An adaptive view of caste differentiation in the neotropical wasp Polybia (Trichothorax) sericea Olivier (Hymenoptera: Vespidae)

ECOLOGY, BEHAVIOR AND BIONOMICS

An adaptive view of caste differentiation in the neotropical wasp Polybia (Trichothorax) sericea Olivier (Hymenoptera: Vespidae)

IC Desuó^I; CB Souza-Galheico^II; SN Shima^I; GMM Santos^II; J D Cruz^II; CC Bichara Filho^II; CTS Dias^III

^IDepto de Zoologia, Instituto de Biociências, UNESP, Rio Claro, SP, Brasil

^IIUniv Estadual de Feira de Santana, Feira de Santana, BA, Brasil

^IIIDepto de Ciências Exatas, Escola Superior de Agricultura "Luiz de Queiroz" - ESALQ, USP, Piracicaba, SP, Brasil

^{Correspondence}

ABSTRACT

The tribe Epiponini comprehends the swarm-founding Neotropical wasps, with several species endemic to Brazil, which are extremely important in studies of social evolution of wasps. The Epiponini diverge in several ways from the definitions of high eusociality, since caste syndromes range from species without morphological caste differentiation to those with complete caste dimorphism, and all species are polygynous. Frequently, indirect studies based on morphometry and physiology are the only solutions to collect data regarding the natural history and caste system in this tribe, since most species are extremely aggressive and build enveloped nests, usually in places of difficult access. We analyzed morphological parameters in seven colonies of the Epiponini species Polybia (Trichothorax) sericea Olivier in different phases of colonial development. Nine body variables were taken and females were classified according to their ovary development and spermathecal contents. The results showed that caste differences in this species are based on a contrast among variables: queens have larger mesosoma and abdomen, but are smaller in head width and wing length. These results suggest that morphological caste differentiation in this species is based mainly on body shape. We considered this combination of characters as being adaptive. We also showed that caste differences varied according to the colony cycle, with more conspicuous differences when queen number is reduced.

Keywords: Social insect, morphological difference, Epiponini

Introduction

One of the most important features of social insects is the division of labor among reproductive and non-reproductive females. Since they can be morphologically distinct, understanding caste differentiation is crucial to evaluate the evolutionary trajectories of sociality in these insects (Wilson 1971, Bourke 1999, Oliveira et al 2007). In fact, the level of sociality is a function of caste differentiation. Commonly, species with a higher degree of caste asymmetries tend to present a well defined reproductive division of labor, and consequently, a higher degree of sociality (Bourke 1999). Highly eusocial insects, such as honeybees, ants and termites, present conspicuous caste dimorphism and pre-imaginal caste determination.

In this aspect, the Neotropical swarm-founding wasps Epiponini (Carpenter 1997) diverge in several ways from the definition generally given for high eusociality, as in this group, caste syndromes range from species without any morphological caste differences to those with complete caste dimorphism (Richards 1978). Moreover, caste differences may vary according to the colony phase, and this variation may be related to the process of queen selection (Noll & Zucchi 2002). Caste determination is also variable and complex, and may be pre or post-imaginal depending on the species (Mateus et al 2004). All Epiponini wasps are polygynous, and queens do not play an important role in the control of social cohesion (Noll & Wenzel 2008). Contrarily to other highly eusocial insects, workers of Epiponini police the reproductive status of queens and eliminate those with lower reproductive potential (West-Eberhard 1973, Nascimento et al 2004). This variation in the number of queens may be important to the genetic composition of the colony: when polygyny levels are high, relatedness among queens is low and, consequently, overall relatedness among workers is also low. Such a scenario would act against social cohesion. An opposite situation occurs when levels of polygyny are low and queens are closely related (Hamilton 1972, Queller et al 1993, Strassmann et al 1998, Nascimento et al 2004). Additionally, a recent study showed through gene-based simulations, that cooperative behavior would be favored by multiple matings (Nonacs 2011). This enigmatic scenario involving the Epiponini draws sociobiologists' attention to the real social status of this group (for further discussion see Bourke 1999 and Noll & Wenzel 2008).

