Behavioral Ecology Advance Access originally published online on August 19, 2007
Behavioral Ecology 2007 18(6):985-993; doi:10.1093/beheco/arm068
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Extrapair paternity and the opportunity for sexual selection in a socially monogamous passerine
Department of Biology, P.O. Box 751, Portland State University, Portland, OR 97207, USA
Address correspondence to A. C. Dolan. E-mail: dolana{at}pdx.edu.
Received 26 October 2006; revised 1 June 2007; accepted 19 June 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
Effects of alternate mating strategies on the opportunity for sexual selection are widely debated, and recent studies have concluded that the effects of extrapair (EP) paternity on the opportunity for sexual selection may have been overstated due to 1) methodological limitations of empirical studies and 2) the potential for males to gain from additional within-pair (WP) reproductive opportunities. We therefore examined the impact of EP paternity on the opportunity for sexual selection in the socially monogamous and single-brooded eastern kingbird (Tyrannus tyrannus). EP paternity was common in all 3 years of our study (61% of 89 broods, 47% of nestlings) and realized reproductive success (EP + WP young) ranged from 0 to 9 young/male/year. A total of 31% of males lost all WP paternity (24% sired neither WP nor EP young, whereas 7% sired EP but not WP young), and variance in male realized reproductive success was more than 9 times greater than that of apparent reproductive success. Nearly half of EP mates were not nearest neighbors, and many were separated by 3 or more territories (>1000 m). EP success was independent of nest defense behavior, but early singing males and males with high song rates were most successful at both a population level and when cuckolders and cuckoldees were compared. EP paternity contributed significantly to the opportunity for sexual selection in kingbirds, and we suggest that this is probably due to the low potential for WP variation in reproductive success, apparent long-distance movements of one or both sexes, and consequent absence of reciprocal cuckoldry.
Key words: dawn song, eastern kingbird, extrapair paternity, opportunity for sexual selection, realized reproductive success, Tyrannus tyrannus.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
Sexual selection may profoundly influence the evolutionary history of a species and, among birds, is widely acknowledged as the basis for the striking size and plumage dimorphisms seen among polygamous and lek-breeding species (Darwin 1871
; Andersson 1994). The strength of sexual selection is directly associated with the difference between the sexes in the variance of reproductive success (Arnold and Wade 1984; Andersson 1994), where variance in reproductive success arises from differences in 1) number of social mates, 2) fecundity of social mates, 3) ability to guard within-pair (WP) paternity, and 4) success at obtaining extrapair (EP) fertilizations (Webster et al. 1995).
Ample evidence has shown that the number of young sired by males of facultatively polygynous species is dependent mostly on number of mates acquired (Webster et al. 1995; Forstmeier 2002; Whittingham and Dunn 2005). Hence, the influence of sexual selection on socially monogamous species was historically thought to be weak because variance in reproductive success was assumed to be low. However, EP paternity is now known to be common among many socially monogamous bird species (Westneat and Sherman 1997; Griffith et al. 2002), and a number of studies have proposed that even low levels of EP paternity can increase the opportunity for sexual selection among socially monogamous birds (Möller and Birkhead 1994; Möller and Ninni 1998; but see Dunn et al. 2001). However, the ability to assess variance in reproductive success requires a near complete sampling of the male population to avoid an inevitable and artificially low denominator (mean reproductive success) in the equation used to calculate standardized variance (Arnold and Wade 1984). Freeman-Gallant et al. (2005) suggested that this shortcoming characterized many of the published studies that have reported estimates of the opportunity for sexual selection, and they showed that a strong inverse relationship existed between estimates of the opportunity for sexual selection and the proportion of males sampled in the available field studies.
Moreover, for biological reasons, even very high levels of EP paternity need not increase variance in male reproductive success within socially monogamous species (Webster et al. 2001; Freeman-Gallant et al. 2005; Whittingham and Dunn 2005). Certainly, the opportunity for sexual selection will be high when a small proportion of males sire a majority of young (e.g., Dunn and Cockburn 1999), but sexual selection has little opportunity to act when EP gains are negated by WP losses as a result of reciprocal cuckoldry between males (i.e., gains = losses; Freeman-Gallant et al. 2005). Moreover, most studies that have identified EP sires place them within 1–2 territories of focal females, indicating that spatial limitations on movements constrain choice of EP sires to nearest neighbors (e.g., Kempenaers et al. 1997; Stutchbury et al. 1997; Langefors et al. 1998; Thusius et al. 2001; but see Woolfenden et al. 2005). WP components of reproductive success may also overshadow the EP components despite moderately high levels of EP paternity when a few males acquire second mates, or variance in female success is high, as for instance when the frequency of second broods differs considerably among females (Stutchbury et al. 1997; Forstmeier 2002; Whittingham and Dunn 2005). Thus, we currently lack strong evidence that EP paternity creates a significant opportunity for sexual selection in socially monogamous species.
