Caste
The term “caste” refers to long-term, stable differences among insect colony members that affect the roles played by individuals in their social group. It was the existence of distinct morphological castes in insect colonies that led Charles Darwin to identify social insects as a major challenge to his theory of evolution by natural selection. Few topics are more central to the study of social insect biology than caste. It is ironic, then, that few topics have generated more controversy and debate among social insect biologists. Beyond the deceptively simple definition just offered, there is little agreement on how the term should be defined, or on what characteristics should be used to identify the castes of individual insects. In part, the controversy exists because published definitions of caste are often not operational. For example, some definitions do not specify whether caste differences must be developmentally fixed and permanent or, alternatively, whether individuals can exhibit caste flexibility. As a result, “caste” has been applied to a wide array of physiological and behavioral phenomena.
The diversity of caste systems between and within evolutionary lineages of social insects may preclude a simple, universal definition of caste. Rather than advocating a single definitional point of view, this article explores the diversity of ways in which the concept has been used and the array of important phenomena caste encompasses in different insect societies.
REPRODUCTIVE CASTES
Differences in Reproductive Function
A fully social or eusocial group is generally understood to exhibit reproductive division of labor. This means that eusocial groups must include some individuals that forgo direct reproduction and instead aid the rearing of the offspring of others in their group. In eusocial insects, the helpers comprise the worker caste and reproductive females are referred to as queens. Termite colonies possess long-lived royal couples (a queen and a king), whereas in eusocial Hymenoptera, males are sometimes referred to as drones. Males in the order Hymenoptera (bees, ants, and wasps) rarely work for their colonies and typically die soon after mating. In contrast, male euso-cial thrips (Thysanoptera) and termites (Isoptera) comprise part of the worker force and participate fully in colony labor.
Social insect species vary according to whether the group’s members are permanently relegated to reproductive versus worker roles and in the degree of fecundity differences between reproducers and workers. There is a general evolutionary trend toward increased reproductive caste specialization as more complex, larger societies evolve from simpler, smaller ones. In some ants, workers lack reproductive organs and are permanently sterile. In most species, however, workers can achieve limited direct reproduction under some conditions.
Morphological Differences
Some species are reproductively monomorphic, and reproductives do not differ significantly in body structure from workers. Many sweat bees and bumble bees, some paper wasps, and even some primitive ants are examples of reproductively monomorphic species. Workers in monomorphic species are often smaller than reproductives, but there can be considerable overlap in body size distributions among the reproductive castes. In some cases, clear physiological differences distinguish workers from reproductives when morphology does not. For example, temperate Polistes paper wasp colonies produce gynes (potential female reproductives) at the end of the summer. Gynes possess enlarged, nutrient-laden fat bodies, not present in female workers, which permit them to overwinter in a quiescent state.
In contrast, consistent reproductive caste differences in body size and shape have evolved in several lineages of social insects. Most eusocial insects with wingless workers, such as ants and termites, retain a morphologically distinct reproductive caste with wings. In species with flying workers, developmental allometry can still result in the production of distinct, nonoverlapping body forms for repro-ductives and workers. Morphologically discrete reproductive castes are found among honey bees, stingless bees, and some paper wasps. Reproductives are often not only larger than workers but also differ in body proportions (hence shape) in ways that suggest specialization in egg laying, such as relatively enlarged abdomens. The degree of morphological differentiation between reproductive castes probably evolves in response to a complex array of natural selection pressures. For example, the degree to which the colony occupies a defensible, long-lasting nest site may in part determine whether the queens can afford to adopt relatively immobile body forms.
Caste Determination: Immature Development and Adult Interactions
Other than an interesting exception in the ant Harpagoxenus sub-laevis, there are no well-documented cases of genetic differences that affect reproductive caste differentiation. Often caste differentiation must depend in part on differential patterns of gene expression during development, particularly in species with distinct caste morphology.
Differences in environmental conditions during immature development can have strong effects on an individual’s caste. Nutritional effects on reproductive caste have been documented in numerous taxa and appear to be widespread, if not universal, among eusocial insects. Differences in the amount of food provided to larvae may underlie many of the differences between reproductives and workers, especially in species exhibiting the common pattern of larger body size for repro-ductives. However, differences in food quality, possibly including the addition of glandular secretions and pheromones, cannot be ruled out. Especially interesting in this regard are those eusocial wasps whose reproductives are smaller than workers (genus Apoica) or identical in size but different in shape (genus Pseudopolybia).
Social interactions among adults may also influence reproductive caste, particularly in species without apparent morphological caste differences. For example, dominance interactions among paper wasp (Polistes) females, which often cooperate to start new colonies, determine which female acts as the sole reproductive. Subordinate Polistes females function as workers.
CASTES IN THE WORKER FORCE
Morphological Castes
DISCRETE WORKER MORPHOLOGY
In all termites and in approximately 10% of ant species, workers exhibit developmental allometry resulting in body shape variation within the worker caste. Interestingly, this type of morphological caste has not been documented in social insects with flying workers, such as bees and wasps. There is typically some association between a worker’s body form and the tasks that she performs. One of the most common types of morphological specialization is the assignment of large workers, called soldiers, to the special role of colony defense. When the colony is threatened by an animal, the soldiers advance and attack, while other workers flee. Often the soldiers uniquely possess heavily armored exoskeletons and some type of weaponry, including enlarged muscular heads, long, piercing mouthparts, or glands that produce defensive chemicals. In other cases, worker’s body shape variation affects the performance of more mundane tasks such as food collection. In army ants (Eciton spp.), longer-legged workers select larger food items to carry back to their colonies. In leafcutter ants (Atta spp.), the largest workers are soldiers, the medium-sized workers cut and transport leaves, and the smaller workers usually remain in the nest to tend the colony’s fungus garden. An ant worker’s body size and shape are fixed upon adult emergence; further growth is not possible. In contrast, some termite workers (Zootermopsis spp.) exhibit considerable caste plasticity, potentially molting among different body forms, and even switching from soldier to nonwinged reproductive castes under certain conditions.
