Distribution and natural infection status of synantrophic triatomines (Hemiptera: Reduviidae), vectors of Trypanosoma cruzi, reveals new epidemiological scenarios for chagas disease in the Highlands of Colombia

Introduction Updating the distribution and natural infection status of triatomine bugs is critical for planning, prioritizing, and implementing strategies to control Chagas disease (CD), especially after vector reduction programs. After carrying out a control program, the Department of Boyaca contains the highest number of Colombian municipalities certified by PAHO to be free of intradomiciliary transmission of Trypanosoma cruzi by Rhodnius prolixus. The present work describes the spatial distribution, natural infection (NI), and molecular characterization of T. cruzi in synanthropic triatomines from the Department of Boyaca in 2017 and 2018. Materials and methods An entomological survey was conducted in 52 municipalities in Boyaca known to have had previous infestations of triatomine bugs. Insects were collected through active searches carried out by technical personnel from the Secretary of Health and community members using Triatomine Collection Stations (PITs-acronym in Spanish). For evaluation of natural infection, triatomines were identified morphologically and grouped in pools of one to five individuals of the same species collected in the same household. DNA derived from the feces of each pool of insects was analyzed by PCR for the presence of T. cruzi using primers flanking the satellite DNA of the parasite. SL-IR primers were used to differentiate TCI from the other DTUs and to identify different genotypes. The distribution of the collected triatomines was analyzed to determine any vector hotspots using spatial recreation. Results A total of 670 triatomine bugs was collected, belonging to five species: Triatoma dimidiata (73.2%), Triatoma venosa (16.7%), Panstrongylus geniculatus (5.7%), Rhodnius prolixus (4.4%), and Panstrongylus rufotuberculatus (0.4%), from 29 of the 52 municipalities. In total, 71.6% of the bugs were collected within houses (intradomiciliary) and the rest around the houses (peridomiciliary). Triatoma dimidiata was the most widely distributed species and had the highest natural infection index (37.8%), followed by T. venosa and P. geniculatus. TcI was the only DTU found, with the TcI Dom genotype identified in 80% of positive samples and TcI sylvatic in the other insects. Spatial analysis showed clusters of T. dimidiata and T. venosa in the northeast and southwest regions of Boyaca. Conclusions After some municipalities were certified free of natural transmission within houses (intradomiciliary transmission) of T. cruzi by R. prolixus, T. dimidiata has become the most prevalent vector present, and represents a significant risk of resurgent CD transmission. However, T. venosa, P. geniculatus, and P. rufotuberculatus also contribute to the increased risk of transmission. The presence of residual R. prolixus may undo the successes achieved through vector elimination programs. The molecular and spatial analysis used here allows us to identify areas with an ongoing threat of parasite transmission and improve entomological surveillance strategies.


Introduction
Updating the distribution and natural infection status of triatomine bugs is critical for planning, prioritizing, and implementing strategies to control Chagas disease (CD), especially after vector reduction programs. After carrying out a control program, the Department of Boyaca contains the highest number of Colombian municipalities certified by PAHO to be free of intradomiciliary transmission of Trypanosoma cruzi by Rhodnius prolixus. The present work describes the spatial distribution, natural infection (NI), and molecular characterization of T. cruzi in synanthropic triatomines from the Department of Boyaca in 2017 and 2018.

Materials and methods
An entomological survey was conducted in 52 municipalities in Boyaca known to have had previous infestations of triatomine bugs. Insects were collected through active searches carried out by technical personnel from the Secretary of Health and community members using Triatomine Collection Stations (PITs-acronym in Spanish). For evaluation of natural infection, triatomines were identified morphologically and grouped in pools of one to five individuals of the same species collected in the same household. DNA derived from the feces of each pool of insects was analyzed by PCR for the presence of T. cruzi using primers flanking the satellite DNA of the parasite. SL-IR primers were used to differentiate TCI from the other DTUs and to identify different genotypes. The distribution of the collected triatomines was analyzed to determine any vector hotspots using spatial recreation. PLOS

Results
A total of 670 triatomine bugs was collected, belonging to five species: Triatoma dimidiata (73.2%), Triatoma venosa (16.7%), Panstrongylus geniculatus (5.7%), Rhodnius prolixus (4.4%), and Panstrongylus rufotuberculatus (0.4%), from 29 of the 52 municipalities. In total, 71.6% of the bugs were collected within houses (intradomiciliary) and the rest around the houses (peridomiciliary). Triatoma dimidiata was the most widely distributed species and had the highest natural infection index (37.8%), followed by T. venosa and P. geniculatus. TcI was the only DTU found, with the TcI Dom genotype identified in 80% of positive samples and TcI sylvatic in the other insects. Spatial analysis showed clusters of T. dimidiata and T. venosa in the northeast and southwest regions of Boyaca.