Caste differentiation in the Epiponini is a complex trait, since a wide range of different patterns is found (Richards 1978, Noll & Zucchi 2002, Noll et al 2004). According to Noll et al (2004), four different syndromes are now recognized in the Epiponini: (1) casteless - no morphological differences in size or shape and the presence of females with all ovary developmental conditions (Shima et al 1998, Mateus et al 2004, Noll et al 2004), (2) without a clear morphological distinction, but physiological differences with the presence of intermediate females (Noda et al 2003, Noll et al 2004, 2010), (3) queens larger than workers, but no shape differences (Noll et al 2004), and (4) caste differences based on shape, with queens smaller than workers in some variables and larger in others (Shima et al 1994, 1996a,b, 2003, Hunt et al 1996, Desuó et al 2011).

The Epiponini wasp Polybia (Trichothorax) sericea Olivier is widely distributed throughout South America, especially in Brazil. The nests are usually globose, reddish brown in color, never very large, and commonly built on branches of various trees. Colony size may vary from a few individuals to thousands (Richards 1978). This species belongs to the most diverse genus of Epiponini, which presents several different syndromes of caste differentiation. There are no studies regarding caste differentiation in this species. Therefore, the aims of this study were (1) to investigate the pattern of caste differentiation of Po. (T.) sericea and (2) to explain this pattern within the framework of an adaptive scenario.

Material and Methods

In this study seven colonies of the Epiponini wasp Po. (T.) sericea were analyzed; they were collected in different phases of the colony cycle. Colonies were collected in the following regions: (I) Mirandópolis/SP/Brazil (20º36'58.8"S; 49º27'54"W), (II) Ribeirão Preto/SP/Brazil (21º12''42'S; 47º48'24''W), (III) Rio Claro/SP/Brazil (22º24'36''S; 47º33'36''W), (IV) Penápolis/SP/Brazil (21º24'59.1"S; 50º04'23"W), (V) Paulo Afonso/BA/Brazil (9º33'13"S;38º29'20"W), and (VI) and (VII) were collected in Senhor do Bonfim/BA/Brazil (10º02'14"S;40º00'43"W). All adult wasps of each colony were immediately killed and fixed in Dietrich's solution and preserved in ethanol 70%.

A sample of 25% of the adult population of each colony was randomly chosen for morphological studies and had their measurements taken under a binocular stereomicroscope (smallest unit = 0.083 mm). Eight body parts were measured to obtain the morphological data (Fig 1): HW (head width), IDM (minimum inter-orbital distance), MSW (mesoscutum width), AL (alitrunk length), T1L (length of tergite 1), T1AH (apical height of tergite 1), T2BW (basal width of tergite 2), and WL (wing length). Colonies I, II, III and IV also had the following variables assessed: T4W (width of tergite 4) and T4H (height of tergite 4). Specimens were dissected to check for insemination and ovary development. According to the ovary condition and insemination, females were regarded as workers or queens.

Statistical analyses were performed using SAS 9.2® statistical package and JMP 7.0 linked to the SAS database. The statistical procedures were as follows: (1) the original data were converted into log ranks to facilitate arranging them in a normal distribution; (2) PROC MEANS was used to calculate basic statistics; (3) PROC ANOVA was used to test univariate differences between the arithmetic means of workers and queens of each colony; (4) PROC DISCRIM was used to find a function which maximized the separation between the castes; a classification matrix of each colony was made considering that there was the same prior probability that one individual belonged to one of the two castes; (5) PROC CANDISC was used to perform a canonical analysis of the original data; setting NCAN = 2 provided two axes and produced a canonical scatterplot to visualize the separation of the castes in the different phases of the colony cycle; (6) PROC STEPDISC was used to perform a stepwise discriminant analysis in order to choose the most important variables that discriminate castes in Po. (T.) sericea. We used the option SLENTRY = 0.15, which specifies the significance level for adding variables. In this case, variables with a p-value greater than 0.15 were removed from the model. The procedures (4), (5) and (6) were employed using the original data, in millimeters.

