Skip to main content
FreeNatural History Note

Surviving in Sympatry: Paragenital Divergence and Sexual Mimicry between a Pair of Traumatically Inseminating Plant Bugs


Reproductive interactions between species can carry significant costs (e.g., wasted time, energy, and gametes). In traumatically inseminating insects, heterospecific mating costs may be intensified, with indiscriminate mating and damaging genitalia leading to damage or death. When closely related traumatically inseminating species are sympatric, we predict selection should favor the rapid evolution of reproductive isolation. Here we report on a cryptic species of traumatically inseminating plant bug, Coridromius taravao, living sympatrically with its sister species, Coridromius tahitiensis, in French Polynesia. Despite their sister-species relationship, they exhibit striking differences in reproductive morphology, with females of each species stabbed and inseminated through different parts of their abdomens. Furthermore, C. tahitiensis is sexually dimorphic in coloration and vestiture, while both sexes of C. taravao share the C. tahitiensis male expression of these traits. These findings support a role for (1) reproductive character divergence and (2) interspecies sexual mimicry in limiting interspecific mating brought about by indiscriminate male mating behavior.

Online enhancement:   video. Dryad data:


Competition for a shared limited resource often precludes the coexistence of closely related species (Hochkirch et al. 2007). This is sometimes resolved by ecological segregation, such as host plant switching in phytophagous insects (Walsh 1867; Berlocher and Feder 2002), and in this manner, several closely related species are able to live in sympatry with minimal contact. Competition for resources, however, is not the only form of interspecific interaction that may impede species coexistence. Often overlooked are interspecific sexual interactions brought about by imperfect or nonexistent species recognition systems, which can also have severe fitness costs for one or all species involved (e.g. wasted time, wasted gametes, decreased reproductive success, physical damage, or even death; Gröning and Hochkirch 2008). This reproductive interference can lead to physical segregation and/or reproductive character displacement (Gröning and Hochkirch 2008; Crampton et al. 2011; Bath et al. 2012). Competition and reproductive interference are, thus, both important components affecting species distribution patterns, as well as potential drivers of species diversification.

On isolated islands, where resources and space are limited, interactions between closely related species may be difficult to avoid, and we may therefore expect heightened selection for the rapid evolution of various reproductive isolating mechanisms. These may include microhabitat segregation (Coyne and Orr 1989; Berlocher and Feder 2002; Drès and Mallet 2002; Matsubayashi et al. 2010); divergence in mate recognition systems (Crampton et al. 2011), and the evolution of genital or gamete incompatibility (Alipaz et al. 2001; Servedio 2001; Bath et al. 2012; Hollander et al. 2013). This may be further exacerbated in species whose mating systems are characterized by a lack of precopulatory mate choice, as in some sexual conflict systems where female reluctance and coercive mating is the norm (Arnqvist 1992).

In the traumatically inseminating plant bug genus Coridromius Signoret (Insecta: Heteroptera: Miridae), males use hypodermic genitalia to stab females in the abdomen during sex, bypassing the female genital tract and inseminating into their body cavity (Tatarnic et al. 2006; Tatarnic and Cassis 2008, 2010). Bed bugs and bat bugs (Insecta: Heteroptera: Cimicidae) also mate in this manner (Carayon 1966), and studies in these insects, have shown that traumatic insemination is costly to females, with damage and infection leading to reduced longevity and lifetime reproductive success (Stutt and Siva-Jothy 2001; Morrow and Arnqvist 2003; Reinhardt et al. 2003). Driven by these costs, females in many traumatically inseminating lineages have evolved elaborate internal and external modifications (paragenitalia) at the new site of insertion, such as copulatory guides and specialized immune organs to restrict damage and fight infection (e.g., bedbugs: Morrow and Arnqvist 2003; Reinhardt et al. 2003; Coridromius: Tatarnic et al. 2006).

