Carbon Dioxide Released from Human Skin: Effect of Temperature and Insect Repellents D. A. CARLSON, C. E. SCHRECK, AND R. J. BRENNER Medical and Veterinary Entomology Research Laboratory, USDA-ARS, P.O. Box 14565, Gainesville, Florida 32604
KEY WORDS Insecta, host-odor, Aedes aegypti, carbon dioxide
T H E ROLE OF CARBON DIOXIDE (COJ
in the at-
traction of mosquitoes to hosts has been the subject of numerous laboratory studies (Rudolfs 1922; Gouck & Gilbert 1962; Mayer & James 1969, 1970) that have shown CO 2 to be an activator rather than an actual attractant. However, field experiments (Schreck et al. 1970, Gillies 1980) have shown that CO 2 can attract mosquitoes to traps. Acree et al. (1968) reported that the major component in material isolated from humans that was active as an attractant for Aedes aegypti (L.) was L(+) lactic acid, together with a small (2.5 ml/min) but essential amount of CO 2 . Gillies (1980) reviewed the role of CO 2 in host-finding by mosquitoes, and concluded that it has two actions: (1) attraction causing orientation toward the host, but causing only activation of mosquitoes in the absence of moving air currents, and (2) a "combined action with warm, moist convection currents at close range and with odor factors at a distance from the host." Several researchers have examined the potential roles of CO 2 in attraction. Frame et al. (1972) measured continuous CO 2 emission from human hands and arms with a Luft-type infrared analyzer, and showed increases in CO 2 output with vigorous exercise. Room air was collected with and without human effluent in large Teflon bags, and then was passed into an olfactometer and evaluated as a mosquito attractant (Price et al. 1979). Addition of CO 2 did not affect the attracThis article reports the results of research only. Mention of a proprietary product does not constitute an endorsement or a recommendation for its use by USDA.
tion of mosquitoes to the airstream until it was well above the physiological range. Gouck & Bowman (1959) sealed arms in a rubber glove for 1 h, and measured the excreted water and CO 2 (the latter by titration). This static procedure used unnatural and uncomfortable circumstances that seriously could affect the measurement of CO 2 ; differences could result when measurements were made at room temperature. We thought that the change in CO 2 evolved from treating skin with commonly used repellents was unlikely to affect mosquito attraction and that any differences in CO 2 output between unequally attractive test subjects could be compensated by using compressed gas. In this study we determined the effect of small changes in temperature on the continuously measured evolution of CO 2 from treated human arms; the effect of three commonly used repellents on CO 2 production; and whether supplemented CO 2 influenced the activity of mosquitoes exposed to air that had passed over hands of subjects unequally attractive to mosquitoes. Materials and Methods CO 2 Detection. All compressed gases were supplied from commercial sources. Flow meters were calibrated before and after tests with a bubblemeter. The concentration of CO 2 was determined with a Luft type infrared analyzer (MSA Model 200 LIRA Analyzer, Mine Safety Appliances, Pittsburgh, Pa.) by continuously passing effluent from a sampling cylinder directly into the analyzer. The detector indicated the concen-
Downloaded from http://jme.oxfordjournals.org/ by guest on June 8, 2016
J. Med. Entomol. 29(2): 165-170 (1992) ABSTRACT Measurement with an infrared analyzer of CO 2 given off by the hands of human volunteers under laboratory conditions showed that they continuously produced CO 2 at the rate of 1.0-1.8 ml/h. Increased production of CO 2 was observed with increase in temperature for all subjects. Treatment of subjects with three insect repellents or ethanol resulted in a short-term drop in CO 2 production, after which it returned to pretreatment levels. Olfactometer studies showed no correlation between the amount of CO 2 produced by hands and the attractancy of the subjects to host-seeking female Ae. aegypti (L.). The supplemental addition of five times the amount of CO 2 given off by the hands did not affect attractancy of subjects to mosquitoes. The amount of CO 2 released by hands is negligable compared to ambient levels of 300 ppm, and it is unlikely to be attractive at this level of release by itself.
