Social Insects

views updated

Social insects

The fascination of social insects

The social insects represent one of evolution's most magnificent, successful, and instructive developments. Ever since the behavior of ants, bees, wasps, and termites was first recorded in antiquity, these insects have exerted a powerful hold on our imaginations.

Three characteristics of social insects account for this interest. The first is the very habit of living in social groups. To biologists, this way of life represents a fascinating evolutionary innovation, yet it also poses a basic dilemma. Contrary to what we may have commonly learned, the essential history of life on Earth has not been the story of the rise and fall of dominant groups of animals like the dinosaurs. Instead, life's history has been the succession of what evolutionary biologists J. Maynard Smith and E. Szathmáry termed "major transitions" in evolution. Without any one of these transitions, living things today would be fundamentally different.

Two of these major transitions are the evolution of sexually reproducing organisms from those that reproduce asexually, and the evolution of multicellular from single-celled organisms. These transitions share a radical reorganization of living matter. In particular, at each transition, existing units coalesced to form larger units. In the process, the original units lost some or all of their power of independent reproduction, and the way genetic information is transmitted was changed. The social insects are the prime examples of the next major transition, from solitary organisms to social organisms. As a result of this change, workers of social insects have largely lost the power of reproducing independently, and genetic information has come to be primarily transmitted from generation to generation via the reproduction of the queens (and kings in termites). Why these workers have foregone their power of reproduction is a basic dilemma that insect sociality poses for biologists.

The second characteristic of social insects, which explains why they command attention, is their overwhelming numerical and ecological dominance. Ants, for example, occur from the Arctic tundra to the lower tip of South America. They swarm in diverse habitats, ranging from desert to rainforest, and from the depths of the soil to the heights of the rainforest canopy. The abundance and ubiquity of social insects give them an ecological importance that is unmatched among land-dwelling invertebrates.

The third and final fascinating characteristic of social insects is the way in which they perform work. Clearly, social insects carry out cooperative tasks, such as building nests, with results that are stunningly complex and precise. Termites build nests that are towering hills of red clay, with intricate internal architecture designed to maintain a cool, stable interior. The nests of honey bees house waxen combs composed of thousands of exquisitely arrayed hexagonal cells. Social wasps construct nests in which parallel tiers of cells are enclosed within a delicate paper globe.

These nest forms point to the ability of social insects to achieve, as a collective, impressive feats of architectural complexity. However, biologists have long puzzled over how social insects complete these complex tasks as a group when individual workers frequently strike us as inept. "It seems to me that in the matter of intellect the ant must be a strangely overrated thing," wrote Mark Twain after watching the fumbling meanderings of a worker carrying a grasshopper's leg back to its nest. The ant's seemingly unnecessary scaling of an obstacle was, Twain went on to say, "as bright a thing to do as it would be for me to carry a sack of flour from Heidelberg to Paris by way of Strasburg steeple." In this chapter, we first review some basic biology of social insects, then consider the three principal characteristics of social insects in turn.

An outline of social insect biology

What are social insects?

The social insects are not a taxonomic group. Instead, the transition to living in societies has evolved independently in several lineages of insects. This happened in the very distant past. For example, ants are thought to have arisen by approximately 80 million years ago, and the termites even earlier. The societies, or colonies, of social insects are distinct from mere swarms or aggregations. Species of insect qualifying as truly social, or eusocial, exhibit three traits. The first, and most important, is a reproductive division of labor. Some individuals (the queens and kings) are fertile and reproduce, whereas others (workers, or neuters) are sterile or semisterile. The workers either do not reproduce, or reproduce to a far lesser extent. The second trait is that the workers perform tasks of benefit to the colony, and in particular, cooperatively

rear the young. The third trait is that the society has some permanence and, specifically, that more than one adult generation coexists in the colony.

