Eusocial species, species with the presence of overlapping generations, cooperative brood care, and a division of labor in reproduction (Carpenter, 1991) are now recognized in several insect orders. Eusociality was once thought to be unique to Hymenoptera and Isoptera and thus most of the what we know about eusociality comes from studies within those two orders. It wasn't until more recently that other animal groups were recognized to contain eusocial members as well. One of these groups is the social aphids (Homoptera).
Social aphids have been found all over the world. One of the most well studied are the bamboo aphids of Japan (Ito, 1989). Recently studies have looked at gall producing aphids from other areas such as Malaysia (Stern et al., 1994) and even North America (Moran, 1993).
Social aphids have introduce a whole new direction in the evolution of eusociality and behavioral ecology. Although they are diploid by nature, they also have the ability to reproduce asexually thus producing genetically identical individuals (Alexander et al., 1991). This leads to many implications for relatedness within a colony and sets the stage for cooperation within a colony of aphids.
There are three major areas presently studied in the behavioral ecology of social aphids. 1) Nesting behavior; All eusocial animals have at least one nest. In fact it has been proposed that nesting behavior is a necessary preadaptation for eusociality (Anderson, 1984). The social aphids either produce galls or live inside hollow plants such as bamboo (Stern, 1994; Aoki 1987). 2) Defense; Many species of social aphids produce a specialized soldier caste. These individuals are both sterile modified to defend the nest from predation (Aoki, 1987; Stern et al, 1993). 3) Life histories; the colonies are able to adjust the production of the various castes throughout the colony cycle (Sunose et et al. 1991, Sakata et al., 1991). Much has been done recently in all three areas of the aphid biology. In this paper I will attempt to unite the many studies on social aphids and discuss what the similarities and differences between species. From that perhaps we can better understand the evolutionary origins of this remarkable social system.
Aoki (1977) made a remarkable find, demonstrating to the world that the aphid, Colophina clematis, produced instars that defend the colony from intruders. Since then many species of social aphids have been discribed in the two families Pemphigidae and Hormaphididae (Stern and Foster, 1996). Eusociality, having overlapping generations, cooperative brood care, and a division of labor in reproduction (Carpenter, 1991) was originally thought to only exist in the Hymenoptera and Isoptera but now are recognized in five insect orders and in mammals (Alexander et al, 1991). Aphids provide a unique prospective to the study of eusociality. Although they are genetically diploid, aphids reproduce both sexually and parthenogenetically. It is in these clonal stages that large colonies are formed comprising of genetically identical individuals. Aphids are the only colonal species that exhibit eusocial behavior. This colonal nature of a colony would create different rules for the colonies life histories. They will lack the parent-worker conflicts seen in other colonies that arise from colony members being genetically unequal. In a clonal society everyone is the genetically identical so such conflicts should be nonexistent.
Stern and Foster (1996) examined the evolutionary of eusociality in aphids and took an evolutionary and systematic approach. They describe the phylogenetic distributions and the relationships of all of the social aphid species known to that data. In this, paper I will take a more behavioral and ecological approach to answering the question, using what is known about the behavior of the social aphids from the level of the castes to the life cycles and life histories of the colony. I will end with a model which may explain how a social aphid colony regulates its soldiers.
The known distribution of social aphids is expanding as new species are being found (Stern and Foster, 1996). Most of the known species are found in Japan. This probably has to do with the fact that most of the species discribed to this point were done by Aoki et al. (Stern and Foster, 1996). Today we recognize social aphid species in Europe (Foster, 1990), Malaysia (Stern et al., 1994) and even the United States (Aoki and Kurosu, 1988; Moran, 1993). Stern and Foster (1996) report many new species found that are currently in press or even in preparation (Table 1).
Social aphids as far as we know are limited to two families, Hornaphididae and Pemphigidae (Stern, 1994). Within that clade it appears that the sociality of the aphids has arisen 6-9 nine separate times (Stern et al., 1994, Stern and Foster, 1996). This is approaching the amount of times eusociality is estimated to have arisen in the social Hymenoptera (Alexander et al., 1991). The study of social aphids is still quite young and that estimate probably will change as we add more species to the growing list.
