186 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 1, March 2011 
have led to increased attention of urban roosts 
(Ward et al. 2006). 
American Crows (Corvus brachyrhynchos) 
roost communally during the winter, often 
forming flocks of >1,000 birds (Emlen 1938, 
Stouffer and Caccamise 1991). Winter roost 
dynamics of American Crows in North America 
have long been the focus of study (Caccamise et al 
1997, Preston 2005). Despite efforts to dissuade 
urban roosting behavior of crows (Gorenzel and 
Salmon 1992, Avery et al. 2008), studies through¬ 
out the range suggest urban winter roosts and 
habitat associations are becoming more common 
(Gorenzel et al. 2000, Marzluff et al. 2001). 
However, why crows may exhibit strong winter 
roost site fidelity and the mechanisms of winter 
roost selection remain elusive (Fiedler 1969, 
Gorenzel and Salmon 1995). Early studies suggest 
congregation of roosts in Ohio may be impacted 
by temperature and wind (Haase 1963). Crow 
roosts have consisted of both deciduous and 
coniferous trees as well as artificial structures 
including buildings and bridges (Stouffer and 
Caccamise 1991, Gorenzel and Salmon 1995, 
Gorenzel et al. 2000). Examination of winter 
roosts of American Crows has occurred for over a 
century (Edwards 1888) but, to our knowledge, 
ours is the first published observation of urban 
ground roosting behavior by the American Crow. 
OBSERVATIONS 
We observed American Crows roosting on the 
ground in Springfield, Ohio on the evening of 12 
January 2010 during a study examining patch- 
occupancy in urban winter crow roosts. Spring- 
field is a city of ~60,000 people in the Miami 
Valley region of southcentral Ohio. Average 
winter temperatures range from -7 to 1° C. The 
crows observed were part of a larger roost 
estimated at 2,500 individuals discovered during 
an ongoing study of urban winter roost selection. 
At 2015 hrs EST, 2.45 hrs after sunset, we noticed 
a roost on the roof of a public library building and 
in nearby trees. Temperatures reached a low of 
-12 C before midnight with winds up to 79 km/ 
hr reported by the National Weather Service in 
Dayton. Road surface temperatures in areas 
susceptible to hazardous road conditions ranged 
from -8 to —16° C overnight according to DOT 
records (Clark County, Ohio Department of 
Transportation, pers. comm.). 
We noticed roosting birds on neighboring 
structures and adjacent trees. The site was 
illuminated by numerous street lights and was 
bordered by industrial and office buildings. Six 
streets and an active railroad track bisected the 
site and crows were observed roosting in parking 
lots near these roost structures. Approximately 
15 cm snow was on the ground but the parking lot 
had been cleared of snow. Drifting snow had 
covered about 50% of the lot. A combined total of 
250-300 crows was ground roosting in three 
separate parking lots. Each roost had small trees 
(DBH <25 cm) bordering the ground roost site 
with additional birds roosting in these trees. 
Most birds in the roost showed little response to 
human activity (i.e., cars driving by, people 
walking nearby, and a train passing within 25 m 
of the roost). Ground-roosting birds remained 
stationary for the duration of our observation 
(~45 min) with most birds exposing as much 
surface area as possible to substrate in what may 
be described as a resting or roosting position. The 
presence of crows was not confirmed the 
following morning, but the stationary nature of 
the crows, their lack of response to human 
activity, the length of the observation, and the 
time frame with respect to sunset, all suggest this 
was not an observation of a staging event. 
DISCUSSION 
Studies suggest American Crows may be 
deriving substantial benefit from close association 
with humans. American Crow populations in 
urban areas achieve higher densities and experi¬ 
ence more rapid population growth yet costs in 
terms of survivorship and reproduction may be 
insufficient to explain this growth (Marzluff et al. 
2001). Despite winter roosts in northern latitudes, 
crows do not possess major physiological modi¬ 
fications tor existence in cold winter climates 
other than large body size and possible benefits 
afforded through nasal feathers (Wunder and 
Trebella 1976). The available evidence suggests 
thermoregulation may be a driving factor in 
American Crow foraging decisions during tem¬ 
peratures below an estimated lower critical 
temperature (Kilpatrick 2003). Our observations 
occurred well below estimated lower critical 
temperatures based on metabolic studies of other 
corvids, and could provide insight into the 
potential benefits of urban associations at northern 
latitudes. Possible thermoregulation benefits from 
potential wind shelter and heat conductance from 
the pavement may explain why crows roosted on 
the ground when trees were available. The road 
