Habitat suitability modeling of mosquito species in Faiyum Governorate, Egypt, using GIS

Background The life cycle and geographic range of mosquitoes are profoundly influenced by weather conditions. In Faiyum Governorate, Egypt, researchers looked at over 42 different mosquito breeding areas in the wild. Culex pipiens, Culex antennatus, Culex perexiguus, Culex theileri, Anopheles multicolor, Anopheles sergentii, Ochlerotatus caspius, Culiseta longiareolata, and Uranotaenia unguiculata, were all represented among the collected mosquito larvae. Using Maxent, we identified the most important bioclimatic factors affecting habitat appropriateness for each species. In order to determine the link between a species’ habitat appropriateness and bioclimatic factors, the computer builds response curves for each bioclimatic factor. All species risk maps, species richness maps, and individual species maps were generated with a Software that build interactive web maps (ArcGIS 10.3). Results Each mosquito species’ range was impacted by four bioclimatic factors. The most influential factors were those linked to temperature and precipitation. Culex antennatus, Culex perexiguus, Culex pipiens, and Ochlerotatus caspius responded best to temperature‑related factors. Precipitation of wettest month (bio 13) was effective in four species ( Anopheles multicolor, Culex antennatus, Culex perexiguus, and Culex pipiens ). Conclusion Areas in the north and west of El‑Fayoum district (Zawyat El‑Karadsah, Beni‑Saleh, and Monshat Abd‑ Allah), in Tamiya in Kafr Mahfouz, in the south of Abshway, especially in Tobhar, in the north of Atssa, in the south and center of Sinnuris have been identified as high‑risk areas and more susceptible to mosquito‑borne diseases.

Valley.Faiyum Governorate is located in western Egypt and is a significant agricultural region.The area's water resources, such as the Nile River and Lake Qarun, make it an ideal mosquito breeding site.Faiyum Governorate is home to a variety of mosquito species, including Anopheles mosquitoes, which are the primary vectors of malaria in this region, and they breed in stagnant water bodies, such as irrigation canals and rice paddies.Culex mosquitoes are known to transmit diseases, such as West Nile virus and filariasis [9][10][11].They may be found in sewage, drainage ditches, and other water sources where they can reproduce.Malaria, dengue fever, and the Zika virus are just a few of the illnesses spread by the Aedes mosquito.They procreate in manmade objects like empty cans, flower vases, and tires [12][13][14].These mosquitoes can spread a variety of diseases, making their presence in Faiyum Governorate a serious public health concern.In terms of malaria, Faiyum Governorate is a significant epicenter because the illness has been present there since antiquity.Polymerase chain reaction (PCR) assays were used to confirm its presence in Egyptian mummies (El-Fayoum) [15].Anopheles mosquitoes are the most important vectors of malaria transmission in the area.Malaria is a serious public health concern in Egypt, and Faiyum Governorate is one of the regions with the highest incidence of the disease [16][17][18].Several public health initiatives, such as the use of insecticide-treated bed nets, repellents, and larvicides to target mosquito breeding grounds, are used to lessen the effects of disease spread by mosquitoes.The absence of clear distribution maps of these mosquito species in the study area reduces the effectiveness of the control strategy.Such maps can easily be generated using data science.Many biologists are motivated to solve environmental problems using data science [19].Biodiversity biological records are the focus of informatics, a subfield of data science.Information on biodiversity may benefit from bioinformatics methods for display, discovery, exploration, and analysis of biological data [20].It relies on digitally preserved taxonomic, biogeographic, and ecological information gathered in the contemporary era [21].Biodiversity informatics in the final set usually produces species distribution models (SDMs).SDMs calculate the relationship between species records at sites and species spatial analyses [22,23].It makes predictions on where species will be found and how communities and ecosystems will change as a result of climate and other environmental shifts [24].Spatial biodiversity patterns may be broken down into their ecological and evolutionary components with the use of species distribution models [25].SDMs have been used in a variety of disciplines, including ecology, parasitology, biogeography, conservation ecology, and evolution, to evaluate species' spatial distributions in recent years [26][27][28][29][30].The maximum entropy model (Maxent) [31] is a crucial resource for modeling species distributions since it generates a jackknife, accommodates limited data, and provides reliable results [22,31,32].Maxent is superior to other ecological niche models for forecasting species distribution when just a few sample places are provided [25,33].Replication is made easier since cross-validation and testing of the model's performance are made possible by repeated subsampling [34].A Geographical Information System (GIS) is a database and set of related software applications that allows users to see and manage geographic data, conduct geographical analysis, and simulate real-world events [35,36].The application of this new technique in mosquito ecology will help in their control and evaluation of their status in different landscapes.The present study aimed to apply SDMs tools to monitor mosquito species and their suitable habitats in El Fayoum by surveying the governorate and assessing the status of these vectors in this important area of Egypt.

