J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 351 http://jad.tums.ac.ir Published Online: December 25, 2018 Original Article Wing Geometry Analysis of Aedes aegypti (Diptera, Culicidae), a Dengue Virus Vector, from Multiple Geographical Locations of Samut Songkhram, Thailand *Tanawat Chaiphongpachara 1, Nattapon Juijayen 2, Kitthisak Khlaeo Chansukh 3 1College of Allied Health Sciences, Suan Sunandha Rajabhat University, Thailand 2Bachelor of Public Health, College of Allied Health Sciences, Suan Sunandha Rajabhat University, Thailand 3Department of Applied Thai Traditional Medicine, College of Allied Health Sciences, Suan Sunandha Rajabhat University, Thailand (Received 21 Feb 2017; accepted 21 July 2018) Abstract Background: Dengue Haemorrhagic Fever (DHF) is a mosquito-borne disease and remains a major public health problem, especially in tropical and temperate countries. Studying wing morphometric of Aedes aegypti as a mosquito vector of DHF can help to better understand biological process of the mosquito adaptation to the environment. We aimed to study the geometric morphometric of Ae. aegypti from multiple geographical areas. Methods: Samples were collected from Samut Songkhram Province in Thailand, including coastal, residential and cultivated areas, by Ovitrap once per month during Oct to Nov 2016. Results: According to size variation analysis of Ae. aegypti, the female mosquito in a cultivated area was significant- ly different from those in the coastal and residential areas (P< 0.05). Whereas male Ae. aegypti in a cultivated area were significantly different from those in a residential area (P< 0.05). The shape variation of both female and male Ae. aegypti from all areas was statistically different (P< 0.05). Conclusion: Normally, living organisms, including mosquitoes, are adapted to their environment. The studied geo- graphical locations affect Ae. aegypti morphology. Keywords: Aedes aegypti, Mosquito vector, Geometric morphometric, Dengue hemorrhagic fever Introduction Dengue Haemorrhagic Fever (DHF) is a mosquito-borne disease and major public health problem in several countries worldwide (1). Ae- des aegypti (Diptera, Culicidae) is the primary mosquito vector of DHF that carries and trans- mits dengue virus to humans. Moreover, Ae. aegypti is closely associated with humans (it is active in the daytime and prefers resting in and around human houses) (2). In addition, Ae. aegypti can be resistant to chemical insecti- cides (3). The combination of these factors is associated with outbreaks of DHF. The geo- graphical environment affects the size and shape of mosquitoes (4). The geographical is relating to the landscape of the earth, where it has an environmental and habitat associated with mosquito vector (5). Thailand is currently experiencing a disease epidemic. Regarding the current DHF situa- tion in Thailand, the Bureau of Vector Borne Disease at the Department of Disease Control (within the Ministry of Public Health) has com- piled retrospective data for recent years. Be- tween 2011 and 2013, the DHF incidences rate were 100.789, 116.24, and 234.86 cases per 100000 population, the number of DHF pa- tients is increasing year on year, and the disease remains a major public health problem in Thai- land. Samut Songkhram is the smallest prov- ince in Thailand. Samut Songkhram is among those provinces with the most DHF cases (671 DHF patients [345.54 per 100000 population] *Corresponding author: Dr Tanawat Chaiphongpachara, E-mail: tanawat.ch@ssru.ac.th elham.jahani56@gmail.com J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 352 http://jad.tums.ac.ir Published Online: December 25, 2018 in 2015). This province consists of a coastal area influenced by tides and having diverse coastal plants, a residential area (high popu- lation density), and a cultivated area (low pop- ulation density). These environments produce morphological variation in Ae. aegypti (6). Stud- ying the morphological variation of the mos- quito helps us have a better understanding of causes, factors, and biological process of the mosquito’s adjustment to the environment (7). Recently, transmission of Dengue virus by Ae. aegypti in any geographic area depends on many factors, including extrinsic features re- lated to the environment and intrinsic factors as- sociated with the virus and vector interaction (6). Therefore, to reduce the presence