Breeding Habitat Preference of the Dengue Vector Mosquitoes Aedes aegypti and Aedes albopictus from Urban, Semiurban, and Rural Areas in Kurunegala District, Sri Lanka

Elimination of vector mosquito larvae and their breeding environments is an effective strategy in dengue disease control. Present study examined larval density and water quality in breeding habitats and container preference of dengue vectors Ae. aegypti and Ae. albopictus. Larval surveys were conducted monthly in urban, semiurban, and rural sites in Kurunegala, Sri Lanka, from January 2019 to December 2021. Larval densities were recorded under the following three categories: type of container (16 types), type of material (6 types), and location (indoor/outdoor). Breeding preference ratios (BPRs) were calculated using Index of Available Containers and the Index of Contribution to Breeding Sites. Out of 19,234 wet containers examined, larval stages were found in 1,043 habitats. Ae. albopictus larvae were in all three areas whereas Ae. aegypti larvae were restricted to urban areas. Highest number of wet containers and highest positivity were reported from urban followed by semiurban. In general, discarded nondegradable items were the most frequent and mostly positive breeding sites. For Ae. aegypti, the most preferred breeding sites were gutters and concrete slabs. Ae. albopictus mostly preferred concrete slabs in urban areas and tyres in semiurban and rural areas. Material types such as rubber and concrete were mostly preferred by Ae. aegypti whereas ceramic was preferred by Ae. albopictus. Although plastic was the most available material type in all study sites, preference to plastic was low except for urban Ae. albopictus. Both species preferred urban indoor breeding habitats although outdoor breeding was preferred by Ae. albopictus in rural areas. Larval densities of Ae. aegypti and semiurban Ae. albopictus significantly correlated with the BPR of the container type and material type. Dengue vector larvae were found in a 6.7–9.4 pH range. Total dissolved solids and alkalinity positively correlated with preference. Information generated can be successfully used in waste management and public education for effective vector control.


Introduction
Dengue is a mosquito-borne disease that has rapidly spread across the world [1].It is caused by the DENV virus of the genus Flavivirus.DENV has the following four serotypes: DENV 1 to DENV 4. Te disease is transmitted through the bites of infected Aedes mosquitoes [2].Ae. aegypti Linnaeus and Ae.albopictus Skuse are considered as the most prominent dengue vector species in the world [2], and their close association to human habitations and acquired human host specifcity have enabled them to display high vector competence giving high infection and transmission rates for dengue and other viruses [3].Tey are considered container breeders and breed in indoor and outdoor settings in a wide variety of natural and man-made water-holding containers, such as discarded plastic containers, plant axils, water storage barrels, cement tanks, and fower pots.Ae. aegypti, which is considered the primary vector of dengue, is highly anthropophilic and thrives close to human inhabiting mainly in urban and periurban areas [4,5] and commonly breeds in artifcial containers in houses [6].Ae. albopictus, the secondary vector of dengue, is a sylvatic species that has recently been adapted to urban and semiurban environments but still tends to breed more often in natural containers, such as tree stumps and coconut shells, and to a lesser extent in artifcial containers [6].
Even though more research studies are needed to explain the factors responsible for the attraction of dengue vectors to diverse breeding sites, many researchers have identifed breeding habitat availability as the key factor that determines the preference [7][8][9][10].In most cases, data from larval surveys are used to calculate the breeding preference ratios based on positivity which may have little to do with productivity.It is important to understand the association between breeding habitat physicochemical parameters and larval productivity.Water quality parameters in breeding environments play a signifcant role in egg hatching and growth of the progeny from larvae to adults [11,12].Because of this, gravid female mosquitoes are sensitive to both biotic and abiotic factors such as organic matter [12], bacteria, phosphate, ammonia, and potassium content [13] of the water during the breeding exercise.Tese factors are known to be closely related to the abundance of larvae and adults in the feld [14].Although presence of heavy metals such as iron, zinc, and copper has been found at various concentrations in Ae. aegypti breeding sites [15,16], their relationship with larval development remains unknown.
Type of the containers, such as buckets, bottles, vases, and tanks, has been searched by many researchers [8][9][10] while some others have used other criteria, such as material type and the capacity [9,17], together with the container type for their studies.None of these studies, however, has incorporated features like attractiveness of female mosquitoes or functional characteristics as infuenced by human actions to breeding sites.Lack of uniformity among container classifcations has made the situation worst in comparing the results and decision-making strategies.
In order to formulate successful vector control strategies, the current study attempted to establish the association between larval density and container preference based on the container availability, container type, materials of container, location, and water quality.

