Background
Malaria remains a major public health issue in the Solomon Islands [
1,
2]. Attempts at eradication in the 1960s and 1970s greatly reduced malaria incidence but the programme was abandoned when it was realized that countrywide eradication was not obtainable after the main vector,
Anopheles farauti, developed behavioural resistance [
3,
4]. With the collapse of vector control, transmission rates resurged until insecticide treated nets (ITN) were introduced in 1992–1993 [
5]. This intervention measure resulted in a reduction in malaria cases from a high of 450 cases per 1000 people in 1992 to 150 cases per 1000 people in 1999 [
6]. More recently, in 2008, the Solomon Islands government refocused the National Vector Borne Disease Control Programme (NVBDCP) to implement intensified countrywide control and regional elimination with financial backing from the Global Fund and AusAID. The key vector control tools are again insecticide treated nets (long-lasting insecticidal nets) and indoor residual spraying (IRS) with pyrethroids rather than DDT. This rejuvenated programme further reduced the countrywide incidence of malaria to 48 cases per 1,000 population in 2011 [
1]. However, there exist large variations in malaria incidence between and within provinces [
2,
6,
7]. One of the most malarious areas in the country is Guadalcanal Province [
1] which has historically been a problem area [
8] and where in 2011 there were 87 cases per 1000 people (Ministry of Health, unpublished data).
Understanding the behaviour of the local vectors is essential for planning vector control activities. The primary vector control tools, ITN and IRS, depend on mosquitoes biting indoors, late in the night and resting indoors after feeding. There are three species of malaria vectors in the Solomon Islands:
Anopheles punctulatus,
Anopheles koliensis and
Anopheles farauti[
9]. The malaria eradication programme of the 1960-70s had reduced the distribution and abundance of
An. koliensis and
An. punctulatus, both of which were late night and highly endophagic biters [
4], leaving
An. farauti as the primary vector of malaria. The bionomics of
An. farauti in Guadalcanal Province was profiled in the early 1990s, prior to the introduction of ITNs. At this time,
An. farauti occurred in large numbers, with peak biting outdoors and early in the night (21.00) and the entomological inoculation rates (EIR) was up to 1,022 infective bites/person/year [
10‐
12]. Recent work in the elimination provinces of the Solomon Islands indicates that this early night, outdoor feeding pattern is still maintained [
13,
14]. Such early night outdoor feeding behaviour of
An. farauti could potentially compromise the efficacy of the vector control programme.
The NVBDCP is driven to reduce malaria transmission in Guadalcanal to improve the livelihood of the residents, but also because large numbers of cases are continually exported to the elimination provinces. The area around Red Beach and Koli Point, about 20 km east of Honiara, was used extensively in the late 1980s to early 1990s to study the bionomics of
An. farauti and to trial the comparative effectiveness of DDT-IRS and pyrethroid ITN [
5,
10‐
12,
15]. More recent faunal surveys have verified that
An. farauti remains very common along the north coast of Guadalcanal [
16,
17]; however for 20 years no studies profiled the bionomics of the vector. During this time frame ITNs were introduced into the area (in 1992–1993) and distribution and re-treatment activities were completed annually by the NVBDCP (Ministry of Health, unpublished data). The annual coverage rates varied depending on the availability of funds and political stability; nonetheless there was a continual presence of ITNs in the area. The hypothesis for this study questioned if the modified feeding behaviour of
An. farauti observed after the use of DDT-IRS had been maintained over time. Such information is fundamental to conducting successful elimination or intensified control programmes.
Discussion
This study observed a distinct seasonality in adult
An. farauti densities. It is important that this seasonality is considered when planning the timing of vector control activities by the NVBDCP; in particular it would be advantageous if annual activities were completed before the peak in biting occurs towards the end of the year. This seasonality reflects the larval ecology of this species and its ability to utilize brackish water lagoons for oviposition [
25,
26]. A study of the larval ecology in the study villages was simultaneously conducted [
17], which demonstrated that larval presence and density also varied seasonally and was primarily driven by rainfall. Larval abundance was highest in the drier months when brackish lagoons formed at the mouth of the streams behind sandbars. In this supporting study [
17], the peak in larval abundance occurred from September to December and would have supported the higher adult densities observed at this time. When rainfall was high (January to April), the sandbars at the mouth of the streams were washed away and in the following month the density of both larvae and adults was lower. The negative association of severe rainfall with reduced larval and adult densities of
An. farauti is supported by previous studies from both Vanuatu [
27] and Papua New Guinea [
28].
In the current study, the populations of
An. farauti in Guadalcanal were observed to feed primarily outdoors and early in the evening. During the original eradication programmes of the late 1960s and early 1970s, it was observed that
An. farauti avoided exposure to DDT-IRS by changing their feeding behaviour [
29]. In Makira-Ulawa Province prior to DDT-IRS use, the percentage of
An. farauti biting before 21.00 was around 40% and there was equal feeding indoors and outdoors [
29]. After DDT-IRS was implemented, the percentage of biting before 21.00 rapidly increased to more than 70% and the majority (66%) of biting shifted to outdoors [
29]. After DDT pressure was eventually removed from the mosquito populations, the modified behaviour of
An. farauti persisted. More recently, over the past five years, this early, outdoor biting has been observed to be sustained across the country in both Temotu [
13] and Isabel Provinces [
14].