The introduction of a clonal species into the eusocial groups has changed the nature of how we view the evolution of sociality (Ito, 1989, Stern and Foster, 1996). Although the haplodiploid hypothesis (Hamilton, 1964) worked well in the beginning of the studies of sociality, the introduction of more diploid species in the eusociality continuum (Sherman et al., 1995) has probably decreased the value of the kin selection argument as the sole means of selecting for sociality (Reeve, 1996). Aphids seem to go beyond kin selection to almost self- selection.
Each individual in a colony will be genetically identical and in a sense they are always behaving for their own genes and not just a subset. These colonies might be more akin to the social Hydrozoans and Ectoprocta. These colonies contain clones which specialize for certain labors. In hydrozoan colonies, there are three basic morphs, gastrozooids which feed the colony, dactylzooids which defend the colony, and gonozooids which reproduce (Hyman, 1941, Brusca and Brusca, 1990 ). In Ectoprocta there are similar specialized zooids, avacularia, vibracula, and gonozooids (Hyman, 1959, Brusca and Brusca, 1990). In these colonies the individuals are connected, or partially connected by a common raced like tube (Brusca and Brusca, 1990).
If clonality favors sociality, then why aren't all aphids eusocial? One possibility could be the colony's lifestyle. In all of these cases the social aphids live in a gall or a segment of Bamboo (Aoki, 1982; Aoki, 1987; Stern and Foster, 1996). Anderson (1984) proposed that all social species should be predisposed in having a nest. A nest is a place where a colony is harboring all of it's resources. This centralizes the reproductive investment of a colony.
Many times the nest serves as the first line of defense of a colony (Starr, 1990), protecting the investment from many outsiders (Anderson, 1984). Termites, mole-rats and some Hymenoptera burrow underground (Starr, 1990). Many wasps create cryptic nests (Starr, 1990). Social aphids produce galls, which are tough pockets artificially induce in a plant by the aphids (Faith, 1979). All of the alates and reproductive destined instars are normally found inside the galls and rarely outside (not including dispersal periods)( Aoki, 1989; Kurosu and Aoki, 1991a). The individuals on the outside are the soldiers which defend the gall from any predator that would destroy this nest and it's contents. Starr (1985) proposed that eusociality can only evolve if the animal could defend it's colony against vertebrate predation. Vertebrates are capable of destroying a nest in a single attack. (Starr, 1985). Although for most species of aphids vertebrates are not the main concern (see Table 2), defending against predators is a must, necessitating the production of soldiers. A species of aphid that produces galls might be predisposed to sociality and the production of soldiers would be an additional necessity to remain social.
There are two basic life cycles of all aphids, those that use only one host species, Monoecious, and those that switch between two hosts during the life cycle, Diecious (Dixon, 1985). Most of the social aphids are Diecious but a few have reverted back to the Monoecious state (Stern, 1994). The aphids will spend most of autumn, winter and spring on a primary host (Dixon, 1985). These hosts are usually trees or woody shrubs. A foundrix will establish a gall and then produce clones (Aoki, 1980b; Dixon, 1985). In dimorphic species (see below) she will produce reproductive destined instars before producing soldiers (Tanaka, 1993). In many cases the only soldiers produced by the aphids are on the primary host (Table 1). In the summer the aphids will produce more migrating individuals, usually alates, and colonize a secondary host (Kurosu and Aoki, 1991a; Dixon,1985) on for the switch is for nutritional reasons (Dixon, 1985). In the spring the trees are a good food source but their nutritional value declines in the summer as they age and produce toxins. Herbaceous plants are just starting out and will flower in the summer thus providing a better additional food source for the aphids. As the secondary host dies the aphids leave this host, returning to the primary one (Dixon, 1985). The colonies on the secondary host rarely produce soldiers (Table 1; Stern and Foster, 1996). In cases where soldiers exist on both the primary and secondary, the soldiers of the primary host colonies are different then those of the secondary colonies (Table 1; Kurosu and Aoki, 1991a). This suggests that the soldiers may have specialized for different defense jobs.
