Anopheles darlingi, the main malaria vector in the Neotropics, has been considered to be highly anthropophilic. However, many behavioral aspects of this species remain unknown, such as the range of blood-meal sources. Barrier screens were used to collect resting Anopheles darlingi mosquitoes from 2013 to 2015 in three riverine localities (Lupuna, Cahuide and Santa Emilia) in Amazonian Peru. Overall, the Human Blood Index (HBI) ranged from 0.58–0.87, with no significant variation among years or sites. Blood-meal analysis revealed that humans are the most common blood source, followed by avian hosts (Galliformes-chickens and turkeys), and human/Galliforme mixed-meals. The Forage Ratio and Selection Index both show a strong preference for Galliformes over humans in blood-fed mosquitoes. Our data show that 30% of An. darlingi fed on more than one host, including combinations of dogs, pigs, goats and rats. There appears to be a pattern of host choice in An. darlingi, with varying proportions of mosquitoes feeding only on humans, only on Galliformes and some taking mixed-meals of blood (human plus Galliforme), which was detected in the three sites in different years, indicating that there could be a structure to these populations based on blood-feeding preferences. Mosquito age, estimated in two localities, Lupuna and Cahuide, ranged widely between sites and years. This variation may reflect the range of local environmental factors that influence longevity or possibly potential changes in the ability of the mosquito to transmit the parasite. Of 6,204 resting An. darlingi tested for Plasmodium infection, 0.42% were infected with P. vivax. This study provides evidence for the first time of the usefulness of barrier screens for the collection of blood-fed resting mosquitoes to calculate the Human Blood Index (HBI) and other blood-meal sources in a neotropical malaria endemic setting.
Anopheles darlingi is the major malaria vector in the Amazon. This species has been commonly described as highly anthropophilic throughout its geographic range, although little is known about its feeding preferences. Scant information is available regarding the origin of An. darlingi blood-meals. In the context of malaria elimination programs, the Human Blood Index (HBI) may provide crucial information regarding mosquito-human contact related to transmission dynamics. Additionally, collection of resting An. darlingi is challenging, mainly because the resting behavior of this species has not been well characterized. Our study, conducted from 2013–2015 in three localities in Loreto Department in the Peruvian Amazon, showed for the first time the efficacy of the barrier screen methodology for collecting recently blood-fed An. darlingi in a neotropical setting for the purpose of identifying the source of their blood-meals. Our data show that An. darlingi feeds on humans, Galliformes, dogs, pigs and goats, and that 30% of the mosquitoes fed on more than one type of host. Despite this opportunistic feeding behavior, however, An. darlingi is primarily anthropophilic. We hypothesize that mosquito population structure is associated with feeding preferences, which may affect the pattern of malaria transmission in the area.
Citation: Moreno M, Saavedra MP, Bickersmith SA, Prussing C, Michalski A, Tong Rios C, et al. (2017) Intensive trapping of blood-fed Anopheles darlingi in Amazonian Peru reveals unexpectedly high proportions of avian blood-meals. PLoS Negl Trop Dis 11(2): e0005337. https://doi.org/10.1371/journal.pntd.0005337
Editor: Paulo Filemon Pimenta, Fundaçao Oswaldo Cruz, BRAZIL
Received: October 14, 2016; Accepted: January 18, 2017; Published: February 23, 2017
Copyright: © 2017 Moreno et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This research was funded by grant U19AI089681 to JMV and grant AI110112 to JEC from National Institutes of Health/National Institute of Allergy and Infectious Diseases (www.niaid.nih.gov). The Biodefense and Emerging Infectious Disease training fellowship grant T32AI05532901 provided partial support for CP. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The Human Blood Index (HBI), formerly known as the anthropophilic index or human blood ratio, is the proportion of recently-fed mosquitoes, usually vector species that have taken a human blood-meal . This index is a very important component of the formulae used to determine vectorial capacity and varies depending on mosquito species, collection area and season or time of collection . From an epidemiological standpoint, it is crucial to be able to accurately identify mosquito blood-meals for studies of transmission dynamics of viral and parasitic pathogens . For example, in Equatorial Guinea, the calculation of this index before and after indoor interventions to reduce malaria did not detect any mosquito behavioral differences, and researchers concluded that control strategies in this region were ineffective . In Central Kenya, anthropophily decreased in An. gambiae after the introduction of long lasting insecticide nets (LLINs) and zooprophylaxis . However, in southern Zambia, after two years of LLIN intervention, the main vector, Anopheles arabiensis, remained highly anthropophilic . In Tanzania the HBI showed a change in the main blood-source in An. arabiensis but not in An. funestus after the use of spatial repellent coils .
Another index to quantify host selection patterns is the incidence of multiple blood-meals from the same host species (cryptic) or from two or more different host species (patent) . Evidence that malarial mosquitoes take partial blood-meals from multiple hosts may be interpreted as interrupted blood-feedings that could increase the probability of both acquiring and transmitting Plasmodium . On the other hand, Burkot and colleagues  contend that fewer gametocytes would be ingested per meal, resulting in lower mosquito infection rates.
Anopheles darlingi, the primary regional malaria vector in the Amazon Basin, is anthropophilic in the Iquitos region , although both human biting rate (HBR) and entomological inoculation rate (EIR) vary widely  depending on the setting [13–15]. The An. darlingi feeding site in this region is exophagic and/or endophagic, depending on local circumstances (e.g., vegetation cover, type of house) and host availability [11, 12, 14,15].
