MHC-associated mating strategies and the importance of overall genetic diversity in an obligate pair-living primate
Mate choice is one of the most important evolutionary mechanisms. Females can improve their fitness by selectively mating with certain males. We studied possible genetic benefits in the obligate pair-living fat-tailed dwarf lemur (Cheirogaleus medius) which maintains life-long pair bonds but has an extremely high rate of extra-pair paternity. Possible mechanisms of female mate choice were investigated by analyzing overall genetic variability (neutral microsatellite marker) as well as a marker of adaptive significance (major histocompatibility complex, MHC-DRB exon 2). As in human medical studies, MHC-alleles were grouped to MHC-supertypes based on similarities in their functional important antigen binding sites. The study indicated that females preferred males both as social and as genetic fathers for their offspring having a higher number of MHC-alleles and MHC-supertypes, a lower overlap with female’s MHC-supertypes as well as a higher genome wide heterozygosity than randomly assigned males. Mutual relatedness had no influence on mate choice. Females engaged in extra-pair mating shared a significant higher number of MHC-supertypes with their social partner than faithful females. As no genetic differences between extra-pair young (EPY) and intra-pair young (IPY) were found, females might engage in extra-pair mating to ‘correct’ for genetic incompatibility. Thus, we found evidence that mate choice is predicted in the first place by the ‘good-genes-as-heterozygosity hypothesis’ whereas the occurrence of extra-pair matings supports the ‘dissassortative mating hypothesis’. To the best of our knowledge this study represents the first investigation of the potential roles of MHC-genes and overall genetic diversity in mate choice and extra-pair partner selection in a natural, free-living population of non-human primates.
KeywordsMate choice MHC class II Microsatellites Pair-living lemur Extra-pair partner Cheirogaleus medius Madagascar
Mate choice is one of the most important evolutionary mechanisms. Increasing evidence indicates that females can improve the viability of their offspring by selectively choosing certain males (Promislow et al. 1998). Females could gain direct benefits from their choice (e.g. provision with resources, superior paternal care or high-quality territories) or receive indirect benefits through genes that confer increased offspring fitness (Andersson 1994; Ryan 1997; Kokko et al. 2003). Genetic benefits can be obtained if the risk of genetic incompatibility between maternal and paternal genomes is minimised by avoiding mating with close kin (inbreeding avoidance), or by increasing the genetic heterozygosity or diversity within the progeny (reviewed by Tregenza and Wedell 2000; Hansson and Westerberg 2002). Thereby, different alternative mechanisms have been proposed. The ‘disassortative mating hypothesis’ states that females choose genetically dissimilar males, which would result in offspring being genetically different from each parent (Zeh and Zeh 1996; Penn and Potts 1999; Tregenza and Wedell 2000; Bernatchez and Landry 2003). Similarity might be measured by the number of shared alleles, by genetic distances between the partners or by the relatedness (Queller and Goodnight 1989). The ‘good-genes-as-heterozygosity hypothesis’ (Brown 1997, 1999) states that females may chose mates with a high level of heterozygosity providing direct benefits but also potentially leading to offspring that are more diverse on average. Candidates for ‘good genes’ also include genes for vigour and health (Hamilton and Zuk 1982), e.g. specific alleles for resistance to various pathogens.
Particularly suited as potential candidates for the genetic basis of mate choice in vertebrates are the genes of the major histocompatibility complex (MHC) (Tregenza and Wedell 2000), a multigene family encoding cell-surface glycoproteins (MHC molecules) that present small peptide antigens to T-cells of the immune system which initiates the appropriate immune response (Klein 1986; Janeway and Travers 2002). Besides its central immunological function, the MHC has been recognized as a possible source of individual specific body odours and therefore provides a cue for individuals to distinguish MHC-identities (reviewed in Penn 2002). The genes of MHC are the most polymorphic loci in the vertebrate nuclear genome (Robinson et al. 2003). The high MHC diversity could potentially be driven by pathogen-mediated selection or by sexual selection which are not mutually exclusive (Apanius et al. 1997; Edwards and Hedrick 1998; Penn and Potts 1999). If heterozygotes are able to mount an immune response against a broader array of pathogens than homozygotes, females should be choosy in order to increase the survival capability of their progeny (Brown 1997; Tregenza and Wedell 2000). Thus, the connection between the MHC and mate-choice is presumably a consequence of the central role of the MHC in the immune system with either heterozygotes or specific or genetically dissimilar alleles being selectively favoured (Penn and Potts 1998; Potts 2002). Evidence for MHC-associated mating strategies have been found in mice (Yamazaki et al. 1976; Egid and Brown 1989; Eklund 1997; Penn and Potts 1998; Roberts and Gosling 2003), birds (Freeman-Gallant et al. 2003), sand lizards (Olsson et al. 2003), fishes (Landry et al. 2001) and also in humans (Wedekind et al. 1995; Ober et al. 1997). Other studies did not find a clear pattern of MHC-associated mating preferences (Eklund 1997; Hedrick and Black 1997; Paterson and Pemberton 1997; Wenink et al. 1998; Garside et al. 2000; Westerdahl 2004; Sommer 2005a) and empirical data from non-model species in natural populations are rare (Sommer 2005b; Piertney and Oliver 2006).
