• View in gallery

    Distributions of anopheline larval habitats in the rainy season and dry season in the study area in Kakamega, western Kenya. m = meters.

  • View in gallery

    Results of K-function analysis on the spatial relationship between anopheline larval habitats and distances to the nearest streams in the rainy season (A) and dry season (B), and results of bivariate K-function analysis comparing the distribution patterns between the rainy and dry seasons (C) and between the Anopheles gambiae complex and An. funestus (D). The bold lines are the observed value of the test statistic L(d) at a given distance d, and the gray lines indicate the 95% confidence intervals. m = meters.

  • View in gallery

    Distribution of anopheline larval habitats with respect to the distance to the nearest stream. m = meters.

  • View in gallery

    Distribution of anopheline larval habitats with respect to elevation. m = meters.

  • View in gallery

    Relationships between area of each land cover type and distance from streams. m = meters.

  • 1

    Roberts JM, 1964. The control of epidemic malaria in the highlands of Western Kenya. I. before the campaign. J Trop Med Hyg 67 :161–168.

    • Search Google Scholar
    • Export Citation
  • 2

    Malakooti MA, Biomndo K, Shanks GD, 1998. Reemergence of epidemic malaria in the highlands of western Kenya. Emerg Infect Dis 4 :671–676.

    • Search Google Scholar
    • Export Citation
  • 3

    Lindblade KA, Walker ED, Onapa AW, Katungu J, Wilson ML, 1999. Highland malaria in Uganda: prospective analysis of an epidemic associated with El Nino. Trans R Soc Trop Med Hyg 93 :480–487.

    • Search Google Scholar
    • Export Citation
  • 4

    Shanks GD, Biomndo K, Hay SI, Snow RW, 2000. Changing patterns of clinical malaria since 1965 among a tea estate population located in the Kenyan highlands. Trans R Soc Trop Med Hyg 94 :253–255.

    • Search Google Scholar
    • Export Citation
  • 5

    John CC, Ouma JH, Sumba PO, Hollingdale MR, Kazura JW, King CL, 2002. Lymphocyte proliferation and antibody responses to Plasmodium falciparum liver-stage antigen-1 in a highalnd area of Kenya with seasonal variation in malaria transmission. Am J Trop Med Hyg 66 :372–378.

    • Search Google Scholar
    • Export Citation
  • 6

    John CC, O’Donnell RA, Sumba PO, Moormann AM, de Koning-Ward TF, Kazura JW, Crabb BS, 2004. Evidence that invation-inhibitory antibodies specific for the 19-kDa fragment of merozoite surface protein-1 (MSP-1 19) can play a protective role against blood-stage Plasmodium falciparum infection in individuals in a malaria endemic area of Africa. J Immunol 173 :666–672.

    • Search Google Scholar
    • Export Citation
  • 7

    Lindsay SW, Martens WJ, 1998. Malaria in the African highlands: past, present and future. Bull World Health Organ 76 :33–45.

  • 8

    Mouchet J, Manguin S, Sircoulon J, Laventure S, Faye O, Onapa AW, Carnevale P, Julvez J, Fontenille D, 1998. Evolution of malaria in Africa for the past 40 years: impact of climatic and human factors. J Am Mosq Control Assoc 14 :121–130.

    • Search Google Scholar
    • Export Citation
  • 9

    Lindblade KA, Walker ED, Onapa AW, Katungu J, Wilson ML, 2000. Land use change alters malaria transmission parameters by modifying temperature in a highland area of Uganda. Trop Med Int Health 5 :263–274.

    • Search Google Scholar
    • Export Citation
  • 10

    Githeko AK, Lindsay SW, Confalonieri UE, Patz JA, 2000. Climate change and vector-borne diseases: a regional analysis. Bull World Health Organ 78 :1136–1147.

    • Search Google Scholar
    • Export Citation
  • 11

    Zhou G, Minakawa N, Githeko AK, Yan G, 2004. Association between climate variability and malaria epidemics in the east African highlands. Proc Natl Acad Sci U S A 101 :2375–2380.

    • Search Google Scholar
    • Export Citation
  • 12

    United Nations, 2003. World Population Prospects: the 2002 Revision. New York: United Nations.

  • 13

    Brooks TM, Pimm SL, Oyugi JO, 1999. Time lag between deforestation and bird extinction in tropical forest fragments. Conservation Biol 13 :1140–1150.

    • Search Google Scholar
    • Export Citation
  • 14

    Whitmore TC, 1997. Tropical forest disturbance, disappearance, and species loss. Laurence WL, Bierregaard Jr RO, eds. Tropical Forest Remnants. Chicago: University of Chicago Press, 2–28.

  • 15

    Walsh JF, Molyneux DH, Birley MH, 1993. Deforestation: effects on vector-borne disease. Parasitology 106 (Suppl):S55–S75.

  • 16

    Patz JA, Graczyk TK, Geller N, Vittor AY, 2000. Effects of environmental change on emerging parasitic diseases. Int J Parasitol 30 :1395–1405.

    • Search Google Scholar
    • Export Citation
  • 17

    Minakawa N, Mutero CM, Githure JI, Beier JC, Yan G, 1999. Spatial distribution and habitat characterization of anopheline mosquito larvae in western Kenya. Am J Trop Med Hyg 61 :1010–1016.

    • Search Google Scholar
    • Export Citation
  • 18

    Minakawa N, Sonye G, Mogi M, Yan G, 2004. Habitat characteristics of Anopheles gambiae s.s. larvae in a Kenyan highland. Med Vet Entomol 18 :301–305.

    • Search Google Scholar
    • Export Citation
  • 19

    Service MW, 1977. Mortalities of the immature stages of species B of the Anopheles gambiae complex in Kenya: comparison between rice fields and temporary pools, identification of predators, and effects of insecticidal spraying. J Med Entomol 13 :535–545.

    • Search Google Scholar
    • Export Citation
  • 20

    Khan B, Ofulla AV, Kariuki DM, Githure JI, Kabiru EW, Martin SK, 1992. Drug sensitivity studies during a highland malaria epidemic in Kenya. Trans R Soc Trop Med Hyg 86 :371–372.

