Odour cues and sexual selection in the three-spined stickleback (Gasterosteus aculeatus)

A large part of the studies carried out on odour-related mate choice in respect to MHC have concentrated on functional aspects of the phenomenon. The ones that show sexual selection for heterozygosity at the MHC concentrate on systems with few strains of MHC I and II congenic laboratory populations of mice and rats. Thus inbreeding avoidance was thought to be the main selective force creating MHC diversity through mate choice. Only a few studies have tried to deal with the full natural diversity of MHC genes. Surveys in natural populations have revealed striking diversity in MHC alleles in many species, with more than 30 alleles per population per locus. In particular when MHC loci are duplicated, which is the case for almost all mammals, the combination of distinct haplotypes becomes very large, so that the chances of choosing a partner with identical MHC haplotypes become very unlikely. This makes inbreeding avoidance as the origin of observed MHC diversity in natural populations questionable. We first investigated the mate choice behaviour of the three-spined stickleback G.aculeatus. A striking variation of MHC class II molecules in the populations studied was found, both in loci and alleles per locus. In a second step we tested whether the variation in number of loci was of any functional significance in mate choice. We found that gravid three-spined sticklebacks prefer males with more MHC class II alleles over individuals that had only few alleles. This was a very interesting finding, since to our knowledge it was the first evidence for such a mate choice strategy where females were counting MHC alleles in their potential partners. We then tried to get an insight into the molecular nature of the substance that is complex enough to code for MHC diversity on the level of an individual and small enough to still be smelled. Thomas Boehm provided us with what he thought was a very plausible candidate to fulfil the two criteria. We tested whether gravid stickleback females would alter their mate preference when either exposed to water stemming from a male, or the same male water plus either one or five peptides that can bind to MHC molecules. The chemical diversity of bound peptide populations is directly determined by the molecular structure of ligand binding grooves of MHC molecules. Therefore, information for individual recognition may be conferred either by the polymorphic MHC molecules themselves or by the chemical properties of their ligands. While addition of peptides to plain water was not causing gravid females tochoose, male water with peptides was always preferred over simple male water. Nongravid females did not make the distinction, giving further evidence for mate choice related behaviour. The peptides we used were not innate to the fish in the experiment and it is very difficult to really test mate choice decisions. Nevertheless, we believe our results show that for fish these rather large entities could be the carrier substance for MHC related mate choice. In a further study we showed that sticklebacks with a moderately high number of about six MHC class II alleles suffered least from tapeworm infection. They had the lowest relative parasite weight, and showed the lowest expression of an innate immune trait, the respiratory burst reaction (RB). This could be adaptive, since immune responses are energetically demanding (Sheldon and Verhulst 1996), and a reduction of RB could save energy which can be directed to other essential functions. During RB reactive oxygen and nitrogen intermediates are produced by the host, to fight the intruding parasite. But oxygen radicals can potentially also harm the host, and not only the parasite. Thus regulation of RB reaction during infection might be vital to the host and only individuals that are able to rely on an efficient adaptive immune system with an optimal number of MHC molecules might be able to do so. We argue here that individuals with an allele diversity close to the optimum may rely on their adaptive immune system to very precisely fight parasitic invasion, and reduce RB which is an effective but dangerous tool to use, as too high levels of RB can harm host tissues (oxidative stress). Reducing RB also means that fewer oxygen scavengers, like carotenoids, are needed to reduce oxidative stress. Since carotenoids are used for breeding ornamentation, males with an optimal number of MHC alleles, which are more resistant to S. solidus infection, have a lower RB and could thus invest more in their sexual ornamentation. Barber et al. (2001) possibly showed such a link in their study where females of the three-spined stickleback chose brightly coloured males, and offspring produced later was more resistant to infection with S. solidus. We then investigated whether other odour cues could be identified in a mate choice situation with our experimental set-up. We therefore let gravid females choose between combinations of odours stemming from males with