Introduction
The intensity of malaria and risk factors associated with transmission is distributed in a highly uneven manner across the globe. Knowledge of malaria entomological parameters is considered a prerequisite for planning effective control interventions (WHO, 2015). Hence, knowing the risk factors of malaria transmission, even at a very local scale is essential. The distribution, abundance, biology and bionomics of Anopheles fauna and their respective trophic behaviors are important factors affecting the risk of malaria transmission (Tchuinkam et al., Reference Tchuinkam, Simard, Lele-Defo, Tene-Fossog, Tateng-Ngouateu, Antonio-Nkondjio, Mpoame, Toto, Njine, Fontenille and Awono-Ambene2010; Tainchum et al., Reference Tainchum, Ritthison, Chuaycharoensuk, Bangs, Manguin and Chareonviriyaphap2014). The intensity of malaria transmission is commonly evaluated in terms of entomological inoculation rate (EIR), the product of the human-biting rate and the proportion of mosquitoes infected with sporozoites (Amek et al., Reference Amek, Bayoh, Hamel, Lindblade, Gimnig, Odhiambo, Laserson, Slutsker, Smith and Vounatsou2012; Abraham et al., Reference Abraham, Massebo and Lindtjorn2017). However, due to high cost, resources and strict ethical limitations, accurate estimation of EIR is rarely documented (Tusting et al., Reference Tusting, Bousema, Smith and Drakeley2014). To solve these limitations, other parameters were progressively incorporated which can efficiently serve as early warning predictors for malaria. Amongst them sporozoite carrier and human blood preference vectors are essential indices for a proxy measurement of malaria transmission (Burkot et al., Reference Burkot, Dye and Graves1989; Kent et al., Reference Kent, Thuma, Mharakurwa and Norris2007). An alternative method to estimate EIR requires accurate detection of infective anthropophagic vector species which can be achieved through a systematic and sequential vector collection and processing without any sampling bias. Apart from the sporozoite carrier, human blood-feeding characteristics of anopheles vectors also contribute indirectly to EIR. These two factors, to our knowledge, are part of the same question: is human blood-feeding behavior by vectors a critical component of malaria transmission? Even though the vector mosquito feeds on human, it cannot be considered a potential malaria transmitter unless it can transmit the plasmodial sporozoite. The aim of this study was to assess how malaria transmission is influenced and can be predicted by local major vectors dynamics. In this study, we perform an entomological assessment on major Anopheline mosquito indicators including abundance, habitat preference, resting and feeding behavior, infectivity rates, and other entomological parameters. We use these indicators to measure the differences in malaria transmission between a vector control intervention and non-intervention scenario in a high endemic region of Kalahandi district of Odisha, India. The ultimate purpose of this work was to prioritize and integrate these key entomological parameters as an early warning system for transmission study in malaria endemic regions.
Materials and methods
Study area and period
The study was conducted over a 2-year period (2016 and 2017) in Belgaon sub-center of Kalahandi district which is characterized by hilly and dense forest area. Kalahandi district covers an area of 7920 km2 and is situated in the south western region of Odisha, between latitude19°3′N and 21°5′N and longitude 82°30′E and 83°74′E. The sub-center is comprised of 22 villages in total, of which 15 villages were assessed during the present survey. The district is hyper endemic for malaria and is rich in two major vector species, Anopheles culicifacies (Giles s.l) and Anopheles fluviatilis (James) (Sharma et al., Reference Sharma, Upadhyay, Haque and Subbarao2004). Deltamethrin 62.5% IRS (Indoor residual spray; >85% coverage) was introduced in Belgaon in 2017, during July which is the monsoon period in the state. Malaria transmission in Kalahandi district is perennial with one sharp peak following the rainy season. Hence, we focused on the wet season (i.e. pre-monsoon, monsoon, and post-monsoon periods).