In Coridromius, mating lacks precopulatory courtship, and males instead pounce on nearby females, with whom they struggle violently in their attempt to mate (video 1). Mating appears indiscriminate, and in some species (including C. tahitiensis) males have occasionally been observed briefly mounting and struggling with other males. These events have been interpreted as misdirected mating attempts (Tatarnic et al. 2006), though they may also confer some benefit to males (see Levan et al. 2009). Since traumatic insemination is costly to females (Stutt and Siva-Jothy 2001; Morrow and Arnqvist 2003; Reinhardt et al. 2003; Benoit et al. 2011), same-sex mating attempts should be costly to “recipient” males as well (Ryne 2009). Evidence for this is found in the bat bug Afrocimex constrictus, in which misdirected matings are common and have led to males also developing functional paragenitalia (Reinhardt et al. 2007). Evidence for this is also found in Coridromius, with the species C. bulbopella, exhibiting paragenitalia in both sexes (Tatarnic et al. 2006; Tatarnic and Cassis 2008).

Video 1. 
Video 1. 

Video 1. Coridromius tahitiensis mating. In C. tahitiensis males and females engage in vigorous mating struggles.

Thus, the costs of damaging mating can extend beyond conspecific females to any recipient of male mating attempts, which could also include other sympatric species. Several species of bed bug and bat bug are sometimes found in sympatry with congeners (Reinhardt and Siva-Jothy 2007), and when heterospecific matings occur, they are often fatal to females (e.g., Ueshima 1964; Reinhardt and Siva-Jothy 2007). In Coridromius, we predict that heterospecific mating costs should be equally high, with males and females of separate species neither physically nor physiologically coadapted to one another. We therefore predict that sympatric species should evolve mechanisms to limit costly interspecific matings.

Here we present evidence of reproductive isolation in a pair of sympatric species of Coridromius sharing the same host plants on Tahiti Island, French Polynesia. Though molecular sequence data show that C. tahitiensis (Tatarnic and Cassis 2008) and the recently discovered C. taravao (Tatarnic and Cassis 2013b) are sister species, females are inseminated through different abdominal regions via entirely different paragenital structures. Additionally, patterns of male and female vestiture and coloration suggest that both sexes of C. taravao may be mimicking the males of C. tahitiensis. We interpret (1) the paragenital divergence as evidence of reproductive character displacement, and (2) the color and vestiture traits as potential interspecies sexual mimicry. Both of these mechanisms could act to limit potentially costly interspecies traumatic inseminations.

Material and Methods

Specimen Collection

Over three trips (May 2011, February 2012, March–April 2012), collecting surveys were conducted across Tahiti and Moorea Islands. These data were supplemented by historical collecting data and from recent collecting efforts by colleagues from the University of California, Berkeley (Brad Baljukian, personal communication). Specimens were collected by beating vegetation, hand collection, and light trap. A summary of all collecting data is available from the Dryad Digital Repository (; Tatarnic and Cassis 2013a).

Morphological Examination

Somatic, genital, and paragenital characters were examined using a Leica MZ205 microscope and a Hitachi TM3000 desktop scanning electron microscope. All specimens were assessed visually for reproductive morphology, color pattern, and vestiture traits. For three-dimensional and cross-sectional visualization of reproductive structures, X-ray microtomography images (tomograms) were produced for both sexes of each species, using an X-Tek RTR-UF225 X-ray source and Roper PI-SCX100:2048 X-ray camera at the Australian National University Department of Applied Mathematics CT facility (Sakellariou et al. 2004). Volume rendering and tomogram visualization were conducted using the software package Drishti 2.1 (Limaye 2012), available freely from

Mating Behavior

To better understand mating behavior in Coridromius, we recorded mating interactions of C. tahitiensis using a Panasonic AG HPX172 HD video camera, with Hoytz +2 and +4 diopter lenses. Over a 6-day period (February 1–6, 2012), interactions between 12 males and 28 females of C. tahitiensis collected from Taravao Plateau were recorded. To film mating interactions, a glass enclosure was constructed using two 7.6 × 12.7-cm glass panes separated by ∼3 cm of matte board, with a cutting of host plant placed within as substrate. The shallow depth of the enclosure was necessary to keep the insects in focus yet still allow them to move freely. All matings and mating attempts were filmed between 7 a.m. and 12 p.m., as males appeared to lose interest in mating after this period. For each mating trial a single female was first introduced into the enclosure, followed by a male approximately 5 min later. Pairs were filmed for 30 min or until the male mounted or attempted to mount the female. The duration of interactions was timed from when individuals were within one body length of one another until physical contact was terminated.

In addition to these matings, a single C. tahitiensis male was observed attempting to mate with a C. taravao female, while confined in a 50-mL collecting tube. This mating attempt was not filmed.

