Lethal and Sublethal Concentrations of Formulated Larvicides Against Susceptible Aedes aegypti

Integrated vector management programs have been shown to be effective at controlling populations of vectors that can transmit pathogens to humans and animals. Morrison et al. (2008) describes several examples where vector control has been at the fore of protecting human populations from contracting arthropod-borne diseases. US Army Surgeon General William Gorgas stopped yellow fever transmission among workers by eliminating Aedes aegypti (L.) larval habitat sites in Havana, Cuba, and during the construction of the Panama Canal (Patterson 1989). The Rockefeller Foundation led the temporarily successful Ae. aegypti eradication program during the 1950s–1960s that eliminated yellow fever and dengue transmission from a majority of Central and South American countries (Gubler 1997). More recently, Singapore and Cuba reduced the risk of dengue transmission by implementing anti–Ae. aegypti legislation and control actions (Boo 2001, WHO 2009). Additionally, the use of biological control agents, such as copepods of the genus Mesocyclops (i.e., predatory crustaceans), is another tool used to induce mortality among immature stages and reduce recruitment of adult mosquitoes. The use of Mesocyclops copepods in a community-based vector control program from 1998 to 2003 resulted in drastic reductions or localized eradication (6 of 9 communes) of Ae. aegypti and dengue in select provinces of Vietnam (Kay and Vu 2005).

Currently, Ae. aegypti control programs primarily use source reduction and chemical controls such as ultra-low volume (ULV) adulticide space sprays and larviciding of artificial containers to control target vectors through the use of natural and synthetic chemicals. Despite best control practices and modern technological advancements, use of insecticides to reduce mosquito populations has challenges. Ineffective treatment, insecticide resistance, and error can result in exposing mosquitoes to sublethal dosages of insecticides that may impact individual phenotypic traits and population ecology by affecting: 1) recruitment to adulthood and survival of adults, 2) reproductive ability of individuals, and 3) frequency changes of resistance alleles of future generations through selection (Moriarty 1969, Yu et al. 2010, Evans et al. 2020).

Vectors can be exposed to sublethal concentrations of insecticides in a variety of ways, from the improper use of an insecticide (i.e., not following the label) to failure to re-treat an environment to maintain a sufficiently high lethal dose. The current study determined the lethal concentration (LC) equivalent of 25% and 50% mortality (LC₂₅ and LC₅₀) of larval Ae. aegypti following exposure to formulated larvicidal products containing Bacillus thuringiensis israelensis de Barjac (Bti), spinosad, temephos, pyriproxyfen, and methoprene. Additionally, we report on observed changes in sex ratio and net growth, as measured by adult wing length, for mosquitoes that survive exposure to these insecticides relative to control conditions.

The mosquitoes used in this study were pyrethroid susceptible Ae. aegypti (Orlando strain [ORD]) obtained from colonies maintained at the USDA-ARS Center for Medical, Agricultural, and Veterinary Entomology (CMAVE), in Gainesville, FL, since the colony was transferred from the USDA Insects Affecting Man and Animals Research Laboratory established in 1952 (Personal communication D. Kline), originally collected in Orlando, FL. This strain has been reported to have some refractoriness to dengue virus infection but is pesticide susceptible (Estep et al. 2017, Souza-Neto et al. 2019). The selection of this strain was made to best observe patterns of increased or decreased dengue virus infection, dissemination, and infected saliva (i.e., transmission) in future studies. The mosquitoes were reared following modified protocols outlined in Gerberg et al. (1994). Mosquito larvae were kept in environmental control chambers (Conviron, C1020, Winnipeg, Manitoba) with a photoperiod of 12:12 h light:dark (LD) and 80% relative humidity (RH) at 28 ± 1°C. Adult colony mosquitoes were kept in screened colony cages measuring 61 × 79 × 61 cm with access to 10% sucrose solution ad libitum.

