The Alarm Pheromones of Social Insects: A Review
By Casandra Lloyd
clloyd@lamar.colostate.edu
Pheromones are utilized by many different animals for communications such as aggregation, mating, food location, kin recognition and warning just to name a few. Alarm pheromones in particular have the simplest structure of all, but are possibly the most poorly understood. They have low molecular weights, are highly volatile and appear to be the least specific of all pheromones. This evolutionary development allows social species to protect their colony when in danger. A few alarm pheromones are species specific while most others are not. This can be seen as an advantage to social species which would be able to “read” the signals of other species that are a threat to them. There are also certain predators that have evolved ways to detect alarm pheromones to locate their hosts. Some animals that have been found to communicate through alarm substances are sea anemone, rats, skunks, hyenas, shrews, deer, schooling fish, tadpoles, mites, aphids and the social insects which include ants, bees, termites and wasps.
In the social insects, alarm pheromones are critical in detecting and escaping dangerous predators and protecting the colony. These chemicals are usually in the form of terpenoids or aliphatic ketones and esters, and can be released from the mandibles, anal glands or stinging apparatuses of insects. The two primary alarm behaviors that a colony can exhibit are to disperse or to move towards the source of alarm in an aggressive manner to attack the intruder. Large amounts of the chemical are necessary to initiate attack and in some species the chemicals are sprayed onto the enemy to tag him as an aggressor.
In many ways the alarm pheromones are a hard class to define. In a bioassay it is difficult to dissociate other stimuli and thus alarm is often confused with defense, recruitment or attraction. The issue of quantifying these types of reactions, although some work has been done to standardize this. The purpose of this article is to review alarm pheromones in social insects over the past 35 years, discuss current research and examine methods for insect control using alarm pheromones.
Alarm pheromones were discovered in the early 1960’s and since then have been reported to be found in a wide array of animals: sea anemone, rats, skunks, hyenas, shrews, deer, schooling fish, tadpoles, mites, aphids and the social insects which include ants, termites, bees and wasps. This review will focus on the social insects, in which alarm pheromones are generally widespread. Recognizing enemies and colony defense is crucial to the survival and future fitness of social insects, particularly those that have numerous predators and are under selective pressure (Dicke and Grostal, 2001). The study of alarm pheromones has recently peaked human interest for its use in developing control methods for insect pests.
Although alarm pheromones have been known for over 40 years, they are probably the least understood of all pheromones. The alarm behavior has been difficult to define but can be generalized into "flight" or "fight" responses. Some characteristic behaviors of alarm are raised heads or antennae, bursts of running, movement to the source of the pheromone, biting and stinging (Hughes et al., 2001). More alarm behaviors can be seen in Table 1. Factors such as pheromone concentration, length of exposure, and context also play a crucial role in the reaction to alarm pheromones. Alarm is often confused with attraction, recruitment and defense and can be hard to distinguish in a bioassay. The function of alarm pheromones is to warn colony members of an enemy, not to ward off intruders (Vander Meer et al., 1998). However, there are some natural enemies that have evolved ways to exploit these signals to find their hosts (Feener et al., 1996 and Allan et al., 1996). Some insects, especially ants and stinging wasps, can “tag” an enemy with the alarm pheromone so that others know it is an aggressor (Harborne, 1993).
Unlike other classes of pheromones, the alarm pheromones are produced in fairly large abundance and appear to be the least specific of all pheromones (Birch, 1974). Some of the compounds are even found in multiple subfamilies and can cause alarm behavior in other species. These chemicals volatize rapidly and usually have a boiling point between 151-263°C. They typically have a molecular weight of less than 200 and are in the C5-C10 range (Vander Meer et al., 1998). The alarm chemicals are frequently in the form of terpenoids, ketones and aldehydes and can be released from the mandibular gland, anal gland, poison gland, Dufour’s gland or frontal gland (Blum, 1969). Little is known about the biosynthesis of terpenes in insects, but most of them seem to be natural products synthesized by the insect as opposed to sequestered compounds them from plant material (Prestwich, 1979).
