Introduction

Aedes aegypti L. and Aedes albopictus (Skuse) mosquitoes that have bitten viral-infected humans transmit the disease to other humans. These mosquitoes are major vectors of arboviruses such as Zika, yellow fever, dengue, and chikungunya1,2,3. Dengue is one of the most dangerous arboviruses, causing high morbidity and mortality rates in several countries around the world including many parts of Thailand2,4. According to a report by the Thai Ministry of Public Health, there were 9084 and 19,380 total dengue cases in Thailand, in 2021 and 2022, respectively, with 9 and 17 fatalities, calculated into a morbidity rate of 0.06% and 0.09%, respectively5. Since vaccines and other drugs have limited effectiveness in controlling dengue, controlling the population of mosquitoes with pesticides is the most effective measure to avoid this contagious disease6.

Controlling the mosquitoes at their embryonic and larval stages is the key strategy in controlling mosquito populations6,7,8. Generally, temephos, a common synthetic organophosphate insecticide, is used extensively around the world, especially in Thailand, for killing mosquito larvae. Unfortunately, its efficacy has been reduced drastically because populations of mosquitoes have developed resistance to it1. Moreover, temephos incurs serious negative side effects. It causes irreversible damage to non-target aquatic predators and humans, thus limiting its use9,10. In this glum context, many researchers have urgently developed alternative strategies that are safer for humans and the environment7.

Because plant essential oils (EOs) are natural substances, they and their phytochemical constituents are good, green alternatives to temephos. They are not harmful or only slightly harmful to mammals and non-target aquatic predators of mosquitoes at a practical pesticidal level, plus they degrade rapidly in the environment3,11. The EOs of Cinnamomum verum (J. Presl.), Citrus aurantium L., and Cymbopogon citratus (Stapf.) were investigated in this study. They have been reported safe for humans and mammals as well as having low toxicity on non-target predators because they have been long used as feed in food industry as well as antibiotic and antioxidant in folk medicine for thousands of years12,13,14.

Regarding plant EOs insecticidal efficacy, they are strongly insecticidal against many insect pests, such as Ae. aegypti (Order Diptera)15, Musca domestica (Order Diptera)16, Haemaphysalis longicornis (Order Ixodidae)17, Pediculus humanus capitis (Order Phthiraptera)18, Spodoptera littoralis (Order Lepedroptera)19, and Periplaneta americana (Order Blattodea)20. Specifically, EOs from C. verum, Coccinia indica, C. citratus, Illicium verum (Hook.f.), and Moringa oleifera (Lam.) and their major constituents (geranial and d-limonene) exhibited strong ovicidal activity against many mosquitoes species like Anopheles indica7, Ae. aegypti21,22,23, and Culex quinquefasciatus24 and housefly (M. domestica)16 with an LC50 ranging from 3.31 to 303,200 ppm. Furthermore, combined formulations of EOs and EO constituents showed even higher efficacy than their individual EOs25,26. For example, Soonwera et al.27 reported that a combined trans-anethol + I. verum EO formulation provided complete mortality against Ae. aegypti and Ae. albopictus larvae. Another group of researchers, Andrade-Ochoa et al.28, showed that a combined trans-cinnamaldehyde + trans-anethol formulation and a combined trans-cinnamaldehyde + (–)-limonene formulation were highly insecticidal against Cx. quinquefasciatus larvae and pupae, synergistically high. Youssefi et al.29 stated that a combination of thymol + carvacrol provided strong ovicidal and larvicidal activities against Cx. pipiens.

To conclude, d-limonene, geranial, and trans-cinnamaldehyde show several dominant activities for mosquito and other insect pest control. Single and combined formulations of d-limonene showed a strong larvicidal activity against Ae. aegypti25 and also shown a strong pupicidal activity against Ae. aegypti and Ae. albopictus27. Geranial showed a strong adulticidal activity against Ae. aegypti, Ae. albopictus, and M. domestica30,31. Single and combined formulations of trans-cinnamaldehyde also showed a strong adulticidal activity against Ae. aegypti and Ae. albopictus4,15.

