https://orcid.org/0000-0002-2025-7837
Figures
Abstract
Background
The invasive mosquito species Anopheles stephensi plays a critical role in malaria transmission, particularly in urban environments. Its ability to thrive in such settings has raised public health concerns, especially as it expands its geographical range. The resurgence of malaria in Iran underscores the challenges posed by this vector, which is further complicated by factors such as climate change and the movement of populations. Understanding the ecological niches of An. stephensi is essential for developing targeted malaria control strategies. This study aims to assess the current and projected distribution of An. stephensi in Hormozgan Province, Iran, under varying climatic conditions.
Method
The study was conducted in Hormozgan Province, Iran, characterized by a hot, arid climate. A database of 96 occurrence points for An. stephensi was compiled through literature searches, which were refined to 72 points to ensure data quality. Environmental and bioclimatic data were sourced from the WorldClim v2.1 database, with a focus on various Shared Socioeconomic Pathways (SSPs). The MaxEnt modeling technique was employed to assess the impact of climate change on the species’ distribution, with model performance evaluated using the two metrics, Area Under the Curve (AUC) and True Skill Statistic (TSS). The analysis aimed to map shifts in suitable habitats under different climate scenarios.
Results
The MaxEnt model predicts a significant decline in environmental suitability for An. stephensi under future climate scenarios, particularly in western and central Hormozgan Province. Habitat loss is prevalent, with stable areas primarily located in Bashagard County. Factors such as altitude and precipitation patterns significantly influence species distribution, with altitude showing the highest impact. The model’s performance, indicated by an average AUC of 0.765 and a TSS of 0.519, demonstrates moderate predictive accuracy for identifying suitable habitats.
Discussion
Despite advancements in malaria control, An. stephensi remains a significant threat in Iran, particularly due to its invasive nature and adaptability to climate change. The study indicates a projected decline in suitable habitats, especially in coastal areas, highlighting the need for adaptive vector control strategies. Bashagard County may serve as a stable refuge, warranting further investigation. The findings emphasize the importance of continuous monitoring and integrating climate projections into public health interventions to effectively combat malaria transmission. Overall, ongoing research is crucial to refining understanding and enhancing malaria control efforts in response to shifting environmental conditions.
Author summary
The mosquito An. stephensi is an important vector of malaria. Unlike many other malaria vectors, it can survive and reproduce in cities, which makes it a growing concern for public health. In recent years, this mosquito has expanded its presence in Iran, raising fears that malaria may spread again, especially as climate change creates new challenges for disease control. In this study, we explored how current and future climate conditions might affect where An. stephensi can live in Hormozgan Province, southern Iran. We used climate data and records of where the mosquito has been found to create maps showing areas that are most suitable for its survival. Our results suggest that many areas, especially in central and western Hormozgan, may become less suitable for the mosquito in the future. However, some regions, particularly Bashagard County, are likely to remain favorable habitats. These findings can guide health authorities in designing more effective and adaptive strategies to reduce malaria transmission.
Citation: Bozorg-Omid F, Abbasi M, Yaghoobi-Ershadi MR, Hanafi-Bojd AA (2026) Suitable ecological niches of invasive malaria vector under present and projected climatic conditions in South of Iran. PLoS Negl Trop Dis 20(3): e0014054. https://doi.org/10.1371/journal.pntd.0014054
Editor: Kevin Shyong-Wei Tan, National University of Singapore, SINGAPORE
Received: June 27, 2025; Accepted: February 18, 2026; Published: March 16, 2026
Copyright: © 2026 Bozorg-Omid 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: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by the Research Deputy of Tehran University of Medical Sciences (Project No. 36251) to A.A.H‑B. The funder 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.
Introduction
In 2023, the World Health Organization (WHO) reported 263 million new cases of malaria globally [1]. Iran effectively executed malaria control measures, resulting in 2021 having no indigenous malaria cases for 4 consecutive years [2]. Nevertheless, a resurgence of the disease was observed in 2022, with confirmed malaria cases rising sharply to around 1,432 locally acquired infections that year (total cases: 5,677). Furthermore, in 2023, the caseload doubled compared to the previous year (total cases: 9,868) [3,4]. Public health officials attribute the outbreaks to several factors, including the presence of foreign nationals, an increase in malaria cases in neighboring Pakistan, poor detection of new cases, heavy summer rains, a focus on the COVID-19 pandemic, and the emergence of invasive Aedes species. This emergence has diverted the healthcare system’s attention and resources away from malaria control efforts, leading to increased vulnerability and resulting in the local transmission of the newly emerged dengue fever in Iran [5]. The regions most impacted in Iran are the south and southeast [3,6].
Anopheles stephensi, a mosquito species originally from the Indian subcontinent and the Persian Gulf, is notable for its role in transmitting malaria parasites, specifically Plasmodium falciparum and P. vivax. Unlike other malaria vectors, it thrives in urban environments and has significantly expanded its range westward since its first detection in Africa in Djibouti in 2012. This species has since been reported in several countries across the Horn of Africa, including Ethiopia, Sudan, Somalia, and Eritrea, as well as in Yemen, Kenya, Ghana, and Nigeria, highlighting its growing geographical distribution and public health implications [7–13]. Recent habitat suitability studies provide further insights into this expansion: Sinka et al. (2020) used spatial models to predict potential urban areas across Africa where An. stephensi could establish, estimating that over 126 million people may be at risk if the spread continues unchecked [14]. Similarly, Samake et al. (2025) analyzed the distribution and genetic diversity of An. stephensi in Kenya, revealing multiple introductions, co-occurrence with native Anopheles species, and environmental factors such as low precipitation and minimal seasonal temperature variation influencing its spread, underscoring the need for ongoing surveillance and predictive modeling [15].
The influence of global climate change on natural ecosystems has been extensively documented, highlighting its profound effects on human welfare and biodiversity [16–18]. The latest valuation report from the United Nations Intergovernmental Panel on Climate Change (IPCC AR5) highlights a significant increase in global land and ocean temperatures, with a rise of 0.89 °C (0.69-1.08 °C) from 1901 to 2012 [19]. This temperature increase is anticipated to have profound implications, including potential changes in suitable ranges for species and the exacerbation of the negative impacts of invasive species in various regions [20,21]. Climate change also affects the El Niño cycle, impacting the transmission of diseases like malaria, dengue, and Rift Valley fever [22]. Moreover, the complex connection between the geographical distribution of malaria, local vectors, and climatic and geographic factors is becoming increasingly clear, directly impacting the local epidemiology of malaria [23,24]. This emphasizes the need to integrate climate change considerations into malaria control strategies [23,25,26].
Efforts to control malaria through surveillance programs, including early diagnosis, treatment, and prevention measures (such as insecticide-treated nets), have shown significant progress in reducing malaria transmission rates [27,28]. The Global Malaria Action Plan aims to eliminate malaria in several countries. However, the impacts of climate change may pose considerable challenges to these efforts, potentially increasing the epidemic potential of malaria in susceptible regions [29,30]. Therefore, maintaining strong surveillance and preparedness is crucial, especially in developing countries with limited resources. As global temperatures continue to rise, the range and behavior of malaria vectors, such as An. stephensi are likely to be significantly affected, potentially leading to shifts in their distribution and posing new challenges for malaria control and elimination efforts [31]. Additionally, a study in the Ecuadorian Amazon reveals that changes in land use and land cover, alongside climate factors, significantly influence malaria risk. This indicates that both land use and climate factors must be considered to fully understand malaria dynamics [32].
The use of bioclimatic data and Geographic Information System (GIS) analysis has become fundamental in predicting the potential habitats of insect vectors across time. Climate change scenarios play a crucial role in understanding how future environmental conditions may alter vector distributions [33,34]. Unlike the earlier Representative Concentration Pathways (RCPs), which focused solely on greenhouse gas (GHG) emission trajectories, Shared Socioeconomic Pathways (SSPs) integrate both climate and societal factors, providing a more comprehensive framework for analyzing future environmental and health-related challenges [35,36]. In the context of vector-borne diseases, including malaria, SSPs are crucial because they allow researchers to model not only how climate change may affect the distribution of disease vectors, but also how socioeconomic conditions may influence transmission, control measures, and adaptation strategies [37].
