Fighting injuries, fig exit, and dimorphism in two species of sycoryctine fig wasp (Chalcidoidea, Pteromalidae)
expand article infoDa-Mien Wong, Anthony Bain§, Shiuh-Feng Shiao, Lien-Siang Chou
‡ National Taiwan University, Taipei, Taiwan
§ National Sun Yat-sen University, Kaohsiung, Taiwan
Open Access


Similar to many vertebrate and invertebrate species, many fig wasp species are fighting other members of their species for mates. Fighting between the males of many non-pollinating fig wasp species involves injuries and fatalities. Studies have shown that large males fight for mates, whereas conspecific small males tend to adopt nonfighting, sneaky behaviors. To analyze male morphs in two non-pollinating fig wasps (Philotrypesis taida Wong & Shiao, 2018 and Sycorycteridea taipeiensis Wong & Shiao, 2018) associated with the fig tree Ficus benguetensis, the head and mandible allometry and injuries were examined as well as the morphologies of their heads and mandibles. Male fig wasps of these two nonpollinating species can be divided into two morphological groups according to their head and mandible shapes. Approximately 88% of the Philotrypesis and 62% of the Sycorycteridea males were injured and no males belonging to largest morphs were decapitated. Moreover, nearly 31% of the Philotrypesis and 45% of the Sycorycteridea males left their natal figs. No difference in injury level or male exit rate between the male morphs was observed. This study reveals slight morphological and behavioral differences that may hint towards different mating strategies among morphs.


Ficus benguetensis, fatal fighting, male dimorphism, male dispersal, Philotrypesis, Sycorycteridea


Fighting is ubiquitous in the animal kingdom (Hardy and Briffa 2013; Huntingford 2013; Vieira and Peixoto 2013), and intraspecific fighting is mainly caused by the need to defend for food and territory (Backwell and Jennions 2004) or for mating opportunities (Baxter et al. 2015). For example, many Cervidae stags fight for the control of the females (Clutton-Brock et al. 1979; McElligott et al. 1998); such activity is common in mammals (Packer 2001; Casey et al. 2015). Fighting is also extremely common in invertebrate species such as cephalopods (Schnell et al. 2015) and weaponless butterflies (Kemp and Wiklund 2001). Hymenopterans, such as ants, wasps, and bees, are well-known to engage in intraspecific fighting, either to protect their colonies (Gloag et al. 2008; Rudolph and McEntee 2015) or to compete for the queen position (Berthelot et al. 2017). Fighting often causes many casualties, especially in reproductive castes of social insects (Kärcher et al. 2013; Heinze et al. 2016). Fatal fighting may occur when the following two conditions are met (Enquist and Leimar 1990): the value of the contested resource (e.g. a limited number of mating partners) is particularly high and the value of future reproduction is extremely low (e.g., the organism’s lifespan is short). Many examples of fatal fights between males have been documented in the hymenopteran superfamily Chalcidoidea: in the genus Melittobia (Hartley and Matthews 2003) and in numerous fig wasp species (Cook 2005). For example, the males of several fig wasp species can only live for up to one day, and the numbers of females within a mating arena in these species are often low (Cook et al. 2017). Thus, the non-pollinating fig wasp model fits the conditions required for the evolution of fatal fighting (Enquist and Leimar 1990).

Fig. wasps are small insects that are obligately associated with plant of genus Ficus (Kjellberg et al. 2005). As larvae, all fig wasp species (pollinators and non-pollinating fig wasps) are confined inside closed inflorescences called figs. Because of the following aspects of their peculiar life cycle, non-pollinating fig wasps provide a unique opportunity for studying non-pollinating male fig wasp fights: (1) Each fig is a spherical enclosed inflorescence containing hundreds of flowers with one single exit (the ostiole), and (2) limited variation in environmental factors, and the number of figs produced by each tree enable easy replication of experiments in a known environment. In each fig, one or a few female pollinators enter receptive fig, pollinate the flowers inside the fig and lay eggs into the fig ovaries. Wasp larvae then develop inside galled flowers over several weeks. Non-pollinating fig wasps mostly colonize figs after pollination and parasitize fig ovules or pollinating or non-pollinating fig wasp larvae; they do not enter the fig but rather use their long ovipositor to lay eggs through the fig wall into ovules (Kerdelhué and Rasplus 1996). Male pollinating fig wasps hatch first and mate with unhatched pollinating female fig wasps.