The eastern kingbird (Tyrannus tyrannus) is a socially monogamous and sexually monochromatic suboscine species in which both parents contribute to the rearing of 3–4 young annually (Murphy 1996; Woodard and Murphy 1999). Based on standard measures, kingbirds fall well within the range of sexual size dimorphism found in other socially monogamous birds, yet male pectoral muscle is 30% larger than that of females (Murphy 2007). Except for the latter, few aspects of their biology suggest that sexual selection may be operating within this system, but roughly 60% of nests from 2 locations contained EP young (McKitrick 1990; Rowe et al. 2001). Kingbirds thus present an ideal opportunity to test for the potential influence of EP paternity on sexual selection in a socially monogamous species because 1) bigamy never occurs, 2) only one brood is raised annually, and 3) pairs remain together for an entire breeding season. Consequently, variance in male reproductive success is limited to their ability to secure WP paternity with a single female and gain EP fertilizations. Using 3 years of data, we partition the variance in male reproductive success to its WP and EP components to evaluate the opportunity for sexual selection and use morphological and behavioral data (nest defense and song) to identify characters that may serve as targets of female mate choice. We demonstrate that EP mating behavior created a strong opportunity for sexual selection and that the most successful EP sires were early singing males.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
Study site and species
Work was conducted between May and August of 2002 through 2004 at the northern end of the Great Basin Desert at Malheur National Wildlife Refuge (MNWR), Harney County, Oregon (43°N, 119°W; 1400 m at sea level). The Donner und Blitzen River runs through the refuge to create an island of riparian and marsh habitat surrounded by desert. The channelized river is narrow (5–10 m) and shallow (<2 m) and is paralleled closely by a gravel road (Center Patrol Road). Kingbirds are Nearctic–Neotropical migrants that arrive at MNWR in mid-May. Females build open-cup nests in late May in trees located along the river and egg laying begins around the first week of June.
Field methods
Survey methods are described in Redmond et al. (2007), but we note here that we are confident that nearly all pairs were located because of the openness of the habitat, restriction of nesting to the narrow riparian strip, and conspicuousness of kingbirds and their nests. All nest locations were recorded with a Garmin global positioning system 72 (±3–4 m). We uploaded the data points onto a PC using the program DNRGarmin (Minnesota DNR 2001), converted it to a shape file (ArcView 3.2a [ESRI Inc 2000]), and measured distances between nests with ArcView extension Nearest Features v. 3.8 (Jenness 2004).
Nests were visited at 2- to 3-day intervals to determine egg-laying dates, clutch size, hatching success, brood size, and fledging success (number of young to fledge). More than 50% of nests failed every year, but a majority of pairs renested, and identical data were collected for replacement nests. Adults were captured using mist nets placed near the nests during the nestling period and were banded with 1 aluminum US Fish and Wildlife service band and 3 colored plastic bands. On capture, we recorded body mass to the nearest 0.1 g (50- or 100-g Pesola scale) and measured relaxed wing chord to the nearest 0.5 mm (stopped wing rule) and tail and tarsometatarus (tarsus) lengths to the nearest 0.1 mm (dial calipers). A small blood sample (
50 µl) was taken from the brachial vein. Blood samples were taken from nestlings between 5 and 13 days of age, were immediately mixed with 1 ml of Longmire's Buffer (Longmire et al. 1988), and stored at 4 °C. Eggs that failed to hatch and nestlings that died in the nest were also collected and assayed to avoid loss of paternity information.
Genetic analysis
Parentage of nestlings was determined by comparing nestling and adult genotypes at 7 microsatellite loci. DNA was extracted from 100 µl of buffered blood with a Qiagen DNeasy extraction kit (Qiagen, Valencia, CA; #69504) using the protocol for whole-nucleated blood. DNA was amplified in 25 µl polymerase chain reactions (PCRs) using GE Healthcare PuReTaq Ready-to-Go PCR beads (GE Healthcare, Piscataway, NJ; #27-9558-01) following conditions specified in Table 1. Primers were labeled with fluorescent tags and microsatellites were visualized and interpreted using fluorescent detection (ABI Genescan; Applied Biosystems, Foster City, CA).
|
Maternity was confirmed by direct comparison of mother and offspring genotypes. Paternity was assessed by direct exclusionary analysis using the nestling's nonmaternal genotype. Putative mothers matched nestlings at every locus, indicating that intraspecific brood parasitism did not occur and that mutation and nonamplyfying alleles did not influence paternity assignment. Nestlings with at least one mismatch with their social father were considered EP, and their genotypes were compared with all other sampled males in the population to identify the genetic father. To be assigned paternity, a nestling and male had to match at all 7 loci. We used CERVUS 2.0 to calculate average exclusion probabilities at each locus and at all loci combined (Marshall et al. 1998
; Slate et al. 2002). Because genetic mothers of the nestlings were known, this represents the probability that a randomly chosen male could be excluded as the genetic father of a nestling. In addition, we assessed deviations from Hardy–Weinberg equilibrium and the probability of null alleles using CERVUS 2.0 that performs a chi-square test using Yates correction for continuity when there is only 1 degree of freedom (df). The Hardy–Weinberg test combined alleles when the expected frequencies were small. In one case this created 1 allelic class, and in this instance we ran a standard chi-square test (Brookfield 1996).
Annual adult return rates were 65% (Murphy MT, unpublished data), and 19 males and 15 females contributed parentage data in more than 1 year. From a statistical perspective, it is appropriate to exclude all but one observation per individual, but from a biological perspective, excluding males from a limited pool of sires, specifically those who survived to breed more than once, seemed inappropriate because these were likely to be high-quality males (as evidenced by their return). Excluding these individuals was likely to create a bias against detecting evidence of an opportunity for sexual selection. For this reason and the fact that repeatability (perLessells and Boag 1987) of EP, WP, and total reproductive success of males (analysis of variance [ANOVA]: P = 0.14, P = 0.22, and P = 0.07, respectively) and females (ANOVA: P = 0.29, P = 0.17, and P = 0.22, respectively) were all nonsignificant, we chose to include all nesting attempts in our analyses. Nonetheless, we recognized the potential statistical problems arising from pseudoreplication and therefore also conducted all analyses with and without males who were present in more than 1 year and report results of tests when the omission of the latter males caused a significant result (P 
0.05) to become nonsignificant. We randomly sampled 1 year for each male, except for cases where incomplete data existed for one of the years.
