Formica Fusca)

bs_bs_banner

Biological Journal of the Linnean Society, 2014, 113, 415–422. With 4 ﬁgures

Evidence for adaptive brain tissue reduction in obligate social parasites (Polyergus mexicanus) relative to their hosts (Formica fusca)

ELISABETH SULGER1, NOLA MCALOON1, SUSAN J. BULOVA1, JOSEPH SAPP2 and SEAN O’DONNELL1*

1Department of Biodiversity, Earth & Environmental Science, Drexel University, Philadelphia, PA 19104, USA 2Department of Ecology & Evolutionary Biology, University of California-Santa Cruz, Santa Cruz, CA 95064, USA

Received 8 May 2014; revised 2 June 2014; accepted for publication 2 June 2014

Brain investment is evolutionarily constrained by high costs of neural tissue. Several ecological factors favour the evolution of increased brain investment; we predict reduced brain region investment will accompany the evolution of organismal or social parasitism when parasites rely on host behaviour and cognition to solve ecological problems. To test this idea we investigated whether brain region investments differed between obligate slave-making Polyergus mexicanus ant workers and their Formica fusca slave workers. Polyergus workers perform little labour for their colonies; enslaved workers of Formica host species forage, excavate nests and tend the brood. We focused on the calyces of the mushroom bodies, central processing brain regions that are larger in social insect workers that perform complex tasks. As predicted we found lower relative investment in mushroom body calyx in P. mexicanus workers than in F. fusca workers; by contrast, enslaved and free F. fusca workers did not differ in mushroom body calyx volume. We then tested whether slave-makers and hosts differed in brain investment among sensory modalities. Polyergus slave-makers employ several unique classes of pheromones during raids, and eye size relative to head size was smaller in P. mexicanus workers than in F. fusca workers. The size of antennal brain tissues relative to visual tissues was greater in Polyergus, both in the peripheral sensory lobes and in the mushroom body calyx, suggesting greater relative investment in antennal processing by slave-makers. © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 415–422.

ADDITIONAL KEYWORDS: antennal lobes – Formicidae – mushroom bodies – slavemaking ants.

INTRODUCTION correlates of evolved reductions in neural tissue. This pattern holds for changes in sensory ecology. The The brain is among the most metabolically and devel- evolution of activity under lower light conditions is opmentally costly organs, and these costs place limits associated with decreases in visual processing tissue on the evolution of greater brain size (Laughlin, 2001; (Barton, Purvis & Harvey, 1995; Catania, 2005; Fujun Niven & Laughlin, 2008; Navarrete, van Schaik & et al., 2012; O’Donnell et al., 2013). We hypothesize Isler, 2011). Greater brain investment is expected to that the evolution of parasitism will lead to reduced follow the evolution of increased cognitive demands. brain investment when parasites exploit host sensory Evolutionary increases in sociality and in foraging on and cognitive systems. Brain reductions in parasites complex resources are associated with the evolution should be restricted to speciﬁc brain regions. Only of increased brain investment (Dunbar & Shultz, those brain regions that govern host sensory or cog- 2007; Farris 2008a; West, 2014). The converse should nitive functions the parasite exploits will be reduced also hold: it should be possible to identify ecological in the parasite (mosaic brain evolution: Chittka & Niven, 2009; Shultz & Dunbar, 2010; Smaers & *Corresponding author. E-mail: [email protected] Soligo, 2013).

