- Open Access
Spatial and temporal distribution of the malaria mosquito Anopheles arabiensis in northern Sudan: influence of environmental factors and implications for vector control
© Ageep et al; licensee BioMed Central Ltd. 2009
- Received: 03 April 2009
- Accepted: 07 June 2009
- Published: 07 June 2009
Malaria is an important public health problem in northern Sudan, but little is known about the dynamics of its transmission. Given the characteristic low densities of Anopheles arabiensis and the difficult terrain in this area, future vector control strategies are likely to be based on area-wide integrated pest management (AW-IPM) that may include the sterile insect technique (SIT). To support the planning and implementation of future AW-IPM activities, larval surveys were carried out to provide key data on spatial and seasonal dynamics of local vector populations.
Monthly cross-sectional larval surveys were carried out between March 2005 and May 2007 in two localities (Dongola and Merowe) adjacent to the river Nile. A stratified random sampling strategy based on the use of Remote Sensing (RS), Geographical Information Systems (GIS) and the Global Positioning System (GPS) was used to select survey locations. Breeding sites were mapped using GPS and data on larval density and breeding site characteristics were recorded using handheld computers. Bivariate and multivariate logistic regression models were used to identify breeding site characteristics associated with increased risk of presence of larvae. Seasonal patterns in the proportion of breeding sites positive for larvae were compared visually to contemporaneous data on climate and river height.
Of a total of 3,349 aquatic habitats sampled, 321 (9.6%) contained An. arabiensis larvae. The frequency with which larvae were found varied markedly by habitat type. Although most positive sites were associated with temporary standing water around the margins of the main Nile channel, larvae were also found at brickworks and in areas of leaking pipes and canals – often far from the river. Close to the Nile channel, a distinct seasonal pattern in larval populations was evident and appeared to be linked to the rise and fall of the river level. These patterns were not evident in vector populations breeding in artificial water sources away from the river.
The GIS-based survey strategy developed in this study provides key data on the population dynamics of An. arabiensis in Northern State. Quantitative estimates of the contributions of various habitat types and their proximity to settlements provide a basis for planning a strategy for reducing malaria risk by elimination of the vector population.
- Land Cover
- Global Position System
- Geographical Information System
- Aquatic Habitat
Malaria is a significant health problem in Sudan, affecting 52% of outpatients and accounting for 9% of all hospitals deaths . Given its diversity and the size of the country, the largest in Africa, it is not surprising that knowledge of malaria transmission dynamics, including the temporal and spatial distribution of vector populations, is generally poor. The Nile valley between Khartoum and the Sudanese/Egyptian border is not one of the most seriously affected areas, but malaria is endemic, and the area is of considerable regional significance and of growing national importance due to large hydroelectric and agricultural development projects.
Results from historical surveys dating back over fifty years, and more recent extensive surveys conducted over the last seven years, have indicated that Anopheles arabiensis is common along the Nile valley north of Khartoum to the northern limit of its distribution between Abri and Akasha, about 200 km south of the Egyptian border (Figure 1). In addition to its (probable) incursion into Egypt in the 1940s, An. arabiensis has been found on occasion as far north as Wadi Halfa (~250 km north of the current Dongola field area). Historically, it may have had a more northerly distribution, but most, if not all, of the more recent records can be attributed to human assisted dispersal [3, 4]. Today, the vector is held in check by poor terrain, less favourable climate and the vector control operations implemented by the joint Egyptian-Sudanese Gambiae Control Project.
Malaria in Northern State presents a difficult control problem that requires an effort disproportionate to the incidence of the disease and the size of the human population at risk. The geographical area affected is not large, but is elongated across difficult terrain with limited accessibility. Anopheles arabiensis is the only vector present, and for most of the year, adults are quite difficult to find. The existing control campaign relies heavily on larviciding and source reduction. The use of insecticide to treat adult mosquitoes is often directed to urban areas. This strategy maximizes the number of people protected, but arguably has more impact on biting nuisance caused by Culex quinquefasciatus Say than on malaria transmission.
The most focused vector control effort has been directed towards pushing the northern limit of An. arabiensis southward along the Nile valley, but forty years of uninterrupted control operations have not moved the boundary much further upstream. The history of vector control along the rest of the Nile within Northern State is patchy, but overall it is clear that the use of the available resources has been largely ineffective. A major contributing factor has been poor understanding of the malaria transmission dynamics and vector bionomics. This is reflected in the limitations of a control programme based largely on insecticide, which is inappropriate for targeting a low-density vector population dispersed widely across difficult terrain.
