, Volume 138, Issue 6, pp 657–665

Are large wattles related to particular MHC genotypes in the male pheasant?


    • Institute for Ecosystem Study, C.N.R.
  • Martina Ammannati
    • Department of Evolutionary Biology “L. Pardi”University of Florence
  • Claudia Magnelli
    • Department of Evolutionary Biology “L. Pardi”University of Florence
  • Alessandro Massolo
    • Department of Ecosystem and Public Health, Faculty of Veterinary MedicineUniversity of Calgary
  • Francesco Dessì-Fulgheri
    • Department of Evolutionary Biology “L. Pardi”University of Florence

DOI: 10.1007/s10709-010-9440-5

Cite this article as:
Baratti, M., Ammannati, M., Magnelli, C. et al. Genetica (2010) 138: 657. doi:10.1007/s10709-010-9440-5


In sexually dimorphic species, partners can assess heritable mate quality by analyzing costly sexual ornaments in terms of their dimension and possibly of their symmetry. In vertebrates an important aspect of genetic quality is the efficiency of the immune system, and in particular the Major Histocompatibility Complex (MHC). If ornaments are honest advertisements of pathogen resistance (good genes), in line with the Hamilton-Zuk hypothesis, a correlation between ornament expression and MHC profiles should exist. We tested this hypothesis in the common pheasant Phasianus colchicus by comparing male ornament characteristics (wattle and spur size, and wattle fluctuating asymmetry) with a portion of exon 2 of the class IIB MHC genes containing 19 putative antigen recognition sites. A total of 8 new alleles was observed in the MHCPhco exon IIB. We found significant differences in the occurrence of MHC genotypes between males carrying large or small wattles. Homozygous genotypes predicted large wattle males more correctly than small wattle males. The association between the dimension of the spur and the occurrence of MHC genotypes was marginally significant, however, we did not find any significant association between MHC genotypes and asymmetry. Our results suggest that female pheasants may use the ornament size as a cue to evaluate male quality and thus choose males carrying particular MHC profiles.


CE-SSCPGood genesHamilton-Zuk hypothesisMHCOrnamentsSymmetry


The hypothesis that mate choice allows individuals to acquire good genes for offspring is now widely accepted. Ornaments that are honest signals of good genes must be both costly and accurately discriminate between low and high quality individuals (Zahavi 1975; Grafen 1990). Since the seminal publication of Hamilton and Zuk (1982), genes related to the development of ornaments are now considered as likely to provide disease resistance (for review Bernatchez and Landry 2003; Mays and Hill 2004; Hunt et al. 2004; Neff and Pitcher 2005; Milinski 2006). A trade-off between the immune system and ornaments is hypothesized. Only individuals in prime physical condition with a very efficient immune systems can afford costly ornaments (Andersson 1994). A related idea is that male attractiveness is associated with the symmetry of bilateral ornaments (Mǿller 1992, 1999, 2006). The degree of asymmetry, measured as fluctuating asymmetry (FA) (Mǿller 1992; Palmer and Strobeck 1992), appears to reflect vulnerability to environmental stress in accordance with Hamilton and Zuk (1982).

The common pheasant is a good model to investigate the association between ornaments and genotypes. In this polygynous species, males do not provide paternal care and are territorial during the breeding season. Male pheasants have multiple ornaments: elongated tail feathers, spurs, ear tufts and wattles. The wattle, which has an important role during male courtship (Johnsgard 1999), is of particular interest. Its size is positively correlated with viability, measured as the ability to evade predators (Papeschi and Dessi-Fulgheri 2003), circulating testosterone levels (Briganti et al. 1999; Papeschi et al. 2000), and linked to early nutritional conditions (Ohlsson et al. 2002). The spur is another sexually dimorphic trait and older individuals have larger spurs (Johnsgard 1999). Females tend to prefer males with longer spurs and they use spur size to evaluate male age (Grahn and von Schantz 1994; von Schantz et al. 1996, 1997). The species is also known to be exposed to a variety of parasites (e.g. coccidia of the genus Eimeria and the roundworm Heterakisgallinarum) showing heritability of parasite resistance and a negative relationship between parasite load and courtship vigour (Hillgarth 1990).