Organismal host/parasite interactions are mirrored (Visicchio et al., 2001). As in other social in many ways by social parasites. Social parasites Hymenoptera, the ant brain is divided among ana- exploit the cooperative social behaviour of their hosts tomically distinct regions that process sensory input (Kilner & Langmore, 2011). We investigated whether from the antennae (mainly olfactory information) and a unique form of social parasitism – slave-making in visual input from the compound eyes (Gronenberg, ants – was associated with a reduction in brain 1999, 2001; Gronenberg & Hölldobler, 1999). We pre- investment. Workers of obligate slave-making ants dicted the greater reliance of slave-makers on chemi- (for example, species in the genus Polyergus) have cal communication relative to their slaves would be limited labour abilities. Polyergus workers do not reflected in brain architecture. Slave-maker workers excavate nests, forage for food, tend brood or feed should have larger brain regions for antennal percep- themselves (Topoff, 1990; D’Ettorre & Heinze, 2001). tion relative to visual brain regions. Polyergus workers capture immature workers of several Formica species, and the enslaved Formica ants perform most tasks for the slave-makers MATERIAL AND METHODS (Hölldobler & Wilson, 1990; D’Ettorre & Heinze, SUBJECT SPECIMEN COLLECTIONS 2001). Specimens were collected in the field at Sagehen Brain regions involved in task performance should Creek Field Station from mid-altitude (approximately be reduced in volume in Polyergus workers relative to 2000 m above sea level) mixed-conifer forests in the their Formica slave workers. We focused on the Sierra Nevada range of California, USA (39°25.92′N, neuropil of the mushroom bodies, the higher-order 120°14.47′W) on 20–22 July 2012. We collected ants processing region of insect brains (Strausfeld et al., from six P. mexicanus raids performed by different 1998, 2009; Davis, 2005; Farris, 2008b). The mush- slave-maker colonies on different host nests. We col- room body calyx increases in volume and neuronal lected 1–16 adult P. mexicanus slave-makers and complexity when social insect workers perform approximately 20 enslaved F. fusca workers from the complex tasks such as foraging (Gronenberg, Heeren same colonies during raids on F. fusca nests. We also & Hölldobler, 1996; Farris, Robinson & Fahrbach, collected approximately 20 F. fusca workers from the 2001; O’Donnell, Donlan & Jones, 2004; Seid, Harris nests that were being raided. The ages and behav- & Traniello, 2005; Jones, Donlan & O’Donnell, 2009; ioural histories of the subjects were not known, but Muscedere & Traniello, 2012). We predicted relative all were performing behaviours typical of mature investment in mushroom body neuropils of Polyergus workers. Polyergus were participating in raids, workers would be reduced relative to their enslaved enslaved F. fusca were active receiving brood and Formica workers. foraging above-ground at their home nests, and free To build on the slave-maker/slave comparison, and F. fusca were defending at the raided nest. Defence as a partial control for shared nest environments, we and foraging are the final tasks performed by social also compared brain architecture between free and insects with age-related division of labour, suggesting enslaved Formica workers. If exposure to distinct all of our subjects were older workers (Wilson, 1971). nest environments caused brain investment differences between slaves and slave-makers, then free and enslaved F. fusca workers should also differ in brain HISTOLOGY AND NEUROANATOMY architecture. Alternatively, if task performance plays We collected neuroanatomical data on slave-makers, a major role in determining mushroom body calyx slaves and free workers from each raiding colony/host size, free and enslaved Formica workers should have colony pair. We analysed brain architecture of N =15 similar brain architecture because they perform P. mexicanus (1–5 per colony), N = 19 enslaved similar sets of tasks, even though they occupy differ- F. fusca (2–5 per colony) and N = 22 free F. fusca ent nests and distinct social environments. workers (N = 3–5 per colony). All ants were collected Although they have a limited task repertoire, directly into aldehyde-based fixative (Prefer; Anatech Polyergus workers perform several specialized behav- Ltd). Workers were stored in fixative at 4 °C for iours related to host exploitation, including scouting approximately 10 months until histological process- for host nests, recruiting nest mates to raids and ing. We cut each ant’s head capsule from the body at attacks on host nests (Topoff, Cover & Jacobs, 1989; the narrow neck-like juncture with the thorax and Topoff, 1990). These specialized behavioural acts may removed the antennae by cutting. Prior to histological involve substantial sensory and cognitive processing. processing we photographed each head capsule lying Slave-maker workers may be more reliant on chemi- flat in a Petri dish from the frontal plane using a cal communication than their Formica host workers digital camera mounted on a dissecting microscope because Polyergus employ several functionally dis- at 24–56× magnification and digital image resolution tinct pheromones during raids on Formica nests of 2048 × 1536 pixels (Fig. 1). We estimated head

Figure 1. Left, digital photographs of head capsules of Formica fusca (A) and Polyergus mexicanus (C) workers. Right, representative stained thin sections through head capsules showing brain tissue of F. fusca (B) and P. mexicanus (D). Left-hand scale bar refers to A and C; right-hand scale bar refers to B and D. Some brain regions are labelled for one hemisphere of the brain in the F. fusca section (B): MB, calyx–mushroom body calyx neuropil; OL, optic lobe; Pr, main body of the protocerebrum; AL, antennal lobe. Corresponding structures are visible in the P. mexicanus section (D).