Results

The composition of the colonies is detailed in Table 1. According to the number of immatures and adults, number of meconia, relative age structure and percentage of queens (see Table 1 for further details), colonies were classified I - Pre-emergence; II, III, IV and V - worker production; VI, VII - male production.

Three "types" of ovary development condition were observed in females of P. (T.) sericea : (1) Type A - characterized by filamentous ovarioles without any oocyte development; (2) Type B - ovarioles with slight oocyte development; and (3) Type C - represented by well developed ovarioles with many mature oocytes. Type B was gradually subdivided into B1, B2 and B3 (Fig 2). Queens were associated with type C ovaries and were the only individuals that presented sperm-filled spermathecae. Non-inseminated females with ovary patterns A or B were considered as workers.

The One-Way ANOVA results revealed that queens were overall larger than workers (Table 2). It is worth mentioning that the variables shown in Table 2 were chosen by stepwise discriminant procedure, and thus represent the most important variables to discriminate castes in P. (T.) sericea. From the analysis of the total-sample standardized canonical scores in Table 2, it is evident that CAN1 is defined by a contrast between head and wing size versus abdominal variables. Queens are characterized as larger individuals (especially in the abdominal variables), with smaller head width and wing length. In colony V, for example, the canonical score was 0.77 for T2BW and -0.57 for HW. This contrast means that for both variables chosen using a stepwise discriminant function, caste differences in this phase were represented mostly by queens with larger T2BW and smaller HW. This assumption holds for most of the colonies (Table 2), except for colony III, in which the canonical score was positive for HW.

Canonical scatterplots (Fig 3) indicated that castes can be separated into queens and workers; however, such separation was less evident in colony I and more conspicuous in colony III (see the overlapping area of Fig 3). Queens have higher values in CAN1 than workers (Fig 3), and linking this information with the canonical scores (Table 2), it is possible to define the morphological differences between queens and workers in each colony phase.

The values of Wilk´s lambda (Table 2) ranged from 0.4 to 0.8, which indicates, according to Noll et al (2004), that the power of discrimination of the variables ranged from moderate to low. Finally, the classification matrix obtained after discriminant function analysis (Table 3) showed that queens were only 100% correctly classified in colonies II and III, indicating that queens from these colonies were more distinct morphologically from workers, resulting in a higher degree of caste differentiation than in the other colonies.

Discussion

The genus Polybia (Lepeletier) is the most diversified group within the Epiponini. Caste differentiation in this taxon is complex and several syndromes are found (Noll et al 2004, Noll & Wenzel 2008). Our data showed that castes of Po. (T.) sericea differed morphologically, however such morphological differences, given by size, were less conspicuous than in other Epiponini genera (Agelaia and Apoica) and species from Polybia, Polybia lilliacea (F.) and Polybia (Trichothorax) ignobilis (Haliday). Nevertheless, queens of Po. (T.) sericea were larger than workers in some of the body variables, yet smaller in others, and similar patterns were also found in other lineages of the Epiponini (Noll et al 2004, Noll & Wenzel 2008), but queens were considerable larger than workers (Wilk´s Lambda ranged from 0.1 to 0.5) (Desuó et al 2011). Based on morphological data sets, Jeanne et al (1995) hypothesized that such pattern of caste differences is related to a differential growth of imaginal discs during the immature period, which results in differences in adult shape, rather than size. However, there are no studies regarding the development of imaginal discs in the Epiponini (Desuó et al 2009, 2011).

