Efficacy of the botanical repellents geraniol, linalool, and citronella against mosquitoes

2 Journal of Vector Ecology June 2009 Efficacy of the botanical repellents geraniol, linalool, and citronella against mosquitoes Günter C. Müller 1, Amy Junnila 2, Jerry Butler 3, Vassiliy D. Kravchenko 4, Edita E. Revay 5, Robert W. Weiss 6, and Yosef Schlein 1 1 Department of Parasitology, Kuvin Centre for the Study of Infectious and Tropical Diseases, The Hebrew University, Hadassah-Medical School, Jerusalem, Israel, 91120 2 Department of Parasitology, McGill University, Macdonald Campus, Ste-Anne-de-Bellevue, Quebec H9X 3V9, Canada 3 Medical-Veterinary Entomology and Nematology Department, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL 32611-0620, U.S.A. 4 Department of Zoology, Tel Aviv University, Tel Aviv, Israel 5 Bruce Rappaport Faculty of Medicine, Technion, Haifa, Israel 6 Entomological Associates, 102 Beaver Mills Road, Julian, PA 16844, U.S.A. Received 25 February 2007; Accepted 2 December 2008 ABSTRACT: We determined the degree of personal protection provided by citronella, linalool, and geraniol in the form of commercially available candles or diffusers, both indoors and outdoors. Under the uniform conditions of the experiments, all substances repelled significantly more mosquitoes than the unprotected control. Furthermore, the repellents tested were more active when in the form of a continuous release diffuser than in candle form. All candles were 88g containing 5% of the active ingredient and all diffusers contained 20g of 100% active ingredient. Indoors, the repellency rate of citronella candles was only 14% while the repellency rate of citronella diffusers was 68%. The repellency of geraniol candles was 50% while the diffusers provided a repellency rate of 97%. No linalool candles were available for study but linalool diffusers repelled mosquitoes by 93%. Outdoors, citronella diffusers placed 6 m from mosquito traps repelled female mosquitoes by 22%, linalool repelled females by 58%, and geraniol repelled females by 75%. Trap catches were significantly reduced again when diffusers were placed 3 m from the traps. We concluded that geraniol had significantly more repellent activity than citronella or linalool in both indoor and outdoor settings. Journal of Vector Ecology 34 (1): 2-8. 2009. Keyword Index: Monoterpene alcohols, mosquito repellent, citronella, geraniol, linalool, Aedes spp., Culex spp., botanical repellents, plant essential oils, Puerto Rico. INTRODUCTION In recent years, botanical insect repellents have become increasingly popular as viable alternatives to synthetic chemical pest repellents because they reputedly pose little risk to the environment or human health. However, the body of scientific literature documenting bioactivity of plant derivatives is sometimes contradicting and lacking in standardized testing protocols. Therefore, we feel there is a need for increased research on the use of natural or herbalbased repellents to ensure quality and determine the most effective means of application and use. N,N-diethyl-3-methylbenzamide (DEET) remains the gold standard of currently available insect repellents, however, there have been case reports of DEET toxicity in the literature (Zadikoff 1979, Snyder et al. 1986, Osimitz and Grothaus 1995, Osimitz and Murphy 1997) and consumer interest in natural alternative repellents is growing rapidly. A broad spectrum of plants and plant essential oils have been tested as potential insect repellents (for a review see Sukumar et al. 1991). Out of those tested, such a large number of claims of repellent activity were made that the United States Environmental Protection Agency (EPA) formed an advisory committee in 2000 to curb inconsistencies in repellent product performance testing. The committee pointed out that many factors play a role in how effective repellents are, including frequency and uniformity of repellent application (Khan et al. 1972, Gabel et al.1976), the number and species of the organisms attempting to bite, the user s inherent attractiveness to blood-feeding insects, age (Muirhead-Thomson 1951), sex (Gilbert et al. 1966) and size of the potential host (Port and Boreham 1980), and the physical activity level of the potential host. As well, the committee made recommendations as to the duration of the tests, statistical significance of sample size, replication of experiments, and rotation of human volunteer subjects (http://www.epa.gov/oscpmont/sap/meetings/2000/april/ freportapril572000.pdf). Specifically, many claims have been made regarding the repellent properties of citronella essential oil and various terpene alcohols (Hwang et al. 1985, Tawatsin et al. 2001, Barnard and Xue 2004). essential oil is derived from different species of Cymbopogon (citronella grass) and contains some industrially important aromatic compounds such as geraniol. and linalool, isomers of each other, are monoterpene alcohols found in many plant essential oils such as citronella and thyme respectively (Choi et al. 2002,