There are several types of soldiers found in the social aphids. Stern and Foster (1996) describe several types of soldiers based on physical characteristics, behavior and which instar they are created from. Although this is a usefully way to look at soldiers for systematics, soldiers can be summarized as monomorphic or dimorphic, cases where defending first instars are morphologically different from reproductive destined instars (Stern et al., 1996). The dimorphic colonies can be further subdivided into those that use legs and stylets and those that use special grasping legs and horns.
Monomorphic colonies Most of the colonies of Pemphigidae are monomorphic and only one species in Hornaphididae, Ceratovacna nekaoshi (Table 1; Aoki and Kurosu, 1988). In these species the first instars are characterized by a prolonged first instar stage and then much shorter times spent at each stage (Akimoto, 1996).
Dimorphic biting instars (Figure 1a) These soldiers tend to be hairier and have longer hind or front legs. The soldiers will not only pierce their victim with stylets, but will also rip into a victim with their legs. (Aoki, 1978a; Ohara 1985a,b). The bite of many of these species is painful to vertebrates (Aoki et al., 1977; Aoki, 1979)
Pseudoscorpoin instars (Figure 1b) These instars have elongated front legs which allow them to grasp their victim. The frontal horns are elongated allowing them to pierce the victim with the horns rather than bite. There are some instars of this type that lack the horns but instead bite the intruder (Aoki, 1978b; Aoki and Akimoto, 1981).
There has been little research done on how the aphids achieve this dimorphism. Since the aphids are clonal it would seem unlikely that genetics has much to do with the decision to be a soldier or a reproductive destined instar. It has been proposed that there might be genetic variation within clones due to genetic crossing over (Cognetti, 1961) but that has been shown not to be the case, and that all colony members are genetically identical (Blackman, 1977; Blackman, 1979).
One study (Fukatsu and Ishikawa, 1992) compared the internal physiology of a normal instar and a soldier. They found that the soldiers lack any sign of a reproductive system and have a large number of fat bodies. Ito (1989) made the observation that the soldiers don't seem to eat but live for a couple of weeks. Perhaps it's these fat stores that allow the soldiers to stay alive (Fukatsu and Ishikawa, 1992). The other internal physiological difference was the presence or absence of an endosymbiont. These symbionts help the aphids produce amino acids that the aphid could not normally produce on it's own. The soldiers were found to lack the endosymbiont while the other first instars had them (Fukatsu and Ishikawa, 1992). Males were also found to lack the symbiont. If a foundrix aphid is able to control whether or not a symbiont is transferred to its offspring perhaps that somehow controls the physiology of the instar. The lack of certain amino acids available to the instar might change the chemical make up of the individual and perhaps alter the physiology and behavior.
What would promote the evolution of soldiers in the first place? The best place to start answering this question would to examine the job of an aphid soldier. The term soldier may be rather unfortunate as there is emerging evidence that the soldiers not only defend but aid in the maintenance of the gall (Aoki, 1980a; Aoki, 1989a; Benton and Foster, 1992). There are many observations of soldier instars cleaning the gall of excess honeydew (Benton and Foster, 1992). The honeydew build up in a gall apparently can be detrimental to the colony (Aoki, 1980a; Benton and Foster, 1992). In an experiment with Pemphigus spyrothecae, Benton and Foster (1992) found that colonies without soldiers did not clean their galls. When the galls with soldiers were positioned so that the galls couldn't be cleaned by the colony the galls did worse than colonies in galls that were able to be cleaned by the soldiers.
One possibility that has not been examined is if there is a division of labor among soldiers. Are defenders and cleaners separate groups or are they the same? This may change the nature of how we view this caste and the over all organization of the colony. Aoki (1978b) found a bimodel distribution in soldier size in Pseudoregma alexanderi but could offer no explanation for it. Perhaps the two sizes represent different castes of soldier aphids. Aoki (1995) reported that the ÒsoldierÓ caste of Tuberaphis leeweni were not very aggressive towards insects or man. Perhaps the ÒsoldiersÓ this species are in fact specialized cleaners and do not act as defenders. These individuals have the specialized hairs associated with cleaning (Aoki,1985)
Soldiers are capable of defending against different types of predators (Table 2). The stylet using soldiers are not only effective against insects but are also effective against vertebrates (Aoki, 1979 ). According to Aoki (1977) the aphids give a very irritating bite. Asteroptryx stracicola was observed to have soldiers roaming around the branches near the gall and attacking vertebrates touching the tree. (Aoki, 1979)., Furthermore, Aoki and Kurosu (1991) and Chou et al. (1985) found evidence of squirrels attacking galls of Ceratoglyphina bambusae. If the galls are cut, a wax is produced by the colony which helps to seal it up again and the gall will heal in 24 hours.. If the soldiers can drive away a squirrel then the gall will be able to heal quickly. Whether or not soldiers can drive away a squirrel has yet to be demonstrated experimentally. However, Aoki (1979) showed that the aphid Astegopteryx styracicola can not only irritate man but will have the same effect on mice.
