Proximate and ultimate factors influencing reproductive skew in the ant species Leptothorax acervorum (Hymenoptera: Formicidae) DISSERTATION ZUR ERLANGUNG DES DOKTORGRADES DER NATURWISSENSCHAFTEN (DR. RER. NAT.) DER FAKULTÄT FÜR BIOLOGIE UND VORKLINISCHE MEDIZIN DER UNIVERSITÄT REGENSBURG vorgelegt von Jürgen Trettin aus Jena im Jahr 2015 Das Promotionsgesuch wurde eingereicht am: 31.03.2015 Die Arbeit wurde angeleitet von: Prof. Dr. Jürgen Heinze Unterschrift: i To Ivana & My Parents ii ACKNOWLEGMENTS I would particularly like to thank... Prof. Dr. Jürgen Heinze, ... for inviting me to work in his lab and for proposing me the thesis topic ... for advice and support whenever needed Prof. Dr. Alfred Buschinger, ... for his help in the field, his advices and 'histories' of myrmecology Prof. Dr. Christoph Oberprieler, PD Dr. Christoph Schubart, Dr. Carsten Löser, Dr. Alireza Keikhosravi, Shobit Agrawal and Nicolas Thiercelin, ... for helpful discussions and suggestions on phylogeography. Dr. Bartosz Walter, ... for the practical introduction to 'measuring behavior' Stephanie Leopold, Andi Sailer, Monika Haubner and Thomas Seyferth, ... for 'enduring me patiently' as co-supervisor of their theses (Bachelor or Zulassungsarbeit) Dr. Ivana Miranda da Silva, Andi Trindl, Masaki Suefuji and many students, ... for advice and help during my laboratory work Dr. Simon Tragust, Dr. Abel Bernadou and the members of the 'horse group' ... for helpful discussions and advice on statistical issues Marion Füßl, Monika Haubner, Sabine Hutschenreuther, Peter Koller, Richard Landstorfer, Stephanie Leopold, Doris Rothgänger, Gudrun Schneider, Ivana Miranda da Silva, Thomas Seyferth, Christiane Wanke, many colleagues and students, ... for assistance during the many field trips for ant sampling Shobit Agrawal, Dr. Cátia Bartilotti, Alireza 'the boss' Keikhosravi, Carsten Löser, Gudrun Schneider and Nicolas Thiercelin, ... for all kinds of support, including fruitful discussions and friendship Grete & Peter Trettin, ... for their constant support, patience and love Die Deutsche Forschungsgemeinschaft (He 1623/25) ... for the financial support ... and finally all members over the years from the Heinze lab! iii Declaration of thesis's composition and author's contribution CHAPTER 1 - Queen dominance and worker policing control reproduction in a threatened ant Jürgen Trettin, Monika Haubner, Alfred Buschinger, Jürgen Heinze Published in: BMC Ecology (2011) 11: 21 Author contributions AB and JH devised the study and took part in field work; JT did most of the collection and, together with MH, the genetic and behavioral studies and analyzed the data. JT and JH wrote the paper. All authors read and approved the final manuscript. CHAPTER 2 - Behavioral Plasticity in Ant Queens: Environmental Manipulation Induces Aggression among Normally Peaceful Queens in the Socially Polymorphic Ant Leptothorax acervorum Jürgen Trettin, Thomas Seyferth, Jürgen Heinze Published in: PLoS ONE (2014) 9(4): e95153. doi:10.1371/journal.pone.0095153 Author Contributions Conceived and designed the experiments: JH, JT. Performed the experiments: TS, JT. Analyzed the data: JT, TS. Wrote the paper: JT, JH. iv TABLE OF CONTENTS SUMMARY ........................................................................................................................... 1 GENERAL INTRODUCTION ............................................................................................... 2 CHAPTER 1 .......................................................................................................................... 6 Queen dominance and worker policing control reproduction in a threatened ant CHAPTER 2 ........................................................................................................................ 24 Behavioral Plasticity in Ant Queens: Environmental Manipulation Induces Aggression among Normally Peaceful Queens in the Socially Polymorphic Ant Leptothorax acervorum CHAPTER 3 ........................................................................................................................ 40 Phylogeography of a socially plastic ant in SW-Europe CHAPTER 4 ........................................................................................................................ 75 Genetic or social environmental basis underlying queen behavioral variation – a laboratory transplantation experiment GENERAL DISCUSSION ................................................................................................... 83 REFERENCES ..................................................................................................................... 87 APPENDIX .......................................................................................................................... 97 v SUMMARY In animal societies beside cooperation also conflict among members can arise. Normally, these conflicts emerge among individuals over their share of group's reproductive output, i. e., reproductive skew. In social hymenoptera, additional conflict over resource allocation can arise between members of different castes. Herein this thesis we focus on the first type of conflict, by studying the proximate and ultimate factors of variation in reproductive skew in the ant Leptothorax acervorum (Hymenoptera: Formicidae). Over large parts of its Holarctic range, the species shows a low skew colony structure (i. e., one or more breeding queens per colony) while populations from its range margin usually express a high skew colony phenotype (i. e., only one breeding queen per colony). To clarify whether queens and/or worker are involved in the establishment of high skew we observed queen and worker behav- ior in colonies from central Spain. Among others, we show that aggression among queens leads to the formation of dominance hierarchies, in which only the top-ranking queen becomes fertile (chapter 1). Are queens able to react flexibly to changes in reproductive skew, as assumed by skew theory? To answer this question, we experimentally manipulated socio-environmental factors to investigate whether queen-queen aggression can be evoked in low skew colonies. Our results show that queens are able to react to changes in skew in behavior rather than evolutionary time (chapter 2). In addition, we performed population genetic and phylogeographic analyses to study the ultimate conditions underlying variation in L. acervorum's social organization (chapter 3). Finally, we conducted a reciprocal transplantation experiment to investigate whether variation in skew can be affected by its social environment or, alternatively, has a genetic basis only (chapter 4). 