166
Vol. 29, no. 2
JOURNAL OF MEDICAL ENTOMOLOGY
beaker containing the solution and the excess permitted to drip off. Ten minutes later, the hand was sealed into the sampling cylinder and tested for CO 2 output as described above. Olfactometer Tests. Test volunteers were four white and one black adult males. The dual port olfactometer system and test methods determined attraction to laboratory reared Ae. aegypti by trapping the responding females (Schreck et al. 1967, Acree et al. 1968, Smith et al. 1970). Two subjects were compared by simultaneously exposing effluent from their hands in the olfactometer, side by side, in a competitive test against 125 3-d-old female Ae. aegypti in a 3-min test. CO2 Supplementation of Host Odor. The hands of two subjects of unequal attractiveness to mosquitoes were compared against Ae. aegypti with and without supplemental CO 2 . The added CO 2 was a reference gas containing 74 ppm CO 2 in nitrogen metered through a polyethylene tube into the olfactometer at a flow rate yielding 7.5 ml/h CO 2 . Ratios of attractiveness (Ar) with CO 2 added to either or both subjects were calculated according to the formula: coo where Pco2 *s * n e proportion of mosquitoes attracted when CO 2 was added to either (or both) subjects, and P o is the proportion of mosquitoes attracted without additional CO 2 . Separate ratios were calculated for each hand of each subject, and were compared statistically using least significant difference analysis of variance procedures. Chemicals. A suspension of mineral oil and ethanol (1:1) and 95% ethanol were used as control treatments. The standard repellents, deet (N,N-3-methylbenzamide [formerly N,N-diethylmeta-toluamide]), dimethyl phthalate (DMP), and 2-ethyl-l,3-hexanediol (EHD) were dissolved in ethanol at 50%, or used from commercial stocks with no dilution. Results Carbon Dioxide Output from Untreated Hands. The mean output of CO 2 from untreated hands of five subjects are given in Table 1. Analysis of variance revealed that differences between morning and afternoon production of CO 2 were not significant (F = 0.45; df = 1, 13; P = 0.51). Similarly, differences between right and left hands were not significant (F = 0.23; df = 1, 13; P = 0.64). However, there were differences among subjects (F = 11.7; df = 4,13; P = 0.0003). Output by subjects NS and TN were similar (26.2 and 24.2 ppm, respectively), and did not differ significantly (Table 1); other subjects produced less CO 2 than NS and TN.
Downloaded from http://jme.oxfordjournals.org/ by guest on June 8, 2016
tration of one component of a mixture by comparing the infrared absorption of gases in the flow chamber versus the calibration chamber. Because gases absorb infrared energy in different spectral regions, CO, CO 2 , H 2 O vapor, NH 3 , and CH 4 can be determined with a sensitivity of a few parts per million (ppm) in the presence of other gases in room air. A strip chart recorded the output of the analyzer directly in parts per million, and was calibrated daily with a reference gas mixture containing 42 ppm of CO 2 in nitrogen. CO2 Collection. The sampling cylinder that held the subject's hand was made of clear polystyrene plastic, 11.7 cm in diameter and 13.8 cm deep. One end of the cylinder was sealed except for two plastic tubes (0.9 cm inner diameter) cemented in holes in the end as an outlet and inlet for the carrier gas. Rubber tubing (0.9 cm inner diameter) was attached to the inlet tube inside the container and coiled around the interior wall. Holes were cut in the rubber tubing so the carrier gas could flush the CO 2 from around the wrist forward to the outlet tube. A thermometer was sealed into the side of the cylinder. A rubber sleeve made from a surgical glove was sealed to the open end of the container. A plastic sleeve was attached over the rubber sleeve as an added barrier to external contamination. Two cold traps and a 0.3-/im filter were located between the sampling cylinder and the analyzer. The first cold trap contained ice and the second ice-salt, to ensure that water given off by the hand was removed from the gas stream and was not condensed in the detector cell. The CO 2 emitted from untreated hands of five subjects was monitored continuously while they were subjected to small changes in temperature. For each sampling, the hand was inserted into the sampling cylinder and the rubber sleeve was sealed tightly around the wrist. The plastic sleeve then was pulled over the rubber sleeve, flushed with nitrogen to remove atmospheric CO 2 , and sealed over the upper wrist. The sampling cylinder was flushed for 4 min with dry, CO2-free nitrogen gas, sealed, and held for 1 min with no gas flow. The system then was attached to the nitrogen source at a flow rate of 1 liter/min, to flush CO 2 emanating from the hand through the cold traps and filter, and into the analyzer. The recorder would show a sharp rise and fall. When stabilized, a 5-min test was begun. At the conclusion of each test, the average reading was recorded as a subject's CO 2 output in parts per million, together with the temperature. Data were subjected to analysis of variance, testing the effects of subjects, hands, and time (SAS Institute 1988). Hands of the subjects each were treated with one of the three standard repellents, either diluted 50% in ethanol or at full strength. Hands were treated by submerging them in a 1-liter
March 1992
CARLSON ET AL.:
CO 2
167
PRODUCTION AND MOSQUITO ATTRACTION
Table 1. CO2 emitted from left and right human hands in morning and afternoon
Subject
Left Morning
NS HG TN DG DS
25.6 ± 19.8 ± 22.2 ± 18.7 ± 21.6 ±
2.0 0.7 1.7 1.5 3.1
Mean ± SD CO 2 emitted by hands, ppm (5 min) Right Afternoon Morning Afternoon 24.7 ± 23.2 ± 27.7 ± 20.0 ± 18.1 ±
1.2 1.4 1.2 0.7 0.6
27.9 ± 20.4 ± 23.2 ± 18.6 ± 18.9 ±
1.9 0.9 2.2 0.6 2.6
26.7 20.4 23.7 21.0 16.9
± ± ± ± ±
2.0 0.9 2.1 1.8 1.5
26.2 dt 2.0A 21.1 dt 1.4BC 24.2 it l.OAB 19.6 ibO.6C 18.8 dt 1.1C
Means followed by the same letter were not significantly different (P > 0.05).