capsule width by measuring the widest point of the the tissue sections, with a digital image resolution of head capsule using the ruler tool to count pixels in 2048 × 1536 pixels (Fig. 1). For each ant worker we ImageJ version 1.46 digital imaging analysis software photographed every section starting at the section (http://rsbweb.nih.gov/ij/); we converted pixels to where brain tissue first became visible. ImageJ version length in millimetres using photographs of a 1-mm 1.46 software was used to quantify the volumes of stage micrometer taken at the same magnifications brain structures. To quantify brain regions on each with the same camera and microscope. section we outlined the target brain regions and used We dehydrated head capsules through an ethanol ImageJ to count the number of image pixels in the series, acetone and then increasing concentrations structure. We converted the pixel counts to area using of plastic resin. Resin comprised 45% (by weight) a photograph of a stage micrometer taken at the same Embed812 [a mixture of bisphenol A/epichchlo- magnification with the same microscope and camera as rohydrin epoxy resin (CAS#25068-38-6) and epoxy a size reference, then multiplied the areas by section modifier (CAS#2425-79-8)], 47% DDSA (dodecenyl thickness to yield a volume estimate. Only brain succinic anhydride), 5.5% DBP (dibutyl phthalate) and neuropils were measured; we did not measure adjacent 2.5% DMP-30 [2,4,6-(tri(dimethylaminoethyl)phenol)]. cell body regions. We measured the volumes of the Individual ant heads were incubated in 0.2 mL resin in following brain sub-regions: parts of the optic lobes pyramid-shaped flexible moulds at 60 °C for 72 h. The (the medulla and the lobula), the antennal lobes, and solid resin pyramids were mounted on acrylic posts. the mushroom body calyx lip and collar (Fig. 1). Other Each head was sectioned into 12-μm-thick slices using brain regions were pooled as an index of brain a rotary microtome with disposable steel histology size: mushroom body peduncle and lobes, central blades. We placed the sections on gelatin-coated micro- complex, and the remainder of the protocerebrum, scope slides and stained the tissue with Toluidine blue. deutocerebrum and tritocerebrum. We cleared in an ethanol series and cover slipped We used the ratio of mushroom body calyx volume/ under transparent mounting medium. brain size index as a measure of investment in mush- We used a digital camera mounted on a compound room body tissue. We used the ratios of antennal light microscope using the 10× objective to photograph processing to visual processing brain region volumes

© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 415–422 418 E. SULGER ET AL. as a measure of relative investment in antennal 0.07 versus visual processing (Gronenberg et al., 1996; 3 O’Donnell et al., 2013). We pooled the antennal lobes 0.06 with the mushroom body calyx lip to estimate antennal brain investment, and the optic lobes and 0.05 mushroom body calyx collar to estimate visual brain 0.04 investment. 0.03

STATISTICS 0.02 All analyses were performed with SPSS version 20 Total brain volume, mm 0.01 software. We used general linear models (GLM) to analyse species and free vs. enslaved worker differ- 0.8 1.0 1.2 1.4 1.6 ences in head capsule width and brain region volume Head width, mm ratios. Because subject workers were not randomly Figure 2. Scatter plots showing relationships between selected from colonies, we accounted for the effects of brain volume and head capsule size for ant workers: free colony of origin as a random factor; colony identity Formica fusca (black), enslaved F. fusca (grey) and was entered first in all statistical models. All Polyergus mexicanus (white). Lines are linear regression F-statistics significance values we report are based on best fits for each worker category computed separately for Type II sums of squares: the effect of a predictor illustration purposes: free F. fusca (black line), enslaved variable in the order it was entered in the statistical F. fusca (grey line) and P. mexicanus (dashed line). model (in other words, after accounting for colony effects). We used Duncan’s multiple range test to make post-hoc comparisons among group means; the 3 0.014 test identified sets of homogeneous groups (in this 0.012 case, species or behavioural classes) whose means did not differ from each other significantly with critical 0.010 alpha = 0.05. The geometric mean of group sizes was used in making post-hoc comparisons because sample 0.008 sizes were unequal. We used analysis of covariance (ANCOVA) to test whether slopes of brain region 0.006 relationships (calyx volume versus brain size) differed between worker categories. 0.004 0.002 Mushroom body calyx volume, mm RESULTS 0.01 0.02 0.03 0.04 HEAD CAPSULE SIZE AND BRAIN VOLUME Brain remainder volume, mm3