In 2015, Loreto Department reported 95% of the total malaria cases in Peru (59,349 of 62,220 total) with Plasmodium vivax as the most prevalent human parasite followed by P. falciparum, with 46,924 and 12,425 cases, respectively . Parker and collaborators  demonstrated that high HBR, EIR, and infectivity of An. darlingi are a signature of remote riverine malaria hot spots and hyperendemicity in certain areas of the Peruvian Amazon, upending previous notions that transmission is hypoendemic throughout the peri-Iquitos region [11,12]. Recent studies also detected very high seasonal HBR and moderate EIR in the peri-Iquitos region [14, 15]. Most malaria cases occur during the rainy season, from December to June  and a correlation was detected between An. darlingi abundance and peak river levels, but there was no significant correlation between river level and malaria case numbers [12, 14, 15]. In this last study, mosquitoes positive for Plasmodium were collected in peridomestic areas within approximately 10 m of the main house entrance, (a caveat being that very few An. darlingi were found indoors despite extensive searching), suggesting that most malaria is transmitted exophagically, where humans have little protection against mosquito bites.
Despite being the dominant malaria vector in Amazonia, few studies have documented the blood-meal sources for An. darlingi. In Amapá state, Amazonian Brazil, an ELISA analysis found that 13.1% of blood-meals were human; most resting An. darlingi had fed on cattle, pigs and dogs . Notwithstanding the relatively low level of HBI, these communities are endemic for malaria, and An. darlingi is considered to be the most effective local vector . In Peru, no studies have been published on the identity of An. darlingi blood-meals, but potential non-human hosts in rural residences near Iquitos include common peridomestic animals, dogs and chickens, and several potential wild mammalian hosts .
Although resting mosquitoes are optimal for calculating HBI, adequate sample sizes can be difficult to obtain in some habitats [18–20]. Little information exists on host preference and resting behavior of An. darlingi. The location of resting sites of An. darlingi could be useful for focal vector control if such mosquitoes are clustered non-randomly in the landscape. The development of barrier screens as a method for collecting anophelines outdoors has been tested successfully in the South East Pacific  and recently in southern Zambia .
This study was designed to address the following questions regarding An. darlingi feeding behavior in the Peruvian Amazon: i) are barrier screens a useful tool to collect resting blood-fed An. darlingi in the area; ii) what is the degree of anthropophily (HBI) in An. darlingi in contrast to more opportunistic behavior; iii) what is the influence of available host biomass and iv) is there evidence of seasonal age-structure in An. darlingi.
This study was approved by the Human Subjects Protection Program of the University of California San Diego, La Jolla, California and by the Ethical Boards of Universidad Peruana Cayetano Heredia and Asociación Benéfica PRISMA, Lima, Peru.
The strategy of the barrier screen method of collecting mosquitoes outdoors is to intercept and capture mosquitoes transiting between blood feeding and resting sites . Two possible scenarios can be identified: 1) intercepting mosquitoes entering a village seeking a blood-meal after emergence or oviposition; and 2) intercepting blood-fed mosquitoes leaving the village and seeking resting sites for egg development (swamp, creek, stream, forest). In this Peruvian study, barrier screens were placed to intercept mosquitoes flying between house-forest and house-river depending on the specific characteristics of the locality. Mosquito collections were performed in three villages in Loreto Department: Lupuna (LUP) and Cahuide (CAH) in the peri-Iquitos area, and Santa Emilia (SEM), in a remote area ~150 km from Iquitos (Fig 1). Detailed descriptions of these villages are in [15, 22]. In 2013, from March to May, a pilot study was conducted using a single screen in LUP and CAH placed at different points within each village (between the creek/river and village houses). Specimens were collected for 4 nights (6PM- 6AM) each month.
Each barrier screen was constructed from a lightweight window screen mesh approximately 15 m long and 2 m high (S1 Fig). Screens were then attached to poles with thin wire. Permission from the inhabitants/owners was obtained prior to any activity, including setting up the barrier screens and performing mosquito collections. Resting mosquitoes from the barrier screens were sampled by manually searching the surface of the screens with a mouth aspirator every hour for 15 minutes on each side, and the location (next to house, forest or river) and height (˃ or ˂ 1m above ground) of mosquitoes was recorded. Mosquitoes were captured and stored by hour of collection and screen side separately. In 2014 (monthly) and 2015 (January-June), the design was slightly modified to include four barrier screens in LUP and CAH to better represent the An. darlingi population in each locality. When multiple screens were used per village, data from each screen was maintained separately. In SEM, a remote village along the Nahuapa River, collections were performed with two barrier screens for two nights in May-June 2014 and May-September 2015. Additionally, in 2015, daytime mosquito collections (6AM-6PM) with barrier screens were performed two days monthly from January-June in LUP and CAH, and from May-July in SEM. Screen orientation, wind speed and direction were recorded for every collection with a Windmate 300 Wind/Weather Meter. A census questionnaire of domestic hosts present in the study villages was performed in October 2014 in LUP and CAH and May 2015 in SEM (S1 Table, Fig 2). Because the first study was performed a year prior and the animal composition could have changed, the questionnaire included a retrospective question to assess the presence of potential past hosts.
Additional animals seen frequently by the inhabitants were rats, toads, snakes and wild rodents.
All specimens collected were morphologically identified using entomological keys [23–25] and abdominal status recorded (unfed, blood-fed or gravid). Mosquitoes were stored and labeled individually with silica gel and placed at 4°C until subsequent analysis.
Estimation of parity and daily survival rate
To estimate the female age composition of the population, in March-April 2014 and February-June 2015 in LUP and CAH a proportion of females were dissected to determine the parity rates per hour, trap and side of trap . Parity is also used as an indicator of mosquito survival under natural conditions. Mosquito longevity (life expectancy) was estimated using Davidson’s methodology (1954) , where is the natural logarithm of the constant P (daily survival rate). (P) was calculated P = , where PR is the ratio of parous mosquitoes and the total number of females dissected, and gc is the duration of the gonotrophic cycle in days . A limitation of this calculation is the assumption of accurate estimates of the length of the gonotrophic cycle. We have assumed that two or more blood-meals are required for the first oviposition and that the temporal feeding pattern is not regular, and therefore, we followed the method of calculations proposed by Garret-Jones and Grab . Various studies have estimated the gonotrophic cycle of An. darlingi to be 2–3 days [29, 30, respectively]. Recently, it was calculated to be 2.19 days in the rainy season and 2.43 in the dry season . Calculations in our study were performed using the 2.19 day estimate based on the timing of our An. darlingi collections (the rainy season).
