Harnessing insect-derived oils for enhanced efficacy of plant-based repellents against disease-transmitting mosquitoes

Abstract

Mosquito-borne diseases such as malaria, dengue, chikungunya, Zika, and yellow fever pose a growing global health threat, worsened by resistance to synthetic pesticides. This study evaluated insect oils from Schistocerca gregaria, Ruspolia differens, and Macrotermes spp. as adjuvants in plant-based mosquito repellents. The formulations were tested for complete protection time (CPT) against Anopheles gambiae, Aedes aegypti, and Culex quinquefasciatus, and volatile organic compounds (VOCs) emitted from treated volunteers were analyzed. Incorporating insect oils doubled CPT compared to essential oils alone. The most effective blend of Macrotermes oil with Cymbopogon nardus (1:1) achieved CPTs of 3.50 h for An. gambiae, 2.56 h for Ae.aegypti, and 2.81 h for Cx. quinquefasciatus, showing repellency comparable to 20% DEET during the first 3.5 h. Blends retained three times more VOCs than plant oils alone, prolonging repellent activity. Key compounds such as geranyl acetate, citronellyl acetate, and geranyl propanoate persisted on skin for over 180 min. These findings highlight the potential of insect oils, particularly from Macrotermes spp., to enhance the efficacy and longevity of plant-based mosquito repellents, offering a sustainable, eco-friendly alternative amid growing pesticide resistance.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-38831-x.

Introduction

Vector-borne diseases (VBDs) account for approximately 20% of the global tropical infectious disease burden, causing nearly 1 million deaths annually, with sub-Saharan Africa disproportionately affected¹. In East Africa, major VBDs such as malaria, lymphatic filariasis, and dengue are primarily transmitted by mosquitoes, with malaria accounting for the highest morbidity and mortality². In Kenya, malaria is responsible for approximately 3.5 million new clinical cases and 10,700 deaths annually³. The principal malaria vectors include Anopheles gambiae s.s (sensu stricto), Anopheles arabiensis, and Anopheles funestus s.l (sensu lato)⁴. According to the World Health Organization, Aedes aegypti transmitted infections, including dengue, chikungunya, Zika, and yellow fever, are rapidly expanding due to urbanization, globalization, and climate variability. Dengue has reached record levels, with over five million reported cases across more than 130 countries, while outbreaks of chikungunya and yellow fever continue to threaten populations in Africa, Asia, and the Americas. Zika transmission remains sporadic but under close surveillance because of its congenital complications. In addition, Culex quinquefasciatus borne diseases such as Lymphatic filariasis and West Nile virus persist in tropical and subtropical regions⁵.

In present-day malaria control practice, vector management primarily relies on chemical insecticides targeting various developmental stages of mosquitoes to reduce disease transmission. In endemic regions, insecticide-treated nets (ITNs) remain the most widely used method for protecting populations from malaria vectors. These nets are typically impregnated with pyrethoids, which the World Health Organization (WHO) recommends because of their high efficacy at low doses, rapid knockdown effect, and relative safety to humans⁴. However, the extensive use of pyrethoids has led to widespread resistance among mosquito populations across Africa, mainly due to mutations at the target site, commonly referred to as knockdown resistance (Kdr)^6–8. Furthermore, many once effective and affordable antimalarial drugs have become largely ineffective against the malaria parasite⁴. These challenges, pesticide resistance, antimalarial drug resistance, and the absence of fully effective vaccines, highlight the urgent need for innovative, sustainable vector management strategies^9,10. Consequently, the most promising yet demanding approach involves integrating multiple, locally adapted control methods based on specific ecological and resource contexts.

Mosquito repellents are among the most effective tools for minimizing mosquito bites and reducing the transmission of mosquito-borne diseases such as malaria, dengue, and yellow fever^11,12. Botanical repellents derived from natural sources such as citronella, neem, eucalyptus, and lemongrass oils are increasingly favored due to their biodegradability, ecological safety, and consumer preference for natural products. However, their practical use is limited by high volatility, rapid degradation under tropical conditions, and poor skin retention, resulting in shorter Complete Protection Times (CPT) compared to synthetic repellents like DEET(N,N-Diethyl-meta-toluamide) or Picaridin¹³. Despite significant advancements in repellent development, synthetic compounds dominate the global market^14–16. DEET, Picaridin, IR3535, and para-menthane-3,8-diol (PMD) are the principal active ingredients (A.I.s) in most commercial formulations, providing extended protection ranging from 5 to 12 h¹⁷. Picaridin provides similar protection (8–10 h) with a more pleasant sensory profile¹⁸, while IR3535 offers moderate protection (4–6 h) and is often used in formulations for children¹⁹. PM, a plant-derived repellent, delivers 5–7 h of protection and is marketed as a natural alternative²⁰. To enhance efficacy, these actives are combined with various adjuvant solvents, emulsifiers, fixatives, and penetration enhancers that improve stability, release rate, and skin adherence. Nevertheless, many of these synthetic additives present drawbacks such as dermal irritation, allergenic potential, volatility, and poor compatibility with natural bioactives¹³. Moreover, increasing environmental and health concerns associated with persistent synthetic chemicals have intensified the demand for eco-friendly, biocompatible alternatives.In response, nature-based formulations have gained momentum. Plant-derived oils and essential oils are used not only as repellents but also as carriers and stabilizers in biobased products^21,22. However, their application remains constrained by physicochemical limitations, particularly their rapid evaporation and oxidation under field conditions²³. These shortcomings have created an urgent need for new adjuvant systems that can enhance the protection time, stability, and bioefficacy of natural repellents while maintaining safety and sustainability^{12,13,23–28}.