creased CO 2 output by 10.6%, whereas ethanol alone reduced CO 2 output by 41.5% (Table 2). Hands treated with full strength deet, EHD, and DMP showed 25, 40, and 16% decreases in CO 2 output, respectively. Effect of Supplemental Carbon Dioxide on Host Attraction. In earlier studies (Smith et al. 1970), subject NS invariably was more attractive to mosquitoes than was subject TN. In our competitive tests (Fig. 2), NS again was always more attractive to Ae. aegypti than TN; without added CO 2 , NS attracted an average of 47% of the mosquitoes, compared with 24% attracted by TN. Yet, CO 2 output of the volunteers was not significantly different—NS, 26.3 ppm (1.58 ml/h), and TN, 24.2 ppm (1.45 ml/h) (t = 0.19, df = 6, equal variances, from Table 1). However, when 7.5
30E Q.
25-
0.01) than the same hand with added CO 2
March 1992
CARLSON ET AL.: CO 2 PRODUCTION AND MOSQUITO ATTRACTION
observed. The emissions of CO 2 measured here (1.8 x 10~5 ml/cm2/min) compared fairly well with those of Frame et al. (1972), even though they did not report whether their subjects were sitting or standing. Frame et al. (1972) found that 3.4 x 10~5 ml/cm2/min of CO 2 was produced from human hands and forearms, with 2.5 times more emitted from the hand (4.6 x 10~5 ml/cm2/ min) than the forearm. The emission rate increased by 2.5 times after vigorous exercise or after soaking the arm in warm water and sealing it in a humid glass chamber at high humidity. We observed the same phenomena in preliminary tests. Our observations are consistent with Gouck & Bowman (1959) who found that the treatment of the subjects' arms with the repellents DMP and EHD reduced the CO 2 output in two subjects and deet reduced it in a third. They concluded that repellents affect CO 2 emission from the skin over short but not extended time periods. Price et al. (1979) concluded that airborne human emanations other than CO 2 and water attracted Anopheles quadrimaculatus Say, and suspected the presence of other biologically active materials beside lactic acid. They tested human emanations collected in large Teflon bags and showed that an unattractive collection was not made attractive by adding moderate amounts of CO 2 or water to either the emanations bag or the control bag. Also, Schreck et al. (1981) reported that residues left by human hands on glass surfaces were attractive to female Ae. aegypti in a dual port olfactometer without the addition of CO 2 . Recently, human hands and foreheads were found to be the most active body surfaces in producing residues on glass that were attractive for several hours to several species of mosquitoes (Schreck et al. 1990). We feel that the amounts of CO 2 produced from these glass surfaces are negligible compared with the nearly 300 ppm of CO 2 present in the ambient airstream, as are the amounts produced by a hand in this study (18 to 32 ppm). The CO 2 released from one hand supplemented by five times that amount still is only =0.017% of the amount of CO 2 released for attraction of biting insects in the field, typically 1 liter/min. (D. L. Kline, personal communication). Thus, decreased CO 2 production does not appear to be a major factor in determining whether mosquitoes will avoid a repellent-treated host. Acknowledgment We thank D. Godwin, H. Gouck, T. Newsome, D. Smith, and N. Smith for their technical assistance.