Murphy (2004) showed that tarsus and bill lengths were significantly repeatable across years. Thus, if we obtained parentage data but lacked morphological data for a male in a given year, we used morphological measurements from the previous year in our analyses of the relationship of WP and EP success to morphology if they were available. Wing chord, although repeatable (Murphy 2004), may change with nutritional history or age, and thus, we did not use wing chord, body mass, nest defense, or song data (see below) in analyses unless they were collected in the same year for which parentage data were available.
Parental defense and song behavior
Kingbirds are notoriously aggressive nest defenders (Davis 1941) and males are more aggressive than females (Davis 1941; Redmond 2005). Therefore, in 2003 and 2004, we documented parental defense behavior to test for relationships between nest defense and WP losses and EP gains of paternity. We used the method of Blancher and Robertson (1982) for quantifying nest defense by measuring responses to the presentation of a taxidermic mount of an American crow (Corvus brachyrhynchos), a nest predator at our study site, once during the incubation period and once when the nestlings were 6–8 days old (between 0900 and 1200 Pacific Daylight Time). To measure nest defense, the mounted crow was attached to a 3-m pole and held by one person within 1 m of the nest for a 5-min test period. Two observers, one for each parent, recorded responses on a scale of 1–5 (1 = call, 2 = approach mount, 3 = hover near mount, 4 = dive at mount, and 5 = strike mount). All birds called and approached, so we used only the number of hovers, dives, and strikes to derive a nest defense index (NDI) score for males and females (NDI = log [(
hover + 1) x ( dive + 1) x ( strike + 1)]; 1 was added to each score to account for zeros). Individual measures of NDI were repeatable within and between years (Redmond 2005).
Kingbirds vocalize throughout the day, but once females arrive, males sing a highly ritualized dawn song (Smith 1966) in the predawn darkness for the entire breeding season from a prominent perch located within 50 m of the nest (Sexton et al. forthcoming). Copulations virtually never occur during hours of daylight (Smith 1966, Murphy MT, personal observation), and we assume that they occur during the dawn song period. Presumably, at least one function of dawn song is to advertise a male's availability to females. In 2003 and 2004, we therefore documented dawn song behavior to assess its possible role in WP and EP choice of sires. The narrow daily window of time in which dawn songs are delivered required a team of 3–5 people to document the behavior, which we did nearly daily from mid-June to late July. We arrived at nests about 2 h before dawn and recorded the time that males began to sing (start time), when they ended (end time), and total song length (length; absolute value of difference between start and end time relative to civil dawn). Nearly all songs were produced within the first 30 min of singing (Sexton et al. forthcoming), and therefore, the number of songs per minute was recorded for the first half-hour of the bout. Not all males sang for the full 30 min, and therefore, we calculated 3 song rates: "30-MinRate" = average song rate per minute for 30 min; "ActualRate" = average song rate per minute for the period of singing; "Peak Rate" = average song rate per minute for the 5-min period of highest output. All males were observed at least twice, and start time and the 3 measures of song rate were highly repeatable for individuals both within and between breeding seasons (Murphy MT, Sexton K, Dolan AC, Redmond LJ, unpublished data). We therefore used the average values for each song behavior from the multiple observations (2–5) of each male in our analyses.
Statistical analyses
To determine whether EP young were randomly distributed among nests, we compared observed numbers of EP young with expected values using a chi-square test. Expected numbers of EP young were computed using a hypergeometric distribution based on population-level variation in clutch size and number of EP and WP young using a SAS script (Neuhauser et al. 2001). All other analyses were performed using STATISTIX 8.0 and SPSS 11.5. Standardized variance in reproductive success (variance/mean2) was computed according to Arnold and Wade (1984), and we used equations from Webster et al. (1995; Table 1) to calculate components of variance and covariance in WP and EP success.
We used least squares linear regression and stepwise multiple linear regression to test for relationships between 3 measures of reproductive success (number of WP, EP, and total young) and male morphology (mass, tarsus, wing chord, bill, and tail lengths), behavior (NDI during incubation and nestling period, and song variables), and ecological circumstance (timing of breeding, breeding density, and breeding synchrony). Timing of breeding was defined as the date on which a female laid her first egg, whereas breeding density was measured as the distance to the nearest neighbor and average distance to the 3 nearest neighbors. Stutchbury and Morton (1995) proposed that EP paternity would be more common among synchronously breeding species because females can better compare males when they are in the same breeding state. Hence, we calculated the breeding synchrony index (Kempenaers 1993) for each female. Unless otherwise stated, all variables retained in the multiple regressions contributed significantly (P 0.05, based on Type III sums of squares) and we report adjusted R2. Pairwise comparisons of morphological and behavioral characters between successful EP males (i.e., cuckolders) and the WP males who lost paternity to them (i.e., cuckoldees) were made using paired t-tests. Sample sizes vary among tests because not all data were available for all males.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
Paternity results
Population size was 52, 59, and 52 pairs in 2002, 2003, and 2004, respectively. Initial nests often failed (64% of 163 nests), mostly due to nest predation (57% of 163 nests). Many pairs who failed also renested (61%) and nearly half (48.4%) bred successfully. The 7 loci yielded 81 alleles (Table 1), resulting in a total exclusionary power of 0.998. The predicted frequency of null alleles at the 7 loci was low, every nestling matched the social female at every locus, and no deviations from Hardy–Weinberg were detected (CERVUS: Marshall et al. 1998; Table 1). Sires were identified for 73%, 76%, and 85% of the offspring sampled in 2002, 2003, and 2004, respectively. Most EP young (246 of 264; 93%) mismatched their social father at 2 or more loci, but 7% of nestlings exhibited only one mismatch (3 in 2002 and 2004 and 12 in 2003). Given the lack of evidence for heterozygote deficiency (the usual result if null alleles exist [Pemberton et al. 1995
]), these nestlings were deemed EP.