In this context, insect-derived oils represent a promising yet underexplored alternative. Edible insects have attracted considerable attention as sustainable sources of protein, lipids, and micronutrients for human food and animal feed^27,29–31. During industrial processing, large quantities of insect oils are produced as by-products, yet these are often discarded or underutilized³². Chemically, insect oils exhibit distinct profiles rich in medium-chain and monounsaturated fatty acids, particularly oleic acid³³ which confer higher oxidative stability than the polyunsaturated fatty acid (PUFA)-rich plant oils prone to rancidity and degradation^34,35. Their semi-solid consistency, favourable saponification values, and emollient properties also make them suitable for cosmetic and pharmaceutical formulations³⁵. These physicochemical and functional characteristics suggest that insect oils could serve not merely as carriers but as bioactive adjuvants capable of enhancing the performance of volatile essential oils.

Despite these promising attributes, the application of insect oils in mosquito repellent development remains largely unexplored. Preliminary studies have shown that insect-derived materials such as chitin, chitosan, and oils can enhance the efficacy and stability of bioactive agents in agricultural and pest management contexts^36,37. Yet, their potential role in extending the protection duration, improving adherence, or stabilizing natural repellents has not been systematically investigated.

Addressing this gap, the present study explores the potential of three insect oils Schistocerca gregaria (desert locust), Ruspolia differens (nsenene), and Macrotermes spp. (termite) as functional adjuvants for formulating plant-based mosquito repellents using Cymbopogon nardus (citronella), Ocimum kilimandscharicum (African camphor basil), and Eucalyptus citriodora (lemon eucalyptus) essential oils. The study aims to evaluate whether integrating insect-derived oils with plant essential oils can improve repellent longevity, stability, and overall efficacy against disease-transmitting mosquitoes.

To test this hypothesis, oils from S. gregaria, R. differens, and Macrotermes spp. were extracted and characterized alongside essential oils from C. nardus, O. kilimandscharicum, and E. citriodora. The oils were analyzed for yield and chemical composition using standard extraction and GC–MS techniques. Subsequently, individual and blended formulations of insect and plant oils were developed to evaluate their physicochemical compatibility and repellent efficacy against major mosquito vectors. This approach aimed to establish whether insect oils can function as bioactive adjuvants that enhance the stability, longevity, and performance of plant-based repellent species.

Results

Yield of essential oils

Among the insect-derived oils, Ruspolia differens exhibited the highest yield (4.5–7.2%), followed by Schistocerca gregaria (2.0–5.0%) and Macrotermes spp. (1.5–3.0%). The superior yield observed in R. differens is attributed to its naturally high fat content, making it a promising candidate for oil extraction in entomophagy and other bio-based applications. In contrast, the yields of plant-derived essential oils, among these, Eucalyptus citriodora produced the greatest yield (1.0–2.0%), followed by Ocimum kilimandscharicum (0.7–1.5%) and Cymbopogon nardus (0.5–1.5%).

Chemical composition of insect oils and plant essential oils

A total of 143 compounds were identified across both insect-derived and plant essential oils. These compounds were classified as alcohols (6), fatty acids (45), monoterpenes (30), sesquiterpenes (32), alkanes (8), sterols (7), aldehydes (2), hydrocarbons (14), ketones (4), and α-tocopherol (1) (Table 1; Supplementary Table S1). The chemical profiles showed distinct differences between sources: plant essential oils were dominated by terpenes, whereas insect oils primarily contained hydrocarbons, fatty acids, and sterols. Both qualitative and quantitative variations were observed in the volatile compositions of the oils. Notably, dodecanoic acid, heptadecanoic acid, ethyl hexadecanoate, n-hexadecanoic acid, and oleic acid were the most abundant components in the insect oils, with S. gregaria showing the highest concentrations (Table 1). Among all compounds, dodecanoic acid exhibited the greatest concentration in S. gregaria at 1874.21 ± 152.32 ng/µL of oil. In the plant essential oils, citronellal, a monoterpene aldehyde, was the most prominent compound, detected at 183.61 ± 38.04 ng/µL in E. citriodora and 100.03 ± 36.59 ng/µL in C. nardus. Common compounds identified in both plant and insect oils included p-xylene, o-xylene, and α-pinene, except in S. gregaria. The compound δ-3-carene was detected in all oil samples except those from S. gregaria and O. kilimandscharicum. Squalene was unique to insect oils. Meanwhile, terpenes such as limonene, linalool, β-myrcene, γ-muurolene, germacrene D, and α-muurolene were characteristic of plant essential oils (Table 1; Supplementary Table S1).

Table 1.

Composition of volatile organic compounds (ng/uL)) in insect and plant oils.

