See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/274248575

Morphological and molecular characterization of the ecological, biological and

behavioural variants of the JE vector Culex tritaeniorhynchus: An assessment
of its taxonomic status

Article in Journal of Vector Borne Diseases · March 2015

CITATIONS READS

10 606

6 authors, including:

Aladu Ramakrishnan Rajavel N Pradeep Kumar

Indian Council of Medical Research Indian Council of Medical Research
51 PUBLICATIONS 1,064 CITATIONS 211 PUBLICATIONS 1,525 CITATIONS

SEE PROFILE SEE PROFILE

Ramalingam Natarajan Perumal Vanamail

35 PUBLICATIONS 645 CITATIONS

SRM Institutes for Medical Science
193 PUBLICATIONS 3,590 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by N Pradeep Kumar on 23 July 2015.

The user has requested enhancement of the downloaded file.

J Vector Borne Dis 52, March 2015, pp. 40–51

Morphological and molecular characterization of the ecological, biological

and behavioural variants of the JE vector Culex tritaeniorhynchus: An
assessment of its taxonomic status

A.R. Rajavel, N. Pradeep Kumar, R. Natarajan, P. Vanamail, A. Rathinakumar & P. Jambulingam

Vector Control Research Centre (ICMR), Indira Nagar, Puducherry, India

ABSTRACT

Background & objectives: Culex tritaeniorhynchus (Diptera: Culicidae), an important vector of Japanese
encephalitis belongs to the Culex vishnui subgroup which includes two other vector species namely, Cx. vishnui
and Cx. pseudovishnui. Many varieties and types of Cx. tritaeniorhynchus have been reported, besides populations
that exhibit behavioural and biological differences. This study was undertaken to find out whether Cx.
tritaeniorhynchus populations exhibiting behavioural and biological variations, and those from different
geographical areas, are comprised of more than one taxon or belong to a single taxon.
Methods: Morphological characterization was done by examining 153 morphological and morphometric characters
in the larval (75), pupal (60) and adult stages (18) of five geographical populations of Cx. tritaeniorhynchus.
Molecular characterization was done by PCR amplification of mitochondrial cytochrome c oxidase (COI) gene
sequences (DNA barcodes) and another hypervariable genetic marker, the ribosomal DNA (16S). One-way ANOVA,
principal component analysis (PCA) and discriminant factor analysis (DFA) were done for statistical analyses
using the statistical package SPSS IBM version 19.0.
Results: Morphological characterization showed that no intraspecific differentiation can be made among the five
geographical populations of Cx. tritaeniorhynchus. Molecular characterization done by DNA barcoding also showed
that the COI sequences of all the five populations of Cx. tritaeniorhynchus grouped into a single taxonomic clade
plus the genetic differentiation among these was non-significant and the overall gene flow among the populations
was very high. Analysis of the ribosomal DNA also confirmed that the Cx. tritaeniorhynchus populations belonged
to a single taxon.
Interpretation & conclusion: Culex tritaeniorhynchus is a taxon that does not involve cryptic species.
Key words Culex tritaeniorhynchus; Culex vishnui subgroup; DNA barcode; morphology; taxonomic status

INTRODUCTION tions of a taxon which have been subsequently identified

as sibling species in the case of anophelines such as
Japanese encephalitis (JE) is a major public health Anopheles fluviatilis and An. subpictus Grassi, 1899
problem in Southeast Asia. The mosquito vectors involved (Diptera: Culicidae). These observations reported in the
in the transmission of this viral infection mainly belong case of Cx. tritaeniorhynchus added a new dimension to
to the Culex vishnui subgroup which includes the three the already complex nature of the members of the sub-
important species, namely Cx. tritaeniorhynchus Giles, group whose identification as adults also remains diffi-
1901, Cx. vishnui Theobald, 1901 and Cx. pseudovishnui cult. These circumstances warranted a morphological and
Colless, 1957 ( Diptera: Culicidae)1. Larvae of these spe- molecular characterization of the geographically and eco-
cies are known to occur in fresh water habitats such as logically variant populations of the important JE vector
paddy fields, ground pools, ponds, wells and ditches while Cx. tritaeniorhynchus of this subgroup.
adults prefer to rest outdoors2. However, in India, Cx.
tritaeniorhynchus was reported to rest predominantly in- MATERIAL & METHODS
doors in Bellary district of Karnataka state3. Yet another Study area
report from Mysore indicated that larvae of Cx. Geographical locations in Bellary district, Mysore and
tritaeniorhynchus inhabiting ground pools varied in some Mandya districts in Karnataka, where biological and
of their morphological characters from those inhabiting behavioural variations have been reported in Cx.
paddy fields4. Such behavioural variations have been in- tritaeniorhynchus formed the primary study sites. JE
dicative of the presence of genetically isolated popula- prone areas such as Warangal and Karim Nagar districts
 Rajavel et al: Taxonomic status of Culex tritaeniorhynchus 41

in Telangana state (erstwhile Andhra Pradesh state) and Mosquito collection in Mysore and Mandya was
areas of repeated epidemic outbreaks of JE such as mainly focused on larval collections because morphologi-
Gorakhpur district of Uttar Pradesh state were the other cal variations have been reported in larvae of JE vector
study sites. All these study sites are known to be erst- from these areas. Larval collection was done in Mysore
while JE epidemic prone areas. In addition, samples from City from Kukarhalli tank area. The vast area adjacent to
West Bengal, Tamil Nadu and Assam were also included the tank was a marsh grassy region with large puddles of
for molecular characterization. water stagnation that formed the larval habitats. Another
Two villages within the Bellary Developmental location of larval collection in the city was the
Block, namely Rupanagudi (15o 03.209' N, 77o 03.738' Vidyaranyapuram farm where fodder grass is cultivated.
E) and Moka (15o 14.027' N, 77o 03.622' E) were selected Larval collection outside the city area was done in the
for mosquito collections because of the reported paddy fields, irrigation channels, ditches and ponds in
behavioural variation of resting indoors (endophilism) in both Mandya and Mysore. Indoor and outdoor resting
Cx. tritaeniorhynchus 3. Mysore (12 o 18.841' N, 76 o collection of adults was made in Chikpalya village in
37.740' E) and Mandya (12o 30.359' N, 76o 50.024' E) Mandya.
districts are the sites where biological variation in the form Adult and larval collections of Cx. vishnui subgroup
of differences in the morphological structures of larvae mosquitoes were made in Warangal and Karimnagar dis-
of Cx. tritaeniorhynchus has been reported4. Chalapalli tricts of Andhra Pradesh during August 2010 and in
(18o 15.061' N, 79o 27.106' E)and Rajapalli (18o 13.744' Gorakhpur district of Uttar Pradesh during September
N, 79o 26.073' E) villages of Chelpur Primary Health Cen- 2010, since peak densities of Cx. tritaeniorhynchus have
tre (PHC), Uppalpalli village (18o 10.774' N, 79o 28.535' been reported during the respective months in these study
E) of Uppal PHC and Medipalli village (18o 03.242' N, sites. Adults were collected from both indoor and out-
79o 27.135' E) of Hasanparthy PHC were the mosquito door resting habitats and larval collections were done in
collection sites in Warangal and Karimnagar districts. paddy fields, ground pools and irrigation channels.
Mosquito collection sites in Gorakhpur were Mahesra (26o Voucher specimens of mosquitoes collected from all the
49.305' N, 83o 21.243' E), Rampurkhud (26o 50.350' N, study areas have been deposited in the mosquito museum
83o 26.183' E), Sekurwa (26o 49.006' N, 83o 20.834' E) at the Vector Control Research Centre, Puducherry5–6.
(Chargawan PHC), Karimnagar (Nagar Nigam) (26 o
48.364' N, 83o 23.042' E), Kusmitakia (26o 44.868' N, Morphological characterization
83o 20.834' E), Shivpur (26o 39.918' N, 83o 29.899' E) Specimens obtained from different study areas formed
(Pipraich PHC) and Rudrapur (26o 39.269' N, 83o 15.009' five different geographical populations, viz. (i) Bellary,
E) (Khajani PHC). (ii) Mysore City, (iii) Mysore outskirts and Mandya, (iv)
Warangal and Karimnagar, and (v) Gorakhpur. Morpho-
Collection of mosquitoes logical characterization was based on the study of larvae,
Field work was carried out during February to pupae, adult females and males of Cx. tritaeniorhynchus
November 2009 in study localities of Bellary, Mysore of all the five populations. Majority of the specimens ex-
and Mandya districts for collection of mosquitoes of the amined were those obtained through associated rearing,
Culex vishnui subgroup. Adult collections from both in- so that the adult could be linked to its larval and pupal
door and outdoor resting sites were made in the morning skins.
between 0900 and 1000 hrs and in the evening between Larva: There are about 186 pairs of seta found in
1700 and 1800 hrs using an oral aspirator and torch light. mosquito larva besides the comb scales, pecten and anal
In Moka and Rupanagudi villages of Bellary district, apart papillae, of which around 48 pairs are commonly used in
from exclusive human dwellings and cattlesheds, most different keys for identification of species. In this study,
of the houses had the cattle together with the inmates shar- 75 characters including length, width and ratios were used,
ing the dwelling, which were considered as mixed dwell- for which 197 IV instar larvae and larval skins were ex-
ings. One man hour collection of indoor resting mosqui- amined.
toes was done in each of these types. Collection of outdoor Pupa: About 118 pairs of seta are present in the mos-
resting mosquitoes was done in the paddy fields, in the quito pupa, of which around 53 pairs of seta, besides the
vegetation along irrigation channels, hay stacks in the trumpet and paddle are commonly used in keys for spe-
open, and cotton fields using an oral aspirator. Larval cies identification. In this study, 60 characters which in-
collection was done in the villages, from the paddy fields, cluded the number, length, width and ratios were exam-
irrigation channels and ditches by using a dipper. ined in 130 specimens.
42 J Vector Borne Dis 52, March 2015