Conclusions
After some municipalities were certified free of natural transmission within houses (intradomiciliary transmission) of T. cruzi by R. prolixus, T. dimidiata has become the most prevalent vector present, and represents a significant risk of resurgent CD transmission. However, T. venosa, P. geniculatus, and P. rufotuberculatus also contribute to the increased risk of transmission. The presence of residual R. prolixus may undo the successes achieved through vector elimination programs. The molecular and spatial analysis used here allows us to identify areas with an ongoing threat of parasite transmission and improve entomological surveillance strategies.

Author summary
Chagas disease is one of the most important tropical diseases in the Americas. Since 2010, Colombia has implemented programs to interrupt the intradomiciliary transmission of T. cruzi by R. prolixus, Colombia's primary vector. Boyaca, located in this country's central region, is one of the most endemic departments for Chagas disease. Control measures have been implemented, and the intradomiciliary transmission of T. cruzi by R. prolixus has been significantly reduced in 24 municipalities, according to PAHO certifications. Currently, the main risks in these certified municipalities are the presence of secondary vectors and certain ecological conditions favorable to triatomines. In the present study, we provide evidence that T. dimidiata and T. venosa have become the most common species. We observed a cluster of these infected species in the northeast and southwest regions of Boyaca. P. geniculatus and P. rufotuberculatus were present but less abundant, and residual R. prolixus populations remain in some certified-free municipalities. The overall natural infection index of secondary vectors was 31%, and spatial analysis identified priority areas for implementing surveillance and control actions. We identified T. cruzi in municipalities above 2,000 meters above sea level, which traditionally are considered nonendemic regions in Colombia. This suggested new epidemiological scenarios for Chagas disease transmission in Colombia's highlands and requires more studies to examine these scenarios.

Introduction
Triatomines (Hemiptera: Reduviidae) are hematophagous insects that play an essential role as vectors of Trypanosoma cruzi, the causative agent of Chagas Disease (CD), which is considered the most important anthropozoonotic infection in Latin America [1]. The parasite T. cruzi presents tremendous genetic diversity and has been divided into six discrete typing units (DTUs, TcI to TcVI) [2], and other genotypes found in bats designed as TcBat [3]. Although the circulation of all DTUs has been described in Colombia, TcI is the most widely distributed DTU in this country [4], and which is then subdivided into two major genotypes associated with transmission cycles; domestic (TcI Dom) and sylvatic (TcI sylvatic) [5,6]. The subfamily Triatominae comprises five tribes, 18 genera, and 154 described species, of which the genera Triatoma, Rhodnius, and Panstrongylus contain the main vectors of T. cruzi to humans [7]. In Colombia, 26 triatomine species have been reported, and Rhodnius prolixus is considered the most important vector due to its infestation indices, high natural prevalence of infection and vectorial capacity [8,9], and its preference to live in human houses. This species has been the main target of vector control programs in Colombia, whose efforts have resulted in the interruption of within household (intradomiciliary) transmission of T. cruzi by R. prolixus in 63 municipalities in 6 departments [9][10][11]. However, other invasive species of the genus Triatoma and Panstrongylus have gained importance in recent years due to their presence in homes, natural infection rates with T. cruzi, and incrimination as vectors [12][13][14].
Colombia has a wide distribution of triatomine bugs reported from 465 municipalities, 95% of which are located below 2,000 meters above sea level (masl), and which are considered endemic for Chagas disease [15,16]. However, T. cruzi-infected secondary species, such as Triatoma dimidiata and Panstrongylus geniculatus, have been found in municipalities from not endemic areas in the departments of Boyaca, Santander, and Cundinamarca suggesting changes in Chagas disease epidemiology [9,17]. Nonetheless, the entomological indices related to the risk of transmission are still unknown for these highland zones. A similar situation has been described in other Andean countries, such as Ecuador, Peru, and Bolivia, where the secondary and sylvatic vectors, Panstrongylus rufotuberculatus and P. geniculatus that were initially restricted to the Amazon basin have been reported to colonize zones located between 2,000 to 3,300 m.a.s.l. [18][19][20][21][22][23].
The department of Boyaca, located in the center of the country, on the eastern Andean region, has 24 municipalities certified free from intradomiciliary transmission of T. cruzi by R. prolixus, making it the department with the largest number of certified T. cruzi-free cities in Colombia [10,11]. Nine triatomine species have been described, in Boyaca, of which five (T. dimidiata, T. venosa, P. geniculatus, P. rufotuberculatus, and Rhodnius pictipes) have been found infected with T. cruzi. This represents a risk of transmission in the municipalities where they are found [8,17]. Recently, CD cases have been reported in areas of Boyaca that do not have R. prolixus and likely involves secondary species. These reports identify the need to update entomological surveillance to include these other vectors after the successes of interruption programs to reduce intradomiciliary transmission of CD [13]. The present study aimed to analyze the spatial distribution, natural infection (NI), and molecular characterization of T. cruzi in synanthropic triatomines from Boyaca between 2017 and 2018.
Triatomines were found in 21 of 26 municipalities using active searches, but were found in only 8 of 26 (30,7%) municipalities where PITS were installed (Table 1).
However, 19 of 24 municipalities that had been certified free of intradomiciliary transmission of T. cruzi by R. prolixus had triatomines (

