The male of Parahiranetis salgadoi Gil-Santana, 2015 is described for the first time, with a redescription of the female of this species based on additional specimens. Comments on possible mimicry and crypsis exhibited by adults and nymphs of this species, respectively, are provided.

Graptocleptes , Harpactorini , Heteroptera , Hiranetis , Neotropics, sexual dimorphism, wasp mimicry

Harpactorinae is the largest subfamily of Reduviidae and is represented by the tribes Apiomerini and Harpactorini in the Neotropical region (Gil-Santana et al. 2015). Harpactorini is the most diversified Reduviidae group with more than 53 recognized genera in the Neotropical region (McPherson and Ahmad 2011, Forero 2011, 2012, Swanson 2012, Gil-Santana 2015, Gil-Santana et al. 2015). Several taxa of Harpactoriniare thought to be involved in mimicry systems with Hymenoptera (Champion 1899, Elkins 1969, Maldonado and Lozada 1992, Hogue 1993, Gil-Santana 2008, Forero and Giraldo-Echeverry 2015, Gil-Santana 2015, 2016). They resemble bees or wasps in general body and wing coloration as well as physical proportions (Haviland 1931, Leathers and Sharkey 2003, Gil-Santana 2008, 2015, 2016). The striking resemblance resulting from this mimicry has frequently led to the confounding of some closely related genera, such as Hiranetis Spinola, 1840 and Graptocleptes Stål, 1866 (Gil-Santana et al. 2013).

Sexual dimorphism has been recorded in several species of Harpactorini. In a number of species belonging to Zelus Fabricius, 1803, for example, males and females differ drastically in size, body configuration, and coloration (Zhang et al. 2016). In addition to the bigger size and larger abdomen of females, males in several genera have larger eyes and/or a basally-thickened third antennal segment. The latter has been considered to be among the diagnostic features at generic level (Stål 1872, Champion 1899, Martin-Park et al. 2012, Gil-Santana et al. 2013, Gil-Santana 2016). However, sexual dimorphism may also be limited to minor differences in coloration and size, as in many species of Zelus (Zhang et al. 2016).

Male genitalia have been found to provide useful diagnostic characteristics for distinguishing species within the genera of Harpactorini (e.g. Elkins 1954a,b, Hart 1975, 1986, 1987, Forero et al. 2008, Zhang et al. 2016). The parameres, medial process of the pygophore, dorsal phallothecal plates, and the struts were the main structures with attributes important at the specific level in studies of Aristathlus Bergroth, 1913 (Forero et al. 2008), Atopozelus Elkins, 1954 (Elkins 1954a), Atrachelus Amyot & Serville, 1843 (Elkins 1954b), Ischnoclopius Stål, 1868 (Hart 1975) and Zelus (Hart 1986, 1987, Zhang et al. 2016). The endosoma contents, such as its processes, were not examined or recorded in most of these studies. Nevertheless, they have been described for one species of Graptocleptes (G. bicolor (Burmeister, 1838); Gil-Santana et al. 2013) and one of Hiranetis (H. atra Stål, 1872; Gil-Santana 2016). These genera are considered to be closely related to each other (Stål 1872, Champion 1899) and to Parahiranetis Gil-Santana, 2015 (Gil-Santana 2015, 2016), allowing for important comparisons of several diagnostic traits of male genitalia.

Parahiranetis is a monotypic genus containing the wasp-mimicking species P. salgadoi Gil-Santana, 2015. This genus was described based on two female specimens collected in the middle of last century (holotype) and in its first half (paratype) in Rio de Janeiro, Brazil (Gil-Santana 2015). Following the publication of the species, the second author (ATS) informed the first author (HRG-S) that she had observed and collected specimens of a wasp-mimicking reduviid in São Paulo, a state neighboring Rio de Janeiro, the type-locality. Surprisingly, the species that she observed was P. salgadoi. Through this opportunity, we were able to obtain both males and females, which allowed us to provide the first description of the male and a redescription of the female of this species and also record the morphological variability occurring in this species. We also provide photographs of live immature and adult specimens. The apparent mimicry and crypsis shown by the adults and nymphs, respectively, are also discussed.