References Cited Acree, F., Jr., R. B. Turner, H. K. Gouck, M. Beroza & N. Smith. 1968. L-Lactic acid: a mosquito at-
Downloaded from http://jme.oxfordjournals.org/ by guest on June 8, 2016
which mosquitoes were attracted to the hand. The nitrogen flow rate of 1 liter/min was used because Acree et al. (1968) showed that filtered air passed over a human hand at this rate was very attractive to Ae. aegypti in an olfactometer. Also, no condensation of moisture in the plastic container was observed at this flow rate. In contrast, static (discontinuous) tests (Gouck & Bowman 1959) quantified CO 2 by titration of condensed moisture collections from a hand and arm sealed for 1 h in a rubber glove of unknown porosity. The quantity of CO 2 produced was correlated with the number of Ae. aegypti attracted to the subjects. The amount of water collected from the cold traps from one subject (NS) in our preliminary studies was 4.0 ml in a 200-min test, or 1.2 ml/h for a hand of =400 cm2 surface, or 300 mg/h per 100 cm2 of skin. This rate was consistent with Gouck & Bowman (1959) who reported that human arms released 117-170 mg of water/h per 100 cm 2 of skin. The production of CO 2 was reduced by 100% ethanol within minutes after treatment, indicating that CO 2 reduction by 50% ethanol-repellent solutions was caused by the ethanol diluent and not the repellent. Treatment with volatile, soonevaporated ethanol appeared to cool the skin surface temporarily, but did not reduce the attractancy of arms to mosquitoes. For this reason, it was used for many years as a diluent in repellent testing and in commercial repellent products. Thus, reduction of C 0 2 emission is apparently not the reason for repellency (or reduced attractancy) when repellents are applied to skin. Variation in CO 2 output between subjects NS and TN was not significant (i.e., the variation in 5 min for each subject exceeded the differences between them). Thus, CO 2 output was not the major determining factor in host attractancy between TN and NS, nor was it probably responsible for variation among the other subjects, because host differences could not be equalized by adding CO 2 . The amount of CO 2 added was ~5 times the calculated average amount given off by two hands as determined in previous tests. This quantity was about 1/30 of the amount injected together with L(+) lactic acid to attract Ae. aegypti in the same olfactometer system (Smith et al. 1970). We feel that the quantity of CO 2 was well below the minimum amount necessary to activate mosquitoes in this system. We have ignored the possible effect of CO 2 in expired air in this system, although we recognize that it may be an important attractant in nature. Our protocol for the test subjects required them to sit quietly in a reclining chair. It became clear during preliminary tests that animated movement, talking, laughing, or telling jokes would cause sharp increases in CO 2 production, nearly doubling its release. We attempted to inhibit such behavior during measurement periods, but some variation in CO 2 production was
169
170
JOURNAL OF MEDICAL ENTOMOLOGY
Rudolfs, W. 1922. Chemotropism of mosquitoes. NJ. Agric. Exp. Sta. Bull. No. 367. SAS Institute. 1988. SAS/STAT users guide. Release 6.03 ed. SAS Institute, Cary, N.C. Schreck, C. E., H. K. Gouck & N. Smith. 1967. An improved olfactometer for use in studying mosquito attractants and repellents. J. Econ. Entomol. 60: 1188-1190. Schreck, C. S., H. K. Gouck & K. H. Posey. 1970. An experimental Plexiglas mosquito trap utilizing CO2Mosq. News. 30: 641-645. Schreck, C. E., D. A. Carlson, N. Smith, G. D. Price, D. Haile & D. R. Godwin. 1981. A material isolated from human hands that attracts female mosquitoes. J. Chem. Ecol. 8: 429-438. Schreck, C. E., D. L. Kline & D. A. Carlson. 1990. Mosquito attraction to substances from the skin of different humans. J. Amer. Mosq. Control Assoc. 6: 406-410. Smith, C. N., N. Smith, H. K. Gouck, D. E. Weidhaas, I. H. Gilbert, M. S. Mayer, B. J. Smittle & A. Hofbauer. 1970. L-Lactic acid as a factor in the attraction of Ae. aegypti (Diptera: Culicidae) to human hosts. Ann. Entomol. Soc. Am. 63: 760-770. Received for publication 24 January 1990; accepted 28 October 1991.
Downloaded from http://jme.oxfordjournals.org/ by guest on June 8, 2016
tractant isolated from humans. Science 161: 1346— 1347. Frame, G. W., W. G. Strauss & H. I. Maibach. 1972. Carbon dioxide emission of the human arm and hand. J. Invest. Dermatol. 59: 155-159. Gillies, M. T. 1980. The role of carbon dioxide in host-finding by mosquitoes (Diptera: Culicidae): a review. Bull. Entomol. Res. 70: 525-532. Gouck, H. K. & M. C. Bowman. 1959. Effect of repellents on the evolution of carbon dioxide and moisture from human arms. J. Econ. Entomol. 52: 1157-1159. Gouck, H. K. & I. H. Gilbert. 1962. Responses of mosquitoes and stable flies to a man in a lightweight rubber diving suit. J. Econ. Entomol. 55: 386-392. Mayer, M. S. & J. D. James. 1969. Attraction of Aedes aegypti (L.): responses to human arms, carbon dioxide and air currents in a new type of olfactometer. Bull. Entomol. Res. 58: 629-642. 1970. Attraction of Aedes aegypti (L.) II. Velocity of reaction to host with and without additional carbon dioxide. Entomol. Exp. Appl. 13: 47-53. Price, G. D., N. Smith & D. A. Carlson. 1979. The attraction of female mosquitoes (Anopheles quadrimaculatus Say) to stored human emanations in conjunction with adjusted levels of relative humidity, temperature and carbon dioxide. J. Chem. Ecol. 5: 383-395.
