68 Journal of Vector Ecology June2011

Comparison of the efficacy of CO2-baited and unbaited light traps, gravid traps, backpack aspirators, and sweep net collections for sampling mosquitoes infected

with Japanese encephalitis virus

Yu-Chen Chen, Chih-Yuan Wang, Hwa-Jen Teng, Chien-Fu Chen, Mi-Chun Chang, Liang-Chen Lu, Cheo Lin, Shu-Wan Jian, and Ho-Sheng Wu

Research and Diagnostic Center, Centers for Disease Control, Taiwan, ROC

Received 13 July 2010; Accepted 2 December 2010

ABSTRACT: Two field studies were conducted to determine the efficacy of mosquito collection methods for species composition, species abundance, and Japanese encephalitis virus infection rates in Taiwan. Traps evaluated included John W. Hock (JH) model UD black light traps, JH model 1012 new standard miniature CDC light traps, JH model 1712 CDC gravid traps, and Taiwan-made Pest-O-Lite light traps. Backpack aspirators and sweep nets were also used to collect the resting population. Culex tritaeniorhynchus in all studies and Mansonia uniformis in the Taipei areas were the two most abundance species collected. Dry ice-baited UD black light traps were effective in regard to species diversity, species abundance, and Japanese encephalitis virus infection rates. The unbaited Pest-O-Lite light traps collected significantly more female mosquitoes than the UD black light traps but performed similarly with regard to species diversity and male mosquito collection. Most mosquitoes collected by Pest-O-Lite light traps were dried and not suitable for virus detection. Dry ice-baited CDC light traps collected significantly fewer mosquitoes than other light traps. Although CO2-baited UD black light traps with octenol attracted more mosquitoes, no statistical significance was found compared to CO2-baited UD black light traps without octenol. Japanese encephalitis viruses were isolated from half of the positive pools in UD black light traps and CDC light traps. Journal of Vector Ecology 36 (1): 68-74. 2011.

Keyword Index: Light traps, Culex tritaeniorhynchus, JE virus.

INTRODUCTION

Of the 132 mosquito species that have been identified in Taiwan, Aedes aegypti (L.), Ae. albopictus (Skuse), Culex tritaeniorhynchus Giles, Cx. annulus Theobald, Cx. fuscocephalus Theobald, and Anopheles minimus Theobald are major local vectors of dengue, Japanese encephalitis, and malaria. Among these diseases, only Japanese encephalitis is considered to be an endemic disease in Taiwan, while the other two diseases are at present considered to be travel-related diseases (Wu et al. 2009).

Mosquito surveillance to determine species composition, species abundance, and pathogen infection rates provides basic information for control strategies and measures of mosquito-borne infectious diseases. Surveillance requires accurate sampling techniques of adult mosquitoes based on the behavior and ecology of the target species. Common techniques used to collect adult mosquitoes include human landing collections, light traps, gravid traps, backpack aspirators, and sweep nets. However, each of these techniques has drawbacks that interfere with its effectiveness and accuracy. Human landing collection, backpack aspirators, and sweeping nets are unfavorable due to their labor-intensiveness, collector factors, and the risk of exposing collectors to pathogen transmission. Light traps attract most mosquitoes but might miss some non-phototactic Culex species (Tsai et al. 1988, Nayar et al. 2001, Kline et al. 2006). A widely used light trap for the surveillance

of vector mosquitoes of West Nile virus is the Centers for Disease Control (CDC) trap, which uses carbon dioxide and a light source to attract mosquitoes. Furthermore, most mosquitoes collected by this trap type are host-seeking mosquitoes that include large quantities of first-time feeders that will underestimate the infection rates of pathogens. An alternative to the light trap for Culex collection is the gravid trap, which is designed to collect gravid female mosquitoes that are attracted to a potential oviposition site containing water high in organic matter (Reiter 1983, 1987). Hence, it assures that the collected mosquitoes blood-fed at least once, increasing the possibility of detecting viral infections in the mosquito population.