The vector population is restricted to within a few kilometres of the Nile by the desert, and migration along the river into Northern State is limited by unsuitable terrain along the Abu Hamed Reach. The population is therefore almost entirely isolated, and this has been confirmed by analysis of variation in microsatellite DNA in samples collected along the Nile from Khartoum to the Third Cataract . The vector problem is, therefore, ideally suited to an area-wide integrated pest management (AW-IPM) programme, and particularly to the use of sterile insect technique (SIT) .
It is evident from a series of preliminary surveys conducted along the length of the Nile between the Third Cataract and Abu Hammed, that the temporal and spatial distribution of An. arabiensis is very dynamic. The Nile flood has always been regarded as of primary significance, but many other factors are evident. To make a decision on appropriate and potentially effective control measures requires a comprehensive understanding of the factors involved and their relative importance. If AW-IPM incorporating SIT is the most suitable and cost effective approach, the detail and accuracy of these data become of paramount importance .
In the current study, the distribution of An. arabiensis adults was considered too sparse to provide an effective starting point for detailed population studies. A survey of the spatial and temporal distribution of larval breeding sites was more practical, although this was not achievable across the length of the area to be targeted for future AW-IPM. Based on previous surveys, two study zones were selected that as far as possible represented the diversity of terrain and land use found throughout Northern State. This paper reports results from larval surveys conducted at these sites between March 2005 and May 2007. The aim of these surveys was two-fold: to provide a dataset that would aid informed decision making on future malaria control strategies, and to provide a basis for predicting and modelling vector population dynamics in order to facilitate detailed planning for future data gathering and ultimately the design, planning and implementation of the control strategy. Results will also help to address significant knowledge gaps with respect to the epidemiology of malaria in these settings.
Study areas were selected on the basis of the wide variety of riverbank conditions and land use/land cover types they contained. Ecological conditions within the field sites are therefore considered to be typical of those over a much larger area.
In both field areas, local agriculture is based on irrigation and is primarily influenced by flooding of the Nile and seasonal changes in temperature. River levels are lowest in March and highest in August and September. Date palms dominate much of the cultivated landscape throughout the year, but the area covered by other crops varies seasonally, being greater in the winter when wheat and beans are grown, than in the summer when these are partly replaced by maize and sorghum. Irrigated agriculture, interspersed with settlements, forms a more or less continuous belt along both sides of the Nile channel but rarely extends further than about 4 km away from the river. In non-irrigated areas the dominant land cover is sand, stone desert or scrub vegetation.
Meteorological and hydrological data
Two solar-powered automatic weather stations (Campbell Scientific; Loughborough, UK) were installed in each field area. Measurements of temperature, relative humidity, wind speed and wind direction were logged at hourly intervals during the period February 2004 to May 2007. Daily data for river level at gauging stations in Kareima and Dongola were obtained from the Ministry of Irrigation, Sudan.
Land cover mapping
Pan-sharpened multispectral QuickBird data (DigitalGlobe, Longmont, CO, USA) were acquired for January 2004 and January 2005 for the Dongola and Merowe field areas, respectively. Geometric and topographical corrections were carried out using Geomatica OrthoEngine software (v. 9.1; PCI Geomatics, Ontario, Canada) using a combination of 101 ground control points, mapped using differential GPS (Trimble GeoExplorer III, Trimble Navigation Ltd, Sunnyvale, CA, USA), and a high resolution digital elevation model derived by air photogrammetry provided by the Merowe Dam Project . Object-oriented land cover classification was carried out using eCognition software (v. 4; Definiens AG, Munich, Germany). Source images were first 'segmented' into image objects using parameters developed and tested on a small representative sub-region. In addition to standard pixel-level spectral data, object characteristics such as shape, context and texture were then incorporated in the classification process [9, 10]. Images were first classified at a high spatial resolution (0.6 m) using 19 predefined land cover types. This classification product was then used to produce a generalized (100 m resolution) map of predominant land cover based on five composite definitions of land cover: 'channel edge', 'inland settlement', 'riverside settlement', 'field-dominated land cover mosaic' and 'tree-dominated land cover mosaic' (Figure 2). Survey blocks not allocated to these classes represented bare desert or waste ground and were not included in the sample universe. For full details of the classification procedures used in this study, including information on adopted class hierarchies, membership functions and classifiers, see [Additional file 1].