Larval Survey.
Only the sites maintained by home owners (residential areas) were investigated.Mosquito larval surveys were conducted at all selected sites following the systematic sampling strategy [19].Every second premise along the locality's inspection route was chosen for sampling.A total of 108 surveys were completed over the study period from all study areas.For one survey, a minimum of 100 locations per site were selected for sampling.At each location, all water-holding containers, in both indoors and outdoors, were inspected for the presence of larvae.A container with at least one immature mosquito was considered as a positive breeding site.Dippers were used to sample 50 mL from larger water-holding bodies, and depending on the water volume, up to three samples were taken from varying depths.Larval density per container was calculated as the number of larvae per 50 mL volume.Surveys were conducted monthly during January 2019 to December 2021 period in all three study sites.
Collected larvae were brought to the water laboratory at the ofce of the regional director of health services, Kurunegala, fxed in 70% ethanol and identifed using the key introduced by Darsie and Ward [20].Water samples were collected separately from each type of breeding habitats for quality analysis.

Characterisation of Dengue Vector Breeding Habitats.
Water-holding positive and potential containers were classifed based on their usage type, material, and placement.Te usage type was classifed into sixteen categories considering the use indicated by the homeowners, i.e., concrete slabs, gutters, water storage barrels water storage cement tanks, tyres, ornamental fower pots, tree holes, leaf axes, bamboo stumps, and clay plots, AC refrigerators, used/nonused commodes and cisterns, discarded degradable items (coconut shells, decaying leaves, kitchen waste, paper waste, damaged paper boxes, etc.), discarded nondegradable items (tins, yoghurt and ice cream cups, bottles, cans, damaged ceramic items, etc.), and covering items.All other breeding habitats were classifed as miscellaneous (Figure 2).Te materials of the containers were classifed into eight categories, i.e., concrete, plastic, natural, clay, ceramic, rubber, glass, and miscellaneous (tin, paper-based material, and so on).Based on the placement, containers were classifed as indoor and outdoor.

Measurement of Water Quality Parameters.
Water samples for analysis were collected according to the container type and the material of the container.Analysis required about 1 L volume from each container type, and for the containers with smaller capacities, water from the same container type was combined to make 1 L volume.Water in the containers coming under miscellaneous category, leaf axis, and tree holes were not analysed due to the low water volumes obtained.Minimum of 3 one liter samples from each container type were analysed.
Six water quality parameters were tested; that is, pH readings were measured with a pH meter (Eutech Instrument, PC 700), free ammonia concentration was measured using Lovibone Nessterizer with Lovibond ® water testing, total alkalinity and chloride concentration were determined using titrimetric methods according to the standard procedures given by the Sri Lanka Standard Institution [21], TDS content was measured using TDS meter (Bench type conductivity, Eutech Instrument), and total iron concentration was determined using a spectrophotometer (UV/VIS Spectrodirect-German).