Previously on northern Guadalcanal, the biting behaviour of
An. farauti was profiled in 1988 [
10]. The previous study was conducted after the DDT-IRS of the original eradication programme had ceased, and also before ITNs were introduced in 1993. In 1988, the peak biting time for
An. farauti was 21.00-22.00 and endophagy was 30% (range 16 to 32%) [
10]. In the current study, conducted 20 years later, the peak biting time was earlier at 19.00-20.00 and endophagy was similar with 36% feeding indoors. As this change to early outdoor biting was already in place prior to the introduction of ITN it is not possible to state unequivocal that this behaviour has been maintained in
An. farauti populations by the introduction of ITNs. However there is evidence that early night feeding was increased in
An. farauti after only 3 weeks of ITN use in an area which had previously had none or very little DDT-IRS [
30]. Also following the implementation of elimination efforts using ITNs and pyrethroids- IRS in Temotu Province, a lower portion of mosquitoes sought meals after 21.00 [
13]. This persistence in early, outdoor biting will allow malaria to be maintained as is evident in the cases of malaria reported over the last 20 years [
1]. As such there is a need to develop an integrated vector control programme which utilizes complementary strategies that consider the subtleties of mosquito ecology to further reduce the density of the local vector populations and associated malaria transmission. The most promising complementary tool which is currently available is larval control; other complementary tools generally remain at the proof of principle stage and there is a need to prioritize research funding to facilitate investigations of potential efficacy before they can be adapted into programmatic use.
Interestingly, there are indications that the behaviour of
An. farauti inside of sprayed houses has also changed [
5]. In the same area of northern Guadalcanal, the blood-feeding success and survival of
An. farauti was assessed in 1989–1991. Of those females which entered houses, collections from exit window traps showed that about half were able to successfully obtain a blood meal in either houses sprayed with DDT (52% fed) or houses with an ITN (43% fed) [
5]. However, the 24-hour mortality rates in these mosquitoes differed significantly, with 98.2% (n = 219) mortality of females collected from houses with ITN, but only 10.1% (n = 24) mortality of females collected from DDT-sprayed houses. This indicated that
An. farauti was able to avoid the insecticides on the walls and leave immediately after obtaining the blood meal. It should be noted that as this time,
An. farauti was susceptible to DDT and mortality when exposed in WHO susceptibility tests was 77% [
5]. Additionally, the susceptibility of
An. farauti to both DDT and pyrethroids was also assessed in 2006 and 100% mortality of wild-caught adults was recorded (Ministry of Health, unpublished data). Whether
An. farauti will show the same behavioural response to IRS with pyrethroids is unknown.
The annual mean parous rate recorded in this study (54.8%) is similar to that recorded 20 years earlier in the area both in unsprayed (55.5%) and DDT-sprayed houses (53.6%) [
5,
10]. However, in 1989–1991 the parous rates in houses provided with ITN were lower (49.9%) than the sprayed and unsprayed houses [
5], indicating that ITNs had an initial impact on the longevity of the
An. farauti populations; but this was not sustained, possibly due to deterioration in net quality and insecticidal efficacy. Interestingly, this survival rate from 2007–08, recorded just prior to the introduction of LLINs, appears to be higher than that recorded in other populations of
An. farauti in the Solomon Islands from 2008 onwards (42% in Temotu [
13] and 41% in Isabel [
14]). In the current study, the parous rate was relatively stable throughout the year and did not fluctuate with the changes in
An. farauti densities. However, parity rate for the first hour of biting (18.00-19.00) was dominated by more nulliparous mosquitoes than the remainder of the night. This has similarly been seen in populations of
An. farauti in Temotu Province [
13] and Central Province (Russell
et al. unpublished data) and
An. punctulatus in Papua New Guinea [
31]. However, for
An. farauti it is unlikely that this phenomenon would have any epidemiological relevance. Considering the early-biting cycle, the majority of human exposure to parous mosquitoes – which are older and have taken multiple blood meals – still occurs before 21.00.
Conclusion
The current study describes the bionomics of the primary malaria vector in the Solomon Islands, An. farauti. Recently, the NVBDCP of the Solomon Islands refocused to implement intensified countrywide control and regional elimination. The key to a successful programme will be understanding, and responding to, the behaviours of the target vector. It was observed that An. farauti has a distinct seasonal profile, with peak activity from October to December, indicating that annual vector control activities should be completed before this period. Importantly, it was observed that the early outdoor biting habit of An. farauti, first observed in the study area in 1988 still persists 20 years later. With this feeding behaviour, the target mosquitoes are able to minimize exposure to ITNs and IRS. Therefore, there is a need to implement complementary tools that provide personal protection or target other bionomics’ vulnerabilities that may exist outside of houses, such as in the larval stages, during mating, sugar feeding or any other aspect of the life cycle. This will not only improve the success of vector control in Guadalcanal, but will reduce the number of cases that are exported to the control provinces.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
HB, JLH, and CCC designed the study; HB, CB, CI, AA, and CCC performed the fieldwork; HB, RDC and TLR analysed the data and wrote the manuscript. All authors read and approved the final manuscript.