Research Article

Influence of Urban Landscapes on Population Dynamics in a Short-Distance Migrant Mosquito: Evidence for the Dengue Vector Aedes aegypti

  • Ryan R. Hemme,

    Affiliation: Department of Biological Sciences, Eck Institute for Global Health, University of Notre Dame, Notre Dame, Indiana, United States of America

  • Clayton L. Thomas,

    Affiliation: Department of Biological Sciences, Eck Institute for Global Health, University of Notre Dame, Notre Dame, Indiana, United States of America

  • Dave D. Chadee,

    Affiliation: Department of Life Sciences, University of the West Indies, St. Augustine, Trinidad and Tobago, West Indies

  • David W. Severson mail

    Affiliation: Department of Biological Sciences, Eck Institute for Global Health, University of Notre Dame, Notre Dame, Indiana, United States of America

  • Published: March 16, 2010
  • DOI: 10.1371/journal.pntd.0000634



Dengue viruses are endemic across most tropical and subtropical regions. Because no proven vaccines are available, dengue prevention is primarily accomplished through controlling the mosquito vector Aedes aegypti. While dispersal distance is generally believed to be ~100 m, patterns of dispersion may vary in urban areas due to landscape features acting as barriers or corridors to dispersal. Anthropogenic features ultimately affect the flow of genes affecting vector competence and insecticide resistance. Therefore, a thorough understanding of what parameters impact dispersal is essential for efficient implementation of any mosquito population suppression program. Population replacement and genetic control strategies currently under consideration are also dependent upon a thorough understanding of mosquito dispersal in urban settings.

Methodology and Principal Findings

We examined the effect of a major highway on dispersal patterns over a 2 year period. A. aegypti larvae were collected on the east and west sides of Uriah Butler Highway (UBH) to examine any effect UBH may have on the observed population structure in the Charlieville neighborhood in Trinidad, West Indies. A panel of nine microsatellites, two SNPs and a 710 bp sequence of mtDNA cytochrome oxidase subunit 1 (CO1) were used for the molecular analyses of the samples. Three CO1 haplotypes were identified, one of which was only found on the east side of the road in 2006 and 2007. AMOVA using mtCO1 and nuclear markers revealed significant differentiation between the east- and west-side collections.

Conclusion and Significance

Our results indicate that anthropogenic barriers to A. aegypti dispersal exist in urban environments and should be considered when implementing control programs during dengue outbreaks and population suppression or replacement programs.

Author Summary

Worldwide, 2.5 billion people are at risk for dengue infection, with no vaccine or treatment available. Thus dengue prevention is largely focused on controlling its mosquito vector, Aedes aegypti. Traditional mosquito control approaches typically include insecticide applications and breeding site source reduction. Presently, novel dengue control measures including the sterile insect technique and population replacement with dengue-incompetent transgenic mosquitoes are also being considered. Success of all population control programs is in part dependent upon understanding mosquito population ecology, including how anthropogenic effects on the urban landscape influence dispersal and expansion. We conducted a two year population genetic study examining how a major metropolitan highway impacts mosquito dispersal in Trinidad, West Indies. As evidenced by significant differentiation using both nuclear and mitochondrial DNA sequences, the highway acted as a significant barrier to dispersal. Our results suggest that anthropogenic landscape features can be used effectively to enhance population suppression/replacement measures by defining mosquito control zones along recognized landscape barriers that limit population dispersal.


Anthropogenic assisted invasions by non-indigenous insect vectors of human disease have and will continue to have profound effects on global health [1]. In addition, anthropogenic land use changes can represent primary drivers of infectious disease epidemics and significantly alter disease transmission dynamics [2]. The dengue and yellow fever vector mosquito, Aedes aegypti, is a remarkably successful invasive species. A highly anthropophilic form likely emerged in North Africa within the past 2–4 thousand years and has subsequently been transported via human efforts to most subtropical and tropical regions worldwide [3],[4].

Approximately two-fifths of the world's population is at risk for dengue infection and an estimated 500,000 people are affected by dengue hemorrhagic fever (DHF) annually, with fatality rates exceeding 20% when proper treatment is unavailable [5]. Dengue is widely distributed in the tropics, occurring in Central and South America, South and Southeast Asia, Africa, the Caribbean, and Pacific regions [6]. During the past 45 years the incidence of dengue infection has steadily increased throughout the globe as greater numbers of people permanently migrate to cities with continued growth and urbanization [7].

Aedes aegypti population dynamics in urban areas is subject to daily as well as seasonal meteorological variability [8]. The interaction between temperature, relative humidity and rainfall impact adult survival and availability of oviposition sites. The goal of A. aegypti control programs is to reduce the population density of adult mosquitoes below a critical threshold where epidemic dengue transmission is unlikely to occur [9]. Vector population suppression programs most often involve the elimination or insecticide treatment of larval habitats that are typically man-made containers located within or around houses. During epidemic outbreaks, ultra low volume (ULV) spraying of insecticides is often used as an emergency control measure to reduce the adult mosquito population [10]. Critical to the long-term success of any A. aegypti population suppression method is the influence of dispersion patterns of adult mosquitoes. A greater understanding of factors limiting adult dispersal would allow health agencies to be more efficient in allocating resources to vector control programs. Moreover, considerable interest exists in developing novel dengue control strategies through the development of genetically modified A. aegypti incapable of transmitting dengue virus (DENV) and their subsequent introduction into the field as part of a population replacement program [11],[12]. A thorough understanding of dispersal behavior in urban environments is essential to successful implementation of any control strategy.