Phylogenetic Analysis

To assess their evolutionary relationship, we conducted a Bayesian phylogenetic analysis of 2,051 aligned base pairs of mitochondrial (16S and CO1) and nuclear (28S subregions D2 and D3–5) sequence data. Taxa selected for the analysis form a monophyletic subclade, identified from an existing phylogeny of genus Coridromius (Tatarnic and Cassis 2010). These include C. variegatus (New Caledonia, Fiji); the three Australian endemics C. chenopoderis, C. monotocopsis, and C. pilbarensis, and C. norfolkensis, a newly described species endemic to Norfolk Island (Tatarnic and Cassis 2013b). Outgroup taxa include the two Southeast Asian species C. chinensis and C. testaceous. Coridromius bicolor from Vanuatu was unavailable for this study, however, based on a previous phylogeny (Tatarnic and Cassis 2010), it is not closely related to either Tahitian species. Sequences were edited and aligned using Geneious Pro 3.5.7 (available from Biomatters at PartitionFinder, version 1.0.1 (Lanfear et al. 2012), was used to select appropriate nucleotide substitution models for the different data set partitions (codon position and gene region), and the phylogenetic analysis was conducted using MrBayes 3.2 (Huelsenbeck and Ronquist 2001; four chains, 2,500,000 generations). Further details can be found in the appendix.



From the material examined, two species of Coridromius were identified, C. tahitiensis and C. taravao. Coridromius tahitiensis is found throughout Tahiti Island and the nearby island of Moorea (15 km away), while C. taravao is presently known only from two sites on Tahiti, which it shares with its congener. At Taravao Plateau (17.777°S, 149.255°W), males and females of both species were collected off the foliage and flowers of Metrosideros collina (J. R. Forster & G. Forster) A. Gray (Myrtaceae) in May 2011 and April 2012. At the base of the “Belvedere Trail,” Rau Ape Aorai, near Papeete (17.551°S, 149.529°W), they were found on the flowers of the introduced weed Schinus terebinthifolius Raddi (Anacardiaceae) in April 2012. We observed no apparent difference in microhabitat use, with both sexes of both species found on the flowers and foliage of their host plants. In total, 29 males and 25 females of C. tahitiensis and 6 males and 28 females of C. taravao were collected directly from their host plants. An additional 17 males and 36 females of C. tahitiensis and 5 females of C. taravao were collected by light trap adjacent to a stand of Metrosideros at Taravao Plateau in February 2012. A further 45 Tahitian Coridromius specimens were obtained on loan from the University of California, Berkeley, all of which were subsequently identified as C. tahitiensis (20 males, 25 females). All other specimens of Tahitian Coridromius known from museum collections have previously been examined by us and identified as C. tahitiensis.

Mating Behavior

Mating behavior in C. tahitiensis is characterized by an apparent lack of precopulatory mate choice. Of the 28 recorded mating trials, in 24 of these, the male approached and attempted to mount the female. Seventeen times the male was able to jump onto the female’s dorsum, with the female struggling immediately upon contact. Mate struggles lasted from less than 1 s to over 3 min ( s; mean ± SD). When mounted, females raised their right hind leg and were sometimes able to obstruct the male from reaching their paragenitalia or else were able to dislodge the male by kicking, shaking, and jumping. In two instances, the female ceased resisting, with both these mating attempts lasting over 2 min. In addition to male avoidance and struggling, some females would lunge at harassing males, with one female even leaping on her suitor, who immediately fled thereafter.

Coridromius taravao was not observed mating. However, one interspecific mating attempt was observed. While transporting specimens of both species in a 50-mL centrifuge tube, one male C. tahitiensis was observed mounting a female C. taravao three times over 2 min. Each time he mounted the female, a struggle immediately ensued, and the pair rapidly disengaged. Subsequent examination of the female revealed no noticeable mating damage.

Morphology: Coridromius tahitiensis

Coridromius tahitiensis, the larger of the two species (body length, males 1.90–2.97 mm, females 2.05–2.76 mm, both sexes) is the only one of 40 Coridromius species known to exhibit consistent and clear sexual color dimorphism (Tatarnic and Cassis 2008). Males are pale tan and green with a bright yellow scutellum, while females are mostly medium to dark brown with darker brown markings on their forewings (fig. 1A). Additionally, C. tahitiensis exhibits sexual dimorphism in abdominal vestiture, with males but not females adorned with a conspicuous patch of long hairlike setae on the right side of the abdomen (fig. 1C, 1D). This setal patch has not been observed in other Coridromius species.