of the vector, it is necessary to study the variability of mosquitoes as a basis for medical ento- mology. Geometric morphometric (GM) is a newly developed morphometric technique for anal- ysis of shapes and sizes of organisms using the principles of geometry (8). Many reports have applied GM to medical mosquitoes (9). GM can also be applied to study morphological varia- tions of organisms and analyze the evolution of wing morphometric (10). As with Aedes spp., the GM technique was used to identify three species in Thailand (Ae. aegypti, Ae. albopictus and Ae. scutellaris), found to be highly capable (11). Molecular techniques are popular and powerful tools for the identification of spe- cies (12). PCR-based methods are often used for species identification because of their sen- sitivity, reliability, and specificity for species identification (13). In addition, molecular meth- ods are also commonly used to study genetic variation across different areas (14). Although high-efficiency molecular techniques can use for species identification and genetic variation of mosquito, these are expensive and require specialized training (9). GM is an attractive al- ternative approach to identification and varia- tion because it is cheap, easy, and fast. We aimed to study the GM to investigate the impact of geography on Ae. aegypti from multiple geographical areas of Samut Songkhram Province, including coastal, res- idential and cultivated areas. Materials and Methods Mosquito collection Larvae of Ae. aegypti were collected in three different geographical locations across Samut Songkhram Province (coastal, residen- tial, and cultivated sites) by ‘ovitrap’ once per month during Oct to Nov 2016. The geograph- ical locations were selected by spatial data on land utilization information in Samut Songkhram from the Department of Provin- cial Administration of Thailand. The coastal area was Bang Cha Kreng Subdistrict (13° 23'31.57"N 100°1'59.36"E), the residential area was Mae Klong Subdistrict (13°24'32.52" N 100°0'41.40"E), and the cultivated area was Jompluak Subdistrict (13°28'23.7"N 99°55' 05.1"E) (Fig. 1). One village in each of the study areas reporting the highest dengue cas- es was selected, according to data from the report of Samut Songkhram provincial Bu- reau of Epidemiology in 2015. Six ovitraps (one trap per house) were set around houses or spaces under houses in each geographical location. Field collected larvae were then reared in the laboratory of College of Allied Health Sciences (Suan Sunandha Rajabhat Uni- versity, Thailand). Emerged adults were mor- phologically identified by illustrated keys to the mosquitoes of Thailand (15). Mosquito preparation Only the right wing of Ae. aegypti was analyzed. The right wings were dissected and mounted on microscope slides with coverslips using Hoyer solution. All of the sample wings were photographed using a Nikon DS-Ri1 SIGHT digital camera connected to a Nikon AZ 100M stereo-microscope (Nikon Corpo- ration, Tokyo, Japan). The images were ana- lyzed using the Morphometric CLIC Program. J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 353 http://jad.tums.ac.ir Published Online: December 25, 2018 Geometric morphometric Fourteen landmarks were digitized (Fig. 2). For these selected landmarks, there is a se- lection criterion that must be clearly visible in order to prevent mistakes when plotting. The precision and measurement error of images was estimated by the ‘repeatability’(R) index (7, 8). Ten images of mosquito wing in each geographical location were randomly chosen for R testing and repeatability of plotting land- mark from random images. After that, both two sets of images were computed as 1-R, with R, which is a Model II one-way ANOVA on re- peated measures (7, 8). The landmark-based GM analyzed the size and shape of the mos- quito. For wing size estimation, the size was measured by estimating the centroid size (CS) defined as the square root of the sum of the squared distances between the center of the configuration of landmarks and each individ- ual landmark (11). Statistical differences be- tween the centroid sizes of the male and fe- male mosquito wings from the different areas were analyzed by non-parametric permutation tests (1000 cycles). For wing shape evaluation, shape variables were measured and analyzed by principal com- ponents (PCs) of the “partial wrap” scores