Fighting among male fig wasps occurs in both pollinating and non-pollinating species (Bean and Cook 2001; Nelson and Greeff 2009) and tend to be fatal (Bean and Cook 2001; Pereira and Prado 2008; Dunn et al. 2015). The fights take place in the fig: more precisely in the central cavity (the lumen) of the fig and during the short period before the fig opens (or is opened by male fig wasps) to let the female fig wasps disperse. A considerable proportion of fighting occurring between individuals results in severe injuries or death (Bean and Cook 2001). Fig. wasp males belonging to the same species can exhibit a range of morphologies, including armored males bearing large falcate mandibles that fight for mates and flattened males that stealthily seek mates (Murray 1990; Cook et al. 1997; Jousselin et al. 2004). Moreover, some species have dispersing morphs such as winged males (Jousselin et al. 2004) and apterous dispersing males (Greeff and Ferguson 1999); however very little is known about the dispersal rate and behavior of the male fig wasps. Theoretically, a male reaching a fig other than its native fig would be able to mate with female wasps that likely are genetically very different.

The aim of this study was to document morphological variations, the fighting outcomes, and the exit rates of two species of non-pollinating fig wasp living on Ficus benguetensis. Preliminary observations of Philotrypesis taida Wong & Shiao, 2018 and Sycorycteridea taipeiensis Wong & Shiao, 2018 males had revealed that: (1) males are fighting inside the figs, (2) fighting may result in death by decapitation, (3) males vary in size, and (4) males may exit figs. On the basis of these observations, this study raised the following questions concerning male interactions among the studied species: (1) How different in size and morphology are these males? (2) How common are injuries and fatal injuries? (3) How often do males exit their natal figs? (4) What are the factors (e.g., fig diameter linked with the fighting space, sex ratio linked with the number of fighters) affecting the level of injury and the exit rate?


Study species

Philotrypesis taida and Sycorycteridea taipeiensis belong to the Sycoryctinae subfamily (Chalcidoidea: Pteromalidae) and are the two non-pollinating fig wasp species associated with the gynodioecious Ficus benguetensis Merrill (Berg 2011; Wong et al. 2018). The trees bear either seed-producing female or wasp-producing “male” figs. Until recently (Segar et al. 2012), the genus Sycorycteridea and three other genera (Arachonia, Sycoryctes, and Sycoscapteridea) were synonymized under the Sycoscapter genus (Bouček 1988).

Sample collection and behavior observation

Thirty ripe unexited figs were haphazardly collected from five F. benguetensis trees in Taiwan from October 2011 to July 2012 (see Suppl. material 1: Table S1). Unripe figs are dark green and firm to the touch, whereas ripening figs gradually change to yellowish green and yield to gentle pressure. The diameters of all sampled figs were measured before the figs were stored at room temperature in transparent plastic pots covered with mesh for 24 h with their ostioles facing upward. A fig’s diameter is an indicator of its size and correlates with the size of the lumen, i.e. the space available for the male fig wasps to fight. Onset of wasp exit was identified by the presence of a hole chewed through the fig-wall. All wasps that exited from their natal figs were collected and stored in 75% ethanol. The figs were then dissected and the wasps inside them were collected. All the collected wasps were then identified, sexed, and counted under a stereomicroscope at 6.3× magnification. Images were captured using a digital reflex camera. Head width (across the eyes) and mandible length were measured using the 1.45a image processing software (National Institutes of Health, USA). Head width, rather than the hind tibia length, was used to estimate body size because limbs are often lost during fighting.

Fighting injuries and exit

The collected Philotrypesis and Sycorycteridea males were examined for their fighting injuries. All observed lacerations and amputations were noted. The injury level of an individual male was estimated according to a rating scale developed for fighting fig wasps (Murray 1987). However, because the behavior of male fig wasps is specialized to the dark environment of a fig, the scale was modified in the present study; loss of antenna was changed to from 0.5 point to 2 points (Table 1). The original scale developed by Murray (1987) focused on the fluid loss, whereas the modified scale considers the necessity for the male wasps to locate females in dark environments by using only their antennae. Mate-seeking ability is severely impaired when an individual loses part or all of its antennae. Hence, it was assigned a heavier weight in the present study. Injury level per fig was estimated according to the following three variables proposed by Murray (1987): (1) lifetime extent of injury (LEI), the injury scores of all males from the same fig divided by the total number of males; (2) injury frequency (IF), the proportion of injured males; and (3) severe injury frequency (SIF), the proportion of injured males with injury scores greater than 8 points.