Vol. 29, no. 2
KEY WORDS Insecta, host-odor, Aedes aegypti, carbon dioxide
T H E ROLE OF CARBON DIOXIDE (COJ
in the at-
traction of mosquitoes to hosts has been the subject of numerous laboratory studies (Rudolfs 1922; Gouck & Gilbert 1962; Mayer & James 1969, 1970) that have shown CO 2 to be an activator rather than an actual attractant. However, field experiments (Schreck et al. 1970, Gillies 1980) have shown that CO 2 can attract mosquitoes to traps. Acree et al. (1968) reported that the major component in material isolated from humans that was active as an attractant for Aedes aegypti (L.) was L(+) lactic acid, together with a small (2.5 ml/min) but essential amount of CO 2 . Gillies (1980) reviewed the role of CO 2 in host-finding by mosquitoes, and concluded that it has two actions: (1) attraction causing orientation toward the host, but causing only activation of mosquitoes in the absence of moving air currents, and (2) a "combined action with warm, moist convection currents at close range and with odor factors at a distance from the host." Several researchers have examined the potential roles of CO 2 in attraction. Frame et al. (1972) measured continuous CO 2 emission from human hands and arms with a Luft-type infrared analyzer, and showed increases in CO 2 output with vigorous exercise. Room air was collected with and without human effluent in large Teflon bags, and then was passed into an olfactometer and evaluated as a mosquito attractant (Price et al. 1979). Addition of CO 2 did not affect the attracThis article reports the results of research only. Mention of a proprietary product does not constitute an endorsement or a recommendation for its use by USDA.
tion of mosquitoes to the airstream until it was well above the physiological range. Gouck & Bowman (1959) sealed arms in a rubber glove for 1 h, and measured the excreted water and CO 2 (the latter by titration). This static procedure used unnatural and uncomfortable circumstances that seriously could affect the measurement of CO 2 ; differences could result when measurements were made at room temperature. We thought that the change in CO 2 evolved from treating skin with commonly used repellents was unlikely to affect mosquito attraction and that any differences in CO 2 output between unequally attractive test subjects could be compensated by using compressed gas. In this study we determined the effect of small changes in temperature on the continuously measured evolution of CO 2 from treated human arms; the effect of three commonly used repellents on CO 2 production; and whether supplemented CO 2 influenced the activity of mosquitoes exposed to air that had passed over hands of subjects unequally attractive to mosquitoes. Materials and Methods CO 2 Detection. All compressed gases were supplied from commercial sources. Flow meters were calibrated before and after tests with a bubblemeter. The concentration of CO 2 was determined with a Luft type infrared analyzer (MSA Model 200 LIRA Analyzer, Mine Safety Appliances, Pittsburgh, Pa.) by continuously passing effluent from a sampling cylinder directly into the analyzer. The detector indicated the concen-
Downloaded from http://jme.oxfordjournals.org/ by guest on June 8, 2016
J. Med. Entomol. 29(2): 165-170 (1992) ABSTRACT Measurement with an infrared analyzer of CO 2 given off by the hands of human volunteers under laboratory conditions showed that they continuously produced CO 2 at the rate of 1.0-1.8 ml/h. Increased production of CO 2 was observed with increase in temperature for all subjects. Treatment of subjects with three insect repellents or ethanol resulted in a short-term drop in CO 2 production, after which it returned to pretreatment levels. Olfactometer studies showed no correlation between the amount of CO 2 produced by hands and the attractancy of the subjects to host-seeking female Ae. aegypti (L.). The supplemental addition of five times the amount of CO 2 given off by the hands did not affect attractancy of subjects to mosquitoes. The amount of CO 2 released by hands is negligable compared to ambient levels of 300 ppm, and it is unlikely to be attractive at this level of release by itself.
