Sampling Methods for Potential Epidemic Vectors of Eastern Equine Encephalomyelitis Virus in Massachusetts

Joumal of the Arnerican Mosquito Control Association, 13(4):342-347, 1997 Copyright @ 1997 by the American Mosquito Control Association, Inc.

SAMPLING METHODS FOR POTENTIAL EPIDEMIC VECTORS OF EASTERN EQUINE ENCEPHALOMYELITIS VIRUS IN MASSACHUSETTS

RAJEEV VAIDYANATHAN'eNo JOHN D. EDMAN

Department of Entomology, University of Massachusetts, Amherst, MA 0IOO3

ABSTRACT. To determine which of 11 trapping methods best sampled populations of 6 potential epidemic vectors of eastern equine encephalomyelitis (EEE) virus, we compared New iersey (NJ) ligirt trap, CtiC tigtrt trap' CDC with octenol, CDC with COr, CDC with CO, plus octenol, American Biophysicso-cor, tigtrt trap wlth flickering light (AB flicker), AB steady light, AB flicker with octenol, AB flicker wiirr Ad- flicker with CO, plus octenol, and 3 resting boxes. There was no significant difference between CDC andAB light traps (p > 0.05). The addition of octenol increased trap catch with both CDC and AB light traps; however, tli. irr"."ur" was not statistically significant (P > 0.05). Both brands supplemented with CO, collected more Aedes canadensis, Coquillettidia perturbans, Culex salinarius, and Anopheles spp. than did unbaited light traps (p < 0.05). The addition of octenol to CO, increased collections of Anopheles spp. and decreased colGctions oi A"d"t spp. and Cq- perturbans at some sites, but these changes were not statistically significant (P > O.05). Resting boxes were best for sampling bloodfed or parous Anopheles spp. The NJ, AB flicker, and AB steady light trips were not effective for sampling potential vectors of EEE virus.

INTRODUCTION jacent lowlands to collect Cs. melanura for EEE virus isolation (D. Buckley, MassachusettsDepart- Culiseta melanura (Coquillet), the enzootic vec- ment of Health, personal communication). Once an tor of eastern equine encephalomyelitis (EEE) virus EEE virus isolation is made, Cor-baited CDC light along the northeastern coast of the United States, traps are used to include more species and to in- rests and oviposits in cold, acidic, red maple-white crease the numbers of mosquitoes collected. cedar swamps and feeds predominantly on passer- Traps used to collect Cs. melanura usually are ine birds. Its narrow host range and limited flight placed in swampy, forested habitat, so collections range preclude it as an epidemic vector of EEE vi often underestimate mammalophagic species that rus of horses and humans. Aedes canadensis (Theo- corrmute from woods to unforested areas to feed. bald), Aedes vexans (Meigen), Coquillenidia per- Therefore, these samples often fail to recover EEE turbans (Walker), Anopheles quadrimacularzs Say, virus from commuter species and provide a limited Anophelespunctipennis (Say), and Culex salinarius profile of virus prevalence and transmission among Coquillet are potential epidemic vectors of EEE vi- species that may be involved in transmission of virus. The EEE virus has been isolated from all 6 rus from avian reservoirs to humans and horses- species(Sudia et al. 1968, Grady et al. 1978, How- We evaluated the ability of CDC light traps, NJ ard et al. 1988, Edman et al. 1993, Ninivaggi and light traps, resting boxes, and a new light trap (AB) Guirgis 1994). They rest in shaded woods during manufactured by American Biophysics@ of James- the daytime but cornmute to more open or transi- town, RI, to estimate the diversity and density of tional habitats around sunset or early evening. They potential epidemic vectors of EEE virus. feed predominantly on mammals, including humans and livestock, but also may feed on enzootic avian reservoirs of arboviruses(LaSalle and Dakin 1982. MATERIALS AND METHODS Weathersbeeand Meisch 1990, Edman et al. 1993). All but Ae. canadensis host seek in open fields or Test sites: The Hockomock Swamp, a red ma- suburban areas during the evening. Aedes canaden- ple-white cedar freshwater swamp in the Taunton sls generally feeds on a variety of smaller hosts River basin of southeasternMassachusetts. is en- along the forest ecotone. zootic for EEE virus and Cs. melanura (A. Decas- New Jersey(NJ) and CDC light traps historically tro, Bristol County Mosquito Control, personal were used to monitor the relative densities and communication). Four sites in this area were se- changing population patterns of Cs. melanura lected for rap comparisons in 1994 and 1995 as (Headlee 1932, Sudia and Chamberlain 1962, Mat- described by Vaidyanathan and Edman (1997). In sumoto and Maxfield 1985). The Massachusetts 1996, we compared traps along the margin of a Department of Health uses unbaited CDC light forested swamp adjacent to a flood plain in Hadley, traps within red maple-white cedar swamps and ad- MA. Experimental design: In 1994 and 1995, we used CDC miniature light traps (Sudia and Chamberlain rPresent address:Department of Parasitology,Hebrew 1962), NJ light traps (Headlee 1932), resting boxes University-Hadassah Medical School, Ein Kerem, Jeru- (Nasci and Edman 1981), and AB light traps for 2 salem91120. Israel. summers. Tfaps, resting boxes, octenol, and CO,