The addition of olfactory attractants in light traps has been found to be effective in the collection of host-seeking mosquitoes in some but not all species (Canyon and Hii 1997, Mboera et al. 2000, McCardle et al. 2004, Shone et al. 2006, Muturi et al. 2007). These attractants include CO2, lactic acids, 1-octen-3-ol, acetone, and carboxylic fatty acids (Takken and Knols 1999), and serve as kairomones or allomones. The most commonly used chemicals were CO2 and octenol, which are considered as long range kairomones. A kairomone is a semiochemical or infochemical that acts between individuals of different species (hosts) to the benefit of the receiver (mosquitoes). Carbon dioxide appears to be universally attractive to mosquitoes (Gibson and Torr 1999). 1-octen-3-ol is a chemical compound present in bovine emanations, which especially attracts various Culex,

Vol.36,no.1 Journal of Vector Ecology 69

Anopheles, and Aedes species, especially when combined with CO2 (Kline 1994). Many of these species are also important vectors of mosquito-borne diseases. Therefore, trapping systems that incorporate combinations of these attractants may enhance surveillance for these diseases and provide a better estimate of disease prevalence in individual vector species.

Mosquito surveillance for estimating abundance and infection rates is problematic. In Taiwan, we used unbaited Pest-O-Lite light traps hung in front of a house to estimate the distribution and abundance of An. minimus. The collected mosquito samples were usually dried, damaged, and not suitable for virus detection. Therefore, finding a trap for collecting suitable mosquito samples to determine both species abundance and infection rates is highly desirable for mosquito surveillance. The objective of our study was to compare different collection methods with and without attractants for species diversity, species abundance, and virus infection rates in mosquitoes to determine which trap is the most effective for mosquito surveillance in Taiwan.

MATERIALS AND METHODS

Study sitesTrapping studies were conducted at human dwellings

of rural areas in southeast Taiwan and in natural wild fields (a natural park and an animal zoo) in Taipei in northern Taiwan. At the wild field sites, there were ponds and birds. The villages of southeastern Taiwan have rice fields and livestock in close quarters to human dwellings.

Evaluation of light traps in villagesTwo light traps, the UD black light trap (Model 1312,

John W. Hock Company, Gainesville, FL) and the Pest-O-Lite light trap (locally manufactured in Taiwan), were evaluated in this study. The UD black light trap uses a 4-W black light tube (320-420 nm) and a 6 VDC battery, whereas the Pest-O-Lite light trap (18.5 cm dia. base and 28 cm dia. cover x 25 cm high) also uses a black light (300-400 nm) but with an AC electric plug. Mosquitoes are attracted by the black light on the top cover and captured by a vacuum mechanism downward into a net bag. From April to September, 2005 and 2006, two to three villages were surveyed each month in southeast Taiwan for a total of 19 villages. At each visit, these two traps were set up outside of the houses overnight approximately 1.5-2.0 m off the ground. On the same day, two teams collected adult mosquitoes inside the houses and their surroundings for one hour between 15:00 and 17:00. Each team included two people with one modified CDC backpack aspirator (Model 1412, John W. Hock Company) and one sweep net. Additionally, the effect of octenol paired with CO2-baited UD black light traps was evaluated from June to September, 2006 inside the villages, village edge, and outside the villages.

Evaluation of light traps and gravid traps in open fieldsThree types of traps, UD black light trap (Model

1312, the same as above), new standard miniature light

trap (Model 1012, John Hock Company), and gravid trap (Model 1712, John Hock Company) were evaluated in this study. The new standard miniature CDC light trap uses an incandescent light (CM47 bulb) source and requires a 6 V battery. Both light traps were baited with dry ice and were hung from the trees approximately 1.5 m off the ground. The gravid traps, collecting gravid females, were placed on the ground in partially shaded areas. Mosquitoes are drawn into the gravid traps from the water surface into a bag by a 6 V fan positioned inside a PVC intake tube which is suspended over a grey pan (22 cm in width x 34 cm in length x 17 cm in depth) containing an oviposition attractant. A yeast infusion (1 kg to 1 liter H2O) incubated for seven days was chosen based as the oviposition attractant based on a previous pilot study. Each month, these three traps were set up randomly overnight from 17:00 to 09:00 the next morning at two sites in the Nature Park and the Animal Zoo, respectively, from April-August, 2006. Additionally, human collections by sweeping nets were conducted to collect the mosquitoes on the same day between 17:30 and 19:30 for 30 min.

