Behavioral Ecology Advance Access originally published online on December 20, 2006
Behavioral Ecology 2007 18(2):399-409; doi:10.1093/beheco/arl096
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Fertilization success and UV ornamentation in blue tits Cyanistes caeruleus: correlational and experimental evidence
Max Planck Institute for Ornithology, Seewiesen, PO Box 1564, D-82305 Starnberg, Germany
Address correspondence to K. Delhey, who is now at the Max Planck Institute for Ornithology, Vogelwarte Radolfzell, Schlossallee 2, D-78315 Radolfzell, Germany. E-mail: delhey{at}orn.mpg.de. A. Peters is now at the Max Planck Institute for Ornithology, Vogelwarte Radolfzell, Schlossallee 2, D-78315 Radolfzell, Germany. A. Johnsen is now at the Natural History Museum, University of Oslo, PO Box 1172, Blindern, N-0318, Oslo, Norway.
Received 25 April 2006; revised 4 October 2006; accepted 27 November 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cases where less ornamented males are favored through sexual selection are rare among birds. Here we show, based on data from 3 consecutive breeding seasons, that male blue tits with less ultraviolet (UV)-ornamented crown feathers sire more offspring. This pattern was mainly driven by the higher success of older, less UV-ornamented males at siring extrapair offspring. The reason behind this relationship is unclear although we hypothesize that being less UV-ornamented may enable adult males to intrude into nearby territories by mimicking juveniles. To test causality, we experimentally manipulated male crown coloration creating 2 groups, one with higher (UV(+) treatment) and one with lower UV reflectance (UV(–) treatment). Contrary to our expectations, UV(–) males were less likely to sire extrapair offspring than UV(+) males. The treatment had no effect on the likelihood of losing paternity in a male's own nest. Because the experimental evidence does not support the observational data, a direct effect of male crown color on extrapair success cannot be confirmed. However, potential pitfalls of this and other such color manipulation experiments, like fading of treatment with time and mismatches between behavior and coloration, call for new improved manipulation techniques and detailed behavioral observations to conclusively test for the effect of blue tit crown coloration on male extrapair success.
Key words: color manipulation, Cyanistes (Parus) caeruleus, extrapair paternity, sexual selection, ultraviolet.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In sexually dimorphic birds, more ornamented males are usually more successful in mate attraction. Generally, those males that express the most extreme form of a particular ornament, such as extravagant plumage or bright coloration, obtain more matings and/or mates of higher quality by either outcompeting other males (male–male competition) or being directly preferred by females (female choice) (Darwin 1871
; Ryan and Keddy-Hector 1992; Andersson 1994). All other things being equal, pairing with many or high-quality females leads to increased breeding success, and sexual selection is thereby mostly directional, toward increased ornamentation.
Conversely, only relatively few examples exist where males with a less exaggerated sexual trait experience higher breeding success through sexual selection. Most of these cases refer to smaller males being more successful by virtue of their increased agility, aerobatic abilities, or reduced energetic requirements (Blomqvist et al. 1997; Balmford et al. 2000; Voigt et al. 2005). In most of these species, sexual selection for smaller males leads also to reversed size dimorphism (Szekely et al. 2004). Examples where males are the larger or more ornamented sex but nonetheless smaller or less ornamented males seem to be favored by sexual selection are rare. In some populations of house sparrows (Passer domesticus) and house finches (Carpodacus mexicanus), for instance, females prefer to mate with less ornamented males because these are willing to invest more in brood rearing (Griffith et al. 1999; Badyaev and Hill 2002). In these cases, male ornamentation is maintained by opposing positive selection on ornament elaboration through alternative mechanisms: more ornamented male house finches are preferred by young inexperienced females and pair early in the year (Badyaev and Hill 2002), whereas more ornamented house sparrows show higher overwinter survival (Griffith et al. 1999). However, in the absence of these opposing selective forces, male ornamentation might become reduced (Saetre et al. 1997) or even completely lost in the course of evolutionary history (Wiens 2001).
Recently, we documented higher annual reproductive success for less ornamented blue tit (Cyanistes caeruleus) males (Delhey et al. 2003). In this species, both sexes display brilliant ultraviolet (UV)/blue crown feathers, but males are more UV ornamented than females (Andersson et al. 1998; Hunt et al. 1998). Several lines of evidence suggest that male UV ornamentation is under direct sexual selection through female choice. In the wild, blue tits have been found to mate assortatively by crown UV coloration (Andersson et al. 1998), whereas choice-chamber experiments revealed a preference for "UV-intact" against "UV-blocked" mates (Hunt et al. 1999). Moreover, experimental and correlational evidences suggest that females perceive highly UV-reflectant males as attractive because they bias brood sex ratios toward sons (Sheldon et al. 1999; Griffith et al. 2003) and invest relatively more in brood rearing and nest defense if paired to more UV-reflectant males (Limbourg et al. 2004; Johnsen et al. 2005). These studies suggested that more UV-ornamented males are preferred, and accordingly in our Viennese study population, we found that more UV-ornamented males achieved higher within-pair paternity. However, at the same time, older, less UV-ornamented males were more successful at siring extrapair offspring (Delhey et al. 2003). Because selection pressure through extrapair success was greater than selection through within-pair success, there was an overall weak negative selection on male UV ornamentation through annual fertilization success.
The aim of this study is 2-fold. First, we expand the correlational analysis of crown color and male reproductive success reported by Delhey et al. (2003) including 2 more years of data. Delhey et al. (2003) was based on data from a single breeding season, and selection pressures may vary among years. Furthermore, one cannot exclude the possibility that reported correlations, even though significant, are due to type I statistical errors. Second, we experimentally test the relationship between ornament expression and within- and extrapair paternity by manipulating male crown coloration. Experimental studies testing the causal link between male ornamentation and paternity are surprisingly scarce (Griffith et al. 2002), and often male ornaments are manipulated far beyond the natural degree of variation (Smith et al. 1991; Johnsen et al. 1998) precluding inferences about the importance of natural variation. In an attempt to avoid this problem, we use a new method to manipulate male crown color largely within the natural range of variation.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study site and general methods
This study was carried out between 2001 and 2003 at Kolbeterberg (48°13'N, 16°20'E), in the outskirts of Vienna, Austria. The study site consists of a
35 ha plot part of a large deciduous mixed forest. The plot contained 220 nest boxes in 2001 and 250 in 2002 and 2003. Each year, from the end of March onward, we regularly checked each nest box for signs of occupancy. Occupied boxes with complete nests (i.e., lined cups) were checked every day to determine laying date and clutch size. After incubation initiation, we left the nests undisturbed until the expected hatching date, when boxes were checked daily again until hatching finished. All unhatched eggs and dead nestlings found in the nest were collected and stored in 70% ethanol. As part of another project, 20 clutches were collected under license (10 in 2001 and 10 in 2003) after one day of incubation, eggs were opened, and embryos stored in ethanol. Broods were inspected at least 3 more times before fledging (which happens around day 19–20 posthatch). A small blood sample (5–25 µl) was drawn from the brachial vein of nestlings at day 9–14 posthatch, and each chick was banded with a numbered metal ring. Adults were captured at the nest box while feeding 8- to 14-day-old chicks. If unbanded, they were fitted with a numbered metal ring and a unique combination of plastic color bands, and a blood sample (10–50 µl) was taken from the brachial vein. Adult birds were sexed by presence or absence of the brood patch and aged as juveniles or adults following Svensson (1992). After banding and blood sampling, we measured the color of the crown (see below).
