|Home | About | Journals | Submit | Contact Us | Français|
Little is known about the presence and distribution of Aedes mosquitoes in northern Tanzania despite the occurence of viruses transmitted by these mosquitoes such as Chikungunya virus (CHIKV) and Dengue virus (DENV) in the region. Adult and larval mosquitoes were collected from rural and urban settings across a wide range of altitudes in the Kilimanjaro Region using the Mosquito Magnet CO2 Trap for collection of adults and old tires for breeding of larvae. Polymerase chain reaction assays were performed on captured adult mosquitoes to detect the presence of CHIKV and DENV. A total of 2609 Aedes aegypti adult mosquitoes were collected; no other Aedes species larvae were found. Mosquito yields were significantly higher in urban settings than rural settings (26.5 vs. 1.9 mosquitoes per day, p = 0.037). A total of 6570 Ae. aegypti larvae were collected from old tires; no other Aedes species larvae were found. Of the 2609 adult mosquitoes collected, none tested positive for CHIKV or DENV. As far as we are aware, this paper reports for the first time the presence of Ae. aegypti in the Kilimanjaro Region of northern Tanzania. Although CHIKV and DENV were not isolated from any of the collected mosquitoes in this study, the apparent absence of other Aedes species in the area suggests that Ae. aegypti is the primary local vector of these infections.
The Aedes mosquitoes are the principal vectors of many arboviruses of human health importance, such as Dengue virus (DENV), Chikungunya virus (CHIKV), and Yellow Fever virus.1,2 In sub-Saharan Africa, Aedes aegypti and Aedes albopictus have been the vectors responsible for outbreaks of arboviral infections among humans in several settings, including coastal Gabon3 and coastal Kenya.4,5 However, in many other parts of sub-Saharan Africa, the presence of these Aedes species and the arboviruses that they transmit has not been studied. One such location is the Kilimanjaro Region of northern Tanzania, where the presence of Aedes species has not been documented. However, a study of severe febrile illness etiology in northern Tanzania in 2007–2008 demonstrated that 7.9% of febrile inpatients had acute CHIKV infection and 9.5% of febrile inpatients had presumptive acute DENV infection.6,7 These finding suggested the presence of at least one Aedes species capable of transmitting these arboviruses to humans. Although Ae. aegypti has been reported to be responsible for DENV epidemics in the low-altitude coastal region of Tanzania,8 the presence of Aedes mosquitoes has not been described in the high altitudes of the Kilimanjaro Region.
In order to identify and characterize the local vector of CHIKV and DENV, we conducted a mosquito trapping study to confirm the presence and identity of Aedes mosquitoes in the Kilimanjaro Region of northern Tanzania and their geographic distribution.
This study was conducted in and around the town of Moshi (human population – 144,000), situated at an altitude of 890 m above sea level in the Kilimanjaro Region (human population >1.4 million) of northern Tanzania. The location was selected to match the catchment area of the study of severe febrile illness etiology.6,7
Mosquito trapping was conducted using the Mosquito Magnet CO2 Trap™ (Woodstream Corporation, Lilitz, PA, USA). Trapping was conducted from 15 March 2012 through 25 May 2012, a period covering portions of both the annual dry and rainy seasons. Mosquitoes were collected at 14 different sites; 6 of these sites were located within Moshi Urban District, and the remaining 8 sites were in rural settings outside of the city. GPS coordinates of each trapping site were collected using the Garmin (Olathe, KS, USA) Etrex 10™ GPS device. Trapping was conducted at a single site each day; the trap was set from sunrise to sunset in well-vegetated, shaded settings near human dwellings to maximize capture success.9–11
Upon collection, mosquitoes were killed by freezing, sorted by species and sex using appropriate taxonomic keys,12,13 and stored at −80°C. At the conclusion of the sampling period, mosquito samples were shipped on dry ice to collaborators at the Duke-National University of Singapore in Singapore for molecular testing.