Female: In the case of adult females, six characters, DNA Polymerase, 5 μl of 10× PCR Buffer, 2.5 mM Mag-
namely colour of erect scales on vertex, anterior 0.7 of nesium chloride, 10 μl Q solution, 1 μl of dNTP mixture
mesonotum (dark or pale brown), mesonotum scales (fine and 1 μl of 10 pmol of forward and reverse DNA primers
or coarse), accessory pale scales on the ventral side of 5'-GGATTTGGAAATTGATTAGTTCCTT-3' and 5'-
proboscis (present or absent), plume scales on wing (nar- AAAAATTTTAATTCCAGTTGGAACAGC-3' along
row or broad), and dark scaling in hind femur were ex- with 2 μl of template DNA7. The reaction parameters were
amined in 1821 specimens. Since, the cibarial characters as follows: Initial denaturation of 5 min (95°C) followed
are generally used in differentiating closely related spe- by 5 cycles of 94°C for 40 sec (denaturation), 45°C for 1
cies of genus Culex, the cibarial armature was dissected min (annealing), and 72°C for 1 min (extension) and 35
and mounted for 29 specimens and five cibarial charac- cycles of 94°C for 40 sec (denaturation), 51°C for 1 min
ters, namely cibarial width and length of larger dorsal (annealing), 72°C for 1 min (extension), and a final ex-
papillae, number of smaller dorsal papillae, length and tension at 72°C for 10 min. The amplified fragments were
number of cibarial teeth were also examined. run on a 1% agarose gel to check the integrity of the frag-
Male: In the case of adult males, seven characters, ments and the PCR product was purified by
namely length of proboscis in relation to palpal segment QIAGENGmbH PCR purification kit. The purified prod-
5, dorsal apical pale band on palpal segment 2, dorsal ucts were eluted to 20 μl of deionized water, and a por-
median pale band on palpal segment 3, ventrolateral tufts tion of it was lyophilized in a Speed Vac concentrator
on palpal segment 3, setae on ventral surface of probos- (Thermo Electron Corporation, Waltham, MA) followed
cis, distal setae (d-f) of subapical lobe and number of fin- by custom sequencing. Both reads (from forward primer
ger like processes on phallosome were examined in 385 as well as reverse primer) were done, and the sequences
specimens. were analyzed as follows. The DNA sequences were sub-
jected to alignment using ClustalW. Sequence divergences
Molecular characterization among individuals were quantified by using the Kimura
Cytochrome c oxidase 1 gene: Altogether, 143 speci- two-parameter (K2P) distance model8. A neighbour join-
mens of Cx. tritaeniorhynchus collected from different ing (NJ) tree of K2P distances was constructed to
parts of the country (Karnataka, Kerala, Andhra Pradesh, provide a graphic representation of the clustering
West Bengal, Tamil Nadu, Assam and Uttar Pradesh) were pattern among different species9-10. These analyses of
subjected to PCR amplification of mitochondrial cyto- the sequences were conducted using MEGA version 3.1
chrome c oxidase (COI) gene sequences (DNA barcodes), software11.
a genetic marker established to be species specific, in or-
der to delineate whether any cryptic species existed among Ribosomal DNA
Cx. triteaniorhynchus. This included 135 specimens from Another hypervariable genetic marker, the ribosomal
the study areas (Bellary 36, Mysore City 22, Mysore out- DNA (16S) was analysed to find out whether the popula-
skirts 25, Warangal 29 and Gorakhpur 23), besides eight tions of Cx. tritaeniorhynchus from the five different study
specimens of Cx. tritaeniorhynchus collected in Kerala, areas exhibit genetic variation. Altogether, 55 specimens
West Bengal, Tamil Nadu and Assam. Also eight speci- of Cx. tritaeniorhynchus collected from the five study ar-
mens of Cx. pseudovishnui and six specimens of Cx. eas were subjected to rDNA analysis. The extracted DNA
vishnui which are the other two species of the Cx. vishnui from the specimens was subjected to PCR using the fol-
subgroup, collected from the study areas were included lowing conditions. The DNA forward and reverse prim-
in the analysis. ers used were 5'- CGCCTGTTTATCAAAAACAT-3' and
About 600 bp of the mitochondrial cytochrome c oxi- 5'-CTCCGGTTTGAACTCAGATC-3' which amplify
dase gene was PCR amplified following the protocol de- about 500 bp of hypervariable ribosomal DNA12. The 50
scribed by Kumar et al7. Total DNA was extracted from μl reactions set up included 5 μl of 10× Buffer, 10 μl of Q
3 legs of the specimens using Genelute Mammalian DNA solution, 2 μl of MgCl2, 1 μl dNTP mixture of 10 mM
Miniprep kit (Sigma-Aldrich, U.S.A). The extracted DNA each, 1.5 U of Taq polymerase, 0.2 pmol of forward and
was dissolved in 30 μl of deionized water. 5' COI region reverse primers and 2 μl of template DNA (Taq PCR core
of mitochondrial DNA was amplified in Eppendorff kit, Qiagen, Germany). The reaction was carried out in
Mastercycler Gradient S (Germany) using following cy- the Eppendorff Mastercycler Gradient S and the PCR
cling conditions. The 50 μl reaction was set up using conditions were as follows. Initial denaturation of 4 min
Qiagen Taq PCR Core kit (Qiagen, Germany). The reac- at 95°C followed by 35 cycles of 50 sec denaturation step
tion components included 1.5 U of Thermus aquaticus of 94°C; annealing of 1 min at 55°C and the extension of
 Rajavel et al: Taxonomic status of Culex tritaeniorhynchus 43

