Female Responses to Intraspecific Brood Parasitism in the Tree Swallow

Abstract

We studied female responses to experimental intraspecific brood parasitism (IBP), or egg-dumping, in Tree Swallows (Tachycineta bicolor). Unlike other species of swallows, Tree Swallow nests are rarely parasitized by conspecifics. We experimentally parasitized nests of Tree Swallows to investigate how females respond to uncertain maternity. Host females accepted a parasitic egg if it was added to the nest within 3 days of the host's first egg (62%). In contrast, the host female buried the parasitic egg (24%) or deserted the nest (14%) when the parasitic egg was added 4 or more days before the host's first egg. The acceptance of parasitic eggs close to the host's own laying date is similar to the behavior reported for other species; however, egg burial and nest desertion appear to be rare as responses to intraspecific brood parasitism. We suggest that the low level of IBP in Tree Swallows has evolved as an indirect consequence of females defending their nest cavity against usurpation.

Individuals can potentially enhance their reproductive success by parasitizing the parental care of conspecifics. In biparental species, males may parasitize the parental care of conspecific males through extrapair fertilizations (EPF), in which one male fertilizes the eggs of another male's mate. In contrast, females may lay their eggs in the nests of conspecifics and parasitize the parental care of both male and female care givers by intraspecific brood parasitism (IBP). Recent studies have shown that EPFs are widespread among taxa and common in many species of birds (reviewed in Gowaty 1996). As a result, male response to uncertain paternity has received much recent attention (Whittingham and Dunn, in press). In contrast, IBP is relatively rare, but also is widespread among taxa and common in some species of birds (Rohwer and Freeman 1989). In this case, both sexes face uncertain parentage and incur the costs of caring for unrelated young, although the costs are likely to be higher for female birds, which generally invest more than males in rearing young (Andersson 1984). Thus, in species with IBP we expect females to evolve responses to uncertain maternity that reduce their probability of investing extensive parental care in unrelated young. Here, we present the results of an experimental study of female responses to IBP in the Tree Swallow (Tachycineta bicolor).

Female responses to IBP are expected to evolve if the costs exceed the benefits of providing care to unrelated young. Female responses to IBP may include: (1) aggressive nest guarding, (2) asynchronous egg laying, (3) parasitic egg ejection, (4) parasitic egg burial, and (5) nest desertion (Petrie and Møller 1991). Although all of these responses have been observed in birds, the type(s) of female response varies among species. The type of the response may depend on the timing of IBP relative to the host's clutch initiation and the frequency of IBP.

IBP is expected to be more common in birds that: (1) lay large clutches, (2) breed in colonies, (3) have intense competition for nest sites, and (4) lack territorial defense (Hamilton and Orians 1965, Yom-Tov 1980). IBP has been detected in five of seven species of swallows and, thus, may be a fairly common female reproductive strategy in Hirundinidae. Many species of swallows are colonial or nest at high densities, and some may be severely nest-site limited because they are secondary cavity nesters (Holroyd 1975). IBP occurs frequently in colonial Cliff Swallows (up to 40% of nests, Hirundo pyrrhonota; Brown and Brown 1989) and Purple Martins (up to 36% of nests, Progne subis; Morton et al. 1990). Female Cliff Swallows are known to eject brood-parasite eggs that are laid more than 4 days before the host begins laying.

We might expect a similar response in Tree Swallows, which are secondary cavity nesters, or cavity adopters (limited by nest site availability, Leffelaar and Robertson 1985), and nest at high densities in both natural cavities and nestbox grids (Barber et al. 1996). Although their breeding ecology suggests that IBP might be common, it has been reported in only 1% to 6% of nests in studies of several different populations (Kuerzi 1941, Lombardo 1988, Stutchbury and Robertson 1987). IBP is much less frequent in Tree Swallows than in other species of swallows, and it is not known how female Tree Swallows respond to IBP.

Methods

We studied Tree Swallows during 1997–2000 at the University of Wisconsin-Milwaukee Field Station near Saukville, Wisconsin (43°23′N, 88°01′W). Tree Swallows nested in two grids of 28 and 30 nestboxes, with a minimum of 28 m between boxes. An additional 24 nestboxes were distributed randomly in between the grids (15–35 m apart). The density of nestboxes in our population was similar to Tree Swallows nesting in natural cavities and to densities in other box-nesting populations (Robertson and Rendell 1990, Barber et al. 1996).