Vol. 34, no. 1 Journal of Vector Ecology 3 Park et al. 2005). Unfortunately, there seems to be little consistency in the experimental protocols used to test these repellents. Consequently, there are major differences in concentration, application, and test methods (i.e., hand-inbox vs field trials) used to determine bioactivity of natural repellents (Sukumar et al. 1991). Moreover, in the case of citronella oil, counter claims of reduced or absent repellent activity have been made (Lindsay et al. 1996, Centers et al. 2002). This study was conducted to help resolve some of these issues and determine the repellent activity of three commonly used natural repellents (citronella, linalool, and geraniol) in candle or diffuser form, both indoors and outdoors. To ensure quality control, the current study adheres to the above-mentioned EPA guidelines, as well as to currently accepted standards for testing insect repellents (Govere and Durrheim 2007, Barnard et al. 2007). MATERIALS AND METHODS The repellency rates of candles and/or diffusers containing a specific type and concentration of active ingredient (Table 1) were determined in both an indoor and outdoor setting. Indoor experiments Aedes aegypti mosquitoes were bred under insectarium conditions following the recommendations of the EPA. Ae. aegypti was used in evening experiments since this species is active in the evening in Puerto Rico as well as in Israel (Muller and Schlein, unpublished data). There are reports on a third peak of indoor biting activity around 18:00 h in Trinidad, parts of Africa, and Indonesia (Chadee and Martinez 2000). Larvae were reared at 27±1º C, relative humidity 80 ±10%, and photoperiod 16:8 hours (light:dark). Adults were 8 days old, fed 10% sucrose, and received no blood meals before the test. Before the experiments, the insects were starved for 24 h. After each test the mosquitoes were discarded. Four authors of this study (, ) served as the volunteer subjects and were therefore fully informed of the nature and purposes of the test and of any reasonably foreseeable physical and mental health consequences. For the trials, the exposed legs (from knee to ankle) of each volunteer were used as a test area. The skin outside the test area was covered with regular clothes to protect it from mosquito bites. Volunteers wore short trousers and longsleeved shirts. Immediately before each trial, the exposed skin on each volunteer was cleaned with 70% isopropyl alcohol. The volunteers were advised to avoid alcohol, caffeine, and fragrance products (e.g., perfume, cologne, hair spray, lotion, etc.) during the entire test period. Experiments were performed on the northern coastal plain of Puerto Rico, during March 2004, south of the city of San Sebastian. The test site consisted of a dormitory building in a church retreat. Four rooms of the same size and shape, with a single screened window, were situated near each other along a corridor. From these four rooms, all furniture was removed except for two chairs. The doors to the other rooms were closed. During the experiments, the air conditioning and the lights were switched off. The average indoor temperature at the time of the experiments was 24.4 ± 0.8º C with a relative humidity of 74-75%. At 18:00 local time, 200 Ae. aegypti females were released in the sealed corridor and allowed to disperse for 30 min. Prior to the mosquito release, a Coleman Mosquito Deleto diffuser (model# 2950-602) containing one of four cartridges, or one of three candles (Table 1) was placed on a chair in the center of each experimental room and turned on or lit. On the other chair in a corner of the room approximately 2 m from the diffuser or candle, opposite the door, a volunteer with exposed legs sat on a chair prepared to collect landing mosquitoes. At 18:30, the doors to each of the experimental rooms were opened. A head-mounted light covered with red foil was used to improve vision while volunteers collected mosquitoes with an aspirator. After the experiment, the mosquitoes in the aspirators were counted and the remaining mosquitoes in the corridor and rooms were removed the following morning with an entomological net and a vacuum cleaner. On the first day of testing, the diffusers or candles were randomly assigned to the volunteers. The candle and diffuser experiments were conducted for three h on nine consecutive nights; nine nights of candles and nine nights of diffusers for a total of 18 nights. To avoid sex and locational bias, the volunteers and diffusers or candles were rotated to different rooms on successive nights to allow each volunteer to test each set-up. Outdoor experiments A readily available residential outdoor mosquito trap was used in lieu of volunteers to determine the effectiveness of the diffuser repellents outdoors. The model chosen was the Lentek MK01 mosquito trap, sold in the U.S., which uses 120V AC electric power and generates CO 2, heat, and moisture by burning propane. Tests were performed in the tropical area of Puerto Rico at the northwestern coastal plain during early to mid- March near the city of Mayaguez. The test site was along a canal surrounded by pasture and farm land. A slow, but repeatedly changing, airflow was observed during the balance of the test period. Experiments were conducted during the dry season and consequently no rainfall occurred. The average temperature at night was 21.5±0.9º C with a relative humidity of 70-75%. No dramatic weather changes occurred during the trial period. To determine the effective range of the repellents, two experiments were conducted in which four diffusers hung on tripods (approximately 80 cm from the ground) were arranged in squares (3 m and 6 m squares) with an MK01 trap placed in the middle. The diffusers were equipped either with citronella, linalool, geraniol, or empty cartridges. Diffuser/trap squares were distanced 100 m apart and were positioned alternately (active ingredient vs empty cartridge). The squares were rotated on consecutive nights to account for locational bias.