Of the 11,000 known ant species, all are eusocial and therefore rely heavily on pheromone communication within the colony (Vander Meer et al., 1998). The glandular source of ant alarm pheromones is usually the mandibular gland but has also been found in the Dufour’s gland and anal glands (Birch, 1974). Table 2 shows some of the known alarm substances from worker ants as well as their glandular source and the effect they have on conspecifics. Table 3 shows some of the typical compounds found in ant alarm pheromones. Despite the common use of alarm pheromones in ants, the responses can differ drastically among species (E.O. Wilson, 1971). Many ants have stingers that can produce chemical defenses, but those that do not, such as the Formicinae and Dolichoderinae, are essentially houses of chemical warfare. For example, species of Formica spray powerful formic acid along with their Dufour’s gland secretions which acts well as a chemical defense (Evans and Schmidt, 1990).
One of the simplest examples of alarm communication is that of Acanthomyops colonies. These are subterranean ants that are large in size and dense in number and therefore it could be speculated that they would respond to alarm pheromones by the “fight” reaction because of the unlikeliness of dispersal. This subfamily releases alarm pheromones from its mandibles as well as the Dufour’s gland and the alarm substance has been found to be composed of terpenes, hydrocarbons and ketones. Workers that are nearby react the fastest and those that are farther away take longer to react. The general response is raised antennae followed by opened mandibles and then running in the direction of the pheromone source. If more alarm pheromones are not released, the signal dies out within a few minutes (E.O. Wilson, 1971).
One of the related ants, Lasius alienus, is smaller and nests under rocks or wood which allows for fast dispersal when disturbed. The principle component of their alarm substance is the same as that of Acanthomyops but instead of running to the source, they frantically run around in no particular direction. They also react to lower concentrations of alarm pheromones than Acanthomyops (E.O. Wilson, 1971).
A more complicated response is found in the harvesting ant, Pogonomyrmex badius, whose active alarm pheromone is 4-methyl-3-heptanone. When the concentration of the pheromone is around 1010 molecules/cm3 it acts only as an attractant, but when the concentration is greater than that, the ants become aggressive and frenzied (E.O. Wilson, 1971). 4-methyl-3-heptanone is also known as a repellant to the predacious fire ant, Solenopsis invicta (Evans and Schmidt, 1990).
The fire ant, Solenopsis saevissima, is known to be aggressive and also has a complex alarm system. When in combat the worker ant releases two substances: one from its head that acts as an alarm pheromone and another from the Dufour’s gland that acts as a trail pheromone to orient other workers to the enemy (E.O. Wilson, 1971).
A paper by Charles Kugler (1979) evaluated the function of the pygidial gland in the ant Pheidole biconstricta, in the subfamily Myrmicinae, and was the first to report that it is a source of alarm pheromone. The pygidial gland is similar to the anal gland and has been found to have three purposes. First, the volatile component acts as a repellant against other ant species and the larger the pygidial gland, the more repellant the ant is. Second, the viscous component acts as a second line of defense against attackers by being applied directly to the enemy and causing gumming and irritation. Lastly the pygidial gland secretes an alarm pheromone that increases running and aggressive behaviors in P. biconstricta. Traniello and Jayasuriya (1981) also found alarm pheromones in the pygidial glands of a rare and primitive ant species, Aneuretus simoni, found only in the tropical rain forests of Sri Lanka. Worker ants of this species show varying degrees of alarm behavior to pygidial gland secretions but no strong responses to mandibular, poison, or Dufour’s glands.
Leaf-cutting ants release an alarm pheromone from their mandibular glands and respond aggressively to them. Many studies have focused on determining the chemical components of this alarm pheromone and mixed results have been found. Hughes et al. (2001) examined two grass-cutting ant species, Atta bisphaerica and Atta capiguara, to see if 4-methyl-3-heptanone, the concluded active alarm releaser, stimulates the same alarm response alone as do crushed ant heads. The results did indeed find that 4-methyl-3-heptanone is the compound that elicits the most alarm behavior from the ants and appears to be a common alarm pheromone compound among ants in general. Nonetheless, the crushed ant heads stimulated greater levels of alarm behavior and this suggests that one or more of the minor compounds may act synergistically with the main compound. This example proves the complexity of chemicals released by ants in nature.
A recent study evaluated three different Crematogaster ant species that live symbiotically with the acacia tree. These ants compete for host trees and all of them produce potent alarm pheromones from their mandibles. Interestingly, when ants are disturbed on one tree, alarm behavior is observed in ants of the same or different species on close-by trees. The study discovered from GS-MS analysis that the alarm pheromone chemical composition is different for each species and is also more complex than any other Crematogaster species reported thus far. In those Crematogaster species that share certain compounds, the concentrations differ dramatically and may be used in species discrimination (Wood et al., 2002).
