Insects face numerous obstacles throughout their development and life as an adult and their   biggest obstacle is avoiding predators.  There are a number of ways insects deal with predators: they may   have bright-distinctive colors or larval forms (caterpillars) that emit an offensive odor and some possess   stinging hairs (loaded with Oxalic Acid), and we already studied noctuid moths and their sensitivity to   ultrasound.  Many insects have developed anti-predator defenses blocking attempts to detect, attack,   capture, and consume them.

  Many Lepidopteran, Hymenopteran, and Dipteran insects form different types of ant-guards.    Some paper wasps, if they are nesting alone, rub the pedicel of the nest with a chemical ant repellent;  while owlfly females apply egg-like chemical defenses around a grass stalk providing protection for   their real eggs; some honeybees coat the branch leading to their nest with a sticky ant trap; and the larva  of  the Neotropical butterfly Eunica bechina (Nymphalidae) use frass chains where they rest immune from  predatory ants that climb the Caryocaraceae (Eunica bechina host plant) because of its nectary secretions.    If the Eunica bechina caterpillars are caught feeding on the leaves they will be killed and removed by   foraging ants.  There are other defenses these neotropical caterpillars use to fend off foraging ants.

 If these Neotropical caterpillars are attacked they can repel their aggressors by regurgitating or   bleeding a toxic substance from their body.  These substances cause ants to exhibit strange behavior.    After aggressive ants come into contact with these substances they are strongly disturbed and attentively  cleaned their mandibles and heads.  While the ants were preoccupied doing this, these caterpillars escaped,  maybe wounded, but still alive.  Under heavy attack by foraging ants, after subsequent bites, Eunica   bechina larva will drop lines/frass chains from a plant and hang on the end of a dragline getting away from  their predators.  The production of frass chains, during larval development, is the major anti-predation  device used by Eunica bechina avoiding all walking predators, especially ants, which have a difficult time attacking caterpillars hanging on these safety havens.

   Key Words: Nymphalidae, larval defense, frass chains, Eunica bechina, Caryocar brasiliense, Myrmecotrophy, Mutualisms.

Introduction:    “The genus Eunica Hubner, includes 45 species and 24 additional subspecies distributed   throughout the Neotropical region, the majority in the Andean Region and the Amazon Basin” (Freitas and  Oliveira, 1992 and Jenkins, 1990).  Eunica bechina magnipunctata occurs in Southeast Brazil, there it   frequents in savanna-like vegetation (Cerrados).  The Cerrados cover two million km^2, which is 25% of   Brazil.  Cerrados vary physiognomically and consist of four main types: (1) forest with more or less   merging canopy, (2) dense scrub of shrubs and trees, (3) open scrub and, (4) open grassland with scattered   shrubs.  The most common plant species in this area is Caryocar brasiliense, which is the host plant for   Eunica bechina magnipunctata.  Eunica was seen on the Caryocar during the rainy season, September to   January, but the highest concentration occurs when the leaves are young, soft, and red in appearance,   which is during September to October (Oliveira and Freitas, 1991).  This host plant, of Eunica bechina,   bears extrafloral nectaries on the outer surface of the sepals with ants being frequent visitors.  Extrafloral   nectaries provide ants sustenance in the Cerrados and there’s evidence that Myrmecotrophy provides an   anti-herbivore tactic in this type of vegetation.  There are several types of ants found here: Azteca (4   species), Pheidole (3 species), and Zacryptocerus (1 species).  Social insects, like ants, are able to defend   themselves and because of this they are able to establish coevolutionary relationships with a plant species.    With ants providing the defense, it’s less exhausting for the plant to manufacture extra nectar and proteins,   satisfying and keeping the ants content, than it would be for the plant to manufacture toxic substances for   its own protection.  So, first to fourth instar larva, of Eunica bechina, construct frass chains, where they   hang impervious from attacks by foraging ants climbing on the Caryocar for its nectary secretions (Freitas   and Oliveira, 1992).  This behavior occurs in other Nymphalidae.

 Discussion: Larval Development

  “Egg (Fig. 1A): yellowish, conical, and flattened at the top, with 12 to 14 longitudinal ridges and   10 to 12 transverse ridges.  Average height 0.76 mm . . ; average diameter 0.72 mm . . . Larvae hatch   five days after oviposition . . .

