Current Biology 19, 740–744, May 12, 2009 ª2009 Elsevier Ltd All rights reserved

DOI 10.1016/j.cub.2009.03.015

Report Communal Nutrition in Ants

Audrey Dussutour1,2,* and Stephen J. Simpson1 of Biological Sciences and Centre for Mathematical Biology The University of Sydney NSW 2006 Australia 2Centre de Recherches sur la Cognition Animale UMR 5169 CNRS Universite´ Paul Sabatier 31062 Toulouse France

1School

Summary Studies on nonsocial insects have elucidated the regulatory strategies employed to meet nutritional demands [1–3]. However, how social insects maintain the supply of an appropriate balance of nutrients at both a collective and an individual level remains unknown. Sociality complicates nutritional regulatory strategies [4–6]. First, the food entering a colony is collected by a small number of workers, which need to adjust their harvesting strategy to the demands for nutrients among individuals within the colony [4–7]. Second, because carbohydrates are used by the workers and proteins consumed by the larvae [7–14], nutritional feedbacks emanating from both must exist and be integrated to determine food exploitation by foragers [4–6, 15, 16]. Here, we show that foraging ants can solve nutritional challenges for the colony by making intricate adjustments to their feeding behavior and nutrient processing, acting both as a collective mouth and gut. The amount and balance of nutrients collected and the precision of regulation depend on the presence of larvae in the colony. Ants improved the macronutrient balance of collected foods by extracting carbohydrates and ejecting proteins. Nevertheless, processing excess protein shortened life span—an effect that was greatly ameliorated in the presence of larvae. Results and Discussion We first aimed to establish whether there is a ratio and a rate of protein and carbohydrate collected that are maintained in the face of variation in the nutritional environment. Accordingly, 20 colonies of green-headed ants, 10 with larvae present from the start and 10 without, were challenged to demonstrate whether they have the capacity to regulate intake of protein and carbohydrate when offered a total of six different twofood choices, varying in the ratio and concentration of protein and carbohydrate. Achieving the same intake of protein and carbohydrate in the face of six different complementary food pairings would prove that ants have the capacity to regulate both protein and carbohydrate collection. The choices were (1) 3:1 versus 1:2 protein to carbohydrate (300 g/l), (2) 3:1 versus 1:2 (200 g/l), (3) 3:1 versus 1:2 (100 g/l), (4) 2:1 versus 1:3 (300 g/l),

*Correspondence: [email protected]

