- Record: found
- Abstract: found
- Article: found
An increased number of individuals of a potential host facilitates non-photic synchronisation in the haematophagous insect Triatoma infestans
research-article
Publication date (Electronic, pub):
31 July 2023
Publication date (Electronic, collection):
2023
Journal:
Memórias do Instituto Oswaldo Cruz
Publisher:
Instituto Oswaldo Cruz, Ministério da Saúde
There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.
Abstract
BACKGROUND
Triatoma infestans (Kissing bug) is the main vector of the parasite causative of Chagas disease in Latin-America. This species shows clear activity rhythms easily synchronised to day-night cycles (photic cycle). The haematophagous nature of these insects lead us to think that they may temporally adapt to the particular activity rhythms of potential hosts (non-photic cycle). Our previous data showed that kissing bugs were weakly affected by the activity-inactivity rhythm of a single host.
OBJETIVE
To determine if by increasing the number of individuals of a potential host, T. infestans could increase the likelihood of synchronisation.
METHODS
Individual activity rhythms of experimental insects, maintained in constant darkness in light-tight cabinets, localised in a room with 24 rodents, were continuously monitored. Another insect group that served as control was maintained in the same conditions but in a room without rodents.
Related collections
Most cited references32
- Record: found
- Abstract: not found
- Article: not found
The chi square periodogram: its utility for analysis of circadian rhythms.
Phillip G. Sokolove, Wayne Bushell (1978)
- Record: found
- Abstract: found
- Article: not found
In search of a temporal niche: environmental factors.
Time as an ecological niche variable or "temporal niche" can be defined in the context of the most prominent environmental cycles, including the tidal cycle, the lunar day and month, the solar day, and the earth year. For the current review, we focus on the 24-h domain generated through the earth's rotation around its axis (solar day). The daily environmental cycles of light and temperature are a dominant ecological factor generating a variety of adaptations among animals. In this review, we describe these adaptations with a special focus on the visual system and on the adaptive plasticity of activity patterns. Our goals are: (1) Underscore the importance of the 24-h time axis as critical variable in the ecological niche. (2) Highlight cases of temporal niche switches at the evolutionary timescale (phylogenetic level). (3) Review temporal niche switching within an individual's lifespan. (4) Evaluate possible underlying mechanisms for temporal niche switching. (5) Describe a new hypothesis of circadian thermoenergetics which may explain several cases of temporal niche switching in mammals. With this, we hope to inspire experiments under natural conditions or more complex laboratory environments, aimed to reveal environmental factors and mechanisms underlying specific temporal programs. Copyright © 2012 Elsevier B.V. All rights reserved.
- Record: found
- Abstract: found
- Article: not found
Strong Host-Feeding Preferences of the Vector Triatoma infestans Modified by Vector Density: Implications for the Epidemiology of Chagas Disease
Introduction Host choice of hematophagous insects mainly depends on relative host abundance and proximity, host defensive behavior, the density of blood-sucking insects, and the spatial and temporal concurrence of hosts and insects [1],[2]. Examples of innate (genetically determined) host-feeding preferences are few, and convincing evidence with both experimental and field support is scarce [1],[3]. Fleas (Xenopsylla conformis) do not have an innate preference but can discriminate between juvenile and adult hosts, and derive a higher reproductive reward when feeding on juvenile hosts [4]. In Lutzomyia longipalpis sandflies, host size was the main determinant of host-feeding choices among a human, a dog and a chicken exposed simultaneously to laboratory-reared sandflies [5], and its feeding success on chickens was density-dependent [6]. For Triatoma infestans bugs [7] and Simulium damnosum blackflies [8], the proportion of insects biting humans was strongly density-dependent. For Glossina palpalis gambiensis tsetse flies, male flies preferred to feed on cattle rather on reptiles in a stable; the host species selected for the second bloodmeal depended on the host encountered for the first bloodmeal, the between-meal interval and the interaction between these two factors [9]. In mosquitoes, acquired feeding preferences are reflected in their tendency to return to the same villages, houses, host species and oviposition sites [10]. A non-homogeneous distribution of vector feeding contacts on the same host species leads to a basic reproduction number of the pathogen (R 0) greater than or equal to that obtained under uniform host selection, a result that still holds when groups of mosquitoes and hosts are highly structured in patches [11],[12]. Triatomine bugs (Hemiptera: Reduviidae) are the vectors of Trypanosoma cruzi, the causal agent of Chagas disease. Triatoma infestans (Klug), the main vector of T. cruzi, is a highly domiciliated species that also occurs in peridomestic structures housing domestic animals [13]. Like most species of triatomine bugs, T. infestans shows eclectic host-feeding patterns [14],[15]. Host proximity has usually been considered more important than host preference for hungry bugs seeking to feed [14]. In laboratory-based host choice experiments of Triatoma sordida (a species typically associated with birds), first-instar nymphs significantly preferred birds to humans [16] whereas fifth-instar nymphs feeding success and bloodmeal size were significantly larger on guinea pigs than on pigeons [17]. Triatoma infestans preferred caged chickens to guinea pigs though not in all replicates [18]. In a simultaneous exposure of four caged vertebrate species to separate groups of fifth-instar nymphs of T. infestans, Triatoma dimidiata and Rhodnius prolixus, none displayed dominant host-feeding preferences among dogs, chickens and opossums but toads were only rarely fed upon [19]. These authors [19] concluded that T. infestans showed a slight preference for dogs in short daytime experiments and a slight one for chickens in overnight trials. No measure of variability in host-feeding choices between the 7–22 replicates for each triatomine species was reported and neither were statistical procedures described. Within the restricted experimental conditions used, the tested triatomine species do not appear to have a fixed or dominant preference for any of the study hosts, and the question whether there are host-feeding preferences between dogs and chickens is still unresolved. In rural areas of the Argentine Chaco, domestic T. infestans blood-fed more frequently on dogs or chickens than on the human hosts or cats with which they shared sleeping quarters [7]. Seasonal host shifts were recorded. In spring-summer bug collections, the proportion of domestic bugs that fed on dogs increased significantly with increasing numbers of dogs and T. infestans in bedroom areas, and decreased as bug feeding frequency on chickens rose. Feedings on cats increased significantly with the number of cats and decreased with the number of dogs in bedroom areas. Dog-fed T. infestans had higher infection prevalence with T. cruzi than bugs feeding on other hosts, but many bugs within a given house fed on up to four different bloodmeal sources in summer [20]. Both domestic dogs and cats acted as a source of T. cruzi infection to other species, including humans, whereas chickens (not susceptible to T. cruzi) contributed strongly to bug population growth [21],[22]. Using molecular typing techniques of T. cruzi, we recently showed that dogs, cats and a large fraction of the T. infestans within a household shared the same parasite sublineage and therefore were connected epidemiologically [23]. Understanding the factors that affect the host-feeding selection patterns of triatomine bugs is crucial to estimating transmission risks and predicting the putative effects of introducing or removing domestic animal hosts or targeting them for control. The recent emergence of pyrethroid resistance in T. infestans in northern Argentina and Bolivia [24], combined with the low effectiveness of standard residual spraying of pyrethroid insecticides in peridomestic structures [25], gave strong impetus to the search for cost-effective, alternative treatments based on the application of powder, topical lotions or insecticide-impregnated collars to the domestic animals themselves [26]–[28]. Whether chickens or dogs would be the preferred targets is one of the questions that motivated the current study, in which we report the first binary discrete host choice experiments of triatomine bugs conducted in small mud-and-thatch huts under natural climatic conditions. We tested whether T. infestans displayed host-feeding preferences between dogs and chickens and between dogs and cats, all unrestrained, and whether the density of vectors per hut modified host-feeding success, blood-engorgement and other vital rates in replicated trials. Analysis of field blood-feeding patterns and laboratory experiments supported the hypothesis that T. infestans would prefer chickens to dogs and dogs to cats, though the evidence regarding chickens and dogs was