Criteria used in scoring injuries of male fig wasps.

Score Description
1 Loss of part or whole tarsus
1 Loss of part or whole mandible
2 Loss of part or whole antenna
2 Loss of part or whole tibia
3 Loss of part or whole femur
4 Loss of part or whole coxa
4 Laceration in abdomen
8 Decapitation

Male exit was estimated as the proportion of males exiting their natal fig, namely the “male exit proportion”, and was calculated as the number of exited males divided by the total number of males per fig.

Data analysis

The sex ratio for each species was calculated as the number of male wasps divided by the total number of wasps within a fig. All statistical tests were performed using the R 3.6.1 statistical software (R Development Core Team 2019). The dataset was not normally distributed so the used tests were nonparametric. To investigate the factors affecting male exit proportion, multiple linear regression (which do not require a normally distributed dataset) was performed, with five variables: average head width, fig diameter, LEI, morph ratio, and wasp sex ratio.


Male morphology and sex ratio

One hundred eighty-two Philotrypesis and 42 Sycorycteridea male wasps were collected. The average sex ratio of the sampled figs was 0.36 ± 0.03 for Philotrypesis and 0.47 ± 0.09 for Sycorycteridea, which was significantly different from the 1:1 ratio for Philotrypesis (t = 4.44, df = 48, P < 0.001) but not for Sycorycteridea (t = 0.39, df = 22, NS). The sex ratio of the two species was not different (t = -1.16, df = 14.03, NS).

Two Philotrypesis male morphologies, namely “atypical” and “typical” morphs, were identified. They were distinguished by the shapes of their mandibles (Fig. 1; alternative images in Suppl. material 2: Fig. S1); the atypical males (7% of the males) have scythe-shaped mandibles, whereas the typical males have three-tooth mandibles. In addition, atypical males were significantly larger than the typical males (Table 2). The largest Philotrypesis male was approximately twice the size of the smallest, and its mandible was approximately three times as long (Table 2; Fig. 2a). The sizes of the two morphs overlapped; the smallest P. taida atypical males were smaller than many of the typical males (Fig. 2a). The Sycorycteridea males were divided into two morphs according to their size (Table 2): the “small” morph, having head widths narrower than 0.425 mm (57% of the individuals), and the “large” morphs, having heads wider than 0.425 mm (Fig. 2b). The largest Sycorycteridea male was approximately thrice the size of the smallest and had a mandible approximately four times as long. Sycorycteridea mandible length and head width were discontinuously distributed. No male with a head width between 0.400 and 0.450 mm was found. The measurement gap between the two morphs represents approximately 12% and 19% of the average male head width and mandible length, respectively.

Figure 1. 

Frontal view of the males of the two studied species and mandible outline (a) a “typical” Philotrypesis male: rounded back of the head and broad mandibles (b) “atypical” Philotrypesis male: head more square and falcate mandibles (c) Sycorycteridea small male, rectangular head (d) Sycorycteridea large male, lantern-shaped head.

Head width and mandible length of each male morph.

Morph N Mean head width Head width range Mean mandible length Mandible length range
Philotrypesis Atypical 13 0.673 ± 0.187 a 0.550–0.719 0.323 ± 0.089 0.242–0.370
Typical 169 0.576 ± 0.044 b 0.366–0.675 0.233 ± 0.018 0.134–0.310
Sycorycteridea Large 18 0.502 ± 0.118 c 0.453–0.562 0.290 ± 0.068 0.243–0.326
Small 24 0.336 ± 0.069 d 0.215–0.388 0.153 ± 0.031 0.078–0.230
Figure 2. 

Relationship between head width and mandible length in a Philotrypesis and b Sycorycteridea. Typical Philotrypesis and small Sycorycteridea are represented by the unfilled circles whereas atypical Philotrypesis and large Sycorycteridea are represented by filled circles.