Monthly larval sampling was carried out using 100 m resolution sampling grids generated from the generalized land cover classification products described above. These grids consisted of a total of 2,446 and 4,559 survey cells (or 'blocks') for the Dongola and Merowe field areas respectively. Individual survey blocks covered an area of 1 ha (100 × 100 m). Each month a random sample of 40 blocks, stratified by predominant land cover type, was generated for each field area. Sections of the field area not easily accessible were excluded from the sample, as were survey blocks where the predominant land cover type was bare desert (white areas of Figure 2). In addition to the random blocks, 20 'static' blocks were surveyed in each field area each month. Static blocks were chosen during preparatory field visits and represented locations where a variety of permanent or seasonal water bodies were known to exist.
Field sampling procedures
Mosquito larva sampling was carried out monthly between March 2005 and May 2007 (full surveys, incorporating random blocks were carried out from May 2005 onwards). For each field visit a map of the static and random survey blocks to be visited was generated and uploaded to Trimble Recon rugged handheld field computers using Trimble PathFinder Office (v. 2.9) and Trimble Terrasync (v. 2.4) mobile GPS/GIS software (Trimble Navigation Ltd). Field computers were connected to Trimble Pocket GPS receivers, enabling fieldworkers to track their vehicle positions in real time and navigate to the target blocks. Within survey blocks, a fieldworker's movements were indicated as a trace on the handheld computer screen, allowing them routinely to assess their coverage of the area to be surveyed .
In each survey block all potential aquatic habitats were assessed for mosquito larvae and their locations recorded by GPS. Where possible, 30 dips were carried out at each water source, but in most cases the number of dips was limited by the size of the water body (median number of dips = 10). Anopheles arabiensis was identified morphologically and the presence of different larval stages recorded. In subsequent analysis An. arabiensis was considered to be present if at least one larva was collected. At each site fieldworkers recorded data on the physical and chemical characteristics of the aquatic habitat (surface temperature, pH, salinity, turbidity, depth, surface area, degree of shading and the presence of aquatic vegetation or algae). Habitat type was assessed on the ground and allocated to one of 20 habitat classes. The general situation of the site in terms of surrounding land use/land cover (e.g. cultivated fields, palm groves, settlement, etc.) was also recorded. All field data were entered directly into the handheld computer using data entry forms developed in Terrasync Professional software (Trimble Navigation Ltd).
ArcGIS and ArcInfo Workstation (v. 9.2; ESRI, Redlands, CA, USA) were used to integrate field data with existing GIS and RS data. All statistical analyses were carried out in Stata (v. 10; StataCorp, College Station, TX, USA). A two-sample test of proportions was used to compare frequencies of larva-positive aquatic habitats by field area and by type of survey block (static or random). Bivariate and multivariate logistic regression models were fitted to identify variables associated with the presence of larvae. Multivariate models, incorporating variables significant at p < 0.2 in the bivariate analysis, were developed using a forward-stepwise procedure and assessed using the likelihood ratio test. Interactions between key variables were assessed, and significant terms were included in the multivariate models. All regression analyses were carried out using a combined dataset from random and static blocks.
Kulldorf's spatial scan statistic was used to test whether sites containing An. arabiensis larvae were distributed randomly over space, and to identify significant spatial clusters . Analysis was carried out in SaTScan software  using a Bernoulli model and a specified upper limit for the window size of 10% of the study population (see  for a more detailed explanation of these terms). Spatial analysis was carried out using data from randomly selected survey points (i.e. data from static blocks were excluded). Temporal patterns in the proportion of aquatic habitats containing larvae were assessed using basic time series plots. Series were smoothed using a locally weighted polynomial regression (lowess) to filter daily variations and aid visual interpretation of seasonal trends.
Total number of aquatic habitats sampled and number positive for An. arabiensis larvae by type of site and field area
Habitats positive for larvae
Habitats positive for larvae
Habitats positive for larvae
Types of aquatic habit sampled and number/proportion positive for larvae
Animal water pool
Observed situation (land cover) of sampled aquatic habitats
Logistic regression modelling for larvae presence/absence
Water body characteristics
Bivariate logistic regression results for presence of An. arabiensis larvae: water body characteristics
Water temperature (°C)
Water depth (cm)
Surface area (m 2 )
Degree of shading (%)
Presence of algae
Presence of vegetation
Bivariate logistic regression results for presence of An. arabiensis larvae: landscape characteristics
Observed land cover
RS-derived land cover
Distance to river (m)
Within the RS-derived land cover classification areas of settlement were arbitrarily sub-divided into localities within or outside a 200 m buffer of the Nile channel ('riverside settlement' and 'inland settlement' respectively). Using inland settlement as a reference class, areas close to the channel edge were associated with a greatly elevated risk of larvae being present (for the two field areas combined, OR = 8.2; 95% CI = 5.4–12.5; p < 0.0001). In Dongola, areas of riverside settlement were significantly more likely to contain larvae than areas of settlement 'inland' – but no significant difference in risk between the two classes was apparent in Merowe. In Merowe, land primarily given over to cultivation ('field-dominated mosaic') was more likely to contain larvae than the reference class, although the significance of this difference was borderline, and no similar effect was observed in Dongola.