From these pieces of studies, our group was inspired to investigate the egg mortality against Ae. aegypti and Ae. albopictus incurred by EOs from C. aurantium, C. citratus, and C. verum, their major constituents, and several of their combinations. In addition, the biosafety of the EO treatment was evaluated against two common, non-target predators of mosquitoes, Poecilia latipinna and Poecilia reticulata fishes.

Results

Essential oils and GC/MS analysis

GC–MS analysis of EOs was necessary because different parts of the three plant species—C. aurantium, C. citratus, and C. verum EOs—gave different EO chemical profiles, and hence can make an accurate efficacy comparison between studies meaningless.

All EOs were pale yellow. Table 1 is a list of the components of essential oils discovered by GC–MS. The highest percentage of extraction yield, at 1.30% v/w, was recovered from C. aurantium EO, followed by from C. citratus EO (1.14% v/w), and from C. verum EO (1.01% v/w). C. aurantium EO, 21 chemical constituents were found to compose 96.67% of its chemical composition. d-limonene (78.15%) was the major constituent. Some other main constituents were linalool (4.80%), δ-3-carene (2.40%), and β-myrcene (2.00%). For C. citratus EO, 9 chemical constituents were found to compose 96.54% of its chemical profile. The major constituent was geranial (45.41%). A few other main constituents were neral (24.80%), 1,8-cineole (10.59%), and geraniol (4.70%). For C. verum EO contained 14 constituents as 97.26% of its chemical profile. Trans-cinnamaldehyde (73.21%) was the major constituent. Some other main constituents were benzyl alcohol (12.87%), cinnamyl acetate (2.50%), and eugenol (2.35%).

Table 1 Chemical composition of the essential oils Cinnamomum verum, Citrus aurantium, and Cymbopogon citratus.
Full size table

Toxicity against target mosquito

The following ovicidal activity indexes for each formulation against the two mosquito species: inhibition rate, 50% Lethal time (LT50), 50% Lethal concentration (LC50), effective inhibition rate index (EII) versus temephos are tabulated in Tables 2 and 3. The table also includes the determined synergistic index (SI) of each formulation. For example, from Table 2, C. aurantium EO at 30,000 ppm alone provided an egg inhibition rate of 78.1%, an LT50 of 55.3 h, and an LC50 of 15,071.7 h, with EII = 2.56 against Ae. aegypti. From Table 3, C. aurantium EO at 30,000 ppm alone provided an egg inhibition rate of 76.3%, an LT50 of 51.7 h, and an LC50 of 16,592.2 h, with EII = 2.59 against Ae. albopictus. Regarding the synergistic index (SI) column, since this formulation is of single C. aurantium EO, the synergistic index of combined formulation is not applicable. Several individual EOs and EO constituents at 30,000 ppm showed a significantly higher efficacy than at a lower concentration. The highest egg inhibition rate of individual EOs was at 91.0% against Ae. aegypti and 93.0% against Ae. albopictus, achieved by C. verum EO. At 30,000 ppm. It provided an LT50 of 29.7 h against Ae. aegypti and 31.1 h against Ae. albopictus. In contrast, at 10,000 ppm, C. aurantium EO provided the lowest egg inhibition rate, at 59.7%, against Ae. aegypti and 53.2% against Ae. albopictus, with an LT50 of 83.6 h and 85.8 h, respectively. The highest egg inhibition rate of EO constituents was 96.7% against Ae. aegypti and 95.2% against Ae. albopictus, achieved by trans-cinnamaldehyde, the major constituent of C. verum EO. At 30,000 ppm, trans-cinnamaldehyde provided an LT50 of 24.3 h against Ae. aegypti and 24.7 h against Ae. albopictus. In contrast, at 10,000 ppm, d-limonene, the major constituent of C. aurantium EO, provided the lowest egg inhibition rate against Ae. aegypti at 76.4% and against Ae. albopictus at 72.1%, with an LT50 of 39.5 h and 36.3 h, respectively. To conclude, C. verum EO exhibited a stronger ovicidal activity (lower LC50) against Ae. aegypti, but trans-cinnamaldehyde was stronger against Ae. albopictus.