The specimens described here are deposited in the Entomological Collection of the National Museum of the Federal University of Rio de Janeiro (Museu Nacional da Universidade Federal do Rio de Janeiro), Rio de Janeiro, Brazil (MNRJ). When citing the text on the labels of a pinned specimen, a slash (/) separates the lines and a double slash (//) different labels. All measurements are in millimeters (mm).

All fieldwork, including observation and collection of the living specimens from the state of São Paulo, was undertaken by the second author (ATS), who also pinned and dried these specimens for subsequent study. She obtained images of live specimens (Figs 43, 59–63) using a Canon EOS 60D digital camera with a Canon EF 100 mm macro lens.

Scanning electron microscopy images (Figs 4–5, 14–16, 18–21, 46–47, 54, 57–58) were obtained by the third author (JO). The male and female adults and adult female external genitalia from P. salgadoi were cleaned in an ultrasound machine. Subsequently, the samples were dehydrated in alcohol, dried in an incubator at 45 ºC for 20 min, and fixed in small aluminum cylinders with transparent glaze. Sputtering metallization was then performed on the samples for 2 minutes at 10 mA in an Edwards sputter coater. After this process, the samples were studied and photographed using a Topcon SM-300 scanning electron microscope, as described by Rosa et al. (2010, 2014).

The photo of the female holotype of Isthmiade braconides (Perty, 1832) (Coleoptera: Cerambycidae: Cerambycinae: Rhinotragini) (Fig. 64), which is deposited in the Bavarian State Collection of Zoology (Zoologische Staatssammlung München, ZSM), Munich, Germany, was kindly provided by Dr Steve Lingafelter (Arizona, USA).

All remaining figures were produced by the first author (HRG-S). The fixed adults, microscopic preparations and genitalia were photographed using digital cameras (Nikon D5200 with a Nikon Macro Lens 105 mm, Sony DSC-W830 and Sony DSC-HX400V). Drawings were made using a camera lucida. For clarity, the vestiture (setation) was omitted in the ink drawings of Figs 23 and 24. Images were edited using Adobe Photoshop 7.0.

Observations were made using a stereoscope microscope (Zeiss Stemi) and a compound microscope (Leica CME). Measurements were made using a micrometer eyepiece. The total length of the head was measured excluding the neck, for better uniformity of this measurement. Antennal segments were cleared in 20% NaOH solution for 72 hours for microscopic examination (Figs 6–10, 12–13, 49–52). Dissections of the male genitalia were made by first removing the pygophore from the abdomen with a pair of forceps and then clearing it in 20% NaOH solution for 24 hours. Following this procedure, the phallus was found to have spontaneously and completely inflated in one male (Figs 26–29). In another three specimens, the phallus was almost completely everted by carefully pulling on the dorsal endosomal processes using a pair of fine forceps (Figs 30–33). The dissected structures were studied and photographed in glycerol.

General morphological terminology mainly follows Schuh and Slater (1995). Currently, there is a lack of consensus about the terminology to be applied to female and male genitalia in Reduviidae (e.g. Rédei and Tsai 2011). Therefore, in order to maintain uniformity with previous works about species of Harpactorini, the terminology of the male and female genitalia structures follows Davis (1969), Forero et al. (2008), Gil-Santana et al. (2013) and Gil-Santana (2016).

Examination of additional specimens, including males, allowed us to provide a more complete description of the adults of Parahiranetis salgadoi. Gil-Santana (2015), for example, had not recorded the sticky substance on some parts of the body and appendages or the white wax-like substance covering groups of setae on the thorax and abdomen (Figs 14–15, 19–20, 53–54, 56), as both were completely absent from the type specimens. Moreover, the sticky substance was present only on some specimens from São Paulo, while the wax-like substance was sometimes absent from portions where it was observed on other specimens (Figs 55–56). Both substances may be lost during manipulation of the individuals and, in the case of the wax-like substance, this may also include loss of the thin fragile setae associated with it (HRG-S pers. obs.). However, the patches of short adpressed setae on the midlateral portions of sternites III and IV were completely covered with wax in some specimens (Fig. 56), while others clearly had no wax and the setae were nonetheless well preserved (Fig. 55). This observation makes it clear that individual differences (at least regarding this trait) can be found even in the same population and at the same observation time, possibly associated to the age and/or behavioral state of each specimen prior collection. Thus, in the type specimens, both substances may have been inadvertently removed; or may have disappeared over time, taking into account the fact that these are old specimens; or alternatively, these specimens may never have had either substance. In any case, it is no longer possible to know the extent to which these substances were present when the specimens were alive.