The SSP framework, when combined with Species Distribution Models (SDMs), provides a valuable tool for predicting the future distribution of malaria vectors under different climate and socioeconomic conditions. SDMs, also known as Ecological Niche Models (ENMs), are widely used to predict the geographic range of species under different environmental conditions [38,39]. These models utilize species occurrence records and environmental variables to identify areas with a high probability of species presence [40]. SDMs have proven particularly useful in assessing the potential spread of invasive species and understanding how climate change may alter habitat suitability [41]. Over the past decade, more than 35 modeling approaches have been developed for generating SDMs, each with its advantages and limitations [39]. Some of the most commonly used models include Maximum Entropy (MaxEnt), Generalized Linear Model (GLM), Random Forest [42], and Generalized Boosting Model (GBM). Among these, MaxEnt has been widely recognized for its effectiveness in modeling species distributions using presence-only data. Comparative studies suggest that MaxEnt often performs as well as, or better than, other methods in predicting suitable habitats for species, particularly in cases where occurrence data are limited [38,39,43,44].
High-resolution, region-specific modeling enhances the reliability of predictions, allowing for better-targeted vector control strategies and public health interventions [45]. By incorporating detailed local environmental and climatic data, these models provide valuable insights into future vector distribution trends, aiding in proactive disease management and mitigation efforts [45]. This approach facilitates evidence-based decision-making for malaria prevention and control strategies, ensuring that interventions are targeted toward the most vulnerable regions [45]. Quantitative modeling techniques rooted in ecological niche theory are primarily utilized to understand the relationship between species and their environment [46]. Ecological niche modeling forecasts a species’ range by leveraging its occurrence data and specific environmental parameters to define its niche within an ecosystem, which is then projected onto geographical landscapes [47]. When focused on organisms, it identifies suitable habitats for each species in areas lacking mosquito population data, enabling evidence-based assessments of malaria transmission risk in regions not currently covered by interventions [48,49]. Forecasting the presence of a vector in particular areas poses a difficulty for numerous disease control initiatives aiming to enhance the planning and execution of control strategies and adaptation measures [50]. The maximum entropy model (MaxEnt v3.4.1) [51] is a species distribution prediction model grounded in the maximum entropy principle that exhibits notable predictive performance compared to various other models for predicting species distribution, particularly in scenarios where species distribution models have been suggested to be accurate with small samples [52]. It often outperforms similar prediction models in such cases [53,54].
The objective of this research was to gather all records of An. stephensi occurrence in Hormozgan Province, South of Iran, over the past five decades. Additionally, the study aimed to forecast the present and future distribution as well as the environmental suitability of these species for the years 2030s, 2050s, and 2070s. Understanding the spatial distribution pattern of these vectors is essential for evaluating disease transmission risk in various regions and for designing effective vector control programs.
Materials and methods
Ethics statement
The protocol for this project has been approved by the ethics committee at the Research Deputy of Tehran University of Medical Sciences, Tehran, Iran, with Ethics Code: IR.TUMS.SPH.REC.1397.297.
Study area
Hormozgan Province, situated in the southern region of Iran between 25°23’ and 28°57’ N, and 52°41’ and 59°15’ E, covers an area of 70,000 km2. It is bordered by the coasts of the Oman Sea and the Persian Gulf, spanning approximately 900 Km in the south (Fig 1). The climate of the province falls within the subtropical category, characterized by warm air as the most prominent feature. Known as one of the hot and arid regions of Iran, it is heavily influenced by semi-desert and desert climates. Summers along the coastal strip are extremely hot and humid, with temperatures occasionally exceeding 52℃. The average annual temperature stands at 27℃. The province experiences a protracted hot season lasting nine months, starting from early March and reaching its peak in July and August. This is followed by a brief three-month cool season, beginning in early December and influenced by cool western air masses. Topographically, the elevation of Hormozgan Province varies considerably, ranging from 5 m below sea level in coastal areas to 3,116 m in the northern highlands. Overall, the climate of Hormozgan Province closely resembles that of desert areas, with extremely low levels of atmospheric precipitation, resulting in a notable lack of vegetation [55].
Map created using ArcGIS Desktop (version 10.5; Environmental Systems Research Institute, Redlands, CA, USA) (www.esri.com).
Occurrence data
To compile a presence location database of Anopheles stephensi, the terms “An. stephensi” and “malaria” were used in a literature search of various online scientific sources (Google Scholar, PubMed, and Web of Science, SID, Irandoc, and Magiran) from 1970 to 2022 (S1 Table). Geographic coordinates were extracted. Following this web search, a total of 96 presence points of An. stephensi species were collected. The dataset included both larval and adult detections. To ensure data quality, duplicate records, defined as entries with the same geographic coordinates reported, were removed [56,57]. Additionally, to minimize spatial autocorrelation and prevent pseudo-replication [56], the data were processed using the spatial rarefaction tool in SDMs Toolbox v2.5 (ArcGIS 10.5) and then imported into ArcMap 10.5 for visualization. Finally, the number of points of presence was reduced to 72 (Fig 1) (S1 File).
Environmental and bioclimatic data
Both topographic (altitude) and bioclimatic data (current (1970–2000) and future (2020–2040, 2041–2060, and 2061–2080) were obtained from the WorldClim v2.1 database (www.worldclim.org) with a spatial resolution of approximately 1 km² [58]. We selected three Shared Socioeconomic Pathways (SSP 1–2.6, SSP2-4.5, and SSP 5–8.5) from the sixth version of the Model for Interdisciplinary Research on Climate (MIROC6), representing scenarios with low, moderate, and high greenhouse gas emission concentrations, respectively. These scenarios were chosen to capture a range of plausible future climate conditions, allowing assessment of how different levels of climate change may influence the distribution and habitat suitability of An. stephensi in Hormozgan Province. All bioclimatic and topographic layers in raster format were imported into ArcGIS 10.5, and the data were clipped using the Hormozgan Province boundary shapefile for regional analysis. To ensure the accuracy and reliability of the analysis, we accounted for multicollinearity among environmental variables. Highly correlated climate variables can introduce biases in statistical models, leading to misleading results. Therefore, collinearity among climatic variables was assessed using the Variance Inflation Factor (VIF) method in the usdm package in R software [59,60]. Variables with a VIF greater than 10 were removed to minimize redundancy and enhance model performance [59]. A final set of 10 uncorrelated variables was selected from an initial pool of 20 layers (19 bioclimatic variables and one altitude layer) for modeling (Table 1).
Species distribution modeling
The MaxEnt modeling technique was employed to assess the effects of climate change on the distribution shifts of An. stephensi in Hormozgan Province [61]. MaxEnt v3.4.3 software was utilized for the analyses. Customized settings were applied for the modeling process, including the selection of 10,000 background points. To evaluate model performance, 80% of the occurrence data were designated for model training (calibration), while the remaining 20% were reserved for model testing (evaluation) with 10 repetitions to ensure robustness of the results. To assess the relative importance of environmental variables in the MaxEnt model, we applied the Jackknife test. This method evaluates each variable individually and in combination with others to determine its contribution to model performance [62].
To illustrate changes in suitable habitat, we qualitatively and quantitatively summarized the distributional shifts in species ranges based on SSP scenarios and time periods. Specifically, “stable”, “gain”, and “loss” areas in species distribution were mapped under future climate change scenarios. The analysis was conducted using R-4.4.2v software, where the continuous habitat suitability outputs were converted into binary presence–absence maps using the Maximum training sensitivity plus specificity (MaxSSS) threshold [63]. This thresholding rule is widely recommended because it provides an optimal trade-off between omission errors (false negatives) and commission errors (false positives), thus improving the ecological realism of the predictions [64]. The gain/loss/stable areas were derived by comparing binary maps with the terra package [65]. This approach allowed for a clear representation of the changes in suitable habitats under different climate scenarios. In other words, it shows how habitat suitability is expected to shift under different climate scenarios, whether the environmental suitability has increased, decreased, or remained stable compared to the current conditions. Maps were visualized using the ggplot2 package [66].