Table 2 Ovicidal effects of EO formulations from Cinnamomum verum, Citrus aurantium, and Cymbopogon citratus EOs and their major constituents and combined formulations on the hatching rate of Aedes aegypti eggs after 48 h of incubation.
Full size table
Table 3 Ovicidal effects of EO formulations from Cinnamomum verum, Citrus aurantium, and Cymbopogon citratus EOs and their major constituents and combined formulations on the hatching rate of Aedes albopictus eggs after 48 h of incubation.
Full size table

Moreover, the ovicidal efficacy of all combined formulations against the eggs of the two mosquito vectors was greater than the efficacy of individual EOs and EO constituents, with a synergistic index (SI) in the range of 0.26–0.49. The highest egg inhibition rate was at 100% against both Ae. aegypti and Ae. albopictus, achieved by the combination of C. verum EO + geranial (2:1) 10,000 ppm. It provided an LT50 of 17.7 h against Ae. aegypti and an LT50 of 16.9 h against Ae. albopictus. In contrast, the combination of C. aurantium EO + geranial (2:1) 10,000 ppm provided the lowest egg inhibition rate, at 81.0% against Ae. aegypti and 80.0% against Ae. albopictus, with an LT50 of 35.8 h and 34.3 h, respectively. To conclude, C. verum EO + geranial (2:1) 10,000 ppm exhibited a stronger ovicidal activity against both Ae. aegypti and Ae. albopictus, in terms of low LC50.

Regarding the effective inhibition rate index (EII), every combination of separate EOs and EO components as well as all combined formulations showed higher than 1.0 EII—they were more toxic to the eggs of Ae. aegypti and Ae. albopictus than 1 ppm temephos was.

Toxicity against non-target aquatic predators

The estimated LC50 values against P. latipinna and P. reticulata, two fish species, of all formulations are summarized in Table 4. The two species were less susceptible to individual EOs and trans-cinnamaldehyde, d-limonene, and geranial EO constituents than every combined formulation, in terms of LC50. The range of LC50 values against the two fish species of those EOs and EO constituents was from 8165.5 to 57,232.5 ppm, while for the combined formulations, the range was 4091.6–5921.3 ppm. On the other hand, the range for 1 ppm temephos was high toxic to both species with LC50 from 298.7 to 526.7 ppm.

Table 4 Effect of Cinnamomum verum, Citrus aurantium, and Cymbopogon citratus EOs and their major constituents and combined formulations against Poecilia latipinna and Poecilia reticulate fishes sharing the same ecological niche of Aedes aegypti and Aedes albopictus.
Full size table

On the biosafety index (BI) are shown in Table 5. All formulations provided a high BI from 1.03 to 9.77, these BIs were higher than 1. Therefore, all formulations were not toxic to both fish species.

Table 5 Biosafety index (BI) against Poecilia latipinna and Poecilia reticulate sharing the same ecological niche of Aedes aegypti and Aedes albopictus, exposed to Cinnamomum verum, Citrus aurantium, and Cymbopogon citratus EOs and their major constituents and combined formulations.
Full size table

Discussion

Regarding extraction yield, the extraction yields of all tested plants were the same or only slightly different from the corresponding yields reported by previous studies20,32. The extraction yield of C. verum EO was 1.05% v/w compared to 1.1% v/w found by Aungtikun and Soonweera4 and Soonweera et al.20; the extraction yield of C. aurantium EO was 1.30% v/w compared to 1.4% v/w found by Bnina et al.33; and the extraction yield of C. citratus EO was 1.14% v/w, compared to 1.2% v/w found by Soonwera et al.20. The slight differences can be attributed to many factors, e.g., the harvesting season, the integrity of the plant species, the adequate plant management, and the degree of fertility of the soil (soil chemicals and relative humidity)20,32.

