Findings

DNA barcoding, as proposed by Hebert et al. [1] assumes that a biological entity is completely separated from its closest relatives by a barcoding gap[2], which means that intraspecific genetic distances (from COI sequences) are never greater than interspecific distances.

Triatoma Laporte (Hemiptera: Reduviidae) is the most diverse genus of Chagas Disease vectors, and accurate identification of species is imperative for the efficiency of vector control programs. The Triatoma genus is divided into species complexes and subcomplexes according to geographic distribution and morphological similarity [3].

Recently, Justi et al. [4] reported that the relationships between species assigned to South American Triatoma subcomplexes could not be untangled with the data in hand. We were then prompted to investigate whether DNA barcoding would be a useful tool for identifying the species within the infestans, matogrossensis, sordida and rubrovaria subcomplexes [3].

Kimura-2-parameter genetic distances [5] were calculated pairwise within each of the above mentionedsubcomplexes (Table 1) using the software MEGA v. 5 [6], and intra and interspecific distances were compared.

Table 1 K2p-distances between species of the Triatoma subcomplexes studied
Full size table

In all subcomplexes we observed at least one intraspecific distance greater than interspecific distances (Table 1). To be considered appropriate to identify species within a group, intraspecific distances must always be greater than interspecific ones [2], and therefore DNA barcoding is not accurate for the species-level identification of South American Triatoma. Moreover, the method fails to account for hybridization events, which are naturally observed in Triatoma[7],[8], and introgression, which is frequent in nuclear DNA [9]. These considerations argue that Hebert et al.’s [1] proposal of cataloguing biodiversity based only on DNA barcoding may severely underestimate it.

Besides that, as highlighted by Dujardin et al.[10], the morphological changes observed in closely related “species”, or “lineages” as we prefer to call them, may have led taxonomists to rush into describing subspecies or species, even genera. Molecular phylogenetic studies are in their infancy in unravelling the evolution of Triatominae, and a comprehensive molecular phylogeny, including more than one specimen for most lineages, was published only in 2014 [4], although several analyses were conducted focusing on small species groups. Taken together, these statements make it clear that further investigations of Triatominae evolution are long overdue, preferably integrating morphological, molecular and ecological data.

Lineage evolution has not occurred, but it is happening now. Concerning lineages designated in the infestans complex (including the subcomplexes studied here), separation is much clearer in terms of morphology than in molecular systematics. In cases where lineages have not reached reciprocal monophyly, defining taxonomic entities is not a straightforward issue [11]. Therefore caution is necessary, especially in a group where accurate identification of taxa is fundamentally linked to public health issues.

Conclusions

Although DNA barcoding is a straightforward approach, it was not applicable for identifying Southern American Triatoma species, which may have diverged recently. Thus, caution should be taken in identifying vector species using this approach, especially in groups where accurate identification of taxa is fundamentally linked to public health issues.

Authors’ information

CD is PhD student funded by Instituto Oswaldo Cruz and SAJ is a Post Doctoral Fellow funded by CNPq.