The mandible length of Philotrypesis males correlated positively with head width (Spearman rank correlation test: ρ = 0.759, df = 180, P = 2.58E-35; Fig. 2a). In addition, significantly positive correlations were independently observed in the atypical morph (Spearman rank correlation test: ρ = 0.714, df = 167, P = 1.09E-27) and typical morph (Spearman rank correlation test: ρ = 0.828, df = 11, P = 0.000473). The mandible length of the Sycorycteridea males exhibited a positive relationship with their head width (Spearman rank correlation test: ρ = 0.939, df = 40, P = 3.82E-20; Fig. 2b). Furthermore, the same positive correlation existed independently for the large males (Spearman rank correlation test: ρ = 0.765, df = 22, P = 1.35E-05) and small males (Spearman rank correlation test: ρ = 0.765, df = 22, P = 1.35E-05).

Fighting injuries

The number of Philotrypesis males of each morph in each injury level was not different (χ²1 = 0.48, NS), same for the Sycorycteridea morphs (χ²1 = 0.00, NS; Table 3). Additionally, no difference was observed in average injury score between the atypical and typical Philotrypesis males (Wilcoxon rank sum test: W = 1096.5, P = 0.99; Table 4) or between the Sycorycteridea male morphs (Wilcoxon rank sum test: W = 197.5, P = 0.64; Table 4). Finally, half of the Sycorycteridea large males were uninjured (Table 3; Fig. 3b), whereas more than half of the atypical Philotrypesis were severely injured (Table 3).

Proportion of injured males.

Philotrypesis Sycorycteridea
typical (N = 169) atypical (N = 13) small (N = 24) large (N = 18)
Uninjured (injury score = 0) 0.11 0.15 0.29 0.50
Minor (injury score <8) 0.46 0.31 0.50 0.33
Severe (injury score ≥8) 0.43 0.54 0.21 0.17

Mean injury score per individual and injury proportion distributed among different body parts.

Philotrypesis Sycorycteridea
typical (N = 169) atypical (N = 13) small (N = 24) large (N = 18)
Mean injury score 7.4 ± 0.6 a 7.1 ± 2.0 a 3.8 ± 0.8 b 3.5 ± 0.8 b
Abdomen 0.01 0.00 0.00 0.00
Antenna 0.37 0.62 0.25 0.28
Coxa 0.31 0.23 0.21 0.11
Decapitation 0.08 0.00 0.04 0.00
Femur 0.49 0.46 0.25 0.22
Tarsus 0.42 0.38 0.29 0.22
Tibia 0.32 0.46 0.17 0.22
Figure 3. 

Proportions of a Philotrypesis and b Sycorycteridea exited and injured males.

Philotrypesis had a significantly higher LEI value (t = 3.11, df = 35, P < 0.01; Table 5) and a higher proportion of injured males than did Sycorycteridea (χ²1 = 15.58, P < 0.01; Table 5); however, the proportion of severe injury was not different between the two species (χ²1 = 7.86, NS). Rates of decapitation were low in Philotrypesis and Sycorycteridea males and all decapitated specimens were small Sycorycteridea and typical Philotrypesis males (Table 4; Fig. 3). The femur and tarsus were the most commonly injured body parts for the Philotrypesis males, whereas for the Sycorycteridea males, the antenna and tarsus were the most commonly injured parts (Table 4).

Average numbers per fig and injury rates for the two non-pollinating fig wasp species (N = 30 figs).

Philotrypesis Sycorycteridea
Mean wasp number 21 ± 5 12 ± 3
Male exit rate 0.31 ± 0.06 0.45 ± 0.13
Lifetime Extent of Injury (LEI) 5.6 ± 0.6 a 2.5 ± 0.7 b
Proportion of injured males (IF) 0.88 ± 0.02 a 0.62 ± 0.07 b
Severe Injury Frequency (SIF) 0.23 ± 0.03 0.10 ± 0.05

The calculation of the injury score differs from Murray’s scoring (1987) from the points allotted to the antenna injuries, increasing their score from 1.5 or 3 points from Murray’s score. Seventy-one (39.0%) Philotrypesis males and 11 (26.2%) Sycorycteridea males had antenna injured. Moreover, 12 Philotrypesis males and two Sycorycteridea males have seen their injury level increased using our scoring method (from minor to severe). The new injury scores were significantly higher than Murray’s scores for Philotrypesis (Wilcoxon signed rank test: V = 2556, P = 3.548e-15) and for Sycoscapteridea (Wilcoxon signed rank test: V = 66, P = 0.001586).

