Ants have played and play a very important role in a number of ecosystem around the world. They are geographically the most widely distributed of the eusocial groups, ranging over virtually all the land outside the polar regions. They are also numerically the most abundant (Wilson, 1980). They have a very sophisticated social organisation and employ the most complex forms of chemical communication of any animals. The diversity of the ants is substantial, far exceeding that of the other social insects and reflecting the manner in which ant species have evolved to saturate a wide range of feeding niches in the soil and vegetation (Hölldobler & Wilson, 1990). This makes the ants a very interesting object of study, both from an ecological and an evolutionary view. I have chosen to study the recruitment behaviour in ants, because this communication behaviour shows several levels of sophistication in the ants, which may be paralleled to an evolutionary approach of social behaviour.

My purpose with this study is to examine whether different recruitment behaviours can be correlated directly to the phylogenetic evolution of the ants (Figure 1), i. e. if the more so called primitive species also have a less sophisticated recruitment behaviour.

The mapping of the phylogenetic tree is not completely unambiguous. There are several slightly different versions of the tree. According to Hölldobler and Wilson (1990), the version shown in Figure 1 is the correct one.

In this study I have concentrated on recruitment behaviour to food, but I also discuss migration behaviour, as this in many cases is closely related to food supply. I have excluded raiding behaviour in ants, as this is a very extensive area, and concentrated on the behaviour the ants perform when they have found a new food source (or nest site) and return to the nest to recruit other ants.

To clarify what social insects really are, I have included a definition of the group according to Wilson (1971).

The "truly" social insects, or eusocial insects as they are sometimes more technically labelled, include ants, all termites, and the more highly organised bees and wasps. These insects can be distinguished as a group by their common possession of three traits: individuals of the same species cooperate in caring for the young; there is a reproductive division of labour, with more or less sterile individuals working on behalf of fecund individuals; and there is an overlap of at least two generations in life stages capable of contributing to colony labour, so that offspring assist parents during some period of their life (Wilson, 1971).

Unless otherwise stated, the contents of this chapter mainly follows Hölldobler and Wilson (1990).

In the majority of cases the origin of communicative systems in animals is based on ritualization, the evolutionary process by which a phenotypic trait is altered to serve more efficiently as a signal. Commonly, the process begins when some movement, anatomical feature, or physiological process that is functional in another context acquires a secondary value as a signal. During ritualization movements (or odours, or visual features) are altered in a way that makes their communicative function still more effective. They acquire support in the form of additionally anatomic structures or biochemical changes that enhance the distinctiveness of the signal. The movements also tend to become stereotyped and exaggerated in form. Finally, the receiving apparatus is modified to detect such ritualised signals with less ambiguity. In the case of trail systems among ants, the chemoreceptors have been modified to detect minute traces of the appropriate pheromone, which often occur in nanogram or even femtogram amounts.

Ritualization can be found in many kinds of signalling: tactile, acqustical and chemical. Chemical alarm communication evidently evolved from chemical defence behaviour. Ants and other social insects use chemical secretions to repel predators and other enemies. In social insects, however, defensive reactions are closely linked with alarm communication, and quite often a single substance serves both functions. Most ant species engage in some form of alarm-defence, in which the same chemicals are used to repel enemies and alert nestmates. A second intergradient category that has been documented is alarm-recruitment. Alarm signals both alarm and attract in some species, while in others the alarm pheromones are combined with odour trails that lead nestmates to or from the source of danger. Many species employ a single alarm-recruitment procedure to alert nestmates to both enemies and prey, and in fact the distinction between the two may be wholly blurred with reference to communication. The signals (chemical and others) used in these behaviours are often the same as the ones used in other recruitment behaviours. The step between alarm communication and communication to food seems to be a relatively short one.