Individual An. darlingi were bisected between the head/thorax and abdomen and DNA was extracted manually using the DNeasy Blood & Tissue kit (Qiagen). A PCR-RFLP protocol was performed to detect the most common host in the area  for all mosquito abdomens in 2013–2015, except for a subsample (60%) of mosquitoes collected in LUP 2014 (due to a extended sample size). In addition, samples were tested for Galliformes (Gallus gallus and turkeys; see census and proportion of chickens; Fig 2, S1 Table) following , rat and didelphis , and monkey . A subsample of the unidentified blood samples was sequenced for the mitochondrial COI gene  and then compared with sequences in GenBank using BLASTn (http://www.ncbi.nmln.nih.gov) or BOLD SYSTEMS v2.5 (http://www.barcodinglife.org). The best match with identity of 95% or above was recorded.
Detection of Plasmodium infection was conducted using real-time PCR of the small subunit of the 18S rRNA, with a triplex TaqMan assay (Life Technologies), as described in . First, DNA was extracted from each specimen of An. darlingi, then the RT-PCR was conducted on pools of DNA of head/thoraces of five mosquitoes, and finally the pools were analyzed for detection of P. vivax and P. falciparum. Specimens from positive pools were tested individually to calculate infection rate (IR).
HBI was calculated as the proportion of mosquitoes fed on a specific host divided by the number of mosquitoes analyzed (mixed blood-meals were added to totals of each host). To adjust the HBI, mosquitoes with unidentified blood-meals were excluded. This index was calculated monthly in each locality and Chi-square (χ2) analyses were performed to compare statistical differences temporally and among sites. Host data recorded in the census was used for the calculation of the forage ratio (wi) [38, 39] and selection index (Bi) , to quantify the preference of mosquitoes for available blood resources. The forage ratio for species i was calculated as , where oi is the proportion of host species i in the blood-meals, and pi is the proportion of available host in the environment. Forage ratios >1.0 indicate preference and < 1.0 avoidance and selection of another host; ~1.0 means neither preference nor avoidance. The selection index Bi was calculated with the formula , where wi is the forage ratio for species i and n is the number of blood sources available.
Wind speed was measured at 6:00pm, 12:00am, and 6:00am each collection night in LUP, CAH, and SEM in 2015. For each collection night, mosquito density was aggregated into four 3-hour collection periods (6-9pm, 9pm-12am, 12-3am, and 3-6am). The wind speed at 6:00pm was assigned to the 6-9pm collection time, the wind speed at 12:00am was assigned to the 9pm-12am and 12-3am collection times, and the wind speed at 6:00am was assigned to the 3-6am collection time. The mosquito density was plotted against wind speed for each collection period at each location (n = 48 collection periods each for LUP and CAH, and 40 collection periods for SEM) using the ggplot2 package in RStudio v0.98.1091 .
A null-model analysis was used to test whether An. darlingi feeding habits were random or structured among the three villages, as in  and . All specimens with identified blood-meals from 2013–2015 for LUP, 2013–2015 for CAH, and 2014–2015 for SEM were included, and specimens with mixed blood-meals were counted once for each host identified in the blood-meal. We calculated a C-score comparing the blood- meal sources of mosquitoes from the three villages using Ecosim 7.0 and we used the R bipartite package  to generate a host-vector quantitative interaction network for the three localities, as in .
Barrier screen mosquito collections
In 2013, all specimens caught on the screens were collected and identified to determine the potential use of screens for collecting not only Anophelinae but also other Culicidae, potential vectors of parasites and arboviruses. A total of 322 mosquitoes in LUP and 514 in CAH were collected in 6 nights (72 h collection) (Table 1); 94.4% (304/18) of mosquitoes collected in LUP and 89.7% (461/53) of all mosquito species in CAH were females. Anopheles darlingi comprised 78.9% and 61.5% of these collections in LUP and CAH, respectively, and Culex quinquefasciatus was the second most common species identified in both localities (Table 1). Only one additional species of anopheline, Anopheles forattini, was identified (in LUP).
With respect to screen position, in LUP 63.4% of the An. darlingi were collected on the side facing the houses (In) and 36.6% on the side facing the creek (Out), although this difference was not significant (Kolmogorov-Smirnov test; p = 0.4). On both sides of the screen, most of the specimens were collected <1m from the ground (Below; Table 2) (range 76.5–90.2%). In CAH, 61.8% of the mosquitoes were collected on the house side and 38.2% on the creek side, and 93.1% and 84.5% (In and Out, respectively) were caught <1 m from the ground. No differences were found between LUP and CAH for side of the barrier screen. Only 1.62% in LUP and 6.57% in CAH of the An. darlingi females were determined by visual inspection to be blood-fed, with no differences between screen sides (Table 3).
In 2014, using multiple barrier screens per locality, a total of 4,593 An. darlingi females were collected in LUP, 175 in CAH and 216 in SEM (Table 2). One specimen of Anopheles dunhami in LUP and eighteen Anopheles benarrochi B in SEM were also identified as in . In LUP, no significant differences were detected between the sides of four screens tested independently. However, when data were grouped over months there was a significant difference between mosquitoes collected on the side of the houses (In) and creek/vegetation side (Out) (Wilcoxon test; p = 0.0313). In CAH, the four barrier screens were not homogeneous, with significant differences in number of mosquitoes collected from each side (K-S; In: p = 0.0082 and Out: p = 0.0356), and when In/Out were compared by month (K-S; p = 0.0022). There were also significant differences between collections in LUP and CAH (K-S, p = 0.0336). In SEM, captures in May (two screens) and in June (four screens), were not significantly different between screens.