    • Search Google Scholar
    • Export Citation
  • 21

    Hightower AW, Ombok M, Otieno R, Odhiambo R, Oloo AJ, Lal AA, Nahlen BL, Hawley WA, 1998. A geographic information system applied to a malaria field study in western Kenya. Am J Trop Med Hyg 58 :266–272.

    • Search Google Scholar
    • Export Citation
  • 22

    Lemmon PE, 1956. A spherical densiometer for estimating forest overstory density. Forest Sci 2 :314–320.

  • 23

    USEPA, 1983. Methods for Chemical Analysis of Water and Wastes. Washington, DC: U.S. Environmental Protection Agency.

  • 24

    APHA, 1992. Standard Methods for the Examination of Water and Wastewater. Washington, DC: American Public Health Association.

  • 25

    Gillies MT, Coetzee M, 1987. A Supplement to the Anophelinae of Africa South of the Sahara. Johannesburg: The South African Institute for Medical Research.

  • 26

    Scott JA, Brogdon WG, Collins FH, 1993. Identification of single specimens of the Anopheles gambiae complex by the polymerase chain reaction. Am J Trop Med Hyg 49 :520–529.

    • Search Google Scholar
    • Export Citation
  • 27

    Koekemoer LL, Kamau L, Hunt RH, Coetzee M, 2002. A cocktail polymerase chain reaction assay to identify members of the Anopheles funestus (Diptera: Culicidae) group. Am J Trop Med Hyg 66 :804–811.

    • Search Google Scholar
    • Export Citation
  • 28

    Ripley BD, 1976. The second-order analysis of stationary point processes. J Appl Probabil 13 :255–266.

  • 29

    Minakawa N, Githure JI, Beier JC, Yan G, 2001. Anopheline mosquito survival strategies during the dry period in western Kenya. J Med Entomol 38 :388–392.

    • Search Google Scholar
    • Export Citation
  • 30

    Minakawa N, Seda P, Yan G, 2002. Influence of host and larval habitat distribution on the abundance of African malaria vectors in western Kenya. Am J Trop Med Hyg 67 :32–38.

    • Search Google Scholar
    • Export Citation
  • 31

    Wiegand T, Moloney KA, 2004. Rings, circles, and null-models for point pattern analysis in ecology. Oikos 104 :209–229.

  • 32

    Gillies MT, De Meillon B, 1968. The Anophelinae of Africa South of the Sahara. Johannesburg: The South African Institute for Medical Research.

  • 33

    Gimnig JE, Ombok M, Kamau L, Hawley WA, 2001. Characteristics of larval anopheline (Diptera: Culicidae) habitats in western Kenya. J Med Entomol 38 :282–288.

    • Search Google Scholar
    • Export Citation
  • 34

    Zar JH, 1999. Biostatistical Analysis. Englewood Cliffs, NJ: Prentice Hall.

  • 35

    Bentley MD, 1989. Chemical ecology and behavioral aspects of mosquito oviposition. Annu Rev Entomol 34 :402–421.

  • 36

    Mogi M, 1981. Population dynamics and methodology for bio-control of mosquitoes. Laird M, ed. Biocontrol of Medical and Veterinary Pests. New York: Praeger, 140–172.

  • 37

    Washburn JO, 1995. Regulatory factors affecting larval mosquito populations in container and pool habitats: implications for biological control. J Am Mosq Control Assoc 11 :279–283.

    • Search Google Scholar
    • Export Citation
  • 38

    Sunahara T, Ishizaka K, Mogi M, 2002. Habitat size: a factor determining the opportunity for encounters between mosquito larvae and aquatic predators. J Vector Ecol 27 :8–20.

    • Search Google Scholar
    • Export Citation
  • 39

    Gimnig JE, Ombok M, Otieno S, Kaufman MG, Vulule JM, Walker ED, 2002. Density-dependent development of Anopheles gambiae (Diptera: Culicidae) larvae in artificial habitats. J Med Entomol 39 :162–172.

    • Search Google Scholar
    • Export Citation
  • 40

    Ye-Ebiyo Y, Pollack RJ, Spielman A, 2000. Enhanced development in nature of larval Anopheles arabiensis mosquitoes feeding on maize pollen. Am J Trop Med Hyg 63 :90–93.

    • Search Google Scholar
    • Export Citation
  • 41

    Ye-Ebiyo Y, Pollack RJ, Kiszewski A, Spielman A, 2003. Enhanced development of larval Anopheles arabiensis by proximity to flowering maize (Zea mays) in turbid water and when crowded. Am J Trop Med Hyg 68 :748–752.

    • Search Google Scholar
    • Export Citation

 

 

 

 

SPATIAL DISTRIBUTION OF ANOPHELINE LARVAL HABITATS IN WESTERN KENYAN HIGHLANDS: EFFECTS OF LAND COVER TYPES AND TOPOGRAPHY

View More View Less
  • 1 Faculty of Medicine, Saga University, Saga, Japan; Kenya Medical Research Institute, Kisumu, Kenya; Department of Biological Sciences, State University of New York at Buffalo, Buffalo, New York

The distributions of anopheline larval habitats were aggregated in valley bottoms in Kenya in both the rainy and dry seasons, although the degree of aggregation was higher in the dry season than in the rainy season. Larvae of the Anopheles gambiae complex larvae were found more frequently in habitats in farmlands and pastures. However, An. funestus larvae were found more frequently in natural swamps and pastures. Canopy cover was the only variable significantly associated with the occurrence of the An. gambiae complex and An. funestus. The average canopy cover was significantly less in the habitats with the An. gambiae complex and An. funestus larvae than those without the anopheline larvae. Thus, land cover types and topographic features showed important effects on the distribution of anopheline larval habitats. These results suggest that clearing riparian forests would improve growing conditions of the An. gambiae complex and An. funestus larvae in Kenyan highlands.