To evaluate the overall effect of distance from river on presence of larvae (i.e. regardless of local land cover), an additional logistic regression model was developed on the basis of distances between each aquatic habitat and the Nile channel, calculated in a GIS (Table 5). Using habitats within 200 m of the channel as a reference class, in Dongola there is a clear and monotonic decline in risk with increasing distance from the Nile (and particularly for distances > 400 m). In Merowe the relationship between risk and distance to channel appears to be more complex -there being no significant increase in risk in areas greater than 800 m from the river.
Spatial distributions of larvae
The spatial distributions of surveyed sites in the two field areas are shown in Figure 2, which also shows the land use categories used to stratify the random survey blocks in each site. To assess whether distinct spatial clusters in the distribution of breeding sites exist, a spatial scan statistic was determined for randomly-selected sites separately for Dongola and Merowe. In Dongola, two clusters consisting of 16 breeding sites (expected = 2.3; relative risk = 8.7; p < 0.001) and 18 breeding sites (expected = 3; relative risk = 7.7; p < 0.001) were identified (Figure 2a). In Merowe (Figure 2b), one major cluster containing 12 breeding sites (expected = 1.9; relative risk = 7.1; p < 0.001) was identified at the north of the field area. Four very small clusters, each containing between four and seven breeding sites were also identified.
Seasonal distributions of larvae
Bar charts in Figure 3 show seasonal patterns in the proportion of aquatic habitats positive for An. arabiensis larvae, together with temporal variations in mean daily temperature and mean daily river gauge (river height) over the two field areas. Distinct seasonal patterns are evident, particularly in areas situated within 200 m of the main channel (Figure 3a) where decreased river height appears to promote breeding. Seasonal variations in areas away from the river are less marked, and percentages of larvae-positive sites are generally much lower in these areas, compared with sites within 200 m of the main channel. (Note however, that the total number of positive habitats found in inland and at riverside sites was roughly similar due to the much larger number of sites sampled inland). During very high river levels (August to November), the proportion of larvae-positive sites was higher in areas away from the river.
Seasonal patterns in the proportion of habitats containing larvae differed between years. In 2005, for example, the level of the Nile fell rapidly from mid-September and by December the river gauge, averaged across the two sites, was 10.8 m and the proportion of aquatic habitats in riverside areas containing An. arabiensis larvae was above 40% (Figure 3a). In 2006, river levels appear to have dropped off relatively slowly in the period October-December; the average river gauge in December was 1.3 m higher than it had been twelve months previously and the proportion of habitats positive for larvae in areas near the Nile was only about 10%.
Overall, analysis of survey results by habitat type and situation suggests that a large proportion of breeding sites were associated with riverside habitats. Across both field areas, 40% of all larvae sampled were from breeding sites located within 100 m of the Nile channel. This finding is consistent with those of other studies of vector breeding in the vicinity of major streams or rivers [15–17]. However, a significant number of positive sites were located away from the river (see Figure 2) and were associated largely with artificial habitats such as leaking pipes and brickworks.
Temporal patterns in the proportion of aquatic habitats containing larvae (Figure 3) are consistent with those previously reported by Dukeen and Omer  for the Nile valley north of Dongola. Dukeen and Omer were able to sample larvae in riverside locations throughout the period November to June, but found that from July increasing river levels quickly reduced the number of suitable habitats until, by August, very few pre-existing breeding sites had escaped inundation. Observed densities of An. arabiensis larvae and adults declined rapidly in riverside areas from July onwards, although immature stages continued to be found in water sources away from the Nile channel. This picture is entirely consistent with the results from the current study: in riverside areas the proportion of aquatic habitats containing larvae was lowest in July to October, but no similar pattern was observed in aquatic habitats located away from the river and therefore not influenced by river height.