166
Vol. 29, no. 2
JOURNAL OF MEDICAL ENTOMOLOGY
beaker containing the solution and the excess permitted to drip off. Ten minutes later, the hand was sealed into the sampling cylinder and tested for CO 2 output as described above. Olfactometer Tests. Test volunteers were four white and one black adult males. The dual port olfactometer system and test methods determined attraction to laboratory reared Ae. aegypti by trapping the responding females (Schreck et al. 1967, Acree et al. 1968, Smith et al. 1970). Two subjects were compared by simultaneously exposing effluent from their hands in the olfactometer, side by side, in a competitive test against 125 3-d-old female Ae. aegypti in a 3-min test. CO2 Supplementation of Host Odor. The hands of two subjects of unequal attractiveness to mosquitoes were compared against Ae. aegypti with and without supplemental CO 2 . The added CO 2 was a reference gas containing 74 ppm CO 2 in nitrogen metered through a polyethylene tube into the olfactometer at a flow rate yielding 7.5 ml/h CO 2 . Ratios of attractiveness (Ar) with CO 2 added to either or both subjects were calculated according to the formula: coo where Pco2 *s * n e proportion of mosquitoes attracted when CO 2 was added to either (or both) subjects, and P o is the proportion of mosquitoes attracted without additional CO 2 . Separate ratios were calculated for each hand of each subject, and were compared statistically using least significant difference analysis of variance procedures. Chemicals. A suspension of mineral oil and ethanol (1:1) and 95% ethanol were used as control treatments. The standard repellents, deet (N,N-3-methylbenzamide [formerly N,N-diethylmeta-toluamide]), dimethyl phthalate (DMP), and 2-ethyl-l,3-hexanediol (EHD) were dissolved in ethanol at 50%, or used from commercial stocks with no dilution. Results Carbon Dioxide Output from Untreated Hands. The mean output of CO 2 from untreated hands of five subjects are given in Table 1. Analysis of variance revealed that differences between morning and afternoon production of CO 2 were not significant (F = 0.45; df = 1, 13; P = 0.51). Similarly, differences between right and left hands were not significant (F = 0.23; df = 1, 13; P = 0.64). However, there were differences among subjects (F = 11.7; df = 4,13; P = 0.0003). Output by subjects NS and TN were similar (26.2 and 24.2 ppm, respectively), and did not differ significantly (Table 1); other subjects produced less CO 2 than NS and TN.
Downloaded from http://jme.oxfordjournals.org/ by guest on June 8, 2016
tration of one component of a mixture by comparing the infrared absorption of gases in the flow chamber versus the calibration chamber. Because gases absorb infrared energy in different spectral regions, CO, CO 2 , H 2 O vapor, NH 3 , and CH 4 can be determined with a sensitivity of a few parts per million (ppm) in the presence of other gases in room air. A strip chart recorded the output of the analyzer directly in parts per million, and was calibrated daily with a reference gas mixture containing 42 ppm of CO 2 in nitrogen. CO2 Collection. The sampling cylinder that held the subject's hand was made of clear polystyrene plastic, 11.7 cm in diameter and 13.8 cm deep. One end of the cylinder was sealed except for two plastic tubes (0.9 cm inner diameter) cemented in holes in the end as an outlet and inlet for the carrier gas. Rubber tubing (0.9 cm inner diameter) was attached to the inlet tube inside the container and coiled around the interior wall. Holes were cut in the rubber tubing so the carrier gas could flush the CO 2 from around the wrist forward to the outlet tube. A thermometer was sealed into the side of the cylinder. A rubber sleeve made from a surgical glove was sealed to the open end of the container. A plastic sleeve was attached over the rubber sleeve as an added barrier to external contamination. Two cold traps and a 0.3-/im filter were located between the sampling cylinder and the analyzer. The first cold trap contained ice and the second ice-salt, to ensure that water given off by the hand was removed from the gas stream and was not condensed in the detector cell. The CO 2 emitted from untreated hands of five subjects was monitored continuously while they were subjected to small changes in temperature. For each sampling, the hand was inserted into the sampling cylinder and the rubber sleeve was sealed tightly around the wrist. The plastic sleeve then was pulled over the rubber sleeve, flushed with nitrogen to remove atmospheric CO 2 , and sealed over the upper wrist. The sampling cylinder was flushed for 4 min with dry, CO2-free nitrogen gas, sealed, and held for 1 min with no gas flow. The system then was attached to the nitrogen source at a flow rate of 1 liter/min, to flush CO 2 emanating from the hand through the cold traps and filter, and into the analyzer. The recorder would show a sharp rise and fall. When stabilized, a 5-min test was begun. At the conclusion of each test, the average reading was recorded as a subject's CO 2 output in parts per million, together with the temperature. Data were subjected to analysis of variance, testing the effects of subjects, hands, and time (SAS Institute 1988). Hands of the subjects each were treated with one of the three standard repellents, either diluted 50% in ethanol or at full strength. Hands were treated by submerging them in a 1-liter
March 1992
CARLSON ET AL.:
CO 2
167
PRODUCTION AND MOSQUITO ATTRACTION
Table 1. CO2 emitted from left and right human hands in morning and afternoon
Subject
Left Morning
NS HG TN DG DS
25.6 ± 19.8 ± 22.2 ± 18.7 ± 21.6 ±
2.0 0.7 1.7 1.5 3.1
Mean ± SD CO 2 emitted by hands, ppm (5 min) Right Afternoon Morning Afternoon 24.7 ± 23.2 ± 27.7 ± 20.0 ± 18.1 ±
1.2 1.4 1.2 0.7 0.6
27.9 ± 20.4 ± 23.2 ± 18.6 ± 18.9 ±
1.9 0.9 2.2 0.6 2.6
26.7 20.4 23.7 21.0 16.9
± ± ± ± ±
2.0 0.9 2.1 1.8 1.5
26.2 dt 2.0A 21.1 dt 1.4BC 24.2 it l.OAB 19.6 ibO.6C 18.8 dt 1.1C
Means followed by the same letter were not significantly different (P > 0.05).
