A novel hypothesis for the origin of the sexual division of labor in termites: which sex should be soldiers?

Sexual specialization and skewed sex ratios of the altruistic castes, especially soldiers, are common in many termite taxa. However, no theoretical or empirical studies have explained the origin of the sexual division of labor in termites. In most termite species, female alates are larger than male alates, and mature queens are much larger than kings, with females under consistent selection for high fertility. Therefore, females usually have the potential to be larger than males. Here, I present a novel preadaptation hypothesis that potential sexual differences in the suitability for the caste give rise to the sexual division of labor, and I provide the first evidence in support of this hypothesis in termites. Defense in Reticulitermes is typically performed by soldiers via mandibular and phragmotic defense in which soldiers with pluglike heads block openings, thus preventing enemies from invading the nest. Phragmotic defense requires that soldiers have heads wide enough to plug nest openings. Therefore, a size threshold for workers that develop into soldiers is a likely adaptation for effective defense. I show that sexual size dimorphism (SSD) and a size threshold for soldiers promote skewed sex ratios. A female-biased soldier sex ratio was observed in species with SSD, whereas there was no bias in soldier sex ratio in species without SSD. Thus, SSD and soldier sex ratio data from several Reticulitermes species support the preadaptation hypothesis.     

The ontogeny of soldiers in Prorhinotermes simplex (Isoptera, Rhinotermitidae)

The ontogeny of the soldier caste in Prorhinotermes simplex is studied with emphasis on the instar origin of soldiers, the duration of the presoldier stage, and changes in external anatomy during the soldier morphogenesis. The soldiers of different instar origin are compared morphometrically and by analysis of the soldier defensive secretion. The soldiers were found to develop from larvae of the second up to the eighth instar, via a short (13–17 days) presoldier stage. The early soldier instars were found exclusively in incipient colonies while the mature colony contains late instar soldiers only. The abrupt change in the external anatomy occurs in two steps. During the larva-presoldier moult, the head increases in length but only slightly in width. During the presoldier-soldier moult, both the length and width of the head increase markedly and the typical cordate shape is attained; the relative size of the pronotum increases considerably. The long falcate mandibles develop mainly during the larva-presoldier moult. One antennal segment is added during these two moults. With increasing instar age of soldiers a relative increase of the pronotum and the head size in its posterior region was observed. No functional differences in external anatomy were found among the six soldier instars. The composition of the frontal gland secretion is similar in the six soldier instars, an extraordinarily high amount of a defensive substance, (E)-1-nitropentadecene, was detected together with (Z,E)-α-farnesene, presumably an alarm substance. The strategy of soldier production is discussed. We conclude that the observed patterns of soldier production in P. simplex correspond well to its ecological strategy of an advanced single-site nester able to migrate and form foraging groups. Received 16 August 2005; revised 21 November and 6 February 2006; accepted 16 February 2006.     

Alarm Reaction of European Reticulitermes Termites to Soldier Head Capsule Volatiles (Isoptera, Rhinotermitidae)

The alarm behavior of worker and soldier castes of four European subterranean termite species of the genus Reticulitermes (R. santonensis, R. lucifugus, R. grassei, R. banyulensis) is described. In petri dish bioassays we presented filter papers with squashed soldier heads as an odor source to trigger a alarm reaction. Untreated filter papers were used as controls. The behavioral patterns were alike for all four species. In the control situation (no alarm), only occasional workers and few soldiers were seen walking in the petri dish. When the odor source was presented, a general state of alertness was observed, showing as higher activity and presence, zigzag running, antennation of nestmates, jittering and jerking, mandible snapping, head-banging, and attraction to the odor source. Workers were attracted within a few seconds (6.9–13.4 s); soldiers followed much later (18.5–64.6 s). Both workers and soldiers inspected the odor source only briefly (workers for 8.7–12.8 s, soldiers for 10.0–19.5 s), with the exception of R. santonensis, where soldiers stayed an average of 209.8 s. The number of individuals, of 50 workers and 3 soldiers, near the odor source averaged 6.0–9.2 workers and 1.2–2.3 soldiers. The presence of workers is crucial: without workers, a significant alarm reaction in soldiers could not be induced. The roles of soldier and worker castes of Reticulitermes species within alarm communication and the origin and nature of possible alarm signals are discussed.     