Modulatory communication

Communication in complex social systems is seldom characterised by a direct, all-or-nothing response. Signals do not always merely "release" behavioural responses of a particular kind, but instead often appear to adjust the behaviour of nestmates toward one another in a manner appropriate to the surrounding environment. In such instances the communication may have relatively low informational content when measured simply by the number of "bits" transferred. According to this interpretation, outwardly inefficient communication systems serve different but no less important purposes than more direct, deterministic systems. They influence the behaviour of receivers, not by forcing them into narrowly defined behavioural channels but by slightly shifting the probability of the performances of other behavioural acts. In this way the degree of accuracy and sophistication of different behaviours can be increased. Modulatory communication can be expected to be more frequent in the most complex animal societies, where many members perform many different tasks at the same time. In this arrangement, a flexible program is required if the work force is to distribute its energy investment among the different tasks in an effective manner. Modulatory communication appears to be a primitive phenomenon in ants and other social insects, e.g. it exist in almost all ants species and probably evolved early in their evolution.

Unless otherwise stated, the contents of this chapter mainly follows Hölldobler and Wilson (1990).

The use of vibrational signals is weakly developed in ants in comparison with communication by pheromones. It often occurs in conjunction with chemical signals. Most but probably not all acoustical signals are transmitted primarily through the soil, nest wall, or some other solid substratum rather than through the air. Two forms of sound projection have been identified, body rapping against the substratum, and stridulation (rubbing one body part against the surface of another).

Several independent investigators have concluded that tactile signal serve several communicative functions, but also, despite their outward complexity, convey only a limited amount of information. Nevertheless, touch by the antenna and forelegs have been firmly implicated in invitation behaviours that entrain most forms of recruitment, e. g. in tandem running.

The use of visual signals in ants is at best very minor and in fact not a single example has yet been solidly documented. This is not the outcome of physical constraint. Many large-eyed ants have excellent vision and are very good at detecting moving objects. In some species, workers can be alerted as they see other ants, but as this is done, the alerting ant also releases chemical substances, so it cannot be proved that the stimuli was sight alone.

Chemical signals is by far the most used form of communication in ants. It is used in all kinds of communication they have. The ants have several different glands with which to excrete different kinds of chemical substances, and have very sensitive chemoreceptors to receive those signals. In this study, most of the substances discussed will be pheromones. A pheromone is usually a glandular secretion, used between individuals within the same species.

Unless otherwise stated, the contents of this chapter mainly follows Wilson (1971).

Recruitment is defined as a communication behaviour that brings nestmates to some point in space where work is required. The category of recruitment is a very loose one, and in particular cases it often cannot be clearly distinguished from alarm and simple assembly. Some ants mixes these functions or have developed intermediate behaviours. When you compare different recruitment behaviours according to speed, experiments show that tandem recruitment is the slowest, mass recruitment the fastest and group recruitment intermediate (Deneubourg, 1983).

Visual communication (recruitment)

Large-eyed ants have excellent form vision and are especially keen at detecting moving objects. Workers generally do not respond to prey insects that are standing still, but run towards them when they start to move. In some species, when a lone worker encounters a prey insect, they dash in erratic circles around it and thereby attract other workers exploring in the close vicinity (Stäger, 1931). Sometimes the sight of a running worker alone is enough to set another worker running. This behaviour is as written above not certain just to involve visual communication, and perhaps it is just a side effect of the chemical communication.

A simple form of chemical recruitment is when an individual ant discharges a alarm pheromone that simply attract other ants in the vicinity. This might happen when an ant attacks a prey so large that it cannot handle it alone. The pheromone can be the same as the one used in alarm signals. This recruitment behaviour might consequently be a felicitous by-product of the alarm system.

A more sophisticated way to use chemical recruitment techniques is tandem running (Figure 2). When an ant finds a food item to large to carry, it runs home and fetches another ant to help it carry the object. The interaction follows a stereotyped sequence. This sequence may differ between different ant species but the main parts are the same. First the scout ant or leader ant makes contact with the follower ant by touching it or by being touched. Then it runs a short distance, coming to a complete halt. The follower ant, now in an excited state apparently due to a secretion released by the leader, runs swiftly behind, makes fresh contact, and drives the leader forward. Other workers approaching the leader is similarly excited, even if the latter is completely immobile at the time. After each contact and subsequent forward drive of the leader, the follower may press immediately behind and move it again until they reach the food find. Often, tandem running involves only two worker ants.