In nestboxes, completed nests range from 2 to 8 cm deep, and the nest cup is usually lined profusely with feathers of other species, although most feathers are added after egg laying begins (Robertson et al. 1992). In our population, completed nests were usually 5 to 6 cm deep (as measured from the side opening of the nestbox), with a few feathers in the nest cup when egg laying began. Nest construction began in late April and continued through early May, and most nests contained 5 or 6 eggs. All nestboxes were opened and inspected daily to record the progress of nestbuilding and egg laying. Thus, levels of disturbance were similar at all nests. Females were captured early in the nestbuilding period, given a USFWS aluminum legband, and marked for individual identification with non-toxic colors on the breast (felt-tip markers), wings or tail (acrylic paint; Dunn et al. 1994). Adult female swallows were classified as second-year (SY) or after-second-year (ASY) on the basis of plumage coloration (Hussell 1983).

In 1997 and 1998 we experimentally parasitized 29 swallow nests (10 in 1997 and 19 in 1998) by adding one egg to the nest prior to the start of egg laying by the resident female (host). All experimental eggs were added to nests between 06:00 and 08:00 when egg laying usually occurs. We chose experimental nests randomly from among the active nests within our population and added parasitic eggs when nest building was essentially completed. In 1997, five nests were parasitized using a viable Tree Swallow egg collected from another nest in the population and five nests were parasitized with a model Tree Swallow egg. Model eggs were made from white Fimo modeling clay (Eberhard Faber Co., Neumarkt, Germany). This nontoxic clay was initially soft and pliable but hardened after heating at 130°C for 20 min. Model eggs were very similar in color (white) and mass (mean ± SD = 1.9 ± 0.1 g) to natural Tree Swallow eggs (1.9 ± 0.1 g). We did not measure the length or width of natural Tree Swallow eggs in our population, but the model eggs were similar in these dimensions (length: 18.5 ± 0.3 mm, width: 13.2 ± 0.4 mm) to those reported for natural Tree Swallow eggs in other populations (average length 18.7 mm, average width 13.2 mm; Robertson et al. 1992). There was no bias in the response of host females (desertion, burial, or acceptance) to natural or model eggs in 1997 (χ²₂ = 2.9, P = 0.2), and, thus, we used only model eggs for the experiments in 1998. All tests were two-tailed, and means are presented ± SE unless noted otherwise.

To examine whether females responded similarly to a non-egg object, we conducted a second experiment in 2000. From 1 to 8 days before egg laying, we placed one “pebble” in each of 13 nests of ASY females. Pebbles were made from blue Fimo modeling clay with the intention that they could be detected easily (i.e., were not camouflaged against the nesting material) and did not resemble the white eggs. Pebbles were 13.8 ± 0.3 mm in diameter and shaped irregularly. Pebbles were added to nests between 06:00 and 08:00 and nests were checked every day until incubation began.

Results

Females experimentally parasitized with a real or model egg responded in one of three ways: (1) accepted the parasitic egg, (2) buried the parasitic egg, or (3) deserted the nest. In 18 of 29 (62%) experimental nests, the host female accepted the parasitic egg and incubated it normally with her clutch (egg laying began within 3 days in all cases). In 7 of 29 (24%) cases, the host female buried the parasitic egg in nesting material and began laying her clutch 4 to 6 days later. In five of these cases, we observed females bringing more nesting material to the nest and subsequently burying the parasitic egg. In the other cases, the female may have pushed the parasitic egg below the nest cup or brought additional nesting material when we were not observing. Either way, the parasitic egg was not present in the nest cup and was not incubated with the female's own clutch.

Four of 29 (14%) host females deserted the nest containing the parasitic egg and initiated a new nest in a nearby vacant nestbox. In all cases desertion occurred within 24 hr of the appearance of the parasitic egg in the nest. There were no other cases of nestbox desertion during this study; thus, desertion appears to have occurred specifically in response to the brood parasite egg. Females that deserted began laying in their new nestbox within 8 to 19 days. Of the 29 experimental females, 21 were ASY and 8 were SY females. There was no difference in female response to parasitism in relation to female age (χ²₂ = 1.0, P > 0.5).

Host females were more likely to accept a parasitic egg if it appeared within 3 days of the laying date of the host's first egg (logistic regression of female response [accept or reject—“buried” and “deserted” combined] on host laying date: χ²₁ = 38.5, P < 0.001). Although desertion of the nest occurred soon after the appearance of the parasitic egg, we cannot rule out the possibility that other factors influenced nest desertion of those four females. Thus, we also did the analysis excluding females that deserted the nest and found a similar result (logistic regression χ²₁ = 29.6, P < 0.001).