creased CO 2 output by 10.6%, whereas ethanol alone reduced CO 2 output by 41.5% (Table 2). Hands treated with full strength deet, EHD, and DMP showed 25, 40, and 16% decreases in CO 2 output, respectively. Effect of Supplemental Carbon Dioxide on Host Attraction. In earlier studies (Smith et al. 1970), subject NS invariably was more attractive to mosquitoes than was subject TN. In our competitive tests (Fig. 2), NS again was always more attractive to Ae. aegypti than TN; without added CO 2 , NS attracted an average of 47% of the mosquitoes, compared with 24% attracted by TN. Yet, CO 2 output of the volunteers was not significantly different—NS, 26.3 ppm (1.58 ml/h), and TN, 24.2 ppm (1.45 ml/h) (t = 0.19, df = 6, equal variances, from Table 1). However, when 7.5
30E Q.
25-
0.01) than the same hand with added CO 2
March 1992
CARLSON ET AL.: CO 2 PRODUCTION AND MOSQUITO ATTRACTION
observed. The emissions of CO 2 measured here (1.8 x 10~5 ml/cm2/min) compared fairly well with those of Frame et al. (1972), even though they did not report whether their subjects were sitting or standing. Frame et al. (1972) found that 3.4 x 10~5 ml/cm2/min of CO 2 was produced from human hands and forearms, with 2.5 times more emitted from the hand (4.6 x 10~5 ml/cm2/ min) than the forearm. The emission rate increased by 2.5 times after vigorous exercise or after soaking the arm in warm water and sealing it in a humid glass chamber at high humidity. We observed the same phenomena in preliminary tests. Our observations are consistent with Gouck & Bowman (1959) who found that the treatment of the subjects' arms with the repellents DMP and EHD reduced the CO 2 output in two subjects and deet reduced it in a third. They concluded that repellents affect CO 2 emission from the skin over short but not extended time periods. Price et al. (1979) concluded that airborne human emanations other than CO 2 and water attracted Anopheles quadrimaculatus Say, and suspected the presence of other biologically active materials beside lactic acid. They tested human emanations collected in large Teflon bags and showed that an unattractive collection was not made attractive by adding moderate amounts of CO 2 or water to either the emanations bag or the control bag. Also, Schreck et al. (1981) reported that residues left by human hands on glass surfaces were attractive to female Ae. aegypti in a dual port olfactometer without the addition of CO 2 . Recently, human hands and foreheads were found to be the most active body surfaces in producing residues on glass that were attractive for several hours to several species of mosquitoes (Schreck et al. 1990). We feel that the amounts of CO 2 produced from these glass surfaces are negligible compared with the nearly 300 ppm of CO 2 present in the ambient airstream, as are the amounts produced by a hand in this study (18 to 32 ppm). The CO 2 released from one hand supplemented by five times that amount still is only =0.017% of the amount of CO 2 released for attraction of biting insects in the field, typically 1 liter/min. (D. L. Kline, personal communication). Thus, decreased CO 2 production does not appear to be a major factor in determining whether mosquitoes will avoid a repellent-treated host. Acknowledgment We thank D. Godwin, H. Gouck, T. Newsome, D. Smith, and N. Smith for their technical assistance.
