Summary
We monitored the time spent at the nest and following behavior of mated tree swallows to determine if males were guarding their mates. The proportion of time spent together at the nest did not decrease significantly between fertile and postfertile periods, and the tendency of males to follow females was not significantly different from that of females to follow males. Following by either sex was infrequent. We suggest that the lack of mate guarding in tree swallows is related to an apparently low probability of extra-pair copulations, which in turn is likely to be due to two factors. Nesting opportunities are limited, perhaps more so for females than males. If the operational sex ratio is skewed towards females, this would not only reduce the risk of extra-pair copulations, but would also select against promiscuous females which would risk being abandoned by their mates. Secondly, under natural conditions, the limited availability of nest sites has selected for territorial defense by both males and females, which may decrease the occurrence of extra-pair copulation. Both factors would lead to relaxed selection for mate guarding behavior.
Methodology
This study was carried out during the summer of 1982 at the Queen's University Biological Station near Chaffey's Lock, Ontario (44 34'N, 76 20'E). The study site consisted of seven hayfields and a nearby marsh in which grids of nest boxes had been established in 1977 or earlier. All boxes were mounted on aluminum poles to prevent predation. Of 139 boxes, 96 were arranged in spiral cells of six boxes each, with boxes located 1, 2, 4, 8 and 16 m from a central box. Because of the territoriality of tree swallows (Robertson and Gibbs 1982), usually only one box within each spiral cell was occupied. Fifty-five clutches were initiated in the study area in 1982.
In four of the hayfields and in the marsh, the presence of at least 20 boxes allowed several pairs of tree swallows to nest. The birds nesting in these areas are referred to as the "dense-nesting" pairs. In each of the remaining three fields, a single box permitted only one pair of tree swallows to nest. These birds are referred to as the "sparse-nesting" pairs.
At the start of the breeding season we captured birds on the nest, or with mist nets, and marked each one with acrylic paint. We used white, red, green and yellow paints on the back, forehead, wings and tail in specific patterns to enable individual identification. Because the sex of the bird was not always obvious upon examination, subsequent behavior during incubation was used to distinguish the sexes (only females incubate).
To determine whether or not mate guarding was occurring, we recorded activities of 17 of the dense-nesting pairs and 3 of the sparse-nesting pairs at the nest box before, during and after the egg-laying period. Observations began once pairs had chosen a nest box and had been color-marked. Observations were made from 7 May to 29 May, by which time most pairs were incubating, and hatching was occurring in the earliest nests.
We watched pairs at varying times on alternate days and the total number of observation periods varied for different pairs of birds. Each nest was observed for one 30-rain period between 0600 and 1200 h on the day of study. Observations were carried out under all conditions except when heavy rain drove the birds from the area. During the observation period, we recorded times of arrival and departure from the nest for each bird. We could then determine how much time the pair spent together at the nest site. Since a pair often defended an entire spiral cell of boxes, arrival of a member of the pair was defined as the moment when a bird perched anywhere within 18 m of the nest box. A departure occurred when the bird left such a perch.
We compared the spatial and temporal relations between the male and female during and after the female's fertile period. The data for dense and sparse-nesting pairs were analyzed separately. Because of small sample sizes and non-normally distributed data, nonparametric statistics (Siegel 1956) were used to analyze the data