INTRODUCTION

Malaria, a major human health threat, occurs globally in tropical and sub-tropical regions. In the east African highlands, the threat is mounting as shown by more frequent malaria outbreaks in areas where malaria was previously rare.14 Unlike their counterparts in malaria-endemic regions, recent studies suggested that the residents of highland areas generally lack immunity to Plasmodium falciparum and are particularly vulnerable to malaria infection.5,6 Several hypotheses have been proposed to explain the increased malaria transmission in the highlands, including land-use changes, global climate changes, increased drug resistance, cessation of malaria control activities, and demographic changes.2,711

In recent years, Kenya has experienced very rapid human population growth. The population of this country has almost doubled since 1980.12 Population growth in the western Kenya highlands is a particularly severe problem, due in part to a lack of family planning and to migration from other areas. As a consequence of dramatic human population increase, there have been unprecedented land-use changes in the highlands. For example, since 1965 the Malava Forest in the Kaka-mega district has been reduced from 600 to less than 100 hectares.13 Most rain forests have been cleared for crop planting, cattle grazing, commercial logging, firewood collection, and housing construction.14

The drastic land-use alteration can promote vector-borne disease transmission in several ways.15,16 First, more human-made aquatic habitats become available for the Anopheles gambiae complex, which is the primary malaria vector in Africa. For example, increased cattle grazing creates more open temporary habitats that may serve as mosquito breeding habitats.17,18 Second, the physical and chemical properties of mosquito larval habitats may change. Third, the microclimate of mosquito larval habitats may change. Habitats well exposed to sunlight may have a higher temperature than those in shaded areas, and thus mosquito larval development rate may be increased. For example, a higher adult density of the An. gambiae complex was found in houses near cultivated swamps than in those near natural swamps in a Ugandan highland site.9 Natural swamps are primarily populated with tall grasses such as papyrus, whereas cultivated swamps are planted with agricultural crops such as maize.18 Since cultivated swamps generally receive more exposure to sunlight than those in natural swamps, the ambient air temperature in the cultivated swamp area was significantly higher than that in natural swamps.9 Moreover, mosquito larval predators may be more prevalent in natural swamps than in cultivated swamps.19 These differences in aquatic habitat characteristics may have important effects on anopheline larval development and survivorship, and on the distribution of suitable larval habitats.

The aim of the present study is to examine the effects of land use/land cover types on the distribution of anopheline larval habitats in a western Kenya highland site. We determined the spatial distribution pattern of larval habitats and analyzed several physical and chemical components in the aquatic habitats that may be affected by land use and may be relevant to mosquito larval development. We also examined the effects of topography on the spatial distribution of larval habitats. The east African highland region contains numerous small valleys and basin-like depressions in a plateau.7 Stagnant water bodies are more likely to form in valley bottoms and basin-like depressions than in the hills.18 Thus topography may be an important factor for the distribution of aquatic habitats in the highlands. Knowledge of the effects of land use/land cover types and topography on mosquito larval habitat distribution is useful for predicting the distribution and abundance of malaria vectors and the impact of land-use practice on malaria transmission, and thus for designing novel strategies for malaria intervention.

MATERIALS AND METHODS

Study area.

We established a study area of approximately 4 km2 along Yala River in the Kakamega district, western Kenya, where epidemics of malaria have claimed several hundred lives in recent years.20 The elevation of the study area ranged from 1,400 to 1,580 meters above sea level. During the period between May 24 and August 31, 2002, the air temperature ranged from 11.4°C to 33.6°C, and the average air temperature was 19.3°C. In 2002, the long rainy season persisted from April through early June, and the dry season started in July and ended in September. In early 2002, 707 houses and 2,217 inhabitants were counted within the study area. The inhabitants live in traditional stick and mud houses with thatch or iron sheet roofs. The area included small patches of indigenous forests along the river and streams. Deforestation and cultivation along the forest edges are ongoing. The major crop grown in this area was maize.

Land cover types.

Prior to the larval habitat survey, we estimated the surface area (meters2) of each land cover type within the study area. Land cover types were classified into five categories: farmlands, forests, pastures, roads, and natural swamps. Farmlands were characterized by the presence of any agricultural crops and bare ground that had been prepared for planting crops. Pastures were grasslands used for grazing. An area with a mixture of grass and shrubs was also classified as pasture. Natural swamps were characterized by the presence of emergent aquatic plants. Forests and natural swamps were the categories with the least modification by anthropogenic activities. A multispectral one-meter ground resolution Ikonos image (Space Imaging, Atlanta, GA) of the study area was acquired on April 12, 2001. Each land cover type was visually digitized using the geographic information system (GIS) software package ArcView 3.3 (Environmental Systems Research Institute, Redlands, CA). The digitized land covers were validated by ground truthing. The surface area of each land cover type was estimated by first projecting the land cover layer into Universal Transverse Mercator 1983 zone 36N and then using an Avenue Script that calculates the surface area in ArcView. All streams (including the Yala River) and roads in the study area were also mapped and digitized using the satellite image in the GIS and validated by ground truthing.

Habitat characterization.

We first counted and categorized all aquatic habitats (excluding running water and water in containers in houses and in tree holes) with respect to land cover types in the long rainy season (May 17 to May 30, 2002) and dry season (August 13 to August 21, 2002). Land cover types, where aquatic habitats were located, were classified into the above five categories. The location and elevation of each habitat was recorded using the global positioning system in differential mode.21 Occurrence of anopheline larvae was examined at each habitat using a standard dipper (size = of 350 mL). Water was dipped up to 20 times. When a habitat was too small to make 20 dips, water was dipped as many times as possible.

We also characterized 130 and 50 randomly selected aquatic habitats (including both anopheline-positive and -negative habitats) with respect to physicochemical variables in the long rainy season and dry season, respectively. The selected habitats were characterized using the following variables: 1) canopy cover, 2) presence of filamentous algae, 3) turbidity, 4) habitat size (volume), 5) pH, and 6) concentrations of ammonium (NH4+), nitrate (NO3), and orthophosphate (PO4−3). Canopy cover was estimated with a spherical densiometer.22 Surface water was sampled and analyzed for the above chemical variables and turbidity following the standard methods of the U.S. Environmental Protection Agency.23,24 The volume (meters3) of an aquatic habitat of 10 evenly spaced grids was established by measuring its length, mean width, and mean depth, and calculated as the product of these three indices.