cal- culated after generalized Procrustes analysis of raw coordinates. Wing shape differences be- tween geographical locations were calculat- ed based on the Mahalanobis distance. Simi- lar to wing size estimation, statistical differ- ences in wing shape were analyzed by non-par- ametric permutation tests (1000 cycles). Neigh- bor-joining (NJ) trees were generated and cal- culated based on Procrustes distances between Ae. aegypti from different locations. Validat- ed reclassification accuracies were estimated for testing variation of Ae. aegypti in each geography location yielded by GM, each in- dividual was reclassified by comparing the shape based on the Mahalanobis distances. Software Data collection and analyses were performed using the various modules of the CLIC ver- sion 97 (Collecting Landmarks for Identifi- cation and Characterization), which is freely available at http://mome-clic.com (7). The fol- lowing modules used COO for landmark col- lection, TET for the transformation of data to be analyzed, MOG for centroid size and shape variables analysis to compute Procrustes dis- tances, PAD for statistical significance anal- ysis of shape variables and to compute Ma- halanobis distances, and VAR for statistical significance analysis of size variables. Results By applying the GM technique, we ana- lyzed 220 samples comprised of 103 female and 117 male mosquitoes. According to ge- ographical classification, 68 analyzed samples were from a coastal area, 82 samples from a residential area, and 70 samples from a culti- vated area (Table 1). For repeatability, com- parison of two sets of repeated measurements for the same images of Ae. aegypti wing for GM testing showed good scores (0.994). Size variation Size variation of Aedes aegypti wings was analyzed from the centroid size average of the wings from the different areas. Classified by sex, female mosquitoes in the residential ar- ea had the highest average (2.06vs 2.04 and 1.95mm) in the coastal and cultivated areas respectively. Similarly, male mosquitoes in the residential area had the highest average (1.62 vs 1.57 and 1.54mm) in coastal and cultivat- ed areas respectively (Table 2). Female mos- quitoes in the cultivated area had significant smaller wing size than those in the coastal and residential area (P< 0.05). Male mosquitoes from the cultivated area had smaller wing size than those from the residential area (P< 0.05) (Table 4). Shape variation After superimposition of the mean land- J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 354 http://jad.tums.ac.ir Published Online: December 25, 2018 mark configuration of the wings of the males and females from the different areas, polygons as connected mean landmark positions were demonstrated. Both polygons as connected mean landmark configurations of female and male mosquito in each different environmen- tal types were clearly different (Fig. 3). Both factor map of female and male Ae. aegypti from landmark-based discriminant analysis by partial wrap showed overlapping of wing shapes in each different geographical type (Fig. 4). Mahalanobis distances of female Ae. ae- gypti wing shapes in cultivated and coastal areas had the highest value (3.80). Similarly, male Ae. aegypti wing shapes in the cultivat- ed and coastal areas had the highest value (both 2.63) (Table 3). Tested by non-parametric per- mutation tests (1000 cycles), there was a sta- tistical difference (P< 0.05) in both male and female of Ae. aegypti mosquitoes in each geographical location (Table 4). The neigh- bor-joining trees, based on the Mahalanobis distances between PCs, separated each sex and each location (Fig. 5). The validated reclassification accuracy scores confirmed separation from shape dif- ferences of Ae. aegypti in each geography location, males and females are slightly dif- ferent. Reclassification scores of male mos- quitoes were 80% to 89%, while these scores for female mosquitoes were 84% to 90%. The highest percentage of reclassification of male mosquito showed of 89% for the culti- vated area, as well as the highest percentages of female mosquito was 90% of cultivated area. The lowest percentages of reclassifica- tion were 73% (male mosquito from the res- idential area) and 75% (female mosquito from the residential area) (Table 5). Table 1. The number of Aedes aegypti used for analysis to classify by sex and geography Sex Number of Aedes aegypti