Chapter 7

PHEROMONES

Harry H. Shorey

Modern man makes minimal use of chemical communication with others of his own species. Thus, he has little intuitive feel for the great reliance placed on this communication mode by much of the remainder of the animal kingdom. Correspondingly, until recent years, man's research into the chemical communication systems used by other animals has lagged behind research into visual and sonic means of communication. However, during the past few decades, advances in microanalytical chemistry, coupled with an increasing recognition that man may profit by controlling the chemical communication systems of certain animal species, has resulted in a great expansion in our knowledge in this previously neglected field.

The term "pheromone" was coined by Karlson and Lüscher (1959), and slight modifications in the definition were proposed by Kalmus (1965). A pheromone is a chemical or a mixture of chemicals which is released to the exterior by an organism and, which, upon reception by another individual of the same species, stimulates one or more specific reactions. Although this review is restricted to a consideration of animals, interorganism communication by pheromones also occurs in many species of plants.

Wilson and Bossert (1963) and Wilson (1963a) separated pheromones into two categories: Releaser pheromones are direct stimulators of behavior, often involving a classical stimulus- response sequence that is mediated entirely by the central nervous system; primer pheromones induce relatively long-lasting physiological changes, especially involving the endocrine and reproductive systems. Primer pheromones, for example, cause acceleration or inhibition of maturation of the reproductive systems of many vertebrate as well as invertebrate species. Because this chapter is concerned with communication systems in which a more immediate response to the chemical stimulus occurs, primer pheromones will not be discussed.

Sometimes when an animal is exposed to a pheromone, a stimulus-respone reflex is not seen. In much of the communication of vertebrates—as well as that of many invertebrates— the pheromone may cause subtler effects. It may, for example, sensitize the animal to other stimuli, such as the sight of a nearby potential mate. In other instances, the pheromone may be used to identify the other individual or to assess the physiological or social status of the other individual, with no overt behavioral reaction resulting at all.

IDENTIFICATION

Individual animals of many species of mammals and fish can determine the identity of other members of the same species by their odors (Bardach and Todd, 1970; Beauchamp, 1973; Müller-Schwarze, 1971; Mykytowycz, 1970). Whether any invertebrates are capable of this identification of individuals is doubtful; however, Linsenmair and Linsenmair (1971) noted that males and females of the desert wood louse, terrestrial crustaceans that pair for life, drive off from their burrows all intruders of either sex except their partners. The identification is apparently based on pheromones.

An ability to identify individual animals on the basis of pheromones alone infers that the pheromone of each individual consists of a complex blend of chemicals that differs in some way from that of other individuals. There is considerable evidence that such complex blends do occur. The chemicals secreted by the anal, chin, and inguinal glands of the rabbit (Oryctolagus cuniculus) vary both qualitatively and quantitatively from individual to individual (Goodrich and Mykytowycz, 1972). The number of different compounds in the pheromone blend may be very high; the castoreum gland of the beaver (Castor canadensis) contains approximately fifty different chemicals (Kingston, 1965).

Pheromones may also be used to identify the physiological or social status of an individual and the social group or colony to which that individual belongs. A semantic problem enters here; in some cases, the pheromone may not be used for actual "identification" of these characteristics, but may simply cause the release of appropriate behavior by the respondent, such as a copulatory reaction by a male when exposed to a pheromone released by a sexually receptive female. Especially in the vertebrates, there is probably a continuum between identification of the status of an individual and stimulation of appropriate behavioral acts by the "identifying" animal.

An extensive use of pheromones by mammals in determining the identity as well as the physiological or social status of their species mates is often inferred by the characteristic behavior displayed when two animals meet. A prominent part of this meeting behavior is the "sniffing" of certain portions of the body, especially the nasal, anal, genital, and various glandular areas (Schloeth, 1956).

Social insects, as well as many mammalian and fish species, differentiate between members of their own social group or colony and those of other social groups by odors (Mykytowycz, 1970; Wilson, 1971). Characteristic behavior may be displayed by the individual perceiving the odors; these responses will be discussed later with reference to behavior associated with territoriality, aggression, and marking of a home range. Mechanisms must exist for the establishment of a common odor on all members of the colony or social group. In certain mammalian species, the animals in a social group—especially the dominant males—mark other individuals within the group with a pheromone. This method of attaining a common group odor is seen in the gliding phalanger (Petaurus breviceps papuanus: Schultze-Westrum, 1965), the rabbit (Mykytowycz, 1965), and the tree shrew (Tupaia belangen: Martin, 1968). Mutual scents of colony members in the social insects may be acquired in a number of ways, including: secretions produced by the queen and distributed throughout the colony; secretions produced by the colony members themselves, with each member having a common genetic complement provided by the queen-mother; the odor of the specific nest material in which the colony is housed; and a common odor due to the specific food on which the colony has fed (Hangartner et al., 1970; Morse, 1972; Wilson, 1963b).

AGGREGATION

An aggregation pheromone causes one or more responding individuals to become localized near the source of pheromone emission. General use of the term "attraction" to describe the process by which aggregation is accomplished should probably be avoided. According to the definition of Dethier et al. (1960), an attractant causes oriented locomotory responses toward the odor source. In addition to attraction, a number of other behavioral mechanisms, operating singly or in conjunction with each other, may bring about aggregation. For example, certain pheromones may inhibit locomotion and thus reduce dispersal of animals away from the vicinity of the odor source (Shorey, 1973).

Some pheromones appear to have no biological function other than promoting aggregation. In other cases, the same pheromone that stimulates the animals to aggregate then assumes an additional role in stimulating appropriate behavior, such as mating or aggression. This latter group of pheromones will be discussed here with reference to their aggregative properties as well as in later sections with reference to the stimulation of the additional behavioral activities.

Mechanisms of Aggregation

Most of the research on the behavioral mechanisms used by animals that aggregate in response to stimulatory odors has been conducted on insects. The movement toward the odor source may take place over tens, hundreds, or even thousands of meters. Many of the claims of great distances are probably exaggerated. For instance, if a male moth is marked and released at a great distance from a pheromone-emitting female, there is no way of knowing what proportion of its flight to the pheromone source might consist of a random, appetitive locomotory behavior, not stimulated by the pheromone.

The flow and turbulence characteristics of air probably destroy any recognizable chemical concentration gradient within a few centimeters of the source of the pheromone. For this reason, most investigators are agreed that the guidance mechanisms used by animals orienting toward an odor source from relatively great distances cannot be based simply on the following of an odor gradient (Bossert and Wilson, 1963; Shorey, 1973). The guidance is probably accomplished by a complex interaction of a number of behavioral mechanisms. One that has been demonstrated for certain flying insects is anemotaxis (wind steering), whereby the animal steers its body axis upwind when it is stimulated by an appropriate odor (Farkas and Shorey, 1974; Kennedy, 1939). An aquatic animal may also orient its body axis into the flowing current; in this case, the phenomenon is called "rheotaxis." Considerable evidence indicates that the upwind or upcurrent orientation is accomplished by the animal's steering its body axis so that the visual field of the substratum over which it is flying or swimming unreels caudally, from front to rear. If the animal were proceeding at an angle to the current, sideslippage would cause the visual field to appear to move at an angle relative to the body axis; presumably, the animal would then make corrective steering reactions that would orient its body axis upcurrent. If the animal loses the odorous current, it may make crosscurrent casts (Traynier, 1968), which probably maximize its likelihood of regaining the odor stimulation.