342 DrceMsen 1997 Seuplnc METHoDs ron EEE Vrnus n MASSACHUSETTS 5+5

Table l. Number of mosquitoescollected with 8 sampling methods in southeasternMassachusetts in 1994 and 1995.

Flicker*COr* Flicker* Mosquito Flicker*CO, octenol CDC octenol Flicker Resting Steady NJ Ae. can' 154 r47 2322203 ) Ae. ve* 6 4 300000 Ae. stim3 16 3l 11 3 4 0 3 2 Ae. darF 1,2t4 763 333 337 109 | 2 5 Anophls5 98 127 29178631 0 Cs. mela6 5l 50 170 52 27 16 5 I Cq. perf 659 2to 212 41 42 0 17 7 Culef 85 t45 902416115 18 Total 2,283 1,477 871 496 208 81 46 38 Nights in use 3t 27 36 34 33 1l 13 ll '74 Mean 55 24 15'1 34 4 SD 104 76 282r1178 )

I Aedes canadensis. 2 Aedes veruns. 3 Aedes stimulans, Ae. fitchii, Ae. excrucians. a Aedes triseriatus, Ae. abserratus, Ae. cinereus, Ae. auriftr, Ae. tivitattus. s Anopheles quadrimculatus, An. punctipennis, An. walkeri. 6 Culiseta melanura. 1 Co quille ttidia p e rturbans. 8 Culex salinarius, Cx. restuans, Cx. territans. were employed as described by Vaidyanathan and males per night, significantly better than either Edman (1997). In 1996, we compared AB flicker, flicker or steady traps (P < 0.05). Table 3 is a sum- AB flicker*octenol, AB flicker*COr, AB mary of trap comparisons and corresponding P-val- flicker*COr*octenol, CDC light trap, CDC*octenol, ues. CDC+CO' and CDC+COr*octenol. The octenol The addition of octenol to a flicker trap increased emission rate was increasedto 2.0 mg/h and traps averageyield from 6 to 15 females per night; how- were operatedfor 15 days in June 1996. ever, the octenol traps were so variable (SD : 21) Data analysis: Nightly meansand speciesmeans that this difference was not significant (P : 0.07). were calculated for each sampling method in 1994 The addition of CO, to a flicker trap significantly and 1995. Only nightly totals and overall trap increased yield from 6 to 74 females per night (P means were calculated for each sampling method < 0.05). The flicker*CO, trap caught significantly in 1996 when AB and CDC light traps sampled more females per night than a flicker+octenol trap equivalent numbers. Data were analyzed as de- (P < 0.O5). The addition of CO, and octenol to a scribed by Vaidyanathan and Edman (1997>. flicker trap significantly increased average trap yield over octenol alone (P < 0.05) but not over CO, alone (P = 0.39). The flicker*CO, and RESULTS flicker*COr*octenol traps were significantly better 1994 and 1995: A total of 5,500 females were than an unbaited CDC light trap (P < 0.05), which collected with 8 sampling methods over 38 nights was significantly better than a flicker*octenol trap (Tables 1 and 2). No significant difference was (P < 0.05). found between AB flicker and steady traps (P : Resting boxes were used 11 nights in August. 0.18) or between flicker and NJ traps (p : 0.10). Numbers from resting boxes were not significantly The CDC light trap collected an average of 24 fe- different from those from CDC (P : O.54),

Table 2. Nightly means for 8 sampling methods for mosquitoes in southeastern Massachusetts for 1994 and 1995.