RT	Compound	RICal*	Mean ± SE, (standard error) (ng/µLof oil)
			Ruspolia differens	Macrotermes spp	Schistocerca gregaria	Ocimum kilimandscharicum	Eucalyptus citriodora	Cymbopogon nardus
Fatty acid
19.32	Dodecanoic acid	1517	9.0 ± 0.74	65.15	1874.21 ± 152.32
21.55	Tetradecanoic acid	1709	8.8 ± 0.65	14.09 ± 2.59	380.51 ± 40.45
24.41	n-Hexadecanoic acid	2013	192.5 ± 101.38	56.20 ± 13.79	1046.22 ± 50.20
Ketone
8.16	o-Xylene	854	4.3 ± 0.45	3.95 ± 0.64		4.87 ± 1.08	4.43 ± 0.96	3.73 ± 0.42
8.17	p-Xylene	854	4.2 ± 0.72	4.46 ± 0.68	3.91 ± 0.48	3.49 ± 0.44	4.53 ± 0.37	3.22 ± 0.12
Monoterpene
9.64	α-Pinene	913	4.5 ± 0.60	3.70 ± 0.38		12.96 ± 2.51	4.53 ± 0.62	3.77 ± 0.35
10.56	β-Myrcene	954				6.41 ± 1.29	6.91 ± 0.55	6.20 ± 0.12
11.23	δ-3-Carene	985	9.0 ± 0.48	7.55 ± 1.16			8.73 ± 0.69	6.64 ± 0.42
11.25	Limonene	985				143.90 ± 91.50	31.13 ± 7.84	16.68 ± 0.75
12.89	Linalool	1077				38.41 ± 21.10	10.88 ± 1.73	7.65 ± 0.43
13.34	Camphor	1107				382.74 ± 136.49
13.42	Citronellal	1111					183.61 ± 38.04	100.03 ± 36.59
14.57	Citronellol	1179					94.96 ± 27.90	53.96 ± 7.46
14.96	Geraniol	1202					117.42 ± 32.81	71.21 ± 7.87
Sesquiterpene
18.02	γ-Muurolene	1415				5.34 ± 0.95	7.29 ± 0.55	5.3 ± 0.35
18.10	Germacrene D	1422				7.18 ± 2.59	9.24 ± 0.95	9.46 ± 0.41
18.30	α-Muurolene	1437				4.71 ± 1.00	8.16 ± 1.28	7.55 ± 0.28
Triterpene
30.96	Squalene	2787	48.0 ± 18.94	25.41 ± 4.66	105.40 ± 9.95

RT- Retention time. Definitive compound identification via authentic standard; if not authenticated then compound identity is tentative by RI match to only one column. Std, comparison with available analytical standards. Mean ± SE (standard error) of triplicate determinations (ng/µL of oil).

Complete protection time of mosquito repellents from different plant essential oils infused with insect oils against An. gambiae

The pure plant EOs exhibited a Complete Protection Time (CPT) of 1.60 ± 0.25, 1.80 ± 0.37, and 0.9 ± 0.10 h for E. citriodora, C. nardus, and O. kilimandscharicum, respectively. On the other hand, the CPT of 0.70 ± 0.20, 0.60 ± 0.10, and 0.60 ± 0.10 h was recorded for insect oils from S. gregaria, R. differens, and Macrotermes spp, respectively. When plant oil from C. nardus was assayed in combination with the insect oils, 1:1 combination with oils from Macrotermes spp offered the maximum CPT of 3.50 ± 0.46 h, which was compared to 20% DEET at 3.75 ± 0.70 h (Fig. 1, Supplementary S2). When plant oil from E. citriodora was assayed in combination with the insect oils, 1:1 combination with oils from Macrotermes spp offered the maximum CPT of 2.90 ± 0.55 h followed by 1:1 combination with oils from S. gregaria with a CPT of 2.80 ± 0.26 h both of which were not significantly different from 20% DEET (Fig. 1). Combinations of insect oils and O. kilimandscharicum offered a CPT of 2.60 ± 0.45 h at 1 1:1 combination with oils from Macrotermes spp, followed closely by 1:1 combination with oils from S. gregaria (2.5 ± 0.44 h) (Fig. 1).

Fig. 1.

Complete protection time (h) of different combinations of essential oils with insect oils against An. gambiae. (a) E. citriodora, (b) C. nardus and (c) O. kilimandscharicum. Distinct letters denote statistically significant differences at p ≤ 0.05, as determined by the Tukey test.

Non-Metric Multidimensional Scaling (NMDS) revealed clear combinational clustering among the treatments. The 20%, 33%, 50%, and 67% combinations grouped closely on the negative side of NMDS1, indicating similar chemical profiles, while the 80% combination (20:80, essential oils: insect oils) separated distinctly on the positive side, corresponding to its low complete protection time (CPT). The repellent efficacy ranking followed the order 50% > 67% > 33% > 20% > 80%, with the 50% combination (1:1 ratio of plant essential oils to insect oils) providing the longest CPT against mosquitoes. These results indicate that balanced mixtures of plant and insect oils produce optimal repellent performance, whereas excessive insect oil content leads to compositional divergence and reduced efficacy, as depicted in Fig. 2.

Fig. 2.

Non-metric multidimensional scaling (NMDS) showing the clustering of different combinations of essential oils and insect oils irrespective of their sources against An. gambiae. Where Euc=Eucalyptus and Citro=citronella.

Further assays with the best combination (1:1, C. nardus: Macrotermes spp) against different mosquito species revealed that the formulation offered a CPT of 3.50 ± 0.46, 2.56 ± 0.32, and 2.81 ± 0.45 h against An. gambiae, Ae. aegypti and Cx. quinquefasciatus, respectively. The CPT values of DEET (positive control), against the three mosquito species, ranged from 3.38 to 3.75 h. The order of CPT per species was ranked as follows: An. gambiae > Cx. quinquefasciatus > Ae. aegypti. Statistically, the CPT of the best combination (1:1, C. nardus: Macrotermes spp) against An. gambiae were comparable to that of 20% DEET (Fig. 3, Supplementary S2).

Fig. 3.