Occurrence data and study area
El-Fayoum is a 6,068.7 Square kilometers (km 2 ) agricultural governorate in Egypt located below sea level.(https:// en.wikip edia.org/ wiki/ Faiyum_ Gover norate).It lies 90 km (km) southwest of Cairo and is connected to the Nile by the Bahr Youssef River, which enters the governorate from the east and branches into other rivers before emptying into Qaroun Lake.The Governorate, which resembles a depression, is situated in the western desert between the latitudes of 29° 02' and 29° 35' N and the longitudes of 30° 23' and 31° 05' E. Sinnuris, Tamiya, El-Fayoum, Atssa, Abshway, and Youssef El-Sedik are the six districts that make up this city.El-Fayoum has a hot, dry climate with infrequent winter rain and abundant sun throughout the year.In El-Fayoum, it is estimated that there are 3845.1 h of sunshine every year.In January, the average temperature is 13.5 Degree Celsius (°C), while in August, it is 30.6 °C.According to climate-data.org (https:// en.clima te-data.org/ africa/ egypt/ faiyum-Gover norate/ faiyum-5569/# weath er), May had the lowest relative humidity (31.46%) and December had the highest (55.31%).In October and November 2020, mosquito larvae were collected from locations where they might develop successfully throughout the governorate, as shown in Fig. 1.The 42 sites that made up the breeding habitats were canals, sewage tanks, agricultural puddles, stagnant water puddles, and wetlands.According to the World Health Organization (WHO) [37], The dipping technique was used to capture the larvae, and further laboratory identification followed Mattingly and Knight criteria [38], Harbach [39], Harbach [40], Savage and Strickman [41], Harbach [42], and Harbach [43] (Fig. 1).

Climatological data
Climate data with a spatial resolution of 30 arc seconds were downloaded from the WorldClim database (http:// www.world clim.org/) (1 km).The present-day weather was calculated using data from 19 different meteorological elements in this investigation (Table 1).We started by utilizing Maxent's jackknifing function to estimate the minimum set of variables necessary for our purposes.Jackknife parameters with zero impact on species distribution were also discovered in this study.Strongly linked parameters with little biological impact on the species were eliminated using the Universal tool in ArcGIS 10.3 SDM Tools (Remove highly correlated variables) [44].Therefore, we chose four important factors for each species to consider while establishing a suitable habitat in the study area.ArcGIS V.10.3 was used to clip and resize all bioclimatic layers to fit inside the boundaries of the Faiyum governorate, and the resulting ASCII grid files were imported into Maxent.Extract by mask in Arc Map was used to collect data for each bioclimatic variable [45].These factors were included in models of species dispersion [46,47].

Habitat suitability modeling
In this study, we used the maximum entropy model (Maxent version 3.3.3k, available at http:// www.cs.princ eton.edu/ wscha pire/ maxent/) since it performed better than other modeling techniques when dealing with limited data sets (see reference [48]).To determine the correlations between a species' habitat appropriateness and bioclimatic parameters, the software assigns a value between 0 and 1 to each bioclimatic variable and creates response curves for each [49].Maxent will automatically delete any entries that are entered twice in the same cell.To train our models, we used 75% of the occurrence data, whereas 25% were used for testing [50].By running the model 10 times for each species and averaging the results, tenfold cross-validation was used to increase  model performance.The area under the curve (AUC) of the receiver operating characteristics (ROC) was used to evaluate the performance of the models [51,52].The AUC may be anywhere from 0 (random discrimination) to 1 (perfect discrimination).

Species richness and risk map
The species richness map was obtained by submitting the presence/absence maps that were generated for each of the five mosquito species using ArcGIS 10.3.This process was performed using the Arc toolbox Map algebra.All recorded data of collected mosquito species (either the five selected species or the remaining species that were not included in the modeling study) were implemented in Maxent software using the 19 bioclimatic variables to produce a risk map that represents the most suitable area for all mosquitoes reported in the governorate [53].