Virus detection in mosquitoesAll mosquitoes collected from 2005 to 2008 by various

methods were stored on dry ice and brought back to the laboratory for species identification and stored at –20o

C for virus detection. A one-step SYBR Green I-based Real-Time Reverse Transcription-PCR Assay was used to test for mosquito Japanese Encephalitis infection rates. The detailed protocol of this method was performed as described by Yang et al. (2010). Mosquito pools were homogenized and clarified by centrifugation. Viral RNAs (70 μl) were extracted from 140 ml of mosquito suspension using the QIAamp viral RNA mini kit (cat. no. 52,906, Qiagen, Hilden, Germany) according to the manufacturer’s instructions. RT-PCR amplification was performed in the Mx4000 quantitative PCR system (Stratagene, CA, U.S.A.). Samples were assayed in a 50 μl reaction mixture containing 10 μl of sample RNA and optimal concentration of the primers by using QuantiTect SYBR Green RT-PCR kit (cat. no. 204, 243, Qiagen, Hilden Germany). The thermal profile consisted of a 30-min RT step at 50° C and 15 min of Taq polymerase activation at 95° C, followed by 45 cycles of PCR. Following amplification, a melting curve analysis was performed to verify the correct product by its specific melting temperature. Furthermore, positive PCR products were sequenced in order to determine disease profile. The maximal number of mosquitoes tested per pool was 50. All positive pools of mosquitoes were subjected to virus isolation.

Statistical analysisAll statistical analyses were performed using Statistica

version 5.0 statistical package (StatSoft Holdings, Inc., Tulsa, OK, U.S.A.). Because the main objective of the study was to compare the effectiveness of the traps and not the difference between sites, data for each site was treated as a block to eliminate site variance. The species diversity and mosquito abundance were analyzed by a completely

70 Journal of Vector Ecology June2011

randomized block design. Statistical analysis was done after log transformation of adult mosquito counts to normalize the distribution and minimize the standard error. Means separation was made using a least significant difference test. Additionally, a paired t-test was used to compare the differences in the mean number of mosquitoes collected by CO2-baited UD black light traps with and without octenol.

RESULTS

Evaluation of light traps in villagesSignificantly more species (P < 0.01), more females

(P < 0.01), and fewer males (P < 0.05) were collected in light traps than by using backpack aspirators (Table 1). For species diversity, UD black light traps (5.4 species per trap) and Pest-O-Lite light traps (4.8 species per trap) were significantly better than backpack aspirators (2.9 species per trap). For female abundance, Pest-O-Lite light traps (77.3 females per trap) were best followed by UD black light traps (49.6 females per trap) and backpack aspirators (15.8 females per trap). For male abundance, backpack aspirators (16.8 males per trap) collected significantly more males than the Pest-O-Lite and UD black light traps (7.7 and 5.1 males per trap). From a total of 1,614 mosquitoes that were collected during the sampling period, 19 mosquito species from the genera Aedes, Anopheles, Armigeres, Conquillettidia, Culex, and Mansonia were captured in Pest-O-Lite light traps (Table 2). Culex tritaeniorhynchus constituted 45.2% of the total collection followed by Ar. subalbatus (22.2%), An. sinensis (13.1%), An. ludlowae (5.8%), An. minimus (3.1%), Cx. quinquefasciatus (2.8%), Cx. annulus (2.0%), An. maculatus (1.3%), and Cx. sitens (1.2%). The other ten species occurred in quantities smaller than 1% of the total collection. In the CO2-baited UD black light traps, 23 species were captured. The most abundant species in these collections was Cx. tritaeniorhynchus (24.4%), followed by Ar. subalbatus (18.3%), Ae. vexans vexans (14.5%), Ae. albopictus (10.6%), An. ludlowae (6.1%), Cx. quinquefasciatus (5.0%), Cx. annulus (4.5%), An. minimus (4.2%), An. maculatus (2.9%), Ae. desmotes (2.5%), An. sinensis (2.3%), and Cx. sitiens (1.9%). The other 11 species were represented in quantities lower than 1% of the total

capture. In backpack aspirators, only seven species were captured. The most abundant species was Ar. subalbatus (49.4%), followed by Cx. quinquefasciatus (30.9%), and Ae. albopictus (18.1%). The other four species were captured in quantities lower than 1%.