(5) 2:1 versus 1:3 (200 g/l), and (6) 2:1 versus 1:3 (100 g/l). Despite being provided with two different nutrient ratios and a 3-fold range of nutrient concentrations, colonies with larvae maintained the ratio and amounts of protein and carbohydrate collected remarkably consistently (Figure 1 and Tables S1 and S2 available online). Nutrient intake differed in three respects according to the presence of larvae. First, colonies with larvae maintained a higher nutrient intake than colonies without larvae, i.e., they ate more food. Second, the regulated P:C ratio was more protein biased for colonies with larvae than for those without larvae. Third, whereas colonies with larvae were able to maintain nutrient intake constantly across a 3-fold range of nutrient dilutions, the colonies without larvae maintained nutrient intake when the diet was diluted from 300 to 200 g/l but were unable to do so when the foods were further diluted to 100 g/l. Hence, the presence of larvae changes the nutritional requirements of the colony but also contributes to the effectiveness of nutritional regulation. Having established that colonies are able to switch among nutritionally imbalanced but complementary foods to maintain the supply of protein and carbohydrate, a second experiment explored the responses of colonies when confined to a single diet containing an excess of one nutrient relative to the other. In this experiment, the ants were forced to ingest foods that were to some extent imbalanced and confront the situation wherein there is conflict between meeting their requirements for protein and carbohydrates. We confined 30 colonies with larvae at the start and 30 colonies without larvae initially to one of five diets varying in P:C (1:3, 1:2, 1:1, 2:1, and 3:1—all at 200 g/l for P + C). Food collected was measured for each colony over 2 day periods across 50 days. The resulting nutrient collection arrays are shown in Figures 2A–2D and indicate three important points. First, the amount of food collected was higher for colonies with larvae initially than for colonies without larvae (Figure 2 and Table S3). Second, colonies without larvae from the beginning of the experiment collected substantial excesses of protein when fed with the highest P:C diet (3:1), presumably in an effort to maintain a constant carbohydrate intake. The tendency for the array to run parallel to the protein axis in Figure 2A indicates that all colonies managed to collect almost the same quantity of carbohydrates regardless of the diet that they were fed. Third, colonies with larvae made greater efforts to maintain protein intake than did those without larvae. Thus, on the lowest P:C diet (1:3), collection of food increased to provide limiting protein (indicated by the pronounced kink upward in the intake array relative to 1:2 in Figure 2B), whereas, on the highest P:C diet (3:1), food collection was reduced (at least up until day 18), hence ameliorating excess protein collection (shown by the kink downward in the intake array relative to 2:1). Beyond day 18 on diet 3:1 and day 32 on 2:1, all larvae had died, indicating a cost to the larvae of excess protein collection (see below), and thereafter, these colonies assumed the intake pattern seen for colonies without larvae, i.e., a large increase in food collection (see arrows on Figure 2D). In the course of the experiment, we noticed that ants were storing food in the form of pellets inside the nest before removing them from the nest and stockpiling them in a waste

Communal Nutrition in Ants 741

A

B

Figure 1. Protein and Carbohydrate Collection Measured during Choice Experiments (A and B) Empty circles and crosses represent the amount of protein (P) and carbohydrate (C) collected for each replicate for colonies without larvae and with larvae present from the outset of the experiment, respectively. Full circles and full squares show the mean values. The ‘‘random’’ outcomes indicate the expected intake trajectories if feeding had occurred indiscriminately between the two foods in a food pairing treatment (dotted black lines). The solid black lines represent the observed intake trajectory. The solid gray lines correspond to the two foods in a food pairing treatment. Note how colonies with larvae converged upon the same point of nutrient collection in all treatments. Colonies without larvae regulated nutrient collection to a lower P:C ratio than did colonies with larvae, reflecting the increased protein needs of those colonies, and were unable to compensate for the most extreme dilution, indicating a role of larvae in colony nutrient regulation. Three-way ANOVAs: choice effect on food collected and P:C ratio, F1,108 = 0.38 (p = 0.534) and F1,108 = 0.09 (p = 0.781); dilution effect on P:C ratio, F2,108 = 0.01 (p = 0.994); larval effect on P:C ratio and food collected, F1,108 = 792.79 (p < 0.001) and F1,108 = 939.12 (p < 0.001); dilution 3 larvae effect on food collected, F2,108 = 178.79 (p < 0.001).

dump. Consequently, ants were not eating all of the food that they were harvesting. We collected, dried, and weighed the pellets from the waste dump at the end of the experiment. The number of stockpiled pellets increased as the ratio of protein to carbohydrate in the diet increased, with little or no waste being stockpiled on the lowest P:C diet (Figures 3, S1, S3, and Table S4). In addition, the number of pellets on the dump was twice as large in colonies with larvae than in colonies without larvae. The question arose as to whether the chemical composition of the waste pellets differed from that of the food. Extraordinarily, the chemical composition of the stockpiled pellets changed systematically with dietary P:C. As the proportion of protein increased in the diet, the proportion of protein in the pellets rose much faster (Figure S2 and Table S5). The concentration of carbohydrate was considerably lower in the pellets than in the diet, whereas the quantity of protein was