The larval bioassay procedure used to assess the lethal concentration (LC) values is derived from the larvicide bioassay procedure used by the World Health Organization (WHO 2005). A sample of Ae. aegypti eggs were volumetrically measured and combined in a 22.375 × 17.5 × 3 in. rectangular pan of water, where they were fed intermittently on a nutrient slurry [larval food] as described by Gerberg et al. (1994) [80 g of a 3:2 ratio of bovine liver powder and brewer's yeast mixed into a slurry with 1800 ml water]. On the 4th day following hatching, 3rd–4th instars were strained using a No. 30 stainless steel sieve (Advantech; Taipei, Taiwan). Fifty larvae were added to 473-ml (16-oz) treatment cups (Clark Inc.; Lancaster, PA). Each cup was then filled with well water, 6 ml of “larval food” over 2 consecutive days (3 ml each day), and a known amount of diluted formulated insecticide. Each experimental replicate (cup) contained 395 ml water, 50 larvae, and the same amount of food to ensure that larvae had as close to similar rearing conditions as possible.

Stock insecticide solutions were made from formulated larvicides in a similar manner to the methods described by Zhao et al. (2017). One of 4 formulated larvicides were diluted into a 0.01% stock solution then applied to each replicate cup of Vectobac®12AS (Valent Biosciences Corp., Libertyville, IL; active ingredient (AI), Bti]; Natular™ 2EC (Clarke Mosquito Control Products, St. Charles, IL, spinosad—a mixture of spinosyn A and D produced by Saccharopolyspora spinosa Mertz and Yao); Abate® (Clarke Mosquito Control Products, St. Charles, IL; AI, temephos); and Altosid® Liquid Larvicide Mosquito Growth Regulator (Central Life Sciences, Dallas, TX; AI, methoprene). Nyguard insect growth regulator (IGR) (McLaughlin Gormley King Company, Minneapolis, MN; AI, pyriproxyfen) was diluted into a 0.001% stock solution. Three to 5 replicates were completed for each concentration and formulated larvicide treatment. Predicted larvicide LC₅₀ rates were used from the technical LC₅₀ derived from a literature review and confirmed against the target strain with a pilot experiment; spinosad (Zhao et al. 2017), Bacillus thuringiensis israelensis (Zhao et al. 2017), pyriproxyfen (Seccacini et al. 2008), temephos (Seccacini et al. 2008), methoprene (Ali et al. 1995).

The 0.01% stock solution was made by combining 10 μl of a formulated larvicide to 100 ml of well water. The 0.001% stock solution was made by combining 1 μl of a formulated larvicide to 100 ml of aged well water. The stock solutions were then stirred for at least 5 min at 400 rpm using a magnetic stirring plate. Gradually increasing aliquots of a stock solution were then added to the prepared treatment cups with larvae.

The larvae were fed and placed in an incubator (Percival; Perry, IA) set to 28°C, 60% RH, with a 12:12 h LD cycle and reared to adulthood. After mosquitoes developed to the pupal stage, cups were covered by the funnel emergence portion (top) of a mosquito breeder (1425; Bioquip, Rancho Dominguez, CA), and the pupae were left to emerge as adults, which were removed and counted. The number of emerged adults was the number that survived, and the value used to assess the LC₂₅ and LC₅₀ values.

A control cup with no formulated insecticide was added to every set of formulated larvicide in order to assess baseline mortality from handling and experimental procedures. If the control cup suffered >20% mortality, the replicate was discarded.

Larvicides assessed were selected from currently available and registered formulated larvicides. The larvicides selected consisted of different active ingredients representing several modes of action in order to assess sublethal effects that may only manifest from that particular mode of action. The formulated larvicides assessed were: Vectobac12AS (11.61% Bti); Natular 2EC (20.6% spinosad); Abate (2% temephos); Altosid Liquid Larvicide Mosquito Growth Regulator (5% methoprene); and Nyguard IGR (10% pyriproxyfen). The active ingredient, Bti, is classified under group 11A, microbial disruptors of insect midgut membranes by the Insecticide Resistance Action Committee (IRAC). They are protein toxins that bind to receptors on the midgut membrane and induce pore formation (IRAC 2022). The active ingredient, spinosad, is classified under group 5, nicotinic acetylcholine receptor (NACHR) allosteric modulators, site I by IRAC. The NACHRs are allosterically activated causing hyper excitation of the nervous system (IRAC 2022). The AI, temephos, is classified under group 1-B, acetylcholinesterase (ACHE) inhibitors by IRAC. They inhibit ACHE and cause hyperexcitation of the nerve (IRAC 2022). The AI methoprene is classified under group 7-A, juvenile hormone mimics (juvenile hormone analogues). The AI pyriproxyfen is classified under group 7-C juvenile hormone mimics (pyriproxyfen). Juvenile hormone mimics prevent and/or disrupt metamorphosis when applied at a juvenile stage (IRAC 2022).

