Extra-Pair Mating Tactics and Vocal Behavior of Female Acadian Flycatchers

Abstract.

In many passerines, rates of extra-pair paternity are high, but relatively few studies have examined the behaviors females use to obtain extra-pair copulations. We studied extra-pair behavior in females of the Acadian Flycatcher (Empidonax virescens), a species in which 58% of females produce extra-pair young and males often sire offspring on distant territories. We used radiotelemetry to document the extent and frequency of females' forays away from their territories and used behavioral observations and playback experiments to test if females' vocalizations advertise their fertility. No radio-tagged females left their territories (n = 12 females, 105 hr of tracking) even during the hour before dawn. Paternity analyses revealed that at least six of these females produced extrapair young. Focal observations revealed that fertile females spent significantly more time calling “chiff,” gave more calls per hour, and called in longer bouts than did incubating females. Extent of females' chiffing during their fertile period was not related to subsequent extra-pair paternity in their nest (n = 12). We tested for a territory-defense function of chiffing by playing chiff calls to females at both the fertile and incubation stages. Females responded to playbacks on their territory by approaching or flying over the speaker, did not increase their chiffing rate significantly, but did give other high-intensity calls. Our results suggest that extra-pair fertilizations of Acadian Flycatchers occur primarily on a female's own territory and that extra-pair males could use females' vocalizations to locate fertile females.

Resumen.

En muchas aves paseriformes, las tasas de paternidad extra-pareja son altas, pero pocos estudios han examinado los comportamientos que usan las hembras para obtener cópulas extra-pareja. Estudiamos el comportamiento extra-pareja en hembras de Empidonax virescens, una especie en la cual el 58% de las hembras produce pichones extra-pareja y en que los machos frecuentemente engendran crias en territorios distantes. Seguimos a las hembras utilizando radio telemetría para documentar el alcance y la frecuencia de las salidas de sus territorios. Además empleamos observaciones de comportamiento y experimentos con reproducción de llamadas previamente grabadas para evaluar si las vocalizaciones de las hembras publicitan su fertilidad. Ninguna de las hembras marcadas con transmisores dejaron sus territorios (n = 12 hembras, 105 hs de seguimiento) incluso durante la hora antes del amanecer. Los análisis de paternidad revelaron que al menos seis de estas hembras produjeron pichones al extra-pareja. Observaciones focales revelaron que las hembras fértiles gastaron significativamente más tiempo realizando llamados tipo “chiff,” emitieron más llamadas por hora y solicitaron más encuentros que las hembras que incubaban. El alcance del llamado “chiff” de las hembras durante sus períodos de fertilidad no se relaciono con una subsecuente paternidad extra-pareja en sus nidos (n = 12). Evaluamos si el llamado “chiff” tiene una función de defensa del territorio mediante la reproducción de llamados a hembras en etapas fértiles y de incubación. Las hembras en sus territorios respondieron a las grabaciones acercándose o volando sobre el parlante, no aumentaron significativamente sus tasas de llamados “chiff,” pero si emitieron otras llamadas de alta intensidad. Nuestros resultados sugieren que las fertilizaciones extra-pareja en E. virescens se dan primariamente en el propio territorio de una hembra y que los machos extra-pareja podrían usar las vocalizaciones de las hembras para localizar a las hembras fértiles.

Introduction

Sexual conflict in animals is perhaps at its highest in socially monogamous species where females obtain fertilizations outside the pair bond. Extra-pair paternity is common in birds (reviewed in Griffith et al. 2002, Westneat and Stewart 2003) and can create strong sexual selection via female extra-pair mate choice and differential male mating success (Stutchbury et al. 1997, Dolan et al. 2007, Webster et al. 2007). Pursuit of extra-pair copulations can benefit females indirectly (genetically) and/or directly (reviewed in Westneat and Stewart 2003). Despite hundreds of papers published documenting extra-pair fertilizations in birds, studies of the behaviors females use to obtain extra-pair copulations are still relatively uncommon (Rubenstein 2007). Females may solicit extrapair males by advertising their fertility (Sheldon 1994, Tarof et al. 2005, Neudorf et al. 2008) and/or pursuing prospective extra-pair males off territory (Smiseth and Amundsen 1995, Neudorf et al. 1997, Double and Cockburn 2000, Tarof et al. 2005). Here we investigate the role of female behavior in the extra-pair mating system of the Acadian Flycatcher (Empidonax virescens), a sexually monomorphic bird with high rates of extra-pair fertilization and long-distance off-territory forays by males (Woolfenden et al. 2005).

