498 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124, No. 3, September 2012 
activity varied by vegetative strata, within core 
areas of territories, in association with canopy 
gaps, or among tree species. 
METHODS 
Study Area. —Lewis Wetzel Wildlife Manage¬ 
ment Area (LW) in Wetzel County near Jackson- 
burg. West Virginia. USA. is 5.418 ha in size with 
elevations of 224-476 in. Two ridges. Hart Ridge 
(—49 ha) and Snake Ridge (—45 ha), were 
surveyed for Cerulean Warblers during the 2005 
breeding season. Both ridges predominantly 
consisted of mature, —70-80 year-old, second- 
growth forest of mixed mesophytic and Appala¬ 
chian oak forest types. However, the two ridges 
differed in vegetative structure and tree species 
composition, partly as a result of past agricultural 
activity on Hail Ridge. 
Hart Ridge was mostly mesie and common tree 
species were sugar maple (Acer saccharin,,). 
northern red oak ( Quercus ruhra), tulip poplar 
(Liriodendron tuUpifera ), white ash (Fraxinns 
americana), hickory ( Carya spp.), black cherry 
(P/7/m/.v serotina ), black walnut (Juglans nigra), 
black locust (Rohinia pseudoacacia), and red 
maple (A. rubnmi). Canopy openings included a 
partially open-canopy trail that followed the 
ndgeline. three maintained wildlife openings 
0. 7-0.32 ha in size on broad sections of the 
ridge, and treefall gaps. 
Snake Ridge had a xerie ridge-top with rocky 
soil and xenc or mesic side-slopes. Co-dominant 
ree species on the ridge-top were chestnut oak 
(Q ' pn,u,s) and scaf I eI oak ( Q . coccinea). Other 
common species were several hickory species 
northern red oak. black oak ((7. ve lutina), white 
oak (Q. alba), white ash. basswood (Tilia 
americana ) tulip poplar, sugar maple, red maple 
Snl CU pm T magnolia Magnolia acuminata). 
bnake R,d ge had no recent anthropogenic dist¬ 
ances at the time of the study but had treefall 
Saps. 
systematically searched 
2005 d„H ° r . :5 " ,8 "’ g males ,n June 
8 n '° m !, ng <060(M 130 We waited 
- min before record,ng locations of males to reduce 
ob^" d ; h ? nspicu r s ioraiions ' we 
survey burs, '""T durin 8 » 30-min 
allows sufficient time for a male to traverse its 
entire territory if it chose to do so, and provides 
biologically independent locations (Barg et ul. 
2005). We used a survey in analyses if a male had 
at least eight unique flagged locations or 15 min of 
observation during the 30-min survey period. 
We recorded activity at the first second of each 
minute the observer had the bird in view. One 
observer counted down the seconds with a stop 
watch while a second observer watched the bird 
through binoculars and identified the activity ai 
the first second of each minute. No activity was 
recorded it the observer did not have the bird in 
view when the minute turned. 
We classed activity into 11 categories (Table It. 
Foraging included all food-related activities such 
as gleaning insects, hopping along a branch 
actively searching for prey, and one observation 
°1 u male beating a caterpillar against a twig. 
Hopping was recorded as a separate activity when 
the bird had its head up while moving along the 
branch and appeared to be using this behavior as a 
means ol locomotion instead of a foraging strategy. 
Wc defined perching as a bird observed sitting still 
and not singing, whereas a bird lhai was perched 
and singing was recorded as singing. Dancing 
was a unique behavior where the male Cerulean 
Warbler elevated his wings slightly without 
opening them and moved them buck and forth in 
a wiggling motion that may function as a courtship 
display (R. P. Detimers and M. R. Lutmerding. 
pers. comm.). Other behaviors, witnessed <3 times 
each, were fanning tail feathers, fluffing out 
feathers, and fluttering wings while perched. These 
behaviors were included in analyses as singing if 
observed while the bird was singing and otherwise 
included as perching. 
We recorded tree species and vegetative strata 
tor each activity when possible. We recorded tree 
•species and vegetative strata hut not activity type 
when the male was obscured by vegetation but its 
location was known. W'e categorized vegetative 
strata as understory, midstory, lower-canopy, mid- 
canopy, und upper-canopy. Each strata represent¬ 
ed 20% increments that were visually estimated 
beginning at the forest floor through the upper 
canopy layer. Field personnel practiced estimating 
•strata before sampling began to ensure consisten¬ 
cy among personnel. 
Mapping .—We revisited flagged male locations 
in late June through July to obtain coordinates of 
each flagged location using global positioning 
system (GPS) units. Coordinates were corrected 