On GC–MS analysis results, the chemical profiles of all tested EOs agreed well with those found in previous studies13,20. The major chemical constituent and key component (the active component12,20) of C. verum EO was trans-cinnamaldehyde (73.21% of the profile), very close to 72.2% reported by Soonwera et al.20; the major chemical constituent and key component13 of C. aurantium EO was d-limonene (78.15%), agreeing well with 73.6% supported by Bnina et al.33; and the major chemical constituent and key component14 of C. citratus EO was geranial (45.41%), agreeing well with 49.4% found by Chanthai et al.34. Nevertheless, some papers report larger differences. For C. verum EO, Chansang et al.15 reported a higher percentage of trans-cinnamaldehyde (90.2% compared to 73.21%); for C. aurantium EO, Zarrad et al.32 reported a higher percentage of d-limonene (87.5% compared to 78.15%); and for C. citratus EO, Brügger et al.35 reported a lower percentage of geranial (31.5% compared to 45.41%). This variation can be attributed to seasonal fluctuations, differences in temperature at the farms where these plants were cultivated, geographic location, ontogenetic variables, the growth stage of the plant at the time of harvest (pest management), and extraction method27,32, and all key components were robustly identified and quantified.

Regarding ovicidal efficacy results, based on LT50, the ranking of inhibition against Ae. aegypti and Ae. albopictus eggs was as follows: (1) C. verum EO + geranial (2:1), (2) geranial + trans-cinnamaldehyde (1.5:1.5), (3) d-limonene + trans-cinnamaldehyde (1.5:1.5), (4) C. citratus EO + d-limonene (2:1), (5) d-limonene + geranial (1.5:1.5), and (6) C. aurantium EO + geranial (2:1). All tested formulations were more effective than temephos, and one formulation was outstanding. The outstanding combination of C. verum EO + geranial (2:1), at the final concentration of 10,000 ppm each, showed the shortest lethal time (LT50 ranging from 16.9 to 17.7 h) and the smallest lethal concentration (2303.5 ppm for 100% mortality). In contrast, temephos showed a lethal time in the range of 60.2–61.3 h, which is longer than that (ranging from 34.3 to 35.8 h) of the least effective EO formulation, C. aurantium EO + geranial (2:1) 10,000 ppm. Previous works4,27 have already established the potent toxicity of several combined EOs and EO constituents against mosquitoes at most stages in their life cycle, except at the egg stage. The effectiveness of the formulations that combined EO and EO constituents and targeted mosquitoes at the egg stage of their life cycles was first reported in this study. Regarding the most effective combined formulation, it was not surprising that it was so effective since its individual components, C. verum EO and geranial, have already been shown to be effective against several pest insects, as presented in the following papers. Nakasen et al.24 supported that C. verum EO at 12.5 ppm showed high ovicidal activity against Cx. quinquefasciatus with a 100% inhibition rate and an LC50 of 3.31 ppm. Soonwera et al.20 reported that C. verum EO had strong ovicidal activity against Periplaneta americana. Dias et al.36 indicated that trans-cinnamaldehyde exhibited a strong insecticidal effect against Mahanarva spectabilis eggs. Trans-cinnamaldehyde showed a toxic effect against the eggs of P. humanus capitis37, and Spodoptera littoralis19. Finally, Castillo-Morales et al.38 reported that geranial provided strong ovicidal activity against Ae. aegypti. Regarding the low efficacy of temephos, It can be inferred that it was low because it was not designed specifically to kill mosquitoes at the egg stage but at the larval stage. The mosquito subjects were a laboratory-selected strain, not field-collected, and hence the larvae have not developed resistance to temephos. Their morbidity was confirmed. The low egg-inhibition activity of temephos is supported by previous works of Puwanard and Soonwera23 and Cotchakaew and Soonwera39, indicating that 1% (w/w) temephos showed an inhibition rate ranging from 9.3 to 34.6% against the eggs of Ae. aegypti and Ae. albopictus, while the EOs showed a 47.0–100% inhibition rate.