Especially interesting and suitable models for investigations of the genetic basis of female mate choice are species living in obligate pair bonds but engaging in extra-pair matings. Extra-pair males only contribute genes to their offspring whereas social mates often provide parental care as well as genes (Mays and Hill 2004). A female has the choice of accepting the genotype of her social mate as a sire of her offspring or choosing among potential extra-pair sires. Moreover, females might be using different rules for different sorts of mates, such that social mates are chosen on the basis of cues that reflect the ability of a male to provide parental care and genetic fathers are chosen on the basis of their genetic quality (Mays and Hill 2004).
We investigated MHC-associated mating strategies and the importance of overall genetic diversity in an obligate pair-living primate, the fat-tailed dwarf lemur Cheirogaleus medius (Cheirogaleidae, Primates), a small nocturnal primate (about 130 g) occurring in the dry deciduous forests of western Madagascar. Males and females live in lifelong pair bonds and usually separate only when one partner dies (Fietz et al. 2000; Fietz and Dausmann 2003). Both sexes take extensive care of their offspring, and this is obligatory for the survival of the offspring. Observations showed that mothers, whose male partners died before or shortly after the birth of the young, were not able to rear their offspring successfully (Fietz 1999a, b; Fietz and Dausmann 2003). Despite of this social obligate monogamy, genetic parentage analysis revealed an extraordinary high rate (44%) of extra-pair young (EPY) which interestingly does not seem to reduce the cuckolded social partners’ care investment into the offspring (Fietz et al. 2000). Since there is no sex dimorphism in this species, and males (at least in captivity) cannot coerce copulation (Foerg 1982), females must search for extra-pair copulations and advantages from this behaviour are expected.
How do females choose social fathers for their offspring which are obligate for successful raising of offspring?
Is there female choice for genetic fathers? Are social fathers and genetic fathers chosen according to the same rules?
Why do females engage in extra-pair copulations? Do genetic differences between social partners of faithful females and cuckolded social partners exist? Do differences between the cuckolded social partner and the genetic father of EPY occur?
Are there differences in the genetic constitution between IPY and EPY?
Materials and methods
Study area and sample collection
Sampling was carried out in western Madagascar, in the dry deciduous Kirindy Forest located about 60 km northeast of Morondava. The research area of about 25 ha (500 × 500 m) is part of a forestry concession of the “Centre de Formation Professionelle Forestière de Morondava, CFPF”. The region is characterised by a strong seasonality with a dry season of about 8 months and a rainy season. C. medius spends about 7 month hibernating during the dry season in this region (Dausmann et al. 2004). Thus, the reproduction is limited to the short rainy season (about Nov/Dec–April) where females give birth to one or two young.
A detailed description of the forest and trapping conditions is given elsewhere (Ganzhorn and Sorg 1996; Fietz 1999a, b; Fietz and Dausmann 2003). Briefly, captures were carried out for four consecutive nights per month using 200 Sherman Traps (7.7 × 7.7 × 30.5 cm) which were placed in 50 m intervals at the intersections of a grid system of trails. The traps were opened in the late afternoon, baited with banana and checked early the next morning. Captured individuals were sexed, anaesthetised for tissue collection and individually marked with subdermally injected transponders. The animals were released at their capture sites in the late afternoon of the same day. In total, tissue samples for genetic analyses from 259 individuals were taken (rainy seasons 1995–2004).
