Conflict management is crucial for social animals to prevent escalation of fights among group members (Aureli et al. 2002). Conflict escalation within social groups can be avoided through the establishment of dominance hierarchies, the use of ritualized appeasement gestures or displays, or chemical signaling (Preuschoft and van Schaik 2000; van Wilgenburg et al. 2005; Ratnieks et al. 2006; Hick et al. 2014). Encounters between social groups are more difficult to prevent from escalating, but if both sides benefit from it, mediating behavioral mechanisms can evolve (e.g., Neat et al. 1998). For example, the ritualized tournaments of desert honeypot ants allow to avoid costly fights between colonies of similar size (Hölldobler 1976). Social insects often rely on chemical information to decide whether an interaction will remain friendly or escalate into a fight. In particular, ants use sophisticated chemical communication systems to differentiate friends from foes (Van Zweden and D’Ettorre 2010). Still, ant societies can fall victim to social parasites, who exploit their altruistic behaviors (Thomas et al. 2005). Social parasites have to gain entry into their well-protected societies, and those interactions often result in escalated conflicts. Aggressive conflicts between groups of antagonistic species are even more difficult to keep from escalating, but if enemies manage to stay undetected, open fights can be avoided, benefitting the parasite, but not the host.

Many social parasites have evolved refined chemical strategies to avoid host recognition. Chemical insignificance is a rare strategy where the parasite lacks recognition cues on its cuticle (Lenoir et al. 2001). Host ants facing these chemically “invisible” parasites will not recognize them and unintentionally integrate them into their colony (e.g., D’Ettorre and Errard 1998). As hydrocarbons also function as desiccation barriers (Gibbs and Rajpurohit 2010), there are clear constraints to this strategy, which might explain why it is so rare. More commonly used is chemical mimicry, where social parasites actively biosynthesize substances to mimic the hosts’ profile (Dettner and Liepert 1994; Lenoir et al. 1997). A restriction for this strategy is that the parasite has to focus on a single host, whose profile it will imitate. Chemical camouflage is more flexible and more common, because social parasites using this strategy acquire the host’s odor through rubbing, allogrooming, or trophallaxis or from nest material (Dettner and Liepert 1994; Lenoir et al. 2001; D'Ettorre et al. 2002; Tsuneoka and Akino 2012). The problem here is the first host contact, when the parasites had as yet no time to acquire host chemicals. Chemical manipulation of host behavior is a fourth and commonly used strategy by which parasites use offensive chemicals to disrupt the host defense systems. While entering host colonies, social parasites use substances released from the Dufour’s, the Pygidial, or the Poison gland, which serve as manipulative alarm signals or chemical weapons (Lenoir et al. 2001) and as appeasement substances to circumvent aggressive escalations, such as attacks directed toward the social parasite (Mori et al. 2000; Brandt et al. 2006)

Slavemaking ants are social parasites that have to successfully invade host colonies recurrently during their life cycle, in contrast to many other social parasites, such as the workerless inquilines, that do so only once in their life. Most slavemakers are obligate parasites that are contingent on their slave workforce, which oversees colony maintenance including caring for the parasites’ offspring. During destructive slave raids, slavemaking ants invade free-living host colonies to steal brood (Foitzik and Herbers 2001). These raids often escalate into violent fights, as hosts fiercely defend their young. Slavemaking ants can show high prevalence, so that the resulting parasite pressure can lead to the evolution of fine-tuned host defense strategies, including enemy recognition, flight, and fighting strategies (Foitzik et al. 2001; Bauer et al. 2009; Jongepier et al. 2014; Kleeberg et al. 2014, 2015). These host defenses might be one of the reasons why slavemakers generally do not employ a sneaking strategy or avoid aggressive escalations during raids. Instead, they enter host colonies forcefully and openly attack adult hosts. The downside of this aggressive strategy is that adult host workers cannot be manipulated to join slavemaking nests; instead, slavemakers rely entirely on the enslavement of the hosts’ brood (Schumann 1992; Mori et al. 2001; Herbers and Foitzik 2002).

Temnothorax longispinosus, Temnothorax curvispinosus, and Temnothorax ambiguus, hosts of the slavemakers Protomognathus americanus, Temnothorax duloticus, and the focal species here, Temnothorax pilagens, generally recognize slavemakers as enemies (Alloway 1990) and respond to invasions with counterattacks (Alloway 1979; Pamminger et al. 2011). Next to aggression as a beneficial anti-parasite defense (Pamminger et al. 2012; Kleeberg et al. 2014), host colonies also react with fast nest evacuation (Alloway 1979; Jongepier et al. 2014). Slavemakers try to undermine these adaptive host responses through behavioral (e.g., nest entrance guarding; Alloway 1979) or chemical strategies, such as manipulation through the use of the Dufour’s gland (Jongepier et al. 2015; Brandt et al. 2006). Slave raids by T. duloticus and P. americanus often escalate in open fights with the consequence that many hosts and slavemakers die during these encounters (Foitzik and Herbers 2001). Here, we focus on the newly described North American slavemaker T. pilagens (Seifert et al. 2014), a close relative of the two other slavemakers (Beibl et al. 2005; note that T. pilagens was then undescribed and referred to as T. spp.), which employs two different raiding strategies as preliminary observations indicate. (a) Peaceful raid: during most raids, neither the slavemaker T. pilagens nor colonies of its two Temnothorax hosts show aggression. Besides stealing the host brood unmolested, the slavemaker carries the adult host workers into its own nest and integrates them into its slave workforce (“eudulosis,” Kutter 1957). So far, eudulosis has only been described in the slavemakers Strongylognathus afer, Formica naefi, and infrequently in Polyergus rufescens, and it is as yet unclear how adult enslavement is accomplished (Kutter 1957; D’Ettorre and Heinze 2001; Sanetra and Guesten 2001). Ants learn their colony odor as young adults and hence attack chemically distinct slavemakers rather than being induced to work for them. This makes eudulosis rather difficult to accomplish and might be a reason why it is so rare. (b) Escalated raid: Occasionally, however, even in T. pilagens, raids can escalate. Preliminary observations indicate that T. pilagens switches its behavior and stings most adult hosts to death. Especially, the peaceful strategy of T. pilagens is clearly distinct from the strategies used by the closely related slavemakers P. americanus and T. duloticus. Eudulosis, that is the reprogramming of adult workers to serve as slaves, is fascinating. However, for the fitness of host workers, which are either victim to a peaceful or an escalated raid, it might make no difference whether they are enslaved or killed, as enslaved host workers do not reproduce (Gladstone 1981).

First, observations of the peaceful raiding strategy of T. pilagens let us predict that this slavemaker is able to prevent raid escalation by outwitting the hosts’ recognition system. Here we analyze how T. pilagens manages the apparent conflict, chemically or behaviorally, to maximize its raiding success. We predict that Temnothorax hosts should show less aggression toward their sympatric T. pilagens slavemaker than toward the other two slavemaking species. To circumvent recognition and aggression, T. pilagens is expected to exhibit chemical adaptations, be it mimicry or insignificance, and we therefore analyzed its cuticular hydrocarbon (CHC) profile and compared it to that of its hosts and to that of the two related slavemakers T. duloticus and P. americanus. Furthermore, we staged raids in the laboratory to unveil the causes and consequences of the two strategies. We hypothesize that raids only escalate if host colonies recognize and attack the slavemaker, which will be counteracted by a killing frenzy of T. pilagens. Avoiding raid escalation and keeping the interaction peaceful should be more beneficial for the slavemaker, as they not only can steal brood unmolested but can moreover exploit adult hosts.