SEARCH

SEARCH BY CITATION

Keywords:

  • cuckoo;
  • host-parasite coevolution;
  • host races;
  • nestmate recognition;
  • Polistes;
  • Polistes atrimandibularis;
  • social parasite

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Parasites that exploit the parental behaviour of several host species may be selected to form distinct host-specific genetic lineages. This process is well documented in bird brood parasites, but not in insect social parasites. Polistes atrimandibularis is the only paper-wasp social parasite known to exploit four host species. It does not form genetically distinct host races according to analyses based on microsatellite loci. Also, there were no size-matching between parasites and host species. Instead, P. atrimandibularis queens seemed to be successful as parasites in this population only when they originated from nests of P. dominulus, the largest species. The other host species are a sink for P. atrimandibularis since adult females emerging from those nests appear too small to usurp colonies themselves. Traits that may help P. atrimandibularis infiltrate multiple species may include its nonaggressive usurpation tactics and its ability to acquire host cuticular hydrocarbon recognition labels.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Parasites that exploit the parental behaviour of their hosts may be engaged in an arms race, so they may be selected to specialize on one or few closely related host species (Davies et al., 1989). A parasite exploiting more than one host species may thus form separate genetic lineages or host races. This process is well documented in bird brood parasites like Cuculus canorus (the common cuckoo) which lays its eggs in other species’ nests, and evolved distinct genetic lineages that mimic the egg pattern of its many different hosts (Brooke & Davies, 1988; Gibbs et al., 2000). However, there is little evidence that the same process occurs in insect social parasites that usurp the colonies of other species and then monopolize most or all of the reproduction (Davies et al., 1989; but see Packer et al., 1995).

Among the more than 200 species of Polistes waSPS currently described (Carpenter, 1997), three are known to be obligate social parasites: Polistes sulcifer (Zimmermann), P. semenowi (Morawitz) and P. atrimandibularis (Zimmermann). All three live in sympatry around the Mediterranean and Caspian Basin, although P. atrimandibularis is more common at high elevations, and the other two species are typically found at low elevations (Cervo & Dani, 1996). How social parasitism evolved in this group has been a matter of discussion (Cervo & Dani, 1996; West-Eberhard, 1996) but it is clear that they are monophyletic and that no parasite is their host's closest relative (Choudhary et al., 1994; Carpenter, 1997). Previous work suggested that each social parasite exploited only one species, or a pair of sister species (Choudhary et al., 1994; see shaded areas in Fig. 1). P. sulcifer and P. semenowi parasitize species in the sister clade to the social parasites: P. dominulus (Christ), and both P. dominulus and P. nimphus (Christ), respectively. P. atrimandibularis, on the other hand, was known to parasitize two more distantly related species P. gallicus (L.), and P. biglumis (L.).

image

Figure 1. Phylogenetic relationships of the three Polistes social parasites and their hosts. Bold squares: hosts described before 2001; dotted squares: recently described hosts of P. atrimandibularis (modified from Carpenter, 1997).

Download figure to PowerPoint

The adaptations requiring host specificity in Polistes may include physical as well as chemical characteristics, because Polistes social parasites control their host colonies (Cervo & Lorenzi, 1996) by aggressive threats (Cervo et al., 1990a), by altering their species-specific mixture of cuticular hydrocarbons to match the odour of the host colony after usurpation (Bagnères et al., 1996; Turillazzi et al., 2000; Fanelli, 2001; Sledge et al., 2001; Lorenzi & Bagnères, 2002; Lorenzi et al., 2004), and possibly also by employing pacification pheromones (Cervo et al., 1990a).

Recently, it was discovered that a lowland population of P. atrimandibularis parasitizes three additional host species that it was not previously known to utilize: P. dominulus, P. nimphus and P. associus (Kohl) (Fanelli et al., 2001; dashed circles in Fig. 1). P. atrimandibularis is so successful in fooling all of these host species that in a few cases the hosts tended and reared the parasite's larvae to maturity weeks after the usurping P. atrimandibularis disappeared from the colony (Fanelli, 2001). This finding challenges the hypothesis that the different Polistes social parasites are specialists on a single clade of hosts. Perhaps P. atrimandibularis manages to parasitize different hosts by forming specialized host races.