Queens with larger T2BW is the ancestral condition of the Epiponini, and also a widespread pattern in this tribe (Noll & Wenzel 2008). In social insects, larger abdomen may be considered a predictable trait for queens, as they are frequently physogastric. However, if queens have larger ovaries and are specialized in laying eggs, then they should present adaptations in their external morphology to support their ovary status. These adaptations, such as distended tergites, should appear primarily in the abdominal region allowing the abdominal cavity to carry a larger ovary. Considering that insects do not grow during adult life, these structural modifications may represent differences in developmental routes in the pre-imaginal stages of queens and workers. These different developmental paths could lead to higher growth of abdominal structures in queens, and then could represent the first step into the development of specialized castes in the Epiponini.

Queens with smaller head width and wing length is a common pattern of caste differences in the Epiponini (Noll & Wenzel 2008). Such variation from the ancestral condition ("egglayer equal to non-egglayer", see Noll & Wenzel 2008) may represent an adaptation to a higher caste specialization. Queens with smaller head width are reported in species with substantial caste differences (Noll et al 2004, Noll & Wenzel 2008). In general, queens spend more time in the colony performing reproductive tasks (egg-laying), whereas workers are mostly responsible for foraging activities (O´Donnell 1991). Foraging involves the ability of interacting with the environment and a better capacity to fly (Lima & Prezoto 2003, Gomes et al 2007). These requirements may change the physiological status of individuals and thus workers may have larger brains and wings. Jones et al (2009) showed that the volume of the mushroom bodies of workers of P. aequatorialis (Zavattari) increases as they progress from in-nest activities to forage. It is noteworthy that the canonical score of HW was positive in colony III, indicating that queens were also larger than workers in this body variable.

Our data also showed that the morphological caste differences of P. (T.) sericea varied according to the colony cycle, and that the differences between queens and workers were apparently larger when the number of queens was low (Table 1, Fig 3-III), a trend already shown by others (Noll & Zucchi 2000, 2002, Pizarro et al 2009, Desuó et al 2011). Therefore, variations in the number of queens and the observed caste differences appear to be related, leading to the conclusion that selection favors larger and more productive queens (queens with higher quantities of fat body and more developed ovaries).

The variance in the number of queens is also an important reproductive trait of the Epiponini, as it affects the genetic relatedness of nestmates within the colony. Usually colonies with a few or a single queen favor the production of new queens while colonies of multiple queens favor the production of males. This occurs because workers are three times more closely related to their sisters than to their brothers in single-queen colonies and equally related to their nieces and nephews in multiple-queen colonies (Strassmann et al 1998). In most social insects the production of daughter colonies are tied to the production of new queens. In the Epiponini, however, there is a disconnection between these two events which allows the Epipoini to form new colonies during the most favorable season, contributing to ecological success of this tribe in the neotropical region (Strassmann et al 1998).

The literature regarding caste in the Epiponini establishes two main strategies for caste determination: (1) pre-imaginal, in which queens form morphological clusters distinct from workers and (2) post-imaginal caste differentiation where castes are more flexible and non-reproductive resemble reproductive adults (Noll et al 2010). Our data clearly indicate that caste determination in P. (T.) sericea is pre-imaginal. The strategy adopted by the studied species, as well by other Epiponini, ensures that workers may have a lower chance to achieve a reproductive status than in species in which caste differences are only slight or absent. In fact, Noll et al (2010) proved that in species with slight caste differentiation, such as Leipomeles dorsata (Fabricius), workers may become reproductively active.

Finally, the pattern of caste differentiation showed in this study seems to operate in other lineages of Epiponini as well, such as Agelaia, Apoica and other Polybia (Pizzarro et al 2009). Some of these lineages are not close phylogenetically to P. (T.) sericea, which corroborates the view that caste evolution in the Epiponini is complex and that the different syndromes of caste differentiation found in this tribe probably evolved several times in the different lineages. (Noll et al 2004, Noll & Wenzel 2008).