Females that seek extra-pair copulations by making off-territory forays may benefit as a result of sampling a larger number of potential mates and/or interacting with particular males without interruption or harassment from their social mates. For most species with extra-pair fertilizations, however, nothing is known about whether or not females pursue extra-pair copulations off territory even though this strategy can be a primary mechanism of female choice and strongly influence male mating success (Wagner 1992). In the Superb Fairy-wren (Malurus cyaneus), for example, female visits to neighboring territories result in extra-pair young being sired by the visited male(s) (Double and Cockburn 2000). In the Hooded Warbler (Wilsonia citrina), female foray rate is positively correlated with the number of extra-pair young in the nest and females make more frequent forays off territory if paired to a male singing at a relatively low rate (Chiver et al. 2008). Similarly, female Common Yellowthroats (Geothlypis trichas) are more likely to visit the territories of more ornamented males, and these males also sired more extra-pair young (Pedersen et al. 2006). An alternative explanation for female off-territory movements is pursuit of food (Gray 1996). Evidence that supports female forays' contributing toward extra-pair copulation includes the observation that in most species females make forays only when fertile (Evans et al. 2008) and a food-supplementation experiment in which well-fed female Northern Cardinals (Cardinalis cardinalis) made more, rather than fewer, forays (Humbird and Neudorf 2008).

An alternative, but not mutually exclusive, female extra-pair copulation behavior is attracting extrapair mates via conspicuous vocalizations. The fertility-advertisement hypothesis (Montgomerie and Thornhill 1989) proposes that fertile females give loud calls to attract potential extra-pair mates onto their territory, allowing them to choose a copulation partner from among competing males. Tobias and Seddon (2002) found that female European Robins (Erithacus rubecula) emit loud and frequent “seep” begging calls to solicit food from their social mate during courtship. In a mate-removal experiment, while on their territory females increased their rate of seep calling, which attracted neighboring males who exchanged food for copulations. Fertile female Hooded Warblers emit more “chip” bouts per hour than do nonfertile (incubating) females (Neudorf et al. 2008), and neighboring males are more likely to visit territories when females are chipping (Stutchbury 1998). Tarof et al. (2005) found that fertile female Least Flycatchers (Empidonax minimus) often give conspicuous “whit” calls near the border of their mate's territory, attracting extra-pair copulation attempts by neighboring males. Female vocalizations are typically considered important in territory defense, courtship feeding, communication between mates, and nest defense (Hobson and Sealy 1989, Wiley 2005), but few studies have compared the vocal behavior of fertile and nonfertile females.

In this study, our objective was to use radiotelemetry and playback experiments to examine the extra-pair mating tactics of female Acadian Flycatchers. This species is socially monogamous and sexually monochromatic, and pairs defend all-purpose territories in mixed deciduous hardwood forests. Woolfenden et al. (2005) found that 58% of Acadian Flycatcher nests contained extra-pair young, and radiotracking revealed that most males made long-distance forays to other territories. About half of the genetically identified extra-pair sires were not immediate neighbors; indeed, extra-pair sires often defended territories far from the nests containing their extra-pair young (Woolfenden et al. 2005). It was unknown whether female Acadian Flycatchers also make off-territory forays, and if so, whether they travel as far as males. Especially while building nests and incubating, female Acadian Flycatchers also give a frequent and spontaneous “chiff” call (Whitehead and Taylor 2002), which could attract intrusions by extra-pair males.

