DISSERTATION DER FAKULTÄT FÜR BIOLOGIE DER LUDWIG-MAXIMILIANS-UNIVERSITÄT MÜNCHEN COEVOLUTION IN THE SLAVEMAKING ANT PROTOMOGNATHUS AMERICANUS AND ITS TEMNOTHORAX HOST SPECIES: INFLUENCE OF PARASITE PRESSURE, BEHAVIORAL ADAPTATIONS AND PATTERNS OF GENE FLOW eingereicht von Alexandra Achenbach München 29 / 01 / 2009 Promotionsgesuch eingereicht am 29.01.2009 Die Arbeit wurde angeleitet und betreut von Prof. Dr. S. Foitzik Prüfungsausschuss: Vorsitz: Prof. Dr. S. Foitzik 1. Gutachter: Prof. Dr. S. Renner 2. Gutachter: Prof. Dr. G. Grupe 3. Protokoll: Prof. Dr. D. Metzler Tag der mündlichen Prüfung: 24.06.2009 ERKLÄRUNG Hiermit versichere ich, Alexandra Achenbach, ehrenwörtlich, dass meine Dissertation selbstständig und ohne unerlaubte Hilfsmittel angefertigt worden ist. Die vorliegende Dissertation wurde weder ganz, noch teilweise bei einer anderen Prüfungskommission eingereicht. Ich habe noch zu keinem früheren Zeitpunkt versucht eine Dissertation einzureichen oder an einer Doktorprüfung teilzunehmen. Unterhaching, den 29.01.2009 Alexandra Achenbach TABLE OF CONTENTS GENERAL INTRODUCTION……………………………………………………………………………….…. 1 SUMMARY: AIMS OF THIS THESIS……………………………………………………………………….….. 8 PUBLICATION 1: Locally-adapted social parasite affects density, social structure and life history of its ant hosts…………………………………………….. 9 Introduction………………………………………………………………………………. 11 Material and Methods………………………………………………………………….. 14 Results…………………………………………………………………….......................... 19 Discussion………………………………………………………………………………….. 29 PUBLICATION 2: First evidence for slave rebellion: Enslaved ant workers systematically kill the brood of their social parasite Protomognathus americanus……………………………………..…….... 34 Introduction……………………………………………………………………………….. 36 Material and Methods………………………………………………………………….. 39 Results…………………………………………………………………….......................... 41 Discussion………………………………………………………………………………….. 44 PUBLICATION 3: Evolutionary arms races within social parasite colonies: Behavioral and chemical strategies of slaves and slavemakers….…………………….………………………………………… 48 Introduction……………………………………………………………………………….. 50 Material and Methods………………………………………………………………….. 53 Results…………………………………………………………………………………….… 56 Discussion………………………………………………………………………………….. 65 PUBLICATION 4: Comparative population structure, gene flow and post-glacial colonization of the social parasite Protomognathus americanus and its host species Temnothorax longispinosus and Temnothorax curvispinosus…………………………………………………………………. 69 Introduction………………………………………………………………………………. 71 Material and Methods………………………………………………………………….. 73 Results……………………………………………………………………………………… 76 Discussion…………………………………………………………………………………. 90 SYNOPSIS…………………………………………………………………………………………………. 95 CONCLUSION…………………………………………………………………………………………….. 98 ZUSAMMENFASSUNG……………………………………………………………………………………… 99 ACKNOWLEDGMENTS……………………………………………………………………………………… 101 CURRICULUM VITAE……………………………………………………………………………………….. 103 REFERENCES……………………………………………………………………………………………..... 106 GENERAL INTRODUCTION 1 GENERAL INTRODUCTION Adaptation to abiotic conditions and biotic factors is certainly one of the main driving forces for the evolution of the enormous number of morphologies, behaviors and life history strategies in the animal kingdom (Futuyma 1986). This is especially true for the adaptation to and the interactions with other species. According to the geographic mosaic of coevolution theory (Thompson 1999b, 2005) the interactions between two species can be influenced by several parameters: The variation of ecological parameters, the composition and history of a local community and the evolutionary potential of the two opponents. In contrast to ecological parameters, species are capable and forced to change rapidly and frequently. Interacting species like competitors, predator and prey or parasite and host thus exert highly variable selection pressures (Thompson 1994, 1999a), which can potentially result in a continuous process of reciprocal adaptations: the so-called “coevolutionary arms race” (Dawkins & Krebs 1979). In the extreme case, this race can even end in the Darwinian extinction of one of its participants due to a time lack in the development of effective counter-adaptations (Darwin 1859). The highly specialized relations of