inconclusive [15],[19],[29]. We also hypothesized that blood-feeding success, engorgement and post-exposure bug weight would be reduced in a density-dependent way [13],[30]. To infer the putative processes accounting for the observed discrepancies, we re-examined the reported host-feeding patterns of domestic T. infestans in the field in light of experimental host choices and the demographic and behavior patterns of domestic animal hosts. Materials and Methods Study Site The trials were carried out in the field station run by the Argentinean National Vector Control Program in Punilla, Province of Córdoba (31°14′S, 64°28′W) in summer (late January) and in early winter (June) 2006. Study location and experimental set-up were previously described [27],[31]. For the present study, six small experimental huts simulating typical mud-and-thatch houses (1.60×0.80×0.80 m with a 40 cm-wide entrance) were built and subdivided into two equally-sized rooms that shared an adobe-bricked wall with loose bricks; this arrangement allowed the bugs to hide and move freely between rooms. The lower third of the middle wall and all of the other walls were plastered on the inside with a 7∶1 mixture of soil and cement, and a cement carpet was added over the floors of beaten earth. A cage of plastic mosquito netting mounted on a metal frame was placed above each hut to prevent bugs from escaping. The six huts were arranged in two rows over a 50 m2 rectangle. Animals Seven mongrel male dogs (approximate age range, 4–7 years; mean weight, 10.8 kg; SD, 3.4; range, 7–15) were used in the trial. All dogs had been exposed to T. infestans and had worn deltamethrin-impregnated collars for a four-month period ending six months before the current experiment [27] but not thereafter. According to the collars' manufacturer, the residual effect of the insecticide should cease within six months or one month after removing the collars; since collar use started >10 months before the first trial, no residual effect was expected to occur at this time. Dogs were vaccinated and dewormed with mebendazole prior to the start of the trial; they were kept in kennels made of chicken wire and a roof and fed twice daily. Chickens (all females; approximate age range, 2–2.5 years; mean weight, 2.4 kg; SD, 0.3; range, 1.9–2.8) of Lohmann breed were identified with a color ribbon and kept separately from other animals in a similar pen. Cats were a female and two male adults (approximate age range, 2.5–4 years; mean weight, 2.7 kg; SD, 0.2; range, 2.5–2.9). Dogs, but not chickens or cats, had previously been exposed to T. infestans bites six months before the first trial [27]. Chickens and cats had not been treated with insecticides. During the trials each animal was stationed individually inside a specific experimental hut at sunset and then released every morning into its specific area within the compound. This study complied with guidelines on research and biological testing activities involving live vertebrate animals from the Institutional Animal Care and Use Committee (IACUC) at FCEN-UBA, which is based on the International Guiding Principles for Biomedical Research Involving Animals developed by the Council for International Organizations of Medical Sciences. The T. infestans bugs used in these experiments were first or third generation from bugs collected in Córdoba, Santiago del Estero and San Luis (Argentina); they had been reared on chickens at the insectary (at 27°C, relative humidity 70%), fed to repletion on the fourth instar, and starved for 2.5 (first trial) to 3.5 months (second trial) prior exposure to the hosts. This long starvation period (>2 months after they molted to fifth instars) normally does not increase bug mortality, and was used to secure that the previous bloodmeals on chickens were completely digested at the time of the trials (i.e., no ‘false positive” bloodmeal). Before release, a 20% sample of triatomine bugs for each trial was weighed individually with an electronic balance (precision, 0.1 mg), and the volume and shape of the bugs' midgut was observed by transparency against a torch light to check their nutritional status semi-qualitatively based on the size of the bug abdomen and occurrence of blood remnants [27],[32]. This classification (by which bugs are scored as unfed, little fed, medium fed, and fully fed) was consistent between observers. All bugs were classified as unfed immediately before the trials. Study Design The first trial was started on late January 2006 (summer) and included six matched dog-chicken pairs, each housed in a different hut. Each pair was randomly assigned to one of two levels of bug abundance (30–31 or 90–91 fifth-instar nymphs of T. infestans); the upper bug