Color measurements
Reflectance of the crown feathers was measured with a S-2000 spectrometer with a DH-2000-FHS deuterium halogen light source (Ocean Optics, Eerbeek, The Netherlands). Spectrometer and lamp were connected through a bifurcated fiber optic probe, fitted at the end with a plastic cylinder to standardize measuring distance and exclude ambient light. The probe was held perpendicular to the surface of the feathers, and we took readings of 5 standardized spots (11.3 mm2 each). Reflectance (R) was calculated relative to a WS-2 white standard using the program Spectra-Win (Top Sensor Systems) in the range between 300 and 700 nm. Raw spectra were imported into a spreadsheet program and smoothed using a running average over a 10-nm interval. Previous studies on blue tit crown coloration (Sheldon et al. 1999; Delhey et al. 2003; Griffith et al. 2003; Limbourg et al. 2004; Johnsen et al. 2005) indicate that patterns of variability in male crown coloration are adequately and meaningfully summarized by the following 2 variables: 1) hue or spectral location, determined as the wavelength of peak reflectance (
Rmax) and 2) the relative amount of UV reflectance or "UV chroma," calculated as reflectance in the UV range (300–400 nm) divided by total reflectance ([R300–400/R300–700] x 100) and expressed as a percentage. A more exaggerated signal is one that shows a reflectance peak shifted further toward or into the UV (i.e., a smaller value of hue) and a higher relative UV reflectance (higher UV chroma) (Andersson 1999; Keyser and Hill 1999, 2000).
Color manipulation experiment
The color treatment was designed to modify crown reflectance within the natural range of variation found in our population. We captured 47 males in March–April 2003, on average 18 days before their mates laid the first egg (range 1–25 days), measured premanipulation color as described above (Color measurements), and manipulated male crown color. Males of both age classes were sequentially allocated to either UV(+) treatment (n = 22), where peak reflectance in the UV was shifted toward shorter wavelengths, or UV(–) treatment (n = 25), where peak reflectance was strongly shifted toward longer wavelengths (Figure 1). To manipulate crown coloration, we used Edding 4500 "T-Shirt Marker" pens (Ahrensburg, Germany). UV(+)-treated males were painted on the crown feathers with a light blue pen (Edding col.10), whereas UV(–) males were first painted with a dark blue pen (Edding col.03) and then with the light blue pen (Edding col.10) on top. To enhance water resistance of both treatments, we applied a coat of silicone-based fly dressing (Balzer Silicone-Fett, Germany) after painting the feathers. After completing the treatment, crown color was measured again. Treated males behaved normally, and their ability to raise the crown feathers was not impaired. We have recently reported the successful use of this same color manipulation method during chick feeding in 2003 (Johnsen et al. 2005). Note that unlike previous manipulations (e.g., Sheldon et al. 1999; Limbourg et al. 2004), we chose not to include a traditional "control" treatment (e.g., males treated only with silicone fly dressing) due to sample size limits caused by the difficulty of capturing a large sample of male blue tits closely before their females started egg laying. Thus, our manipulation was aimed at maximizing differences in reflectance between both treatment groups that should in turn increase the power to detect (if present) a significant difference in paternity success among treatments.
|
Before manipulation, the 2 treatment groups did not differ in crown color (all P > 0.35). After treatment, UV(+) males had a more UV-shifted hue (UV(+) = 353 nm, standard error [SE] = 1.4, range: 342–371; UV(–) = 401.5 nm, SE = 2.6, range: 367–434; t = 16.2, P < 0.001) and higher UV chroma (UV(+) = 33%, SE = 0.2, range: 31–35; UV(–) = 31%, SE = 0.3, range: 28–34; t = –6.58, P < 0.001) than UV(–) birds. When comparing postmanipulation color with premanipulation values, UV(+) males presented a more UV-shifted crown hue after treatment, whereas UV(–) males had a less UV-shifted hue (UV(+): mean difference = –9.9 nm, SE = 1.4; UV(–): mean difference = 36.8, SE = 2.6; paired t-test, both P < 0.001). Postmanipulation UV chroma, however, was lower than premanipulation values for both treatments (UV(+): mean difference = –1.8, SE = 0.2; UV(–): mean difference = –4.0, SE = 0.3; paired t-test, both P < 0.001). This was due to the fact that UV chroma is computed relative to achromatic brightness, which was increased in both treatments due to the coating with silicon fly dressing (Figure 1), a common effect of using this substance for waterproofing. Despite the fact that UV(+) males were less UV chromatic than before manipulation, they had a significantly higher UV chroma than UV(–) males (see above). After manipulation, the color variables were largely within the natural range of variation (Johnsen et al. 2005
) of this population except for 2 UV(–) males that had a more long-wave hue (418 and 434 nm) than the upper limit (410 nm) of the natural range. To assess the durability of the treatment, we recaptured 16 males on average 15 days postmanipulation (range 7–25 days). UV(–) males (n = 9) were still significantly less UV chromatic (paired t-test, t8 = 4.20, P = 0.003) and had a less UV-shifted hue (paired t-test, t8 = –3.179, P = 0.013) than before treatment. Crown reflectance of the UV(+) males (n = 7) was no longer significantly different from pretreatment values (P > 0.30). Nonetheless, the 2 treatment groups still differed significantly in crown hue (t14 = 2.178, P = 0.047), with UV(–) males having a less UV-shifted hue than UV(+) males. To account for this fading of the treatment, we included the time (in days) between treatment and laying date as a covariate in the analysis (see below).