Model evaluation
Various metrics have been proposed to evaluate the performance of habitat suitability models [44]. However, there is no consensus on which metrics are most appropriate for assessing model performance. Therefore, we employed two widely used metrics to evaluate the generated habitat suitability model: TSS [67,68] and Area Under the Receiver Operating Characteristic Curve (AUC) [69]. AUC values were obtained directly from MaxEnt outputs, while TSS values were calculated using the presenceAbsence package in R-4.4.2v software [70]. According to previous studies, TSS ranges from −1 to +1, where +1 represents perfect model performance and a value of 0 indicates predictions no better than random. AUC values range from 0 to 1, with 0.5 indicating random performance and values approaching 1.0 reflecting increasingly better model performance [71,72]. Maps were visualized using the ggplot2 package [66].
Result
Environmental suitability and area gain and loss patterns
The MaxEnt model projected that environmental suitability for An. stephensi declines significantly under future climate change scenarios. This decline becomes progressively more severe over time. The findings suggest that by the end of the century, a large portion of currently suitable areas will no longer provide optimal environmental conditions for the species. The most notable changes in environmental suitability occur in the western and central parts of the province, where Bandar Abbas, Bandar Khamir, and the islands of Qeshm, Kish, and Abu Musa exhibit a sharp decline in suitability. These areas are expected to become increasingly unsuitable due to changing climatic conditions, as reflected in the progressive reduction of suitable habitat patches over time. Conversely, suitable habitats in Bashagard County would be stable areas. Notably, under SSP5-8.5-2070s, Bashagard is the only county that will remain suitable (Fig 2).
Map created using ArcGIS Desktop (version 10.5; Environmental Systems Research Institute, Redlands, CA, USA) (www.esri.com).
The gain and loss maps provide further insights into the spatial redistribution of suitable habitats. The overall trend points to a declining habitat suitability in most of the province, with only a few areas showing potential for habitat gain. Red-colored areas indicate habitat loss, meaning that these regions are currently suitable but are expected to become unsuitable in the future. This is particularly evident in the western and central regions. On the other hand, the gain areas are observed primarily in the eastern and southern parts of the province, particularly in Bashagard. These areas exhibit an expansion of suitable habitats over time, particularly under the SSP5-8.5 scenario, which projects the highest levels of habitat gain. Conversely, the SSP1-2.6 scenario consistently predicts the least habitat gain across different time periods, suggesting that lower-emission scenarios may result in fewer spatial shifts in habitat suitability (Fig 3).
The Map was generated using R-4.4.2v, with raster processing performed using the terra package and visualization using the ggplot2 package.
The bar chart illustrates the proportions of Stable, Gain, and Loss across different SSP scenarios. The dominant category in all SSPs is Loss, consistently occupying the largest portion, ranging between approximately 55% and 85%. The highest percentage of loss is seen in the 2070s and 2050s, especially in the SSP5-8.5 and SSP2-4.5 scenarios, each having nearly 75–85% of the total. The Stable category varies significantly, with the 2030 period, especially SSP1-2.6, showing the highest stability at around 30%. The Gain category remains relatively small in all cases, with percentages generally ranging between 5% and 15%. The highest gain percentage appears in the 2030s and the SSP5-8.5 scenario, where it reaches around 15%. On the other hand, the lowest stability is observed in the 2070s and the SSP5-8.5 scenario, with only about 5% remaining stable. Overall, the chart emphasizes that most scenarios experience overwhelming loss, with only minor portions remaining stable or gaining (Fig 4).
The Y-axis shows the climate scenarios (SSP1-2.6, SSP2-4.5, SSP5-8.5) in different periods, and the X-axis shows the percentage of habitat categorized as gain, loss, or stable. Colors indicate habitat change: green = stable, yellow = gain, and red = loss. The Map was generated using R-4.4.2v, with visualization using the ggplot2 package.
An analysis of changing suitable areas, measured in square kilometers, across various SSP scenarios reveals significant trends in environmental suitability, stability, gain, and loss areas over time. The projected environmental suitability decreases, starting at approximately 9,200 km2 in the 2030s and dropping to 2,490 km2 in the 2070s, SSP5-8.5 scenario. Stability consistently declines from 6,300 km2 in the 2030s, SSP1-2.6 scenario, to 1,052 km2 from the 2070 period (SSP5-8.5 scenario). Although gain areas initially peak at approximately 3,500 km2 in SSP5-8.5 from the 2030s, they diminish significantly in subsequent scenarios, reaching a low of 900 km2 in the 2070s, SSP1-2.6 scenario. Conversely, loss areas increase over time, beginning at approximately 10,635 km2 in the SSP1-2.6 scenario from the 2030 period and rising to 18,704 km2 in the 2070s, SSP2-4.5 scenario. This indicates a pattern of decreasing suitability, stability, and gain, with an increase in environmental loss, especially in more extreme scenarios (Fig 5).
Projected environmental suitability: suitable area (km²) under current conditions and three climate scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5) for 2030, 2050, and 2070; Gain, Loss, and Stable: areas of habitat expected to increase, decrease, and areas expected to remain stable, respectively, under the same scenarios, in comparison with environmental suitability. The Map was generated using R-4.4.2v, with visualization using the ggplot2 package.
Analysis of variable contributions
Based on the permutation importance values in Fig 6, ALT (altitude) has the highest impact on the model with 22.9%, followed closely by bio14 (precipitation of the driest month) at 19.8% and bio18 (precipitation of the warmest quarter) at 18.5%. bio3 (isothermality) also plays a significant role with 13.5%, while bio19 (precipitation of the coldest quarter) and bio15 (precipitation seasonality) have moderate importance at 9.9% and 9.6%, respectively. Other variables, such as bio10 (mean temperature of the warmest quarter) and bio9 (mean temperature of the driest quarter), have lower permutation importance at 3.8% and 1.1%, while bio8 (mean temperature of the wettest quarter) has the least impact with only 0.1% (Fig 6). These results indicate that certain environment-related variables strongly influence the model’s predictions. Based on the jackknife results using test gain, ALT (altitude) stands out as the single most informative variable when used alone, indicating it provides substantial predictive power by itself. Meanwhile, bio3 (isothermality) leads to the greatest decrease in gain when omitted, suggesting it offers unique information not captured by the other variables (Fig 7).
Bio3 = Isothermality ((Mean Diurnal Range/ Temperature Annual Range) ×100), Bio4 = Temperature seasonality (SD × 100), Bio8 = Mean Temperature of Wettest Quarter, Bio9 = Mean Temperature of Driest Quarter, Bio10 = Mean Temperature of Warmest Quarter, Bio14 = Precipitation of Driest Month, Bio15 = Precipitation seasonality (CV), Bio18 = Precipitation of Warmest Quarter, Bio19 = Precipitation of Coldest Quarter, Alt = Altitude (m).
The Y-axis shows environmental variables, and the X-axis represents model gain during testing. Colors indicate the contribution of each variable: green = variable excluded from the model, blue = model with only that variable, red = model with all variables included. Variables with taller blue bars demonstrate strong predictive power when used alone, while shorter green bars indicate that excluding the variable substantially reduces the model’s predictive accuracy, highlighting its critical importance.