density level was chosen because it had revealed negative density-dependent effects on domestic bug host-feeding patterns whereas the lower one did not [7]. The trial was replicated on three successive nights in the absence of any artificial source of light. Each host species was housed in a different room. Hosts were rotated among huts and between rooms every night, so that each individual host was matched with a different individual of the other host species during each of the three nights, and each individual room housed alternate host species in successive nights. Before the hosts were stationed within the huts at 8 p.m., the bugs were placed in a box with holes on the central wall at mid-day, and recovered after dismantling the movable parts of each hut on the next morning at 8 a.m. On recovery, all insects were immediately brought to the insectary, counted, scored for degree of engorgement, kept for 2 days post-recovery and then weighed (to allow them to approach the body weight plateau after eliminating the surplus of water in the bloodmeal), put in a vial labeled with a unique identifier for each bug, and then frozen at −20°C until dissection and bloodmeal identification. A subsample of 20 bugs not exposed to the hosts (control bugs) was frozen at −20°C at the same time as the recovered bugs to check whether there was any residual chicken bloodmeal in them. Given the experimental setup, we exclude the possibility that the small proportions of lost bugs escaped from the caged huts, and assume that lost bugs were most likely eaten by hosts. The proportion of blood-fed bugs is defined as the number of fed bugs (including little-fed, medium-fed and fully-fed bugs) plus bugs in the unfed nutritional class that later were ELISA-reactive to the test host species, relative to the total number of bugs examined for nutritional status; “fed” is therefore a composite category adjusted for bloodmeal reactivity among unfed bugs. The proportion of engorged bugs is defined as the sum of bugs medium fed and fully fed relative to the total number of fed bugs. The second trial, conducted in June 2006 (late fall), included three pairs of dog-cat and used the same protocol as the first trial except that hosts were stationed in the huts at 6 p.m. One replicate could not be finished properly because the cat fled away at the outset; this replicate was excluded from all calculations and analysis. Temperature and relative humidity inside the huts were measured using data loggers (Hobo H08, Onset) inserted into the thatched roofs of both rooms and on the outside wall of a hut in the first trial, and on each of the three huts in the second trial. In the dog-chicken trial, mean internal temperatures from 8 p.m. (sunset) to 8 a.m. over the three trial nights were 22, 24 and 21°C, respectively; the mean temperature difference between rooms within a hut (dog-to-chicken) in the stated period ranged from −0.2 to +0.7°C. In the dog-cat trial, mean (minimum, maximum) internal temperatures from 6 p.m. to 8 a.m. over the three huts in each trial night were 8.8 (5.4, 13.3), 7.6 (3.7, 12.2), and 10.0°C (7.8, 14.9), respectively. The mean temperature difference between rooms within a hut (dog-to-cat) averaged over the three huts for each trial night was −0.52 (SD, 0.46), +0.09 (SD, 0.27), and −0.41°C (SD, 0.53). Identification of Bloodmeals Standardization of the direct ELISA assay was based on previous procedures [33],[34] and the ELISA reagents' manufacturer manual (Kirkegaard & Perry Laboratories (KPL) Inc., Gaithersburg, MD). Preparation of bloodmeal samples Bloodmeal contents of laboratory-reared T. infestans nymphs fed only on goat, dog, cat, chicken and human and then frozen at −20°C were used to standardize the ELISA assay. The individual bloodmeals were prepared for testing by cutting the thorax transversally at the level of the third pair of legs, and then by expressing the bloodmeal out of the crop into a labeled vial. The bloodmeal contents were weighed, diluted from 1∶5 to 1∶50 with 0.01 M phosphate buffered saline (PBS), pH 7.4, and then frozen at −20°C until testing. ELISA procedure Peroxidase conjugates for identifying dog, chicken and cat blood meals obtained from KPL were reconstituted according to the manufacturer's instructions in 1∶1 of distilled water and glycerol. Bloodmeals were diluted in PBS (1∶5,000) and 50-µl volumes of the dilution were added to wells of polyvinyl chloride, mid-binding flat-shaped bottom, 96-well microtiter plates (Greiner Bio-one). Control sera were diluted 1∶60,000. The plates were covered with an acrylic plate and incubated at 37°C for 1 h, or at 4°C for 15–20 h. Well contents were aspirated and the plate was tapped against a stack of paper towels to remove