348 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 2. June 2011 
METHODS 
Study Area, —The three study sites were along 
the northern coast of the Gulf of Mexico: (I) 
Grand Chenier. Cameron Parish. Louisiana; (2) 
Hackberry Ridge. Cameron Parish. Louisiana; and 
(3) Smith Point, Chambers County, Texas. Grand 
Chenier was a more mature and diverse coastal 
forest than the other two sites, whereas Hackberry 
Ridge had a relatively low canopy and consisted 
predominantly of sugarberry (Celtis laevigata) 
trees. The Smith Point site had many live oaks 
(Quercus virginiana) with thick understory veg¬ 
etation (Barrow et al. 2000). 
Each study site consisted of a “disturbed” plot 
with reduced understory, primarily due to cuttle 
grazing, and a “control” plot where the under¬ 
story had not been affected. All control plots had 
significantly more suheanopy and understory veg¬ 
etation than disturbed plots (Barrow et al. 2000: 
figure 5). Each plot was to be a 100 X 300-m 
rectangle, but plots of this size could not be 
obtained at all sites because of limited availability. 
Disturbed and control plots were adjacent at Grand 
Chenier, but separated by I km al Hackberry Ridge 
and by 100 m at Smith Point. 
The long axis of each plot was oriented East- 
West. roughly parallel to the coastline. We 
established grids marked every 25 m with flags 
that delineated the boundaries of many small 
blocks and formed several transect lines within all 
study plots. 
FieUl Procedures.—Grand Chenier was sam¬ 
pled from 15 April to 15 May 1993. Each site was 
systematically sampled for a week from 10 March 
to 17 May 1994 to equalize effort among sites. 
Five additional days (19-21 Mar. 21-22 Apr) 
were spent at Smith Point in 1995, because this 
site had the least data collected from the previous 
year. 
Some species are known to have individuals 
that overwinter at our sites and they were included 
in the analysis. These included Yellow-rumped 
Warbler ( Dendroica coronata), Ruby-crowned 
Kinglet (Regulus calendula), and Blue-headed 
Vireo (Vireo solitarius). as well as some individ¬ 
uals of Common Yellowthroat (i Geothlypis tri- 
chas), Blue-gray Gnatcatcher (Polbptila caeru- 
lea), and White-eyed Vireo (Vireo griseus). 
Foraging behavior of birds in terrestrial habitats 
has been divided into five basic components: 
search, “attack,” “foraging site,” “food” 
and “food handling” (Remsen and Robinson 
1990). Search behavior can be further divided into 
“scanning" and “movement" (O’Brien el al. 
1990). “Scanning," the action of the head and 
eyes to spot prey, is not included in our study 
because it is difficult to quantify. We focused on 
search movements used to locate food or for 
substrates that contain food. “Search" ends once 
food or food-hiding substrates are observed ind 
attacked (Remsen and Robinson 1990i. Classifi¬ 
cation of search movements was modified from 
Remsen and Robinson (1990) as: 
'hop’- movements made only with the legs, 
‘flutter’- movements made mainly with the 
legs, but with the support of the wings, and 
‘fly’- movements made by flapping the wings. 
In our experience, a bird flies only when the 
distance between two perches is substantially 
greater than its body length. Jander (1975) divided 
flying into two types: flights within patches and 
flights between patches. A patch could be a tree or 
a group of connected trees. Flights between 
patches were used to cross gaps and are for 
traveling rather than for search. We distinguished 
between the two types of flights beginning in 
1994, and only within-patch flights were included 
in the analysis. 
Search data were recorded as we traversed the 
study plots. Attempts were made to equalize 
sampling effort in each part of the plot. We tried 
not to observe individuals of the same gender of a 
particular species at the same spot or within a 
mixed-species foraging flock to avoid collecting 
data from the same bird. Repeated sampling of 
individuals should be rare because most Nearctic- 
Ncotropic migrants are known to depart the night 
of their arrival (Gauthreaux 1971. 1972: Moore 
and Kcrlinger 1987; Kuenzi et al. 1991). 
We used “focal sampling” and ‘‘continues 
recording" following Martin and Bateson (199-’)• 
We quietly followed each bird encountered and 
entered observations info a tape recorder, 
recorded species, time of day, presence ot migrant 
fallout (ca. 10-fold increase or greater from 
previous day), every search movement observed, 
and the distances the bird moved (in cm: we u>eJ 
the length ot the bird to estimate these distances'. 
Only one distance estimate was made for f 30 * 1 
type ot search movement for an individual- ^ L 
stopped recording once the bird went beyond our 
view, and sequence and duration information on 
search movements was later transcribed fro' 11 
the tape. 