In 2015, in LUP, 1,019 female mosquitoes were collected, 233 in CAH and 277 in SEM. Most specimens were captured resting < 1m from the ground with little variation among years and sites (Table 2).
Differences in mosquito density by time of collection and side of barrier screen were tested (Fig 3) with time of collection split into four three-hour periods (6-9pm, 9pm-12am, 12-3am, and 3-6am). In both LUP and CAH in 2015, there was a significant difference in the distribution of mosquito collection location (side of screen) by time period (Kruskal-Wallis p < 0.0001 for both sites), with higher proportions of mosquitoes found on the In (facing house) side of the screen from 9pm-12am and 12-3am than from 6-9pm and 3-6am. In LUP and CAH in 2013 and 2014, and in SEM in 2015, there was no significant difference in mosquito density by time of collection (Kruskal-Wallis p>0.05).
*Significant difference in the distribution of mosquito collection location by time period (Kruskal-Wallis p < 0.0001).
Plots of mosquito density against wind speed for each locality in 2015 are shown in Fig 4. Overall, there was a negative but non-significant correlation between mosquito density and wind speed (Pearson’s r = -0.09, p = 0.3). The correlation between mosquito density and wind speed was also negative in LUP (Pearson’s r = -0.25, p = 0.1) and SEM (Pearson’s r = -0.27, p = 0.09), but was positive in CAH (Pearson’s r = 0.14, p = 0.34) (Fig 4).
Mosquitoes were collected from 6pm-6am from January-June 2015 in CAH and LUP and May-September 2015 in SEM. Linear regression of mosquito density on wind speed shown for each location (CAH: Pearson’s r = 0.14, p = 0.34; LUP: Pearson’s r = -0.25, p = 0.1; SEM: Pearson’s r = -0.27, p = 0.09).
To investigate the diurnal behavior of An. darlingi, barrier screen collections were performed in LUP and CAH from January to June, and in SEM from May to June from 6AM to 6PM twice January-June 2015. In LUP a total of 59 An. darlingi were collected during this period and female activity was reported from 6AM to 9AM and from 2PM to 5PM. In CAH, the number of collected specimens was 23, with an activity similar to LUP. In SEM, 33 mosquitoes were collected, with an extension of the flying activity until 8AM, and beginning again in the evening at 4PM. In LUP, 20.3%, in CAH, 34.8% and in SEM 54.5% of diurnal An. darlingi specimens were collected on the house side (In).
Variation in parity and daily survival rate
A total of 583 An. darlingi females from LUP were dissected in 2014 (12% of the total) and 19 in CAH (11%); in 2015, n = 633 in LUP (62%) and n = 153 (65%) in CAH were dissected. The monthly mean parity rate in LUP in 2015 was ~ 55% (range 45.6–66.7) and in CAH it was ~ 51% (range 27.8–64.5) (Table 4). No significant differences were found between months or between localities, although in February, the rate was slightly higher compared to June. Mosquito age in LUP in March—April 2014 was 7.47 and 14.21 days, respectively, whereas in 2015 it ranged from 14.21–23.90 days. In CAH, mosquitoes collected in March 2014 were estimated to survive 14.98 days, and between 3.73–20.24 days in 2015 (Table 4).
Blood-meal source identification
Blood-meal source was determined for 4,417 An. darlingi females (S2 Table). A total of 3,214 mosquitoes from LUP, 729 from CAH and 474 from SEM were analyzed. Single-host blood-meals were the highest percentage among the blood-meals detected (69.98%) and human was the most common blood source (42.5%), followed by Galliformes (25.1%) and dog (1.42%; Fig 5). Only 4% of the samples could not be identified to blood-meal source. Multiple blood-meals were found in 1,272 mosquitoes and accounted for 30% of the blood- meals, with 1,262 double feeds in the three localities, and triple feeds (n = 10) only identified in LUP.
In total, seventy-three samples with non-identified blood-meal source by PCR-RFLP, were sequenced for 16S ribosomal DNA  and mammalian cytochrome-b . Only ten were identified as of human origin with the 16S protocol, whereas 23 were consistent with human for cytochrome-b.
The distribution of blood-meal source in An. darlingi presented little temporal or spatial variation. Evaluation of the proportion of feeds on single different hosts showed that in LUP, no significant differences between years were detected by one-way ANOVA analysis; paired Wilcoxon-tests were not significant when comparing years 2013–2014 with 2015 or 2013 and 2014. In CAH, no significant differences between the years 2013–2014, 2014–2015 or among the 3 years were found. In SEM, a non-parametric Mann-Whitney test was not significant comparing 2014 and 2015. For locality comparison, data from the same years and different localities were compared. In 2013, there were no significant differences between LUP and CAH, and in 2014 and 2015 a one-way ANOVA test did not show differences between sites.
HBI was calculated monthly (S3 Table) and annually (Table 5) per locality. In 2013, no significant differences were detected in LUP or CAH. Mean HBI per year was non-significant among localities (LUP, CAH, SEM) and years 2014–2015.
The Forage Ratio and Host Selection Index were calculated, accounting for single and multiple blood-meals (Table 6). Humans were the preferred source, closely followed by Galliformes, in all three settings for both years. When the Forage Ratio was analyzed, the weight per host was used instead of the numerical presence at the site  (S4 Table), Galliformes were by far the preferred host, with humans as the second most favoured. For example in LUP, the Galliforme forage ratio ranged from 10.35 to 17.96 and the human forage ratio from 0.58–0.72. The null model test indicated that the mosquito feeding patterns were aggregated among the localities, indicating that diet overlapped more than expected between the localities, although this finding was only marginally significant (C-score: 0.33, p = 0.08). The quantitative interaction network of blood-meal source by locality (Fig 6) supported patterns of organization based on the above-mentioned trophic preferences (humans and Galliformes) from the three mosquito populations (LUP, CAH, SEM).
Network is based on the analysis of blood-meal source for 4,417 An. darlingi females collected from 2013–2015.