In group recruitment the recruiter lays a trail while returning from the food source to the nest, and is then needed to guide a group of recruits along the trail towards the food source chemically (Deneubourg, 1983) (Figure 3).

As a worker ant heads home after discovering a food source, it walks at a slower, more deliberate pace. It holds its body close to the ground and lays an odour trail consisting of pheromone released from the hind gut. While laying a trail, the worker sometimes loops back in the direction of the food find, but only for short distances, before turning nestward again. If another worker is contacted, the homebound worker turns towards it. It may do no more than rush against the encountered worker but sometimes (in some ants) the reaction is stronger: it climbs partly on top of the worker and, in some instances, shakes its body lightly but vigorously. The movement may represent an modulatory signal that enhances the effect of the pheromone, but it does not appear to impart any essential information about the food find, because contacted workers do not exhibit trail-following behaviour obviously different from those not contacted. Moreover, the pheromone is by itself sufficient to induce immediate and full trail following when laid down in artificial trails (Hölldobler & Wilson, 1990). Most workers encountering a freshly laid trail respond at once by following it outward from the nest. The workers do not follow a liquid odour trail on the ground. Instead they move through the vapour created by diffusion of the pheromone in the air. There is a space, within which the pheromone is detected by the ants. As follower workers travel through this "vapor tunnel", they sweep their antennae from side to side, evidently testing the air for odourant molecules.

The odour trail system is the most evolved of all communication systems in ants. Trail communication has evidently evolved, at least in some groups of ants, from tandem running. Intermediate stages exist, in where the leader ant does not halt and wait to be touched, while the follower drops behind several centimetres and seems to orient along a short-lived odour trail. There are also occasions in tandem running when several workers follow the leader, instead of just one. These intermediate stages are not at all rare in ants. In would seem to be only a short step in evolution from trail-guided procession where the followers must stay close behind the leader, to typical trail following, where followers are guided by odour alone over long distances in the absence of a trail layer.

Mass recruitment is when information are transmitted from one group of individuals to another group of individuals. There are several ways by which this could be achieved. One example is in Solenopsis saevissima (subfamily Myrmicinae), where the number of workers leaving the nest is controlled by the amount of trail substance being emitted by foragers already in the field. The number of individuals drawn outside the nest is a linear function of the amount of the substance presented to the colony as a whole. Under natural conditions this relation results in the adjustment of the outflow of workers to the level needed at the food source. The number of workers increase until it gets crowded and workers have to return without any food and consequently without laying a trail. As a result, the number of workers at food masses tends to stabilise at a level which is a linear function of the area of the food mass. The workers can also adjust their traillaying to the quality of the food find (Figure 4). This mass communication of quality is achieved by a "electorate" response in which the individuals choose whether to lay trails after inspecting the food find. The more desirable the food find, the higher the percentage of positive response, the greater the trail-laying effort by individuals, the more trail pheromone presented to the colony, and hence, the more the newcomer ants that emerge from the nest.

The mass recruitment mechanism probably includes the following components: (1) an ability of ants to detect the concentration of pheromone on the trail and to relate it to the attractiveness of the food; (2) modulation by individual ants of the amount of trail-pheromone released in accordance with the quality of the food and the concentration of pheromone already present on the trail; (3) an ability of the ants to measure the density of ants on a trail at a food source; and (4) an effect of the duration of starvation on the trail laying activity of the individuals and probably also on the responsiveness of the ants towards the trail pheromone. This would provide a highly sensitive recruitment system which would enable the ants to determine whether to follow a new trail or not, depending on their motivation to collect food from known source which they might already be exploiting, and depending on the concentration of trail pheromone on a new trail. This would enable the colony to concentrate on the best food available (Jaffe & Howse, 1979).