The Major-Histocompatibility-Complex (MHC) is often viewed as one of the best examples of adaptive genes (Klein 1986; Andersson 1994; Trowsdale and Parham 2004; Danchin and Pontarotti 2004; Kasahara et al. 2004; Kelley et al. 2005). The MHC is a multigene family of co-dominant and extremely polymorphic genes at the interface between the immune system and infectious diseases. MHC diversity at particular loci often results from specific host-parasite co-adaptation cycles, suggesting that the MHC can be used to measure the genetic basis of mate choice based on costly ornaments. However, to our knowledge only five studies have reported a correlation between MHC genotypes and ornaments: tarsal spurs in the common pheasant Phasianus colchicus (von Schantz et al. 1996, 1997), antlers in the white-tailed deer Odocoileus virginianus (Ditchkoff et al. 2001), snood in the turkey Meleagris gallopavo (Buchholz et al. 2004), breeding colouration in the three-spined stickleback Gasterosteus aculeatus (Jäger et al. 2007) and the train length in the peacock Pavo cristatus (Hale et al. 2009). Ornament size generally reflected particular MHC genotypes in all the above studies. In birds, the MHC of the fowl Gallus gallus domesticus is the only well-characterized avian MHC (Guillemot et al. 1988; Kroemer et al. 1990), but a similar MHC organization seems to characterize the common pheasant (Jarvi and Briles 1992; Jarvi et al. 1996; Wittzell et al. 1994, 1998, 1999; von Schantz et al. 1996). The pheasant MHC (MHCPhco) is highly polymorphic with two unlinked gene clusters, which correspond to those found in the fowl (BLB and YLB genes, for the black grouse, Tetrao tetrix, see Strand et al. 2007), both comprising class I and IIB genes (3–4 genes). The exon 2 of the B complex coding for the β1 chain is the most variable part of MHC class IIB genes and it is commonly used for population genetic analyses (Ditchkoff et al. 2001; Oliver and Piertney 2006; van Oosterhout et al. 2006; Zhang et al. 2006).

In the present study we investigated the relationships between ornaments and MHC variability in the common pheasant. We compared the ornament dimensions (wattle size, tarsal spur size) and the wattle fluctuating asymmetry in groups of male pheasants with similar genetic profiles defined on the basis of a portion of the exon 2 of the class IIB MHC genes. We hypothesized that individuals, exposed to parasites and characterized by large ornaments, are expected to have different MHC profiles from those with small ornaments, in line with the Hamilton-Zuk hypothesis.

Materials and methods

Animals and sampling procedures

We studied 79 yearling pheasant males (10 months old). At this age pheasants are sexually mature (Johnsgard 1999). The animals were bred at the Montepaldi farm, University of Florence. To avoid the effects of domestication and inbreeding, every year this farm introduces 20 first generation males, caught in the wild from different areas of Tuscany as breeders (for a total of 40). Microsatellite markers show that these individuals were unrelated (Baratti et al. in preperation). After hatching, the pheasants were kept under controlled temperature and at 40 days were transferred to large aviaries with a density of 0.1 ind./m2. Food and water were provided ad libitum.

Biometric measures

Measurements were carried out between January and February 2002 during the reproductive and territorial phase of males, when inter-male aggressiveness occurs. We used the vertical wattle length as a size indicator (MVW Mean Vertical Wattle: averaging left and right measures), as the height and the area of the wattle are highly correlated in pheasants (Papeschi et al. 2000; Ohlsson et al. 2002). The wattle was measured at the points of maximum vertical height. We also measured the right spur length (S) and tarsus length (T). The tail was not considered because in captive animals it is often worn or broken. All the measures were taken to the nearest 0.1 mm. The precision of wattle measurements, estimated using both the standard deviation (SD) and the coefficient of variation (CV) of a series of 10 intra-individual measurements, was considered acceptable for our study purposes (SD = 0.681 mm; CV = 2.156%).