4 Journal of Vector Ecology June 2009 Diffusers and traps were operated from one h before average sunset (18:32) to one h after average sunrise (06:40). There were 12 repetitions on consecutive nights. Traps were emptied and mosquitoes counted daily. The field trial test protocol and site selection were in accordance with EPA published guidelines (Govere and Durrheim 2007, Barnard et al. 2007). Statistical analyses Statistical analysis was carried out using the GraphPad Prism 4.0 statistical package. The numbers of mosquitoes in the rooms used in indoor tests were analyzed using multivariate ANOVA technique with the mean as withinsubject variable. Experiments with diffusers and candles were analyzed separately. Following a significant F score, post-hoc tests (Tukey s Honestly Significant Difference) were used to further distinguish between groups. All differences were considered statistically significant if p<0.05. The number of mosquitoes in the traps used in outdoor tests were analyzed using the ANOVA technique. Following a significant F score, the Tukey s Honestly Significant Difference post-hoc test was used to further distinguish between groups. All differences were considered statistically significant if p < 0.05. The total number of mosquitoes caught in traps and aspirators were converted to mean ± standard deviation. Mean totals were converted to percentage of the unprotected controls by using the following equations: for indoor experiments, the repellency rate (%) = [l - (biting attempts treated / biting attempts control)] x 100. For outdoor experiments, the repellency rate (%) = [l - (total caught treated / total caught control)] x 100. RESULTS After each overnight use, candles and diffusers were weighed to determine the average volume of material expended. The volume expended was the same among the candles, and the same among diffuser cartridges, regardless of composition. The 100% active ingredient diffusers lost an average total weight of only 0.1 g per h, whereas the 5% candles lost an average of 2.2 g per h (Table 1). The candles are only 5% and therefore contain 0.05 g of active ingredient per 1.0 g of candle wax but lose more volume than the diffusers. Assuming that the active ingredient is uniformly distributed throughout the candle wax, the candles would have released 0.11 g of active ingredient per h, which is nearly the same as the amount of active ingredient released by diffusers (0.1 g). Indoor experiments Within the first hour, about 90% of the mosquitoes, which were ready to feed, entered one of the rooms with the volunteers. During the following two hours, very little movement was observed. After three hours, there were no more feeding attempts. On average, geraniol, linalool, and citronella diffusers all provided significant protection to volunteers compared to controls (Figure 1) F (4,5) = 645.6, p<0.0001. Among the diffusers, the geraniol and linalool equipped units provided significantly more protection than citronella equipped units (Tukey-Kramer Test p<0.05). Overall, however, the geraniol diffuser was about twice as effective as the linalool diffuser and about five times as effective as the citronella diffuser in protecting a person from Ae. aegypti indoors. Of the 1,800 female mosquitoes released for the indoor diffuser experiment (200 at each of nine repetitions), 1,103 (61%) tried to feed on the unprotected person (control). Of all the released mosquitoes, 13% did not try to bite any of the volunteers and were removed in the following mornings. Near a citronella diffuser there were 355 feeding attempts, near a linalool diffuser 74 attempts, and near a geraniol diffuser 38 attempts. When the number of mosquitoes collected by the non-protected control volunteer was compared to the number collected by the protected volunteers, it was found that the repellency rate of citronella diffusers was 68%, linalool diffusers 93%, and geraniol