Even though female genitalia structures can be observed by OM, SEM has several advantages over optical. For this purpose it was possible to view the three dimensional external shape of a structure (in this case the female genitalia) in the same image. Electron microscopy also allows us to focus on many details of the structure. After checking them by OM, observations described the species specific features, and discarded the polymorphic ones.
Among the 18 Rhodnius species [2, 12, 13], the female external genitalia for R. amazonicus, R. dalessandroi, R. paraensis and R. zeledoni were not studied because specimens were unavailable.
The relevance of the male genitalia in specific identification of triatomines has beeen widely used by many authors, including Lent and Jurberg [18] to describe new species. On the other hand the female genitalia, also evaluated by other authors [6, 9], has shown itself less valuable due to the hard task of the dissection technique, unlike the male genitalia. The first publication in 2010 using SEM involving non-dissected insects of non-closely related triatomine species (P. megistus, P. herreri, R. prolixus, R. colombiensis, T. infestans and T. vitticeps), clearly showed that these species can be distinguished by this features [19]. However, in 2012 the validity of this approach was confirmed for closely related species (R. robustus and R. montenegrensis) [13]. In an unpublished masters thesis, Simone Caldas Neves used the same approach to distinguish a recently described species (T. jatai[23], in the theses called T. n.sp.) closely related to T. costalimai[24].
Given the difficulties for specific distinctions in the Rhodnius species [4, 9, 12] this study was performed to increase the number of morphological traits that can be used to identify the species in this genus.
For this study, details previously published for external female genitalia traits in four species were reconsidered; the species included were R. colombiensis, R. prolixus[19], R. montenegrensis and R. robustus[13].
The key presented and summarized in Table 2 was developed using the most evident traits to identify the 14 Rhodnius species using the external female genitalia. Three groups of species (a, b, and c) were formed according to characteristics of the dividing line between the seventh and eighth dorsal segments, which is a visible and perceptible feature (Figure 1A-N). However, given the information verified by the 42 figures, the 14 species can be identified using traits on the dorsal, posterior, and ventral sides as either isolated or associated characteristics.
In the key, R. montenegrensis and R. robustus separation is based on the dorsal, posterior, and ventral sides (Figures 1I,J, 2I,J, 3I, and J). Though eight species from group a and four species from group c were characterized based on their posterior and ventral sides, it is important to note that these species also include characteristics on the dorsal side, as with R. nasutus and R. neivai, which have distinct tenth segment shapes (Figure 1D,E).
Thus, the dorsal side shows that the eighth, ninth and tenth segments are also distinct among the 14 species evaluated (Figure 1A-N).
Based on a posterior perspective, the 14 species can be distinguished by the dimension and shape of the eighth gonocoxites and gonapophyses, the ninth and tenth segments as well as the dividing lines between the eighth and ninth as well as the ninth and tenth segments. This perspective shows that the posterior portion of the tenth segment is concave at the end in 11 species (Figure 2B-E,G,I-N); in R. pictipes and R. stali, this segment is straight at the end (Figure 2F,H), while in R. domesticus, it is semi- circular (Figure 2A).
A ventral perspective shows that the 14 species have distinctive lines at the end of the seventh segment as well as shapes and dimensions for the gonapophyses, the gonocoxites, as well as the ninth and tenth segments. From this perspective, 10 species have a dividing line between the seventh segment and eighth gonocoxites as well as gonapophyses, which is curved at the sides and convex in the middle (Figure 3C-H,J,L-N). In the remaining four species, this line is curved (Figure 3A,B,J, and K). In six species, the eighth gonocoxites meet in a triangular shape (Figure 3A,B,G, and I-K); in the other eight species, this point is non-triangular (Figure 3C-F, H, and L-N). The eighth gonapophyses are triangular in 12 species (Figure 3A,E,G, and I-N) and rod-shaped in two species (Figure 3F,H).
After combining the results herein on the external female genitalia for these 14 Rhodnius species using the five complexes established by Carcavallo et al. [20] for this genus, the following factors can be considered.
A comparative analysis of the external female genitalia characteristics for the R. dalessandroi complex was impossible because only R. milesi specimens were examined.