Polyergus mexicanus slave-makers and F. fusca Figure 3. Scatter plots showing relationships between workers in our sample differed significantly in mushroom body calyx volume and remaining brain volume head capsule width, although there was substantial (brain size correction) for ant workers: free Formica fusca overlap between species in this index of body size (black), enslaved F. fusca (grey) and Polyergus mexicanus < (Fig. 2, F2,49 = 50.13, P 0.001; Duncan’s multiple (white). Lines are linear regression best fits for each range post-hoc test: P. mexicanus had wider heads worker category computed separately for illustration pur- than both slave and free F. fusca, which did not differ poses: free F. fusca (black line), enslaved F. fusca (grey from each other). However, these size differences line) and P. mexicanus (dashed line). were not associated with overall brain size differences. Total brain volume increased with head width SPECIES DIFFERENCES IN MUSHROOM BODY SIZE across all species (F1,46 = 17.10, P < 0.001), but total brain volume did not differ significantly between Formica fusca workers invested more in the mush-

P. mexicanus and F. fusca workers (F2,49 = 0.11, room body calyx (relative to the rest of the brain)

P = 0.90). Furthermore, the slope of the brain volume than P. mexicanus slave-makers (Fig. 3; F2,48 = 7.13, relationship with head width was similar across P = 0.002; post-hoc univariate contrasts: enslaved P. mexicanus and both F. fusca worker classes, sug- F. fusca differed from P. mexicanus, P = 0.002, free gesting brain size scales similarly to body size in F. fusca differed from P. mexicanus, P = 0.001). Mean these species (Fig. 2; ANCOVA non-signiﬁcant species mushroom body calyx volume was 30.3% of the rest of by head width interaction term F2,46 = 0.02, P = 0.99). brain volume in P. mexicanus workers, 15% lower

A 0.55 B 0.16

0.50 0.14

0.12

0.45 3 0.10 0.40 0.08 0.35 volume, mm 0.06

Mean eye height mm 0.30 0.04 0.25 Antennal processing regions 0.02 0.6 0.8 1.0 1.2 1.4 1.6 0.02 0.04 0.06 0.08 0.10 Head width mm Visual processing regions volume, mm3

Figure 4. A, scatter plot showing the relationship of eye height to head width for ant workers: free Formica fusca (black), enslaved F. fusca (grey) and Polyergus mexicanus (white). B, scatter plots showing the relationship of total volume of antennal processing brain regions with total volume of visual processing brain regions for ant workers: free Formica fusca (black), enslaved F. fusca (grey) and P. mexicanus (white). than the 35.7% mean relative mushroom body calyx reduced task performance repertoire of Polyergus volume in F. fusca slave workers. Mushroom body workers relative to their enslaved Formica host calyx volume increased with brain size at a lower workers (Topoff et al., 1985b; Topoff, 1990). Task per- rate in P. mexicanus slave-makers than in F. fusca formance, especially foraging, is associated with hosts (Fig. 3, ANCOVA interaction term F2,50 = 7.23, increased mushroom body volume in many social P = 0.002). Hymenoptera, including ants (Withers, Fahrbach & Robinson, 1995; Gronenberg et al., 1996; Farris et al., 2001; O’Donnell et al., 2004). Lower mushroom body SENSORY STRUCTURES IN THE BRAIN AND EYE SIZE calyx investment by Polyergus workers may be per- Polyergus workers invested more in antennal process- mitted by their reliance on host workers for complex ing relative to visual processing tissue than both free task performance. Studies comparing independently and enslaved F. fusca workers (antennal lobe/optic evolved sets of slaves and slave-makers (D’Ettorre & < lobe ratio, F2,49 = 9.87, P 0.001; Duncan’s multiple Heinze, 2001) are needed to test the generality of this range test indicated P. mexicanus differed from both pattern. In contrast, we found no evidence for differ- slave and free F. fusca, which did not differ from each ences in brain architecture between enslaved and other in the antennal/optic lobe ratio; mushroom body free-living F. fusca workers. The comparison of free calyx lip/collar ratio, F2,49 = 6.58, P = 0.003; Duncan’s and enslaved host workers is important because it multiple range test indicated P. mexicanus differed suggests the reduced mushroom bodies of Polyergus from both slave and free F. fusca, which did not differ did not result from common exposure to the slave- from each other in the lip/collar ratio). Polyergus maker nest environment, although there may be dif- workers had smaller eyes relative to head capsule ferences in neuronal activity and function between size than F. fusca workers (slaves and free workers free and enslaved F. fuscsa workers that were not < pooled) (Fig. 4A; F1,48 = 80.07, P 0.001). The species reﬂected in brain anatomy (Lihoreau, Latty & difference was mainly driven by greater antennal Chittka, 2012). structure size in P. mexicanus (Fig. 4B; F. fusca slaves We found relative investment in antennal process- and free workers pooled; F1,54 = 12.60, P = 0.001); the ing brain tissues was greater for Polyergus than for species did not differ signiﬁcantly in the volume of Formica workers. The reduced cost of Polyergus visual processing brain tissues (Fig. 4B; F1,54 = 2.10, mushroom body investment may permit increased P = 0.15). growth of other brain regions such as antennal processing tissues. Despite their limited task performance activities, Polyergus workers do have well- DISCUSSION developed, albeit specialized, cognitive abilities. We demonstrated reduced investment in mushroom Slave-making ant workers scout for host colonies to bodies for social parasite ants relative to their host raid and scouts lead raid parties back to the potential workers. We predicted this pattern based on the slave sources they discovered (Alloway, 1979; Topoff

