Ecologic niche modeling and potential reservoirs for Chagas disease, Mexico

Abstract

Ecologic niche modeling may improve our understanding of epidemiologically relevant vector and parasite-reservoir distributions. We used this tool to identify host relationships of Triatoma species implicated in transmission of Chagas disease. Associations have been documented between the protracta complex (Triatoma: Triatominae: Reduviidae) with packrat species (Neotoma spp.), providing an excellent case study for the broader challenge of developing hypotheses of association. Species pairs that were identified coincided exactly with those in previous studies, suggesting that local interactions between Triatoma and Neotoma species and subspecies have implications at a geographic level. Nothing is known about sylvatic associates of T. barberi, which are considered the primary Chagas vector in Mexico; its geographic distribution coincided closely with that of N. mexicana, suggesting interaction. The presence of this species was confirmed in two regions where it had been predicted but not previously collected. This approach may help in identifying Chagas disease risk areas, planning vector-control strategies, and exploring parasite-reservoir associations for other emerging diseases.

Figure 1

Modeled geographic distributions of Neotoma woodrats in mainland Mexico. (A) black = Neotoma fuscipes, red = N. phenax, green = N. mexicana (note that this distribution includes those of other, less widely distributed, species), yellow = N. goldmani, blue = N. palatina, orange = N. angustapalata. (B) red = N. albigula, yellow = N. micropus, green = N. albigula and N. micropus.

Figure 2

Modeled geographic distributions of members of the protracta species complex: red = Triatoma barberi, yellow = T. p. woodi, green = T. sinaloensis, blue = T. p. protracta, black = T. peninsularis, and pink = T. p. zacatecensis. Only areas predicted for each species at the highest level of confidence (all best-subsets models agree) are shown.

Figure 3

Modeled geographic distributions of Triatoma peninsularis (red) and Neotoma lepida (black), showing the tight geographic correspondence between the distribution of insect and host mammal. Almost all (93.8%) of the distribution area of T. peninsularis overlaps the distribution area of N. lepida.

Figure 4

Frequency distribution of areas of overlap between Triatoma and Neotoma species, expressed as percent of total modeled area (at highest predictive level) of the Triatoma.

Figure 5

Modeled geographic distribution of Triatoma barberi, shown with known occurrence points used to create and test the ecologic niche model. Dark red = 100% of best-subsets models predict presence, medium red = 75% of best-subsets models predict presence, light red = 50% of best-subsets models predict presence, lightest red = any best-subsets model predicts presence.