Complete protection times (CPT) using a 1:1 C. nardus:Macrotermes spp against the three mosquito species (Ag = An. gambiae, Ae = Ae. aegypti and Cx = Cx. quinquefasciatus). MO=Mineral oil. CN = C. nardus oil. TM = marcrotermes sp oil. Distinct letters denote statistically significant differences at p ≤ 0.05, as determined by the Tukey test.

The percentage repellence rate of the best formulation (1:1, C. nardus: Macrotermes spp) with the highest CPT against the thee mosquito species revealed a decrease over time, from 100% until 180 min to 93.4% at 270 min, from 99.9% at 150 min to 87.9% at 270 min and from 98.8% at 150 min to 85.1% at 270 min against An. gambiae, Cx. quinquefasciatus and Ae. aegypti, respectively (Fig. 4).

Fig. 4.

Repellency of different mosquito’ species when exposed to the 1:1 C. nardus: Macrotermes spp formulation. Ag = An. gambiae, Ae = Ae. aegypti and Cx = Cx. quinquefasciatus.

Furthermore, a comparative study was carried out involving the combination of C. nardus with mineral oil (1:1) and mineral oil alone (100%) against An. gambiae. Low repellency was recorded when mineral oil was tested individually and in combination with C. nardus oil. Repellency of binary mixtures containing equimolar 1:1 C. nardus: Macrotermes spp and 1:1 C. nardus: Mineral oil combinations indicated the difference in the average CPT (Table 2).

Table 2.

Effect of fixative/adjuvant on the complete protection time (CPT) against An. gambiae.

Formulation	Fixative/adjuvant (%)	Average protection (mean ± SE)
100% Cymbopogon nardus oil	0	1.80 ± 0.37b
100% Macrotermus spp oil	0	0.60 ± 0.10a
100% Mineral oil	0	0.50 ± 0.00a
50%. C. nardus oil /Macrotermus spp oil	50% TMT	3.50 ± 0.46c
50% Mineral oil (MOL)/ C. nardus oil	50% MOL	1.60 ± 0.88b
20%DEET/Ethanol	80%Ethanol	3.75 ± 0.70c

Different superscript letters in the same column signify significant difference at P < 0.05 (Tukey test).

Analysis of VOCs collected on human skin when applied with the formulated combination

The general analysis of the most effective repellent formulation, i.e. 1:1 C. nardus: Macrotermes spp, portrayed the key volatile organic compounds released from the skin. High retained amounts of esters such as geranyl acetate, geranyl propanoate, were identified with moderate citronellyl acetate, which indicated a heightened 3-fold increase in concentration (Table 3). The absence of these esters in the bare skin, Macrotermes spp oil and presence in C. nardus oils, and the observed increase in 1:1 in C. nardus and Macrotermes spp oil combination indicate that they are crucial for the observed repellence against the three mosquito species.

Table 3.

The amounts of esters formed after the combination.

Compounds	1:1 C. nardus/Macrotermes spp	C. nardus	Macrotermes spp	Bare skin
Citronellyl acetate	✔✔✔	✔✔	✘	✘
Geranyl acetate	✔✔✔	✔	✘	✘
Geranyl propanoate	✔✔✔	✔	✘	✘

The tick (✔) symbol indicates present, while the cross (✘) means absence.

The rate of release of the volatiles over time was also determined. As shown in Table 4; Fig. 5, the volatile abundance varied over time and with the treatment applied. The optimal release of VOCs was recorded in the order: C. nardus: Macrotermes> DEET > C. nardus > Bare skin > Macrotermes (Table 4).

Table 4.

Ranking of volatile release rate over 3 h by the repellent formulations and DEET.

Fig. 5.

Volatile organic compounds released by the formulations over time. “TMT” =Macrotermes spp oil, CTN= Cymbopogon nardus oil.

There was a high abundance of VOCs trapped at t = 0, ranging between 50 and 55 compounds, when essential oils from C. nardus and its combination with Macrotermes were applied. This was contrary to applying 20% DEET, Macrotermes, and bare skin, which revealed < 10 compounds at t = 0. It was also evident that the release rate of volatiles from C. nardus alone was higher, which dropped sharply to two compounds in 2 h. This was contrary to the action of DEET and C. nardus: Macrotermes combination, where some compounds persisted beyond 3 h (Fig. 5).

Discussion and conclusion

In our pursuit to develop improved insect-oil-based formulations of plant-based mosquito repellents, the study characterized the plant essential oils and the insect oils. While both insect and plant oils have value, insect sources generally offer higher extraction yields, which may make them more suitable for applications that require a substantial oil volume. The findings are supported by results obtained by Cheseto et al.³¹. However, differences in chemical composition, volatility, and bioactivity between insect and plant oils must be considered for applications like repellents or pharmaceuticals. The compositional differences observed in the GC-MS profiles of monoterpenes and alcohols dominating in essential oils and fatty acids and sterols in insect oils align with earlier studies^37,38. This suggests that the chemical profiles are characteristic of their respective biological sources. Essential oils from C. nardus, O. kilimandscharicum, and E. citriodora and their combinations have been used as mosquito repellent agents¹¹. The higher levels of fatty acids and sterols in the insect oils from S. gregaria, R. differens, and Macrotermes spp highlight their potential to be used as fixatives due to their high molecular weight and non-toxicity, as suggested by Verheyen et al.³⁹for oils from mealworms. Numerous field tests and lab studies have shown that repellent plant oils can interact synergistically and occasionally antagonistically with other adjuvant oils^40,41. The ecological implications of using essential oils and, particularly, their potential non-target effects have been extensively reported. For instance, the major constituents of C. nardus, O. kilimandscharicum, and E. citriodora, such as citronellal, citronellol, camphor, and 1,8-cineole, undergo rapid photodegradation and microbial breakdown, reducing risks of long-term contamination or bioaccumulation⁴². Nonetheless, their broad-spectrum bioactivity can impose unintended ecological effects on non-target organisms. Studies show that essential oil vapors may deter or impair pollinators such as honeybees and butterflies by disrupting olfactory cues and foraging behavior, potentially affecting pollination services⁴³. Similarly, predatory and parasitic insects contributing to natural pest regulation may be repelled or killed at high exposure levels⁴⁴. Therefore, while these essential oils offer a promising, biodegradable vector control option, ecologically sensitive application practices, including optimized dosing, controlled-release formulations, and habitat-based targeting, are essential to minimize non-target impacts and preserve ecosystem integrity. Nonetheless, the impact of combining essential oils with insect-derived oils on insect repellence has not yet been documented.