Another fascinating aspect of alarm pheromones is their exploitation by natural enemies to locate hosts. Communication between conspecifics is used by a variety of predators for foraging and is therefore acting as a pheromone as well as a kairomone. One paper by Allen et al. (1996) found that the cursorial spider Habronestes bradleyi, a specialist predator of the meat ant Iridomyrmex purpureus, is able to cue into the alarm pheromone of the ant. 6-methyl-5-hepten-2-one is the alarm pheromone component released by alarmed and injured ants and was shown to attract these spiders. Another paper was released the same year by Feener et al. (1996) demonstrating that the parasitoid, Apocephalus paraponerae, is able to detect alarm pheromones released from the mandibular glands of the giant tropical ant, Paraponera clavata. It has long been known and is not surprising that some parasitoids have evolved the ability to utilize chemical cues from their hosts. The two alarm pheromone products that A. paraponerae are attracted to were identified as 4-methyl-3-heptanone and 4-methyl-3-heptanol. The male parasitoids are attracted to these chemicals for feeding and location of female mates whereas the female parasitoids seek out these chemicals in order to lay their eggs in the injured worker ants and feed from their wounds. Even though their alarm communication may attract predators, the cost may be small when ants occur in large numbers and the selection for modification of this trait is probably weak.
Alarm pheromones in termites were first discovered over 40 years ago by Ernst, yet research in this area is lagging far behind that of other social insects. Termite alarm pheromones are often in the form of monoterpenes and are only produced by soldiers. The monoterpenes also function as toxicants and irritants to enemies that come in contact with them, but their main function is to invoke alarm behavior. The soldiers secrete the pheromone with a defensive substance from their frontal glands and those soldiers that lack highly evolved frontal glands, such as Microcerotermes turneri or Gnathamitermes perplexus, most likely do not produce alarm secretions (Prestwich, 1979). One report found that Hodotermes mossambicus excretes an alarm pheromone in its fecal material, but this is the only report of an alarm pheromone source other than the frontal gland in termites (Vander Meer et al., 1998).
Some termites produce two enantiomers of a monoterpene and only one of the enantiomers is able to induce alarm. For example, the alarm pheromone of Nasutitermes princeps is a-Pinene but only the (+)-enantiomer causes an alarm behavior. On the other hand, Nasutitermes costalis only responds to the (-)-enantiomer of its alarm pheromone, carene. Furthermore, workers and soldiers respond in different ways to the same alarm substance (Vander Meer et al., 1998).
Occasionally the chemicals of a termite soldiers alarm substance can be different from colony to colony. It is usually the diterpenes that vary within species and the monoterpenes are typically constant. It should be noted that not all chemically defended termites produce alarm pheromones (Vander Meer et al., 1998). Table 4 lists some of the alarm pheromones produced by termite soldiers.
Honey bees (Apis spp.) are perhaps the most studied insects for social behavior of all the social insects. It has been known among beekeepers for many years that once you have been stung by a honey bee, others will attack shortly after. Beekeepers also know that smoke can mask the bees alarm pheromone and prevent the beekeeper from being noticed. Charles Butler (1609) was the first to recognize that the attack signal was a chemical released from the stinger of the bee (E.O. Wilson, 1971). The alarm pheromone is only found in worker bees and only when they have reached guarding behavior (Vander Meer et al., 1998). However, it is interesting to note that individual foraging workers do not respond to alarm pheromones and that as the group size increases, the reaction of individual bees increases logarithmically (Moritz and Burgin, 1987). The alarm pheromone is located in the membranes surrounding the stinger but the bee does not have to sting the enemy in order to release the pheromone. The first response after perceiving a potential enemy is for the bee to open its sting chamber and release the alarm pheromone to recruit other bees from the colony. If the intruder is a big enough threat, the bee will sting its enemy. Typically the barbed stinger is detached from the bee along with the poison gland and Dufour’s gland that are still attached to the stinger. There have been many studies to determine factors that elicit defense in honey bees (Vander Meer et al., 1998).
