Journal of Heredity 2003:94(6):457–463
DOI: 10.1093/jhered/esg100
Ó 2003 The American Genetic Association
Population Genetic Structure of
Anopheles arabiensis Mosquitoes in
Ethiopia and Eritrea
S. R. G. NYANJOM, H. CHEN, T. GEBRE-MICHAEL, E. BEKELE, J. SHILILU, J. GITHURE, J. C. BEIER,
AND G. YAN
From the Institute of Pathobiology and Department of Biology, Addis Ababa University, Addis Ababa, Ethiopia (Nyanjom,
Gebre-Michael, and Bekele), Department of Biological Sciences, State University of New York at Buffalo, Buffalo, NY 14260
(Chen and Yan), International Center for Insect Physiology and Ecology (ICIPE), Nairobi, Kenya (Shililu and Githure), and
Department of Tropical Medicine, Tulane University Medical Center, New Orleans, LA 70112 (Beier).
Address correspondence to Guiyun Yan at the address above, or e-mail: [email protected].
Abstract
This study examined the population genetic structure of the major malaria vector, Anopheles arabiensis mosquitoes, in Ethiopia
and Eritrea. Ethiopia and Eritrea have great geographical diversity, with high mountains, rugged plateaus, deep gorges, and
rolling plains. The plateau is bisected diagonally by the Great Rift Valley into the Northwestern Highlands and the
Southeastern Highlands. Five A. arabiensis populations from the Northwestern Highlands region and two populations from
high-altitude sites in the Great Rift Valley were genotyped using six microsatellite markers to estimate the genetic diversity
and population genetic structure of A. arabiensis. We found that A. arabiensis populations from the Northwestern Highlands
and the Great Rift Valley region showed a similar level of genetic diversity. The genetic differentiation (FST) of the five
mosquito populations within the Northwestern Highlands region was 0.038 (P , .001), while the two populations within the
Great Rift Valley showed little genetic differentiation (FST ¼ 0.007, P , .01). The degree of genetic differentiation between
the Northwestern Highlands region and the Great Rift Valley region was small but statistically significant (FST ¼ 0.017, P ,
.001). The population genetic structure of A. arabiensis in the study area did not follow the isolation-by-distance model (r2 ¼
0.014, P . .05). The low FST estimates for A. arabiensis populations in Ethiopia and Eritrea are consistent with the general
population genetic structure of this species in East Africa based on other molecular markers.
Malaria is a major public health problem in Africa. The World
Health Organization (WHO) estimates that there are about
300–500 million clinical malaria cases annually and that 1–1.5
million people die of malaria each year. Anopheles arabiensis and
Anopheles gambiae are the principal vectors of malaria in subSaharan Africa, but in some areas, such as the Great Rift
Valley in East Africa, A. arabiensis is the predominant malaria
vector species (Minakawa et al. 2002). A. arabiensis is better
adapted to dry environments than A. gambiae (Lindsay et al.
1998). Currently the most efficient malaria prevention
method is to reduce human vector contacts using insecticide-impregnated bednets (Goodman et al. 2001; WHO
1993). The mosquito vectors have been developing resistance
to insecticides in areas where insecticide-impregnated bednets
and pesticide sprays targeting agricultural pests are intensively
applied (Chandre et al. 1999; Curtis et al. 1998). Information
on the population genetic structure of the anopheline vectors
is useful for predicting the spread of insecticide-resistance
genes. In addition, this information is critical for designing
rational strategies for spreading parasite-inhibiting genes
through releasing transgenic mosquitoes to reduce malaria
transmission, an approach currently under active investigation (Atkinson and Michel 2002; Crampton et al. 1994).
Ethiopia and Eritrea are located in the horn of Africa.
About 75% of the two countries is malarious, with 65% of
the population at risk of infection (WHO 1998). The major
malaria vector is A. arabiensis; Anopheles funestus and Anopheles
nili play a secondary role (Abose et al. 1998). Cytological
analyses of A. arabiensis established that 2Rb and 3Ra
inversions occur in high frequencies within the Great Rift
Valley and in the southern part of Ethiopia (Mekuria et al.
1982; Nigatu et al. 1994). Even though some degree of
insecticide resistance has been encountered in different
localities, insecticide spray and insecticide-impregnated bednets are the main vector control method (WHO 1998).
Different genetic markers have been used to study the
457
Journal of Heredity 2003:94(6)
mountains are generally in the west and southeast regions of
Ethiopia. The Rift Valley region is of lower altitude (115 to
þ700 m above sea level) and is dotted by lakes and high
mountains. The present study included A. arabiensis
populations from the Northwestern Highlands of Ethiopia
and Eritrea, and the Rift Valley in Ethiopia.
Materials and Methods
Study Area
Figure 1.