Population structure of A. aegypti is complex, varies by region and scale, and can be influenced by environment and geography [13][20]. Urban estimates of genetic differentiation have varied in part due to environmental conditions and dispersal patterns [21][24]. Typically, adult A. aegypti mosquitoes travel relatively short distances of up to ~100 m, although longer dispersal estimates of ~800 m have been observed [25][29]. In Queensland, Australia, a mark-release-recapture study reported that A. aegypti would readily cross smaller, quieter roads, but significantly fewer crossed a major highway near the release point, and concluded that busy roads may have impeded dispersal [30]. Similar observations were made with bumblebees (Bombus impatiens and B. affinis), where Bhattacharya et al. [31] reported high fidelity between the bumblebees and their foraging sites and that they would rarely cross nearby roads or railways. These observations indicate the possibility that habitat fragmentation due to roads or other anthropogenic environmental manipulations may act as significant barriers to migration of A. aegypti and other insects.

In urban environments anthropogenic landscape features can result in habitat fragmentation and thereby influence dispersal patterns of mosquitoes. Ultimately, these features can affect the flow of genes conditioning vector competence and insecticide resistance. In the current study, we report evidence of limited A. aegypti movement across an expansive 4 lane split highway in an urban environment in Trinidad, West Indies, as evidenced by mitochondrial and nuclear molecular analyses.

Materials and Methods

Sample collections

A. aegypti larvae were collected within urban breeding sites along a 900 m length of Uriah Butler Highway (UBH) on the east and west sides in the Charlieville neighborhood of Chauguanas. UBH is the major north-south highway in western Trinidad, extending from San Fernando in the south to east of Port of Spain (Figs. 1 & 2). The distance between buildings on the east and west sides of UBH ranged from ~80 m to ~130 m and the two sides were connected by a walking overpass and on-off ramps on both ends of the sampling area. Charlieville is a diverse urban neighborhood with mixed commercial, industrial, and residential buildings clustered closely together.


Figure 1. Uriah Butler Highway.

Samples were collected along an ~900 m stretch of the highway. The distance across the highway is ~65 m.


Figure 2. Collection sites were located along Uriah Butler Highway in the Charlieville neighborhood of Chagaunas, Trinidad, West Indies.

This is the major north-south highway from Port of Spain to San Fernando.


Larvae were collected during October of 2006 and 2007 with assistance from field technicians within the Insect Vector Control Division at the Ministry of Health. In 2006, larvae were collected from 18 larval habitat sites and in 2007 collections were taken from nine sites. Our samples were collected from diverse, but typical, larval habitats including water storage drums, folded sheets of commercial plastic, small buckets, and neglected and disused auto parts. Larvae were preserved in ethanol, and carried to the University of Notre Dame for genotyping.

DNA extraction

DNA was extracted from the mosquito samples using a standard phenol-chloroform method [32]. In 2006, 147 larvae were genotyped; 58 from the western side of UBH and 89 on the eastern side while in 2007, DNA was extracted from a total of 83 larvae; 33 from the west and 50 from the east.

Mitochondrial gene amplification and haplotype determination

A 710-basepair region of the cytochrome c oxidase subunit I gene (CO1) was amplified using CO1-specific universal primers [33]. Five µl of template DNA was amplified by polymerase chain reaction (PCR) in a 25.0 µl reaction containing 1X Taq buffer (10 mM KCl, 2 mM Tris, pH 9.0, 0.02% TritonX), 1.5 mM MgCl2, 0.4 mM each dATP, dCTP, dGTP, dTTP, 5 pmoles of each primer, and 1 U Taq DNA polymerase. The thermocycle conditions were 94°C for 1 min, 4 cycles at 94°C for 1 min, 45°C for 1.5 min, followed by 34 cycles at 94°C for 1 min, 50°C for 1.5 min, and 72°C for 1 min with a final extension period at 72°C for 5 minutes.

Haplotypes were identified by examining banding patterns using Single Strand Conformation Polymorphism (SSCP) as per [34]. Briefly, 5.0 µl of PCR product was mixed with 3.0 µl of Denaturing Loading Mix (DLM), that consisted of 0.1 ml of 1N NaOH, 9.5 ml formamide, 0.005 g of bromophenol blue, 0.005 g of xylene cyanol, and brought up to 10 ml with ddH20. The mixture was denatured at 95°C for 5 min and snap cooled on ice. Approximately 7.0 µl of PCR product-DLM were used for electrophoresis on 42×33 cm 5% polyacrylamide gels for 3–4 hours at 30 milliamps. Gels were stained using silver nitrate solution with protocol adapted from Promega's GenePrint® STR Systems (Promega U.S., Madison, WI) to visualize DNA banding patterns. Confirmation of haplotypes was accomplished by sequencing 5–10 individuals of each haplotype from both strands using the same CO1 primers.