Data Analysis.
For each breeding habitat, three indices were calculated based on the three container grouping categories.Te Index of Aailable Containers (IACs) was calculated as the total number of containers of a particular type from each category divided by the total number of containers in the residences.Te Index of Contribution to Breeding Sites (ICBSs) was calculated as the number of positive containers from each category divided by the total number of positive containers in the residences.Te breeding preference ratio (BPR) was calculated as the ratio of ICBS to IAC for each category [22].A value less than one indicates that the category is not attractive to female mosquitos, whereas values more than one indicate that the category is exploited.A value of one would suggest that the particular container category is used in the same proportion as it is available [22].
Te Pearson correlation and linear regression analysis was utilized to determine the container variables related to the abundance of Ae. aegypti and Ae.albopictus larvae.Oneway ANOVA and Tukey's pairwise comparison tests were used to make comparisons.Variation of each of the physicochemical characteristics between breeding sites was determined by one-way ANOVA.Pearson's correlation coefcient (r) analysis was used to explore the association between physicochemical parameters and mosquito breeding preference.Multivariate studies, which included principal component analysis (PCA) and cluster analysis, were carried out using XLSTAT Version 2012.2.03 to investigate the link between each variable and the group based on similarity levels.

Prevalence of Containers and Teir Positivity for Dengue
Vector Larvae.During the period of January 2019 to December 2021, 14,452 premises were investigated from all three study sites.Highest number of water-holding containers (wet containers) were reported from the urban area (n � 10,578; 54.73%) followed by the semiurban (n � 4,424; 23.07%) and rural (n � 4,232; 22.2%) areas.Dengue vector larval stages were found in 1043 (5.4%) of the water-holding containers.Ae. aegypti was reported only from the urban area, whereas Ae. albopictus was found from urban, semiurban, and rural areas.Highest positivity was reported from the urban area (n � 603; Ae. aegypti only � 114; Ae. albopictus only � 477; both species � 12), followed by semiurban areas (n � 247; Ae. albopictus only) and rural areas (n � 193; Ae. albopictus only).Overall mixed breeding percentage was 0.03%.

Prevalence according to the Type of Container.
Index of Available Containers (IACs) for diferent container types is shown in Figure 3. On average, highest IAC was reported for discarded nondegradable items (n = 4423, 23.01%) followed by clay pots (n = 3269, 16.91%).In urban areas, discarded nondegradable items had the highest IAC, which was signifcantly higher (df = 14, F = 11.23,p � 0.002) than that reported for other two areas (Figure 3).In semiurban areas, highest IAC was for clay pots and it was signifcantly higher (df = 14, F = 13.23,p � 0.02) than that for other areas.In rural areas, water storage barrels had the highest IAC which was signifcantly higher (df = 14, F = 14.67, p � 0.003) than that reported for other two areas (Figure 3).Larval density (larvae per 50 ml volume of water) of all container types signifcantly correlated with their respective IAC values, and the association was stronger for urban Ae.albopictus (p � 0.001, r = 0.78  6 Journal of Tropical Medicine reported for discarded nondegradable items (urban 29.32 ± 1.57, semiurban 28.5 ± 5.7, rural 24 ± 3.5) (Figure 3).For both vector species, larval density of all container types signifcantly correlated (p < 0.05) with their respective ICBS values (for urbanAe.aegypti, r = 0.831; for urbanAe.albopictus, r = 0.933; for semiurban, r = 0.922; and for rural, r = 0.874).

Prevalence according to the Type of Material.
Plastic was the most available material type of wet containers in all three areas (Figure 4 4).Larval density signifcantly correlated with ICBS for all material types for all three study area Ae. albopictus (p � 0.001, r � 0.90) but not for urban Ae.aegypti.

Prevalence according to the Place of Container.
Indoor IAC was signifcantly lower than the outdoor IAC (df � 1, F � 12.3.p � 0.001) (Figure 5) for each area.No signifcance diference was observed among the indoor IACs and among the outdoor IACs in all three areas.Ae. albopictus larval density signifcantly correlated (p < 0.05) with respective indoor IAC values in semiurban and rural areas (r > 0.9).