Plasmodium mosquito infection
A total of 5,387, 362 and 455 mosquitoes in LUP, CAH and SEM, respectively, collected on barrier screens, were tested for Plasmodium. The Infection rate (IR) of mosquitoes varied among sites and seasons, ranging from 0.20–3.85 in LUP, 0.51–14.3 in CAH and 0–2.04 in SEM (Table 7). A logistic regression model analysis determined that IR was significantly higher in CAH (p = 0.02) and SEM (p = 0.003) vs. LUP. No specimens from the diurnal collections in the three localities (n = 116) were positive for P. vivax, independent of the collection season.
Ours is the first study to conclusively demonstrate that An. darlingi readily feeds on Galliformes. Overall, the feeding preference of An. darlingi in the Peruvian Amazon is more variable than previous studies have assumed. In addition, a consistent pattern of blood-meal source was observed at each site every year of collection: mosquitoes feeding only on humans, only on chickens, or on both hosts. This consistency could suggest the co-occurrence of different subpopulations within a metapopulation, with local adaptation as the main driving force.
A single metapopulation was initially detected in An. darlingi in the Iquitos area with AFLPs  and microsatellite markers . However, using 2x the number of microsatellites, a population replacement event was detected between 2006 and 2012 and two subpopulations were detected, one significantly more prevalent in highway compared with riverine habitat . This recent genetic structure could explain some of the heterogeneity in feeding preferences of An. darlingi among localities [45, 46]. Additional studies, focused on intrinsic host preference, vector density and social practices of the human population might elucidate the basis for the described behavior and whether some An. darlingi populations are under selective pressure for host preference or whether this pattern is strongly correlated with host availability.
Similar HBI across the dry and rainy seasons and between populations infers that mosquitoes maintain their host preference behavior independent of local ecological conditions. In an earlier investigation of HBI of An. darlingi in riverine villages in Amapá State, Brazil , researchers reported high among-village variance (HBI 0.131–0.435) and ~10% of mixed blood-meals overall, mainly from cattle and pigs. In contrast, in our study, there was virtually no variance in HBI among localities, HBIs were higher (0.58–0.79) and ~30% of blood-meals were mixed, with Galliformes as the primary alternate host. Because HBI is an integral parameter of the vectorial capacity formula (the daily rate of malaria transmission from a single infected human, assuming every bite from an infected mosquito leads to transmission) , our data suggest that An. darlingi is a more effective vector in the peri-Iquitos area compared with Amapá state, Brazil. Curiously, in Tanzania, An. arabiensis avoids, and may be repelled by, the volatiles of chickens . Subgenera Nyssorhynchus (An. darlingi) and Cellia (An. arabiensis) were estimated to have diverged ~94 million years ago ; therefore their olfactory responses are expected to have evolved differentially.
The present study provides evidence of the successful use of barrier screens to collect blood-fed An. darlingi mosquitoes in Amazonian Peru. Initially, in 2013, we conducted preliminary barrier screen collections with Procopack aspirators in LUP and CAH from 5 to 8 AM for 6 days/collection in March-May in at least 10 houses each time, but only one An. darlingi specimen was caught. Interestingly, in Iquitos the Procopack effectively collected indoor resting Culicidae including Aedes aegypti and Culex pipiens complex . One explanation for our failure to find An. darlingi using the Procopack despite extensive searching could be due to its singular resting and biting behavior in this region.
Anopheles darlingi resting behavior varies across its range : in Venezuela, Guyana  and in Brazil, in Matto Grosso and in southern Amazonas [52, 53] it rests during the day inside houses (endophily). In contrast, in Suriname, using exit traps, a peak departure from the dwelling was observed at sunrise  and in Brazil An. darlingi was resting indoors only at night . In Amapá state, Brazil, resting mosquitoes were collected after sunrise (6AM-7AM) under houses and in peridomestic vegetation . In French Guiana, no resting An. darlingi were collected indoors after pyrethroid spray, from pit-shelters or in the shade in the peridomestic area . In our study, overall differences detected between screen sides may reflect the relative nearness of screens to houses, resulting in the interception of a higher proportion of blood fed An. darlingi leaving the peridomestic area, compared with questing females, entering the village from numerous resting and/or breeding sites. In CAH, we hypothesize that additional differences among screens and between months could result from a much smaller population of An. darlingi intercepted in this village. Our results constitute a major accomplishment: the use of barrier screens in this setting to overcome the difficulty of performing host-independent sampling for determining blood-meal sources.
The success of individual mosquito blood-meal identification in this study (range of 92.2–99.3%), was remarkably high when compared to visually blood-fed mosquitoes (0.92%-14.44%). When analysis is restricted only to the latter, information from partial blood-meals or partially digested blood is missed, leading to underestimation of the proportion of host sources (up to 18.7%); hence, a miscalculation of HBI . One limitation of our study was the lack of identification of potential wild animal hosts; use of novel targeted high-throughput sequencing  would rectify this.
In LUP, the age of the mosquito population at each time point is enough to sustain the sporogonic cycle of P. vivax (range 7.24–9.13 days; calculated by the Moshkovsky method in ), whereas in CAH the population is, in general, younger, but with non-dangerously aged mosquitoes only in May and June. The proportion of young females might be explained by differential dispersal and aggregation of different age classes of An. darlingi populations, as previously reported for An. farauti in Papua New Guinea . Use of 2.19 days of the gonotrophic cycle  could have produced a miscalculation in the age parameter. For instance, gravid females may experience delays while searching for suitable oviposition sites or there could be variation in extrinsic environmental conditions within this population of An. darlingi . Because of the natural development of the parasite within the mosquito, a longer life-span is related to a higher potential to transmit malaria . Parity is also associated with seasonality, i.e., mosquitoes generally survive longer during the rainy season [62,63], but see .
