Introduction Aedes albopictus (Skuse) (Diptera: Culicidae), the Asian tiger mosquito, is an aggressive, strongly anthropophilic, exophagic, and exophilic mosquito. As an important vector of dengue fever, chikungunya disease, and yellow fever, Ae. albopictus has emerged as a global public health threat [1]–[3]. Ae. albopictus is indigenous to both tropical and temperate regions of Southeast Asia and islands of the western Pacific and Indian Oceans, but it has recently expanded its range to every continent except Antarctica [4], [5]. Unlike wetland mosquito species that oviposit and develop in habitats that are large, predictable, and easy to identify, Ae. albopictus is difficult to locate and control because this species utilizes small, different types of habitats including small containers and spare tires [6]–[8]. Ae. albopictus originated at the edges of forests and bred in natural habitats (e.g., tree holes, bamboo stumps, and bromeliads) and was previously considered a rural vector [9]. However, this species has adapted well to urban environments with larvae now breeding in artificial containers (e.g., tires, cemetery urns, and water storage containers) and has become the most important and sometimes sole vector in urban areas [8], [10], [11]. Ae. albopictus is found almost everywhere, especially in urban areas in southern and southwestern China [12]–[15]. The frequent outbreaks of dengue fever in the cities in southern (mainly Guangdong province) and southeastern coastal (mainly Fujian and Zhejiang provinces) China in the past few decades have caused serious public health concerns [16]–[18]. Although Aedes albopictus is described as a minor vector of dengue and possibly chikungunya in the world, it is emerging as a major dengue vector in China and was responsible for most outbreaks of dengue in China [19], [20] and chikungunya in 2010 in Guangdong, China [21]. Similar to other mosquito vectors, Ae. albopictus needs aquatic habitats to breed and develop, and therefore, it is sensitive to environmental changes [8], [22], [23]. Destruction of breeding habitats is an important strategy to reduce the Aedes mosquito population; eliminating suitable breeding habitats reduces larval development and thus the adult mosquito population. Equally importantly, environmental changes, such as changes in temperature, affect habitat productivity, larval and adult development times, and survival, which in turn directly and indirectly affect disease transmissibility [24]–[32]. Urbanization refers to the increasing population of urban areas. Urbanization predominantly results in the physical growth of urban areas, leading to environmental changes. Urbanization is a global trend that results from economic development. Asian countries including China and India, countries in Southeast Asia, and African countries such as Nigeria are the fastest growing areas in the world, and the unprecedented movement of people into these areas is predicted to intensify in the future [33]. Many problems have emerged as a result of urbanization, including environmental pollution, crowding, and the destruction of natural ecology. The socioeconomic effects of urbanization have been extensively studied by socio-ecologists [34]–[36]; however, the ecological effects and their impact on vector biology and vector-borne infectious disease transmission remain unclear. Most dengue fever outbreaks occur in the urban areas of China, and these outbreaks have become more frequent over the past decade [18], [20], [37]. There is an accelerating trend of urbanization in China; will this process of urbanization accelerate dengue fever outbreaks? Changes in environmental conditions as a result of urbanization may directly and/or indirectly affect the ecology of mosquitoes, e.g., larval habitat availability and suitability, development, and survivorship. Because Ae. albopictus has invaded Europe (e.g., Italy and France) and the Americas (e.g., USA), which increases the global vulnerability to dengue fever outbreak, therefore, it is crucial to evaluate its adaptations to urban environments. We hypothesized that urbanization increases Ae. albopictus larval habitats and survivorship and accelerates the development of larvae and adults. This study explored the ecology of Ae. albopictus in different settings (urban, suburban, and rural) in the Great Guangzhou area, China. Field surveys of larval habitat availability, larval development and adult mosquito life-table experiments were conducted in semi-natural conditions to test the hypothesis. Materials and Methods Study areas The field surveys of larval habitat availability and semi-natural condition larval development and adult mosquito life-table experiments were carried out in Guangzhou, the capital city of Guangdong province, China. Guangzhou is the largest city in southern China, and it is located in the Pearl River Delta, where numerous cities form a Canton-Macao-Hong Kong economic development zone. The annual average temperature in Guangzhou is 21.6°C, and its annual rainfall is approximately 1,980 mm. This climate is ideal for the development and reproduction of Ae. albopictus. The city has experienced rapid expansion during the recent regional economic development. Several major dengue fever outbreaks have occurred in this area since 