The major groups of eusocial insects are to be found in the Hymenoptera (ants, bees, wasps, and relatives) and the Isoptera (termites). The ants (Formicidae) are all eusocial, and represent the most species-rich group of social insects, with over 9,000 described species. The majority of bees are not social, but the group includes such well-known eusocial groups as the honey bees (Apini), stingless bees (Meliponinae), and bumble bees (Bombini), as well as less-conspicuous eusocial groups such as the sweat bees (Halictinae). There are about 1,000 eusocial species of bees. Eusocial groups among the wasps (Vespidae) include the hover wasps (Stenogastrinae), the paper wasps (Polistinae), and the yellowjackets and hornets (Vespinae). There are around 1,200 eusocial species of wasps. The termites (Isoptera), like the ants, consist only of eusocial species, and number over 2,800 species. Eusocial representatives also occur in at least two other insect orders, the Hemiptera (bugs and aphids) and the Thysanoptera (thrips). All eusocial hemipterans and thrips live in plant galls and possess a defensive "soldier" morph. Eusocial hemipterans consist of 60 aphid species in the two subfamilies, Hormaphidinae and Pemphiginae; about six species of eusocial thrips have been described in the family Phlaeothripidae.

The structure and life cycle of insect societies

The typical social insect colony is a family. In ants, bees, and wasps, the workers are all females. Males die after mating with queens and so take no part in social life. The colony therefore consists of a queen, her adult worker daughters, and the brood (eggs, larvae, and pupae) produced by the queen. The job of the workers is to rear the brood to produce new workers, new queens, and males. The new queens and males are collectively referred to as reproductives, or sexuals. The workers carry out their job by foraging for food, building or defending a nest, and feeding and tending the brood. Termite colonies have the same basic structure, except that workers are of both sexes and the queen is accompanied by a reproductive male, a king.

The life cycle of the typical social insect begins with the formation of a colony by a mated queen (in the ants, bees, wasps) or by a queen and king (in termites). The colony then enters a growth phase, with the founding individual or pair producing worker offspring that gather resources for the colony and thus allow it to expand. Once the colony is sufficiently populous, the reproductive brood of new queens and males is produced. As adults, these leave the colony, mate or pair, and disperse to found new colonies of their own. The mating and dispersal flights of ants can involve the sudden summertime emergence of thousands of winged queens and males ("flying ants"), which attract attention, and occasionally dismay, in the cities of northern Europe and North America.

There are many variations on this basic colony structure and life cycle. Colonies may be started by a single queen, as in the garden ant (Lasius niger), or multiple queens, as in the desert ant (Messor pergandei). Mature colonies may be headed by one queen, again as in the garden ant, or by more than one queen, as in the red ant (Myrmica rubra). Likewise, in termites, mature colonies may be headed by just a single royal pair or several such pairs.

Queens may mate with one male or many males. For example, queens of the red imported fire ant (Solenopsis invicta) always mate with one male; whereas queens of the domestic honey bee (Apis mellifera) mate on average with about 17 males, and queens of A. nigrocincta from Southeast Asia mate, on average, with the extraordinarily high number of 40 males.

Colonies may reproduce, not by releasing winged sexuals, but by splitting in two, as in the wood ants (Formica spp.). They may occupy a single nest site or, as in the acorn-nesting ant (Leptothorax longispinosus), several neighboring nests. Colonies may be annual or perennial. Those of bumble bees (Bombus spp.) are annual, founded by a single queen in spring. Before the year is over, she produces workers and reproductives and then dies. She leaves behind the new generation of mated queens to found fresh colonies in the following year. Colonies of honey bees are perennial. They are founded by a queen and a swarm of workers and, over the next two to three years, the queen produces reproductives each year before dying.

Queens of perennial social insects are probably the longestlived of all insects, with the record being held by a captive garden ant queen that lived 29 years. Insect colonies also vary enormously in average size. At one extreme, some halictine bee colonies contain just a handful of workers. At the other, colonies of the army ants and driver ants of tropical America and Africa (Eciton and Dorylus spp.) have a workforce numbering hundreds of thousands or even millions.

Social insects exhibit two additional important variations in their social structure. The first is social parasitism. Some species of ants, bees, and wasps do not produce workers, and queens subsist as parasites in the nests of other social insects, in the way that cuckoos parasitize the labor of nesting birds. One example of this practice is found in ants of the genus Epimyrma. After mating, a queen enters a colony of a related species and kills the existing queen by the macabre method of straddling her back and slowly throttling her. The workers, fatally to their interests, seem unaware of the death of their queen, and go on to rear eggs laid by the parasite.