Using microsatellite loci developed for P. dominulus (Henshaw, 2000), we looked for evidence of population subdivision along three host lines, as well as for inbreeding as a result of assortative mating within host lines. We also measured the sizes of successful parasites in the two most common lowland host species, and we measured the daughter parasites that were produced in three host species to determine whether there was selection for morphological specialization. We detected no significant host-related population structure, and no size-matching between parasites and host species, but we found that successful parasites on nests were bigger than the parasite brood that emerged from the hosts P. gallicus and P. associus, but not P. dominulus.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Collections and size measurements

As in the other two Polistes social parasites, P. atrimandibularis usurps its host colonies in late spring, prior to the emergence of the first workers. The parasite usurper soon starts to lay its own eggs in the colony, and will usually stay in the host colony until the end of its life cycle. The newly emerged parasites usually remain on the host colony until late summer, and then migrate to mountain peaks where they mate and overwinter (Cervo & Dani, 1996). We collected a total of 58 P. atrimandibularis females in a plain in northwestern Tuscany (Fanelli et al., 2001), including 47 newly emerged females and 11 usurpers. Usurpers and newly emerged parasites can be distinguished by differences in their behaviour as well as by the worn wing tips of usurpers. The sample of newly emerged females was collected between 10 and 20 August 1999, and consisted of 23 females that emerged from eight nests of P. gallicus, 9 females that emerged from four nests of P. dominulus and 15 females that emerged from three nests of P. associus. The sample of usurpers was collected on 1 August 2001 and consisted of eight usurpers on P. gallicus colonies and three on P. dominulus colonies. These usurpers were all collected after their brood had started to emerge, indicating they were all successful social parasites. These sample sizes are large considering that Polistes social parasites are rare (Cervo & Dani, 1996).

All individuals were stored at −20 °C. Measurements were taken under a binocular microscope. The maximum head width and the length of the medial cell of the wing were quantified with an optical ruler, which allowed a maximum error of ca 1%. A unique function of size measure was extracted by Principal Component Analysis (93.6% of total variance explained). An anova with a nested design was then employed to test the effect of host species and of nest of origin within host species. The post hoc LSD tests were Bonferroni-corrected for multiple comparisons.

Genetic analyses

We genotyped 32 newly emerged female parasites: 15 parasites from four nests of P. gallicus, nine from four nests of P. dominulus, and eight from three nests of P. associus. All individuals were genotyped at six polymorphic microsatellite loci developed for the nonparasitic species Polistes dominulus: Pdom2, Pdom7, Pdom121, Pdom122, Pdom127b and Pdom140 (Henshaw, 2000). With the exception of Pdom2, which had only three alleles, all the loci had seven to nine alleles. We extracted, amplified and scored the DNA following standard laboratory techniques (Strassmann et al., 1996).

To better understand the history of usurpation on each nest and the mating system of P. atrimandibularis in this lowland population, we assessed the kin structure of the newly emerged parasites in each nest. We estimated relatedness among the newly emerged parasites using the computer programme Relatedness 5.06 (Goodnight & Queller, 1999) and assessed the number of matrilines necessary to account for their genotypes based on Mendelian principles of inheritance. For relatedness estimations, we calculated the background allele frequencies from all of the genotyped individuals because there was no evidence for population differentiation (see below). Colonies were weighted equally for allele frequency calculations. Standard errors and 95% confidence intervals were calculated by jackknifing over loci for estimates within colonies and over colonies for population estimates.