Another mechanism that might be used in distance orientation to a pheromone source is aerial odor-trail following (Farkas and Shorey, 1974). The cloud of molecules extending down-current from the odor source often can be visualized as forming a three-dimensional trail. Many species of flying insects exhibit lateral zigzag or sinusoidal oscillations across the breadth of the trail as they proceed toward the odor source. Some experimental evidence indicates that these oscillations may be an important factor in enabling the flying insect to maintain contact with the odorous trail. Perhaps appropriate turns, back toward the central axis of the trail, are stimulated when the insect senses that it is entering a lower pheromone density as it diverges to the right or left of the axis. This method for aerial trail following, then, would be analogous to a mechanism proposed for animals that follow terrestrial trails (discussed below). Possibly the phenomenon of aerial trail following, operating to cause the animal to maintain itself within the boundaries of the trail, interacts with anemotaxis, to orient each turn in an upcurrent direction.

The progressively increasing odor concentration encountered as an animal approaches a pheromone source may stimulate other behavioral reactions that result in the animal's localization near the source. The speed of forward progress may be reduced, and ultimately locomotion may be inhibited, as the animal encounters the higher concentrations (Bennett and Borden, 1971; Farkas and Shorey, 1974; Wood, 1970). Also, the animal may become sensitized by the high concentrations to respond to other environmental stimuli. For instance, a number of insect species, when flying in air containing a high concentration of an aggregation pheromone, tend to approach and land on vertical objects such as the trunks of trees (references in Farkas and Shorey, 1974). Presumably, this behavior is adaptive, because the pheromone emitter is likely to be located on such a vertical object.

When the animal is very close to the pheromone source and in a well-defined chemical concentration gradient, chemotactic mechanisms may steer it toward the highest pheromone concentration. Males of many moth species, when stimulated by the high pheromone concentrations found near receptive females, steer their bodies first to one side and then to the other, with wings vibrating. The air pulled by the vibrating wings, from front to rear and over the antennae, probably gives good cues as to the direction of the site of pheromone emission (Schneider, 1964).

Finally, visual orientation to the body of the pheromone emitter itself may be stimulated when a male insect is exposed to a high concentration of female pheromone (Daterman, 1972; Shorey and Gaston, 1970; Traynier, 1968).

Terrestrial odor trails differ from aerial odor trails in that the odor molecules do not emanate from a single point source. Rather, the pheromone emitter deposits the chemical as a series of discrete spots or as a continuous streak as it moves along the substratum. Such pheromone trails are widely used in aggregation behavior. Some examples, which will be discussed in later sections, include the pheromone trail deposited by a snake crawling along the ground, by a hoofed mammal through contact with the ground during locomotory behavior, and by workers of many ant species when they return to their nest after finding food. Mechanisms of terrestrial trail following have been best studied in ants. The following ants detect the small active space caused by the volatilization of the chemical into the air above the trail. Some evidence indicates that they maintain contact with the trail as they run along it by chemotactic mechanisms, which tend to steer them toward the central axis of the trail whenever they diverge to one side (Hangartner, 1967). Thus, the behavior of ants (and of dogs) following terrestrial odor trails is frequently characterized by a zigzag progression (Wilson, 1962; Wilson and Bossert, 1963).

Role of Aggregation Pheromones in Animal Behavior

As mentioned earlier, aggregation is typically a precursor of further behavioral activities such as aggression or mating. Those pheromones that stimulate both aggregation as well as the further behavioral activities will be considered in later sections. This section will be involved with pheromones whose roles are essentially restricted to causing aggregation, with further behavioral activities within the aggregations apparently being caused primarily by other stimuli.

Cellular slime molds live as single-celled amoebae, ingesting bacteria, during the feeding phase of their life cycle. After the food supply is exhausted, some of the amoebae of Dictyostelium discoideum (the most studied species) emit pulses of a pheromone identified as cyclic 3,'5' adenosine monophosphate (Konijn et al., 1968). The pheromone causes nearby amoebae to move toward the source and to emit pulses of the same chemical. The aggregating amoebae form a multicellular organism, which rises above the substratum and differentiates into a stalk and an apical fruiting body (Bonner, 1970).

Many animal species use pheromones to locate their home or nest site. Limpets (Mollusca) of a number of species have a specific area on a rock to which they return after feeding expeditions. Although not yet conclusive, considerable evidence indicates that the limpets find their "homes" by following a pheromone trail laid down by them on the substratum during previous excursions. Similarly, a number of social Hymenoptera (bees, ants, and wasps) recognize their own nest sites and are stimulated to enter the nest entrance by odors characteristic of their specific colony (Butler et al., 1969; Hangartner et al., 1970). The odor, which has sometimes been referred to as "hive atmosphere," may originate partly from pheromones produced by the colony mates and partly from the general odor of their food and nest material.

In some animal species that do not have a permanent home or resting site, pheromones may stimulate the formation of temporary aggregations in which the individuals spend certain times of the day or the year. Thus, pheromones promote aggregations of a number of arthropod species in sheltered places during their inactive time of day (Friedlander, 1965; Levinson and Bar Ilan, 1971). In certain coccinellid beetles (Hodek, 1960) and snakes (Dundee and Miller, 1968; Noble and Clausen, 1936), pheromones play a role in promoting aggregations prior to the time of hibernation. Noble and Clausen (1936) found that the common brown snake follows the trails laid by previous snakes moving toward hibernation sites in suitable cavities. Once in the cavities, a combination of visual and pheromonal stimuli maintains cohesiveness of the aggregations.

A number of marine animals that are sessile as adults produce free-swimming larvae. At the proper state for settling, the larvae often give a gregarious response to other members of their species. Thus, a pheromone released by barnacles that are successfully colonizing a surface causes larvae of their species to explore and settle on the same surface (Crisp and Meadows, 1963).

Many aquatic animals move together in aggregations called "schools." The function of the schools is not well understood; they might be of benefit in locating suitable food sources, for mutual protection, and in bringing together the sexes prior to mating. Vision is apparently the major sense involved in maintaining schools of fish. However, schooling fish have also been found to respond to the odor of conspecifics; this pheromone-induced aggregation may be a major mechanism for maintaining schooling behavior at night (Hemmings, 1966; Wrede, 1932).

Pheromones are characteristically used by the worker caste of social insects (bees, ants, and termites) to recruit colony mates to a supply of food (Moser and Blum, 1963; Stuart, 1970; Wilson, 1962, 1963b). There are many variations in the recruitment behavior; however, a typical sequence of events for many ant species is as follows: If an ant finds a source of food too large to transport to the nest in one piece, it returns to the nest while depositing a trail of pheromone on the ground. Other ants in the nest are stimulated to follow the trail. After arriving at the food source, they carry pieces to the colony while also depositing pheromone on the substratum. Thus, the terrestrial trails are fortified as long as food is available. If no food remains, the returning ants deposit no pheromone, and the trail fades out through volatilization. Some trails are very transitory; this is the case with the imported fire ant, which mainly forages for small items of food within a few meters of its nest (Wilson, 1962). Other ant species have highly persistent trails; an example is the Texas leaf-cutting ant, which may harvest leaves from plants hundreds of meters from its nest for several months (Moser and Blum, 1963).