Social structure and the defensive role of soldiers in a eusocial bamboo aphid,Pseudoregma bambucicola (Homoptera: Aphididae): A test of the defence-optimization hypothesis

A complete understanding of the evolution of sociality in aphids requires a detailed knowledge of the patterns of soldier investment in their ecology. The eusocial bamboo aphidPseudoregma bambucicola has a morphologically specialized first-instar soldier caste. The proportion of soldiers was positively correlated with colony size. Within a colony, soldiers were evenly distributed among subcolonies; within each subcolony, however, their distribution was biased toward peripheries which were exposed to many predators. Field experiments introducing natural enemies such asEupeodes confrater (Diptera: Syrphidae) andSynonycha grandis (Coleoptera: Coccinellidae) revealed that the survival rate of these predators was negatively correlated with the density of soldiers, suggesting that soldiers can more or less defend their colonies by killing or removing a range of natural enemies. Observations suggest that large mature colonies attract more predators than newly established small colonies and that, within a colony, the predators attack each subcolony regardless of its position on bamboo shoots. This implies the presence of a positive correlation between colony size and predation risk. Thus, the investment in soldiers seems to reflect the attacking pattern of predators within a colony. These results agree with the defence-optimization hypothesis in soldier investment ofP. bambucicola colonies.     

Caste allometries in the soldier-producing aphidPseudoregma alexanderi (Hormaphididae: Aphidoidea)

Summary Colonies of the aphidPseudoregma alexanderi produce morphologically-specialized first-instar larvae, termed soldiers, that defend the colony from predators. The environmental cues and physiological mechanisms governing soldier production are currently unknown. Here we present a morphometric study of soldiers and normal first-instar larvae ofP. alexanderi. Several morphological features (fore-leg length and width, hind-leg length, and horn length) plotted against body length display relationship that are similar to a sigmoidal curve. We found further support for an earlier finding that soldiers fall into two size categories, majors and minors, although both types of soldiers appear to follow the same allometry. The patterns of allometry in the soldier-producing aphids are very different from those found in other social insects and do not easily fit into the traditional categorization of allometries. We present two simple alternative models of soldier development as a framework for guiding future studies of the mechanisms of soldier production.     

Inter- and intraspecific alarm response in the termiteMacrotermes subhyalinus (Rambur)

Summary In laboratory studies we analysed the reaction of aMacrotermes subhyalinus (Rambur) colony in contact with a small group of test termites placed at the head of a gallery system. As test termites we usedMacrotermes subhyalinus of the same nest,Macrotermes subhyalinus of a foreign nest and the related sympatric speciesMacrotermes bellicosus (Smeathman).Contact with an introduced group of foragers evokes a significant recruitment of workers and minor soldiers. The intensity of the response depends on the test groups we introduced. The highest recruitment results withM. bellicosus andM. subhyalinus from a foreign nest. During the recruitment the increase in the number of minor soldiers is greater than the increase in the number of workers, and as a result, the relative proportion of the soldiers in the population leaving the nest is significantly higher than before.Direct observations at the place of contact show thatM. bellicosus individuals are immediately attacked and killed. In the case ofM. subhyalinus from a foreign nest, not all individuals are killed; some are only intensively examined with the antennae or seized with the mandibles.M. subhyalinus from the same nest are rarely attacked. Mostly they are antennally inspected or enter the gallery system unhindered.The recruitment of termites from the nest is released by major workers which, after contact with the test termites, run back into the nest and provoke an alarm by tactile stimuli. On their way back into the nest the major workers, in all probability, lay a pheromone trail which leads the outgoing alarmed termites towards the place of contact.     