Flicker* Flicker* COr+ Flicker+ Sample CO, octenol CDC octenol Flicker Steady NJ Resting All species 74 55 24 15 7 4 4 3 Aedesl 45 35 l0 ll 4 1 I 0 Coquillettidiaperturbans 21 8 6 1 I I I 0 Culexsalinarius 5 8 5 2 | 1 2 0 Anophel.ef35l10006

I Aedes canadensis, Ae. vemns, Ae. stimulm s.1., Ae. abserratrc, Ae- triseriatus, Ae. trivittatus, Ae. aurifer. 2 Anopheks punctipennis, An. qmdrimculatrc, An. walkeri. JounNn- op rnp AusnrcAN Moseurro CoNTRoLAssocranoll Vor-.13, No.4

Table 3. Pairwise comparisons using the Wilcoxon 0.05) significantly increased trap catch over un- signed rank test between sampling methods for all baited AB traps. Octenol by itself did not signifi- species, 1994 and 1995. If P > 0.05, then difference is cantly increase trap catches by AB flicker traps (P not significant; the probability that 2 methods are the : 0.41). Tfaps baited with only same is >SVo.If P < 0.05, then difference is significant. CO, collected sig- Two-tail P-values are given for normal approximation nificantly greater numbers than only octenol (P < with continuity correction. 0.05). A combination of octenol and CO, did not significantly increase yield over CO, alone (P = P- 0.05) but did significantly increase yield over oc- Comparison of methods value tenol alone (P < 0.05). Significant difference present (Tfap I > trap 2) An average of 6 female Anopheles spp. were col- CDC vs. flicker 0.00 lected nightly from resting boxes, the highest av- Flickerf CO, vs. flicker 0.00 erageof the 8 sampling methods(Table 2). A night- Flicker+CO, vs. flickerf octenol 0.00 ly average of 5 females were collected from Flickerf COr*octenol ys. flicker*COr*octenol; however, resting boxes col- flickerfoctenol 0.0r lected more bloodfed or parous females. Resting CDC vs. AB*steady 0.00 boxes were significantly better than CDC traps (P FlickerfCO, vs. CDC 0.00 < 0.05) and flicker*octenol (P < 0.05). Flicker+COr+octenol vs. CDC 0.00 The addition of octenol, COr, and COr*octenol to AB No significant difference flicker traps did not signiflcantly increase collec- Flicker vs. steady 0.18 tions of Anopheles (all P > 0.05). Trap catch from Flicker vs. NJ 0.10 flicker*CO, was not significantly different from Flicker vs. flickerf octenol 0.07 flicker*octenol (P : 0.12). A combination of CO, Flicker*CO, vs. Flickerf COr*octenol 0.39 and octenol significantly increased trap catch over octenol alone (P < 0.05) but had no significant effect over CO, alone (P : 0.09). flicker*CO, (P = O.92), or flicker*COr*octenol 1996: A total of 16,318 females were collected traps (P : 0.06) operated during this same period. with 8 sampling methods operatedfor 15 nights (fa- Resting boxes were compared only with corre- ble 4). On the basis of 1994-95 dat4 NJ and steady sponding light trap nights. light traps and resting boxes were not used in 1996. Trap specificity (1994 and 1995): The addition The CDC light trap collected an average of 13 fe- of octenol to an AB flicker light trap did not sig- males per night, not significantly different from AB nificantly increase the average yield of female Ae- flicker (P : 0.53). The CDC*octenol trapscollected des species(P : O.I2; Table 2). A combination of an average of 25 females per night, not significantly CO, and octenol significantly increased yield (p < different from flicker*octenol (P : 0.85). The ad- 0.05) over octenol alone but not over CO, alone (P dition ofoctenol to an AB flicker light trap increased : 0.94). averageyield from 9 to 19 (SD = 21) females per Aedes vexans numbers were low both years, so night (P : 0.07). There was also no significant dif- we analyzed Ae. canadensis independently. There ference between CDC+CO, and flicker+CO, 1f : was no significant difference between CDC and 0.84). However, CDC*COr*octenol was signifi- flicker (P : 0.86), flicker and flicker*octenol (P : cantly better than AB flicker*COr*octenol (P < 0.19), or flicker*CO, and flicker*COr*octenol (P 0.0s). : O.23). The addition of CO, to a flicker light trap The addition of octenol to a CDC light trap in- sampled significantly more Ae. canadensisfemales creased average yield from 13 to 25 females per than octenol alone (P < 0.05). Resting boxes, NJ night; howeve! the octenol traps were so variable light trap, and steady light trap were ineffective in (SD : 48) that this difference was not significant sampling Ae. canadensis. (P : O.24). The addition of CO, to a CDC trap Yields of female Cq. perturbans from flicker, signiflcantly increased yield from 13 to 319 females steady,NJ, and flicker*octenol light traps were not per night (P < 0.05). The CDC+CO, trap caught significantly different from one another (all P > significantly more females per night than 0.05; Table 2). Tlrc addition of octenol to the AB CDC*octenol (P < 0.05). The addition of CO, and flicker trap did not significantly affect trap catch (P octenol to a CDC trap significantly increased av- : O.42), whereas the addition of CO, to the AB erage trap yield over octenol alone (P < 0.05) but flicker trap significantly increased trap catch (P ( not CO, alone (P : O.23). 0.05). A combination of CO, and octenol signifi- Trap specificity (1996): There was no significantly increased AB trap catch over octenol alone cant difference between unbaited AB and CDC (P < 0.05) but had no significant effect over CO, light traps for any species (Table 5). Aedes cana- alone (P : 0.61). densis, Aedes stimulans (Walker) s.l.g., and dark- Flicker*COr* octenol traps collected an average legged Aedes were equally represented by of 8 female Cx. salinariu,s per night, the highest for flicker*COr, flicker*COr*octenol, CDC+COr, 8 sampling methods (P < 0.05; Table 2). The ad- and CDc+COr*octenol. More Aedes spp. were dition of CO, (P < 0.05) and COr*octenol (P ( sampled with CO, plus octenol vs. CO, alone, but SllrplNc Msrsoos ron EEE Vtnus nv MASSAcHUSETTS 345