In all 3 years, at least 59% of nests contained one or more EP young (Table 2), and no differences existed among years (2 x 3 contingency table; P = 0.958; n = 89). Pooled across years, 61% of nests contained at least one EP nestling and 47% of all nestlings were sired by EP males (Table 2). There was no difference in the frequency of EP young in first (34 of 59) and replacement nests (20 of 30; Fisher's exact test, P = 0.494). EP young were not randomly distributed among nests: more nests contained all or no EP young than expected by chance (
2 = 35.47, P < 0.001, df = 14; Figure 1). Distance between females and EP sires ranged from 67.2 to 15 359 m (
= 1779.8 m, median = 404.1, standard deviation [SD] = 3246.7, n = 39), and at least one territory separated nearly half (18/40) of EP mates. Nearest neighbor distances averaged between 250 and 300 m in all years, and of the EP partners separated by at least one territory, distance between partners averaged 3517 m (SD = 4357.9, n = 17).
|
|
Nearly a third of males (27 of 89) gained EP fertilizations, but more than half of these males also lost paternity within their own nests (15 of 27). Of the males who sired no EP young (62 of 89), nearly a quarter either sired no WP young (21 of 89) or all WP young (23 of 89), whereas the remainder sired only a portion of their social mate's brood (18 of 89). Among years, 16–36% (mean = 24%) of the males sired no young (either WP or EP) and between 20% and 44% (mean = 31%) of males sired no offspring within their nest (27 of 89). The most successful males sired EP young and at least some (9 of 89) or all (12 of 89) WP young. Males that sired EP young, and at least some WP young, averaged 91% higher genetic reproductive success than males who sired all WP young but no EP young (23 of 89).
Opportunity for sexual selection
Apparent male reproductive success (the number of young in a male's nest that survived until blood sampling) ranged from 1 to 5 nestlings (mean = 2.97). The number sired (realized success) averaged just more than 2 (annual range: 1.8–2.5) but varied between 0 and 9 (Table 3). EP success accounted for 46% of the standardized variance in reproductive success (annual range: 36–48%), whereas WP contributions accounted for 42% (annual range: 31–51%). Standardized covariance of WP and EP success was moderate and positive (12% of the total variance, Table 3). The variance in realized male reproductive success averaged over the 3 years was 9.4 times greater than for apparent reproductive success (Fmax test, P < 0.001), whereas among years it ranged from 7.3 to 15.6, indicating that EP paternity contributed substantially to overall genetic reproductive success and that an opportunity for sexual selection existed in all years. Removal of males who were sampled more than once resulted in the same conclusions: variance in realized reproductive success was 6.2 times greater than apparent reproductive success.
|
The largest contribution to the variance in genetic reproductive success of EP males was the number of EP mates gained by a male (33% of the standardized variance of male success; Table 4). Nearly all the variance in male WP success was determined by the proportion of a female's clutch that he sired (Table 4). Female quality (=number of eggs produced) contributed very little to variance in reproductive success (Table 4), which is not surprising because 89% of females laid either 3 or 4 eggs (n = 199 first and replacement clutches).
|
Predictors of WP and EP success: breeding date, breeding synchrony, and density
Among first clutches, the social mate of males who sired EP young initiated their WP clutches approximately 5.5 days before males who failed to sire EP young (F = 7.31, P = 0.009, n = 54). Consequently, the breeding synchrony index of the social mates of successful EP males was lower than that of males who failed to sire EP young (F = 5.7, P = 0.02, n = 53). Excluding the males sampled in a second year resulted in qualitatively similar results that were not statistically significant (breeding date, F = 1.97, P = 0.168; synchrony index, F = 1.59, P = 0.218; n = 31 for both). The occurrence of EP young was unrelated to nest density as the distance to the nearest nest did not differ between males who did (
= 313.8, SD = 374.7, n = 53) or did not ( = 275.7, SD = 164.0, n = 34; t = 0.56, P = 0.52) lose paternity. Similarly, nearest neighbor distance of males who did ( = 261.0, SD = 158.8, n = 28) or did not ( = 361.9, SD = 359.1, n = 59) sire EP young did not differ (t = 1.01, P = 0.32). Identical conclusions resulted for both the latter comparisons when we used the average distance to the 3 nearest neighbors and when males sampled in more than 1 year were excluded (results not shown).
Predictors of WP and EP success: morphology and behavior
Correlation analyses of WP success and morphology of individual males failed to detect any significant relationships (strongest correlation with wing chord: r = 0.154, n = 87, P = 0.154). However, EP success was significantly and positively correlated with tarsus length (Figure 2, r = 0.374, n = 87, P = 0.0004). Realized success (WP + EP) was also significantly correlated with tarsus length for the entire sample of males (r = 0.220, n = 87, P = 0.040) but not after we omitted males that returned to breed in more than 1 year (r = 0.032, P = 0.925, n = 62).
|
WP success was positively associated with NDI during incubation (r = 0.342, n = 43, P = 0.025; without second sample for returning males, r = 0.295, P = 0.065, n = 38) but not to NDI during the nestling period (r = 0.185, n = 38, P = 0.270) or to song behaviors (for all, P
0.11). Number of EP young was unrelated to NDI score during either time period (P 0.29), but number of EP young correlated inversely with StartTime (r = –0.508, n = 55, P < 0.001; Figure 3) and positively with 30-MinRate (r = 0.307, n = 55, P = 0.023; without second sample for returning males, r = 0.265, P = 0.083, n = 44). Thus, males who began singing early relative to sunrise sired the most EP young. Among males who sired at least one EP offspring, timing of dawn song accounted for 65.4% of the variation in EP success (P < 0.001, n = 18; Figure 3).