Some ant species lay trails as they proceed outwards, in search of food. These have been called "exploratory trails" by Wilson (1963b) and are typical of army ants (genus Eciton). Army ants are true nomads, without a permanent nest. Their massive columns, either radiating from a central area at which the colony is temporarily bivouacked or during emigration of the entire colony, are guided by pheromone deposited by the individual workers as they proceed outwards.

Certain ants practice "slavery." They raid the nests of related species, kill or repel the workers, and capture the pupae, which they transport back to their nest. Slave-making ants of the Formica sanguinea group direct their raids to selected nests of other species by odor trails. Once in the raided nest, the ants discharge large amounts of pheromone consisting in part of decyl alcohol, dodecyl alcohol, and tetradecyl alcohol. The pheromone attracts further raiders and at the same time causes the defenders of the raided nest to disperse (Regnier and Wilson, 1971).

Workers of a number of stingless bee species, especially in the genera Melipona and Trigona, release from their mandibular glands large quantities of odor when they have located a suitable source of food. The pheromone attracts other workers, which collect the food and transport it to the nest (Lindauer and Kerr, 1960). Certain species make trails that are partly terrestrial and partly aerial. A scout bee of Trigona postica, after having collected nectar or pollen, flies from the flowers toward the nest, stopping at frequent intervals and leaving at each spot an odor mark from her mandibular glands. Recruited workers follow the resulting odor path to the food source. Another variation in stingless bee recruitment behavior is found in Lestrimelitta limao, a species that steals food from nests of other stingless bees. The first invading bees are almost invariably killed, resulting in the liberation of citral from their mandibular glands (Blum et al., 1970). This pheromone attracts large numbers of additional invaders, which raid the nest. Citral also causes a complete disruption in the social organization of the raided colony, with the resident bees often abandoning their nest.

A complex of pheromones secreted by the workers and the queen is used during swarming behavior of the honeybee (Butler and Simpson, 1967; Morse and Boch, 1971). Swarming bees usually form a cluster on a branch or some other substrate at least once during their migration from the parent hive to a new nesting site. Aggregation of the workers and the queen at a clustering site or at a new nest site is stimulated by a pheromone emitted by the scout bees that locate that site. The swarming bees are also attracted by a pheromone from the queen, 9-oxo-trans-2-decenoic acid, which enables them to relocate her if they should lose her. Another pheromone emitted by the queen, 9-hydroxy-trans-2-decenoic acid, acts as a behavioral stabilizer for the worker bees, causing them to alight and group into a quiet cluster when near her.

In many beetle species, either one sex or the other emits a pheromone that causes the aggregation of beetles of both sexes. In most cases, the biological function served by the aggregation has not been determined (Abdullah, 1965; Levinson and Bar Ilan, 1967). However, one might infer that two major functions often served are the enhancement of mating through the bringing of beetles of both sexes into close proximity, and the bringing of beetles to a suitable food source detected by the pheromone-emitting individual. Eisner and Kafatos (1962) assume that the large aggregations of Lycus loripes beetles that occur in response to a pheromone emitted by the males confer a survival benefit. These beetles are distinctively colored and possess a chemical defense mechanism against potential predators. A predator that has had a distasteful experience with one individual in the aggregation would have a reduced likelihood of attacking other individuals.

Aggregation of both sexes in response to a pheromone emitted by one sex is characteristic of the ambrosia beetles and bark beetles (family Scolytidae). These beetles are infrasocial, with large numbers colonizing a suitable host tree. The tree that is designated for colonization may be dead, dying, weakened, or healthy, depending on the species of beetle involved. Either males or females, depending on the species, initially invade a host tree. A pheromone, often consisting of a medley of chemicals, is secreted by these invaders, either when they first arrive at the tree or after they have constructed an entry tunnel through the bark and have commenced feeding on the phloem. This pheromone, often in association with volatile chemicals from the tree itself, causes the approach of others of both sexes.

Typically, the responding beetles of the same sex as the initial invaders also establish entry tunnels and release a pheromone. Beetles of the other sex, upon arriving at the tree, may also release a pheromone and enter the tunnels constructed by the initially invading sex. The beetles mate within the tunnels, and their offspring establish feeding galleries under the bark. Changes in the concentration and the blend of pheromone chemicals released as the tree is progressively attacked by more beetles may determine the course of colonization, being at first stimulatory and later inhibitory to aggregation. The complex of chemicals released early in the colonization process stimulates beetles to approach from a distance and often inhibits their tendency toward further locomotion after they land on the tree (Bennett and Borden, 1971; Vité and Pitman, 1969; Wood, 1970).

Later in the colonization process, chemicals may be released that cause beetles not to approach the tree or not to be arrested in their locomotion if they should land on the tree. This inhibition of aggregation is often seen when responding beetles of the opposite sex from that of the initial invaders enter the tunnels made by the latter (Rudinsky et al., 1973; Vité and Renwick, 1971). Inhibition of further colonization also appears to be due to the increasingly high concentration of previously attractive chemicals released by the aggregating beetles. Apparently, the concentration often rises to a sufficiently high level to stimulate beetles to orient visually to other, nearby trees (Gara and Coster, 1968). Thus, adjacent trees are often colonized.

Only a few additional examples of pheromone-mediated aggregation behavior will be given here so as to indicate the great diversity of biological functions that might be served by such behavior. Workers of certain termite species lay pheromone trails from a point of disturbance, such as a breach or an invader in the nest, to other places within the nest. Other termites follow the trail to the site of disturbance and then display appropriate behavior depending on the stimulus nature of the disturbance itself (Stuart, 1967). Young larvae of snails of the genus Crepidula aggregate in response to a pheromone released by mature females. They attach to a female and are induced, through the action of a primer pheromone produced by her, to transform into males, which later inseminate her (Gould, 1952). Freshwater pulmonate snails (Physa acuta) follow trails made by conspecifics when they move to the surface of the water to replenish their air supplies (Wells and Buckley, 1972). A pheromone, 2-methoxy-5-ethylphenol, is given off from the feces of migratory locusts. The chemical serves both as an aggregation pheromone, promoting gregarious behavior of the immature locusts, and as a primer pheromone, inducing the physiological changes that are characteristic of the migratory phase of this species (Nolte et al., 1973).

DISPERSION

Pheromones that cause dispersion of conspecifics can be subdivided into several fairly distinct categories, dealing with maintenance of optimal interindividual spacing, inhibition of tendencies to aggregate, dispersion during times of danger, and maintenance of the integrity of territories.