Paternity analysis
We used 8 polymorphic microsatellite markers (Pca3, 7, 8, and 9 [Dawson et al. 2000], Pocc1 and 6 [Bensch et al. 1996], Phtr3 [Fridolfsson et al. 1997], and PK11 [Tanner SM, Richner H, Schuenperli D, unpublished data; European Molecular Biology Laboratory accession number: AF041465]) to determine the paternity of 2430 offspring (including dead nestlings and embryos from unhatched eggs) from 227 broods. We followed a standard polymerase chain reaction (PCR) protocol using about 20 ng genomic DNA, 0.25 U of Taq DNA polymerase (Promega), and 1.5 mM MgCl2. The PCR profiles had the following annealing temperatures: 53 °C (Pca8, Phtr3), 55 °C (Pca3), 56 °C (Pocc1), 57 °C (Pca9, Pocc6), and 60 °C (Pca7, PK11). Amplified fragments were resolved on an ABI Prism 310 Genetic Analyzer (Applied Biosystems).
The combined probability of exclusion (Jamieson 1994) for the marker set was >0.999. Paternity was excluded if two or more loci showed mismatches between putative fathers and offspring. In 2051 cases, there was no (n = 1987) or one mismatch (n = 73) with the social father. For these offspring, the average probability of false inclusion (Jeffreys et al. 1992) was 2.15 x 10–4 ± 8.19 x 10–4 standard deviation (SD) (range 4.77 x 10–3 to 9.44 x 10–10; mismatched locus excluded from calculation). We therefore conclude that they were sired by the social male and that the single mismatches were due to mutations or typing errors. The remaining offspring showed two or more mismatches with the putative father and were thus sired by extrapair males. For 215 of the extrapair offspring, an alternative male matched the paternal genotype completely. Fifteen offspring showed a single mismatch with the putative extrapair father, and in 12 of these cases, the male was the unequivocal sire of another offspring in that brood. For these 15 offspring, the average probability of false inclusion was 2.90 x 10–5 ± 9.71 x 10–5 SD (range 3.79 x 10–4 to 1.61 x 10–8; mismatched locus excluded from calculation), and we therefore assigned them to their putative extrapair fathers. Overall, the probability of false inclusion for the 230 assigned extrapair offspring was 2.79 x 10–5 ± 2.09 x 10–4 SD (range 3.07 x 10–3 to 7.59 x 10–10).
Statistical analysis
We constructed generalized linear models using SPSS 12. Full models were simplified by excluding variables in order of decreasing significance until only terms with P < 0.1 remained in the model. Excluded variables were included in the final model one by one to confirm their lack of significance. Throughout the paper parametric tests were used when variables were normally distributed, otherwise we employed their nonparametric equivalents. All tests are 2 tailed.
Correlational data
The aim of this part of the study was to test whether the relationships between within- and extrapair success and male crown color observed in 2001 (Delhey et al. 2003) were consistent when adding data from 2 more years. Because hue and UV chroma are highly correlated (r = –0.76, P < 0.001, n = 127), we analyzed their effects in separate models. Data from the 3 years were pooled, and for males that bred in more than one year, we randomly included only one breeding attempt in the analyses to avoid pseudoreplication. In case of polygyny, only the primary brood of a male was used (Johnsen et al. 2001). Polygyny is rare in our study population (less than 10% of males, Delhey K, Peters A, Kempenaers B, unpublished data), and including the offspring from secondary females as part of within-pair success does not change the results (not shown). Polygynous and monogamous males did not differ in crown color (Mann–Whitney test: hue, Z = –0.61, P = 0.54; UV chroma, Z = –0.22, P = 0.82; nmonogamous = 116, npolygynous = 11). As male color differed between years (Delhey and Kempenaers 2006, see also Figure 2), we standardized (mean = 0, SD = 1) hue and UV chroma for each year separately. We used ordinal and binary logistic regressions (Thomson et al. 1998) to analyze the effect of color on the following dependent variables: clutch size, likelihood of being cuckolded, likelihood of being a cuckolder, proportion of offspring sired in the own brood, number of sired extrapair offspring, and total fertilization success. Full models included study year and male age (juvenile or adult) as factors and male crown color as a covariate. Additionally, we tested for the interaction between male age and color. The apparent between-year consistency of the significant relationships between color and fertilization success was tested by including a posteriori the corresponding color x year interactions in the final model and by visual inspection of the data.
|
Experimental data
First, we used univariate tests to compare breeding success (date of first egg and clutch size) and fertilization success (proportion of within-pair offspring, number of extrapair young, and total fertilization success) between UV(–) and UV(+) males. However, the effect of UV reflectance manipulations may depend on confounding variables such as premanipulation color (Sheldon et al. 1999
, see Results), male age, or the delay between treatment and the start of laying (due to fading of the treatment, see above). Therefore, we included these terms as explanatory variables in the models with breeding and fertilization success as dependent variates and treatment (UV(+) and UV(–)) as a factor. We also tested the interactions age x treatment, age x premanipulation color, and treatment x premanipulation color (see Predictions for the experiment in Results for rationale). |
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Correlational results
We genotyped a total of 1877 offspring from 172 unmanipulated broods. The proportion of broods containing at least one extrapair offspring was higher in 2001 (65%, 33/51) than in 2002 (56%, 47/84) or 2003 (56%, 21/37), but these differences were not significant (
22 = 1.07, P = 0.58). On average, males sired 85.2% (SE = 1.4; median = 91.3%, range 0–100%) of the offspring in their own nest, and there was little variation between study years (2001: mean = 84.6 ± 2.4%, median = 90.0%, range = 27–100; 2002: 85.1 ± 2.2%, median = 91.6%, range = 0–100; 2003: 86.1 ± 3.0%, median = 91.6%, range = 15–100). Clutch size, after controlling for the effects of year and laying date, correlated weakly with male crown color (Table 1). Females paired to less UV males (higher hue, lower UV chroma) tended to lay larger clutches. This effect was independent of laying date (in this population early laid clutches are usually larger, Table 1) because excluding this variable from the model did not change the effect of male color (hue, P = 0.047; UV chroma, P = 0.073), and crown color is unrelated to laying date (both color variables, P > 0.6).
|
Within-pair paternity was not consistently related to male crown color in the 3 study years (Table 1, Figure 2a). There was a trend that cuckolded males, irrespective of age, had a less UV-shifted hue than males with full paternity in their broods, but the difference was small (Table 1). Although annual differences in this relationship were not sufficiently large to support a significant hue x year interaction (
22 = 2.60, P = 0.27), the trend for higher within-pair success for more UV males was clearly absent in 2002 and very weak in 2003 (Figure 2a). Additionally, there was no relationship between the likelihood of being cuckolded and UV chroma (Table 1). Likewise, the proportion of within-pair paternity did not correlate significantly with male crown color (Table 1).