Response curves
Curves in Fig 8 show how each environmental variable is associated with the predicted probability of An. stephensi presence. The response curve analysis indicates that higher temperatures and precipitation levels are predicted to be associated with higher probabilities of the presence of An. stephensi. Among these factors, altitude shows an inverse relationship with the predicted probability of presence, with the highest predicted probability recorded at 200 meters. Isothermality (Bio3) varies between 32% and 44%, with the predicted probability peaking at 36% before declining. The mean temperature of the warmest quarter (Bio10) is predicted to influence the probability of presence, fluctuating between 20°C and 34°C, with the highest predicted probability at 33°C. Similarly, the mean temperature of the wettest quarter (Bio8) follows the same positive association, ranging from 5°C to 30°C, with a peak predicted probability at 17°C. Precipitation of the driest month (Bio14) falls within -1 mm to 9 mm, where the predicted probability of presence increases beyond 3 mm, peaking at 9 mm. Precipitation of the coldest quarter (Bio19) ranges between 80 mm and 180 mm, with the highest predicted probability observed at 140 mm. Conversely, precipitation of the warmest quarter (Bio18) is predicted to be associated with lower probabilities of presence, varying from 0 mm to 700 mm. Precipitation seasonality (Bio15) fluctuates between 55 and 120, with predicted probabilities showing two peaks at 60 and 115. Temperature seasonality (Bio4) ranges from 500 to 800, with the highest predicted probability at 550 (Fig 8).
The curves (blue) show the mean response of the 10 replicate MaxEnt runs [8] and the mean + /- one standard deviation. The x-axis shows environmental variables (see Table 1), and the y-axis denotes the predicted probability of species presence on the cloglog scale, where cloglog transforms the model output into an occurrence probability between 0 and 1.
Model performance
The model’s predictive performance was evaluated using the TSS and the AUC. Across ten replicate runs, the average TSS was 0.519, while the average AUC was 0.765, indicating consistent model performance (Fig 9).
The Map was generated using R-4.4.2v, with visualization using the ggplot2 package.
Discussion
Despite significant progress in malaria control measures over the past decades, malaria remains a persistent problem in the southern provinces of Iran [4]. Monitoring vectors is a crucial aspect of planning, executing, and assessing vector control measures. Gaining a comprehensive understanding of the distribution of malaria vector populations across a country is an invaluable asset in combating malaria transmission [73]. With global temperatures rising, understanding and mitigating the impact of climate change on malaria vectors is crucial to ensuring effective public health measures and preserving natural ecosystems. Among malaria vectors, An. stephensi stands out due to its invasive nature and its ability to thrive in rapidly changing environments [10,11]. Its growing presence in new regions underscores the urgent need for continuous entomological surveillance and proactive vector control strategies. Given the dynamic nature of malaria vectors and the influence of climate change on their distribution, predictive modeling can offer valuable insights for designing effective control strategies. In this study, we employed species distribution modeling using MaxEnt to assess the current and future distribution of An. stephensi in Hormozgan Province, Iran, providing valuable insights into how climate change may reshape its habitat suitability.
Our projection indicates that habitat loss is the dominant trend for An. stephensi across all SSP scenarios, with a particularly sharp decline observed under the SSP5-8.5 scenario in the 2070s. Conversely, in colder regions, climate change is expected to increase the suitability of habitats for this species. Studies have predicted that a warming temperature is enabling An. stephensi to thrive in once inhospitable areas, thereby expanding their geographic range [74,75]. With climate change and rising temperatures, the thermal conditions suitable for this species are likely to shrink further, reducing habitat suitability in Hormozgan. In contrast, in cooler regions of the world, warming may expand the potential range of this vector, highlighting the context-dependent effects of climate change on An. stephensi distribution [74,75]. Therefore, it is crucial to consider regional variations when assessing the future risks associated with climate change. The projected reduction of suitable habitats in Bandar Abbas, Bandar Khamir, and the islands of Qeshm, Kish, and Abu Musa suggests that coastal and low-altitude areas (Fig 3) may become increasingly unsuitable. Conversely, Bashagard County, despite being lowland, appears to retain more stable environmental conditions and may serve as a potential refugium for An. stephensi. This resilience suggests that specific environmental conditions in Bashagard may buffer against climate change effects, warranting further investigation into its microclimatic and ecological characteristics. The species exhibits notable ecological plasticity, being able to exploit a variety of domestic and peri-domestic habitats, including storage tanks, overhead tanks, and discarded containers [76,77]. However, the reliance on human-made water containers also implies that anthropogenic factors, such as urbanization, water management practices, and waste disposal, will strongly influence local population dynamics. Consequently, even in regions where climatic conditions become less favorable, the species may maintain viable populations if suitable artificial habitats are available. Nonetheless, the species remains sensitive to environmental conditions, particularly temperature, which can influence its survival and reproduction. Previous studies in Iran have shown that An. stephensi is sensitive to temperature, surviving within a range of approximately 10.3 °C to 38 °C and requiring around 187 cumulative degree-days for development [78]. These physiological thresholds may help explain why some areas remain suitable for the species despite broader climatic changes. While climate change may create some new suitable habitats, especially under high-emission scenarios like SSP5-8.5, the MaxEnt model estimated the loss area as more than the gain area for An. stephensi in Hormozgan province. So that the total area of suitable habitats drops significantly over time from 6,300 km² in the 2030s to just 1,052 km² in the 2070s under SSP5-8.5. Furthermore, as the species may shift to higher latitudes [79] or neighboring provinces, further research is essential, especially in southern provinces, to assess the potential impacts of new SSP scenarios on habitat availability and vector control strategies.
Moreover, projections show that suitable habitats for most malaria vector species in Iran are expected to decline in the future [80–82]. However, global trends present a more complex picture. While studies predict an expansion of malaria vectors in China and Central America in the coming decades [83,84], the availability of suitable habitats for these vectors is expected to decline in regions such as Africa and South America [85]. Previous studies have generally suggested that habitat suitability for An. stephensi remains relatively stable under climate change in Iran [80,81,86]. In contrast, our findings indicate a significant decline in suitability over time. This difference highlights the importance of conducting localized studies at smaller spatial scales because national-scale assessments might overlook region-specific variations and more detailed ecological characteristics. Additionally, another potential reason for these contrasting results is the use of different climate scenarios. All of the previous studies in Iran have relied on RCP scenarios [80,81,86], whereas our study employs SSP scenarios. Unlike RCPs, SSPs incorporate both climate projections and socio-economic factors, which may provide a more comprehensive assessment of future environmental suitability. In other words, SSP scenarios have emerged as advanced climate modeling frameworks that account not only for greenhouse gas concentration levels but also for the broader socio-economic drivers of climate change, including population dynamics, economic development, land use patterns, and energy transitions [35]. Moreover, the referenced studies did not account for potential correlations between climatic variables and incorporated all 19 layers in their analysis. Additionally, discrepancies in findings may stem from differences in the selected global climate models (GCMs), as each model produces distinct regional climate change projections. In our study, we utilized the SSP-MIROC6 model, whereas previous studies were based on the RCP-BCC-CEM2-MR model, which may have contributed to the observed variations in habitat suitability predictions. These methodological differences may partly explain the contrasting results, underscoring the need for continued research at multiple spatial scales to ensure accurate assessments of vector distribution dynamics [87].
As highlighted by Ryan et al. (2023), the anticipated spread of this vector necessitates proactive monitoring and targeted strategies to mitigate transmission risks in areas projected to become suitable under future climate scenarios. This underscores the importance of ongoing entomological surveillance, as some studies suggest that climate change may not uniformly affect vector distribution, indicating a complex relationship of ecological factors that permit further investigation [86–88]. Understanding the shift in malaria patterns due to changes in habitat suitability requires a comprehensive approach that accounts for multiple epidemiological factors. While habitat change may influence vector distribution, it is crucial to recognize that malaria transmission in this province is not solely driven by a single vector species. Previous studies have shown that Anopheles fluviatilis s.l., An. culicifacies s.l., An. dthali, and An. superpictus s.l. play the main role in malaria transmission in Hormozgan province [89]. Indeed, the environmental suitability of other vectors must be thoroughly assessed to gain a complete understanding of transmission dynamics. Moreover, malaria transmission is shaped by factors beyond vector presence in Iran, especially in this province. Effective malaria control requires implementing targeted interventions in high-risk areas, including preventing imported cases and identifying asymptomatic carriers [90,91]. Additionally, more research is needed to assess the shifting patterns of malaria transmission and adapt control strategies accordingly. This knowledge will enable the development of evidence-based vector control strategies. National and local malaria programs must proactively identify areas at risk of transmission to ensure effective intervention and sustainable disease control. In other words, effective malaria control strategies must go beyond vector surveillance and habitat assessment, incorporating socio-demographic factors and human movement patterns to develop more targeted interventions.