Genotyping of T. cruzi
Only TcI was found, and both genotypes (TcI Dom and TcI sylvatic) were detected (S2A and S2B Fig). 80% (49/61) of the analyzed sequences belonged to TcI Dom and 20% (12/61) to TcI sylvatic. 85% of the sample sequences from T. dimidiata corresponded to TcI Dom, while the rest was TcI sylvatic. Regarding P. geniculatus, an equal number of samples were TcIDom and TcI sylvatic (50% (3/6) each), and the sequences obtained from P. rufotuberculatus belong to the TcI sylvatic. We did not find a mixture of both genotypes. Geographically, TcI sylvatic was only located in Soata, El Espino, and Miraflores, while TcI Dom was present in eight municipalities (S2A and S2B Fig).

Ethics statement
Ethical approval (Act No 113 of 2017) for analyzing animal species was obtained from the Antioquia University's animal ethics committee. All infested houses were sprayed with insecticide by SHB for the elimination of triatomine bugs.

Study area
Boyaca is located in the center of Colombia, in the Andes: 04˚39'10" and 07˚03'17" North, 715 7'49" and 74˚41'35" West. This department is made up of 123 municipalities, which are PLOS NEGLECTED TROPICAL DISEASES divided into 13 provinces: Centro, Gutierrez, Lengupa, Libertad, Marquez, Neira, Norte, Occidente, Oriente, Ricaurte, Sugamuxi, Tundama, and Valderrama. The topography of this area, shaped by the eastern mountains, determines the presence of diverse landscapes and climatic conditions, from highlands (paramos) to hot climate areas (located less than 500 m.a.s.l.). This department has a mean annual temperature range of 22˚C to 9.7˚C and a mean annual rainfall of 645 mm. This study was carried out between March 2017 and November 2018 in 52 municipalities in nine provinces with a prior history of infestation with triatomines, according to the Boyaca Department Health Service (BDHS) ( Table 1).