Overall, our study provides unreported information of the blood-meal preferences of An. darlingi in the peri-Iquitos area, which will be the base-line to compare potential changes in the behavior of these mosquito populations. HBI, together with other malaria metrics such as HBR or EIR, should be taken into consideration for surveillance and epidemiological studies of malaria transmission.
Construction and set up of the barrier screens in Iquitos, Peru; (A, B, C, F): 2 m high and 15m long. Screens were examined hourly by flashlight and resting mosquitoes captured by aspiration (D, E).
S1 Table. Census of domestic and wild animals in the study localities 2013–2015.
Rats, toads, snakes and wild rodents were other animals frequently observed by the inhabitants.
S2 Table. Summary of An. darlingi blood-meal sources per year per locality.
S3 Table. Monthly variation of Human Blood Index (HBI) for An. darlingi in three sites.
S4 Table. Forage ratio (FR) of An. darlingi using host biomass.
Mean weight of hosts was: human (65kg), dog (25 kg), chicken (1.5kg), turkey (13.1 kg), pig (90kg), goat (45 kg).
S1 Dataset. Summary of mosquito collections by barrier screens methodology.
Collection site and dates of collection, mosquito species identification and side and height of the barrier screens are designated for each mosquito used in the analysis.
We would like to thank Eliseo Ramirez, José Manuel Reyna, Victor Pacaya, David Arimuya, and Hercules Maytahuari for their assistance in the field. We appreciate the enthusiastic support of the communities of Cahuide, Lupuna and Santa Emilia (Loreto Department). We are grateful to Dirección Regional de Salud (DIRESA, Iquitos, Loreto) for collaboration and facilitating logistics in Loreto. We are grateful to the Applied Genomic Technologies Core at the Wadsworth Center, New York State Department of Health for the sequencing of the samples.
- Conceptualization: MM JEC.
- Formal analysis: MM SAB CP.
- Funding acquisition: JEC JMV.
- Investigation: MM MPS SAB CP AM CTR JEC.
- Methodology: MM JEC SAB MPS CP.
- Resources: JEC JMV.
- Visualization: MM SAB CP JEC.
- Writing – original draft: MM JEC.
- Writing – review & editing: MM JEC MPS SAB CP AM CTR JMV.
- 1. World Health Organization. Terminology of malaria and of malaria eradication. 1963. Geneva.
- 2. Garrett-Jones C. The human blood index of malaria vectors in relation to epidemiological assessment. Bull World Health Organ. 1964;30: 241–61. pmid:14153413
- 3. Garrett-Jones C, Boreham PFL, Pant CP. Feeding habits of anophelines (Diptera: Culicidae) in 1971–78, with reference to the human blood index: a review. Bull Entomol Res. 1980;70: 165–185.
- 4. Pappa V, Reddy M, Overgaard HJ, Abaga S, Caccone A. Estimation of the Human Blood Index in Malaria Mosquito Vectors in Equatorial Guinea after Indoor Antivector Interventions. Am J Trop Med Hyg. 2011;84: 298–301. pmid:21292902
- 5. Kaburi JC, Githuto JN, Muthami L, Ngure PK, Mueke JM, Mwandawiro CS. Effects of long-lasting insecticidal nets and zooprophylaxis on mosquito feeding behaviour and density in Mwea, central Kenya. J Vector Borne Dis. 2009;46: 184–90. pmid:19724081
- 6. Fornadel CM, Norris LC, Glass GE, Norris DE. Analysis of Anopheles arabiensis blood feeding behavior in southern Zambia during the two years after introduction of insecticide-treated bed nets. Am J Trop Med Hyg. 2010;83: 848–53. pmid:20889878
- 7. Maia MF, Kreppel K, Mbeyela E, Roman D, Mayagaya V, Lobo NF, et al. A crossover study to evaluate the diversion of malaria vectors in a community with incomplete coverage of spatial repellents in the Kilombero Valley, Tanzania. Parasit Vectors. 2016;Aug 15;9:451. pmid:27527601
- 8. Boreham PF, Garrett-Jones C. Prevalence of mixed blood-meals and double feeding in a malaria vector (Anopheles sacharovi Favre). Bull World Health Organ. 1973;48: 605–14. pmid:4544148
- 9. Garrett-Jones C. Prognosis for interruption of malaria transmission through assessment of the mosquito's vectorial capacity. Nature 1964;204: 1173–5. pmid:14268587
- 10. Burkot TR, Graves PM, Paru R, Lagog M. Mixed blood feeding by the malaria vectors in the Anopheles punctulatus complex (Diptera: Culicidae). J Med Entomol. 1988;25: 205–13. pmid:3404539
- 11. Vittor AY, Gilman RH, Tielsch J, Glass G, Shields T, Sanches Lozano W, et al. The effect of deforestation on the human-biting rate of Anopheles darlingi, the primary vector of falciparum malaria in the Peruvian Amazon. Am J Trop Med Hyg. 2006;74: 3–11. pmid:16407338
- 12. Reinbold-Wasson DD, Sardelis MR, Jones JW, Watts DM, Fernandez R, Carbajal F, et al. Determinants of Anopheles seasonal distribution patterns across a forest to periurban gradient near Iquitos, Peru. Am J Trop Med Hyg. 2012;86: 459–63. pmid:22403317
- 13. Parker BS, Paredes Olortegui M, Penataro Yori P, Escobedo K, Florin D, Rengifo Pinedo S, et al. Hyperendemic malaria transmission in areas of occupation-related travel in the Peruvian Amazon. Malar J. 2013;12: 178. pmid:23724869
- 14. Turell MJ, Sardelis MR, Jones JW, Watts DM, Fernandez R, Carbajal F, et al. Seasonal distribution, biology, and human attraction patterns of mosquitoes (Diptera: Culicidae) in a rural village and adjacent forested site near Iquitos, Peru. J Med Entomol. 2008;45: 1165–72. pmid:19058644
- 15. Moreno M, Saavedra MP, Bickersmith SA, Lainhart W, Tong C, Alava F, et al. Implications for changes in Anopheles darlingi biting behaviour in three communities in the peri-Iquitos region of Amazonian Peru. Malar J. 2015;30;14:290.