Nuclear marker amplification and polymorphism detection

A total of 11 markers including nine microsatellite loci [35],[36] and two Single Nucleotide Polymorphism (SNP) loci [37] were used for genotyping. PCR amplification was performed with genomic DNA isolated from individual mosquitoes in 25 µl volumes as described above. PCR reactions for microsatellite loci AC2, AG2, AG7, A10, B19, CT2, H08, M201, M313 were performed under the following conditions: 94°C for five minutes, followed by 30 cycles of 94°C for 1 min, 60°C anneal for 1 min, 72°C extension for 2 min, and a final 72°C extension for 10 min. PCR conditions for SNP loci LF178 and RT6 were performed at: 94°C for 5 minutes, followed by 39 cycles at 94°C for 20 sec, 55°C for 20 sec, and 72°C for 30 sec, and a final 72°C extension for 10 min. SNP products LF178 and RT6 were digested with RsaI and MnlI respectively and size fractionated in 3% agarose gels and visualized with ethidium bromide under UV light.

Polymorphisms in microsatellite loci were resolved and analyzed using a Beckman-Coulter CEQ8000 and Beckman-Coulter CEQ8000 software. Briefly, microsatellites were amplified using dye-labeled primers (Sigma Proligo, Sigma-Aldrich Inc., St. Louis, MO) and pooled into groups of 3 loci. Pools consisted of 0.4 µl of 400 bp standard, and 30.0 µl of standard loading solution (Beckman-Coulter Inc., Fullerton, CA) with 1.0 µl of diluted amplified product added to each well.

Data analysis

Conformance with Hardy-Weinberg equilibrium (HWE), gametic disequilibrium between pairs of loci in each population, and inbreeding coefficients (FIS) were computed on FSTAT version [38]. Presence of null alleles was examined using Micro-Checker [39]. Mitochondrial sequences were aligned using SEQMAN from the Lasergene package (DNASTAR Inc., Madison, WI) and analyzed using DnaSP [40]. Population structure was examined using locus by locus Analysis of Molecular Variance (AMOVA) for nuclear markers and mtDNA haplotypes were analyzed using standard AMOVA on Arlequin version 3.1 [41].


mtDNA CO1 polymorphism

The mitochondrial CO1 gene showed sequence polymorphisms among the samples. A total of 3 haplotypes were identified within the 177 individuals in our test samples (Fig. 3). Mitochondrial CO1 haplotypes-1 and 2 were the most common across populations and years, accounting for ~42% and ~52% of total known haplotypes. Haplotype-3 comprised ~5% of the total, but was unique to the eastern side of the road in 2006 and 2007 samples and was never detected on the western side of UBH. Haplotype-2 was the only haplotype detected from the 2007-west population (Fig. 3). Haplotype frequencies were compared spatially (east compared to west) and temporally (2006 compared to 2007). There was significant differentiation between mosquitoes collected on the east side of the road and those collected on the west side of the road in 2006 and 2007 (Table 1). Temporally, the 2006-west population differed from the 2007-west population, however on the east side of the road there was no genetic differentiation between years. Estimated FST values were moderate to large, ranging from 0.042 to 0.390. Our data suggested relatively lower FST values for spatial samples than temporal samples. The east and west side samples show FST 0.172 (year 2006) and 0.390 (year 2007) and the 2006 and 2007 samples show FST 0.249 (east) and 0.49 (west).


Figure 3. Frequency of unique COI mtDNA haplotypes recovered from the Charlieville neighborhood on the east and west side of Uriah Butler Highway in 2006 and 2007.


Table 1. Analysis of molecular variance using mtDNA CO1 haplotypes.


Nuclear marker polymorphism

Five out of 44 tests (11%) were significant for deviation from Hardy-Weinberg equilibrium after Bonferroni correction (Table 2). Deviations in expected heterozygosity were due to heterozygosity deficits in locus B19 (2006-east and 2006-west collections), and locus M313 (2006-east and 2007-east collections). Deviation at locus M201 in the 2007-west population was a result of heterozygote excess. Null alleles were identified at loci LF178 and RT6 in the 2006-east and 2007-east populations, respectively and in locus A10 in both 2007 populations, but the loci were in HW equilibrium. Evidence for null alleles was present in locus M313 in all but the 2007-west population, and locus B19 had high levels of null alleles in all populations, and both B19 and M313 deviated from HW equilibrium. Gametic disequilibrium analysis revealed significant disequilibrium between loci H08 and A10 and loci H08 and AG7. The H08 and AG7 loci are physically linked on chromosome II, while locus A10 is located on chromosome III. Due to the presence of null alleles at loci B19, M313, and A10, and gametic disequilibrium with locus H08, these markers were removed prior to AMOVA.


Table 2. Summary of variation at 9 microsatellite and 2 SNP loci by collection.


Four of the five microsatellites had private alleles that were specific to either the 2006-east, 2006-west, 2007-east or 2007-west collections. Neither SNP locus had private alleles. Markers AG2 and CT2 had alleles that were present in a population at a frequency between 5–10% (Fig. 4 & 5). In marker AG2, 9 of the 16 alleles were private to at least one collection, however most had a frequency <3%. Allele AG2-O was present only in the 2006-west collection at 6.5%. Allele AG2-D was present in 2006-east, 2006-west, and in 2007 was only present in the 2007-east collection at a frequency of 10.9% (Fig. 4). Similarly allele AG2-N was present in both 2006 collections (east and west), but unlike allele AG2-D was absent in the 2007-east collection and present in the 2007-west collection at 10.3% (Fig. 4). Six alleles were found in CT2, 3 were private to at least one collection (Fig. 5). Allele CT2-D was present in the 2006-east (5.4%) collection, 2007-east (4.2%), and 2007-west (3.9%), but was not present in the 2006-west collection (Fig. 4). In the remaining 2 microsatellites (AC2 and AG7) alleles that were private in at least one collection existed, but the frequency ranged from ~2% to ~4% (Fig. 5 & 6). Overall, the amount of variation contained between populations was small, ranging from 0.2% to 1.4% of the total variation. AMOVA found small but significant FST estimates between collections on the east and west side of UBH in 2006 and 2007, ranging from 0.011 to 0.021 (Table 3).