Overall genetic variability was assessed on the basis of seven microsatellite loci. The PCR and analysis of the microsatellite loci was conducted using an extended dataset with the primers and conditions as described in Fietz et al. (2000). Adaptive variability was studied in the highly polymorphic MHC-DRB exon 2 (171 bp, without primer) which includes the functionally important antigen-binding sites (ABS) (Brown et al. 1988, 1993). The PCR amplification was carried out using the primers JS1 and JS2 as described in Schad et al. (2004, 2005) which were originally designed for the closely related lemur species Microcebus murinus but were also successfully applied in different rodent species (Froeschke and Sommer 2005; Harf and Sommer 2005; Meyer-Lucht and Sommer 2005). To genotype the individuals, we used single-strand conformation polymorphism analysis (SSCP) (Orita et al. 1989a, b; Meyers and Bull 2002). PCR products were loaded on 15% polyacrylamide gels following the manufacturer’s instructions (ETC, Elektrophoresetechnik, Kirchentellinsfurt, Germany). Runs were carried out using a horizontal cooling electrophoresis system (Amersham Pharmacia Biotech, Freiburg, Germany). After electrophoresis the gels were fixed and silver stained to visualise the resulting bands. Bands were re-arranged and classified into alleles. Each allele was sequenced bidirectionally. Therefore, at least two different samples of each allele were cut out of the gel and dissolved in 1× TBE-buffer. Subsequent reamplification was carried out under the same PCR conditions as before but with a reduced number of cycles. Cycle sequencing was performed using a dye terminator kit (Applied Biosystems, Foster City, CA) on an Applied Biosystems sequencer (Model 3100) following the manufacturer’s protocol. Details on the molecular techniques are outlined in Sommer et al. (2002), Sommer (2003) and Schad et al. (2004).
Analyses and statistical treatment
Paternity analysis was performed by the application of microsatellites for all young for which information about their social parents was available from field observations (details in Fietz et al. 2000) using the software CERVUS (Marshall et al. 1998). As measuring units of the overall individual genetic diversity we used the observed individual microsatellite heterozygosity (Hobs) and the genetic distance between microsatellite alleles (d² value). Hobs was calculated by dividing the number of heterozygous microsatellite loci per individual by the total number of typed loci. d² results from the squared difference in repeat units between two alleles at a locus averaged over all typed loci and was calculated as d² = 1/n Σn (ai − aj)², where ai and aj are the length in repeat numbers of each allele at a locus averaged over n typed loci (Slate and Pemberton 2002; Coltman and Slate 2003). As a measurement of inbreeding avoidance we quantified genetic relatedness between individual males and females using the relatedness measure of (Queller and Goodnight 1989) implemented in the computer program SPAGeDi (Hardy and Vekemans 2002).
MHC class II DRB sequences were edited and aligned using MEGA 3 (Kumar et al. 2004). Individual levels of MHC allele diversity were calculated by the mean number of amino acid differences. Similarity between the MHC genotypes of two individuals was assessed by the least possible sum of pairwise amino acid differences including all alleles of the individuals (Ekblom et al. 2004).
Even though the highly polymorphic MHC genes clearly play a crucial role in the immune response, their great diversity is a major obstacle in distinguishing allele-specific effects. Collecting sample numbers sufficient for definitive results is often not feasible. In humans, a new approach to reduce this problem was recently proposed by classifying alleles to supertypes based on shared antigen binding similarities. With respect to MHC class II DRB alleles, it was primarily developed in the context of vaccine design since supertypes comprise alleles with similar binding properties (Sette and Sidney 1999). Overlapping peptide binding capabilities of different MHC class I and class II alleles have been demonstrated in several human studies (e.g. Bertoni et al. 1997; Southwood et al. 1998; Trachtenberg et al. 2003; Lund et al. 2004). Briefly, in C. medius we checked for the presence of codon sites affected by positive selection (positively selected sites, PSS) which are the putative antigen binding sites. Therefore, we conducted a maximum-likelihood analysis by using the programme CODEML (included in PAML version 3.14 software package) (Yang 1997). In the next step, we defined MHC-supertypes by applying amino acid sequence based hierarchical clustering including all identified PSS as proposed by Doytchinova and Flower (2005). Details on the application of this approach in investigations on the adaptive value of MHC-supertypes in parasite resistance in C. medius are described in Schwensow et al. (2007).