We calculated the value of asymmetry of the wattle (the spur was not considered, as frequent accidental breaks occurring in captivity may bias the results) following the equation (suggested by Kellner and Alford 2003):
$$ I_{\text{Asymmetry}} = (X_{\text{right}} - X_{\text{left}} )/[(X_{\text{right}} + X_{\text{left}} )/2] $$
where X = vertical wattle size.

Genetic samples

Blood samples were taken from each individuals and stored into tubes containing EDTA and kept at −80°C until DNA extraction. Genomic DNA was isolated from whole blood using a Puregene D-5000A isolation kit (Gentra Systems, Minneapolis, USA).

Amplification of a section of the MHCPhco class II β second exon was performed by PCR as in Wittzell et al. (1994). We tested three different pairs of oligonucleotides (INT3-DR1, 471-DR1, B44-DR1; Wittzell et al. 1994). As we did not obtain specific PCR products for INT3-DR1, 471-DR1, we used the forward primer B44 (5′-ACCCAGCAGGTGAGGCATGTG) and the reverse primer DR1 (5′-GCTCCTCTGCACCGTGAAGGA); the first being complementary to the 5′ end of the exon, and the second to the 3′ end.

MHC polymorphism screening

To detect the presence of mutations by the electrophoresis variance of the molecules, we used the screening of genetic polymorphisms by the capillary electrophoresis-single strand conformation polymorphism technique, CE-SSCP (Kourkine et al. 2002).

The samples were prepared as follows: 1 μl of PCR product was added to 10.5 μl formamide, 0.5 μl 0.3 N NaOH and 0.75 μl GeneScan Standard TAMRA 500, then denatured for 2′ at 90°C and cooled on ice. Samples were then loaded onto a 47 cm length, 50 μm diameter capillary filled with a polymer containing 10% glycerol and 1× AB Buffer 310 and the concentration was tested at 3 and 4%. The electrophoresis conditions showed that a 15 kV EP voltage and the 4% polymer concentration-30°C hotplate temperature (24 min) combination was best for peaks separation with run conditions at optimal levels. To allow the differentiation of the two single strands, we used 5′-end labeled primers for both DNA strands, with different dyes (HEX for the forward and TET for the reverse). As an additional internal standard, we added to each electrophoresis run a labelled 6-FAM PCR product of a known sequence.

We selected homozygous individuals for direct sequencing using an ABI kit (Big Dye Terminator Sequencing v. 2.0-ABI PRISM, PE Biosystem). We chose different heterozygotes, cloning them using a pGEM®-T Easy Vector System (Promega). All sequences were aligned and compared to those previously described for the this species (Wittzell et al. 1994; von Schantz et al. 1996).

Due to problems in generating PCR- and cloning artefacts in analyses of polymorphic gene complexes, we decreased the number of cycles till a reliable amplification fragment was visible (30 cycles) and established criteria to avoid misinterpretation of alleles. An allele was considered to be real only when it occurred in more than one clone, from the same or from different individuals, and when it was supported by two independent PCRs. As a control, clones were also run in CE-SSCP analysis.

All CE-SSCP analyses and sequencing were performed by a 310 ABI® automated sequencer. GeneScan software version 3.7 (PE Applied Biosystems) was used to visualize CE-SSCP profiles.

Phylogenetic analyses

We tested for the evolutionary model that best fit our data by MODELTEST v. 3.04 (Posada and Crandall 1998). Based on the Akaike Information Criterion (AIC), we selected the JC + I model as the best model (Jukes and Cantor 1969). A Bayesian analysis (BI) was performed with MrBayes ver. 3.1 (Ronquist and Huelsenbeck 2003) using the parameters of the substitution model suggested by Modeltest 3.7: 1,000,000 generations with the tree sampled every 100 generations. The summaries of the Bayesian inference relied on 20,000 samples (from two runs). A consensus tree was constructed by MrBayes 3.1 from the remaining trees after burning 1,000. Support for nodes was assessed with the posterior probabilities of reconstructed clades as estimated by MrBayes 3.1 (Ronquist and Huelsenbeck 2003). Representative sequences of avian MHC IIB exon 2 genes were searched on GenBank ( to infer the monophyly of pheasant alleles (Table 1). A sequence from a human class IIB gene was used as an outgroup (Genbank Acc. No AB046526).
Table 1