  First instar larvae (Fig. 1B): head translucent brown, body translucent yellow changing to pale red   after feeding (due to visible intestinal contents), legs and prolegs translucent yellow; maximum length   3mm; average width of head capsule 0.42 mm . . , average duration 2.6 days . . . The distribution of   setae in the first instar larva are  given in Fig. 2A.

  Second instar larva (Fig. 1C): head black with two short stubby horns; body pale brown with short   conical scoli; maximum length 16 mm; average width of head capsule 0.67 mm . . ; average length of the   horn 0.33 mm . . , average duration 1.5 days . . .

  Third instar larva (Fig. 1D): Head black with white warts  and with two long diverging horns   armed with accessory spines in the middle and ending distally in a knob crowned with short spines; body   dark brown with several scoli; maximum length 12 mm; average width of head capsule 1.26 mm . . ;   average length of the horn 2.36 mm . . ; average duration 2.4 days . . .

  Fourth instar larva (Fig. 1E): Head as in third instar; body dark brown with pale brown lateral   stripe; maximum length 20 mm; average width of head capsule 1.95 mm . . ; average length of the horn   4.56 mm . . ; average duration 3.7 days . . .

  Fifth instar larva (Fig. 1F): Head as in fourth instar; body brown, dorsal region dark brown with  white lines and stripes, ventral region varies from yellow to orange or red, sublateral stripe orange or pale   yellow, legs dark brown and prolegs red except the anal prolegs, (shiny black), scoli black with yellow and   red dots.  The placement of the scoli in the body is shown in Figure 2B.  Maximum length 40 mm; average   width of head capsule 3.28 mm. . ; average length of the horn 6.62 mm. . ; average duration 6.33 days   . . .  Prepupa assumes a ‘J’ position, fixed on the substrate by the anal prolegs and abundant silk.  There is   no great change in color.

  Pupa ( Fig. 1G, H): Green, purple or yellowish, changing to brown and gray or green after one or   two days; spiracula inconspicuous light brown; a dorsal indentation separates abdomen from thorax.    Abdominal segments are mobile; average size 2.2 cm . . , average duration 8.7 days . . . ” (Freitas and   Oliveira, 1992).

 ************* Figures come from Freitas and Oliveiras, 1992 *******************

 Natural History

 Females of Eunica bechina lay their eggs, which are yellowish, conical, and flattened at the top,   individually on small shrubs of Cayocar (Freitas and Oliveira, 1991 & 1992).  The females oviposit mainly   on young leaves, but also on, shoot tips, petioles, and stems of the Caryocar.  The topography of these eggs   exhibit 12-14 longitudinal ridges and 10-12 transverse ridges and usually is 0.76 mm high and there   diameter is 0.72 mm (Freitas and Oliveira, 1991).  To protect these eggs from predators on the ground, they   have vast distributions on the plant: from three to 150 cm above ground (Freitas and Oliveira, 1992).    Helping to provide nutrition to these young caterpillars, they will eat a small portion of their eggshell and   their main diet will consist of young leaves of the Caryocar.  Early instars of Eunica bechina (1st to 4th)   construct frass chains allowing them to escape from predators.  Pupation doesn’t occur on the host plant,   but on shrubs in close proximity.  The adult butterflies usually don’t fly any higher than three meters off the   ground and there appears to be an agonistic behavior and chases occurring between males exhibiting   territoriality (Freitas and Oliveira, 1992).  Sap from tree wounds, along with decaying fruits and mud   puddles, which provide necessary nutrients, are the main sources of food for males.

 General Biology

  “ . . . detailed description of the biology and behavior of Eunica bechina immature . . . The   distribution of setae in the first instar larvae is very similar to the ‘primitive’ pattern of Nymphalidae   . . .  The distribution pattern of the scoli of the fifth instar larvae and pupae is like that of Nica flavilla   . . ; however, the spine distribution pattern in N. flavilla is quite distinctive.  The pupae of  Eunica suggest   that the genus is among the Callicorini, as stated by Otero (1990), who places Eunica in the most advanced   branch of Callicorini, paraphyletic with Temenis and Nica.  However, Oteros results are based on only  eight characters of adult morphology . . . On the other hand, Harvey (1991) proposes that Eunica is   related with Myscelia, Catonephele, Nessaea, Cybdelis and Libythina . . . and the paleotropical genus   Sallya . . . Further study of immatures of other Eurytelinae genera may help solve some systematic   problems in this group, as has been done for the Ithomiinae . . . ” (Freitas and Oliveira, 1992).