Table 4. Number of mosquitoes collected with 8 sampling methods in western Massachusetts in 1996. Flicker+ Flickert Flicker* cor+ cDc+ cDC+CO,+ Mosquito Flicker octenol CO, octenol cDc octenol CDC+CO, octenol Ae. canl 8 l3 426 529 99290 335 Ae. ve* 0 0 4 o2 I 10 Ae. stim3 l 7 130 r36 o2149 r63 Ae. darka .A 47 t,o77 984 t9 33 r,225 t,ol7 Anophls5 J 20 ll3 257 610 164 211 Cs. mela6 I 0 0 7 00 5 0 Cq. perf 93 r90 r,253 2,166 145 302 ? R55 r,523 Culef f 10 84 85 914 l02 27 Total r39 287 3,O87 4,168 188 372 4,79r 3,286 Nights in use l5 15 15 l3 l5 l5 l5 9 Mean 9 l9 206 321 l3 25 319 365 SD 9 2l 226 204 18 48 288 202

I Aedes canadensis. 2 Aedes vexans. 3 Aedes stimulans, Ae. ftchii, Ae. excmcians- 4 Aedes triseriatus, Ae. abserratus, Ae. cinereus, Ae. aurifer, Ae. trivitattus. 5 Anophleles qrcdimculatus, An. punctipennis, An. walkzri. 6 Culiseta melanura. 1 C oq ui llettidia pe rturbans. 3 Culex salimrias, Cx. restuans, Cx. territans- this difference was not signiflcant (P > 0.05). The Table 6 is a summary of trap comparisons and addition of octenol to AB and CDC light traps did correspondingp-values. not significantly increase numbers of Aedes spp. too sampled. Aedes vexans populations were Poorly DISCUSSION represented to predict anything. Addition of octenol to AB and CDC light traps The AB light traps with a steady light source increased collections of Cq. perturbansi howevet, were comparable with those with a flickering light this increasewas not significant (Table 5). Addition source and were not used the 2nd summer (Table of CO, to AB and CDC light traps significantly 3) becauseflicker uses less battery power. Flicker increasedcollections of Cq. perturbans (P < 0.05). and CDC light traps also consistently collected The combination of CO, and octenol did not sig- higher total numbers than NJ light ffaps. Size, con- nificantly increasecollections over CO2alone in ei- venience, and greater trap yield clearly favor AB ther AB or CDC light traps. flicker or CDC light traps over NJ traps for extend- Culex salinarir,s was sampled best by light traps ed surveillance. supplementedwith COr. There was no significant Resting boxes were poor indicators of Aedcs spp. increase by adding octenol to CO2, and in some atmdCq. perturbans abundance but were the best cases,this actually decreasedthe number of host- indicators of Anopheles spp. in southeasternMas- seeking females sampled. All Anopheles spp. were sachusetts;resting boxes were not used in western better represented in light traps supplemented with Massachusetts.Resting boxes tended to collect en- both CO, and octenol vs. either CO, alone or oc- gorged or parous females and provided more blood- tenol alone (P < 0.05). fed Anopheles from which to recover virus.

