https://orcid.org/0000-0003-0710-5086
Figures
Abstract
Background
Global warming, urbanization and the resumption of global population movements post-COVID-19 have made the prevention and control of dengue and its vector transmission more challenging. To tackle this issue, this study evaluated and predicted the suitable habitats of Aedes albopictus and Aedes aegypti in China. This study aims to identify the key influencing factors, analyze patterns of habitat expansion and contraction, and explore regions in China at risk of dengue fever outbreaks in the future.
Methodology/principal findings
This study utilized mosquito distribution data from 2010 to 2023 and employed Maxent to map the distribution of suitable habitats. The key influencing factors and response curves were further analysed, and patterns of habitat expansion and contraction were investigated. The findings reveal that the main variables affecting the distributions of Ae. albopictus and Ae. aegypti are annual precipitation, annual mean temperature, and land use and land cover changes. The suitable habitat for Ae. albopictus shows a significant northward expansion trend, reaching large areas in Shandong and Henan province by 2030. The suitable habitat area for Ae. albopictus is increasing annually and can reach approximately 2.38 million square kilometers by 2030. Compared to the outbreak year of dengue fever in China in 2019, the suitable habitat area for Ae. albopictus in 2030 will increase by approximately 17.06%, with a growth of 2.57% in the sum of high-risk and medium-risk suitable habitat areas. In contrast, the suitable habitats of Ae. aegypti are primarily concentrated in Guangdong, Hainan and Yunnan Provinces.
Conclusion/significance
This study compared the potential changes in the distributions of suitable habitats for Ae. albopictus and Ae. aegypti in 2014, 2019, and 2023 and predicted suitable habitats for 2030, as well as contraction and expansion trends in the suitable habitats of Ae. albopictus. The findings aim to identify regions at risk of future dengue fever outbreaks in China, providing a scientific basis for public health authorities to develop effective dengue prevention and control strategies.
Author summary
Dengue fever is one of the most significant mosquito-borne diseases in tropical and subtropical regions. In June 2024, the World Health Organization (WHO) Pathogens Prioritization framework identified Orthoflavivirus denguei as a priority pathogen in all six WHO regions. In 2014 and 2019, China experienced severe dengue outbreaks. This study employs climate and environmental-economic variables in China in 2014, 2019, 2023 and 2030 to explore the main influencing factors of mosquito-borne transmission and the trend of habitat expansion and contraction, with the goal of assessing the risk areas for future dengue outbreaks in China. Research shows that the main factors that influence the distributions of Ae. albopictus and Ae. aegypti are the annual mean temperature, annual precipitation and urban/industrial/residential land. Aedes albopictus is the main vector of dengue fever transmission in China, and its suitable habitat will expand to the north and be widely distributed in Shandong Province by 2030. The suitable habitats of Ae. aegypti were mainly concentrated in Guangdong, Hainan and Yunnan Provinces. The results of this study provide a scientific basis and early warning system for identifying and predicting potential high-risk areas for dengue fever epidemics in China in the past 10 years and in the future in 2030.
Citation: Xie Y, Wang B, Chen Q, Wei H, Ke Y, Xie F, et al. (2025) Forecasting high-risk areas for dengue outbreaks in China: A trend analysis of Aedes albopictus and Aedes aegypti distributions from 2014 to 2030. PLoS Negl Trop Dis 19(7): e0013237. https://doi.org/10.1371/journal.pntd.0013237
Editor: Maira Aguiar, Basque Center for Applied Mathematics: Centro Vasco de Matematicas Aplicadas, SPAIN
Received: December 2, 2024; Accepted: June 12, 2025; Published: July 9, 2025
Copyright: © 2025 Xie et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its Supporting Information files.
Funding: This study was partly supported by the National Natural Science Foundation of China (82404329); the Postdoctoral Fellowship Program of China (GZC20240890); the NHC Key Laboratory of Parasite and Vector Biology (National Institute of Parasitic Diseases, Chinese Center for Disease Control and Prevention, NHCKFKT2024-3); Public Health Talent Cultivation Support Program (2023–2025); Hubei Province Public Health Leading Talent Program (2021–2025); Fujian Provincial Department of Science and Technology Program, grant number 2023Y0004; Xiang'an Innovation Laboratory/National Key Laboratory for Infectious Disease Vaccine Development Science and Technology Project, Grant No. 2024XAKJ0100003. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Dengue fever is one of the most significant mosquito-borne diseases in tropical and subtropical regions. Dengue cases surged worldwide in 2023 and 2024 [1,2]. The World Health Organization (WHO) assigned the highest level of emergency (grade 3) to the global increase in dengue cases [3].
China experienced severe outbreaks of dengue fever in 2014 and 2019 [4], indicating the potential for large-scale transmission of dengue fever in the country. With respect to global change, urbanization has directly promoted the spread of mosquito vectors, and the recovery of global population mobility post the coronavirus disease 2019 (COVID-19) pandemic has sharply increased the risk of dengue fever import into China, presenting greater challenges for its prevention and control. Even more concerning is the expanding distribution of Aedes albopictus, the main vector of dengue fever in China. Currently, Ae. albopictus has been reported in provinces and autonomous regions such as Liaoning, Hebei, Shanxi, Shaanxi, and Sichuan [5]. Given the combination of mosquito vector expansion and increased risk of imported cases, along with the lack of widespread dengue fever vaccination, the situation for dengue prevention and control in China has become more complex and urgent.