In order to investigate MHC disassortative mating we generated a random model of female choice where we let each female choose randomly for 15,000 times between all males of the respective year to generate a null distribution. Subsequently we compared the observed values to the simulated values of random female choice. This analysis was carried out using RESAMPLINGSTATS 5.0.2 (Resamplingstats, Inc., Arlington, VA, USA). As the MHC DRB locus is duplicated in C. medius, the degree of heterozygosity was assessed by the individual number of different MHC-alleles (Ekblom et al. 2004) and MHC-supertypes, respectively. We also generated a random model of female choice with respect to heterozygosity in the same way as described above. To investigate the question why females engage in extra-pair copulations we compared (for every mate choice event) both cuckolded social partners to not-cuckolded social partners and to the known genetic fathers of EPY. Statistical tests were performed by SPSS version 11.5 (SPSS Inc., Chicago, IL, USA). Whenever our data fulfilled the requirements we used parametric tests, otherwise we applied non-parametric tests. Calculations are two-tailed and based on a significance level of α = 0.05.
Microsatellite typing: identification of mate choice events
Two hundred and forty seven individuals were typed at seven and 12 individuals at six microsatellite loci, respectively. Thereby, we were able to identify 43 mother–offspring pairs. For all offspring also information about their social fathers was available. Twenty six (60.5%) of the offspring were sired by their social fathers (intra-pair young, IPY), whereas 17 (39.5%) were sired by extra-pair males (extra-pair young, EPY). In addition, one extra-pair mate choice event was documented by a vaginal plug. In sum, we investigated 44 mate choice events of 20 reproducing females from the years 1995 through 2004. Eight of them produced offspring as well with their own social mate as with an extra-pair mate (plus one infidelity documented by a vaginal plug).
A total of 75 adult males were investigated, 21 of them were the social partners of the 20 reproducing females (one social partner of a female died and was replaced). Sixteen males were identified as genetic fathers. In 11 cases the social father was excluded as genetic father but the genitor is unknown.
The DRB exon 2 sequences of 149 MHC-genotyped individuals were highly variable. We found 74 polymorphic nucleotide positions leading to 50 different alleles. The nucleotide sequences are deposited in GenBank (Accession numbers EF194225–EF194272). The alleles differed at 1–42 nucleotide positions (average = 21.67, SE = 2.46). No indels were detected. Within one individual we found at least two and at most four different alleles indicating a duplication of the locus. Since no evidence for a pseudogene like nonsense or stop codons changing the reading frame were found we assume that both loci are functional. Also analysis of dN/dS ratios in classical antigen-binding sites and non-antigen binding sites supported this assumption (Schwensow et al. 2007). Each nucleotide sequence was transformed into a unique amino acid sequence. On the amino acid level 33 of the 57 positions were variable and alleles differed at between one and 24 sites (average = 14.42, SE = 2.05). We defined 11 MHC class II supertypes based on significant PSS using a hierarchical amino acid sequence-based clustering considering physiochemical similarities between all identified alleles (see Schwensow et al. (2007) for details).
No correlation between the number of MHC-alleles and neutral heterozygosity (Hobs: Spearman correlation, ρ = 0.81, ns) as well as between the number of MHC-alleles and D² (Spearman correlation, ρ = 0.23, ns) could be detected.
Female choice for social fathers
Results of all tests for female mate choice
The relatedness (calculated from the microsatellite data set) between females and the chosen social fathers of their offspring did not differ from the relatedness of females to the randomly assigned males (observed = −0.03, simulated = −0.02, ns; Table 1).
Whereas the d²-value (calculated from the microsatellite data set) indicated no difference between chosen social fathers and randomly assigned males (observed = 51.16, simulated = 54,33, ns), the observed heterozygosity (Hobs) was significantly higher in social fathers than in randomly assigned males (observed = 0.78, simulated = 0.71, P = 0.001; Table 1, Fig. 1c).
Female choice for genetic fathers
No differences in the number of shared MHC-alleles with the female (observed = 0.26, simulated = 0.29, ns) and in pairwise amino acid differences (observed = 10.48, simulated = 9.21, ns) between chosen genetic fathers and randomly assigned males were observed (Table 1). However, genetic fathers shared less MHC-supertypes with the female than randomly assigned males (observed = 0.58, simulated = 0.85, P = 0.005; Table 1).
