INTRODUCTION

The origin and the maintenance of a sterile caste has been one of the main problems of the theory of natural selection, dubbed Darwin's dilemma by West-Eberhard (1996). Excepting a few cases, workers are partly sterile, so they combine typical behaviors of the sterile caste (like food collection, brood care, defense, and nest construction) with some type of reproduction, originating the various patterns found in the Hymenoptera (Bourke, 1988). In fact, as previously predicted by kin selection theory, workers' sterility is conditionally expressed and reveals the variation of genetic interests of the colony (Queller and Strassmann, 1998).

Even though the population size of the mature colonies can be determined at least partially by ecological factors, changes in the number of individuals in a colony can have very important social consequences (Jeanne, 1991; Alexander et al., 1991; Bourke, 1999). The most important of these is the predicted change in the reproductive potential of the workers. As suggested by Bourke (1999), as the colony size increases, workers experience a decrease in their chances of becoming reproductive substitutes, so they increase mutual reproductive inhibition: worker policing (Ratnieks, 1988). Once the workers' reproductive potential decreases, the level of reproductive dimorphism between castes increases (Wilson, 1971; Michener, 1974; Oster and Wilson, 1978; Hölldobler and Wilson, 1990; Alexander et al., 1991; Wheeler, 1991). That sort of morphological skew (Bourke, 1999) would help explain why societies composed of a few individuals have small differences between castes, and those with many individuals present a more pronounced distinction. For these reasons, small societies would be characterized by a direct conflict between reproduction and caste determination. In contrast, conflicts in larger societies should be predominantly over brood composition, and the members of these societies should be relatively more “resigned” to the manipulation of their castes (Bourke, 1999). In this way, colony size deserves an ampler consideration as a determinant, like kin structure, social complexity, workers' reproductive potential, levels of caste differentiation, and the nature of social conflicts (Bourke, 1999).

The swarm-founding epiponine wasps represent an ideal subject for studying morphological skew because caste differentiation differs from null to complete dimorphism, and worker reproduction is widespread (Noll et al., 2021). In two cases, morphological skew theory applies to these wasps. Some species present small colony size, and slight or indistinct morphological differences and all individuals present consistent ovarian development⁴ [Parachartergus smithii (Mateus et al., 1997), Pseudopolybia vespiceps (Shima et al., 1998), Chartergellus communis (Mateus et al., 1999), Brachygastra augusti (Baio et al., 2004)]. Other species present larger colonies, castes quite distinct based on allometric differences and worker sterility [Agelaia. flavipennis (Evans and West-Eberhard 1970), A. areata (Jeanne and Fagen, 1974), A. vicina (Sakagami et al., 1996; Baio et al., 1998), A. pallipes and A. multipicta (Noll et al., 1997a), Protonectarina sylveirae (Shima et al., 1996a; Tanaka et al., 2010), Polybia scutellaris (Noll et al., 1997b), Epipona guerini (Hunt et al., 1996), Apoica flavissima (Shima et al., 1994) and A. pallens (Jeanne et al., 1995)]. However, at first glance, two other unusual patterns cannot fit into morphological skew theory. Some species have large colony sizes, allometric caste differences, and the presence of uninseminated egg layers [Protopolybia exigua and P. acutiscutis, Simões (1977), Naumann (1970)], while others have small colony sizes, low caste differentiation, and worker sterility [Metapolybia aztecoides (West-Eberhard, 1978)].

FIGURE 1.

Representative scheme of the seven measures for morphometric analyses. Head: HW, head width, IDm, minimum interorbital distance; Wing: WL, partial length of the forewing; Mesosoma: MSW, width of mesoscutum, AL, alitrunk length; Metasoma: T1L, length of gastral tergite I, and T2BW, basal widths of tergite II. Modified from Noll and Zucchi, 2004.

The genus Agelaia Lepeletier, 1836, is a conspicuous part of the social wasp fauna in much of tropical America (Jeanne, 1991), presenting 31 extant species and one fossil species recorded from Dominican amber (Andena et al., 2024; Carpenter and Grimaldi, 1997). Species vary in features such as nest architecture and number of individuals. There are species with small colonies, and species such as Agelaia vicina, with colonies harboring hundreds of thousands of individuals (Zucchi et al., 1995). There is a clear dimorphism between queens and workers (Cooper, 2000). In general, compared with workers, queens are larger, the dorsal pronotal carina, when developed, is blunter; the valvula shorter and with a narrower, hyaline border and tergum I wider (Richards, 1978; Noll et al., 1997a; Cooper, 2000). Agelaia species typically build their nests in cavities, subterranean or arboreal (Wenzel, 1998). Since the nests are hidden in the majority of species, the presence of an envelope is not the common pattern. Agelaia areata and A. flavipennis build an exposed nest of a single spiral comb with the cells on the inside so that the outermost part of the comb functions as an envelope (Jeanne, 1973; Cooper, 2000). A true envelope is found only in A. timida and A. baezae.