We evaluated population structure, including differentiation between host races, and inbreeding, using the hierarchical F-statistics calculated in the computer programme Relatedness 4.2c, as well as an Analysis of Molecular Variance (AMOVA) as implemented in the computer programme Arlequin (Schneider et al., 2000). For the AMOVA analysis, we selected one individual from each nest because genotypes within nests are not independent. Relatedness 4.2c accounts for this nonindependence in its calculations, so all of the samples were used for analyses with this programme.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Host-related size differences

Both nest of origin and host species affected the size of the newly emerged parasite adults (Nested anova: n = 47, host species: F2 = 16.220, P < 0.001, nest of origin (nested in host) F12 = 6.076, P < 0.001) (Fig. 2). The significance of the nest of origin suggests that sizes must be in part a phenotypic consequence of environmental parameters linked to nest location (average temperature, humidity, food availability, and other environmental variables). However, the effects of nest position and host species are not independent because the different host species have different nesting-site preferences (Pardi, 1941; Fanelli, Pers. Obs.). The female parasites raised by P. dominulus were bigger than the ones raised by P. associus and P. gallicus (anovapost hoc LSD test Bonferroni-corrected: P < 0.001 in both cases), while these latter two were undistinguishable (P = n.s.).

image

Figure 2. Sizes of P. atrimandibularis that successfully usurped nests of P. gallicus and P. dominulus, and sizes of female brood emerged from three different host species. For display purposes, the overlapping points have been separated by adding or subtracting 0.05 mm from the medial cell length.

Download figure to PowerPoint

The adult females that usurped nests of P. dominulus were not different from those that usurped nests of P. gallicus (independent-samples t-test: t9 = −0.7, P = n.s.). However, overall, usurping parasites were bigger than the newly emerged daughters (t56 = 4.355, P < 0.001). In particular, they were bigger than the parasite brood raised by P. gallicus and P. associus (anovapost hoc LSD test Bonferroni-corrected P < 0.001 in both cases) but undistinguishable from the parasite brood raised by P. dominulus (anovapost hoc test Bonferroni-corrected P = n.s.) (Fig. 2). In summary, there was no matching of sizes between P. atrimandibularis and its different host species. However, only the bigger parasites were successful usurpers in any species, and they were significantly bigger than the brood raised on P. gallicus and P. associus, but equal in size to the brood raised on P. dominulus. Based on the size-function extracted by PCA, only one individual raised in a P. associus colony and two individuals raised in a P. gallicus colony fell in the size class of successful usurpers.

Among the brood that emerged from all hosts, the mean head width was 3.24 mm ± 0.02SE and the mean wing length was 6.33 mm ± 0.05SE. The selection differential (Lande & Arnold, 1983; Brodie & Janzen, 1996) for being a successful lowland usurper was s = 0.527 for head width (t-test not assuming equal variances t = 7.357, d.f. = 36.173, P < 0.001) and s = 0.531 for wing length (t = 4.64, d.f. = 21.478, P < 0.001).

Population structure and relatedness

We found limited, but inconclusive evidence for host specialization. Inbreeding (FIS) was −0.0723 and did not differ from 0, while FST was 0.0567 and was marginally significantly different from 0 when the standard error was calculated by jackknifing over loci (Table 1). However, FST was not significantly different from 0 when the standard error was calculated by jackknifing over nests (Table 1), nor was it significantly different from 0 in the AMOVA analysis (Table 2). Thus, there was no evidence to support assortative mating nor does it seem likely that there was genetic differentiation among host lineages, particularly given that the size data indicate that parasites emerging from species other than P. dominulus are not successful.

Table 1.  Inbreeding coefficients of the brood of P. atrimandibularis raised by three different host species.
CoefficientValueSE95% CI (loci)95% CI (nests)
  1. Confidenceintervals were calculated by jackknifing over loci and over nests (total n = 32, nine colonies).

Individuals (FIS)−0.07230.06970.1790.222
Host species (FST)0.05670.02160.05580.065
Population (FIT)−0.01150.07190.18480.096
Table 2.  The AMOVA for parasite brood emerged from different host species.
Variance componentVarianceTotal (%)φ-statisticsP
  1. Results include the partitioning of the variance, estimates of standard F-statistics and significance estimates (total n = 32, nine colonies).

Individuals (FIS)−0.11837−5.68−0.06001n.s.
Host species (FST)0.109645.270.052650.055
Population (FIT)2.09091100.42−0.00420n.s.