The second important variation in social structure is unicoloniality. Populations of social insects typically exist as a set of distinct colonies that are tightly defended against all intruders, including workers of the same species. In unicolonial ants and termites, the social boundaries that maintain the distinctiveness of colonies have disappeared, and the population is a loose network of nests among which individuals appear to be interchangeable. In its introduced, European range, the Argentine ant (Linepithema humile) occurs in a vast unicolonial population, or "supercolony," which runs for over 3,728 mi (6,000 km) along the Mediterranean coast from Spain to Italy. Unicoloniality represents the emergence of another grade of sociality among insects. In fact, to the extent that unicolonial populations represent groupings of whole societies, we see in them the beginnings of a further major transition in evolution.

Caste in the social insects

The queens and workers of social insects are often referred to as castes. The distinction between castes may be behavioral or morphological. For example, in paper wasps (such as Polistes spp.), the queen and her workers do not differ morphologically. Instead, the queen maintains her position as the principal or only egg layer, by aggressively dominating the workers. By contrast, in vespine wasps (such as Vespula spp.), the queen is larger than the workers, differs from them morphologically, and is rarely aggressive. The most extreme caste dimorphism between queens and workers is found in ants and termites, especially in species with very large colonies. The queens of termites are among the largest. With a massively enlarged abdomen, a queen Macrotermes can reach 5.5 in (14 cm) long and 1.4 in (3.5 cm) wide, and be capable of laying 30,000 eggs per day. In general, queens, specialized for dispersal and egg laying, have a relatively large thorax and abdomen. The workers, specialized for labor, lack these features and are relatively small. In ants and termites, workers are always wingless, whereas young queens bear wings that they shed after their dispersal or mating flight.

Workers may themselves form separate subcastes. In the leaf cutter ants (Atta spp.), there are at least four distinct size-classes of worker in each colony, varying 200-fold in body weight. Each subcaste performs the tasks most suited to its body size. The largest, the major workers, cut green leaves from plants outside the nest. The smallest, the minor workers, who labor in the narrow tunnels deep within the nest, insert the chopped-up leaf pieces into the growing mass of the symbiotic fungus on which these ants feed. In other ants, major workers act as a defensive subcaste, the so-called soldiers. Only ants and termites exhibit morphological worker subcastes. However, even in species without them, worker subcastes exist in a temporal sense, since workers tend to change their jobs as they grow older. Typically, young workers perform tasks inside the nest such as feeding and cleaning the brood, and old workers perform external tasks such as defense and foraging.

The differences between the castes are, with few or even no exceptions, not based on genetic differences. Instead, whether a female egg develops into a queen, worker, or worker subcaste depends on how she is reared, and particularly on her nutrition. For example, in the honey bee, queens are reared in large cells and receive a special diet from workers, the so-called royal jelly. Workers are reared in small cells and do not receive royal jelly. At the genetic level, development as either caste must depend on the differential expression of genes present in all females. In other words, honey bee larvae in large cells that are fed royal jelly switch on one set of genes and develop as queens. Larvae treated otherwise fail to switch on these genes, or switch on an alternative set, and develop as workers. Seven of the genes involved in this process have been isolated and identified in honey bees, but very little is known about corresponding genes in other social insects.

The evolution of social behavior in insects

The origin of sociality

Explaining the origin of sociality in insects has been one of the great successes of evolutionary theory. Biologists generally agree that the likeliest pathway by which insect sociality arose was through offspring remaining in the nest as adults to help rear their mother's young. Many Hymenoptera exist (so-called solitary or nonsocial species) in which females feed their larvae themselves, unaided by workers. Some populations of halictine bees contain both nests of this sort and nests in which the parent is assisted in brood-rearing by her earliestemerging daughters. We can therefore observe, in a single population, the likely state of affairs immediately before and immediately after the appearance of sociality. What, though, might be the evolutionary mechanism driving this process?