1 min at 72°C. The final elongation step provided was tal of 4048 adult specimens were obtained as study mate-
72°C for 7 min. rial. Of these, 387 specimens were collected as adults and
342 adults emerged from larval collections were mounted
Statistical analysis on minuten while the others were preserved in pill boxes.
Data on different morphological characters examined From the larval collections done, 928 larvae, larval skins
were subjected to one-way ANOVA for assessing the dif- and pupal skins were mounted on slides.
ferences among the populations of Cx. tritaeniorhynchus In Warangal and Gorakhpur, Cx. tritaeniorhynchus,
and post-hoc test was used to make pair-wise compari- Cx. pseudovishnui and Cx. vishnui were obtained in adult
son of each characteristic using Bonferroni method. Prin- collections, with Cx. tritaeniorhynchus being the predomi-
cipal component analysis (PCA) was done to derive small nant species in both the areas. The per man hour density
number of factors, of many observed characteristics. (PMD) of resting Cx. tritaeniorhynchus in Warangal was
Number of factors were considered based on eigen value 30.22 in human dwellings, 88.67 in cattlesheds and 417.8
of >1. Discriminant factor analysis (DFA) was done to in outdoors and in Gorakhpur it was 28, 36.62 and 318.02
find the variables, which were most useful for discrimi- respectively. In both the areas, high density of the spe-
nating between individuals of different populations. In cies was found in outdoor habitats, indicating the exophilic
discriminant factor analysis, Pillai’s trace statistics were behaviour. Density of Cx. pseudovishnui and Cx. vishnui
performed to determine whether the clusters were sig- was very low in both the areas, being < 1 per man hour in
nificantly different from each other. Within-class covari- human dwellings and cattlesheds. In outdoor habitats,
ance, matrices were assumed to be equal and prior prob- PMD of Cx. pseudovishnui was 1.05 and 1.39 while that
abilities were taken into account while performing the of Cx. vishnui was 0.52 and 2.31 in Warangal and
discriminant analysis. For all statistical tests p <0.05 was Gorakhpur respectively. In Warangal, larvae of Cx.
considered statistical significance. All these analyses were tritaeniorhynchus were obtained mostly from paddy fields
performed using statistical package for social sciences, while in Gorakhpur it was from the ground pools and
SPSS IBM version 19.0. vacant plots. Very high density (250 per dip) of larvae of
For molecular characterization, the PCR reactions Cx. tritaeniorhynchus was observed in a vacant plot in
were ran on a 1.5% agarose gel. The amplified fragments Karimnagar village of Gorakhpur where JE case had been
were purified using Qia quick PCR purification kit recorded. A total of 830 adults of Cx. tritaeniorhynchus
(Qiagen, Germany) and sent for custom sequencing. The obtained from different collection sites in Warangal and
DNA sequences were aligned in ClustalW software and 860 adults from different collection sites in Gorakhpur
were subjected to NJ phylogenetic analysis in MEGA 4.0 were mounted on minuten, while the rest were preserved
software11. The DNA sequence polymorphism was ana- in pill boxes. From larval collections made, 200 and 260
lyzed in DNASP 5.013. larvae, larval skins and pupal skins from Warangal and
Gorakhpur were mounted on slides.
RESULTS
Morphological characterization
In Bellary, all the three species of the Cx. vishnui Larva: Of the 75 morphological characters examined
subgroup, Cx. tritaeniorhynchus, Cx. pseudovishnui and in larvae, 66 characters, excluding those which were simi-
Cx. vishnui were collected and Cx. tritaeniorhynchus was lar in all the populations, were considered for morpho-
found to be the predominant species. Among the indoor logical analysis of the five populations based on one-way
resting habitats, resting density of Cx. tritaeniorhynchus ANOVA. Among the 66 characters, only 35 were sig-
was in the order of cattleshed > mixed dwelling > human nificantly different in the one-way ANOVA (Table 1).
dwelling, with per man hour densities of 315, 273.43, The pair-wise comparison of morphological characters
and 41.11, respectively. The overall indoor resting den- using the Bonferroni test revealed that characters such as
sity of this species in Bellary was high (202.15 per man seta 1-A branch, 1-A length, antenna length, antenna width
hour) compared to the outdoor resting density (17 per at middle, antennal index, head/antenna width at point of
man hour). The density of Cx. pseudovishnui and Cx. attachment of 1-A ratio, 5-C branch, 5-C length, 6-C
vishnui was < 1 per man hour in all the resting habitats. length, 7-C length, 7-C branch, 9-C branch, 13-C branch,
In the larval collections done in Bellary, only Cx. 6-I length, 7-I length, 1-III length, 1-IV length, 1-V
tritaeniorhynchus was obtained. In larval collections made length, 1-VI length, 6-IV length, 6-VI length, number of
in Mysore and Mandya, larvae of Cx. tritaeniorhynchus, comb teeth, number of denticles in the apical pecten teeth,
Cx. vishnui, and Cx. pseudovishnui were obtained. A to- siphon length, siphon width at middle, siphon width at
44 J Vector Borne Dis 52, March 2015

Table 1. Morphological analysis of larvae of Culex tritaeniorhynchus of different geographical populations using one-way ANOVA test

Character Bellary Mysore City Mysore Warangal Gorakhpur F p-value

Mean (SD) Mean (SD) outskirts Mean (SD) Mean (SD) <0.05
Mean (SD)