For host females that deserted their parasitized nestbox, it was not possible to determine whether desertion prolonged the time to laying of their first egg or whether they were just further away from their first egg date than the other females and had more time to investigate empty nestboxes and build a new nest. Of the four host females that deserted and renested, three of them nested in a neighboring nestbox (within 40 m of the original nestbox). The other female renested in a box 56 m away from her original nestbox.

In Cliff Swallows, parasitized females reduce the number of eggs they lay, and thus incur a cost from being parasitized (Brown 1984). To determine whether this occurred in Tree Swallows, we examined the clutch size of unmanipulated females in relation to females in the different parasite response groups. Clutch sizes of females that accepted (5.7 ± 0.1 eggs) or buried (5.3 ± 0.4 eggs) parasitic eggs were similar to those of unmanipulated (5.5 ± 0.1 eggs) females (Mann-Whitney U- tests, n₁ = 18, n₂ = 26, U = 195, P = 0.3; n₁ = 7, n₂ = 26, U = 75.5, P = 0.45, respectively). In contrast, females that deserted the parasitized nest and renested had significantly smaller clutch sizes (4.5 ± 0.3 eggs) than unmanipulated females (Mann-Whitney U- test n₁ = 4, n₂ = 26, U = 16, P < 0.02). Thus, Tree Swallow hosts did not reduce their clutch size in response to parasitism; however, females that deserted and renested incurred a cost because they nested later in the season when clutch size was smaller.

Female Tree Swallows responded differently to the non-egg pebbles than they did to real or model eggs. Most females ejected the pebble from the nestbox, a response which did not occur with parasitic eggs. Of 13 females, 10 ejected the pebble from the nestbox, 2 buried the pebble, and 1 accepted it and began incubation of her clutch with the pebble among the eggs. In all cases, ejection of the pebble from the nestbox occurred within 24 hr (prior to the next nest check). Ejection of the pebble was confirmed when the nest was removed and examined following fledging. There were no cases of nest desertion by the host female in response to the non-egg object.

Discussion

Female Tree Swallows either accepted, buried, or deserted conspecific parasitic eggs, whereas they generally ejected non-egg objects (irregularly shaped pebbles). Thus, we controlled for the disturbance caused by placing a foreign object in the nest (egg or pebble). The difference in response to experimental eggs and pebbles suggests that females can recognize eggs, but why they did not also eject the experimental eggs is unknown. The most important factor influencing female response was the timing of the parasitic egg relative to the host's first egg. Parasitic eggs appearing within 3 days of the host's first egg were accepted, whereas parasitic eggs appearing 4 or more days before the host's first egg were buried or the nest was deserted. The acceptance of parasitic eggs close to the host's own laying date is similar to the behavior reported for other species of birds, including other swallows. However, when Tree Swallows rejected parasitic eggs they differed markedly from other species in their type of response.

Tree Swallows are similar to other species of swallows in their criteria for the acceptance of parasitic eggs from conspecifics. During the pre-laying period, female Cliff Swallows only accepted eggs appearing less than 4 days before the host's first egg (Brown and Brown 1989). Similarly, in Barn Swallows (Hirundo rustica, 16.5% of broods with IBP), experimental brood-parasite eggs were accepted only when they appeared during laying or incubation, and not when they appeared during the last 5 days of the pre-laying period (Møller 1987). An artificial egg addition experiment in Sand Martins (Riparia riparia; Alves and Bryant 1998) showed that most (91%) parasitic eggs were accepted when they appeared during or after the onset of laying, but not when they appeared during the pre-laying period (58% rejected). However, it was not stated how long prior to the host's first egg the artificial parasitic eggs were added, and, thus, it is difficult to compare to our study.

Eggs can be rejected by ejection, burial, or nest desertion. The most obvious differences between Tree Swallows and other species of swallows related to how they rejected the parasitic eggs. When rejecting parasitic eggs, both Cliff and Barn Swallows ejected the egg from the nest. Alves and Bryant (1998) did not state the method Sand Martins used to reject parasitic eggs during the pre-laying period; however, brood parasite females did eject a host egg from the nest when adding their own egg, so it follows that hosts would certainly be capable of ejecting a parasitic egg as well. We did not observe egg ejection in Tree Swallows and, to our knowledge, it has not been reported elsewhere. We do not believe that our results are an artifact of using artificial eggs as they were similar in color, size, and mass to natural Tree Swallow eggs, and none of the natural eggs were ejected in the 1997 experiment.