1980, and Ae. albopictus is the sole dengue vector [13], [38], [39]. Therefore, Great Guangzhou is an ideal place to study the impacts of urbanization on Ae. albopictus. The study was conducted in three areas that represented urban, suburban, and rural settings in Guangzhou (Figure 1 and Figure S1). Each study area was approximately 1.8 km2. The distance between each area was approximately 24 km. Tonghe (113°19′E, 23°11′N, 31 m above sea level (a.s.l.)) is an urban area with a population density of >3,000 people/km2. The land use types are primarily residential and commercial buildings and public services such as schools and hospitals, filled with trees and grasses. Liangtian (113°23′E, 23°21′N, 25 m a.s.l.) is a suburban area with a population density of approximately 1,000 people/km2, and land use includes a mixture of residential, manufacturing, and farmland. Dengcun (113°33′E, 23°30′N, 42 m a.s.l.) is a rural area and has a population density of 80%. The number of larvae and pupa were counted and the larval stages identified daily. Emerged adult mosquitoes were counted daily, and their sexes were determined. The experiments began in October and were conducted simultaneously in all study sites. Adult mosquito life tables Mosquitoes used for adult life-table experiments were all F0 individuals who originated from different habitats in the study areas. Newly emerged ( 0.05). 10.1371/journal.pntd.0003301.g003 Figure 3 Mean density of adult Aedes albopictus (adults/trap/night) in the three study sites from July to November. Square root transformed data were used, and the 95% confidence interval is shown as a bar. Life table analysis of immature Ae. albopictus Ae. albopictus adult emergence rates were significantly different among urban, suburban, and rural areas regardless of in natural habitats or with food supplement groups (χ2-test, all P 0.05) (Figure 4B). 10.1371/journal.pntd.0003301.g004 Figure 4 Survival rate and development time of Aedes albopictus larvae in urban, suburban and rural areas. A and B: survival rate; A: natural habitat; B: food supplement. C and D: development time; C: natural habitat; D: food supplement. Values are the mean ±95%CI. For the natural habitat group, larval development time in urban areas was significantly shorter than that in both the suburban and rural areas (male F = 19.0, d.f. = 2, 92, P 50% shorter in control groups than it was in natural habitat groups in all study sites (Figure 4, Table S1). The larval stage-specific development time were shown in Figure 5. Young larval (1st and 2nd instar) and pupa developed significantly faster in urban areas than that in suburban and rural areas (Tukey HSD test, all P 0.05) (Figure 5). 10.1371/journal.pntd.0003301.g005 Figure 5 Stage survival rate and stage development time of Aedes albopictus larvae in urban, suburban and rural areas. A and B: stage survival rate; A: natural habitat; B: food supplement. C and D: stage development time; C: natural habitat; D: food supplement. Values are the mean ±95%CI. Life table analysis of adult Ae. albopictus From August to September, the life span of female adult mosquito was significantly longer in urban areas than that in suburban and rural areas, but the difference in median survival time between suburban and rural areas was insignificant (Figure 6, Table S2). Adult male mosquito survival time was significantly different among study sites (χ2 = 17.4, d.f. = 2, P 0.05) (Table S2). Survival curves were similar in females between urban and suburban areas but different from those in rural areas (Figure 6C and 6D). 10.1371/journal.pntd.0003301.g006 Figure 6 Survivorship curve of adult Aedes albopictus in different seasons. The left panel is for August to September, and the right panel is for October and November. The top panel (A and B) is for male, and the bottom panel (C and D) is for female. From October to November, the median survival of adult female mosquitoes in urban and suburban areas was significantly longer than in rural areas but the difference between urban and suburban areas was insignificant (Figure 6, Table S2). The median survival of males was significantly different among the three sites (χ2 = 181.1, d.f. = 2, P<0.001), with the longest and shortest survival times in urban and suburban areas, respectively (Figure 6, Table S2). The average outdoor temperature in urban areas (24.8±2.6°C) was significantly higher than that in suburban (22.8±3.6°C) and rural areas (21.9±2.4°C), and there was no difference in temperature between suburban and rural areas (Table S2). There was no significant difference in the relative humidity among the three areas. (F = 1.9, d.f. = 2, 134, P = 0.15). Mean daily survival rates were similar in all study sites but differed between males and females (Figure 6, Table S2). Survival curves were similar in females between rural and suburban areas but very different in those from urban areas, which showed prolonged survivorship (Figure 6D). Factors influencing the presence of immature Aedes albopictus Stepwise logistic regression revealed that six factors were significantly associated with the presence of immature mosquitoes in the study sites (Table 3). Habitat Ae. albopictus larval presence rate was significantly greater in the urban area than in suburban and rural areas (OR = 1.71, P<0.001), and the suburban area was significantly higher than the rural area (OR = 1.67, P<0.001). The presence of Ae. albopictus larvae was significantly in negative correlation with habitat water depth (OR = 0.03, P<0.001); whereas, it showed positive correlation with habitat water surface area (OR = 3.95, P<0.001). The presence of Aedes larvae was significantly greater in clean water than that in tinted or polluted water (OR = 1.889, P<0.001), and greater in tinted water than polluted water (OR = 1.78, P = 0.034). Shading (regard less of fully shaded or half-shaded), compare to open area, was positively affecting the presence of Aedes larvae (OR = 2.29, P<0.001). The presence of Ae. albopictus larvae was also positively correlated with habitats that have leaves on water surfaces (OR = 2.25, P<0.001), and with habitats that have soil and moss substrates (OR = 1.71, P<0.001). 10.1371/journal.pntd.0003301.t003 Table 3 Factors that were significantly associated with the presence of immature Aedes albopictus. Term Sub-term Estimate Chi-square P Odds ratio (95% CI) Intercept - 0.80 109.10 <.0001 Surface area (m2) - 0.14 17.70 <.0001 3.95 [2.11, 7.60] Water depth (m) - −0.04 108.54 <.0001 0.03 [0.01, 0.05] Study area Urban vs. Suburban and Rural 0.27 89.94 <.0001 1.71 [1.53, 1.91] Suburban vs. Rural 0.25 47.55 <.0001 1.66 [1.44, 1.92] Water clearance Clear vs. tinted and polluted 0.32 20.75 <.0001 1.88 [1.44, 2.48] Tinted vs. polluted 0.29 4.51 0.0338 1.77 [1.05, 3.03] Surface type Leaf vs. no substrate, soil, moss, and sand 0.41 189.44 <.0001 2.25 [2.01, 2.53] Soil and moss vs. no substrate and sand 0.27 57.19 <.0001 1.71 [1.49, 1.97] Canopy cover Full shade vs. full sun and partial shade 0.41 178.50 <.0001 2.29 [2.02, 2.58] Full shade vs. partial shade 0.30 69.75 <.0001 1.81 [1.57, 2.07] Note: A stepwise logistic regression was used. The presence of immature Aedes albopictus was used as the dependent variable, and variables that were insignificant at the 0.05 level are not included in the table. Discussion Outbreaks of dengue fever in China were reported in Hainan province and southern Guangdong province in the 1980s and have been reported in Zhejiang province in 2004, illustrating a 2,000 km expansion from subtropical to temperate areas over 30 years [16]. Among these outbreaks, Ae. albopictus was the only vector reported [13], [20], [44]. Although the causes of dengue fever outbreaks are multi-factorial, environmental changes such as urbanization may be one of the leading factors. We found that in urban areas, there are more Ae. albopictus habitats. In addition, urban areas promoted faster larval and pupal development, and higher larval-to-adult survival rate compared to rural areas. Ae. albopictus mosquito is strongly anthropophilic and has a higher blood-feeding rate in urban areas, where human population density is great, than that in rural areas [29], making it a more susceptible vector in urban areas. Because there is no effective drug therapy or vaccine for dengue fever, vector population control is by far the only effective method for reducing dengue virus transmission. In this context, understanding the vector ecology and biology is essential for developing dengue control strategies. Unfortunately, it is unclear how urbanization impacts the ecology of Ae. albopictus, and the lack of this key knowledge hinders disease control efforts. We found that, in the similar sampling area, the total number of potential habitats and the number of Aedes-positive habitats were significantly higher in urban areas than in suburban and rural areas. Urban areas have 10-fold higher human population density and more frequent human activities than do suburban and rural areas, leading to a larger number of artificial containers such as abandoned tires, disposable food tins, and flowerpots, which are all favorable breeding habitats for Ae. albopictus [7], [23], [45]. Larger size and higher density of human populations also mean more opportunities for Ae. albopictus blood feeding. Previous study found that Ae. albopictus has a higher blood-feeding rate in urban areas than in rural areas, most likely due to host availability [29]. Additionally, the existence of stable and abundant artificial containers produced by human activities serve as larval development sites, facilitating large mosquito densities in urban areas [11]. Urbanization shifts mosquito breeding sites from natural habitats to artificial habitats. These artificial habitats are usually small containers such as used tires and disposable containers, and they are often directly exposed to sunlight. Therefore, the water temperature in these habitats is higher than in rural areas. In our study sites, the average water temperature of urban habitats was 5°C higher than in suburban and rural areas. Similarly, vegetation changes and land use changes in urban areas may affect the radiation budget and energy balance of the land surface and thus may modify the microenvironments, e.g., food sources that enhance larval survival. These changes facilitate the development of immature Ae. albopictus, i.e., shorten the larval-to-adult development time and enhance the larva survival rate. Our findings are consistent with other studies conducted in different countries and for different mosquito