The mass communication system clearly represents an advanced evolutionary grade. In order to identify the more primitive recruitment behaviours from which this system may have evolved, less sophisticated modes of recruitment communication must be discovered and characterised. Tandem running, used variously during emigration and recruitment to food, is generally considered to be one such mode. Only a single nestmate is recruited at a time, and the follower has to stay in direct antennal contact with the leader. Comparative studies have revealed several strikingly convergent pathways in the evolution of mass communication. On the other hand, morphological and behavioural findings indicate that communication by chemical trails in ants is considerably more diverse overall than researcher had assumed. Ten different anatomical structures have been identified in various species of ants as sources of trail pheromones. It is obvious that trail communication has evolved several times independently. Even in the same subfamily the mechanisms and anatomical structures for trail communication have diverged considerably (Hölldobler & Wilson, 1990).

One of the methods of recruitment used by ants is extremely direct: worker are merely dragged or carried to the target area by their nestmates (Figure 5). This behaviour is almost wholly limited to emigrations to new nest sites. It is generally true among ant species that, if a colony is forced to move from one nest site to another, the most active and experienced workers carry the brood and, especially during the later stages of the emigration, pull or bodily carry other adult members to the new nest site.

In the higher ants, adult transport has evolved into an elaborate, stereotyped form of communication. There are also many variations of this behaviour, even among the subfamilies. Because of this, it is evident that adult transport has evolved considerably within the ants and shows at least a rough correspondence with the principle taxonomic groupings. The behaviour pattern is nevertheless not found in every ant species.

Among the primitive ants (e.g. Myrmecia and some ponerines), workers usually transport only aged and ailing individuals, callow workers, nest queens, and males. The transported

individual also does not cooperate in any way by folding the appendages or suchlike and is simply dragged along. In more advanced species, the carried ant folds his or her appendages in a "pupal position" that makes the ant easier to carry (Hölldobler & Wilson, 1990).

The basic transporting behaviour has been adapted to new ends by a few ants species. In some ants (e.g. Manica rubida) it is used to remove alien workers from the colony territory. Interestingly enough, the subdued aliens respond with the same submissive behaviour as nestmates.

When colonies move from one nest sit to another, the new site is chosen by scout workers which then lay odour trail back to the old nest. The pheromone used in this behaviour is very often the same as in the odour trails that lead to a food find. Other workers are drawn out by the pheromone. They investigate the site and, if satisfied, add their own pheromone to the trail, much as in mass recruiting. In this fashion the number of workers travelling back and forth builds up exponentially. In time the brood is transferred, the queen walks over, and the emigration is complete.

The modes of transport in emigration are based upon some of the most highly evolved communicative techniques in the ants, and they vary greatly among different species.

One interesting thing is that the complex communication mediating colony emigration is more widespread among the phylogenetic groups than is recruitment to food. It also appears to be more primitive, in other words precedent to the latter behaviour, since it occur in some primitive species (e. g. Myrmecia) unaccompanied by any form of food recruitment. Emigration is usually organised by a minority of the workers, who are either "elite", that is, individuals who work hard at many tasks, or moving specialists (Hölldobler & Wilson, 1990).

The trail pheromone used in emigration, through the mass effect, provides a control that is more complex than might have been assumed from knowledge of the relatively elementary individual response alone (Hölldobler & Wilson, 1990).

The contents of this chapter mainly follows Hölldobler & Wilson (1990).

A number of studies on ants have made it clear that motor displays and tactile signalling play an important role during recruitment communication in many ants species. It appears, however, that during the course of evolution these signals became less important with the increasing sophistication of the chemical recruitment system. This may in part have to do with colony size. The larger the mature colony size among ant species, the more reliance is placed on trail pheromone as opposed to motor displays in the initiation of recruitment. Also, chemical trails become more important than tandem running in the orientation of the follower ants.

