756 
THE WILSON JOURNAL OF ORNITHOLOGY • Vo/. 123, No. 4. December 2011 
the number of eggs and/or nestlings. Empty nc 
were assumed to have been successful (often w 
egg eft after hatching) if the dry skin that flal 
off nestlings was present in the nest w ith minin 
disturbance, or if dependent fledglings wt 
present in the vicinity of the nest. Measuremei 
of irrigation water depth (Apr to Jun) and lieit 
from nest bottom to ground were recorded as w 
as the direction of the entrance to nests. Hum 
scent and disturbances were minimized by usi 
reeds instead of hands to touch nests, and I 
visiting each nest for < 10 min. 
We used the point quadrant (a steel wire of 
percentToT" a sharpcnecl li P> to measu 
land 0996) e ^e ‘urn T** Su,he 
vertically l 0 '; h T ' ,ng f0d «uc 
being 50 cm in length) around the nest, ane 
readings were taken at the position of each holein 
the bar. The presence (touching a grass leaf I m 
absence readings from the 16 -point quadrant 
were summed to give a score for the entire (rami 
(max = 10). The mean was obtained from 
readings of tour sides based on the assumption 
that percent cover was unchanged during the 
breeding period: care was taken to not disturb the 
vegetation when placing the point quadrant or 
taking readings. 
We used regular sampling methods to measure 
reed and grass height: the quadrant size was 
0.25 nr’ for grasses and the radius of the sample- 
circle was 1.5 m for reeds. Five points (4 corners 
and I center) were selected in a square or an 
inscribed square of the sampling circle, each with 
four grass plants or two reeds per point. Those 