Table 5. Nightly meansfor 8 sampling methods for mosquitoesin western Massachusettsfor 1996. Flickerf CDC+ Flickert Flicker* C02+ cDc+ co,+ Sample Flicker octenol CO, octenol CDC octenol CDC*CO, octenol

All species 9 19 206 321 13 25 3r9 365 Aedescanadensis I I 28 4tl 11937 OtherAedest 2 4 81 86 I 2 92 r32 Coquillettidia perturbans 6 13 84 167 10 20 190 169 Culexspp.2 0 I 6 7l173 Anopheles3 0 I 8 200 1 11 23 1Aedes vexans,Ae. stimulanss.1., Ae. abserratus,Ae. triseriatus, Ae. tivitattus, Ae. aurifer, 2 Culex salinarius, Cx. restuans, Cx, teffitans. 3Anopheles punctipennis, An. quadrimculatus, An. walkeri. 346 JoURNALoF THE AMERTcANMoseurro CoNrnol AssocrerroN VoL. 13,No.4

Table 6. Pairwise comparisonsusing the Wilcoxon females per night (P < 0.05) for 1994-95; the av- signed rank test between sampling methods for all erage increasedfrom 9 to 258 (SD : 226) in 1996. species,1996. If P > 0.O5,then difference is not The addition of CO, to a CDC light trap increased significant; the probability that 2 methods are the same averageyield from (SD = is )SVo. If P < 0.05, then difference is significant. Tfvo- 12 to 319 288) females tail P-values are given for normal approximation with per night (P < 0.05). Previous studies have found continuity correction. that the addition of CO, to light traps increases trap catches of Aedes spp., Cq. perturbans, Cx. salinar- P. ius, and An. quadrimaculatus and provides a more Comparison of methods value accurate estimate of nuisance mosquito problems Significant difference present (Trap I > trap 2) (Newhouse et al. 1966, Carestia and Savage 1967, CDC + COr*octenol vs. Magnarelli 1974, Buckley et al. 1994). We found flicker*COr*octenol 0.03 that the addition of CO, significantly increased trap CDC+CO, vs. CDC 0.00 catch over AB flicker and CDC light traps alone CDC*COrtoctenol vs. CDC 0.01 for Aedes sW., Cq, perturbans, Cx. salinarius, and No significant difference Anopheles spp. (all P > 0.05). Light traps supple- CDC vs. flicker 0.53 mented with CO, collected approximately lO--29 CDCf octenol vs. flicker*octenol o.85 times the numbers of mosquitoes collected by un- CDC+CO, vs. flicker*CO, 0.84 baited traps. BecauseAB and CDC light traps col- CDC vs. CDC*octenol o.24 lected similar numbers of target species, both traps supplemented with CO, are equally useful in arbovirus surveillance. The AB light traps and CDC light traps sampled The AB and CDC light traps supplemented with roughly the same number of female mosquitoes CO, were significantly better than traps supple- whether they were unbaited or supplemented with mented with octenol (P < 0.O5). Tlrkken and Kline CO, octenol, or both. Although 1994-95 data show (1989) found that octenol emitted at 1.57-2.26 that the unbaited CDC light trap was significantly mg/h attracted mosquitoes in numbers similar to better than an unbaited flicker light trap, results CO, released at 20O mymin in the Everglades. We from 1996, when densities were higher, found no found that octenol increases trap catch relative to difference. Also, in 1994 and 1995, we sampled in unbaited traps, but emissions between O.5 and 2.0 Cs. melanura habitat, where unbaited CDC light mg/h attract signiflcantly fewer host-seeking fe- traps took the highest numbers of Cs. melanura. ln males than CO, released at 4fi)-500 mVmin. 