Mosquito sampling sites in Ethiopia and Eritrea.
population genetic structure of A. gambiae and A. arabiensis,
including chromosomal inversions, allozymes, random
amplified polymorphic DNA (RAPD), mitochondrial
DNA (mtDNA) sequences, and microsatellite markers
(Coluzzi et al. 1979, 1985; Lanzaro et al. 1995, 1998;
Lehmann et al. 1996b, 1997, 2003; Toure et al. 1998).
Microsatellites are tandem repeats of two to five nucleotides
and have been regarded as excellent molecular markers for
studying population genetic structure because they are very
abundant throughout the genome, codominantly inherited,
usually neutral, and relatively easy to score at low cost
(Bowcock et al. 1994; Estoup et al. 1995; Goldstein and
Schlotterer 1999). Previous studies suggest that A. arabiensis
populations generally show a lower level of genetic
differentiation than A. gambiae. For example, significant
genetic differentiation (FST ¼ 0.072–0.100) was found for
A. gambiae populations between western Kenya and coastal
Kenya using microsatellite markers; hence it was inferred
that the Great Rift Valley is a major gene flow barrier for
A. gambiae (Lehmann et al. 1999, 2003), but no significant
genetic differentiation was detected for A. arabiensis
populations from the two regions using the same loci
(Kamau et al. 1999). Similarly Donnelly and Townson (2000)
did not detect significant genetic structure for A. arabiensis
populations within Malawi and Sudan; however, a high level
of genetic differentiation was found for A. arabiensis
populations from two islands that are 240 km apart in the
Indian Ocean (FST ¼ 0.080–0.215) (Simard et al. 1999).
In this study we examined the population genetic
structure of A. arabiensis from Ethiopia and Eritrea, a region
that has not been well studied. Ethiopia and Eritrea have
great geographical diversity, with high mountains, rugged
plateaus, deep gorges, and rolling plains. The Great Rift
Valley extends from the southwest to the northeast, and high
458
Ethiopia is characterized by a high, mountainous, central
plateau bisected diagonally by the Great Rift Valley into the
Northwestern Highlands and the Southeastern Highlands.
The plateau is rugged and interspersed with towering
mountains and deep chasms. Some high mountains exist in
the Great Rift Valley. Ethiopia can be divided into four
major zones: the central highlands, the western lowlands
(one-third of the country), the northeastern lowlands (hot
and humid), and the Denakil lowlands (semidesert). Eritrea is
located north of Ethiopia, and the country is divided into
three major ecological zones: the western lowlands (700–
1500 m above sea level); the highlands, with average
elevation of 2000 m above sea level; and the coastal plains
(0–300 m above sea level), which experience scanty rainfall.
The distribution of humans and anopheline mosquitoes is
not continuous across this region, but generally clustered on
the high-elevation areas where rainfall is abundant.
Mosquito Collection
Mosquitoes were collected in four localities (Arba Minch,
Debre Zeit, Gambela, and Inda Sellase) in Ethiopia and three
localities (Dasse, Hiletsidi, and Adibosqual) in Eritrea (Figure
1). Gambela, Inda Sellase, Dasse, Hiletsidi, and Adibosqual
are in the Northwestern Highlands region, while Arba Minch
and Debre Zeit are located in the Great Rift Valley region.
The altitude and the annual average maximum and minimum
temperatures and rainfall of each collection locality are listed
in Table 1. Mosquitoes were collected using the indoor
pyrethrum spray catch method (WHO 1975). Samples from
the sites in Eritrea were collected from October to
November 1999, and Ethiopian samples were collected
from September to October 2001. A. gambiae sensu lato were
morphologically separated from other anopheline species
(A. pharoensis, A. nili, A. funestus, and A. christyi) and culicine
mosquitoes, preserved in 95% ethanol, and kept at 208C.
DNA Extraction and Species Identification
Genomic DNA was extracted from individual female mosquitoes according to Yan et al. (1997). Because A. gambiae
and A. arabiensis cannot be distinguished by morphology,
the rDNA–polymerase chain reaction (PCR) method was
used to determine the species identity for specimens within
the A. gambiae species complex (Scott et al. 1993). If a
specimen failed in the initial PCR amplification, the specimen was tested in the second or third PCR until a success-
Nyanjom et al. Population Genetic Structure of Anopheles arabiensis
Table 1.
Description of mosquito sampling sites in Ethiopia and Eritrea
Annual mean of
maximum/minimum
temperature (8C)
Annual rainfall
(mm)
Number
of houses
Number of
mosquitoes
analyzed
Collection period
1009
905
1107
507
17
23
15
17
60
60
60
60
Sept.–Oct.
Sept.–Oct.
Sept.–Oct.
Sept.–Oct.