1 GENERAL INTRODUCTION In group living animals not only cooperation but also conflict among group members can emerge. Usually, these conflicts in groups arise over share of reproduction among its members, as genetically non-identical members may have divergent interests about their share of group's reproductive output (e. g. Keller & Reeve 1994, Hager & Jones 2009 and contribu- tions therein, Port & Kappeler 2010). Moreover, in animal societies with unusual genetic family structures (e. g. social Hymenoptera), additional social conflicts over the allocation of resources towards one offspring sex or female larvae's development into sexuals or workers can emerge between individuals of different or same caste (see Heinze 2004, 2010, Ratnieks et al. 2006, Strassmann & Queller 2007 for a detailed discussion). Normally, efficient mechanisms of conflict resolution such as punishment, policing and dominance help to regulate egg laying and development in societies in which all females are morphologically identical and potentially capable of reproducing (Monnin & Ratnieks 2001, Ratnieks et al. 2006). In species with a clear queen-worker diphenism, workers normally refrain from laying eggs in response to the odor of a fertile queen (Bourke 1988, Hammond & Keller 2004), presumably because they otherwise risk to be attacked by their nestmates. Herein, we focus on the first type of social conflict: the partitioning of reproduction among totipotent members of an animal society, i. e. reproductive skew (Vehrencamp 1983, Reeve & Keller 2001, Port & Kappeler 2010). High skew emerges if one or a few individuals dominate reproduction in a group (e.g. in eusocial insects where normally only one female, the queen, dominates reproduction), while low skew exists in groups where reproduction is 2 equally shared among members (e. g. in banded mongooses [Mungos mungo], Sherman et al. 1995, Gilchrist et al. 2004). Reproductive or optimal skew theory provides an explanatory framework for how repro- ductive skew is adjusted and thus reproductive conflicts are resolved among group members (e. g. Vehrencamp 1983, Reeve & Keller 2001, Ratnieks et al. 2006, Nonacs & Hager 2011, Trubenová & Hager 2012). The theory includes several optimal skew models, each one with its own assumptions and predictions that can be distinguished into two main classes of models. Briefly, in 'transactional models' reproductive skew is the outcome of reproductive transactions between dominant and subordinate group members, mediated either by the subordinates' threat to leave the group (concession models) or by the dominants' threat to expel them from the group (restraint models). In contrast, 'compromise models' assume that reproductive skew is the outcome of a competition over reproduction between group members, which intensity is mediated only by the costs it imposes on group productivity (see Johnstone 2000, Nonacs & Hager 2011, Reeve & Shen 2013, Kappeler et al. 2013 for detailed reviews). Key parameters in most skew models that determine the degree of reproductive skew within groups are the genetic relatedness among its members, their respective fighting strength, the impact on group productivity and ecological constraints on breeding indepen- dently (e. g. Keller & Reeve 1994, Johnstone 2000, Nonacs & Hager 2011). However, despite the initial success, advance and proliferation of skew models empirical evaluating and testing of their assumptions and predictions have lagged behind that theoretical development over the last two decades (e. g. Magrath & Heinsohn 2000, Kokko 2003, Nonacs & Hager 2011), even leading to strong resistance against more model variants (Reeve & Shen 2013). For instance, skew models assume that group members adaptively adjust skew to their current situation. Alternatively, group members might be genetically predisposed to either fight or tolerate a rival reproductive, i. e. reproductive skew might be an evolutionary (genet- ically constraint) rather than a behavioral response to environmental change (Kokko 2003). 3 In addition, an adjustment of skew in behavioral time might be limited by further types of mutually non-exclusive constraints, such as developmental, social or ecological constraints (Kappeler et al. 2013). Indeed, whether queens are capable of flexibly reacting to changed environmental conditions (Kokko 2003) has rarely been investigated. Nevertheless, while most studies of reproductive skew in mammalian societies have found support for compromise models (Port & Kappeler 2010, Kappeler et al. 2013), studies of skew in social insect societies usually found support for the transactional types of model (e. g. Reeve & Keller 2001). In particular, the latter models predict that high reproductive skew should be evolutionary stable when the costs of dispersal (and hence becoming reproductive away from the established nest) and relatedness among nestmates are high (e. g. Keller & Reeve 1994). In addition, models predict that higher levels of skew should be associated with higher levels of within-group aggression and vice versa (e. g. Bourke & Heinze 1994, Reeve & Keller 2001). Support for both predictions can be found in Formicoxine ants, in which high skew species are indeed specialists of patchy habitats and queen-queen aggression is involved in the establishment of reproductive hierarchies. In contrast, low skew species living in homogenous habitats such as the extensive boreal forests of Northern hemisphere, don't show any aggressive establishment of hierarchies and finally contribute equally to colony's offspring (Bourke & Heinze 1994, Heinze 2010). For our studies we chose as a model the Formicoxine ant species Leptothorax acervorum (FABRICIUS 1793) for which reproductive skew appears to vary with habitat characteristics. The species is widely distributed over large parts of the northern hemisphere (Francoeur 1983, Seifert 2007, Czechowski 2012) where in extensive coniferous forests its colonies show a low skew social structure (facultative polygyny, Buschinger 1968, Bourke 1991, Heinze et al. 1995a & b). In contrast, the species expresses a high skew colony structure (functional monogyny) in patchy habitats (i. e., on sun exposed slopes in Alaska, in light clearings in 4