We compared MHC-supertype frequencies of the whole population to the frequency of each supertype in the group of known genetic fathers. No differences were found (χ²-tests, all ns, data not shown).
No differences in the relatedness between females and chosen genetic fathers versus randomly assigned males could be detected (observed = −0.03, simulated = −0.02, ns).
Genetic fathers displayed a significantly higher overall heterozygosity than randomly assigned males (observed = 0.77, simulated = 0.71, P = 0.002; Fig. 3c). The d²-value did not differ between these groups (observed = 50.03, simulated = 53.82, ns).
No differences in the individual numbers of MHC-alleles and MHC-supertypes between cuckolded and not cuckolded social fathers were identified. The same was true for the comparison of cuckolded social fathers and the genetic fathers of EPY (Table 1).
All comparisons revealed no effects of specific supertypes on female cuckoldry behaviour (χ²-tests, all ns, data not shown).
We did not find differences in the relatedness (Queller and Goodnight 1989) to the females between cuckolded and not-cuckolded social fathers. Also cuckolded social fathers and genetic fathers did not differ in this measurement (Table 1).
No differences in terms of overall heterozygosity could be detected between cuckolded social fathers and not-cuckolded social fathers or the extra-pair genetic fathers, respectively (Table 1).
Comparison IPY and EPY
We also tested for differences between IPY and EPY in terms of (1) MHC-heterozygosity (number of MHC-alleles or MHC-supertypes, respectively), (2) individual MHC-diversity (mean distance between all individual alleles), (3) number of MHC-alleles or MHC-supertypes distinct from the mother, (4) differences in the number of MHC-alleles or MHC-supertypes to the mother, (5) microsatellite heterozygosity (Hobs), (6) d²-value and (7) relatedness to the mother. Neither comparison revealed any differences between IPY and EPY (all tests ns, data not shown).
In monogamous species, especially when pairs are staying together for their whole life, mate choice should be crucial for the reproductive success and fitness of the partners. The main objective of this study was to test for female mate choice mechanisms in a pair-living primate with obligate biparental care but a high rate of extra-pair paternity. We investigated possible mate choice mechanisms regarding the choice of the social fathers of the offspring, the choice of genetic fathers and examined possible female strategies explaining cuckoldry. Finally, we tested for differences between IPY and EPY.
Our extended dataset to the one used by Fietz et al. (2000) confirmed the high rate of EPY in C. medius (40% this study; 44% previous study). In the allied rock-wallaby (Petrogale assimils) (Spencer et al. 1998), the alpine marmot (Marmota marmota) (Goossens et al. 1998; Cohas et al. 2006), and the African wild dog (Lycaon pictus) (Girman et al. 1997) extra-pair paternity rates of 10–33% have been observed. In C. medius, we found evidence that females do select both, the social and the genetic fathers of their offspring, respectively, on the basis of genetic quality mainly in terms of heterozygosity. Chosen males (both social and genetic fathers of offspring) had a higher number of MHC-alleles, a higher number of MHC-supertypes and a higher degree of overall heterozygosity (Hobs) measured by microsatellites. Thereby, we did not find evidence for linkage between the neutral diversity and adaptive MHC-variability since microsatellite diversity and the number of MHC-alleles was not correlated. The importance of overall heterozygosity as a strong fitness predictor has been indicated by both theoretical and empirical studies (reviewed by Hansson and Westerberg 2002). We therefore assume that females generally chose heterozygous males in order to achieve enhanced genetic diversity in their offspring, according to both, MHC-diversity and neutral-variability whereas the later marker might be linked to other fitness-relevant loci. Moreover, females engaged in extra-pair mating shared a significant higher number of MHC-supertypes with their chosen social fathers than faithful females. Thereby, the mutual relatedness (Queller and Goodnight 1989) and neutral parameters in general had no influence on females’ selection of extra-pair partners. Thus, we found evidence that in the first place mate choice is performed to maximise genetic variability of offspring as predicted by the ‘good-genes-as-heterozygosity hypothesis’ (Brown 1997). Contrarily to evidence from sticklebacks (Gasterosteus aculeatus) we did not find support fort the ‘allele counting hypothesis’ (Reusch et al. 2001; Wegner et al. 2003) which assumes that female choice aims to achieve an optimal (i.e. intermediate) diversity for the offspring to ensure optimal parasite resistance. To date it is not clear whether this selection pattern is confined to species with a relatively flexible genomic architecture such as teleosts with haplotype variation in their MHC locus duplication numbers, or whether it represents a more general feature that has been overlooked in previous studies (Reusch et al. 2004). In mammals, the appearance of multiple MHC class II DRB loci with variable loci numbers between individuals has been described in rhesus macaques (Macaca mulatta, Doxiadis et al. 2001) and California sea lions (Zalophus californicus, Bowen et al. 2004) but no relationship between the total number of unique DRB genes and the presence of cancer has been identified (Bowen et al. 2005). In hairy-footed gerbils (Gerbillurus paeba) which possess two functional DRB loci a possible relationship between the number of MHC-alleles and parasite load was detected in form of an increased parasite burden in individuals with three alleles compared to individuals with four alleles (Harf and Sommer 2005). This is in accord with the theoretical background which assumes that animals containing more MHC-alleles should be able to recognize and react against a broader array of different parasites (Doherty and Zinkernagel 1975).