Repellence of mosquitoes by the combination of oils from C. nardus and Macrotermes spp., comparable to 20% DEET, may result from the interaction of their dominant chemical components. These interactions could produce esters with odors unfavorable to mosquitoes. Alternatively, the effect may arise from masking the CO₂ released by humans treated with the blend. The possible mechanism involves reduction of citronellal through hydrogenation to a primary alcohol, followed by reaction of this alcohol with a fatty acid carboxyl group (fixative) to form an ester via a nucleophilic acyl substitution process (Fig. 6). DEET and other compounds such as N, N-diethyl phenylacetamide (DEPA), N, N-diethyl benzamide (DEBA), and allethin are used as commercial mosquito repellents⁴⁵. These compounds are all carboxylic acid derivatives classified as amides except allethin, which is an ester. High retained amounts of esters such as geranyl acetate, geranyl propanoate were identified with moderate citronellyl acetate, which indicated a 3-fold increase in concentration. This agrees with the current study, which could be due to the formation of ester compounds, including geranyl acetate geranyl propanoate, and citronellyl acetate (Table 3). The elevated levels of fragrant esters resulting from the combination are key in influencing insect behavior by masking carbon dioxide (CO₂), which usually serves as an attractant for biting insects like mosquitoes.

Fig. 6.

The plausible mechanism for the ester formation.

The CO₂ component is a primary cue for many vectors, signaling the presence of a potential host. The identified mechanism of action would be through the formation of sweet-smelling esters that mask CO₂, which are known mosquito attractants⁴⁶. The increase in aromatic esters formed can also interfere with the perception of host-attractant signals, exciting a receptor for competing behavior^47,48.

However, in addition to repelling or killing the mosquitoes, they could prevent human-mosquito interaction and prevent the transmission of malaria parasites or other vector born disease, ensuring that such protection is long enough to avoid mosquito bites and thereby disease transmission. A previous study found that incorporating citronella oil into a β-cyclodextrin complex for controlled release resulted in a Complete Protection Time (CPT) of only 1.8 h, below the recommended minimum of 2.0 h⁴⁸. To achieve longer-lasting protection, the use of fixatives that enhance the strength and persistence of fragrance compounds has been suggested⁴⁹. In this context, combining plant and insect oils was explored as an alternative strategy for developing repellent systems with controlled release of volatile organic compounds, addressing the limited long-term effectiveness of existing personal mosquito protection products¹².

In this study, the insect oils used as fixative, especially oils from Macrotermes spp, amplified the duration of the fragrance from 1.8 to 3.75 h, thus making it a potent candidate for the formulation of plant-based mosquito repellents. An essential characteristic of a fixative is that it must be safe for human skin. This study aligns with earlier research indicating that oils from certain edible insects, such as Schistocerca gregaria (desert locust) and Ruspolia differens (nsenene), are non-toxic and contain high levels of essential omega-3 fatty acids and antioxidants like vitamin E and flavonoids³¹. Available evidence suggests that purified insect oils are non-irritant and biocompatible, though formal dermal toxicity and sensitization testing remain necessary before use as scaled adjuvants in mosquito repellents.The effect of 1:1 combination of C. nardus: Macrotermes spp against the three mosquito species revealed a varied species-specific repellence, of which An. gambiae was most effectively repelled (Fig. 3). Our results highlight the need for future investigations to elucidate the reasons for these differences, specifically at the antennal signaling level.

From the obtained results, synergism was observed in terms of the protection time when insect oils and plant essential oils were combined. The repellence efficiency of the individual plant or insect oils was found to be lower than compared of the combinations. Citronella, a well-known mosquito repellent⁵⁰, showed an increased CPT in the presence of oils from Macrotermes spp as a fixative. The results indicated that a 1:1 combination of C. nardus: Macrocentrus spp had a repellence of 93.4% at 4 h and CPT of 210 min when tested against An. gambiae, which closely followed 20% DEET that portrayed > 90% repellence for 4 h with a CPT of over 270 min, highlighting the potential of oils from Macrotermes spp. as safe adjuvants. The low CPT for DEET, as compared to other studies, could suggest that repeated exposure may lead certain mosquito populations to develop behavioral or sensory tolerance to DEET⁵¹. This combination of essential oils with insect oils increased the protection time by a two fold factor. The observed synergism is comparable to the interaction between a repellent and a non-pyrethroid insecticide that was reported by Pennetier et al.⁵². Thus, combining plant and insect oils offers a more potent alternative for improved vector control in cases where disease vectors are becoming resistant.