Seta 1-A branch 28.1 (3.16) 27.7 (2.13) 28.1 (2.58) 28.4 (1.97) 26.4 (1.94) 2.778 0.029
Seta 1-A length 0.22 (0.02) 0.22 (0.02) 0.25 (0.08) 0.23 (0.02) 0.21 (0.02) 4.947 0.001
Antenna length 0.56 (0.03) 0.56 (0.03) 0.56 (0.03) 0.54 (0.02) 0.53 (0.02) 8.887 0.0
Antenna width 0.07 (0.01) 0.07 (0.01) 0.08 (0.01) 0.08 (0.00) 0.08 (0.01) 4.469 0.002
Antenna index 7.87 (1.46) 7.83 (0.87) 7.56 (0.88) 6.82 (0.38) 6.99 (0.77) 8.436 0.0
Seta 1-A width 0.05 (0.0) 0.05 (0.0) 0.05 (0.0) 0.06 (0.0) 0.05 (0.0) 10.417 0.0
Head ratio 14.7 (1.12) 14.4 (1.28) 14.2 (1.20) 12.6 (0.94) 13.7 (1.24) 9.736 0.0
Seta 5-C branch 3.24 (0.48) 3.40 (0.45) 3.20 (0.42) 3.34 (0.54) 3.08 (0.37) 2.924 0.022
Seta 5-C length 0.49 (0.04) 0.51 (0.04) 0.52 (0.05) 0.46 (0.03) 0.47 (0.03) 13.219 0.0
Seta 6-C length 0.56 (0.04) 0.59 (0.04) 0.59 (0.05) 0.52 (0.03) 0.53 (0.03) 19.459 0.0
Seta 7-C branch 7.83 (0.98) 8.20 (1.08) 7.62 (1.05) 8.52 (1.01) 8.20 (0.97) 3.771 0.006
Seta 7-C length 0.51 (0.03) 0.53 (0.03) 0.53 (0.04) 0.50 (0.03) 0.49 (0.02) 10.205 0.0
Seta 13-C branch 2.97 (0.67) 3.03 (0.30) 2.99 (0.31) 3.04 (0.27) 3.31 (0.51) 2.929 0.022
Seta 13-T branch 8.12 (1.24) 8.32 (1.63) 7.91 (1.41) 8.98 (1.40) 8.83 (1.71) 2.922 0.022
Seta 6-I length 0.76 (0.06) 0.78 (0.05) 0.76 (0.06) 0.74 (0.04) 0.73 (0.05) 4.289 0.002
Seta 7-I length 0.57 (0.04) 0.59 (0.04) 0.56 (0.04) 0.53 (0.04) 0.55 (0.04) 13.191 0.0
Seta 1-III length 0.24 (0.03) 0.23 (0.02) 0.23 (0.04) 0.24 (0.02) 0.21 (0.03) 4.296 0.004
Seta 1-IV length 0.30 (0.03) 0.31 (0.02) 0.30 (0.03) 0.30 (0.03) 0.26 (0.03) 11.458 0.0
Seta 1-V length 0.35 (0.04) 0.38 (0.02) 0.34 (0.04) 0.35 (0.03) 0.33 (0.03) 4.473 0.003
Seta 1-VI length 0.41 (0.04) 0.44 (0.03) 0.39 (0.04) 0.40 (0.03) 0.37 (0.05) 6.137 0.0
Seta 6-IV length 0.48 (0.04) 0.49 (0.06) 0.49 (0.06) 0.46 (0.03) 0.46 (0.04) 3.021 0.019
Seta 6-V length 0.60 (0.04) 0.59 (0.06) 0.58 (0.06) 0.56 (0.05) 0.56 (0.04) 3.649 0.007
Seta 6-VI length 0.67 (0.05) 0.68 (0.07) 0.67 (0.06) 0.64 (0.05) 0.65 (0.04) 2.585 0.039
Comb scales 35.7 (3.29) 36.6 (4.34) 33.7 (4.20) 34.8 (3.27) 34.0 (2.12) 4.043 0.004
No. of denticles in the 7.88 (1.09) 7.70 (1.45) 7.94 (1.31) 6.98 (1.0) 6.73 (0.99) 8.58 0.0
distal pecten teeth
Siphon length 1.79 (0.19) 1.81 (0.19) 1.81 (0.17) 1.71 (0.14) 1.66 (0.13) 5.724 0.0
Siphon width at base 0.29 (0.02) 0.28 (0.02) 0.29 (0.02) 0.29 (0.01) 0.29 (0.02) 3.341 0.011
Siphon width at middle 0.19 (0.03) 0.18 (0.02) 0.19 (0.02) 0.19 (0.01) 0.20 (0.02) 3.394 0.010
Siphon index at base 6.28 (0.82) 6.57 (0.61) 6.37 (0.67) 5.97 (0.59) 5.83 (0.75) 7.512 0.0
Siphon index at middle 9.53 (1.97) 9.94 (1.51) 9.39 (1.38) 8.84 (1.03) 8.41 (1.33) 6.232 0.0
Saddle length 0.33 (0.02) 0.32 (0.03) 0.33 (0.02) 0.32 (0.01) 0.33 (0.01) 5.431 0.0
Saddle width 0.38 (0.02) 0.35 (0.03) 0.36 (0.07) 0.36 (0.07) 0.37 (0.02) 2.510 0.043
Siphon/Saddle ratio 5.38 (0.48) 5.73 (0.46) 5.44 (0.49) 5.39 (0.43) 5.04 (0.39) 12.196 0.0
Anal papillae length 0.26 (0.06) 0.26 (0.05) 0.33 (0.08) 0.31 (0.04) 0.36 (0.05) 10.135 0.0
Anal papillae width 0.06 (0.02) 0.06 (0.01) 0.06 (0.02) 0.07 (0.01) 0.07 (0.02) 3.356 0.015

F—Fishers’s test statistics based on one-way ANOVA.

base, siphon index at base, siphon index at middle, saddle factors explained 73.03% of the total variability in the
length, saddle width, siphon/saddle ratio and anal papil- data. The first factor (F1) explained 19.57% of the total
lae index were shorter in length and number of branches variability, while the second factor (F2) explained 11.28%.
were lesser in Gorakhpur and Warangal populations than Similarly, the third factor (F3) and fourth factor (F4) ex-
in Mysore City, Mysore outskirts and Bellary populations. plained 6.05% and 5.13% of the total variability, respec-
However, between Gorakhpur and Warangal populations, tively. Altogether, the first four factors explained 42.02%
the Warangal population had longer/higher number of of the total variability. Characters such as seta 6-V length,
branches. 6-IV length, 6-III length; 6-VI branch, 6-I length, 5-C
length, antennal length, 6-C length, 7-I length and saddle
Multivariate analysis of the larval populations width showed high magnitude on F1. Siphon index at base,
Principal component analysis (PCA) of the morpho- siphon/saddle ratio, siphon index at middle, siphon length,
logical analysis extracted five factors with eigen-values siphon width at base, and siphon width at middle showed
of more than one. The Kaiser-Meyer-Olkin (KMO) mea- high magnitude on F2. Seta 13-T branch, number of den-
sure of sampling adequacy was 0.566. Cumulatively, these ticles on apical pecten teeth, number of pecten teeth, seta
 Rajavel et al: Taxonomic status of Culex tritaeniorhynchus 45

9-C branch, and seta 7-C branch showed high magnitude Larval habitat-wise analysis
on F3. Antenna width and antennal index showed high Besides, analysis of the morphological characters of
magnitude on F4. DFA could distinguish five significant different geographical populations, analysis was done
clusters for Mysore City, Mysore outskirts, Bellary, based on larvae obtained from different larval habitats.
Warangal and Gorakhpur populations (Pillai’s trace = The mean values of the characters considered for mor-
0.923, F = 1.57, p <0.012). Characters that showed high phological analysis showed significant difference among
factor loading on F1 included length of setae 6-I, 6-III, 6- the seven habitats, namely ground pools, irrigation ca-
IV, 6-V, 5-C, 6-C, 7-I, antenna width and saddle width, nals, paddy fields, rock pools, seepages, sewage and
while characters that showed high factor loading on F2 swamps, based on one-way ANOVA. Among the 66 char-
included only siphon length, siphon index at middle, si- acters, only 18 were found to be significantly different
phon index at base, and siphon/saddle ratio. (Table 2). The pair-wise comparison of morphological
characters using Bonferroni test revealed that characters
Analysis based on variables related to length and indices such as antenna length, siphon length, siphon index at
Principal component analysis was also done by con- base, siphon index at middle, and siphon/saddle ratio were
sidering only the variables related to length and indices. shorter in ground pool than irrigation canals, paddy fields,
The KMO measure of sampling adequacy was 0.672. rock pools, seepages, sewage and swamps; length of seta
Cumulatively, these factors explained 76.16% of the total 5-C and 7-C shorter in ground pools, irrigation canals
variability in the data. The first factor (F1) explained and paddy fields than rock pools, seepages, swamps and
30.31% of the total variability, while the second factor sewages; length of seta 6-C shorter in ground pools and
(F2) explained 16.96% of the total variability; paddy fields than irrigation canals, seepages, sewage and
together the first two factors explained 47.27% of the swamps; number of comb scales higher in sewage and
total variability. Characters such as length of seta 6-III, 6- swamps than ground pools, irrigation canals and paddy
IV, 6-V, 6-I, 7-I, antenna length, saddle width, and siphon fields.
width at base showed high magnitude on F1. Siphon index
at base, siphon/saddle ratio, siphon index at middle, and Pupa
siphon length showed high magnitude on F2. The analysis Of the 60 morphological characters examined in pu-
showed that increase in length is not specific to siphonal pae, 42 characters, excluding those which were similar in
characters alone but also for other body setae. all the populations, were considered for morphological
analysis of the five populations based on one-way
Analysis based on variables related to number ANOVA. Among the 42 characters only 25 were signifi-
Principal component analysis by considering only the cantly different in the one-way ANOVA (Table 3). The
variables related to number of branches of setae was also pair-wise comparison of morphological characters using
performed. The KMO measure of sampling adequacy was Bonferroni test showed that trumpet length of Gorakhpur
0.526. Cumulatively, these factors explained 68.21% of and Mysore City populations was similar and Mysore out-
the total variability in the data. The first factor (F1) ex- skirts and Warangal populations was similar while Bellary
plained 13.71% of the total variability, while the second population had trumpet length greater than all other popu-
factor (F2) explained 8.81% of the total variability, the lations. Trumpet width of Bellary and Warangal popula-
third factor (F3) explained 7.67%, the fourth factor (F4) tions was similar while other populations significantly
explained 6.91% and fifth factor (F5) explained 6.17%; differed from each other. Seta 5-CT branch number of
together the first five factors explained 43.27% of the to- Gorakhpur and Mysore outskirts populations was similar
tal variability. Characters such as number of branches of while other populations significantly differed from each
seta 13-T, seta 8-C, 9-C, number of pecten teeth and num- other. Paddle width of Bellary and Mysore outskirts popu-
ber of denticles on apical pecten teeth showed high mag- lations was similar while other populations significantly
nitude on F1. Branch number on seta 1-IV, III, and V differed from each other. Seta 9-VIII branch number of
showed high magnitude on F2. Branch number of seta Mysore City and Mysore outskirts populations was simi-
13-C and 10-C showed high magnitude on F3. Branch lar while other populations significantly differed from
number of seta 6-I, and 7-I showed high magnitude on each other. Seta 5-VII branch number of Gorakhpur and
F4. Branch number of seta 1-VI and 7-C showed high Bellary populations was similar while other populations
magnitude on F5. It showed there was no significant dif- significantly differed from each other.
ference among the five populations in relation to the num- Mysore City population differed significantly from
ber of branches of the various setae. Warangal population with respect to branch number of
46 J Vector Borne Dis 52, March 2015