In addition, the effect of octenol was also evaluated at the same sites. Although octenol (in addition to CO2 in UD black light traps) attracted more mosquitoes (males, females, and total), no statistically significant difference was found (P > 0.05). The mean number of species, males, females, and total attracted were 3.4 species, 4.8 males, 57.5 females and 62.3 mosquitoes per trap in CO2-baited UD black light traps, and 4.6 species, 7.1 males, 87.4 females, and 94.5 mosquitoes in CO2-baited UD black light traps paired with octenol, respectively.

Evaluation of light traps and gravid traps in open fieldsSignificant differences among the five mosquito-

collecting methods were found in respect to species diversity, mosquito abundance, female quantity, and male quantity (P < 0.01) (Table 3). For species diversity, light traps (7.7, 5.6, and 5.9 species per trap in downward or upward CO2-baited UD black light traps and CDC light traps, respectively) collected significantly more species than gravid traps (1.6 species per trap) and sweeping nets (3.1 species per trap). For mosquito abundance, significantly more mosquitoes were collected by light traps (119.4-191.7 mosquitoes per trap-night) than sweeping nets (47.4 mosquitoes per night) and gravid traps (11.3 mosquitoes per trap-night). Higher numbers of female mosquitoes were collected in light traps (114.4-187.9 females per trap-night) than sweeping net (31.9 females per night) and gravid traps (9.7 females per trap-night). Significantly more males were collected by sweeping nets (13.5 per trap-night) than light traps (1.5-5.0 per trap-night).

From a total of 13,124 mosquitos that were collected during the sampling period, twenty-seven mosquito species from the genera Aedes, Anopheles, Armigeres, Conquillettidia, Culex, Heizmannia, Mansonia, Tripteroides, and Uranotaenia were captured (Table 4). In UD black light traps, Culex tritaeniorhynchus constituted 66.7-73.6% of the total collection in UD black light traps,

Variables nCollection methods

F2,36 PPest-O-Lite light trap

UD black light trap Backpack aspirator

Species 19 4.8 a 5.4 a 2.9 b 5.8 <0.01

Males 19 7.7 a 5.1 a 16.8 b 9.2 <0.01

Females 19 77.3 a 49.6 b 15.8 c 4.3 <0.05

Total 19 85.0 a 54.7 a 32.6 a 3.0 >0.05

Table 1. Numbers of mosquito species, males, and females collected in backpack aspirators, Pest-O-Lite light traps, and UD black light traps in human dwellings from 2004 to 2005.

Means within each row (species, sex, and total) having the same letter are not significantly different at P ≤ 0.05.

Vol.36,no.1 Journal of Vector Ecology 71

Table 2. Total number (percentage) of mosquito species caught in various collecting methods in villages of southern Taiwan from 2004 to 2005.

*14 species included Aedes desmotes, Ae. lineatopennis, Ae. albolateralis, Anopheles tessellates, Anopheles species, Armigeres omissus, Coquillettidia crassipes, Culex bitaeniorhynchus, Cx. fuscanus, Cx. hayashii, Cx. neomimulus, Cx. nigropunctatus, Cx. culiciomyia sp., and Mansonia uniformis.