In this context, controlling the transmission vectors of dengue fever, Ae. albopictus and Aedes aegypti, is crucial. However, owing to China’s vast territory, which spans three climate zones, and the significant differences in climate and terrain, mosquito monitoring is challenging. This makes vector control and dengue prevention particularly difficult. Therefore, comprehensively assessing and dynamically monitoring the distributions of Ae. albopictus and Ae. aegypti in China, and updating dengue vector distribution maps, have become important for improving dengue surveillance and early warning capabilities, as well as formulating targeted control strategies.
To address these problems, this study employs an ecological niche model and utilizes the latest climate data and environmental-economic related variables to evaluate the distributions of Ae. albopictus and Ae. aegypti in China. Given that 2014 and 2019 witnessed significant dengue fever outbreaks in China, Maxent models were built to map the suitable habitat distribution and potential trend changes in Ae. albopictus and Ae. aegypti in 2014, 2019, and 2023. Furthermore, the model was used to predict the suitable habitat distribution for these mosquitoes in 2030.
The objective of this study was to pinpoint potential high-risk areas for future dengue outbreaks. The findings will offer scientific insights to public health professionals, aiding in the development of targeted mosquito vector management and monitoring strategies. Additionally, these findings will provide scientific backing for public health departments to devise strategies aimed at preventing and controlling dengue. Ultimately, these efforts will ultimately guide the implementation of effective public health interventions to control the transmission of dengue fever.
2. Methods
The technical approach of this study is shown in Fig 1. In short, we divided the Aedes mosquito distribution data into three time periods and collected the corresponding climate variables and environmental-economic variables. Then, we constructed model A, model B, and model C based on different combinations of variables. Finally, we used AUC, response curves, and so on to evaluate and explain the role of the variables, and described the contraction and expansion of the suitable habitats in 2014, 2019, 2023, and 2030.
2.1. Data collection
2.1.1. Mosquito data.
The occurrence data for Ae. albopictus and Ae. aegypti in this study were obtained from two primary sources. The first source is a comprehensive and systematic literature search of databases such as CNKI, SinoMed, and PubMed, which yielded mosquito coordinates from relevant literature published between 2010 and 2023. The second source is the Global Biodiversity Information Facility (GBIF) database (https://www.gbif.org/), which contains numerous global occurrence records of species. We filtered for Ae. albopictus and Ae. aegypti coordinate records in China from 2010 to 2023.
Considering the long-term trend changes in variables over time, species behavior changes, and the number of species points, we focused more on the distribution of newly added species points over five years for each study year. This approach better reflects the distribution trend changes of mosquitoes. We divided the mosquito coordinates into three time periods: 2010–2014, 2015–2019, and 2019–2023. There are 1,794 records of Ae. albopictus and 407 records of Ae. aegypti.
2.1.2. Variable acquisition.
The climate variables include the annual mean temperature (AMT) and annual precipitation (APP) for the years 2014, 2019, 2023, and 2030. These data were obtained from the National Earth System Science Data Center (https://www.geodata.cn/index.html) [6–8]. For the future year of 2030, we selected the SSP245 scenario for the projections. The 2000 digital elevation model (DEM) was obtained from the Resource and Environmental Science Data Platform (RESDC: https://www.resdc.cn/). Land use and land cover change (LUCC) data for 2013, 2018, and 2023 were obtained from the Resource and Environmental Science Data Platform (RESDC: https://www.resdc.cn/doi:10.12078/2018070201). Since LUCC data for 2014 and 2019 were not included in this dataset and considering the small changes in LUCC over the years, we used LUCC data from the previous years, 2013 and 2018, as substitutes to ensure the continuity and reliability of the analysis. LUCC data for 2030 were obtained from the Global Change Research Data Publishing & Repository (https://geodoi.ac.cn/WebCn/Default.aspx) [9], also under the SSP245 scenario. LUCC data result from the combined effects of natural conditions and human activities (e.g., urbanization of residential areas). As the only categorical variable, we reclassified LUCC into six categories (arable land, forestland, grassland, urban/industrial/residential land, unused land, and water bodies) to standardize the values. These categories represent the different land use types and coverage changes in China. The map was obtained from GaryBikini (https://zenodo.org/records/10624971) [10].
All variables were clipped and adjusted to the geographical region of China via ArcGIS 10.8 and then resampled to a uniform resolution of 1 km × 1 km.
2.2. Data analysis
2.2.1. Establishment of the Maxent model.
Maxent 3.4.3 (http://biodiversityinformatics.amnh.org) is widely used to assess and predict suitable habitat distributions of species because of its inherent stability and high accuracy [11]. Among various modelling methods, the Maxent model has demonstrated outstanding predictive performance in species niche modelling [12]. In short, the model uses the Maximum Entropy method to assess the probability of species occurrence. By using the known distribution locations of species and corresponding explanatory variables, the Maxent model analyzes the ecological requirements of the species, mapping them to different geographical areas to predict the potential distribution of the species in a specific region. For a more detailed understanding of the model’s principles and usage, please refer to its official website (http://biodiversityinformatics.amnh.org) and articles [13–15].
To differentiate and assess the impacts of various types of variables and their interactions on the prediction of habitat suitability, we constructed three models based on the original variables: climate model A, which included climate variables (AMT and APP); environmental-economic model B, which included LUCC and DEM variables; and all-factor model C, which included all the variables (Table 1).
In this study, 75% of the species distribution points were set as training points, and the remaining 25% were set as testing points (randomly selected by the program). Each model was repeated 10 times, and the average value was taken as the final prediction result.
