Characterization of mosquito larval habitats and assessment of insecticide-resistance status of Anopheles gambiae senso lato in urban areas in southwestern Ghana

Andreas A Kudom, Ben A. Mensah, Thomas Kwaku Agyemang

Journal of Vector Ecology · 2012 · 46 citations · 23 references

DOIFull text

Open access

Concepts

Abstract

The study was carried out to characterize potential larval habitats in the city of Sekondi with the aim of assessing the relative importance of anthropogenic and natural water bodies as larval habitats. Insecticide-resistance status of Anopheles gambiae senso lato in the southwestern part of the coastal savannah zone in Ghana was also assessed against four different classes of insecticides. Larval surveys were carried out in two communities that are separated by a lagoon. Although the lagoon was a potential mosquito larval habitat, we showed that it was not an important mosquito breeding site. The major larval habitats were anthropogenic, resulting from human behavior. Some of the organically polluted breeding sites were inhabited by both An. gambiae s.l. and Culex quinquefasciatus larvae. The data also showed that An. gambiae s.l. has currently developed a strong resistance to DDT and pyrethroid insecticides in southwestern Ghana, where the species was reported to be susceptible about a decade ago. The use of insecticides in households was implicated as a possible cause of the development of resistance among An. gambiae s.l. populations in the area. The management of insecticide resistance among malaria vectors needs urgent attention if insecticide-treated materials can continue to be used for malaria control. There has been renewed interest in the problem of urban malaria in Sub-Saharan Africa (SSA) in recent years. It is estimated that about 200 million urban dwellers in SSA are currently at risk of malaria and 6–28% of annual malaria cases in Africa occur in urban environments (Keiser et al. 2004). The current strategy of the Ghana National Malaria Control Programme involves the use of insecticide-treated nets (ITN) and indoor residual spraying (IRS) of insecticide. Over the past decade, there has been a tremendous increase in the number of households protected by ITN and IRS. The percentage of households owning at least one ITN increased from 2% in 2000 to 47% in 2010, but there is neither evidence of a reduction in malaria cases nor a decrease in reported deaths (WHO 2010). This situation calls for a re-evaluation of the strategies for effective and sustainable vector management. Larval control through source reduction and routine application of larvicides was a key intervention in eradicating malaria in many parts of the world (Kitron and Spielman 1989, Utzinger et al. 2001, Killeen et al. 2002), but these methods of control have been largely neglected in recent decades in SSA, partly because of the perceived difficulty of identifying larval habitats in rural areas. It is generally thought that the abundance of clean, sun-lit, and shallow bodies of water increase the number of anopheline larval habitats in rural areas, but the absence of suitable habitats and increased water pollution generally inhibit the development of anopheline larvae in urban areas, resulting in fewer anopheline larval habitats (Chinery 1984). Consequently, a number of malaria control options, such as modification of river boundaries, draining of swamps, and insecticide application to open water bodies in urban areas, become possible and cost-effective (Robert et al. 2003). However, most urban areas in SSA may not have such marked decreases in mosquito-breeding habitats due to poor housing, lack of sanitation, and poor drainage systems. Broken pipes and pools formed at construction sites were found to be major Anopheles larval habitats in Accra (Klinkenberg et al. 2008). Keating and colleagues (2004) also found similar habitats in unplanned and poorly-drained areas in urban Kenya. Urban agriculture has been found to create additional mosquito-breeding habitats in urban areas. These major characteristics of many urban areas in Ghana and other countries in SSA could make it very difficult to apply vector control measures that target larvae. The numerous anthropogenic habitats do not only increase the rate of breeding of the vector, and thus human-vector contact, but can also increase the use of insecticides by urban households. Besides economic implications, household use of insecticides has been implicated in the development of resistance in An. gambiae in many West African countries (Elissa et al. 1993, Diabate et al. 2002a). In addition to the anthropogenic habitats, there are natural water bodies scattered around most urban areas that can support mosquito breeding. This makes potential larval habitats numerous and larval control seems a difficult option in mosquito control. Therefore, this study