increasingly higher in the stockpiled waste than in the food as dietary P:C rose. Therefore, ants were manipulating the diet collected, extracting the carbohydrates, and rejecting excess protein in the form of pellets. The presence of larvae affected the extent of this manipulation; in particular, colonies with larvae were more effective in voiding excess collected protein than colonies without larvae when supplied with proteinbiased diets (compare the shaded triangles in Figures 3A and 3B). When adjusted for this postcollection manipulation of diet composition and ejection of unwanted food residues onto the waste dump (nutrients collected minus nutrients rejected), the intake array for colonies with larvae shifted, such that estimated intake of protein varied less across the range of dietary P:C values than did the array based on food collected (Figure 3B). For protein-biased diets, colonies were still supplied with an excess of protein, but not as large as the excess predicted from the food collected. In colonies without larvae from the outset, the corrected intake array was not substantially changed relative to the array for collected nutrients (Figure 3A). When the same adjustments for manipulation of dietary composition were made to foods collected during the choice experiment (Figure 1), the intake target shifted to align with a ratio closer to 1:1.5 for colonies with larvae but 1:2 for colonies without larvae (open red circles in Figure 3). Finally, we measured various performance indicators for the colonies on different diets. Ant mortality increased substantially at the two highest P:C ratios (2:1 and 3:1) (Figures 4A and 4B and Table S6). Surprisingly, this pattern was considerably less pronounced in the presence of larvae from the beginning of the experiment. The number of larvae produced per colony was also a function of dietary P:C. Colonies without larvae at the beginning had produced the most larvae by day 50 when confined to diet 1:2. Colonies with larvae present from the beginning had a higher number of larvae at the end of 50 days than did colonies in which larvae were absent initially (Figure 4C). Larvae introduced into the colony before the beginning of the experiment died after 2–3 weeks when colonies were fed with the most protein-biased diets but survived and metamorphosed in colonies fed with carbohydrate-biased diets (Figure 4D). Regulation of nutrient intake requires assessment of the nutritional quality of available foods and sensors indicating the nutritional state of the regulating entity. In the case of ants, the assessment and collection of food is undertaken on behalf of the colony by the foraging ants, which can be specialized in collecting different food types [17]. The major sink for collected protein is growing larvae, whereas worker ants require mainly carbohydrate for energy. Accordingly, when larvae were present in the colony, ants not only collected more food in total, but protein comprised a higher proportion of the macronutrients collected. This change in the colony-level intake target indicates clearly a feedback emanating directly or indirectly from larvae to foragers. More than this, however, if larvae were present, ants regulated macronutrient intake more precisely when offered choices of nutritionally complementary foods. Notably, in the presence of larvae, ants compensated for a 3-fold change in the concentration of nutrients in the diet by making commensurate adjustments to the amount of food collected. However, when larvae were absent from the beginning of the experiment, ants did not increase food collection on the most diluted foods. In a previous study [18], we showed that, in the face of changes in the concentration of sugar solution, ants were better able to

Current Biology Vol 19 No 9 742

A

B

C

D

regulate their carbohydrate intake when larvae were present rather than when larvae were absent. Not only did ants regulate macronutrient collection more precisely in the presence of larvae, but they also survived better when fed high-protein diets. Excess protein ingestion in relation to requirements has recently been shown to shorten life span in other insects [19, 20]. It has been shown in other ant species that carbohydrates may be digested extraorally because carbohydrase activity is present in salivary and mandibular glands, whereas protein needs to be ingested because protease activity is restricted to the midgut [21–23]. Salivary digestion of carbohydrate in the oral cavities of worker