Figure 1. 
Figure 1. 

AF, Comparison of Coridromius tahitiensis (left) and Coridromius taravao (right). A, Female and male C. tahitiensis exhibit sexual color dimorphism, while C. taravao (B) do not, with both sexes resembling male C. tahitiensis. Lateral views depict females of each species. CF, Ventral views of abdomens. In C. tahitiensis, males (D) exhibit a patch of hairlike setae on the right side of the body, absent in females (C). Both females (E) and males (F) of C. taravao also exhibit a patch of setae on their right side. G, Bayesian phylogenetic reconstruction of Austral-Pacific Coridromius species based on 16S, CO1, and 28S sequence data. Numbers adjacent to nodes are posterior probabilities. Scale bar indicates number of substitutions per site.

The male genitalia of C. tahitiensis are typical of the genus, with the scythe-like left paramere coupled with the aedeagus to form the intromittent organ (fig. 2A). When not in use, the left paramere rests within the genital capsule, concealed beneath the wings. When deployed, the paramere is extended caudally, braced in a groove in the posterior margin of the genital capsule (fig. 2A).

Figure 2. 
Figure 2. 

X-ray tomographs of Coridromius tahitiensis and Coridromius taravao. Both C. tahitiensis females (A) and C. taravao (B) females exhibit elaborate modifications associated with mating on the right side of their body. In C. tahitiensis, females are inseminated through a funnel-shaped copulatory opening at the trailing edge of the first visible abdominal segment, indicated by arrow (C). This opens into a sclerotized chamber, which is shown as a red spot in the translucent tomograph (D). In C. taravao, there is no paragenital opening behind the first abdominal segment (E). Instead, the first segment is heavily sclerotized, particularly along its swollen anterior margin (F), indicated by arrow. This anterior margin is hooked under the trailing edge of the mesially deflected metepimeral lobe of the thorax, which also interlocks with the margin of the forewing, shown intact (G) and in cross section (H), indicated by arrow. The exact site of insemination in C. taravao is unknown, but may be on the opposite side of the body, as evidenced by a depression on the first abdominal segment, behind the left metepimeral lobe, indicated by arrow (I). Male genitalia of C. tahitiensis (J) and C. taravao (K) differ mainly in size. Though structurally very similar, the intromittent organ (left paramere) of C. tahitiensis (J) is longer than that of C. taravao (K). Orange scale bars in CH = 0.2 mm.

Coridromius tahitiensis females are inseminated on the right side of the abdomen through specialized paragenitalia at the posterior margin of the first visible abdominal segment, manifested externally as a heavily sclerotized, funnel-shaped guide (fig. 2A, 2C). During mating, the male paramere is guided through this opening and into an invaginated, sclerotized chamber (fig. 2D), the epithelium of which sometimes bears melanized scars, indicative of past mating wounds (N. Tatarnic, personal observation). The male paramere punctures the wall of this chamber to inject sperm into a spherical membrane-bound organ, which we believe is analogous to the mesospermalege of bedbugs—a specialized organ containing phagocytic cells, which functions to reduce the risks of infection from mating (Morrow and Arnqvist 2003; Reinhardt et al. 2003).

Morphology: Coridromius taravao

Coridromius taravao is slightly smaller than C. tahitiensis, though size ranges overlap for both sexes of both species (body length, males 1.80–2.02 mm, females 1.83–2.20 mm, males, 10 females). In C. taravao, all males examined () and 23 of the 27 females are colored identically to C. tahitiensis males (three females from the Belvedere Trail and one from Taravao Plateau are brown as in C. tahitiensis females; fig. 1A). Again, similar to C. tahitiensis males, both sexes of C. taravao exhibit elongate hairlike setae on the right side of the abdomen (fig. 1E, 1F).

The male genitalia of C. taravao are similar to those of C. tahitiensis, though considerably smaller (∼3/5×; fig. 2K). This is consistent with their smaller body size.