Extrapair success was mainly affected by male age, older males being more successful at siring extrapair offspring (Table 2). On average, adult males sired 1.4 (SE = 0.3, n = 45) extrapair offspring against 0.4 (SE = 0.1, n = 80) sired by juveniles (Mann–Whitney test, Z = –3.14, P = 0.002), and 40% (18/45) of adult males sired at least one extrapair offspring against 16% (13/80) of juveniles (12 = 8.71, P = 0.003). Crown hue seemed to have an effect on extrapair success, although in an age-dependent manner, whereas there was no relationship between crown UV chroma and extrapair success (Table 2). The significant interaction hue x age (Table 2) and visual inspection of the data (Figure 2b) suggested that the relationship between hue and extrapair success was different in the 2 age classes. Considering each age class separately showed a clear relationship between extrapair success and hue among adults, but not among juvenile males. Adult males that sired extrapair offspring had a less UV-shifted hue than adult males that did not sire extrapair young (t-test, t53 = –2.41, P = 0.019; Figure 2b), but this was not the case among juvenile males (t-test, t89 = 0.10, P = 0.92; Figure 2b). Moreover, the number of extrapair offspring sired correlated positively with hue among adult (rs = 0.33, P = 0.012, n = 55) but not juvenile males (rs = 0.01, P = 0.93, n = 91). These differences between juvenile and adult males were relatively consistent over the 3 study years as indicated by the nonsignificant year x hue x age interaction and visual inspection of Figure 2b (likelihood of siring extrapair offspring: 22 = 1.51, P = 0.47; number of extrapair offspring: 22 = 3.32, P = 0.19).
|
Pairwise comparisons between cuckolding males and the males they cuckolded confirmed that cuckolders are on average older than the males they cuckold (Wilcoxon paired test, Z = –3.4, P = 0.001, n = 49). Overall, there were no differences in color between cuckolder and cuckolded males (paired t-test—hue: t = 0.78, P = 0.44; UV chroma: t = –1.17, P = 0.24, n = 48). However, in those cases where the cuckolder was an adult male (n = 33), it had a less UV-shifted hue (paired t-test, t = 2.19, P = 0.036) and lower UV chroma (paired t-test, t = –2.26, P = 0.030) than the male it cuckolded. This was not the case when the cuckolder was a juvenile male (paired t-test—hue: t = –1.25, P = 0.23; UV chroma: t = 1.13, P = 0.28; n = 15).
Total fertilization success, the sum of within-pair and extrapair offspring, was mainly influenced by male age (B (SE) = –1.23 (0.35); 12 = 12.49, P < 0.001). Adult males sired on average 10.8 (SE = 0.6) offspring against 9.4 (SE = 0.3) sired by juveniles (Mann–Whitney, Z = –2.58, P = 0.010). This was mainly due to adult males' higher extrapair success (see above) because the number of within-pair offspring did not differ significantly between age classes (adults: 9.5 (SE = 0.4); juveniles: 9.0 (SE = 0.3); Mann–Whitney: Z = –1.37, P = 0.17). After controlling for age differences, total fertilization success was negatively related to the expression of UV coloration (hue: B (SE) = 0.55 (0.17), 12 = 10.53, P = 0.001; UV chroma: B (SE) = –0.35 (0.16), 12 = 4.27, P = 0.038). The interaction between age x color was not significant, indicating that this pattern was similar for both age classes (P > 0.17). Similarly, the color x year interaction was nonsignificant for both color variables (P > 0.45).
Predictions for the experiment
Based on the above correlations, we could make the following predictions regarding the experimental results. 1) If the pattern found in 2001 (Delhey et al. 2003) and the weak trend in the overall data set (Figure 2a) reflect a female preference for more UV-ornamented social mates, males in the UV(+) treatment should experience (slightly) higher within-pair breeding success than UV(–) males. 2) If being less UV ornamented directly improves the chances of adult males to sire extrapair offspring, we would expect adult UV(–) males to have higher extrapair success. These predictions would be supported by significant treatment effects and a significant treatment x age interaction, respectively. However, these relatively straightforward predictions could become more complicated if the response to treatment depends on the color prior to treatment. Indeed, Sheldon et al. (1999) showed that a reduction in crown UV reflectance had a much stronger effect on brood sex ratio when males had been more UV reflective before the treatment, resulting in a negative correlation between male premanipulation color and experimental brood sex ratio. Hence, for within-pair paternity, we might likewise expect such a negative relationship, in both treatment groups. UV(–) males that were more UV reflectant prior to treatment would experience a greater decline in attractiveness and a more marked decrease in within-pair success. On the other hand, in the UV(+)-treated group naturally less UV-ornamented males would experience a greater boost in presumed attractiveness and therefore a greater increase in within-pair paternity. Similarly, the effect of premanipulation color could also affect extrapair success. A significant negative effect of premanipulation color on male fertilization success could lend support to these predictions.
Experimental results
We obtained paternity estimates for 23 UV(–), 19 UV(+), and 37 unmanipulated males in 2003. In total, we typed 836 offspring in these broods. Extrapair offspring accounted for 15% (126/836) of the typed offspring, and overall extrapair offspring were found in 60% of the broods (48/79). Three males (one UV(–) and 2 UV(+) males) sired none of the offspring in their broods, and the proportion of sired offspring ranged from 0% to 100% with a mean of 83% (SE = 3).
