Summary
Sexually selected infanticide, the killing of young to secure a mating opportunity, appears to be rare in birds, although it is relatively common in mammals. To examine the frequency of occurrence and the conditions under which male tree swallows, Tachycineta bicolor, commit infanticide, resident males were removed from 15 nests shortly after the eggs hatched. Of seven replacement males, two were directly observed killing nestlings, and three were strongly suspected of killing nestlings. Three of the infanticidal males proceeded to mate; one with the widowed female, and two with a new female. A subadult female floater was also observed committing infanticide at a nest where both residents were present. For male tree swallows, a nest site is essential for acquiring a mate. Intense intrasexual competition among males for nest sites results in a floating population of males with no mating success. It is argued that infanticide in male tree swallows is a sexually selected behaviour, because unmated males that are infanticidal obtain a nest site and therefore have a higher probability of mating than males that refrain from infanticide.
Methodology
This study was conducted at the Queen's University Biological Station, 50 km north of Kingston, Ontario, where we have grids of nest boxes in four open hayfields. Nest boxes are placed 1.5 m high on aluminium posts to reduce predation. Boxes used in this study were 28 m from the nearest box along the diagonal, and 40 m from adjacent boxes along the rows and columns of the grid. Grids of boxes have been in place since 1976, and the population has been stable at about 55-65 nesting pairs for the past 5 years.
Most resident males and females were captured early in the season using mist nets or nest-box traps (Stutchbury & Robertson 1986), banded with Canadian Wildlife Service numbered leg bands, and uniquely marked on the wings and tail with acrylic paint to allow recognition of individual birds.
Males were removed within 3 days of hatch at five nests in 1985. In 1986, males were removed from 10 nests within 2 days of hatch. Removals were accomplished by shooting the resident male since we knew of no way to keep an aerial insectivore alive in captivity for the time required for this experiment. Experimental nests were located in all four grids and were not adjacent to other experimental nests. Only nests with hatch dates at the peak of hatching for this population (early June) were included in this experiment.
In 1985, we observed experimental nests for 1 h on the day of the removal and periodically thereafter, depending on whether the nestlings had been injured. In 1986, we observed experimental nests for 1 h per day for 5 days following the removal, and then every other day. We made additional observations at nests where there was evidence of injury or where nestlings had disappeared. Removals and observations were done in the morning when activity was highest. During observations, we recorded the presence and behaviour of all intruders and replacement males, and the response of the resident female to intruders. Subadult (1-year-old) female intruders were distinguished by their brown-blue plumage as opposed to the iridescent blue plumage of adult males and females (Hussell 1983). Male intruders were distinguished on the basis of their behaviour (Cohen 1984). Males were considered to be replacements if they were seen persistently at the nest site. The nestlings were checked every day for injuries.