A

B

Figure 2. Cumulative Protein and Carbohydrates Collected by Ants Colonies were provided with one of five diets that differed in their ratio of protein to carbohydrate at 2 day intervals over 50 days (the results were adjusted to ant mortality). Each diet is represented as a dotted line in the protein/carbohydrates plane in (A) and (B). Within each time interval, the nutrient collection points are connected with lines to and from collection arrays, which demonstrate the nutrient balancing strategy. The inserted images in (A) and (B) show examples of the amount of food left after 2 days. The amount of diet collected by ants at 2 day intervals over 50 days is shown in (C) and (D). Four-way ANOVAs: time effect on the amount of food collected, F24,960 = 106.82 (p < 0.001); time 3 larvae effect, F24,960 = 8.49 (p < 0.001); time 3 diet effect, F96,960 = 6.66 (p < 0.001); time 3 larvae 3 diet effect, F96,960 = 3.14 (p < 0.001); larvae effect, F1,40 = 63.00 (p < 0.001); diet effect, F4,40 = 17.48 (p < 0.001); larvae 3 diet effect, F4,40 = 0.60 (p = 0.661).

ants, either during collection or within the nest, followed by discarding unwanted protein-rich food stock to an external waste dump, resulted in ants improving the macronutrient balance of collected food. Why did ants discard more excess protein when larvae were present from the outset than when they were absent? One would expect that colonies without larvae, which do not need as much protein, would have rejected a larger proportion of excess collected protein on high-protein diets. Their failure to do so indicates once again that the larvae are important in providing nutritional feedbacks to workers. Workers, unlike larvae, have a limited ability to digest bulky proteinaceous foods in the midgut because of a combination of their narrow waist (petiole) separating the thorax from the abdomen [7, 24, 25] and producing only very small amounts of proteases in their midguts [22, 25]. Larvae, in contrast, are capable of

Figure 3. Protein and Carbohydrates Collected and Ingested, i.e., Not Taken out and Placed on a Waste Dump by Ants after 50 Days

(A and B) Ants were provided with one of five diets differing in their ratio of protein to carbohydrate over 50 days. Triangles show the amount of C and P rejected from the food collected. Note how colonies with larvae from the outset were more effective than colonies without larvae in manipulating the composition of collected food by retaining the limiting nutrient and rejecting the excess nutrient in collected food (shaded triangles). The intake target collected is the amount of carbohydrates and protein collected during the choice experiment. The intake target ingested is the intake target once the correction for carbohydrate extraction from the foods collected in the choice experiments is made. Results are presented as intake per ant to adjust for mortality. Three-way ANOVA: diet effect on the amount of stock, F4,40 = 78.61 (p < 0.001); larvae effect, F1,40 = 96.20 (p < 0.001). Four-way ANOVAs with repeated measures: manipulation effect on C and P quantity in the stock, F1,40 = 4068.94 (p < 0.001) and F1,40 = 755.13 (p < 0.001); manipulation 3 larvae effect 3 diet effect on C quantity, F1,40 = 8.45 (p < 0.001).

Communal Nutrition in Ants 743

A

C

Figure 4. Colony Performance

B

Ants were provided with one of five diets differing in their ratio of protein to carbohydrate over 50 days. (A and B) Mortality for colonies with and without larvae from the beginning of the experiment. (C) Number of larvae produced after 50 days. (D) Number of larvae that hatched after 50 days in colonies in which 60 larvae were introduced at the beginning of the experiment. Three-way ANOVAs: diet effect on the number of dead ants, F4,40 = 25.99 (p < 0.001); larvae 3 diet effect, F4,40 = 1.28 (p = 0.292); larvae effect, F1,40 = 7.56 (p = 0.009); diet effect on the number of larvae, F4,40 = 47.42 (p < 0.001); larvae 3 diet effect, F4,40 = 2.29 (p = 0.076); larvae effect, F1,40 = 11.56 (p = 0.002). One-way ANOVA: diet effect on number of ants that hatched, F4,29 = 82.81 (p < 0.001).