Differences in the female paragenitalia, however, are more dramatic. In C. taravao, females are not inseminated between the first two abdominal segments on the right side of the abdomen. Instead, the first abdominal segment is heavily sclerotized, particularly along its swollen anterior margin, which interlocks with the trailing edge of the thorax (the metepimeral lobe; fig. 2B, 2E–2H). The metepimeral lobe is also coupled with a flange on the ventrolateral surface of the forewing (fig. 2G, 2H). These traits—the heavy sclerotization and interlocking of the metepimeron with both the abdomen and forewing—appear to restrict access to the right side of the body, including the thoracic-abdominal junction, a region where several related Coridromius species are inseminated (e.g., mating scars in C. pilbarensis females are found in the thoracic-abdominal membrane behind the right metacoxa, while in C. norfolkensis, C. bicolor, and C. variegatus, scars are found at the upper right corner of the abdomen: N. Tatarnic, personal observation). In C. taravao, we have yet to identify a paragenital sinus or find any evidence of scarring on the right side of the body. However, we have identified a putative paragenital guide on the opposite side of the abdomen, formed by a deep depression on the first visible sternite adjacent to the mesially deflected metepimeral lobe (fig. 2I).

Phylogenetic Relationships

Our phylogenetic analysis shows that C. tahitiensis and C. taravao are sister species, with their nearest relative the Australian endemic C. pilbarensis (fig. 1G). This echoes our previous morphology-based phylogeny of the genus, which placed C. tahitiensis as sister to C. pilbarensis (Tatarnic and Cassis 2010). Extensive collecting throughout French Polynesia has revealed no additional species beyond those we have identified, and because of Tahiti’s extreme isolation (over 6,000 km from Australia, the nearest mainland), repeat colonization events are unlikely. These data lead us to predict that these species have both evolved on Tahiti from a common ancestor.


Closely related species with similar ecological needs seldom coexist in sympatry without some mechanism in place to prevent interbreeding. In phytophagous insects, specialization on different host plants allows related species to live sympatrically and has even been invoked as a likely mechanism for sympatric speciation (Berlocher and Feder 2002; Drès and Mallet 2002; Matsubayashi et al. 2010). Alternatively, reproductive interference may lead to reproductive isolation through reproductive character displacement (Brown and Wilson 1956; Coyne and Orr 1989, 1997; Kameda et al. 2009; Crampton et al. 2011; Hollander et al. 2013). In C. tahitiensis and C. taravao, isolation via host specialization does not occur (though it may have in the past), and by sharing the same host plants, they are likely to come into regular contact with one another. As mating in Coridromius is characterized by aggressive premating struggles, a lack of precopulatory courtship, occasional misdirected mating attempts, and damaging mating, we believe that the potential for reproductive interference between C. tahitiensis and C. taravao should be high. Based on our observations, we propose two mechanisms by which these species maintain reproductive isolation despite their sympatric existence: (1) reproductive character divergence and (2) sexual mimicry.

Reproductive Character Divergence

Perhaps the strongest evidence of reproductive isolation comes from the reproductive morphology of these two species. Despite sharing a common ancestry, their divergence in reproductive traits, particularly between females (e.g., C. tahitiensis [fig. 2C, 2D] vs. C. taravao [fig. 2E–2I]), is extreme, with females of each species inseminated through different parts (and possibly different sides) of the body (fig. 2C and 2D vs. 2I). Furthermore, C. taravao exhibits morphological modifications that appear to be barriers against insemination through the right side of the abdomen. These female paragenital differences, combined with the differences in male intromittent organ size, may function as mechanical barriers to interspecific copulation. For example, the reinforced armature on the right-hand side of female C. taravao could protect against puncturing by the longer intromittent organ of C. tahitiensis males, while the relatively deep paragenital sinus of female C. tahitiensis may exclude piercing by the shorter intromittent organ of C. taravao males.

Traits associated with reproduction are often more divergent between species when they share sympatric distributions than when they are allopatric (Coyne and Orr 1989, 1997; Konuma and Chiba 2007; Hollander et al. 2013), with their diversification partly attributed to reproductive interference (Konuma and Chiba 2007; Hollander et al. 2013). This appears to be the case with Coridromius: within the Austral-Pacific subclade, only these two Tahitian species are known to exhibit true sympatry (geographically overlapping Australian species feed on separate host plants and are thus functionally isolated (Tatarnic and Cassis 2008), and they are the only two species in the group to exhibit elaborate paragenitalia. A cursory survey of Coridromius species suggests that this pattern may be widespread across the genus, as incidences of paragenital complexity and divergence tend to occur in areas where multiple species coexist (Tatarnic and Cassis 2008, 2010). However, due to limited sampling, further collecting in these undersampled areas (e.g., Borneo, New Guinea) will be needed to test this trend.

