Beason et al. • BLACK SWIFT MIGRATION AND WINTERING AREAS 
5 
isws 
SOUTW 7*WW 
FIG. 2. Northern Black Swifts marked in Colorado 
wintering distribution kernel density contours (red = 50%. 
green = 15%. blue = 90%) from 28 September-12 October 
2009 to 9-20 .May 2010. 
broadleaved forest regularly flooded. Areas of 
mosaic cropland/vegetation, mosaic forest shrub- 
land/grassland, closed to open shrubland or grass¬ 
land. bare areas, and water bodies represented 
<2% use. 
DISCUSSION 
We documented the timing of fall and spring 
migration, wintering area, and spring migration 
paths for three Northern Black Swifts using 
geolocators. The return dates to breeding sites 
after spring migration and dates of the initiation of 
fall migration correlate with data collected in 
other research (Hirschman et al. 2007). Wintering 
area locations were previously unknown, and only 
sporadic reports existed which did not fully 
delineate migratory paths. 
The highest kernel density estimates indicate 
the three birds wintered almost entirely within the 
State of Amazonas, Brazil. The clustering of the 
three individuals exhibited by nearest-neighbor 
analysis might be expected based on the birds 
seeking a physiologically optimal climate, pre¬ 
ferred habitat, abundant prey, or other factors 
within a certain geographic area. Amazonas is 
composed almost entirely (—98%) of lowland 
rainforest at elevations between 34 and 116 m. 
The area is sparsely populated with a density of 
2.05 inhabitants/km 2 with 78% of die population 
in cities (IBGE 2011). Amazonas has an equato¬ 
rial tropical rainforest climate with annual rainfall 
of 1.50-2.50 m and all months have a mean 
precipitation of at least 60 mm (IBGE 2011). The 
average temperature per day per year is 26.7 C 
(23.3-31.4 Cl with high humidity (Brasil Travel 
Guide 2011). 
The three birds tracked in this study represent 
only a small geographical subset of the Northern 
Black Swift population and further studies are 
needed to delineate more completely the full 
extent of the wintering distribution of this 
subspecies. The three birds wintered in the same 
general area, suggesting a high level of connec¬ 
tivity between breeding and wintering popula¬ 
tions. Stutchbury et al. (2009) found a similar 
connectivity for Wood Thrush (Hylocichla mus- 
lelina), not previously documented for migratory 
songbirds. The large w intering areas may retied a 
temporal movement noted for each bird.'The data 
indicate a trend for each bird to be in the eastern 
portion of the kernels in October with gradual 
movement west in April and May. The most likely 
explanation for this replicated non-random change 