Table 2. Morphological analysis of larvae of Culex tritaeniorhynchus from different larval habitats using one-way ANOVA test

Character Ground pool Irrigation canal Paddy field Rock pool Seepage Sewage Swamp F p-value
Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) <0.05

Antenna length 0.53 (0.02) 0.55 (0.03) 0.55 (0.02) 0.58 (0.02) 0.57 (0.02) 0.54 (0.01) 0.56 (0.03) 7.68 0.0
Antenna width 0.08 (0.007) 0.07 (0.008) 0.08 (0.008) 0.08 (0.004) 0.07 (0.007) 0.08 (0.006) 0.07 (0.008) 3.16 0.01
Antenna index 6.79 (0.63) 7.58 (0.80) 7.36 (1.26) 7.44 (0.34) 7.94 (0.81) 6.76 (0.43) 7.92 (0.83) 5.15 0.0
Seta 5-C length 0.47 (0.03) 0.49 (0.05) 0.49 (0.04) 0.53 (0.03) 0.52 (0.04) 0.50 (0.03) 0.52 (0.04) 5.73 0.0
Seta 6-C length 0.53 (0.03) 0.56 (0.05) 0.54 (0.04) 0.56 (0.03) 0.58 (0.04) 0.58 (0.04) 0.59 (0.04) 10.10 0.0
Seta 7-C length 0.49 (0.02) 0.50 (0.04) 0.51 (0.03) 0.54 (0.03) 0.53 (0.04) 0.52 (0.04) 0.53 (0.03) 7.45 0.0
Seta 13-C branch 3.36 (0.53) 3.1 (0.92) 2.95 (0.34) 3.08 (0.58) 3.03 (0.21) 3.20 (0.44) 3.02 (0.29) 2.69 0.02
Seta 6-I length 0.74 (0.04) 0.77 (0.07) 0.75 (0.05) 0.75 (0.05) 0.76 (0.06) 0.73 (0.05) 0.78 (0.05) 2.95 0.01
Seta 7-I branch 1.92 (0.23) 2 2 2 2 2.10 (0.22) 2.01 (0.07) 3.52 0.0
Seta 7-I length 0.54 (0.04) 0.57 (0.05) 0.55 (0.04) 0.57 (0.04) 0.57 (0.03) 0.56 (0.02) 0.59 (0.04) 7.28 0.0
Comb teeth 33.48 (2.03) 34.72 (3.16) 34.71 (3.72) 35.07 (2.30) 35.54 (3.61) 38.80 (3.33) 36.47 (4.43) 2.76 0.01
Siphon length 1.62 (0.14) 1.82 (0.22) 1.77 (0.15) 1.77 (0.2) 1.78 (0.17) 1.88 (0.18) 1.81 (0.19) 4.34 0.0
Siphon width at base 0.29 (0.02) 0.30 (0.02) 0.28 (0.02) 0.27 (0.01) 0.28 (0.01) 0.28 (0.02) 0.28 (0.02) 4.69 0.0
Siphon width at middle 0.21 (0.02) 0.20 (0.02) 0.19 (0.02) 0.19 (0.02) 0.19 (0.02) 0.21 (0.008) 0.18 (0.02) 5.64 0.0
Siphon index at base 5.57 (0.68) 6.17 (0.86) 6.23 (0.67) 6.55 (0.92) 6.43 (0.62) 6.75 (0.34) 6.55 (0.63) 7.06 0.0
Siphon index at middle 7.96 (0.99) 9.04 (1.88) 9.39 (1.60) 9.58 (1.71) 9.55 (1.08) 9.04 (0.75) 9.99 (1.54) 5.65 0.0
Saddle length 0.33 (0.01) 0.34 (0.03) 0.33 (0.02) 0.34 (0.03) 0.33 (0.02) 0.31 (0.03) 0.32 (0.03) 2.76 0.0
Siphon/Saddle ratio 4.91 (0.35 5.43 (0.52) 5.43 (0.43) 5.15 (0.32) 5.45 (0.48) 6.10 (0.24) 5.69 (0.46) 11.28 0.0

F—Fishers’s test statistics based on one-way ANOVA.

Table 3. Morphological analysis of pupae of Culex tritaeniorhynchus of different geographical populations using one-way ANOVA test

Character Bellary Gorakhpur Mysore City Mysore outskirts Warangal F p-value

Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) <0.05

Trumpet length 0.60 (0.047) 0.56 (0.03) 0.56 (0.04) 0.59 (0.07) 0.59 (0.05) 4.76 0.001
Trumpet width 0.12 (0.01) 0.13 (0.01) 0.10 (0.01) 0.11 (0.01) 0.12 (0.01) 12.12 0.0
Trumpet ratio 5.14 (0.41) 4.63 (0.46) 5.41 (0.59) 5.44 (0.78) 4.86 (0.39) 7.46 0.0
Seta 5-CT branch 4.86 (0.82) 4.45 (0.57) 4.57 (0.72) 4.43 (0.49) 5.25 (0.81) 4.84 0.001
Sect 6-CT branch 3.43 (0.71) 3.77 (0.79) 2.98 (0.61) 3.17 (0.68) 3.35 (0.72) 4.04 0.004
Seta 10-CT branch 7.21 (2.20) 7.38 (1.30) 6.13 (1.40) 6.42 (1.73) 8.09 (2.09) 4.89 0.001
Seta 12-CT branch 3.79 (0.56) 3.45 (0.50) 3.47 (0.53) 3.67 (0.73) 3.94 (0.58) 2.67 0.035
Seta 1-II branch 21.05 (4.96) 19.55 (6.97) 17.64 (4.48) 19.39 (3.95) 20.11 (3.44) 2.47 0.048
Seta 1-III branch 10.67 (1.96) 10.42 (2.49) 9.49 (1.85) 10.57 (1.77) 11.44 (2.08) 3.88 0.005
Seta 1-IV branch 8.69 (1.48) 9.45 (2.23) 8.07 (1.46) 8.52 (1.58) 9.32 (1.80) 2.94 0.023
Seta 1-V branch 7.75 (0.87) 8.00 (1.58) 7.22 (1.25) 7.31 (1.04) 8.69 (1.06) 6.13 0.0
Seta 1-VI branch 6.70 (0.67) 6.59 (1.30) 6.16 (1.12) 6.19 (0.97) 7.06 (0.95) 3.37 0.012
Seta 1-VII branch 4.65 (0.61) 4.67 (0.72) 4.38 (0.59) 4.25 (0.49) 5.03 (0.66) 5.75 0.0
Seta 5-IV branch 5.61 (0.94) 5.35 (0.58) 4.96 (0.84) 4.82 (0.52) 5.37 (0.91) 3.84 0.006
Seta 5-V branch 2.10 (0.41) 2 (0.0) 1.95 (0.18) 2 (0.15) 2.22 (0.43) 3.75 0.007
Seta 6-V branch 5.34 (0.53) 5.83 (0.81) 4.97 (0.67) 5.32 (0.78) 6.08 (0.73) 10.21 0.0
Seta 6-VI branch 5.45 (0.79) 5.88 (0.64) 5.15 (0.66) 5.26 (0.85) 6.03 (0.81) 5.76 0.0
Seta 7-III branch 5.72 (0.81) 6.33 (0.29) 5.05 (1.07) 5.07 (0.90) 5.53 (0.90) 3.12 0.018
Seta 9-VIII branch 6.95 (0.76) 6.64 (0.80) 6.44 (0.99) 6.45 (1.32) 7.60 (1.50) 3.55 0.01
Seta 4-VIII branch 2.61 (0.46) 2.63 (0.61) 2.33 (0.46) 2.52 (0.54) 2.71 (0.45) 2.89 0.025
Paddle width 0.49 (0.05) 0.51 (0.04) 0.45 (0.04) 0.49 (0.05) 0.50 (0.07) 5.7 0.0
Paddle ratio 1.46 (0.16) 1.49 (0.12) 1.60 (0.13) 1.48 (0.12) 1.45 (0.11) 7.73 0.0
Seta 5-VII branch 1.95 (0.21) 1.95 (0.15) 1.73 (0.42) 1.81 (0.36) 2.03 (0.26) 3.54 0.009
Seta 8-VI branch 3.82 (0.87) 4.25 (0.58) 3.40 (0.92) 3.83 (0.80) 4.66 (0.69) 8.61 0.0
Seta 8-VII branch 4.02 (0.75) 4.71 (1.25) 3.52 (0.69) 3.83 (0.70) 4.44 (0.66) 8.5 0.0