Triatomine collection and processing
The 52 municipalities with historical reports of triatomine bugs in the department of Boyaca were divided into two groups: the first consisted of 24 municipalities certified to be free of intradomiciliary transmission of T. cruzi by R. prolixus in 2010 and 2017, and those that during the study period were developing activities to obtain the certification in 2019 [10,11]. Cubara and Maripi municipalities were included in this group even though they were not certified but had recent reports of R. prolixus (Table 1). For this group, one entomological sampling per year was carried out through active searches in the urban and rural areas of each municipality by trained agents from the Boyaca Department Health Service (BDHS) vector control program following the National Protocols of Entomological Surveillance [17]. In brief, synanthropic triatomines were sought in indoor and outdoor ecotopes for 30 min each; flashlights were used to look into cracks and crevices throughout the fabric of buildings walls, behind pictures of the walls, behind furniture, in closets, and especially, under bedding material. In addition, BDHS health workers left plastic pots in which householders were asked to collect any triatomines that they found. These insects were brought by householders to the Triatomine Recollection Stations (PITs-acronym in Spanish) established by BDHS to monitor the presence of triatomines. A total of 20,529 households were visited once every year during the study period ( Table 1). The second group consisted of the 26 remaining municipalities, where only the regular community surveillance was carried out (Table 1). Insects were collected in intradomicile and peridomicile environments by residents who sent them to the PITs of each municipality. All entomological samples sent to the PITS from all the municipalities were placed inside plastic containers labeled with the date, municipality, and site of collection (peridomicile or intradomicile), and then transported to the departmental public health laboratory where insects were identified using taxonomic keys [24].

DNA extraction
Specimens were grouped into pools of one to five individuals based on their species and household of collection. For each pool, we dissected triatomines in a biological safety cabinet to avoid contamination. The feces were diluted in 300 μL of sterile PBS pH 7.2 and were used for DNA extractions. Genomic DNA was extracted from 200 μL of feces using the Genomic DNA purification kit (DNeasy Blood & Tissue kit Qiagen, Germantown, USA) following the manufacturer's instructions.

Molecular detection of T. cruzi infection
All DNA preparations were screened for the presence of T. cruzi using a conventional PCR targeting satellite DNA [25,26]. The PCR was performed in a final volume of 25 μL containing 40-50 ng of genomic DNA, 1X of buffer, 0.04 mM of dNTP, 1.5 mM MgCl2, 0.4 μM of each primer (TCZ1 and TCZ2), and 0.05 U of Taq polymerase (Invitrogen, California, USA). The thermal cycling conditions were as follows: pre-heating at 95˚C for 15 min, 40 cycles at 95˚C for 10 s, 55˚C for 15 s, and 72˚C for 10 s in a thermal cycler [25]. Amplification products were electrophoresed on a 1.5% agarose gel stained by ethidium bromide and visualized under UV light. Samples were considered positive for T. cruzi when a band of 188 bp was observed in the gel.

Molecular characterization of T. cruzi
Positive T. cruzi samples were analyzed for molecular discrimination of TcI DTU of the other DTUs based on the amplification of the spliced leader intergenic region (SL-IR) gene using the primers TCC, TC1, and TC2, as previously reported [27,28]. The PCR was performed in a final volume of 25 μL containing 40-50 ng of genomic DNA, 1X of a buffer, 0.25 mM of dNTP, 2 mM MgCl2, 0.4 μM of each primer, and 0.05 U of Taq polymerase (Invitrogen, California, USA). The thermal cycling conditions were as follows: pre-heating at 94˚C for 5 min, 35 cycles at 94˚C for 30 s, 55˚C for 30 s, and 72˚C for 45 s in a thermal cycler, and a final extension at 72˚C for 5 min. We also used part (PCR directed to the SL-IR region only) of the algorithm implemented by Hernández et al. [29]. Amplification products were run on a 1.5% agarose gel, stained by ethidium bromide, and visualized under UV light. TcI was identified in the samples when a band of 350 bp was observed in the gel. For TcI positive samples, SL-IR PCR products were purified and sequenced at the Macrogen sequencing service, Seoul, South Korea. All nucleotide sequences were edited and aligned using CLUSTALW as implemented in BioEdit v.7.1.9 [30], and the microsatellite motif of the spliced leader gene (positions ranking between~14 to~40) was omitted as suggested [31]. The highest nucleotide identity value of the sequences based on optimal global pairwise alignments of each SL-RI sequence against reference strains reported for Colombia [5] was calculated in BioEdit v.7.1.9 [30].

Detection of PCR inhibition
The presence of PCR inhibitors was examined in PCR-negative feces. In brief, 1 μL of DNA from negative feces were mixed with 10 ng of DNA from the Gal 61 T. cruzi strain in the same reaction tube to reference the amplification of satellite DNA of T. cruzi, as described above. The PCR was considered as inhibited when the 188 bp fragment was not detected in the gel.