© 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113, 415–422 420 E. SULGER ET AL. et al., 1985a; Topoff, 1990). Workers employ a battery not clear (Boerner & Kruger, 2008; Corfield et al., of chemical compounds during raids on Formica 2013). Slave-makers and their hosts are well suited nests. Polyergus use trail pheromones and propa- to comparative analysis because slave-makers are ganda substances that disrupt Formica colony closely phylogenetically related to their hosts (Brady defences (Visicchio et al., 2001). This suggests greater et al., 2006; Moreau et al., 2006; Buschinger, 2009). In reliance on chemical communication and chemical part this pattern may be due to the fact that ecologi- cognition accompanied the evolution of Polyergus cal needs and behavioural abilities of slaves must slave-making. Reduced eye size in P. mexicanus prob- closely match those of their slave-makers so that ably reflects lower use of visual information, although enslaved host workers can successfully rear slave- greater investment in Polyergus antennal brain maker offspring. There are approximately 50 known tissues, rather than decreased visual tissue size, dis- species of obligate slave-making ants among the over tinguished slave-makers from F. fusca hosts. 14 000 described ant species and obligate slave- Formica slaves shared social and nest environ- making evolved independently over ten times in ments with their slave-makers, yet the slaves had diverse ant taxa (D’Ettorre & Heinze, 2001). The brain architecture resembling their free conspecifics. independent origins of obligate slave-making involve These patterns suggest the reduced mushroom body convergent reductions in slave-maker task reper- calyx investment of Polyergus slave-makers was asso- toires, with enslaved workers performing most tasks. ciated with their reduced task repertoires. However, We predict there will be reductions in slave-maker our data cannot establish whether the mushroom mushroom body calyx investment relative to the body and antennal investment differences resulted enslaved workers in independently derived slave- from brain developmental plasticity in response to maker ant taxa. different task performance experiences, or whether the brain structure differences result from fixed ACKNOWLEDGEMENTS (experience-independent) species patterns of brain growth. Studies tracking brain architecture changes Several anonymous reviewers made helpful com- with worker age and experience could be used to ments on earlier versions of the manuscript. Beatriz distinguish between evolved and plastic mechanisms Nobua Behrmann, Mallory Hee and Norah Saarman of brain structure development. The slave-maker/ assisted with field research. Funding was provided slave species differences in mushroom body calyx by Drexel University STAR undergraduate research investment were not explained by brain or body size fellowships (to E.S. and N.M.), and by NSF grant differences, and therefore were not a result of differ- 1209072 and Drexel University startup-funds (to ential brain region allometry (O’Donnell et al., 2013). S.O’D.).