Maintenance of Optimal Spacing

At particular times during the life cycle of some animal species, pheromones are apparently released by all the individuals and maintain an optimal distance between them. Amoebae of Dictyostelium discoideum were discussed earlier with reference to their behavior of aggregating after completion of feeding. Bef ore the aggregation phase, the amoebae secrete a pheromone that repels others of the species. The pheromone thus causes the individual amoebae to remain dispersed and to optimally utilize the available food (Bonner, 1970). After the food has been depleted and the amoebae have come together into centers of aggregation, each center releases a pheromone that further maintains optimal spacing by inhibiting the formation of nearby centers. As a fruiting mass rises on a stalk from each center, above the substratum, the same spacing pheromone controls the direction of growth of the stalk so that it is maximally displaced from nearby stalks.

Mature larvae of a number of moth species that infest grain, when they meet, deposit on the substratum a pheromone from their mandibular glands (Mudd and Corbet, 1973). The pheromone causes the larvae to disperse, thereby increasing the likelihood that they will reach an area of low population density before they spin their cocoons and pupate.

Female flour beetles, when at high population densities, distribute themselves uniformly throughout their food medium. A pheromone secreted by the females themselves, as well as a pheromone secreted by their larvae, repels the females (Naylor, 1965). The resulting spacing is presumably adaptive, resulting in optimal colonization of the environment.

Inhibition of Tendency to Aggregate

Adults of many insect species aggregate in areas containing suitable food for their young. The females of certain of these species, after depositing their eggs in the food material, mark the external surface of the material with a pheromone that inhibits other females from depositing eggs in the same area. This behavior has been seen in certain hymenopterous parasites that lay their eggs in insects of other species (Rabb and Bradley, 1970) and in the apple maggot fly, which lays its eggs in various fruits (Prokopy, 1972).

When not receptive for mating, a female ground beetle (Pterostichus lucublandus) normally runs away from males. If pursued by a male, the female discharges a blast of liquid toward him from the tip of her abdomen. The male stops running, makes cleaning movements of his face and antennae, becomes uncoordinated, and may fall into a coma for several hours. Kirk and Dupraz (1972) feel that this defensive behavior may be used by the female while laying eggs. The male, which would eat the eggs of the species, is immobilized until the female completes oviposition and covers the entrance to the egg chamber.

Dispersion during Times of Danger

Pheromones released when an animal is threatened with danger or when injured are generally called "alarm pheromones." For many species, their primary role is the stimulation of dispersion of conspecifics. For other species, the alarm pheromone may induce aggregation plus aggressive behavior; these responses will be considered in a later section. And for still other species, the alarm pheromone may induce either dispersion or aggregation plus aggression, depending on the environmental context in which the animals perceive the pheromone.

Aquatic snails of nineteen species have been found to exhibit escape reactions when they perceive the juices of crushed conspecifics. The behavior usually consists of dropping from the substratum to the bottom of the water, followed by burrowing into the bottom material. However, certain air-breathing species react by crawling up out of the water (Snyder, 1967). Likewise, sea urchins (Diadema antillarium) move rapidly away from an area containing the juices of crushed conspecifics; they often mount on their ventral spines and "race" away for one or two meters (Snyder and Snyder, 1970). Alarm reactions are stimulated in a number of schooling fish species by a pheromone that is released when the skin of a conspecific is injured. The reactions are extremely variable, depending on the species, and range from avoidance of the area containing the pheromone and dispersal of the school to an increase in the cohesiveness of the school (Bardach and Todd, 1970; von Frisch, 1941). An increase in school cohesiveness might at first seem maladaptive; however, Pitcher (1973) indicates that certain predator fish are less successful in attacking schools of fish than individual prey.

Pheromones released by many animals when they are threatened or attacked by an enemy serve a dual function. They not only cause conspecifics to disperse but also may serve a defensive function, deterring or repelling the predator. Examples are seen in earthworms (Ressler et al., 1968), a number of aphid species (Bowers et al., 1972), the plant bug (Dysdercus intermedins: Calam and Youdeowei, 1968), the beetle Blaps sulcata (Kaufmann, 1966), and ants of a number of species.

Ant alarm pheromones are highly variable in their effect, depending on the species and the location and behavioral activities of the ants when they perceive the pheromone. At the nest, the pheromone typically triggers attraction and attack behavior (considered later). However, away from the nest, at the feeding place, ants perceiving the pheromone typically disperse (Maschwitz, 1966). Even at the nest, those species that construct small or diffusely distributed colonies often disperse (Wilson and Regnier, 1971).

Mammalian alarm pheromones may also be highly variable in their effect, depending to a large extent on the context in which they are perceived and on the previous experiences of the responders. However, repellency or dispersion is one of the possible effects, as demonstrated by the reactions of mice to the odor of stressed conspecifics (Rottman and Snowdon, 1972).

Territories

Individuals of many vertebrate species band together in social groups that occupy clearly demarked territories. Each territory is occupied solely by a given group and is defended against all entering foreign conspecifics. Occupied territories may be designated by the occupiers by means of visual, sonic, or chemical cues. Chemical cues are especially used by the mammals, which deposit odorous pheromone secretions on objects within the territory, especially near the borders of the territory. This scent marking seems to be especially advantageous, because ownership of the territory is advertised even when the residents are absent (Jones and Nowell, 1973). Male mice deposit a pheromone from the coagulating glands with the urine. Foreign males are deterred from investigating the marked areas; if an intruder enters a marked territory, the pheromone reduces his aggressive tendencies, thus tipping the scales in favor of a victory by the resident and resulting in the ultimate flight of the intruder (Jones and Nowell, 1973). Similar territorial behavior is seen in a number of other mammal species, including the rabbit (Mykytowycz, 1968), the tree shrew (Martin, 1968), and the black-tailed deer (Müller-Schwarze, (Müller-Schwarze, 1972).

SEXUAL BEHAVIOR

A group of diverse chemicals are collected under the term "sex pheromones." They are produced by either males or females and stimulate one or more behavioral reactions in the opposite sex. The reactions lead directly or indirectly to mating. The indirect reaction most frequently observed is aggregation of the opposite sex near the pheromone-emitting sex. After the two sexes have come together, a variety of other reactions involved with courtship or copulatory behavior may occur under the influence of sex pheromones. In some animal species, a pheromone might stimulate only the aggregation reactions, and other, nonpheromonal stimuli may be involved in stimulating close-arange courtship and copulatory responses. In other species, pheromones might only be involved in courtship and copulatory behavior after the sexes have been brought together in response to other stimuli. In still other species, the same pheromone that causes the approach of a mating partner then stimulates appropriate close-range behavioral responses as well.

Aggregation Prior to Sexual Behavior

The role of pheromones released by females at or before the time of their receptivity for mating in causing male aggregation has been documented so frequently in the animal kingdom that only a few representative examples will be given here:

Phylum Protozoa

A ciliate (Rhabdostyla vernalis: Finley, 1952).

Phylum Aschelminthes

Rotifers (Gilbert, 1963), and a variety of free-living, plant-parasitic, and animal-parasitic nematode species (Greet et al., 1968; Salm and Fried, 1973).