|
Thus, timing of dawn song appeared to be particularly important to EP success, but timing of song was correlated with other song and morphological traits. Males who began singing early tended to sing at a higher rate (r = 0.569, P < 0.001, n = 57), for longer periods (r = 0.613, P < 0.001, n = 57), and had longer tarsi (r = 0.302, P < 0.024, n = 57; without the returning males, r = 0.086, P = 0.362, n = 43). We therefore conducted a stepwise multiple regression to evaluate potential contributions of morphological, song, and nest defense variables to EP success. Timing of dawn song was the first variable to enter the regression of EP (r = 0.401, P = 0.014, n = 39) and total reproductive success (r = 0.318, P = 0.023, n = 39), and after accounting for its effect, no other variable contributed significantly. A similar multiple regression of WP success showed that WP success was related only to NDI during incubation (r = 0.285, P = 0.043, n = 39), but this result became nonsignificant when second samples for returning males were omitted (r < 0.001, P = 0.18, n = 28).
Comparisons of cuckolder and cuckoldee
Pairwise comparisons of successful EP sires (cuckolders) to the males from whom they gained paternity (cuckoldees) suggested that song behavior and morphology differed between groups. Morphological comparisons of the 40 pairs for whom data for both males were available indicated that cuckolders had longer tarsi (and possibly longer tails; Table 5). Behavioral differences for pairs with song data available for both males were more pronounced (Table 5). Cuckolders started singing earlier and sang faster than cuckoldees. On the other hand, nest defense behavior (NDI score) did not differ between cuckolders and cuckoldees (Table 5).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
General characteristics
On average, EP young occur in 16% of nests of socially monogamous bird species breeding in north temperate regions (Westneat and Sherman 1997
; Griffith et al. 2002; Spottiswoode and Möller 2004). EP paternity is most common among migratory passerines (Arnold and Owens 2002), but even for this ecological group, eastern kingbirds are an extreme case; 61% of nests contained EP young and 47% of all offspring were the result of EP matings (Table 2). Virtually identical values have been reported for eastern kingbird populations from Michigan (McKitrick 1990) and New York (Rowe et al. 2001), indicating that eastern kingbirds consistently engage in cryptic polygamy.
EP fertilizations were not evenly distributed among broods (see also Rowe et al. 2001) or among males. This was the case in all 3 years and it created big winners and big losers, and as a consequence, EP paternity accounted for nearly half of the variance in male reproductive success. The skew in reproductive success is strikingly exemplified by the fact that 31% of males provisioned broods in which they had no genetic investment (range among years, 20–44%). Rowe et al. (2001) likewise reported that 30% of male kingbirds failed to sire any young in nests from New York. The large contribution of EP paternity to variance in male reproductive success underscores the point that EP mating behavior is not merely an alternative mating strategy involving a minority of males but is instead a primary selective force that defines the eastern kingbird mating system.
High rates of nest failure are also characteristic of kingbirds and other open-cup nesting species for which parentage data have been reported (e.g., Webster et al. 2001; Whittingham and Dunn 2005). High offspring mortality prior to sampling due to nest predation may increase the WP component of variance in reproductive success and, when unaccounted for, may possibly lead to an increase in the apparent importance of the EP components of variance in reproductive success. However, Whittingham and Dunn (2005) pointed out that unless either the WP or the EP components of reproductive success are differentially affected by offspring mortality prior to sampling, their relative contributions to overall variance in reproductive success should be unaffected. This is likely the case in kingbirds as Murphy (2004) showed that nest predation was not a repeatable individual character but instead occurred stochastically throughout the population. Comparison of the opportunity for sexual selection that we measured for kingbirds (9.4) to estimates made for other temperate zone breeding and open-cup nesting passerines (range = 1.4–1.7; see Whittingham and Dunn 2005), all of whom experience frequent nest predation, suggests strongly that EP mate choice created a significant opportunity for sexual selection in kingbirds.
Whittingham and Dunn (2005) and Freeman-Gallant et al. (2005) argued convincingly that earlier reports (see Introduction) of the opportunity for sexual selection among facultatively polygynous and socially monogamous species may have been overestimated, in part because of the inadequate sampling of the male population (Freeman-Gallant et al. 2005). We sampled roughly 75% or more of the male population in every year, and therefore, our estimate of the opportunity for sexual selection was unaffected by this concern. Why then does EP paternity create a high opportunity for sexual selection in kingbirds when equally high or higher levels of EP paternity in other species does not (Forstmeier 2002; Freeman-Gallant et al. 2005; Whittingham and Dunn 2005)? We attribute the differences to at least 3 features of kingbird biology. First, nearly 90% of females lay just 3 or 4 eggs, and none raise second broods. Second, bigamy has never been seen in 22 years of field work (Murphy MT, personal observation), and therefore, males have no opportunity to acquire more than one mate. Thus, there is little potential for high WP variability in reproductive success among socially bonded pairs. The third factor is that kingbirds do not limit EP mating opportunities to nearest neighbors. Nearly half of EP fertilizations occurred between individuals that were separated by at least 1 territory, and a surprisingly large number involved males and females separated by 3 or more territories (20%). We do not know whether males, females, or both sexes move to acquire copulations, but like Acadian flycatchers (Empidonax virescens; Woolfenden et al. 2005), interactions of male and female kingbirds are not limited to nearest neighbors. Hence, reciprocal exchange of paternity among neighboring males is not common, and this opens the door for some males to achieve very high success. The latter does not appear to be the case in the majority of other species studied to date (see Introduction).