Male exit

Most of the male fig wasps exited within 2 h after the first wasp had emerged from the fig. No difference in the proportion of exited males was observed between the two non-pollinating species (χ²1 = 0.00, NS; Table 5). However, the two species differed in terms of the injury scores of exited males. First, the average injury scores of exited and philopatric Philotrypesis males were 3.1 (range: 0–8) and 4.2 (range: 0–18), respectively, indicating significant variation between the two groups (Wilcoxon rank sum test: W = 4329.5, P = 0.001145). The test for Sycorycteridea values was not significant (Wilcoxon rank sum test: W = 229.5, P = 0.1595): 1.33 (range: 0–5) for exited males and 2.33 (range: 0–11) for philopatric males. Moreover, the severe injury proportion was lower for exited males than for philopatric males among Philotrypesis males (χ²1 = 6.25, P < 0.05) but not among the Sycorycteridea males (χ²1 = 0.00, NS).

The multiple linear regression model contained five variables (average head width, fig diameter, LEI, morph ratio, and sex ratio) was significant, showing that the five variables collectively explain the exit rate. In addition, the average Philotrypesis male head width was the main contributor of the model (Table 6).

Result of the multiple linear regression model with the male exit proportion as dependent variables.

Philotrypesis (R2 = 0.47* F = 3.37) Sycorycteridea (R2 = 0.52 F = 1.28)
β t β t
Average head width 4.05 2.36* 1.15 0.28
Fig. diameter -0.02 -1.40 -0.01 -0.22
LEI -0.02 -0.85 -0.07 -1.12
Morph ratio -0.49 -1.09 -0.48 -0.67
Wasp sex ratio -0.32 -0.90 -0.93 -1.61


Our results show that the males of each of the two studied species had two clearly defined morphs. The Philotrypesis males could not be categorized according to their size only because some of the atypical males were smaller than the largest typical male: The shapes of the heads and mandibles were the discriminative morphological features. The shapes of the head and the mandibles were also discriminative for the morphs of the Sycorycteridea males but these males could also be segregated according to their size only. Such dimorphism (discontinuous distribution of male size) has never been observed before in male fig wasps. Moreover, injuries were found in every morph of both species, and it seems that belonging to a specific morph does not affect the probability of injury. Nevertheless, morph type may affect the severity of injury; no decapitated males belonging to the atypical Philotrypesis or large Sycorycteridea morphs were found in this study. However, the absence of decapitated males could be a random effect of the sampling as only the little number of atypical Philotrypesis has been found.

In contrast to other Philotrypesis species (Cook et al. 1997; Jousselin et al. 2004), no winged males have been found in P. taida. The dimorphism of the Philotrypesis morphs in the present study is similar to that of congeneric species living on F. septica in Australia (Cook and Bean 2006) and F. rubiginosa (Moore et al. 2009). In other words, in the three Philotrypesis species whose morphometries have been analyzed, atypical males have longer mandibles than typical males (for a given head size). The main difference between the Taiwanese species and Australian species is that the Australian Philotrypesis morphs are more clearly separated by their head and mandible morphometry (Cook and Bean 2006; Moore et al. 2009) (Fig. 2a). To the best of our knowledge, no other study has described the morphometry of the Philotrypesis male morphs.

The Sycorycteridea morphs in this study were uniquely morphometrically defined; no continuous distribution of mandible length or head width was noted, in contrast to Australian Sycoscapter species living on F. macrophylla (Bean and Cook 2001) or F. rubiginosa (Moore et al. 2008). Genera Sycorycteridea and Sycoscapter are closely related but phylogenetically distinct (Segar et al. 2012). Dimorphism seems to be expressed differently among the genera of a single hymenopteran group, with three distinct morphometric manners to be dimorphic according to Cook and Bean (2006). Our study added a fourth manner among Sycorycteridea morphs, namely discrete size-dependent dimorphism. The diversity of the allometric dimorphism has also been documented in other insect genera such as Onthophagus dung beetles (Emlen et al. 2005), Oxyporus rove beetles (Hanley 2001), and hymenopteran insects (Danforth and Desjardins 1999).

