Fig 1.
Conceptual diagram of the triatomine microcosm in Amazonian palm trees across a sylvatic-domestic gradient.
Palm crowns provide microhabitats that support sylvatic vertebrate hosts and palm-dwelling triatomine colonies. The diagram depicts three palm-based landscapes along a deforestation and human-use gradient: forested landscapes, where triatomines are primarily associated with T. rangeli; human-modified landscapes, where both T. rangeli and T. cruzi may occur; and highly anthropized landscapes near human dwellings, where T. cruzi can predominate. The figure was created by the authors using Inkscape v1.4, with icons adapted from the Noun Project.
Fig 2.
Study area and landscape context of triatomine sampling in Cruzeiro do Sul, Acre, Brazil.
Land-use and land-cover map showing forested (green) and anthropized areas (yellow). Circles represent the 22 landscape units surveyed for triatomines. Solid black circles indicate landscapes sampled in 2022; solid pink circles indicate landscapes sampled in 2024, and combined black-and-pink circles indicate landscapes sampled in both 2022 and 2024. The map was created by the authors using ArcMap 10.8.2 based on open-access datasets from IBGE (administrative boundaries) and MapBiomas (land-use and land-cover), along with field-collected GPS data.
Table 1.
Number of triatomines collected per landscape site and year, infection with T. cruzi or T. rangeli, and associated landscape characteristics, Cruzeiro do Sul, Acre state, Brazil.
Fig 3.
Fine-scale landscape structure and palm-household distances at representative sampling sites in Cruzeiro do Sul, Acre, Brazil.
Panels show four sites across a gradient of forest cover within 3 km2 landscape units: (A) A14 (~65% forest cover), where only T. rangeli-infected triatomines were detected; (B) A20 (~20% forest cover), with both T. rangeli and T. cruzi infections; and (C–D) A5 and A7 (~10% forest cover), where only T. cruzi–infected triatomines were detected. Yellow palm icons indicate palm trees sampled in 2022, and pink icons indicate those sampled in 2024; selected households are shown in brown. Maps were produced by the authors in ArcMap 10.8.2 using orthomosaics generated from original drone imagery (DJI Mavic 3) collected during fieldwork and processed by the authors.
Table 2.
Effects of landscape structure on triatomine infection probability.
Fig 4.
Predicted probability of T. cruzi infection in Rhodnius across a gradient of forest cover (%) at short and long palm-household distances.
Short and far distances correspond to the 10th and 90th percentiles of palm-household distances (i.e., 53 and 301 meters, respectively). Solid lines represent posterior medians and dashed lines indicate 95% credible intervals.