diffusers 97%. and citronella candles also provided significant protection to volunteers compared to controls (Figure 2) F (3,6) = 101.9, p<0.0001. Overall, however, the geraniol candle provided significantly more protection than the citronella candle, (Tukey-Kramer Test p<0.05) the geraniol being about five times as effective as the citronella in protecting a person from Ae. aegypti feeding indoors. Of the 1,800 female mosquitoes released for the indoor candle experiment (200 released at each of nine repetitions), 908 (50%) tried to feed on the unprotected volunteer. Of all the released mosquitoes, 14% did not try to bite any of the volunteers and were removed in the following mornings. Near a citronella candle there were 533 feeding attempts and near a geraniol candle there were 106 attempts. When the number of mosquitoes caught by the non-protected control volunteer was compared to the number of mosquitoes caught by the protected volunteers, it was found that the repellency rate of citronella was 41% while the repellency rate of geraniol was 88%. Commercially available geraniol and citronella were more effective in protecting volunteers in diffuser form than in candle form. Outdoor experiments The following common mosquito species were encountered in the study: Culex nigripalpus, Ae. aegypti, Ae. mediovittatus, and Ae. sollicitans. Most of the mosquitoes caught in the 3 m 2 and 6 m 2 experiments were Aedes species; the rest were Culex. There was no significant difference in the response of the two genera towards linalool or geraniol., linalool, and citronella in combination with the Coleman Mosquito Deleto Diffuser significantly reduced mosquito capture (Figure 3; 3 m 2 F (3,9) = 16.2 p< 0.0001; 6 m 2 F (3,44) = 39.3, p < 0.0001). When arranged in a 3x3 m square, geraniol reduced the number of mosquitoes caught by the MK01 mosquito trap by 95.5%, linalool by 88.4%, and citronella by 65.6%. When arranged in a 6x6 m square, geraniol reduced the number of mosquitoes caught by the MK01 mosquito trap by 75%, linalool by 58%, and

Vol. 34, no. 1 Journal of Vector Ecology 5 Table 1. Details of equipment used in experiments. All batteries, cartridges, and candles were changed daily. Active Ingredient and Release Type Concentration (%) Weight (g) Amount Active Ingredient (g/g) Total loss of Weight per h (g) Loss of Active Ingredient per h (g) Composition of Active Ingredient Supplier 1 ** Diffuser 100% 20g 1.0 0.1 0.1 23.18% Nature's Alkamy 34.79% l 11.19% Citronellol 0.72% Linalool 30.12% All Other* 1 Linalool Diffuser 100% 20g 1.0 0.1 0.1 95.54% Linalool Coleman 4.6% Isomers 2 Diffuser 100% 20g 1.0 0.1 0.1 95.54% Fasst Products 4.6% Isomers Diffuser 0% n/a 0.0 n/a n/a n/a Coleman ** Candle 5% 88g 0.05 2.2 0.11 23.18% Wal-Mart 34.79% l 11.19% Citronellol 0.72% Linalool 30.12% All Other* Candle 5% 88g 0.05 2.2 0.11 95.54% Fasst Products 4.6% Isomers Candle 0% 88g 0.0 n/a n/a n/a Wal-mart * oil is comprised of more than 80 closely related terpenes, alcohols, and aldehydes (PMRA 2007). Only relevent components are listed here. ** Type "Java" from Cymbopogon winterianus. n/a: not applicable. 1 Oil. 2 Impregnated low density polyethelene pellets.

6 Journal of Vector Ecology June 2009 # Mosquitoes 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Corridor * Linalool Figure 1. Average number of mosquitoes caught after indoor diffuser experiments (N=9) ± standard deviation. All repellents were significantly more repellent than the control (F(4,5) = 645.6, p < 0.0001). * significantly more repellent than all others (Tukey-Kramer Test p<0.05). # Mosquitoes 120 110 100 90 80 70 60 50 40 30 20 10 0 * Corridor Figure 2. Average number of mosquitoes caught after indoor candle experiments (N=9) ± standard deviation. All repellents were significantly more repellent than control (F(3,6) = 101.9, p < 0.0001). * significantly more repellent than all others (Tukey-Kramer Test p<0.05).