Interspecific Interactions Between Coptotermes formosanus and Reticulitermes flavipes (Isoptera: Rhinotermitidae) in Laboratory Bioassays

Experiments were conducted to examine competitive interactions between the Formosan subterranean termite, Coptotermes formosanus Shiraki (FST), and the eastern subterranean termite, Reticulitermes flavipes (Kollar) (EST), using groups of termites with different worker:soldier proportions. Experiments were conducted using three connected test chambers: an FST chamber, an unoccupied center chamber, and an EST chamber. When groups of FST were comprised of 20% soldiers versus 2% EST soldiers, only 8% of center chambers were occupied exclusively by EST. When groups of FST were comprised of 10% soldiers versus 1% EST soldiers, 44% of center chambers were occupied exclusively by EST. When the only food source was located in the center chamber, 60% of center chambers were occupied by both species. FST did not completely displace EST in any of these experiments.     

Buying time for colony mates: the anti-predatory function of soldiers in the eusocial aphid Ceratovacuna japonica (Homoptera, Hormaphidinae)

Eusocial aphids produce sterile individuals (soldiers) that specialize in colony protection. Killing predators is often considered the main function of soldiers. In this study, we tested the effect of harassment by soldiers of a eusocial aphid, Ceratovacuna japonica (Homoptera, Hormaphidinae), on predation by this species’ natural enemy, the larvae of Atkinsonia ignipicta (Lepidoptera, Stathmopodidae). We experimentally introduced some aphids and a predator to petri dishes and compared the survivorship of first-instar reproductives in the presence and absence of soldiers. We showed that soldiers can reduce the rate of predation on their colony mates without killing the predators. When predators encountered soldiers, they did not attempt to prey on them. Instead, they evaded them and often started to make a nest by spinning silken threads. The soldiers, in contrast, waved their forelegs and attacked the predator, and they sometimes succeeded in grasping the predator’s body. Because the predator used its mandibles to remove any soldier that succeeded in grasping its body, the soldier did not kill the predator. The reduction of predation was apparently caused by the delay of predation resulting from the harassment behavior of the soldiers. In eusocial aphids, a defensive strategy that delays predation may buy the soldiers’ colony mates time to reproduce or to escape from the predator.     

Agonistic behavior correlated with hydrocarbon phenotypes in dampwood termites,Zootermopsis (Isoptera: Termopsidae)

The dampwood termite genus ZootermopsisEmerson contains three recognized species with four distinct and consistent hydrocarbon phenotypes. Agonistic behaviors among nonreproductive insects from colonies of the same and different hydrocarbon phenotypes were observed in the laboratory. Various combinations of soldier versus nymphs, pseudergate versus pseudergate, and soldier versus soldier encounters were used in experimental trials. Soldiers or pseudergates seldom attack individuals of the same hydrocarbon phenotype. Z. angusticollis(Hagen) (phenotype II) is typically aggressive toward phenotype III of Z. nevadensis(Hagen) but not always aggressive against phenotype I of Z. nevadensis.The variation in response is dependent on which castes are placed in the bioassay arena: soldier versus soldier bouts result in consistent aggression, while pseudergate versus pseudergate or soldier versus nymphs contacts do not. Both pseudergates and soldiers of Z. laticeps(Banks) (phenotype IV) respond agonistically toward the other three phenotypes: Z. angusticollis (II) and Z. nevadensis(I and III). Although hydrocarbon phenotypes I and III, both Z. nevadensis,are morphologically indistinguishable, agonistic behavioral responses between phenotype I and phenotype III are not equivalent to I versus I or III versus III behavioral responses. The I versus III engagements, regardless of the castes involved, display a greater proportion of avoidance and aggressive responses than I or III intraphenotype encounters. We interpret the lack of avoidance or aggressive behavior within each of the two phenotypes of Z. nevadensisand the significant avoidance and aggressive behavior between phenotypes as definite evidence of discrimination between disparate hydrocarbon phenotypes. These agonistic bioassays along with data on distinct hydrocarbon patterns and geographic distributions serve as the basis for creating two subspecies of Z. nevadensis: Z. n. nevadensis(Hagen) and Z. n. nuttingiHaverty and Thorne, ssp. nov.     