Acknowledgments

To Mr. Jaime R. Somera for illustrations in Fig 2. We also thank CNPq for the financial support. CBSG received a scholarship from FAPESB (BOL0248/2009), GMMS received a productivity fellowship from CNPq (309711/2009-6).

Received 14 February 2011 and accepted 09 August 2011

Edited by Fernando B Noll - UNESP

• Bourke AFG (1999) Colony size, social complexity and reproductive conflict in social insects. J Evol Biol 12: 245-225.
• Carpenter JM (1997) A note on the names of paper wasps tribes (Insecta:Hymenoptera:Vespidae). Nat Hist Bull Ibaraki Univ 1: 5-16.
• Desuó IC, Murakami ASN, Oliveira VC, Shima SN (2009) Overview of caste differentiation in the Polistinae, emphasizing the Neotropical swarm-founding Polistinae (Hymenoptera, Espidae, Epiponini). Sociobiology 53: 851-889.
• Desuó IC, Shima SN, Oliveira ACL, Gomes B, Dias CTS (2011) Caste asymmetries in the Neotropical swarm-founding wasp Polybia (Trychothorax) ignobilis (Hymenoptera, Vespidae, Epiponini). Sociobiology 57: 51-69.
• Gomes L, Gomes G, Oliveira HG, Morlin Jr JJ, Desuó IC, Silva IM, Shima SN, Von Zuben CJ (2007) Foraging by Polybia (Trichothorax) ignobilis (Hymenoptera, Vespidae) on flies at animalcarcasses.  Revta Bras Entomol 51: 389-393.
• Hamilton WD (1972) Altruism and related phenomena, mainly in the social insects. Annu Rev Ecol Syst 3: 193-232.
• Hunt JH, Schmidt DK, Mulley SS, Williams MA (1996) Caste dimorphism in Epipona guerini (Hymenoptera, Vespidae): Further evidence for larval determination. J Kansas Entomol Soc 69: 362-369.
• Jeanne RL, Graf CA, Yandell BS (1995) Non-size-based morphological castes in a social insect. Naturwissenschaften 82: 296-298.
• Jones TA, Donlam, NA, O'Donnell S (2009) Growth and pruning of mushroom body Kenyon cell dendrites during worker behavioral development in the paper wasp, Polybia aequatorialis (Hymenoptera: Vespidae). Neurobiol Learn Mem 92: 485-495.
• Lima MAP, Prezoto F (2003) Foraging activity rhythm in the Neotropical swarm-founding wasp Polybia platycephala sylvestris  Richards, 1978 (Hymenoptera: Vespidae) in different seasons of the year. Sociobiology 42: 645-752.
• Mateus S, Noll FB, Zucchi R (2004) Caste flexibility and variation according to the colony cycle in the swarm-founding wasp, Parachartergus fraternus (Gribodo) (Hymenoptera: Vespidae, Polistinae). J Kansas Entomol Soc 77: 470-483.
• Nascimento FS, Tannure-Nascimento IC, Zucchi R (2004) Behavioral mediators of cyclical oligogyny in the Amazonian swarm-founding wasp Asteloeca ujhelyii (Vespidae, Polistinae, Epiponini). Insectes Soc 51: 17-23.
• Noda SCM, Shima, SN, Noll FB (2003) Morphological and physiological caste differences in Synoeca cyanea (Hymenoptera, Vespidae, Epiponini) according to the ontogenetic development of the colonies. Sociobiology 41: 547-570.
• Noll FB, Gomes B, Lima ACO, Mateus S, Wenzel JW (2010) Castes in the Neotropical social wasp Leipomeles dorsata (Fabricius) (Hymenoptera: Vespidae): a window for workers achieving a new status in the colony. Neotrop Entomol 39: 549-554.
• Noll FB, Wenzel JW (2008) Castes in the swarming wasps: "queenless" societies in highly social insects. Biol J Linn Soc 93: 509-522.