nest, males without a nest and females. We found that males were highly preferred. This was shown by the fact that they were preferred even when having no nest and only few MHC class II alleles. A fact which we have shown before, that should render them less attractivefor gravid females, assuming that female three-spined sticklebacks produce MHC related odour, too. In a following study we looked at the growth curve patterns of S. solidus in its second intermediate host, the three-spined stickleback. We followed parasite growth within the stickleback host for 17 weeks post infection and looked at the parasite growth of the different infection levels. We were blind concerning the sex of the fish since secondary sexual traits were not yet expressed. A recent study (Reimchen and Nosil 2001) showed a female bias in prevalence and intensity of S. solidus in sticklebacks. It also found a diet divergence which would lead to a differential exposure to the parasite. The first intermediate hosts of S. solidus are mainly pelagic copepods. They conclude that ecological differences may be important components of biased parasitism in natural populations. Actually, their data show that females are probably more exposed to parasites than males, but it does not mean that infection really is sex-biased. The only way to know whether there exists a sensitivity bias is to perform an experimental test. The present study provides useful information on this question. In our trials with experimental infections, we found an overall sex bias towards more heavily infected males. Taking these results into account, males might constrain themselves to the benthic environment as a consequence of a high risk of infection. This would mean that the different feeding habits of the two sexes is a consequence, and not a cause, of parasitic infection. In a field study Heins, Baker and Martin (2002) looked at S.solidus infections in the three-spined stickleback of two lakes in Canada and did not find conclusive evidence for neither a RC nor LHS strategy. In our experimental study, we found trends going in the direction of adjustment of growth to other conspecifics present by the parasite (LHS hypothesis). Parasite indexes between level 1 and level 3 were almost significantly different. Since the mortality of infected copepods, the first intermediate host of S.solidus, was very high, our results have to be reinforced with further trials to be conclusive. The information which is gathered by inspecting animals during their inspection visits to a potential predator is still unknown. We believe we deliver the first evidence that sticklebacks adjust their behaviour to the hunger status of a pike after inspection of the latter. Test fish that had been assigned to a close inspection distance (close approach to the predator) started foraging earlier after they had inspected a satiated pike compared to those that inspected a hungry pike. Fish assigned to far inspection distances could not assess the pike`s hunger state as precisely. This was shown by asignificant interaction between hunger state of the pike and inspection distance on the time to start foraging. Conclusions drawn from three pikes as predator stimulus should be interpreted with caution. However, our sophisticated experimental design allowed for a rigorous testing of the hypothesis.Table of contents 1. General Introduction: Sexual selection and mate choice 2. Outline of the thesis 3. Chapters Chapter 1 Female sticklebacks count alleles in a strategy of sexual selection explaining MHC polymorphism. Reusch, T. B. H., M. A. Häberli, P. B. Aeschlimann and M. Milinski. 2001. Nature 414:300-302. Chapter 2 Female sticklebacks recognize peptides as odour cues during MHC related sexual selection. Häberli M. A., P. B. Aeschlimann, M. Milinski and T. Boehm. Submittable manuscript Chapter 3 MHC diversity, mate choice and the optimal immune response. Joachim Kurtz, Martin Kalbe, Peter B. Aeschlimann, Michael A. Häberli, K. Mathias Wegner, Thorsten B. H. Reusch and Manfred Milinski. Nature Genetics in review Chapter 4 Male traits influence odour based mate choice in the three-spined stickleback (Gasterosteus aculeatus). Häberli M. A., P. B. Aeschlimann. Manuscript submitted for publication. 4. Summary and Discussion 5. Acknowledgements 6. Appendix: Chapter 1: Threat sensitive feeding strategy of immature sticklebacks (Gasterosteus aculeatus) in response to recent experimental infection with the cestode Schistocephalus solidus. P. Aeschlimann, M. Häberli and M. Milinski. 2000. Behavioral Ecology and Sociobiology 49:1-7. Chapter 2: Female sticklebacks Gasterosteus aculeatus use self-reference to optimise MHC allele number during mate selection. Peter B. Aeschlimann, Michael A. Häberli, Thorsten B. H. Reusch, Thomas Boehm and Manfred Milinski. Manuscript submitted for publication. Chapter 3: Consequences of MHC based female choice for parasite resistance in offspring. Peter B. Aeschlimann, Michael A. Häberli, Thorsten B. H. Reusch, K. Mathias Wegner and Manfred Milinski. Manuscript 7. Curriculum Vita