TABLE 1.

Morphometric differences. Means, t-test for difference between queens and workers of seven characters used for discriminating the castes of Agelaia timida. Head: HW, head width, IDm, minimum interorbital distance; Wing: WL, partial length of the forewing; Mesosoma: MSW, width of mesoscutum, AL, alitrunk length; Metasoma: TIL, length of gastral tergite I, and T2BW, basal widths of tergite II. * = All values statistically significant (P <0.01); N.S. = not statistically significant.

Here, we test the application of morphological skew theory (Bourke, 1999) in epiponines by reporting caste dimorphism in Agelaia timida Cooper, 2000, and comparing it with other Agelaia and other epiponine species. We also provide a description of a nest of A. timida.

MATERIAL AND METHODS

The analyzed colony of Agelaia timida were collected in Petit Saut, French Guiana (AMNH_ HYM 00000494). The colony was in a mature stage, characterized by the presence of different-aged brood (workers) and at least one adult generation (Noll and Zucchi, 2000, 2002), and all the adult wasps were fixed in alcohol. Forty females, from a total of 53, were randomly selected for measurements and dissections. Seven body parts (fig. 1) were measured under a binocular microscope with an ocular micrometer (smallest unit = 0.0875 mm): head width (HW), minimum interorbital distance (IDm), width of mesoscutum (MSW), alitrunk length (AL), length of gastral tergite I (T1 L), basal width of tergite II (T2BW), and partial length of the forewing (WL). Ovarian condition (number of ovarioles and development of oocytes) and insemination were determined by dissection under a stereomicroscope. The presence of sperm cells was confirmed by microscope.

Before statistical analysis, data were converted by log transformation in order to avoid problems of variance. Two groups, those with ovarian development and insemination (queens) and the remaining females (workers and intermediates), were determined for statistical purposes. Means and standard deviations were calculated from the seven morphological measurements. Bonferroni t-test was used for mean comparisons. The contribution of each variable to caste discrimination was examined using discriminant function analysis with the stepwise method (Rao, 1973).

In order to detect a correlation between caste differences and colony size, Mahalanobis distances (Anderson, 1958) from several epiponines from the literature were used. Mahalanobis distance between the group centroids is similar to the standard Euclidean distance measure, except that it accounts for the correlations between variables. The larger the differences, the farther are the respective groups apart from each other and the more discriminatory power our current model possesses for discriminating between the respective two groups. Statistical analyzes were performed using Statistica software (v. 12.5). A detailed description of a nest of Agelaia timida is also given. The nest is deposited at American Museum of Natural History (Nest 091203-1). It was collected in 2009 in Petit-Saut, French Guiana, by A. Dejean.

FIGURE 2.

Dispersion diagram showing the differentiation among females with the three types of ovarian development recognized (predicted groups according Mahalanobis distance values).

RESULTS

Ovary Development and Spermathecal Contents

The adult population comprised 53 females, of which 40 were examined. The ovariole number was always three in each ovary, and three types of ovarian development were documented: type A (n = 20) with filamentous ovarioles, which had no visible oocytes, or with some very small oocytes (workers); type B (n = 13) bearing some young oocytes or with one or more mature oocytes in each ovariole (intermediates); type C (n = 7) with well-developed and very long ovarioles with at least one mature egg, which was contorted inside the metasoma. Insemination was confirmed only in females with type C ovaries, i.e., queens.

TABLE 2.

Discriminant analysis among castes. Classification results for group comparisons through discriminant analysis in Agelaia timida (predicted groups according Mahalanobis distance values).

Nest Architecture

Agelaia timida is one of the few Agelaia species that build a true envelope surrounding their nests (also A. baezae), which sets them apart from most species within the genus (Cooper, 2000).

The nest was built on the surface of a tropical plant leaf, and another leaf was used as part of an envelope (fig. 3), which has not been previously reported for the genus (Wenzel, 1998). The vegetable fibers seem to come from the same type of plant, which differs from other Agelaia, which may use several types of plants (Wenzel, 1998). Compared to the envelope of A. baezae, the envelope of A. timida seems to be more fragile.

FIGURE 3.