2.2.2. Variables analysis.
Generalized variance-inflation factors (GVIFs) are calculated with vif function in car R package, with GVIF greater than 10 considered highly collinearity [16]. No variables exhibited high collinearity (S3 File).
The jackknife test reflects the effect of sequentially excluding variables and using variables alone [17]. The larger the gain obtained when only a single variable is used, the more information that a variable contains that is not present in other variables [18], which contributes more to the model’s predictive power.
2.2.3. Model evaluation and division of habitat suitability.
We imported the variables into the Maxent model according to the intended model structure. The area under the receiver operating characteristic curve (AUC) was used to assess the predictive performance of the model. The larger the AUC value is, the better the model’s predictive ability: excellent (0.90–1.00), very good (0.80–0.90), good (0.70–0.80), fair (0.60–0.70), and poor (0.50–0.60) [19].
Considering both the model AUC values and the distribution and quantity of species occurrence points, we used the Aedes mosquito distribution data from 2015 to 2019 along with relevant variables to train the model, which was then used to predict the habitat suitability for Aedes mosquitoes in 2030. The suitability maps for Ae. albopictus and Ae. aegypti were generated in ArcGIS 10.8, where values ranged from 0 to 1, with higher values indicating greater suitability for mosquitoes. To more intuitively display the regional suitability risk, we categorized the habitat suitability values into four risk levels: 0.00–0.05 for uninhabitable habitat, 0.05–0.25 for low suitability habitat, 0.25–0.50 for medium suitability habitat, and 0.50–1.00 for high suitability habitat.
3. Results
3.1. Distribution
Using ArcGIS 10.8, the distribution points of Ae. albopictus and Ae. aegypti were visualized and displayed on a map of China, as shown in Fig 2. The results indicate that Ae. albopictus has a relatively wide distribution and is found in southern, eastern, and central China, with some scattered presence in the northern regions. In contrast, the distribution of Ae. aegypti in China is limited and sparse and is concentrated mainly in Guangdong, Yunnan, and Taiwan.
The map was obtained from GaryBikini (https://zenodo.org/records/10624971).
3.2. Model performance evaluation
In species distribution modelling, the AUC value is a commonly used metric for comparing the performance of different models. After 10 repetitions, all the tested models demonstrated excellent (AUC > 0.9) predictive performance. The results are shown in S4 File, where all-factor Model C consistently had the highest AUC values across all years.
3.3. Evaluation of influencing factors
To fully consider the interactions between variables, we analysed the years with the highest AUC in all-factor Model C (2023 for Ae. albopictus and 2019 for Ae. aegypti).
The percentage contribution values indicate that, for Ae. albopictus, APP, AMT and LUCC contributed the most to improving model performance, accounting for more than 90% (92.6%) of the total. For Ae. aegypti, AMT and LUCC were the most significant contributors to improving model performance, accounting for more than 90% (91.3%) (Fig 3).
A. Variable percent contribution to Aedes albopictus. B. Variable percent contribution to Aedes aegypti.
Considering all the jackknife test results (Fig 4), the impacts of the variables on Ae. albopictus and Ae. aegypti were consistent. When only a single variable was used for modelling, AMT, APP and LUCC had the greatest impact on the model’s predictive ability. Among these, the exclusion of LUCC had the most significant impact on the model’s overall performance.
Panels A, B, and C represent the test AUC, test gain, and regularized training gain of Aedes albopictus. Panels D, E, and F represent the test AUC, test gain, and regularized training gain of Aedes aegypti. Light green indicates the impact on the model when this variable is excluded, and light blue indicates the independent impact of this variable to the model.
Therefore, considering both the percentage contribution values and the jackknife test results, the main influencing variables for both Ae. albopictus and Ae. aegypti are APP, AMT and LUCC.
3.4. Response curve
Based on the Maxent output results and the reasonable range of variables, we generated response curves for the three most significant influencing variables (Fig 5). The mean + /- one standard deviation is an indicator that describes the central tendency and degree of dispersion of the results. Aedes albopictus is most active at around 25°C, with a minimum temperature threshold of 12.5°C (Fig 5A), while Ae. aegypti requires a relatively higher minimum temperature of around 17.5°C (Fig 5D). The predicted probability of both mosquito species increases with rising APP (Fig 5B and 5E). In terms of LUCC, urban/industrial/residential land had the greatest impact on the predicted probability of Aedes mosquitoes (Fig 5C and 5F).
Panels A, B and C show the AMT, APP, and LUCC response curves of Aedes albopictus. Panels D, E and F shows the AMT, APP, and LUCC response curves of Aedes aegypti. These curves show how each variable affects the Maxent prediction. The curves show how the predicted probability of presence changes. The red curves show the mean response of the 10 replicate Maxent runs, and the blue range represents the mean + /− one standard deviation.
3.5. Habitat suitability
On the basis of different variable combinations in models A, B and C, different models reflect the impact of considering AMT and APP, LUCC and DEM, as well as the comprehensive consideration of all factors affecting habitat suitability for Aedes mosquitoes.