1996, we sampled near a forested marsh in western The CDC+COr*octenol traps collected signifi- Massachusetts in an area ideal for Aedes, Anophe- cantly greater numbers of host-seeking females les, and, Cq. perturbaw but not for Cs. melanura. than flicker*COr*octenol (P Because the CDC light trap no longer sampled Cs. CDC+COr+octenol traps collected some species melanura, no significant difference was found be- not found in other traps, such as Aedes implicatus tween the total number of mosquitoes collected in Vockeroth, Aedes intrudens Dyar, and Wyeomyia it and in the AB flicker light trap. smithii (Coquillet). Overall, there was no signifi- The AB flicker*octenol and CDC+octenol traps cant difference between CDC+COr+octenol were not significantly different from one another (p and CDC+CO, (P O.23) or between : 0.85). Both traps had a high standard deviation flicker+COr+octenol and flicker*CO, (P : O.39). and neither increase was significant. Still, the trend Light traps supplemented with CO, alone may suggeststhat octenol could play a role in mosquito collect lower numbers overall than traps with CO, sampling, especially for mosquitoes that feed on plus octenol, but CO, alone was more effective for large mammals. sampling Ae. canadensis and Cq. perturbans (Table Previous authors have reported an increase in 2). Tlte addition of octenol to CO2 may increase collections of Cq. perturbans, Cx. salinarius, arrd catch size for some species and decreaseit for oth- Aedes spp. collected with an emission of 3.0 mg/h ers (Kline et al. 1990a, 1990b).Studies have shown of octenol to CDC light traps (Kline et al. 1990a, that light traps supplemented with a combination of 1991b).We used octenol emission of 0.5 mg/tr add- CO, plus octenol have collected greater numbers of ed to AB flicker light traps in 1994 and 1995, Aedes spp., Cq. perturbans, and Cx. salinarius which may have been inadequate to increase trap (Kline et al. 1990a, 1991b). We sampled lower catch significantly. We increased emission rate to numbers of Ae. canadensis and Cq, perturbans with 2.O mglh in 1996 and did sample more mosquitoes, this combination in southeasternMassachusetts. In emphasizing the role of this volatile material in the western Massachusetts,we sampled more Ae. can- host-seeking behavior of some species, but the in- adensis and Cq. perturbans per night with the same creasestill was not significant. combination. This chemical combination also has The AB flicker*CO, and CDC+CO, trap yields yielded ambiguous results with An. quadrimacula- were not significantly different from one another (P rus (Kline et al. 1990a, 1990b, l99la). We noticed : 0.84). The addition of CO, to an AB flicker trap collections of Cx. salinarius and Anopheles spp. tn- increased average yield from 7 to 74 (SD : 10a) creased signiflcantly with this combination, but DECEMBER 1997 Snraplwc Mpnroos FoR EEE Vnus nt MASSAcHUSETTS 347 sample sizes varied dramatically from site to site, for mosquitoes associated with irrigated rice fields rn emphasizing the importance of regional differences Arkansas. J. Am. Mosq. Control Assoc. 7:165-169. R. Wood and J. A. Cornell. 1991b. In- in sampling and the danger of exffapolating light Kline. D. L., J. effects of l-octen-3-ol and carbon dioxide on (and population) teractive trap data from one site to another. rnosquito (Diptera: Culicidae) surveillance and control. BecauseCx. salinariur and the 2 Anophelesspecies J. Med. Entomol. 28:254--258. Qtunctipennis and quadrimaculatus) have similar Kline, D. L., J. R. Wood and C. D. Morris. 1990a. Eval- flight ranges and nocturnaVseasonalperiodicities, a uation of l-octen-3-ol as an attractant for Coquillettidia light trap supplemented with CO, and octenol could perturbans, Mansonia spp. and Culex spp. associated be more effective for sampling these 3 species in with phosphate mining operations. J. Am. Mosq. Con- particular. trol Assoc.6:605-611. L., W. Takken, J. R. Wood and D. A. Carlson. AB light traps appear to sample Kline. D. The CDC and 1990b. Field studies on the potential of butanone, car- about the same numbers of host-seeking females bon dioxide, honey extract, l-octen-3-ol, llactic acid whether unbaited or supplemented with octenol or and phenols as attractants for mosquitoes. Med. Vet. COr. Only CDC light traps supplemented with both Entomol. 