Besides gall eating vertebrates, aphids are also preyed on by many insect larvae (Table 2). In many cases the soldiers have been observed to destroy eggs of possible predators (Ohara, 1985a; Aoki and Akimoto, 1981). The soldiers were very effective against most types of larvae. Foster (1990) compared the survival of Colophina arma colonies with soldiers and those without soldiers. He set up artificial colonies all with equal numbers of aphids but some colonies lacked a soldier cast. The colonies with soldiers were able to survive predation events while undefended colonies were killed by the predators. This demonstrated that soldiers were able to defend the nests. One Syrphid fly, Metasyrphus confrater modified it's behavior in response to the soldiers. A female will lay her eggs on old spider webs near a gall. This makes them inaccessible to the roaming soldiers (Ohara, 1985a,b). Soldiers of this species are able to kill the eggs not on webs (Ohara, 1985a) and early instar larvae but are ineffective against later instars (Ohara, 1985b). Tanaka and Ito (1995) found weak evidence that larger colonies of Psuedoregma bambucicola were able to defend against a predator Eupeodes confrater more efficiently then smaller colonies. They even found evidence to suggest that a Eupeodes female will aviod laying eggs near a large colony.
Models Akimoto (1996) created a model to examine what factors might have effected the evolution of soldiers and what conditions might favor the selection for dimorphic individuals. The model was tested using computer simulations. He found that a number of factors would favor an increased first instar span, as seen in the monomorphic aphid colonies. These factors included increased predation, low birthrates, colony duration, and density effects. In his simulations low birthrates also predicted polymorphism in the first instar stage. By having polymorphic first instar, the soldiers could maintain the long first instar span while the other morph could have a normal short first instar span. This allows the reproductive destined instars to mature faster and still satisfy the conditions for a long first instar span. The longer the colony duration the more soldiers were produced in both the monomorphic and dimorphic simulations.
There is a small amount of data supporting parts of this model. First instars of the monomorphic species tend to last longer then the the other stages (Aoki, 1980; Akimoto, 1992). This span difference increase from generation to generation (Akimoto, 1996). An interesting test to this model would be to compare the birthrates of nonsocial aphids with those of social aphids. This model would predict the non-social aphids to have a higher birthrate. Akimoto pointed out that the social galling Pemphigus aphids tend to remain in their gall stage longer that non social Pemphigus species (Aoki and Kurosu, 1988b; Akimoto, 1996).
Most of the published work up to this point is the discovery and description of the casts on the social aphid colony. Very few studies have actually looked at the life histories of aphid colonies, and fewer still have offered a model to explain how an aphid colony might regulate it's casts. Two main studies have been done in the former area. Moran (1993) investigated the life cycle of the monomorphic social aphid Pemphigus obesinymphae. She found that the amount of soldiers increase from August to late September and then decreased in October. This decline is usually followed by the production of alates which fly to the secondary host to over-winter.
Sunose et al. (1991) also found a fluctuation of soldiers during the colony life cycle in the dimorphic aphid Pseudoregma bambucicola. The number of soldiers increases from September to November and then declines from late October to mid-April. In early May the alates will fly out of the galls and move to the secondary host (Kurosu and Aoki, 1991a. They return to the primary host in August (Sunose et al, 1991; Kurosu and Aoki, 1991a) The predation pressures were found to be the highest in the periods between the months of August and November, the same time the soldiers are maximized (Sunose et al, 1991). Whether or not the increase of soldiers in this time period is simply due to the predation pressure is not clear.
