Irritancy and repellency of Aedes aegypti (Diptera: Culicidae) to insecticides and implications for vector control operations

Deon Canyon (PhD)

Tropical Infectious and Parasitic Diseases Unit, School of Public Health and Tropical Medicine, James Cook University, Townsville Qld 4811, Australia. Deon.Canyon@jcu.edu.au

The possible repellent effect of insecticides on aedines was suggested as an important potential factor in Aedes aegypti (Linnaeus) eradication programs due to Ae. aegypti’s anthropophilic and endophilic nature which makes it an ideal target for domestic insecticide application (McClelland, 1967; Surtees, 1967). Surtees (1967) suspected that insecticide spraying would have the effect of depressing oviposition. This was supported by Jacob (1969) who found that tire breeding sites containing ovicides repelled ovipositing females, by Von Windeguth et al. (1971) who found that larvicides may have induced excito-repellent responses in ovipositing mosquitoes in Florida, by Verma (1986) who found that several synthetic pyrethroids repelled ovipositing Ae. aegypti, and by Moore (1977), who directly linked the repellent effect to insecticide concentration. Contrary results were recorded for insect growth regulators where methoprene briquettes were observed to attract oviposition (Carroll, 1979).

This study investigated malathion, temephos, permethrin, methoprene and Bacillus thuringiensis israelensis (Bti) for their irritant and repellent properties. The importance of this work stems from the probability that the use of a larvicide with irritant or repellent properties in control operations could result in greater dispersal of the mosquito population through relocation of gravid or ovipositing females to other areas with untreated sites.

Materials and methods

Insecticide avoidance behavior was not measured using standard excito-repellency methods because the aim was to observe oviposition responses of Ae. aegypti gravids to the presence of insecticides. Oviposition responses of c. 25 week-old female Ae. aegypti (4 replicates) were investigated for each insecticide. Two consecutive tests were conducted for each insecticide, corresponding to the first and second oviposition cycles, by introducing the author’s legs into the bed nets to provide a blood source on the first and fifth days of the experiment. Eggs were collected and counted on the fifth and ninth days. The 4 oviposition test replicates of each insecticide were carried out in untreated bednets (2.43m³) in an environmentally controlled experimental room (25° ± 1°C, 60 ± 5% RH, 16:8 L:D) (Fig. 1).

Fig. 1 – Experimental enclosures.

Oviposition sites were treated with insecticides at the following LC₉₉ doses: malathion (1.042 mg/L), temephos (0.193 mg/L), permethrin (0.0093 mg/L), methoprene (20 mg/L) and Bti, 6AS VectoBac^® (10mg/L). These doses were estimated from a Townsville colony established in 1995 (Canyon and Hii, 1998) and maintained according to the procedures of Foster (1980). Malathion, permethrin, temephos and methoprene were appropriately diluted with 90% ethanol and Bti with water to obtain the desired concentrations. One insecticide at a time was tested to avoid competitive insecticide odors. A control beaker containing unfiltered tap water (300 ml) and a beaker with insecticide treated tap water (300 ml) was placed into each bednet. Pieces of filter paper were placed on the sides of each beaker as an oviposition substrate. Eggs laid on the filter paper and on the solution surface were counted since it was possible that irritancy might affect oviposition site preference. Beaker solutions were not prepared freshly for the second oviposition cycle because the aim was to simulate non-renewed bodies of water in the field and investigate the persistent or possibly temporal effect of these insecticides.

The paired t-Test was used to compare the number of eggs laid in treated and control beakers in each oviposition cycle. The same test was used on ratios, calculated as the number of eggs laid in treated beakers divided by the number of eggs laid in control beakers, to analyze differences between the two oviposition cycles.

Results

Irritancy

Oviposition patterns in response to malathion were observed to be different between treated and control beakers. Eggs were laid in clumps in a narrow band above the water line in the control beaker, as is typical for Ae. aegypti, but were laid individually in a very broad band from the water line to the top of the filter paper in the treated beaker. The latter pattern was not as distinct or was absent in the other insecticide treated beakers.