Figure 4. Distribution of alleles for microsatellite locus AG2 in 2006 and 2007.

*Denotes private alleles with a frequency of >5%.


Figure 5. Distribution of alleles for microsatellite locus CT2 in in 2006 and 2007.

*Denotes private alleles with a frequency of >5%.


Figure 6. Distribution of alleles for microsatellite locus AC2 in 2006 and 2007.


Table 3. Analysis of molecular variance using nuclear microsatellite and SNP markers.



Aedes aegypti is highly adapted to a peridomestic environmental niche which has enabled it to spread throughout most large tropical cities in the world [42]. After confirmation that A. aegypti was the primary vector of yellow fever virus and dengue virus, disease prevention programs focused on the control of the mosquito vectors [43]. The pinnacle of dengue control began with the decision by PAHO to eradicate A. aegypti from the western hemisphere using a top down control structure and effective use of insecticides [42],[44],[45]. Incidences of A. aegypti transmitted diseases were greatly reduced along with the distribution of A. aegypti in the late 1950s to mid-1970s. However, in the late-1970s control programs were disbanded in part to financial considerations and the realization that unless a global campaign to eradicate A. aegypti was undertaken any attempts to eliminate the mosquito from the Americas would be unsuccessful due to the increased frequency and speed of air travel and other transportation options capable of transporting A. aegypti eggs and adults [46],[47]. Further complicating eradication efforts was the emergence of resistance to insecticides in the mid-1950s [46]. As a result in areas where A. aegypti was once eliminated, reinfestation and outbreaks of dengue eventually followed where vigilant vector surveillance and control was not implemented [43],[47],[48]. In addition, the availability of a highly effective vaccine for yellow fever likely contributed to a decline in active mosquito surveillance programs in areas certified as A. aegypti free.

Our micro-geographic analysis of genetic variability in A. aegypti from Trinidad was studied using microsatellite and SNP nuclear markers, and mitochondrial CO1 sequences. The distribution of the 3 CO1 haplotypes is strongly indicative of UBH acting as a barrier to dispersal. We were unable to detect haplotype-3 on the west side of the road in either 2006 or 2007. This is interesting because the distance between collections on the east and west side of UBH ranged from ~80 m to ~130 m, which given dispersal estimates ranging from 100 m to 800 m, should not have limited adult dispersal potential and mosquitoes with haplotype-3 would likely be expected to have colonized the west side of UBH unless they were unable to successfully transect the highway. FST estimates from AMOVA also revealed significant differentiation between populations collected on the east and west side of UBH in 2006 and 2007. Results from nuclear marker analysis therefore showed the same pattern of differentiation as the mitochondrial sequence data; however the magnitude of the FST and amount of variation was much lower (Table 3). Two explanations, neither mutually exclusive may explain the discrepancy in magnitude between the class of markers. The first is due to the preservation of diversity as a result of A. aegypti utilizing heterogeneous larval habitats in Trinidad. A large number of alternative habitats that are suitable for mosquito production may have gone undetected by surveyors in the Charlieville neighborhood as indicated by the large numbers of alleles detected (Fig. 4, 5, 6 & 7). Of the 25 containers from which we collected larvae, 10 were from containers other than water storage drums. In Trinidad water storage drums are the main source of A. aegypti production, however surveys have shown that A. aegypti mosquitoes will utilize a wide range of permanent and semi-permanent containers and the types of containers used can depend upon the degree of urbanization [49],[50]. The second explanation is the existence of homoplasy in the microsatellites which has been proposed as an explanation for observed differences in differentiation between SNP and mtDNA marker in A. aegypti populations in Venezuela [13]. If present, homoplasy would underestimate the amount of differentiation between the collections [51].


Figure 7. Distribution of alleles for microsatellite locus AG7 in 2006 and 2007.


We also observed temporal differences in haplotype frequency between 2006 and 2007, although no significant population structure was found between 2006 and 2007 on the east side of UBH (Table 1). In all four groups of mosquitoes there was a decrease in the frequency of CO1 haplotype-1 and increase in haplotype-2 (Fig. 3). One possible explanation for this observation is that the resulting change in haplotype frequencies is a consequence of vector control efforts conducted periodically by the Ministry of Health in Trinidad. Changes in haplotype frequency could also be due to normal temporal fluctuations in the mosquito population. If the former is the primary cause affecting haplotype frequency one would expect to see reduced heterozygosity in the nuclear markers, which was not observed. Although there was not a physical barrier separating the 2006 and 2007 collections the significant differentiation observed between the 2006 west and 2007 west population was not unexpected, as previous examinations of temporal variation in A. aegypti populations have also reported changes in genetic differentiation over seasons, which were most likely due to changes in mosquito density, availability of oviposition sites, and the type of environment [21],[23],[52]. In Phnom Penh, Cambodia, genetic differentiation in A. aegypti populations was influenced by seasonality and environment type (urban vs. suburban vs. rural), with significant differentiation occurring within the city [23]. The authors suggested that ideal urban conditions, including an abundant supply of hosts and oviposition sites, limited the need for dispersal by adult mosquitoes. In suburban areas differentiation was in part dependent upon the physical environment with variations in human density, availability of running water, and rural versus residential developments impacting the population structure.