Perhaps the presence of predators might also select for an increase of time spent in the gall phase of social aphids. If predators are around when the aphids would normally leave the gall, then it might pay to spend a longer time in the gall phase and ultimately shift the production of alates during a period without high predation pressures. This would then create a need for a lag time that could be solved by increasing the first instar span to which would favor the production of soldiers as predicted by Akimoto's (1996) model.
The first model established to look at soldier investment was created by Stern and Foster (1996). This model tried to predict the optimal cast ratios based on level of colonal mixing, predation, soldier efficiency and birth rate. Stern and Foster (1996) looked at the effects of predation, birthrate and clonal mixing on the allocation of soldiers in a colony. They found through computer simulations that as, with Akimoto, increased predation and lower birthrates also favor an increase in soldier production. The lower levels of clonal mixing still favored soldiers but high levels did not. Evidence for colonal mixing is not seen in the literature.
Stern and Foster (1996) admitted the oversimplification of the model. However there were a few important details that they left out. One important observation that has been seen in many colonies is the intolerance of soldiers to alates approaching the colony. In most cases the soldiers will kill any non-soldier aphid outside the colony even kin and members of their own colony (Sakata and Ito, 1991; Aoki, 1991; Carlin et al, 1995 ). Although baffling at first the lack of kin recognition makes sense in light of the natural history of the animals. Normally most of the reproductive destined the colony members tend to remain protected in the gall (Kurosu and Aoki, 1991b). If an alate approached the colony then most likely that individual is not a part of the colony.
Although Stern and Foster (1996) mentioned a clone-eye view they really never separated clones from other eusocial insects. Relatedness in an aphid colony is 100% (Blackman,1982). Any intruding reproductive would drop that value significantly. It would seem that the best evolutionary strategy would be to be intolerant to any outsider regardless of relatedness. In a simplistic view, the aphid system might be similar an immune system of an organism. Any foreign individuals should be removed. In Ectoprocta the specialized zooids, avacularia and vibracula, are specifically produced to remove any foreign matter including settling larvae (Hyman, 1959). As with the social aphids they probably don't show any kin recognition abilities. Although the disadvantage to this system is that the soldiers could kill kin by mistake, there is also a decreased chance that social parasites could successfully evolve in this system. In fact, there is no report of any social parasites in eusocial aphids.
In an aphid colony there are two basic castes. One is the reproductive caste. This caste could be subdivided into the alate sexuals and the instar crawlers. In many cases colonies will produce both (Aoki, 1980b). The second cast is the working cast. This manly includes the soldiers but also contains the cleaners (Stern and Foster, 1996). Although the morphology is apparently the same there may be some form of division of labor between soldiers and cleaners. Since such evidence does not exist, I'll just deal with soldiers for the purposes of this model.
The soldiers serve two basic functions, 1) protect the colony from predation and 2) protect the colony from immigration of outsiders. This division would lead to the following predictions: Soldier number should increase under increased predation pressure and increased immigration pressure. The in-gall population could effect the soldier population in two ways. As the colony is growing one could expect the soldier number to increase. There comes a point where crowding effects might stimulate emigration morphs to appear, and then one could expect the soldier population to decrease. There is a lot of evidence that crowding can effect -12- the production of migrating castes in aphids (Dixon, 1985). Since soldiers do not recognize kin (Aoki, 1991; Carlin et al, 1995), it would be advantages to have the soldiers absent from the colony when reproductives emerge. So the increase in the number of individuals leaving the gall should necessitate a decrease in soldier number. This would predict a bell curve shaped soldier population over the colony cycle (Figure 2a). In the beginning of the cycle, the soldier population should increase. It will reach a peak at a point where the colony finds it self at a certain crowding point and then the population of the soldiers should decrease. The predation and immigration pressures would determine the boundaries of the curve. The greater the predation pressure the the higher the curve (Figure 2b), producing a greater the amount of soldiers at the peak. The immigration pressure would have similar effects. As mentioned above the length of time the predation pressure exists could cause an lengthening in the amount of time spent in the gall by the colony. This would decrease the birthrate of the colony and would increase the time the soldiers are present (Figure 2c).
