What we have figured out about social insects?

  • This section is still under development, but here are some fun things we have done.
  • Click on picture to see some footage of various species of Polises
  • A tropical wasp, Mischocyttarus mexicanus, has arrived and established in Houston and San Antonio from Florida(Carpenter et al. 2009).
  • Unicolonial ants lack colony boundaries and so are expected to be short-lived as species and to lose worker traits (Helantera et al. 2009).
  • Polistes dominulus in its native Tuscany does not appear to have status badges (Cervo et al. 2008).
  • Polistes wasps do not remove the larvae of social parasites from the nest (Cervo et al. 2008).
  • In Polistes wasps, queen is not a pacemaker who activates workers to forage; instead they are self-organized (Jha et al. 2006).
  • In Polistes, social parasites are monophyletic and do not form host races (Fanelli et al. 2005, Choudhary et al. 1994, Carpenter et al. 1993).
  • We describe the phylogenetic relationships of the polistine wasps (Arevalo et al. 2004).
  • In Polistes gallicus males are commonly produced by workers at the end of the season after the queen dies – do the workers hasten her end? (Strassmann et al. 2003).
  • Nest foundresses in Polistes carolina have a dominance hierarchy based on order of arrival; they nest only with relatives, and do not behave in ways predicted by skew theory (Seppä et al. 2002).
  • Multiple mating is rare in social insects, occurring only in highly derived species (Strassmann 2001). This facilitated the origin of eusociality in Hymenoptera.
  • Social insects overwhelmingly help relatives reproduce. We discovered in the social wasp Polistes dominulus 35% of helpers are unrelated to the queen, apparently helping for the chance to inherit the nest and workers (Queller et al. 2000).
  • Foundresses of Polistes carolina do not preferentially feed their own larvae, which is an indication of the absence of within-colony kin discrimination (Strassmann et al. 2000).
  • Unrelated ant foundresses cooperate at colony foundation, then fight to the death for the colony resource (Bernasconi and Strassmann 1999).
  • Cooperation in social insects can be explained by kin selection. Benefits and costs can be broken down into those for fortress defenders and those for life insurers (Queller and Strassmann 1998).

What we have figured out about stingless bees

The social insects provide one of the best models for studying the evolution of cooperative behavior. Kin selection theory explains why a social insect might help relatives instead of trying to reproduce alone, but a more difficult problem remains unsolved. Individuals, once in groups, should often be selected to prefer being a reproductive rather than a helper. But if this is so, why is reproduction so often limited to one queen (or a few queens), particularly in large colonies where the queen cannot hope to physically dominate all her subordinates? A likely hypothesis is that average or collective worker interests determine who reproduces, including the possibility that workers often supress each other.

One test of this theory to date is in the honey bee, where extreme multiple mating by the queen causes workers to be more related to her sons than to the sons of other workers. The Strassmann/Queller group studied worker preferences with respect to male production in a related group, the stingless bees (Apidae: Meliponinae). They have previously found that stingless bee queens are generally singly mated, which causes workers to be less related to the queen’s sons than to each others¹ sons. The theory of collective worker control therefore predicts that stingless bees will differ from honey bees: workers should lay the male-destined eggs in stingless bees. We found that this was the case, with complications.

For each species, genotypes of highly variable microsatellite loci were used to confirm that the queen is singly mated, and to determine the fractions of males derived from the queen and from workers. These results were related to behavioral observations of conflict during the cell provisioning and oviposition process of each species, a period of highly ritualized behaviors that varies extensively among species. We confirmed the hypothesis that behavioral conflict is highest in species where both queens and workers produce some of the males, less high in species where both parties can produce males but only one does, and lowest in species where workers cannot produce males.

  • Predicted conflict between workers and the queen over male parentage because of relatedness asymmetries in stingless bees actually occurs and has different outcomes in different species, including workers opening closed cells to sneak in their own eggs (Tóth et al. 2002a, 2002b, 2003, 2004).
  • Stingless bee queens mate with a single male, in contrast to honeybee queens, which mate with many males. In accord with theory, stingless bee workers tolerate virgin queens in the nest, attempt to produce the males, and a daughter queen leaves the nest during swarming (Peters et al. 1999).