Phylum Arthropoda

A variety of species of crabs (Kittredge et al., 1971; Ryan, 1966), amphipods (Dahl et al., 1970), copepods (Katona, 1973), spiders (Dondale and Hegdekar, 1973), ticks (Berger, 1972), mites (Beavers and Hampton, 1971; Cone et al., 1971), and a very large number of species of insects (see Jacobson, 1972, and Shorey, 1973, for partial reference lists).

Phylum Chordata

Various species of amphibians (Cedrini, 1971) , snakes (Noble, 1937), and fish (Tavolga, 1956; Timms and Kleerekoper, 1972), and many species of mammals (Beauchamp, 1973; Davies and Bellamy, 1972; Lindsay, 1965; Schein and Hale, 1965).

Only a few examples of unusual pre-mating aggregation behavior will be mentioned here. Females of the spider Pardosa lapidicina secrete a pheromone into the silken strand that they leave on the substrate as they move about; males use the pheromone-treated strands in locating the females (Dondale and Hegdekar, 1973). The queen honeybee mates sequentially with a number of drones while flying several meters above the ground; therefore, the drones must orient to an aerial trail of pheromone that is emitted from a moving source (Butler and Fairey, 1964). In the solitary bee (Adrena flavipes), an odor characteristic of the female's nest site, rather than one released by the female directly, causes males to search for females in that area (Butler, 1965). In a number of termite species, after reproductives of both sexes leave the nest in a dispersal flight, tandem pairs, each consisting of a male and a female, form on the ground. A male apparently recognizes a female by her pheromone and follows her closely as she runs off to a suitable area, where they excavate a cell in the soil or wood (Stuart, 1970). In some animals, including the crab-hole mosquito (Downes, 1966) and certain mites (Beavers and Hampton, 1971; Cone et al., 1971) and crabs (Kittredge et al., 1971; Ryan, 1966), the males are attracted to the vicinity of the female before she molts into her reproductive state; they remain in attendance for some time awaiting her emergence. In some crabs the pheromone that induces male aggregation is apparently identical to the molting hormone, crustecdysone (Kittredge et al., 1971). The male crab is stimulated to grasp the female with his chelae, hold her, and carry her beneath his body for some days before she molts, at which time copulation occurs.

In a number of animal species, the male releases a pheromone that attracts the female from a distance. Examples include:

Phylum Aschelminthes

Several nematode species (Bonner and Etges, 1967; Salm and Fried, 1973).

Phylum Arthropoda

A large number of insect species (see Jacobson, 1972, and Shorey, 1973, for partial reference list).

Phylum Chordata

Mammals: mice (Davies and Bellamy, 1972) and rats (Carr et al., 1965).

Male bumble bees lay down scent trails on their mating flights. They fly around a circuit or loop, stopping at various points and depositing scent marks. A number of drones fly around the same circuit, continuously fortifying the scent marks. The queen bumble bees are apparently attracted to these scent paths and mate with the drones flying along them (Lindauer and Kerr, 1960).

In the ant Camponotus herculeanus, a male-produced pheromone has no aggregative properties per se, although it results in aggregation of the two sexes. The males take off from the nest in a swarming flight and then release from their mandibular glands a pheromone that stimulates the females to fly. Thus, the pheromone synchronizes the timing of flight of the two sexes (Hölldobler and Maschwitz, 1965).

Both males and females of some animal species emit pheromones that stimulate aggregation of the opposite sex. Examples include:

Phylum Aschelminthes

Certain nematodes (Bonner and Etges, 1967; Salm and Fried, 1973).

Phylum Arthropoda

Certain plant bugs (Mitchell and Mau, 1971) and beetles (August, 1971).

Phylum Chordata

Mammals: mice (Davies and Bellamy, 1972) and rats (Carr et al., 1965).

Courtship and Copulation

Often, the same pheromone that stimulates the approach of a male to a female from a distance then induces his close-range courtship or copulatory behavior. This dual function of the pheromone is seen in such diverse groups as rotifers (Gilbert, 1963), insects (Shorey, 1973, and included references), fish (Losey, 1969; Tavolga, 1956), and mammals (Lindsay, 1965; Schein and Hale, 1965). Receptive females of many mammalian species deposit the pheromone with their urine (Beauchamp, 1973; Davies and Bellamy, 1972; Schein and Hale, 1965). The urine marks not only act as focal points for distance attraction of males but also at close range increase the intensity of courtship behavior patterns. In the gobiid fish Bathygobius soporator, a pheromone secreted by the female induces the approach of males, a change in the males' coloration to the "courtship phase," and courtship reactions composed of rapid fanning and gaping movements (Tavolga, 1956).

Among a number of insect species, the progression of behavioral steps involved in the approach of a male to the vicinity of the pheromone source and the following stimulation of his courtship and copulatory behavior are arranged in a hierarchy, with each successive step requiring a higher concentration for its release than the previous one (Bartell and Shorey, 1969; Traynier, 1968). For example, Bartell and Shorey (1969) found that the sequence of behavior exhibited by a group of males of the light brown apple moth that were stimulated by the odor of extracted female pheromone included antennal movements, whole-body movements, initiation of flight, orientation toward the pheromone source, and a complex series of activities near the source, including attempted copulation with another male moth. A greater than 100,000fold increase in pheromone concentration was required to elicit the final step (copulatory attempts) as compared with the first step (antennal movements). The concentration required to elicit upwind orientation was intermediate.

In some cases, the female-produced pheromone apparently causes no aggregation behavior, and the only evident role for the pheromone is in the stimulation of courtship displays or direct copulatory reactions. The only reported response of males of the nematode Nematospiroides dubius following stimulation by the female sex pheromone is a flaring of the copulatory opening (Marchant, 1970). A male hermit crab, Pagurus bernhardus, perceives the pheromone upon contact with the exoskeleton of a receptive female. He then grasps the rim of the female's shell aperture with his minor cheliped and pulls her around for some hours or days, until she presumably arrives at the appropriate state for mating (Hazlett, 1970). In certain Diptera, including the sheep blowfly (Bartell et al., 1969) and Drosophila melanogaster (Shorey and Bartell, 1970), the odorous female pheromone apparently lowers the threshold for visually guided courtship behavior, which is directed toward a nearby fly. Additional pheromone stimulation, apparently obtained through direct contact, releases the copulatory behavior of the male flies. Male mosquitoes (Aedes spp.) are attracted by the sounds produced by the vibrating wings of flying females. The males are stimulated to copulate when they perceive a pheromone through direct contact with the females (Nijholt and Craig, 1971). Male tortoises (Geochelone spp.) also initiate courtship or copulatory attempts following pheromone stimulation (Auffenberg, 1965). As with much pheromone behavior, the stereotyped behavioral response is often displayed when the pheromone stimulus is presented in the wrong context; male tortoises have been observed to mount such inappropriate objects as a head of lettuce over which a female had recently clambered.