F—Fishers’s test statistics based on one-way ANOVA.

seta 12-CT, seta 1-III, seta 1-IV, seta1-VI, seta 5-CT, seta spect to trumpet width, branch number of seta 6-CT, seta
4-VIII, and seta 5-VII. In characters such as trumpet width, 6-V, seta 8-VI, seta 8-VII, and paddle width, Mysore City
branch number of seta 5-CT, seta 10-CT, seta 1-V, seta population was significantly different from Gorakhpur
1-VI, seta 6-V, seta 6-VI, seta 9-VIII, and seta 8-VI, both population. Paddle ratio of Mysore City population was
Mysore City and Mysore outskirts populations were sig- significantly different from all other populations. With
nificantly different from Warangal population. With re- respect to 22 characters related to number of branches,
 Rajavel et al: Taxonomic status of Culex tritaeniorhynchus 47

Mysore City, Mysore outskirts and Bellary populations generated along with an outgroup is provided (Fig. 1). In
were similar having lesser number than the Gorakhpur the analysis it was found that all the COI sequences (143)
and Warangal populations, but between the Gorakhpur of Cx. tritaeniorhynchus grouped into a single taxonomic
and Warangal populations, the Gorakhpur population had clade. Also, the other two species included in the study
lesser number. branched as separate two taxonomic clusters. The statisti-
cal significance of this branching pattern was found to be
Multivariate analysis of the pupal populations highly significant (the interior Branch test with 1000 rep-
Principal component analysis of the morphological lications showed the statistical significance of the clades
analysis extracted five factors with eigen-values of more to be 99% for Cx. tritaeniorhynchus; Cx. pseudovishnui and
than one. The KMO measure of Sampling Adequacy was Cx. vishnui) indicating these to be three distinct species. The
0.662. Cumulatively, these factors explained 67.18% of K2P genetic distance (d) recorded for these three species
the total variability in the data. The first factor (F1) ex- were found to be negligible (0.0073; 0.0080 and 0.0088
plained 31.64% of the total variability, while the second respectively) indicating the specimens included
factor (F2) explained 11.74% of the total variability; to- under each category to be of single taxonomic units. The
gether the first two factors explained 43.38% of the total interspecific ‘d’ was found to be 0.0768 between Cx.
variability. Characters such as branch number of seta tritaeniorhynchus and Cx. pseudovishnui; 0.0586 between
1-V, seta 1-VI, seta 1-IV, and seta 1-III showed high mag- Cx. tritaeniorhynchus and Cx. vishnui. The interspecific ge-
nitude on F1. Trumpet ratio and trumpet width showed netic distance between Cx. pseudovishnui and Cx. vishnui
high magnitude on F2. F1 was significantly different in was recorded to be 0.0591. These interspecific values also
all five populations, whereas F2 was different in all five denote these three taxonomic clades to be three distinct spe-
populations except the trumpet width, which was similar cies categories. Besides, the study clearly indicated the
in Bellary and Warangal populations. DFA could distin- utility of DNA barcode approach to distinguish these three
guish five significant clusters for Mysore City, Mysore morphologically similar species, which are important vec-
outskirts, Bellary, Warangal and Gorakhpur populations tors of Japanese encephalitis.
(Pillai’s trace = 0.447, F = 2.37, p <0.001). Characters Also, the five different populations of Cx. tritaenio-
that showed high factor loading on F1 included branch rhynchus from Bellary, Mysore City, Mysore outskirts,
number of seta 1-V, 1-VI, 1-IV, 1-VII, and 1-III, while Warangal and Gorakhpur were subjected to phylogenetic
characters that showed high factor loading on F2 included analysis to understand the genetic variability in geographi-
trumpet width and trumpet ratio. cally isolated populations. The K2P genetic distance (d)
values between the populations were found to be negli-
Female gible. The genetic differentiation estimated among these
Morphological characters examined in adult five populations (Fst = 0.00500) was found to be non-
females did not show any variation between the different significant (χ2 = 347.92; p = 0.6103) and the overall gene
populations. For most of these characters including that of flow (Nm) among the populations computed was found
the accessory pale scales on the ventral side of proboscis, to be very high (99.51). The maximum genetic differen-
over 90% of the specimens of Mysore City, Mysore out- tiation was recorded among Mysore outskirts and
skirts, Bellary, Warangal and Gorakhpur populations were Gorakhpur populations (Fst = 0.03996), which also re-
similar. For characters related to the cibarial armature, corded the least gene flow (Nm = 12.01). However, this
one-way ANOVA showed that only the cibarial width highest genetic differentiation value recorded among these
significantly (p <0.05) differed among the populations. populations was also not significant genetically (χ2 =
‘ 34.643; p = 0.5331). The above observations indicate
Male negligible genetic variability among the five populations
All the morphological characters examined in adult of Cx. tritaeniorhynchus.
males including the length of proboscis in relation to pal-
pal segment 5, and genital characters of distal setae (d-f) Ribosomal DNA (16S) analysis
of subapical lobe, and number of finger like process on The K2P genetic distance and mean nucleotide di-
phallosome, were similar in the Mysore City, Mysore out- versity (pi) recorded for 55 specimens analyzed were
skirts, Bellary, Warangal and Gorakhpur populations. found to be very negligible (0.0007 and 0.00067 respec-
tively). The phylogenetic tree using NJ analysis of K2P
Molecular characterization genetic distances is given (Fig. 2). Thus, this hypervariable
Cytochrome c oxidase 1 gene: The phylogenetic tree genetic marker also confirmed that Cx. tritaeniorhynchus
48 J Vector Borne Dis 52, March 2015

Fig. 2: Phylogenetic (Neighbour-joining) analysis of the hyper

variable ribosomal DNA sequences of Culex tritaeniorhynchus
from different study areas; BLR—Bellary; GKP—Gorakhpur;
MND—Mysore outskirts; MYS—Mysore City; WGL—
Warangal.

specimens collected and analyzed from different geo-

graphical localities in the study belonged to a single taxo-
nomic unit.