Total Coastal Area Residential Area Cultivated Area Female 103 32 40 31 Male 117 36 42 39 Total 220 68 82 70 Table 2. Means of wing centroid size of Aedes aegypti classified by sex and geography Sex Geography n Means±SD (mm.) Range (Min-Max) Female Coastal Area 32 2.04±0.11 1.93-2.15 Residential Area 40 2.06±0.21 1.85-2.27 Cultivated Area 31 1.95±0.16 1.79-2.11 Male Coastal Area 36 1.57±0.18 1.39-1.75 Residential Area 42 1.62±0.15 1.47-1.77 Cultivated Area 39 1.54±0.14 1.40-1.68 n= Number of Aedes aegypti Table 3. Mahalanobis distances between wing shapes of Aedes aegypti classified by sex and geography Females Males Geography Coastal Area Residential Area Cultivated Area Coastal Area Residential Area Cultivated Area Coastal Area 0.00 0.00 Residential Area 2.53 0.00 2.06 0.00 Cultivated Area 3.80 3.13 0.00 2.63 1.79 0.00 J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 355 http://jad.tums.ac.ir Published Online: December 25, 2018 Table 4. Statistical significance of size and shape differences of Aedes aegypti by non-parametric permutation tests (1000 cycles) Females Males Geography Coastal Area Residential Area Cultivated Area Coastal Area Residential Area Cultivated Area Size Coastal Area 0.00 0.00 Residential Area 0.66 0.00 0.15 0.00 Cultivated Area 0.01* 0.00* 0.00 0.35 0.02* 0.00 Shape Coastal Area 0.00 0.00 Residential Area 0.00* 0.00 0.00* 0.00 Cultivated Area 0.00* 0.04* 0.00 0.02* 0.00* 0.00 *Statistically significant (P< 0.05) Table 5. Validated reclassification accuracies of male and female of Ae. aegypti in each geography location Geography Percentage of reclassification Male Female Coastal Area 80% (29/36) 84% (27/32) Residential Area 73% (31/42) 75% (30/40) Cultivated Area 89% (35/39) 90% (28/31) Fig. 1. Map of Aedes aegypti collection sites in different geographical locations (1= Bang Cha Kreng Subdistrict as Coastal area, 2= Mae Klong Subdistrict as residential area, and 3= Jompluak Subdistrict as cultivated area) J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 356 http://jad.tums.ac.ir Published Online: December 25, 2018 Fig. 2. Aedes aegypti wing showing the 14 landmarks used in the morphometrics analysis Fig. 3. Superimposition of the mean landmark configurations of Aedes aegypti in different areas [Coastal area (blue), Residential area (red), Cultivated area (green)]. Top: females of Ae. aegypti, bottom: males of Ae. aegypti Fig. 4. Factor map from landmark-based discriminant analysis by partial wrap for females (A) and males (B) of Ae- des aegypti classified by geographical locations J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 357 http://jad.tums.ac.ir Published Online: December 25, 2018 Fig. 5. Neighbor-joining trees for shape based on GM analyses of male and female Aedes aegypti from different geographical locations Discussion Here, GM was used to study the wing mor- phometric of Ae. aegypti in Samut Songkhram Province, Thailand. Two-hundred-twenty Ae. aegypti mosquitoes were collected from three different geographical sites. Size variation of mosquito Females of Ae. aegypti have a larger cen- troid wing size than males, which is in line with other studies (10). Moreover, we detected a statistical difference in size variation of the fe- male Ae. aegypti. The mosquitoes in the cul- tivated area were significantly different from those in the coastal and residential areas (P< 0.05). In addition, male Ae. aegypti were sig- nificantly different from those in the residen- tial area (P< 0.05). The difference between Ae. aegypti in the cultivated area and the other areas might be a result of environmental fac- tors, such as temperature, food quality and quantity, and suitable habitat (16). In fact, we did not study the environmental factors that affect Ae. aegypti. However, population den- sity differences and the number of households in the cultivated area might have contributed to our findings. Other studies have addressed the relationships between population and house- hold number and Ae. aegypti numbers (2). Ur- ban areas are more suitable for Ae. aegypti than rural and sub-rural areas (17). Residen- tial areas support the viability and breeding of Ae. aegypti, the mosquitoes can easily feed on human blood within houses, and there are many water containers (breeding sites for the DHF vector) (2). High nutrient levels produce larger mosquitoes (18). However, here we did not detect significant