Newly emerged daughter parasites in the same colony were related by 0.687 ± 0.125 CI on average, which is not statistically different from 0.75 expected for haplodiploid full sisters. Likewise, all of the daughter genotypes in each colony were consistent with a single matriline. This indicates that the P. atrimandibularis usurpers were singly mated, and that no usurpation from a second parasite took place, or that if it did, all brood of the previous usurper were removed.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Polistes atrimandibularis has recently been found to parasitize three lowland host species not previously recorded. This study finds no consistent evidence for the development of host-related genetic differentiation in P. atrimandibularis. There was no evidence for morphological differences between successful usurpers on different hosts. Nor could we find evidence for genetic differences or assortative mating by parasites reared by different hosts (FST was small and not significant by two of three tests). Despite the lack of specialization, these parasites succeeded in reproducing in their host colonies.

Polistes social parasites may be better at fooling several species into taking care of their young than are cuckoos because of the nature of the recognition signal. Birds that are hosts to cuckoos recognize their eggs, which have a specific and fixed pattern. A parasitizing cuckoo must match that pattern, so there is strong selection for host race formation. Social waSPS recognize their nestmates by their cuticular hydrocarbon profiles and these can be acquired and manipulated after the parasite joins the colony (Bagnères et al., 1996; Turillazzi et al., 2000; Fanelli, 2001; Sledge et al., 2001; Lorenzi, 2003; Lorenzi et al., 2004). This means that usurpation in waSPS is inherently more plastic than usurpation in cuckoos. It should be easier for waSPS to parasitize multiple species and to do so without the formation of host races.

Successful usurpers of all studied lowland host species were bigger than nearly all of the newly emerged brood, indicating selection for increasing size. Only 3 of 28 females of P. atrimandibularis emerging from nests of P. gallicus or P. associus fell in the size category of successful usurpers (Fig. 2). P. dominulus is the largest host, and has larger colonies (Fanelli, 2001; Cervo, unpublished data). The brood produced in this host species matched the size of successful lowland usurpers, so selection is acting on this lowland population of parasites to increase its average sizes and to prefer the host P. dominulus.

The most likely explanation for the disappearance of smaller individuals from the population is that they are less successful host usurpers. However, on a few nests in the population the brood of P. atrimandibularis was in rare cases raised to maturity weeks after the usurping parasite had disappeared (Fanelli, 2001). We have no information on the sizes of these latter early-disappearing usurpers, so we cannot exclude that these were the smaller parasites, which would thus have enjoyed a higher fitness than our data suggests. Moreover, we cannot exclude that the smaller parasites do not die but have dispersal abilities correlated to their size (see Vogt et al., 2000), so that after the winter diapause they remain at higher altitudes and invade nests close to their over-wintering sites. However, P. dominulus is a typical lowland host, so we would still expect the larger individuals invading P. dominulus to increase in frequency in the lowland population.

Our selection hypothesis could be contradicted if the parasite females that usurp the suboptimal hosts P. gallicus and P. associus in the lowlands produced males rather than females, and if small males were successful in attracting mates. However, there is no evidence that parasites produced more males on P. gallicus and P. associus than on P. dominulus (Fanelli, 2001). Moreover, the males of all social parasites including P. atrimandibularis occupy mating territories and aggressively defend them against male competitors (Beani, 1996). Hence, larger sizes are presumably associated with higher fitness in males, and a parasite female that produced smaller males on the suboptimal host species would still incur a fitness cost.

Polistes atrimandibularis did not seem to preferentially use P. dominulus as a host in the lowlands, even though it was the only host where P. atrimandibularis was likely to rear a successful next generation. Indeed, the opposite seemed to be the case because P. dominulus was the most abundant host in the area, but was successfully invaded at lower rates than other species (Fanelli, 2001; Fanelli et al., 2001). Our data are so sparse that we cannot determine whether or not P. atrimandibularis was simply less successful at invading P. dominulus. There may be a maladaptive lack of preference for the most profitable host species because P. atrimandibularis has only recently invaded this lowland environment and is not at equilibrium. If the recent invasion hypothesis is true, it underscores a main conclusion of this study, that successful host exploitation by Polistes social parasites requires less specialization than had been thought.