In riverside areas larval populations appear to increase rapidly once river levels begin to fall significantly in November. Pools in riverbank areas, left behind by the receding channel, are most common at this time of year and it seems likely that it is this increased availability of suitable larval habitats for An. arabiensis that is most important in determining larval density and adult abundance. The fact that these months coincide with a period of relatively cool daytime temperatures and relatively humid conditions is also likely to be highly significant in terms of adult survivorship and vectorial capacity . Data for malaria admissions in Northern State suggest case numbers peak in May, and this is consistent with the population dynamics of the vector described here.
A detailed appraisal of the results reported here will help to define the future of vector control in Northern State, but a discussion of two non-exclusive options may suffice to illustrate the critical importance of the current study. The main findings confirm the importance of naturally occurring riverside larval breeding sites and their seasonal predictability, but non-persistence. In contrast, artificial breeding sites appear to have a less predictable spatial pattern, and while the overall contribution to the total population is much lower, the sites can be present at any time of the year and may be long-lasting. This suggests that a two-pronged strategy may be most appropriate: a short seasonal but intense control effort launched from the river combined with a year-long, less intense offensive in and around the villages approached from the roads and desert tracks.
A massively more effective and efficient control programme than the current one based on larviciding and source reduction would be one option. The logistics of such a programme would be challenging and very costly. The use of boats to access riverside breeding sites is in fact the preferred strategy adopted by the Gambiae Control Project north of the Third Cataract, but it is difficult to conceive how this could be achieved at the necessary scale. The riverside breeding sites appear and disappear in quick succession; treated sites may be replaced by new nearby sites in a matter of days. Workers would need to visit all areas along a roughly 500 km reach of the Nile every week for up to two months and be prepared to search for sites within at least 200 m of the riverbank. This could be dealt with a by a smaller task force and through education of residents. Access is, however, only part of the problem. There is also still the inefficiency of transporting and applying insecticide to target a low-density population and the issues of environmental hazards and impact on non-target species.
SIT may offer an attractive alternative that overcomes most of the disadvantages of insecticides. The discovery that most larval breeding sites occur within 500 m of the river suggests that delivery of sterile males from boats would effectively cover most of the target area without the need for workers to move far from the riverbank. On the other hand, it is likely that the non-seasonal larval breeding sites located further from the river would seriously undermine this strategy. Strategic planning should therefore extend to a review of alternative delivery methods and their integration within the programme.
This study illustrates the potential benefits of using geospatial tools (GIS, RS, GPS) for planning and carrying out entomological surveys. Through their use it was possible to rapidly assemble a detailed and extensive dataset on An. arabiensis breeding sites using relatively limited resources. In terms of data representativeness, the stratified random sampling incorporated in this study has clear advantages over more orthodox 'purposeful' sampling commonly adopted in other studies.
The combination of the current dataset and remotely sensed data for the whole field site will help provide a comprehensive picture of the vector population, but some limitations to this approach were revealed. For example the association of larval breeding sites with poorly maintained pipes, pumps and irrigation channels is not something that can be easily extended outside of the study areas. Pipes can be associated with houses, but not all houses have plumbing, not all houses are occupied and many areas practice water conservation throughout much of the year by limiting availability to certain hours of the day. Therefore, future studies will need to identify alternative ways of capturing this information.
How critical the non-seasonal larval breeding sites are to sustaining the Northern State An. arabiensis population also has to be addressed. The low numbers of vectors away from the river may act as the seedbed for large increases in the population when the level of the Nile falls. This is not a new observation – in 1945 Shousha reported that: "during most of the year there are relatively few breeding places for An. gambiae in areas of basin irrigation. However, there are enough to produce a quantity of An. gambiae sufficient to cause very serious epidemics of malaria" . If that is true, it suggests that a focused effort on the elimination of larval breeding sites during the middle of the year would have a dramatic impact on the overall population.
Integrative geospatial analysis using GIS, RS and GPS enabled rapid and effective assembly of entomological data to aid the design of an AW-IPM vector elimination programme incorporating SIT.
The majority of An. arabiensis breeding is associated with the main Nile River channel and subject to its seasonal fluctuations. Larval numbers fall with increasing river level due to flushing of breeding sites; the opposite occurs with falling levels, when numerous breeding sites are created and provide optimal habitat.
AW-IPM necessitates coverage of inland (man-made) sites in which breeding occurs year-round; such sites constitute the focal points from which populations build up when the Nile provides suitable habitat.
An SIT campaign will benefit from targeted elimination of inland breeding sites during the hot dry season (July-October), followed by sterile male releases during the more conducive season (November onwards).