 Ant-Plant Mutualisms

  The way in which plants manipulate ants, and the way ants manipulate plants, is a very subtle, but   complex process.  The mutualistic response is usually do to selective pressures (Beattie, 1985).  These   mutualisms occur so often and are so varied in different types of insects that only a few have been   described.  There are four basic interactions between ants and plants: (1) ants protect the plant from   herbivores and other enemies, (2) ants feed plants essential nutrients, (3) they help with the dispersal of the   plants seeds and fruits, and (4) they aid in pollination (Beattie, 1985).  By aiding the plants, ants get   rewards, which can be, nest sites or food.  In plant growth, survivorship and fecundity, the aid of ants   affects plant fitness.  Many plants will produce sugar droplets, which provides the ants a reward in driving   off herbivores (Bentley, 1976).  “ . . . the controversy has been between two opposing groups: the   ‘protectionists’ who supported the idea that ants visiting the nectaries protect the plant from herbivorous   animals, and the ‘exploitationists’ who felt that the plants ‘ . . . have no more use of their ants than dogs do   their fleas . . .” (Bentley, 1976).  Having researched this subject enough, the view of the “exploitationists”   is wrong.  Ants that visit extrafloral nectaries of a plant, without an obligate ant-plant mutualism, still   decreased herbivore damage (Bentley, 1976).  “It is suggested that the ant-cherry relationship is a   facultative mutualism and that nectar production is timed . . . to maximize the chance of successful ant   predation . . .” (Tilman, 1978).  Extrafloral nectaries will increase plant reproduction strongly where there’s   ant density and the potential of herbivores is high (Barton, 1986).  I don’t want to give the wrong idea:    when there’s a caterpillar on the plant an attending ant will come and get rid of it.  This is quite the   contrary; there are many ants that take care of both the plant producing extrafloral nectaries and the   caterpillars.  We will get into ant-caterpillar mutualisms and defenses a little later.

 Ant-plant mutualisms aren’t something that has spontaneously materialized over the past few   years.  This relationship goes as far back as the Cretaceous period where fossils of plants and ants have   been examined.  This relationship has been going strong for 100 million year (Beattie, 1985).  As plants   have evolved they have come up with ingenious ways to protect themselves.  They have evolved various   types of ant attractants, which include special cavities for nesting, small epidermal food bodies, and   extrafloral nectar (Beattie, 1985).  Ants that forage on plants remove a variety of prey items, like insects   and other invertebrates that are herbivores or seed predators (Beattie, 1985).  These plants providing   nesting for ants, have swellings on branches or in the trunk where nests occur, and also may have thorns for   protection.  In this plant species these thorns are for protection, usually against birds or other vertebrates,   but there’s a sensitivity factor, once a herbivore gets onto the plant this vibration triggers swarms of ants,   from their nest, to rid of the predator.  This process happens rapidly to make sure no damage is caused to   the plant.  Ants will go as far as to enter tunnels of leaf-mining beetles to kill the tiny occupants (Beattie,   1985).  Caterpillars aren’t the only insects that ants will remove; they rid of stinkbugs (Pentomidae) and if a   large vertebrate, like a giraffe, comes to snack on the foliage of the plant the ants will swarm over the face   of the giraffe, biting its tongue and face profusely, to drive it away.  Ants are even used as biological   control agents.

  Ants, as biological control agents, are used in pine plantations all over China (Beattie, 1985).    “For example, a single nest of Polyrhacis dives is estimated to harvest about 2000 caterpillars of the pine   defoliator Dendrolimus punctatus per day” (Beattie, 1985 and Hsiao, 1980).  In Europe, several ant species   are used for “forest hygiene.”  “ . . . Formica polyctena, which is a major predator of the green oak leaf   roller (Tortrix viridana) reduces defoliation by this moth so successfully that it is used to control it . . . It   has been estimated that a medium-sized nest of this ant with a territory of 0.27 hectares harvest   approximately six million prey items in a year, many of the herbivores and seed predators . . . Skinner   and Whittaker . . . showed the Formica rufa significantly reduced the populations of lepidopteran larvae   living in Sycamore and Oak Trees in English Woodlands” (Beattie, 1985).  Ants are used as biological   control agents everywhere.

