Relationships between hormones, physiological performance and immunocompetence in a color-polymorphic lizard species, Podarcis melisellensis

Abstract

Species with alternative phenotypes offer unique opportunities to investigate hormone–behavior relationships. We investigated the relationships between testosterone, corticosterone, morphology, performance, and immunity in a population of lizards (Podarcis melisellensis) which exhibits a color polymorphism. Males occur in three different color morphs (white, yellow, orange), providing an opportunity to test the idea of morphs being alternative solutions to the evolutionary challenges posed on the link between hormones, morphology, performance, and immunity. Morphs differed in bite force capacity, with orange males biting harder, and in corticosterone levels, with yellow males having lower levels than orange. However, morphs did not differ in testosterone levels or in the immunological parameters tested. At the individual level, across morphs, testosterone levels predicted size-corrected bite force capacity, but no relation was found between hormone levels and immunity. Our results do not support the testosterone-based polymorphism hypothesis and reject the hypothesis of a trade-off between testosterone and immunity in this species, but provide a mechanistic link between testosterone and a sexually selected performance trait.

Introduction

In the past three decades, it has become increasingly evident that reproductive tactics may vary not only between sexes, but also among individuals of the same sex (Gross, 1996, Brockmann, 2001). Alternative reproductive phenotypes exhibited by males or females of a species can be genetically fixed or plastic, and typically involve a combination of behavioral, morphological, physiological, and life history characteristics. Circulating sex hormone levels can play an important role, either through organizational or activational effects, in triggering the expression of these alternative reproductive morphs (Moore et al., 1998, reviewed in: Oliveira et al., 2008). It has been documented that in male vertebrates, high levels of circulating androgens can correlate with dominant, territorial behavior and enhance the expression of secondary sexual traits that advertise the individual's social ranking (reviews in Oliveira, 2004, Hau, 2007, Moore and Marler, 1987). How circulating levels of testosterone result in dominant behavior is less clear. They may exert their influence primarily by affecting the brain, rendering the animal more aggressive and/or more motivated to take risks. For example, in the house sparrow, testosterone reduces responsiveness to pain stimuli, which may promote the willingness to engage in aggressive encounters (Hau et al., 2004). In all likelihood, these hormones can also affect the morphology and physiology of the individual in such a way that it may be better equipped to deal with the challenges of living a dominant life (Husak et al., 2007). For instance, it can be expected that testosterone will promote the development of morphological or physiological features that aid in fighting rivals, guarding territories, or dealing with an increased predation risk. Accordingly, testosterone is known to affect lower-level morphological and physiological traits that may contribute to performance (e.g. number of muscle units: Tobin and Joubert, 1991; enzymatic activity in muscles and neurons: Luine et al., 1980; contractile properties: Girgenrath and Marsh, 2003; muscle mass: Sidor and Blackburn, 1998; Huyghe et al., unpublished data). However, the evidence for testosterone affecting relevant whole-animal performance is scant and equivocal. In addition to testosterone, a rise in corticosterone levels, commonly occurring in response to stressors in vertebrates, may affect sexually selected traits (Husak and Moore, 2008) and reproductive behavior (Creel, 2001). However, the effects of stress and corticosterone on aggressive and territorial behavior are variable (reviewed in Moore and Jessop, 2003, Goymann and Wingfield, 2004).

Part of the appeal of an androgen-based model of reproductive polymorphism is that the costs and benefits of androgens may help explain why different morphs can co-exist in the first place. Testosterone has been called a ‘double-edged sword’ that stimulates development of characteristics used in a reproductive context, driven by sexual selection, but at the same time may reduce immunocompetence (Folstad and Karter, 1992), and thus survival (natural selection) (Marler and Moore, 1988). Although it is as yet unclear whether testosterone affects immunocompetence directly (Folstad and Karter, 1992) or induces behavioral changes that increase the likelihood of infection (Owen-Ashley et al., 2004, Roberts et al., 2004), parasite loads and testosterone levels can co-vary in natural populations (Klukowski and Nelson, 2001, Saino and Moller, 1995, Mougeot et al., 2005, Perez-Orella and Schulte-Hostedde, 2005). Whereas some studies have revealed a clear positive link between experimentally elevated testosterone levels and parasite infestation (Olsson et al., 2000, Klukowski and Nelson, 2001, Cox and John-Alder, 2007), some show an opposite result where elevated testosterone levels did not result in a higher parasite load (Oppliger et al., 2004). Reproductive morphs could be seen as alternative solutions to an evolutionary trade-off between parasite load susceptibility (survival) and access to mates (reproductive success). However, as far as we know, no previous study has evaluated the prediction that male reproductive morphs should differ in parasite load.

In lizards, as in many vertebrates, testosterone is an important mediator of the expression of secondary sexual traits that differentiate males from females. In addition, testosterone levels affect the degree of trait expression among males. In male Sceloporus lizards seasonal color development is activated by testosterone (Rand, 1992, Cox et al., 2008). Similarly, in Uta stansburiana, males with orange dewlaps are ultra-territorial and have a higher testosterone level than the other two color morphs (Sinervo et al., 2000). On the other hand, organizational hormonal mechanisms in early development differentiate between territorial and non-territorial male morphs in Urosaurus ornatus (Moore et al., 1998). Experimental elevation of testosterone levels has been shown to improve running endurance capacity in lizards (Klukowski et al., 1998, Sinervo et al., 2000), but Husak et al. (2006a) found no relationship between sprint speed and natural variation in testosterone levels in the lizard Crotaphytus collaris. However, a positive relationship between plasma testosterone concentrations and bite performance was found in Anolis carolinensis, but only because of mutual intercorrelation with body size (Husak et al., 2007). In Gallotia galloti, a positive relationship was found between baseline testosterone levels and bite force capacity, but experimental elevation of circulating testosterone levels did not result in enhanced performance (Huyghe et al., unpublished data).

Color-polymorphic species provide an excellent opportunity to study the role of androgens in the mechanisms of origin and maintenance of these polymorphisms, and at the same time to test the immunocompetence handicap hypothesis. The continued existence of the different morphs in a population implies that each reflects an alternative strategy to deal with natural and sexual selection pressures. In this paper, we examine whether levels of circulating testosterone and corticosterone differ among the three male color morphs of the lacertid lizard Podarcis melisellensis. Males of this species can have an orange, yellow, or white ventral color, and there is circumstantial evidence that orange males are behaviorally dominant over the two other morphs (Huyghe et al., 2007). We also evaluate individual relationships between circulating hormone levels, whole-animal performance (bite force) and immunity (parasite load and T cell-response). Since orange males are more aggressive than the other morphs, we predict that orange males will have higher testosterone and higher corticosterone levels, as we hypothesize that this polymorphism is androgen mediated and might be reflected in alternative reproductive tactics. With regard to the immunocompetence handicap hypothesis, we expect that these same orange males will have suppressed immunity resulting in a larger parasite load and smaller PHA response compared to the other morphs.