Breeding Preference Ratio for Dengue Vectors in Urban,
Semiurban, and Rural Areas.For Ae. aegypti, the most preferred breeding container types were gutters (BPR � 3.89) and concrete slabs (BPR � 3.86).For Ae. albopictus, the most preferred were concrete slabs (BPR � 1.98) in urban areas and tyres in semiurban (BPR � 4.56) and rural (BPR � 1.66) areas.Tyres and discarded nondegradable items had >1 BPR value in all study sites for Ae.albopictus (Table 1).
With regard to the material types of the containers, rubber (BPR � 2.25) followed by concrete (BPR � 1.43) had high BPR value for Ae.aegypti.For Ae. albopictus, ceramic (urban � 1.34; semiurban � 2.40; rural � 1.04) and rubber (urban � 1.48; semiurban � 3.64; rural � 1.97) had >1 BPR values in all study sites.Although the plastic material type was the most available in all study sites, plastic had <1 BPR in all study areas except for urban Ae.albopictus (Table 1).Breeding preference ratio (BPR) was >1 only for urban indoor breeding habitats.In rural areas, outdoor breeding habitats had >1 BPR for Ae.albopictus (Table 1).

Water Quality Parameters.
Water quality analysis results showed that TDS content of the breeding sites ranged from 60 (±12.7)mg/L in commodes and cisterns to 1547 ± 374 mg/L in tyres.Free ammonia level ranged from 0.08 (±0.001) mg/L in bamboo stumps to 0.271 (±0.24) mg/L in discarded degradable habitats.Total alkalinity range was 38 (±13.9)mg/L in commodes and cisterns to 182.8 (±36.2) mg/L in tyres.Total iron ranged from 0.04 (±0.02) mg/L in clay pots to 0.56 (±0.14) in gutters.Te highest chloride concentration was from AC refrigerators (632 ± 528 mg/L) whereas bamboo stumps, covering items, discarded items, gutters, and slabs reported zero chloride concentrations.Dengue vector larvae were found in a wide pH range (6.7 ± 0.23 to 9.4 ± 0.5) (Table 2).
According to material type of containers, highest TDS, pH, free ammonia, and total alkalinity were from rubber.Highest iron content was from concrete breeding habitats.Highest chloride content was from plastic and rubber (Table 3).
Principal component analysis (PCA), using the physicochemical parameters of diferent breeding habitats according to the type and the material of containers, was carried out.A biplot analysis was used to visualize water variable correlations with container categorization (Figure 6).Results showed that water storage barrels, slabs, ornamental fower plots, tyres, and nonused commodes signifcantly and positively correlate with TDS.Gutters, bamboo stumps, covering items, discarded nondegradable, discarded-degradable items, and clay plots signifcantly associated with both TDS and alkalinity.AC refrigerators signifcantly associated with chloride concentration (Figure 6(a)).Tree major clusters of containers were obtained based on the agglomerative hierarchical clustering (AHC) technique (Figure 6(b)).
All the container types, except AC refrigerators, and all the material types, except glass, signifcantly and positively correlated with TDS and alkalinity of water (Figures 6(a