Vol. 34, no. 1 Journal of Vector Ecology 7 # of female mosquitoes 250 200 150 100 50 0 Linalool Linalool 20 3 m ft square 10 6 m ft square Figure 3. Average number of female mosquitoes caught in MK01 mosquito traps (N=12) ± standard deviation. Active compounds were significantly more repellent than control (3 m square: F(3,44) = 39.3, p<0.0001) (6 m square: F(3,9) = 16.2 p<0.0001). citronella by 22%. All compounds provided significantly more protection (Tukey Kramer Test p > 0.05) when placed closer to the MK01 trap. DISCUSSION In a previous study (Lindsay et al. 1996), 3% citronella candles and 5% citronella incense in an outdoor setting increased the repellency rate over the unprotected control by only 42.3 and 24.2%, respectively. In this study, similar results were obtained; a 5% citronella candle increased the repellency rate indoors by only 14%. However, the continuous release citronella diffusers increased the repellency rate indoors by 68%. Since the amount of active ingredient expended by candles and diffusers is nearly identical, one possible explanation for this difference is that the citronella candle releases its compounds with heat from the flame and this may cause the active ingredients to concentrate and move vertically away from the candle with the heat. The diffuser, on the other hand, continuously disperses an active ingredient horizontally. Only diffusers, not candles, were tested for repellency outdoors; the results of which, for citronella, were poor (21%). in candle form used indoors, or in diffuser form used outdoors, does not reduce mosquitoes or mosquito bites to a degree acceptable to the EPA or conventional recommended standards for repellent testing (Govere and Durrheim 2007, Barnard et al. 2007). Currently accepted guidelines require candles, coils, vaporizing mats, or other such products to provide at least a 50% repellency rate to make a reliable claim that the product repels mosquitoes (Govere and Durrheim 2007). However, citronella diffusers when used indoors are fairly repellent, possibly because airflow indoors is minimal and the repellent molecules are not diluted by airflow (i.e., building up vapor pressure). Human forearm tests of topically applied linalool report a 91.7% repellency rate towards Cx. pipiens (Park et al. 2005). No linalool candles were available for evaluation but we report here that linalool diffusers provided a 93% repellency rate indoors and a repellency rate of 58% outdoors. These rates, especially indoors, support the claim that linalool repels mosquitoes (Govere and Durrheim 2007). Studies conducted on the repellent effect of geraniol on mosquitoes are limited and use varying measurements of repellency. One laboratory study claims that 25% geraniol applied topically repels mosquitoes for 3.2 to 4.8 h (Barnard and Xue 2004) and still others claim geraniol repels termites (Blaske and Hertel 2001), human body lice (Mumcuoglu et al. 1996) and some species of flies (McGuire 1984). Under the uniform conditions of our experiments, geraniol repelled significantly more mosquitoes than citronella or linalool, both indoors (97% repellency with diffusers; 50% with candles) and outdoors (95.5% repellency with diffusers at 3 m, 75% repellency with diffusers at 6 m). According to the data collected in the present study, geraniol diffusers, used indoors, fit the repellency criteria outlined by Govere and Durrheim (2007) and Barnard and colleagues (2007), and also strongly suggest that geraniol provides much greater indoor and outdoor protection coverage than the other tested products (Figures 1 and 2). When diffusers were tested outdoors, all compounds repelled better when placed closer to the traps (Figure 3),

8 Journal of Vector Ecology June 2009 suggesting the repellent activity of botanical products can be improved by moving closer to the source compound. We also conclude that diffusers are more effective than candles when used indoors. Although DEET remains the gold standard for repelling mosquitoes, the human health and ecological consequences of the use of synthetic mosquito control compounds are becoming of greater concern to consumers. At least fifty years of sustained struggle to control mosquitoes has resulted in cases of toxicity to non-target organisms (Croft and Brown 1975), insecticide resistance (Brown 1986), and ecological hazards (Hayes and Laws 1991). Among alternative control strategies, the use of non-toxic plant essential oils such as citronella and essential oil derivatives such as linalool and geraniol are being more widely considered for both industrial and household uses. REFERENCES CITED Barnard, D.R. and R. Xue. 2004. Laboratory evaluation of mosquito repellents against Aedes albopictus, Culex nigripalpus, and Ochlerotatus triseriatus (Diptera: Culicidae). J. Med. Entomol. 