What we figured out about tropical swarm-founding wasps

The social insects provide some of the most extraordinary and puzzling products of evolution. They live in colonies in which most individuals serve as workers, giving up their own reproduction in order to help one or a few queens. The degree of cooperation and integration that occurs in some species has led to their colonies being likened to organisms. Yet social insect colony-mates, unlike the cells of an organism, are not genetically identical, so it is much more difficult to explain their extreme self-sacrifice and integration.

Kin selection theory has successfully explained why a social insect might help instead of trying to reproduce alone, but a more difficult problem remains unsolved. Given that an individual decides to join a group, it should often be selected to prefer being a reproductive, to be helped rather than to be a helper. But if this is so, how is anarchy avoided, particularly in large colonies where the queen cannot hope to physically dominate all her subordinates? The most promising hypothesis appears to be collective worker control. Even if each worker might prefer to reproduce, under certain conditions they are predicted to suppress each other in favor of the queen. Though this hypothesis is theoretically appealing, it has not been widely tested. We have taken advantage of extensive past work on a species of social wasp, Parachartergus colobopterus, to test collective worker control hypotheses in three distinct contexts. First, do workers choose who lays the haploid male eggs? Second, do workers choose which of the numerous queens will survive to lay the eggs that will become reproductive females? Third, do workers choose the timing and the identity of new queens?

The methods  employed are three. Behavioral observations, recorded on videotape, are made for colonies with individually-marked wasps. These focus on behaviors that show who is reproducing (egg laying) and behaviors that might influence who reproduces, (such as egg eating, feeding, and aggression). The observation colonies are collected and brought back to the lab for DNA microsatellite analysis. Microsatellites are genetic loci that are both numerous and highly variable. They are used in two ways. First, they allow precise estimates of relatedness that will be used to predict who workers should prefer to reproduce. Second, are used to assign maternity, that is, to determine who does reproduce, and how much. Together these three types of data determine what workers should prefer, whether their behavior is consistent with that preference, and whether that preference is actually achieved.

  • Workers keep young females from becoming queens by chewing on them and dominating them (Platt et al. 2004).
  • We found that the African species Polybioides tabidus exhibits cyclical oligogyny where queens are only produced in the single queen stage, just as we found in neotropical wasps (Henshaw et al. 2000).
  • Workers and queens concur in preferring queens to produce males, but differ in sex ratio preferences in Polybioides tabidus. We found indirect evidence that workers kill some male brood to favor their own sex ratio interests over those of the queen (Henshaw et al. 2002).
  • Pupae that would have become workers can become queens if all the current queens are removed, demonstrating caste plasticity (Strassmann et al. 2002).
  • Worker and queen relatedness interests agree on queen male production in Parachartergus colobopterus. Queens, not workers, do produce the males (Henshaw et al. 2000).
  • Queens do not orchestrate worker behavior or suppress others from becoming queens in Parachartergus colobopterus; instead queens function only as egg layers, essentially as the colony’s ovaries (Herman et al. 2000).