The observed fixative was deciphered by analyzing the volatile organic compounds released from human skin over time after applying the individual oils and their different combinations. The trapped volatiles from both untreated skin and skin treated with various combinations of C. nardus and Macrotermes spp. oils were analyzed to identify the olfactory signals that influence the host-seeking behavior of mosquito vectors responsible for transmitting major diseases such as malaria. When bare skin was examined, 24 volatile compounds were found, 14 when Macrotermes spp oils were applied, 100 when C. nardus oil was applied, and 120 when a 1:1 combination of C. nardus/Macrotermes spp oils was applied. The presence of carboxylic acids, aldehydes, alcohols, and ketones dominated the bare skin volatile profile, as previously reported⁵³. The findings align with reports indicating that the number of bare skin compounds is fewer, around 40 compounds⁵⁴. Citronella oil-related compounds included geraniol, limonene, terpene-ol, sabinene, pinene, and cymene, and were second in order of dominance⁵⁵. The observed differences in the quantities of trapped volatile organic compounds were attributed to variations in the air permeability of the repellents at typical room temperatures. Additionally, the emission patterns of VOCs from bare skin vary between cold and warm seasons, as noted by Zhang and Ruan⁵⁶. Since the average room temperature in this experiment was controlled at 24 °C, the repellent formulations could be ranked according to how quickly volatility decreased in this case, by the end of 4 h, as C. nardus/Macrotermes spp oils > DEET > C. nardus> Bare SKIN> Macrotermes spp. This suggests that a 1:1 combination of C. nardus:Macrotermes spp oils could be a viable repellent organic compound carrier, with a protection time that is comparable to 20% DEET during the first 4 h of exposure, although DEET typically offers longer protection (8–10 h)¹⁸.

Upon evaluation at various time intervals, it was observed that certain formulations exhibited greater volatility during the initial hours and subsequently demonstrated a moderate yet consistent volatile release. Citronella oil was no exception, releasing a significant number of volatiles in the first hour but significantly lessening later. This could explain its low repellence beyond one 1 h as compared to when combined with termite oil. Throughout the test time, DEET, the positive control, released volatiles steadily and continuously. Combining termite oil with Citronella oil released volatiles sharply for the first hour, followed by controlled volatile release over an extended duration, as observed with DEET. After 3 h, the volatile release rate of the citronella–termite oil combination was slightly higher than that of DEET (Fig. 5), indicating greater stability, which may be attributed to ester formation.

The analysis revealed a diverse array of 143 compounds identified from insect and plant oils, highlighting the complexity of their chemical profiles. This composition includes various terpenes (43.4%) and fatty acids (31.5%), suggesting potential biochemical roles for these compounds. The presence of different functional groups, such as sterols (4.9%), alcohol (4.2%), and aldehydes (1.4%), further underscores the potential for these oils to be explored for their biological and pharmacological applications. Repellent activity against An. gambiae was observed in both plant essential and insect oils. However, there was enhanced repellence when essential and insect oils were used in combination, especially at 1:1 ratios. Therefore, the established plant-insect-based formulation offers a promising avenue for enhancing the stability and effectiveness of repellent products⁵⁷. This combination extended the duration of protection, both through controlled volatile release and the formation of ester compounds that can offer a sweet aroma. The aroma can confuse or deter these insects and can enhance user acceptance. The findings pave the way for creating safer, more effective plant-based mosquito repellent formulations as alternatives to synthetic products. This compelling interaction boosts both the effectiveness and duration of plant-based repellents, offering potential for their inclusion in vector control strategies.

Materials and methods

Insect oils

Edible insects, Schistocerca gregaria (desert locust) were mass-reared at the International Centre of Insect Physiology and Ecology (icipe), Nairobi, Kenya (1.2219° S, 36.8967° E; 1600 m a.s.l.). The locusts were fed on wheat seedlings, maize leaves, and wheat bran. Upon reaching the adult stage, they were harvested for processing. Samples of sun-dried winged termites (Macrotermes spp.) were purchased from local markets in Bungoma, Kenya (0.5695° N, 34.5584° E), while Ruspolia differens (nsenene) were obtained from markets in Uganda. All insect samples were oven dried and processed by cold pressing, a preferred method for safer and sustainable oil extraction. Dried insects were finely ground and pressed mechanically using a screw-type expeller at ambient temperature without solvent addition. The crude oils were filtered through Whatman No. 1 paper to remove debris and stored in amber vials at 4 °C. Oil yield (%) was determined as the weight of oil extracted relative to the dry sample weight³¹. Each extraction was performed in triplicate using independent sample batches to ensure reproducibility and enable statistical analysis.

Plant collection and extraction of essential oils

The aerial parts of O. kilimandscharicum and C. nardus were collected in July 2022 from Isecheno Village, Kakamega County, Western Kenya (0°17′ N; 34⁰ 45′ E), while E. citriodora was collected in April 2022 in Karen, Nairobi County (S 1° 19’ 15.0708”, E 36° 41’ 5.7696”). The sample specimens were identified by a plant Taxonomist, Mr. Simon Mathenge (posthumous). The voucher specimens OK/KAK/01/22, and OK/KAK/02/22 and EC/NRB/01/22 of the plants were deposited at the Herbarium of the National Museums of Kenya. Essential oils were extracted by hydro-distillation from dried aerial parts (500 g) of O. kilimandscharicum, E. citriodora, and C. nardus using a Clevenger apparatus. Plant material was distilled in a 1:3 (w/v) ratio with distilled water for 4 h. The essential oils were collected from the water–hexane layer using GC-grade hexane (≥ 99.9%, Sigma-Aldrich, St. Louis, USA) for separation, then dried over anhydrous sodium sulfate. The solvent was removed in vacuo, and the oil yield was recorded. Each extraction was performed in triplicate using independent plant batches to ensure reproducibility and statistical validity. The oils were stored in amber glass vials at 4 °C until analysis.