However, the ability to parasitize P. dominulus may reflect ancestral abilities. We can now say that all three social parasites parasitize their sister clade (P. dominulus and P. nimphus), which supports the view that social parasitism not only arose once in these waSPS, but that the original social parasite used its closest relative as host, in accord with Emery's Rule (Lowe et al., 2002). As both social parasites and clades evolved and speciated, P. semenowi and P. sulcifer remained on the original host clade. We now know that P. atrimandibularis has also remained partially on this clade, or at least retained the ability to recolonize it.

The most profound shift in host use was in the ability of P. atrimandibularis to parasitize species outside the original host clade: P. biglumis, P. gallicus, and P. associus. As noted above, the ability to change hydrocarbon profiles may have facilitated such a shift, but this does not explain why the other two social parasites have not similarly expanded their host range. One possibility is that their larger sizes make them less able than P. atrimandibularis to utilize smaller hosts. Another possibility is the peculiar behaviour, which characterises P. atrimandibularis. This parasite, differently from the other two, not only uses a nonaggressive usurpation tactic (Cervo et al., 1990a) but also contributes to the rearing of immature brood by plundering larvae and pupae from other nests in the surroundings (Cervo et al., 1990b). Whatever the explanation, it is clear that P. atrimandibularis has the ability to defeat host defenses and usurp an unusually broad set of hosts, and it appears to do so with a generalized strategy rather than through a set of specialized strategies adapted to each host.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We whish to thank Stuart A. West and an anonymous referee for useful comments that helped us to improve the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • Bagnères, A.G., Lorenzi, M.C., Dusticier, G., Turillazzi, S. & Clément, J.L. 1996. Chemical usurpation of a nest by paper wasp parasites. Science 272: 889892.
  • Beani, L. 1996. Lek-like courtship in paper waSPS: ‘‘a prolonged, delicate, and troublesome affair’’. In: Natural History and Evolution of Paper-WaSPS (S.Turillazzi & M. J.West-Eberhard, eds), pp. 113125. Oxford University Press, Oxford.
  • Brodie, E.D. & Janzen, F.J. 1996. On the assignment of fitness values in statistical analyses of selection. Evolution 50: 437442.
  • Brooke, M. de L. & Davies, N.B. 1988. Egg mimicry by cuckoos Cuculus canorus in relation to discrimination by hosts. Nature 335: 630632.
  • Carpenter, J.M. 1997. Phylogenetic relationships among European Polistes and the evolution of social parasitism (Hymenoptera: Vespidae; Polistinae). In: The Origin of Biodiversity in Insects: Phylogenetic Tests of Evolutionary scenarios (P.Grandcolas, ed.), Mém. Mus. Natn. Hist. Nat. 173: 135161.
  • Cervo, R., Lorenzi, M.C. & Turillazzi, S. 1990a. On the strategies of host nest invasion in three species of Sulcopolistes, social parasites of Polistes waSPS. Actes Coll. Insect. Soc. 6: 6974.
  • Cervo, R., Lorenzi, M.C. & Turillazzi, S. 1990b. Sulcopolistes atrimandibularis social parasite and predator of an alpine Polistes (Hymenoptera, Vespidae). Ethology 86: 7178.
  • Cervo, R. & Dani, F.R. 1996. Social parasitism and its evolution in Polistes. In: Natural History and Evolution of Paper-WaSPS (S.Turillazzi & M. J.West-Eberhard, eds), pp. 104112. Oxford University Press, Oxford.
  • Cervo, R. & Lorenzi, M.C. 1996. Inhibition of host queen reproductive capacity by the obligate social parasite Polistes atrimandibularis (Hymenoptera, Vespidae). Ethology 102: 10421047.
  • Choudhary, M., Strassmann, J.E., Queller, D.C., Turillazzi, S. & Cervo, R. 1994. Social parasites in polistine waSPS are monophyletic: implications for sympatric speciation. P. Roy. Soc. Lond. B Bio. 257: 3135.