Model A revealed that the suitable habitats for Ae. albopictus were concentrated in southern provinces, such as Guangdong, Hainan, Guangxi, and Taiwan, and gradually expanded northwards over time to regions such as Anhui, Jiangsu, and Shandong (Fig 6). The suitable habitats for Ae. aegypti are concentrated mainly in southern coastal provinces, including Guangdong, Guangxi, Hainan, and Taiwan (Fig 7). Model B, which considers LUCC and DEM variables, indicates that the suitable habitats for Ae. albopictus are primarily distributed in the eastern and southern provinces of China (Fig 6). The suitable habitats for Ae. aegypti are less stable, showing significant variation across years, but are also concentrated mainly in the eastern and southern provinces (Fig 7). Model C, which integrates AMT, APP, LUCC and DEM, revealed that the suitable habitats for Ae. albopictus gradually expanded from 2014 to 2030. Guangdong, Hainan, Guangxi, and Taiwan show consistently high suitability across all time points. Additionally, the suitable habitats have continuously expanded northwards over time, and by 2030, they have spread to most areas of Shandong and Henan (Fig 6). This model indicates that the suitable habitats for Ae. aegypti are relatively limited and are primarily concentrated in southern regions such as Guangdong, Hainan, and Yunnan (Fig 7).
The habitat suitability values ranged from 0 to 1, with higher values indicating greater suitability for mosquitoes. The map was obtained from GaryBikini (https://zenodo.org/records/10624971).
The habitat suitability values ranged from 0 to 1, with higher values indicating greater suitability for mosquitoes. The map was obtained from GaryBikini (https://zenodo.org/records/10624971).
The habitat suitability results (model C) indicate that the suitable habitats of Ae. albopictus are widely distributed in China, while the habitat suitability of Ae. aegypti are relatively limited, confined to a few areas in the south (Guangdong, Hainan, Yunnan, and Taiwan). Therefore, the following sections will mainly demonstrate the classification and assessment of the suitable habitats of Ae. albopictus, and the classification of the suitable habitat risk levels of Ae. aegypti is detailed in S1 Fig.
3.6. Classification of risk levels in terms of habitat suitability
From a temporal perspective, all three models for Ae. albopictus show a continuous expansion of suitable habitat risk levels and ranges over time (Fig 8). In contrast, Ae. aegypti exhibited some fluctuations, with the risk levels and ranges being relatively consistent in 2014, 2019 and 2030 but expanding in 2023 (S1 Fig).
0.00–0.05 indicates uninhabitable habitat, 0.05–0.25 indicates low suitable habitat, 0.25–0.50 indicates medium suitable habitat, and 0.50–1.00 indicates high suitable habitat. The map was obtained from GaryBikini (https://zenodo.org/records/10624971).
Compared with the results from the models (Fig 8), Model A, which considers climate variables, reveals a broader suitable habitat range with higher risk levels in central, southern, and eastern China and southwestern provinces such as Guizhou and Yunnan. Model B, which incorporates LUCC and DEM, predicts suitable habitats that are primarily distributed in eastern and southern China and even some parts of northeastern China. Model C, which integrates all the factors, reflects the comprehensive suitable habitat ranges for Ae. albopictus.
In terms of the distribution differences between the two species of mosquitoes (Figs 8, and S1), all three models for Ae. albopictus present relatively large suitable habitat ranges. In Model C, which integrates all the factors, the high-risk suitable habitats are primarily concentrated in Taiwan, Hainan, and Guangdong. However, some high-risk areas are gradually expanding northwards, reaching Shandong, Jiangsu, and Henan Provinces (Fig 8). In contrast, the suitable habitat range for Ae. aegypti in Model C was narrower, with high-risk suitable habitats scattered in parts of Hainan, Yunnan, and Guangdong (S1 Fig).
3.7. Expansion and contraction
In Model C, the changes in habitat suitability risk levels and the expansion and contraction of Ae. albopictus during 2014–2019, 2019–2023, and 2023–2030 are shown in Fig 9. The distribution of Ae. albopictus habitat suitability clearly tends to expand northwards, with large areas extending to Shandong by 2023 and 2030.
A. The risk level changes, expansion and contraction of the suitable habitats for Aedes albopictus during the three time periods of 2014–2019, 2019–2023, and 2023–2030. B. The percentage of suitable and uninhabitable habitat areas for Aedes albopictus in each year, as well as the suitable habitat areas for each risk level (10,000 km2). The map was obtained from GaryBikini (https://zenodo.org/records/10624971).
From a quantitative perspective, the suitable habitat area for Ae. albopictus is generally increasing annually and can reach approximately 2.38 million square kilometers by 2030. Compared to the outbreak year of dengue fever in China in 2019, the suitable habitat area for Ae. albopictus in 2030 will increase by approximately 17.06%, with a growth of 2.57% in the sum of high-risk and medium-risk suitable habitat areas (Fig 9).
4. Discussion
This study compared and predicted the changes in habitat suitability risk levels and ranges for the dengue fever vectors Ae. albopictus and Ae. aegypti in China for the years 2014, 2019, 2023, and 2030 to explore regions in China that may experience dengue outbreaks in the future.
From a temporal perspective, this study uses the latest climate and environmental-economic variables to assess the suitable habitats for Aedes mosquitoes, which is consistent with previous studies [20,21]. The high suitable habitats predicted for Ae. albopictus, such as Guangdong, Guangxi and Hainan, are regions in China with a high burden of dengue fever [4]. According to the habitat suitability results in 2014, the suitable habitats for Aedes mosquitoes were mainly concentrated in southern and southeastern China, which aligns with the outbreaks of dengue fever in Guangdong, Guangxi and other provinces in 2014 [22]. In 2019 [23], outbreaks occurred in Guangdong, Zhejiang, Jiangxi, and Hainan Provinces, and the outbreaks in Zhejiang and Jiangxi also corresponded with the predicted expansion and contraction of suitable habitats. Zhuowei Li [24] analysed the spatial and temporal distributions of dengue fever from 2019 to 2023 and reported that most of the southern regions of China are at risk of dengue transmission, with Southwest China and South China being key risk areas. This finding is also consistent with our 2019 and 2023 suitability results. In addition, cases of dengue fever in China have been reported in Shandong, Henan and Chongqing Provinces [25–27], which is also consistent with our prediction. According to the study by Zhao [28], there were cases of dengue fever in Shandong Province in 2023. On the basis of the results of suitable habitats for Ae. albopictus and the expansion and contraction of these areas, Shandong Province has gradually transitioned from having scattered suitable habitats to having large suitable habitats for Ae. albopictus, including some high-risk suitable habitats. The suitable habitat range for Ae. albopictus is generally increasing annually, and the high-risk and medium-risk habitat areas will gradually increase from 2014 to 2030. These results collectively indicate that the risk of dengue fever outbreaks can be reasonably predicted through the distribution of Aedes mosquitoes.