BODY SIZE EFFECTS
Even in monomorphic species, body size differences can influence the tasks that workers perform. In some species, larger-bodied workers dominate their smaller nest-mates (Polistes metricus, P. fuscatus, and P. dominulus), and dominance status in turn affects the tasks a worker performs. In some bumble bees (Bombus spp.), however, larger workers are more likely to perform certain tasks such as foraging to collect food for the colony, independently of obvious worker aggression.
Behavioral Castes
Workers can be assigned to behavioral castes when they specialize on a subset of the tasks that the colony needs. In some eusocial insect species such as Bombus and Polistes, workers exhibit a great deal of flexibility, switching among tasks often, and behavioral castes are weakly defined. In honey bees (Apis mellifera) and swarm-founding paper wasps (Polybia spp.), on the other hand, workers specialize more consistently.
AGE OR TEMPORAL POLYETHISM
Changes in task specialization as workers age are among the best-studied factors that influence workers’ behavioral caste. “Age” or “temporal polyethism” refers to an ordered, predictable sequence of task specializationsthrough which an adult worker passes as it ages. Typically, species with temporal polyethism exhibit centripetal development: workers begin by working deep inside the nest, close to the queen(s) and brood; they later perform tasks at the nest periphery; and they finally move further out to perform risky tasks such as foraging and nest defense (Fig. 1). This centripetal pattern of development is remarkably similar among the diverse eusocial insect species that exhibit well-developed temporal polyethism. Workers usually follow the same sequence of task specializations, but individuals vary in their rate of passage through the sequence. Changes in hormone titers, such as juvenile hormone, have been implicated in determining the rate of temporal polyethism in Apis and Polybia.
FIGURE 1 Temporal or age polyethism in the paper wasp Polybia aequatorialis. Data were collected on 130 individually marked, known-age workers. Two measures of worker activity at three task sets are plotted against worker age (solid lines: mean rate of task performance; dashed lines: number of workers performing the task). Note the typical centripetal developmental sequence: in-nest tasks (mostly nest cleaning) are followed by building on the nest exterior, and later by foraging (leaving the nest and returning with food and building materials).
Workers’ relative age influences social status and task performance in some species. In the paper wasps P. exclamans. the first-emerging (and consequently the oldest) workers in the nest tend to socially dominate their nestmates, a pattern referred to as gerontocracy, which is independent of body size variation. In this case, age influences workers’ behavioral caste in a static way, rather than in a dynamic way as in temporal polyethism.
INDIVIDUAL DIFFERENCES AND SPECIALIZATION
Superimposed on broader patterns of division of labor, such as body size or age effects, workers sometimes exhibit idiosyncratic specialization on tasks. For example, Apis and Polybia foragers often specialize by collecting one of the several materials their colony needs to function. Such specialization may benefit the colony by increasing the efficiency or reliability of task performance.
Genetic Effects Genetic relatedness among the offspring in the worker force is highest when the workers are born to a single reproductive female, which has mated with a single male. Some social insect species exhibit mating behavior or social structure that decreases the genetic relatedness among the offspring workers within colonies. When queens mate with several different, unrelated males (polyandry), or when several reproductive females are present in the colony (polygyny), workers can find themselves sharing a nest with a combination of more closely and more distantly related individuals. In a number of polyandrous and polygynous species, including Apis spp. and stingless bees (genus Partamona), several species of ants, and Polybia spp., workers that are more genetically similar have been found to specialize on similar tasks.
Experience and Learning A predicted benefit of task specialization is that workers can improve performance as they gain experience. There is evidence that some insect workers learn to perform tasks more effectively with experience. Bumble bees collecting nectar and pollen from complex flowers learn to do so more rapidly after repeated attempts to handle a given type of flower. Polybia foragers are less likely to return from foraging trips empty-handed as they gain foraging experience.
COLONY-LEVEL INVESTMENT
Investment in Growth and Maintenance versus Reproduction
One of the major challenges that faces growing organisms is the developmental decision of how many resources to invest in growth and how many to invest in reproduction. Insect colonies can be treated as organisms in this sense, since each colony must decide how much it will invest in different castes (i.e., in workers vs reproduc-tives). To the extent that colonies are reproductive units, optimality theory predicts that natural selection will favor colonies that allocate their limited resources efficiently into different castes. Many insect societies segregate the production of workers (early in colony development) from the production of new queens and males (later in colony development).
Worker Caste Ratios
Insect colonies appear to behave in an adaptive manner by adjusting their worker caste ratios to meet current colony needs. Production of different worker castes reflects a trade-off between the costs and benefits of producing and maintaining workers of different kinds. As ant colonies with morphologically specialized workers grow in size, their amount of investment in large-bodied workers increases, and many eusocial insects produce tiny nanitic workers early in colony development. Colonies of the ant Pheidole pallidula increase their rate of production of soldiers when exposed to potential competitors. Similar colony flexibility is apparent in age-caste distributions. In honey bees and paper wasps, if the level-of-colony need for foragers changes, some workers accelerate or reverse their behavioral development, performing the age-atypical tasks that are in greatest demand. Identifying the mechanisms that link individuals’ developmental plasticity with the level-of-colony need remains as a central challenge in the study of caste.