References Cited Acree, F., Jr., R. B. Turner, H. K. Gouck, M. Beroza & N. Smith. 1968. L-Lactic acid: a mosquito at-
Downloaded from http://jme.oxfordjournals.org/ by guest on June 8, 2016
which mosquitoes were attracted to the hand. The nitrogen flow rate of 1 liter/min was used because Acree et al. (1968) showed that filtered air passed over a human hand at this rate was very attractive to Ae. aegypti in an olfactometer. Also, no condensation of moisture in the plastic container was observed at this flow rate. In contrast, static (discontinuous) tests (Gouck & Bowman 1959) quantified CO 2 by titration of condensed moisture collections from a hand and arm sealed for 1 h in a rubber glove of unknown porosity. The quantity of CO 2 produced was correlated with the number of Ae. aegypti attracted to the subjects. The amount of water collected from the cold traps from one subject (NS) in our preliminary studies was 4.0 ml in a 200-min test, or 1.2 ml/h for a hand of =400 cm2 surface, or 300 mg/h per 100 cm2 of skin. This rate was consistent with Gouck & Bowman (1959) who reported that human arms released 117-170 mg of water/h per 100 cm 2 of skin. The production of CO 2 was reduced by 100% ethanol within minutes after treatment, indicating that CO 2 reduction by 50% ethanol-repellent solutions was caused by the ethanol diluent and not the repellent. Treatment with volatile, soonevaporated ethanol appeared to cool the skin surface temporarily, but did not reduce the attractancy of arms to mosquitoes. For this reason, it was used for many years as a diluent in repellent testing and in commercial repellent products. Thus, reduction of C 0 2 emission is apparently not the reason for repellency (or reduced attractancy) when repellents are applied to skin. Variation in CO 2 output between subjects NS and TN was not significant (i.e., the variation in 5 min for each subject exceeded the differences between them). Thus, CO 2 output was not the major determining factor in host attractancy between TN and NS, nor was it probably responsible for variation among the other subjects, because host differences could not be equalized by adding CO 2 . The amount of CO 2 added was ~5 times the calculated average amount given off by two hands as determined in previous tests. This quantity was about 1/30 of the amount injected together with L(+) lactic acid to attract Ae. aegypti in the same olfactometer system (Smith et al. 1970). We feel that the quantity of CO 2 was well below the minimum amount necessary to activate mosquitoes in this system. We have ignored the possible effect of CO 2 in expired air in this system, although we recognize that it may be an important attractant in nature. Our protocol for the test subjects required them to sit quietly in a reclining chair. It became clear during preliminary tests that animated movement, talking, laughing, or telling jokes would cause sharp increases in CO 2 production, nearly doubling its release. We attempted to inhibit such behavior during measurement periods, but some variation in CO 2 production was
169
170
JOURNAL OF MEDICAL ENTOMOLOGY
Rudolfs, W. 1922. Chemotropism of mosquitoes. NJ. Agric. Exp. Sta. Bull. No. 367. SAS Institute. 1988. SAS/STAT users guide. Release 6.03 ed. SAS Institute, Cary, N.C. Schreck, C. E., H. K. Gouck & N. Smith. 1967. An improved olfactometer for use in studying mosquito attractants and repellents. J. Econ. Entomol. 60: 1188-1190. Schreck, C. S., H. K. Gouck & K. H. Posey. 1970. An experimental Plexiglas mosquito trap utilizing CO2Mosq. News. 30: 641-645. Schreck, C. E., D. A. Carlson, N. Smith, G. D. Price, D. Haile & D. R. Godwin. 1981. A material isolated from human hands that attracts female mosquitoes. J. Chem. Ecol. 8: 429-438. Schreck, C. E., D. L. Kline & D. A. Carlson. 1990. Mosquito attraction to substances from the skin of different humans. J. Amer. Mosq. Control Assoc. 6: 406-410. Smith, C. N., N. Smith, H. K. Gouck, D. E. Weidhaas, I. H. Gilbert, M. S. Mayer, B. J. Smittle & A. Hofbauer. 1970. L-Lactic acid as a factor in the attraction of Ae. aegypti (Diptera: Culicidae) to human hosts. Ann. Entomol. Soc. Am. 63: 760-770. Received for publication 24 January 1990; accepted 28 October 1991.
Downloaded from http://jme.oxfordjournals.org/ by guest on June 8, 2016
tractant isolated from humans. Science 161: 1346— 1347. Frame, G. W., W. G. Strauss & H. I. Maibach. 1972. Carbon dioxide emission of the human arm and hand. J. Invest. Dermatol. 59: 155-159. Gillies, M. T. 1980. The role of carbon dioxide in host-finding by mosquitoes (Diptera: Culicidae): a review. Bull. Entomol. Res. 70: 525-532. Gouck, H. K. & M. C. Bowman. 1959. Effect of repellents on the evolution of carbon dioxide and moisture from human arms. J. Econ. Entomol. 52: 1157-1159. Gouck, H. K. & I. H. Gilbert. 1962. Responses of mosquitoes and stable flies to a man in a lightweight rubber diving suit. J. Econ. Entomol. 55: 386-392. Mayer, M. S. & J. D. James. 1969. Attraction of Aedes aegypti (L.): responses to human arms, carbon dioxide and air currents in a new type of olfactometer. Bull. Entomol. Res. 58: 629-642. 1970. Attraction of Aedes aegypti (L.) II. Velocity of reaction to host with and without additional carbon dioxide. Entomol. Exp. Appl. 13: 47-53. Price, G. D., N. Smith & D. A. Carlson. 1979. The attraction of female mosquitoes (Anopheles quadrimaculatus Say) to stored human emanations in conjunction with adjusted levels of relative humidity, temperature and carbon dioxide. J. Chem. Ecol. 5: 383-395.
Vol. 29, no. 2