Comparing the years 2014, 2019, 2023, and 2030, the suitable habitats of Ae. albopictus will further expand northward in 2030. In addition to the southern regions such as Guangdong, Guangxi, Hainan, and Yunnan, which often report dengue fever cases, northern provinces like Shandong and Henan will also have large areas of new suitable habitats, and there will also be some high-risk level suitable habitats. The results of this study predict the areas in China where dengue fever may outbreak in the future through the analysis of the suitable habitats of Aedes mosquitoes, providing a reference for the prevention and control of dengue fever.
From the perspective of modelling, Model C, which takes into account all the factor variables, may provide a more comprehensive prediction of the suitable habitat range. However, it is worth noting that, through the aforementioned analysis of influencing factors, AMT and APP are two important factors affecting the distribution of suitable habitats for Aedes mosquitoes. Therefore, the results of Model A may also represent a warning that the addition of suitable habitat areas in Model A, compared with the results of Model C, may also lead to the spread of dengue fever cases in these areas. Notably, the prediction results for Ae. aegypti in Models B and C for 2023 show a relatively peculiar situation compared with other years. Model B and Model C in 2023 showed a greater expansion compared to other years for Ae. aegypti. However, taking into account the collected species distribution data and comparing it with other years, we tend to believe that this is a slightly biased result. This may be because, during the period from 2019 to 2023, there were fewer distribution points of Ae. aegypti species (only 24 records), and they were distributed across various types of land use and cover change (LUCC). Since both Model B and Model C take LUCC into account, coupled with the small number of species distribution points, the suitable habitat results for this period showed a significant expansion compared to other years. If we disregard the prediction results for Ae. aegypti in 2023, the prediction results for 2014, 2019, and 2030 all show a high degree of consistency in Model C (S1 Fig).
Analyze the differences between Ae. albopictus and Ae. aegypti from the perspective of suitable habitat. Some studies have shown that dengue fever outbreaks in China are mainly caused by Ae. albopictus [29,30], and this can also be seen from the results of the suitable habitat ranges of Ae. albopictus and Ae. aegypti in this study. The suitable habitat area for Ae. albopictus is generally increasing annually. Compared to the outbreak year of dengue fever in China in 2019, the suitable habitat area in 2030 will increase by approximately 17.06%, with a growth of 2.57% in the sum of high-risk and medium-risk suitable habitat areas. This may be due to the strong adaptability of Ae. albopictus to environmental and climate changes. Analysis of the response curves revealed that its minimum temperature requirement is approximately 5°C lower than that of Ae. aegypti, which may result in a broader distribution area for Ae. albopictus. In contrast, the suitable habitats of Ae. aegypti in China are relatively limited.
The analysis of influencing factors revealed that APP, AMT and LUCC are important factors affecting the habitat suitability of Ae. albopictus and Ae. aegypti. These findings are consistent with those of other studies. Mosquito density is closely related to local climate conditions [31], and high temperatures and precipitation each year promote mosquito activity [32]. Climate directly regulates mosquito density through temperature and precipitation [33]. Temperature not only affects the development rate, mortality, and adult lifespan of mosquitoes but also influences virus replication and transmission in vectors [34]. Aedes prefer to lay eggs and breed in water-filled containers, so changes in precipitation can affect mosquito density [35,36]. Recent studies have shown that winter temperatures have the most significant effect on the presence of Ae. albopictus, followed by summer precipitation [37]. Among these variables, excluding the LUCC variable has the most significant effect on model prediction according to the jackknife test. These findings suggest that LUCC might play a crucial role in providing complementary information for species survival and distribution. Even though the prediction performance of this variable alone may not be the highest, its role in the model is likely irreplaceable and may offer key insights that other variables cannot provide. The high-risk habitat areas for Aedes mosquitoes predicted in this study are located mainly in coastal cities and other economically developed regions. Urban/industrial/residential land has a much greater influence on the distribution of Aedes mosquitoes than other land use types do. Studies have shown that socioeconomic factors are related to immature populations of Ae. albopictus [38,39] and that population density and urbanization level can affect mosquito distribution [40].