Once LC₂₅ and LC₅₀ values were calculated for a larvicide, additional bioassays were conducted to observe and assess sublethal effects. Similar to the bioassay procedure described above, larvicide bioassays involved the exposure of 50 3rd–4th instar Ae. aegypti larvae in a replicate cup and 395 ml of water with a dilute larvicide value closest to the calculated LC₅₀. Larvae were fed on 2 consecutive days a diet of 3 ml of larval food. On the 3rd day, a funnel was fitted to the treatment cups to capture emerged adults as described earlier. After all pupae had either emerged or died, the adults were collected, sexed, counted, and had their wings removed and measured. Bioassays were placed in an incubator (Percival; Perry, IA) set to 28°C, 60% humidity, with a 12:12 h LD cycle during rearing to adulthood. Bioassays with >10% mortality in the controls were discarded and repeated.

Wing dissection and preparation was carried out by dissecting a wing from the adult body while viewing through a stereomicroscope (Meiji Techno; Saitama, Japan) and attaching it to a layer of double-sided sticky tape (Scotch, 3M; Saint Paul, MN) with No. 5 fine jeweler's forceps. Wing measurements were recorded using a digital microscope (Keyence; Osaka, Japan) internally calibrated, and confirmed against a 1 mm slide micrometer with divisions at 0.01 mm (AmScope; Irvine, CA). Wing length was measured from the axillary incision to the apex, not including the fringe scales.

To assess the influence of an LC₅₀ larvicide treatment on mosquito wing length, only wing lengths from mosquitoes that emerged from cups that generated between an LC₄₀ and LC₆₀ mortality were used for comparison between different sexes and between treatments within those sexes, and from replicates whose control mortalities did not exceed 10%. First, wing length measurements across the male and female sample sets were tested for normality through visual assessment of plotted distributions, then a Shapiro–Wilk test for normality found the wing length measurement not normally distributed by sex (male, P = 1.15 × 10⁻⁹; female, P < 2.2 × 10⁻¹⁶).

Because the wing length measurement deviated from the assumption of normality, we conducted a Kruskal–Wallis test to assess whether median wing lengths differed from one another between replicates within a treatment. If they were different, a Conover–Iman test of multiple comparisons using rank sums to identify statistically significant differences in median wing length between treatment replicates was run to identify the replicate that was significantly different, all measurements from that replicate set were discarded, and a new replicate was conducted to ensure that at least 3 replicates were assessed from the same treatment. Multiple comparisons using rank sums were used to identify statistically significant differences in median wing length between treatments separated by sex.

Finally, a test of equal or given proportions was conducted to compare the proportion of males to the proportion of females that emerged pooled across all replicates of a specific treatment (LC₅₀ larvicide treatments and control).

The concentration curves for each formulated insecticide are illustrated in Fig. 1A–F. The LC₂₅ and LC₅₀ for each formulated larvicide are described in Tables 1 and 2, respectively. Control mortality did not exceed 10% in all formulated larvicide sets assessed, and therefore no correction was needed. The calculated LC₅₀ in μg/liter of Natular 2EC (26.15) fell between the 23.97 and 29.96 μg/liter bioassay concentrations, which yielded an average proportion of survivorship between 0.48 and 0.44. The calculated LC₅₀ in μg/liter of Abate 4E (3.12) fell between the 3.00 and 3.59 μg/liter bioassay concentrations, which yielded an average proportion of survivorship between 0.53 and 0.32. The calculated LC₅₀ in μg/liter of Altosid (0.43) fell between the 0.39 and 0.52 μg/liter bioassay concentrations, which yielded an average proportion of survivorship between 0.59 and 0.35. The calculated LC₅₀ in μg/liter of Nyguard (0.02) fell between the 0.02 and 0.03 μg/liter bioassay concentrations, which yielded an average proportion of survivorship between 0.56 and 0.03. The calculated LC₅₀ in ITU/liter of VectoBac12AS (496.7) fell between 383.7 and 511.6 ITU/liter bioassay concentrations, which yielded an average proportion of survivorship between 0.61 and 0.40.