In tropical forests, ants have been used as biological control agents for many crops, the majority   are cocoa and coconuts.  Ant species: Oecophylla longinoda, Tetramorium accleatus, and many other   species will remove vast quantities of orthopterans and hemipterans that damage host plants (Beattie,   1985).  Forests in Mexico have shown large decreases in crop predators because of ants.  These ants are   able to remove these predators with great efficiency wiping out between 82 and 98% of predators.

  Ants help reduce losses to herbivores and seed eaters, but there can be a cost involved with this   protection resulting from the omnivory of ant species (Beattie, 1985).  Ants will kill herbivores during the   same time they’re cultivating homopterans for honeydew.  This behavior has caused a huge loss in   production in many citrus groves, coconut plantations, and cornfields (Beattie, 1985).  “Indiscriminate   predation by ants on insects beneficial to plants and plant enemies may result in the Voltera Effect; that is,   that ants reduce the numbers of the natural predators and parasites of plant enemies as well as the enemies   themselves.  Under these circumstances the population sizes of the herbivores or seed predators can recover   faster than those of their predators and parasites.  Thus, indiscriminate predation by ants can lead   eventually to serious damage to plants” (Beattie, 1985).  This is the case in man-made vegetation, but in   undisturbed areas their mutualism with plants works well.

  There are four ways Myrmecotrophy has been intensified and protection has been made better.    “(1) Ants nest in a wide variety of plant structures most of which are abundant but fortuitous, such as dead   twigs and abandoned insect galls.  However, some structures are specialized and appear to have evolved   specifically as nest sites for ants, (2) a few plant species produce specialized food bodies for ants on the   leaves, petioles, or stems, (3) extrafloral nectaries are present on many plant species, and (4) relationships   with homopterans and lepidoterans may have evolved to such a degree that the losses of plant fluids may be   outweighed by gains in ant protection (Beattie, 1985).  What this does is keeps ants on a specific tissue   when that plant needs the protection.  We have gone deeply into Myrmecotrophy and now lets look at Ant-  Caterpillar Mutualisms and then the defense tactics used by caterpillars and other insects to fend off   foraging ants.
 
 

Ant-Caterpillar Mutualisms

  Ants not only have mutualisms with plants, but they have them with caterpillars too.  Caterpillars   feed on plants, and what they ingest helps form sweet secretions.  These sweet secretions are produced to   establish the ant-caterpillar mutualism.  Caterpillars are aloud to feed as long as they form these secretions   for ants.  Ants contently feed on these secretions and provide protection for caterpillars.  “Ants protect   larvae against attacks of predatory wasps, but not against tachinid parasitoids” (DeVries, 1991).  With   several ants swarming around a feeding caterpillar, a predatory wasp or parasitoid, won’t take a chance of   laying there eggs on the caterpillar because ants will kill or remove anything affecting their mutualism with   caterpillars.  “Among butterflies only larvae of the monophyletic lycaenoids (the Lycaenidae plus   Riodinidae) are known to possess specialized secretory organs for associating with ants” (DeVries, 1991).    This is a wide spread behavior, although only two families exhibit it.  “The widespread taxonomic   occurrence of these organs on larval lycaenids . . . suggest that myrmecophily was an integral part of the   evolutionary history of the lycaenoids” (DeVries, 1991).  “Two ideas have been advanced to explain the   evolution of specialized ant-organs and the nature of myrmecophily in butterflies.  The ‘mutualism’   hypothesis . . . holds that ants protect larvae from predators and parasitoids in exchange for secretions   provided by the larval ant-organs.  The ‘appeasement’ hypothesis . . . maintains that, because ants are   generally marauding predators and abundant in many habitats . . , larvae offer appeasement rewards to   them via the ant-organs to avoid being killed” (DeVries, 1991).  Some Lycaenid larvae have thick cuticles,   which is used as armor, to protect them against ant attacks because ants don’t affect some parasitoids.            “ . . . the secretions produced by larval ant-organs are used as ‘bribery’ to avoid being attacked or killed by   ants” (DeVries, 1991).  The ant-caterpillar mutualism can be summarized into four categories: (1) ants that   are primarily predators, (2) ants that are herbivores and seedeaters, (3) ants that are scavengers, and (4)   ants that, first and foremost, harvest secretions (DeVries, 1991).  In studies done by DeVries, he found   there were three basic outcomes between ants and Thisbe irenea: “(1) An ant species may ignore larvae   (actively or passively), being neither mutualist nor predator.  Fourteen of 26 ant taxa showed this response.    (2) An ant species may attack larvae, injuring or killing them, thereby being a predator.  Four ant taxa responded in this manner.  (3) An ant species may investigate larvae and accept rewards from the ant-  organs.  If the ant species continues to accept these rewards and attends the larvae; it thereby becomes a   mutualist potentially capable of defending its resources against intruders.  Six ant taxa responded in this   way” (DeVries, 1991).  From these observations, ants from all subfamilies ignored larvae or were specialist   predators or general scavengers; families Ponerinae and Dolichoderinae killed larvae or were secretion   foragers, respectively; and the subfamilies of Ponerine, Formicine, Myrmicine, and Dolichoderine were   secretion foragers (DeVries, 1991).  Not only do caterpillars produce secretions for ants, but also they have   calls that attract ants as well.