600
500
550
10
10
10
36
45
50
Oct.–Nov. 1999
Oct.–Nov. 1999
Oct.–Nov. 1999
Country
Locality
Altitude
(m)
Ethiopia
Arba Minch
Debre Zeit
Gambela
Inda Sellase
1602
1850
539
2182
28/13
18/6
34/19
25/7
Eritrea
Dasse
Hiletsidi
Adibosqual
680
450
1100
35/20
36/22
28/15.5
ful amplification was achieved. An individual was scored as
unknown if it could not be identified by three PCR tests.
Microsatellite Loci and Genotype Scoring
Six microsatellite markers developed for A. gambiae were
used to genotype A. arabiensis. Other anopheline species
were not studied because of insufficient samples. These
microsatellite markers, including AGXH1D1 and
AGXH131 on chromosome X, AG2H46 and AG2H79
on chromosome 2, and AG3H29C and AG3H33C on
chromosome 3, can be readily amplified using A. arabiensis
genomic DNA as a template (Lehmann et al. 1996b;
Zheng et al. 1996). The cytological locations of the
microsatellite markers in relation to A. arabiensis chromosomal inversions are unknown. Microsatellite analyses
were conducted on 39–60 individuals per population
(Table 1).
A Li-Cor Model 4200 Automated DNA Analyzer (LiCor Inc., Lincoln, NE) was used for gel electrophoresis.
For the apparatus to detect PCR products, one primer in
every pair of microsatellite primers must be fluorescently
labeled. To reduce the cost associated with synthesis of
fluorescently labeled primers, we used the ‘‘tailed primer’’
method (Oetting et al. 1998; Sharakhov et al. 2001); that is,
the forward primer for each microsatellite locus was
synthesized with an additional 19 bp sequence (59
CACGACGTTGTAAAACGAC 39) added to the 59 end
of the primer. A third primer with the same 19 bp
sequence was directly labeled with the fluorescence and
was used as the sole type of labeled primer for the
detection of all microsatellite alleles. The tailed primer
method reduced the cost of oligonucleotide synthesis by
more than 80%. The 10 ll PCR reaction contained 13 Taq
buffer, 0.2 mM dNTPs, 1.5 pmol forward and reverse
primers, 1.5 pmol fluorescently labeled 19 bp sequences,
1.5 mM MgCl2, 1.0 lg BSA, 1.0 unit Taq polymerase, and
about 20 ng genomic DNA. Cycling conditions in an MJ
Research PTC-220 thermocycler were 35–40 cycles of
948C for 30 s, 558C for 30 s, and 728C for 45 s. Allele sizes
were determined using Gene ImagIR computer software
(Scanalytics Inc. 1998). The allele sizes used in the analysis
were true allele sizes that were adjusted for the 19 bp tail in
the forward primer.
2001
2001
2001
2001
Data Analysis
Microsatellite polymorphism was measured by the number
of alleles and heterozygosity at each locus. Using the
probability test in the GENEPOP computer program
(Raymond and Rousset 1995), conformance with HardyWeinberg equilibrium (HWE) was tested for each locus and
population and the Bonferroni correction was applied for
multiple comparisons. We further determined whether
distortion from HWE resulted from deficient or excessive
heterozygosity using the FIS statistics and probability test.
Because the probability test is robust to low allele
frequencies, rare alleles were not pooled. Variations in
heterozygosity among the populations were analyzed
following the method of Weir (1990), using the analysis of
variance (ANOVA) with subpopulations, individuals, loci
and interactions of loci, and individuals as factors. All factors
were treated as random effects except loci. The Fisher exact
test was performed to detect genotypic linkage disequilibrium
for pairs of loci in each population (Raymond and Rousset
1995). Population genetic structure was examined using
Wright’s F-statistics (FST) using the FSTAT computer
program (Goudet 1995). The standard deviations (SD) of
the F-statistics were obtained for each locus by a jackknife
procedure over all the alleles. These were used to test the
statistical significance. RST statistics, analogous to FST but
specifically for microsatellite markers (Slatkin 1995), were
not calculated because RST is particularly sensitive to the
assumptions underlying microsatellite evolution (Gaggiotti
et al. 1999), but the mode of microsatellite evolution in
anopheline mosquitoes is unknown. In addition, the microsatellite markers used in this study were isolated from A.
gambiae; converting microsatellite allele size to the number of
tandem repeats based on A. gambiae genome sequence is not
appropriate. Genetic differentiation between regions (e.g.,
Ethiopian versus Eritrean populations, and Northwestern
Highlands and the Great Rift Valley populations) was
calculated by nested ANOVA.
The isolation-by-distance model of population genetic
structure was tested by linear regression of pairwise FST/
(1 FST) against the natural logarithm of straight-line
geographical distance between population pairs (Rousset 1997).