Avian taxa included in the phylogenetic analysis with the GenBank Accession Numbers



Genbank Acc No

Homo sapiens



Gallus gallus



Phasianus colchicus (DAB*)























Andropadus virens



Aphelocoma coerulensis



Petroica australis



Passer domesticus



Melanospiza richardsoni



Statistical analyses

Chromas v. 2.01 (Technelesyum Pty Ltd, Australia) was used to correct chromatograms. The sequences, aligned with CLUSTAL X version 1.83 (Thompson et al. 1997) were translated into aminoacids using MEGA version 3.1 (Kumar et al. 2004).

The ratio of non-synonymous (dN) to synonymous (dS) substitutions was calculated in MEGA 3.1, using the Nei-Gojobori method with the Jukes-Cantor correction for multiple substitutions (Nei and Gojobori 1986) and with 1,000 bootstrap replicates to obtain standard errors. The location of the putative PBR codons was inferred from published MHC sequences described in Westerdahl et al. (2000). The PBR codons are expected to experience positive Darwinian selection (Nei and Kumar 2000), with dN/dS ratios in excess of unity. By contrast, purifying selection is generally found to govern sequence variation outside the PBR, probably owing to functional constraints of the protein. The Z-test, as implemented in MEGA, was performed to verify positive selection acting on the MHC fragment. The D Tajima neutrality test (Tajima 1989) was also performed to test the hypothesis that all mutations were selectively neutral. The significance level of the D test was set at 0.10 as suggested by Aguilar et al. (2006). Genetic distance among all MHCPhCo alleles was calculated based on Kimura’s two-parameter model (Kimura 1980).

Differences in wattle and spur size in MHC genotypes were explored by two different statistical approaches. In the first approach, we performed an analysis of variance (Searle 1982) with the mean vertical wattle (MVW) or the spur length (S) as dependent variables and the genotypes as fixed factors. Then, an ANCOVA was carried out with the tarsus length as a covariate. The sums of squares of an effect in the design were computed using type III in which the sums of squares adjusted for any other effects that were not contained and orthogonal to any effects (if any) that were contained. Relations between wattle asymmetry and MHC genotypes were estimated by ANOVA (Sokal and Rohlf 1995).

In the second approach, as females likely use a rule of thumb (Dawkins 1976; Janetos and Cole 1981) to easily distinguish among males with different ornament sizes, we divided the individual values of each measurement into three groups on the basis of percentiles (33 and 66%) and then we only used the two extreme groups of each distribution. For this analysis only genotypes belonging to the first and the third groups were considered. Thereafter, we used Logistic Regression analysis (Hosmer and Lemeshow 1989) to compare small (wattle < 34.0 mm; spur < 10.2 mm; asymmetry < 0.031) vs. large (wattle > 35.9 mm; spur > 12.0 mm; asymmetry > 0.067) ornament characteristics (dimension and asymmetry) on the basis of their genetic profiles (genotypes or alleles). Models were tested by Omnibus tests of model coefficients, and model fitting was estimated by the Nagelkerke’s R2 (Nagelkerke 1991). We also estimated model fitting by the number of correctly classified cases, using the “leave one out” classification procedure that consisted in developing a model leaving 1 case out from the sampling set and testing the model to verify if it was correctly classified; the same procedure is repeated for each case. The Fisher exact test was used to compare the frequency of large and small wattle individuals in homozygosis and heterozygosis (Sokal and Rohlf 1995).

All statistical analyses were computed using the Statistical Package for Social Sciences ver. 15.05 (SPSS®). Within the ANOVA and the ANCOVA analyses we included only those genotypes that occurred in at least 2 individuals.


MHC analyses

All 79 individuals were analysed by CE-SSCP, but 9 pheasants were excluded from the subsequent analyses as their CE-SSCP profiles were unclear even after repeated electrophoresis runs. The remaining 70 individuals, with one or two alleles, were genotyped. All specimens that showed the same CE-SSCP profile consistently revealed the same MHC genotype after sequence analysis (i.e. technique was consistent for genotype scoring of our samples).