Sexual Mimicry

Sexual color dimorphism is known in several insects, though it is thought to be extremely rare in the Heteroptera (Punzalan et al. 2008a). Although our phylogenetic reconstruction confirms that the brown female morph represents the ancestral condition for both sexes, we cannot explain the differences in color morph expression between the sexes of the two species from the phylogeny alone. Below we discuss possible scenarios by which these might have evolved and are maintained.

First, let us consider C. tahitiensis. In the absence of male courtship or any other overt form of sexual signaling, we do not believe the sexual color dimorphism of C. tahitiensis functions as a signal in female mate choice or intrasexual competition. One possibility is that the green and yellow color of C. tahitiensis males facilitates mate recognition, reducing accidental male-male copulations. If color does facilitate mate recognition, then resemblance to C. tahitiensis males by both sexes of the smaller C. taravao could afford this species a degree of protection from accidental mating attempts by the larger C. tahitiensis males. Sexual mimicry by females to avoid mating harassment is known in several taxa (e.g., butterflies: Cook et al. 1994; dragonflies and damselflies: Sherratt and Forbes 2001; Gosden and Svensson 2007; Sherratt 2008); however, to our knowledge, sexual mimicry between species to avoid interspecific copulation has never been documented.

While a signaling function is commonly invoked to explain sexual color dimorphism, other nonsignaling mechanisms, including viability selection, cannot be ruled out. For example, sexual color dimorphism has been associated with differences in niche use (Selender 1966), the thermoregulatory needs of males (Punzalan et al. 2008b) and females (Ellers and Boggs 2004), protection from ultraviolet radiation (Cooper 2010), and even differences in immunity (Joop et al. 2006). It is possible, for example, that males and females of C. tahitiensis forage slightly differently, or, because of differences in immune stress brought about by traumatic mating, females exhibit more melanization. In this study, sex and species differences in microhabitat use are not apparent, with both sexes of both species observed on the same parts of their host plants. Though this provides tentative support to our mimicry hypothesis, it does not mean that such differences do not exist. Until we know more about the biology of these species, we refrain from making any definitive statements on either the origin or maintenance of the observed color morph patterns.

Lending further support to the sexual mimicry hypothesis is the peculiar abdominal vestiture exhibited by male C. tahitiensis and both sexes of C. taravao. As above, this trait is not known in any other species. Based on its location on the abdomen and the mating position of these insects, we propose that this vestiture trait functions as either a male sex recognition cue or as a mating impediment. Thus, in C. tahitiensis, if a male were to mount another male, as he probes with his genitalia for the paragenital opening, he will instead be confronted by this patch of setae, identifying his partner as a male. By extension, a male C. tahitiensis mounting either sex of C. taravao would be faced with a similar setal patch, which would misidentify the other individual as another male. In this manner, both sexes of C. taravao would be misclassified as conspecific males and reduce heterospecific mating. Furthermore, if C. taravao are indeed inseminated on the other side of the body, this setal patch should have no effect on within-species matings. Male gender-specific tactile cues in traumatically inseminating insects, and their mimicry by females, have also been described in bed bugs (Reinhardt et al. 2007), who apparently cannot distinguish sex of conspecifics without genital contact (Stutt and Siva-Jothy 2001; Siva-Jothy 2006). Occasional male-male mountings have been observed in several Coridromius species (N. Tatarnic, personal observation), suggesting that they too may have difficulty distinguishing their partner’s sex before contact.


The presence of any two sister species living sympatrically and sharing the same resources requires some mechanism to explain their continued coexistence. This is especially true when mating is indiscriminate and physically damaging, as in Coridromius. Here we provide multiple lines of evidence to explain how C. tahitiensis and C. taravao may be able to persist in sympatry. Future comparisons between sympatric and allopatric populations of both species along with phenotypic manipulations of the traits discussed herein (e.g., artificial modification of color patterns: Kemp 2007, 2008; microsurgical trait manipulation: Polak and Rashed 2010) are needed to test the adaptive hypotheses we have proposed. Future studies should also be undertaken to ascertain the efficacy of sex recognition by males, as what we interpret as maladaptive, indiscriminate mating may in effect be adaptive (e.g., Levan et al. 2009). Our findings suggest that reproductive interactions between closely related species, combined with indiscriminate male mating behavior, could play an important role in driving diversification across this genus, and may help explain the overwhelming diversity of genital forms we see in other animals as well.