Following the identification of the LC₅₀ concentrations for each treatment, an LC₅₀ concentration was applied to larval bioassays to assess the sublethal effects on the adults, specifically wing length and sex proportion. The applied LC₅₀ concentrations for each treatment assessed for sublethal effects were: Abate (3.00 μg/liter; 25 μl of stock solution in 400 ml), Altosid (0.45 μg/liter; 35 μL of stock solution in 400 ml), Natular (23.97 μg/liter; 400 μl of stock solution in 400 ml), Nyguard (0.02 μg/liter; 10 μl of stock solution in 400 ml), and VectoBac (511.6 ITU/liter; 1600 μl of stock solution in 400 ml).

Median male wing length (2314 um) and median female wing length (3028 um) differed significantly (χ² = 561.73, df = 1, P < 0.0001). Therefore, the wing lengths were separated by sex to allow for treatments to be evaluated. Median wing length differed significantly between treatments for both males (χ² = 139.68, df = 5, P < 0.0001) and females (χ² = 195.28, df = 5, P < 0.0001).

Differences in wing length between treatments separated by sex are shown in Fig. 2 (males) and Fig. 3 (females). Post hoc analysis of median wing lengths between treatments in males and females is described in Table 3. There was no statistically significant difference between control larvae and larvae exposed to Altosid LC₅₀; however, all other LC₅₀ treatments were statistically significantly different to the control. Percentage differences in wing lengths between treatments separated by sex are described in Table 4.

The ratio of male:female adults that emerged differed by treatment and is described in detail in Table 5. There was no significant difference in the emerged adult male:female ratio from control larvae (P = 0.92) and larvae exposed to Altosid (P = 0.31), Nyguard (P = 0.052), and VectoBac P = 0.60). There was a significant difference in the emerged adult male:female ratio from larvae exposed to Abate (P = 0.001) and Natular (P = 0.0005).

We were able to determine the LC₂₅ and LC₅₀ values for 5 formulated larvicides (Abate 4E, Altosid SSR, Natular 2EC, Nyguard IGR, and VectoBac 12AS) in susceptible Ae. aegypti after exposure as 3rd–4th instars. We observed sublethal effects on the wing length and sex proportion that may go unnoticed in mass larval bioassays within the same container (Alto and Lord 2016). Furthermore, the observed sublethal effects are discussed with respect to their potential impact on vectorial capacity, and the ecological importance is highlighted.

The calculated LC₅₀ values are similar to or less than those values found in the literature referenced below. The calculated LC₅₀ of Natular (spinosad; 26.15 μg/liter) was less than ½ that of the calculated LC₅₀ (spinosad; 0.06 mg/liter) of a wild population of Ae. aegypti collected from Chiapas, southern Mexico (Antonio et al. 2009). The calculated LC₅₀ of Abate (temephos; 3.12 μg/liter [0.003 ppm]) was 20× less than that of the calculated LC₅₀ (temephos; 0.06 ppm) of a colony of Rockefeller strain Ae. aegypti maintained by the National Centre for Disease Control (NCDC) in Mettupalayam, Coonoor, India (Muthusamy and Shivakumar 2015). The calculated LC₅₀ of Altosid (methoprene; 0.43 μg/liter) was ca. 6.5× less than the calculated LC₅₀ (Metoprag 20CE; methoprene; 2.83 mg/liter) of a colonized population of Rockefeller strain Ae. aegypti maintained in Brazil (Braga et al. 2005). The calculated LC₅₀ of Nyguard (pyroproxifen; 0.02 μg/liter) was 10× more susceptible than the calculated LC₅₀ 1.1 × 10⁴ (1.0 × 10⁴–1.1 × 10⁴) mg/liter of a colonized population of Ae. aegypti Bora strain obtained in French Polynesia (Darriet and Corbel 2006). The calculated LC₅₀ of VectoBac12AS (Bti; 496.7 ITU/liter) was ca. 1.3× more than the calculated LC₅₀ (1.86 ppm, giving a potency of 382.95 ITU/mg) of a colonized population of Ae. aegypti BKK1 strain obtained in Thailand (Fansiri et al. 2006). These observed sublethal concentrations and documented concentrations from the literature that span a variety of susceptible strains identify the ORD strain as a strain sensitive to pesticides.