More detail on some topics

  • Microsatellites and kinship: Studies of social behavior badly needed a genetic marker that could be used to identify relatives, reveal extra-pair parentage, and quantify population structures. DNA fingerprinting showed promise, but was not easily quantified. We quickly realized that DNA microsatellite loci, also called simple tandem repeats, were the ideal markers for behavioral studies, because they are common, Mendelian, and easily visualized. The challenge is in identifying unique primers to amplify single loci that vary in repeat numbers. In a series of papers, we publicized the strength of the method, showed how to find primers, and showed what the markers could reveal about social wasp and bee sociality.
  • Kin selection and social insects: In this review we introduce the concepts of fortress defense and life insurance as the two main reasons sterile castes have evolved. Fortress defenders like termites, aphids, thrips, naked mole rats, and social shrimp, live in their food. Defense of the nest, brood, and food source is paramount. If there are morphological castes, the first to evolve will be defensive soldiers. Life insurers, by contrast, forage outside the nest for food. If one individual dies, something that is common for external foragers, another can take over rearing the half-grown brood. Wasps, bees, and ants show this kind of social advantage. The first caste to evolve is likely to be foragers.  Hymentoptera have pre-adaptations for sociality which include maternal care of brood, a powerful sting, originally turned on prey, but which could be used as defense. In the face of these advantages and preadaptations, we argue that haplodiploidy is unlikely to be important for the evolution of eusociality in hymenoptera.
  • The rarity of multiple mating in social hymenoptera: In this paper we review the literature on multiple mating by females in social hymenoptera. We make the argument that single mating is pervasive and ancestral. Multiple mating is rare and found only in a few highly social ants and bees where its advantage seems to come from increasing genetic variability in the brood, a gain that may help with disease avoidance, or caste differentiation. Interestingly, the highly social wasps of the neotropics have multiple queens, not multiple mates of a single queen.
  • Unrelated helpers in a social insect: We discovered that in Tuscany Polistes dominulus wasps begin nests with unrelated females about a third of the time, so we were interested in why a female would help a non-relative.  The females that joined non-relatives had sisters in the population. They were not smaller than other females. They did not join later than other females. They did not show more aggression towards the original queen. They are able to recognize relatives. We hypothesize that they join non-relatives as a bet-hedging strategy geared towards acquiring nest sites for the future. The wasps nest in highly sheltered locations that are hard to find.
  • Mate number, kin selection, and social conflicts in stingless bees and honeybees: Colonies of stingless bees and honeybees are both headed by single queens, but in honeybees these queens mate many times in a short time while we here document that stingless bees mate only once. This means that there should be more conflict among workers over male production in stingless bees than in honeybees. We show that this is the case in this and other papers.
  • Evolutionary implications of early male and satellite nest production in Polistes exclamans: A queen who lays eggs and dominates the workers heads a paradigm social wasp colony. We discovered that in this Texas species of wasp, workers sometimes leave their natal nest and begin a satellite nest. The queen herself may also begin a satellite nest. Satellite nests are initiated from thriving nests. They attract additional workers. If either nest is destroyed by a predator, the second nest is joined by all the survivors. They are typically built in hidden locations and are a way of surviving through a very long growing season.
  • Gerontocracy in the social wasp Polistes exclamans Queens often die during the long nesting season in Texas in open nesting wasps like Polistes exclamans, though less so in hole-nesting wasps like Polistes carolina. In a series of experiments where we removed the queen, then waited to see which worker would become the new queen, we discovered that there is a strict hierarchy: the oldest worker becomes the new queen. This is interesting because it means that foraging does not take the worker off the potentially reproductive path.
  • The challenge of multiqueen tropical wasps Bill Hamilton viewed tropical wasps as very challenging for the theory of kin selection because they have so many queens. If they all reproduce, and have mated with different males, then genetic relatedness of workers to brood would be so low that very high advantages to sociality would be required to offset the cost of helping distantly related young. We showed that relatedness is not low in these multiple queen colonies because each queen mates only once, and new queens are only produced when queen number has dropped to one. This means the queens on a nest are related as full sisters. The pattern, called cyclical oligogyny, is universal for the neotropical epiponine swarm-founding wasps, and also for the African polybioides wasps.
  • Caste and conflict in Polistes wasps Primitively eusocial wasps have workers with the potential to become queens. We showed that in Polistes exclamans a worker can mate and become a fully functional queen if the original queen is removed. Furthermore, we found that it was the oldest worker that becomes queen, even if she has previously been an active worker. This is a remarkable degree of caste plasticity. It is a kind of hierarchical convention, something we also found among foundresses. In several species, we found that it is the first to arrive at a nest site that becomes queen, even if later arrivers are larger. This kind of hierarchy, or queuing reduces conflict. In another experiment we demonstrated that newly emerging females became workers if they had larvae to care for, and hibernating future queens if they encountered no larvae to tend. This experiment definitively shows that caste is not fixed before the adult stage. A related experiment with an epiponine wasp showed that caste was plastic there too.
  • Social parasites in Polistes A social parasite is a closely related species that takes over the queen role in colonies and uses the workers to care for her brood. There are three species of social parasites in Polistes, all circum-Mediterranean. We have shown that they are monophyletic, then diverged to parasitize different species in slightly different ways.