It is further evident that at least the Formicinae ants evolved chemical trails by means of ritualization of defecation. A comparative study has revealed that in many species workers do not defecate randomly but visit specific locations preferentially. Beside certain sites within the nest, they favour nest borders, "garbage dumps", and the border of trunk trails that lead to permanent food sources or connect multiple nest entrances. We may suppose that in the Formicinae hindgut material first became an important cue in home range orientation and then was transformed into a more specific orienting and stimulating signal used during recruitment.

Another trend of increasing sophistication among the ants generally is the use of a mix of chemical components, each of which serves a different role in the recruitment process.

Yet another trend toward complexity is to vary the signals given and their intensity in order to produce graded messages. Workers in ant species like Solenopsis invicta and Formica oreas adjust the amount of pheromone in their odour trails according to the quality of the food find.

Unless otherwise stated, the contents of this chapter mainly follows Hölldobler and Wilson (1990).

The recruitment system of Camponotus sericeus exemplifies the more elementary evolutionary grade of tandem running in the Formicinae (Figure 2). The first scouting ant typically fills her crop and returns to the nest, touching the tip of her abdomen to the ground for short intervals. As she does this, she deposits chemical signposts with material from her hindgut. Inside the nest she performs short-lasting fast runs and contacts several nestmates. When the ant leaves the nest, most of the encountered ants try to follow. However, only the ant keeping the closest antennal contact succeeds in accompanying her out of the nest and to the food source. Most of the followers who reach the food source in this way soon return to the nest and recruit other ants. Experiments have shown that the hindgut trail laid by homing scouts has no recruiting effect by itself. Only experienced ants follow the trail, and then exclusively for orientation. Similarly, during the tandem running the trail pheromone appears to serve no significant communicative function. The leader and follower are bound together by tactile signals and pheromones on the surface of the body.

The next higher organisational level of recruitment is group recruitment. This behaviour has been observed in several Camponotus species e.g. Camponuotus socius (Figure 3). Scouts in this species use chemical signposts around newly discovered food sources and lay a trail with hindgut contents from the food source to the nest. The trail pheromone alone, however, does not induce recruitment to any significant extent. Inside the nest the recruiting ant performs a waggle display when facing nestmates head-on. Nestmates are alerted to this and follow the recruiter to the food source. Only workers stimulated in this way follow the recruiter along a laid trail. The presence of the leader ant is still essential for a complete performance, however. Similar behavioural patterns are employed during recruitment to new nest sites. A difference here is that males also respond to the signals.

The next level is represented by species in which workers stimulated by a motor display follow the trail to the food source even in the absence of the recruiting ant. In Formica fusca, the hindgut trail an successful scout lays to the nest has no primary stimulating effect. After the scout has performed a vigorous waggle display inside the nest, nestmates rush out and follow the trail to the food source without additional cues from the recruiter.

Camponotus pennsylvanicus represents the next level. Here too, the nestmates can be alerted by the display and follow the trail. However, workers encountering the odour trail follow it even without being mechanically stimulated by the scout ant. Nevertheless, the number of ants responding is higher if a scout is first allowed to stimulate her nestmates with motor displays.

Finally, from the C. pennsylvanicus level it is only a short step to chemical mass communication of the Solenopsis kind, where the pheromone is the overwhelming prevalent recruitment signal and the outflow of foragers is controlled by the amount of pheromone discharged.

In this connection it is also interesting to mention the African weaver ant, Oecophylla longnoda, which utilise no less than five recruitment systems to draw nestmates from the nest to the foraging area. The multiple recruitment systems of these ants are the most complex thus far discovered in the ants. They practice e.g. mass recruitment and have a very high level of sophistication and accuracy in their supply of information (Hölldobler & Wilson, 1978).