Dispersal range is an important aspect of dengue transmission and much research has been conducted attempting to determine how far A. aegypti adults travel, however large variations in daily and lifetime dispersal rates have been reported. Larger estimates of dispersal have reported mosquitoes traveling >800 m [29],[53]. Many studies using mark-release-recapture methods have reported a shorter flight range of A. aegypti [25], [54][56]. Examining mean distance traveled (MDT) and the flight range within which 50% (FR50) and 90% (FR90) of mosquitoes travel, as opposed to maximum distance traveled may be a more epidemiologically important parameter [29]. In a Kenyan village, McDonald and others [25] recaptured a majority of mosquitoes within the house they were released over 12 days. Marked mosquitoes released in a tire dump in New Delhi, India dispersal ranged between 50–200 m, but most were recaptured within 50 m of the release point [57]. Similarly, Muir and Kay [54] reported females having a MDT of 56 m and FR90 of 108 m. Released mosquitoes tended to cluster around houses with some dispersal towards adjacent houses and mosquitoes released on the perimeter of villages moved towards the center of the village [26], [29], [58][60]. The relatively large numbers and duration of DENV infected females captured in houses with confirmed dengue cases in Merida, Mexico may further indicate high fidelity between A. aegypti mosquitoes and place of pupal emergence [61].

Results from both classes of makers show strong evidence of limited gene flow across UBH, effectively fragmenting the populations on the east and west side of the highway. Mosquito dispersal patterns are nonrandom and influenced by environmental factors as reported by Sheppard et al. [62] and Hausermann et al. [63] in A. aegypti mosquitoes using mark-release-recapture methods. Furthermore, Chadee [64] indicated that prevailing weather patterns may potentially influence dispersion. Range of dispersal is dependent upon a mosquito's ability to remain in flight and the availability and abundance of shelter, food sources, hosts for blood meals and suitable oviposition sites [62]. Suitable host availability may reduce dispersal as reported by Suwonkerd et al. [65] where fewer A. aegypti mosquitoes exited a hut when a human host was present than with controls consisting of a dog or no host. Edman et al. [66] reported that when an abundance of suitable oviposition sites were available dispersion of female A. aegypti mosquitoes was reduced. Although the distance across the highway is well within dispersal estimates for A. aegypti, lack of cover and shade may have made UBH a harsh environment for mosquitoes to transect. This is supported by Tun-Lin et al. [67] who reported shade as a significant factor impacting the presence of A. aegypti in premise surveys and Russell et al. [30] reported that released A. aegypti dispersal patterns were nonrandom with more mosquitoes being recaptured along a corridor with heavy shading from trees and vegetation. Furthermore, oviposition sites were most likely minimal, even along peripheral ditches and nonexistent blood meal hosts may have dissuaded migration across UBH and prevented a stepping stone model of colonization from occurring over UBH.

Ecological features including accessible water and availability of oviposition sites, vegetation patterns, humidity, and housing density contribute to determining the distribution of A. aegypti mosquitoes. The effects of topographic features of urban environments are not fully understood, however Reiter et al. [27] noted that buildings were not an impediment to A. aegypti flight. Our results indicate that urban landscape features do contain barriers to dispersal, and thereby affect the population structure of mosquitoes. This information could be used by vector control agencies to more efficiently target mosquito populations for suppression. Control programs can divide an urban area into zones of control along landscape features that are large enough to impede dispersal. This technique allows for the possibility of local elimination of A. aegypti moquitoes, barring or at least minimizing re-infestation due to the active transportation of the mosquito. Furthermore, during dengue outbreaks control agencies can more accurately target areas of higher risk along these same control zones.

Difficulties in vaccine development [68] and the sequencing of the entire A. aegypti genome [69] have shifted some research efforts to preventing illness by developing applications that make use of transgenic mosquitoes incapable of transmitting the virus. Central to the successful use of transgenic mosquitoes to replace competent vector populations or to effect population suppression/elimination is a thorough understanding of A. aegypti bionomics, answering the basic questions of how many mosquitoes need to be released, where is the best place for them to be released, and when should they be released [70]. Results from early efforts using the sterile insect technique (SIT) to eliminate mosquito populations could have been improved or were negatively influenced by incomplete knowledge of adult mosquito dispersal behavior in A. aegypti and Culex fatigans [71],[72]. Anthropogenic landscape features may therefore have profound effects on the implementation of traditional as well as proposed novel genetic mosquito control programs. Yakob et al. [73] explored the dynamics of population suppression dynamics with SIT and insects engineered to carry a dominant lethal gene (RIDL). Mathematical models indicated that dispersion parameters for A. aegypti were fundamental in the success of replacement efforts and that enhanced connectivity treatment and peripheral populations could result in increased densities of wild-type mosquitoes as a consequence of SIT programs. Natural and anthropogenic barriers may actively influence, either positively or negatively depending on the strategy, the success of population replacement or population reduction by limiting the effective range of gene flow. Understanding the role of landscape features on population dispersal is likely critical to achieving success with any A. aegypti control strategy.