Comparison of the number of eggs laid on the filter paper (side) and the solution (surface) suggested that each insecticide elicited a very different response (Table 1).

The side oviposition site was preferred over the surface in all tests, however, differences were observed in the degree of surface laying. Relatively fewer eggs were laid on control surfaces than malathion, temephos, permethrin and methoprene treated surfaces in both oviposition cycles. The opposite was observed for Bti where more eggs were laid on control surfaces than treated surfaces. No difference was observed between control and temephos-treated beakers.

Avoidance

Significantly more eggs were laid in beakers treated with malathion, temephos, methoprene and Bti than controls in oviposition cycle 1, and significantly more eggs were laid in temephos, methoprene and Bti control beakers than in treated beakers in oviposition cycle 2. No significant difference between treated and control beakers was observed in the permethrin tests (Table 2).

Analysis of the ratio of eggs laid in each ovicycle revealed that significantly more eggs were laid in malathion, temephos, methoprene and Bti treated beakers than control beakers in cycle 1. This was reversed in cycle 2 when control beakers for the same insecticides received significantly more eggs than treated beakers. No significant change in oviposition preference between the two cycles was observed in the permethrin tests. Figure 2 depicts the log-transformed ratios of the number of eggs in beakers treated with various insecticides and the number of eggs in control beakers - a positive value indicating consequently a preference of the treated beaker.

Table 1.  Ratio of the sum of eggs laid on the filter paper/the sum of eggs laid on the solution surface in two consecutive oviposition cycles in control and insecticide treated beakers. A higher number indicates more laid on the filter paper and less on the surface.

Table 2 – Significant differences (P values) between the mean number of eggs oviposited in insecticide-treated and control beakers in two consecutive oviposition cycles.

Fig. 2.  Log-transformed ratio of the number of eggs laid in treated beakers / the number of eggs laid in control beakers in two consecutive oviposition cycles (Cycle 1: à, Cycle 2: ¨). Positive values indicate a preference for treated beakers. P values resulting from the comparison of cycles 1 and 2 for each insecticide are displayed above the figure.

Discussion

It is now clear that some insecticides have a significant impact on the oviposition preferences of gravid Ae. aegypti. Responses can be classed as those relating to irritation, which may result in avoidance and those relating to repellence, which definitely result in avoidance.

Irritancy

Observational comparisons of the patterns created by eggs laid by ovipositing gravids on filter paper suggested that malathion was the only insecticide which was capable of causing a tangible degree of physical irritation. The preference for treated beakers in the first oviposition cycle and the lack of any significant difference between treated and control beakers in the second cycle indicated that this level of irritation was insufficient to cause avoidance. 

The comparison of the number of eggs laid on side (Fig. 3) and surface areas was somewhat inconsistent between insecticides (Table 1).

Fig. 3 – Samples of egg-laden paper showing a preference for untreated water on the right and a more diffuse egg-laying pattern on the left in the presence of malathion.

These results are difficult to interpret since the same concentration of solution was present in both areas, however the data suggest that a behavioral change had occurred.

Avoidance

All insecticide-treated beakers, with the exception of permethrin, initially attracted ovipositing gravids at the end of the first gonotrophic cycle. This suggested two possibilities: the presence of a volatile semiochemical with attractant properties or insecticide recognition after prior exposure. In the first oviposition cycle, the beakers with insecticides containing ethanol, a well-known attractant (Skinner et al., 1965), and Bti (containing unknown volatiles), elicited oviposition responses from the mosquitoes even after 4 days of evaporation. After 7-8 d of evaporation at the end of the second gonotrophic cycle, oviposition site preference was reversed and gravids avoided ovipositing in treated beakers in favor of control beakers. It was inferred that either the attractant components in all insecticides had simultaneously ceased to exert an influence due to evaporation and an innate behavioral response to the insecticide itself was observed or that the cohort of mosquitoes previously exposed to an insecticide recognized its presence and ‘learned’ to avoid it in subsequent cycles. If the latter was the case, the concept of repellency as it applies to insecticides should be reviewed because repellency is irrelevant when avoidance is due to recognition of insecticide treated sites.