parasites and their hosts represent ideal model systems for the study of coevolution. Parasites were shown to influence their host species in various aspects (e.g. Anderson & May 1979, Freeland 1979, 1983; Price et al. 1986) and to dramatically reduce the fitness of infected host individuals up to complete sterilization. In the case of microparasites, this severe impact is due to the asymmetric evolutionary potentials of the opponents caused by differences in population sizes, mutations rates and generation times (e.g. viruses, bacteria, fungi). In contrast, macroparasites are much more similar to their host species and often show comparable evolutionary potentials (e.g. arthropods, helminthes). In the case of brood parasites the similarity between parasite and host is even greater since only closely related host species potentially fulfill both behavioral and nutritional needs of the parasitic young. Most energy acquired during an animal’s adult life is channeled into reproduction and it is thus not surprising that strategies evolved to lower the costs of brood care. This is especially true for species with extensive parental care, such as mammals, birds and social insects. One way to reduce these costs is brood parasitism, where the parasite exploits the brood care behavior of its host species. Brood parasitism has been found in a variety of different taxa (e.g. Boulton & Polis GENERAL INTRODUCTION 2 2002, Sato 1986, Brooke & Davies 1988, Hölldobler & Wilson 1990). During the last decades, predominantly avian brood parasites, such as cuckoos and cowbirds, have received great attention and the potential occurrence of a coevolutionary arms race has been studied intensively in several species (Davies & Brooke 1989, Lotem & Rothstein 1995, Rothstein 1990, Soler & Soler 2000, Soler & Møller 1990). These avian brood parasites have been shown to lay their eggs in nests of other bird species and to thereby avoid the cost of parental care. Their avian host species thus suffer both the cost of brood care for the unrelated offspring and the loss of their own eggs due to the destructive behavior of the parasite young. Social parasites represent a highly analogous system and reduce the cost of parental care by exploiting the brood care behavior and the entire social system of members of their own or another socially living species (Davies et al. 1989). Generally this form of brood parasitism can be found among social insects and, within the Hymenopterans, it is especially common in ants. Remarkably, in ants, more than 250 of the over 12,000 known ant species live a social parasitic life style (Buschinger 1986, Hölldobler & Wilson 1990) and these ants exploit their host species either temporarily or permanently. In ants, there are mainly two different forms of social parasites: inquilines, which have lost the worker caste and produce only sexuals to be raised by the hosts, and slavemakers, in which the queen produces her own workers that conduct slave raids on neighbouring host nests. Since its first detailed description in the year 1810 (Huber), the latter behavior in ants has fascinated both scientists and the general public. Slavemaking ants are social parasites with an extremely interesting life style (Buschinger 1986, Hölldobler & Wilson 1990, Brandt et al. 2005a). Usually, these parasites obligatorily depend on related ant species for brood care, foraging, and nest defense and their specific morphology and behavior allows them to find and subdue colonies of their host species for their purposes. Slavemaking ants are characterized by a large body size, strong mandibles, thick cuticle, a broad postpetiole and antennal scrobes, which are used to protect the antennal scapes during fights. The life cycle of a slavemaking ant begins, when a mated parasite queen conquers a host nest, expels or kills the host queen(s) and adult workers and takes over the remaining worker brood (Fig. 1) (Alloway 1979, Alloway 1980, Wesson 1939, Wilson 1971). These first slave workers, which soon eclose from the usurped brood, learn to accept the slavemaker queen during the first days (Goodloe & Topoff 1987, Jaisson 1975, LeMoli & Mori 1982, LeMoli & Passetti 1977) and GENERAL INTRODUCTION 3 subsequently fulfill all tasks of colony maintenance and brood care. Since enslaved