622 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol 124, No. 3. Sep,ember 2012 
80 
75 
g 70 
& 65 
co 
g 60 
x> 
55 
o 
ot 50 
-o 45 
c 
=> 40 
35 
30 
Used 
Mature Mature mix. Forested Sapling Shrub 
deciduous conifer wetland clearcut clearcut 
wetlands santa d'ZnT, 8 ?'^ *"***"»** *>•«■"<* ''»«<• malure mixed-decid„ou S -conifcr tom toml 
Z p 'dencJ l em r S S W ™ nUttd “»• '«*»"> used bv Ovenbird Dedpngs after 
boarS a, TO random Z '",u T M,nncsn,a ' 'P™ -f board obscured) was measured using a profile 
mean SE aZSZci es„Z v T T ml 58 "“W* '“‘“i"" 8 ' Diamonds, boxes, and whisker nrpresem 
mean, Sb, and 95 h Cl, respectrvely. Letters represent significantly diflbnm groups at a = 0 05 
our ground-based telemetry capabilities using 
fixed-wing aircraft following standard aerie 
telemetry methods (Meeh 1983). We located biro 
from the air and communicated those locations vi 
cellular lext message (upon landing) to a groun 
crew, which continued tracking the bird on fool 
We ceased monitoring each fledgling when th 
fledgling died, the transmitter fell off, or tin 
signal was lost and could not be found for 3 day 
from the ground and after one telemetry flight 
We assumed transmitter signals lost during Thi: 
period were due to transmitter expiration because 
they were near expected battery expiration dates 
we were usually able to recover transmitters aftei 
predation events (Streby and Andersen 2011 ). 
We recorded cover type at each fledgling daily 
location and recorded the location with a hand-held 
global positioning system (GPS) unit (100 points 
averaged to improve accuracy). Ovenbird post- 
fledging habitat use and survival have been 
associated with use of areas with denser understory 
vegetation than that of their nesting locations (King 
et al. 2006), and the fledglings we monitored were 
within 2-m of the ground during nearly all 
observations. We used a profile-board method 
modified from MacAnhur and MacArthur (1961) 
to estimate understory vegetation density at tledg- 
hng locations every fourth day (based on logistical 
constraints) and at random locations within each 
used, cover type. We used a 2 X 0.25-m board 
d.v,ded into eight. 0.25 X 0.25-m squares. One 
investigator held the boani vertically (ground to 2 n, 
above ground) while a second investigator stood 
° a ™dom direction (azimuths chosen 
consecutively from an electronically produced list of 
random numbers between I and 360) and estimated 
the percent of each square obscured by vegetation. 
We turned the board 90 degrees and repealed the 
process. We used the mean of 16 estimates (8 from 
each direction) us a single estimate of understory 
vegetation density at each location. 
We measured straight-line distances between 
subsequent daily fledgling locations using global 
information system (GIS) software. We used 
aerial photographs and cover-type layers (U S. 
Forest Service, Chippewa National Forest) in GIS 
software to measure the proportional availability 
of cover types within a 5-km radius from the 
center of each study site. Space available to 
individual animals is difficult to accurately assess 
for wild populations (Acbischer et al. 1993). 
However, increasing or decreasing the radius ol 
our study areas by 1 km. and moving the center of 
each study area 1 km in each cardinal direction, 
had little effect (< 2% change for all cover types) 
on cover-type composition. This method assumes 
all locations within the study area are available to 
each individual on all days of the study. Some 
birds monitored moved -3 km within 2 days of 
independence from adult care, and all cover types 
were present <1 km from any location in our 
study area, suggesting all cover types were 
similarly accessible by all birds on all days of 
the study. 
Data Analysis .—We reclassified cover types 
delineated in U.S. Forest Service cover-type 
layers into five categories: (1) mature forest. (2> 
forested wetlands, (3) sapling-dominated clearcuts 