excess liquid. Each well was then washed three times with PBS containing 0.01% Tween 20 (PBS-Tween 20), followed by the addition of 200 µl of the blocking buffer (BB; PBS plus 3% milk; San Regin no-fat milk) and incubated at 37°C for 1–2 h. Meanwhile, the antibody-enzyme conjugate (antihost IgG-peroxidase) was prepared by adsorbing the corresponding antisera in PBS plus 0.05% milk at 37°C for 1 h (AbC-BB). Adsorptions for each antihost IgG involved approximately 16 µl of each heterologous serum for each 8 ml of specific AbC-BB solution [33]. For the domestic hosts of T. infestans, we used the following adsorptions: antichicken-IgG with goat and cat sera (because no cross-reaction with other mammal sera was observed at our test conditions, and to minimize pre-incubation time); antidog-IgG with cat, chicken and human sera, and anticat-IgG with dog, human and chicken sera. Well contents were aspirated after 1 h, washed three times with PBS-Tween 20, and left at −20°C if they were scheduled for use on the following days. For the immune reaction, 50 µl of the AbC-BB solution were added to each well and incubated at 37°C for 4 h. Well contents were aspirated again and washed three times with PBS-Tween 20. For the enzymatic reaction, 100 µl of ABTS (2,2′-azino-di[3-ethyl benzthiazoline sulfonate]) peroxidase substrate (KPL Inc.) left for 1 h at room temperature prior to utilization were added to each well and left at room temperature for 10–15 min. The reaction was stopped by adding 100 µl of SDS stop solution 1× (KPL Inc.) to each well. Absorbance at 405 nm was determined with an ELISA reader (Bio-rad 680 XR). Each microtiter plate contained test and control samples of the target host species and four negative controls (i.e., bloodmeal contents of heterologous host species -goat, dog, cat, and chicken- in PBS). Samples were considered positive if absorbance values exceeded the maximum OD of four negative control bugs (cut-off) plus three times the standard deviation of the mean. The indeterminate zone was taken to include 10% around the cut-off. Samples were tested in duplicates, and the outcome was considered valid if the mean of both duplicates of each sample did not differ by more than 15%. Sensitivity and specificity of the direct ELISA for the target host species was evaluated by testing bugs fed to repletion separately on a dog, cat and chicken, killed 7–30 days after feeding and held frozen at −20°C. Bloodmeal contents were weighted and diluted in PBS to 1∶5,000. A total of 44 bloodmeals on each of the three test host species was tested. The sensitivity of the direct ELISA test was 100% for all three host species. Specificity was 100% for dog and cat, and 97% for chicken. Bloodmeal testing In the dog-chicken trial, all bloodmeals from low-vector density replicates and a randomly chosen sample (>50%) of the blood-fed insects pertaining to high-vector density replicates were tested by ELISA. The latter procedure was followed because of the very large number of fed bugs available for testing, and the large preference for dogs observed. In the dog-cat trial, all bloodmeals from insects recovered were tested by ELISA (including bugs scored unfed) to allow for the possibility that bugs achieved very small bloodmeals on cats. Data Analysis The data collected were entered in an Access database. Feeding indices (FI) were calculated as the ratio of the number of bugs that fed on a given host species X to the number of bugs that fed on the matched host species Y (whether or not the bugs that fed on X fed on Y and vice versa). As only one host of each host species was present we did not need to correct for the number of hosts [35]. Four related measures of blood gain by the bugs were used: i) feeding success, a binary variable measuring the likelihood of blood-feeding on any one or on both host species inside the hut (i.e., overall feeding success: fed bugs relative to the number of bugs recovered alive or dead), or on a specific host species as determined by ELISA (i.e., host choice); ii) engorgement (a binary variable including medium-fed and fully-fed bugs: engorged bugs relative to the number of fed bugs); iii) nutritional status (a categorical variable with four levels), and iv) post-exposure bug weight (a continuous variable, measured two days after host exposure). Engorgement and post-exposure bug weight measure the amount of blood imbibed overall or on a given host species. Exact 95% confidence intervals (95% CI) for mean vital rates (i.e., binary variables) were based on the binomial distribution. The effect size on several binary response variables was estimated by fitting random-effects logistic regression models clustered by hut to the data using the command xtlogit in Stata 9.1 [36]. The use of random-effects models addresses the fact that insects within a hut roughly share the same environment and other undetermined characteristics that may create dependencies between responses within the same cluster of observations. We tested for significant (P 0.8). Vector density adjusted for occasion effects did not modify significantly any of the vital rates in both trials (Table 1). 