We thank Sarah Epstein, Stephen “Billy” Deonarine, Lester James, Larry Smith, and Steve Gittens for their help with collecting mosquitoes in Trinidad. In addition we thank Becky deBruyn and Diane D. Lovin for their help and support in the laboratory.

Author Contributions

Conceived and designed the experiments: RRH DDC DWS. Performed the experiments: RRH CLT DDC DWS. Analyzed the data: RRH. Wrote the paper: RRH DDC DWS.


  1. 1. Lounibos LP (2002) Invasions by insect vectors of human disease. Annu Rev Entomol 47: 233–266.
  2. 2. Patz JA, Daszak P, Tabor GM, Aguirre AA, Pearl M, et al. (2004) Unhealthy landscapes: policy recommendations on land use change and infectious disease emergence. Environ Health Persp 112: 1092–1098.
  3. 3. Tabachnick WJ (1991) Evolutionary genetics and the yellow fever mosquito. Am Entomol 37: 14–24.
  4. 4. Centers for Disease Control (2008) Dengue Fever. Available from​ndex.htm.
  5. 5. World Health Organization (2009) Dengue and dengue haemorrhagic fever. Available from​s/fs117/en/.
  6. 6. Thomas SJ, Strickman D, Vaughn DW (2003) Dengue epidemiology: virus epidemiology, ecology, and emergence. Adv Virus Res 61: 235–289.
  7. 7. Gubler DJ (2002) The global emergence/resurgence of arboviral diseases as public health problems. Arch Med Res 33: 330–342.
  8. 8. Halstead SB (2008) Dengue virus – mosquito interactions. Annu Rev Entomol 53: 273–291.
  9. 9. Service MW (1992) Importance of ecology in Aedes aegypti control. SE Asian J Trop Med 23: 681–690.
  10. 10. Lloyd LS (2003) Strategic report 7: Best practices for dengue prevention and control in the Americas. Environmental Health Project, Office of Health, Infectious Diseases and Nutrition, Bureau for Global Health, US Agency for International Development, Washington, DC.
  11. 11. Sperança MA, Capurro ML (2007) Perspectives in the control of infectious diseases by transgenic mosquitoes in the post-genomic era – a review. Mem I Oswaldo Cruz 102: 425–433.
  12. 12. James AA (2005) Gene drive systems in mosquitoes: rules of the road. TRENDS Parasitol 21: 64–67.
  13. 13. Urdaneta-Marquez L, Bosio C, Herrera F, Rubio-Palis Y, Salasek M, et al. (2008) Genetic relationship amony Aedes aegypti collections in Venezuela as determined by mitochondrial DNA variation and nuclear single nucleotide polymorphisms. Am J Trop Med Hyg 78: 479–491.
  14. 14. da Costa-Ribeiro MCV, Lourenço-de-Oliveira R, Failloux AB (2007) Low gene flow of Aedes aegypti between dengue-endemic and dengue-free areas in southeastern and southern Brazil. Am J Trop Med Hyg 77: 303–309.
  15. 15. Paupy C, Chantha N, Huber K, Lecoz N, Reynes JM, et al. (2004) Influence of breeding sites features on genetic differentiation of Aedes aegypti populations analyzed on a local scale in Phnom Penh municipality of Cambodia. Am J Trop Med Hyg 71: 73–81.
  16. 16. Huber K, Le Loan L, Chantha N, Failloux A-B (2004) Human transportation influences Aedes aegypti gene flow in southeast Asia. Acta Trop 90: 23–29.
  17. 17. García-Franco F, Muñoz MdL, Lozano-Fuentes , Fernandez-Salas I, Garcia-Rejon J, et al. (2002) Large genetic distances among Aedes aegypti population along the south pacific coast of Mexico. Am J Trop Med Hyg 6: 594–598.
  18. 18. Gorrochotegui-Escalante N, Munoz MdL, Fernandez-Salas , Beaty BJ, Black WC (2000) Genetic isolation by distance among Aedes aegypti populations along the northeastern coast of Mexico. Am J Trop Med Hyg 62: 200–209.
  19. 19. Gorrochotegui-Escalante N, Gomez-Machorro C, Lozano-Fuentes S, Fernandez-Salas I, Munoz MD, et al. (2002) Breeding structure of Aedes aegypti populations in Mexico varies by region. Am J Trop Med Hyg 66: 213–222.
  20. 20. Yan G, Chadee DD, Severson DW (1998) Molecular population genetics of the yellow fever mosquito: evidence for genetic hitchhiking effects associated with insecticide resistance. Genetics 148: 793–800.
  21. 21. da Costa-Ribeiro MCV, Lourenço-de-Oliveira R, Failloux AB (2006) Geographic and temporal genetic patterns of Aedes aegypti populations in Rio de Janeiro, Brazil. Trop Med Int Health 11: 1276–1285.
  22. 22. Paupy C, Chantha N, Reynes J-M, Failloux A-B (2005) Factors influencing the population structure of Aedes aegypti from the main cities in Cambodia. Heredity 95: 144–147.