- 16. MINSA (Ministerio de Salud del Peru). Sala de Situacion de Salud SE 32. Boletin Epidemiologico. 2016. Available from: http://www.dge.gob.pe/portal/docs/vigilancia/boletines/2016/32.pdf
- 17. Chuquiyauri R, Paredes M, Penataro P, Torres S, Marin S, Tenorio A, et al. Socio-demographics and the development of malaria elimination strategies in the low transmission setting. Acta Trop. 2012;121: 292–302. pmid:22100446
- 18. Zimmerman RH, Galardo AK, Lounibos LP, Arruda M, Wirtz R. Bloodmeal hosts of Anopheles species (Diptera: Culicidae) in a malaria-endemic area of the Brazilian Amazon. J Med Entomol. 2006;43: 947–56. pmid:17017232
- 19. Galardo AK, Arruda M, D'Almeida Couto AA, Wirtz R, Lounibos LP, Zimmerman RH. Malaria vector incrimination in three rural riverine villages in the Brazilian Amazon. Am J Trop Med Hyg. 2007;76: 461–9. pmid:17360868
- 20. Burkot TR, Russell TL, Reimer LJ, Bugoro H, Beebe NW, Cooper RD, et al. Barrier screens: a method to sample blood-fed and host-seeking exophilic mosquitoes. Malar J. 2013;12: 49. pmid:23379959
- 21. Stevenson JC, Simubali L, Mbambara S, Musonda M, Mweetwa S, Mudenda T, et al. Detection of Plasmodium falciparum Infection in Anopheles squamosus (Diptera: Culicidae) in an Area Targeted for Malaria Elimination, Southern Zambia. J Med Entomol. 2016;.
- 22. Lainhart W, Bickersmith SA, Nadler KJ, Moreno M, Saavedra MP, Chu VM, et al. Evidence for temporal population replacement and the signature of ecological adaptation in a major Neotropical malaria vector in Amazonian Peru. Malar J. 2015;14:375. pmid:26415942
- 23. Forattini OP. Entomologia Medica. Faculdade de Higiene e Sáude Publica, São Paulo; 1962.
- 24. Faran ME, Linthicum KJ. A handbook of the Amazonian species of Anopheles (Nyssorhynchus) (Diptera: Culicidae). Mosquito System. 1981;13:1–81.
- 25. Consoli RA, Lourenco-de-Oliveira R. Principais mosquitos de importância sanitária no Brasil. Editora Fiocruz, Fundação Oswaldo Cruz; 1994.
- 26. Detinova T. Age-grouping methods in Diptera of medical importance, with special reference to some vectors of malaria. Monogr Ser World Health Organ. 1962;47:13–191. pmid:13885800
- 27. Service MW. Mosquito Ecology: Field Sampling Methods. Elsevier Science Publ., Essex, 988 pp. 1993.
- 28. Garrett-Jones C, Grab B. The assessment of insecticidal impact on the malaria mosquito's vectorial capacity, from data on the proportion of parous females. Bull World Health Organ. 1964;31:71–86. pmid:14230896
- 29. Roberts DR, Alecrim WD, Tavares AM, McNeill KM. Field observations on the gonotrophic cycle of Anopheles darlingi (Diptera: Culicidae). J Med Entomol. 1983;20:189–192. pmid:6842526
- 30. Charlwood JD, Wilkes TH. Studies on the age-composition of samples of Anopheles darlingi Root (Diptera: Culicidae) in Brazil. Bull Entomol Res. 1979;69:337–342.
- 31. Barros FSMd Honório NA, Arruda ME. Survivorship of Anopheles darlingi (Diptera: Culicidae) in Relation with Malaria Incidence in the Brazilian Amazon. PLoS ONE. 2011;6(8): e22388. pmid:21857927
- 32. Fornadel CM, Norris DE. Increased endophily by the malaria vector Anopheles arabiensis in southern Zambia and identification of digested bloodmeals. Am J Trop Med Hyg. 2008;79(6):876–80. pmid:19052296
- 33. Ngo KA, Kramer LD. Identification of mosquito bloodmeals using polymerase chain reaction (PCR) with order-specific primers. J Med Entomol. 2003;40(2):215–22. pmid:12693851
- 34. Ribeiro G Jr, Gurgel-Gonçalves R, Reis RB, Santos CG, Amorim A, Andrade SG, et al. Frequent house invasion of Trypanosoma cruzi-infected triatomines in a suburban area of Brazil. PLoS Negl Trop Dis. 2015;Apr 24;9(4)
- 35. de Carvalho GC, Malafronte Rdos S, Miti Izumisawa C, Souza Teixeira R, Natal L, Marrelli MT. Bloodmeal sources of mosquitoes captured in municipal parks in São Paulo, Brazil. J Vector Ecol. 2014;Jun;39(1):146–52. pmid:24820567
- 36. Navia-Gine WG, Loaiza JR, Miller MJ. Mosquito-Host Interactions during and after an Outbreak of Equine Viral Encephalitis in Eastern Panama. PLoS ONE. 2013;8(12): e81788. pmid:24339965
- 37. Bickersmith SA, Lainhart W, Moreno M, Chu VM, Vinetz JM, Conn JE. A sensitive, specific and reproducible real-time polymerase chain reaction method for detection of Plasmodium vivax and Plasmodium falciparum in field-collected anophelines. Mem Inst Osw Cruz. 2015;110(4):573–6.
- 38. Hess AD, Hayes RO, Tempelis CH. The use of the forage ratio technique in mosquito host preference studies. Mosq News. 1968;28:386–389.
- 39. Savage RE. The relation between the feeding of the herring off the east coast of England and the plankton of the surrounding waters. Fishery Investigation, Ministry of Agriculture, Food and Fisheries. 1931; Series 212:188.