D

protein digestion both extraorally through high protease levels in labial gland secretions and in the midgut [25]. This difference between workers and larvae explains why preys are fed to ant larvae in an undigested state [26]. Our results suggest that the ants may further overcome the deleterious effects of excess proteins by getting the larvae to process them, confirming the hypothesis that larvae are not passive recipients of nutrition but, rather, a protein digestive organ for the colony [4, 11, 27, 28]. Excess dietary protein, nevertheless, resulted in poor larval performance (both survival and production). Ants that were offered a choice of foods regulated their collection of macronutrients to a ratio of protein to carbohydrate that appeared to maximize neither ant life span nor larval survival and production. However, once the correction for carbohydrate extraction from the foods collected by the ants in the choice and no-choice experiments is made, the macronutrient balance shifts toward the protein-to-carbohydrate ratio that best supported larval production and sustained maximal larval life span, though it is still somewhat protein biased. The reason for this partial misalignment between the ratio of nutrients ingested and that which provided maximum performance under no-choice conditions is unclear but is likely to relate to the fact that the choice data were collected during a shorter period (2 day periods for each of 6 diet pairings) than were the performance results (50 days), and the target protein-tocarbohydrate ratio may well have changed over time. A striking division of labor has occurred in ants regarding nutritional regulation. First, the foragers, responding to the relationship between colony needs and the nutritional environment, harvest food and prepare it by extracting carbohydrates. Second, protein in the preprocessed food is digested by the larvae, used for larval growth, and presumably fed back to sustain the protein requirements of workers [4, 7, 27]. Thereafter, residual excess protein is removed from the colony and dumped. The nutritional interdependence among the different actors within the colony is fully consistent with the proposed importance of nutrition in the maintenance of

sociality, in caste determination, and in the initial evolution of sociality [29–31]. More broadly, our results provide an example of how nutrient-specific interactions between individuals can lead to complex collective behaviors, just as they can explain collective behavior in simpler group living organisms [32–35]. Experimental Procedures Full details of the Experimental Procedures are provided in the Supplemental Experimental Procedures. Species Colonies of monomorphic green-headed ants (Rhytidoponera sp.) [36] were collected in Sydney, Australia. Choice Diet Experiment In the choice experiment, we housed 20 colonies of 250 ants in experimental nests [18, Supplementary Data]. Before starting the experiment, we added 100 larvae to 10 colonies. We prepared 12 foods varying in both the P:C ratio and the total concentration P + C [37]. The 20 colonies were provided with the following food choices, presented in random order at 2 day intervals: 3:1 versus 1:2 (300 g/l), 3:1 versus 1:2 (200 g/l), 3:1 versus 1:2 (100 g/l), 2:1 versus 1:3 (300 g/l), 2:1 versus 1:3 (200 g/l), and 2:1 versus 1:3 (100 g/l). Diet consumption was measured as described below. No-Choice Diet Experiment In the no-choice experiment, we housed 60 colonies of 250 workers in experimental nests [18]. Before beginning the experiment, we added 60 larvae to 30 colonies. We prepared five diets differing in their content of protein (P) and carbohydrates (C) [37]. The 5 P:C ratios used were 3:1, 2:1, 1:1, 1:2, and 1:3. The total concentration of protein and carbohydrates (P + C) was 200 g/l. Every 2 days for 50 days, 6 colonies with larvae and 6 colonies without received 8 mg of one of the 5 diets in a Petri dish. The Petri dish was weighed every 2 days before it was placed in the foraging arena and again after it was removed. The quantities of carbohydrates and protein present in the food stored by the ants and in the diet provided to the colonies were measured for each colony with the Phenol-sulphuric assay and the Bradford assay, respectively. The number of larvae was estimated every week and measured accurately after 50 days. The number of dead ants within each colony was counted every 2 days, and corpses were removed. Supplemental Data Supplemental Data include Supplemental Experimental Procedures, three figures, and seven tables and can be found with this article online at http:// www.cell.com/current-biology/supplemental/S0960-9822(09)00818-5.