The analysis of variables underscores the importance of altitude and precipitation patterns in determining habitat suitability for An. stephensi. Moreover, response curve analysis further elucidates the complex relationships between environmental factors and species distribution. Altitude was identified as the most influential variable, suggesting that An. stephensi thrives in lower-altitude regions, which may explain its projected decline in coastal areas. Altitude plays a pivotal role, as species exhibit different responses at various altitudes. Following that, precipitation-related factors, particularly rainfall during the driest months, play a critical role, underscoring the sensitivity of An. stephensi to hydrological fluctuations. In this region, dry-season rainfall often occurs as flash floods, which subsequently create suitable breeding habitats for this vector. This is likely due to the formation of residual water pools after flooding, providing optimal conditions for oviposition and larval development. This discovery suggests that the optimal temperature and precipitation played a critical role in determining the spatial distribution of An. stephensi. Anopheles and Plasmodium are both influenced by temperature sensitivity. Given that mosquitoes are ectothermic organisms, temperature plays a crucial role in the developmental and mortality rates of each life stage [78,92]. Moreover, the gonotrophic cycle in adult female mosquitoes, which involves the blood meal-egg laying process, is also influenced by temperature fluctuations [93]. A notable correlation was observed between the monthly density of An. stephensi larvae and adults, and the levels of precipitation and mean temperature [94].
Recent studies have similarly identified key environmental factors influencing the distribution of malaria vectors, particularly An. stephensi [80,87]. These studies highlight that altitude, precipitation, and temperature are crucial variables impacting An. stephensi distribution [87]. Specifically, precipitation seasonality and annual mean temperature were noted as significant contributors to habitat suitability in Iran [80]. Similar modeling approaches for other Anopheles species, such as An. farauti in Australia has identified temperature, atmospheric moisture, and elevation as critical factors limiting coastal distribution [95]. This alignment with existing literature underscores the importance of altitude and precipitation in understanding the ecological niche of An. stephensi.
Moreover, isothermality (bio3) plays a crucial role in predicting An. stephensi distribution in the current study, as indicated by the significant decrease in model gain when it is omitted. This suggests that bio3 provides unique information that other climatic variables fail to capture. Isothermality represents the balance between daily and annual temperature fluctuations, which can directly impact the survival and adaptability of species, particularly An. stephensi. Our findings indicate that isothermality (bio3) values ranged between 32 and 44%, with the highest species response observed at 36%. It shows that the An. stephensi favors habitats where daily temperature fluctuations are relatively small compared to annual variations, possibly reducing thermal stress and enhancing survival rates. Given that climate change may alter temperature stability across different regions, shifts in bio3 could impact the species’ physiology [78]. Anopheles and Plasmodium are both influenced by temperature sensitivity. Given that mosquitoes are ectothermic organisms, temperature plays a crucial role in the developmental and mortality rates of each life stage [92,96]. Moreover, the gonotrophic cycle in adult female mosquitoes, which involves the blood meal-egg laying process, is also influenced by temperature fluctuations [93]. A notable correlation was observed between the monthly density of An. stephensi larvae and adults, and the levels of temperature [94]. A decrease in isothermality might expose the vector to harsher temperature fluctuations, potentially limiting its range, while an increase could create new suitable habitats. These results highlight the need to incorporate temperature stability into predictive modeling to improve our understanding of vector ecology and enhance malaria control strategies. Further research should explore how bio3 interacts with other climatic and ecological factors to refine species distribution predictions.
The Area Under the Curve (AUC) is a widely recognized metric in species distribution modeling, valued for its ability to assess predictive performance objectively [80,87]. In addition to AUC, the TSS was calculated to provide a complementary measure of model accuracy. In this study, the MaxEnt model demonstrated moderate predictive capacity for distinguishing suitable from unsuitable habitats for An. stephensi. This level of performance aligns with previous research findings [71,97–99], reinforcing the reliability of MaxEnt in ecological modeling. Various factors, including sample size, spatial resolution, choice of environmental predictors, and the method used for selecting pseudo-absence data, can influence the accuracy of predictive models. Refining these methodological aspects can help mitigate potential biases and enhance the reliability of distribution modeling. By optimizing data selection and model parameters, researchers can improve predictive performance and generate more precise insights for ecological and epidemiological applications [100,101].
Despite the valuable insights gained from this study, several limitations must be acknowledged. First, in presence-only species distribution modeling, pseudo-absence data are often used when true absence records are unavailable, providing a simple way to contrast presence points and prevent model convergence issues. However, the use of pseudo-absence data can introduce errors in predicted distributions, particularly if the selection of points does not accurately represent actual habitat conditions. This potentially can lead to over- or underestimation of the species’ range and reduce model accuracy, especially when presence data are unevenly distributed across geographic regions [100]. In this study, due to the limited availability of true absence data, background points were used, which is acknowledged as a limitation of the modeling approach.
Second, while the SSP scenarios provide a comprehensive look at potential future climates, they do not account for all possible local environmental changes, such as land-use alterations, which may further affect malaria vector distribution. Additionally, the study focuses primarily on one vector species, An. stephensi, while other species of Anopheles mosquitoes can influence their distribution through interspecific interactions or competition in the region [79]. Moreover, the dispersal of vectors can occur both naturally and through human-mediated transport. Natural dispersal is influenced by the flight capacity of mosquitoes, which can be incorporated into modeling frameworks to estimate local spread. Human-assisted dispersal, such as movement along roads, trade routes, or with transported goods, can facilitate long-distance introductions into new areas [102]. Incorporating these aspects into distribution models can enhance the spatial accuracy and predictive power of An. stephensi habitat suitability projections. Additionally, incorporating more detailed local environmental variables, such as land-use change and human infrastructure, would improve the model’s accuracy [103]. However, the primary scope of the present study was to assess the role of climatic conditions in shaping the potential ecological niches of An. stephensi in southern Iran. Due to the lack of reliable, spatially explicit data on human movement and public health interventions for this region, these factors could not be included in the current modeling framework.
Given the invasive nature of An. stephensi, it is also important to consider its capacity for further adaptation. Gene flow across populations could introduce novel heritable traits that enhance its ability to colonize diverse habitats. Population-genetic studies on An. stephensi and other malaria vectors (e.g., An. gambiae) have shown that even geographical barriers such as mountain ranges may not prevent gene flow, with genetic drift and human-mediated transport contributing to differentiation and spread [104,105]. Such findings emphasize that genetic connectivity and adaptive potential should be accounted for in long-term management strategies. In practical terms, longitudinal monitoring of distribution shifts in response to climate and anthropogenic changes will be invaluable for informing adaptive control measures. Regions such as Bashagard, which appear climatically stable in our projections, should be monitored closely as potential refuges for An. stephensi. Public health authorities may prioritize these areas for proactive surveillance.
Conclusion
This study provides valuable insights into the impact of climate change on the distribution of An. stephensi in Hormozgan Province, Iran. Our findings highlight the potential decline in suitable habitats for this malaria vector under future climate scenarios, particularly in coastal areas. These changes underscore the urgency of adapting vector control strategies to the evolving environmental conditions. While the MaxEnt model offers strong predictive performance, further research is necessary to refine our understanding of how climate and ecological factors interact to shape vector populations. The findings emphasize the importance of monitoring malaria vectors continuously and integrating climate change projections into public health interventions. By doing so, we can better prepare for the challenges posed by climate change and protect vulnerable populations from malaria transmission.
Supporting information
S1 Table. Summary of Anopheles stephensi related data sources in Hormozgan province.
https://doi.org/10.1371/journal.pntd.0014054.s001
(DOCX)
S1 File. Occurrence records of Anopheles stephensi in Hormozgan Province, Iran (1970–2022).
https://doi.org/10.1371/journal.pntd.0014054.s002
(CSV)
Acknowledgments
We express our gratitude to the Health Deputy, Hormozgan University of Medical Sciences, as well as the entomologists at Bandar Abbas Health Center, for their invaluable assistance in facilitating the field operations of our research. We also extend our heartfelt thanks to the residents of the villages under study for their wholehearted cooperation with our team of researchers.