Nest of Agelaia timida built on the surface of a leaf, and another leaf was used as part of an envelope. A, Frontal and B, back views (considering the surface where the brood comb is initiated).

Another difference is related to the entrance to the nest. In the nests of A. baezae, the entrance is always positioned at the end of the envelope facing downward (toward the ground). In A. timida, on the other hand, the entrance is positioned almost in the central area of the envelope. However, three nests of A. timida were previously described by Cooper and, in two of them, the entrance was positioned at the distal end (Cooper, 1986, 2000).

It is interesting to note that a nest with similar characteristics was described and illustrated by Wenzel (1998) but attributed to a species of Marimbonda (today synonymized with Leipomeles).

DISCUSSION

In neotropical swarm-founding wasps, caste differences can be arranged along a spectrum ranging from taxa in which queens and workers are externally similar, lacking morphological differences, to others with fairly distinct caste attributes (Richards, 1978; Jeanne, 1980; da Silva et al., 2021).

According to several authors (reviewed in Noll et al., 2004) caste differentiation in the Epiponini is most developed in Agelaia. Most of the species of this genus present the clear-cut case, in which morphological differences between castes are constant, and queens are always distinct from workers throughout the colony cycle (Noll et al., 2020). The results for A. timida differ strikingly from other previously studied species. The clear-cut pattern is also found in other Epiponini like Apoica and Polybia dimidiata (Noll et al., 2004). Still, it was never found in species with small colonies (a few dozen individuals) such as Agelaia timida (Noll et al., 2020). Thus, A. timida fits into morphological skew theory, since the species form small colonies with low caste differentiation and nonsterility of workers.

Agelaia timida presents intermediate females, which is exceptional for the genus as it was previously found only in A. lobipleura (Richards, 1978), a species that belongs to the same clade as A. timida in the phylogeny of the genus (Andena et al., 2024). The occurrence of laying workers in epiponines (intermediates) suggests reproduction may not be entirely the charge of queens. Even though these non-inseminated layers have mainly been found in species with low caste dimorphism, they have also been found in species with caste differences (Noll et al., 2004).

FIGURE 4.

Mahalanobis distance values versus colony size among species of Epiponini using discriminant function analysis. Abbreviations: Ac, Agelaia cajennensis; Ap, Angiopolybia pallens; Ba, Brachygastra augusti; Cc, Chartergellus communis; Cs, Clypearia sulcata; Ld, Leipomeles dorsata; Ma, Metapolybia docilis; Nc, Nectarinella championi; Pv, Pseudopolybia vespiceps; Ss, Synoeca surinama.

The low value of Mahalanobis distances obtained for A. timida (3.47), which indicates low differentiation between castes, strongly contrasts with the high values obtained for other Agelaia species, such as A. vicina (207.15 and 176.7; Baio et al., 1998), A. pallipes, and A. multipicta (124.67 and 110.99, respectively; Noll et al., 1997a) (fig. 4, table 3). A low value of Mahalanobis distances was also found for other species, such as: Pseudopolybia vespiceps (0.99; Shima et al., 1998), Polybia dimidiata (6.76; Shima et al., 1996b), Protopolybia exigua (4.95, Noll et al., 1996) and Apoica flavissima (11.27; Shima et al., 1994).

Richards (1978) in his book The Social Wasps of the Americas states that the only Stelopolybia (= Agelaia) known to make nests with envelopes are A. areata (Say) and A. flavipennis (Ducke), however, the envelope in these cases are the backs of the cells playing the role of protecting other parts of the comb. Cooper (1986) found and described two nests of Agelaia covered with a real envelope, both attached to the underside of leaves. On that occasion he identified the nests as belonging to A. cajennensis (F.). Later, Cooper (2000) corrected the identification, saying that in fact those nests were A. timida and not A. cajennensis.

The combination of less markedly distinct castes, plus the presence of a real envelope in Agelaia timida is an interesting aspect for the evolution of Agelaia. Considering that A. timida is part of the sister clade of all other species of Agelaia (Andena et al., 2024), it can be suggested that the characteristics observed in A. timida may be plesiomorphic, as we can partially observe in the ground plan of Epiponini, similar to what is found in Angiopolybia (Noll et al., 2021). In this scenario, the other clade of Agelaia species might have an ancestor that evolved into a very rigid caste system, with clear physiological distinction, large populations, and loss of envelope.

TABLE 3.

Mahalanobis distance values for Epiponini species. Colony size and obtained values of Mahalanobis distance for Agelaia timida and 30 other species of Epiponini (data from literature).