CONCLUSION REFERENCES The evolutionary success of parasites derives in part Alloway TM. 1979. Raiding behaviour of two species of from their ability to co-opt host resources for parasite slave-making ants, Harpagoxenus americanus (Emery) and beneﬁt, potentially relaxing demands for parasite Leptothorax duloti cus Wesson (Hymenoptera: Formicidae). investment in some tissues. A common theme in para- Animal Behaviour 27: 202–210. site evolution is reduction of mass and complexity of Barton RA, Purvis A, Harvey PH. 1995. Evolutionary organs, and entire organ systems are lost in some radiation of visual and olfactory brain systems in primates, parasitic taxa (Chernin, 2000). Neural tissue is bats, and insectivores. Philosophical Transactions of the among the most metabolically and developmentally Royal Society B: Biological Sciences 348: 381–392. Boerner M, Kruger O. 2008. Why do parasitic cuckoos have costly, and the central nervous system is under strong small brains? Insights from evolutionary sequence analyses. investment constraints (Laughlin, 2001; Niven & Evolution 62: 3157–3169. Laughlin, 2008; Navarrete et al., 2011). Neural tissue Brady SG, Schultz TR, Fisher BL, Ward PS. 2006. Evalu- reduction in parasites is expected when parasites rely ating alternative hypotheses for the early evolution and on host cognition and sensation. Because social para- diversiﬁcation of ants. Proceedings of the National Academy sites can exploit a host’s sensory and behavioural of Sciences of the United States of America 48: 18172–18177. capabilities, reduced brain tissue investment can Buschinger A. 2009. Social parasitism among ants: a review accompany the evolution of social parasitism as well (Hymenoptera: Formicidae). Myrmecological News 12: 219– as organismal parasitism. The possibility of brain 235. reduction in social parasites has been explored in Catania KC. 2005. Evolution of sensory specializations in birds. The evolution of brood parasitism in birds insectivores. The Anatomical Record. Part A, Discoveries in was associated with reduced brain size in several Molecular, Cellular, and Evolutionary Biology 287: 1038– clades, but the adaptive reasons for this pattern are 1050.