Mosquito species Pest-O-Lite light traps

CO2-baited UD black light traps Backpack aspirator

Cx. tritaeniorhynchus 730 (45.2) 254 (24.4) 6 (1.0)

Ar. subalbatus 358 (22.2) 190 (18.3) 306 (49.4)

An. sinensis 213 (13.2) 24 (2.3) 0

An. ludlowae 94 (5.8) 63 (6.1) 0

An. minimus 50 (3.1) 44 (4.2) 1 (0.2)

Cx. quinquefasciatus 45 (2.8) 52 (5.0) 191 (30.9)

Cx. annulus 33 (2.0) 47 (4.5) 2 (0.3)

An. maculatus 21 (1.3) 30 (2.9) 1 (0.2)

Cx. sitiens 20 (1.2) 20 (1.9) 0

Ae. albopictus 13 (0.8) 110 (10.6) 112 (18.1)

Cx. fuscocephala 11 (0.7) 4 (0.4) 0

Ae. vexans vexans 6 (0.4) 151 (14.5) 0

Other species* 20 (1.2) 51 (4.9) 0

Total 1,614 (100.0) 1,040 (100.0) 619 (100.0)

Variable n

CO2-baited UD black light traps

CO2-baited

CDC light traps

Gravid traps

Sweeping nets F4, 88 P

Downward Upward

Species 23 7.7a 5.6a 5.9a 1.6b 3.1b 7.6 <0.01

Males 23 3.8a 4.3a 5.0a 1.5a 13.5b 3.8 <0.01

Females 23 187.9a 129.5a 114.4ab 9.7c 31.9bc 4.8 <0.01

Total 23 191.7a 133.8a 119.3a 11.3b 47.3b 4.6 <0.01

Table 3. Mean number of mosquitoes collected in UD black light traps, CDC light traps, and gravid traps during April-August 2007.

Means within each row (species, sex, and total) having the same letter are not significantly different at P ≤ 0.05.

72 Journal of Vector Ecology June2011

Mos

quito

spec

ies

Dow

nwar

d C

O2-b

aite

d U

D b

lack

ligh

t tra

psU

pwar

d C

O2-b

aite

d U

D b

lack

lig

ht tr

aps

CO

2-bai

ted

CD

C li

ght

trap

sG

ravi

d tr

aps

Swee

p ne

t

Cule

x tr

itaen

iorh

ynch

us2,

941

(66.

7)2,

264

(73.

6)98

2 (3

5.8)

13 (4

.4)

263

(24.

2)

Man

soni

a un

iform

is96

4 (2

1.9)

512

(16.

6)1,

378

(50.

2)18

9 (6

4.1)

425

(39.

0)

Cx. a

nnul

us14

8 (3

.4)

140

(4.5

)15

8 (5

.8)

1 (0

.3)

7 (0

.6)

Aede

s alb

opic

tus

109

(2.5

)47

(1.5

)13

2 (4

.8)

22 (7

.5)

245

(22.

5)

Cx. m

imul

us60

(1.4

)7

(0.2

)14

(0.5

)0

(0.0

)6

( 0.6

)

Cx. r

ubith

orac

is56

(1.3

)42

(1.4

)17

(0.6

)3

(1.0

)13

(1.2

)

Cx. q

uinq

uefa

scia

tus

47 (1

.1)

22 (0

.7)

20 (0

.7)

13 (4

.4)

89 (8

.2)

Oth

er sp

ecie

s*84

(1.9

)36

(1.2

)44

(1.6

)18

(6.1

)41

(3.8

)

Tota

l4,

409

(100

.0)

3,07

8 (1

00.0

)2,

745

(100

.0)

295

(100

.0)

1,08

9 (1

00.0

)

Tabl

e 4.

Tot

al n

umbe

r (pe

rcen

tage

) of m

osqu

ito sp

ecie

s cau

ght i

n va

riou

s col

lect

ing

met

hods

in T

aipe

i, Ta

iwan

, dur

ing

Apr

il-A

ugus

t, 20

07.

*19

spec

ies i

nclu

ded

Aede

s pen

ghue

nsis,

Ae.

vexa

ns v

exan

s, An

ophe

les si

nens

is, A

n. te

ssell

ates

, Arm

iger

es su

balb

atus

, Coq

uille

ttidi

a cr

assip

es, C

ulex

bic

ornu

tus,

Cx.

bita

enio

rhyn

chus

, Cx.

fusc

anus

, Cx.

hal

ifaxi

i, Cx

. inf

antu

lus,

Cx. m

urre

lli, C

x. p

allid

otho

rax,

Cx.

pse

udov

ishnu

i, H

eizm

anni

a ta

iwan

ensis

, Mim

omyi

a lu

zone

nsis,

Trip

tero

ides

ba

mbu

sa, T

r. ar

anoi

des,

and

Ura

nota

enia

niv

iplu

ra.