Thus the equation below could predict the number of soldiers at a given time.
s = - pt((C)2)+p
where s = # of soldiers
C= amount of crowding
p = predation pressure +immigration pressure
pt = duration of predation pressure
If plotted over time this equation would create a bell shaped curve. The crowding effect, C, will first increase the soldier population and then a peak will be reached where the crowding will signal the aphids to to prepare to leave the gall by producing emigrating morphs. Predation and immigration pressures will just alter the shape of the curve but will not prevent the drop in soldiers once the alates are being produced. The predation effects can extend the the amount of time soldiers exist. This would simulate the extension of the gall stage we see in many social aphids (Akimoto, 1996). If predation is high then the aphids could wait out the predator and leave the gall when the predation pressure has decreased. This extended time would favor a lower birthrate, to prevent the crowding effect from taking over too fast. Both Akimoto (1996) and Stern and Foster (1996) have demonstrated that low birthrates favor the production of soldiers. It may have been that the earlier species of aphids were responding to the presence and absence of predators during the year and adjusting their birthrate and emerged when the predation pressure was low.
The fluctuations found by Moran (1993) and Sunose et al. (1991) support this models predictions. In both cases the soldier production was high when predation was high and was low once the emigrating individuals were leaving the colony. I also examined the data provided by Aoki (1980b) on Colophina arma. This aphid produces two morphs of first instars, soldiers and midget, (which was described by Aoki (1980b) as an emigrating morph). I plotted the soldier population against the midget population and found there was a significant (p < 0.001, Spearman's Correlation Coefficient) decrease in soldiers as the midget population increased (Figure 3). This was based on a small number of observations. Future studies could compare the number of alates with the number of soldiers. The prediction of this model would be that as the alates increase the soldier number should decline. Another comparison would be to compare the predation pressure with the length of time the species of social aphid spends in it's gall. The percentage of time the aphids spend in the gall phase should increase with the amount of predation pressure. If the predation pressure is high for a long enough time it might cause the species to revert to a monecious cycle as seen with some of the Pemphigus species (Moran, 1993).
Social aphids could provide us with a unique perspective on the evolution of eusociality. Since they are clonal a student of social aphid biology does not have to be concerned with intercolonal conflicts. The interest of the colony should be the same for all members. There are many types of soldiers produced by these colonies. These may reflect the conditions under which the colonies evolved. Species with piercing soldiers might have to defend against invertebrate and vertebrate predators (Aoki, 1979). Others seem ineffective against vertebrates and more specialized for deterring insects. Due to the fact the soldiers are aggressive towards other species and conspecifics the colony must regulated soldier production in response to the ecological factors and the crowding of the colony. Predation would favor the increase in soldiers and the inability of the soldiers to recognize kin would select for a decrease in soldier production when the alates are ready to leave. This would create a cyclical pattern of soldier populations seen by the few natural history studies (Sunose et al, 1991; Moran, 1993).
Akimoto, S. 1992. Shift in life-history from reproduction to defense with colony age in the galling aphid Hemipodaphis persimilis producing defensive first instar larvae. Res. Pop. Ecol. 34: 359-372.
Akimoto, S. 1996. Ecological factors promoting the evolution of colony defense in aphids: computer simulations. Ins. Soc. 43: 1-15
Alexander, R.D., K.M. Noonan, and B.J Crespi. 1991. The evolution of eusociality. in P.W. Sherman, J.U.M Jarvis, and R.D. Alexander eds. The Biology of the Naked Mole-rat. Princeton University Press. 3-44.
Anderson, M. 1984. The evolution of eusociality. Annual Review of Ecology and Systematics. 15:165-189.
Aoki., S. 1977. Colophina clematis (Homoptera, Pemphigidae), an aphid species with soldiers. Kontyu, 45: 276- 282.
Aoki., S., S. Yamane, and M. Kiuchi. 1977. On the biters of Astegopteryx styracicola (Homoptera, Aphidoidae). Kontyu. 45: 563-570.
Aoki S. 1978a. Two pumphids with first instar larvae attacking predatory intruders (Homoptera, Aphidoidae). New. Entomol. 27: 7-12.
Aoki S. 1978b. Notes on the Pseudoscorpion-like larvae of Pseudoregma alexanderi (Homoptera, Aphidoidae). Kontyu. 46: 433-438.