References
- 1.
World Health O. World malaria report 2024: addressing inequity in the global malaria response. Geneva: World Health Organization; 2024.
- 2.
World Health Organization. World malaria report 2022. World Health Organization; 2022.
- 3. Nikpour F, Vatandoost H, Hanafi-Bojd AA, Raeisi A, Mirolyaie A, Mojahedi A-R, et al. Long-lasting residual efficacy of Actellic®300CS and Icon®10CS on different surfaces against Anopheles stephensi, an invasive malaria vector. Trop Med Int Health. 2024;29(9):781–91. pmid:39081142
- 4. Oshaghi MA, Abbasi M, Hanafi-Bojd AA. Host preference of Anopheles stephensi mosquitoes for blood feeding in south of Iran: Insights from Multiplex-PCR analysis. Trop Med Int Health. 2025;30(2):125–34. pmid:39707593
- 5. WHO. Disease outbreak news; dengue – Iran (Islamic Republic of Iran). 2024. Accessed 2024 July 22. https://www.who.int/emergencies/disease-outbreak-news/item/2024-DON526
- 6. Heydarifard Z, Heydarifard F, Mousavi FS, Zandi M. Dengue fever: a decade of burden in Iran. Front Public Health. 2024;12:1484594. pmid:39507666
- 7. Tadesse FG, Ashine T, Teka H, Esayas E, Messenger LA, Chali W, et al. Anopheles stephensi mosquitoes as vectors of Plasmodium vivax and falciparum, Horn of Africa, 2019. Emerg Infect Dis. 2021;27(2):603–7. pmid:33496217
- 8. Balkew M, Mumba P, Dengela D, Yohannes G, Getachew D, Yared S, et al. Geographical distribution of Anopheles stephensi in eastern Ethiopia. Parasit Vectors. 2020;13(1):35. pmid:31959237
- 9. Ochomo EO, Milanoi S, Abong’o B, Onyango B, Muchoki M, Omoke D, et al. Detection of anopheles stephensi mosquitoes by molecular surveillance, Kenya. Emerg Infect Dis. 2023;29(12):2498–508. pmid:37966106
- 10.
WHO. Vector alert: anopheles stephensi invasion and spread in Africa and Sri Lanka. World Health Organization; 2022. https://www.who.int/publications/i/item/9789240067714
- 11.
WHO. Partners convening: a regional response to the invasion of Anopheles stephensi in Africa. Geneva: World Health Organization; 2023. https://www.who.int/publications/i/item/9789240075535
- 12. Emiru T, Getachew D, Murphy M, Sedda L, Ejigu LA, Bulto MG, et al. Evidence for a role of Anopheles stephensi in the spread of drug- and diagnosis-resistant malaria in Africa. Nat Med. 2023;29(12):3203–11. pmid:37884028
- 13. Hanafi-Bojd A, Abbasi M, Yaghoobi-Ershadi M, Vatandoost H, Oshaghi M, Hazratian T, et al. Resistance status of main malaria vector, Anopheles stephensi Liston (Diptera: Culicidae) to insecticides in a malaria Endemic Area, Southern Iran. Asian Pac J Trop Med. 2019;12(1):43.
- 14. Sinka ME, Bangs MJ, Manguin S, Coetzee M, Mbogo CM, Hemingway J, et al. The dominant Anopheles vectors of human malaria in Africa, Europe and the Middle East: occurrence data, distribution maps and bionomic précis. Parasit Vectors. 2010;3:117. pmid:21129198
- 15. Samake JN, Athinya DK, Milanoi S, Ramaita E, Muchoki M, Omondi S, et al. Spatial distribution and population structure of the invasive Anopheles stephensi in Kenya from 2022 to 2024. Sci Rep. 2025;15(1):19878. pmid:40481090
- 16. Shivanna KR. Climate change and its impact on biodiversity and human welfare. ProcIndian Natl Sci Acad. 2022;88(2):160–71.
- 17. Malhi Y, Franklin J, Seddon N, Solan M, Turner MG, Field CB, et al. Climate change and ecosystems: threats, opportunities and solutions. Philos Trans R Soc Lond B Biol Sci. 2020;375(1794):20190104. pmid:31983329
- 18. Weiskopf SR, Rubenstein MA, Crozier LG, Gaichas S, Griffis R, Halofsky JE, et al. Climate change effects on biodiversity, ecosystems, ecosystem services, and natural resource management in the United States. Sci Total Environ. 2020;733:137782. pmid:32209235
- 19. Woolway RI, Jennings E, Shatwell T, Golub M, Pierson DC, Maberly SC. Lake heatwaves under climate change. Nature. 2021;589(7842):402–7. pmid:33473224
- 20. Ricciardi A, Hoopes MF, Marchetti MP, Lockwood JL. Progress toward understanding the ecological impacts of nonnative species. Ecol Monographs. 2013;83(3):263–82.
- 21. Root TL, Price JT, Hall KR, Schneider SH, Rosenzweig C, Pounds JA. Fingerprints of global warming on wild animals and plants. Nature. 2003;421(6918):57–60. pmid:12511952
- 22. Fouque F, Reeder JC. Impact of past and on-going changes on climate and weather on vector-borne diseases transmission: a look at the evidence. Infect Dis Poverty. 2019;8(1):51. pmid:31196187
- 23. Salahi-Moghaddam A, Khoshdel A, Dalaei H, Pakdad K, Nutifafa GG, Sedaghat MM. Spatial changes in the distribution of malaria vectors during the past 5 decades in Iran. Acta Trop. 2017;166:45–53. pmid:27826012
- 24. Salahi-Moghaddam A, Turki H, Yeryan M, Fuentes MV. Spatio-temporal prediction of the malaria transmission risk in Minab district (Hormozgan Province, Southern Iran). Acta Parasitol. 2022;67(4):1500–13. pmid:35951221
- 25.
WHO. El Niño and health global overview. World Health Organization; 2016. https://cdn.who.int/media/docs/default-source/climate-change/who_el_nino_and_health_global_report_21jan2016.pdf
- 26. Wickremasinghe R, Wickremasinghe AR, Fernando SD. Climate change and malaria a complex relationship. UN Chronicle. 2012;47(2):21–5.
- 27. Fountain A, Ye Y, Roca-Feltrer A, Rowe AK, Camara A, Fofana A, et al. Surveillance as a core intervention to strengthen malaria control programs in moderate to high transmission settings. Am J Trop Med Hyg. 2022;108(2_Suppl):8–13. pmid:35895588
- 28. Li J, Docile HJ, Fisher D, Pronyuk K, Zhao L. Current Status of Malaria Control and Elimination in Africa: Epidemiology, Diagnosis, Treatment, Progress and Challenges. J Epidemiol Glob Health. 2024;14(3):561–79. pmid:38656731
- 29.
Malaria GPTRB. The global malaria action plan for a malaria free world. Roll Back Malaria Partnership; 2008.
- 30. Caminade C, Kovats S, Rocklov J, Tompkins AM, Morse AP, Colón-González FJ, et al. Impact of climate change on global malaria distribution. Proc Natl Acad Sci U S A. 2014;111(9):3286–91. pmid:24596427
- 31.