Model performance contribution of bioclimatic variables
With AUCs between 0.77 and 0.94, the models for all five mosquito species outperformed chance.The effect of each bioclimatic variable used in the generation of the models was tabulated using the jackknife test (Table 2 and Fig. 2).The response curves for the most effective climatological parameters contributing to the distribution of each species are shown in Fig. 3.The curves show how the occurrence of each species is correlated with each variable.Month).The least productive biovariable in this study was bio 2 (Mean Diurnal Range).
In the case of Culex pipiens (Fig. 2c), bio 3 (isothermality) had the greatest impact, followed by bio 5. (Max Temperature of Warmest Month).These two factors explain 81.6% of the variation.Among the most useful metrics, bio 12 was the least useful, while bio 13 (Wet Month Precipitation) ranked third (Annual Precipitation).
Bio 13 (Precipitation of Wettest Month) and bio 16 (Day length of Day) were the two most useful parameters in controlling Anopheles multicolor (Fig. 2d) (Precipitation of Wettest Quarter).In addition, this species ranked third for bio 5 (Highest Temperature in Hottest Month) and fourth for bio 10 (Warmest Quarterly Average Temperature).
Figure 2 e shows that the most influential climatic factor for Ochlerotatus caspius was the mean temperature of the driest quarter (bio 9), followed by annual precipitation (bio 12), the maximum temperature of the hottest month (bio 5), and the minimum temperature of the coldest month (bio 3) (isothermal).

Habitat suitability modeling
According to the acquired results of the selected species distribution modeling, their distribution patterns were similar to their real situation throughout the governorate (Fig. 7).For Culex antennatus, the most current suitable habitats appeared in Atssa (Abo Gandeer, Garado, Menia El-Heet) and El-Fayoum (the center of the city, Manashi El-Khatib, Belal Bin Rabah) districts, while in Tamyia (Kafr Mahfouz, Dar El-Salam, Sersena), Sinnuris (the center of the city, Fedmeen, Tersa, Behmo), and Abshway (Senaro, El-Agmeen) districts habitats were less suitable.

Discussion
The governorate of El-Fayoum is geographically distinct.It is distinguished by its gentle inclination and proximity to sea level.In addition, most people are employed in agriculture and related industries [54].The mild environment and high relative humidity [55] that favor mosquito breeding in this region causes the season of disease transmission to be extended to eight months each year, from roughly the end of March to the end of November [56].
Culex pipiens, Culex antennatus, Culex theileri, Culex perexiguus, Anopheles multicolor, Anopheles sergentii, Ochlerotatus caspius, Culiseta longiareolata, and Uranotaenia unguiculata were among the mosquito species discovered in this study.The five genera to which these species belong are Aedes, Anopheles, Culex, Culiseta, and Uranotaenia.These are the most prevalent genera in Egypt, according to numerous research [17,40,56,57].Most of these diseases are also important disease vectors for both humans and animals.Consider the genus Culex, which includes species such as Culex pipiens, which is thought to be the vector of numerous arboviruses, including the West Nile virus.Louis encephalitis, rift valley disease, and Sindbis viruses pose hazards to human life [58].Wuchereria bancrofti, a filarial nematode that causes lymphatic filariasis, is transmitted by Cx. pipiens [59,60].According to Salamah et al. [61], the primary vector of lymphatic filariasis in Egypt is female Cx.pipiens.It can act as a hepatitis C virus (HCV) vector because an experimental study by El-Kholy et al. [62] demonstrated that Cx. pipiens can contract the disease, the virus reproduces inside the insect body, and the infected insect bites humans.
The current predicted habitat suitability modeling for Culex pipiens was high in many regions, especially in the Abshway and Sinnuris districts, which is consistent with the results of larval collection, as these districts recorded a high abundance of Cx. pipiens larvae.This species was the most frequent type and was found in different breeding habitats, especially in agricultural canals [61,63].Moreover, Cx. pipiens tolerated high degrees of pollution, as it was found with high densities in stagnant water habitats consistent with the findings of Ibrahim et al. [64] (particularly those related to human activities).Generally, breeding sites with the highest larval density were closest to houses, which agrees with Madewell et al. [65], who stated that there was an inverse relationship between the distance to the nearest house and larval abundance.Adult female mosquitoes require a blood meal for oviposition.It may be fed on mammalian blood meal especially human [66] or mixed blood meal [67].In this context, the female preferred to stay close to its host for blood feeding and then chose the nearest breeding site for oviposition; therefore, breeding sites that were close to humans and animals were characterized by high larval density.
Ochlerotatus caspius was predicted to be high in many regions, such as Qarun Lake, Tamiya, El-Fayoum (Beni-Saleh), and areas with high seepage and sewage water [64,68].On the other hand, Cx. antennatus showed high suitability in Atssa, in agreement with the collected samples, as this district had the highest larval density of this species.Oringanje et al. [69] found that polluted breeding habitats encouraged the spreading of more Culicine species but Anopheline species did not prefer this type of habitats.This agrees with our results, as the majority of Culicine breeding sites were polluted.Anopheles sergentii larvae were found at two sites characterized by shallow, clean, freshwater, but some Anopheles multicolor breeding sites were polluted, similar to the observations of El-Hefniet al. [56].
Mosquito species were unequally distributed throughout the governorate.Sites displaying mosquito richness are vulnerable to disease transmission El-Hefni et al. [57].The risk map shows that the desert part of the governorate displays no risk.On the other hand, highly populated agricultural areas display a very high risk of diversity and variety in breeding sites.
For bioclimatic variables, precipitation-related variables and temperature-related variables had the highest effects on all species distributions.High summer temperatures likely have a role in the distribution of the three species evaluated (Culex antennatus, Culex perexiguus, and Ochlerotatus caspius), as shown by the driest quarter mean temperature (bio 9).Temperature plays an important role in mosquito development and density because it is a cooled-blooded animal.Development is accelerated by increasing temperature until a definite limit is reached, mortality begins to increase, and the appropriate temperature for development is approximately 25-27 °C [70].In addition, Loetti et al. [71] found that temperature between 7 and 25 °C had an inverse relationship with the time needed for larval development.According to research conducted by Mamai et al. [72], when temperatures were elevated to 27 °C, the number of larvae rose but the size of the emerging adult dropped.This, in turn, demonstrates how climate change has affected the distribution of different species [24].
Mosquito-borne diseases are a significant public health concern throughout Faiyum Governorate and Egypt as a whole.It is imperative that measures be taken to reduce mosquito populations and stop disease spreading for the sake of public health and safety.

