642 
THE WILSON JOURNAL OF ORNITHOLOGY . Vol 123. No. 3. September 2011 
differences in behavior among subspecies of M. asio. 
Received 14 July 2010. Accepted 1 February 2011. 
The first systematic observations of the rela¬ 
tionship between sunset and evening departure of 
a pair of Eastern Screech-Owls ( Megascops asio) 
were made by Allard (1937). He reported that, 
from 19 March to 21 May, the departures of both 
the male and female averaged 7.8 min after sunset 
on clear evenings and 7.4 min before sunset on 
cloudy and stormy evenings in the Washington. 
D. C. area. He also noted the male left earlier than 
the female and that, during the nestling period, 
both sexes left the box earlier with respect to 
sunset than during the pre-hatching period and 
suggested this was due to the food need of the 
hatchlings. Gehlbach (1994) reported that, during 
the December to February period, males in central 
Texas departed from their nest boxes 15.2 ± 
4.1 min after sunset and that a female left 3 to 
24 min after her mate. He also made light-level 
readings at the times of a male's departures from a 
roost box and a roosting tree and found the means 
ol the readings to be independent of cloud cover. 
More recently, he found that stored food in the 
nest box had a greater effect on timing of 
departure than cloud cover (F. R. Gehlbach, pers. 
comm.). Relationships between nest departures 
and sunset have also been reported for other owls, 
e.g.. Long-eared Owl (Asio otus ) (Wijnandts 
1984), Flammulated Owl (Otus Jlammeolus ) 
(Hayward 1986), Ural Owl ( Stri.x uralensis) 
(Korpimaki and Huhlala 1986), Short-cared Owl 
(Asio flammeus) (Bosakowski 1989), and Mexican 
Spotted Owl (Strix occidentalis lucida) (Delaney 
et al. 1999). 
I attracted Eastern Screech-Owls (Megascops 
asio) to a nest box in my back yard in a heavily 
wooded section of Falmouth. Massachusetts 
(41 32' N, 70°36' W) starting in 1999. I recorded 
the consistency of the timing of their evening 
departures with respect to sunset and the differ¬ 
ences, if any, between individuals, clear and 
overcast evenings, males and females, and 
between roosting and nesting individuals. I also 
examined differences among the departure times 
of birds in Texas, the Washington, D.C. area, and 
my Massachusetts data. 
METHODS 
I recorded 514 nest-box departures at dusk 
between 18 April 1999 and 21 May 2010.1 did not 
want to disturb the birds and had no way of 
marking them individually. I used the following 
criteria to distinguish between males and females: 
(1) color morphs (all but one pair were of different 
colors). (2) positions during mating, which was 
observed six limes. (3) residence outside (males) 
or inside (females) the box during incubation and 
nestling periods, and (4) pitch of their vocaliza¬ 
tions. females have distinctly higher pitch (Miller 
1934, van der Weyden 1975, Cavanagh and 
Ritchison 1987, Gehlbach 1994). Unpaired indi¬ 
viduals could be distinguished by (1) color 
morphs (I had gray, brown [lighter and darker 
shades] and rufous individuals in residence, and 
(2) behavioral differences. In confrontations with 
gray squirrels ( Sciurus carolinensis) at the 
entrance to the box, some roosting (not nesting) 
individuals fled the box, while others stood their 
ground and chased the squirrels away. These 
behavioral differences were matched by differ¬ 
ences in color morphs in several cases and are a 
valid criterion, even for individuals of the same 
color. 1 conclude, based on these criteria, that of 
the 26 birds I observed separately at different 
times over the 12 years of the study. 1 had a 
minimum of 12 different individuals in residence 
and, except for 32 instances, was able to 
distinguish the gender of birds departing the nest 
box. The 32 exceptions were of individuals that 
roosted in the box for brief periods, did not have a 
mate, and did not vocalize. 
RESULTS 
The first resident owl in some years, usually a 
male, failed to attract a mate or abandoned the 
box. Pairs were resident in 1999, 2000. 2001. 
2002, and 2010. Two nestlings fledged in 2000. 
and four each in 2002 and 2010. The hatchlings 
fell prey to a common raccoon (Procxon tutor) in 
1999. and in 2001 the eggs must have been 
infertile, as the female prolonged incubation tor 
41 days. 
The times of 514 nest-box departures (Fig. I) 
cluster near local sunset as given in a web site ot 
the U.S. Naval Observatory in Washington. D.C. 
The data cover 7.5 calendar months, a few days 
short of the full amplitude of the annual daylight 
cycle. The sunset curves for all years, including 
leap years, are identical. No owls were resident in 
the nest box between 8 June and 22 October of 
any year. Both adult and young owls roosted 
outside the box after fledging and no adults ever 
returned to the box until late October at the 