4:383-391. chemical attractants were significantly better than LaSalle, M. W. and M. E. Dakin. 1982. Dispersal of Ca- AB light traps with this combination. The CDC lex salinnrius in southwestern Louisiana. Mosq. News cost somewhat less, but AB traps have the 42:543-55O. traps Relative abundance and parity no Magnarelli, L. A. 1974. following potential advantages:1) collect almost of mosquitoes collected in dry-ice baited and unbaited nontarget insects (e.g., moths, beetles, crane flies, CDC miniature light traps. Mosq. News 35:350-353. wasps); 2) are easier to assembleand disassemble; Matsumoto. B. M. and H. K. Maxfield. 1985. A com- 3) are lighter, more compact, and easier to suspend; parison of female Culiseta melanura captured in New 4) are easier to connect to COr gas cylinders, or Jersey and CDC light traps in southeastern Massachu- add dry ice or octenol packets to; and 5) use less setts. J. Am. Mosq. Control Assoc. l:90-92. and operational analysis battery power (flicker light option). Morris, C. D. 1981. A structural of diurnal resting shelters for mosquitoes @iptera: Cu- licidae). J. Med. Entomol. 18:419-424. REFERENCES CTTED Nasci, R. S. and J. D. Edman. 1981. Blood-feeding patterns of Culiseta melanura (Diptera: Culicidae) and as- Buckley, D., R. Timperi, M. Tobin and B. Werner. 1994. sociated sylvan mosquitoes in southeastern Massachu- Comparison of three trapping methods used to collect setts eastern equine encephalitis enzootic foci. J. Med. mosquitoes in Massachusetts. J. Fla. Mosq. Control As- Entomol. 18:493-5O0. soc. 65:8-l l. Newhouse, V. E, R. W. Chamberlain, J. G. Johnston and Carestia, R. R. and L. B. Savage. 1967. Effectiveness of W D. Sudia. 1966. Use of dry ice to increase mosquito carbon dioxide as a mosquito attractant in the CDC catches of the CDC miniature light trap. Mosq. News miniature light trap. Mosq. News 27:9O-92. 26:30-35. Edman, J. D., R. Timperi and B. Wemer. 1993. Epide- Ninivaggi,D. andS. Guirgis.1994. Mosquito control and (EEE) miology of eastern equine encephalitis in Massachu- eastem equine encephalitis in Suffolk County, 4fth Annu. Meet. Northeastem setts. J. Fla. Mosq. Control Assoc. 64:84-96. NY in 1994. Proc. Assoc., Portsmouth, NH, p. 30. Grady, G. F, H. K. Maxfield, S. W. Hildreth, R. Timperi, Mosq. Control Sudia, W. D. and R. W. Chamberlain. 1962. Battery-op- R. Gilfillan, B. J. Rosenau, D. B. Francy, C. H. Calisher, erated light-trap, an improved model. Mosq. News 22: L. C. Marcus and M. A. Madoff. 1978. Eastern equine t26-129. encephalitis in Massachusetts, 1957-1976: a prospec- Sudia. W. D.. R. W. Chamberlain and P H. Coleman. tive study centered upon analyses of mosquitoes. Am. 1968. Arbovirus isolations from mosquitoes collected J. Epidemiol. lO7 :17 O-178. in south Alabama, 1959-1963, and serologic evidence Headlee, T. J. 1932. The development of mechanical of human infection. Am. J. Epidemiol. 87l-112-126. fauna some equipment for sampling the mosquito and Takken, W. and D. L. Kline. 1989. Carbon dioxide and results of its use. Proc. N. J. Mosq. Exterm. Assoc. 19: l-octen-3-ol as mosquito attractants. J. Am. Mosq. Con- t06-r26. trol Assoc. 5:311-316. Howard,J. J., C. D. Morris,D. E. Emordand M. A. Gray- Vaidyanathan, R. and J. D. Edman. 1997. Sampling with son. 1988. Epizootiology of eastem equine encephalitis light traps and human bait in epidernic foci for eastern virus in upstate New York, USA. VII. Virus surveil- equine encephalomyelitis virus in southeastern Massa- lance 1978-85, description of 1983 outbreak, and series chusetts. J. Am. Mosq. Control Assoc. 13:348-355. conclusions. J. Med. Entomol. 25:.5Ol-514. Weathersbee, A. A. and M. V. Meisch. 199O. Dispersal Kline. D. L.. D. A. Dame and M. V. Meisch. 1991a. Eval- of Anopheles quadrimaculatls (Diptera: Culicidae) in uation of l-octen-3-ol and carbon dioxide as attractants Arkansas ricelands. Environ. Entomol. 19:.961-965.