differences between the coastal and residential areas for either male or female Ae. aegypti. The might be because the coastal area of Samut Songkhram is a tour- ist attraction and, therefore, similar to the res- idential area (tourists plus water containers provide food and breeding sites). Mosquitoes from the residential area were largest (female, 2.06±0.21, male, 1.62±0.15). In 2015, the Samut Songkhram Provincial Bu- reau of Epidemiology reported that this resi- dential area produced more DHF patients than the other two tested areas. The body size of a female mosquito is correlated with fecundi- ty, larger females lay more eggs during the first gonotrophic cycle (18). Moreover, male mosquito size correlates with total sperm num- bers within a male and the number transferred to females (19). Thus, we would expect a rel- atively high density of DHF vector in the res- idential area, which is a factor of disease risk in areas of dengue virus transmission (20). Shape variation of mosquito By visualized factor mapping of the land- mark-based discriminant analysis (partial wrap), we detected separated and overlapping areas in the wing shape morphospace of male and fe- male Ae. aegypti classified geographical loca- tion. After being tested with non-parametric per- mutation tests (1000 cycles), we found that the males and females wing shapes were statisti- cally different in all geographical areas (P< 0.05), likely because of environmental factors J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 358 http://jad.tums.ac.ir Published Online: December 25, 2018 (e.g., wind current and weather) (21). In each geographical location of Samut Songkhram Province, there were different environmental factors affecting Ae. aegypti shape, for ex- ample, storms and wind in the coastal area. The size and shape of Ae. aegypti and Ae. albopictus were examined from Nakhon Nayok Province of Thailand, Cucuta municipality of Colombia, and Florida and Hawaii in the Unit- ed States and detected significant geographic differentiation (10). Although we found no statistical difference in size variation in some areas, we detected shape variation across all groups. This might be because the mosquitos are adapted to their noticeably different environments. Geography has an impact on Aedes mosquito in each ge- ographical location (22). Mosquito wing shape is correlated with mosquito population den- sity and food quality. Studying these varia- tions helps to better understand how Ae. ae- gypti adapts to its environment. In addition, such morphology data is important for tax- onomy studies (11). Samut Songkhram is a small province. Therefore, our ability to detect difference here suggests that GM is a very effective technique for studies addressing mosquito variation. Our validated reclassification accuracies scores demonstrate the efficiency of the separation in each geographical location. Thus, GM should be considered an alternative method for stud- ying the variation of mosquito vectors. Conclusion GM it is a newly-developed morphomet- ric technique used to classify medical insect types and study morphology variants (8). GM is advantageous because it is easy to use, is low-cost, and rapid. GM does not require high entomological skills (7). GM is an effective tool for studying morphology variants of Ae. aegypti. Aedes aegypti from different locations differ in size and shape, which is important for understanding its local adaptation. Such knowledge might help to control mosquitoes in DHF endemic areas. Acknowledgements We would like to thank College of Allied Health Science, Suan Sunandha Rajabhat Uni- versity, Thailand for their kind support of our research. This work was supported by Suan Sunandha Rajabhat University, Bangkok, Thai- land. The authors declare that there is no con- flict of interests. References 1. Kyle JL, Harris E (2008) Global spread and persistence of dengue. Annu Rev Microbiol. 62: 71–92. 2. Rodrigues M de M, Marques GRAM, Serpa LLN, Arduino M de B, Voltolini JC, Barbosa GL, Andrade VR, de Lima VLC (2015) Density of Aedes aegypti and Aedes albopictus and its associa- tion with number of residents and me- teorological variables in the home en- vironment of dengue endemic area, São Paulo, Brazil. Parasit Vectors. 8(1): 115. 3. Ishak IH, Jaal Z, Ranson H, Wondji CS (2015) Contrasting patterns of insecticide resistance and knockdown resistance (kdr) in the dengue vectors Aedes ae- gypti and Aedes albopictus from Ma- laysia. Parasit Vectors. 