  “Butterfly caterpillars produce calls that appear to play a role in maintaining symbiotic   associations with ants” (DeVries, 1990).  We know the only two families involved in these symbioses are   Riodinidae and Lycaenidae, and that these symbioses are those caterpillars provide sugary and amino acid   secretions or semiochemicals for ants, as long as ants provide protection.  A few species of caterpillars have   secretory organs and vibratory papillae to communicate with ants.  There are three theories that DeVries   hypothesized about: “(1) the vibratory papillae of riodinid caterpillars function as organs for producing   acoustic calls, (2) caterpillars unable to produce calls are experimentally shown to attract fewer ants and (3)   comparative data suggest that caterpillar calls have evolved at least three times, always as part of symbiotic   associations with ants” (DeVries, 1990).  The caterpillars producing these calls don’t do it on a continual   basis, they only produce vibratory calls when they’re walking or when an onlooker harasses them (i.e.   ant, parasitoid, etc.).  Removing the vibratory papillae didn’t matter because when they molt new papillae   are formed.  These papillae don’t evolve until the third instar, so first and second instar caterpillars are   mute (DeVries, 1990).  “Our current understanding of insect communication suggests that acoustical   signals evolved in response to courtship and rivalry, mate recognition, short distance communication   between colony members of social insects, or as defenses.  In most of the cases presented here . . ,   caterpillar calls occur in the context of mutualistic interactions with ants, yet calls may even be maintained   in the face of parasitic interactions.  For example, the caterpillars of at least three species of Maculinea . . .   live inside Myrmica ant nests to feed on the larval brood of their host ant species.  Hence, this study points   to the possibility that under selection for symbiotic associations, the calls of one insect species have   evolved to attract other, distally related insect species (DeVries, 1990).

 Larval Defense

  Given a chance, ants will attack and kill any insect/larva that’s foreign to them and their host   plant.  Because of this insects and their larva have come up with schemes to deceive and fend off all types   of predators (i.e. ants in this specific case).  One way to avoid mano-a-mano confrontations with predators   are to have noxious tissues, stinging hairs, and foul exudates, boiling sprays, and sticky secretions,   disgusting excretions, painful injections, or repellent regurgitates (Alcock, 1998).  Many other larva employ   other tactics to avoid confrontations.  Some caterpillars will burrow a hole into a leaf giving them free reign   to both sides of the leaf.  On one side, they have constructed a silk protective sheath that ants can’t   penetrate.  When these caterpillars come into contact with an ant they scurry to there burrow and access   their safety haven on the other side of the leaf.  The unusual aspect here is that the ant won’t follow the   caterpillar through its little hole to the other side of the leaf.  Other caterpillars will thrash wildly when   attacked, but this can agitate an ant and make the ant more ferocious.  Few species of caterpillars   (Nymphalidae) construct draglines from the leaf and these prevent them from being attacked.