On egg morphology, its SEM images in Figs. 1 and 2 show damages to the exochorionic meshwork and tubercles of the outer cells on exochorion cuticle (external chitin layer) with cell borders, papillae, and aeropyles40,41. Moreover, the cell borders and papillae as well as aeropyles were covered with a layer that was assumed to be an oil layer, which would explain the ovicidal mechanism of the EO, discussed in the paragraph below.

Figure 1
figure 1

Scanning electron micrographs of Aedes aegypti eggs: (A,B) non-treated egg, intact exochorionic cuticle with cell borders (b), papillae (p), and aeropyles (a), morphological damage to exochorionic cuticle after treated with d-limonene (C), geranial (D), and trans-cinnamaldehyde (E).

Full size image
Figure 2
figure 2

Scanning electron micrographs of Aedes albopictus eggs: (A,B) non-treated egg, intact exochorionic cuticle with cell borders (b), papillae (p), and aeropyles (a), morphological damage to exochorionic cuticle after treated with d-limonene (C), geranial (D), and trans-cinnamaldehyde (E).

Full size image

Regarding the mechanisms of ovicidal action, as stated in the paragraph above, the aeropyles seemed to be blocked by an oil layer, making respiration difficult or impossible. This respiration inhibition mechanism has been reported by Khedr et al.19 C. verum EO induced mortality of embryo and egg by forming a thin film of oil over the outer egg surface and blocking the egg respiration by sealing the aeropyles. Nakasen et al.24 concluded that C. verum EO destroyed chitin wall by the oil penetrating the eggshell pore leading to embryo death. Another possible mechanism of action is the mechanism that trans-cinnamaldehyde, the major constituent of C. verum EO, acts on the egg. Trans-cinnamaldehyde reduces the ATPase activity in the cell membrane of the respiratory system and inhibits the enzymes involved in cytokinesis as well as retards juvenile hormone production and cell growth in the immune system of mosquito4,42. In short, C. verum and trans-cinnamaldehyde act mainly on the respiratory system of mosquito eggs. Contrarily, geranial affects the egg's neurological system. Geranial inhibits the acetylcholinesterase (AChE) enzymes of neural cells and neuroreceptors30. Castillo-Morales et al.38 concluded that geranial penetrates through the serosal cuticle of an embryo and disturbs the embryogenesis process. Hence, the synergistic effect of the combination may stem from the fact that both substances acted along two different pathways, reinforcing one another.