Discussion
Dengue fever (DF) has emerged as a serious public health concern in Sri Lanka, with an alarming increase in the number of reported cases.Development of resistance against synthetic insecticides has become a serious global issue threating insecticide-based vector control programmes [23,24].Considering the risk of resistance, cost efectiveness, environmental acceptance, and long-term infuence, dengue  Journal of Tropical Medicine vector control eforts in Sri Lanka are primarily focused on larval source reduction [25][26][27].Terefore, establishing an extensive knowledge on breeding environments and vector preference for breeding places has become crucial.In Sri Lanka, several research studies have been carried out on the availability and positivity of Ae. aegypti and Ae.albopictus breeding habitats and their characteristics [25,[28][29][30][31][32].Te current study focused on diferent aspects of mosquito breeding habitats including habitat preference, chemical nature of the breeding water, and their association with larval density which are important for risk assessment and to develop efective vector control strategies [2].Te study design was characterized by its diverse breeding habitat exploration and attention to sample size in order to provide a comprehensive understanding of dengue mosquito breeding preferences.
Our survey results from rural, semiurban, and urban environments revealed that Ae. albopictus is the predominant species in all three areas and Ae.aegypti is restricted to urban areas.Prevalence of the arboviral vector Ae.aegypti in urban areas has been attributed to its dependence on human dwellings for blood meals [33,34].Higa et al. [35] demonstrated that Ae. albopictus was more associated with natural habitats (e.g., tree holes, bamboo stumps, and bromeliads) and considered it a rural vector.Our results also showed that Ae. albopictus prefers outdoor breeding.However, this species has now become adapted to urban environment also, breeding in artifcial containers to become the predominant vector in urban environment [36][37][38].Replacement of Ae. aegypti with Ae. albopictus in urban environments has been observed in many countries throughout the world [36][37][38][39].Most probable reason for this replacement is the successful competitiveness of Ae. albopictus over Ae. aegypti for breeding habitats [40,41].
Te present study demonstrated that discarded nondegradable items were the most frequent and mostly positive breeding sites in all the study areas, confrming previous reports available in the literature [10,42].Larval productivity increases with the availability of the waste containers, and thus appropriate measures should be taken for waste management.Asian productivity organization [43] explained that waste generation is linked with socioeconomic factors, which are expected to difer between urban and rural communities.Although life style patterns difer between urban and rural areas, it appears that the generation of waste amount in rural areas difers quantitatively but not qualitatively from that in urban areas.Terefore, positivity varied across the three study sites, with discarded nondegradable objects being more positive in the urban than the other two.Waste collection is considerably lower in rural regions than in urban where rapid population expansion, industrialisation, urbanization, and increased consumptions take place.Dharmasiri and

Journal of Tropical Medicine
Dharmasiri [44] identifed several challenges, including inefcient waste segregation, poor waste collection mechanisms, and lack of public commitment on waste management in urban areas.Tus, the prevailing system of waste collection, transportation, and disposal is believed to be an issue that needs to be resolved.In this context, awareness through education and changing the attitudes of the public can help to establish proper and sustainable waste management practices.
Current study results highlight the abundance of discarded clay pots in semiurban areas.Many of these pots are used to pack curd commercially, and once the curd is consumed, pots are discarded to the environment by consumers.Tese empty clay pots are used for variety of purposes, including providing drinking water for pet animals, and subsequently turn in to mosquito breeding grounds.Water storage barrels were identifed as major mosquito breeding habitats in rural areas.Lack of an adequate pipe borne water delivery system has compelled the people live in rural areas to store water in these barrels.Public awareness and education about vector breeding environment and proper water delivery system to rural areas will reduce vector breeding incidence.Plastic was identifed as the material which largely contributes for diverse types of breeding habitats.Legislations should be introduced to minimise the use of plastic materials as a part of vector control strategies.Although rubber did not show a signifcantly higher prevalence, it was the most preferred breeding material as it contributes to tyres.Low prevalence could positively associate with breeding preference if the breeding habitat contributes to a high larval density [12,45].
Although previous studies had specifed that TDS levels of Aedes breeding habitats are low [27,[46][47][48], our results revealed that dengue vectors can survive in a broad range of TDS and the level of TDS positively correlates with BPR of Ae. aegypti and Ae.albopictus.Although the larvae were found in a wide range of pH, it has been reported that sites with high pH due to free ammonia are not ideal for mosquito breeding and survival, and a neutral pH range from 6.8 to 7.2 at breeding sites is preferred by mosquitoes [49].Surviving in a wide range of TDS levels, pH and chloride concentrations, may refect the adaptive nature of two vector species.Brackish water tolerance of both Ae.aegypti and Ae.albopictus has also been reported previously [50,51].Wang et al. [52] suggested that human interventions such as organic and nutrient pollution make a major impact on the water quality of mosquito breeding sites.However, diferences in water quality could also be linked to the nature of the usage of the container, natural, artifcial, or material source [53,54].
Te current investigation demonstrated the most available container types and their relationship with water quality metrics, and vector breeding preference in urban, semiurban, and rural settings in a district where dengue incidence is high.Our data provide valuable information to formulate proper waste management plans, public education, and awareness programmes for an efective vector control.