The occurrence of extra-pair matings however supports the ‘MHC-disassortative mating hypothesis’ (Zeh and Zeh 1996; Penn and Potts 1999; Tregenza and Wedell 2000; Bernatchez and Landry 2003). Extra-pair partner selection is performed to minimise the overlap with female’s MHC-supertype constitution with the potential benefit of an improved resistance to infectious diseases. A recent study in the same study population revealed a strong adaptive value of the individual MHC-constitution in parasite resistance whereas no association between neutral overall individual genetic diversity and parasite load could be detected (Schwensow et al. 2007). Negative effects of sharing either certain MHC alleles or the whole MHC haplotype between mates have also been indicated in humans by significantly elevated rates of foetal loss (Ober et al. 1998). MHC-associated mating preferences for heterozygotes and associated fitness benefits were found in several species (reviewed by Penn 2002). For example, male MHC-heterozygous rhesus macaques (Macaca mulatta) have an increased reproductive success (Sauermann et al. 2001). In humans, heterozygosity at MHC loci is associated with increased resistance to hepatitis and HIV infections (Thursz et al. 1997; Carrington et al. 1999) and with resistance/susceptibility to parasite infections in the African striped mouse (Rhabdomys pumilio, Froeschke and Sommer 2005).
In humans, a strong positive relationship between MHC-heterozygosity and the level of allele sharing was found (Roberts et al. 2005a). The authors state that if male heterozygosity is valued by females this results in the reduction of the pool of dissimilar males. Consequently, females would need to become choosy for genetically dissimilar males (Roberts et al. 2005a). This means, that if a female prefers heterozygotes, and for some reasons only heterozygotes who share alleles with her are available, she would somehow be in a conflict which might be suitable to explain extra-pair offspring. So far, evidence for MHC-associated extra-pair mate choice has derived from different bird species. Freeman-Gallant et al. (2003) found in Savannah sparrows (Passerculus sandwichensis) that young (but not older) females avoided similar males and that MHC similarity between young mates predicted EPY in first broods. In Seychelles warblers (Acrocephalus sechellensis), females were more likely to obtain extra-pair paternity when their social mate had a low MHC-diversity and the diversity of the extra-pair male was higher than that of the cuckolded male (Richardson et al. 2005). In contrast to the study in humans (Roberts et al. 2005a), we observed in the primate C. medius no correlation between the level of allele sharing and male heterozygosity. Females preferred males as social fathers for their offspring and as well as genetic fathers with a higher degree of heterozygosity (number of MHC alleles, number of MHC supertypes and overall heterozygosity) than the average of all other available males. Social fathers sharing a higher number of MHC-supertypes with their females were more often cuckolded which also provides evidence for disassortative mate choice. This suggests that females may improve heterozygosity for their offspring by extra-pair mating and thus ‘correct’ if heterozygous social partners are genetically to similar to themselves. Currently, we do not know if females actively search for males with certain MHC-constitutions or if they generally search for extra-pair copulations and the observed pattern is a result of cryptic female choice (reviewed in Dorak et al. 2002; Wedekind et al. 2004) or sperm competition (reviewed in Birkhead and Møller 1998; Wedekind et al. 2004). There is evidence that both post-copulatory genetic compatibility selection and directional sexual selection for ‘good genes’ can operate within the same species (Evans and Magurran 2000; Tregenza and Wedell 2000; Colegrave et al. 2002). Nevertheless, our data suggest that in C. medius females might ‘correct’ for genetic incompatibility by extra-pair mating since no differences between EPY and IPY were observed. In addition to indirect benefits based on genetic attributes mate choice might also be driven by direct benefits provided by paternal care which might play an important role in this species.