First, we used radiotelemetry to determine if female Acadian Flycatchers make forays off territory to pursue extra-pair copulations. Second, we examined the fertility-advertisement hypothesis by testing the prediction that (1) the frequency of chiff calls should decrease after the onset of incubation when females are no longer fertile, and (2) the frequency of extra-pair fertilizations should be positively correlated with frequency of female chiff vocalizations. The alternative hypothesis that the chiff functions primarily in territory defense also predicts that chiff frequency is highest during fertile period because this corresponds with recent territory establishment. Therefore, we conducted a playback experiment to test if females respond aggressively to female chiff vocalizations on their territory.

Methods

We conducted this study at the Hemlock Hill Biological Research Area (41° 46′ N, 79° 56′ W) near Cambridge Springs, Pennsylvania, from May to July 2004–2006. The Acadian Flycatcher is a small (13–14 g) neotropical migrant that typically forms socially monogamous pair bonds (<5% polygyny), and most pairs are single brooded. Only the female builds the nest and incubates, but both sexes feed the young. In our study area (∼125 ha) Acadian Flycatcher territories were arranged linearly along one of seven streams in a moist, low-lying area within a 180-ha fragment of mixed deciduous—hemlock forest (Fig. 1; see Woolfenden et al. 2005 for more details). We monitored all territories (35–40 per year) on this study site via nest searching and banded about 75% of the territorial males annually. Territories averaged approximately 1 ha in size, and territories were often separated by unoccupied forest.

Males first began to settle on our study site during the first two weeks in May; females arrived 1–2 weeks later. We mapped territory boundaries by observing male song and border disputes (e.g., chases, counter-singing). We visited each territory twice per week to determine territory residency and the pairing status of breeding adults. Pairing status (mated or not) was determined by male—female interactions and the presence of females nesting on the territory (Whitehead and Taylor 2002). We captured males in mist nets by using conspecific song playback accompanied by a decoy. We banded adults (n = 124) with a single U.S. Geological Survey aluminum band on one leg and a unique combination of colored bands on the other leg. We measured body mass, tarsus length, and flattened wing length.

Radiotelemetry of Females

To quantify females' off-territory foray behavior, we captured 12 individuals in 2004 and 2005 during nest building and fitted each with a BD-2 (0.6 g) radiotransmitter (Holohil Systems, Ltd.). Radiotransmitters were attached to each bird with a legloop harness made from cotton embroidery thread (Rappole and Tipton 1991). This protocol has been used successfully to track male Acadian Flycatchers without adverse effects on the birds' physical condition or locomotion (Woolfenden et al. 2005). None of the tagged females were depredated during this study.

An R-1000 receiver (Wildlife Materials International) and hand-held three-element Yagi antenna were used to track females during focal telemetry sessions. We followed individual females during their fertile period for 30–180 min at a time, once per day if weather permitted, and total tracking time averaged 8.7 hrper fertile female (range 4–12.5 hr). We defined the fertile period as beginning 5 days before the first egg and ending with laying of the penultimate egg. Female Acadian Flycatchers lay one egg daily for up to a total of 3 eggs per clutch (2.6 ± 0.5 eggs, range 2–3 eggs), so the fertile period lasted 6–7 days depending on the size of the clutch. Tracking began 2 days after the radiotransmitter was attached to allow the bird time to adjust to the transmitter (Woolfenden et al. 2005). Tracking took place between 05:00 and 18:00 Eastern Daylight Time, mostly between 05:00 and 12:00 (Table 1).

To avoid disturbance of normal behavior during telemetry, we maintained a distance of approximately 30 m from the focal female. We determined locations with a compass and 50-m grid reference system (Woolfenden et al. 2005) and recorded locations continuously. We noted any instance when a female left her territory. For analysis, we conservatively defined an off-territory foray as a bird traveling at least 50 m beyond the territory's boundary (Woolfenden et al. 2005). We also quantified female chiff vocalizations (see below), interactions with other Acadian Flycatchers, and any other conspicuous behaviors (e.g., chases, copulations). We recovered four radiotransmitters by capturing females during the incubation or nestling stage; the remaining transmitters fell off during these stages.