  23. 23. Paupy C, Chantha N, Vazeille M, Reynes J-M, Rodhain F, et al. (2003) Variation over space and time of Aedes aegypti in Phnom Penh (Cambodia): genetic structure and oral susceptibility to a dengue virus. Genet Res 82: 171–182.
  24. 24. Huber K, Le Loan L, Hoang H, Ravel S, Rodhain F, et al. (2002) Genetic differentiation of the dengue vector, Aedes aegypti (Ho Chi Minh City, Vietnam) using microsatellite markers. Mol Ecol 11: 1629–1635.
  25. 25. McDonald PT (1977) Population characteristics of domestic Aedes aegypti (DipteraL Culicidae) in villages on the Kenya coast: II. Dispersal within and between villages. J Med Entomol 14: 49–53.
  26. 26. Trpis M, Hausermann W (1986) Dispersal and other population parameters of Aedes aegypti in an African village and their possible significance in epidemiology of vector-borne diseases. Am J Trop Med Hyg 35: 1263–1279.
  27. 27. Reiter P, Amador MA, Anderson RA, Clark GG (1995) Dispersal of Aedes aegypti in an urban area after blood feeding as demonstrated by rubidium-marked eggs. Am J Trop Med Hyg 52: 177–179.
  28. 28. Colton YM, Chadee DD, Severson DW (2003) Natural skip oviposition of the mosquito Aedes aegypti indicated by codominant genetic markers. Med Vet Entomol 17: 195–204.
  29. 29. Harrington LC, Scott TW, Lerdthusnee K, Coleman RC, Costero A, et al. (2005) Dispersal of the dengue vector Aedes aegypti within and between rural communities. Am J Trop Med Hyg 72: 209–220.
  30. 30. Russell RC, Webb CE, Williams CR, Ritchie SA (2005) Mark-release-recapture study to measure dispersal of the mosquito Aedes aegypti in Cairns, Queensland, Australia. Med Vet Entomol 19: 451–457.
  31. 31. Bhattacharya M, Primack RB, Gerwein J (2003) Are roads and railroads barriers to bumblebee movement in a temperate suburban conservation area? Biol Conserv 109: 37–45.
  32. 32. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press.
  33. 33. Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotech 3: 294–299.
  34. 34. Bosio CF, Harrington LC, Jones JW, Sithiprasasna R, Norris DE, et al. (2005) Genetic structure of Aedes aegypti populations in Thailand using mitochondrial DNA. Am J Trop Med Hyg 71: 434–442.
  35. 35. Chambers EW, Meece JK, McGowan JA, Lovin DD, Hemme RR, et al. (2007) Microsatellite isolation and linkage group identification in the yellow fever mosquito Aedes aegypti. J Hered 98: 202–210.
  36. 36. Slotman MA, Kelly NB, Harrington LC, Kitthawee S, Jones JW, et al. (2007) Polymorphic microsatellite markers for studies of Aedes aegypti (Diptera: Culicidae), the vector of dengue and yellow fever. Mol Ecol Notes 7: 168–171.
  37. 37. Severson DW, Meece JK, Lovin DD, Saha G, Morlais I (2002) Linkage map organization of expressed sequence tags and sequence tagged sites in the mosquito Aedes aegypti. Insect Mol Biol 11: 371–378.
  38. 38. Goudet J (1995) FSTAT (vers. 1.2): a computer program to calculate F-statistics. J Hered 86: 485–486.
  39. 39. van Oosterhout C, Hutchinson WF, Wills DPM, Shipley PF (2004) Microchecker: software for identifying and correcting genotyping errors in microsatellite data. Mol Ecol Notes 4: 535–538.
  40. 40. Librado P, Rozas J (2009) DnaSP v5: A software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451–1452.
  41. 41. Excoffier L, Laval G, Schneider S (2006) Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evol Bioinform 47–50.
  42. 42. Gubler DJ (1997) Chapter 2: Dengue and dengue hemorrhagic fever: its history and resurgence as a global public health problem. Gubler DJ, Kuno G, eds. Dengue and dengue hemorrhagic fever. pp. 1–22. London, CAB International,.
  43. 43. Gubler DJ (1998) Resurgent vector-borne diseases as a global health problem. Emerg Infect Dis 4: 442–450.
  44. 44. Slosek J (1986) Aedes aegypti mosquitoes in the Americas: a review of their interaction with the human population. Soc Sci Med 23: 249–257.
  45. 45. Schliessmann DJ, Calheiros LB (1974) A review of the status of yellow fever and Aedes aegypti eradication programs in the Americas. Mosq News 34: 1–9.
  46. 46. Camargo S (1967) History of Aedes aegypti eradication in Americas. Bull World Health Organ 36: 602–603.
  47. 47. Gubler DJ (1989) Aedes aegypti and the Aedes aegypti-borne disease control in the 1990s: top down or bottom up. Am J Trop Med Hyg 40: 571–578.
  48. 48. Gubler DJ (2005) The emergence of epidemic dengue fever and dengue hemorrhagic fever in the Americas: a case of failed public health policy. Rev Panam Salud Publica 17: 221–224.