Chernin J. 2000. Parasitology. New York: Taylor & Francis. Lihoreau M, Latty T, Chittka L. 2012. An exploration of Chittka L, Niven J. 2009. Are bigger brains better? Current the social brain hypothesis in insects. Frontiers in Physiol- Biology 19: 995–1008. ogy 3: 442. Corfield JR, Birkhead TR, Spottiswoode CN, Iwaniuk Moreau CS, Bell CD, Vila R, Archibald SB, Pierce NE. AN, Boogert NJ, Gutiérrez-Ibáñez C, Overington SE, 2006. Phylogeny of the ants: diversification in the age of Wylie DR, Lefebvre L. 2013. Brain size and morphology angiosperms. Science 312: 101–104. of the brood-parasitic and cerophagous Honeyguides Muscedere ML, Traniello JFA. 2012. Division of labor in (Aves: Piciformes). Brain Behavior and Evolution 81: 170– the hyperdiverse ant genus Pheidole is associated with 186. distinct subcaste- and age-related patterns of worker brain D’Ettorre P, Heinze J. 2001. Sociobiology of slave-making organization. PLoS ONE 7: e31618. ants. Acta Ethologica 3: 67–82. Navarrete A, van Schaik CP, Isler K. 2011. Energetics and Davis RL. 2005. Olfactory memory formation in Drosophila: the evolution of human brain size. Nature 480: 91–93. from molecular to systems neuroscience. Annual Review of Niven JE, Laughlin SB. 2008. Energy limitation as a selec- Neuroscience 28: 275–302. tive pressure on the evolution of sensory systems. Journal of Dunbar RIM, Shultz S. 2007. Evolution in the social brain. Experimental Biology 211: 1792–1804. Science 317: 1344–1347. O’Donnell S, Clifford MR, DeLeon S, Papa C, Zahedi N, Farris SM. 2008a. Evolutionary convergence of higher order Bulova SJ. 2013. Brain size and visual environment brain centers spanning the protostome deuterostome bound- predict species differences in paperwasp sensory processing ary. Brain Behavior and Evolution 72: 106–122. brain regions (Hymenoptera: Vespidae, Polistinae). Brain Farris SM. 2008b. Structural, functional and developmental Behavior and Evolution 82: 177–184. convergence of the insect mushroom bodies with higher O’Donnell S, Donlan NA, Jones TA. 2004. Mushroom body brain centers of vertebrates. Brain Behavior and Evolution structural plasticity is associated with temporal polyethism 72: 1–15. in eusocial wasp workers. Neuroscience Letters 356: 159– Farris SM, Robinson GE, Fahrbach SE. 2001. Experience- 162. and age-related outgrowth of intrinsic neurons in the mush- Seid MA, Harris KM, Traniello JFA. 2005. Age-related room bodies of the adult worker honeybee. Journal of Neu- changes in the number and organization of synapses in the roscience 21: 6395–6404. lip region of the mushroom bodies in the ant Pheidole Fujun X, Hu K, Zhu T, Racey P, Wang X, Sun Y. 2012. dentata. Journal of Comparative Neurology 488: 269–277. Behavioral evidence for cone-based ultraviolet vision in Shultz S, Dunbar RIM. 2010. Species differences in execu- divergent bat species and implications for its evolution. tive function correlate with hippocampus volume and Zoologia 29: 109–114. neocortex ratio across nonhuman primates. Journal of Com- Gronenberg W. 1999. Modality-specific segregation of input parative Psychology 124: 252–260. to ant mushroom bodies. Brain Behavior and Evolution 54: Smaers JB, Soligo C. 2013. Brain reorganization, not 85–95. relative brain size, primarily characterizes anthropoid Gronenberg W. 2001. Subdivisions of Hymenopteran mush- brain evolution. Proceedings of the Royal Society B 280: room body calyces by their afferent supply. Journal of 20130269. Comparative Neurology 436: 474–489. Strausfeld NJ, Hansen L, Li Y, Gomez RS, Ito K. 1998. Gronenberg W, Heeren S, Hölldobler B. 1996. Age- Evolution, discovery and interpretations of arthropod mush- dependent and task-related morphological changes in the room bodies. Learning and Memory 5: 11–37. brain and the mushroom bodies of the ant Camponotus Strausfeld NJ, Sinakevitch I, Brown S, Farris SM. 2009. floridanus. Journal of Experimental Biology 199: 2011– Ground plan of the insect mushroom body: functional and 2019. evolutionary implications. Journal of Comparative Neurol- Gronenberg W, Hölldobler B. 1999. Morphologic represen- ogy 513: 265–291. tation of visual and antennal information in the ant brain. Topoff H. 1990. Slave-making ants. American Scientist 78: Journal of Comparative Neurology 412: 229–240. 520–528. Hölldobler B, Wilson EO. 1990. The ants. Cambridge, MA: Topoff H, Cover S, Jacobs A. 1989. Behavioral adaptations Harvard University Press. for raiding in the slave-making ant, Polyergus breviceps. Jones TA, Donlan NA, O’Donnell S. 2009. Growth and Journal of Insect Behavior 2: 545–556. pruning of mushroom body Kenyon cell dendrites during Topoff H, Inez-Pagani M, Goldstein M, Mack L. 1985a. worker behavioral development in the paper wasp, Polybia Orientation behavior of the slave-making ant Polyergus aequatorialis (Hymenoptera: Vespidae). Neurobiology of breviceps in an oak-woodland habitat. Journal of the New Learning and Memory 92: 485–495. York Entomological Society 93: 1041–1046. Kilner RM, Langmore NE. 2011. Cuckoos versus hosts in Topoff H, Inez-Pagani M, Mack L, Goldstein M. 1985b. insects and birds: adaptations, counter-adaptations and out- Behavioral ecology of the slave-making ant Polyergus comes. Biological Reviews 86: 836–852. breviceps in a desert habitat. Southwest Naturalist 30: 289– Laughlin SB. 2001. Energy as a constraint on the coding and 295. processing of sensory information. Current Opinions in Neu- Visicchio R, Mori A, Grasso DA, Castracani C, Le Moli F. robiology 11: 475–480. 2001. Glandular sources of recruitment, trail, and

propaganda semiochemicals in the slave-making ant Wilson EO. 1971. The insect societies. Cambridge, MA: Polyergus rufescens. Ethology Ecology and Evolution 13: Harvard University Press. 361–372. Withers GS, Fahrbach SE, Robinson GE. 1995. Effects of West RJD. 2014. The evolution of large brain size in birds is experience and juvenile hormone on the organization of the related to social, not genetic, monogamy. Biological Journal mushroom bodies of honey bees. Journal of Neurobiology 26: of the Linnaean Society 111: 668–678. 130–144.