was carried out to characterize potential larval habitats in urban communities in Sekondi, Ghana, with the aim of assessing the relative importance of anthropogenic and natural water bodies as larval habitats. The Metropolitan Health Department in Sekondi does periodic larviciding and insecticide spraying of potential breeding sites with the aim of controlling mosquitoes/malaria in the area. Thus, one of the objectives of this study was to find out if mosquitoes indeed bred at the sites considered as potential habitats. In this study, Essei lagoon was considered a natural water body and anthropogenic habitats refer to those created by human activities, such as choked gutters and ponds constructed for agricultural activities. Insecticide-resistance status of An. gambiae s.l. from the area and other urban areas in the coastal savannah zone was also assessed. Sekondi-Takoradi Metropolitan Area, with Sekondi as the administrative capital, occupies the southeastern part of the Western Region of Ghana. It has a land area of 385 km2 and is characterized by an equatorial climate. Temperatures are high with an average of 22° C. It has a mean annual rainfall of 2,350 mm, with a major rainy season in May and June and minor rains occurring between September and November. Larval surveys were carried out in two communities, Bakano and Bakakyir, which are separated by the Essei lagoon. Mosquitoes were collected from the coastal savannah zone in southwestern Ghana. Four contrasting localities were selected on the basis of variation in agricultural production, use of insecticides, and/urban and peri-urban settings. Some mosquitoes were also collected from one locality in the forest zone. The localities were Sekondi-Takoradi, an urban vegetable growing area, and Cape Coast, an area without major urban agriculture, both in the coastal savannah zone. Shama-Ahanta (coastal savanna) and Twifo Praso (forest) are peri-urban areas and mostly engaged in crop farming. Larval surveys were carried out in Bakano and Bakakyir once every month from December, 2010 to March, 2011 during the dry season. All open water bodies were taken as potential breeding sites and explored. Essei lagoon was divided into nine sections and a larval survey was carried out in each section. The presence or absence of larvae was determined after 10 to 15 dips. Where larvae were found, five dips each measuring 200 ml (volume of the dipper) from different areas of the water body were taken at 10-min intervals. Larval density was determined by taking the average number of mosquito larvae from the five dips. Overall larval density was determined by also taking the average number of mosquito larvae in the four-month sampling period. Habitats where mosquito larvae were found throughout the four-month period were described as productive mosquito habitats. Furthermore, productive habitats with fewer than five larvae per 200 ml were described as unimportant larval habitats. Likewise, sites with larval density that ranged from five to 25 larvae per 200 ml were described as important and those with more than 25 larvae as very important larval habitats. Anopheline larvae were identified morphologically as described in Gillies and Coetzee (1987). Habitats were characterized using a standard format for each site, recording the presence of predators (mainly fish), water quality (pH), electrical conductivity (EC), and turbidity, and whether the site was anthropogenic or natural. The pH, salinity, and EC were measured using a portable meter (YSI 63; Yellow Springs Inc., U.S.A.). The locations of larval habitats were determined using a hand-held global positioning system receiver, which gaves a positional accuracy within 3 m. These habitats were then plotted on a map of Essei Lagoon using ESRI's ArcMap 9.2 software. Mosquitoes were collected as larvae during the rainy season, May–June 2011, and brought to the laboratory for emergence of adults. Kisumu, a susceptible strain of An. gambiae, was used as a reference. Adult susceptibility assays were carried out using WHO discriminating dosages of four insecticides from different chemical classes: Permethrin 0.75%, DDT 4%, Malathion 5%, and Bendiocarb 0.1%. Mortality resulting from tarsal contact with treated filter paper was measured using WHO test kits. Each test of a batch of 20 to 25 mosquitoes was replicated four to five times and tests with untreated paper were run in parallel as a control. All tests were done on two to five-day-old, non-blood-fed, female mosquitoes from different collection periods to account for inter-batch variability. Knockdown was recorded after 1 h and final mortality was recorded 24 h post-exposure (WHO 1998). Data from the larval habitat surveys were collected during the dry season, thus most of the transitory habitats were dry and water flow in Essei lagoon was very slow. Approximately 15 