  Eunica bechina caterpillars construct frass chains as their main defensive mechanism, but this   isn’t their only mechanism for defense.  Ants attacking Eunica bechina caterpillars usually are tending   larvae of Lycaenidae and Riodinidae (Freitas and Oliveira, 1992).  This behavior, of dropping lines from a   plant, is a technique used by arthropods that live closely to aggressive ants (Freitas and Oliveira, 1992).    “The frass chains constructed by the larvae may diminish their predation/removal by ants, since the latter   were never observed climbing on this structure.  The behavior of resting or taking refuge on frass chains is   analogous to that exhibited by some Heliconiini larvae, which rest at the end of tendrils or on ‘island-like’   leaf segments” (Benson et al; 1976 and Freitas and Oliveira, 1992).  Constructing frass chains occurs in   first to fourth instar larvae of Eunica bechina.  Constructing drag lines isn’t a common occurrence and   when ants attack a caterpillar they usually, about 50% of the time are removed.  During 47 encounters   between ants and caterpillars, ants attacked caterpillars about 77% of the time.  The number of caterpillars   removed during this trial is 43% and the number of caterpillars’ dropping/falling off of the leaf occurred    14% of the time (Freitas and Oliveira, 1992).  These numbers tell you that unless caterpillars have a solid   defense mechanism that allows them to get away from the ants they will be attacked and removed.  “In   three instances the caterpillars suspended themselves on the end of a silken line for approximately 20   minutes before climbing back to the leaf (Freitas and Oliveira, 1992).  Not many caterpillars exhibit this   behavior and it’s strange because ants can’t climb down on these draglines, so it keeps caterpillars safe   from foraging ants.  These frass chains don’t provide complete protection from every walking or flying   predator.  Some insects have developed the ability to walk down these chains and lay their eggs on the   cuticle of the hanging larvae.  The most fascinating adaptation here is a wasp/parasitoid that pulls up the   chain, with a larva hanging on the end, and lays its eggs on the caterpillar.  What this wasp/parasitoid does   is uses its feet, just like pulling up a string, it pulls it up, holds it with its foot and continues to pull up the   line until its substrate is there to lay its eggs on.  There are other defenses that this specific caterpillar uses.

  Dropping frass chains is Eunica Bechinas primary defense mechanism, but it employs several   other minor ones to ensure its livelihood.  Eunica bechina have long, pointy scoli that keeps parasitoids   off of them and more importantly the parasitoids eggs from developing inside the caterpillars’ cuticle.    Another defense behavior exhibited by them is wild thrashing (beat reflex) to initially startle or intimidate a   predator, but if the predator is an ant, this usually infuriates an ant and causes subsequent attacks.  A fourth   defensive tactic used, “ . . . Eunica caterpillars frequently regurgitate and/or bleed, a behavior shown to   effectively repel their aggressors who may end up abandoning the larvae.  After successive bites on the   larvae, attacking ants frequently exhibit strong disturbance behavior and vigorously clean their mandibles,   antennae and head” (Freitas and Oliveira, 1992).  “Regurgitation is also common in butterflies . . . Just   after attacking the caterpillars, ants receiving this fluid exhibited strong disturbance (walking erratically   and shaking the body) and conspicuously cleaned their mandibles and head.  Ant deterrence can also occur   from bleeding by the injured caterpillars . . .” (Freitas and Oliveira, 1992).

 I first thought that frass chains were made of silk, but this isn’t the case.  The term “frass” is a   pelletlike excrement of insects, especially from caterpillars and woodborers.  From this definition, we   know that, the construction of these chains depends on the food eaten by the caterpillar.  In Eunica bechina,   they feed mainly on Caryocar brasiliense and you have to assume that there’s material in the plant or in the   Digestive tract of this caterpillar, that allows them to construct these drag lines.  Whether the material   comes from the plant or there’s machinery inside the caterpillar to form these frass chains, I don’t know.  I   couldn’t find any literature on the exact production of frass chains.  I’m also unsure of one other aspect of   these frass chains, this is where they come out of the caterpillar.  They either have a specialized structure on   their back end or it comes out of their rectum/anus.  This brings up another question and that is if it comes   out of a specialized organ on the posterior end, no problem, but if frass chains are excrement from their   digestive tracts are frass chains the way Eunica bechina caterpillars remove waste?  I also, didn’t find a   description or any literature to answer this question.  Many insects and larva expend a lot of their energy   forming anti-predator devices, but if they didn’t they wouldn’t survive.  They have one primary reason to   expend this energy and that is to reproduce.  An insect doesn’t care if it only lives one day, as long as it can   reproduce, or lay its eggs or whatever it has to do to increase its species overall fitness.  Insects see greater   fitness with greater numbers.  This means that the more of them there are the more they can reproduce.    This behavior isn’t something an insect thinks about, it’s innate, an internal driven force that makes them so   efficient.