A very widely accepted answer is kin selection. Kin selection theory, devised by the outstanding evolutionary biologist W. D. Hamilton in the 1960s, assumes that social behavior (here the altruistic rearing of brood by sterile workers) is subject to genetic variation. In fact, genes that affect social behavior have been isolated and sequenced in the red imported fire ant. Next, kin selection theory proposes that genes for altruism can be naturally selected if aid is directed at relatives (kin). This is because relatives of an altruist are, by definition, likely to share its genes for altruism. Therefore, by rearing enough relatives, workers can add sufficient genes for altruistic behavior to the next generation to make up for those they fail to add by not producing offspring themselves. This is why the basic social structure of a social insect colony is a simple family in which workers rear their mother's sexual offspring, which are the workers' siblings. As we have seen, in some species there are several queens per colony, and in others queens mate multiply. Their colonies are extended, rather than simple families. Such cases are assumed to be secondary developments in which the dilution in relatedness among individuals has been compensated for by greater efficiency among workers in rearing broods.

Hamilton's kin selection theory is equally applicable to the social Hymenoptera and the termites. In both groups the first societies are likely to have involved individuals remaining in the nest to help rear siblings. There has been a great deal of discussion about a secondary facet of Hamilton's theory, which highlighted the unusual genetic system exhibited by the Hymenoptera, termed haplodiploidy. In haplodiploid organisms, females are diploid (have two sets of chromosome per cell) and develop from fertilized eggs; males are haploid (have one set of chromosomes per cell) and develop from unfertilized eggs. This leads to an unusually high relatedness among sisters. Female offspring of the same mother and father can receive one of two genes from their mother at each locus (position on a chromosome). But they receive only one gene from their father, as being haploid, he has only one gene copy per locus. This raises their average chance of gene-sharing from the usual one-half (as found in wholly diploid organisms) to three-quarters.

Hamilton speculated that the especially high relatedness among haplodiploid sisters accounted for the concentration of independently evolved social lineages in the Hymenoptera and for the fact that workers in the Hymenoptera are always females. This has been termed the haplodiploidy, or three-quarters relatedness, hypothesis. The hypothesis has proved difficult to test and remains unproven. The reason is that features of the Hymenoptera other than the three-quarters relatedness of sisters may have led to the group's propensity to evolve societies based on female-only workers. Separating the contribution of these features from that of haplodiploidy itself requires a good knowledge of the evolutionary tree (phylogeny) of the nonsocial and social Hymenoptera and the traits associated with different points along the tree. As more is learned about these, it should become clearer whether the haplodiploidy hypothesis is upheld.

Conflict inside social insect colonies

A close observer of life inside a bumble bee colony, especially toward the end of the colony cycle, will notice that there is a great deal of friction among the members of the society. Workers jostle the queen, eat her eggs, and attempt to lay their own eggs. The queen retaliates by acting aggressively toward the workers and trying to eat the workers' eggs. Such a colony departs from the traditional picture of a social insect

society in two ways. The first is that the workers are not completely sterile. This is possible because in bumble bees and many other social Hymenoptera, the workers possess ovaries but have lost the ability to mate. Due to haplodiploidy, this means the workers can lay eggs, but because they are unfertilized, these eggs always develop into males. The second way in which the colony is unlike a traditional insect colony is that it is clearly not an internally harmonious society. In some cases, aggressive, reproductive bumble bee workers even kill the queen, their own mother. This is the antithesis of how social insect workers are supposed to behave.

A great strength of the theory of kin selection is that it offers not just a reason for why a society evolves in the first place, but also an explanation of why conflict occurs within existing societies. The essential reason behind conflict is that different sets of individuals within the society are related to the reproductive brood to different degrees. Since natural selection leads them to act toward the reproductives according to their relatedness, conflict over reproductive decisions will therefore arise. In the bumble bees and other social Hymenoptera sharing their social structure, with a single, once-mated queen and workers capable of egg laying, a kin-selected conflict arises over male parentage. Queens are more closely related to the sons they produce and so favor these males being reared. Workers are more closely related to the sons they produce and so favor them being reared. The result is a prolonged squabble over egg laying. The conflict peaks toward the end of the colony cycle because, at this point, the colony moves from rearing workers (over whose production there is little disagreement, because both queen and workers favor the colony growing to be large and productive) to rearing sexuals (over which, for male sexuals, there is disagreement). Workers killing their own mother, which permanently prevents the queen from interfering with the workers' reproductive attempts, is an extreme, but logical, outcome of this conflict over male parentage.