On the biosafety of non-target aquatic predators of mosquito eggs, the combined EO formulation was deemed safe for P. latipinna and P. reticulata, two species of predator fish, because its BI was more than 1 and its high lethal concentration (LC50). EOs are also generally considered safe for other arthropods and fishes43. Other authors have supported the conclusion that EOs are safe for non-target organisms. Alsalhi et al.43 supported that trans-cinnamaldehyde showed very low toxicity on Gambusia affinis (LC50 = 3960.6 ppm). Nwanade et al.17 reported that trans-cinnamaldehyde provided a less toxic effect on Tenebrio molitor (LC50 = 28.4 μL/mL). In addition, Hýbl et al.44 indicated that C. zeylanicum EO did not show toxicity against honey bee, Apis mellifera (LC50 = 4.542 μL). Sabahi et al.45 reported that C. citratus EO was not toxic to A. mellifera (LD50 = 53,304.0 μg/mL). It has also been shown that geranial, the major constituent of C. citratus EO, had a low negative effect on a predatory bug, Podisus nigrispinus (LD50 = 25.56 μg/insect−1)35. In contrast, temephos is highly toxic to several non-target organisms e.g., Acilius sulcatus, Anisops bouvieri, and G. affinis with LC50 ranging from 0.957 to 4.817 ppm43. Chellappandian et al.11 reported that temephos showed a highly toxic effect against aquatic mosquito larvae predator, Toxorhynchites splendens. Along the same line, USA EPA46 concluded that temephos showed highly acute toxicity to risk quotients for freshwater fish: the LC50 against rainbow trout was 3490 ppb. Similarly, in this study, 1% (w/w) temephos showed a high level of toxicity to two fish species, P. latipinna and P. reticulata with LC50 ranging from 298.7 to 526.7 ppm. Furthermore, temephos resists degradation and accumulates in the environment at a high level, thus harming non-target organisms46. On the contrary, EOs and their constituents are natural substances that degrade quickly in the environment and do not accumulate in the environment, hence much safer for the environment. More than one BI for all formulations verified that those formulations were absolutely safe for these non-target aquatic predators. The mortality rate after the treatment of the eggs was much higher than the mortality rate against the fishes (Fig. 3). Most importantly, both EOs from C. verum and C. citratus as well as their major constituents do not exhibit cytotoxicity activity on human fibroblast cells47,48 and show high LD50 value on mammals49,50. They have long been used as a food ingredient, cosmetics, and folk medicine12,14,51. They quickly degraded in the environment50,51. On the other hand, temephos is toxic to the nervous systems of humans. It can cause Alzheimer’s disease as reported by Martins Laurentino et al.52. Because of its efficacy and safety, the combined formulation of C. verum EO + geranial should be developed as a natural insecticide for controlling the eggs of Ae. aegypti and Ae. albopictus to replace commercial synthetic insecticides. However, to develop the combined formulation into a commercial product (a spray or drops of solution into the water), it is still necessary to investigate other factors that affect to mortality of mosquito vectors and their eggs, e.g., a field study and a study of the post-application temperature effect.

Figure 3
figure 3

Mortality rates of the combined formulations against the eggs of Aedes aegypti and Aedes albopictus, compared to those against non-target predators of mosquitoes, Poecilia latipinna and Poecilia reticulate fishes.

Full size image

Materials and methods

Plant collection

All plants were obtained under national and international guidelines. The plants were collected under the supervision and permission of the School of Agricultural Technology, KMITL. All of the authors complied with all local and national guidelines.

Dried barks of C. verum were purchased from a local Chinese pharmacy in Thailand (Chao Krom Poe Dispensary, Bangkok, Thailand). Peels of C. aurantium fruit were obtained from a farm in Nakhon Ratchasima province, Thailand (14° 58′ 47.6400″ N/102° 5′ 51.9756″ E). Fresh stems of C. citratus were obtained from a farm in Chanthaburi province, Thailand (12° 36′ 34″ N/102° 06′ 16″ E) in July–October 2021. All plant species were identified by Mr. Tanapoom Moungthipmalai, a herbal specialist at the KMITL herbal museum, and some of the specimens were kept at the KMITL herbal museum, School of Agricultural Technology, KMITL.

Essential oil extraction

Plant part (1000 g) was washed and extracted by hydro-distillation in 2000 mL of distilled water at 100 °C for 5 h. The rate of distillation was two drops of EO per second. The EO was then filtered and stored in a tea color bottle at 4 °C.