The statistical significance of the regression was tested using the
Mantel test with 2000 permutations. The pairwise geographical
distance of the sampling sites ranged from 76 to 877 km.
459
Journal of Heredity 2003:94(6)
Table 2. Number of alleles (n), observed heterozygosity (Ho), and inbreeding coefficients (FIS) of microsatellite markers in seven
A. arabiensis populations from Ethiopia and Eritrea
Arba Minch
Chromosome Locus
n
X
AGXH1D1 2
AGXH131 4
2
AG2H46
AG2H79
3
AG3H29C 2
AG3H33C 7
Average
9
3
Ho
He
Debre Zeit
FIS
n
Ho
Gambela
He
FIS
n
Inda Sellase
Ho
He
FIS
n
Ho
He
FIS
0.040 0.038 0.028 3 0.040 0.039 0.018 2
0.380 0.331 0.149 5 0.380 0.326 0.168 4
0.040 0.039 0.028
3
0.360 0.301 0.174** 3
0.040 0.039 0.021
0.410 0.342 0.201
0.430 0.458
0.250 0.325
0.062 10 0.440 0.506 0.131* 9
0.233 3 0.031 0.032 0.018 3
0.430 0.468 0.082
0.290 0.284 0.022
10
3
0.380 0.405
0.320 0.322
0.100 0.092 0.086 2 0.060 0.057 0.047 2
0.310 0.362 0.144 7 0.400 0.395 0.014 7
0.120 0.108 0.109
0.370 0.392 0.057
3
7
0.068 5 0.225 0.226 0.003 4.5 0.268 0.265 0.008
4.5 0.252 0.268
Dasse
Hiletsidi
0.063
0.008
0.117 0.130 0.113
0.370 0.405 0.088
4.8 0.273 0.274 0.004
Adibosqual
Chromosome
Locus
n
Ho
He
FIS
n
Ho
He
FIS
n
Ho
He
FIS
X
AGXH1D1
AGXH131
3
4
0.030
0.150
0.029
0.190
0.019
0.214
4
4
0.030
0.220
0.030
0.231
0.012
0.048
3
4
0.020
0.240
0.020
0.273
0.006
0.121
2
AG2H46
AG2H79
9
4
0.130
0.130
0.244
0.155
8
4
0.060
0.170
0.250
0.192
0.763***
0.115
6
4
0.350
0.200
0.343
0.279
0.020
0.286
3
AG3H29C
AG3H33C
3
7
0.150
0.170
0.175
0.156
0.144
0.089
3
8
0.150
0.250
0.220
0.250
0.321
0.001
3
9
0.250
0.190
0.233
0.220
0.073**
0.139
Average
5
0.127
0.158
0.181
5.2
0.147
0.196
0.277
4.8
0.208
0.228
0.094
0.473***
0.164
* P , 0.05, ** P , 0.01, *** P , 0.001.
Results
Species Identification
A total of 550 mosquitoes in the A. gambiae species complex,
collected from seven localities in Ethiopia and Eritrea, were
identified as A. arabiensis in the PCR identifications. Neither
A. gambiae nor A. quadriannulatus were detected.
and 0.363 for Debre Zeit, Dasse, and Hiletsidi, respectively.
Exact tests revealed that only four pairs of loci out of a total
of 105 possible pairs (3.8%) were in linkage disequilibrium
(data not shown), suggesting a very low level of linkage
disequilibrium among the six loci studied.
Population Genetic Structure and Isolation by Distance
Genetic Diversity, HWE, and Linkage Disequilibrium
Population genetic diversity can be measured by the number
of alleles and expected heterozygosity. Loci AGX1D1 and
AG3H29C were the least polymorphic markers, while loci
AG2H46 and AG3H33C were the most polymorphic (Table
2). The average number of alleles over the six markers ranged
from 4.5 to 5.2 per locus and did not vary significantly among
the populations (F ¼ 0.05, df ¼ 6, P . .05). Among the
seven studied populations, the expected heterozygosity
ranged from 0.158 to 0.274 per locus. The observed
heterozygosity over the six loci did not differ significantly
(F ¼ 0.53, df ¼ 6, P . .05). The genotype frequencies at
several loci did not conform to HWE. Of the 42 tests (6
loci 3 7 populations) for conformance to HWE at the locus
level within populations, 13 tests (31%) showed significant
distortion from HWE, and the majority (11 tests) were due
to significant heterozygote deficiency (Table 2). Locus
AG2H46 showed significant heterozygote deficiency in
three of seven populations, likely resulting from the presence
of null alleles (Lehmann et al. 1996a). The maximum
likelihood estimate of null allele frequency is 0.063, 0.229,
460
The seven populations exhibited remarkably similar allele
distributions and shared the same predominant alleles at four
loci (AGXH1D1, AGXH131, AG2H79, and AG3H29C;
data not shown; microsatellite allele frequency data are
available to interested parties upon request). The genetic
differentiation of the four northern populations (Adibosqual,
Hiletsidi, Dasse, and Inda Sellase) (FST ¼ 0.046, P , .001)
was sixfold higher than the three southern populations
(Gambela, Debre Zeit, and Arba Minc) (FST ¼ 0.007, P ,
.01; Table 3), despite the fact that the four northern sites are
more tightly clustered. The genetic differentiation (FST)
between populations from the northern and southern
regions was 0.018 (P , .05). Very small genetic differentiation existed for the two populations in the Great Rift Valley
(Debre Zeit and Arba Minch) (FST ¼ 0.007, P , .01; Table
3), and between the Northwestern Highlands and the Great
Rift Valley regions (FST ¼ 0.017, P , .001).