species [24], [30], [46]–[48]. Compared with the food supplemental group, we found that added food sources significantly affect the developmental time and survival rate of immature mosquitoes, which implies that the habitat types in different areas may affect larval development differently due to the difference in availability of nutrients. However, the effects were more pronounced in urban areas than in suburban and rural areas, implying that other factor such as water temperature may play a more important role than food supply in urban areas. These results demonstrated that larvae develop and survive better in urbanized areas, in other words, Aedes larvae is better adapted to urban environment. Similar to a study conducted in the United States [49], the urban area had a higher pupal and larval density than other two areas; thus, the urban habitats had a higher capacity to support larval development. The reason might be that urban areas had less predators, more nutrition from a “dirtier” environment, or even less drift from agricultural insecticides. Pupal productivity is a good indicator of the abundance of adult mosquitoes [50]–[52]. The surveillance of adult mosquitoes in this study supports this conclusion, i.e., urbanization leads to a higher population density of adult Ae. albopictus. Higher mosquito density does not necessarily lead to increased disease transmission if adult mosquitoes have a very short life span. We found that both male and female mosquitoes in urban areas had the longest life spans. This result may be due to environmental factors such as air temperature and humidity. The average temperature in urban areas is higher than in suburban and rural areas. Longer adult life spans may enhance disease transmission, although the exact correlation between vector capacity and adult life span needs to be further explored. In this study we fed the adult mosquito with 10% sugar solution without blood, which might have led to exerted stress on the females during multiple gonotrophic cycles and affected the longevity of the female mosquitoes. We observed that the mortality of adult mosquitoes in rural area changed dramatically around day 15, because the air temperature in rural area showed a 5.7°C increase from day 11 to day 15 compared to the first 11 days. This drastic increase in temperature might have influenced the mortality rate of the adult mosquito. In our survey, we found that the distribution of immature Ae. albopictus was not random. Habitat surface area, canopy coverage, water turbidity, water depth, and substrate type were all important factors influencing habitat selection. These findings confirmed other studies reporting preferences for urban areas [11], [49], shaded containers [49], clean water [53], water with foliage [49], and larger surface area [49], [54]. These results illustrated the complex ecology of Ae. albopictus, which makes controlling this mosquito species difficult in light of its recent global expansion. In conclusion, the results of this study indicated that urbanization has a significant impact on the ecology of Aedes albopictus. In the urbanizing and urbanized area, the changed environment became more suitable for the growth and development of Ae. albopictus, the condensed population produced more kinds of containers for larval habitats and more blood sources for adult replication. This might be the reason for quick adaptation of Ae. albopictus in urban areas. The epidemic of dengue is largely dependent on vector population. Developing countries such as China and other Southeastern Asian countries experiencing rapid urbanization are under sustained risk of dengue outbreaks. Supporting Information Figure S1 Landscape of study areas in Guangzhou, Guangdong province, China. A and B: Urban; C and D: Suburban; E and F: Rural. (TIF) Click here for additional data file. Figure S2 Weekly temperature, humidity, and half-month precipitation data in urban, suburban and rural areas in 2013. A: Temperature; B: Relative humidity; C: Precipitation in one half month. A and B: Values are the mean ± standard error. (TIF) Click here for additional data file. Figure S3 The most abundant Aedes albopictus breeding habitats in the three study sites. Urban area: (1) aquatic plant, (2) plastic bucket, (3) disposable food tin, (4) gutter, (5) tire; suburban area: (6) tire, (7) disposal food tin, (8) clay pottery, (9) plastic bucket, (10) plastic basin; rural area: (11) clay pottery, (12) plastic bucket, (13) disposal food tin, (14) plastic basin, (15) building tool. (TIF) Click here for additional data file. Table S1 Development of Ae. albopictus larvae. Note: Values are the mean ± standard deviation. Values in the same column within the same experimental group connected with the same letter indicate a significant difference at the 5% level. (DOCX) Click here for additional data file. Table S2 Life table analysis of Ae. albopictus adults. Note: Values are the mean ± standard deviation. Values in the same column connected with the same letter indicate a significant difference at the 5% level within the same experimental group. (DOCX) Click here for additional data file. File S1 Original data of the manuscript. Data used for making the figures and tables for the manuscript. (RAR) Click here for additional data file.