  • Davies, N.B., Bourke, A.F.G. & Brooke, M. de L. 1989. Cuckoos and parasitic ants: interspecific brood parasitism as an evolutionary arms race. Trends Ecol. Evol. 4: 274278.
  • Fanelli, D. 2001. Un parassita sociale sui generis: il generalismo e l'integrazione chimica e comportamentale di Polistes atrimandibularis (Hymenoptera, Vespidae). Tesi di Laurea in Scienze Matematiche, Fisiche e Naturali, Università di Firenze.
  • Fanelli, D., Cervo, R. & Turillazzi, S. 2001. Three new host species of the social wasp parasite, Polistes atrimandibularis (Hymeoptera, Vespidae). Insect. Soc. 48: 352354.
  • Gibbs, H.L., Sorenson, M.D., Marchetti, K., Brooks, M.de L., Davies, N.B. & Nakamura, H. 2000. Genetic evidence for female host-specific races of the common cuckoo. Nature 407: 183186.
  • Goodnight, K.F. & Queller, D.C. 1999. Computer software for performing likelihood tests of pedigree relationship using genetic markers. Mol. Ecol. 8: 12311234.
  • Henshaw, M.T. 2000. Microsatellite loci for the social wasp Polistes dominulus and their application in other social waSPS. Mol. Ecol. 9: 21552157.
  • Lande, R. & Arnold, S.J. 1983. The measurement of selection on correlated characters. Evolution 37: 12101226.
  • Lorenzi, M.C. & Bagnères, A.G. 2002. Concealing identity and mimicking hosts: a dual chemical strategy for a single social parasite? (Polistes atrimandibularis: Hymenoptera: Vespidae). Parasitology 125: 507512.
  • Lorenzi, M.C. 2003. Social wasp parasites affect the nestmate recognition abilities of their hosts (Polistes atrimandibularis and P. biglumis, Hymenoptera,Vespidae). Insect. Soc. 50: 8287.
  • Lorenzi, M.C., Cervo, R., Zacchi, F., Turillazzi, S. & Bagneres, A.G. 2004. Dynamics of chemical mimicry in the social parasite wasp Polistes semenowi (Hymenoptera: Vespidae). Parasitology 129: 643651.
  • Lowe, R.M., Ward, S.A., Crozier, R.H. 2002. The evolution of parasites from their hosts: intra- and interspecic parasitism and Emery's rule. P. Roy. Soc. Lond. B Bio. 269: 13011305.
  • Packer, L., Dzinas, A., Strickler, K. & Scott, V. 1995. Genetic differentiation between two host ‘‘races’’ and two species of cleptoparasitic bees and between their two hosts. Biochem. Genet. 33: 97109.
  • Pardi, L. 1941. Ricerche sui Polistini 2. Nota sui Polistini della campagna pisana. Atti della società Toscana di Scienze Naturali. 50: 37.
  • Schneider, S., Roessle, D., Excoffier, L. 2000. ARLEQUIN, Version 2.000: a software for population genetic data analysis. Genetics and Biometry Laboratory, University of Geneva, Switzerland.
  • Sledge, M.F., Dani, F.R., Cervo, R., Dapporto, L. & Turillazzi, S. 2001. Recognition of social parasites as nestmates: adoption of colony-specific host cuticular odours by the paper wasp parasite Polistes sulcifer. P. Roy. Soc. Lond. B Bio. 268: 22532260.
  • Strassmann, J.E., Solis, C.R., Peters, J.M. & Queller, D.C. 1996.Strategies for finding and using highly polymor phic DNA microsatellite loci for studies of genetic relatedness and pedigree. In: Molecular Zoology: Advances, Strategies and Protocols (J.Ferraris & S.Palumbi, eds), pp. 528549. Wiley-Liss, New York.
  • Turillazzi S., Sledge M.F., Dani F.R., Cervo R., Massolo A. & Fondelli L. 2000. Social Hackers: integration in the host chemical recognition system by a paper wasp social parasite. Naturwissenschaften 87: 172176.
  • Vogt, J.T., Appel, A.G. & West, M.S. 2000. Flight energetics and dispersal capability of the fire ant, Solenopsis invicta Buren. J. Insect Physiol. 46: 697707.
  • West-Eberhard, M.J. 1996. Wasp societies as microcosmos for the study of development and evolution. In: Natural History and Evolution of Paper-WaSPS (S.Turillazzi & M. J.West-Eberhard, eds), pp. 290317. Oxford University Press, Oxford.