Our study shows that suicidal biting as an antipredator defense occurs in stingless bees. We believe this is the first clear demonstration of suicidal biting by any insect worker. The results support our general impression from casual observations that Trigona species are particularly defensive and even suicidal. The three Trigona species led the nine other species in all four aggression measures in the flag test, had the most painful bites, and had the largest proportion of self-sacrificial individuals in the suicide bioassay. The most suicidal was T. hyalinata in which almost all the bees tested (83 %) suffered fatal damage rather than disengage their mandibles from the flag. However, self-sacrifice was not confined to Trigona as it occurred in a significant proportion (7–23 %) of the test bees in the three other species submitted to the suicide bioassay (P. helleri, S. depilis, T. clavipes). This represents a new example of convergent evolution with other suicidal insect workers. Because levels of within-species self-sacrifice may be low, suicidal behavior may be difficult to detect and, therefore, potentially more widespread than previously thought. However, our results make it clear that not all stingless bees have suicidal biting. In fact, three of our study species never attacked the target flag at all.
For suicidal worker defense to evolve, the inclusive fitness benefits gained over non-suicidal defense, in terms of repelling intruders, must be greater than the costs incurred due to reduced worker numbers. Bees which attack more often, in greater numbers, with shorter latencies, and for longer durations will presumably be more effective at repelling the current attack and deterring potential future attacks (Schmidt 1990). Furthermore, in committing self-sacrifice through their jaw clamping behavior, stingless bees can immobilize or kill intruding insects (Grüter et al. 2012) and cause longer-lasting pain to vertebrate predators, preventing further attack on their colonies.
Higher levels of colony defense are likely to increase both colony survival and the mortality risk to the defender. Natural selection should, therefore, favor an optimal level of defense, where colony survival is traded off against the future value of the defender to the colony (Andersson et al. 1980). The optimal level of defense should increase with colony size because the colony contains more kin and is thus of greater value. In social insects, colony size can be large, 10,000s of individuals in Trigona for example (Roubik 2006), while the reproductive value of workers is low. Furthermore, many social insects, including stingless bees, exhibit age polyethism (Sommeijer 1984), where the risky tasks such as guarding are performed by the older workers with shorter life expectancies (Tofilski 2002). These factors can lead to a very high investment in defense and, potentially, the decision to commit self-sacrifice (de Catanzaro 1986; Brown et al. 1999). Eusocial insects could thus be described as having an exaptation for self-sacrificial behavior.
In social insects, each additional worker adds proportionally less to colony fitness (Michener 1964; Nonacs 1991). The relative costs of sacrificing workers will thus be less in large colonies than small ones. Stingless bee species vary greatly in colony size (Wille 1983). A suicidal attack of 20 bees from a 10,000-strong Trigona colony versus a 100-strong Melipona colony represents a loss of 0.2 versus 20 % of the total worker population. In small colonies, the costs of mass attacks, especially those involving suicidal behavior, would likely be greater than the benefits from improved defense. Our results support this theory, as the most aggressive and self-sacrificial species in the study (Trigona) were those with the largest colonies (Roubik 2006). Attacking intruders singly does not represent an effective defensive strategy because the pain and damage per bite, while unpleasant, is unlikely to drive an intruder away. Mass attacks are therefore required to offer a more robust defense. A formal analysis of colony size and aggression is not within the scope of this paper, as only 12 species were studied, of which suicidal behavior was only observed in six. This does, however, raise two important questions: First, is the mean colony size of a species a good predictor of aggression and self-sacrificial behavior? Second, is there intraspecific plasticity in aggression and self-sacrifice between small and large colonies?
Bites from the most aggressive species, namely the three Trigona species, were the most painful. The non-aggressive species which could not be provoked into attacking in the flag tests could not be provoked into biting human skin either. Larger species also tended to be more painful, but the only species to bite larger than the three Trigona species, M. scutellaris, was only mildly painful. Closer examination of the mandibles revealed that the Trigona species possessed serrated mandibles bearing sharp teeth. This morphological specialization presumably allows Trigona to cause more pain and damage to intruders, and as pain was correlated with overall aggression, it suggests that these mandibles are adaptations which enhance colony defense.