A Mantel test revealed no significant correlation between
pairwise FST/(1 FST) against the natural logarithm of
pairwise geographical distance (r2 ¼ 0.014, P ¼ .15),
suggesting that the population genetic structure of A.
Nyanjom et al. Population Genetic Structure of Anopheles arabiensis
Table 3.
FST estimates of A. arabiensis populations from Ethiopia and Eritrea
Locus
between northern of 5 populations
between western
of all 7
of 4 northern of 3 southern and southern
in western
of 2 populations highlands and
populations populations
populations
regions
highlands
in Rift Valley
Rift Valley
AGXH1D1
AGXH131
AG2H46
AG2H79
AG3H29C
AG3H33C
0.002
0.011*
0.042***
0.019**
0.152***
0.007**
0.000
0.021*
0.060***
0.027**
0.142***
0.012
0.001
0.004
0.016**
0.003
0.001
0.004**
0.060
0.001
0.018*
0.009*
0.101***
0.001
0.002
0.013**
0.048***
0.018**
0.145***
0.008**
0.004
0.008
0.021**
0.004
0.001
0.000
0.001
0.002
0.017***
0.017*
0.073***
0.004
Overall
0.034***
0.046***
0.007**
0.018*
0.038***
0.007**
0.017***
* P , 0.05, ** P , 0.01, *** P , 0.001.
arabiensis in Ethiopia and Eritrea did not conform to the
isolation-by-distance model.
Discussion
This study confirmed that A. arabiensis is the predominant
malaria vector species in Ethiopia and Eritrea. A. arabiensis
populations from the Northwestern Highlands region and
the Great Rift Valley region had similar levels of genetic
diversity as measured by the number of alleles and observed
heterozygosity. Small, but statistically significant genetic
structure was identified between Ethiopian and Eritrean
populations, and between the Northwestern Highlands and
the Great Rift Valley regions. The population genetic
structure of A. arabiensis in Ethiopia and Eritrea did not
conform to the isolation-by-distance model.
In our mosquito sampling from Ethiopia and Eritrea,
A. pharoensis, A. nili, A. funestus, and A. christyi were also
found, but A. gambiae and A. quadriannulatus were not detected. This result is consistent with previous findings that A.
arabiensis is the predominant vector responsible for malaria
transmission in Ethiopia (Abose et al. 1998; Donnelly et al.
2001; Lulu et al. 1991; Mekuria et al. 1982; Nigatu et al. 1994).
Climatic factors, such as precipitation and temperature, are
important determinants of the range and relative abundance
of individual species of the A. gambiae complex (Lindsay et al.
1998). A. gambiae is usually the predominant species in
saturated environments, but A. arabiensis is more common in
arid areas. Although climatic conditions are directly
associated with elevation (temperature decreases and rainfall
generally increases with elevation), elevation is not a good
predictor of malaria vector species distribution. For example,
in western Kenya, A. gambiae but not A. arabiensis was found
in areas with an elevation greater than 1500 m above sea
level, and A. gambiae was more abundant than A. arabiensis in
the basin region of Lake Victoria, which has an elevation of
1200–1300 m above sea level (Minakawa et al. 2002; Shililu et
al. 1998). In the present study, no A. gambiae was found in
sampling sites with an elevation of 450 to 2182 m above sea
level in Ethiopia and Eritrea. Moisture index, calculated as
the ratio of precipitation to potential evapotranspiration, is
a better predictor of A. gambiae and A. arabiensis species
distribution (Lindsay et al. 1998; Minakawa et al. 2002).
Genetic diversity of the seven A. arabiensis populations
used in this study was similar, as measured by the number of
alleles and observed heterozygosity. Because our study and
others used several common microsatellite markers, it is
possible to compare these results. Our estimation of genetic
diversity was similar to the Kenyan population (Kamau et al.