- 40. Manly BFJ, McDonald LL, Thomas DL. Resource selection by animals. Statistical design and analysis for field studies London, Chapman and Hall; 1993.
- 41. R Development Core Team. R. A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2008. ISBN 3-900051-07-0, URL http://www.R-project.org.
- 42. Chaves LF, Harrington LC, Keogh CL, Nguyen AM, Kitron UD. Blood feeding patterns of mosquitoes: random or structured? Front Zool. 2010;Jan 21;7:3. pmid:20205866
- 43. Dormann CF, Fründ J, Blüthgen N, Gruber B. Indices, graphs and null models: analyzing bipartite ecological networks. Open Ecol. 2009;J 2: 7–24.
- 44. Pinedo-Cancino V, Sheen P, Tarazona-Santos E, Oswald WE, Jeri C, Vittor AY, et al. Limited diversity of Anopheles darlingi in the Peruvian Amazon region of Iquitos. Am J Trop Med Hyg. 2006; 75: 238–45. pmid:16896125
- 45. Mirabello L, Vineis JH, Yanoviak SP, Scarpassa VM, Povoa MM, Padilla N, et al. Microsatellite data suggest significant population structure and differentiation within the malaria vector Anopheles darlingi in Central and South America. BMC Ecol. 2008;8:3. pmid:18366795
- 46. Oliveira-Ferreira J, Lourenço-de-Oliveira R, Deane LM, Daniel Ribeiro CT. Feeding preference of Anopheles darlingi in malaria endemic areas of Rondônia state—Northwestern Brazil. Mem Inst Oswaldo Cruz. 1992;87: 60l–602
- 47. Jaleta KT, Hill SR, Birgersson G, Tekie H, Ignell R. Chicken volatiles repel host-seeking malaria mosquitoes. Malar J. 2016 Jul 21;15(1):354. pmid:27439360
- 48. Moreno M, Marinotti O, Krzywinski J, Tadei WP, James AA, Achee NL, Conn JE. Complete mtDNA genomes of Anopheles darlingi and an approach to anopheline divergence time. Malar J. 2010 May 14;9:127. pmid:20470395
- 49. Vazquez-Prokopec GM, Galvin WA, Kelly R, Kitron U. A new, cost-effective, battery-powered aspirator for adult mosquito collections. J Med Entomol. 2009;46(6):1256–1259. pmid:19960668
- 50. Hiwat H, Bretas G. Ecology of Anopheles darlingi Root with respect to vector importance: a review. Parasit Vectors. 2011;4: 177. pmid:21923902
- 51. Giglioli G. Biological variations in Anopheles darlingi and Anopheles gambiae; their effect on practical malaria control in the neotropical region. Bull World Health Organ. 1956;15(3–5):461–71. pmid:13404433
- 52. Bustamante FM. Considerações sôbre sertos problemas especiais relacionados com a erradicação da malária no Brasil. Rev Bras de Malariol Doenças Trop. 1959;11:9–17.
- 53. Roberts DR, Alecrim WD, Tavares AM, Radke MG. The house frequenting, host-seeking and resting behavior of Anopheles darlingi in Southeastern Amazonas, Brazil. J Am Mosq Control Assoc. 1987;3: 433–441. pmid:3504928
- 54. Rozendaal JA. Biting and resting behavior of Anopheles darlingi in the Suriname rainforest. J Am Mosq Control Assoc. 1989;5: 351–358. pmid:2584968
- 55. Deane LM, Causey OR, Deane MP. Notas sôbre a distribuição e a biologia dos anofelinos das regiões nordestina e amazônica do Brasil. Rev Serv Espec Saude Publ. 1948;1: 827–965.
- 56. Girod R, Gaborit P, Carinci R, Issaly J, Fouque F. Anopheles darlingi bionomics and transmission of Plasmodium falciparum, Plasmodium vivax and Plasmodium malariae in Amerindian villages of the Upper-Maroni Amazonian forest, French Guiana. Mem Inst Oswaldo Cruz. 2008;103(7): 702–10. pmid:19057822
- 57. Das S, Henning TC, Simubali L, Hamapumbu H, Nzira L, Mamini E, et al. Southern Africa ICEMR Team. Underestimation of foraging behaviour by standard field methods in malaria vector mosquitoes in southern Africa. Malar J. 2015 Jan 21;14:12. pmid:25927429
- 58. Logue K, Keven JB, Cannon MV, Reimer L, Siba P, Walker ED, et al. Unbiased Characterization of Anopheles Mosquito Blood Meals by Targeted High-Throughput Sequencing. PLoS Negl Trop Dis. 2016 Mar 10;10(3):e0004512. pmid:26963245
- 59. Charlwood JD, Birley MH, Dagoro H, Paru R, Holmes PR. Assessing survival rates of Anopheles farauti (Diptera: Culicidae) from Papua New Guinea. J Anim Ecol. 1985;54:1003–1016.
- 60. Gillies MT. Methods for assessing the density and survival of blood-sucking Diptera. Annu Rev Entomol. 1974;19:345–62. pmid:4592890
- 61. Warrell DA, Gilles HM, Essential Malariology, Arnold, London, UK, 4th edition, 2002.
- 62. Charlwood JD. Survival rate variation of Anopheles farauti (Diptera: Culicidae) between neighboring villages in coastal Papua New Guinea. J Med Entomol. 1986;23: 361–365. pmid:3735339
- 63. Birley MH, Charlwood JD. The effect of moonlight and other factors on the ovipositon cycle of malaria vectors in Madang, Papua New Guinea. Ann Trop Med Parasitol.1989;8: 415–422.
- 64. León W, Valle J, Naupay R, Tineo E, Rosas A, Palomino M. Comportamiento estacional del Anopheles (Nyssorhynchus) darlingi Root 1926 en localidades de Loreto y Madre de Dios, Peru 1999–2000. Rev Peru Med Exp Salud Pública. 2003;20:22–7.