The question of water quality and its effect on oviposition behavior in the second cycle is only of potential concern where Bti is concerned since it can be assumed that the water quality dropped due to the metabolism of microbes decomposing the Bti formulation. This is unlikely to have been of any consequence since the same pattern of oviposition behavior was observed for malathion, temephos and methoprene which contain no organic material for microbes to metabolize. If water quality was a significant factor a similar response should have been observed for all insecticides. This was not the case, however, so responses were more probably due to insecticide presence or absence. Since oviposition responses were observed to vary temporally over the 9 days of testing, the value of studies by Carroll (1979) and Mather and Defoliart (1983), who assessed oviposition every 10 and 7 days respectively, are questionable. The latter study may have recorded a lack of repellency due to a combination of avoidance and attraction data. The results from this study do not directly contradict the results obtained by Moore (1977) and Verma (1986), which indicate that malathion, temephos and permethrin treated oviposition sites repel Ae. aegypti, since these studies fail to mention the delay between insecticide preparation and experimental commencement. Also of concern in the latter study was the close proximity of control and treated containers since vapors moving in either direction may have influenced results.

The wide-ranging and consistent results from this study suggest that insecticide based Ae. aegypti control operations not using synthetic pyrethroids may face problems. Application of some insecticides may encourage oviposition in treated containers for the first few days post-treatment and post-insecticide preparation. The opposite corollary of this is that if the period of insecticide effectiveness expires within the period of attractancy due to environmental factors such as rainfall, which stimulates egg hatch rates, a greater number of larvae would be recruited than in the period prior to spraying. Alternately, the avoidance of treated sites observed one-week post treatment could result in treated and often-frequented oviposition sites being avoided in favor of untreated sites. Whether areas would benefit from more regular treatments to counter this repelling effect is unknown. The possible extent of mosquito dispersal due to treatments applied more than 8 days apart also requires attention. The challenge to industry is to develop a non-repellent insecticide with persistent attractant properties to address innate behavioral responses. The inclusion of a substance causing delayed mortality may be useful in addressing insecticide recognition. Alternatively, a switch to synthetic pyrethroids may be the solution since no significant avoidance behavior was observed in response to oviposition sites with permethrin treatment.

References

Carroll MK. (1979) Methoprene briquets as an attractant for gravid Aedes aegypti (L.). Mosquito News 39:680-681.

Foster W A. (1980) Colonization and maintenance of mosquitoes in the laboratory, pp. 103-151. In: Malaria Vol. 2 (ed. by J.P. Kreier). Pathology, vector studies and culture, Academic Press, London.

Jacob WL. (1969) Simulated field tests with ovicides against Aedes aegypti eggs in tires and cans. Mosquito News 29:402-407.

McClelland GAH. (1967) Problems of interaction and density in mosquitoes in relation to Aedes aegypti eradication. Proceedings of the Californian Mosquito Control  Association 35:94-95.

Moore CG. (1977) Insecticide avoidance by ovipositing Aedes aegypti. Mosquito News 37:291-293.

Skinner WA, Tong H, Pearson T, Strauss W, Maibach H. (1965) Human sweat components attractive to mosquitoes. Nature 207:661-662.

Surtees G. (1967) The distribution, density and seasonal prevalence of Aedes aegypti in West Africa. Bulletin of the World Health Organization 36:539-540.

Verma KVS. (1986) Deterrent effect of synthetic pyrethroids on the oviposition of mosquitoes. Current Science 55:373-375.

Von Windeguth DL, Eliason DA, Schoof HF. (1971) The efficacy of carbaryl, propoxur, Abate and methoxychlor as larvicides against field infestations of Aedes aegypti. Mosquito News 31:91-95.