Vocalizations and Playback Experiments

To determine the role of female vocalizations in fertility advertisement or territory defense in the Acadian Flycatcher, we made behavioral observations of fertile and incubating females (3 hr at each stage) and conducted a playback experiment. For the natural observations, we monitored vocalizations of fertile (n = 12 radio-tagged, 9 not tagged) and incubating (n = 5 tagged, 11 not tagged) females. We quantified the following variables: total number of chiffs given per hour, percentage of time chiffing, number of chiffing bouts per hour, mean bout length, and average chiff rate (chiffs min^-1) during bouts. Chiff bouts were periods of consecutive chiffing separated from other such periods by at least 2 min.

In 2006, we conducted a playback experiment by using two high-quality recordings of local undisturbed fertile females made with a Sennheiser MKH 104 omni-directional microphone mounted in a parabolic reflector, diameter 44 cm. Female call notes are difficult to record at high quality, and therefore we elected to use only two recordings rather than compromise the sound quality. We used a Nagra III tape recorder set at a recording speed of 38.1 cm sec^-1 and then copied the recordings to a Sony Proll cassette recorder to make two separate endless-loop playback tapes. To produce sequences of chiffs matching the intervals used by females under natural circumstances we inserted silent intervals between chiffs to produce a natural rate (6 min^-1). Playback experiments were performed on 10 non-radio-tagged females between 0630 and 1200 Eastern Daylight Time during their fertile and nonfertile (incubation) stages in a paired design. Each female was retested at least 2 days after the onset of incubation. For these incubating females, we waited for the female to come off her nest naturally before beginning the playback experiment.

During playback experiments, chiff recordings were played for 5 min through an SME-AFS amplified field speaker driven by a Sony cassette tape player. The recording used for a given experiment was chosen randomly to avoid pseudoreplication or order effects. For each experiment, the speaker was placed on the ground within the territory, 40 m from the female and nest, and aimed toward the subject. Playback began after 5 min of undisturbed observation. The observer stood concealed behind vegetation at least 20 m from the speaker and quantified female behavior during the 5 min of playback. An additional 5 min of observations were recorded after playback. We measured female chiff rate (min^-1), the occurrence and rate (min^-1) of “wheeu” calls (Wiley 2005), and the distance (m) of closest approach to the speaker. The wheeu call is a distinct but infrequently heard female vocalization that appears to be a high-intensity scolding call (Mumford 1964, Wiley 2005).

Paternity Analysis

During the three breeding seasons we collected a 5- to 50-µL blood sample by brachial venipuncture from 14 social pairs (four in 2004, three in 2005, seven in 2006), 15 neighboring males, seven females, and 27 nestlings (from 12 families) for microsatellite paternity profiling. Social parents were identified on the basis of field observations of nest defense and nestling feeding, and brood size averaged 2.3 ± 0.8 nestlings (range 1–3). For three nests we could not obtain a DNA sample from the social male, so the nestlings were omitted from paternity analysis. Each adult was represented once in our paternity data except for one female sampled in two years that paired with a different male on a different territory each year and two males that were socially polygynous.

Blood samples were stored at 4 °C in 1 mL of Queen's lysis buffer (Seutin et al. 1991), and DNA was extracted by standard protocols (see also Woolfenden et al. 2005). Because of unpredictable degradation of nestlings' DNA (Conrad et al. 2000), 500µL blood-buffer aliquots were required in extractions to yield sufficient DNA template. We used touchdown polymerase chain reaction (TD-PCR) amplification of six hypervariable microsatellite loci. Two loci (EMIZ01, EMIZ27) were isolated from Least Flycatchers (Tarof et al. 2002), and four (SAP22, SAP32, SAP53, SAP66) were developed from Eastern Phoebes (Sayornis phoebe; Watson et al. 2002). The forward primer for SAP53 and reverse primer for SAP66 received an 18-bp M13 modification (TGTAAAACGACGGCCAGT). Locus SAP66 was not used the first year of the study.