10.1371/journal.pntd.0000447.t001 Table 1 Mean vital rates (lost, dead, fed, and engorged) of fifth-instar nymphs of T. infestans in two host choice experiments. Trial Day No. of replicates No. of bugs released Percentage of bugs lost dead fed engorged Dog-chicken 1 6 360 8.1 0.0 98.5 48.2 2 6 361 5.3 0.0 99.4 47.1 3 6 360 6.9 0.8 98.2 45.3 Sub-total 18 1081 6.8 0.3 98.7 46.8 OR for vector density effects 0.63 0.26 1.13 95% confidence interval 0.05–7.57 0.03–2.11 0.79–1.62 Dog-cat 1 2 120 4.2 5.0 64.0 39.7 2 3 211 1.4 3.8 64.1 52.3 3 3 210 2.4 1.9 82.8 40.2 Sub-total 8 541 2.4 3.3 71.4 44.4 Total 26 1622 5.3 1.3 89.1 46.2 OR for vector density effects 0.34 3.85 0.62 0.55 95% confidence interval 0.07–1.74 0.50–29.62 0.35–1.08 0.30–1.00 In the dog-chicken trial, we observed that most of the bugs were recovered from the thatched roof of the dog's room, followed by the adobe bricks in the mid-wall; the fewer bugs recovered from the chicken's room were in the thatched roof. Of all the bugs with identified bloodmeals, 81.8% had feedings on dogs and 24.0% on chickens. The dog-to-chicken mean feeding index was 7.0 (95% CI, 3.7–10.3). The total mean percentage of insects that fed on dogs only (75.0%, 95% CI, 71.5–78.3%) was significantly higher than that on chickens only (18.0%, 95% CI, 15.1–21.1%) (Fig. 1A). Both feeding choices were highly significantly correlated (r = 0.88, P 0.1). When each replicate was taken separately, dogs were significantly preferred over chickens in 16 of 18 replicates (binomial test, P≤0.001 in 13 replicates and P 0.2). Two of the cats frequently allowed the bugs to blood-feed on them though with large variations between nights. In the excluded replicate that had no cat, 93% of the released bugs were recovered and 79% of them were fed on dog only, with no feeding on cat detected. The relationship between proportional host body weight and host-feeding preferences in both trials is shown in Fig. 3. Two different patterns were obtained. The proportion of bugs that fed on dogs and proportional dog body weight were unrelated in the dog-chicken trial, whereas a significant relationship was found in the dog-cat trial (OR = 1.21; 95% CI, 1.03–1.44, P = 0.022) where exclusion of an outlier value gave a stronger relationship (OR = 1.32; 95% CI, 1.20–1.45, P 0.4) vector density effects (n = 600, χ2 = 44.3, P 0.6) (R 2 = 0.606, n = 583, P 30%) and predominance of males (range, 67–85%) in northern Argentina [51]. Although 50%–68% of the dogs and cats were reported to sleep in domestic sites (albeit with undetermined frequency), they were not restrained or confined and many roamed freely within and around the village to forage for all or part of their food. Dogs were sometimes reported to free-range in small packs at night (probably in oestrus groups with several males), whereas 56% of the cats were reported to stray in the forest to hunt. These qualitative features appear to be common in resource-poor rural settings in northern Argentina where Chagas disease is hyperendemic, and where there is little demarcation of property lines. In other rural locations, adult male dogs and cats have a larger home range and longer duration of activity than females; dogs display a particularly marked crepuscular free-ranging behavior with two peaks of daily activity [52],[53]. These peaks of activity mostly coincide with the main host-seeking periods of T. infestans bugs, and would reduce host availability and the likelihood of dog-vector encounter at such periods. We infer that current domestic bug host-feeding patterns reflect limited, heterogeneous dog exposures, and there would be ample room for increased feedings on dogs with increased dog exposure. The long-standing controversy on the role of cats in the domestic transmission of T. cruzi infection revolved around the notion that their well-known nocturnal activity pattern and apparent intolerance to bug bites would reduce the likelihood that domestic bugs blood-fed on cats [29]. However, several studies showed that domestic T. infestans and other triatomine species blood-fed on cats as frequently as or more than on dogs [7],[20], a pattern that also emerged in our host choice experiment. To the apparent inconsistencies between different pieces of evidence already discussed [21], we add geographic variations in pet