We amplified a section of MHC class IIB exon 2 between codons 17 and 83, containing 19 possible binding sites for foreign peptide presentation (PBR) (out of the 24 PBR sites included in the entire exon IIB). We identify 8 MHC alleles that were never described before for the species (Phco-Tus1/8, also called in the text and in the Fig. 2a–h), which formed 15 different genotypes. The exon portion translates into a sequence length of 66 aminoacids (Table 2).
Table 2

Alignment of partial MHC IIB aminoacid sequences of the eight alleles, with the GenBank Access Numbers compared with Gallus (as reported in Table 1).

Shaded columns identify the peptide binding regions (PBR) sites

The proportion of non-synonymous substitutions (dN) compared to synonymous ones (dS) was detected considering PBR and non-PBR regions. PBR codons displayed a significantly higher dN than non-PBR codons (0.309 vs. 0.048). In the PBR codon sequence, the dN exceeded the dS (dN/dS = 1.64), but not in non-PBR sites (dN/dS = 0.80), although the Z-test for inequality of dN and dS in both regions was not significant (PBR: P = 0.239; non-PBR: P = 0.683). The D test was not significant in the non-PBR (D = 0.82, P > 0.10), but significant in the PBR codons at a level of 0.10 > P > 0.05 (D = 1.84).

The K2P distances between the different alleles ranged from 2% (PhCo-Tus1 vs. PhCo-Tus2) to 15% (PhCo-Tus1 vs. PhCo-Tus8). In the phylogenetic analysis the eight alleles were grouped in a monophyletic clade. The analysis included the eight novel alleles and the already published pheasant alleles (as in Wittzell et al. 1999; Fig. 1).
Fig. 1

Bayesian majority rule (50%) consensus tree of exon2 of MHC class IIB gene sequences, with the posterior probability values at the nodes. Other avian MHC exon IIB sequences (as reported in Table 1) are included

MHC genotypes versus ornament size and symmetry

The ANOVAs, performed with the wattle and the spur as the dependent variables and the MHC genotypes as fixed factors and the ANCOVA with the tarsus length as covariate, only showed marginally significant differences in ornaments between the different genotypes (wattle, ANOVA: F8,55 = 1.812, P = 0.094, ANCOVA: F8,53 = 1.857, P = 0.086; spur, ANOVA: F8,55 = 1.895, P = 0.079, ANCOVA: F8,53 = 1.882, P = 0.082). Further, no significant differences in ornament dimensions were detected among different MHC alleles, except for the wattle, which was only marginally significant (wattle, ANOVA: F7,132 = 1.879, P = 0.078, ANCOVA: F7,130 = 1.762, P = 0.100; spur, ANOVA: F7,132 = 1.381, P = 0.219, ANCOVA: F7,130 = 1.376, P = 0.221).

On the other hand, the logistic regression analysis showed that different genotypes characterised small and large wattle pheasants (χ2 = 24.069, df = 12, P = 0.020; Nagelkerke R2 = 0.529; Fig. 2), but only a marginal significance was detected considering the alleles (χ2 = 13.200, df = 7, P = 0.067). The genotype characteristics were sufficient for correctly classifying 85.2% (23/27) of large, but only 61.9% of small wattle individuals (13/21). Overall, the 75.0% of cases were correctly classified. A marginal significance was found when comparing small vs. large spur individuals for the genotypes (χ2 = 17.362, df = 11, P = 0.098), but not for alleles (χ2 = 8.930, df = 6, P = 0.178). Homozygote genotypes were predominantly represented by large wattle individuals (8 out of 9: χ2 = 5.444, df = 1, P < 0.039), while there were no significant differences in the occurrence of heterozygosis associated with wattle dimension (Hetero: 20 small vs. 19 large: χ2 = 0.026, df = 1, P = 1.000).
Fig. 2

Number of individuals (Y-axis) carrying MHC genotypes (X-axis). In white: small wattle group (first percentile group < 33% = 34.0 mm; n = 21); in black: large wattle group (third percentile group > 66% = 35.9 mm; n = 27). Alleles names a–h corresponds to Phco-Tus1/8

The ANOVA performed with asymmetry of the wattle as the dependent variable and genotypes/alleles and wattle dimensions as fixed factors did not show a significant effect of MHC genotypes (F = 1.271, df = 8, P = 0.280) or alleles (F = 1.038, df = 7, P = 0.408) on asymmetry. The logistic regressions also showed no significant association between the wattle asymmetry and the genotypes (χ2 = 13.164, df = 12, P = 0.357), or with the alleles (χ2 = 4.024, df = 7, P = 0.777).