8(1): 181. 4. Raharimalala FN, Ravaomanarivo LH, Ravelonandro P, Rafarasoa LS, Zouache K, Tran-Van V, Mousson L, Failloux AB, Hellard E, Moro CV, Ralisoa BO, Mavingui P (2012) Biogeography of the two major arbovirus mosquito vectors, Aedes aegypti and Aedes albopictus (Dip- tera, Culicidae), in Madagascar. Parasit Vectors. 5: 56. 5. Ferraguti M, Martínez-de la Puente J, Roiz J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 359 http://jad.tums.ac.ir Published Online: December 25, 2018 D, Ruiz S, Soriguer R, Figuerola J (2016) Effects of landscape anthropization on mosquito community composition and abundance. Sci Rep. 6: 29002. 6. Hidalgo K, Dujardin JP, Mouline K, Dabiré RK, Renault D, Simard F (2015) Sea- sonal variation in wing size and shape between geographic populations of the malaria vector, Anopheles coluzzii in Burkina Faso (West Africa). Acta Trop. 143: 79–88. 7. Dujardin JP (2011) Modern morphometrics of medically important insects. Genet Evol Infect Dis. 473–501. 8. Dujardin JP (2008) Morphometrics applied to medical entomology. Infect Genet Evol. 8(6): 875–90. 9. Jaramillo-O N, Dujardin JP, Calle-Londoño D, Fonseca-González I (2015) Geomet- ric morphometrics for the taxonomy of 11 species of Anopheles (Nyssorhynchus) mosquitoes. Med Vet Entomol. 29: 26– 36. 10. Henry A, Thongsripong P, Fonseca-Gon- zalez I, Jaramillo-Ocampo N, Dujardin JP (2010) Wing shape of dengue vectors from around the world. Infect Genet Evol. 10(2): 207–214. 11. Sumruayphol S, Apiwathnasorn C, Ruang- sittichai J, Sriwichai P, Attrapadung S, Samung Y, Dujardin JP (2016) DNA barcoding and wing morphometrics to distinguish three Aedes vectors in Thai- land. Acta Trop. 159: 1–10. 12. Ma Y, Li S, Xu J (2006) Molecular iden- tification and phylogeny of the Macula- tus group of Anopheles mosquitoes (Dip- tera: Culicidae) based on nuclear and mi- tochondrial DNA sequences. Acta Trop. 99: 272–280. 13. Walton C, Somboon P, O’Loughlin SM, Zhang S, Harbach RE, Linton YM, Chen B, Nolan K, Duong S, Fong MY, Vythilingum I, Mohammed ZD, Trung HD, Butlin RK (2007) Genetic diversi- ty and molecular identification of mos- quito species in the Anopheles macula- tus group using the ITS2 region of rDNA. Infect Genet Evol. 7: 93–102. 14. Manni M, Gomulski LM, Aketarawong N, Tait G, Scolari F, Somboon P, Gug- lielmino CR, Malacrida AR, Gasperi G (2015) Molecular markers for analyses of intraspecific genetic diversity in the Asian Tiger mosquito, Aedes albopic- tus. Parasit Vectors. 8: 188. 15. Rattanarithikul R, Harrison BA, Panthusiri P, Coleman RE (2005) Illustrated keys to the mosquitoes of Thailand. I. Back- ground, geographic distribution, lists of genera, subgenera, and species, and a key to the genera. Southeast Asian J Trop Med Public Health. 36(SUPPL. 1): 1–80. 16. Ruangsittichai J, Apiwathnasorn C, Dujar- din JP (2011) Interspecific and sexual shape variation in the filariasis vectors Mansonia dives and Ma. bonneae. In- fect Genet Evol. 11(8): 2089–2094. 17. Valdez MRWD (2017) Mosquito species distribution across urban, suburban, and semi-rural residences in San Antonio, Texas. J Vector Ecol. 42: 1–5. 18. Briegel H (1990) Metabolic relationship be- tween female body size, reserves, and fe- cundity of Aedes aegypti. J Insect Phys- iol. 36(3): 165–172. 19. De Jesus CE, Reiskind MH (2016) The importance of male body size on sperm uptake and usage, and female fecundity in Aedes aegypti and Aedes albopictus. Parasit Vectors. 9(1): 447. 20. Chaiphongpachara T, Pimsuka S, Saisanan Na Ayudhaya W, Wassanasompong W (2017) The application of geographic in- formation system in dengue haemor- rhagic fever risk assessment in Samut Songkhram Province, Thailand. Int J GEOMATE. 12(30): 53–60. 21. Motoki MT, Suesdek L, Bergo ES, Sallum MAM (2012) Wing geometry of Anoph- eles darlingi Root (Diptera: Culicidae) in five major Brazilian ecoregions. In- J Arthropod-Borne Dis, December 2018, 12(4): 351–360 T Chaiphongpachara et al.: Wing Geometry Analysis … 360 http://jad.tums.ac.ir Published Online: December 25, 2018 fect Genet Evol. 12(6): 1246–1252. 22. Morales Vargas RE, Ya-umphan P, Phu- mala-Morales N, Komalamisra N, Dujar- din JP (2010) Climate associated size and shape changes in Aedes aegypti (Dip- tera: Culicidae) populations from Thai- land. Infect Genet Evol. 10(4): 580–585.