In the habitat characterization with the physicochemical variables, anopheline larvae were collected and preserved in 95% ethanol from each positive habitat, and third and fourth instar larvae were identified to the An. gambiae complex, the An. funestus complex, and the other anopheline species based on a morphologic key using a microscope.25 Larvae of the An. gambiae complex and the An. funestus complex were further analyzed with the ribosomal DNA (rDNA) polymerase chain reaction (PCR) method.26,27 First and second instar larvae were subjected to rDNA-PCR analysis for species identification directly.

Data analysis.

We determined whether the distribution pattern of anopheline larval habitats was clustered in the study area in each season using point pattern spatial analysis (Ripley’s K-function).28 In this analysis, we only used the data from the anopheline-positive habitats. When it was clustered, we determined whether topography influenced the distribution pattern of anopheline larval habitats using two geographic variables, the distance of each habitat to the nearest stream and its elevation. The Kolmogorov-Smirnov one-sample test was used to determine whether anopheline larval habitats were confined toward valley bottoms in each season. The distance of each larval habitat to the nearest stream was estimated in the GIS. Similarly, we examined the relationships between elevation and the distribution of anopheline habitats. We further determined whether the occurrence (presence/absence) of anopheline larvae was significantly correlated with these geographic variables and season using logistic regression analysis.

Since anopheline breeding sites are generally reduced during the dry season,29,30 we suspected that the reduction is reflected in the difference in habitat distribution patterns between the dry season and rainy season. We used bivariate point pattern analysis to determine whether the distribution of anopheline-positive habitats in the dry season was more confined than in the rainy season using the statistical software package Programita.28,31 When it was significantly clustered, we used a t-test to compare the average distances from the nearest streams and the average elevations of anpheline-positive habitats between the seasons.

Since occurrence of the An. funestus complex larvae is generally limited to larger, semipermanent or permanent habitats with aquatic vegetation,30,32,33 its habitat distribution pattern may be different from the An. gambiae complex. We determined whether the distribution of the An. funestus complex-positive habitats was significantly more clustered than that of the An. gambiae complex using bivariate K-function analysis. When it was significantly clustered, we compared the average distances of larval habitats to the nearest streams and the average elevations of larval habitats between the two species complexes. In the analysis for the species complexes, the data from both seasons were included because the number of anopheline habitats during the dry season was not large enough to perform a separate analysis.

Associations between the occurrence of anopheline larval habitats and the land cover types were examined for each season using the chi-square test. The Tukey-type multiple comparison test was used to test differences in proportion of larval occurrence among the land cover types.34 The association was also examined for each species complex. In the analysis for each species complex, the data from both seasons were included together and road was excluded because few positive sites were found within this land cover type. Logistic regression analysis was used to examine the association between occurrence of each species complex and environmental variables. We further examined whether the characteristics of larval habitats (including both positive and negative sites) varied significantly among the land cover types using two-way analysis of variance for continuous variables and chi-square analysis for categorical variables. Tukey multiple comparison tests were used for post-hoc analyses. The statistical significance was 0.05 and adjusted with the Bonferroni correction if necessary. All continuous variables were log-transformed.

RESULTS

Species composition.

Among the 1,021 anopheline larvae collected, 43 (4.2%) were An. implexus, and 987 larvae belonged to the An. gambiae complex and the An. funestus complex that were subjected to further molecular identification. We found 585 (57.3%) An. gambiae, 73 (7.1%) An. arabiensis, and 122 (11.9%) An. funestus s.s. larvae. A total of 207 (20.3%) anopheline larvae were not identified due to PCR amplification failure. Because the proportion of An. arabiensis was small, we pooled An. arabiensis and An. gambiae in the following analyses.

Relationships with topography.

During the rainy season survey, 767 aquatic habitats were identified in the study area, and anopheline larvae were found at 277 habitats (36.1%; Table 1 and Figure 1). The number of aquatic habitats was reduced to 190, and 48 habitats (25.3%) were anopheline-positive during the dry season. Figure 2 shows measures of the observed L(d) and the 95% confidence interval plotted for values of interpoint distance ranging between 0.1 and 1.4 km. For clustering to be significant at the P < 0.05 level, the observed L(d) must lie above or below the 95% confidence interval lines. We found that anopheline larval habitats were significantly clustered up to 1.4 km during the rainy season, and clustering peaked at 0.7–0.9 km (Figure 2A). During the dry season, larval habitat clustering occurred up to 1.0 km and peaked at 0.7 km (Figure 2B). Bivariate K-function analysis found that the magnitude of the K-function estimate of larval habitats for the dry season was significantly greater than that for the rainy season at distances up to 0.3 km and approximately 0.6 km (Figure 2C). That is, larval habitat during the dry season was significantly more clustered compared with the rainy season at the distances. Similarly, An. funestus larval habitats were significantly more clustered than the An. gambiae complex habitats in the rainy season (Figure 2D).

Anopheline larval habitats were generally clustered near the streams (Figures 1 and 3). In the rainy season survey, 84.4% of anopheline-positive habitats were located within 100 meters from streams, whereas 93.6% of positive habitats were distributed within 100 meters in the dry season (Kolmogorov-Smirnov one-sample test: Z = 7.63, n = 277, P < 0.001 for the rainy season and Z = 2.66, n = 47, P < 0.001 for the dry season). Larval habitats were also clustered in the lower area; 75.1% and 100% of anopheline-positive habitats were found in the valley bottom within 1,400–1,420-meters range of elevation during the rainy season and dry season, respectively (Figure 4). The average distance of anopheline larval habitats to the nearest stream in the dry season (mean = 54.8 meters, SE = 4.9) was significantly shorter than in the rainy season (mean = 67.0 meters, SE = 4.9, t = 4.35, P = 0.038). The average elevation of anopheline-positive habitats in the rainy season was significantly higher than in the dry season (rainy season: mean = 1,415.7 meters, SE = 0.68, n = 277 and dry season: mean = 1,408.0 meters, SE = 1.63, n = 48, t = 4.35, degrees of freedom [df] = 323, P < 0.001). Logistic regression analysis showed that the occurrence of anopheline larvae was significantly correlated with elevation and season (Table 2).

