Geological Changes of the Americas and their Influence on the Diversification of the Neotropical Kissing Bugs (Hemiptera: Reduviidae: Triatominae)

Abstract

Background: The family Reduviidae (Hemiptera: Heteroptera), or assassin bugs, is among the most diverse families of the true bugs, with more than 6,000 species. The subfamily Triatominae (kissing bugs) is noteworthy not simply because it is the only subfamily of the Reduviidae whose members feed on vertebrate blood but particularly because all 147 known members of the subfamily are potential Chagas disease vectors. Due to the epidemiological relevance of these species and the lack of an efficient treatment and vaccine for Chagas disease, it is more common to find evolutionary studies focusing on the most relevant vectors than it is to find studies aiming to understand the evolution of the group as a whole. We present the first comprehensive phylogenetic study aiming to understand the events that led to the diversification of the Triatominae.

Methodology/principal findings: We gathered the most diverse samples of Reduviidae and Triatominae (a total of 229 Reduviidae samples, including 70 Triatominae species) and reconstructed a robust dated phylogeny with several fossil (Reduviidae and Triatominae) calibrations. Based on this information, the possible role of geological events in several of the major cladogenetic events within Triatominae was tested for the first time. We were able to not only correlate the geological changes in the Neotropics with Triatominae evolution but also add to an old discussion: Triatominae monophyly vs. paraphyly.

Conclusions/significance: We found that most of the diversification events observed within the Rhodniini and Triatomini tribes are closely linked to the climatic and geological changes caused by the Andean uplift in South America and that variations in sea levels in North America also played a role in the diversification of the species of Triatoma in that region.

Fig 1. Bayesian phylogeny obtained with six fossil calibrations (B2).

Numbers above branches indicate posterior probabilities greater than 0.50. Red branches indicate the cladogenetic events tested based on the literature. Blue branches indicate vicariant events recovered in our analyses and tested. Green arrow indicates the Triatominae fossil calibration used.

Fig 2. Time estimates and linear regression for the three Bayesian estimates.

Red dots and line refer to the estimates using one fossil calibration (B1); blue dots and line refer to the estimates using six fossil calibrations (B2); black dots and line refer to the estimates using seven fossil calibrations (B3); Y axis represents the width (delta time interval) of the interval estimated and X axis the mean age estimated for each node in the phylogeny.

Fig 3. Vicariant events (black and red horizontal lines) plotted against posterior distribution (95% HPD) estimated for the nodes.

(a) pictipes group / prolixus group–Pebas System; (b) R. neivai / R. domesticus–Acre System; (c) venosa clade / other Triatomini–Western Cordillera uplift; (d) T. venosa / Triatoma sp. 2 –last Andean uplift; (e) T. maculata / infestans group–Pebas System (red horizontal line) and Acre System (black horizontal line); (f) T. bruneri / megistus group–GAARlandia (red horizontal line) and rafting with Capromyinae (black horizontal line); (g) T. tibiamaculata / P. megistus–old Amazonian-Atlantic Forest pathway; (h) (Linshcosteus sp. + T. rubrofasciata) / phyllosoma group–Bering Land Bridge; (i) T. sanguisuga / phyllosoma group (part)–high sea level. Red distribution line refers to the estimates using one fossil calibration (B1); blue distribution line refers to the estimates using six fossil calibrations (B2); black distribution line refers to the estimates using seven fossil calibrations (B3). The vicariant events tested and the corresponding geological events are the following: (a) pictipes group / prolixus group–Pebas System; (b) R. neivai / R. domesticus–Acre System; (c) venosa clade / other Triatomini–Western Cordillera uplift; (d) T. venosa / Triatoma sp. 2 –last Andean uplift; (e) T. maculata / infestans group–Pebas System and Acre System; (f) T. bruneri / megistus group–GAARlandia and Rafting with Capromyinae; (g) T. tibiamaculata / P. megistus–old Amazonia-Atlantic Forest pathway; (h) (Linshcosteus sp. + T. rubrofasciata) / phyllosoma group–Bering Land Bridge; (i) T. sanguisuga / phyllosoma group (part)–high sea level (Fig 3a–3i and Table 3).

Fig 4. Other hypothesis testing.

Events (black and red horizontal lines) plotted against posterior distribution (95% HPD) estimated for the nodes. Red distribution line refers to the estimates using one fossil calibration (B1); blue distribution line refers to the estimates using six fossil calibrations (B2); black distribution line refers to the estimates using seven fossil calibrations (B3); (a) Rhodniini MRCA–Pebas System; (b) pallescens group MRCA–Isthmus of Panama (red horizontal line) and last Andean uplift (Pliocene—black horizontal line); (c) (spinolai clade + geniculatus clade + rubrofasciata clade) estimated ages—Fossils Triatoma dominicana and Panstrongylus hispaniolae ages (black horizontal line) and GAARlandia land bridge (red horizontal line).