 
56.02 ± 13.03 
2 
93 
-5.396 
<0.001 
Number of live reeds 
69.70 ± 93.56 
103 
923 
-3.247 
0.001 
Height of live reeds 
158.68 ± 18.36 
71 
891 
-3.935 
<0.001 
Number of dry reeds 
36.13 ± 46.66 
168.5 
988.5 
-1.893 
0.058 
Height of dry reeds 
138.94 ± 21.01 
113 
933 
-3.052 
0.002 
Grassbirds prefer drier ground over deeper water 
levels. 
Predation of nests also influenced nest success 
as shown by Cain et al. (2006) for other 
passerines. Claws and other remains were found 
in failed nests (as also reported by Du et al. 1959), 
and eggs were regularly lost in five nets; nest 
conditions remained the same after the eggs were 
predated. We believe the harvest mouse ( Micro - 
mys minutus pallets) was the predator. Harvest 
mice are classified as a species of ‘least concern’ 
in China, and their nests typically are constructed 
where the vertical structure of the vegetation is 
significantly denser than average (Bence et al. 
2003). 
The AIC c weights (w,) did not converge after 
the third variable entered the model due to 
complete separation of data points but cover, 
water depth, and number of dry reeds influenced 
nest height. Marsh Grassbirds avoided threats 
from Northern Harriers by using areas with 
efficient cover. Flooding may cause Marsh 
Grassbirds to raise their nests after irrigation. 
Marsh Grassbirds preferred dry reeds for perches. 
and live reeds grow in greater quantity in drier 
reed stands (Spearman’s r = 0.482, P < 0.01). 
Total reed cover at these sites is more dense, 
taller, and nest height becomes higher. 
Live and total grasses at nests with entrances 
were significantly (P < 0.01) less than those with 
no entrance. Marsh Grassbirds built their nests 
where cover was considered to conceal their nests 
better from avian predators. 
CONSERVATION IMPLICATIONS 
Dry-grass patches should be preserved for 
Marsh Grassbirds (De Schauensee 1984), reeds 
and grasses should be harvested every 2 years 
(Valkama et al. 2008), and burning should be 
prohibited after reed harvest with some rice fields 
being allowed to revert to reed-grass fields. Marsh 
Grassbirds adjusted to deeper irrigation water by 
raising nest height when possible. However, lower 
nest height (the lowest was 11.69 cm in 2009), 
may cause nests to be flooded in the rainy season 
(Jul and Aug). Thus, shallow irrigation with water 
depth <8 cm (= the lowest nest height), should 
be implemented before 16 May (nest initiation). 
TABLE 5. Changes in number of Marsh Grassbirds and reedbed management over 4 years, Yalu River Estuary, China. 
2006 
2007 
2008 
2009 
Number of nests 
5 
35 
0 
13 
Successful nests 
2 
7 
0 
3 
Male individuals 
93 
70 
19 
61 
"'ater depth (mean, cm) 
6.4 
4.4 
31.5 
0 
Irrigation time 
18 May 
29 Apr 
8 Apr 
5 Jun 
All grass cutting in winter 
no 
no 
yes 
no 
Turning up the soil in spring 
Dominant live grass/sedge 
no 
C. epigejos 
no 
C. epigejos 
yes 
no 
J. gracillimus, S. planicuhnis 