 
References:

-  Ackery P.R.  (1988) Hostplants and classification: a review of nymphalid butterflies.  33: 95-203.

-  Alcock John  (1998) Adaptive Responses to Predators.  Animal Behavior.  9: 289-341.

-  Barton A.M.  (1986) Spatial Variation in the Effect of Ants on an Extrafloral Nectary Plant.  67(2):
    495-504.

-  Beattie A.J.  (1985) The evolutionary ecology of ant-plant mutualisms.  1-177.

-  Bentley B.L.  (1976) Plants Bearing Extrafloral Nectaries and the Associated Ant Community:
     Interhabitat differences in the reduction of Herbivore Damage.  57: 815-820.

-  Borror D.J, Johnson N.F, Triplehorn C.A.  (1989) Order Lepidoptera: Butterflies and Moths.
    An Introduction to the Study of Insects.  34: 588-661.

-  Brandao C.R.F, Oliveira P.S.  (1991) The ant community associated with extrafloral nectaries in the
     Brazilian cerrados.  In: C.R. Huxley and D.F. Cutler (eds.), Ant-Plant Interactions.  Oxford
     University Press, Oxford.  Pgs. 198-212.

-  Costa F.M.C.B, Oliveira-Filho A.T, Oliveira P.S.  (1992) The role of extrafloral nectaries in Qualea
     grandiflora (Vochysiaceae) in limiting herbivory: an experiment of ant protection in cerrado
     vegetation.  Ecological Entomology.  17: 363-365.

-  DeVries P.J.  (1991) Mutualism between Thisbe irenea butterflies and ants, and the role of ant Ecology
     In the evolution of larval-ant associations.  43: 179-195.

-  DeVries P.J.  (1990) Enhancement of Symbioses between Butterfly Caterpillars and Ants by
     Vibrational Communication.  Science.  248: 1104-1106.

-  Freitas A.V.L, Oliveira P.S.  (1992) Biology and Behavior of the Neotropical butterfly Eunica bechina
With special reference to larval defence against ant predation.  Journal of Research on Lepidoptera.  31: 1-11.

-  Freitas A.V.L, Oliveira P.S.  (1991) Hostplant record for Eunica bechina magnipunctata
     (Nymphalidae) and observations on oviposition sites and immature biology.  30 (1-2): 140-141.

-  Heads  P.A, Lawton  J.H.  (1985)  Bracken, ants and extrafloral nectaries. III.  How insect herbivores
     avoid ant predation.  10: 29-42.

-  Hsiao  K.J.  (1980)  The use of biological agents for control of the pine defoliator,  Dendrolimus
    punctates (Lepidoptera, Lasiocampidae), in China.  Protection Ecology.  2: 297-303.

-  Jenkins D.W.  (1990)  Neotropical Nymphalidae VIII.  Revision of Eunica.  Bulletin of the Allyn
     Museum.  131: 1-177.

-  Schemske D.W.  (1982) Ecological Correlates of a Neotropical Mutualism: Ant Assemblages at
         Costus Extrafloral Nectaries.  63: 932-941.

-  Skinner G.J, Whittaker J.B.  (1981) An experimental investigation of inter-relationships between
     the wood-ant (Formica rufa) and some tree-canopy herbivores.  Journal of Animal Ecology.
 50: 313-326.

-  Stehr F.W.  (1987) Immature Insects.  Order Lepidoptera.  26: 288-596.

-  Tilman D.  (1978)  Cherries, Ants and Tent Caterpillars: Timing of Nectar Production in Relation to
     Susceptibility of Caterpillars to Ant Predation.  59(4): 686-692.