Current Biology Vol 19 No 9 744

Acknowledgments A.D. was supported by a postdoctoral grant from The University of Sydney. S.J.S. was supported by an ARC Federation Fellowship. We thank Z. Osborn for helping us collect ants and N. Soran and F. Clissold for helping us with the chemical analysis. We also thank D. Rubenstein for his comments on the manuscript. Received: January 9, 2009 Revised: February 12, 2009 Accepted: March 2, 2009 Published online: April 2, 2009 References 1. Raubenheimer, D., and Simpson, S.J. (1999). Integrating nutrition: A geometrical approach. Entomol. Exp. Appl. 91, 67–82. 2. Simpson, S.J., and Raubenheimer, D. (2000). The hungry locust. Adv. Stud. Behav. 29, 1–44. 3. Simpson, S.J., Sibly, R.M., Lee, K.P., Behmer, S.T., and Raubenheimer, D. (2004). Optimal foraging when regulating intake of multiple nutrients. Anim. Behav. 68, 1299–1311. 4. Cassill, D.L., and Tschinkel, W.R. (1999). Information flow during social feeding in ant societies. In Information Processing in Social Insects, C. Detrain, J.L. Deneubourg, and J.M. Pasteels, eds. (Basel, Switzerland: Birkha¨user Verlag), pp. 69–81. 5. Seeley, T.D. (1995). The Wisdom of the Hive: the Social Physiology of Honey Bee Colonies (Cambridge, Massachusetts: Harvard University Press). 6. Schmickl, T., and Crailsheim, K. (2004). Inner nest homeostasis in a changing environment with special emphasis on honey bee brood nursing and pollen supply. Apidologie (Celle) 35, 249–263. 7. Ho¨lldobler, B., and Wilson, E.O. (1990). The Ants (Cambridge, Massachusetts: Harvard University Press). 8. Brian, M.V., and Abbott, A. (1977). The control of food flow in a society of the ant Myrmica rubra L. Anim. Behav. 25, 1047–1055. 9. Sorensen, A.A., Busch, T.M., and Vinson, S.B. (1985). Control of food influx by temporal subcastes in the fire ant, Solenopsis invicta. Behav. Ecol. Sociobiol. 17, 191–198. 10. Sorensen, A.A., and Vinson, S.B. (1981). Quantitative food distribution studies within laboratory colonies of the imported fire ant, Solenopsis invicta Buren. Insect. Soc. 28, 129–160. 11. Markin, G.P. (1970). Food distribution within laboratory colonies of the Argentine ant, Iridomyrmex humilis (Mayr). Insect. Soc. 17, 127–158. 12. Howard, D.F., and Tschinkel, W.R. (1981). The flow of food in colonies of the fire ant, Solenopsis invicta: A multifactorial study. Physiol. Entomol. 6, 297–306. 13. Abbott, A. (1978). Nutrient dynamics of ants. In Production Ecology of Ants and Termites, M.V. Brian, ed. (London: Cambridge University Press), pp. 233–244. 14. Cassill, D.L., and Tschinkel, W.R. (1999). Regulation of diet in the fire ant, Solenopsis invicta. J. Insect Behav. 12, 307–328. 15. Portha, S., Deneubourg, J.L., and Detrain, C. (2002). Self-organized asymetries in ant foraging: A functional response to food type and colony needs. Behav. Ecol. 13, 776–781. 16. Creemers, B., Billen, J., and Gobin, B. (2003). Larval begging behaviour in the ant Myrmica rubra. Ethol. Ecol. Evol. 15, 261–272. 17. Quinet, Y., and Pasteels, J.M. (1996). Spatial specialization of the foragers and foraging strategy in Lasius fuliginosus (Latreille) (Hymenoptera, Formicidae). Insect. Soc. 43, 333–346. 18. Dussutour, A., and Simpson, S.J. (2008). Carbohydrate regulation in relation to colony growth in ants. J. Exp. Biol. 211, 2224–2232. 19. Lee, K.P., Simpson, S.J., Clissold, F., Brooks, R., Ballard, J.W.O., Taylor, P.W., Soran, N., and Raubenheimer, D. (2008). Lifespan and reproduction in Drosophila: New insights from nutritional geometry. Proc. Natl. Acad. Sci. USA 105, 2498–2503. 20. Maklakov, A.A., Simpson, S.J., Zajitschek, F., Hall, M., Dessman, J., Clissold, F., Raubenheimer, D., Bonduriansky, R., and Brooks, R.C. (2008). Sex-specific fitness effects of nutrient intake on reproduction and lifespan. Curr. Biol. 14, 1062–1066. 21. Ayre, G.L. (1967). The relationships between food and digestive enzymes in five species of ants (Hymenoptera: Formicidae). Can. Entomol. 99, 408–411.