• Noll FB, Zucchi R (2000) Increasing caste differences related to life cycle progression in some neotropical swarm-fouding polygynic wasps (Hymenoptera, Vespidae, Epiponini). Ethol Ecol Evol 12: 43-65.
• Noll FB, Zucchi R (2002) Caste and influence of colony cycle in swarm-founding Polistinae wasps (Hymenoptera, Vespidae, Epiponini). Insectes Soc 48: 1-13.
• Noll FB, Wenzel JW, Zucchi R (2004) Evolution of castes in neotropical swarm-founding wasps (Hymenoptera, Vespidae, Epiponini). Am Mus Novit 3467: 1-12.
• Nonacs P (2011) Monogamy and high relatedness do not preferentially favor the evolution of cooperation. BMC Evol Biol 11: 58.
• O'Donnell S (1991) Worker age, ovary development, and temporal polyethism in the swarm-founding wasp Polybia occidentalis (Hymenoptera: Vespidae). J Insect Behav 14: 201-213
• Oliveira AL de, Noll FB, Mateus S, Gomes B (2007) Castes and asynchronous colony cycle in Polybia bistriata (Fabricius) (Hymenoptera: Vespidae). Neotrop Entomol 36: 817-827.
• Pizarro LC, Noll FB, Mateus S (2009) Caste differences in Pseudopolybia compressa (Hymenoptera: Vespidae: Polistinae). Rev Bras Zool 26: 624-628.
• Queller DC, Negrón-Sotomayor JA, Hughes CR, Strassmann JE (1993) Queen number and genetic relatedness in a Neotropical wasp, Polybia occidentalis Behav Ecol 4: 7-13.
• Richards OW (1978) The social wasps of the Americas excluding the Vespinae. London, British Museum (Natural History), 580p.
• Richards OW, Richards MJ (1951) Observations on the social wasps of South America (Hymenoptera, Vespidae). Phil Trans R Soc Lond B Biol Sci 102: 1-170.
• Shima SN, Noll FB, Zucchi R (2003) Influence of colony cycle on physiological and morphological caste variation in the perennial Neotropical swarm-founding social wasp, Protonectarina sylveirae (Hymenoptera:Vespidae:Epiponini). Sociobiology 42: 449-446.
• Shima SN, Noll FB, Zucchi R, Yamane S (1998) Morphological caste differences in some neotropical swarm-founding Polistinae wasps VI. Pseudopolybia vespiceps, with preliminary considerations on the role of the intermediate females in social organization of the Epiponini (Hymenoptera, Vespidae). J Hymenopt Res 7: 280-295.
• Shima SN, Yamane S, Zucchi R (1994) Morphological caste differences in some neotropical swarm-founding Polistinae wasps I. Apoica flavissima (Hymenoptera, Vespidae). Jpn J Entomol 64: 811-822.
• Shima SN, Yamane S, Zucchi R (1996a) Morphological caste differences in some neotropical swarm-founding Polistinae wasps II. Polybia dimidiata (Hymenoptera, Vespidae). Jpn J Entomol 64: 131-134.
• Shima SN, Yamane S, Zucchi R (1996b) Morphological caste differences in some neotropical swarm-founding Polistinae wasps III. Protonectarinae Sylveirae (Hymenoptera, Vespidae). Nat Hist Bull Ibaraki Univ 45: 57-67.
• Strassmann JE, Goodnight KF, Klinger CJ, Queller DC (1998) The genetic structure of swarms and the timing their production in the queens cycles of Neotropical wasps. Mol Ecol 7: 709-718.
• West-Eberhard M J (1973) Monogyny in polygynous social wasps. In Proceedings of the VII Congress of I.U.S.S.I. London, p.396-403.
• Wilson EO (1971) The insect societies. Cambridge, The Belknap Press of Harvard University, 548p.

Publication Dates

• Publication in this collection
  06 Feb 2012
• Date of issue
  Dec 2011

History

• Accepted
  09 Aug 2011
• Received
  14 Feb 2011