Correlates of realized (WP + EP) success
WP mating success was unrelated to morphology or song behavior, but a low proportion (10%) was associated with nest defense behavior: defensive males tended to sire more WP young. Why this was the case is unclear, but further study is necessary to establish why defensive (i.e., parental) males tended to sire more of their social partner's young.
In contrast, EP success was unrelated to male nest defense but correlated with song performance and male morphology. Timing was especially important, with the most successful males being those who began their daily dawn song the earliest. Poesel et al. (2006) reported similar findings for blue tits (Parus caeruleus). Data presented elsewhere (Murphy MT, Sexton K, Dolan AC, Redmond LJ, unpublished data) establish that breeding date and timing of the initiation of kingbird dawn song were positively correlated (i.e., males paired to late breeding females began singing closer to dawn), and in other species, early spring arrival has been associated with high male quality, strongly expressed secondary sexual characters, and mating success (Lozano et al. 1996; Möller et al. 2003). And although we have no information on arrival and breeding dates for the years of our study, data from subsequent years confirm that early arriving male kingbirds are the first to form pair bonds and breed (Cooper NW, Murphy MT, unpublished data). We suggest therefore that the earliest singers in the predawn period were high-quality individuals. Our data also show that early and rapidly singing males had long tarsi, wings, and tails (Murphy MT, Sexton K, Dolan AC, Redmond LJ, unpublished data), and we propose that song provided reliable information on male size and overall quality.
Kingbirds are suboscine passerines and song is presumably innate (Kroodsma 1984; Kroodsma and Konishi 1991). From our experience, song varies little in structure among males (Dolan AC, personal observation), and the question of how song provides reliable information on male quality becomes relevant. We suggest that it is through aspects of performance (timing and rates) that are energetically demanding. Laboratory estimates of the metabolic cost of singing vary, but they are high in some species (e.g., Eberhardt 1994; but see Oberweger and Goller 2001). Feeding experiments have also demonstrated enhanced song performance among supplemented males (Gottlander 1987; Cuthill and Macdonald 1990; Berg et al. 2006). Kingbirds initiate dawn song after a 6- to 7-h fast and at a time when air temperatures average between 4 ° and 8 °C (Dolan AC, unpublished data), well below the kingbird's thermoneutral zone (Yarbrough 1971). Dawn song thus takes place at a point when kingbirds are probably at a daily energetic low, and we propose that it provides honest information on male quality.
Timing of dawn song and song rates of kingbirds vary considerably among individuals and are significantly repeatable both within and between seasons (Murphy MT, Sexton K, Dolan AC, Redmond LJ, unpublished data). Hence, dawn song combines the properties of high variability and high repeatability that are essential for conveying individually specific details regarding singers. Pairwise comparisons of cuckolder and cuckoldee provided additional evidence of the importance of song for EP mating success: cuckolders outperformed cuckoldees in 4 of 5 song behaviors (Table 5). Suboscine song is free of the effects of learning on song development (Kroodsma 1984; Kroodsma and Konishi 1991), and song performance presumably integrates a male's developmental history and current nutritional state. Dawn song performance may thus serve as a measure of quality that either females use to seek EP partners for indirect genetic benefits or it may reflect a male's vigor and capacity to acquire EP matings. Analyses of EP mating behavior for female birds indicate that little evidence for indirect benefits exist (Griffith et al. 2002; Arnqvist and Kirkpatrick 2005) and are outweighed by direct costs (e.g., loss of paternal care; Arnqvist and Kirkpatrick 2005). Hence, the more likely explanation may be that selection on males for EP mate seeking is the driver in this system. We do not know if it is males, females, or both sexes that search for EP mates, but high-quality males, either through coercion of females or domination of other males, gain access to fertile females. This does not rule out the possibility that females seek EP matings, but her behavior may emerge as an epiphenomenon of strong selection on males (Arnqvist and Kirkpatrick 2005). Much more work is needed to measure both indirect benefits and direct costs simultaneously in a single population to test the hypothesis that EP mating behavior in females is best viewed as making the "best of a bad job" (Westneat and Stewart 2003).
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
American Ornithologists' Union and the American Museum of Natural History (A.C.D.).
|
ACKNOWLEDGEMENTS |
|---|
We thank Cal and Alice Eltshoff for allowing us to intrude on their lives for more than 2 months each year. We also thank the staff at Malheur National Wildlife Refuge, in particular Rick Roy, for their support and Kelly Hoffman and Rick Ernst for assisting in the collection of the song data. Two anonymous reviewers made very helpful comments on earlier versions of the manuscript.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONFUNDINGREFERENCES |
|---|
Andersson M. Sexual selection (1994) Princeton (NJ): Princeton University Press.
Arnold KE, Owens IPF. Extra-pair paternity and egg dumping in birds: life history, parental care and the risk of retaliation. Proc R Soc Lond B (2002) 269:1263–1269.[Medline]
Arnold SJ, Wade MJ. On the measurement of natural and sexual selection: applications. Evolution (1984) 38:720–734.[CrossRef][Web of Science]
Arnqvist G, Kirkpatrick M. The evolution of infidelity in socially monogamous passerines: the strength of direct and indirect selection on extrapair copulation behavior in females. Am Nat (2005) 165:S26–S37.[CrossRef][Web of Science][Medline]
Berg ML, Beintema NH, Welbergen JA, Komdeur J. Singing as a handicap: the effects of food availability and weather on song output in the Australian reed warbler Acrocephalus australis. J Avain Biol (2006) 36:102–109.