Females mated to UV(+) males laid their first egg earlier than females mated to UV(–) or unmanipulated males (Table 3). Surprisingly, given that in this population earlier clutches are usually larger (Table 1), UV(+) males tended to have smaller clutches than UV(–) males (Table 3). Overall, there were no significant differences in the proportion of sired offspring between UV(–) and UV(+) males (Table 3, Figure 3a). Seven out of 23 males (30%) were cuckolded in the UV(–) treatment and 8 out of 19 (42%) in the UV(+) treatment (12 = 0.61, P = 0.43).
|
|
Extrapair success tended to be higher for UV(+) than for UV(–) males because the former sired more extrapair offspring (Table 3, Figure 3b). Indeed, only one out of 23 UV(–) males sired extrapair offspring, against 5 out of 19 UV(+) males (
12 = 4.10, P = 0.043). Total fertilization success did not differ between treatments (Table 3). Multivariate models containing potentially confounding variables largely confirmed the results from the previous analyses (Table 4). Treatment, age, color pretreatment or the time interval between treatment, and the date of first egg had an effect neither on the proportion of within-pair offspring nor on the likelihood of being cuckolded (Figure 3a). The number of extrapair offspring and the likelihood of siring extrapair offspring were weakly influenced by treatment and by male age, older and UV(+) males having higher extrapair success (Figure 3b). Neither the time interval between treatment and the date of first egg nor the color pretreatment contributed significantly to these models (Table 4). Because only one UV(–) and only one juvenile male sired extrapair offspring, this precluded the inclusion of the interactions pretreatment color x treatment and pretreatment color x age in the models. However, there were no obvious trends between color pretreatment and number of extrapair offspring within the subgroup of UV(+) adult males (hue: rs = 0.28, P = 0.41; UV chroma: rs = –0.18, P = 0.60; n = 11).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Correlations
Based on a large sample collected over 3 breeding seasons, we found largely no (or a very weak) relationship between crown color and within-pair fertilization success. Extrapair fertilization success was correlated with male crown color but only among adults, with less UV-ornamented males being more successful at siring extrapair offspring (Table 1, Figure 2b). This negative correlation between UV ornamentation and extrapair success plus a weak negative correlation with clutch size resulted in an overall negative relationship between total fertilization success and the degree of male ornamentation in both age classes: less UV-ornamented blue tit males sired more offspring, confirming the result from our previous study (Delhey et al. 2003
).
Why do older, less UV males sire more extrapair young?
In many species of passerine birds, adult males are much more successful at siring extrapair offspring than juveniles (Griffith et al. 2002). This pattern has been interpreted as female preference for high-quality males, which have proven their viability (Brooks and Kemp 2001) but could also reflect a change in behavioral strategies as males age (Johnsen et al. 2003; Kleven et al. 2006). Hence, higher extrapair success of adult males could be due to a combination of male tactics and female preference (Westneat and Stewart 2003). Indeed, in the blue tit, both male- and female-driven pursuit of extrapair copulation seem to exist because females have been observed to undertake around-dawn extrapair forays (Kempenaers et al. 1992, own observations) and males often intrude into other males' territories (Foerster and Kempenaers 2005). If adult males invest more time in extrapair behaviors and/or are preferred by females, being less UV may be beneficial for older males either 1) because females directly prefer less UV-ornamented males as extrapair partners or 2) because being less UV facilitates intrusions into territories of other males. Alternatively, 3) crown coloration may not be causally involved in extrapair success but correlate with an unknown trait that enhances extrapair success. If this were the case, we would expect to see no effect of color manipulation on male extrapair success (see discussion of Experimental results below). We now focus on the first 2 possibilities.
The hypothesis of direct female preference for less UV adult males in this population finds some support in the pattern of brood sex allocation. Females paired to adult males produce more male biased broods with decreasing UV ornamentation of their mate, whereas this pattern is reversed in females paired to juvenile males (Delhey K, Peters A, Johnsen A, Kempenaers B, unpublished data). According to the theory of sex allocation based on mate attractiveness, females should produce more male offspring if paired to a sexually attractive male because sons would benefit by inheriting their father's attractiveness (West and Sheldon 2002). If this applies to the blue tit, it would mean that females find less UV-ornamented older males more attractive. However, it remains puzzling then why females are not more faithful to less UV-ornamented adult males or even to adult males in general (Table 1).
Alternatively, adult blue tits might benefit from being less UV ornamented through increased success at intruding into neighboring territories. Male blue tits usually respond extremely aggressive toward taxidermic mounts simulating an intruder. However, aggressiveness was considerably reduced if the mounts were manipulated to reduce crown UV reflectance (Alonso-Alvarez et al. 2004). Hence reduced UV ornamentation might be a desirable attribute among older males seeking extrapair copulations in foreign territories. In lazuli buntings (Passerina amoena), for instance, dull juveniles are more likely to be tolerated by older males because they pose little danger to their paternity (Greene et al. 2000) and likewise juvenile blue tits are far less proficient at obtaining extrapair paternity (Table 1). Interestingly, those adult males that sire extrapair offspring resemble UV-dull juveniles (Figure 2b): age differences in crown color are only evident among males that did not sire extrapair offspring (hue juveniles: 387.9 nm, SE = 1.4, n = 67; hue adults: 378.3 nm, SE = 1.8, n = 27; t-test for unequal variances, t = 3.99, P < 0.001) and not among males that did sire extrapair offspring (hue juveniles: 388.5 nm, SE = 3.0, n = 13; hue adults: 385.4 nm, SE = 3.5, n = 18; t-test, t = 0.54, P = 0.59). This lends support to the idea that UV-dull adult males could be mimicking juveniles or low-quality individuals to deceive territorial males about their true status when roaming for extrapair copulations.
What do more UV-ornamented males gain?
Our results imply that there is little to be gained from being a highly UV-ornamented blue tit male in our study population. Although the likelihood of being cuckolded was lower for more UV-ornamented males in one of 3 years, the advantage is small compared with the increase in extrapair success associated with decreased UV ornamentation. The obvious question then becomes: what selects for UV coloration in this species? Whereas the present study addresses selection through the number of sired offspring, there are other important selective forces operating in the life of a male blue tit, such as surviving and obtaining and keeping a territory and a (high quality) mate.
A survival advantage appears an unlikely candidate for a selective force maintaining crown UV ornamentation because annual survival of juvenile and adult males (and females) in this blue tit population is unrelated to their crown color (Delhey and Kempenaers 2006). On the other hand, an advantage of highly UV-ornamented males in obtaining a territory or a mate could result in a strong selection pressure for ornament exaggeration (Siefferman and Hill 2005), potentially surpassing the more subtle effects of extrapair success. Although there is a suggestion that crown coloration is involved in male–male interactions (Alonso-Alvarez et al. 2004), what attributes of a male determine the outcome of territorial interactions in the blue tit is unknown. Similarly, we lack information on which males were unsuccessful at finding a mate because they are not part of the breeding population and remain undetected. Additionally, more ornamented males could also benefit by obtaining a higher quality mate. In another blue tit population, assortative mating by crown coloration was observed (Andersson et al. 1998), implying mutual mate choice based on crown ornamentation as an indicator of quality (Hunt et al. 1999). However, in the 3 years of our study, we found no evidence of assortative mating by crown coloration, and female coloration did not reflect female quality in terms of reproductive output (Delhey K, Peters A, Kempenaers B, unpublished data). Finally, female differential allocation (Sheldon 2000) based on male crown color may also contribute to the success of more ornamented males. Indeed, experimental studies in 2 populations show that females invest relatively less in brood rearing if paired to males whose crown UV reflectance has been experimentally reduced (Limbourg et al. 2004; Johnsen et al. 2005). In general, it seems that detailed behavioral observations will be required to establish potential benefits of UV crown reflectance in blue tits, be it through female preferences or increased success at agonistic interactions.