Factors affecting the proportion of sterile soldiers in growing aphid colonies

The proportion of sterile soldiers in an aphid colony is positively correlated with colony size. Assuming logistic growth of the aphid colony, Aoki and Kurosu (Insect Soc 50:256–261, 2003) presented an inequality that determines, for any colony size, whether a soldier or a reproductive will be added to the colony. To put it in words, if the marginal defensive efficacy of a soldier, multiplied by the number of reproductives, is larger than the mean productivity of reproductives without defense by that soldier, the soldier will be produced; if not, a reproductive will be produced. Based on Aoki and Kurosus inequality, we carried out simulations to determine whether the proportion of soldiers increased with colony size. Given a constant level of depredation per aphid and a constant number of predators, proportion of soldiers continued to increase with colony size unless a single soldier was very effective or unless carrying capacity was very large. Given a constant number of nongluttonous predators and a decreasing level of depredation per aphid, proportion of soldiers soon began to decrease after a peak. However, given an increasing number of nongluttonous predators to keep a constant level of depredation per aphid, proportion of soldiers again continued to increase. These results confirmed the argument that the proportion of soldiers can increase with colony size under a wide range of realistic assumptions.     

Fire ant predation on native and introduced subterranean termites in the laboratory: effect of high soldier number in Coptotermes formosanus

Abstract.

• 1 Native Reticulitermes spp. Holmgren and the introduced Coptotermes formosanus Shiraki (Isoptera: Rhinotermitidae) were exposed to predation in the laboratory by Solenopsis invicta Buren (Hymenoptera: Formicidae), a known predator of Reticulitermes.
• 2 In one experiment, both taxa contained the 2.4% soldiers commonly found in Reticulitermes. In a second experiment, Reticulitermes contained 2.4% and C.formosanus a more normal 18.3% soldiers.
• 3 When soldier proportions were equal, C.formosanus soldiers suffered less of a decrease in survivorship due to predation compared with Reticulitermes soldiers. Little difference was found between taxa in the effect of predation on termite workers or in the effect of termites on predator survivorship.
• 4 When soldier proportions were unequal, both C.formosanus castes suffered less from predation than those of Reticulitermes. Again, little difference was found in the effect of termites on ants.
• 5 C.formosanus showed less movement into new containers when soldier number was abnormally low, compared with both taxa with normal caste proportions.

     

Castes in humivorous and litter-dwelling neotropical nasute termites (Isoptera,Termitidae)

Summary The developmental pathways of the neuter castes were studied in three species of Nasutitermitinae from central Panama. The humivorousSubulitermes denisae andCoatitermes clevelandi display several primitive traits: absence of sex dimorphism, representation of both sexes among workers and soldiers, and occurrence of successive worker instars. The litter-dwellingVelocitermes barrocoloradensis has a more complex caste system: female larvae are larger than males and give rise to the large workers, which constitute the bulk of the work force; male larvae proceed to soldiers through a small worker or a special larval instar. The resulting soldier caste is polymorphic. These results support previously formulated hypotheses regarding a link between humivorous diet and reduced polymorphism on the one hand, and between forest-floor foraging and large continuous size variation among soldiers on the other. Whereas the caste systems ofSubulitermes andCoatitermes probably represent a primitive condition,Velocitermes shares derived traits withNasutitermes and the other fully nasute genera previously studied. I therefore hypothesize that ancestors with these advanced features may have spread from the neotropics and be at the origin of most nasute genera, including humivorous taxa, present in other regions.     