We acknowledge the contribution of our colleagues who participated in the larval surveys. We thank the Sudan Atomic Energy Commission, Remote Sensing Authority, Federal Ministry of Health, Northern State Ministry of Health, and Ministry of Irrigation for their collaboration. We also acknowledge all members of Tropical Medicine Research Institute (TMRI) for their help and support.
Financial and technical support: This work received technical support from the International Atomic Energy Agency (IAEA) under the regional project RAF 5/052, and financial and technical support from the Ministry of Science and Technology and the National Centre for Research, Sudan.
- FMOH: Annual health statistical report 2006. 2006, Khartoum: Federal Ministry of Health, Republic of SudanGoogle Scholar
- Shousha AT: The eradication of Anopheles gambiae from Upper Egypt 1942–1945. Bull World Health Organ. 1948, 1 (2): 310-352.Google Scholar
- Dukeen M, Omer S: Ecology of the malaria vector Anopheles arabiensis Patton (Diptera: Culicidae) by the Nile in northern Sudan. Bull Entomol Res. 1986, 76: 451-467.View ArticleGoogle Scholar
- Lewis DJ: The anopheline mosquitoes of Sudan. Bull Entomol Res. 1956, 47: 475-494.View ArticleGoogle Scholar
- Azrag RS: Microsatellite and mitochondrial genetic differentiation of Anopheles arabiensis Patton (Diptera: Culicidae) along the river Nile. 2008, Khartoum: University of KhartoumGoogle Scholar
- Benedict MQ, Robinson AS: The first releases of transgenic mosquitoes: an argument for the sterile insect technique. Trends in Parasitology. 2003, 19 (8): 349-355.View ArticlePubMedGoogle Scholar
- Cox J, Vreysen MJB: Use of GIS and spatial analysis in AW-IPM progarmmes that integrate the sterile insect technique. The Sterile Insect Technique Principles and Practice in Areawide Integrated Pest Management. Edited by: Dyck VA. 2005, Berlin: Springer VerlagGoogle Scholar
- Merowe Dam Project. [http://merowedam.gov.sd]
- Baatz M, Benz U, Dehghani S, Heynen M, Höltje A, Hofmann P, Lingenfelder I, Mimler M, Sohlbach M, Weber M, Willhauck G: eCognition Professional: User Guide 4. 2004, München: Definiens Imaging GmbHGoogle Scholar
- Hay GJ, Marceau DJ, Dube P, Bouchard A: A multiscale framework for landscape analysis: Object-specific analysis and upscaling. Landscape Ecology. 2001, 16 (6): 471-490.View ArticleGoogle Scholar
- Cox JS: The role of Geographic Information Systems and spatial analysis in area-wide vector control programmes. Area-Wide Control of Insect Pests: From Research to Field Implementation. Edited by: Vreysen MJB, Robinson AS, Hendrichs J. 2007, Dordrecht, The Netherlands: Springer, 199-210.View ArticleGoogle Scholar
- Kulldorff M, Nagarwalla N: Spatial disease clusters: detection and inference. Statistics in Medicine. 1995, 14 (8): 799-810.View ArticlePubMedGoogle Scholar
- SaTScan. Software for the spatial, temporal, and space-time scan statistics. [http://www.satscan.org/]
- Brooker S, Clarke S, Njagi JK, Polack S, Mugo B, Estambale B, Muchiri E, Magnussen P, Cox J: Spatial clustering of malaria and associated risk factors during an epidemic in a highland area of western Kenya. Trop Med Int Health. 2004, 9 (7): 757-766.View ArticlePubMedGoogle Scholar
- Bøgh C, Clarke SE, Jawara M, Thomas CJ, Lindsay SW: Localized breeding of the Anopheles gambiae complex (Diptera: Culicidae) along the River Gambia, West Africa. Bull Entomol Res. 2003, 93 (04): 279-287.View ArticlePubMedGoogle Scholar
- Majambere S, Fillinger U, Sayer DR, Green C, Lindsay SW: Spatial Distribution of Mosquito Larvae and the Potential for Targeted Larval Control in The Gambia. Am J Trop Med Hyg. 2008, 79 (1): 19-27.PubMedGoogle Scholar
- Sogoba N, Doumbia S, Vounatsou P, Baber I, Keita M, Maiga M, Traore SF, Toure A, Dolo G, Smith T, Ribeiro MC: Monitoring of Larval Habitats and Mosquito Densities in the Sudan Savanna of Mali: Implications for Malaria Vector Control. Am J Trop Med Hyg. 2007, 77 (1): 82-88.PubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.