Experimental results
Within-pair paternity was not significantly affected by the experiment. Neither treatment nor any other variable included in our models had an effect on the proportion of within-pair paternity or the likelihood of being cuckolded (Tables 3 and 4; Figure 3a). The lack of treatment effect coupled with the inconsistent relationship between crown UV reflectance and within-pair paternity in the correlational data set (Table 1) undermines the hypothesis that females are more faithful to more UV-ornamented males or that these males are more successful at defending their paternity (Delhey et al. 2003). However, limitations of the experiment (see below), the positive relationship in one year (Delhey et al. 2003), and the fact that selection pressures often vary considerably between years and populations (this study, Griffith et al. 2003) make general dismissal of selection of male blue tit crown color through within-pair paternity premature. Nonetheless, it seems that selection on male color through this component of fertilization success is weak at best.
Extrapair success was affected by treatment, with UV(+) males being more likely to cuckold other males and hence sire more extrapair offspring than UV(–) males (Figure 3b). However, this pattern is the opposite of what we had expected based on the observed relationships (Table 1). If crown color indeed has a direct effect on extrapair success, adult males from the UV(–) treatment should have been more successful at siring extrapair offspring. Alternatively, we hypothesized that crown color may be negatively correlated to a, yet unidentified, variable that increases a male's chances of siring extrapair offspring. Under this scenario, we would expect no treatment effect on extrapair paternity. The results from the experiment therefore do not fit with either of these predictions. It could be argued that treatment effects were weak, and hence, the result could be a statistical artifact (type I error). However, note that the treatment effect on extrapair success seemed to be stronger than the effect of age (Table 4), a male attribute known to be very important for extrapair success in this and other species (Griffith et al. 2002).
Conflicting results: confounded correlations or experimental artifact?
Lack of agreement between correlational and experimental results is often assumed to suggest that correlational results do not represent causation but are driven by a third, unknown, correlated variable. What could be determining male fertilization success if not crown color? Blue tits show several other ornaments, for instance, the yellow breast that could potentially influence male paternity success and trade-off with crown color. However, the evidence available to us suggests that there is no relationship between male fertilization success and carotenoid ornamentation nor a negative correlation between the coloration of the breast and the crown (Delhey K, Kempenaers B, unpublished data). Alternatively, song output may determine male paternity success. Indeed, in a Belgian population, those males that sung longer strophes had higher extrapair success (Kempenaers et al. 1997). However, the lack of correlation between male color and song output in the blue tit reported in 2 recent studies (Dreiss et al. 2006; Parker et al. 2006) makes a trade-off between these traits unlikely. Finally, and although on average adult males are more UV ornamented than juveniles (Delhey and Kempenaers 2006), the pattern of less UV-ornamented adult males having high extrapair success could be merely a by-product of the success of very old males (older than 2 years). Indeed, these oldest males in the population tend to be less UV ornamented than second year males (Delhey and Kempenaers 2006) and are highly proficient at siring extrapair offspring (77% of these males sired extrapair offspring, n = 9). However, although this undoubtedly contributes to the difference in coloration between successful and unsuccessful extrapair adult males, this difference in color is already present if we only look at second year males (hue sired extrapair young = 384.9, SE = 3.7, n = 11; hue did not sire extrapair young = 377.6, SE = 1.6, n = 25; t = –2.2, P = 0.041). Thus, so far it remains unknown which traits enable male blue tits to increase their fertilization success.
Although correlational results necessarily depend on experimental support for confirmation of causality (Kempenaers and Sheldon 1997; Milinski 1997), solely relying on experimental evidence can also be problematic. Indeed, the results from the experiment highlight the fact that previous knowledge about correlational patterns is necessary for a meaningful interpretation of the outcome of phenotypic manipulations. Had we performed the experiment without previous information on the relationship between extrapair paternity and crown color, we would have concluded that less UV-reflectant males are avoided and/or more UV males preferred by females as extrapair mates. The logical conclusion would then have been that blue tit male crown UV reflectance is under direct sexual selection through extrapair mating success. Although it remains a possibility that, everything else being equal, females prefer more UV-ornamented males as extrapair partners, this potential preference appears insufficient to counter the overriding advantage at siring extrapair young we observed in older, less UV males.
Furthermore, manipulating a bird's phenotype may have unexpected consequences that could confound experimental results. For example, when birds have multiple ornaments that show some degree of intercorrelation (Møller and Pomiankowski 1993; Andersson et al. 2002), manipulating only one ornament may cause mismatches between otherwise correlated traits (Sheldon et al. 1999). In blue tits, besides the conspicuous crown feathers, males also show UV/blue colored wing coverts and tail feathers, and color expression is correlated among these patches (Sheldon et al. 1999; Delhey K, Kempenaers B, unpublished data). These correlations would be disrupted by the manipulation of the crown plumage, an inherent problem of ornament color manipulation. Additionally, mismatches between manipulated appearance and behavior can cause unexpected experimental results. For example, enlarging the badge of status in Harris sparrows (Zonotrichia querula) led to increased dominance rank only if coupled with simultaneous testosterone implants, to also increase dominance behavior (Rohwer S and Rohwer FC 1978). In our blue tit population, male crown coloration correlates with natural testosterone levels in an age-dependent manner, with testosterone levels increasing with increasing UV ornamentation in juveniles and decreasing with increasing UV color in adults (Peters et al. 2006). Hence, changing male color without manipulating male testosterone at the same time could have confounded our experimental results. Other problems with current experimental approaches of color manipulation are that experimental treatments often go beyond the natural range of variation, for example, using sun block to manipulate UV reflectance (Johnsen et al. 1998; Sheldon et al. 1999), or that treatments cannot perfectly match the spectral shape of natural colors as in the present experiment (Figure 1). In our case, the inclusion of a control treatment (for instance, by applying only silicone fly dressing) would have allowed us to compare these males with UV(+), UV(–), and untreated males and thus assess potential artifacts due to imperfect color manipulations. Future studies should include such control treatments. Finally, fading of treatment with time as in the present study will also contribute to the experimental problems. Therefore, being aware of the potential weaknesses of the experimental approach used is a crucial step toward the meaningful interpretation of its results.