Behavioral role differentiation in the primitively eusocial waspBelonogaster juncea juncea (Hymenoptera: Vespidae)

With a view toward describing behavioral variability among individuals of the primitively eusocial speciesBelonogaster juncea juncea, we recorded the time-activity budget spent on five behavioral categories (foraging, building, feeding, inactivity, and reproduction) by 52 individuals belonging to four postemergence colonies. A principal-components analysis coupled with a hierarchical cluster analysis enabled us to discern four behavioral roles. The reproductive role is reserved to one individual (functional monogyny) and the workers’ role is differentiated into foragers, builders, and guards. The foragers are females that spend the most time (82.6% of their time) foraging, supplying the nest with prey load and liquid matter. The builders are individuals (with 41.5% of their time off the nest) that tend to bring pulp into the nest and then undertake building activities. The guards are those females that spend the most time (79.7% of their time) being inactive on the nest.     

Egg care by termite soldiers

The handling and transportation of eggs by soldiers of Prorhinotermes simplex after a nest disturbance is described. The soldiers, apart from performing alarm and defensive activities, tried to handle the eggs on the open surface. They often succeeded in lifting an egg or even a group of eggs, and then searched for a nest entrance. They either entered the nest to come out without the egg(s) within a while, or they passed the egg(s) onto a pseudergate waiting in the opening. As a result, all eggs were rapidly evacuated. Laboratory experimentation showed that pseudergates and nymphs try to deposit the eggs in a dark shelter, whereas the soldiers tend to transfer the eggs to nestmates and drop them after a while if no other termite accepts them. Received 25 February 2005; revised 26 April 2005; accepted 3 May 2005.     

The influence of juvenile hormone analogue on soldier differentiation in the higher termite,Macrotermes michaelseni*

ABSTRACT. The influence of topically-applied or vapour-administered juvenile hormone analogue (ZR-515, methoprene (JHA)), on soldier differentiation in a higher termite species, Macrotermes michaelseni (Sjostedt) (Isoptera; Macro-termitinae) was examined. It was found that even male third larvae, which under natural conditions develop exclusively into major workers, could be induced by JHA to differentiate into presoldiers. Intermediate forms between workers and soldiers were produced from some of the treated individuals. Administered in vapour form, JHA induced a higher proportion of presoldiers and a lower proportion of intermediate forms than it did when administered topically. The rate of presoldier formation was found to be JHA-dose-dependent. Biometric analysis showed that male presoldiers have the largest head capsule size, but the smallest mandibular index. The possible mechanism of soldier induction is discussed.     

Reproductive schedule and factors affecting soldier production in the eusocial bamboo aphid Pseudoregma bambucicola (Homoptera, Aphididae)

Summary: The reproductive characteristics of the soldier-producing aphid Pseudoregma bambucicola were studied in Kagoshima, Southern Japan, to know the factors affecting soldier production of eusocial aphids. The soldier proportion in aphid colonies was highest from October to November. In some large colonies, soldiers were observed in all seasons except in July when colony size was relatively small. Multiple regression analysis showed that the colony size was a principal factor affecting soldier proportion throughout a year. Other social or environmental factors such as aphid composition, host plant conditions and predator abundance were not always significant. Rearing experiments revealed that large colonies (̿,000 individuals) produced soldiers in almost all seasons while small colonies (<1,000) never produced any soldiers. The caste-production schedule of adult females was examined in the field. When solitary females produced both castes, they usually produced normal nymphs first and then soldiers. Females from large colonies tended to produce more soldiers in the earlier period of their lifetime, whereas females from newly established small colonies produced no or only a few soldiers at later times. The average number of soldiers and normal nymphs produced consecutively by a single female was >10 and >20, respectively. Because they have a small number of ovarioles (<15 on average), females should alter caste production within the same ovarioles according to changes in environmental conditions. Artificial removal or introduction of predators and reduction of colony size did not affect soldier production over two successive generations, revealing maternal effects on soldier production. Females cannot shift caste production quickly in response to changes in predator abundance and colony size. This is probably due to early developmental determination of castes within the mother's body.     