In C. medius, females engaged in extra-pair copulations do not seem to risk a loss of paternal care of their social partner. A theoretical framework for the evolution of male parental care and female multiple mating has been summarised by Ihara (2002). One possible explanation could be that cuckolded males may simply not be able to detect the relatedness of the young. Despite that mice and rat are able to distinguish MHC constitutions (Yamazaki et al. 1979; Brown et al. 1987) and imprinting evidently is the mechanism to learn the own smell (Eklund 1997; Penn and Potts 1998), this might, for some reasons, not be true for this natural population of C. medius.
Different cues might be used by females to assess the genetic quality of males. It could be possible that heterozygosity at key or at many loci is indicated by certain traits which, in turn, are favoured in mate choice. In humans, females have been found to prefer the smell of MHC-heterozygous males (Thornhill et al. 2003) and faces of MHC-heterozygous males has been judged more attractive by woman than faces of man that are homozygous at the investigated loci (Roberts et al. 2005b). Recent studies have revealed such mate choice-relevant traits such as antler size in white-tailed deer (Odocoileus virginianus) (Ditchkoff et al. 2001), crown colour in blue tits (Paraus caerulus) (Foerster et al. 2003) and song repertoire in sedge warblers (Acrocephalus schoenobaenus) (Marshall et al. 2003). So far, such external traits have not been identified in C. medius. In addition to external fitness-related traits the cue used in MHC-based mate choice might also be based on odour which allows to distinguish MHC-identities (reviewed in Penn and Potts 1998, 1999; Yamazaki et al. 1998; Eggert et al. 1999). Peptides/MHC complexes that are not retained at the cell surface but instead are released into the extracellular space might appear in the urine and other body secretions and be used for interindividual communication (Singh et al. 1987; Singh 1998). In mammals, the vomeronasal organ is essential in odour-based social recognition by detecting pheromones and other chemosignals that carry information about gender, sexual and social status, dominance hierarchies, and individualities, but it has been difficult to define the molecular nature of these chemosignals. Recent studies provided evidence that MHC class I peptides serve as chemosensory signals in the vomeronasal organ by which individual MHC genotype diversity can be used as a relatedness marker and may influence social behaviour (Leinders-Zufall et al. 2004). Such olfactory cues might be especially important in nocturnal species such as C. medius.
Choice for heterozygous males is not necessarily mutual exclusive to a choice of males displaying certain, advantageous alleles. Host–parasite coevolutionary mechanisms might simultaneously select for mate choice for frequency-dependent advantageous ‘good genes’. Ekblom et al. (2004) found specific MHC-lineages (but not MHC-alleles itself) to be advantageous with respect to mating success. However, these MHC-lineages are based on phylogenetic similarities of MHC-alleles which do not necessarily reflect functional similarities like MHC-supertypes. In C. medius, we observed an increased frequency of MHC-supertype 1 alleles in male social fathers. A previous study in the same study population of C. medius revealed that certain supertypes are correlated with high or low intestinal nematode burden (Schwensow et al. 2007). MHC-supertype 1 is one of the most frequent supertypes in the population and it is correlated with high parasite burden whereas the rare MHC-supertype 7 was identified to be correlated with low parasite burden. Thus, female choice for supertype 1 seems puzzling on the first glance, but can be explained in context with current theory and assumptions on ‘frequency-dependent-selection’ (Clarke and Kirby 1966; Doherty and Zinkernagel 1975). MHC-alleles that are more resistant to parasites cause an advantage to the host and spread out through the population. This increases selection on parasites to evade recognition by these common alleles. As the parasite antigenicity changes, the relative fitness of the common host genotypes decreases and provides a selective advantage to new, rare MHC alleles to which the parasites are not yet adapted to. In our study one explanation would be to assume that MHC-supertype 1 indeed has been advantageous until a very recent antigenicity change occurred in the much faster evolving parasites. As C. medius bond for life, the resulting time lag in the mate choice response would explain the prevalence of MHC-supertype 1 in male social fathers although the selective advantage of MHC-supertype 1 has recently been lost. If this was true we would expect that cuckoldry is MHC-supertype-dependent. Another possibility is, that supertype 1-alleles might be protective against pathogens others than the investigated gastrointestinal parasites. In this case, supertype 1 could have been selectively favoured although it is disadvantageous in terms of protection against the investigated gastrointestinal parasites. However, we did not observe a higher frequency of MHC-supertype 1 in males per se, but only in social fathers. Moreover, intestinal nematode burden did not differ between females and males though there is evidence in a number of species that androgenic hormones (e.g. testosterone) have immuno-suppressive effects which might explain why males suffer from higher parasite burden (Grossman 1984; Alexander and Stimson 1988; Travi et al. 2002). Ongoing studies will focus on the functional importance of MHC-supertypes in mate choice decisions.