Data analysis
To estimate the risk of infection, the following entomological indices were determined in the municipalities of group one: the infestation index (number of houses infested by triatomines / number of houses examined x 100); the colonization index (houses with triatomine nymphs/ number of houses positive for triatomine × 100), the density index (number of triatomines captured / number of houses examined); the crowding index (number of triatomines captured / number of houses with triatomines) and natural infection index (number of infected pools/ total number of analyzed pools x 100) according to WHO (2010) [32]. Proportional comparisons were performed using the Chi-square test (χ 2 ) with a P < 0.05considered as significant. Data analyses were performed using SPSS v.18.0 statistical software.

Geospatial analysis
For the map construction of the distribution of the triatomines, points of the samples localization were visualized in a Geographic Information System (GIS) in the Quantum GIS software version 3.4 Madeira (https://www.qgis.org/en/site/forusers/download.html), using the continental, national, and State boundaries (Shapefiles), extracted from the free and open source DIVA-GIS site (https://www.diva-gis.org/gdata). Coordinates were recorded from the WGS 84 Datum (World Geodetic System 1984) geodetic coordinate system. Mapping of the triatomine distribution in the urban and rural area of the 29 municipalities was done from the study area. For this, the relative distribution of the collected and examined triatomine species, infection index, infestation index, and TcI Dom genotype and sylvatic distribution indicate the total number of the collected and examined specimens, whereas pies display the percentage of collected and examined species. The spatial data were analyzed in a Geographic Information System (GIS) platform using the open-source (Quantum GIS) QGIS (v.3.4 Madeira), a free and open-source geographic information system software.

Local indicators for the spatial association: potential clusters of the distribution of Triatomine species, infestation and infection indices
The Global Moran's I index was calculated to verify spatial autocorrelations between the distribution of triatomine species, infection indices, and infestation indices to identify transmission hotspot areas. To identify localized clusters, the local Moran's I index was calculated. For the univariate Local Moran (five species of triatomine: T. dimidiata, T. venosa, P. geniculatus, R. prolixus, and P. rufotuberculatus; T. cruzi infection index and infestation index) bivariate (five species of triatomine, T. cruzi infection, genotypes, and infestation index x relief), defined the neighborhood matrix using the weights manager tool. The criterion adopted was the queentype contiguity, with 1 value order of contiguity. Regions with common borders were considered neighboring. To index value significance demonstration, ie, the index value was not randomly obtained, the pseudo-significance test was performed with 999 permutations, and the 0.05 p-value was adopted. In this context, the Moran's I index tests the null hypothesis of spatial independence in the entire area (Moran's I = 0), generating a global I value potentially varying from −1 to +1: negative values indicate an inverse correlation (dispersion) while positive values suggest a direct correlation (clustering) [33,34]. Local Moran's I analysis was performed to determine local spatial association patterns. In addition to showing the dispersion diagram, as in IGM, the local index produces significance and cluster maps to determine the presence, or not, of spatial dependence. The Local and Global Moran's indices were calculated using GeoDa software (1.12) (GeoDa Center for Geospatial Analysis and Computation, Arizona State University, Tempe, AZ, USA).