Although some male-produced pheromones stimulate the approach of females from a distance, the primary role for most such pheromones is in courtship. The details of the courtship behavior vary greatly from species to species. In certain millipede (Haacker, 1971) and cockroach (Barth and Lester, 1973) species, the male displays his pheromone glands when a female is nearby. The female is stimulated to feed on the glandular secretions, thus causing her to cease locomotion and to be positioned properly so that the male can copulate with her. In some tephritid fruit flies (Fletcher, 1969), as well as in swine (Hafez et al., 1962), a complex of chemical and sonic signals produced by the male causes the female to assume the mating position. The female fruit flies assume a stance with the reproduction segments extruded and directed toward the pheromone source, while pigs undergo an "immobilization reflex." The swine pheromone consists of a complex of 16- unsaturated C19 steroids, some of which are concentrated in the boar's sweat glands and salivary glands (Gower, 1972). A practical use for these chemicals has been found in animal breeding practices. Two of the chemicals, 5a-androst-16en-3-one and 3a-hydroxy-5a-androst-16-ene, have been incorporated in an aerosol that is dispensed at the female pig prior to artificial insemination, causing her to assume the immobilization reflex and thus be positioned properly for artificial insemination (Melrose et al., 1971).

Unlike the well-known pheromones of female moths, which often attract males from considerable distances, the pheromones produced by males of many moth and butterfly species have the subtle (to humans) function of arresting locomotion of the female after the male has come in close proximity to her (Barth, 1958). The male-produced pheromones might also have a role in lowering the threshold of the female for accepting the male in copulation. Males of many butterfly species follow the females in a visually oriented aerial "dance." While airborne, they distribute their pheromone over the female's antennae from specialized glandular areas, usually located on the abdomen or wings. The female is thus stimulated to alight, whereupon the male may dispense more pheromone over her antennae before attempting to copulate with her (Myers and Brower, 1969; Pliske and Eisner, 1969). In many moth species, the males that have been attracted to the vicinity of the female evert brushlike scent-dispensing structures at the moment that they are about to attempt copulation (Aplin and Birch, 1968; Clearwater, 1972). The male moth pheromone, like that of butterflies, apparently functions mainly as a locomotory arrestant, inhibiting the normal flight response of the female.

Physiological Influences

Animals do not usually produce sex pheromones at a constant level, nor is their responsiveness to pheromones at a constant level, during their entire span of adult life. Rather, physiological mechanisms ensure that the pheromone communication system is operative at the appropriate stage of sexual maturity. For example, there is a direct link in rats and mice between the titre of estrogen in females and androgen in males (signifying reproductive maturity) and the attraction exerted by either sex for the opposite one. Likewise, the responsiveness of male and female rats or mice to the odor of sexually mature individuals of the opposite sex is linked to these same gonadal hormones (Davies and Bellamy, 1972; Caroom and Bronson, 1971).

Females of some insect species, exemplified by cockroaches, have repeated reproductive cycles analogous to those of mammals. Pheromone production in these insects has been found to be under the control of the juvenile hormone, produced in the corpora allata glands. On the other hand, hormonal control has not been implicated in pheromone production by some species of moths that are short-lived as adults and mate and lay their eggs within the first few days of adult life (Barth and Lester, 1973).

An additional physiological mechanism often controls the time of day at which pheromone communication between the sexes takes place. Most animals have their sexual activities compartmentalized into a time of day that is characteristic for a given species. In a number of insect species, the timing of both the release of pheromone by one sex and the maximal responsiveness to the pheromone by the opposite sex is controlled by a circadian rhythm that is entrained by the animal's previous exposure to the alternating cycles of light and dark during each twenty-four-hour day (Sower et al., 1970; Traynier, 1970).

Environmental Influences

Temperature, light intensity, wind velocity, and the nature of the surrounding vegetation play an important role in regulating pheromone communication between males and females of many animal species. Suboptimal levels of these factors may lead to greatly reduced communication efficiency or even to the abolishment of pheromone communication. A few examples, based on insects, will be given below.

In quantitative studies of moth species that mate at night, it has been found that a light intensity higher than that of full moonlight inhibits the tendency of females to release their pheromone (Sower et al., 1972) and of males to respond to the pheromone (Shorey and Gaston, 1964).

Wind velocity has multiple effects. Too low a velocity may create conditions in which a stable aerial trail of odor molecules cannot form, and too high a velocity may impede upwind orientation by responding flying insects. Thus, there are distinct upper and lower limits for many species within which successful communication can take place (Shorey, 1974). Females of certain moth species apparently can detect whether the air velocity is appropriate for pheromone communication. Females of the cabbage looper moth reduce their emission of pheromone when the velocity approaches the lower and upper limits of zero and four meters per second, respectively (Kaae and Shorey, 1972). The potential distance over which pheromone communication can occur is also profoundly affected by air velocity. The active space of the aerial pheromone trail, i.e., the volume within which the pheromone molecular density is above the behavioral threshold of the responder, is greatly reduced as the air velocity is increased (Bossert and Wilson, 1963). Using data concerning the release rate of pheromone by females, the average dilution of the pheromone molecules as they move downwind, and the threshold of pheromone molecular concentration needed to stimulate male responses, Sower et al. (1973) calculated that the theoretical maximum pheromone communication distance for the cabbage looper moth is approximately 200 meters at a wind velocity of 0.3 meter per second and 50 meters at a velocity of 3 meters per second.

Some insect species are relatively monophagous, and it would appear adaptive for mating to take place on the same host plant species that is suitable for egg deposition and larval development. Thus, in certain moth species, either the females release their pheromone only when they sense the odor of the appropriate host plant (Riddiford, 1967) or male attraction to the pheromone is potentiated by the odor of the appropriate host plant (Brader-Breukel, 1969).

The spatial zone within which pheromone communication takes place may be influenced by the surrounding vegetation. For example, male moths are often best attracted to female pheromone sources that are located near the level of the top of the vegetative surface, whether the vegetation is cabbages or forest trees (Kaae and Shorey, 1973; Miller and McDougall, 1973); presumably, this is the elevation at which evolutionary selective processes have caused the males to search for female-produced odor trails.

Reproductive Isolation

Reproductive isolation among closely related animal species is often achieved by mechanisms that prevent individuals of one species from responding to the pre-mating communication signals emitted by the opposite sex of another species. The mechanisms may be related to the specificity of the signal itself, with the context of the message emitted causing appropriate responses by members of the opposite sex of the same species only. Or, alternatively, an identical communication signal might be used by more than one species, but the various species might be isolated geographically or according to different habitats or by different seasons or times of day during which pre-mating communication occurs. With regard to pheromones, these mechanisms for maintenance of reproductive isolation have been most studied in the moths, and the following discussion will be restricted to this group of animals.

The pheromones emitted by females of most moth species attract only males of the same species. There are, however, some exceptions to this statement (Barth, 1937; Götz, 1951), in which reproductive isolation mechanisms that operate at close range, after the male has been attracted to the vicinity of the wrong female, might be expected to operate. Until recently, most investigators believed that most female moths attract males of their own species only because of the utilization of a unique species-specific chemical in the pheromone communication of each species. To a certain extent, this molecular specificity hypothesis appears to be correct. Typically, any minor modification in the molecular structure of a sex pheromone results in a great loss of biological activity (Gaston et al., 1972; Schneider, 1967).