Another well-studied form of kin-selected conflict in the social Hymenoptera is conflict between the queen and workers over the sex ratio (ratio of new queens to males). This arises because, due to their comparatively high relatedness to females, workers favor rearing a more female-biased sex ratio than is favored by the queen. This conflict has different outcomes according to the practical power held by each party. In some ants, the queen can apparently control the caste of female eggs, presumably by adding to the eggs substances that affect differential gene expression during their subsequent development. In this way, a queen can force the workers to rear the sex ratio optimal for her. In other (perhaps most) ants, the queen lacks the ability to determine the caste of eggs. The workers achieve their optimal sex ratio either by selectively killing the extra males in the queen's brood or by channeling more female larvae into development as queens.

Several other types of kin-selected conflict are known from social insects. In fact, potential conflict is present whenever more than one individual is capable of laying eggs (conflict over male parentage being a special case of this). The ponerine ant (Dinoponera quadriceps) is a species that has lost its morphological queen caste. One worker (the alpha) dominates the others as the chief reproductive, and is capable of producing both female and male eggs because, unusually, she remains able to mate. She uses her sting to smear any high-ranking challenger with a chemical that induces the other workers to pin down the upstart for as long as several days. As a result, the immobilized challenger loses her high rank. Both the alpha and the other workers gain from this behavior, because each is less closely related to the brood that a successful challenger would produce than to the alpha's brood. Such collective suppression, or "policing," of selfish behavior can also be provoked by the conflict over male parentage. It is likely to be a general mechanism for maintaining reproductive harmony. Hence, social insect colonies are indeed the integrated organizations that they are popularly supposed to be. But close investigation shows that this state of affairs is conditional. It depends, first on a high degree of common interest within the society (due to relatedness) and, second on social mechanisms that keep costly conflicts in check.

The ecological success of social insects

The habit of social living has clearly made the social insects extremely successful. By insect standards, they are not exceptionally species-rich, although in some habitats their local species diversity may be relatively high. But social insects are exceptionally abundant. The renowned evolutionary biologist and conservationist E. O. Wilson estimates that, although making up approximately 2% of known insect species, social insects form as much as 50% of the world's insect biomass.

As a result of their abundance and their diversity of diets and habits, social insects have a huge impact on other species and on the structure and functioning of ecological communities. They are outstanding "ecological engineers," in that their activities alter the physical environment and affect the availability of resources to other organisms. Such activities include pollination of flowers (by bumble bees, stingless bees, and honey bees), the dispersal of seeds (by harvesting ants), the cropping of green vegetation (by leaf cutter ants), the turning of soil (by soil-dwelling ants and termites), the consumption of wood and grass (by termites), and the predation of other insects (by predatory ants and wasps). Biologists have highlighted a certain set of "ecosystem services" supplied by living organisms on which the world's ecological health, and hence the economic health of human society, depends. Of these services, social insects contribute to at least three, namely pollination, soil formation, and the biological control of harmful arthropods. For example, the annual economic value of honey bee pollination in the United States has been estimated to lie between $1.6 and $5.7 billion. There are many reasons for valuing and conserving social insects, but this fact alone highlights the dangers stemming from any loss of pollinating bees from ecological communities.

Social insects are also partners in a series of unique symbioses with a vast range of other organisms. These include bacteria (in the leaf-cutting ants), fungi (in the leaf-cutting ants and some termites), and flowering plants (some ants only live inside special structure on certain plant species, which they protect from the attacks of herbivorous insects). These symbioses also include those with other arthropods (many mite species live symbiotically inside ant nests) and other insects (lycaenid butterflies are reared inside ants' nests as mutualistic guests or, in some species, as unwelcome parasites).