Mos

quito

spec

ies

UD

bla

ck li

ght t

raps

CD

C li

ght t

raps

PO L

ite li

ght t

raps

**G

ravi

d tr

aps

Hum

an c

olle

ctio

n an

d sw

eepi

ng n

ets

Tota

l(P

ositi

ve

rate

%)

Tota

l po

ols*

Posit

ive

pool

s (is

olat

es)

Tota

l po

ols

Posit

ive

pool

s(is

olat

es)

Tota

l po

ols

Posit

ive

pool

s(is

olat

es)

Tota

l po

ols

Posit

ive

pool

sTo

tal

pool

s

Posit

ive

pool

s (is

olat

es)

Cx. t

ritae

nior

hync

hus

475

12 (6

)61

2 (1

)89

1 (0

)26

099

10 (6

)75

0 (3

.3)

Cx. f

usco

ceph

alus

190

00

41

(0)

00

00

23 (4

.3)

Cx. a

nnul

us83

023

06

03

07

012

2 (0

.0)

Oth

er sp

ecie

s67

20

215

034

014

30

810

1,14

5 (0

.0)

Tota

l1,

249

12 (6

)29

92

(1)

133

2 (0

)17

20

187

10 (6

)2,

040

(1.3

)

Tabl

e 5.

Japa

nese

Enc

epha

litis

viru

s inf

ectio

n in

fiel

d-co

llect

ed m

osqu

itoes

in T

aiw

an b

y va

riou

s col

lect

ion

met

hods

, 200

6-20

08.

*Up

to 5

0 m

osqu

itoes

per

poo

l.**

Mos

quito

es te

sted

wer

e se

lect

ed u

nder

a m

icro

scop

e.

Vol.36,no.1 Journal of Vector Ecology 73

followed by Ma. uniformis (16.6-21.9%), Cx. annulus (3.7-4.5%), and Ae. albopictus (1.5-2.5%). The remaining 21 species occurred in small quantities. In CDC light traps, 22 species were captured. The most abundant species in these collections were Ma. uniformis (50.2%), followed by Cx. tritaeniorhynchus (35.8%), Cx. annulus (5.8%), and Ae. albopictus (4.8%). The remaining 17 species were represented in quantities lower than 1% of the total capture. In gravid traps, only 11 species were captured. The most abundance species were Ma. uniformis (64.1%), followed by Ae. albopictus (7.5%), Cx. tritaeniorhynchus (4.4%), and Cx. quinquefasciatus (4.4%). The remaining seven species were represented in small quantities. In sweeping nets, 17 species were collected. The most prevalent species were Ma. uniformis (39.0%), Cx. tritaeniorhynchus (24.2%), and Ae. albopictus (22.5%), followed by Cx. quinquefasciatus (8.2%) and Ar. subalbatus (1.6%).

Virus detectionA total of 39,295 mosquito adults (17,968 females and

644 males) were grouped by species, place, and date into 2,040 pools (Table 5). Overall, only 3.3% of the 750 pools of Cx. tritaeniorhynchus (28,202 females and 63 males) and 4.3% of the 23 pools of Cx. fuscocephalus (349 females and 63 males) were found to be positive for Japanese Encephalitis virus by RT-PCR. Twelve positives were collected from UD black light traps and two positives were collected from CDC light traps. The highest virus isolation rate for positive pools was 60.0% for mosquitoes collected by human collection with collection tubes and sweeping nets, followed by mosquitoes collected in UD black light traps and CDC light traps (50.0%). No isolation was successful for mosquitoes collected in Pest-O-Lite light traps.