We wish to thank J.-Y. Meyer (Délégation à la Recherche, Tahiti) for logistical support and host plant identifications, B. Baljukian (University of California, Berkeley) for additional specimens and collection data, and the Gump Field Station, Moorea. We also thank M. Turner and T. Senden for producing the X-ray tomographs and J. Sinclair and K. Umbers for assistance with molecular sequencing. This manuscript was significantly improved by comments from R. Bonduriansky, M. Herberstein, and M. Kasumovic. This research was funded through a grant from the Australia and Pacific Science Foundation to N.T. and G.C.

Field Collecting, French Polynesia

Over three trips we sampled insects from Tahiti Island. Specimens were collected predominantly by beating vegetation, with supplemental collection via light trapping. For a video tutorial on vegetation beating, visit the Planetary Biodiversity Plant Bug Inventory website ( Specimens examined for this study were collected from two localities:

Tahiti Island, Tahiti Iti, Taravao Plateau, 17.777°S, 149.255°W, 877 m asl. Specimens were collected off the flowers of Metrosideros collina (Myrtaceae; March 30–April, 7, 2012; February 1–6, 2012; May 15–25, 2011). During February 2012 specimens were also collected by light trap at the same location.

Tahiti Island, Rao Ape Aorai (base of “Belvedere Trail”), 17.551°S, 149.529°W, 331 m asl. Specimens collected off the flowers of Schinus terebinthifolius (Anacardiaceae; March 30–April 7, 2012).

Phylogenetic Analysis

Taxon Sampling

Taxa for the phylogenetic analyses were selected based on an existing phylogeny of the genus Coridromius (Tatarnic and Cassis 2010). These include C. variegatus (New Caledonia, Fiji), the three Australian endemics C. chenopoderis, C. monotocopsis, and C. pilbarensis, as well as a new, recently described species from Norfolk Island. Outgroup taxa include two Southeast Asian species: C. chinensis and C. testaceous. Coridromius bicolor from Vanuatu was unavailable for this study; however, based on our previous phylogenetic analysis, it is not closely related to either of the Tahitian species.

DNA was extracted using DNEasy Tissue Kits ( Extractions and polymerase chain reaction (PCR) amplifications were conducted at the University of New South Wales (UNSW) and the Australian Museum (AM). Because of the bugs’ minute size, whole specimens were used for DNA extraction. Voucher DNA is stored in the UNSW insect collection. Specimen voucher and collection data are supplied below.

Coridromius chenopoderis. Voucher ID C002: Australia: Western Australia: Burrup Conservation Area, near Karratha, 20.578°S, 116.795°E, 17 m asl, June 3, 2004, N. Tatarnic and S. Lassau (coll.), ex. Atriplex amnicola (Chenopodiaceae).

Coridromius monotocopsis. Voucher ID C004: Australia: New South Wales: Royal National Park, Wattamola Beach car park, 34.133°S, 151.113°E, 29 m asl, April 12, 2004, G. Cassis, P. Tinerella, N. Tatarnic, N. Velez, and K. McLachlan (coll.), ex. Monotoca elliptica.

Coridromius pilbarensis. Voucher ID C003. Australia: Western Australia: Pilbara Region, base of Mount Nameless, adjacent to gate. 22.727°S, 117.749°E, June 5, 2004, M. Bulbert, N. Tatarnic and S. Lassau (coll.), ex. Acacia sp. (Fabaceae).

Coridromius variegatus. Voucher ID C001: French Polynesia: New Caledonia: Province Sud: Les Bois du Sud campground, 22.172°S, 166.760°E, 200 m asl, April 23–25, 2005, G. Cassis, M. Wall, N. Tatarnic, and G. Monteith (coll.), ex. Phyllanthus sp. (Euphorbiaceae).

Coridromius norfolkensis. Voucher ID H308: Australia: Norfolk Island: Ball Bay, end of Marsh Road, 29.048°S, 167.980°E, 69 m asl, April 22, 2011, N. Tatarnic and A. Namyatova (coll.), ex. Schinus terebinthifolius (Anacardiaceae).