Identification of essential oil constituents through GC/MS

Samples of C. aurantium, C. citratus, and C. verum EOs were analyzed by an Agilent 6890 N gas chromatograph GC–MS at the central Laboratory, KMITL, following our previous protocol27. Serving as the mobile phase is 1 mL per min flow of helium (99.99%). To start, 0.2 μL of extract in ethyl alcohol solution (split ratio = 1:100) of each EO was injected into the column. A 5973-N mass spectrometer (using an HP-5 MS fused silica capillary column (30 m × 0.25 mm ID with 0.25 m film thickness of 5% phenyl-methylpolysiloxane coating), an electron ionization system with 70 eV electron energy (30–500 m z−1), and an Agilent 6890-N gas chromatograph (USA) made up the GC–MS system. The column temperature was programmed to increase gradually from room temperature to 50 °C and stay there for 2 min. The column temperature was then increased to 200 °C and maintained there for 3 min at a rate of 10 °C min−1. In the final stage, the column temperature was raised to 260 °C at a rate of 15 °C min−1 and held there for 20 min. The injector and detector temperature were held at 270 °C. The total running time was 40 min. A mass spectra search program with Wiley 7 N library was used for identifying all components of EOs. The mass spectra of peaks were compared with those stored in Adams53 and NIST 1754 libraries. Temperature-programmed retention indices (RI) were determined using n-alkanes (C7–C30). The experiment was performed in three replicates.

Source and purity of reagents

Cinnamaldehyde (98% pure), a major constituent of C. verum EO, d-limonene (96% pure), a major constituent of C. aurantium EO, and geranial (96% pure), a major constituent of C. citratus EO together with standard n-alkanes (C7–C30) were supplied by Sigma-Aldrich company (USA). Temephos (1 ppm), the positive control, was obtained from Thailand’s Government Pharmaceutical Organization (GPO) (Pathum Thani, Thailand). Ethyl alcohol (95% v/v) was supplied by Thailand’s Liquor Distillery Organization (Chachoengsao, Thailand). All chemicals used in this study were reagent-grade.

Insect maintenance

The mosquito eggs used in this experiment were freshly laid eggs of mosquitoes of two species, Ae. aegypti and Ae. albopictus, reared in the Entomology laboratory at the School of Agricultural Technology, KMITL. They were reared under the conditions of 26.5 ± 2 °C temperature, 75.0 ± 2% RH, and an 11 ± 13 h photoperiod. Female adult mosquitoes were fed with 2.5% glucose solution + 2.5% multivitamin syrup solution and periodically blood-fed via membrane by an artificial membrane method1. The first generation of eggs was used in various experiments.

Toxicity against target mosquito

Ovicidal activity bioassay was performed on the eggs following the method of Puwanard and Soonwera23. The eggs used for this bioassay were stored at 26.5 ± 2 °C for 7 days after female mosquitoes had laid their eggs on a Whatman No.1® filter paper. Eggs were selected under a stereomicroscope (Nikon® Type 102): abnormal eggs were discarded, and normal eggs were collected for the bioassay. For each mosquito species, 25 eggs were suspended in 99 mL of distilled water in a 150 mL plastic cup. A treatment was added to the cup: 1 mL of each EO formulation. This assay was done in ten replicates for each treatment, with positive, negative, and neutral controls: 1 ppm temephos (based on the recommendation of Thailand’s Government Pharmaceutical Organization (GPO) for destroying mosquito larvae), 70% (v/v) ethyl alcohol, and pure water, respectively. The numbers of hatched larvae at 30 min, 1, 6, 24, and 48 h post-treatment were observed and recorded because it was easier and more practical to count live larvae than to count dead eggs under a stereomicroscope. The percentage inhibition rate after 48 h was determined by the formulas23 below,

$$\mathrm{Hatching\, rate }\left(\mathrm{\%}\right)= \left[\left(\frac{\mathrm{NE}}{\mathrm{NT}}\right)\times 100\right],$$
(1)
$$\mathrm{Inhibition\, rate }(\mathrm{\%}) = 100 -\mathrm{ hatching\, rate }(\mathrm{\%}),$$
(2)

where NE is the total number of hatched eggs and NT is the total number of eggs.

The effective inhibition rate index (EII) as a comparative efficacy index between an EO and temephos, was determined by the formula4 below,

$$\mathrm{EII }= [\mathrm{\% \,inhibition\, rate\, of\, each\, EO\, formulation}/\mathrm{\% \,inhibition\, rate\, of\, temephos}].$$
(3)

EII < 1 indicates that the EO formulation was not as effective as temephos; EII = 1 indicates that the EO formulation was as effective as temephos; and IRI > 1 indicates that the EO formulation was more effective than temephos.