As more sex pheromones are identified, it becomes apparent that an identical chemical may often be produced by the females of a number of different species. Various mechanisms are apparently used by those species that "share" a common pheromone chemical to ensure the responsiveness of males to females of their own species only. One of these mechanisms is concentration specificity (Kaae et al., 1973a; Klun and Robinson, 1972). For instance, females of both the cabbage looper moth and the alfalfa looper moth apparently produce cis-7-dodecenyl acetate as their principal pheromone chemical. High release rates of this chemical, equivalent to those released by living cabbage looper females, attract primarily males of that species, while low release rates, apparently equivalent to those released by living alfalfa looper females, attract primarily males of that species.

For some closely related moth species, the qualitative, as well as the quantitative, blend of two or more chemicals constituting the pheromone of each species ensures species-specific responses. Females of more than one species may produce an identical chemical, attractive to males of all the species involved; but the females of each species may release additional chemicals in a pheromone blend. Some of the chemicals may potentiate the activity of the blend for males of the correct species, and others may cause an inhibition of responsiveness of males of the related species (Comeau and Roelofs, 1973; Klun et al., 1973). For a number of species, the same chemical both potentiates the activity for males of the correct species and inhibits responsiveness of males of the incorrect species. Many variations on this general theme are unfolding. In some cases, no single chemical in a pheromone blend will attract males when evaporated into the environment by itself. The multiple components of the blend are essential for biological activity, and specificity might be obtained by the related species' utilizing different, characteristic proportions of identical multiple components (Tamaki et al., 1973).

As mentioned earlier, some related species that are potentially cross-attractive may achieve reproductive isolation through spatial or temporal separation. For example, the circadian rhythms of timing of pheromone emission by females and of maximal pheromone-responsiveness by males are typically characteristic for a given species and often cause communication to be compartmentalized into a time of day that is different from that of a related species (Comeau and Roelofs, 1973; Kaae et al., 1973b).

STIMULATION OF AGGRESSION

Two general categories are recognized in which the pheromone released by one animal stimulates aggressive behavior by others of the species: the aggression may be directed either toward the pheromone-releasing animal itself or toward an animal of another species that is designated in some way by the pheromone emitter.

Aggression Directed toward a Conspecific

Among the vertebrates, dominant individuals in a social hierarchy frequently react to the pheromone released by subordinates by exhibiting aggressive behavior. Likewise, members of a social group that defend a particular territory may be stimulated to attack intruders when they detect a pheromone emitted by the strange individuals.

Pheromone-induced aggression against con- specifics has been best studied in mice, which establish territories as well as social hierarchies. The odor of a strange (to the social group) male increases the aggressive behavior of other males (Haug, 1971; Mackintosh and Grant, 1966). One pheromone is incorporated in the urine, and its production is under the control of androgen; thus, castrated males receive less aggression than do normal males or castrated males treated with androgen (Lee and Brake, 1972). Mugford and Nowell (1971) found that male mice also produce an aggression-promoting pheromone in the preputial glands; these glands apparently secrete the material directly to the exterior.

The simulation of male aggression by the odor of a strange male has been described for a number of other mammalian species, including guinea pigs (Beauchamp, 1973) and the rabbit (Mykytowycz, 1968).

Although a pheromone associated with the urine of female mice inhibits aggression by males (see below), the same or another urine- associated pheromone incites other females to respond aggressively to a strange female (Haug, 1972).

Pheromones are used in aggressive encounters among fish of certain species. The blind goby lives in burrows made by the shrimp Cal- lianassa affinis (MacGinitie, 1939). A male and female remain paired for life. They recognize gobies invading their burrow by means of a pheromone. The resident male will fight to the death with an invading male, as will the resident female with an invading female. The yellow bullhead exhibits an elaborate social behavior, establishing both territories and hierarchies within its territories (Bardach and Todd, 1970). The pheromone from a strange fish invariably elicits an attack response from the occupier of a territory, and the pheromone from a subordinate fish elicits attack from a more dominant fish within a hierarchy.

Aggression against conspecifics is also stimulated by pheromones in certain social insect species. Honeybee colonies typically contain a single queen. If more than one queen is present, olfactory stimuli emanating from each of them stimulate aggression by the others (Riedel, 1972). Ants generally defend their nest and the area around it from conspecific ants of other colonies. In Formica fusca, the stimulus evoking aggression against foreign ants is believed to be an odor that is distinct for each colony (Wallis, 1963).

In some cases, the release of a pheromone may constitute aggression rather than incite it. Perception of a pheromone may in such cases lead to intimidation of the perceiving animal and possibly stimulate it to flee. Thus, Ralls (1971) observed that many mammals increase their frequency of marking the environment with a pheromone when they are intolerant of and dominant to other members of the same species and when they are likely to win if they attack. Such a situation occurs in connection with territoriality, but it also occurs in other social situations. Especially frequent marking occurs when there is reason to infer that the animal is motivated to aggression. An example is the "stink fights" that occur between males of the ring tailed lemur (Evans and Goy, 1968). Two males in a tree, engaged in aggressive interaction, direct pheromone odors toward each other from a variety of sources. A pheromone secreted by wrist glands is rubbed onto the tail. The males then wave their tails toward each other. In addition, a volatile pheromone is deposited on branches from the palms (palmar marking). First one male palmar marks and then the other, with pauses between. The more aggressive male moves forward and the other retreats. Also, the more aggressive male palmar marks branches that the other male has marked. It appears likely that the pheromone signals interact with other stimuli produced by the males in symbolizing their state of aggressiveness.

The use of pheromones as an aggressive stimulus and to advertise dominance is probably very common among mammals. The advertising is often done even in the absence of an antagonist, with the pheromone being deposited on objects in the environment from specialized skin glands or from glands associated with urine or feces. Circumstantial evidence for this role of pheromones is obtained from numerous morphological and behavioral observations. The largest glands and the highest rates of secretion of pheromones, as well as the highest frequencies of marking objects with a pheromone, are found in the dominant individuals in social hierarchies (Bronson and Marsden, 1973; Müller Schwarze, 1972; Mykytowycz, 1968).

Finally, some additional comments are necessary concerning the inhibition of aggression among mammals. As mentioned earlier, the aggressive tendencies of animals that are foreign to a territory or subordinate within a social hierarchy are likely to be inhibited by the pheromone of a territory occupier or a more dominant animal, respectively. Also, animals that are familiar with each other and that coexist within a society are typically inhibited from aggression toward one another, with males being especially inhibited from exhibiting aggression toward females. This inhibition has been most studied in mice, where it has been found to be attributable, at least in part, to pheromones released by the various individuals (Dixon and Mackintosh, 1971; Haug, 1971; Mugford and Nowell, 1971).

Aggression Directed toward an Individual of Another Species

Pheromones that are used to elicit aggressive behavior of conspecifics toward an individual of another species are commonly found in the social Hymenoptera. Sometimes the biological function is the stimulation of colony mates to attack a prey designated by the pheromone or to plunder the colony of another species. This function was considered earlier. In other cases the pheromone, now referred to as an alarm pheromone, is involved with defense of the nest and its surroundings from attack by an enemy.