ownership and keeping practices, and heterogeneities in host behavior and exposure between demographic subgroups of cats (by age, sex and reproductive status). In our rural study area, kittens and young pups were frequently kept indoors all day long. Cats avoided exposure to hot weather by resting indoor in the cooler, dark mud-and-thatch houses during the daytime. Under such conditions, residents of highly infested houses sometimes reported to be attacked by T. infestans nymphs when resting on the floor indoors at noon (unpubl. observations). Age- and sex-dependent host activity patterns modify host exposure to domestic vectors, which combined with individual host heterogeneities and other factors, explain the large variability in feeding patterns observed between households and studies. Some aspects of our study design limit the interpretation of results. Lack of direct observations on host attractiveness and defensive behavior limits the interpretation of the observed host-feeding preferences in terms of underlying mechanisms. Although both trials had the same initial density of bugs per hut, different temperatures affected attack, feeding and other vital rates; therefore, we refrained from establishing quantitative comparisons between trials and focused on within-trial outcomes. Bloodmeal size was not recorded directly, but nutritional status and post-exposure weight were strongly positively correlated; both may be considered valid surrogate indices of bloodmeal size given that pre-exposure bug weight was very similar between trials. Post-exposure bug weight should be increased roughly by 50% to estimate the actual weight immediately after feeding because of the rapid loss of excess water during the first day or two post-feeding depending on temperature. This rough calculation and comparison with laboratory-based estimates of bloodmeal size suggests that most of the bugs blood-fed to repletion in the summer trial. Because only the study dogs had been bitten by T. infestans before the experiments, differences in their background experience with bugs might modify the bugs' preference for dogs. This effect may not be too serious, considering that naïve chickens have an innate, stable tendency toward tolerating or defending from triatomine bites even after a rest time between exposures [42]. The individual study dogs also expressed some idiosyncratic behaviors against bugs over time. Despite the fact that the study dogs had been exposed to T. infestans bites six months before the first trial (i.e., were presumably immunized against bug salivary antigens), most of the dog-fed bugs blood-engorged close to repletion. Therefore, it seems very unlikely that the dogs' immune response reduced substantively vector host-feeding success or engorgement. Immune reactions to bug saliva on the skin of previously sensitized chickens facilitated vector blood-feeding [54] but we have not observed such skin reactions in dogs. It is unlikely that previous dog-bug contacts may have influenced the various steps involved in the host-selection process prior to engorgement; whether the degree of bug engorgement was modified by prior exposure of dogs remains to be determined. Unlike chickens found indoors in rural villages, the study hens were not brooding and their behavior in response to bugs may be potentially different. Implications for Disease Control and Modeling Increased host tolerance implies increased residence and feeding times on the host, which in turn will increase fitness by increasing the overall rate at which blood is obtained, eggs are produced, and survival per feeding attempt [55]. The nutritional quality of blood may differ substantively between host species of R. prolixus [56], with chicken blood having half the hematocrit than mammals and much lower hemoglobin or plasma protein than dogs [57]. Therefore, the aggregate fitness implications of host choices remain to be established. Of note, the host bloodmeal choice variable includes a survival component because it was measured on recovered, fed bugs. Because bloodmeal size increases the probabilities of T. infestans emitting dejecta sooner [58], ingesting trypanosomes and becoming infected [21], it follows that preferred, tolerant hosts such as dogs will seriously increase transmission rates relative to other domestic hosts. By virtue of allowing larger bloodmeals, the likelihood of dogs being repeatedly contaminated with bug feces and eventually superinfected with various parasite strains would be increased. The large frequency of unmixed dog bloodmeals shown by T. infestans in some field locations further suggests that a strong, stable link between individual dogs or groups of dogs and groups of bugs occurs in some