Synergistic index (SI) is an efficacy comparison index between a combined formulation and the corresponding individual EO or individual EO constituent. It was calculated by the following formula31,

$${\text{SI}} = [{\text{LT}}_{{{5}0}} {\text{\,of\, combined\, formulation}}/\left( {{\text{LT}}_{{{5}0}} {\text{\,of\, individual\, EO}} + {\text{LT}}_{{{5}0}} {\text{of\, individual\, EO\, constituent}}} \right)].$$
(4)

SI < 1 indicates synergistic; SI > 1 indicates antagonistic; and SI = 0 indicates not either one.

Toxicity against non-target aquatic predators

The experimental methods and procedures were performed in accordance with the guidelines and regulations of the National Research Council of Thailand guide for the care and use of laboratory animals and approved by the King Mongkut’s Institute of Technology Ladkrabang of animal care and use committee. This study was carried out in compliance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines.

The effect of individual EOs, EO constituents, and combined formulations against non-target aquatic predators, P. latipinna and P. reticulata, was analyzed with a modified technique reported by Rajeswary et al.55. The test used four concentrations (i.e., 500, 1000, 2500, and 5000 ppm) of treatment that corresponded to the estimated LC50 against the two mosquito species. Both fish species were purchased from a farm in Nakhon Pathom province, Thailand. They were separately kept in a glass container containing 10.5 L of water at 35 ± 2 °C and 77 ± 5% RH. With the registration number KDS2021/002 (August 2nd, 2021), the King Mongkut's Institute of Technology Ladkrabang's Ethics committee had approved each bioassay used in this study. One adult P. latipinna or P. reticulata was put in a glass jar containing 99 mL of water and contaminated with a treatment at a specified concentration. Four replicates were done for each treatment with positive control. Data on mortality and swimming sluggishness were recorded for 5 days post-treatment.

The biosafety index (BI) was determined by the formula55 below,

$${\text{BI}} = \left[ {{\text{LC}}_{{{5}0}} {\text{\,of\, non-target\, aquatic\, predators}}/{\text{LC}}_{{{5}0}} {\text{\,of\, target\, vector\, species}}} \right].$$
(5)

BI > 1 indicates that the EO formulation was safe for the non-target organisms, and BI < 1 indicates that the EO formulation was not safe for non-target organisms.

Egg morphology and observation

After 48 h of treatment, the morphology of the external surface of the eggs of each mosquito species that underwent a treatment or control was observed under scanning electron microscopy (SEM) at the Scientific and Technological Research Equipment Centre, Chulalongkorn University, Thailand. Samples were placed in a fixative, 2.5% glutaraldehyde for 30 min in 0.1 M phosphate buffer. Thoroughly washed with the same buffer, the eggs were dehydrated by soaking in a series of alcohol solutions in water (30, 50, 70, and 95%). Each 1-h soaking process with an alcohol solution was replicated three times with an automatic tissue processor. Then, the eggs were dried with a CO2 critical point drier. Each dehydrated sample was mounted on a stub coated with gold–palladium and examined with a JSM-5800 LV (Tokyo, Japan) SEM. Photographs of the egg surface morphology were taken.

Statistical analysis

The design of the experiments was completely randomized. Analysis of variance (ANOVA) and Duncan’s multiple range test at P < 0.05 were applied to the mortality data of mosquito eggs. The treatment time to produce 50% egg mortality (LT50) was determined by probit analysis. The eggs were observed at 30 min, 1, 6, 24, and 48 h after the treatment. The concentration of a treatment that provided 50% mortality (LC50) against mosquito eggs was determined. The tested concentrations were 10,000 and 30,000 ppm for individual EOs and 5000, 10,000, and 30,000 ppm for individual EO constituents. The LC50 values against the two species of fish were determined at 500, 1000, 2500, and 5000 ppm. SPSS Statistical Software Package version 22 was the statistical software package used.