Mandibular teeth, however, may have also evolved in response to other selective pressures. Stingless bees use a variety of materials to construct their nests including resin and soil (Wille 1983). Toothed mandibles may aid in the acquisition and manipulation of such materials, as Trigona mandibles are similar in appearance to those of mason bees (e.g., Osmia bicornis: Megachilidae) and reminiscent of the fossorial forelegs of mole crickets (Gryllotalpidae). However, P. helleri nests are composed largely of soil but this species possesses only a single small tooth on each mandible. Trigona are also known to aggressively defend foraging patches against other bees, using their mandibles to harass, bite, and kill competitors (Johnson and Hubbell 1974; Nagamitsu and Inoue 1997). The vulture bee Trigona hypogea feeds on carrion and fruit in place of pollen and nectar, and mandibular teeth may facilitate foraging on such alternative food sources (Roubik 1982). While the Trigona species in the present study are not obligatory necrophagous, Wille (1983) suggested that they may turn to carrion when pollen sources are scarce.
Although the serrated mandibles of Trigona are a morphological feature that almost certainly enhances the effectiveness of their biting defense, their self-sacrifice is primarily behavioral in nature through simply refusing to let go. Bees in the suicide bioassay were often so engrossed in their attack on the flag that they made no attempt to evade the brush or forceps. Suicidal biting differs from most previously known examples of self-sacrifice in worker insects in lacking a morphological mechanism that guarantees mortality. Honey bee sting autotomy and autothysis in Camponotus spp. ants both nearly always result in the death of the worker (Hermann 1971; Shorter and Rueppell 2012). The stingless bees in our study, however, showed a gradation in suicidal behavior, and in all but T. hyalinata, mortality was less than 50 %.
Non-aggressive stingless bee species should not be thought of as defenseless, as biting is only one of a wide variety of defensive adaptations (Kerr and de Lello 1962). For example, when provoked during the flag tests, guards from the non-aggressive species in our study retreated from the nest entrance rather than confront the flag. These species tended to be those with very small entrances relative to their body sizes (Couvillon et al 2008a). This strategy represents the opposite of a mass attack, where intruders must combat guards singly in a narrow space. Several species in our study, most notably the mildly aggressive Frieseomelitta varia and moderately aggressive T. clavipes, frequently deposited sticky, odorous resins on the flag. While this behavior would have little effect on a vertebrate predator, it is likely very effective at immobilizing other stingless bees and may be similar in function to secretions found commonly in ants and termites (Prestwich 1979; Bordereau et al. 1997; Davidson et al. 2012).
Killing or disabling intruders is especially important in defending the nest against robbing by other stingless bees, as allowing scout robber bees to successfully scout can result in mass attacks on the colony and potentially far greater costs than the loss of a few suicidal workers. Biting defense is seen in the conflicts between one of our study species, T. angustula, and the obligate robber bee Lestrimelitta limao. Despite a large size disadvantage, T. angustula guards are able to clamp onto the wings of L. limao for long durations. This prevents the robber from flying and returning to its own colony but often results in the death of the T. angustula worker (Grüter et al. 2012). This is paralleled by the thermal defense displayed by honey bee workers against scouts of the Asian giant hornet Vespa mandarinia (Ono et al. 1995).
Our study has shown a wide range in the aggressive, defensive behavior of stingless bees. The presence of suicidal defensive biting in half our study species indicates that this behavior is potentially a widespread defensive strategy. In our experience, the three Trigona study species will almost invariably attack any human standing within a few meters of a nest entrance, often within seconds. Workers attack the head but also other parts of the body. So tenacious and unpleasant is the attack that the victim is forced into a hasty retreat. Bees are especially difficult to remove from hair, and if a bee is removed and released, it usually returns to the head immediately and resume its attack. The only recourse for the victim, therefore, is to flee and kill the bees to stop the attack.