We amplified genomic DNA in 10-µL TD-PCR reactions by using an Eppendorf Mastercycler Gradient thermal cycler under the following conditions: initial denaturation at 94 °C for 2 min; 13 cycles of 20 sec at 94 °C, 20 sec annealing (T_a; EMIZ1 60–54 °C; EMIZ27 65–59 °C; SAP22, 32, 53 67–61 °C; SAP66 65–60 °C), 30 sec at 72 °C (interval 1); 36 cycles of 20 sec at 94 °C, 20 sec T_a (EMIZ01 55 °C; EMIZ27 59 °C; SAP22 65 °C; SAP32 61 °C; SAP53 61 °C; SAP66 60 °C), 30 sec at 72 °C (interval 2); 11 cycles of 20 sec at 94 °C, 20 sec T_a (EMIZ01 53 °C; EMIZ27 58 °C; SAP22 64 °C; SAP32 60 °C; SAP53 60 °C; SAP66 59 °C), 30 sec at 72 °C (interval 3); followed by a final extension at 72 °C for 5 min. Each tube contained 3.6 µL water, 1.0 µL 10× TSG PCR buffer (Bio Basic, Inc.), 2.2 µL 25-mM MgCl₂ (MBI Fermentas), 0.2 µL 10-mM dNTPs, 0.2 µL, 10-µM fluorescently labeled Beckman Coulter WellRed forward primer (250 nm scale, HPLC purification; Integrated DNA Technologies), 0.2 µL 10-µM reverse primer, 0.08 µL 5 units µL^-1 TSG DNA polymerase (Bio Basic, Inc.), and 2.5 µL 1:10 dilution DNA (197.1 ± 187.4 ng mL^-1).

The PCR products were visualized on 1.0 % agarose gels stained with 5 µL ethidium bromide (5 mg mL^-1; BioShop Canada, Inc.) to determine amplification prior to fragment analysis. This protocol yielded dye-labeled amplicons suitable for genotype scoring in an automated DNA sequencer. Samples from adults were more likely to be amplified (70%–90% of individuals per locus) than those from nestling (41–91% individuals per locus). We determined individual genotypes by using the Beckman Coulter CEQ 8000 Genetic Analysis System with a 400-bp size standard, 99.5% deionized formamide (Sigma), and version 8.0 fragment-analysis software. Individuals were genotyped twice for repeatability. The six markers yielded high allelic variation (12–34 alleles), high levels of observed heterozygosity (0.61–0.99), and probabilities of exclusion (P_E) ranged from 0.652 to 0.903. The overall P_E was 0.999 with both social parents known.

To identify potential extra-pair young in nests of breeding pairs we first used CERVUS 3.0.3 for Windows (Kalinowski et al. 2007); the genotypes of suspected extra-pair young were then checked manually against the genotypes of their parents to confirm their extra-pair status. Offspring were typed at three to six loci, and extra-pair young (n = 16) were defined as those that mismatched the male social mate at two or more loci. We defined young as within pair (WPY) if they matched the social parents at all loci genotyped or had one paternal mismatch. We conservatively viewed single paternal mismatches (n = 7) with the attending male as lab artifacts. Next, we used CERVUS to identify candidates for true sires of these extra-pair young. We included the putative male (social father) and all other males sampled each year as candidate sires. We set parentage criteria identify the “most likely” and “second most likely” sires of each offspring. We used the following simulation parameters: 10 000 cycles (default); 75% of candidate sires in population sampled; minimum number of loci typed (2); 81.6% of loci typed (from allele frequencies); 1% genotyping error rate (default). We had no confirmed cases of intraspecific brood parasitism and so considered the female attending the nest as the genetic mother of all young sampled.

Statistical Analyses

We tested for normality and homogeneity of variance and found data were not normally distributed, so we used nonparametric tests in analyses. We used SPSS 16.0 for statistical analyses and express means with SDs. All tests were two-tailed, and we used α = 0.05 to evaluate statistical significance.