Relationships with land cover types.

Farmlands were the major land cover type, covering 59.6% of the study area; this was followed by pastures (20.0%) and forests (17.9%; Table 1). Swamps had little coverage (1.4%). Farm, forest, and pasture were almost equally dominant in the area within 50 meters from streams (Figure 5); however, the proportion of farm to the other land cover types increased with an increase in distance from streams. More than one-third (35.9%) of the total forest cover was present in the area within 50 meters from streams. The aquatic habitats in forests had significantly higher canopy coverage than the habitats in farmlands, pastures, and swamps (Table 3). The proportion of habitats with filamentous algae was significantly higher in farmlands and swamps than in forests and pastures. However, the nitrogen and phosphate contents, pH level, and turbidity of aquatic habitats were not significantly varied among the land cover types.

Anopheline-positive habitats were found most in farmlands (39.0%), followed by pastures (26.7%), during the rainy season (Table 1). During the dry season, positive habitats occurred most in the swamps (41.7%), followed by farmlands (33.3%). The average size of habitats in swamps was significantly larger than in the other land cover types (Table 3). The association between the occurrence (presence/absence) of anopheline larvae and land cover types was statistically significant during the rainy season (χ2 = 36.69, df = 4, P < 0.001), but not significant during the dry season (χ2 = 5.13, df = 4, P = 0.274; Table 1). In particular, the proportion of positive habitats to negative habitats was highest in the farmlands and pastures in both the rainy season (range = 42.8–44.1%) and the dry season (range = 34.0–35.3%), but lowest in road ditches (16.4% and 11.1% for the rainy and dry seasons, respectively).

We found that the larvae of the An. gambiae complex were found more frequently in farmlands and pastures than in larval habitats in the forests (χ2 = 10.53, df = 3, P = 0.015), while An. funestus larvae were found more frequently in swamps and pastures than in farmlands and forests (χ2 = 9.85, df = 3. P = 0.020; Table 4). Among the eight physicochemical variables examined, canopy cover was the only factor significantly associated with the occurrence of the An. gambiae complex and An. funestus larvae (Table 5). The average canopy cover was significantly less in the An. gambiae complex–positive habitats and in the An. funestus–positive habitats than in the negative habitats.

DISCUSSION

The present study demonstrated that topographic features and land cover types play important roles in the spatial distribution of anopheline larval habitats. Anopheline larval habitats were generally located near the valley bottom or streams in our study area, which was a typical highland site in western Kenya characterized by hill-valley topography. Larval habitat distribution became more confined toward the valley bottoms during the dry season. Topographic features affect the formation of aquatic habitats. For example, stagnant aquatic habitats are more prevalent in valley bottoms than on hills because it is more difficult for water to accumulate on hill slopes due to surface runoff, whereas in the valley bottom stagnant aquatic habitats are formed by means of surface runoff from uphill, and from springs and groundwater seepage. Since the groundwater level is lowered during the dry season, the distribution of stagnant aquatic habitats becomes more confined toward valley bottoms. In our study area, most parts of roads are in the mid-slope of hills where rainwater drains quickly. As a result, in our study area puddles associated with roads were short-lived; in contrast, anopheline larvae are often found in habitats associated with roads in the lower area of the Lake Victoria Basin.17,32 The Lake Victoria Basin is generally more flat, and anopheline larval habitats are widely dispersed.29,30

The results from this study suggest that land cover types affect availability and suitability of aquatic habitats for anopheline larvae. Among the physicochemical variables of aquatic habitats that we examined, we found that canopy cover was the only factor significantly associated with the occurrence of the An. gambiae complex and An. funestus larvae. Unshaded aquatic habitats occurred more often in farmlands and pastures than forests. Therefore, the An. gambiae complex larval habitats were more frequently observed in farmlands and pastures. Larval habitats of the An. gambiae complex also commonly occur in temporary habitats in swamp margins in valley bottoms.18 It is interesting to note that the number of anopheline larval habitats in farmlands and pastures was greatly reduced in the dry season, mostly due to low availability of aquatic habitats. Farmers stop drawing water to farmlands from streams through irrigation ditches during the dry season, and there is little rainfall. Anopheles funestus also prefer habitats in unforested areas, but they tend to occur more often in larger, semipermanent or permanent habitats with emergent plants.30,32,33 The confined distribution of An. funestus larval habitats observed in the present study is probably due to limited distribution of longstanding aquatic habitats.

Frequent occurrence of the An. gambiae complex larvae in temporary and sunlit habitats may result from a combination of several factors. At least three plausible explanations for this phenomenon are suggested. First, the An. gambiae complex females preferentially select small, open habitats for oviposition.35 Second, larval predation is less prevalent in temporary habitats than in large, permanent habitats.19,3638 Third, the An. gambiae complex is a typical r-strategist (a species that is opportunistic and reproduces rapidly when density-independent limiting factors are not present), exploiting the increased resources of warmer, open habitats that tend to produce more algae (the main food source for the An. gambiae complex) than do shaded habitats.39 Since maize is a dominant crop in this study area, maize pollen, which is an excellent food source, may also enhance survivorship of the An. gambiae complex in open habitats in farmlands.40,41 Moreover, warmer temperatures encountered in small and open habitats during daytime hours shorten larvae-to-pupae development time, and subsequently larval mortality due to desiccation is reduced.32 The An. gambiae complex may have evolved to exploit these favorable conditions by selecting small and open habitats for oviposition.

Our results suggest that land use practices affect the growing conditions of anopheline larvae and habitat availability in the African highlands. Patches of forest still remain along streams in the Kenyan highlands such as our study area, and clearing these riparian forests will create more larval habitats for the malaria vectors. Assessed from the perspective of larval ecology, our results support the hypothesis that clearing forests leads to an increased incidence of malaria.15 This study also demonstrated that mapping the distribution of anopheline larval habitats may be used as a foundation for larval control. Our finding on the aggregated distribution of larval habitats in the highland sites suggests that larval control can be targeted to aquatic habitats in farmlands and pastures near streams or valley bottoms. Larval control may be more effective if implemented during the dry season when the distribution of larval habitats is more strictly confined toward valley bottoms.