DISCUSSION

A total of 153 morphological and morphometric char-

acters in the larval (75), pupal (60) and adult stages (18)
of Cx. tritaeniorhynchus, were examined in this study to
Fig. 1: Phylogenetic (Neighbour-joining) analysis of mitochondrial find out whether Cx. tritaeniorhynchus populations
cytochrome C oxidase sequences of Culex vishnui subgroup
from different study areas; BLR—Bellary; GKP—Gorakhpur;
exhibit behavioural and biological variations, and those
MND—Mysore outskirts; MYS—Mysore City; WGL— from different geographical areas, are comprised of more
Warangal. than one taxon or belong to a single taxon.
 Rajavel et al: Taxonomic status of Culex tritaeniorhynchus 49

ANOVA results indicated that several morphologi- In another study, Cx. tritaeniorhynchus populations
cal characters differ significantly among the five popula- from Mysore City (ground pools) and outside Mysore
tions of Cx. tritaeniorhynchus studied. Larvae from (paddy fields) were differentiated as two varieties, based
Bellary, Mysore City, and Mysore outskirts had setae with on morphological variations in comb scales, pecten teeth,
lesser number of branches/shorter in length compared to denticles on the apical pecten tooth, siphon index, anal
those from Gorakhpur and Warangal. But no relationship gills index and the length of basal hair tufts on siphon4.
could be established in terms of geographical distribu- Mysore City (ground pools) population was named as va-
tion from the southern to northern region because riety one having shorter siphon index, shorter anal gills
Gorakhpur population had the lesser number/shorter index, less comb scales and longer basal hair tufts on si-
length compared to Warangal population. Principal com- phon, while outside Mysore (paddy fields) population was
ponent analysis (PCA) revealed that among the larval named as variety two having longer siphon index, longer
characters, variables related to the body hair accounted anal gills index, more comb scales and shorter basal hair
significantly for variations among the populations rather tufts on siphon. Analysis of the same two populations
than the variables related to the siphon and the saddle. collected by us also showed that ground pool larvae had
PCA done independently for variables related to number shorter siphon index, length of siphon shorter, less comb
and variables related to length and indices showed that scales and width of siphon at middle greater, and com-
there was no significant difference among the five popu- pared to larvae of paddy fields. But larvae collected in
lations in relation to the number of branches of the vari- swamp (Mysore City) showed longer siphon index, length
ous setae, but it differed significantly in relation to length of siphon longer, and more comb scales compared to lar-
and indices. In larval chaetotaxy, based on which larval vae in paddy fields indicating that characters related to
description of a species is done, emphasis is given to the length can be influenced by the habitats. Further, habitat-
branch number of various body setae rather than on the wise analysis of the larvae from all the five study areas
length of the setae, for the reason that the former is con- showed that, of the 18 variables found to be significantly
sidered as fixed characters, while the latter is subject to different between the various larval habitats, 16 variables
variability. In the discriminant factor analysis also, the are related to length and indices, which again substanti-
characters that were responsible for differentiating the ates that these are not reliable for differentiating the Cx.
five populations as five different clusters were those re- tritaeniorhynchus populations into varieties.
lated to length and indices of the setae and not the num- The range values of the different morphological char-
ber of branches. It is, therefore, evident that there is no acters of larvae recorded in our study, as well as the ear-
basis to consider that any of the five populations of lier studies4, 14, were compared with that of taxonomic
Cx. tritaeniorhynchus is different from the others. descriptions and larval chaetotaxy tables available for Cx.
Earlier reports on morphological variations in Cx. tritaeniorhynchus of Thailand, Oriental region, Japan and
tritaeniorhynchus were all based on very few characters Korea and South-western Asia2, 15-17. It was found that
examined in the larvae. Significant morphological differ- values obtained in our study to that of the other two
entiation in four populations collected from Bellary, earlier studies4, 14 were all within the range provided
Cuddalore, Pune and the Microbial Containment Com- by these authors, showing that any conclusion based
plex laboratory culture in India, based on siphon, saddle on variations in the larval characters in considering
and pecten teeth related variables has been reported14. Cx. tritaeniorhynchus as variant populations is not
However, all the 11 characters considered by them for reasonable.
analysis were only of length and indices related to siphon, In systematic studies it is important to include all life
saddle and anal gills. The present study has shown that stages because evidence of relationships cannot always
these variables are not reliable for differentiating the popu- be found readily in the same stage18–19. While earlier stud-
lations. This is also substantiated by the values obtained ies on morphological variations in Cx. tritaeniorhynchus
for the siphon index in the two studies, while the earlier have not considered the pupal stage, the present study is
study found the siphon index to be highest in Cuddalore the only one which fulfils this requirement, as the pupal
population ranging from 9.42–16 compared to other popu- stage of Cx. tritaeniorhynchus of different geographical
lations including that of Bellary, this study showed that populations has been extensively examined for morpho-
the siphon index of Bellary population ranged from 6.45– logical characters. Discriminant function analysis of the
16 indicating that Cuddalore population cannot be differ- characters distinguished five significant clusters for
entiated from Bellary population based on the siphon Mysore City, Mysore outskirts, Bellary, Warangal and
index. Gorakhpur populations, the characters contributing to the
50 J Vector Borne Dis 52, March 2015

variability being the branch number of seta 1-V, 1-VI, 1- the mounting of the specimen. It should also be noted that
IV, 1-VII, and 1-III. But for these characters also, Mysore summorosus, which was described as a subspecies of Cx.
City, Mysore outskirts and Bellary populations were simi- tritaeniorhynchus mainly based on this character has been
lar having lesser number than the Gorakhpur and synonymized with Cx. tritaeniorhynchus and is now not a
Warangal populations, while between the Gorakhpur and valid taxon24.
Warangal populations, the Gorakhpur population had The results of the morphological characterization done
lesser number. Further, it was seen that the values ob- in the present study by examining a total of 153 charac-
tained for different characters in our study were all within ters in the larva, pupa, female and male of Cx. tritaenio-
the range provided for pupal chaetotaxy of Cx. rhynchus of the Bellary, Mysore City, Mysore outskirts,
tritaeniorhynchus2, 17, 20. Thus, none of the five popula- Warangal and Gorakhpur populations showed that all
tions can be treated as specifically different from the oth- populations is of the same species and no intra-specific
ers. differentiation can be made. This is conclusively proved
In the adult stage, the females of the different popu- by the molecular characterization done by DNA barcoding
lations of Cx. tritaeniorhynchus did not show any varia- in which it was found that the COI sequences of all the
tion in the morphological characters studied. More than five populations of Cx. tritaeniorhynchus grouped into a
90% of the specimens examined were similar for these single taxonomic clade and the genetic differentiation
characters. Among the cibarial armature characters also, among these five populations was found to be non-sig-
many were similar, and though the cibarial width differed nificant, and the overall gene flow among the popula-
significantly among the populations, it was found that tions was found to be very high. Further, the ribosomal
Bellary and Warangal populations were similar for this DNA, which is a hypervariable genetic marker, also con-
character. Further, it was found that our results of the firmed that Cx. tritaeniorhynchus specimens collected and
cibarial characters were comparable to that of the base- analyzed from different geographical localities in the study
line data of the same characters provided for Cx. belonged to a single taxonomic unit.
tritaeniorhynchus in Taiwan and the Oriental region2, 21. Extensive study of the morphological characters and
In Thailand, Cx. tritaeniorhynchus of the different their analysis done in the present study has made it clear
provinces were categorized into three types based on the that for a given species one population can differ from
number of cibarial teeth, siphon index ratio and mating another population in few morphological characters, but
behaviour22. But their results for the three types show unless this variation is consistent and is significantly out-
overlapping values for these characters. With regard to side the known chaetotaxic range of that character for
the cibarial teeth, both type A and type C have < 16 teeth, that species, there is no rationale to categorize the popu-
while the siphon index ratio is shorter, moderate and long lations into Variety or Type.
and the mating behaviour is moderate stenogamous, non- The present study has also shown that the endophilic
stenogamous and moderate stenogamous respectively in resting behaviour of Cx. tritaeniorhynchus in Bellary re-
type A, type B and type C. Their observation that besides ported earlier is not a function of genetic variability in
the morphotaxonomic studies, other studies such as cyto- the population3. Besides, the arid conditions of Bellary to
genetic and biochemical studies and cross mating were which these authors have attributed the indoor resting,
also carried out but were not found significantly useful as our observations indicated that cohabitation of cattle along
markers to separate strains, shows that there is no basis to with humans in the same dwelling (mixed dwelling) at-
consider different populations of Cx. tritaeniorhynchus tracts this zoophilic species to the cattle and also pro-
as types A, B and C. vides a suitable humid condition for resting inside the
In the case of males also, none of the characters stud- mixed dwellings that constitute the majority of dwellings
ied were found to vary between the different populations in the villages of Bellary. This is substantiated by the re-
of Cx. tritaeniorhynchus. The male genitalia characters con- sults of our resting densities which showed that resting
sidered to be of importance in taxonomic differentiation density in the mixed dwellings was 6.65 times higher than
also showed no evidence to consider that any of the five that in exclusive human dwellings.
populations is different from the others. The intraspecific The phylogenetic analysis of the three species of
variations reported in Culex vishnui subgroup species in- the vishnui subgroup namely, Cx. tritaeniorhynchus,
cluded one variant in Cx. tritaeniorhynchus which was Cx. pseudovishnui and Cx. vishnui showed that
based on the length of the finger-like processes on the lat- Cx. pseudovishnui and Cx. vishnui are more closely
eral plate of phallosome23. Measurement of this character related. Very low densities of these two species in
is highly subjective as it could vary with the condition of almost all the study areas may be a reflection of this close
 Rajavel et al: Taxonomic status of Culex tritaeniorhynchus 51