For each real time polymerase chain reaction (RT-PCR) assay, five mosquitoes of the same species and site were pooled. DENV and CHIKV RNA were extracted from the sample using TRIzol® RNA isolation reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol and reverse transcription was performed using Invitrogen Superscript III First Strand Synthesis System™ (Life Technologies, Carlsbad, CA, USA), according to manufacturer’s instructions. RT-PCR assays for DENV and CHIKV were carried out with the LightCycler 480 SYBR Green I Master kit™ (Roche Diagnostics, Penzberg, Germany) using previously described primers and cycling conditions.14,15
Used tires of uniform size served as artificial breeding sites and were placed in five locations in and around Moshi, including two urban locations within the town limits and three rural locations outside of town. Tires were filled with clean water and set in shaded vegetated areas to mimic the natural breeding sites for Aedes mosquitoes.11 Tires were first placed in January 2012. For the ensuing 20 weeks, third and fourth stage larvae and pupae were collected from these breeding sites on a weekly basis. These larvae were subsequently reared in an insectary, and identification of species was performed using taxonomic keys12,13 once the larvae had fully matured into adult mosquitoes. In order to reduce the risk of adult emergence from the artificial breeding sites, larvae were collected at least once per week and subsequently stored in sealed containers.
Urban collection sites were defined as those within the limits of the Moshi Urban District. Rural collection sites were defined as those within the outlying villages and forests of Moshi Rural District and Hai District. Because mosquito trapping was performed for a different number of days at different sites, we defined ‘mosquito abundance’ as the average number of mosquitoes caught per day at each site, to allow for uniform comparison. High altitude sites were defined as those at altitudes ≥900 m above sea level; low altitude sites were defined as those at altitudes <900 m above sea level. Differences in mosquito abundance across groups were compared using Mood’s median test.
This study was approved by the Kilimanjaro Christian Medical University College (KCMU College) Research Ethics committee.
Characteristics of the 14 collection sites are described in Table Table1.1. Six (43%) of the sites were in urban locations, and the median (range) altitude of these sites was 823.5 (720, 1369). A map of the collection sites is displayed in Fig. Fig.11.
Trapping was conducted for a total of 40 days. A total of 2609 Aedes mosquitoes were collected from the 14 sites, and all of the collected mosquitoes were Ae. aegypti. No Ae. albopictus or forest-dwelling Aedes mosquitoes were collected from any trapping site. Of the collected mosquitoes, 1618 (62%) were female.
Table Table22 lists the total number of Aedes mosquitoes caught from each site as well as the mosquito abundance for each site, measured in mosquitoes per day. The median (range) mosquito abundance was 3.3 (0, 219.3) mosquitoes per day.
The median (range) mosquito abundance among urban sites was 26.5 (0.5, 219.3) mosquitoes per day, compared to 1.9 (0, 14.0) mosquitoes per day among rural sites; this difference was statistically significant (p = 0.037). The median (range) mosquito abundance among high altitude sites was 10.8 (0, 219.3) mosquitoes per day, compared to 2.3 (0, 57) mosquitoes per day among low altitude sites; this difference was not statistically significant (p = 0.298).
Table Table33 lists the five sites used for artificial mosquito breeding. A map of these breeding sites is displayed in Fig. Fig.1;1; the breeding sites represented a mix of urban and rural locations in both high altitude and low altitude settings. A total of 6570 Aedes larvae were collected from these breeding sites. All of the collected larvae were Ae. aegypti; no other Aedes larvae were collected from any of the breeding sites. Table Table33 lists the total number of larvae collected from each site. The median (range) larval collection per site was 613 (130, 4693). The highest larvae yield was obtained in a high-altitude urban site, Kilimanjaro Christian Medical Centre (KCMC). Because of the small number of breeding sites, statistical analyses comparing larvae yields among sites of different altitudes or urban vs. rural settings were not pursued.
None (0%) of the 2609 adult Ae. aegypti mosquitoes collected in this study tested positive for CHIV or DENV by RT-PCR.
To our knowledge, this study is the first documenting the presence of Ae. aegypti in the Kilimanjaro Region of northern Tanzania. Despite collection of both adult and larval Aedes mosquitoes from a variety of urban and rural locations in both high and low altitude settings, no other Aedes species were collected during this study. Although we were unable to detect CHIKV and DENV in the collected mosquitoes, the apparent absence of Ae. albopictus or forest-dwelling Aedes species suggests strongly that Ae. aegypti is the principal vector of these viruses in the region. This finding stands in contrast to the results of studies conducted in the Congo, Gabon, Kenya, and elsewhere in sub-Saharan Africa4,16–18 where Ae. albopictus was found to have supplanted Ae. aegypti as the primary vector of CHIKV infection. Moreover, the presence of Ae. aegypti in the region suggests the possibility of outbreaks of other arboviral infections, such as Yellow fever and West Nile virus disease, in the future.