DISCUSSION

The efficacy of different mosquito collection methods varied in terms of species diversity, species abundance, and Japanese Encephalitis virus infection rates. According to our trap data, the CO2-baited UD black light traps were efficient for surveillance of species diversity, species abundance, and Japanese Encephalitis virus infection rates. The unbaited Pest-O-Lite light trap was also efficient for surveillance of mosquito species diversity and species abundance but not for virus infection rates. CO2-baited CDC light traps also worked well in regard to species diversity (Ma. uniformis, Cx. tritaeniorhynchus, Cx. annulus, and Ae. albopcitus). The yeast-baited gravid trap were fair in collecting mostly Ma. uniformis.

In our studies, Cx. tritaeniorhynchus and Ma. uniformis were the most abundant species. The former species is a cosmopolitan mosquito occurring in a wide range of larval habitats, including vegetated lakes and rivers, natural ponds, rice fields, agricultural ponds, and temporary pools. This species, an opportunist, may feed on pigs, humans, and birds and is highly attracted to lights. In our study, few Cx. quinquefasciatus were collected in light traps and gravid traps compared to backpack aspirators and sweeping

nets. In general, mosquitoes respond to the traps using visual, olfactory, and tactile cues. The light traps used in this study attracted females with visual cues with and without olfactory cues from CO2, whereas gravid traps attract with olfactory cues from ovipositional attractants. However, other visual cues, such as pan color, pan size, and water color of the gravid trap are also important in the orientation of mosquitoes in oviposition behavior in a variety of mosquito species (Allan and Kline 2004, Allen et al. 1987, Beehler et al. 1993, Bentley and Day 1989). Oviposition by Cx. quinquefasciatus occurs under low light conditions, typically with most activity within two hours after sunset (Nayar 1982, Schreiber et al. 1989, Beehler et al. 1993).

In our study, most positive pools were found in Cx. tritaeniorhynchus females, which were captured on June through July by UD black and CDC light traps. In contrast to other studies, no positive pools were collected in our gravid traps. The overall infection rate of West Nile virus (2.3 per 1,000 mosquitoes) from gravid traps was nearly 33 times greater than the infection rate (0.1 infected mosquitoes per 1,000 mosquitoes) detected from light traps (Williams and Gingrich 2007). Our gravid traps collected fewer mosquitoes (295) than other collection methods (1,089-4,409), with only 13 Cx. tritaeniorhynchus females collected. Therefore, the use of gravid traps for surveillance of Japanese encephalitis vectors needs further study.

Acknowledgments

We thank the Taipei Animal Zoo and the Guandu Nature Park for their kind assistance in collecting mosquitoes in their fields. This study was supported in part by scientific research grants from the Centers for Disease Control, Taiwan in 2005 and 2006 (DOH94-DC-2015, DOH95-DC-2012 and DOH96-DC-2008).

REFERENCES CITED

Allan, S.A. and D. Kline. 2004. Evaluation of various attributes of gravid female traps for collection of Culex in Florida. J. Vector Ecol. 29: 285-294.

Allen, S.A., J.F. Day, and J.D. Edman. 1987. Visual ecology of biting flies. Annu. Rev. Entomol. 32: 297-316.

Beehler, J.W., J.G. Millar, and M.S. Mulla. 1993. Synergism between chemical attractants and visual cues influencing oviposition of the mosquito, Culex quinquefasciatus (Diptera: Culicidae). J. Chem. Ecol. 19: 635-644.

Beehler, J.W., J.P. Webb, and M.S. Mulla. 1993. Spatial and circadian oviposition patterns in an urban population of Culex quinquefasciatus. J. Am. Mosq. Contr. Assoc. 9: 385-388.

Bentley, M.D. and J.F. Day. 1989. Chemical ecology and behavioral aspects of mosquito oviposition. Annu. Rev. Entomol. 34: 401-421.

Canyon, D. and J.L. Hii. 1997. Efficacy of carbon dioxide, 1-octen-3-ol, and lactic acid in modified Fay-Prince traps as compared to man-landing catch of Aedes

74 Journal of Vector Ecology June2011

aegypti. J. Am. Mosq. Contr. Assoc. 13: 66-70.Gibson, G. and S.J. Torr. 1999. Visual and olfactory

responses of haematophagous Diptera to host stimuli. Med. Vet. Entomol. 13: 2-13.