Many studies have tested the possible role of selection in shaping MHC diversity (cf. Bernatchez and Landry 2003). The long term effects of natural and sexual selection are apparent, for example, from higher non-synonymous vs. synonymous substitution rates observed in peptide presenting domains of MHC proteins (e.g. Van der Walt et al. 2001; Bernatchez and Landry 2003). The high dN/dS ratios for PBR have been attributed to the action of balancing selection, probably mediated by parasites. We provided evidence for balancing selection on the MHC of pheasants with the finding of a much higher dN value in the 19 PBR sites (0.302) compared to the non-PBR sites (0.048). Considering all alleles, the dN/dS ratio at the 19 PBR sites in pheasants was 1.56, while at the non-PBR was <1 (0.80), although the Z-test was not significant for either gene portions. Two MHC class IIB genes, DAB1 and DAB2, arose through a duplication before the split between chicken and pheasant (20MYA), and they differed mainly in the 3′UT region, but with no transcriptional differences (Wittzell et al. 1999). We found eight different PhCo-MHC sequences, each with a distinct amino acid structure at the PBR sites.

The number of MHC alleles found here is similar to that reported for some threatened birds species (Zhang et al. 2006; Bollmer et al. 2007), but lower than that reported for other birds (Ekblom et al. 2007). Limited numbers of MHC alleles were commonly observed for animal species that have gone through known population bottlenecks (Hedrick et al. 1999, 2000; Hoelzel et al. 1999, Oliver and Piertney 2006; Zhang et al. 2006), as is likely to be the case for the pheasant, which is subject to severe hunting in Italy.

We cannot exclude that the two MHCPhco class IIB loci were amplified (DAB1 and DAB2, for details see Wittzell et al. 1999). Some of the alleles (PhCoTUS1, a, PhCoTUS2, b, PhCoTUS6, g) presented a low nucleotide divergence from DAB04 (corresponding to DAB2*04 in Wittzell et al. 1999), and grouped together in the phylogenetic tree. On the other hand, the presence of multiple alleles (>2 per individual) in our samples was suspicious only for the low percentage (11%) found in those nine individuals presenting unclear CE-SSCP profiles. It is important to note that all clear and reliable profiles did not show more than four peaks per individual, suggesting that only one class IIB locus was amplified. One cloned individual, out of the nine removed from the analysis, presented six different sequences, but after alignment we considered them to be cloning artefacts that have been demonstrated for MHC loci in other species (Longeri et al. 2002).

It is generally accepted that females choose high quality males through ornament assessment and thus assure good genes and high survival for their offspring (Hamilton and Zuk 1982; Ohlsson et al. 2002; Penn 2002; Radwan 2008). Studies of MHC variation and phenotypic traits of non-model vertebrates are rare (von Schantz et al. 1996, 1997; Ditchkoff et al. 2001; Buchholz et al. 2004; Jäger et al. 2007; Hale et al. 2009). This scarcity of published studies, despite extensive theoretical development, may be due to confounding variables that affect the physical condition of individuals and hence the expression of their ornaments (Bernatchez and Landry 2003; Piertney and Oliver 2006).

Nevertheless, in studies regarding association between MHC and phenotypic traits, significant positive correlations between morphological traits (spur, snood and train length, antler development, breeding colouration) and MHC genotypes were detected. In particular, female pheasants prefer long-spurred males, probably to assure the offspring with good genes for disease resistance (Von Schantz et al. 1997). Using ornaments for mate selection, females will mate with healthier males, gaining direct benefits, or, as far as immune function is heritable, indirect benefits.