The first component of the honey bee alarm pheromone was identified in 1962 as isopentyl acetate. It was later determined that this was only one component of the alarm substance. Esters and ketones of certain chain lengths seem to be effective as alarm pheromones in Apis mellifera whereas hydrocarbons, aldehydes, acids and alcohols are not (Boch and Shearer, 1971). Although many chemicals can induce alarm, almost all honey bees respond more to an intact stinger than to any one compound. All species of Apis have alarm pheromones and the compounds are generally similar among the common honey bee species with the exception of Apis laboriosa, the giant Himalayan honey bee (Vander Meer et al., 1998). The components of the honey bee alarm pheromone can be seen in Table 5. The only difference between European bees and the Africanized “killer” bees is the concentration of the alarm pheromone (Harborne, 1993). Africanized bees secrete alarm pheromones with the same amount of isopentyl acetate but with more 2-nonanol and decyl acetate than other bees which leads to more aggressive defense and more stings to the enemy. Those honey bee species that have open nests tend to have alarm pheromones that persist longer, such as 2-decenyl acetate (Vander Meer et al., 1998).
An elegant study by Wager and Breed (2000) set out to determine whether or not the alarm pheromone of Apis mellifera contains compounds that orient other bees to the tagged target. Using a t-maze they tested each of the individual chemicals as well as mixtures found in the alarm pheromone to see their effect. They found that isopentyl acetate only increases flight activity, a mixture of chemicals (isopentyl acetate, butyl acetate, 1-hexanol, 1-butanol, 1-octanol, hexyl acetate, and 2-nonanol) served to recruit bees from the interior of the colony, and another mixture (isopentyl acetate, 1-hexanol, octyl acetate, n-pentyl acetate, butyl acetate) heightened response to moving objects. They only found one compound that was able to orient bees in the maze, 2-nonanol, but they concluded that there is no well defined function for each of these compounds and that most likely the signal redundancy acts to diminish the chances of a false response.
Alarm in the Wasps
The only wasps that are considered social and exhibit alarm pheromone communication fall into the family Vespidae, although not all wasps in this family are social. Within this family, there are two groups: the vespine wasps (Vespinae) also known as yellowjackets and hornets, and the Polistine wasps (Polistinae) or paper wasps. Table 6 shows the wasps that have been identified as having and responding to alarm pheromones, although many more have been speculated to exist. Maschwitz (1964) was the first to report the existence of a wasp responding to an alarm pheromone. Although most social insect will attack an intruder, there are some species of wasps which will retreat into their nests such as Protopolybia fuscatus and Polybia emaciata (Vander Meer et al., 1998).
As of yet there have only been three species of wasps in which the alarm pheromone has been identified and all of them occur in the Vespinae subfamily: Vespa crabo, Vespula squamosa and Vespula maculifrons (Sledge et al., 1999). All alarm pheromones so far have been found within the venom glands of worker yellowjackets and hornets except for a few reports of Vespula squamosa and Vespula vulgaris workers responding to crushed heads of conspecifics (Vander Meer et al., 1998 and Landolt et al., 1999). Reed and Landolt (2000) found that when Vespula squamosa applies its alarm pheromone to an enemy, the alarm pheromone persists for up to 15 hours which is very uncharacteristic of alarm pheromones in general. This long lasting pheromone may aid in detecting predators, such as skunks and raccoons, when they return to the wasp colony. There have also been a small number of reports stating that the alarm pheromone may be released before the attack, which would be a much more advanced method of communication (Evans and Schmidt, 1990).
More recent data now suggest that alarm pheromone is also found in the Polistinae. The first evidence for this was presented by Jeanne (1981) with the venom of Polybia occidentalis. Since then there have been several other reports of alarm pheromones found in the venom of workers, but none have been identified yet (Dani et al., 2000). Alarm pheromones are also found in some independent-founding paper wasps. For example, Polistes canadensis females will attack when they detect alarm pheromones from conspecifics as will Polistes exclamans and Polistes fuscatus. However, alarm pheromones are certainly not common among independent-founders (Vander Meer et al., 1998).
A study by Moritz and Burgin (1987), briefly mentioned in the previous section, conducted an experiment to quantify the reaction of social wasps to their own alarm pheromones. What they found was that wasps react in the opposite manner as honey bees to increasing group size. Isolated, single wasps responded more to their alarm pheromone than those that were in large groups. They also found that wasps with larger bodies typically have lower defense activity than smaller species.