Sublethal effects are still a poorly understood topic regarding mosquito control given the range of mosquito species, variability of application type, and diversity in modes of action. But progress in research is being made that eschews a greater appreciation of the multifaceted impact that larvicide exposure has on the surviving adult populations. Our findings demonstrate that larvicides with various modes of action can have a similar impact on emerging adults, for instance the wing lengths or the sex ratio of emerged adults. Our finding that wing lengths of adult males and females were smaller and significantly different from larvae exposed to Abate, Natular, Nyguard, and VectoBac than to the control treatment was unexpected. Contrary to the hypothesis by Agudelo-Silva and Spielman (1984), Wilson et al. (1990), and Alto et al. (2008) that larvicide applications may release larval survivors from competition and lead to larger size adults, susceptible Ae. aegypti larvae reared within LC₅₀ concentrations of Abate (temephos), Natular (spinosad), Nyguard (pyroproxifen), and VectoBac (Bti) at 28°C were significantly smaller than untreated controls reared at the same temperature and density. No difference was observed between Altosid (methoprene) exposed and control Ae. aegypti. These results support findings by Moura et al. (2020), who observed significantly smaller adult Ae. Aegypti exposed to pyriproxyfen as larvae than controls; and by Judd (2018) who observed no difference in wing lengths between Culex quinquefasciatus Say exposed to methoprene. However, these results are not supported by Gowelo et al. (2020), who observed significantly larger Anopheles coluzzi (Coetzee and Wilkerson sp. N.) when exposed to Bti, and by Kelada et al. (1981), who observed significantly larger Cx. pipiens L. when exposed to methoprene. Robert and Olson (1989) documented mixed effects from exposing Cx. quinquefasciatus to methoprene, finding males developed significantly longer wings, but females were significantly smaller. Judd (2018) documented no difference in wing length of Cx. quinquefasciatus exposed to spinosad. To our knowledge, this is the first observation of adult mosquito size associated with sublethal concentrations of abate (temephos).

It is unclear why adults of larvae exposed to some larvicides were small. It contradicts findings by Gowelo et al. (2020) and Antonio et al. (2009), who observed larger adult mosquitoes of larvae exposed to spinosad and Bti. Based on their explanations; 1) larger mosquito larvae may be better capable to cope with stress induced by larvicide exposure and 2) larvicides reduce larval density, which is associated with increased adult body size (Gimnig et al. 2002, Scott and Takken 2012). One possible explanation for why adults of larvae exposed to larvicides were smaller is that stress may induce less movement, less feeding, slower development, and therefore smaller adult body size as observed in mosquitoes exposed to predators, a different type of stress, but stress nonetheless (Lounibos et al. 1993, Roux et al. 2015).