Chemical analysis of insect oils and plant essential oils using GC-MS

Plant essential oils were diluted by dissolving 1 µL of each sample in 1 mL of GC-grade dichloromethane (≥ 99.8%) (Sigma-Aldrich, St. Louis, MO, USA) and analyzed in triplicate using gas chromatography–mass spectrometry (GC–MS) in splitless mode. Fatty acid content in insect oils (20 mg per sample) was analyzed as fatty acid methyl esters (FAMEs) following the protocol of Cheseto et al. (2020)³¹. Each sample was treated with 500 µL of sodium methoxide solution (15 mg mL⁻¹ in dry methanol), vortexed for 1 min, sonicated for 5 min, and incubated at 60 °C for 1 h. The reaction was quenched with 100 µL of deionized water, vortexed, and FAMEs were extracted with 1000 µL of GC-grade hexane (≥ 99.8%, Sigma-Aldrich, St. Louis, MO, USA). Extracts were centrifuged at 14,000 rpm for 5 min, and the supernatant was dried over anhydrous sodium sulfate. A 1.0 µL aliquot was injected for GC–MS analysis.

The GC-MS system consisted of a 7890 A gas chromatograph coupled with a 5975 C mass selective detector (Agilent Technologies, USA) and used a (5%-phenyl)-methylpolysiloxane capillary column (HP5 MS, 30 m × 0.25 mm i.d., 0.25 μm). Helium served as the carrier gas at 1.25 mL/min. The oven temperature was programmed to range from 35 °C to 285 °C: initially held at 35 °C for 5 min, then ramped at 10 °C/min to 280 °C and maintained at this temperature for 20.4 min. The ion source and quadrupole temperatures were maintained at 230 °C and 180 °C, respectively. Mass spectra were acquired using electron impact (EI) ionization at 70 eV, with fragment ions scanned across 40–550 m/z in full-scan mode. Filament delay was set to 3.3 min.

Compound identification was based on comparison with MS reference libraries, including Chemecol, NIST (05, 08, 11), and ADAMS. For relative quantification, serial dilutions (2.25–1000 ng/µL) of an external standard of α-pinene and α-humulene (98% purity) were analyzed using the same GC-MS parameters. Calibration curves were generated using peak areas against the concentrations:

1

2

These equations were used to quantify compounds with retention times below Eq. (1) or above 16 min Eq. (2), respectively. Retention indices (RI) were calculated using the Van den Dool and Kratz method with C₅–C₃₁ alkanes and compared with published values^58,59. Select compounds were further confirmed by comparing retention times with authentic standards under identical GC-MS conditions.

Mosquito rearing

Aedes aegypti, Anopheles gambiae s.s. (Ifakara strain from Tanzania), and Culex quinquefasciatus mosquitoes were reared under controlled conditions in the insectary at the International Centre of Insect Physiology and Ecology (icipe), Duduville, Nairobi, Kenya. Eggs were incubated in water at 27 °C with 80% relative humidity, maintained under a 12-h light/dark cycle. During development, larvae were transferred as needed using an eyedropper. Mosquitoes were reared using Tetramin^® fish food (Tetra GmbH, Melle, Germany). Eggs were submerged in water to facilitate hatching, after which the larvae were separated from any unhatched eggs and transferred to fresh distilled water. Once pupation started, pupae were isolated from the remaining larvae and placed in clean water within rearing cages to develop into adults. Adult female mosquitoes (5–7 days old) were identified morphologically and collected using an aspirator, and placed in a separate cage, where they were sustained on a 10% sucrose solution until experimentation. Before testing, 5–7-day-old blood-feeding naïve females were starved for 18 h, following an initial feeding period on a 6% glucose solution.

Mosquito repellent test

The repellent effectiveness of essential oils, insect oils, and their combinations was evaluated against Anopheles gambiae, Aedes aegypti, and Culex quinquefasciatus. Preliminary screening of the essential oils and formulations for repellency was carried out at concentrations of 10⁻⁵, 10⁻³, and 10⁻¹ g/mL, following the WHO arm-in-cage protocol⁶⁰. Eight malaria-free human volunteers confirmed through screening and found to have little to no sensitivity to mosquito bites or the test oils, provided informed consent, and were selected for the study. They were advised to refrain from using lotions, perfumes, oils, or scented soaps for 24 h before the experiments. Sixteen 50 × 50 × 50 cm cages, each containing 100 starved female mosquitoes, were prepared. Insect and plant oils were standardized before blending based on equal mass (w/w) to ensure consistent formulation ratios. A 1 mL test solution (containing 0.4 mL of the mixture in 1 mL of ethanol) was applied to each volunteer’s forearm, covering the area from wrist to elbow, while the rest of the hand was shielded with a glove. As a positive control, 20% DEET in the same volume was applied to the opposite forearm of the same volunteer; interchangeably, light white mineral oil (Rajol WT 80, India) served as a negative control. Assays were done during daytime under red light to mimic darkness and reproduce mosquitoes’ natural nocturnal biting behavior. After the mosquitoes were released, the untreated (control) arm was inserted into the cage for 3 min, during which the number of mosquitoes that landed on the arm was recorded. The treated arm was then exposed under identical conditions, and landings were again counted. Each test sample was evaluated at progressively increasing concentrations, starting from the lowest dose. All repellence assays were conducted with at least three independent biological replicates, each comprising three technical repeats per treatment. Repellence (R) at each concentration was determined using data from eight replicates and calculated according to Eq. (3), following the method outlined by Fradin and Day⁶¹

3

where C represents the number of mosquito bites on the untreated (control) arm, while T denotes the number of bites on the treated arm.