Possibly due to such difficulties in designing methods to realistically manipulate male phenotype, surprisingly few studies have experimentally tested the link between male plumage ornamentation and paternity success. Numerous studies have described correlations between the expression of male ornaments and within- and/or extrapair success (see review in Griffith et al. 2002). However, only 5 studies on 3 species have experimentally tested for a direct causal link between male plumage ornamentation and paternity success (tail length and breast color in barn swallows [Hirundo rustica] [Smith et al. 1991; Saino et al. 1997; Safran et al. 2005], throat patch UV reflectance in bluethroats [Luscinia svecica] [Johnsen et al. 1998], and epaulet size and body blackness in red-winged blackbirds [Agelaius phoeniceus] [Westneat 2006]). Moreover, in 2 of these cases, the experimental manipulation did not confirm the relationships inferred from the correlational data (Smith et al. 1991; Johnsen et al. 1998, 2001). One explanation is that in these studies, the manipulations resulted in a trait outside the natural range of variation. Hence, experimental evidence supporting the often-assumed causal link between paternity success and male ornamentation in birds is still largely missing.
|
ACKNOWLEDGEMENTS |
|---|
We are thankful to Agnes Tuerk, Emmi Schlicht, Karin Hieke, Kim Teltscher, and Mihai Valcû for help with fieldwork and to Sandra Grainer and Kim Teltscher for performing paternity analysis. The suggestions by 2 anonymous reviewers improved earlier drafts of the manuscript. We are also thankful to Raphael-Thomas Klumpp and Alfred Fojt from the Institute of Silviculture, Vienna, for providing access to their facilities in the study area and finally to Hans Winkler and Dustin Penn from the Konrad Lorenz Institute for Comparative Ethology, Vienna, and to the Beranek Family for logistic support. This study was supported by the Alexander von Humboldt Foundation (Fellowships to A.P. and A.J.) and the Max Planck Society.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Alonso-Alvarez C, Doutrelant C, Sorci G. (2004) Ultraviolet reflectance affects male-male interactions in the blue tit (Parus caeruleus ultramarinus). Behav Ecol 15:805–809.
Andersson M. (1994) Sexual selection. (Princeton University Press, Princeton (NJ)).
Andersson S. (1999) Morphology of UV reflectance in a whistling-thrush: implications for the study of structural colour signalling in birds. J Avian Biol 30:193–204.
Andersson S, Örnborg J, Andersson M. (1998) Ultraviolet sexual dimorphism and assortative mating in blue tits. Proc R Soc Lond B Biol Sci 265:445–450.[CrossRef]
Andersson S, Pryke SR, Örnborg J, Lawes MJ, Andersson M. (2002) Multiple receivers, multiple ornaments, and a trade-off between agonistic and epigamic signalling in a widowbird. Am Nat 160:683–691.[CrossRef]
Badyaev AV and Hill GE. (2002) Paternal care as a conditional strategy: distinct reproductive tactics associated with elaboration of plumage ornamentation in the house finch. Behav Ecol 13:591–597.
Balmford A, Lewis MJ, Brooke MD, Thomas ALR, Johnson CN. (2000) Experimental analyses of sexual and natural selection on short tails in a polygynous warbler. Proc R Soc Lond B Biol Sci 267:1121–1128.[Medline]
Bensch S, Price T, Kohn J. (1996) Isolation and characterization of microsatellite loci in a Phylloscopus warbler. Mol Ecol 5:150–151.
Blomqvist D, Johansson OC, Unger U, Larsson M, Flodin LA. (1997) Male aerial display and reversed sexual size dimorphism in the dunlin. Anim Behav 54:1291–1299.[CrossRef][ISI][Medline]
Brooks R and Kemp DJ. (2001) Can older males deliver the good genes? Trends Ecol Evol 16:308–313.[CrossRef][Medline]
Darwin C. (1871) The descent of man and selection in relation to sex. (John Murray, London).
Dawson DA, Hanotte O, Greig C, Stewart IRK, Burke T. (2000) Polymorphic microsatellites in the blue tit Parus caeruleus and their cross-species utility in 20 songbird families. Mol Ecol 9:1941–1944.[CrossRef][Medline]
Delhey K, Johnsen A, Peters A, Andersson S, Kempenaers B. (2003) Paternity analysis reveals opposing selection pressures on crown coloration in the blue tit (Parus caeruleus). Proc R Soc Lond B Biol Sci 270:2057–2064.[Medline]
Delhey K and Kempenaers B. (2006) Age differences in blue tit Parus caeruleus plumage colour: within-individual changes or colour-biased survival? J Avian Biol 37:339–348.[CrossRef]
Dreiss A, Richard M, Moyen F, White J, Moller AP, Danchin E. (2006) Sex ratio and male sexual characters in a population of blue tits, Parus caeruleus. Behav Ecol 17:13–19.