1999), but considerably lower than the populations in other
East African countries, for reasons that are unclear. For
example, the observed heterozygosity of A. arabiensis
populations in nine localities along a 4500 km transect from
Sudan to Mozambique ranged from 0.47 to 0.73 for locus
AG3H33C and from 0.527 to 0.732 for AG2H46 (Donnelly
and Townson 2000). In comparison to one Ethiopian
population used by Donnelly and Townson (2000), our
current study with comparable sample sizes detected one to
two fewer alleles and 30% less observed heterozygosity in
these two loci. That fewer alleles were detected in our study
was not likely caused by noncanonical alleles that have either
gone undetected or been binned into other classes, because
deviation from HWE was observed in only 5 of 42 tests
(12%). Further, the heterozygosity deficit occurred only in
locus AG2H46, a locus where null alleles are known to occur
in A. gambiae (Lehmann et al. 1996a).
We observed small, but statistically significant genetic
structure for A. arabiensis populations in Ethiopia and Eritrea
(FST ¼ 0.034), and between the Northwestern Highlands and
the Great Rift Valley regions (FST ¼ 0.017). The extent of
genetic differentiation between the Northwestern Highlands
and the Great Rift Valley regions was about half of the
within-population variation of the Northwestern Highlands
region (FST ¼ 0.038). The particular geographic location of
these two sites in the Great Rift Valley may cause this
phenomenon: These two sites are located in the extension of
the mountainous plateau of the Northwestern Highlands
area, as indicated by the high elevation of the sites, and thus
there is no obvious physical barrier between these two
particular study sites in the Great Rift Valley and the
Northwestern Highlands. The geographic distance between
the northern and southern populations should not be a major
factor because the population genetic structure of A.
arabiensis in the study area did not follow the isolation-bydistance model.
The low FST estimates with A. arabiensis populations in
461
Journal of Heredity 2003:94(6)
this study are consistent with the general population genetic
structure pattern of A. arabiensis based on microsatellite and
other molecular markers. Donnelly and Townson (2000)
reported an FST value of 0.035 for nine A. arabiensis
populations spanning 4500 km from north Sudan to south
Mozambique and found that the population genetic structure
is consistent with the isolation-by-distance model. Kamau
et al. (1999) showed no significant genetic structure of A.
arabiensis populations between western and coastal Kenya
that are separated by 700 km distance and the Great Rift
Valley. Besansky et al. (1997) reported that the A. arabiensis
populations from east Africa (Kenya), south Africa, and west
Africa (Senegal) had an FST of 0.038 by allozyme and 0.044
by mtDNA sequence variation. The largest FST (0.080–
0.215) reported in the literature was a comparison of A.
arabiensis populations from Senegal and the east outer islands
(Madagascar, and Reunion and Mauritius Islands) (Simard et
al. 1999). The authors proposed that historical events of drift
rather than mutation were probably the forces generating
genetic divergence between these populations, because
homogenization of the gene pool by migration is restricted
by the ocean.
Two hypotheses can explain the low genetic differentiation of A. arabiensis populations across the African
continent. The first is that there may not be significant gene
flow barriers for A. arabiensis because it is widespread in
Africa. The second is that A. arabiensis populations in Africa
resulted from recent expansions and the large effective
population size (Donnelly et al. 1999; Lehmann et al. 1996b;
Taylor et al. 1993). Even if rugged mountains and deep
chasms in our study area are indeed a significant gene flow
barrier for the mosquitoes, the effect on the population
genetic structure may not be apparent because genetic
heterogeneity may be caused by historical factors, such as
population expansion, rather than by current migration.
Thus gene flow inferred from the FST statistics should be
interpreted with caution.
Bowcock AM, Ruiz-Linares A, Tomfohrde J, Minch E, Kidd JR, and
Cavalli-Sforza LL, 1994. High resolution of human evolutionary trees with
polymorphic microsatellites. Nature 368:455–457.
Chandre F, Darrier F, Manga L, Akogbeto M, Faye O, Mouchet J, and
Guillet P, 1999. Status of pyrethroid resistance in Anopheles gambiae sensu
lato. Bull WHO 77:230–234.
Coluzzi M, Petrarca V, and Di Deco MA, 1985. Chromosomal inversion
intergradation and incipient speciation in Anopheles gambiae. Biol Zool 52:45–
63.
Coluzzi M, Sabatini A, Petrarca V, and Di Deco MA, 1979. Chromosomal
differentiation and adaptation to human environments in the Anopheles
gambiae complex. Trans R Soc Trop Med Hyg 73:483–497.
Crampton JM, Warren A, and Lycett GJ, 1994. Genetic manipulation of
insect vectors as a strategy for the control of vector-borne disease. Ann
Trop Med Parasitol 88:3–12.