Entomological collection
Entomological collections were conducted on a monthly basis from January 2016 to December 2017, where 2016 was denoted as the non-intervention period, and 2017 as the IRS intervention period. During the first 2 months, houses were arbitrarily chosen as there was no prior information on vector abundance. The coordinates of each house were recorded using differential GPS. Both indoor and outdoor mosquito collections were carried out twice per day in each village of Belgaon sub-center during the morning (5:00–8:00 am) and evening (6:30–9.00 pm). A total of 15 houses from each village were randomly selected. Collections were performed for a period of 15 days per month on every alternate day, by 6–7 trained insect collectors. Outdoor resting mosquitoes were collected from a variety of environments including dry pots, under the eaves of huts, water pipes, undersides of bridges, tree bases, tree holes, piles of fallen leaves, cracks and holes in the ground, small ridges under rocks and granaries. The indoor resting mosquito collection was carried out from cattle sheds (CS) and human dwellings (HD). Mosquitoes were collected for 15 min from each habitat using an oral aspirator and a flash light. Human landing collection was omitted due to ethical issues, and instead, alternative methods to evaluate entomological parameters as accurately as the human landing catch were introduced. These included:
Chemical method: Indoor resting mosquitoes were collected using the pyrethrum spray catch between the hours 05:00 and 09:00 am.
Physical/Mechanical method: Mosquitoes from houses were collected by window exit traps set at 19.00 and collected the next morning between 5.00 and 6.00. Three CDC light traps were run simultaneously, positioned a few meters away from the bed where dwellers slept protected by bed nets. These were set at 19.00 h and collected at 6.00 the following morning in three selected houses.
Biological method: Instead of using humans as an attractant, outfits having body odor as an attractant such as shirts, smelly socks, and worn clothes were used to attract anthropophilic Anopheline species in the selected houses (with no inhabitants) of each village. These were used as ‘mosquito magnets’ to collect all human attracting/anthropophilic vectors.
Female Anopheles mosquitoes collected using these methods were collected and monthly averages were calculated over all study houses in each village.
Mosquito identification
Mosquitoes were first sorted according to their generic groups. The Anopheles mosquitoes were identified taxonomically following the key developed by Barraud Reference Barraud1934 (The fauna of British India). Female An. culicifacies were sorted according to their gonotrophic status including unfed (U), blood-fed (F), semi-gravid (SG/HG), or gravid (G). The gravid/semi-gravid appearance of abdomens demonstrates the resting stage, while those fully fed and unfed guts are of seeking stage (WHO, 1975). The ratio of resting to seeking stages for An. culicifacies showed their resting behavior.
Human blood meal and sporozoite detection in An. culicifacies
Each individual mosquito was dissected into two parts; the head thoracic and abdominal parts and were stored in separate 1.5 ml tubes. The genomic DNA was isolated from the corresponding body parts following the protocol developed by Barik et al. (Reference Barik, Hazra, Prusty, Rath and Kar2013). A multiplex PCR (polymerase chain reaction) was carried out to screen human blood and the presence of plasmodial sporozoite following the methodology of Rath et al. (Reference Rath, Prusty, Barik, Das, Tripathy, Mahapatra and Hazra2017).
Mathematical representation
Here, for the first time, we have applied the basic set theorem using the Venn diagram (fig. 1) for better understanding of entomological indices and their relationships (Venn, Reference Venn1880).
Venn diagram showing the distribution of sporozoite and human blood-fed-positive major Anopheline vector population in the study area at a given time period.
A = Total study area.
N A = Total number of An. culicifacies collected in area A in a total time period of ‘T’.
X A = Total number of An. culicifacies detected with sporozoite positive in area ‘A’ in a total time period of ‘T’.
Y A = Total number of An. culicifacies detected with human blood positive in area ‘A’ in a total time period of ‘T’.
In the Venn diagram, the number of An. culicifacies with both sporozoite and human blood positive in the area ‘A’ and time period ‘T’ were represented as: XA ∩ YA
Similarly, the number of An. culicifacies with only sporozoite positive in area ‘A’ and time period
‘T’: XA–(XA ∩ YA)
Number of An. culicifacies with only human blood positive in area ‘A’ and time period ‘T’:
YA–(XA ∩ YA)
[Number of An. culicifacies neither detected with sporozoite or human blood meal in area ‘A’ and time period ‘T’: NA–(XA U YA)]
Proportion of only sporozoite-positive An. culicifacies species in area ‘A’ and time period ‘T’
$$ = \displaystyle{\matrix{{\rm No}{\rm. of}\,An.{\rm} culicifacies\,{\rm detected\ with\ sporozoite} \hfill \cr \quad {\rm only\ in\ area\ {\lsquo}}A{\rm {\rsquo}\ during\ the\ time\ period\ of\ {\lsquo}}T{\rm {\rsquo}} \hfill} \over \matrix{\; {\rm Total\ number\ of}\,An.{\rm} culicifacies\,{\rm analyzed} \hfill \cr \quad {\rm in\ area}\ \lsquo A\rsquo \ {\rm during\ the\ time\ period\ of\ {\lsquo}}T {\rsquo} \hfill}} $$
$$ = \displaystyle{{{\bf X}_{\bf A}\, - \,\lpar {{\bf X}_{\bf A}\cap {\rm} {\bf Y}_{\bf A}} \rpar } \over {{\bf N}_{\bf A}}} = {\bf m}$$Proportion of only human blood-positive An. culicifacies species in area ‘A’ and time period ‘T’
$$ = \displaystyle{\matrix{{\rm No}{\rm. of}\,An.{\rm} culicifacies\,{\rm detected\ with\ human\ blood} \hfill \cr \quad {\rm only\ in\ area\ {\lsquo}}A{\rm {\rsquo}\ during\ the\ time\ period\ of\ {\lsquo}}T{\rm {\rsquo}} \hfill} \over \matrix{{\rm Total\ number\ of}\,An.{\rm} culicifacies\,{\rm analyzed} \hfill \cr \quad {\rm in\ area\ {\lsquo}}A{\rm {\rsquo}\ during\ the\ time\ period\ of\ {\lsquo}}T {\rsquo} \hfill}} $$
$$ = \displaystyle{{{\bf Y}_{\bf A} - {\rm} \lpar {{\bf X}_{\bf A}\cap {\rm} {\bf Y}_{\bf A}} \rpar } \over {{\bf N}_{\bf A}}} = {\bf n}$$Proportion of both sporozoite and human blood-positive An. culicifacies species in area ‘A’ and time period ‘T’
$$ = \displaystyle{\matrix{{\rm No}{\rm. of}\,An.{\rm} culicifacies\,{\rm detected\ with\ both\ sporozoite} \hfill \cr \quad {\rm and\ human\ blood\ meal\ in\ area\ {\lsquo}}A{\rm {\rsquo}\ during} \hfill \cr \quad {\rm the\ time\ period\ of\ {\lsquo}}T{\rm {\rsquo}} \hfill} \over \matrix{{\rm Total\ number\ of}\,An.{\rm} culicifacies\,{\rm analyzed\ in\ area} \hfill \cr \quad {\rm {\lsquo}}A{\rm {\rsquo}\ during\ the\ time\ period\ of\ {\lsquo}}T {\rsquo} \hfill}} $$
$$ = \displaystyle{{{\bf X}_{\bf A}\,\cap \,{\bf Y}_{\bf A}} \over {{\bf N}_{\bf A}}}\, = \,{\bf mn}$$(Almost all An. culicifacies collected were PCR analyzed for sporozoite and/or human blood meal detection)
In the present investigation, the study area ‘A’ included all the selected villages of Belgaon sub-center and the entomological parameters m, n, and mn were calculated on a monthly basis (time period ‘T’ = 1 month) for both intervention and non-intervention year.
EIR is the commonly used parameter for measuring malaria transmission which is the product of sporozoite rate and human biting rate. Apart from ethical issues associated with human subjects, the method also needs further standardization. Although socio-economic, climatic, and other environmental factors play a major role in transmission dynamics, the key players in transmission are the primary mosquito vectors and their biological attributes (Bigoga et al., Reference Bigoga, Manga, Titanji, Coetzee and Leke2007). Here we have evaluated three indices ‘m’, ‘n’, and ‘mn’ (obtained from i, ii, and iii) to correlate with malaria transmission in the study area, Belgaon of Kalahandi district. Characterization of these three parameters will enable a better understanding of malaria endemicity in other locations.
Epidemiological investigations
The epidemiological data for sub-center, Belgaon was collected from the National Vector Borne Disease Control Programme (NVBDCP), Kalahandi, a surveillance system to measure the malaria incidence on the basis of blood smear examination at Primary Health Centres (PHC). It uses various malariometric indices to measure the malaria incidence, morbidity as well as mortality by compiling the data from the village level and aggregates it to the district-level data. Total malaria positive cases from the sub-center Belgaon were collected on a monthly basis, i.e. January 2016 to December 2017. Data were arranged month and/or season wise in order to estimate further association with various entomological parameters. The proportion of malaria cases in each month throughout the study period (both intervention and non-intervention year) in the selected villages of Belgaon sub-center was calculated as:
$$P_x = \displaystyle{{{\rm Total\ malaria\ cases\ per\ month}} \over {{\rm Total\ malaria\ cases\ throughout\ the\ year}}}$$(P x = proportion of malaria cases in a given month x; x = January to December).
Data analysis
The data were analyzed with Microsoft Excel 2010 spread-sheets (Microsoft Corp., Redmond, WA, USA). The density of mosquitoes was expressed as the number of female mosquitoes collected per man hour. One-way analysis of variance was used to analyze the variation in mean densities among different Anopheles species collected and between IRS intervention and non-intervention periods. Two independent groups of An. culicifacies density were compared during Deltamethrin IRS intervention and non-intervention periods (i.e. in the years 2017 and 2016) and between different habits/habitats using the Mann–Whitney U test. Similarly, an unpaired t-test was used to compare human blood-feeding behavior of An. culicifacies in a different seasonal period as well as Deltamethrin IRS intervention and non-intervention periods. In order to determine the entomological indices, linear regression plots were generated between the earlier derived three entomological parameters ‘m’, ‘n’, and ‘mn’ vs. proportion of monthly malaria morbidity (P x). This occurred during Deltamethrin IRS intervention and non-intervention periods to examine the correlations between these indices and malaria transmission dynamics. The P-value < 0.05 was considered statistically significant.
Results
Malaria is endemic throughout the year in Kalahandi district; however, the intensity of transmission is observed at its peak during the rainy season during the pre-monsoon (April–May), monsoon (June–July), and post-monsoon (August–September). During this period, it was essential to identify and characterize dominant malaria vector while assessing the transmission efficacy.
Anopheles culicifacies was found to be the primary vector throughout the study period though a number of other non-vector species were collected in greater densities (Supplementary fig. 1). Furthermore, a significant reduction in per man hour density was observed only for An. culicifacies during IRS intervention period in comparison to the non-intervention period (P < 0.05). Hence, based on its vectorial attributes, a correlation with malaria transmission in Belgaon was evaluated.
Anopheles culicifacies vector density in different habits and habitats during Deltamethrin IRS intervention and non-intervention periods
Indoor and outdoor mean density of An. culicifacies is depicted in fig. 2. Overall, a significant difference in mean density was observed during the post-monsoon period of IRS intervention and non-intervention years (indoor: U = 0, z = 2.50, P = 0.006, outdoor: U = 0.5, z = 2.40, P = 0.0082). However, there was no significant difference in the densities during the pre-monsoon and monsoon periods. Similarly, a significant variation in mean vector density in CS and HD was observed during the post-monsoon seasons in two different study periods (CS: U = 0, z = 2.506, P = 0.006; HD: U = 0.5, z = 2.40, P = 0.006) (fig. 3). On the other hand, the same was not observed for the pre-monsoon and monsoon periods.
Indoor and outdoor Anopheles culicifacies vector density (per man hour) during the pre-monsoon, monsoon, and post-monsoon periods of Deltamethrin IRS non-intervention (a) and intervention years (b) in Belgaon. Error bar represents standard error of means.
Resting habitat preference of Anopheles culicifacies vector density (per man hour) during the pre-monsoon, monsoon, and post-monsoon periods of Deltamethrin IRS non-intervention (a) and intervention years (b) in Belgaon. Error bar represents standard error of mean. HD, human dwelling; CS, cattle shed.
Evaluation of resting behavior of An. culicifacies vectors from their gonotrophic condition during Deltamethrin IRS intervention and non-intervention years
Resting behavior may be defined as the ratio between the resting and seeking stages. During the pre-monsoon period, An. culicifacies showed a higher tendency to rest outside, while during monsoon they showed a slight endophilic behavior in both the intervention and non-intervention periods. A deviation in resting behavior was observed during the post-monsoon period wherein the vectors demonstrated a higher tendency to rest inside during a non-intervention period when compared to an intervention period (table 1).
Gonotrophic condition of female Anopheles culicifacies (both indoor and outdoor) based on two different collection periods in Belgaon sub-center of Kalahandi district
F, fresh fed; UF, unfed; G, gravid; HG/SG, half gravid/semi-gravid.
Numerical values in parentheses indicate the total number of An. culicifacies sorted according to abdominal condition.
a Deltamethrin 62.5 Indoor Residual Spray (IRS) was introduced in Kalahandi (Belgaon) in the year 2017.
Variation in human blood-feeding rates of An. culicifacies during Deltamethrin IRS intervention and non-intervention periods
Vectors were analyzed for their blood-feeding patterns, with humans as the blood source, during the intervention and non-intervention study periods. A significant difference in the vector percentages was observed between the pre- and post-monsoon as well as the monsoon and post-monsoon periods. The same was not observed between the pre-monsoon and monsoon periods (fig. 4). In 2016, the percentage of anthropophagic vectors was highest during the post-monsoon period, whereas during the post-monsoon period of 2017 (i.e. post-IRS period), this percentage dropped down to the lowest level. However, a significant difference (t = 8.77 and P < 0.00001) was observed between the percent of anthropophagic vectors during the post-monsoon season of Deltamethrin IRS intervention and the non-intervention periods.
Human blood-fed Anopheles culicifacies vector density (%) during the pre-monsoon, monsoon, and post-monsoon periods of Deltamethrin IRS non-intervention (a) and intervention years (b) in Belgaon. Whiskers indicate full range of data. Top and bottom lines of boxes indicate 25–75% interquartile ranges, horizontal line within the boxes indicate medians. S, significant; NS, not significant.
Infection rate of An. culicifacies in Belgaon sub-center during Deltamethrin IRS intervention and non-intervention periods
Before the Deltamethrin strategy (i.e. during the non-intervention period 2016), overall sporozoite/infection rate was found to be 22, 28, and 32.7%, respectively, during the pre-monsoon, monsoon, and post-monsoon periods. Upon intervention with Deltamethrin (i.e. 2017), a major reduction in the infection rate was observed. The sporozoite rates were 12.2, 21.2, and 12%, respectively, during the pre-monsoon, monsoon, and post-monsoon periods (fig. 5).
Infection rate of Anopheles culicifacies during the pre-monsoon, monsoon, and post-monsoon periods of Deltamethrin IRS intervention (2017) and non-intervention years (2016) in Belgaon.
[The multiplex PCR results for sporozoite and human blood meal identification in An. culicifacies for IRS intervention and non-intervention periods were depicted in Supplementary fig. 2 and Supplementary table 1.]
A snapshot of the epidemiological profile of malaria in Belgaon sub-center during Deltamethrin IRS intervention and non-intervention periods
During the non-intervention period, the total malaria cases in the selected areas of Belgaon during the pre-monsoon, monsoon, and post-monsoon season were estimated to be 50, 160, and 198, respectively. Malaria cases during the IRS intervention year were approximately 73, 250, and186, respectively, for the pre-monsoon, monsoon, and post-monsoon season (fig. 6).
Total malaria-positive cases during the pre-monsoon, monsoon, and post-monsoon periods of Deltamethrin IRS intervention (2017) and non-intervention years (2016) in Belgaon.
Association between entomological parameters and malaria transmission during Deltamethrin intervention and non-intervention years
A direct positive association was observed between the entomological parameters (such as vector density, sporozoite rate, human blood-feeding rate) and the respective epidemiological data during Deltamethrin IRS intervention and non-intervention years. The extent of correlation was evaluated through the establishment of linear regression plots. A strong positive association was observed between the proportion of malaria morbidity (P x) and the proportion of An. culicifacies having both sporozoite and human blood meal (mn) during both intervention and non-intervention years. The associations were calculated at P < 0.00001; r = 0.96 for non-intervention and P = 0.000105; r = 0.89 for intervention periods. In contrast to this, positive correlations were observed between vectors with either sporozoite (P = 0.022; r = 0.65 for non-intervention and P = 0.04; r = 0.59 for intervention periods) or human blood meal carrier (P = 0.058; r = 0.56 and P = 0.022; r = 0.65 for intervention)(m or n). The Pearson correlation coefficients (r) for all above correlations are depicted in fig. 7. Results indicated a significant and positive correlation except between proportions of malaria cases vs. only anthropophagic vectors during IRS intervention period.
Linear regression plot showing association between the proportions of malaria cases (Y-axis) and the proportion of both sporozoite and human blood carrier (a1, b1), only sporozoite carrier (a2, b2) and only human blood-fed (a3,b3) Anopheles culicifacies (X-axis) during Deltamethrin IRS non-intervention and intervention periods in Belgaon, respectively. r = Pearson correlation coefficient.
Discussion
Despite tremendous successes in control operations, malaria still remains a long-lasting global health threat (Cohen et al., Reference Cohen, Moonen, Snow and Smith2010). Malaria elimination efforts are hampered by the lack of accurate transmission data (De Silva and Marshall, Reference De Silva and Marshall2012). Here we present a highly reliable and accurate tool for measuring transmission intensity in a region where malaria is endemic. Key determinants of malaria transmission dynamics include vector profiles, environmental, ecological, socio-economic, and climatic factors, all of which affect the efficacy of control interventions (Drakeley et al., Reference Drakeley, Schellenberg, Kihonda, Sousa, Arez, Lopes, Lines, Mshinda, Lengeler, Armstrong, Tanner and Alonso2003; Sutherest, Reference Sutherest2004; Kelly-Hope et al., Reference Kelly-Hope, Hemingway and McKenzie2009). Although Anopheles vector dynamics are the major factor affecting the risk of disease transmission, many parameters fail to provide realistic estimates of transmission intensity. The intensity of malaria transmission is commonly represented as the EIR which represents the product of human biting rate (i.e. the number of bites per person per day by vector mosquito species) measured by human landing catch and the sporozoite rate (i.e. proportion of vectors that carry plasmodial sporozoite in their salivary glands; Harada et al., Reference Harada, Ishikawa, Matsuoka, Ishi and Suguri2000). Apart from the cost and ethical concerns with human landing collections, many other factors influence the EIR. These include location, environmental, and climatic factors as well as methods used to measure the inoculation rate. The main goal of disease control is to develop effective control strategies which reduce pathogen transmission and ultimately reduce contact between vectors and humans. The EIR and other related indices are found to be contradictory to this goal, as they depend on human participation in transmission studies. To address this issue, a range of entomological indices across endemic areas could be used and explored as a measure of transmission patterns and their intensity, done through systematic sampling. In an attempt to improve methodologies for capturing Anopheline vectors, several collection methods have been conducted. Here we establish our own preferences to Anopheline collection methods and compare the entomological parameters with transmission intensity between IRS intervention and non-intervention scenarios. Human landing catches that provide information essential to evaluating and understanding malaria transmission dynamics can be replaced by collecting anthropophagic vector species within a given area. Though the human blood index is a measure of feeding behavior of a mosquito, this approach is precluded by the challenge of getting accurate quantitative information on each host in that locality. Thus, we have developed an indirect method, which involved the collection of blood-fed Anopheles vectors to measure feeding rates. Malaria transmission refers to Anopheline vectors actively transmitting malaria infections in human populations. The intensity of malaria transmission is related to the frequency at which a human population may be exposed to the bite of an infected Anopheline vector, thus leading to a plausible infection with plasmodial parasites. Hence, sporozoite rate quantifies the infection rate as well as its propensity to transmit malaria within a human population (Sahu et al., Reference Sahu, Gunasekaran, Krishnamoorthy, Vanamail, Mathivanan, Manonmani and Jambulingam2017).
In the present survey, the combination of vector collections from indoor and outdoor places helped to demonstrate our behavioral framework. Moreover, the vector dynamics study between two different periods (before and after IRS treatments) also provided the efficacy of the vector control intervention and its impact on malaria transmission. Although, An. culicifacies and An. fluviatilis are considered to be the primary vectors of malaria in endemic districts of Odisha, An. culicifacies was found to be the most abundant vector of human malaria in rural areas of Kalahandi during the study period. Anopheles culicifacies contributes to more than 60% of the reported malaria cases in India (Anvikar et al., Reference Anvikar, Shah, Dhariwal, Sonal, Pradhan, Ghosh and Velecha2016). The purpose of this study was not limited or confined to a particular vector species or location. It was rather to evaluate and correlate the entomological parameters of the most abundant malaria vector/s in a locality with malaria transmission dynamics. In order to establish such a correlation, a thorough and systematic Anopheline fauna sampling must be carried out initially and an accurate entomological profile in a malaria endemic created. Our vector profile revealed a high proportion of An. culicifacies species amongst all potential vector species during the IRS non-intervention period while the IRS intervention strategy significantly reduced their population. The overall result suggests that post-IRS intervention (i.e. the post-monsoon period) has inevitably decreased An. culicifacies outdoor population. We infer that IRS has targeted those vectors which commonly prefer biting outdoors but tend to rest indoors after feeding. Further investigations on the exact habitat preferences revealed CS as the primary resting location during the pre- and post-monsoon periods without IRS. However, during the post-monsoon period, An. culicifacies were more likely to be observed in human-dwellings. There appear few research records of Odisha where, amongst the five sibling species, An. culicifacies E was reported to have the highest anthropophilic index, although generally considered to be predominantly zoophilic in nature (Das et al., Reference Das, Das, Patra, Tripathy, Mohapatra, Kar and Hazra2013). After the IRS program (during the post-monsoon or post-IRS months), noticeable differences in vector abundances among the two biotopes were observed. Zoo-prophylaxis, a strategy to target an alternative host to divert malaria vectors away from human beings, is a potential approach to prevent malaria transmission (Waite et al., Reference Waite, Swain, Lynch, Sharma, Haque, Montgomery and Thomas2017). Hence, indoor spraying in HD might shift the vector's behavioral plasticity toward CS during the post-IRS period. Furthermore, a drastic reduction in vector density in HD between IRS intervention years was observed, thus implying the effectiveness of Deltamethrin in these mosquito populations
Our study did not establish any direct relationship between these behavioral attributes with transmission pattern. To ameliorate this issue, easily calculable unambiguous and highly predictive entomological indices were developed as early predictors of transmission dynamics. Amongst these indices, primary preference was given to human blood fed An. culicifacies vectors which showed a sudden decrease in their density after IRS intervention (i.e. during the post-monsoon period). Similarly, the percentage of human-fed vectors was drastically lower than the previous non-intervention year. The higher percentage of blood-fed mosquitoes indicates a higher tendency to feed on human hosts. During human landing collection, mosquitoes are collected when they attempt to land on human volunteers, however these vectors cannot be considered for calculating human biting rate unless they land and bite human hosts. The ultimate purpose of that bite is to consume human blood and transfer plasmodial sporozoite subsequently. Instead of collecting vectors from HD, anthropophagic/human blood-feeding vectors should be targeted for surveillance. We found a drastic reduction in sporozoite rate and infective vector population after Deltamethrin intervention. Lastly, when the epidemiological profile of malaria was analyzed in the study area, a decrease in malaria morbidity was observed during the post-IRS or post-monsoon periods. However, during the non-IRS wet season period, there was an increasing trend in malaria morbidity and infection/sporozoite rate.
As the final predictor of our study was not obtained, we investigated the association between major vectorial parameters (such as sporozoite carrier/infective mosquitoes and/or anthropophagic vectors with malaria incidence) during the IRS intervention and non-intervention periods. Infective anthropophagic vectors established a strong positive correlation with malaria morbidity in comparison to infective or anthropophagic vector species during both the study periods. It was also notable that the relationship was also positive but moderate with either of the two carriers. In spite of vector control measures, the former entomological index (i.e. both sporozoite and human blood carrier vectors) was able to maintain a strong association with malaria transmission dynamics.
Conclusion
Accurate estimations of the dominant vector population, its behavior, transmission intensity, and targeted intervention tools are the basic requirements for effective malaria control. Current research suggests that malaria transmission will not change pattern and intensity; rather, the entomological attributes of the primary vectors are the key determinants predicting transmission. The baseline evidence between entomological indices and malaria transmission dynamics provided here could be used as an early warning system for outbreak prediction in India and elsewhere where malaria transmission is common.
Limitations and recommendations
A larger sample size and wide study area selection would have enabled a more realistic detection of the relationship between entomological parameters with the transmission of malaria. Further studies are required to comprehend the role of secondary vectors in malaria transmission as present estimates of vectorial parameters may underestimate the true picture in the study area. Practically it is also difficult to investigate the species composition and their association with transmission pattern. Despite the study limitations, it provides two things at a time, first a strong association between entomological parameters and malaria transmission through a simple yet efficient procedure, and second, the impact of Deltamethrin intervention strategy on entomological and epidemiological profile.
Following research priorities are recommended for further refinement and improvement of related studies:
• Research may not necessarily involve human subjects.
• Introduction of new collection methods for anthropophilic anopheles vectors should be carried out as a replacement for the human landing collection that is able to provide parameters essential to evaluating and understanding malaria transmission dynamics.
• In order to assess a true relationship between vector indices and malaria transmission, an unbiased and routine vector monitoring and sampling is the first and ultimate requirement.
• Better knowledge of vector ecology is required.
• Every hard core inaccessible pocket should be included while studying their transmission patterns. In case of absence of any epidemiological data, previous records on disease or data from nearby areas should be collected as supportive information.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0007485319000725.
Acknowledgements
The authors gratefully acknowledge the logistic support provided by Dr S. Pati, Director of Regional Medical Research Centre, and Bhubaneswar, India for the study. We also extend our thanks to KIIT School of Biotechnology, KIIT University, Bhubaneswar, India.
Financial support
This work was supported by a research grant from the Indian Council of Medical Research, New Delhi [6/9-7(107)/2015-ECD-II].
Conflict of interest
The authors have no conflicts of interest.