  49. 49. Chadee DD (2004) Key premises, a guide to Aedes aegypti (Diptera: Culicidae) surveillance and control. B Entomol Res 94: 201–207.
  50. 50. Chadee DD, Rahaman A (2000) Use of water drums by human and Aedes aegypti in Trinidad. J Vector Ecol 25: 28–35.
  51. 51. Avise JC (2004) Molecular Markers, Natural History, and Evolution, 2nd edn. Sinauer Associates, Inc., Sunderland, Massachusetts.
  52. 52. Huber K, Le Loan L, Hoang TH, Tien TK, Rodhain F, et al. (2002) Temporal genetic variation in Aedes aegypti populations in Ho Chi Minh City (Vietnam). Heredity 89: 7–14.
  53. 53. Honório NA, Silva WC, Leite PJ, Gonçalves JM, Lounibos LP, et al. (2003) Dispersal of Aedes aegypti and Aedes albopictus (Diptera: Culicidae) in an urban endemic dengue area in the state of Rio de Janeiro, Brazil. Mem I Oswaldo Cruz 98: 191–198.
  54. 54. Muir LE, Kay BH (1998) Aedes aegypti survival and dispersal estimated by mark-release-recapture in northern Australia. Am J Trop Med Hyg 58: 277–282.
  55. 55. Maciel-de-Freitas R, Codeço CT, Lourenço-de-Oliveira R (2007) Body size-associated survival and dispersal rates of Aedes aegypti in Rio de Janeiro. Med Vet Entomol 21: 284–292.
  56. 56. Maciel-de-Freitas R, Codeço CT, Lourenço-de-Oliveira R (2007) Daily survival rates and dispersal of Aedes aegypti females in Rio de Janeiro, Brazil. Am J Trop Med Hyg 76: 659–665.
  57. 57. Reuben R, Yasuno M, Panicker KN. Studies on the dispersal of Aedes aegypti at two localities in Delhi. World Health Organization, WHO/VBC/72.388.
  58. 58. Maciel-de-Freitas R, Neto RB, Gonçalves JM, Codeço CT, Lourenço-de-Oliveira R (2006) Movement of dengue vectors between human the human modified environment and an urban forest in Rio de Janeiro. J Med Entomol 43: 1112–1120.
  59. 59. Getis A, Morrison AC, Gray K, Scott TW (2003) Characteristics of the spatial pattern of the dengue vector, Aedes aegypti, in Iquitos, Peru. Am J Trop Med Hyg 69: 494–505.
  60. 60. Tsuda Y, Takagi M, Wang S, Wang Z, Tang L (2001) Movement of Aedes aegypti (Diptera: Culicidae) released in a small isolated village on Hainan Island, China. J Med Entomol 38: 93–98.
  61. 61. Garcia-Rejon J, Loroño-Pino MA, Farfan-Ale JA, Flores-Flores L, Rosedo-Paredes ED, et al. (2008) Dengue virus infected Aedes aegypti in the home environment. Am J Trop Med Hyg 19: 940–950.
  62. 62. Sheppard PM, MacDonald WW, Tonn RJ, Grab B (1969) The dynamics of an adult population of Aedes aegypti in relation to dengue haemorrhagic fever in Bangkok. J Anim Ecol 38: 661–702.
  63. 63. Hausermann W, Fay RW, Hacker CS (1971) Dispersal of genetically marked female Aedes aegypti in Mississippi. Mosq News 31: 37–51.
  64. 64. Chadee DD, Doon R, Severson DW (2007) Surveillance of dengue fever cases using a novel Aedes aegypti population sampling method in Trinidad, West Indies: the cardinal points approach. Acta Trop 104: 1–7.
  65. 65. Suwonkerd W, Mongkalangoon P, Parbaripai A, Grieco J, Achee N, et al. (2006) The effect of host type on movement patterns of Aedes aegypti (Diptera: Culicidae) into and out of experimental huts in Thailand. J Vector Ecol 31: 311–318.
  66. 66. Edman JD, Scott TW, Costero A, Morrison AC, Harrington LC, et al. (1998) Aedes aegypti (Diptera: Culicidae) movement influenced by availability of oviposition sites. J Med Entomol 35: 578–583.
  67. 67. Tun-Lin W, Kay BH, Barnes A (1995) The premise condition index: a tool for streamlining surveys of Aedes aegypti. Am J Trop Med Hyg 53: 591–594.
  68. 68. Wiwanitkit V (2009) Dengue vaccines: a new hope? Hum Vaccines 5: Epub ahead of print. PMID: 19337028.
  69. 69. Nene V, Wortman JR, Lawson D, Haas B, Kodira C, et al. (2007) Genome sequence of Aedes aegypti, a major arbovirus vector. Science 316: 1718–1723.
  70. 70. Craig GB (1970) Genetic control of insect vectors of disease. Act IV Congr Latin Zool 1: 15–28.
  71. 71. Pal R (1974) WHO/ICMR programme of genetic control of mosquitoes in India. The use of genetics in insect control, Elsevier, New York, NY.
  72. 72. Morlan HB, McGray EM, Kilpatrick JW (1962) Field tests with sexually sterile males for control of Aedes aegypti. Mosq News 22: 295–300.
  73. 73. Yakob L, Alphey L, Bonsall MB (2008) Aedes aegypti control: the concomitant role of competition, space and transgenic technologies. J Appl Ecol 45: 1258–1265.