According to Maschwitz (1966), the glands in hymenopterous species that produce alarm pheromones are invariably associated with the organs of defense. Thus, the pheromones are secreted by various glands that occur in the vicinity of the mandibles or the sting. In fact, a number of ant and bee species have multiple glands that may secrete pheromones separately or simultaneously to incite defensive behavior by conspecifics.

The alarm pheromone often has a dual function, not only to attract colony mates and lower their threshold for attack behavior but also to act as a poisonous or repellent chemical that deters the enemy (Bergström and Lofqvist, 1973; Wilson and Bossert, 1963). The behavior stimulated by alarm pheromones may be arranged in a hierarchal sequence, with each behavioral step requiring a higher concentration. Thus, Moser et al. (1968) found that workers of the Texas leaf- cutting ant, when exposed to a low concentration of the pheromone 4-methyl-3-heptanone, raise their heads and antennae. At higher concentrations, they are stimulated to follow the molecular gradient to its source, and at still higher concentrations to become very active and to open their mandibles.

A honeybee worker that is disturbed at the hive releases an alarm pheromone, isopentyl acetate, from its sting chamber. Other bees are attracted by the pheromone and are stimulated to fly and to attack objects in the vicinity of the hive (Boch and Shearer, 1965). The attack behavior is guided by visual cues from the enemy (Maschwitz, 1966). Upon stinging the enemy, the recruited bees release isopentyl acetate from their sting glands and another pheromone, 2-heptanone, from their mandibular glands (Shearer and Boch, 1965). These pheromones stimulate and direct the attack of additional bees and are probably responsible for the phenomenon well known to beekeepers that more than one bee will often sting in the same spot.

FAMILIARIZATION WITH HOME RANGE

Mammals of many species deposit scent marks from specialized skin glands or from glands that are associated with urine or feces as they move about within their territories. This marking behavior is also seen in many species that do not defend well-defined territories but that have a home range within which they normally confine their activities. The precise role of the marking behavior has been little studied. The behavior may, in fact, serve a variety of interacting functions, which are listed in the following section. A number of investigators feel that a very important function served by scent marking in many species is the maintenance of the animal's familiarity with its home-range environment (Mykytowycz, 1970; Ralls, 1971). Thus, many mammals become especially active in marking objects with scent when they are displaced into a strange environment (Goddard, 1967; Martin, 1968).

As well as "reassuring" the animal that it is in a familiar environment, the pheromone scent marks may aid in orienting its activities in that environment (Goddard, 1967; Wynne-Edwards, 1962). Many mammals habitually follow the same paths as they move about within their home range (Wynne-Edwards, 1962); scent is deposited along the trails, sometimes by glands on the feet (as in ruminants: Bourlière, 1954) and sometimes by other methods. The black rhinoceros follows fecal scent trails (Goddard, 1967). The rhinos deposit feces in piles, which are located randomly over the home range. Any one pile may be used by a number of individuals. Before defecating, they sniff the pile and may sweep it with the anterior horn and shuffle through it with the feet. After defecating, they kick at the pile with their hind feet. The odor is apparently distributed by the feet as the rhino moves through its home range, creating a trail that can be followed by others of the species.

COMPLEX SOCIAL BEHAVIOR

Many of the behavioral activities of animals that live in organized societies appear to be regulated by pheromones. There have been few critical studies of this behavior. Mainly anecdotal evidence indicates that, within mammalian societies, pheromones often relate information concerning individual identity, group membership identity, age, social status, sex, and reproductive state. Also, pheromones often are part of a stimulus complex associated with "greeting" between animals, submission, dominance, attention seeking, gregariousness, signaling of danger or distress, trail following, territorial behavior, and identification of home range (Mykytowycz, 1970).

Likewise, many of the behavioral activities of social insects appear to be regulated by pheromones. Some of the more poorly understood behaviors are probably caused by the so-called surface pheromones, which may be absorbed on the body surfaces and be detected on contact or at extremely close range. These pheromones include the colony odors, the caste-recognition scents, and the releasers of grooming and food exchange (Wilson, 1971). The regulation of insect social organization is probably due in large part to pheromones released by the queen. Workers of a number of species of bees, ants, and wasps engage in "retinue" behavior. They encircle the queen, lick her body, and touch her with their antennae (Gary, 1970; Wilson, 1963b). The workers apparently are stimulated to engage in retinue behavior by pheromones secreted by the queen, and through this behavior they probably obtain from the queen the primer pheromones that inhibit development of their ovaries and control other aspects of their physiology and behavior.

Evolution of Pheromone Communication

Pheromone communication must have arisen very early during the evolution of primitive plants and animals. Indeed, the chemicals used for signaling between the cells of multicellular animals—the hormones and synaptic transmitter substances—probably evolved from the chemicals used for communication between individual free-living cells (Haldane, 1954). After all, a multicellular organism is an aggregation of cells designed to allow intercellular signaling, based mainly on chemicals. Subsequent evolutionary processes led to the elaboration of specialized groups of cells constituting glands for the synthesis and release of pheromones, as well as specialized sensory devices for the perception of pheromones; these developments enabled communication between multicellular animals. The course that this evolution has followed must vary considerably among different animal groups.

Similarities between the chemicals used as pheromones and as hormones by many animals are striking. Terpenoid or steroid compounds are used for these functions by a great variety of vertebrate and invertebrate species, with the same chemical sometimes having both functions in a given species. In various crab species, the process of evolution of hormonal systems from primitive pheromonal systems appears to have reversed (Kittredge et al., 1971). The female crabs release a sex pheromone into the water shortly before they molt. Apparently, the molting hormone, crustecdysone, functions also as the sex pheromone. Evolutionary processes probably led to relatively minor modifications of a preexisting hormone system to fulfil the pheromone communication function. These modifications included elaboration of a system for the release of the chemical into the external environment and of an external system on the antennules of the males for detection of the chemical.

Most pheromones have, perhaps, been derived by natural selection from metabolites that were initially produced for some other function. In mammals, the integumental glands that originally supplied wax or mucus to the skin have probably often been elaborated in this way to produce pheromones that are used for a variety of communication functions (Wynne-Edwards, 1962).

The aggregation pheromones released by a number of bark beetle species are remarkably similar to the terpenoid resins of the host trees that they attack (Hughes, 1973). The host terpenoids are in some cases attractive to the beetles, and attraction to the host may have been the only aggregative factor in primitive bark beetles. It seems reasonable that various beetle species later evolved mechanisms whereby metabolites of the host resins (thus, the resins modified chemically within the insects) stimulated aggregation of conspecifics. Similarly, Shorey et al. (1969) noted that males of two species of fly that congregate on certain food materials are stimulated by the odor of the food to court and attempt copulation with nearby flies. This response to a host odor may represent the primitive situation. Later stages of evolution may have resulted in metabolites of the food chemicals, produced by the insects themselves, fulfilling the sexual communication function. As pointed out by Moore (1967), the animals could now be independent of the original food or habitat odor. Further evolution of the scent-producing and receptor mechanisms would lead to the pheromones, often highly specific in terms of species as well as in their biological effect, that are characteristic of many animals today.