The sticky, viscous substance that covers the integument of some species of Harpactorini, particularly their fore and mid legs, is produced by specialized epidermal glands. These species use it to catch their prey (Zhang and Weirauch 2013a). Using comparative phylogenetic methods, Zhang and Weirauch (2013b) showed that species of Harpactorini that use a ‘sticky trap predation’ strategy have evolved fore femora that are more slender and longer than those of “non-sticky” bugs. Among some other members of Harpactorini, sticky glands were recorded in species of two genera that are considered close to Parahiranetis: Hiranetis braconiformis (Burmeister, 1835) and Graptocleptes sp. (Zhang and Weirauch 2013a). Interestingly, in the phylogenies obtained by Zhang and Weirauch (2013b), species of the wasp-mimicking genera Graptocleptes, Hiranetis, Myocoris, Neotropiconyttus and Xystonyttus were nested in the same clade, among other clades that together make up the “sticky bug clade” within Harpactorini.

Body parts covered with patches of setae with whitish wax-like material have been registered in some Harpactorini species, such as Cosmoclopius curacavensis Cobben & Wygodzinsky, 1975 (Cobben and Wygodzinsky 1975), Harpactor angulosus (Lepeletier & Serville, 1825) (Pikart et al. 2014), various species of Heza Amyot & Serville, 1843 (Maldonado 1976) and Sphedanolestes zhengi Zhao, Ren, Wang & Cai, 2015 (Zhao et al. 2015). However, it is clear from the present study that both the sticky and the wax-like substance may be absent when specimens are examined and described, and thus the extent of their existence may remain unknown. The presence or absence of the sticky glands that produce the sticky substance has been shown to correlate with the phylogeny of Harpactorini and faster rates of evolution among “sticky bugs”, in comparison with other reduviids (Zhang and Weirauch 2013b). Similarly, records of the presence or absence of a wax-like substance may turn out to be an additional feature of systematic or taxonomic importance, in the same way as suggested for the “extensive sericeous areas on the abdominal sterna” of Heza ventralis Stål, 1872 (Maldonado 1976). Therefore, future studies on Harpactorini should include careful handling of the specimens after collection, to avoid unintentional removal of these substances from their bodies. It is also recommended that this information should be included in the records and/or descriptions whenever present.

Differences in body size and in the structure and vestiture of the third antennal segment were clear-cut enough to be considered as sexually dimorphic in P. salgadoi. Despite the small number of measurements obtained for the third antennal segments (because they were broken and lost in some specimens), adults can be sexed readily with the naked eye, by observing these two traits. Firstly, females were larger than males in almost all the morphological characteristics measured. Secondly, the third antennal segment was longer in males than in females, both in absolute and in relative terms (i.e. total length and the ratio between the third and first antennal segments, respectively). Moreover, in males, the third antennal segment was distinctly thickened, approximately in its basal two-thirds, with the thicker portion clearly separated in relation to the thinner, distal portion (Figs 1, 5, 9, 11–12). In females, it showed little or no thickening (Figs 42, 47–48, 51). Two of the sexual differences pointed out in P. salgadoi (i.e. females larger than males and the latter with third antennal segment thickened) are concordant with several observations in the literature (Champion 1899, Martin-Park et al. 2012, Gil-Santana et al. 2013, Gil-Santana 2016). Additionally, the thickened portion of the third antennal segment in males was completely covered by very short, stiff, adpressed, blackish setae (Figs 9–10, 12), which were absent in females (Fig. 51). Small differences in coloration were also recorded (see above); however, more specimens need to be examined in order to clarify whether such differences might be due to sexual dimorphism.