Conclusions
Ae. albopictus is the predominant vector found in all three urban, semiurban, and rural areas while Ae.aegypti is limited to urban areas.Discarded nondegradable items were the most prevalent container type, and plastics were the prominent material type in all study sites.Although the prevalence was low, tyres were the highest preferred breeding site for both species.Both vectors were present in a wide variety of water quality conditions showing their high adaptability.Information gathered can be used to formulate successful waste management plans, public education programmes, and efective vector control practices.

Figure 1 :
Figure 1: Map of the study locations in the Kurunegala district, Sri Lanka.Breeding habitats of Ae. aegypti and Ae.albopictus were studied in three localities (Bandarnayakapura, Galgamuwa town area, and Buluwala).

Figure 3 :
Figure3: Container availability (given by Index of Available Containers (IACs)) and their positivity (given by Index of Contribution to Breeding Sites (ICBSs)) for dengue vector species (A: Ae. aegypti; B: Ae. albopictus) for the urban, semiurban, and rural study areas.Diferent letters indicate signifcant diferences (p < 0.05) among the habitats (a-d) and among the groups (A-E lines above the groups) by analysis of variance (ANOVA).WSBs: water storage barrels, WSC tanks: water storage cement tanks, CSs: concrete slabs, OFPs: ornamental fower pots, THs: tree holes, leaf axis: LA, AC: air conditioning refrigerators, CIs: covering items, D-d: discarded degradable, D-non: discarded nondegradable, CPs: clay pots, C and C: commodes and cisterns, BSs: bamboo stumps, and M: miscellaneous.
) and 6(c)).Two major clusters were obtained based on the Journal of Tropical Medicine similarity of water quality of both the container type and based on the similarity of water quality of the container material (Figures 6(b) and 6(d)).AC refrigerators (container type) and glass (container material) gave smaller clusters with isolation showing their distinctive water quality profles compared to others.According to the Pearson correlation coefcient analysis carried out, Aedes larval density significantly associates (p < 0.05) with TDS (r � 0.31), alkalinity (r � 0.38), and total iron concentration (r � 0.16).

Figure 4 :
Figure 4: Container availability (IAC) according to the material type and their positivity (ICBS) for dengue vector species (A: Ae. aegypti; B: Ae. albopictus) in the urban, semiurban, and rural study sites in Kurunegala district.Diferent letters indicate signifcant diferences (p < 0.05) among the habitats (a-d) and among the groups (A-D lines above the groups) by analysis of variance (ANOVA).

Figure 5 :
Figure 5: Container availability (IAC) according to place of containers and their positivity (ICBS) for dengue vector species (A: Ae. aegypti; B: Ae. albopictus) for the selected urban, semiurban, and rural areas in Kurunegala district.Diferent letters indicate signifcant diferences (p < 0.05) among the habitats (a-c) and among the groups (A-B lines above the groups) by analysis of variance (ANOVA).

Figure 6 :
Figure 6: Principal components analysis (PCA) and agglomerative hierarchical clustering/dendrogram.(a, c) Correlation between water quality parameters and various container types and materials.(b, d) Classifcation of distinct container based on agglomerative hierarchical clustering (AHC) methodologies (clustering dendrogram) on the container type and the material of the containers.

Table 1 :
Correlation between larval densities of dengue vectors (A: Ae. aegypti and B: Ae. albopictus) and breeding preference ratio (BPR) according to container categories in urban, semiurban, and rural study sites of Kurunegala district.

Table 2 :
Water quality parameters of dengue vector mosquito breeding container types (mean ± SE).Diferent letters in the same column indicate signifcant diferences according to the Kruskal-Wallis H test.

Table 3 :
Water quality parameters of dengue vector mosquito breeding containers according to the material type of containers (mean ± SE).Diferent letters in the same column indicate signifcant diferences between diferent breeding habitats in each water quality parameter according to the Kruskal-Wallis H test.