Curtis CF, Miller JE, Hodjati MH, Kolaczinski JH, and Kasumba I, 1998.
Can anything be done to maintain the effectiveness of pyrethroidimpregnated bednets against malaria control? Philos Trans R Soc Lond
Biol 353:1769–1775.
Donnelly MJ, Cuamba N, Charlwood JD, Collins FH, and Townson H,
1999. Population structure in the malaria vector, Anopheles arabiensis Patton,
in East Africa. Heredity 83:408–417.
Donnelly MJ, Licht MC, and Lehmann T, 2001. Evidence for recent
population expansion in the evolutionary history of the malaria vectors
Anopheles arabiensis and Anopheles gambiae. Mol Biol Evol 18:1353–1364.
Donnelly MJ and Townson H, 2000. Evidence for extensive genetic
differentiation among populations of the malaria vector, Anopheles arabiensis,
in East Africa. Mol Biol 9:357–367.
Estoup A, Garnery L, Solignac M, and Cornuet J, 1995. Microsatellite
variation in honeybee (Apis mellifera L.) populations: hierarchical genetic
structure and test of the infinite allele and stepwise mutation models.
Genetics 140:679–695.
Gaggiotti OE, Lange O, Rassmann K, and Gliddon C, 1999. A comparison
of two indirect methods for estimating average levels of gene flow using
microsatellite data. Mol Ecol 8:1513–1520.
Goldstein DB and Schlotterer C, 1999. Microsatellites: evolution and
applications. New York: Oxford University Press.
Goodman CA, Mnzava AE, Dlamini SS, Sharp BL, Mthembu DJ, and
Gumede JK, 2001. Comparison of the cost and cost-effectiveness of
insecticide-treated bednets and residual house spraying in KwaZulu, Natal,
South Africa. Trop Med Int Health 6:280–295.
Acknowledgments
Goudet J, 1995. FSTAT, version 1.2: a computer program to calculate Fstatistics. J Hered 86:485.
We thank the personnel of the Malaria Control Bureaus at Gambella, Inda
Sellase, Debre Zeit, and Arba Minch, Ethiopia, and the Malaria Control
Program of Eritrea for assisting mosquito collection. Two anonymous
reviewers provided constructive suggestions. This work was supported by
NIH grants U19 AI45511, R01 AI50243, and D43 TW01505.
Kamau L, Mukabana WR, Hawley WA, Lehmann T, Irungu LW,
Orago AAS, and Collins HH, 1999. Analysis of genetic variability in
Anopheles arabiensis and Anopheles gambiae using microsatellite loci. Insect Mol
Biol 8:287–297.
References
Lanzaro GC, Toure TY, Carnahan J, Zheng L, Dolo G, Traore S, Petrarca
V, Vernick KD, and Taylor CE, 1998. Complexities in the genetic structure
of Anopheles gambiae populations in west Africa as revealed by microsatellite
DNA analysis. Proc Natl Acad Sci USA 95:14260–14265.
Abose T, Ameneshewa B, Teklehaimanot A, Tulu AN, and Petrarca V,
1998. Cytogenetic studies on the Anopheles gambiae complex species in
Ethiopia. Ethiop J Health Dev 12:81–83.
Lanzaro GC, Zheng L, Toure TY, Traore SF, Kafatos FC, and Vernick KD,
1995. Microsatellite DNA and isozyme variability in a west African
population of Anopheles gambiae. Insect Mol Biol 4:105–112.
Atkinson PW and Michel K, 2002. What’s buzzing? Mosquito genomics
and transgenic mosquitoes. Genesis 32:42–48.
Lehmann T, Besansky NJ, Hawley WA, Fahey TG, Kamau L, and Collins
FH, 1997. Microgeographic structure of Anopheles gambiae in western Kenya
based on mtDNA and microsatellite loci. Mol Ecol 6:243–253.
Besansky NJ, Lehmann T, Fahey GT, Fontenille D, Braack EO, Hawley
WA, and Collins FH, 1997. Patterns of mitochondrial variation within and
between African malaria vectors, Anopheles gambiae and An. arabiensis,
suggesting extensive gene flow. Genetics 147:1817–1828.
462
Lehmann T, Hawley WA, and Collins FH, 1996a. An evaluation of
evolutionary constraints on microsatellite loci using null alleles. Genetics
144:1155–1163.
Nyanjom et al. Population Genetic Structure of Anopheles arabiensis
Lehmann T, Hawley WA, Grebert H, Danga M, Atieli F, and Collins FH,
1999. The Rift Valley complex as a barrier to gene flow for Anopheles gambiae
in Kenya. J Hered 90:613–621.
Scott JA, Brogdon WG, and 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.