Rhodnius pictipes and R. stali, which compose the R. pictipes complex, include traits that join and exclude similar species. These traits include a dividing line between the seventh and eighth segments from the dorsal perspective, rectangular shape of the tenth segment from the posterior perspective, as well as ninth and tenth segment features (Figures 1F,H, 2F, and H). From the ventral perspective, similarities were observed for the line that divides the seventh segment from the eighth gonocoxites and gonapophyses as well as the shape of the eighth gonocoxites and tenth segments (Figures 1F,H, 2F,H, 3F, and H). However, each of these two species maintained distinguishing characteristics. From the dorsal perspective at the intersection between the lines that separate the seventh segment and connectives, R. pictipes include 1 + 1 triangles that are not evident in R. stali. The dorsal perspective also shows that R. stali comprise salient and straight lateral 1 + 1 edges on the ninth segment, while in R. pictipes, these edges are wide (Figure 1F,H). From the ventral perspective, this difference is clarified by the shape of the ninth segment end portion, which is wide in R. pictipes and straight in R. stali (Figure 3F,H.)
For the R. prolixus complex, which comprises R. prolixus, R. domesticus, R. nasutus, R. neglectus, R. robustus, and, more recently, R. montenegrensis, the analyses confirmed that the connecting traits include the shape of the ventral line that divides the seventh segment from the eighth gonocoxites and gonapophyses, except for R. montenegrensis. The species include distinguishing traits on the three sides studied. A distinct trait was verified for R. domesticus, which was the only species of the 14 that does not include transversely grooved cuticles (Figure 1A). Importantly, R. neglectus and R. prolixus, which are particularly difficult to distinguish, do have distinguishing characteristics on the dorsal side; the line that separates the seventh and eighth segments is completely curved in R. prolixus, while in R. neglectus, it is curved on the sides and lightly convex in the middle (Figure 1G,M). On the posterior side, the difference between these two species evident through the dividing line between the ninth and tenth segments, which is circular in R. prolixus and oval at the beginning in R. neglectus but widens on the posterior sides (Figure 2G,M). The ventral side of R. prolixus shows a laterally expanded ninth segment, which was not observed in R. neglectus (Figures 3G,M).
Rhodnius colombiensis, R. ecuadoriensis and R. pallescens, which comprise the R. pallescens complex, include a set of ninth and tenth segments on the posterior side that form an isosceles triangle (Figure 2B,L, and N). This configuration clearly connects these three species. The distinction between such species is particularly evident on the ventral side, where the posterior portion of the eighth gonocoxites and gonapophyses is projected on the ninth segment in R. ecuadoriensis; however, the gonocoxite is not projected in R. colombiensis. This structure has a different shape in the two species (Figure 3B,L). In R. pallescens, the dividing line that separates the seventh segment from the eighth gonocoxites and gonapophyses is elevated in the middle; the elongated shape of the ninth and tenth segments also distinguishes this species from R. colombiensis and R. ecuadoriensis (Figure 3N,L, and B).
The studies performed herein facilitate identification and descriptions for three distinguishing features on the dorsal side, four features on the posterior side and five features on the ventral side of the external female genitalia in the 14 Rhodnius species, which distinguish the species. Using such findings, a key was developed to aid in distinguishing the 14 Rhodnius species. There is no record in the literature for intraspecific polymorphism for the female genitalia; however, this characteristic has not been widely explored. Given that, we do not discard the possibility of intraspecific variation for field populations and some of the features elected shall be confirmed with a much larger sample of field material. Therefore, we consider the possibility of further adjustments in the key. On the other hand, all descriptive studies in the morphology field also have this limitation. For triatomines, morphological variations, not described before, were detected for Rhodnius nasutus in Ceara [25], and Triatoma rubrovaria in Rio Grande do Sul [26].
The morphological traits for the external female genitalia validate four of the five complexes proposed by Carcavallo et al. [20] for the genus Rhodnius. They also specifically distinguish the species in the complexes. The fifth complex proposed by Carcavallo et al.[20], which also includes the species R. milesi was impossible to evaluate due to a lack of R. dalessandroi specimens.
It is important to highlight that the results herein corroborate the observations in Rosa et al. [19], which considered these traits for P. herreri, P. megistus, R. colombiensis, R. prolixus, T. infestans and T. vitticeps and validated the taxonomy. However, the study herein is a continuation of the study on the posterior ventral abdominal segments for the female nymphs in the fifth nymphal stage from six Triatominae species in Rosa et al.[27], wherein specifically distinguishing characteristics were identified.
Finally, it will necessary to clarify whether the features of the female genitalia described by SEM may also be observed through OM by non-experts. In addition, given the possible polymorphism for field populations we recommend: (i) first using to the traditional classification and then (ii) make use of this study for the specific confirmation.