Results

Radiotracking and Paternity

None of the 12 fertile females that we radiotracked left their territory while being tracked (8.7 hr per female, 105 hr of tracking; Table 1, Fig. 1), even during the hour before dawn. Females typically remained near their nest (<30 m) and within the boundaries of the territory; they were usually out of sight when not visiting their nest. In three instances, females moved to the territory boundary but did not enter the neighbor's territory. During tracking, we observed five instances of aggressive interactions between two individuals other than our focal female, which typically concluded with an aggressive chase by one of the birds. The sex of the intruder, however, could not be determined because it arrived without prior vocalization, making its sex impossible to distinguish in this sexually monochromatic species.

In 2001 and 2002, Woolfenden et al. (2005) radiotracked nine male Acadian Flycatchers at various nest stages (nest building/egg-laying, incubation, and nestling) during the morning (07:00–12:00). They documented 24 off-territory forays in 54 hr of radiotracking, and males' foray rate averaged 0.4 ± 0.6 forays hr-1. In our study males' foray rate was significantly higher than females' (Mann—Whitney U-test, z = 3.2, P = 0.001). Woolfenden et al. (2005) found that six of nine males made off-territory forays, significantly higher than the 0 of 12 females that we observed making such forays (Fisher's exact test: P = 0.002), even though we tracked females twice as long as males and included periods before dawn and after noon.

Six of seven (86%) radiotracked females from whose nestlings DNA samples were available produced extra-pair young, and 11 of 14 nestlings (78.6%) were extra-pair (Table 1). We identified two extra-pair sires, which matched the extrapair young at all five loci and accounted for two extra-pair young of radiotracked females; these males held territories 1.0 and 1.2 km away from the female (Fig. 1). Of all nests sampled (n = 14), including those of untracked females, 64% contained at least one extra-pair young, and 16 of 28 (57%) young were the result of extra-pair fertilizations, a rate similar that found previously at this site (Woolfenden et al. 2005).

Vocalizations and Playback Experiments

During focal watches, fertile females vocalized 51% of their time and gave a median of 160 chiffs hr^-1 (range 21–685). Females' chiffing patterns in the fertile (nest building/egg laying) and incubation stages differed significantly (Fig. 2). The percentage of time spent chiffing (Mann—Whitney U-test, z = 3.6, P = 0.0), the number of chiffs per hour (z = 2.8, P = 0.005), and the average bout duration (z = 3.5, P = 0.0) were significantly higher during the fertile than during the nonfertile stages. Fertile females' number of chiffs per minute during a chiffing bout did not differ significantly from incubating females' (z = 0.5, P = 0.65; Fig. 2e). We also conducted a paired test (n = 13 females) to compare female behavior at various nest stages and found the same three variables (percent of time chiffing, chiffs per hour, bout length) to be highly significantly different (Wilcoxon paired signed-ranks tests, P < 0.01), whereas bouts per hour and chiffs per minute were not (P > 0.10).

Males were not radio-tagged during this study, so we could not quantify male movements with respect to female vocalizations. Typically, males continued to sing while their mate was chiffing and remained 30–40 m from the female, even when she was fertile. Similarly, we could not quantify the male intrusion rate because male Acadian Flycatchers are very difficult to see during off-territory forays; they remain in the subcanopy, move quickly, and do not vocalize (Woolfenden et al. 2005).

Female chiffing behavior during the fertile period was not significantly correlated with the percentage of extra-pair young in the nest (percent time chiffing: r_s = -0.01, n = 12, P = 0.78; bouts per hour: r_s = -0.33, n = 12, P = 0.29; bout length: r_s = 0.03, n = 12, P = 0.94). Because of nest predation, our samples were relatively small and power to test this relationship was low. Nevertheless, we can rule out a strong positive relationship and found it striking that two of the five females that chiffed >300 times per hour when fertile produced no extra-pair young.