Blancher PJ, Robertson RJ. Kingbird aggression: does it deter predation? Anim Behav (1982) 30:929–930.[CrossRef][Web of Science]
Brookfield JFY. A simple new method for estimating null allele frequency from heterozygote deficiency. Mol Ecol (1996) 5:453–455.[CrossRef][Medline]
Cuthill IC, Macdonald WA. Experimental manipulation of the dawn and dusk chorus in the blackbird Turdus merula. Behav Ecol Sociobiol (1990) 26:209–216.[Web of Science]
Darwin C. The descent of man and selection in relation to sex (1871) 2nd ed. Chicago: WB Conkey Company.
Davis DE. The belligerency of the kingbird. Wilson Bull (1941) 53:157–168.
Dunn PO, Cockburn A. Extrapair mate choice and honest signaling in cooperatively breeding superb fairy-wrens. Evolution (1999) 53:938–946.[CrossRef][Web of Science]
Dunn PO, Whittingham LA, Pitcher TE. Mating systems, sperm competition, and the evolution of sexual dimorphism in birds. Evolution (2001) 55:161–175.[CrossRef][Web of Science][Medline]
Eberhardt LS. Oxygen-consumption during singing by male carolina wrens (Thryothorus ludovicianus). Auk (1994) 111:124–130.[Web of Science]
Environmental Systems Research Institute. ArcView v. 3.29 (2000) Redlands (CA): ESRI Inc.
Forstmeier W. Factors contributing to male mating success in the polygynous dusky warbler (Phylloscopus fuscatus). Behaviour (2002) 139:1361–1381.[CrossRef]
Freeman-Gallant CR, Wheelwright NT, Meiklejohn KE, States SL, Sollecito SV. Little effect of extrapair paternity on the opportunity for sexual selection in savannah sparrows (Passerculus sandwichensis). Evolution (2005) 59:422–430.[CrossRef][Web of Science][Medline]
Gottlander K. Variation in song rate of the male pied flycatcher Ficedula hypoleuca: causes and consequences. Anim Behav (1987) 35:1037–1043.[CrossRef][Web of Science]
Griffith SC, Owens IPF, Thuman KA. Extra pair paternity in birds: a review of interspecific variation and adaptive function. Mol Ecol (2002) 11:2195–2212.[CrossRef][Medline]
Jenness J. Nearest features (nearfeat.avx) extension for Arcview 3.x, v. 3.80 [Internet] (2004) Flagstaff (AZ): Jenness Enterprises. Available from: http://www.jennessnt.com/arcview/Nearestfeatures.html.
Kempenaers B. The use of a breeding synchrony index. Ornis Scand (1993) 24:84–84.[CrossRef]
Kempenaers B, Geert RV, Dhondt AA. Extrapair paternity in the blue tit (Parus caeruleus): female choice, male characteristics and offspring quality. Behav Ecol (1997) 8:481–492.
Kroodsma DE. Songs of the alder flycatcher (Empidonax alnorum) and willow flycatcher (Empidonax trailii) are innate. Auk (1984) 101:13–24.[Web of Science]
Kroodsma DE, Konishi M. A suboscine bird (eastern phoebe Sayornis phoebe) develops normal song without auditory feedback. Anim Behav (1991) 42:477–488.[CrossRef][Web of Science]
Langefors A, Hasselquist D, Torbjörn S. Extra-pair fertilizations in the sedge warbler. J Avian Biol (1998) 29:134–144.[CrossRef]
Lessells CM, Boag PT. Unrepeatable repeatabilities: a common mistake. Auk (1987) 104:116–121.[Web of Science]
Li SH, Brown JL. High frequency of extrapair fertilization in a plural breeding bird, the Mexican jay, revealed by DNA microsatellites. Anim Behav (2000) 60:867–877.[CrossRef][Web of Science][Medline]
Longmire JL, Lewis AK, Brown NC, Buckingham JM, Clark LM, Jones MD, Meincke LJ, Meyne L, Ratliff RL, Ray FA. Isolation and molecular characterization of a highly polymorphic centromeric tandem repeat in the family Falconidae. Genomics (1988) 2:14–24.[CrossRef][Medline]
Lozano GA, Pereault S, Lemon RE. Age, arrival date and reproductive success of male American redstarts Setophaga ruticilla. J Avian Biol (1996) 27:164–170.[CrossRef]
Marshall TC, Slate J, Kruuk LEB, Pemberton JM. Statistical confidence for likelihood-based paternity inference in natural populations. Mol Ecol (1998) 7:639–655.[CrossRef][Medline]
McKitrick MC. Genetic evidence for multiple parentage in eastern kingbirds Tyrannus tyrannus. Behav Ecol Sociobiol (1990) 26:149–156.[Web of Science]
Minnesota Department of Natural Resources. DNRGarmin 4.2.1 (2001) St Paul (MN): Minnesota DNR.
Möller AP, Birkhead TR. The evolution of plumage brightness in birds is related to extrapair paternity. Evolution (1994) 48:1089–1100.[CrossRef][Web of Science]
Möller AP, Brohede J, Curevo JJ, deLope F, Primmer C. Extrapair paternity in relation to sexual ornamentation, arrival date, and condition in a migratory bird. Behav Ecol (2003) 14:707–712.
Möller AP, Ninni P. Sperm competition and sexual selection: a meta-analysis of paternity studies of birds. Behav Ecol Sociobiol (1998) 43:345–358.[CrossRef][Web of Science]
Murphy MT. Eastern Kingbird (Tyrannus tyrannus). In: The Birds of North America, No. 253. Philadelphia (PA): The Academy of Natural Sciences—Poole A, Gill F, eds. (1996) Washington (DC): The American Ornithologists' Union. 1–23.