Table 1

Total surface area (m2) of each land cover type and occurrence of habitats found during the rainy season and dry season*

FarmForestPastureRoadSwamp
* The letters following the proportion of anopheline-positive sites indicate the results of the Tukey-type multiple comparison tests. The values with the same letter were not statistically significant at the P = 0.05 level after adjustment with the Bonferroni correction.
† The total does not include the surface area coverage of the river and streams within the study area, which was 75,477 m2. Therefore, the actual size of the study area was 4,066,624 (m2).
Total surface area (m2) (%)†2,377,923 (59.6)715,984 (17.9)799,061 (20.0)42,942 (1.1)55,237 (1.4)
Rainy season
    Number of anopheline-positive sites (%)108 (44.1)a44 (34.4)b74 (42.8)ac23 (16.4)d28 (34.6)bc
    Number of anopheline-negative sites (%)137 (55.9)84 (65.6)99 (57.2)117 (83.6)53 (65.4)
    Total number of sites (%)245 (100)128 (100)173 (100)140 (100)81 (100)
Dry season
    Number of anopheline-positive sites (%)†16 (34.0)a5 (29.4)ab6 (35.3)a1 (11.1)b20 (20.0)b
    Number of anopheline-negative sites (%)31 (66.0)12 (70.6)11 (64.7)8 (88.9)80 (80.0)
    Total number of habitats (%)47 (100)17 (100)17 (100)9 (100)100 (100)
Table 2

Results of logistic regression analysis determining whether the occurrence of anopheline larvae was correlated with three independent variables, the distance of a habitat from the nearest stream, elevation, and season

PresenceAbsenceP
* Values for distance and elevation are means ± SE.
† Values for season are numbers of habitats (%).
Distance (m)*58.6 ± 3.360.8 ± 2.40.147
Elevation (m)*1,414.5 ± 0.81,418.2 ± 0.6< 0.001
Season†< 0.001
    Rainy277 (36.1%)490 (63.9%)
    Dry48 (25.3%)142 (74.7%)
Total number of habitats325632
Table 3

Characteristics of aquatic habitats found in different land cover types*

VariablesFarmForestPastureSwamp
* Values are the mean ± SE except for the proportions (%) of habitats with filamentous algae. The letters following the numerical values indicate the results of Tukey-type multiple comparison tests. The values with the same letter were not statistically significant at the P = 0.05 level after adjustment with the Bonferroni correction. NTU = nephelometric turbidity units.
Canopy cover (%)33.8 ± 4.1a87.5 ± 2.4b21.6 ± 3.9a18.4 ± 2.8a
Habitat size (m3)18.0 ± 7.9a25.4 ± 10.0a11.5 ± 4.0a70.5 ± 14.4b
Habitat with algae (%)31.8a10.5b6.8b25.9a
NH4+ (mg/L)0.89 ± 0.16a2.00 ± 0.60a1.03 ± 0.18a1.99 ± 0.82a
NO3 (mg/L)1.11 ± 0.35a0.66 ± 0.40a0.62 ± 0.22a0.17 ± 0.11a
PO4−3 (mg/L)0.38 ± 0.10a0.31 ± 0.00a0.18 ± 0.06a0.18 ± 0.06a
ph6.3 ± 0.1a6.1 ± 0.1a6.2 ± 0.1a6.1 ± 0.0a
Turbidity (NTU)98.5 ± 21.0a85.9 ± 12.7a58.5 ± 10.0a95.2 ± 11.9a
Number of sites44384454
Table 4

Occurrence of Anopheles gambiae complex and An. funestus larval habitats in different land cover types*

FarmlandForestPastureSwamp
* The letters following the proportion of the An. gambiae complex–or An. funestus-positive sites indicate the results of the Tukey-type multiple comparison tests. The values with the same letter were not statistically significant at the P = 0.05 level after adjustment with the Bonferroni correction.
An. gambiae complex (%)19 (43.2%)a7 (18.4%)b21 (47.7%)a17 (31.5%)a
An. funestus larvae (%)2 (4.6%)a1 (2.6%)a11 (25.0%)b10 (18.5%)b
Total number of habitats44384454
Table 5

Characteristics of habitats with/without the Anopheles gambiae complex and An. funestus larvae, and results of logistic regression analysis determining which variables were correlated with the occurrence of the mosquitoes*

An. gambiae complexAn. funestus
VariablesPresenceAbsencePPresenceAbsenceP
* Values are the mean ± SE except for the proportions (%) of habitats with filamentous algae. NTU = nephelometric turbidity units.
Canopy cover26.4 ± 3.643.7 ± 3.40.00220.7 ± 5.740.1 ± 2.80.021
Habitat size (m3)25.6 ± 7.438.2 ± 7.50.52528.9 ± 8.334.5 ± 6.30.07
Habitat with algae (%)18.819.80.7699.026.80.135
NH4+ (mg/L)1.09 ± 0.151.71 ± 0.430.8080.81 ± 0.231.60 ± 0.320.243
NO3 (mg/L)0.13 ± 0.040.15 ± 0.040.9780.15 ± 0.070.14 ± 0.030.991
PO4−3 (mg/L)0.27 ± 0.060.26 ± 0.050.9370.25 ± 0.110.27 ± 0.040.384
ph6.3 ± 0.16.2 ± 0.00.0756.2 ± 0.16.2 ± 0.00.527
Turbidity (NTU)79.4 ± 12.288.2 ± 9.10.99449.6 ± 7.290.5 ± 8.20.089
Number of sites7410626154
Figure 1.
Figure 1.

Distributions of anopheline larval habitats in the rainy season and dry season in the study area in Kakamega, western Kenya. m = meters.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 1; 10.4269/ajtmh.2005.73.157

Figure 2.
Figure 2.