This study makes attempts based on existing research. Unlike most similar studies that rely on WorldClim (https://www.worldclim.org) bioclimatic variables [41–44], which cover only large time spans (such as 1970–2000, 2021–2040, 2041–2060, 2061–2080, and 2081–2100), critical bioclimatic variables for the period from 2001 to 2020 are lacking. This limits the accurate assessment of recent trend changes, as well as comparisons with the dengue outbreak years in China (2014 and 2019). This study integrates the latest climate data from China and precisely selects relevant climate and environmental-economic variables for 2014, 2019, 2023, and 2030, analyzing the habitat suitability distributions of Ae. albopictus and Ae. aegypti at these time points. The variable data used in this study, for 2014, 2019, 2023, and 2030, not only reviews and analyzes the past outbreak years but also assesses the recent and future situations. By comparing with the outbreak years, it can provide risk alerts and warnings to deal with potential future dengue outbreaks. In addition, regarding the selection of species distribution point data, most studies [45–47] use long-term species records, while we focus on the species distribution points within each five-year period. This may reduce the impact of species distribution changes caused by long-term environmental changes, social changes, intervention measures, and other factors. At the same time, it may better capture the distribution changes of Aedes mosquitoes in recent years.
This study has certain limitations. First, GBIF and literature data may tend to record species distributions in urban areas, so these data may over-represent urban areas and under-represent natural or rural habitats. This uneven data distribution does indeed affect the reliability of the model and the accuracy of predictions. This leads the model to overestimate the distribution of Aedes mosquitoes in urban areas and underestimate the distribution in natural or rural habitats when making predictions. Although we have tried our best to conduct a comprehensive search of GBIF and literature data to achieve as complete a species record as possible, this objective limitation still exists. Second, due to the limited records of Ae. aegypti in certain periods, this may lead to an incomplete understanding of its actual distribution and habitat preferences. This incompleteness of the data may affect the accuracy of the model’s predictions of suitable habitats, and thus affect the reliability of the predicted trends. Third, the calculation of the suitable habitat area is influenced by factors such as coordinate system selection and resolution. The suitable habitat area provided in this study is not very accurate and is mainly used for predicting the trends of contraction and expansion. Future research should be dedicated to collecting more comprehensive and representative Aedes mosquito distribution data, or improving the methodology, in order to enhance the accuracy and reliability of predictions.
This study provides important scientific evidence for the development of vector-borne disease prevention and control strategies. This study not only has reference value for the prevention and control of dengue fever but also offers insights for the control of other vector-borne diseases. On the basis of the results of this study, monitoring Ae. aegypti and Ae. albopictus in regions with high habitat suitability risk levels, as identified by the models, and establishing effective monitoring and risk assessment programs are recommended. This will enable timely tracking of mosquito vector dynamics and further validate the predictive accuracy of the models. Additionally, the increased amount of monitoring data contributes to improving the precision and applicability of the models. Future research can build on this study to further develop seasonal dynamic models and incorporate more relevant variables, especially socioeconomic factors and human activities.
5. Conclusions
This study compared the potential changes in the distributions of suitable habitats for two mosquito species, Ae. albopictus and Ae. Aegypti, across the year 2014, 2019, and 2023. Furthermore, it predicted the suitable habitats for these species in the year 2030. The findings of the study unveil that the primary factors influencing the distributions of Ae. albopictus and Ae. aegypti, which are the main vectors of dengue fever in China, encompass annual precipitation, annual mean temperatures, as well as land use and land cover changes. Notably, the suitable habitat for Ae. albopictus demonstrates a prominent trend of expansion towards the north, anticipated to encompass vast regions in Shandong and Henan Province by 2030. The suitable habitat area for Ae. albopictus is generally increasing annually. When compared to the year 2019, which witnessed an outbreak of dengue fever in China, the suitable habitat area for Ae. albopictus in 2030 will increase by approximately 17.06%. Additionally, there will be a 2.57% growth in the sum of high-risk and medium-risk suitable habitat areas. In contrast, the suitable habitats of Ae. aegypti are primarily concentrated in Guangdong, Hainan and Yunnan Provinces. The findings of this study provide a scientific basis for vector-borne disease prevention and control, particularly in dengue fever control, and offer valuable information for identifying and predicting high-risk areas for potential future outbreaks of dengue fever, thereby promoting the precise implementation of public health policies.
Supporting information
S1 File. Aedes albopictus records from 2010 to 2023.
https://doi.org/10.1371/journal.pntd.0013237.s001
(XLSX)
S2 File. Aedes aegypti records from 2010 to 2023.
https://doi.org/10.1371/journal.pntd.0013237.s002
(XLSX)
S3 File. Generalized variance-inflation factors (GVIFs).
https://doi.org/10.1371/journal.pntd.0013237.s003
(DOCX)
S5 File. The data of Aedes aegypti for calculating GVIF.
https://doi.org/10.1371/journal.pntd.0013237.s005
(CSV)
S6 File. The data of Aedes albopictus for calculating GVIF.
https://doi.org/10.1371/journal.pntd.0013237.s006
(CSV)
S7 File. Aedes aegypti Jackknife test result.
https://doi.org/10.1371/journal.pntd.0013237.s007
(XLSX)
S8 File. Aedes albopictus Jackknife test result.
https://doi.org/10.1371/journal.pntd.0013237.s008
(XLSX)
S9 File. Response curve data of Aedes aegypti.
https://doi.org/10.1371/journal.pntd.0013237.s009
(XLSX)
S10 File. Response curve data of Aedes albopictus.
https://doi.org/10.1371/journal.pntd.0013237.s010
(XLSX)
S1 Fig. Distribution of habitat suitability classifications for Aedes aegypti.
0.00–0.05 indicates uninhabitable habitat, 0.05–0.25 indicates low suitable habitat, 0.25–0.50 indicates medium suitable habitat, and 0.50–1.00 indicates high suitable habitat. The map was obtained from GaryBikini (https://zenodo.org/records/10624971).
https://doi.org/10.1371/journal.pntd.0013237.s012
(TIF)
Acknowledgments
The authors extend their gratitude to all CDC staff members who contributed their expertise and support to this project.