Murphy MT. Intrapopulation variation in reproduction by female eastern kingbirds Tyrannus tyrannus: the impacts of age, individual performance, and breeding site. J Avian Biol (2004) 35:252–261.[CrossRef]
Murphy MT. A cautionary tale: cryptic sexual size dimorphism in a socially monogamous passerine. Auk (2007) 124:515–525.[CrossRef][Web of Science]
Neuhauser M, Forstmeier W, Bretz F. The distribution of extra-pair young within and among broods—a technique to calculate deviations from randomness. J Avian Biol (2001) 32:358–363.[CrossRef]
Oberweger K, Goller F. The metabolic cost of birdsong production. J Exp Biol (2001) 204:3379–3388.
Pearson T. Polygyny and extra-pair paternity in a population of southwestern willow flycatchers (Empidonax traillii extimus) (2002) Flagstaff (Arizona): Northern Arizona University.
Pemberton JM, Slate J, Bancroft DR, Barrett JA. Nonamplifying alleles at microsatellite loci: a caution for parentage and population studies. Mol Ecol (1995) 4:249–252.[Medline]
Poesel A, Hangsjorerg PK, Foerster K, Johnsen A, Kembenaers B. Early birds are sexy: male age, dawn song and extrapair paternity in blue tits Cyanistes (formerly Parus) caeruleus. Anim Behav (2006) 72:531–538.[CrossRef][Web of Science]
Redmond L. Breeding habitat selection, nest reuse and nest defense of Eastern Kingbirds in SE Oregon (2005) Portland (Oregon): Portland State University.
Redmond LJ, Murphy MT, Dolan AC. Nest reuse by Eastern Kingbirds: adaptive behavior or ecological constraint? Condor (2007) 109:463–468.
Richardson DS, Jury FL, Dawson DA, Salgueiro P, Komdeur J, Burke T. Fifty Seychelles warbler (Acrocephalus sechellensis) microsatellite loci polymorphic in Sylviidae species and their cross-species amplification in other passerine birds. Mol Ecol (2000) 9:2226–2231.[Medline]
Rowe DL, Murphy MT, Fleischer RC, Wolf PG. High frequency of extra-pair paternity in Eastern Kingbirds. Condor (2001) 103:845–851.
Sexton K, Murphy MT, Radmond LJ, Dolan AC, Forthcoming. Dawn song of eastern kingbirds: intrapopulation variability and sociobiological correlates. Behaviour.
Slate J, Visscher PM, MacGregor S, Stevens D, Tate ML, Pemberton J. A genome scan for quantitative trait loci in a wild population of red deer (Cervus elaphus). Genetics (2002) 162:1863–1873.
Smith WJ. Communications in the genus Tyrannus (1966) Cambridge (MA): Publication of the Nuttall Ornithological Club No. 6.
Spottiswoode C, Möller AP. Extrapair paternity, migration, and breeding synchrony in birds. Behav Ecol (2004) 15:41–57.
Stutchbury BJM, Morton ES. The effect of breeding synchrony on extra-pair mating systems in songbirds. Behaviour (1995) 132:675–690.
Stutchbury BJM, Piper WH, Neudorf DL, Tarof SA, Rhymer J, Fuller G, Fleischer RC. Correlates of extra-pair mating success in hooded warblers. Behav Ecol Sociobiol (1997) 40:119–126.[CrossRef][Web of Science]
Tarof SA, Ratcliffe LM, Boag PT. Polymorphic microsatellite loci for assigning parentage in least flycatchers (Empidonax minimus). Mol Ecol Notes (2001) 1:146–148.[CrossRef][Web of Science]
Thusius KJ, Peterson KA, Dunn PO, Whittingham LA. Male mask size is correlated with mating success in the common yellowthroat. Anim Behav (2001) 62:435–446.[CrossRef][Web of Science]
Watson CJW, Beheler AA, Rhodes OE. Development of hypervariable microsatellite loci for use in eastern phoebes (Sayornis phoebe) and related Tyrannids. Mol Ecol Notes (2002) 2:117–118.[CrossRef][Web of Science]
Webster MS, Chuang-Dobbs HC, Holmes RT. Microsatellite identification of extrapair sires in a socially monogamous warbler. Behav Ecol (2001) 12:439–446.
Webster MS, Pruett-Jones S, Westneat DF, Arnold SJ. Measuring the effects of pairing success, extra-pair copulations and mate quality on the opportunity for sexual selection. Evolution (1995) 49:1147–1157.[CrossRef][Web of Science]
Westneat DF, Sherman PW. Density and extra-pair fertilizations in birds: a comparative analysis. Behav Ecol Sociobiol (1997) 41:205–215.[CrossRef][Web of Science]
Westneat DF, Stewart IRK. Extra-pair paternity in birds: causes, correlates, and conflict. Annu Rev Ecol Evol Syst (2003) 34:365–396.[CrossRef]
Whittingham LA, Dunn PO. Effects of extra-pair and within-pair reproductive success on the opportunity for selection in birds. Behav Ecol (2005) 16:138–144.
Woodard JD, Murphy MT. Sex roles, parental experience and reproductive success of eastern kingbirds, Tyrannus tyrannus. Anim Behav (1999) 57:105–115.[CrossRef][Web of Science][Medline]
Woolfenden BE, Stutchbury BJM, Morton ES. Male Acadian flycatchers, Empidonax virescens, obtain extrapair fertilizations with distant females. Anim Behav (2005) 69:921–929.[CrossRef][Web of Science]
Yarbrough CG. The influence of distribution and ecology on thermoregulation of small birds. Comp Biochem Physiol A (1971) 39:235–266.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||