Results of K-function analysis on the spatial relationship between anopheline larval habitats and distances to the nearest streams in the rainy season (A) and dry season (B), and results of bivariate K-function analysis comparing the distribution patterns between the rainy and dry seasons (C) and between the Anopheles gambiae complex and An. funestus (D). The bold lines are the observed value of the test statistic L(d) at a given distance d, and the gray lines indicate the 95% confidence intervals. m = meters.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 1; 10.4269/ajtmh.2005.73.157

Figure 3.
Figure 3.

Distribution of anopheline larval habitats with respect to the distance to the nearest stream. m = meters.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 1; 10.4269/ajtmh.2005.73.157

Figure 4.
Figure 4.

Distribution of anopheline larval habitats with respect to elevation. m = meters.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 1; 10.4269/ajtmh.2005.73.157

Figure 5.
Figure 5.

Relationships between area of each land cover type and distance from streams. m = meters.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 1; 10.4269/ajtmh.2005.73.157

*

Address correspondence to Noboru Minakawa, Faculty of Medicine, Saga University, 5-1-1 Nabeshima, Saga 849-8501, Japan. E-mail: minakawa@post.saga-med.ac.jp

Authors’ addresses: Noboru Minakawa, Faculty of Medicine, Saga University, 5-1-1 Nabeshima, Saga 849-8501, Japan, E-mail: minakawa@post.saga-med.ac.jp. Stephen Munga, Francis Atieli, Andrew K. Githeko, Kenya Medical Research Institute, Kisumu, Kenya, E-mails: SMunga@kisian.mimcom.net, Fatieli@kisian.mimcom.net, and agitheko@kisian.mimcom.net. Emmanuel Mushinzimana, Guofa Zhou, and Guiyun Yan, Department of Biological Sciences, State University of New York, Buffalo, NY 14260, E-mails: Emushinzimana@kisian.mimcom.net, gzhou2@acsu.buffalo.edu, and gyan@buffalo.edu.

Acknowledgments: We thank H. Atieli, B. Ndenga, M. Okonji, S. Ajuma, W. Miheso, and P. Lutiali for field assistance, U. Kitron and J. Clennon for assistance with statistical analyses, M. Ombok for assistance with the global positioning system in differential mode, L. Carson for critical comments, and T. Wiegand for providing the statistical software package Programita.

Financial support: This work was supported by National Institutes of Health grants R01 AI 50243 and D43 TW01505.

REFERENCES

  • 1

    Roberts JM, 1964. The control of epidemic malaria in the highlands of Western Kenya. I. before the campaign. J Trop Med Hyg 67 :161–168.

    • Search Google Scholar
    • Export Citation
  • 2

    Malakooti MA, Biomndo K, Shanks GD, 1998. Reemergence of epidemic malaria in the highlands of western Kenya. Emerg Infect Dis 4 :671–676.

    • Search Google Scholar
    • Export Citation
  • 3

    Lindblade KA, Walker ED, Onapa AW, Katungu J, Wilson ML, 1999. Highland malaria in Uganda: prospective analysis of an epidemic associated with El Nino. Trans R Soc Trop Med Hyg 93 :480–487.

    • Search Google Scholar
    • Export Citation
  • 4

    Shanks GD, Biomndo K, Hay SI, Snow RW, 2000. Changing patterns of clinical malaria since 1965 among a tea estate population located in the Kenyan highlands. Trans R Soc Trop Med Hyg 94 :253–255.

    • Search Google Scholar
    • Export Citation
  • 5

    John CC, Ouma JH, Sumba PO, Hollingdale MR, Kazura JW, King CL, 2002. Lymphocyte proliferation and antibody responses to Plasmodium falciparum liver-stage antigen-1 in a highalnd area of Kenya with seasonal variation in malaria transmission. Am J Trop Med Hyg 66 :372–378.

    • Search Google Scholar
    • Export Citation
  • 6

    John CC, O’Donnell RA, Sumba PO, Moormann AM, de Koning-Ward TF, Kazura JW, Crabb BS, 2004. Evidence that invation-inhibitory antibodies specific for the 19-kDa fragment of merozoite surface protein-1 (MSP-1 19) can play a protective role against blood-stage Plasmodium falciparum infection in individuals in a malaria endemic area of Africa. J Immunol 173 :666–672.

    • Search Google Scholar
    • Export Citation
  • 7

    Lindsay SW, Martens WJ, 1998. Malaria in the African highlands: past, present and future. Bull World Health Organ 76 :33–45.

  • 8

    Mouchet J, Manguin S, Sircoulon J, Laventure S, Faye O, Onapa AW, Carnevale P, Julvez J, Fontenille D, 1998. Evolution of malaria in Africa for the past 40 years: impact of climatic and human factors. J Am Mosq Control Assoc 14 :121–130.

    • Search Google Scholar
    • Export Citation
  • 9

    Lindblade KA, Walker ED, Onapa AW, Katungu J, Wilson ML, 2000. Land use change alters malaria transmission parameters by modifying temperature in a highland area of Uganda. Trop Med Int Health 5 :263–274.

    • Search Google Scholar
    • Export Citation
  • 10

    Githeko AK, Lindsay SW, Confalonieri UE, Patz JA, 2000. Climate change and vector-borne diseases: a regional analysis. Bull World Health Organ 78 :1136–1147.

    • Search Google Scholar
    • Export Citation
  • 11

    Zhou G, Minakawa N, Githeko AK, Yan G, 2004. Association between climate variability and malaria epidemics in the east African highlands. Proc Natl Acad Sci U S A 101 :2375–2380.

    • Search Google Scholar
    • Export Citation
  • 12

    United Nations, 2003. World Population Prospects: the 2002 Revision. New York: United Nations.

  • 13

    Brooks TM, Pimm SL, Oyugi JO, 1999. Time lag between deforestation and bird extinction in tropical forest fragments. Conservation Biol 13 :1140–1150.

    • Search Google Scholar
    • Export Citation
  • 14

    Whitmore TC, 1997. Tropical forest disturbance, disappearance, and species loss. Laurence WL, Bierregaard Jr RO, eds. Tropical Forest Remnants. Chicago: University of Chicago Press, 2–28.

  • 15

    Walsh JF, Molyneux DH, Birley MH, 1993. Deforestation: effects on vector-borne disease. Parasitology 106 (Suppl):S55–S75.