Discussion
After 24 municipalities in Boyaca were certified by PAHO to be free from intradomiciliary transmission of T. cruzi through programs aimed to eliminate R. prolixus [10,11,35], we carried out an entomological survey to characterize the spatial distribution of secondary vectors found inside household units (intra-and peridomiciliary environments) in this area. The potential risk of T. cruzi transmission by secondary vectors in municipalities located above 2,000 m.a.s.l. constitutes a new scenario that must be included in regional CD elimination programs. The entomological, molecular, and spatial analyses revealed that T. dimidiata and T. venosa are the most abundant vectors with clustered distributions. Residual populations of R. prolixus in these municipalities certified by the PAHO indicate that the vector elimination programs must be strengthened and continued. The distribution of species, their role as vectors of T. cruzi, and infestation and infection indexes will allow us to identify priority areas for subsequent surveillance and control measures.
Triatoma dimidiata has traditionally been considered a vector of secondary importance in Colombia [9,36]. However, recent studies in areas certified by PAHO to be free of intradomiciliary transmission of T. cruzi by R. prolixus that the epidemiological importance of this species in parasite transmission may be increasing. Triatoma dimidiata was the most common (73.2%), most widely distributed (44.8%), and showed the highest levels of prevalence (37.8%) of triatomine species collected in Boyaca 2017 and 2018 (Fig 1 and Table 2). This species has moved into houses; 72.5% of specimens were collected inside homes, and it has high colonization and crowding indices in the northeast provinces (Norte, Gutierrez, and Valderrama). These data are supported by similar entomological indices reported in Capitanejo and Macaravita (Santander); and Soata and Tipacoque (Boyaca) between 2006 and 2008, using the same methodologies [37]. Finally, the spatial analysis showed that this species has a clustered distribution in the northeast (Norte province), and contributes to T. cruzi transmission (S1B Fig).
Triatoma dimidiata is a species with major epidemiological importance in Mexico, Central America, Ecuador, and some departments of Colombia [38][39][40][41][42]. Moreover, an essential epidemiological role for this species has been reported in Guatemala and Nicaragua, where it has maintained T. cruzi transmission after R. prolixus eliminating [43].
Triatoma dimidiata is distributed throughout eleven countries and has been reported from a wide range of altitudes, ranging from 0 to 3,100 m.a.s.l. [23]. Most reports of T. dimidiata from Central America and Ecuador are from areas located below 1,800 m.a.s.l. [23,44]. Our survey and spatial analysis results suggest that a different situation seems to exist in Colombia because this species is of epidemiological importance in municipalities located above 2,000 m. a.s.l. Other authors have reported the infection of T. dimidiata with Trypanosoma spp. and T. cruzi in municipalities in Boyaca, Cundinamarca, and Santander located above this altitudinal range [9,17]. Our results indicate that T. dimidiata could become established as a major intradomiciliary vector of T. cruzi, but further studies must evaluate T. cruzi transmission to humans in these highland areas.
Triatoma venosa has been considered a secondary vector species in Colombia due to the frequency with which they are reported inside home dwellings and peridomicile areas [17]. This sylvatic vector is principally distributed in the Andean region at altitudes between 125 to 2,700m.a.s.l. [23]. It has been recorded in eight departments and 87 municipalities where it has been related to acute outbreaks of CD [9,17]. Here, we show that T. venosa is the second most widely distributed triatomine with natural infection of 5.5%, and with a cluster in their distribution in southwest provinces: Lengupa, Neira, and Oriente (Fig 3). In our study, 17% of T. venosa collected were nymphs, and 71.4% were collected in intradomiciliary areas. Considering that previously ecoepidemiological studies in Boavita (Boyaca) showed high mobility of species between peridomicile and intradomiciliary environments [45], the results in the current study could be a consequence of T. venosa mobility in the peridomicile area or a result of a peculiar domiciliation of this triatomine species in this department [9,17]. Therefore, we suggest that both scenarios should be explored.
Panstrongylus geniculatus has been considered recently as a relevant vector for T. cruzi due to its geographic distribution, its record of domiciliation, and association with outbreaks of oral T. cruzi infections in Colombia and Venezuela [20][21][22]. This species, reported in 18 countries, has a wide altitudinal range of 0 to 3,842 m.a.s.l. [23]. Here, we report its presence in ten municipalities in Boyaca, three of which are located above 2,000 m.a.s.l. and found this species infected with T. cruzi. However, we did not find a clustered distribution [9,17]. The overall prevalence of infection in this species was 36% (S1C Fig), which was lower than that detected in other endemic areas of Colombia (58.8%) [23]. Molecular genetics studies of the population of P. geniculatus showed that this species has a monophyletic origin and contains four clades with no eco-epidemiological differences between them [46]. In this sense, the population found in Boyaca could have the same epidemiological relevance as Venezuela or other Colombian regions where this species has been involved in oral transmission outbreaks [46].