potential anthropogenic habitats were found, however, 86.7% were breeding mosquitoes (actual anthropogenic habitats). Almost 80% of the actual habitats were choked gutters and the remaining 30% were wells and ponds. About 30.8% of the actual habitats bred mosquitoes throughout the four-month survey (productive habitats; B1–4). Larvae of An. gambiae s.l. were found in wells used for irrigation on vegetable farms and a cluster of ponds used for palm kernel oil production were inhabited by both An. gambiae s.l. and Culex quinquefasciatus. Cx. quinquefasciatus were also found breeding in choked gutters. Larval density was extremely high; for example site B1 recorded an average of 65 larvae per 200 ml of water collected. Mosquito larvae were found in only one location on Essei lagoon (A3). This location had been disturbed as a result of a previous bridge construction. Larval density was, however, very low: one larva per 200 ml of water collected. Physicochemical properties varied along the lagoon and the surrounding larval habitats. As expected, salinity in Essei lagoon increased as it got closer to the sea; salinity at A3, farther away from the sea, measured 0.4 ± 0 parts per thousand (ppt) while A1, which is closer to the sea, measured 17.4 ± 2.3 ppt (Table 1). The EC on Essei lagoon differed spatially and temporally and mostly depended on the type of activity at each location. However, the anthropogenic habitats (B1–4) had almost constant EC (Table 1). The highest EC value of 1751 ± 47 μS/cm with turbidity of 180 parts per million (ppm) was recorded at A4 where wastes from sawmill operations and sewage entered the lagoon. The pH value for the entire lagoon was 7.7 ± 0.05. A total of 1,772 An. gambiae s.l. was exposed to permethrin, DDT, malathion, and bendiocarb and Figure 1 shows their resistance status in various locations in southwestern Ghana. Percentage mortality recorded for the control was less than 5%. Complete mortality was observed after exposure of the An. gambiae Kisumu strain to discriminating dosages of all the insecticides. Susceptibility/resistance of mosquitoes was similar in coastal savannah and in the forest zones. However, the mosquitoes were highly susceptible to malathion and bendiocarb. Resistance status of Anopheles gambiae in southwestern Ghana against discriminating dosages of DDT (4%), permethrin (0.75%), malathion (5%), and bendiocarb (0.1%). Although Essei lagoon, a natural water body, was a potential mosquito larval habitat, the study showed that it was not an important mosquito breeding site. The major larval habitats came about as a result of human behavior or influence. These kinds of habitats have been reported by many other studies in other parts of Ghana and SSA. (Chinery 1984, Klinkenberg et al. 2008, Keating et al. 2004). This study also revealed that ponds polluted with palm kernel oil were inhabited by both An. gambiae s.l. and Culex quinquefasciatus larvae. Although polluted waters are not traditionally important breeding sites for An. gambiae s.l., findings from Sekondi, together with other studies (Chinery 1984, Keating et al. 2004, Sattler et al. 2005, Awolola et al. 2007), could be an indication of a change of Anopheles sp. breeding requirements in urban settings. The issue of occurrence of anthropogenic habitats in urban areas has not been adequately addressed, particularly in the context of malaria control due, perhaps, to the fact that anthropogenic habitats in urban areas mostly breed Cx. quinquefasciatus. The interest of this study is to emphasize that currently, An. gambiae s.l. can breed in anthropogenic habitats (polluted or unpolluted), and therefore such habitats also need attention for successful malaria control programs. Physicochemical properties were similar in potential or actual larval habitats. Additionally, mosquito larvae shared habitats with similar physicochemical properties. Sattler et al. (2005) reported that a pH value of less than 7.3 was associated with high culicine larvae. In contrast, this study found high culicine larval densities in breeding sites with pH values of 7.6 and above. The findings of this study agree with those of Antonio-Nkondjio et al. (2011), who found no strong correlation between vector distribution and the physicochemical properties of breeding sites. The influence of physicochemical properties on the distribution of mosquito larvae needs further investigation. The most important factors that may have excluded mosquito larvae from almost the entire length of Essei lagoon were water movement and predators, notably fish. Though water flow was very slow because of the dry season, at least it prevented mosquitoes from breeding. However, if not checked, the dumping of waste into the lagoon from commercial activities, such as sawmills along the banks of the lagoon, may obstruct water flow in the near future. Mosquito