The results of the present study reported the existence of a relation between MHC profiles and sexual ornaments, suggesting that genotypes predicted large wattle males more correctly than small wattle males (85.2 vs. 61.9% respectively). These results advise that female pheasants could use wattle dimensions to evaluate male quality choosing males carrying particular MHC profiles. The association between MHC genotypes and large wattles, a testosterone-dependent trait related to male fitness (Briganti et al. 1999; Papeschi and Dessi-Fulgheri 2003), suggests that the wattle could be a good gene indicator and consequently a good candidate for female mate choice. In fact, Papeschi (1998) described female preference for large wattles of territorial males. However, we found that the correlation between MHC genotypes and ornament size was most clear when considering variant classes. If we considered the trait as a continuous variable, we obtained only marginal significance results. This difference makes sense if, as it is apparently the case, females use a rule of thumb to evaluate this ornament, which undergoes daily and seasonal size variations.

According to many studies, female choice for MHC good genes and/or complementary genes is fundamentally different: choice for ‘good genes’ assumes additive gene action, while choice for compatibility assumes over-dominance (i.e. non-additive genetic action) (Lehmann et al. 2007). The occurrence of large wattles mostly in homozygous individuals seems to sustain the hypothesis of a good gene model of mate choice, with an association between large ornamented males and good genes. As expected from a traditional good-genes scenario, AA males should be in better condition compared to heterozygous males, suggesting an advantage of the homozygote (Lehmann et al. 2007).

At first glance our data appear not to fully corroborate previous studies which showed a significant correlation between pheasant MHC variation and spur dimension (von Schantz et al. 1996; 1997). However, we noted a marginally significant correlation between spur dimension and MHC variation. It is reasonable that both ornaments are related to MHC variation and that females by choosing mates using spur and wattle are selecting non-random MHC genotypes.

We were not able to detect any significant relationship between symmetry (as an indicator of developmental stability) and MHC profiles. Possibly, farm bred pheasants do not experience substantial developmental stresses, because they live under controlled temperature in early life and have life-long access to ad libitum food and water. Further, our result could be also a consequence of the fact that we used only one gene in our analysis. In fact, fluctuating asymmetry is a character whose genetic determination remains still largely unknown and little evidence is available regarding specific genes that drive asymmetry (Leamy and Klingenberg 2005; Polak and Taylor 2007). FA levels in various characters seem to be influenced by dominance and epistatic interactions among different genes affecting these or other characters (Leamy and Klingenberg 2005; Johnson et al. 2008).

The study of genetic variation at selective loci can provide evidence of important adaptive processes (Hedrick et al. 2001). Although other genes are now available for this kind of studies, MHC remains the best choice (Hedrick 1999; Bernatchez and Landry 2003). MHC variability is maintained by pathogen-driven selection and reproductive mechanisms (disassortative mating, preference for heterozygosity, and post-copulatory control; see Sommer 2005; Milinski 2006; Piertney and Oliver 2006). In birds, correlations between male ornaments and male quality are particularly important for conservation and management purposes. The common pheasant is a game species, particularly appreciated all over Europe where it is widely farm-bred for restocking the various wild populations, and it is subject to a wide range of hunting pressures and environmental conditions. Identification of high quality farm-bred individuals adapted to survive in the wild is thus a priority.

Although our understanding of the MHC is improving, we are still far from fully understanding the genetic characteristics of males selected by females, particularly in birds (Mays et al. 2008). MHC genes apparently undergo sexual selection for both good genes and compatible genes (Mays and Hill 2004; Neff and Pitcher 2005). Future studies will likely need to focus on clarifying the interactions among these factors, to put further light on the evolution of sexual selection.


We thank R. Stanyon, G. Bertorelle, M. Ciuffreda and S. Mona for comments and suggestions. We thanks the CIBIACI sequencing centre for optimizing the CE-SSCP technique. This research was supported by grants from the Italian Ministry (PRIN/2005). We also thank G. Sanders for English revision.

Copyright information

© Springer Science+Business Media B.V. 2010