Adult Ae. aegypti mosquitoes were found to be more abundant in urban locations; this result is consistent with the mosquito’s well-documented predilection for urban environments around the globe.19–22 This study did not demonstrate a correlation between Ae. aegypti abundance and altitude, although a more comprehensive study including an analysis of ecologic factors such as vegetation and rainfall is needed to fully elucidate the relationship between altitude and mosquito abundance. Although Ae. aegypti invasion to altitudes greater than 1000 m above sea level has traditionally been considered to be very rare,23 both adult and larval Ae. aegypti were collected from sites at altitudes above 1000 m during this study. This finding is in accord with the results of recent studies from the western hemisphere documenting the presence of large numbers of Ae. aegypti well above this altitude.24 To our knowledge, this is the first report of the invasion of Ae. aegypti to altitudes above 1000 m in the African continent.
CHIKV and DENV were not detected in any of the collected mosquitoes. There are many possible explanations for this finding. First, this study may have been conducted during an inter-epidemic period or period of low CHIKV and DENV transmission, during which time the prevalence of CHIKV and DENV among Ae. aegypti is likely to be very low. Because the etiology of febrile illness study was conducted five years prior to this study, the prevalence of human CHIKV and DENV infection during the mosquito collection period is not known. Second, inadvertent mishandling of the collected mosquitoes during storage and shipping, such as a break in the cold chain, could have compromised our ability to detect CHIKV and DENV when the samples arrived in Singapore.
This study had several limitations. First, mosquito collection was conducted during a relatively brief three-month period. A year-long study would be more useful in detecting temporal trends in mosquito abundance as well as potentially detecting species that were not abundant for whatever reason during the study period. Second, this study did not collect additional data about surrounding vegetation, presence of open containers, and site-specific temperatures which would have allowed for a more robust analysis of the determinants of variation in mosquito abundance across sites. Third, mosquito identification was performed using pictorial taxonomic keys rather than using molecular identification. Although pictorial taxonomic keys are generally sufficient to identify major Aedes species such as Ae. aegypti and Ae. albopictus, molecular identification is the most accurate method and prevents missing species not included in taxonomic keys. Fourth, this study relied on a binary definition of urban and rural settings based on administrative boundaries. Local population density data at each trapping site would have allowed for a more detailed analysis of the interaction between mosquito abundance and population density but unfortunately such comprehensive population density data were not available. Fifth, viral PCR studies were only performed on adult mosquitoes in this study, not on larval mosquitoes. As adult mosquitoes can travel great distances, viral studies on larvae can potentially better characterize local transmission of pathogens.
In conclusion, this study documented for the first time the presence of Ae. aegypti in the Kilimanjaro Region of northern Tanzania, even at altitudes above 1000 m. Other Aedes species were not found during this study, suggesting that Ae. aegypti is likely the principal vector of recently described human CHIKV and DENV infection in the region. Additional study is needed to confirm this finding. Our findings suggest several avenues for further research. First, a more exhaustive year-long trapping study with concomitant temperature, rainfall, and vegetation data would allow for a more detailed analysis of spatial and temporal trends in Ae. aegypti populations. Second, comparing temporal trends in Ae. aegypti populations to trends in human chikungunya infections in the region could further illuminate pathogen–vector–host dynamics. Such data could ultimately be used to plan and implement vector control strategies to reduce the local burden of arboviral disease.
None of the authors have any conflicts of interest to declare.
Support for this study was provided by the Training Health Researchers into Vocational Excellence in East Africa (THRiVE) research grant [grant number 087540] funded by the Wellcome Trust. JAC received support from the joint US National Institutes of Health-National Science Foundation Ecology of Infectious Disease program and the UK Economic and Social Research Council and Biotechnology and Biological Sciences Research Council (R01TW009237).
The authors wish to thank the KCMC, the KCMC-Duke collaboration for the support of mosquito collections and storage, and KCMU College insectary staff for their support in rearing mosquitoes. The authors further thank Charles Masenga, Augustine Mtui, Godfrey Kweka, and Levynson Lewis who served as field assistants.