Conclusion
This study is a baby step toward a fuller understanding of the ecology of significant mosquito species in Egypt.Further surveillance is needed to better evaluate the entire mosquito fauna in the region.In addition, the investigation of GIS and its modeling tools to predict the pattern of such distribution and abundance under different climate change scenarios is required.

Fig. 2
Fig. 2 Jackknife test of the most effective bioclimatic variables contributing to the distribution model of the five mosquito species throughout Faiyum Governorate.Bio 2: Mean Diurnal Range, bio 3: Isothermality, bio 5: Max Temperature of Warmest Month, bio 7: Temperature Annual Range, bio 9: Mean Temperature of Driest Quarter, bio 10: Mean Temperature of Warmest Quarter, bio 12: Annual Precipitation, bio 13: Precipitation of Wettest Month, bio 16: Precipitation of Wettest Quarter

Fig. 3
Fig. 3 The response curve of the most effective bioclimatic variables in the distribution of a Culex antennatus, b Culex perexiguus, c Culex pipiens, d Anopheles multicolor, e Ochlerotatus caspius

Fig. 7 Fig. 8 4 :Fig. 9
Fig. 7 suitability for: a Culex antennatus, b Culex perexiguus, c Culex pipiens, d Anopheles multicolor, and e Ochlerotatus caspius.The map here shows the presence of suitable habitat for each species where the red color represents the confirmation of present of each species through Fayoum territory, while the green indicates the complete absence of the species

Table 1
Environmental parameters that were used to predict the current habitat suitability distribution of mosquito species Variables DescriptionBio 1 Annual mean temperature Bio 2 Mean diurnal range (mean of monthly max temp -min temp)

Table 2
Table 2 shows percentages of bioclimatic variables used in modeling mosquito distribution throughout Faiyum Governorate.By reading Fig. 2a, it showed that bio 9 (Mean Temperature of Driest Quarter) is the best parameter for Culex antennatus.Bio 16 (Wettest Quarter Precipitation) and bio 13 (Wettest Month Precipitation) tied for second and third place in terms of impact, respectively, while bio 7 (Annual Temperature Range) ranked last.Biotype 9 (Mean Temperature of Driest Quarter), Biotype 12 (Annual Precipitation), and Biotype 13 (Day length) were the most important environmental factors for Culex perexiguus (Fig. 2b) (Precipitation of Wettest Percentages of bioclimatic variables used in modeling mosquito distribution throughout Faiyum Governorate