Another aspect of the ecological success of social insects is their extraordinary ability to invade areas outside their normal range, to which they have been introduced by human activity. No fewer than seven of the 14 species recognized as the world's most persistent invasive insects are social insects. These seven are the yellow crazy ant (Anoplolepis gracilipes), the Argentine ant, the big-headed ant (Pheidole megacephala), the red imported fire ant, and the little fire ant (Wasmannia auropunctata), along with the European wasp (Vespula vulgaris) and the termite Coptotermes formosanus.

One key to the success of social Hymenoptera as invasive species is the ability of a single queen to found a colony whose many reproductives can then seed the invasion of a large island or even a continent. In ants, another key is undoubtedly unicoloniality. A unicolonial network of nests can carpet fresh territory without competition between conspecific nests that might otherwise limit the ants' spread. Many, if not all, of the successful invasive ant species are indeed unicolonial. Once established in new territory, even introduced social insects that are not unicolonial can spread rapidly and reach very high numbers. For example, the European wasp (V. vulgaris) increased its range in New Zealand by 22 mi (35 km) per year and has attained the extremely high density of 12 nests per acre (30 nests per hectare).

Self-organization in social insects

Part of the great ecological success of social insects can be attributed to their sophisticated means of communicating with one another. In addition, by weight of numbers, a colony can overpower enemies and monopolize food sources. It can also perform many tasks simultaneously, and one set of workers can finish tasks that another set has left incomplete. Social insects communicate with their nest mates principally by releasing pheromones, but one exception is the famous dance language of honeybees. Here, a returning scout bee directs its fellow foragers to a flower patch by means of a dance whose orientation and duration indicate the bearing of and distance to the patch.

Communication among workers may be used to coordinate tasks, but there is very little evidence that any one individual, even the queen, directs the overall work of the colony. Control of colony activity is decentralized. As we have seen, social insects as a group create complex structures and complete sophisticated tasks at the same time as individual workers seem largely rudderless in their activity. These complex structures include not only nests, but also more transient creations such as the elaborately branching raiding columns of army and driver ants. Social insects considered as working collectives therefore pose a paradox. How can the evident complexity of their works be achieved by the group in the absence of central control or a high degree of individual competence?

The answer to this paradox—self-organization—was established by biologists in the 1970s and 1980s. Self-organization is the process (verified by mathematical simulations and even by experiments with teams of robots) by which relatively unthinking and autonomous units, following a few simple rules, can achieve complex tasks when they interact collectively. Other entities made of many interacting parts, such as brains and computers, probably act in a self-organized way in some respects. The price that self-organized systems pay for their way of operating is a level of error and inconsistency among their individual subunits. Indeed, the errors made by social insect workers in their task performance may even form the raw material from which they discover new ways of collectively achieving goals, in the way that genetic errors (mutations) provide the raw material from which natural selection can craft complex adaptations. This is why social insects may individually seem inefficient and even stupid, but the colony as a whole can achieve something that appears to be the result of planned, intelligent design. Working in the emerging field known as "swarm intelligence", which owes much to observations of the way social insects do things, computer scientists have applied the principles of self-organization to running several types of complex systems required by human economic activity. These include telephone networks, Internet search systems, factory production lines, and even route-planning for fleets of commercial trucks. What is lost by tolerating a level of individual error is more than made up for by the robustness, flexibility, and economic efficiency of the system as a whole. In this way, we continue to learn from social insects in ways never imagined by earlier observers of these enthralling creatures.

Resources

Books

Abe, T., D. E. Bignell, and M. Higashi, eds. Termites: Evolution, Sociality, Symbiosis, Ecology. Dordrecht, The Netherlands: Kluwer Academic Publishers, 2000.

Agosti, D., J. D. Majer, L. E. Alonso, and T. R. Schultz, eds. Ants: Standard Methods for Measuring and Monitoring Biodiversity. Washington, DC: Smithsonian Institution Press, 2000.

Bolton, B. A New General Catalogue of the Ants of the World. Cambridge, MA: Harvard University Press, 1995.

Bourke, A. F. G. "Sociality and Kin Selection in Insects." In Behavioural Ecology: An Evolutionary Approach, edited by J. R. Krebs and N. B. Davies. Oxford: Blackwell, 1997.