Some slight differences were also noted between the type specimens and the specimens from São Paulo. These are very probably due to intraspecific variation. However, measurements of the head and the ratio of head length to maximum width across the eyes are important for separating Parahiranetis from related genera, especially Hiranetis (Gil-Santana 2015, 2016). Gil-Santana (2015) measured the length of the head including the neck and recorded that it was 1.6 to 1.8 times as long as width across the eyes (in the paratype and holotype, respectively). In the present study, in which head measurements excluded the neck, the head was recorded in the type specimens as being 1.47 to 1.65 times as long as width across the eyes (paratype/holotype). In parallel, this ratio was approximately 1.3 to 1.4 in the specimens from São Paulo. Therefore, the main diagnostic characteristics that separate Hiranetis, Graptocleptes, and Parahiranetis according to Spinola (1840a,b), Stål (1866, 1872), Maldonado and Lozada (1992), Gil-Santana et al. (2013), Gil-Santana (2015, 2016) and this work, are the following:

1) Graptocleptes: Head elongate, approximately 1.1 to 1.3 times as long as width across eyes, sparsely setose; legs thicker; fore femur shorter than head and pronotum together and of uniform thickness.

2) Hiranetis: Head gibbous, large, approximately as long as width across eyes, densely setose on ventral and post-ocular portions; legs elongated, slender; fore femur subequally longer than head and pronotum together, thicker basally.

3) Parahiranetis: Head elongate, approximately 1.3 times to 1.7 as long as width across eyes, densely setose on ventral and post-ocular portions; legs elongated, slender; fore femur subequally longer than head and pronotum together, thicker basally.

The male genitalia of P. salgadoi showed some similarities to those of G. bicolor (Gil-Santana et al. 2013) and H. atra (Gil-Santana 2016), such as: (i) parameres similar in shape and somewhat similar in vestiture; and (ii) presence of a pair of elongate, parallel, flat, weakly sclerotized endosomal processes. However, the latter were recorded with different locations and shapes in each of these species. Comparatively, they are somewhat more elongated, straighter and closer to each other in P. salgadoi (Figs 26–28, 31–33, 38, 41). The pygophores of all three species have a somewhat large medial process that is medially rounded at the apex, but in G. bicolor and H. atra it is subtriangular in shape, while in P. salgadoi, it is somewhat arrow-shaped, with the lateral margins acutely pointed (Figs 22, 24, 29). U-shaped and median subspherical endosomal processes that are very similar to those of P. salgadoi (Figs 33, 38–40) were recorded in H. atra. The distal third of the endosoma of P. salgadoi, except in the median portion, is almost completely covered by numerous larger spiny somewhat sclerotized processes (Fig. 37), while in the endosoma of G. bicolor, groups of larger similar spiny processes were also recorded. However, in the latter species, these processes were recorded as occurring mostly at the midportion, apically or subapically, and were more sclerotized and clearly triangularly shaped.

Moreover, the male genitalia of P. salgadoi differ from those of both of those species in some parts. Firstly, the basal plate arms of the articulatory apparatus are clearly longer in this species (Figs 33–34). Secondly, there is a very remarkable difference concerning the struts and especially the dorsal phallothecal plate. The latter has already been recorded as distinctive between G. bicolor and H. atra, but the differences are even more accentuated in P. salgadoi. The shape of the dorsal phallothecal plate and the portions of it (including the lateral large lobes) that are covered with very numerous spiny processes in the middle third, along with the racket-shaped flat sclerite extending towards the apex with lateral large spines, are striking features and may be unique for this species (Figs 30–33, 35–36). Thus, in agreement with previous studies (Elkins 1954a, b, Hart 1975, 1986, 1987, Forero et al. 2008, Zhang et al. 2016), the features of the male genitalia of P. salgadoi that should especially be taken into consideration for comparative purposes are the shape of the medial process of the pygophore and the features of the phallothecal plate.