Kline, D.L. 1994. Olfactory attractants for mosquito surveillance and control: 1-octen-3-ol. J. Am. Mosq. Contr. Assoc. 10: 280-287.

Kline, D.L., M. Patnaude, and D.R. Barnard. 2006. Efficacy for four trap types for detecting and monitoring Culex spp. in North Central Florida. J. Med. Entomol. 43: 1121-1128.

Mboera, L.E.G., B. Knols, M. Braks, and W. Takken. 2000. Comparison of carbon-dioxide baited trapping systems for sampling of outdoor mosquito populations in Tanzania. Med. Vet. Entomol. 14: 257-263.

McCardle, P., R.E. Webbe, B.B. Norden, and J.R. Aldrich. 2004. Evaluation of five trapping system for the surveillance of gravid mosquitoes in Prince Georges County, Maryland. J. Am. Mosq. Contr. Assoc. 20: 254-260.

Muturi, E.J., J. Mwangangi, J. Shililu, S. Muriu, B. Jacob, C.M. Mbogo, G. John, and R. Novak. 2007. Evaluation of four sampling techniques for surveillance of Culex quinquefasciatus (Diptera: Culicidae) and other mosquitoes in African rice agroecosystems. J. Med. Entomol. 44: 503-508.

Nayar, J.K., M.D. Young, and D.J. Forrester. 1982. Experimental transmission by mosquitoes of Plasmodium hermani between domestic turkeys and pen-reared bobwhites. J. Parasitol. 68: 874-876.

Nayar, J. K., N. Karabatos, J.W. Knoght, M. Godsey, J. Chang, and C.J. Mitchell. 2001. Mosquito hosts of arboviruses from Indian River County, Florida during 1998. Fla. Entomol. 84: 376-379.

Reiter, P. 1983. A portable battery-operated trap for collecting gravid Culex mosquitoes. Mosq. News 43: 496-498.

Reiter, P. 1987. A revised version of the CDC gravid mosquito trap. J. Am. Mosq. Contr. Assoc. 3: 325-327.

Schreiber, E.T., J.P. Webb, J.E. Hazelrigg, and M.S. Mulla. 1989. Bionomics of adult mosquitoes associated with urban residential areas in the Los Angeles basin, California. Bull. Soc. Vector Ecol. 14: 301-318.

Shone, S.M., G.E. Glass, and D.E. Norris. 2006. Targeted trapping of mosquito vectors in the Chesapeake Bay Area of Maryland. J. Med. Entomol. 43: 151-158.

Takken, W. and B.G. Knols. 1999. Odor-mediated behavior of Afrotropical malaria mosquitoes. Annu. Rev. Entomol. 44: 131-157.

Tsai, T.F., G.C. Smith, M. Ndukwu, W.L. Jakob, C.M. Happ, L.J. Kirk, D.B. Francy, and K.J. Lampert. 1988. Entomologic studies after a St. Louis encephalitis epidemic in Grand Junction, Colorado. Am. J. Epidemiol. 128: 285-297.

Williams, G.M. and J.B. Gingrich. 2007. Comparison of light traps, gravid traps, and resting boxes for West Nile virus surveillance. J. Vector Ecol. 32: 285-291.

Wu, J.W., H.J. Teng, C. Lin, C.Y. Wang, D.P. Liu, and H.S. Wu. 2009. Recent distribution of vector mosquitoes and epidemiology of the diseases they transmitted in Taiwan. Med. Entomol. Zool. 60: 241-252.

Yang, C.F., C.F. Chen, C.L. Su, H.J. Teng, L.C. Lu, C. Lin, C.Y. Wang, P.Y. Shu, J.H. Huang, and H.S. Wu. 2010. Screening of mosquitoes using SYBR Green I-based real-time RT-PCR with group-specific primers for detection of Flaviviruses and Alphaviruses in Taiwan. J. Virol. Methods 168: 147-151.