TcI was the only DTU found in P. geniculatus. These results are contrary to reports in different Colombia regions, where P. geniculatus was infected with TcI, TcII, TcIII, TcIV, and TcV [14,47,48]. These differences could be related to this species' blood-meal source in the other departments, where its capacity to adapt to different food sources has been reported [14]. Finally, the presence of both genotypes of TcI in this species suggests its role in transporting wild populations of the parasite into houses [13,49].
Panstrongylus rufotuberculatus is considered a sylvatic vector in Colombia, where it is present in eight departments and 26 municipalities [9,17]. This species is found in intradomiciliary and peridomiciliary areas in the lowland areas and above 3,600 m.a.s.l. in Bolivia [22]. P. rufotuberculatus has been incriminated as a CD vector in the Andean and coastal regions of Ecuador [50]. In contrast, in Colombia, it has been considered a major epidemiological risk factor in Amalfi (Antioquia), where it was reported as the second most common triatomine caught inside dwellings [51]. Here we report the presence of P. rufotuberculatus in two municipalities from the study area. However, this species has been reported in this department's other six cities (S1E Fig) [9,17]. These results, including its high infection rate and its exclusive detection inside homes, suggest a high potential transmission risk in municipalities where this vector is present. As far as we know, this is the first time that TcI genotypes of T. cruzi have been identified in P. rufotuberculatus, showing its participation in the sylvatic cycle.
Rhodnius prolixus has been considered for many years to be the most important vector of T. cruzi in Colombia. This species was introduced accidentally into Boyaca and has been reported in 26 municipalities of this department between 2002 to 2014 [17]. Here we show its presence in 10.3% of the cities (Cubara, Labranzagrande, and Paya), supporting the data on the success of control programs and the virtual elimination of this species. However, the vector's presence in two municipalities (Labranzagrande and Paya) that had been certified free of R. prolixustransmitted T. cruzi endangers the overall success of CD control programs. Although R. prolixus in this study were not infected with T. cruzi, the high crowding index detected in municipalities where this was the only species found (Cubara) (Tables 1 and 2 and S1D Fig) demonstrates the remarkable capacity of this species to colonize and establish intradomiciliary populations in this region of Colombia [8]. Reinforcing insecticide spraying and vector surveillance in these municipalities are necessary to prevent vectorial transmission of T. cruzi in cities where residual populations of this vector were detected.
In Colombia, municipalities located at heights above 2,000 m.a.s.l. have traditionally been considered as non-endemic for CD [15,16]. The national protocol for entomological surveillance and vector control of CD suggests continued surveillance to identify triatomine vectors in areas below the 2,300 m.a.s.l. [52]. However, we found populations of T. dimidiata and P. geniculatus infected with T. cruzi in municipalities located higher than the cut-off altitude. The entomological indices and the presence of TcI Dom suggest a regular domestic cycle in these municipalities. We suggest expanding the altitudinal range to improve our understanding of the true entomological range of these vectors and future research to determine intradomiciliary transmission above 2,000 m.a.s.l. in Colombia. The presence of triatomines in highland municipalities may be related to climate change, as suggested for other triatomine species [53][54][55]. Further research about climate change effects should be studied for this species.
Several limitations of the present study should be acknowledged, including (i) we could not carry out an active search in all the municipalities with a history of triatomine infestation, thus preventing comparisons and homogeneous sampling (ii) the differences in triatomine collections between 2017 and 2018 were affected by administrative problems that did not allow the early incorporation of health workers of BDHS, affecting the number of operational personnel to collect insects and the activation of PRT. (iii) The blood sources of synanthropic triatomines have not yet been evaluated, and future trials should be carried out to understand the risk and contact events between triatomines and residents. (iv) The present work was not accompanied by a seroprevalence study of the residents whose homes we collected T. cruzi positive insects. Even though BDHS has carried out screenings in school children under eight years of age at the municipality level, there are no data from residents outside the school ages. New serological studies targeted towards residents of infested homes should be carried out to have a broader picture of the secondary vectors' risk. Finally, further studies are needed in these municipalities to determine seasonal variations in the abundance of secondary vectors to prioritize the effectiveness of control programs.
The interruption of intradomiciliary T. cruzi transmission by R. prolixus in some areas of Colombia has been a major advance in CD control. However, this is an anthropozoonotic disease with transmission carried out by many vector species. The sustainability of successful management programs requires continuous evaluations of entomological data of all vectors that could exploit the available niche after the elimination of common triatomines such as R. prolixus. A sustained surveillance program is essential pre and post primary vector eradication efforts.