Many species of Ponerinae hunt alone. There is no co-operation among these foragers, either through the transfer of information about the location of new sources of prey, or through direct assistance during the killing and retrieving of prey (Peeters, 1997). However, this is not true about all ponerines. Some ponerine ants use elaborate systems of recruitment and co-operative hunting. The occurrence of simple or complex hunting strategies does not reflect phylogenetic relationships (Peeters & Crewe, 1987). Colony size (together with ecological considerations and the extent of prey specificity) seems to be an important factor. Ponerine species with small colonies are not likely to raid in groups. In Amblypone, species that do not recruit have colonies with only one or two dozen workers, but one species from the reclinata group, which is able to retrieve prey co-operately, has 100 workers per colony (Ito, 1993). Similarly, the entire spectrum of foraging strategies is exhibited in Leptogenys sp.. The large colonies tend to be group or swarmraiders, whereas the other species search for prey solitary, and, after encountering prey, recruit a group of nestmates of retrieve the prey alone (Maschwitz & Schönegge, 1983).

In Leptogynes chinensis single workers lay a light trail while foraging. Successful scouts return to the nest and perform recruiting runs. By this means groups of workers are recruited and led to a feeding place by a scout. In experiments, when the scout has been removed such group will still, though slightly disturbed, find their way to the prey (Maschwitz & Schönegge, 1983).

In the group recruitment of Amblypone sp., scout workers can stimulate nest workers without direct contact, and the recruited workers can trace the trail and arrive in the target area without the guidance of the scout worker. Often, however, the scout worker lead the group of recruited ants to the food source (Ito, 1993). This is also the case in Megaponera foetens, in which the scout ant leads a recruited group to the prey (Longhurst & Howse, 1979).

The Ponerinae is a phylogenetically ancient taxon with many species and various morphological types (Maschwitz & Schönegge, 1983). Although many of the ponerines tend to have simpler form of recruitment, like tandem running or simple group recruitment, the ponerines can perform anything from simple recruitment by just chemically "calling" other nestmates in the surroundings to (in some species) mass recruitment (Baroni Urbani, 1993).

Baroni Urbani formulates three hypotheses that have been discussed in the scientific world concerning the evolution of the ants recruitment of nestmates to food or to a new nest site. The three hypotheses are the following: Hypothesis 1: Increasingly efficient recruitment behaviours exhibited by different ant species have been shaped by or are correlated with ant phylogeny. Hypothesis 2: Increasingly efficient recruitment behaviours represent necessary evolutionary steps independently followed during the evolution of different ant clads. Hypothesis 3: Different efficient recruitment behaviours have been selected in a convergent way among different species by similar populations/environmental constraints. The analysis made by Baroni shows that differences in recruitment behaviour do not directly reflect phylogeny, but rather represents species-level adaptations to environmental conditions (Baroni Urbani, 1993). This conclusion is supported by the fact that the morphological differences (e.g. what gland are used to produce the pheromone) between species that uses very similarly recruitment behaviours, are greater than they should have been if the behaviour had evolved in direct correlation to ant phylogeny. In fact, trail communication seem to have evolved several times independently in the ants (Wilson, 1971). Several other studies also support this view. Deneubourg (1983) discusses the differences between accuracy in ant species to recruit, that is, how many of the recruited ants that actually reach the food source and return food to the nest. He found that according to the degree of accuracy, recruitment strategies cannot be ranked along a linear scale of evolution, but their efficiency is strongly dependent of environmental and social parameters (Deneubourg, 1983).

In my comparative study, I have found that although the more sophisticated behaviours seem more common in the subfamily Formicinae than in Ponerinae, the full scale of the behaviours exist in both groups. This leads to the conclusion that even if the recruitment behaviour evolved from the more simple chemical communications, to tandem communication, to group communication and finally to mass communication, there is no indication that this could be directly traced to the level of "primitivness" in the ants, i.e. to their level in the phylogenetic tree.