The proportion of males and females that emerged is another sublethal effect observed in this study. Sex ratio was influenced in Abate and Natular, as described in Table 4. These observations led to the finding that some larvicides significantly alter the sex ratio of males and females that emerged compared to the control—generating greater numbers of females than males. These results were supported by findings from Robert and Olson (1989) and Aguilera et al. (1995), who documented female biased sex ratios from Cx. quinquefasciatus exposed to propoxur, chlorpyriofos, and methyl-pirimiphos. Male sex ratio bias has been documented to occur in Ae. aegypti following exposure to Bti by Wang and Jaal (2005) and Flores et al. (2004) but was not observed in this study. Furthermore, female sex ratio bias has been documented in Ae. aegypti following exposure to pyriproxyfen, but this experiment did not observe a significant difference between pyriproxyfen and control sex ratios (Loh and Yap 1989). Finally, mixed sex ratios have been observed from mosquitoes following exposure to methoprene. Specifically, more male An. dirus Peyton and Harrison emerge following methoprene exposure (Sithiprasasna et al. 1996), but more female Cx. quinquefasciatus emerge following methoprene exposure (Robert and Olson 1989). The current study observed no difference in sex ratios between Altosid and the control. To our knowledge, this is the first documentation of female sex bias from Ae. aegypti exposed as larvae to LC₅₀ concentrations of Natular (spinosad).

Female biased sex ratio is a sublethal effect that could complicate vector control efforts, given female mosquitoes are those targeted for control. The mechanism by which the sex ratio shifts is not understood, but we speculate it is due to a combination of survival factors including size and sex-linked enzyme expression conserved across diverse animals (rats, Kato and Yamazoe 1992; copepods, Kadiene et al. 2020; tea mosquito bugs, Helopeltis theivora,Roy and Prasad 2018).

Based on these results, vectorial capacity of susceptible Ae. aegypti when exposed to a larvicide as larvae might be altered from changes in both the proportion (i.e., number of females that emerge), as well as the vector size (based on wing length). Observations by a number of researchers have correlated wing length to fecundity and adult body mass in mosquitoes, specifically Ae. albopictus (Skuse) and Ae. aegypti, as outlined by Armbruster and Hutchinson (2002). Smaller female mosquitoes as observed by Alto et al. (2008) have an increased rate of biting, but are considered less effective vectors as observed by Sumanochitrapon et al. (1998). However, enhanced pathogen infection and dissemination may be influenced by exterior stressors experienced in the juvenile state, such as those observed and investigated by Alto et al. (2005).

Relative to body size variation and its role in vector transmission, Juliano et al. (2014) observed 2 competition-related hypotheses that influence vectorial capacity: competition-susceptibility and competition-longevity. Where the competition-susceptibility hypothesis postulates that small females are more susceptible to infection and predicts that frequency of infection should decrease with size, the competition-longevity hypothesis postulates that small females have lower longevity and lower probability of becoming competent to transmit the pathogen and thus predicts that frequency of infection should increase with size (Juliano et al. 2014). The results of Juliano et al. (2014) indicate that longevity is not linearly related to adult size. Instead, it is quadratically related, the small and large size adult females living short lives and the intermediates living longest. Additionally, it was observed that infection frequency increased as size increased, peaking between 3.0 and 3.5 mm. These results confound those results by Alto et al. (2005) but support those of Sumanochitrapon et al. (1998).

Ultimately, these results confirm that sublethal concentrations have a significant effect on susceptible adults that survive juvenile larvicide exposure, generating potentially more and smaller adult female vectors. This combination of a higher proportion of females and smaller size is serious, given that an increase in the number of females increases the density of vectors, and their smaller size can potentially make them more efficient vectors if the observations by Alto et al. (2005) are supported. Based on these findings, it is essential for vector control programs to properly apply larvicides and monitor resistance to those larvicides they use in order to ensure that populations are controlled, preventing the production of small female vectors.

For maintenance of the colony of mosquitoes, we thank H. Brown, J. Mollet, and T. Carney at the United States Department of Agriculture—Agricultural Research Service—Center for Medical, Agricultural, and Veterinary Entomology (USDA-ARS-CMAVE) insectaries. This research was supported by the USDA-ARS and the US Department of Defense (DoD) Deployed War-Fighter Protection Program (DWFP). Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by USDA, DoD, the Florida Army National Guard, the University of Florida, or the DWFP. The USDA is an equal opportunity provider and employer. The findings and conclusions in this report are those of the author(s) and do not necessarily represent the views of the Centers for Disease Control and Prevention, the USDA, or the Uniformed Services University.