relationship between these species compared to the Cx. quences. J Mol Evol 1980; 16: 111–20.
tritaeniorhynchus which is found to be predominant. 9. Saitou N, Nei M. The neighbour-joining method: A new method
for reconstructing phylogenetic trees. Mol Biol Evol 1987; 4:
The COI sequences (DNA barcodes) of the three spe- 406–25.
cies, Cx. tritaeniorhynchus, Cx. pseudovishnui and Cx. 10. Hajibabaei M, Janzen DH, Burns JM, Hallwachs W, Hebert PD.
vishnui deposited in the Gene Bank is an additional mo- DNA barcodes distinguish species of tropical Lepidoptera. Proc
lecular tool which is now available for identification of Natl Acad Sci U.S.A. 2006; 103: 968–71.
11. Kumar S, Tamura K, Nei M. MEGA3: Integrated software for
these different taxa. While morphological identification
molecular evolutionary genetics analysis and sequence alignment.
of these species can be done with keys already available, Brief Bioinform 2004; 5: 150–63.
the problem faced in identification of rubbed specimens 12. Shouche YS, Patole MS. Sequence analysis of mitochondrial 16S
that have lost the scales, can now be resolved with this ribosomal RNA gene fragment from seven mosquito species.
new molecular tool. J Biosci 2000; 25: 361–6.
13. Librado P, Rozas J. DnaSP v5: A software for comprehensive
analysis of DNA polymorphism data. Bioinformatics 2009; 25:
ACKNOWLEDGEMENTS 1451–2.
14. Kanojia PC, Paingankar MS, Patil AA, Gokhale MD, Deobagkar
The study was supported by the DST, Govt. of India, DN. Morphometric and allozyme variation in Culex
tritaeniorhynchus mosquito populations from India. J Insect Sci
Grant No. SR/SO/AS-79/2007 awarded to the first
2010; 10: 138.
author. 15. Bram RA. Contribution to the mosquito fauna of Southeast Asia.
II. The genus Culex in Thailand (Diptera: Culicidae). Contrib
REFERENCES Am Entomol Inst (Ann Arbor) 1967; 2: 1–234.
16. Tanaka K, Mizusawa K, Saugstad ES. A revision of the adult
1. Reuben R, Tewari SC, Hiriyan J, Akiyama J. Illustrated keys to and larval mosquitoes of Japan (including the Ryukyu Archi-
species of Culex (Culex) associated with Japanese encephalitis pelago and the Ogasawara Islands) and Korea (Diptera: Culi-
in Southeast Asia (Diptera: Culicidae). Mosq Syst 1994; 26: 75– cidae). Contrib Am Entomol Inst (Ann Arbor) 1979; 16: 1–987.
96. 17. Harbach RE. The mosquitoes of the subgenus Culex in south-
2. Sirivanakarn S. Medical entomology studies. III. A revision of western Asia and Egypt (Diptera: Culicidae. Contrib Am Entomol
the subgenus Culex in the Oriental region (Diptera: Culicidae). Inst (Ann Arbor) 1988; 24: 1–240.
Contrib Am Entomol Inst 1976; 12: 1–272. 18. Reinert JF. Comparative anatomy of the female genitalia of gen-
3. Kanojia, PC, Geevarghese G. First report of high-degree era and subgenera in tribe Aedini (Diptera: Culicidae). Part XIII.
endophilism in Culex tritaeniorhynchus (Diptera: Culicidae) in Genus Ochlerotatus Lynch Arribalzaga. Contrib Am Entomol
an area endemic for Japanese encephalitis. J Med Entomol 2004; Inst (Ann Arbor) 2002; 33: 1–111.
41: 994–6. 19. Belkin JN. The mosquitoes of the South Pacific (Diptera: Culi-
4. Fakoorziba MR, Vijayan VA. Variation in morphological char- cidae); v 2. Berkeley and Los Angeles: University of California
acters of Culex tritaeniorhynchus (Diptera: Culicidae), a Japa- Press 1962; p. 608 and 412.
nese encephalitis vector at Mysore, India. J Commun Dis 2003; 20. Tanaka K. Studies on the pupal mosquitoes of Japan (10) Culex
35: 206–13. (Culex) (Diptera: Culicidae). Jpn J Syst Ent 2004; 10: 9–42.
5. Rajavel AR, Natarajan R, Vaidyanathan K, Soniya VP. A list of 21. Chen CY. Studies on morphology of the cibarium in culicine
the mosquitoes housed in the mosquito museum at the Vector mosquitos: I. Eight species of culicines common in the Taipei
Control Research Centre, Pondicherry, India. J Am Mosq Con- area, Taiwan. J Formosan Med Assoc 1972; 71: 282–91.
trol Assoc 2005; 21: 243–51. 22. Sucharit SK, Surathin K, Shrestha SR. Vectors of Japanese en-
6. Rajavel AR, Natarajan R, Vaidyanathan K, Jambulingam P. Sys- cephalitis virus (JEV): Species complexes of the vectors. South-
tematic list of the species added to the mosquito museum at the east Asian J Trop Med Public Health 1989; 20: 611–20.
Vector Control Research Centre, Pondicherry, India. J Am Mosq 23. Kaur S, Monika Airi, Tewari PK. Intraspecific variations in three
Control Assoc 2011; 27: 8–14. vector species of Culex vishnui subgroup (Diptera: Culicidae)
7. Kumar NP, Rajavel AR, Natarajan R, Jambulingam P. DNA based on male genitalia. Entomon 2009; 34: 123–36.
barcodes can distinguish species of Indian mosquitoes (Diptera: 24. Colless DH. Notes on the culicine mosquitoes of Singapore.
Culicidae). J Med Entomol 2007; 44: 1–7. II. The Culex vishnui group (Diptera: Culicidae), with descrip-
8. Kimura M. A simple method for estimating evolutionary rate of tions of two new species. Ann Trop Med Parasitol 1957; 52:
base substitutions through comparative studies of nucleotide se- 87–101.

Correspondence to: Dr A.R. Rajavel, Scientist ‘D’, Vector Control Research Centre (ICMR), Indira Nagar, Puducherry–605 006, India.
E-mail: arrvel@yahoo.com

Received: 7 August 2014 Accepted in revised form: 28 October 2014

View publication stats