All this seem to lead to the conclusion that the phylogenetic tree alone cannot explain why different ant species behave like they do. In fact environmental and social factors seem to have had a much larger impact on the evolution of recruitment behaviour. According to optimality theory, each selected foraging strategy represents the best achievable balance of costs and benefits that maximises the net energetic yield to the colony (Detrain & Deneubourg, 1995). If some ants live in an environment where the food comes in large patches that disappear quickly, they should evolve a system in which a scout can recruit many nestmates at once to best exploit the food source. If food sources are scattered in many small spots, they should retrieve the food individually or in small groups. As a matter of fact, this is the pattern found in several studies, e.g. Deneubourg, Detrain. Detrain & Deneubourg (1995) found in their study of Pheidole pallidula (subfamily Myrmicinae) that large individual prey induced a more intense response in the ants than a small pile of scattered prey. From their findings, they were able to draw a model over the decision-making process in ants during prey scavenging (Figure 6).

Success in prey-carrying encourage the worker to move on and lay a weak trail on its way back to the nest. This could lead to a slow and progressive monopolisation of the source when small food items are numerous. Failure to retrieve the prey item results in a shortened stay at the food source followed by an intense trail recruitment. Recruited ants then gather around the food and try to retrieve it collectively. If they are unsuccessful, they tear up the large prey item and suck its haemolymph. According to this algorithm, the outcome at the colony level of the best suited recruiting patterns results from crude estimates of prey size and simple decision rules at the individual level (Detrain & Deneubourg, 1995). Individual decisions based upon environmental factors here seem to play a major role in the choice of scavenging and recruiting strategies. This is a very interesting thing to consider, because in the end, all species, families, subfamilies or small groups consist of a number of individuals.

Baroni Urbani, C. (1993). The diversity and evolution of recruitment behaviour in ants, with a discussion of the usefulness of parsimony criteria in the reconstruction of evolutionary histories. Insectes Sociaux, 40: 233-260.

Deneubourg, J.L. (1983). Probabilistic Behaviour in Ants: A Strategy of Errors? Journal of Theoretical Biology, 105:259-271.

Detrain, Claire & Deneubourg, Jean-Louise (1997). Scavenging by Pheidole pallidula: a key for understanding decision-making systems in ants. Animal Behaviour, 53: 537-547.

Hölldobler, Bert & Wilson, Edward O. (1978). The Multiple Recruitment of the African Weaver Ant Oecophylla longinoda (Latreille) (Hymenoptera: Formicidae). Behavioural Ecology and Sociobiology, 3: 19-60.

Hölldobler, Bert & Wilson Edward O. (1990). The Ants. Springer-Verlag. United States of America.

Ito, F. (1993). Observation of group recruitment to prey in a primitive ponerine ant, Amblyopone sp. (reclinata group) (Hymenoptera: Formicidae). Insects Sociaux, 40: 163-167.

Jaffe, K. & Howse, P.E. (1979). The mass recruitment system of the leaf cutting ant, Atta cephalotes (L.). Animal Behaviour, 27: 930-939.

Longhurst, C. & Howse, P.E. (1979). Foraging, recruitment and emigration in Magaponera foetens (fab.) (Hymenoptera: Formicidae) from the Nigerian Guinea savanna. Insectes Sociaux, 26: 204-215.

Maschwitz, U. & Schönegge P. (1983). Forage communication, nest moving recruitment, and prey specialization in the oriental ponerine Leptogenys chinensis. Oecologia, 57: 175-182.

Peeters, Christian (1997). Morphological ´primitive` ants: comparative review of social characters, and the importance of queen-worker diphormism. In: The Evolution of Social Behaviour in Insects and Arachnids. Cambridge University Press. Cambridge.

Peeters, C. & Crewe, R. (1987). Foraging and recruitment in ponerine ants: solitary hunting in the queenless Ophthalmopone berthoudi (Hymenoptera: Formicidae). Psyche 94:201-214.

Wilson, Edward O. (1971). The Insect Societies. The Belknap Press of Harvard University Press. Cambrigde, Massachsetts.

Wilson, Edward O. (1980). Sociobiology. The Belknap Press of Harvard University Press. Cambridge, Massachusetts and London, England.