host workers are unable to reproduce new slave workers, the steady supply of slaves has to be ensured in a different way. For this purpose, slavemaker workers raised during the following years, conduct recurrent slave raids on neighboring host colonies and steal new host worker brood (e.g. Alloway 1979). We consequently find a seemingly peaceful coexistence of species within a slavemaker colony but the colony survival often obligatorily depends on the collaboration of the enslaved host workers. These host workers rear both parasite and raided host brood and thus control the productivity of the social parasite nest. Young mated slavemaker Slavemaker colony with social parasite brood and raided host Colony take-over by slavemaker queen and nest Host colony Slavemaker Slave raids on neighbouring Coexistence of slavemakers host colonies by slavemaker and enslaved host workers Host Figure 1: Life cycle of a slavemaking ant. Graphic by Susanne Foitzik Only 50 ant species of the 200 ant social parasites are active slavemakers (D’Ettorre & Heinze 2001, Hölldobler & Wilson 1990) and this exceptional life style evolved several times independently among the subfamilies Myrmicinae and Formicinae (Buschinger 1990, Buschinger et al. 1980, Hölldobler & Wilson 1990, Stuart & Alloway 1983). Particularly the Myrmicinae tribe of the formicoxenine ants seems to be a hot spot in the evolution of slavemaking ants with at least six independent origins. Among this tribe, slavemakers were shown to originate from non-parasitic Formicoxenini at different points of times (Beibl et al. 2005). The strength of reciprocal adaptations varies with the specificity of its participants and inevitably peaks in the relationship between obligate parasites and GENERAL INTRODUCTION 4 their hosts (Futuyma & Slatkin 1983, Thompson 1994, Thompson 1999a). Despite of this and despite of the strong analogy to avian brood parasites, the fascinating relationship of insect social parasites and their hosts was not investigated under a coevolutionary perspective for a long time as it was assumed that social parasites have won the arms race against their hosts (Gladstone 1981, Grasso et al. 1992). Yet, recent studies on the association between slavemaker presence and host demographic and genetic structure have revealed strong evidence for an ongoing coevolutionary arms race (Foitzik & Herbers 2001a, Herbers & Foitzik 2002, Foitzik et al. 2003, Hare & Alloway 2001) and thus fiercely questioned this view. In particular, empirical studies on the slavemaking ant Protomognathus americanus and its host species have demonstrated the strength of selection pressure, this social parasite exerts on its hosts. P. americanus is a tiny myrmicine ant, which obligatorily exploits three related species of the genus Temnothorax. These ants nest in hollow acorns, hickory nuts and small twigs on the ground of mixed deciduous forests throughout the north-eastern part of the American continent (Fig. 2). Figure 2: P. americanus worker (left), T. longispinosus worker (left middle), T.curvispinosus worker (right middle), T. ambiguus worker (right). Pictures by Miriam Brandt The strong impact of this social parasite on its host species largely stems from the frequent and destructive slave raids, with host colonies suffering from 2-10 successful raiding attacks per year and social parasite colony, and a low post- raid survival (Foitzik et al. 2001, Foitzik & Herbers 2001a, Blatrix & Herbers 2003). In addition, host colonies in the vicinity of the social parasite were shown to exhibit a changed demography and investment strategies. These host colonies were smaller, more frequently monogynous and their production rather focussed on dispersing sexuals than on colony maintenance (Foitzik & Herbers 2001b, Herbers & Foitzik 2002). Yet, a GENERAL INTRODUCTION 5 direct and causal relationship between P. americanus presence and host nest density, demography and investment strategies was as yet lacking. Publication 1 is trying to fill this gap with empirical data on a large-scale and long-term field experiment, manipulating parasite density in field. According to the geographic mosaic theory (Thompson 1994, 1999b, 2005), the nature and outcome of species interactions can vary greatly across and between regions. Earlier