References
- 1. Venkatesan P. Global upsurge in dengue in 2024. The Lancet Infectious Diseases. 2024;24(10):e620.
- 2. Burki T. Risk of dengue spread is high globally, says WHO. Lancet Infect Dis. 2024;24(3):e161–2. pmid:38401565
- 3. eClinicalMedicine. Dengue as a growing global health concern. EClinicalMedicine. 2024;77:102975. pmid:39649133
- 4. Ni H, Cai X, Ren J, Dai T, Zhou J, Lin J, et al. Epidemiological characteristics and transmission dynamics of dengue fever in China. Nat Commun. 2024;15(1):8060. pmid:39277600
- 5. Liu XB, Liu QY. Surveillance and risk warnings for dengue - China, 2016-2019. China Cdc Weekly. 2020;2(24):431–7.
- 6. Peng S, Ding Y, Liu W, Li Z. 1 km monthly temperature and precipitation dataset for China from 1901 to 2017. Earth Syst Sci Data. 2019;11(4):1931–46.
- 7. Peng S, Ding Y, Wen Z, Chen Y, Cao Y, Ren J. Spatiotemporal change and trend analysis of potential evapotranspiration over the Loess Plateau of China during 2011–2100. Agricultural and Forest Meteorology. 2017;233:183–94.
- 8. Ding Y, Peng S. Spatiotemporal change and attribution of potential evapotranspiration over China from 1901 to 2100. Theor Appl Climatol. 2021;145(1–2):79–94.
- 9. Zhang T, Cheng C, Wu X. Mapping the spatial heterogeneity of global land use and land cover from 2020 to 2100 at a 1 km resolution. Sci Data. 2023;10(1):748. pmid:37898602
- 10.
Gary B. China admin division shp. Zenodo; 2024. https://doi.org/10.5281/zenodo.10624971
- 11. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol Model. 2006;190(3–4):231–59.
- 12. Phillips SJ, Dudík M. Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography. 2008;31(2):161–75.
- 13. Elith J, Phillips SJ, Hastie T, Dudík M, Chee YE, Yates CJ. A statistical explanation of MaxEnt for ecologists. Diversity and Distributions. 2011;17(1):43–57.
- 14. Phillips SJ, Anderson RP, Dudík M, Schapire RE, Blair ME. Opening the black box: an open‐source release of Maxent. Ecography. 2017;40(7):887–93.
- 15. Merow C, Smith MJ, Silander JA Jr. A practical guide to MaxEnt for modeling species’ distributions: what it does, and why inputs and settings matter. Ecography. 2013;36(10):1058–69.
- 16. Lu Z-H, Liu C, Chen Y-J, Chen Y-J, Lei X-N, Cai L-J, et al. Gestational Exposure to PM2.5 and Specific Constituents, Meconium Metabolites, and Neonatal Neurobehavioral Development: A Cohort Study. Environ Sci Technol. 2024;58(23):9980–90. pmid:38819024
- 17. Vulu F, Futami K, Sunahara T, Mampuya P, Bobanga TL, Mumba Ngoyi D, et al. Geographic expansion of the introduced Aedes albopictus and other native Aedes species in the Democratic Republic of the Congo. Parasit Vectors. 2024;17(1):35. pmid:38279140
- 18. Luo J, Ma Y, Liu Y, Zhu D, Guo X. Predicting Polygonum capitatum distribution in China across climate scenarios using MaxEnt modeling. Sci Rep. 2024;14(1):20020. pmid:39198562
- 19. Premkumar K, Vaishnav V, Singh S. Geospatial analysis to assess distribution patterns and predictive model for endangered Lilium mackliniae Sealy: to support management decisions. Environ Monit Assess. 2023;195(9):1036. pmid:37572170
- 20. Liu B, Gao X, Ma J, Jiao Z, Xiao J, Hayat MA, et al. Modeling the present and future distribution of arbovirus vectors Aedes aegypti and Aedes albopictus under climate change scenarios in Mainland China. Sci Total Environ. 2019;664:203–14. pmid:30743113
- 21. Li Y, An Q, Sun Z, Gao X, Wang H. Distribution areas and monthly dynamic distribution changes of three Aedes species in China: Aedes aegypti, Aedes albopictus and Aedes vexans. Parasit Vectors. 2023;16(1):297. pmid:37633953
- 22. Yue Y, Liu X, Xu M, Ren D, Liu Q. Epidemiological dynamics of dengue fever in mainland China, 2014-2018. Int J Infect Dis. 2019;86:82–93.
- 23. Sang S, Yue Y, Wang Y, Zhang X. The epidemiology and evolutionary dynamics of massive dengue outbreak in China, 2019. Front Microbiol. 2023;14:1156176. pmid:37138627
- 24. Li Z, Huang X, Tian T, Li AQ, Du S, He G, et al. Epidemiological characteristics of reported dengue fever cases in China, 2019-2023. China Tropical Medicine. 2024;24(8):925–30.
- 25. Liu H, Liu L, Cheng P, Yang L, Chen J, Lu Y, et al. Bionomics and insecticide resistance of Aedes albopictus in Shandong, a high latitude and high-risk dengue transmission area in China. Parasit Vectors. 2020;13(1):11. pmid:31918753
- 26. Lun X, Wang Y, Zhao C, Wu H, Zhu C, Ma D, et al. Epidemiological characteristics and temporal-spatial analysis of overseas imported dengue fever cases in outbreak provinces of China, 2005-2019. Infect Dis Poverty. 2022;11(1):12.