Bourke, A. F. G., and N. R. Franks. Social Evolution in Ants. Princeton: Princeton University Press, 1995.

Camazine, S., J.-L. Deneubourg, N. R. Franks, J. Sneyd, G. Theraulaz, and E. Bonabeau. Self-Organization in Biological Systems. Princeton: Princeton University Press, 2001.

Choe, J. C., and B. J. Crespi, eds. The Evolution of Social Behavior in Insects and Arachnids. Cambridge: Cambridge University Press, 1997.

Crozier, R. H., and P. Pamilo. Evolution of Social Insect Colonies. Oxford: Oxford University Press, 1996.

Hamilton, W. D. Narrow Roads of Gene Land. Vol. 1, The Evolution of Social Behaviour. Oxford: W. H. Freeman/Spektrum, 1996.

Hölldobler, B., and E. O. Wilson. The Ants. Berlin: Springer-Verlag, 1990.

Keller, L., ed. Queen Number and Sociality in Insects. Oxford: Oxford University Press, 1993.

Maynard Smith, J., and E. Szathmáry. The Major Transitions in Evolution. Oxford: W. H. Freeman, 1995.

Michener, C. D. The Bees of the World. Baltimore: Johns Hopkins University Press, 2000.

Ross, K. G., and R. W. Matthews, eds. The Social Biology of Wasps. Ithaca, NY: Comstock Publishing Associates, 1991.

Seeley, T. D. The Wisdom of the Hive. Cambridge: Harvard University Press, 1995.

Williams, D. F., ed. Exotic Ants: Biology, Impact, and Control of Introduced Species. Boulder, CO: Westview Press, 1994.

Wilson, E. O. Success and Dominance in Ecosystems: the Case of the Social Insects. Oldendorf/Luhe, Germany: Ecology Institute, 1990.

Periodicals

Allen-Wardell, G., et al. "The Potential Consequences of Pollinator Declines on the Conservation of Biodiversity and Stability of Food Crop Yields." Conservation Biology 12 (1998): 8–17.

Bonabeau, E., M. Dorigo, and G. Theraulaz. "Inspiration for Optimization from Social insect Behaviour." Nature 406 (2000): 39–42.

Chapman, R. E., and A. F. G. Bourke. "The Influence of Sociality on the Conservation Biology of Social Insects." Ecology Letters 4 (2001): 650–662.

Crespi, B. J., D. A. Carmean, L. A. Mound, M. Worobey, and D. Morris. "Phylogenetics of Social Behavior in Australian Gall-Forming Thrips: Evidence from Mitochondrial DNA Sequence, Adult Morphology and Behavior, and Gall Morphology." Molecular Phylogenetics and Evolution 9 (1998): 163–180.

Evans, J. D., and D. E. Wheeler. "Differential Gene Expression Between Developing Queens and Workers in the Honey Bee, Apis mellifera." Proceedings of the National Academy of Sciences 96 (1999): 5575–5580.

Giraud, T., J. S. Pedersen, and L. Keller. "Evolution of Supercolonies: the Argentine Ants of Southern Europe." Proceedings of the National Academy of Sciences 99 (2002): 6075–6079.

Hammond, R. L., M. W. Bruford, and A. F. G. Bourke. "Ant Workers Selfishly Bias Sex Ratios by Manipulating Female Development." Proceedings of the Royal Society of London, Series B. 269 (2002): 173–178.

Krieger, M. J. B., and K. G. Ross. "Identification of a Major Gene Regulating Complex Social Behavior." Science 295 (2002): 328–332.

Monnin, T., F. L. W. Ratnieks, G. R. Jones, and R. Beard. "Pretender Punishment Induced by Chemical Signalling in a Queenless Ant." Nature 419 (2002): 61–65.

Palmer, K., B. Oldroyd, P. Franck, and S. Hadisoesilo. "Very High Paternity Frequencies in Apis nigrocincta." Insectes Sociaux 48 (2001): 327–332.

Passera, L., S. Aron, E. L. Vargo, and L. Keller. "Queen Control of Sex Ratio in Fire Ants." Science 293 (2001): 1308–1310.