We measured female vocalization rates and distance to the stimulus during playback of chiff calls on the territory. Chiff playbacks commenced at a distance of 40 m from the subject, and females responded by flying closer to the speaker, both when fertile 10.4 ± 4.3 m (range 5–20 m) and nonfertile 20.6 ± 16.8 m (range 3–40 m). The approach distance did not differ statistically by nest stage (Wilcoxon paired signed-rank test: z = 1.4, P = 0.153), but 3 of 10 incubating females did not approach the playback. We recorded four instances of fertile females flying over, and none of nonfertile females. We did not observe attempts at extra-pair copulation during chiff-playback experiments.

Prior to the playback, the chiff rate ranged from 4.5 to 7 min^-1 (Fig. 3a) for both fertile females and incubating females (off their nests), rates similar to those recorded during our longer observation periods (Fig. 2). The rate did not increase significantly during or after the playback (Fig. 3a) for fertile females (Friedman one-way ANOVA: χ² = 0.6, df = 2, P = 0.74) or nonfertile females (χ² = 4.7, df = 2, P = 0.10). The rate for fertile and incubating females did not differ significantly either during playback (Wilcoxon paired signed-ranks test: z = 0.9, P = 0.39) or after playback (z = 1.4, P = 0.16).

We found that (Fig. 3b) for both fertile (Friedman 1-way ANOVA: χ2 = 10.4, df = 2, P = 0.006) and incubating females (χ² = 10.0, df = 2, P = 0.007) wheeu calls increased significantly during playback. Fertile females did not produce this call significantly more often than incubating females during (Wilcoxon paired signed-ranks test: z = 0.9, P = 0.37) or after (z = 1.6, P = 0.11) the playback. Females never sang in response to playbacks of chiff calls. We heard females calling wheeu on only three occasions during our focal observations, indicating this call is rarely given under normal circumstances.

Discussion

Many studies have related extra-pair paternity to male or female characteristics (e.g., song, coloration, size, age; reviewed in Griffith et al. 2002), but few examine how individuals actually obtained those copulations (Kleven et al. 2006). Females make off-territory forays to visit extra-pair males in all passerines studied to date with radiotelemetry (Smiseth and Amundsen 1995, Neudorf et al. 1997, Double and Cockburn 2000, Tarof et al. 2005, Pedersen et al. 2006, Evans et al. 2008, Humbird and Neudorf 2008), so it surprising that female Acadian Flycatchers rarely, if ever, leave their territories during their fertile period. Most radiotracked females we studied obtained extra-pair fertilizations, which implies that extra-pair copulations were a result of male intrusions onto the territory.

Our radiotracking of fertile female Acadian Flycatchers sampled various times of the day and nearly doubled the tracking effort for males (Woolfenden et al. 2005), yet we observed no female trips off territory. In contrast, male Acadian Flycatchers left their territory 0.4 ± 0.6 times per hour and traveled up to 1500 m to visit other territories (Woolfenden et al. 2005). Off-territory forays are costly in terms of time, energy, and risk of injury (Stutchbury 1998, Norris and Stutchbury 2002, Pedersen et al. 2006); a female producing extra-pair young may also risk reduced parental care from her social mate (Dixon et al. 1994, Weatherhead et al. 1994, Hoi-Leitner et al. 1999). Social polygyny does occur in our population (<5%), and off-territory forays could be costly to females if they increase the probability that another female will settle onto the territory (Dale and Slagsvold 1995).