Aoki S. 1979. Further observations on Astegopteryx styracicola (Homoptera, Aphidoidae), and aphid species with soldiers biting man. Kontyu. 47:99- 104
Aoki, S. 1980a. Occurrence of a simple labor in a Gall aphid, Pemphigus dorocola (Homoptera, Aphidoidae). Kontyu. 48: 71-73.
Aoki, S. 1980b. Life cycles of two Colophina aphids (Homoptera, Pemphigidae) producing soldiers. Kontyu 48:464-476.
Aoki, S. 1982. Soldiers and altruistic dispersal in aphids. pp 154-158 in in Breed M. C., Michener C.D., and H. E. Evans eds. The Biology of Social Insects. Westview Press. Boulder, Colorado
Aoki, S. 1987. Evolution of sterile soldiers in aphids. In. Y. Ito, J.L. Brown, and J. Kikkawa eds. Animal Societies: Theories and Facts. Japan Science. Society Press. pp 53-65.
Aoki, S. 1989a. Soldiers of Astegopteryx (Homoptera, Aphidoidae) clean their gall. Jpn. J. Ent. 57: 407-416
Aoki, S. 1989b. Can root gererations of Pemphigus (Homoptera, Aphidoidae) grow in a poplar gall? Jpn. J. Ent. 57: 205-209.
Aoki, S. 1991. Aphid soldiers discriminate between soldiers and non-soldiers, rather than between kin and non- kin, in Ceratoglyphina bambusae. Anim. Behav. 42: 865-866.
Aoki, S.1992. No attack on conspeciefcs by soldiers of the gall aphid Ceratoglyphina bambusae(Homoptera) late in the season.
Aoki, S. 1995. Tuberaphis leeuweni (Homoptera), a tropical, monoecious, gall- forming aphid with soldier like nymphs. Jpn. J. Ent. 63: 77-86
Aoki, S. and S. Akimoto. 1981. Observations on Pseudoregma alexanderi (Homoptera, Pemphigidae), an aphid species producing pseudoscorpion- like soldiers on bamboos. Kontyu 49: 355-366.
Aoki,S. and U Kurosu. 1988. Secondary Monoecy of a North American gall aphid, Pemphigus monophagus (Homoptera, Aphidoidae). Kontyu. 56: 394-401
Aoki, S. and U. Kurosu. 1991a. Galls of the soldier-producing aphid Ceratoglyphina bambusae broken by vertebrates (Homoptera, Aphidoidae). Jpn. J. Ent. 59: 743-746
Aoki, S. and U. Kurosu. 1991b. Galls generations of the soldier-producing aphid Pseudoregma bambucicola (Homoptera). Jpn. J. Ent. 60: 359-368
Benton, T.G. and W. A. Foster. 1992. Altruistic housekeeping in a social aphid. Proc. R. Soc. Lond. B. 247: 199- 202.
Blackman, R. L. 1979. Stability and variation in aphid lineages. Biol. J. Linn. Soc. Lond. 11: 199-202
Blackman, R. L. 1987. Reproduction, cytogenetics and developement. in Minks A. K. and P. Harrewin eds. Aphids: Their Biology , Natural Enemies and Control vol 2. Elsevier, Amsterdam
Carlin, N. F., D. S. Gladstein, A. J. Berry, and N. E. Pierce.1995. Absence of kin discrimination behavior in a soldier producing aphid Ceratovacuna japonica (Hemiptera: Pempgidae; Cerataphidini)
Carpenter, J. M. 1991. Phylogenetic relationships and the origin of social behavior in Vespidae. In Ross. K. C., and R. W. Matthews. The Social Biology of Wasps. Cornell University Press. 7-32.
Cognetti, G. 1961. Citogenetica della partenogenenis negli Afidi. Archivo Zool. Ital. 46: 89-122
Dixon, A. F. G. 1985. Aphid Ecology. Blackie. Glascow.
Faith D. P. 1979. Strategies of gall formation in Phemphigus aphids. NY Entom Soc. 87: 21-37.