Although alarm pheromones have not yet been used to control social insects, the possibility is there. Alarm pheromones could be used to rapidly disperse those insects that react in a “flight” response. The pheromone would need to be fairly species specific so that other insects that have a “fight” response will not react to it. (E)-b-farnesene is the alarm pheromone in aphids that stimulates movement in conspecifics when released. Recently this alarm pheromone has been used against the Russian wheat aphid. The aphids will remove their feeding tubes and crawl away when detecting this pheromone (Becker, 2000). This advance is only one of the first steps in understanding how to utilize alarm pheromone communication for insect control.
Literature Cited
Allan, R.A., Elgar, M.A. and R.J. Capon. 1996. Exploitation of an ant chemical alarm
signal by the Zodariid spider Habronestes bradleyi Walckenaer. Proc. Royal Soc. London B. Biology 263(1366), 69-73.
Becker, H. 2000. Alarm pheromone knocks off Russian wheat aphids. Agricultural
Research. Pp.22.
Birch, M.C. 1974. Pheromones. Amsterdam: North_Holland Publishing Company.
Blum, M.S. 1969. Alarm pheromones. Annual Review of Entomology 14, 57-81.
Boch, R. and D.A. Shearer. 1971. Chemical releasers of alarm behaviour on the
honey-bee, Apis mellifera. Journal of Insect Physiology 17, 2277-2285.
Butler, C. 1609. The Feminine Monarchie. On a Treatise Concerning Bees, and the
Due Ordering of them. Joseph Barnes: Oxford.
Dani, F.R. et al. 2000. Chemical characterization of the alarm pheromone in the
venom of Polybia occidentalis and of volatiles from the venom of P. sericea. Physiological Entomology 25, 363-369.
Dicke, M. and P. Grostal. 2001. Chemical detection of natural enemies by
arthropods: An ecological perspective. Annual Reviews of Ecology and Systematics 32, 1-23.
Evans, D.L. and J.O. Schmidt. 1990. Insect Defenses: Adaptive Mechanisms and
Strategies of Prey and Predators. Albany: State University of New York Press.
Feener, D.H. et al. 1996. Specialized parasitoid attracted to a pheromone of ants.
Animal Behavior 51, 61-66.
Harborne, J.B. 1993. Ecological Biochemistry. London: Academic Press Limited.
Hughes, W.O.H. et al. 2001. The response of grass-cutting ants to natural and
synthetic versions of their alarm pheromone. Physiological Entomology 26, 165-172.
Jeanne, R.L. 1981. Alarm recruitment, attack behaviour, and the role of the alarm
pheromone in Polybia occidentalis (Hymenoptera: Vespidae). Behavioral Ecology and Sociobiology 9, 143-148.
Kugler, C. 1979. Alarm and defense: a function for the pygidial gland of the
myrmicine ant, Pheidole biconstricta. Annals of the Entomological Society of America 72(4), 532-536.
Landolt, P.J., Reed, H.C. and R.R. Heath. 1999. An alarm pheromone from heads of
worker Vespula squamosa (Hymenoptera: Vespidae). Florida Entomologist 82(2), 356-359.
Maschwitz, U. 1964. Alarm substances and alarm behavior in social Hymenoptera.
Nature 204, 324-327.
Moritz, R.F.A. and H. Burgin. 1987. Group response to alarm pheromones in social
wasps and the honeybees. Ethology 76, 15-26.
Prestwich, G.D. 1979. Chemical defense by termite soldiers. Journal of Chemical
Ecology 5(3), 459-479.
Reed, H.C. and P.J. Landolt. 2000. Application of alarm pheromone to targets by
southern yellowjackets (Hymenoptera: Vespidae). Florida Entomologist 83(2), 193-196.
Sledge, M.F. et al. 1999. Venom induces alarm behaviour in the social wasp
Polybioides raphigastra (Hymenoptera: Vespidae): an investigation of alarm behaviour, venom volatiles and sting autonomy. Physiological Entomology 24, 234-239.
Traniello, J.F.A. and A.K. Jayasuriya. 1981. Chemical communication in the
primitive ant Aneuretus simoni: The role of the sternal and pygidial glands. Journal of Chemical Ecology 7(6), 1023-1033.
Vander Meer, R.K. et al. 1998. Pheromone Communication in Social Insects.
Boulder: Westview Press.
Wager, B.R. and M.D. Breed. 2000. Does honeybee sting alarm pheromone give
orientation information to defensive bees? Annals of the Entomological Society of America 93(6), 1329-1332.