The injury frequencies (IF) of P. taida and S. taipeiensis in this study are the highest ever documented in any non-pollinating fig wasp species (88% and 62% for Philotrypesis and Sycorycteridea respectively). In congeneric species, very few fight winners were injured, whereas 22% of Philotrypesis losers and 73% of Sycorycteridea losers were injured (Moore et al. 2008). As examples for other genera, in Walkerella, the IF was approximately 17% (Wang et al. 2010), and in Idarnes wasps, it was approximately 40% (Pereira and Prado 2008). In our study, morph type had no effect on IF for either of the studied species; however, for another species of Philotrypesis for a given mandible size, atypical males were less injured (Moore et al. 2009). The IF of non-pollinating wasps living in F. benguetensis figs is higher than that of other species; thus, environmental factors might be key for determining the intensity of fights. Indeed, IF is often affected by the number of conspecific males (Murray 1987; Pereira and Prado 2005, 2008; Wang et al. 2010), fight duration (Moore et al. 2008, 2009), fig diameter (Pereira and Prado 2005, 2008), male size (Moore et al. 2008; Pereira and Prado 2008), and number of females (Pereira and Prado 2008). Furthermore, injuries can occur only after a certain number of males have hatched (Wang et al. 2010). On the basis of our data, we cannot discriminate between hypothetical explanatory factors. The severe injury frequency (SIF) values of P. taida and S. taipeiensis in this study are within the ranges presented in the literature; SIF was low (1–3%) in Idarnes wasps (Pereira and Prado 2008), approximately 25% in Sycoscapter wasps (Bean and Cook 2001), and approximately 44% in Sycoscapter fight losers (Moore et al. 2008). Also, 11% of Philotrypesis losers-four times less than the SIF of our studied Philotrypesis species-were severely injured (Moore et al. 2008). This difference may be due to the changes we made to Murray’s injury score (1987), increasing the value of damaged antennae.

Although studies on the dispersal (exit from a natal fig) of non-pollinating wingless male fig wasps are extremely limited, the few published studies have revealed a wide range of exit rates; 27%–71% of males of non-pollinating fig wasp species living in F. ingens figs exit their native figs (Greeff and Ferguson 1999), as do 17% of Sycoscapter australis (Bean and Cook 2001) and 29% of Walkerella sp. ex. F. benjamina male wasps (Wang et al. 2010). The exit rate of the Philotrypesis males from F. benguetensis (31%) is very close to that of Philotrypesis (35%) from F. ingens (Greeff and Ferguson 1999). In addition, species from the Sycoryctini tribe exhibit a wide range of exit rates: 17% for Sycoscapter australis on F. macrophylla in Australia (Bean and Cook 2001), 45% for Sycorycteridea sp. 1 on F. benguetensis in Taiwan; and 71% for Sycoryctes sp. on F. ingens in South Africa (Greeff and Ferguson 1999). These numbers reveal highly variable exit rates among studied species in the literature, and even among species living in the same figs (F. benguetensis and F. ingens wasps). This feature may be a strong indicator that exit rate is not primarily driven by environmental factors but by genetic or maternal factors. Even in a single wasp species, different morphs can have very distinct exit rates; almost all typical P. taida typical males exit figs, whereas very few of atypical male ones do. This situation is similar to that of the non-pollinating wasp species Otitesella longicauda, which has two morphs. The dispersing morph (called digitata) always exits figs, whereas the other morph (called religiosa) almost never leaves its natal fig (Greeff and Ferguson 1999). The exit rates of different morphs in a single species do not always differ extremely; for example, in contrast to the Otitesella rotunda digitata and religiosa morphs having an exit rates of 70% and 27%, respectively, the small and large Sycorycteridea sp. 1 have exit rates of 58% and 42%, respectively, and the black and pale morphs of Walkerella sp. ex. F. benjamina have low exit rates of 36% and 23%, respectively (Wang et al. 2010). Differences in exit rates between intraspecific morphs show that dispersal is likely not influenced by environmental cues but by maternal choices, and these choices may be influenced by environmental factors. Nevertheless, some exit rate values fit the morphological adaptations of the males, such as the digitata males (attracted to light and strong legs) and black Walkerella males (darker males may better resist to dehydration). However, the smaller morphs of P. taida and S. taipeiensis exited most frequently; for Philotrypesis, it may be that the largest males of the typical morph exited (i.e., the head width is the only significant factor to explain the exit rate). The fact of smaller morphs exiting more frequently is somewhat counterintuitive because figs of F. benguetensis are patrolled by numerous ants (Lin et al. 2016) and are, on average, more frequently and more severely injured (for Philotrypesis) than the larger morphs; thus, their chances of successful dispersal are low. Smaller males may exit as an alternative strategy to avoid more competitive fighting males inside figs, and because figs can be closely clustered together (Suppl. material 2: Fig. S2), exited males may find females in neighboring figs. Finally, it seems that males with severe injuries are less inclined to exit figs than are males with minor injuries.