Historically, only the pattern of yellowish or straw-colored hemelytra with a median transverse black band has received attention in regards to the supposed mimicry between Harpactorini and species of hymenopteran Ichneumonidae and Braconidae (Champion 1899, Haviland 1931, Maldonado and Lozada 1992, Hogue 1993, Leathers and Sharkey 2003, Hespenheide 2010). However, some species that have similar patterns of blackish wings with yellowish ‘pterostigmata’ arise as possible candidates for mimetic complexes, including P. salgadoi (Gil-Santana 2015). These are the wasp-mimicking harpactorines Graptocleptes bicolor, G. haematogaster (Stål, 1860), and an undescribed species of Hiranetis as well as the ichneumon wasp Pedinopelte gravenstii (Guérin-Méneville, 1826) and the cerambycid beetle Isthmiade braconides (Perty, 1832) (Fig. 64). All of these have been recorded from southeastern Brazil, particularly in the states of Rio de Janeiro and/or São Paulo (harpactorines: Stål 1860, 1872, Gil-Santana et al. 2013, Gil-Santana 2015; ichneumon wasp: De Santis 1980; cerambycid beetle: Zikán and Zikán 1944, Zajciw 1972, Monné 1993, Monné and Bezark 2009).

Figures 63–64. 

63 Parahiranetis salgadoi, live specimens; two male adults and fifth instar nymphs on a tree trunk; red arrows point to nymphs 64 female holotype of Isthmiade braconides (Perty, 1832) (Coleoptera: Cerambycidae: Cerambycinae: Rhinotragini), deposited in the ZSM, Munich, Germany (courtesy of Steve Lingafelter).

Although there are records of color variation in some wasp-mimicking Harpactorini, at least in the species with the pattern of darkened or blackish hemelytra with yellowish pterostigmata, there is no variation in this pattern. The yellowish pterostigmata is always present (e.g. Gil-Santana et al. 2013, this study). Concerning this point, it is appropriate to correct an assertion by Gil-Santana et al. (2013). In the latter report, it was alleged that Stål (1872) would have “recorded variation of the coloration of the corium of hemelytra of Graptocleptes bicolor, describing it as either paler (“Var. b. - pallidior”) or completely blackish (“bb. Corio toto nigricante”)”. This was a double misinterpretation of Stål’s text. The statement “Var. b. - pallidior” did indeed refer to G. bicolor, but it was in relation to a “variety” with paler general coloration, and this did not necessarily relate to its hemelytra. Regarding the second assertion, an accurate reading of Stål’s paper makes it clear that “Corio toto nigricante” was considered to be a characteristic of another species: Graptocleptes sanguineiventris Stål, 1862. Moreover, just above this, Stål (1872) included the previous alternative “b.”, through which the characteristic of “Corio apice vel prope apicem pallescente vel croceo” was attributed to another species (G. gastricus (Stål, 1860), G. haematogaster and G. bicolor). This is a clear description of the pale to yellowish spot on the distal portion of the corium of the hemelytra in all these three species. Therefore, the alternative “bb.” (corium completely blackish) was clearly stated as opposite to this latter and was diagnostic of another species (G. sanguineiventris).

In contrast to the adults, nymphs of P. salgadoi were quite inconspicuous on the tree trunks where they naturally occurred (Figs 59–63). Their cryptic coloration probably has a protective function, which is present to some extent in almost every family of Heteroptera (Schuh and Slater 1995). Ecological studies might improve the current understanding of the contrasting strategies adopted by different life stages in P. salgadoi. Additionally, as emphasized by Gil-Santana (2015), it is necessary to elucidate which species or groups of insects sharing the same color pattern as in P. salgadoi (i.e. blackish to reddish coloration with yellowish ‘pterostigmata’ on wings and/or yellowish markings on legs) are involved in possible mimicry complexes.

We thank the staff of the Base Ecológica da Serra do Japi for their logistic support, and Steve Lingafelter for the photo of the holotype of Isthmiade braconides presented here as Fig. 64. The third author (JO) thanks João Aristeu da Rosa (UNESP/FCFAR) for all his support and teaching of MEV techniques. We are also very grateful to Leonidas-Romanos Davranoglou (University of Oxford, United Kingdom), Tadashi Ishikawa (Tokyo University of Agriculture, Japan), Daniel R. Swanson (University of Illinois at Urbana-Champaign, USA) and Guanyang Zhang (Arizona State University, USA), for their valuable comments and suggestions.