Studies on feeding and trophallaxis in the subterranean termite Odontotermes formosanus using rubidium chloride

Differences in feeding and trophallaxis among castes of the fungus‐cultivating higher subterranean termite Odontotermes formosanus Shiraki (Isoptera: Termitidae) were determined using rubidium (Rb) as a tracer. In the feeding study, workers and third instars fed directly on Rb‐treated filter paper and acquired the highest levels of Rb content. Although fifth and sixth instars obtained the Rb content over the marked level, the feeding ability of the caste was very poor. However, soldiers did not feed on Rb‐treated filter paper. In the trophallaxis study, worker, soldier, and fifth‐ and sixth‐instar recipients paired with Rb‐fed worker donors had a significantly higher Rb content than those paired with control donors. Results indicate that soldiers are completely dependent on workers for nutrition. Workers and fifth and sixth instars may obtain nutritional material from foraging workers. When third‐instar recipients were paired with Rb‐fed worker donors, they did not have a significantly higher Rb content than when they were paired with control donors, suggesting that the frequency of worker/third instar trophallaxis was very low. Transfer efficiency from the four trophallactic combinations ranged from 1.2 (worker donors to third‐instar recipients) to 12.9% (worker donors to soldier recipients). Accordingly, nutritional material from foraging workers could be transferred to non‐foraging workers, soldiers, and fifth and sixth instars by trophallaxis; however, third instars rarely obtain nutritional material from foraging workers, because they are capable of feeding within the nest.     

Physical contact between soldier and worker is essential in soldier self-regulation of Coptotermes formosanus (Isoptera, Rhinotermitidae)

In a termite society, the soldier proportion is regulated at a species-specific level with a seasonal fluctuation. A feedback mechanism in soldier proportion regulation is well recognized, while the proximate means by which the presence of soldiers inhibits the new soldier formation from workers remains mostly elusive. In the present study, some possible means for this inhibition were tested with the Formosan subterranean termite, Coptotermes formosanus Shiraki. The results showed that neither soldier-originated volatiles nor non-volatile chemicals without physical contact were responsible for the inhibition of new soldier formation. The essentiality of physical contact between workers and soldiers in soldier regulation was established. However, by physical contacts, as the soldier-replacement experiment showed, workers could not individually recognize or “count” soldiers to maintain a set soldier proportion. The possible feedback mechanisms for soldier regulation are discussed.     

Agonistic Behavior of the Termite <Emphasis Type="Italic">Prorhinotermes canalifrons</Emphasis> (Isoptera: Rhinotermitidae)

While intercolonial and interspecific agonism has been broadly documented in termites, detailed studies of individual behavior and its differentiation among castes are scarce. We studied the responses of non-reproductive castes of Prorhinotermes canalifrons (pseudergates, presoldiers, soldiers of different age and artificial intercastes) to conspecific and heterospecific aliens. Each caste displayed a specific behavioral repertoire. Soldiers, but to a lesser extent pseudergates as well, are responsible for colony defense. The behavioral repertoire of the soldier appears on the second day after exuviation and becomes complete on the third day, well before its defensive secretion is present. The behavior of artificial intercastes induced by a JH analog is intermediate between that of pseudergates and soldiers. Reactions to heterospecific aliens were much quicker and more violent than reactions to conspecifics, which suggests that individuals experience difficulties in identifying conspecific aliens.     

Impact of a juvenile hormone analogue on the anatomy and the frontal gland secretion of Prorhinotermes simplex (Isoptera: Rhinotermitidae)

In termites, juvenile hormone plays a key role in soldier differentiation. To better understand the evolutionary origin of the soldiers, we studied the external and inner morphology of pseudergate-soldier intercastes and neotenic-soldier intercastes formed artificially by the application of juvenile hormone analogue in Prorhinotermes simplex. A majority of these intercastes had a soldier phenotype, whereas the inner anatomy had an intermediary form between two castes or a form specific to intercastes. Our experiments showed that traits of neotenics and soldiers can be shared by the same individuals, although such individuals do not exist naturally in P. simplex, and they have not been reported in other species but in some Termopsidae. Our results reinforce the hypothesis that soldiers may have emerged from soldier neotenics during the evolution of termites.