Considering the high probability of being captured by ants (Lin et al. 2016), the cost of dispersal seems very high; however, because a fig regularly contains few mothers (in three Sycoscapter species) and because 10%–22% of females are born in figs without conspecific males (Cook et al. 2017), mating in one’s non-natal fig would be extremely rewarding from a genetic perspective. The dispersal strategy could be adopted by less successful fighting males, but not by the smallest males.


The findings of this study confirm the type of dimorphism that has been described in other Philotrypesis species and reveal a new dimorphism type in the Sycoryctinae subfamily. Some intermorph values found in this study could be used as insights for behavioral differences between morphs related to aspects such as fighting or exiting figs. Our research provides basic information that could lead to behavioral studies of these males and the environmental factors that may affect the choices of fig wasp mothers to produce any morph at an oviposition site.


The authors thank Prof. Tzeng Hsy-Yu from National Chung Hsing University for generous guidance on study methods. The authors would also like to express deepest appreciation to Prof. Hsu Yu-Ying and Prof. Yang Jeng-Tze from National Taiwan Normal University and National Chung Hsing University, respectively, for their valuable suggestion and discussion. This research is supported by National Science Council of Taiwan (99-2923-B-002-001-MY3).


  • Berg CC (2011) Corrective notes on the Malesian members of the genus Ficus (Moraceae). Blumea – Biodiversity, Evolution and Biogeography of Plants 56: 161–164.
  • Berthelot K, Portugal FR, Jeanson R (2017) Onset of fights and mutual assessment in ant founding queens. Journal of Experimental Biology 220: 750–753.
  • Bouček Z (1988) Australasian Chalcidoidea (Hymenoptera): a Biosystematic Revision of Genera of Fourteen Families with a Reclassification of Species. CAB International Institute of Entomology, Wallingford, 832 pp.
  • Casey C, Charrier I, Mathevon N, Reichmuth C (2015) Rival assessment among northern elephant seals: evidence of associative learning during male-male contests. Royal Society Open Science 2: 150228.
  • Cook JM (2005) Alternative mating tactics and fatal fighting in male fig wasps. In: Fellowes MDE, Holloway GJ, Rolff J (Eds) Insect Evolutionary Ecology. Royal Entomological Society, UK, 83–109.
  • Cook JM, Compton SG, Herre EA, West SA (1997) Alternative mating tactics and extreme male dimorphism in fig wasps. Proceedings of the Royal Society of London Series B 264: 747–754.
  • Cook JM, Reuter C, Moore JC, West SA (2017) Molecular markers reveal reproductive strategies of non-pollinating fig wasps. Ecological Entomology 42: 689–696.
  • Danforth BN, Desjardins CA (1999) Male dimorphism in Perdita portalis (Hymenoptera Andrenidae) has arisen from preexisting allometric patterns. Insectes Sociaux 46: 18–28.
  • Dunn DW, Jandér C, Lamas AG, Pereira RAS (2015) Mortal combat and competition for oviposition sites in female pollinating fig wasps. Behavorial Ecology 26: 262–268.
  • Emlen DJ, Hunt J, Simmons LW (2005) Evolution of sexual dimorphism and male dimorphism in the expression of beetle horns: Phylogenetic evidence for modularity evolutionary lability and constraint. The American Naturalist 166: 42–68.
  • Gloag R, Heard TA, Beekman M, Oldroyd BP (2008) Nest defence in a stingless bee: What causes fighting swarms in Trigona carbonaria (Hymenoptera, Meliponini). Insectes Sociaux 55: 387–391.