41: 726-730. Barnard, D.R., U.R. Bernier, R. Xue, M. Debboun, and M. Govere 2007. Standard methods for testing mosquito repellents. In: M. Debboun, S.P. Frances, and D. Strickman (eds). Insect Repellents: Principles, Methods and Uses. pp. 103-110. CRC Press, Boca Raton, FL. Blaske, V.U. and H. Hertel. 2001. Repellent and toxic effects of plant extracts on subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 94: 1200 1208. Brown, A.W.A. 1986. Insecticide resistance in mosquitoes: pragmatic review. J. Am. Mosq. Contr. Assoc. 2: 123-140. Chadee D.D. and R Martinez. 2000. Landing periodicity of Aedes aegypti with implications for dengue transmission in Trinidad, West Indies. J. Vector Ecol. 25: 158-163. Choi, W.S., B.S. Park, S.K. Ku, and S.E. Lee. 2002. Repellent activities of essential oils and monoterpenes against Culex pipiens pallens Coquillett. J. Am. Mosq. Contr. Assoc. 18: 348-351. Croft, B.A. and A.W.A. Brown. 1975. Responses of arthropod natural enemies to insecticides. Annu. Rev. Entomol. 20: 285-335. Gabel, M.L., T.S. Spencer, and W.A. Akers. 1976. Evaporation rates and protection times of mosquito repellents. Mosq. News. 36: 141-146. Gilbert, I.H., H.K. Gouck, and N. Smith. 1966. Attractiveness of men and women to Aedes aegypti and relative protection time obtained with DEET. Fla. Entomol. 49: 53-66. Govere, M. and D.N. Durrheim. 2007. Techniques for evaluating repellents. In: M. Debboun, S.P. Frances, and D. Strickman (eds). Insect Repellents: Principles, Methods and Uses. pp. 147-159. CRC press, Boca Raton, FL. Hayes, W.J. and E.R. Laws Jr. (eds.) 1991. Handbook of Pesticide Toxicology. Volume l: General Principles. 497 pp. New York, NY, Academic Press. Hwang, Y., K. Wu1, J. Kumamoto, H. Axelrod, and M.S. Mulla. 1985. Isolation and identification of mosquito repellents in Artemisia vulgaris. J. Chem. Ecol. 11: 1297-1306. Khan, A.A., H.I. Maibach, and D.L. Skidmore 1972. A study of insect repellents: effect of temperature on protection time. J. Econ. Entomol. 66: 437-438. Lindsay, L.R., G.A. Surgeoner, J.D. Heal, and G.J. Gallivan 1996. Evaluation of the efficacy of 3% citronella candles and 5% citronella incense for protection against field populations of Aedes mosquitoes. J. Am. Mosq. Contr. Assoc. 12: 293-294. McGuire, T.R. 1984. Learning in three species of diptera: The blow fly Phormia regina, the fruit fly Drosophila melanogaster, and the house fly Musca domestica. Behavior Genetics 14: 479-526. Muirhead-Thomson, R.C. 1951. The distribution of anopheline mosquito bites among different age groups. Br. Med. J. 1: 1114-1117. Mumcuoglu, K.Y., R. Galun, U. Bach, J. Miller, and S. Magdassi. 1996. Repellency of essential oils and their components to the human body louse, Pediculus humanus humanus. Entomol. Exp. Appl. 78: 309-314. Osimitz, T.G. and R.H. Grothaus. 1995. The present safety assessment of DEET. J. Am. Mosq. Contr. Assoc. 11: 274-278. Osimitz, T.G. and J.V. Murphy. 1997. Neurological effects associated with use of the insect repellent N,N-diethylm-toluamide (DEET). J. Toxicol. Clin. Toxicol. 35: 435-441. Park, B.S., W.S. Choi, J. Kim, K. Kim, and S.E. Lee. 2005. Monoterpenes from thyme (Thymus vulgaris) as potential mosquito repellents. J. Am. Mosq. Contr. Assoc. 21: 80-83. Port, G.R. and P.F.L. Boreham.1980. The relationship of host size to feeding by mosquitoes of the Anopheles gambiae Giles complex (Diptera: Culicidae). Bull. Entomol. Res. 70: 133-144. Snyder, J.W., R.O. Poe, J.F. Stubbins, and L.K. Garrettson. 1986. Acute manic psychosis following the dermal application of N,N-diethyl-m-toluamide (DEET) in an adult. Clin. Toxicol. 24: 429-439. Sukumar, K., M.J. Perich, and L.R. Boobar.1991. Botanical derivatives in mosquito control: a review. J. Am. Mosq. Contr. Assoc. 7: 210-237. Tawatsin, A., S.D. Wratten, R.R. Scott, U. Thavara, and Y. Techadamrongsin. 2001. Repellency of volatile oils from plants against three mosquito vectors. J. Vector Ecol. 26: 76-82. Zadikoff, C.M. 1979. Toxic encephalopathy associated with use of insect repellent. J. Pediatr. 95: 140-142.