Foerster K and Kempenaers B. (2005) Effects of testosterone on male-male competition and male-female interactions in blue tits. Behav Ecol Sociobiol 57:215–223.[CrossRef]
Fridolfsson AK, Gyllensten UB, Jakobsson S. (1997) Microsatellite markers for paternity testing in the willow warbler Phylloscopus trochilus: high frequency of extra-pair young in an island population. Hereditas 126:127–132.[CrossRef]
Greene E, Lyon BE, Muehter VR, Ratcliffe L, Oliver SJ, Boag PT. (2000) Disruptive sexual selection for plumage coloration in a passerine bird. Nature 407:1000–1003.[CrossRef][Medline]
Griffith SC, Örnborg J, Russell AF, Andersson S, Sheldon BC. (2003) Correlations between ultraviolet coloration, overwinter survival and offspring sex ratio in the blue tit. J Evol Biol 16:1045–1054.[CrossRef][ISI][Medline]
Griffith SC, Owens IPF, Burke T. (1999) Female choice and annual reproductive success favour less-ornamented male house sparrows. Proc R Soc Lond B Biol Sci 266:765–770.[CrossRef]
Griffith SC, Owens IPF, Thuman KA. (2002) Extra pair paternity in birds: a review of interspecific variation and adaptive function. Mol Ecol 11:2195–2212.[CrossRef][Medline]
Hunt S, Bennett ATD, Cuthill IC, Griffiths R. (1998) Blue tits are ultraviolet tits. Proc R Soc Lond B Biol Sci 265:451–455.[CrossRef]
Hunt S, Cuthill IC, Bennett ATD, Griffiths R. (1999) Preferences for ultraviolet partners in the blue tit. Anim Behav 58:809–815.[CrossRef][ISI][Medline]
Jamieson A. (1994) The effectiveness of using co-dominant polymorphic allelic series for (1) checking pedigrees and (2) distinguishing full-sib pair members. Anim Genet 25:37–44.[Medline]
Jeffreys AJ, Allen MJ, Hagelberg E, Sonnberg A. (1992) Identification of the skeletal remains of Josef Mengele by DNA analysis. Forensic Sci Int 56:65–76.[CrossRef][ISI][Medline]
Johnsen A, Andersson S, Ornborg J, Lifjeld JT. (1998) Ultraviolet plumage ornamentation affects social mate choice and sperm competition in bluethroats (Aves, Luscinia s. svecica)—a field experiment. Proc R Soc Lond B Biol Sci 265:1313–1318.[CrossRef]
Johnsen A, Delhey K, Schlicht E, Peters A, Kempenaers B. (2005) Male sexual attractiveness and parental effort in blue tits: an experimental test of the differential allocation hypothesis. Anim Behav 70:877–888.[CrossRef]
Johnsen A, Lifjeld JT, Andersson S, Örnborg J, Amundsen T. (2001) Male characteristics and fertilisation success in bluethroats. Behaviour 138:1371–1390.[CrossRef]
Johnsen A, Lifjeld JT, Krokene C. (2003) Age-related variation in mate-guarding intensity in the bluethroat (Luscinia s. svecica). Ethology 109:147–158.[CrossRef]
Kempenaers B and Sheldon BC. (1997) Studying paternity and paternal care: pitfalls and problems. Anim Behav 53:423–427.[CrossRef]
Kempenaers B, Verheyen GR, Dhondt AA. (1997) Extrapair paternity in the blue tit (Parus caeruleus): female choice, male characteristics, and offspring quality. Behav Ecol 8:481–492.
Kempenaers B, Verheyen GR, Vandenbroeck M, Burke T, Vanbroeckhoven C, Dhondt AA. (1992) Extra-pair paternity results from female preference for high-quality males in the blue tit. Nature 357:494–496.[CrossRef]
Keyser AJ and Hill GE. (1999) Condition-dependent variation in the blue-ultraviolet coloration of a structurally based plumage ornament. Proc R Soc Lond B Biol Sci 266:771–777.[CrossRef]
Keyser AJ and Hill GE. (2000) Structurally based plumage coloration is an honest signal of quality in male blue grosbeaks. Behav Ecol 11:202–209.
Kleven O, Marthinsen G, Lifjeld JT. (2006) Male extraterritorial forays, age and paternity in the socially monogamous reed bunting (Emberiza schoeniclus). J Orn 147:468–473.[CrossRef]
Limbourg T, Mateman AC, Andersson S, Lessells CM. (2004) Female blue tits adjust parental effort to manipulated male UV attractiveness. Proc R Soc Lond B Biol Sci 271:1903–1908.[Medline]
Milinski M. (1997) How to avoid seven deadly sins in the study of behavior. Adv Study Behav 26:159–180.
Møller AP and Pomiankowski A. (1993) Why have birds got multiple sexual ornaments? Behav Ecol Sociobiol 32:167–176.[ISI]
Parker TH, Barr IR, Griffith SC. (2006) The blue tit's song is an inconsistent signal of male condition. Behav Ecol 17:1029–1040.
Peters A, Delhey K, Goymann W, Kempenaers B. (2006) Age-dependent association between testosterone and crown UV coloration in male blue tits (Parus caeruleus). Behav Ecol Sociobiol 59:666–673.[CrossRef]
Rohwer S and Rohwer FC. (1978) Status signalling in Harris sparrows: experimental deception achieved. Anim Behav 26:1012–1022.[CrossRef]
Ryan MJ and Keddy-Hector A. (1992) Directional patterns of female mate choice and the role of sensory biases. Am Nat 139:S4–S35.[CrossRef]
Saetre GP, Moum T, Bures S, Kral M, Adamjan M, Moreno J. (1997) A sexually selected character displacement in flycatchers reinforces premating isolation. Nature 387:589–592.[CrossRef]
Safran RJ, Neuman CR, McGraw KJ, Lovette IJ. (2005) Dynamic paternity allocation as a function of male plumage color in barn swallows. Science 309:2210–2212.
Saino N, Primmer CR, Ellegren H, Moller AP. (1997) An experimental study of paternity and tail ornamentation in the barn swallow (Hirundo rustica). Evolution 51:562–570.[CrossRef]
Sheldon BC. (2000) Differential allocation: tests, mechanisms and implications. Trends Ecol Evol 15:397–402.[CrossRef][Medline]
Sheldon BC, Andersson S, Griffith SC, Örnborg J, Sendecka J. (1999) Ultraviolet colour variation influences blue tit sex ratios. Nature 402:874–877.[CrossRef]
Siefferman L and Hill GE. (2005) UV-blue structural coloration and competition for nestboxes in male eastern bluebirds. Anim Behav 69:67–72.[CrossRef]
Smith HG, Montgomerie R, Poldmaa T, White BN, Boag PT. (1991) DNA fingerprinting reveals relation between tail ornaments and cuckoldry in barn swallows, Hirundo rustica. Behav Ecol 2:90–98.
Svensson L. (1992) Identification guide to European passerines. (Lars Svensson, Stockholm (Sweden)).
Szekely T, Freckleton RP, Reynolds JD. (2004) Sexual selection explains Rensch's rule of size dimorphism in shorebirds. Proc Natl Acad Sci USA 101:12224–12227.
Thomson DL, Furness RW, Monaghan P. (1998) The analysis of ordinal response data in the behavioural sciences. Anim Behav 56:1041–1043.[CrossRef][ISI][Medline]
Voigt CC, Heckel G, Mayer F. (2005) Sexual selection favours small and symmetric males in the polygynous greater sac-winged bat Saccopteryx bilineata (Emballonuridae, Chiroptera). Behav Ecol Sociobiol 57:457–464.[CrossRef]
West SA and Sheldon BC. (2002) Constraints in the evolution of sex ratio adjustment. Science 295:1685–1688.
Westneat DF. (2006) No evidence of current sexual selection on sexually dimorphic traits in a bird with high variance in mating success. Am Nat 167:E171–E189.[CrossRef]
Westneat DF and Stewart IRK. (2003) Extra-pair paternity in birds: causes, correlates, and conflict. Annu Rev Ecol Evol Syst 34:365–396.
Wiens JJ. (2001) Widespread loss of sexually selected traits: how the peacock lost its spots. Trends Ecol Evol 16:517–523.[CrossRef]