22. Ricks, B.L., and Vinson, S.B. (1972). Digestive enzymes of the imported fire ant, Solenopsis richteri (Hymenoptera: Formicidae). Entomol. Exp. Appl. 15, 329–334. 23. Josens, R.B., Farina, W.M., and Roces, F. (1998). Nectar feeding by the ant Camponotus mus: Intake rate and crop filling as a function of sucrose concentration. J. Insect Physiol. 44, 579–585. 24. Glancey, B.M., Vander Meer, R.K., Glover, A., Lofgren, C.S., and Vinson, S.B. (1981). Filtration of microparticles from liquids ingested by the red imported fire ant, Solenopsis invicta Buren. Insect. Soc. 28, 395–401. 25. Petralia, R.S., Sorensen, A.A., and Vinson, S.B. (1980). The labial gland system of larvae of the imported fire ant, Solenopsis invicta Buren: Ultrastructure and enzyme analysis. Cell Tissue Res. 206, 145–156. 26. Cassill, D.L., Butler, J., Vinson, S.B., and Wheeler, D.E. (2005). Cooperation during prey digestion between workers and larvae in the ant, Pheidole spadonia. Insect. Soc. 52, 339–343. 27. Went, F.W., Wheeler, J., and Wheeler, G.C. (1972). Feeding and digestion in some ants (Veromessor and Manica). Bioscience 22, 82–88. 28. Sorensen, A.A., Kamas, R.S., and Vinson, S.B. (1983). The influence of oral secretions from larvae on levels of proteinases in colony members of Solenopsis invicta Buren (Hymenoptera: Formicidae). J. Insect Physiol. 29, 163–168. 29. Hunt, J.H. (2007). The Evolution of Social Wasps: History, Dynamics, and Paradigm (New York: Oxford University Press). 30. Hunt, J.H., and Amdam, G.V. (2005). Bivoltinism as an antecedent to eusociality in the paper wasp genus Polistes. Science 308, 264–267. 31. Page, R.E., and Amdam, G.V. (2007). The making of a social insect: Developmental architectures of social design. Bioessays 29, 334–343. 32. Mas, F., and Kolliker, M. (2008). Maternal care and offspring begging in social insects: Chemical signalling, hormonal regulation and evolution. Anim. Behav. 76, 1121–1131. 33. Costa, J.T. (2006). The Other Insect Societies (Cambridge, MA: Harvard University Press). 34. Choe, J.C., and Crespi, B.J. (1997). The Evolution of Social Behaviour in Insects and Arachnids (New York: Cambridge University Press). 35. Simpson, S.J., Sword, G.A., Lorch, P.D., and Couzin, I.D. (2006). Cannibal crickets on a forced march for protein and salt. Proc. Natl. Acad. Sci. USA 103, 4152–4156. 36. Ward, P.S. (1981). Ecology and life history of the Rhytidoponera impressa group (Hymenoptera: Formicidae). I. Habitats, nest sites, and foraging behavior. Psyche (Stuttg.) 88, 89–108. 37. Dussutour, A., and Simpson, S.J. (2008). Description of a simple synthetic diet for studying nutritional responses in ants. Insect. Soc. 55, 329–333.