households, thereby increasing transmission of T. cruzi back and forth from dogs to bugs and creating a transient partial refuge for other host species (a zooprophylactic effect). In most households, however, the frequency of mixed bloodmeals on dogs is high during spring-summer, and because domestic host species and bugs are more connected the flux of parasites between them is enhanced. Selective host choice amplified by a greater feeding success on diseased or infected hosts will increase the basic reproduction number of T. cruzi (though with possibly depressed prevalence and incidence as the outbreak follows through) compared with the base case represented by homogeneous contact rates [11],[12],[59]. An increase in dog or cat availability or accessibility in domestic areas will increase the rate of bug feeding on them which in turn will exert non-linear effects on R 0 through the squared biting rate term. When the proportion of insects feeding on a given host species (i.e., humans) varies with the relative abundance of non-human (i.e., dogs, cats, chickens) and human hosts and with the ratio of vectors to hosts, as our studies have shown, the relationship between R 0 and host blood indices is predicted to be strongly non-linear [2]. This implies that different tactics that seek to reduce vector abundance will exert very different impacts on parasite transmission depending on the exact relationship between R 0 and the vector-to-host ratio. The empirical evidence further supports the prediction that removal of dogs from bedroom areas will strongly decrease domestic bug population size, transmission rates and human incidence of infection [22]. Heterogeneities in vector feeding rates and in host exposure and infection will tend to create ‘hot’ and ‘cold’ spots of transmission, which can be used to target more accurately and efficiently host species and individuals accounting for most of the risk. Application of pyrethroid-impregnated dog collars, causing reduced repellency but increased bug mortality for extended periods [27], are predicted to strongly reduce domestic bug population size and transmission rates. The various layers of heterogeneity involving dogs in rural endemic areas, including household aggregation of infection, infectiousness to bugs and exposure patterns [21],[51], can be used when designing control measures. For increased impact, collars or other similar tools should be preferentially applied to those dogs that are infected with T. cruzi and/or highly infectious to bugs and that are also closely associated with domestic sites (e.g., pups, females in reproductive state or restrained dogs). Such dogs can be turned into baited lethal traps, though a thorough cost-effectiveness assessment of such tactics is needed before large-scale field application. Other possible applications are to use dogs as baited sentinels of bug presence through the use of its immune response to salivary antigens for serologic surveillance during a bug elimination campaign [60] and as sentinels of parasite transmission [51]. Supporting Information Alternative Language Abstract S1 Translation of the abstract into Spanish by Ricardo E. Gürtler (0.02 MB DOC) Click here for additional data file.
Author and article information
Contributors
Veronica Sandra Valentinuzzi/:
ORCID: http://orcid.org/0000-0003-2223-1769
Role: Both authors contributed equallyRole: Approved the final manuscript
Luciana Beatriz Abrahan: Role: Both authors contributed equallyRole: Approved the final manuscript
Journal
Journal ID (nlm-ta): Mem Inst Oswaldo Cruz
Journal ID (iso-abbrev): Mem Inst Oswaldo Cruz
Journal ID (publisher-id): mioc
Title:
Memórias do Instituto Oswaldo Cruz
Publisher:
Instituto Oswaldo Cruz, Ministério da Saúde
ISSN
(Print):
0074-0276
ISSN
(Electronic):
1678-8060
Publication date
(Electronic, pub):
31
July
2023
Publication date
(Electronic, collection):
2023
Volume: 118
Electronic Location Identifier: e220289
Affiliations
[1
]Centro Regional de Investigaciones Científicas y de Transferencia Tecnológica de La
Rioja, Universidad Nacional de La Rioja, Servicio Geológico Minero Argentino, Universidad
Nacional de Catamarca, Consejo Nacional de Investigaciones Científicas y Técnicas,
Entre Ríos y Mendoza, La Rioja, Argentina
Author notes
Author information
Veronica Sandra Valentinuzzi/
http://orcid.org/0000-0003-2223-1769
Article
Other ID: 01118DOI: 10.1590/0074-02760220289
PMC ID: 10392893
SO-VID: 0b3fbe8e-1820-45f8-a9ef-5454447758e7
License:
This is an open-access article distributed under the terms of the Creative Commons Attribution License
History
Date
received
: 21
December
2022
Date
accepted
: 19
June
2023
Page count
Figures: 3,
References: 30
Categories
Subject:
Research Article
Data availability: