Mosquito biting activity on humans & detection of Plasmodium falciparum infection in Anopheles stephensi in Goa, India

Knowledge of bionomics of mosquitoes, especially of the disease vectors, is necessary to plan appropriate vector avoidance and control strategies. For example, information on biting activity of vectors during the night hours in different seasons could help in choosing personal protection measures that would prevent human-mosquito contact. Earlier a study on the biting activity of disease vectors using human baits was conducted in the coastal urban areas of Goa, India1. Similar studies have generated valuable information on biting patterns of different vector species on bovine and human baits in India and neighbouring countries2–5. These studies point to the seasonal and temporal variations in mosquito biting rhythms. We undertook this study from October 2005 to September 2006 with two main objectives: (i) to determine the composition of mosquito fauna biting on humans and variation in their biting activity during different phases of night and seasons; and (ii) to measure sporozoite infection rates in Anopheles stephensi, which is considered primary malaria vector species, and other anophelines to find the existence of any secondary malaria vector species in Goa, India

Material & Methods

Study area: Goa is situated on the western coast of India with a population of 1.34 million6 residing in two districts and 11 sub districts comprising of 15 towns and 398 villages. The study was conducted in these ecotypes covering both urban and rural areas of Goa within the jurisdiction of the Urban Health Centres of Panaji, Margao and Vasco and Primary Health Centres Candolim, Bicholim, Cansarvarnem, Ponda, Betki, Cortalim, Quepem, Canacona, Valpoi and Sanguem.

Meteorological data: Weather data were obtained from the Regional Meteorological Observatory, Goa, and monthly means of various weather parameters viz., maximum and minimum temperatures, rainfall and relative humidity were worked out for the pre-monsoon (January-May), the monsoon (June-September), and the post-monsoon periods (October to December).

Mosquito collection: A preliminary survey was conducted to select suitable human dwellings for all-night bait collections in the study areas. The inhabitants of the dwellings were pre-informed about the purpose of mosquito collections and their informed consent was taken after study protocol was approved by Institutional Ehics Committee of National Institute of Malaria Research, New Delhi. Collections of female mosquitoes landing on human volunteers were carried out indoors from 1800 to 0600 h during 85 nights. Mosquitoes were etherized and identified in the laboratory to species level following the keys7–10.

Entomological inoculation rate (EIR): All female anophelines were dried and stored individually in plastic vials containing dried silica gel under cold conditions (0-4 ° C). These were tested by sporozoite enzyme-linked immunosorbent assay11 using antibodies to circumsporozoite proteins (CSP) of Plasmodium falciparum, P. vivax 210 and P. vivax 247. End point results were read visually and confirmed at 450 nm using a Vmax kinetic microplate reader (Molecular Devices Corporation, Sunnyvale, CA, USA). The mean absorbance values of the five known negative controls was 0.207 (range: 0.119 - 0.305) for P. falciparum; 0.215 (range: 0.157 - 0.273) for P. vivax 210; and 0.095 (range: 0.092 - 0.103) for P. vivax 247. Twice the mean negative value was taken as cut-off value in this study, which has been earlier found to be most reliable12. Hence, samples with values above twice the mean of negative values were considered positives for the CSP of concerned Plasmodium species. EIR was calculated for Panaji and Goa using formula EIR = Number of landing mosquitoes/person/night × sporozoite rate.

Statistical analysis: Hourly data of landing mosquitoes of all the nights in each month were pooled by species and averages were worked out. The hourly data of number of mosquitoes collected from 1800 to 0600 h were pooled to study landing trends in different phases of night. Also to study seasonal biting trends, species- wise landing collections of various months were pooled for pre-monsoon, monsoons and post-monsoon period. Z - test of proportions was applied to study differences of vectors landing in different seasons. To study biting trends during the night in the entire study period covering all the collection hours and in different seasons species-wise, William's mean (M_w) was computed which is a modified geometric mean that compensates for the zero values, and is calculated using the formula Mw = anti log [sum (x+1)/N] – 1, where Mw = the modified geometric mean, x = number of mosquitoes caught during different hours and N = the number of observations5.

The human volunteers and mosquito collectors were explained about the purpose, procedure and risk of infection during the collection and informed consent was obtained in writing from them. Volunteers and mosquito collectors were administered a weekly prophylactic dose of 600 mg chloroquine in a single dose. The study protocol was approved by the e0 thics Committee of National Institute of Malaria Research, New Delhi.

Results

A total of 4,191 mosquitoes were collected during 85 nights from 14 different localities with an overall mean mosquito landing rate of 49.3 mosquitoes per person/night. These mosquitoes belonged to 5 genera and 23 species, including the malaria vectors An. stephensi Liston (55, 1.31%), An. fluviatilis James (75, 1.79%), and An. culicifacies Giles (32, 0.76%); the filariasis vectors Culex quinquefasciatus Say (1710, 40.8%) and Mansonia uniformis Theobald (76, 1.81%); the Japanese encephalitis (JE) vectors Cx. vishnui Theobald, Cx. pseudovishnui, Cx. tritaeniorhynchus and Cx. gelidus (1262, 30.1%); and the dengue and chikungunya vectors Aedes aegypti L. and Ae. albopictus (Skuse) (63, 1.5%). Of the total 4,191 female mosquitoes, 3,273 (78.1%) were known vectors of different mosquito borne diseases and the remaining (918, 21.9%) were of non vector species (Table I). The mean number of female mosquitoes landing per person per night for malaria vectors was 1.89, filariasis vectors 20.9, JE vectors 14.83, and dengue/chikungunya vectors 0.73. The number of landing mosquitoes collected on human bait was maximum between 0300-0600 h (1622; 38.7%) followed by 1800 and 2100 h (1059; 25.3%), 2100-2400 h (830; 19.8%), and least (680; 16.2%) during 2400 to 0300 h (Table I).

Table I Total number (%) of landing mosquitoes collected on human bait during various phases of 85 all night (1800 - 0600 h) in 14 different localities of Goa

Malaria vectors

An. stephensi - The maximum numbers of An. stephensi females were caught from 0300-0600 h (23; 41.8%) and the least during 2400-0300 h (6; 10.9%) (Table I). The trends of landing mosquitoes based on William's mean values also suggested that distinct peak activity of An. stephensi was in the early hours of the morning. However, the temporal difference in biting behaviour was apparent during different seasons. A distinct peak in biting activity in the early hours of the morning was seen in the pre-monsoon months while in the monsoons there was accelerated biting between 2000 to 2400 h and in the early morning hours from 0300 to 0600 h. In the post-monsoons, a small peak in the evening was followed by another at midnight. Of the total 55 An. stephensi collected, 15 (27.3%), 17 (38.9%) and 23 (41.8%) were collected during pre-monsoons, monsoons and post-monsoon seasons, respectively. However, the seasonal differences were found to be non significant (Table II).

Table II Seasonal landing populations of mosquitoes on human baits and inter- seasonal comparison

Of the 831 female anophelines, evaluated for CSP by ELISA, only 2 An. stephensi females were found to contain P. falciparum sporozoites. Among anophelines tested, the plasmodial infection rate was 0.2 per cent, while amongst An. stephensi alone, it was 3.6 per cent. The two infected females were among 18 An. stephensi collected from Panaji city. The EIR of P. falciparum was calculated as 18.11 for Panaji and 2.35 for entire Goa.

An. fluviatilis - Similar to An. stephensi, the peak number of females of An. fluviatilis were caught during early hours of the morning from 0300 to 0600 (30; 40%) and least during 2400 to 0300 h (10; 13.3%). An. fluviatilis showed bimodal peaks one from 2100-2300 h and a higher peak from 0300 to 0600 h during the monsoons while maximum biting was noticed from 2100 to 2200 h during post-monsoon period and biting during monsoons was very less. Of the total 75 An. fluviatilis collected, most (54, 72%) were collected during pre-monsoon which were significantly more than 5 (6.7%) collected during monsoons and 16 (21.3%) collected during post-monsoons (Table II). None of the 75 An. fluviatilis were positive for Plasmodium infection during sporozoite ELISA studies.

An. culicifacies - An. culicifacies was found biting throughout the night at variable frequencies (Table I). There was no definite biting trend shown by the species although during the post-monsoons, maximum biting occurred before 0200 h. Maximum number of landing An. culicifacies mosquitoes were collected during post-monsoon i.e. 25 of 32 (78.1%) which were significantly more than the other two seasons (Table II). Of 32 An. culicifacies collected, none were found positive for plasmodium infection during sporozite ELISA studies.

Filariasis vectors: Cx. quinquefasciatus - This species was captured throughout the year though maximum landing females were collected from 0300-0600 h (698; 40.8%) and the least (280; 16.4%) during 2400-0300 h (Table I). Biting activity of Cx. quinquefasciatus increased gradually from 0200 h and peaked between 0500 and 0600 h both during pre- and post-monsoon seasons although peak in the post-monsoon season was comparatively much higher. Of the 1710 Cx. quinquefasciatus mosquitoes, 636 (37.2%) were collected during the pre-monsoon, 200 (11.7%) during the monsoon and 874 (51.1%) during the post-monsoon and the difference between the different seasons was significant (Table II).

Mansonia uniformis: William's mean suggested that there was no time preference for biting shown by M. uniformis in general although during the post-monsoon biting pattern showed variability temporally and biting increased past mid night during this season. M. uniformis was found in much lesser numbers as compared to Cx. quinquefasciatus. This vector of Brugian filariasis was more prevalent during post monsoon season when maximum number of these mosquitoes i.e. 54 (71.1%), followed by 20 (26.3%) during pre- monsoon and only 2 (2.6%) were encountered during the monsoon period and the difference between the seasons was significant (Table II).

Japanese encephalitis vectors: Cx. tritaeniorhynchus along with other 2 JE vectors viz., Cx. vishnui and Cx. gelidus were found feeding throughout the night with maximum biting occurring from 0300-0600 h and the least during 2400-0300 h (Table I). Another suspected vector of JE in this area, Cx. pseudovishnui was found feeding in negligible numbers with biting activity between 1800 and 2100 h (3; 75%) and from 2100-2400 h (1; 25%). Of the four JE vectors, Cx. tritaeniorhynchus, showed distinct peak in biting activity after 0300 h during the post-monsoon while during pre- and post- monsoons there was no such distinct peak. In the case of Cx. vishnui there was much more biting activity up to midnight compared with later hours during the pre-monsoon months while during the monsoon there was very less landing populations and no definite choice of biting hours. During the post-monsoon however, there were two clear peaks one from 2100 to 2200 h and other from 0500 to 0600 h. The third JE vector Cx. gelidus showed trimodal biting pattern during the pre-monsoon, the first peak from 2000 to 2100 h, the second from 2300 to 2400 h and the third much higher at 0500-0600 h. This species showed higher biting rate before 2100 h and decline thereafter. During the post-monsoon months there were three peaks like the pre-monsoon months, the first up to 2100 h, the second from 0100 to 0200 h and much higher peak from 0400 to 0500 h. Overall, significantly more number of JE vectors viz., Cx. tritaeniorhynchus, Cx. vishnui and Cx. gelidus were collected during the post-monsoon season being 408 (55.9%), 206 (63.6%) and 124 (60.5%), respectively than the pre-monsoon i.e. 217 (29.8%), 103 (31.8%) and 71 (34.6%), respectively.

Dengue and chikungunya vectors:

Although Aedes aegypti and Ae. albopictus are primarily diurnal species, they were also found active biting human host especially during the early phase of night and early morning hours as compared to the second and third phases of the night. Although Ae. aegypti did not show any preferred time, it was encountered in all the three seasons in small numbers i.e. 11 (42.3%), 5 (19.2%) and 10 (38.5%), respectively and the difference was found to be non significant (Table II). Ae. albopictus was only found during post-monsoon season and of the 37 landing mosquitoes collected, the species showed crepuscular feeding up to 2100 h and then during early hours from 0300 to 0600 h.

The non vector anophelines showed temporal biting variations during different seasons although overall maximum biting occurred from 0300-0600 h and least during 2400-0300 h (Table I). These species of anophelines were significantly more prevalent during the pre-monsoon compared with other two seasons (Table II). During the pre-monsoon months, the biting started during the first phase of night but declined after 2300 h and picked up after 0200 and peaked from 0500 to 0600 h. During the monsoon, the populations of these anophelines declined drastically, whereas during the post-monsoon, the pattern of biting appeared quite similar to pre-monsoon months.

The non vector culicines were most prevalent during post-monsoon i.e. 134 (53.8%) followed by significantly less numbers i.e. 96 (38.6%) in the pre-monsoon and 19 (7.6%) in the monsoon. The seasonal biting pattern revealed that during the pre-monsoon months there were two peaks, one in the early phase of night and second lower peak during early morning. During the monsoon, the populations declined significantly although maximum biting was noticed during early morning hours. During the post-monsoon months, there were two prominent peaks one crepuscular and the second in the early morning.

Overall, about half of the mosquitoes were collected during post-monsoon season i.e. 2156 (51.4%) followed by pre-monsoon 1586 (37.8%) and the least i.e. 449 (10.7%) during the monsoon and this difference was statistically significant (Table II).

Discussion

The study revealed that the vectors of malaria, filariasis and Japanese encephalitis were actively feeding on humans throughout the year between 1800 h to 0600 h. Differences were observed in the biting rhythms of different vector species during different phases of the night and seasons. The data also revealed that vector biting was the least in the monsoon months compared with other seasons which could be due to the flushing effect of heavy rainfall on immature populations although the temperature and humidity conditions were ambient during this season.

An. stephensi, a primary malaria vector, was found in the urban coastal localities of Goa viz., Panaji, Candolim, Margao and Cortalim. It was found actively pursuing the host throughout the night with distinct peak activity in the early hours of the morning and in all the three seasons. However, a similar study1 carried out in Goa had observed peak biting of An. stephensi from 2100 to 2400 h. The biting habits of An. stephensi are found to be multi-modal or arrhythmic216. The previous studies3–5 also suggest that the biting activity of the same species in a particular location can be influenced by sleeping behaviour of the host, microclimatic conditions and the lunar cycle.

An. fluviatilis was reported earlier as a main vector of malaria in the rural eastern hilly regions of Goa13. In the present study, this species was collected in the entire scotophase and in all seasons from both hills and foothills of Sanguem and Valpoi in the east, Quepem in the south central and Canacona in the south western parts of Goa. An. culicifacies was recorded earlier from the middle sub coastal belt of Goa13. In the present study also, this species was captured biting from the three sub coastal areas of Bicholim, Cansarvarnem and Quepem and did not appear to play any significant role in malaria transmission. The species also did not show much seasonal and temporal variability. Earlier studies had however, shown that the feeding patterns of An. culicifacies were highly variable and dependent on local ecology, particularly the climatic factors14–16.

The landing catches of Cx. quinquefasciatus was the most predominant of all the species. The brugian filariasis vector M. uniformis was found comparatively in much lesser numbers than Cx. quinquefasciatus. In the absence of Brugia malayi infections in Goa, this species has only nuisance potential.

The JE vectors pose a serious challenge to the public health in Goa as sporadic outbreaks of JE have been frequently reported especially in the sub coastal belt of Goa. Of all the JE vectors, Cx. tritaeniorhynchus was found to be the predominant vector species in Goa. This and other two species, viz., Cx. vishnui and Cx. gelidus were also similarly active throughout the night irrespective of seasons. However, the fourth Japanese encephalitis vector, Cx. pseudovishnui was collected in negligible numbers.

Ae. aegypti and Ae. albopictus are well-known vectors of dengue and chikungunya respectively in urban and rural areas in India. Small numbers of both these vectors were active after dusk and around dawn. Although these two vectors are known to be primarily diurnal species and Goa being endemic to both dengue and chikungunya, the present study suggests that preventive measures undertaken against other vector species would also be partly effective against these vectors especially during the early hours of the evening and morning when their biting was noticed.

The incrimination studies have confirmed that only An. stephensi is the primary malaria vector in Goa as two of the 831 female anophelines, evaluated for CSP by ELISA, were females of this species found positive for P. falciparum sporozoites. Panaji continues to be high malaria risk area in Goa as both the positive females were captured from the city where EIR was also found to be quite high in previous two studies, An. stephensi was incriminated as malaria vector from Panaji and Candolim in Goa1718.

A study on the biting rhythm of mosquitoes2 postulated the possibility of genetic factors influencing the biting behaviour of mosquitoes and variations in the degree of anthropophily and endophagy among different populations of anophelines and aedine species of mosquitoes. Other studies have indicated that feeding pattern of disease vectors varies widely and is dependent on climatic factors1920. Vector mosquito population densities and the extent of their contact with human hosts are important factors in determining the transmission rates of vector borne diseases4. Hence, the circadian rhythm of biting cycles is of great epidemiological significance.

The vector-host contact can be prevented by using available personal protection methods such as repellents, proper clothing and long lasting insecticide nets (LLINs). The latter act as mechanical and insecticidal barrier which can repel or knockdown mosquitoes upon contact thus providing effective protection against vector borne diseases during most part of the night21. In the present study, about a quarter of all mosquitoes were found biting in the evening and early hours of the night. During this period, application of suitable repellents and proper clothing can provide protection against vectors.