10 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124. No. 1. March 2012 
of migration. The Kirtland’s Warbler, a species 
designated as federally endangered in the United 
States and Canada, has a small core breeding 
range (—71km’), and breeds almost exclusively 
in one distinct habitat type, young stands of jack 
pine (Pinas hanksiana) on sandy oulwash soils in 
the northern Lower Peninsula of Michigan (May- 
field 1960. Donner et al. 2008 ). Small numbers of 
birds also breed in similar habitat in Michigan's 
Upper Peninsula, Wisconsin (Probst et al. 2003, 
Trick et al. 2008), and southern Ontario (Richard 
2008). The total number of singing male Kirt- 
land's Warblers was estimated to be 1,750 in 2010 
(Elaine Carlson, pers. comm.). T his warbler is one 
of a few passerine species for which it is feasible 
to search breeding areas for individuals color- 
banded in wintering areas because locations of all 
singing males are mapped during an annual 
population survey (Probst et al. 2005); there arc 
a relatively small number of birds and sites to 
check for arriving color-banded individuals. 
Our objective was to document the duration of 
spring migration for individual Kirtland’s War¬ 
blers based on field observations of departure and 
arrival times of uniquely color-banded birds. To 
our knowledge, these estimates of spring migra¬ 
tion duration of Kirtland’s Warblers arc the first 
derived from observations of the same color- 
banded individuals in wintering and breeding 
areas. They also provide comparative data for 
estimates of migration rates and duration gener¬ 
ated from other methods. 
METHODS 
Teams based in The Bahamas, Michigan. 
Wisconsin, and Ontario searched the Kirtland's 
Warbler winter and breeding habitat to estimate 
duration of migration as part of a coordinated 
effort to study linkages between winter and 
summer ranges. We captured 232 Kirtland’s 
Warblers from 2002 to 2010 in mist nets and 
color-banded birds at several sites (Wunderle et 
al. 2010) within 30 km of each other on southern 
Eleuthera, The Bahamas (—25 N. 76 W; Fig. I 
Many of the banded individuals were observei 
repeatedly during a given wintering season. Wi 
visited Eleuthera sites with several color-bandei 
Kirtland’s Warblers three to 15 times/season fron 
mid-April through I May 2003-2010 to documen 
a date as close to departure as possible. Many o 
these warMcrs show winter site fidelity (Syke 
and Clench 1998. Wunderle et al. 2010), bu 
relocating birds in winter is challenging because 
some individuals move to different sites, and are 
difficult to locate and identify in the Lhick shrubby 
habitats. 
The Kirtland’s Warbler breeding habitat was 
searched for color-banded birds from 2003 to 
2010. This species often shows breeding site 
fidelity (Walkinshuw 1983, Mayfield 1992, Bo- 
cetti 1994; SMR, unpubl. data), like many 
territorial migrant landbirds. although a few birds 
disperse to new sites between years (Walkinshuw 
1983; DNE and KRH, unpubl. data). We could 
often estimate arrival dates of individuals return¬ 
ing to their territories the following spring by 
checking sites occupied by banded warblers in 
previous years. Our searches focused on a subset 
of banded birds that we were able to relocate 
between mid-to-late April in The Bahamas, for 
which we had georeferenced their breeding 
territories in the Lower Peninsula of Michigan in 
previous years. 
We monitored territories for a minimum of 
30 min/day every I to 3 days from early May 
through 30 May or until the bird was found. 
Observers walked through an area encompassing 
roughly a 200-400 m radius around the georefer¬ 
enced site during these visits, searching for color- 
banded birds. Data for birds with known territo¬ 
ries were supplemented with records of individ¬ 
uals for w hich we had recorded last observation 
dates from Eleuthera in mid-to-late April and 
then opportunistically observed the same bird 
in Michigan while systematically searching for 
returning territory holders. Field work was ini¬ 
tiated each spring, soon after the first confirmed 
arrival dale of a Kirtland's Warbler in the 
breeding areas. 
We calculated each individual's duration of 
migration, and average distance covered per day 
from these departure and arrival date estimates. 
Duration was estimated by calculating the number 
of nights between the last observation date in The 
Bahamas (assuming the bird left on the night of 
the last date it was observed) and the night 
previous to the first observation in Michigan. The 
interval between these dates represents the 
maximum duration of migration. Actual times of 
migration could be less than the durations 
reported here because an individual may not have 
departed immediately following the last Bahamas 
observation. In addition, individuals may not have 
been observed on their first day of arrival in 
Michigan, especially males that did not sing 
within our sampling period or females. Average 