  • Hanley RS (2001) Mandibular allometry and male dimorphism in a group of obligately mycophagous beetles (Insecta: Coleoptera: Staphylinidae: Oxyporinae). Biological Journal of the Linnean Society 72: 451–459.
  • Hartley CS, Matthews RW (2003) The effect of body size on male-male combat in the parasitoid wasp Melittobia digitata Dahms (Hymenoptera: Eulophidae). Journal of Hymenoptera Research 12: 272–277.
  • Heinze J, Bernahrd S, Zieschank V (2016) Massacre of the innocents in Cardiocondyla thoracica: Manipulation by adult ant males incites workers to kill their immature rivals. Entomological Science 19: 239–244.
  • Huntingford FA (2013) Animal Conflict. Springer Science & Business Media, UK.
  • Jousselin E, van Noort S, Berry V, Rasplus J-Y, Rønsted N, Erasmus JC, Greeff JM (2008) One fig to bind them all: Host conservatism in a fig wasp community unraveled by cospeciation analyses among pollinating and non pollinating fig wasps. Evolution 62: 1777–1797.
  • Jousselin E, van Noort S, Greeff JM (2004) Labile male morphology and intraspecific male polymorphism in the Philotrypesis fig wasps. Molecular Phylogenetics and Evolution 33: 706–718.
  • Kärcher MH, Menezes C, Alves DA, Beveridge OS, Imperatriz-Fonseca VL, Ratnieks FLW (2013) Factors influencing survival duration and choice of virgin queens in the stingless bee Melipona quadrifasciata. Naturwissenschaften 100: 571–580.
  • Kemp DJ, Wiklund C (2001) Fighting without weaponry: a review of male-male contest competition in butterflies. Behavioral Ecology and Sociobiology 49: 429–442.
  • Kerdelhué C, Rasplu J-Y (1996) Non-pollinating Afrotropical fig wasps affect the fig-pollinator mutualism in Ficus within the subgenus Sycomorus. Oikos 75: 3–14.
  • Kjellberg F, Jousselin E, Hossaert-McKey M, Rasplus J-Y (2005) Biology ecology and evolution of fig-pollinating wasps (Chalcidoidea Agaonidae). In: Raman A, Schaefer W, Withers TM (Eds) Biology Ecology and Evolution of Gall-inducing Arthropods. CRC Press, US, 539–572.
  • McElligott AG, Mattiangeli V, Mattiello S, Verga M, Reynolds CA, Hayden TJ (1998) Fighting tactics of fallow bucks (Dama dama Cervidae): Reducing the risks of serious conflict. Ethology 104: 789–803.
  • R Core Team (2019) R: A language and environment for statistical computing. R Foundation for Statistical Computing Vienna Austria.
  • Rudolph KP, McEntee JP (2015) Spoils of war and peace: Enemy adoption and queen-right colony fusion follow costly intraspecific conflict in acacia ants. Behavioral Ecology 27: 793–802.
  • Segar ST, Lopez-Vaamonde C, Rasplus J-Y, Cook JM (2012) The global phylogeny of the subfamily Sycoryctinae (Pteromalidae): Parasites of an obligate mutualism. Molecular Phylogenetics and Evolution 65: 116–125.
  • Tomkins JL (1999) Environmental and genetic determinants of the male forceps length dimorphism in the European earwig Forficula auricularia L. Behavioral Ecology and Sociobiology 47: 1–8.
  • Vieira MC, Peixoto PE (2013) Winners and losers: a meta-analysis of functional determinants of fighting ability in arthropod contests. Functional Ecology 27: 305–313.
  • Wang Z-J, Peng Y-Q, Compton SG, Yang D-R (2010) Reproductive strategies of two forms of flightless males in a non-pollinating fig wasp under partial local mate competition. Ecological Entomology 35: 691–697.
  • Wong D-M, Bain A, Chou L-S, Shiao S-F (2018) Description of two new species of fig wasps (Chalcidoidea: Pteromalidae: Sycoryctinae) associated with Ficus benguetensis. Taiwania 63: 155–162.