Complete Protection Time (CPT) was defined as the duration until the first mosquito landed on or bit the treated arm. To measure CPT, each of the 8 volunteer’s treated right arms was placed into the test cage for 3 min. If no bites were observed, the arm was reintroduced at 30-minute intervals until the first bite occurred⁴⁸.

Determination of volatile organic compounds (VOCs) from the skin

Ex-situ collection of exposed hand part volatiles

A simultaneous experiment was carried out to evaluate the levels of volatile organic compounds (VOCs) emitted from untreated skin and skin treated with different oil formulations, tracking their variation over time. The collection of volatiles was performed ex-situ alongside the behavioral assays. For each sample, sterilized medical-grade cotton wool pieces (0.2 g) were prepared and placed in 15 ml Falcon tubes using sterilized forceps, ready for skin swabbing. A 3 cm × 3 cm template was used to define several swabbing areas on both hands of the volunteer, ensuring consistency across samples. Oil formulations were prepared in different concentrations in 50 mL vials. Initially, baseline (blank or bare skin) VOCs were collected by swabbing the marked areas with clean cotton pieces. Following this, 1,000 µL of the designated oil formulation, consisting of 400 µL of sample oils (mixed 1:1) diluted to 1,000 µL with ethanol, was evenly applied to the skin. A second swab was then immediately collected from the same marked area. Additional swabs were collected at 30-minute intervals at separate delineated spots to track VOC variation over time for 4 h. The headspace volatiles captured on cotton wool swabs were transferred into a 13 mL autosampler glass vial (Shimadzu, Kyoto, Japan). Retention of volatiles was assessed by collecting headspace samples followed by GC–MS analysis.

Chemical analysis of swabbed VOCs body samples

The cotton swabs containing headspace volatiles were placed in a 250 mL Quick-Fit glass chamber (Agricultural Research Service, Gainesville, FL, USA). Activated charcoal-filtered and humidified air was passed over the samples at 340 mL min⁻¹ using a push-pull Gast pump (Gast Manufacturing Inc., Benton Harbor, MI, USA). Volatiles were collected on pre-cleaned Porapak Q traps (30 mg, polydivinylbenzene resin, 80–100 mesh; Supelco, Bellefonte, PA, USA) and subsequently eluted with GC-grade dichloromethane (DCM). The outlet flow rate was maintained at 170 mL min⁻¹ using a Vacuubrand CVC2 vacuum pump (Wertheim, Germany) for 4 h. The trapped compounds were then eluted with 200 µL of DCM into 2 mL glass vials containing 250 µL glass inserts (Supelco, Bellefonte, PA, USA) and analyzed immediately by GC-MS. Volatile collection was performed in triplicate across multiple batches of oil mixtures and untreated skin. Quantification of headspace volatile compounds was achieved using calibration curves of α-pinene and α-humulene (purity > 98%, Sigma-Aldrich^® Solutions, St. Louis, MO) prepared in concentration ranges between 2.25 and 1000 ɳg/µl. The Eq. (1) from α-pinene was used to semi-quantify compounds with retention times below 16.0 min, while the Eq. (2) for α-humulene semi-quantified compounds with retention times equal to or above 16.0 min^62,63 The measured concentrations were converted to ng of VOC emitted per h (ng/VOC/h) following the method of Miano et al.⁶³. Compound identification was based on comparison with MS reference libraries, including Chemecol, NIST (05, 08, 11)⁶⁴ and ADAMS (2005)⁵⁸. For relative quantification, serial dilutions (2.25–1000 ng/µL) of an external standard of α-pinene and α-humulene (98% purity) were analyzed using the same GC-MS parameters. Calibration curves were generated using peak areas against the concentrations.

Statistical analysis

Data were expressed as mean ± standard error (SE). Differences among treatments (CPTs, individual oils, and DEET) were analyzed using one-way analysis of variance (ANOVA), followed by the Student–Newman–Keuls (SNK) post hoc test for mean separation at p < 0.05. Data analysis was conducted using R software version 4.3.2⁶⁵. and PAST (Paleontological Statistics) version 3.12⁶⁶. Significant differences were indicated by different letters, which represent pairwise p-value groupings from the Tukey test. Protective efficacy (PE) values were calculated from four replicates, while repellence dose (RD₅₀) values were determined via probit analysis^67,68. Results were visualized on non-metric multidimensional scaling (NMDS), using Bray–Curtis by PAST software (version 4.06b).

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

BT, CT, SK, JBO: conceived and designed the project. JBO, GO, conducted the laboratory experiments and bioassays, GO, XC, HMO, CMM, JBO: acquired and analyzed GC-MS data. XC, HMO, GO, CMM, and JBO: carried out data analysis. JBO, CMM, HMO, XC, SS, and CT: contributed to the writing of the original draft. BT, CT, SK supervised the work. CT, SK, BT, SS: provided financial resources. All authors reviewed the article and approved the submitted version.

Data availability

All data generated or analysed during this study are included in this published article [and its supplementary information files].

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

All methods were performed in accordance with the relevant guidelines and regulations. This study was approved by the Kenya Medical Research Institute, KEMRI SERU Secretariat (approval no KEMRI/RD/22).