For males, the opportunity for sexual selection via extra-pair paternity can be high (Whittingham and Dunn 2005, Dolan et al. 2007, Webster et al. 2007), so in most species males' effort to compete for extra-pair copulations is also high (Stutchbury 1998). For female Acadian Flycatchers, however, the costs of seeking forays off territory presumably outweigh the benefits of gaining extra-pair copulations. Fertile females must devote time to nest building, and the energy required to make eggs likely increases the cost of off-territory forays over that for males. In other species female forays are typically restricted to nearby territories (Neudorf et al. 1997, Tarof et al. 2005, Evans et al. 2008), and in our study area this restriction would limit most female Acadian Flycatchers to visiting only one or two males because territories are often aligned linearly along streams (Fig. 1). Female Acadian Flycatchers that remain on their territory can nevertheless interact with potential extra-pair sires from a large pool of visiting males (Woolfenden et al. 2005). We suggest that the spatial distribution of territories has favored long-distance forays by males (Woolfenden et al. 2005), which in turn means that most females can interact with extra-pair males without making costly forays themselves. We predict limited off-territory forays by females of other socially monogamous species in which males make long-distance forays, such as riparian species like the Eastern Kingbird (Tyrannus tyrannus; Dolan et al. 2007). Females' forays may also be reduced in populations occupying fragmented habitat. In the Hooded Warbler, for example, in isolated forest fragments males make longer forays and females make more limited forays in comparison to those living in continuous forest where territories are close together (Norris and Stutchbury 2002).

Female Acadian Flycatchers could use chiff calls to advertise their location on a territory and fertility status to neighboring males. Some fertile female Acadian Flycatchers gave chiff calls hundreds of times per hour, and the pattern of calling of fertile and incubating females differed dramatically. In many ways, in the Acadian Flycatcher females' chiffing behavior is similar to what Neudorf et al. (2008) described in female Hooded Warblers, which initiate chip calls spontaneously away from their nest and in the absence of predators or humans. Our previous work on the Acadian Flycatcher showed that extra-pair males target intrusions onto territories where females are fertile (Woolfenden et al. 2005), and we suggest that the frequent chiff of the female conveys information, intentionally or not, about her location and fertility status to any male within earshot. Males from distant territories cannot monitor a female's fertility continuously, so loud and persistent chiffing could benefit females by increasing the frequency of visits by potential extra-pair sires. We did not find a correlation between females' chiff rate and extra-pair paternity, but a more direct test of the fertility-advertisement hypothesis would be to test if chiffing is correlated with the frequency of intrusions by radiotracked males (Stutchbury 1998).

Females continued to chiff even during incubation, when they are not fertile, suggesting that chiffing also has a territorial function. Females' territory defense against other females is relatively subtle and often difficult to observe, possibly explaining why it remains relatively understudied (Amundsen 2000). Female Acadian Flycatchers responded aggressively to playback of chiff calls on their territory by moving closer to the stimulus and by giving high-intensity wheeu calls. Females may be aggressive to other females to prevent them from usurping their territory or settling as a second female. It is also possible that females direct aggressive calls to other females to deter their mates from engaging in extra-pair copulations (Wagner 1992).

The Acadian Flycatcher's mating system is superficially similar to that of a congener, the Least Flycatcher, as both have high levels of extra-pair paternity and conspicuous female vocalizations (Tarof et al. 2005). However, in the Least Flycatcher females make frequent off-territory forays and territories are clustered, in response to opportunities for extra-pair mating, a situation termed a “hidden lek” (Wagner 1992). In contrast, Acadian Flycatcher territories are established in similar locations from year to year and are not clustered, perhaps because females do not pursue extra-pair copulations off territory. Interspecific differences in extra-pair tactics are not well understood because so few studies have carefully documented the behavior of both males and females or have compared populations with differing spatial arrangements of territories (Evans et al. 2009). Clearly it is critical to quantify females' extra-pair tactics in other species in order to understand the costs and benefits of female pursuit of extra-pair copulations, why off-territory foray behavior varies by species, and the extent to which female tactics determine male success in gaining extra-pair fertilizations.

Acknowledgments

We thank Gene Morton for his comments, advice, and sound recordings and for use of the Hemlock Hill Biological Research Area. Bonnie Woolfenden provided critical help early in the study. We also thank our field assistants Shidan Murphy and Victoria Kennedy, who trekked through the woods for many hours and counted thousands of individual chiff calls. This research was funded by a Discovery Grant, Research Tools and Instruments Grant, and Canada Research Chair Award to BJMS from the Natural Sciences and Engineering Research Council of Canada.

Literature Cited