SHORT COMMUNICATIONS 
609 
TABLE 1. Hypotheses and predictions commonly used lo explain why an 
adult would perform infanticide. 
Hjpothcsis 
Prediction 
Citation 
Sexual selection 
Increases mating opportunity and 
allows one to raise own offspring 
drivers 1972. Hrdy 1979) 
Resource competition 
Increases one's own offspring survival 
in time of limited resources 
(Hrdy 1979. Sherman 1981. 
Ebenspcrger 1998) 
Exploitation 
Increases fitness directly via exploiting the 
infant (i.e.. cannibalism or play mothering) 
(Lancaster 1971. Fox 1975. 
Hrdy 1979. Quiatt 1979) 
Parental manipulation 
Increases survival of healthy offspring 
by killing unhealthy offspring 
(Low 1978) 
We observed 12 and 13 day-old nestlings on IK 
and 19 July 2011 being pecked and pushed out ot a 
nest (# 106) by the female at a nearby nest (# 37). 
Nest # 106 had four nestlings and nest # 37 had 
recently (within 3 days) Hedged three young; the 
two nests were ~2 in apart. The barn manager saw 
three nestlings full from nest # 106 onto the 
concrete below on the evening of IS July 2011. 
Upon seeing the first nestling fall, she watched as 
an adult pushed out the other two. All three 
survived the fall, hut two died shortly after. The 
third was kept overnight and taken to an animal 
rehabilitation center the next morning. We watched 
as the female from nest # 37 pecked and pushed out 
the fourth nestling at —0900 hrs on 19 July 2011. It 
appeared the nestling was dead prior to the fall, 
likely due to a large puncture wound on its body. 
We expected the female from nest U 37 to mate 
with the male from nest # 106, or alternatively, 
that the pair from nest # 37 would take over nest 
# 106 for their second brood. 1 lowever. no activity 
was seen for I week; 8 days after the nestlings 
were killed, the pair from nest # 37 initiated a 
clutch in a newly built nest within 30 cm of their 
original nest, and Hedged four young. A new pair 
laid eggs in nest # 106. but later abandoned the 
clutch; the original pair from nest # 106 was not 
seen for the remainder of the breeding season. 
DISCUSSION 
Infanticide is typically a behavior that increases 
an individual's fitness, either directly through 
increased access to mates, or indirectly through 
increased resources for offspring (Hrdy 1979). 
Thus, infanticide without apparent fitness benefits 
is much more difficult to explain. Any titness 
benefits gained by the female, direct or indirect, in 
our observations arc difficult to ascertain ( Table 2). 
Evidence for sexually-selected infanticide has been 
documented for male Barn Swallows, but this 
behavior has not been observ ed for females. Female 
infanticide has been documented in other passer¬ 
ines. particularly in species where male mates are 
limited (Veiga 1990, Chek and Robertson 1991, 
Hansson et al. 1996. Veiga 2004). It is possible the 
female from nest # 37 killed the nestlings from nest 
# 106 in an attempt to usurp the male from that nest 
for a future breeding attempt, but it seems unlikely 
given the pair from nest # 37 remained intact lor a 
second nesting attempt. However, the pair from nest 
# 106 was not observed for the remainder of the 
season and. while completely speculative, perhaps 
the female’s intention of usurping the male was not 
realized due to his disappearance. 
The architecture of the particular barn made it 
difficult to monitor active nests in the indoor 
arena and we do not know whether birds captured 
but not associated with a nest were truly floaters 
(males and females), or nesting in the arena. We 
observed 43 females and 50 males in the barn- 
captured in mist nets or detected as unhanded 
adults attending nests. This ratio suggests a 
surplus of males, which further contradicts the 
hypothesis of sexually-selected infanticide by 
females as reported for other socially monoga¬ 
mous passerines with biparental care (Veiga 1990, 
Chek and Robertson 1991, Veiga 2004). 
Alternative explanations could involve limited 
resources, either nests or food.. Limited availabil¬ 
ity of nest sites is unlikely given the number of 
available potential sites in the barn. There were 
two other active nests in close proximity to nest 
# 37 that remained undisturbed, and the pair 
renested within 30 cm of their original nest (# 37) 
where they continued to tolerate these neighbors. 
The pairs at these undisturbed nests may have 
been better nest guarders (Moller 1988). but we 
did not observe any differences in nest attendance 
(JKH and ALT, pers. obs.). Limited availability of 
food resources is also unlikely. We have seen high 
nestling mortality and predation resulting in low 
reproductive success in past years. However, the 