  • 16

    Patz JA, Graczyk TK, Geller N, Vittor AY, 2000. Effects of environmental change on emerging parasitic diseases. Int J Parasitol 30 :1395–1405.

    • Search Google Scholar
    • Export Citation
  • 17

    Minakawa N, Mutero CM, Githure JI, Beier JC, Yan G, 1999. Spatial distribution and habitat characterization of anopheline mosquito larvae in western Kenya. Am J Trop Med Hyg 61 :1010–1016.

    • Search Google Scholar
    • Export Citation
  • 18

    Minakawa N, Sonye G, Mogi M, Yan G, 2004. Habitat characteristics of Anopheles gambiae s.s. larvae in a Kenyan highland. Med Vet Entomol 18 :301–305.

    • Search Google Scholar
    • Export Citation
  • 19

    Service MW, 1977. Mortalities of the immature stages of species B of the Anopheles gambiae complex in Kenya: comparison between rice fields and temporary pools, identification of predators, and effects of insecticidal spraying. J Med Entomol 13 :535–545.

    • Search Google Scholar
    • Export Citation
  • 20

    Khan B, Ofulla AV, Kariuki DM, Githure JI, Kabiru EW, Martin SK, 1992. Drug sensitivity studies during a highland malaria epidemic in Kenya. Trans R Soc Trop Med Hyg 86 :371–372.

    • Search Google Scholar
    • Export Citation
  • 21

    Hightower AW, Ombok M, Otieno R, Odhiambo R, Oloo AJ, Lal AA, Nahlen BL, Hawley WA, 1998. A geographic information system applied to a malaria field study in western Kenya. Am J Trop Med Hyg 58 :266–272.

    • Search Google Scholar
    • Export Citation
  • 22

    Lemmon PE, 1956. A spherical densiometer for estimating forest overstory density. Forest Sci 2 :314–320.

  • 23

    USEPA, 1983. Methods for Chemical Analysis of Water and Wastes. Washington, DC: U.S. Environmental Protection Agency.

  • 24

    APHA, 1992. Standard Methods for the Examination of Water and Wastewater. Washington, DC: American Public Health Association.

  • 25

    Gillies MT, Coetzee M, 1987. A Supplement to the Anophelinae of Africa South of the Sahara. Johannesburg: The South African Institute for Medical Research.

  • 26

    Scott JA, Brogdon WG, Collins FH, 1993. Identification of single specimens of the Anopheles gambiae complex by the polymerase chain reaction. Am J Trop Med Hyg 49 :520–529.

    • Search Google Scholar
    • Export Citation
  • 27

    Koekemoer LL, Kamau L, Hunt RH, Coetzee M, 2002. A cocktail polymerase chain reaction assay to identify members of the Anopheles funestus (Diptera: Culicidae) group. Am J Trop Med Hyg 66 :804–811.

    • Search Google Scholar
    • Export Citation
  • 28

    Ripley BD, 1976. The second-order analysis of stationary point processes. J Appl Probabil 13 :255–266.

  • 29

    Minakawa N, Githure JI, Beier JC, Yan G, 2001. Anopheline mosquito survival strategies during the dry period in western Kenya. J Med Entomol 38 :388–392.

    • Search Google Scholar
    • Export Citation
  • 30

    Minakawa N, Seda P, Yan G, 2002. Influence of host and larval habitat distribution on the abundance of African malaria vectors in western Kenya. Am J Trop Med Hyg 67 :32–38.

    • Search Google Scholar
    • Export Citation
  • 31

    Wiegand T, Moloney KA, 2004. Rings, circles, and null-models for point pattern analysis in ecology. Oikos 104 :209–229.

  • 32

    Gillies MT, De Meillon B, 1968. The Anophelinae of Africa South of the Sahara. Johannesburg: The South African Institute for Medical Research.

  • 33

    Gimnig JE, Ombok M, Kamau L, Hawley WA, 2001. Characteristics of larval anopheline (Diptera: Culicidae) habitats in western Kenya. J Med Entomol 38 :282–288.

    • Search Google Scholar
    • Export Citation
  • 34

    Zar JH, 1999. Biostatistical Analysis. Englewood Cliffs, NJ: Prentice Hall.

  • 35

    Bentley MD, 1989. Chemical ecology and behavioral aspects of mosquito oviposition. Annu Rev Entomol 34 :402–421.

  • 36

    Mogi M, 1981. Population dynamics and methodology for bio-control of mosquitoes. Laird M, ed. Biocontrol of Medical and Veterinary Pests. New York: Praeger, 140–172.

  • 37

    Washburn JO, 1995. Regulatory factors affecting larval mosquito populations in container and pool habitats: implications for biological control. J Am Mosq Control Assoc 11 :279–283.

    • Search Google Scholar
    • Export Citation
  • 38

    Sunahara T, Ishizaka K, Mogi M, 2002. Habitat size: a factor determining the opportunity for encounters between mosquito larvae and aquatic predators. J Vector Ecol 27 :8–20.

    • Search Google Scholar
    • Export Citation
  • 39

    Gimnig JE, Ombok M, Otieno S, Kaufman MG, Vulule JM, Walker ED, 2002. Density-dependent development of Anopheles gambiae (Diptera: Culicidae) larvae in artificial habitats. J Med Entomol 39 :162–172.

    • Search Google Scholar
    • Export Citation
  • 40

    Ye-Ebiyo Y, Pollack RJ, Spielman A, 2000. Enhanced development in nature of larval Anopheles arabiensis mosquitoes feeding on maize pollen. Am J Trop Med Hyg 63 :90–93.

    • Search Google Scholar
    • Export Citation
  • 41

    Ye-Ebiyo Y, Pollack RJ, Kiszewski A, Spielman A, 2003. Enhanced development of larval Anopheles arabiensis by proximity to flowering maize (Zea mays) in turbid water and when crowded. Am J Trop Med Hyg 68 :748–752.

    • Search Google Scholar
    • Export Citation

Author Notes

Reprint requests: Dr. Noboru Minakawa, Faculty of Medicine, Saga University, 5-1-1 Nabeshima, Saga 849-8501, Japan.
Save