WHO. Global technical strategy for malaria 2016–2030, 2021 update. Geneva: World Health Organization; 2021. https://www.who.int/publications/i/item/9789240031357
- 32. Navas ALA, Janko MM, Benítez FL, Narvaez M, Vasco LE, Kansara P, et al. Impact of climate and land use/land cover changes on malaria incidence in the Ecuadorian Amazon. PLOS Clim. 2024;3(4):e0000315. pmid:39027120
- 33. Sofaer HR, Barsugli JJ, Jarnevich CS, Abatzoglou JT, Talbert MK, Miller BW, et al. Designing ecological climate change impact assessments to reflect key climatic drivers. Glob Chang Biol. 2017;23(7):2537–53. pmid:28173628
- 34. Marques R, Krüger RF, Cunha SK, Silveira AS, Alves DMCC, Rodrigues GD, et al. Climate change impacts on Anopheles (K.) cruzii in urban areas of Atlantic Forest of Brazil: challenges for malaria diseases. Acta Trop. 2021;224:106123. pmid:34480869
- 35. O’Neill BC, Carter TR, Ebi K, Harrison PA, Kemp-Benedict E, Kok K, et al. Achievements and needs for the climate change scenario framework. Nat Clim Chang. 2020;10(12):1074–84. pmid:33262808
- 36. Gurney KR, Kılkış Ş, Seto KC, Lwasa S, Moran D, Riahi K, et al. Greenhouse gas emissions from global cities under SSP/RCP scenarios, 1990 to 2100. Global Environ Change. 2022;73:102478.
- 37. Hussain AMH, Alzeer AMY, Alsomali AMH, Muyini HA hmed M, Madkhali JAA, Alkharaan BSR, et al. Assessing the impact of climate change on vector-borne diseases: a systematic review of current evidence. Egypt J Chem. 2024;67(13):1209–20.
- 38.
Franklin J. Mapping species distributions. Cambridge University Press; 2010. https://doi.org/10.1017/cbo9780511810602
- 39. Urbina-Cardona N, Blair ME, Londoño MC, Loyola R, Velásquez-Tibatá J, Morales-Devia H. Species distribution modeling in Latin America: a 25-year retrospective review. Trop Conserv Sci. 2019;12.
- 40. Elith J, Leathwick JR. Species distribution models: ecological explanation and prediction across space and time. Annu Rev Ecol Evol Syst. 2009;40(1):677–97.
- 41. Hansen SE, Monfils MJ, Hackett RA, Goebel RT, Monfils AK. Data-centric species distribution modeling: Impacts of modeler decisions in a case study of invasive European frog-bit. Appl Plant Sci. 2024;12(3):e11573. pmid:38912123
- 42. Hawkins BA, Field R, Cornell HV, Currie DJ, Guégan J-F, Kaufman DM, et al. Energy, water, and broad-scale geographic patterns of species richness. Ecology. 2003;84(12):3105–17.
- 43. Norberg A, Abrego N, Blanchet FG, Adler FR, Anderson BJ, Anttila J, et al. A comprehensive evaluation of predictive performance of 33 species distribution models at species and community levels. Ecol Monograph. 2019;89(3).
- 44.
Guisan A, Thuiller W, Zimmermann NE. Habitat suitability and distribution models: with applications in R. Cambridge: Cambridge University Press; 2017.
- 45. Chang T, Choi S, Jun H, Chai J-Y, Song SH, Kim S, et al. Nowcasting vector mosquito abundance and determining its association with malaria epidemics in South Korea. Transbound Emerg Dis. 2025;2025:9959287. pmid:40302739
- 46. Lestina J, Cook M, Kumar S, Morisette J, Ode PJ, Peairs F. MODIS imagery improves pest risk assessment: a case study of wheat stem sawfly (Cephus cinctus, Hymenoptera: Cephidae) in Colorado, USA. Environ Entomol. 2016;45(6):1343–51. pmid:28028080
- 47. Elith J, Leathwick JR. Species distribution models: ecological explanation and prediction across space and time. Annu Rev Ecol Evol Syst. 2009;40(1):677–97.
- 48. Peterson AT, Martínez-Campos C, Nakazawa Y, Martínez-Meyer E. Time-specific ecological niche modeling predicts spatial dynamics of vector insects and human dengue cases. Trans R Soc Trop Med Hyg. 2005;99(9):647–55. pmid:15979656
- 49. Escobar LE, Romero-Alvarez D, Leon R, Lepe-Lopez MA, Craft ME, Borbor-Cordova MJ, et al. Declining prevalence of disease vectors under climate change. Sci Rep. 2016;6:39150. pmid:27982119
- 50. Carvalho BM, Rangel EF, Vale MM. Evaluation of the impacts of climate change on disease vectors through ecological niche modelling. Bull Entomol Res. 2017;107(4):419–30. pmid:27974065
- 51.
Phillips SJ, Schapire RE. Maxent software for modeling species niches and distributions. 2024.
- 52. Hernandez PA, Graham CH, Master LL, Albert DL. The effect of sample size and species characteristics on performance of different species distribution modeling methods. Ecography. 2006;29(5):773–85.
- 53. Sarma RR, Munsi M, Ananthram AN. Effect of climate change on invasion risk of Giant African Snail (Achatina fulica Férussac, 1821: Achatinidae) in India. PLoS One. 2015;10(11):e0143724. pmid:26618637
- 54.
Nephin J, Gregr EJ, St. Germain C, Fields C, Finney JL. Development of a species distribution modelling framework and its application to twelve species on Canada’s Pacific Coast. Canadian Science Advisory Secretariat; 2020.
- 55. Iran SCo. Iran statistical yearbook 2024. https://www.amar.org.ir/english#10291052-national-statistics
- 56. Boria RA, Olson LE, Goodman SM, Anderson RP. Spatial filtering to reduce sampling bias can improve the performance of ecological niche models. Ecol Model. 2014;275:73–7.
- 57. Brown JL. SDMtoolbox: a python‐based GIS toolkit for landscape genetic, biogeographic and species distribution model analyses. Methods Ecol Evol. 2014;5(7):694–700.
- 58. Fick SE, Hijmans RJ. WorldClim 2: new 1‐km spatial resolution climate surfaces for global land areas. Intl J Climatol. 2017;37(12):4302–15.
- 59.
Quinn G, Keough M. Experimental design and data analysis for biologists. 2023.
- 60.
Naimi B. Usdm: uncertainty analysis for species distribution models. 2015;1:1–12.
- 61. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol Model. 2006;190(3–4):231–59.
- 62. Phillips SJ. A Brief Tutorial on Maxent. 2017. Accessed 2025 February 21. https://biodiversityinformatics.amnh.org/open_source/maxent/
- 63. Freeman EA, Moisen G. PresenceAbsence: AnRPackage for Presence Absence Analysis. J Stat Soft. 2008;23(11).
- 64. Liu C, White M, Newell G. Selecting thresholds for the prediction of species occurrence with presence‐only data. J Biogeo. 2013;40(4):778–89.
- 65. Hijmans RJ, Barbosa M, Bivand R, Brown A, Chirico M, Cordano E, et al. terra: spatial data analysis. https://rspatial.github.io/terra/
- 66.
Wickham H. ggplot2: elegant graphics for data analysis. Springer Publishing Company, Incorporated; 2016.
- 67. Allouche O, Tsoar A, Kadmon R. Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS). J Appl Ecol. 2006;43(6):1223–32.
- 68.
Frans VF. True skill statistic (TSS) calculation across multiple. Dept. of Fisheries and Wildlife CfSIa, Sustainability MSU, USA. Maxent Runs. 2018.
- 69. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology. 1982;143(1):29–36. pmid:7063747
- 70. Freeman EA, Moisen G. PresenceAbsence: AnRPackage for presence absence analysis. J Stat Soft. 2008;23(11).
- 71. Duan R-Y, Kong X-Q, Huang M-Y, Fan W-Y, Wang Z-G. The predictive performance and stability of six species distribution models. PLoS One. 2014;9(11):e112764. pmid:25383906
- 72. FIELDING AH, BELL JF. A review of methods for the assessment of prediction errors in conservation presence/absence models. Envir Conserv. 1997;24(1):38–49.