studies have found evidence for local adaptation of the cuticular hydrocarbon profile of P. americanus (Foitzik et al. 2001, Brandt & Foitzik 2004, Brandt et al. 2005a), which matched the chemical odor of sympatric hosts. Therefore it was especially interesting, if the social parasite is also locally adapted on an ecological scale. Publication 1 is addressing this question by including a special cross-fostering design with two geographically separated ant communities into the set-up (e.g. Foitzik et al. 2001, Herbers & Foitzik 2002, Brandt & Foitzik 2004, Foitzik et al. 2004). Within the parasite-host system of P. americanus, preliminary studies and Publication 1 have impressively demonstrated the strong selection pressure exerted by this social parasite. This strong parasite pressure has inevitably led to the evolution of host defenses, especially in the context of slave raids and thus proven an ongoing coevolutionary arms race between P. americanus and its host species (Alloway 1990, Mori et al. 1991, Foitzik & Herbers 2001a, Herbers & Foitzik 2002). Apart from the already mentioned changes in demography and investment strategy, host defenses include enemy recognition and better fighting abilities as well as the fast evacuation and escape of the attacked host colonies (e.g. Alloway 1990, Foitzik et al. 2001, Brandt et al. 2005a). All these adaptations are defense strategies that help to avoid falling victim to slave raids or the colony take-over by a mated social parasite queen. Yet, counter-parasite strategies after successful establishment of parasitism have been argued to be impossible to be selected for (Gladstone 1981), because enslaved host workers seemed to gain no fitness benefit by rebelling against their oppressors. Publication 2 intensely questions this view and documents the killing of social parasite brood by enslaved host workers. The survival of a social parasite colony entirely depends on the collaboration of enslaved host workers. Apart from rearing social parasite brood, enslaved host workers are confronted with a steady supply of raided host brood representing the future slave workforce. Hence, the acceptance and brood care for alien raided brood is a critical point in the parasitic life style of slavemaking ants. Enslaved host workers could also develop defence strategies in this context. Publication 3 thus aims GENERAL INTRODUCTION 6 to enlighten the potential evolution of a rejection behavior in the host and empirically investigates the acceptance of alien host and parasite pupae of different castes by Temnothorax workers in parasite colonies and in non-parasitized host colonies (Fig. 3). The rejection of raided alien host brood was so far excluded, since the brood acceptance of segregated host workers was shown to be either enhanced by the presence of the social parasite (Alloway 1982), or enforced by the exposure of enslaved host workers to allospecific odors during a critical learning period after their eclosion (Jaisson 1975, LeMoli & Passetti 1977, LeMoli & Mori 1982, Goodloe & Topoff 1987). Yet, behavioral studies on the brood acceptance in naturally composed field colonies, as in Publication 3, have as far been lacking. M W Q M W Q M W Q Figure 3: Male (M), worker (W) and queen (Q) pupae of the host species T. longispinosus, T. curvispinosus and the social parasite P. americanus. Furthermore, Publication 3 investigates a potential chemical recognition mechanism, which Temnothorax workers might use to discriminate between pupae of different castes and species. In social insects, chemical signals e.g. cuticular hydrocarbons and gland secretions are known to be widely used in the recognition and discrimination of nestmates, sex and species (Howard 1993). As demonstrated, P. americanus and its Temnothorax host species are caught in an ongoing coevolutionary arms race of reciprocal counter adaptations which were at least in part already shown to significantly vary across and between geographical regions (Foitzik et al. 2001, Brandt & Foitzik 2004, Brandt et al. 2005a, Foitzik et al. in press). Both the coevolutionary arms race and the local adaptation of P. americanus on a chemical and an ecological scale are clearly based on the evolutionary potentials of the interacting species. Unlike asymmetric host-parasite