larvae were found in one location where there had been considerable human interference due to bridge construction. This implies that Essei lagoon could become an important mosquito larval habitat if the flow of water is obstructed. Informal interaction with the farmers and other residents in the communities near the lagoon revealed that they thought the lagoon was a major mosquito breeding site but were not aware that the ponds or the wells were breeding mosquitoes. Neither could they identify mosquito larvae when shown to them. This is not surprising because Kudom and Mensah (2010) found similar inadequate knowledge of mosquito life cycles even among tertiary level students in Cape Coast, Ghana. About a decade ago, An. gambiae s.l. from southwestern Ghana was reported to be susceptible to pyrethroid insecticides, particularly permethrin (Kristan et al. 2003). Data from this study, however, show a strong resistance of An. gambiae s.l. to permethrin and DDT. This confirms the rapid spread of insecticide resistance among malaria vectors in SSA, especially against pyrethroid insecticides. This phenomenon has been detected across most countries in SSA (Chandre et al. 1999, 2000), since Elissa and colleagues (1993) first reported pyrethroid resistance of An. gambiae in the Ivory Coast. Cross-resistance of An. gambiae s.l. to DDT and pyrethroids has been attributed to the kdr gene, conferring knockdown resistance (Fanello et al. 2003, Weill et al. 2000, Diabate et al. 2002b). Low mortality rate of An. gambiae to DDT and permethrin in this study may also be attributed to the kdr gene, conferring knockdown resistance. Yawson et al. (2004) showed evidence of high frequency of kdr mutation in An. gambiae from southwestern Ghana but did not carry out insecticide bioassays. Therefore, this study provides some support for the presence of the high frequency of kdr mutation among An. gambiae in the area. The susceptibility of An. gambiae population to malathion and bendiocarb probably rules out a metabolic resistance mechanism and absence of Ace.1 mutation in the study area. The frequency of resistance of An. gambiae to DDT and permethrin and susceptibility to malathion and bendiocarb was similar in the study sites, though the sites were in different ecological settings and had different levels of agricultural activity and pesticide use. The common characteristic in the study sites were high numbers of anthropogenic habitats, which is mostly associated with high populations of mosquitoes. Excessive use of agricultural pesticide and household use of insecticide has been implicated in the development of resistance in An. gambiae in many West African countries, including Ghana (Ellisa et al. 1993, Diabate et al. 2002a, Akobgeto et al. 2006, Corbel et al. 2007, Boakye et al. 2009). We have the opinion that the use of agricultural insecticides may have played a minor role in the development of resistance in this study due to different levels of agricultural activity and pesticide use. Rather, domestic use of insecticides in the form of aerosol sprays or coils may have been the major cause of resistance in mosquitoes in the study sites. In order not to jeopardize the efficacy of insecticide treated nets through insecticide resistance, there is a need to reduce insecticide use, especially at the household level. One way is to reduce the mosquito population, especially in urban areas where most of the important larval habitats are anthropogenic as reported both here and by other workers (Keating et al. 2004, Klinkenberg et al. 2008). Such habitats can easily be managed through human behavior change, proper waste management, and proper construction of drains. Habitat-based modelling of impacts of mosquito larval interventions have been shown to have a great potential in reducing transmission intensity and incidence of malaria (Gu and Novark 2005), especially when combined with other interventions. In summary, our study found that not all potential breeding sites in the study area were important mosquito breeding sites and most Anopheles habitats were anthropogenic. The level of pyrethroid resistance observed in An. gambiae s.l. populations in the study area can have a negative impact on the nation's malaria control program. To be effective, malaria control programs should also target Culex mosquitoes because previously exclusive Culex habitats currently also support Anopheles. People will resort to household insecticides for protection against mosquito bites, whether from Anopheles or Culex, and this could continue to drive resistance to a higher level. Larval control and the prevention of formation of larval habitats can be a strong tool in managing both Culex and Anopheles in urban areas. Larval habitat prevention can only be achieved when people have adequate knowledge of the life history of mosquitoes.

References

23