Foster, W. A. Experimental evidence for the effective and altruistic colony defense against natural predators by soldiers of the gall-forming aphid Pemphigus spyrothecae (Hemiptera: Pemphigidae), having sterile soldiers. Behav. Ecol. Sociobio. 27: 421-430.
Fukatsu, T. and Hajime Ishikawa. 1992. Soldier and male of an eusocial aphid Colophina arma lack endosymbiony: implications for physiological and evolutionary interaction between host and symbiont. J. Ins. Physiol. 38:1033-1042.
Hamilton, W.D. 1964. The genetical evolution of social behavior. J. Theoret. Biol. 7:1-52.
Ito,Y. 1989. The evolutionary biology of sterile soldiers in aphids. TREE. 4: 69- 73
Kurosu U. S Aoki. 1988. Monomorphic first instar larvae of Colophina clematicola (Homoptera, Aphidoidae) attack preditors. Kontyu. 56: 867- 871.
Kurosu, U. and S. Aoki. 1991a. The gall formation, defenders and life cycle of the subtropical aphid Astegopteryx bambucifoliae.(Homoptera). Jpn. J. Ent. 59: 375-388
Kurosu, U. and S. Aoki. 1991b.Incipient galls of the soldier producing aphid Ceratoglyphina bambusae(Homoptera). Jpn. J. Ent. 59: 663-669
Moran N. A. 1993. Defenders in the North American aphid Pemphigus obesinyphae. Ins. Soc. 40: 391-402
Ohara, K. 1985a. Observations on the oviposition behavior of Metasyrphus confrater (Diptera, Syrphidae) and the defensive behavior of soldiers of Pseudoregma bambucicola (Homoptera, Pemphigidae). ESAKIA. 23: 99- 105
Ohara, K. 1985b. Observations on the prey-predator relationship between Pseudoregma bambucicola (Homoptera, Pemphigidae) and Metasyrphus confrater (Diptera, Syrphidae) with special reference to the behaviour of t he aphid soldiers. ESAKIA. 23:107-110
Reeve, H. K. 1993. Haplodiploidy, eusociality and absence of male parental and alloparental care in Hymenoptera: a unifying genetic hypothesis distinct from kin selection theory. Phil. Trans. R. Soc. Lon. 342: 335- 352.
Sakata, K. and Y. Ito. 1991. Life history charactoristics and behavior of the bamboo aphid, Pseudoregma bambucicola (Hemiptera: Pemphigidae) having sterile soldiers. Ins. Soc. 38: 317-326
Starr, C. K., 1985. Enabling mechanisms in the origin of sociality in the Hymenoptera- The stings the thing. Annals of the Entomological Society of America. 78: 836-840.
Starr, C. K. 1990. Holding the fort: colony defense in some primitively social wasps. in Evans, D. L. and J. O. Schmidt. Insect defenses. State University of New York Press, Albany. 421-461.
Stern, D. L. 1994. A phylogenetic analysis of soldier evolution in the aphid family Hormaphididae. Proc. R. Soc. Lond. 256: 203-209
Stern, D. L., S. Aoki, and U. Kurosu.1994. A test of geometric hypothesis for soldier investment patterns in the gall producing tropical aphid Cerataphis franssene (Homoptera, Hormaphididae). Ins. Soc. 41: 457- 460.
Stern, D. L., and W. A. Foster. 1996. The evolution of sterile soldiers in aphids. Biol Rev. 71: 27-79
Stern, D. L., A. Moon, and C. Martinez del Rio. 1996. Caste allometries in the soldier-producing aphid Pseudoregma alexanderi (Hormaphididae: Aphidoidae). Ins. Soc. 43: 137-147.
Sunose, T., S. Yamane, T. Kiyoshi, and K Takasu. 1991. What do the soldiers of Pseudoregma bambucicola (Homoptera: Pemphigidae) defend? Jpn. J. Ent. 59: 141-148
Tanaka, S. Reversal of caste production schedule in a eusocial aphid, Pseudoregma koskunensis. Naturewissen. 81, 411-413
Tanaka, S. and Y. Ito. 1995. Interrelationships between the eusocial aphid, Pseudoregma bambucicola, and its Syrphid predator, Eupeodes confrater. Jpn. J. Ent. 63: 221-228