Coridromius tahitiensis. Voucher ID H205: French Polynesia: Society Islands: Tahiti Island: Tahiti Iti, Taravao Plateau, 17.777°S, 149.255°W, 877 m asl, May 19, 2011, G. Cassis and N. Tatarnic (coll.), Metrosideros collina (Myrtaceae).

Coridromius taravao. Voucher ID H305: French Polynesia: Society Islands: Tahiti Island: Tahiti Iti, Taravao Plateau, 17.777°S, 149.255°W, 877 m asl, May 19, 2011, G. Cassis and N. Tatarnic (coll.), ex. Metrosideros collina (Myrtaceae).

Coridromius chinensis. Voucher ID NT01: China: Yunnan Province: Xishuangbanna Tropical Botanical Garden, edge of gallery forest, 21.932°N, 101.245°E, 547 m asl, May 26, 2006, N. Tatarnic (coll.), ex Macaranga sp. (Euphorbiaceae).

Coridromius testaceous. Voucher ID NT12: China: Yunnan Province: Xishuangbanna Tropical Botanical Garden, edge of gallery forest, 21.932°N, 101.245°E, 547 m asl, May 26, 2006, N. Tatarnic (coll.), ex Macaranga sp. (Euphorbiaceae).


A total of 892 aligned mitochondrial nucleotides were amplified from the 16S (450 bp) and CO1 (442 bp) regions, while 1,159 aligned nuclear nucleotides were amplified from two 28S subregions: D2 (637 bp) and D3–5 (522 bp). The primers used are shown in table A1:

PCR protocol was as follows: one initial denaturation at 94°C for 3 min; and then denaturation: 30 s at 94°C, annealing: one cycle each at 55°, 53°, 51°, 49°, and 47°C, and 35 cycles at 45°C for 30 s; extension: 72°C for 45 s; and a final extension of 72°C for 10 min (MJ Research PTC 100 Thermocycler). The PCR product was visualized on a 2% agarose gel to ensure single bands of the expected size and then sent to Macrogen ( for cleanup and sequencing.

Concatenation of forward and reverse sequences and multiple sequence alignments were performed in Geneious Pro 3.5.7 ( Alignments were conducted using the Geneious Alignment method (default settings, five alignment iterations), followed by manual editing. Before analysis, PartitionFinder, version 1.0.1 (Lanfear et al. 2012), was used to select appropriate nucleotide substitution models for the different data set partitions (codon position and gene region). Data partitions were set for each gene region as well as each codon position for CO1. From this four best models were selected for our six subset partitions: HKY + G (16S); JC + I (CO1 position 2, D3–5); HKY + G (CO1 position 3); and K80 + I (CO1 position 1, D2). Phylogenetic analysis with the above partitions and nuclear substitution models was conducted using MrBayes 3.2 (Huelsenbeck and Ronquist 2001; four chains, 11,000,000 generations, burn-in 5,000). Genbank accession numbers are provided in table A2.

Table A1. 

Primers used for DNA amplification

RegionPrimerPrimer sequence (5′–3′)Source
CO1C1-N-2609GAAATACTGCTCCTATGGATADamgaard and Sperling 2001
CO1C1-J-2183CAACATTTATTTTGATTTTTTGGDamgaard and Sperling 2001

Note. AMNH = American Museum of Natural History.

View Table Image
Table A2. 

Genbank accession numbers

Taxon (Coridromius spp.)16SCO128S, D2 subregion28S, D3–5 subregion
C. chinensisKF005578KF036255KF036264
C. testaceousKF005579KF036246KF036254KF036263
C. monotocopsisKF005580KF036249KF036257KF036268
C. chenopoderisKF005581KF036248KF036256KF036267
C. pilbarensisKF005582KF036252KF036258KF036269
C. tahitiensisKF005583KF036253KF036261KF036270
C. taravaoKF005584KF036247KF036262KF036271
C. variegatusKF005585KF036250KF036260KF036265
C. norfolkensisKF036251KF036259KF036266

Literature Cited

An evening storm rolls in over the Taravao Plateau, Tahiti Island (photo credit: Nikolai Tatarnic).

Editor and Associate Editor: Mark A. McPeek