- 27. Yue Y, Liu Q, Liu X, Zhao N, Yin W. Dengue Fever in Mainland China, 2005-2020: A Descriptive Analysis of Dengue Cases and Aedes Data. Int J Environ Res Public Health. 2022;19(7):3910.
- 28. Zhao J, Zhou Y, Chen S, Qiu J, Yang X. Incidence trends and spatiotemporal distribution characteristics of dengue fever in China. China Tropical Medicine. 2024;24(8):920–4.
- 29. Sun JM, Lu L, Wu HX, Yang J, Xu L, Sang SW. Epidemiological trends of dengue in mainland China, 2005-2015. International Journal of Infectious Diseases. 2017;57:86–91.
- 30. Luo L, Jiang L-Y, Xiao X-C, Di B, Jing Q-L, Wang S-Y, et al. The dengue preface to endemic in mainland China: the historical largest outbreak by Aedes albopictus in Guangzhou, 2014. Infect Dis Poverty. 2017;6(1):148. pmid:28934991
- 31. Morin CW, Comrie AC, Ernst K. Climate and dengue transmission: evidence and implications. Environ Health Persp. 2013;121(11–12):1264–72.
- 32. Wang T, Fan Z-W, Ji Y, Chen J-J, Zhao G-P, Zhang W-H, et al. Mapping the Distributions of Mosquitoes and Mosquito-Borne Arboviruses in China. Viruses. 2022;14(4):691. pmid:35458421
- 33. Abdalgader T, Banerjee M, Zhang L. Spatially weak syncronization of spreading pattern between Aedes albopictus and dengue fever. Ecol Model. 2022;473:110123.
- 34. Reinhold JM, Lazzari CR, Lahondère C. Effects of the Environmental Temperature on Aedes aegypti and Aedes albopictus Mosquitoes: A Review. Insects. 2018;9(4):158. pmid:30404142
- 35. Lowe R, Gasparrini A, Van Meerbeeck CJ, Lippi CA, Mahon R, Trotman AR, et al. Nonlinear and delayed impacts of climate on dengue risk in Barbados: A modelling study. PLoS Med. 2018;15(7):e1002613. pmid:30016319
- 36. Roiz D, Rosà R, Arnoldi D, Rizzoli A. Effects of temperature and rainfall on the activity and dynamics of host-seeking Aedes albopictus females in northern Italy. Vector Borne Zoonotic Dis. 2010;10(8):811–6. pmid:20059318
- 37. Liu H, Huang X, Guo X, Cheng P, Wang H, Liu L, et al. Climate change and Aedes albopictus risks in China: current impact and future projection. Infect Dis Poverty. 2023;12(1):26. pmid:36964611
- 38. Dowling Z, Ladeau SL, Armbruster P, Biehler D, Leisnham PT. Socioeconomic status affects mosquito (Diptera: Culicidae) larval habitat type availability and infestation level. J Med Entomol. 2013;50(4):764–72. pmid:23926774
- 39. LaDeau SL, Leisnham PT, Biehler D, Bodner D. Higher mosquito production in low-income neighborhoods of Baltimore and Washington, DC: understanding ecological drivers and mosquito-borne disease risk in temperate cities. Int J Environ Res Public Health. 2013;10(4):1505–26. pmid:23583963
- 40. Zahouli JBZ, Utzinger J, Adja MA, Müller P, Malone D, Tano Y, et al. Oviposition ecology and species composition of Aedes spp. and Aedes aegypti dynamics in variously urbanized settings in arbovirus foci in southeastern Côte d’Ivoire. Parasit Vectors. 2016;9(1):523. pmid:27682270
- 41. Hussain SSA, Dhiman RC. Distribution Expansion of Dengue Vectors and Climate Change in India. Geohealth. 2022;6(6):e2021GH000477. pmid:35769847
- 42. Portilla Cabrera CV, Selvaraj JJ. Geographic shifts in the bioclimatic suitability for Aedes aegypti under climate change scenarios in Colombia. Heliyon. 2020;6(1):e03101. pmid:31909268
- 43. Cardoso-Leite R, Vilarinho AC, Novaes MC, Tonetto AF, Vilardi GC, Guillermo-Ferreira R. Recent and future environmental suitability to dengue fever in Brazil using species distribution model. Trans R Soc Trop Med Hyg. 2014;108(2):99–104. pmid:24463584
- 44. Liu Q, Zhang H-D, Xing D, Jia N, Du Y-T, Xie J-W, et al. The predicted potential distribution of Aedes albopictus in China under the shared socioeconomic pathway (SSP)1-2.6. Acta Trop. 2023;248:107001. pmid:37634685
- 45. Varamballi P, Babu N N, Mudgal PP, Shetty U, Jayaram A, Karunakaran K, et al. Spatial heterogeneity in the potential distribution of Aedes mosquitoes in India under current and future climatic scenarios. Acta Trop. 2024;260:107403. pmid:39278522
- 46. Johnson TL, Haque U, Monaghan AJ, Eisen L, Hahn MB, Hayden MH, et al. Modeling the Environmental Suitability for Aedes (Stegomyia) aegypti and Aedes (Stegomyia) albopictus (Diptera: Culicidae) in the Contiguous United States. J Med Entomol. 2017;54(6):1605–14. pmid:29029153
- 47. Abdelkrim O, Samia B, Said Z, Souad L. Modeling and mapping the habitat suitability and the potential distribution of Arboviruses vectors in Morocco. Parasite. 2021;28:37. pmid:33861197