- 73. Adeogun A, Babalola AS, Okoko OO, Oyeniyi T, Omotayo A, Izekor RT, et al. Spatial distribution and ecological niche modeling of geographical spread of Anopheles gambiae complex in Nigeria using real time data. Sci Rep. 2023;13(1):13679. pmid:37608210
- 74. Liu Q, Wang M, Du Y-T, Xie J-W, Yin Z-G, Cai J-H, et al. Possible potential spread of Anopheles stephensi, the Asian malaria vector. BMC Infect Dis. 2024;24(1):333. pmid:38509457
- 75. Mordecai EA, Ryan SJ, Caldwell JM, Shah MM, LaBeaud AD. Climate change could shift disease burden from malaria to arboviruses in Africa. Lancet Planet Health. 2020;4(9):e416–23. pmid:32918887
- 76. Taylor R, Messenger LA, Abeku TA, Clarke SE, Yadav RS, Lines J. Invasive Anopheles stephensi in Africa: insights from Asia. Trends Parasitol. 2024;40(8):731–43. pmid:39054167
- 77. Zhou G, Zhong D, Yewhalaw D, Yan G. Anopheles stephensi ecology and control in Africa. Trends Parasitol. 2024;40(2):102–5. pmid:38142196
- 78. Abbasi M, Oshaghi MA, Mehdi Sedaghat M, Hazratian T, Rahimi Foroushani A, Jafari-Koshki T, et al. Development of a degree-day model to predict the growth of Anopheles stephensi (Diptera: Culicidae): implication for vector control management. Environ Entomol. 2023;52(6):1126–38. pmid:37738476
- 79. Lawlor JA, Comte L, Grenouillet G, Lenoir J, Baecher JA, Bandara RMWJ, et al. Mechanisms, detection and impacts of species redistributions under climate change. Nat Rev Earth Environ. 2024;5(5):351–68.
- 80. Pakdad K, Hanafi-Bojd AA, Vatandoost H, Sedaghat MM, Raeisi A, Moghaddam AS, et al. Predicting the potential distribution of main malaria vectors Anopheles stephensi, An. culicifacies s.l. and An. fluviatilis s.l. in Iran based on maximum entropy model. Acta Trop. 2017;169:93–9. pmid:28188765
- 81. Hanafi-Bojd AA, Vatandoost H, Yaghoobi-Ershadi MR. Climate change and the risk of malaria transmission in Iran. J Med Entomol. 2020;57(1):50–64. pmid:31429469
- 82. Hanafi-Bojd AA, Sedaghat MM, Vatandoost H, Azari-Hamidian S, Pakdad K. Predicting environmentally suitable areas for Anopheles superpictus Grassi (s.l.), Anopheles maculipennis Meigen (s.l.) and Anopheles sacharovi Favre (Diptera: Culicidae) in Iran. Parasit Vectors. 2018;11(1):382. pmid:29970145
- 83. Ren Z, Wang D, Ma A, Hwang J, Bennett A, Sturrock HJW, et al. Predicting malaria vector distribution under climate change scenarios in China: challenges for malaria elimination. Sci Rep. 2016;6:20604. pmid:26868185
- 84. Fuller DO, Ahumada ML, Quiñones ML, Herrera S, Beier JC. Near-present and future distribution of Anopheles albimanus in Mesoamerica and the Caribbean Basin modeled with climate and topographic data. Int J Health Geogr. 2012;11:13. pmid:22545756
- 85. Drake JM, Beier JC. Ecological niche and potential distribution of Anopheles arabiensis in Africa in 2050. Malar J. 2014;13:213. pmid:24888886
- 86. Saberi N, Raeisi A, Gorouhi MA, Vatandoost H, Bozorg Omid F, Hanafi-Bojd AA. Modeling the effect of climate change on the distribution of main malaria vectors in an endemic area, Southeastern Iran. Iran J Public Health. 2023;52(5):1061–70. pmid:37484730
- 87. Ismail RBY, Bozorg-Omid F, Osei JHN, Pi-Bansa S, Frempong KK, Ofei MK, et al. Predicting the environmental suitability for Anopheles stephensi under the current conditions in Ghana. Sci Rep. 2024;14(1):1116. pmid:38212448
- 88. Ryan SJ, Lippi CA, Villena OC, Singh A, Murdock CC, Johnson LR. Mapping current and future thermal limits to suitability for malaria transmission by the invasive mosquito Anopheles stephensi. Malar J. 2023;22(1):104. pmid:36945014
- 89. Oshaghi MA, Yaaghoobi F, Vatandoost H, Abaei MR, Akbarzadeh K. Anopheles stephensi biological forms; geographical distribution and malaria transmission in malarious regions of Iran. Pak J Biol Sci. 2006;9(2):294–8.
- 90. Raeisi A, Gouya MM, Nadim A, Ranjbar M, Hasanzehi A, Fallahnezhad M, et al. Determination of Malaria Epidemiological Status in Iran’s Malarious Areas as Baseline Information for Implementation of Malaria Elimination Program in Iran. Iran J Public Health. 2013;42(3):326–33. pmid:23641411
- 91. Shahandeh K, Basseri HR. Challenges and the path forward on malaria elimination intervention: a systematic review. Iran J Public Health. 2019;48(6):1004–13. pmid:31341841
- 92. Tuno N, Farjana T, Uchida Y, Iyori M, Yoshida S. Effects of temperature and nutrition during the larval period on life history traits in an invasive malaria vector Anopheles stephensi. Insects. 2023;14(6):543. pmid:37367359
- 93. Beck-Johnson LM, Nelson WA, Paaijmans KP, Read AF, Thomas MB, Bjørnstad ON. The effect of temperature on Anopheles mosquito population dynamics and the potential for malaria transmission. PLoS One. 2013;8(11):e79276. pmid:24244467
- 94. Hanafi-Bojd A, Abbasi M, Rahimi Foroushani A, Jafari-Koshki T, Pakdad K, Vatandoost H. The impact of climatic variables on the population dynamics of the main malaria vector, Anopheles stephensi Liston (Diptera: Culicidae), in southern Iran. Asian Pac J Trop Med. 2020;13(10):448.
- 95. Sweeney AW, Beebe NW, Cooper RD, Bauer JT, Peterson AT. Environmental factors associated with distribution and range limits of malaria vector Anopheles farauti in Australia. me. 2006;43(5):1068–75.
- 96. Abbasi M, Oshaghi MA, Mehdi Sedaghat M, Hazratian T, Rahimi Foroushani A, Jafari-Koshki T, et al. Development of a degree-day model to predict the growth of Anopheles stephensi (Diptera: Culicidae): implication for vector control management. Environ Entomol. 2023;52(6):1126–38. pmid:37738476
- 97. Hanafi-Bojd AA, Vatandoost H, Yaghoobi-Ershadi MR. Climate change and the risk of malaria transmission in Iran. J Med Entomol. 2020;57(1):50–64. pmid:31429469
- 98. Estallo EL, Sangermano F, Grech M, Ludueña-Almeida F, Frías-Cespedes M, Ainete M, et al. Modelling the distribution of the vector Aedes aegypti in a central Argentine city. Med Vet Entomol. 2018;32(4):451–61. pmid:30027565
- 99. Oppel S, Meirinho A, Ramírez I, Gardner B, O’Connell AF, Miller PI, et al. Comparison of five modelling techniques to predict the spatial distribution and abundance of seabirds. Biol Conserv. 2012;156:94–104.
- 100. Wisz MS, Hijmans RJ, Li J, Peterson AT, Graham CH, Guisan A. Effects of sample size on the performance of species distribution models. Divers Distribut. 2008;14(5):763–73.
- 101. Tsoar A, Allouche O, Steinitz O, Rotem D, Kadmon R. A comparative evaluation of presence‐only methods for modelling species distribution. Divers Distribut. 2007;13(4):397–405.
- 102. Ahn J, Sinka M, Irish S, Zohdy S. Modeling marine cargo traffic to identify countries in Africa with greatest risk of invasion by Anopheles stephensi. Sci Rep. 2023;13(1):876. pmid:36650185
- 103. Couper LI, Farner JE, Caldwell JM, Childs ML, Harris MJ, Kirk DG, et al. How will mosquitoes adapt to climate warming?. Elife. 2021;10:e69630. pmid:34402424
- 104. Sharma V, Dube M, Gakhar SK. Gene flow in malaria vector Anopheles stephensi (Diptera: Culicidae) across the Aravalli Hills in North‐west India. Entomol Res. 2010;40(2):85–96.
- 105.
Lanzaro G, Tripet F. Gene flow among populations of Anopheles gambiae: a critical review. 2004.