To conclude, in C. medius we found evidence that mate choice is predicted in the first place by the ‘good-genes-as-heterozygosity hypothesis’ whereas the occurrence of extra-pair matings supports the ‘dissassortative mating hypothesis’. To the best of our knowledge this study represents the first investigation of the potential roles of MHC-genes and overall genetic diversity in mate choice as well as extra-pair partner selection in a natural population of a pair-living non-human primate.
We are grateful to the “Commission Tripartite” of the Malagasy Government, the “Laboratoire de Primatologie et des Vertébrés de l´Université d´Antananarivo”, the “Parc Botanique et Zoologique de Tsimbazaza”, the “Ministère pour la Production Animale” and the “Département des Eaux et Forêts” for their collaboration and permission to work in Madagascar. Many thanks to the ‘Centre de Formation Professionnelle Forestière de Morondava’, B. Rakotosamimanana, R. Rasoloarison, and L. Razafimanantsoa for logistical support, and to the German Primate Centre (DPZ) for the opportunity to work at the field station. We thank I. Tomaschweski for technical assistance in the lab, A. Hapke and H. Zischler for introducing microsatellite analyses and J. Ganzhorn for unflagging support. Two anonymous reviewers provided helpful comments on a former version of this manuscript. This study was made possible by the German Science Foundation (So 428/4-1, So 428/4-2).
- Andersson M (1994) Sexual selection. Princeton, New JerseyGoogle Scholar
- Birkhead TR, Møller AP (1998) Sperm competition and sexual selection. Academic Press, LondonGoogle Scholar
- Bowen L, Aldridge BM, DeLong R, Melin S, Buckles EL, Gulland F, Lowenstine LJ, Stott JL, Johnson ML (2005) An immunogenetic basis for the high prevalence of urogenital cancer in a free-ranging population of California sea lions (Zalophus californianus). Immunogenetics 56:846–848PubMedCrossRefGoogle Scholar
- Brown JL (1997) A theory of mate choice based on heterozygosity. Behav Biol 8:60–65Google Scholar
- Colegrave N, Kotiaho JS, Tomkins J (2002) Mate choice or polyandry: reconciling genetic compatibility and good genes sexual selection. Evol Ecol Res 4:911–917Google Scholar
- Ganzhorn JU, Sorg JP (1996) Ecology and economy of a tropical dry forest in Madagascar. In: Primate Report 46, GöttingenGoogle Scholar
- Janeway CA, Travers P (2002) Immunology. Spektrum Akademischer Verlag GmbH, OxfordGoogle Scholar
- Klein J (1986) Natural history of the major histocompatibility complex. Wiley & Sons, New YorkGoogle Scholar
- Ryan MJ (1997) Sexual selection and mate choice. In: Krebs JR, Davies NB (eds) Sexual selection and mate choice. Blackwell, Oxford, pp 179–202Google Scholar
- Schwensow N, Fietz J, Dausmann K, Sommer S (2007) Neutral versus adaptive variation in parasite resistance: importance of MHC-supertypes in a free-ranging primate. Heredity. doi: 10.1038/sj.hdy.6800993