Lehmann T, Hawley WA, Kamau L, Fontenille D, Simard F, and Collins
FH, 1996b. Genetic differentiation of Anopheles gambiae populations from
east and west Africa: comparison of microsatellite and allozyme loci.
Heredity 77:192–208.
Sharakhov IV, Braginets O, Mbogo CN, and Yan G, 2001. Isolation and
characterization of trinucleotide microsatellites in African malaria mosquito
Anopheles funestus. Mol Ecol Notes 1:289–292.
Lehmann T, Licht M, Elissa N, Maega BTA, Chimumbwa JM, Watsenga
FT, Wondji CS, Simard F, Hawley WA, 2003. Population structure of
Anopheles gambiae in Africa. J Hered 94:133–147.
Lindsay SW, Parson L, and Thomas CJ, 1998. Mapping the ranges and
relative abundance of the two principal African malaria vectors, Anopheles
gambiae sensu stricto and An. arabiensis, using climate data. Proc R Soc Lond
Ser B Biol Sci 265:847–854.
Shililu JI, Maier WA, Seitz HM, Kubasu SS, and Orago AS, 1998. Use of
polymerase chain reaction to identify Anopheles gambiae (Diptera: Culicidae)
sibling species. J Appl Entomol 122:461–464.
Simard F, Fontenille D, Lehmann T, Girod R, Brutus L, Gopaul R,
Dournon C, and Collins FH, 1999. High amounts of genetic differentiation
between populations of the malaria vector Anopheles arabiensis from west
Africa and eastern outer islands. Am J Trop Med Hyg 60:1000–1009.
Slatkin L, 1995. A measure of population subdivision based on microsatellite allele frequencies. Genetics 139:457–462.
Lulu M, Nigatu W, Gezahegn T, and Tilahun D, 1991. Inversion
polymorphisms in Anopheles arabiensis Patton in five selected localities from
east, south and southwest Ethiopia. Insect Sci Applic 12:375–378.
Taylor CE, Toure YT, Coluzzi M, and Petrarca V, 1993. Effective
population size and persistence of Anopheles arabiensis during the dry season
in west Africa. Med Vet Entomol 7:351–357.
Mekuria Y, Petrarca V, and Tesfamariam T, 1982. Cytogenetic studies on
the malaria mosquito Anopheles arabiensis Patton in the Awash Valley,
Ethiopia. Parassitologia 24:237–243.
Toure YT, Petrarca V, Traore SF, Coulibaly A, Maiga HM, Sankare O, Sow
M, Di Deco MA, and Coluzzi M, 1998. Distribution and inversion
polymorphism of chromosomally recognized taxa of the Anopheles gambiae
complex in Mali, West Africa. Parassitologia 40:477–511.
Minakawa N, Sonye G, Mogi M, Githeko A, and Yan G, 2002. The effects
of climatic factors on the distribution and abundance of malaria vectors in
Kenya. J Med Entomol 39:833–841.
Nigatu W, Petros B, Lulu M, Adugna N, Wirtz R, and Tilahun D, 1994.
Some aspects of malaria prevalence, vector infectivity and DDT resistance
studies in Gambella region, southwestern Ethiopia. Ethiop J Health Dev
8:1–10.
Oetting WS, Armstrong CM, Ronan SM, Young TL, Sellers TA, and King
RA, 1998. Multiplexed short tandem repeat polymorphisms of the Weber
8A set of markers using tailed primers and infrared fluorescence detection.
Electrophoresis 19:3079–3083.
Weir BS, 1990. Genetic data analysis. Sunderland, MA: Sinauer Associates.
WHO, 1975. Manual on practical entomology in malaria. Part II. Methods
and techniques. No. 13. Geneva, Switzerland: WHO.
WHO, 1993. Global malaria control. Bull WHO 71:281–284.
WHO, 1998. Re-orientation and definition of the role of malaria vector
control in Ethiopia. WHO/MAL/98.1085. Geneva, Switzerland: WHO.
Yan G, Christensen BM, and Severson DW, 1997. Comparisons of genetic
variability and genome structure among mosquito strains selected for
refractoriness to a malaria parasite. J Hered 88:187–194.
Raymond M, and Rousset F, 1995. An exact test for population
differentiation. Evolution 49:1280–1283.
Zheng L, Benedict MQ, Cornel AJ, Collins FH, and Kafatos FC, 1996. An
integrated genetic map of the African human malaria vector mosquito,
Anopheles gambiae. Genetics 14:941–952.
Rousset F, 1997. Genetic differentiation and estimation of gene flow from
F-statistics under isolation by distance. Genetics 145:1219–1228.
Received February 6, 2003
Accepted August 30, 2003
Scanalytics, Inc., 1998. Gene profiler user’s manual, version 10.98. Fairfax,
VA: Scanalytics, Inc.
Corresponding Editor: R. C. Woodruff
463