Burying eggs may be an adaptation to reusing limited nesting cavities in species that are unable to eject foreign eggs. Of the species of swallows discussed here, Tree Swallows are the only cavity adopters, whereas Sand Martins excavate burrows and Cliff and Barn Swallows build mud nests. Cavity adopters depend on other species to excavate cavities and as a result are often nest-site limited (Holroyd 1975). In Tree Swallows, competition among females for nest sites is intense, often resulting in fights and nest usurpation (Leffelaar and Robertson 1985, Stutchbury and Robertson 1987); thus, egg burying may facilitate the reuse of limited nesting cavities. American Coots (Fulica americana; Arnold 1987) are one of the few other species in which hosts respond to intraspecific brood parasitism by burying eggs. Egg burying was detected in coots in response to the experimental addition of many eggs in one nest during the host's laying period. Although egg burying is a common response to interspecific brood parasitism (Rothstein 1990), it appears to be rare when birds parasitize individuals of their own species.

Nest desertion also is an uncommon response to intraspecific brood parasitism, especially in altricial species. In our study, three of the four desertions occurred in 1997, when nestbox occupancy was lower (37%) than in 1998 (87%). Thus, desertion may have been influenced by the availability of unoccupied nestboxes. In both years, the parasitized females that deserted nests all found vacant nestboxes nearby and renested; none forfeited nesting for the entire season. Desertion may be the most costly response for Tree Swallows, as females that renested had significantly smaller clutches (mean of 1.0 fewer eggs) than other experimental females or unmanipulated females (probably as a consequence of nesting later in the season). Thus, the benefits of desertion are unclear, as the costs of rebuilding the nest and relaying appear similar to, if not greater than, the cost of raising one parasitic nest-ling.

In Cliff Swallows, IBP is often a successful strategy (parasitic eggs are accepted) when hosts are parasitized within 3 days of the start of laying by the host (Brown and Brown 1989, 1998). Female brood parasites also have greater overall reproductive success than hosts (Brown and Brown 1998). Given that we have found a similar “acceptance window,” why is IBP not more common in Tree Swallows?

The low rate of IBP in Tree Swallows could be related to nesting density. For example, IBP occurs more often in large than small colonies of Cliff Swallows (Brown and Brown 1989) and Barn Swallows (Møller 1987), suggesting that increased nesting density facilitates IBP. In Cliff Swallows, most successful intraspecific brood parasites lay their eggs in the nests of adjacent neighbors or in nests within 1 m of their own (Brown and Brown 1989). Similarly, Alves and Bryant (1998) suggested that a slightly higher rate of nest parasitism in Sand Martins (4% of nests) than House Martins (Delichon urbica, 0% of nests, Riley et al. 1995, Whittingham and Lifjeld 1995) occurred because Sand Martins nest in larger and denser colonies than House Martins. In contrast, density does not appear to influence IBP in studies of Tree Swallows nesting in both natural cavities and nestboxes (Dunn et al. 1994, Barber et al. 1996).

Alternatively, the low level of IBP in Tree Swallows may be influenced by female-female aggression. In the competition for nest sites, female Tree Swallows are sometimes injured or killed defending a cavity against usurpation by other females (Leffelaar and Robertson 1985, Lombardo 1986). Losing a nest site and, thus, the opportunity to breed for the entire season is likely to be much more costly than the addition of a brood parasite egg to the clutch. Thus, it seems likely that the intense aggression displayed in defending the nest site throughout the pre-laying, laying, and incubation periods (Dunn and Hannon 1991) may have the indirect consequence of also defending against potential brood parasites.

In summary, Tree Swallows have evolved a set of responses to intraspecific brood parasitism even though it occurs at very low frequencies. The rejection responses of Tree Swallows are more variable than other species of swallows. Differences between and within species in their response to parasitism may be influenced by factors such as nest type and nest density. As a contrast to Tree Swallows, it would be interesting to study female responses to intraspecific brood parasitism in another cavity adopting species that nests at greater density, such as the Purple Martin, in which up to 36% of first year females experience IBP (Morton et al. 1990).

We thank Lisa Belli, Ethan Clotfelter, and Kevin Thusius for help in the field. The UWM Field Station staff facilitated this work, and we are particularly grateful to Lou Nelson for building the nestboxes. This research was supported in part by a UWM Graduate School Research Award and NSF grant IBN-98-05973.

Author notes

[email protected]