Introduction

The western honey bee, Apis mellifera L. (Hymenoptera: Apidae), is the most important pollinator of a wide variety of agricultural crops and the main source of bee products such as honey, propolis, pollen and royal jelly (Abd El-Wahab et al. 2021; Flores et al. 2021). In recent decades, dramatic honey bee colony losses are recorded to be linked mainly to several biotic and abiotic stressors like habitat loss, pesticides exposure, poor nutrition, beekeeping practices, climatic changes and invasive pathogens (Glenny et al. 2017; Almecija et al. 2020). Amongst its parasites, the cosmopolitan ectoparasitic mesostigmatid mite Varroa destructor Anderson and Trueman (Acari: Varroidae) is marked as the major pathological threat (Giacobino et al. 2014; Erban et al. 2015; Alburaki et al. 2018). This mite affects directly both adult bees and brood by feeding on their hemolymph and fat body tissue (Ramsey et al. 2019; Slowinska et al. 2019; Traynor et al. 2020; Piou et al. 2023) and acts as a viral vector, which leads to the colony’s collapse within 2–3 years (Mondet et al. 2020). Besides, on an individual scale, varroosis induces very harmful effects getting to malformation, dehydration of the body and weight drop, lifespan decrease, olfactory sensation problem, immunosuppression as well as alteration of flying and foraging ability (Noël et al. 2020; Sabahi et al. 2020; Khan and Ghramh 2021).

As a consequence of these dramatic effects and losses, various synthetic acaricides have been largely used to fight Varroa mite such as pyrethroids (tau-fluvalinate and flumethrin), organophosphate (coumaphos), and formamidine (amitraz) (Li et al. 2017; Riva et al. 2019). Nevertheless, it should be mentioned that their high efficacy is faded by V. destructor resistance development, the contamination of beehive products by chemical residues of these acaricides, as well as detrimental effects on honey bee health and the environment (Riva et al. 2019; Noël et al. 2020). Recently, natural compounds, especially organic acids, herbal essential oils (EOs) and derivatives have attracted a lot of attention as potentially safe, eco-friendly and efficient alternatives to conventional acaricides (Vilarem et al. 2021). Indeed, EOs are believed to have a broad-spectrum efficacy with ovicidal, larvicidal, adulticidal and ectoparasite-repellent effects (Abbas et al. 2018). The application of EOs is reported to trigger disorders in insect and mite activities, which include antifeedant activity, growth and reproduction deficiency, breakage of cuticle as well as exuviation and respiration repression (Ben Chaaban et al. 2019). Ramzi et al. (2017) found that certain EOs tend to enhance the honey bee grooming behavior. Notably, thymol is a good example of a bioacaricide that is used by beekeepers to treat honeybee colonies infested by V. destructor known for its mite antifeedant and repression of reproduction and growth activities (Lindberg et al. 2000; Singh 2014).

As a source of herbal EOs, several plants have been used to extract efficient and active organic compounds worldwide and in Algeria as well. Among these Algerian plants, we can distinguish the genera Artemisia (Asteraceae) and Ruta (Rutaceae). Artemisia is reportedly comprising > 500 diverse species, most of which have been recognized as an important source of EOs (Hu et al. 2019). Overall, Artemisia species are characterized by multiple biological properties including antibacterial, antifungal, antiviral (Abad et al. 2012), antiprotozoal, anthelmintic, anti-inflammatory, cytotoxic, antitumor, insecticidal and repellent activities (Ivănescu et al. 2021). Godara et al. (2013, 2015) reported that extracts from Artemisia absinthium exhibit acaricidal activity against the ticks Rhipicephalus sanguineus and Hyalomma anatolicum. Furthermore, a high mortality rate was recorded by the use of Artemisia sieberi extract to control V. destructor (Ashrafi Parchin et al. 2012). Field trials of extracts from Artemisia annua and A. absinthium against V. destructor revealed a potent miticidal effect (Rasool et al. 2017; Allabergenova et al. 2021). EOs extracted from A. annua and Artemisia verlotiorum revealed a significant repellent activity against Varroa under laboratory conditions (Conti et al. 2020).

The genus Ruta comprises 40 species, highly rich in EOs that contain mainly aliphatic ketones, e.g., 2-undecanone and 2-nonanone (Nahar et al. 2021). Ruta spp. have been widely identified as antiparasitic, stomachic, digestive, vermifuge and molluscicidal (Ferhat et al. 2014), antiseptic, antispasmodic, anthelmintic (Haddouchi et al. 2013), insecticidal and larvicidal agents (Boutoumi et al. 2009; Majdoub et al. 2014). Castagnino Laércio and Orsi (2012) demonstrated that Ruta graveolens volatile oil is effective to reduce the honey bee colonies infestation by V. destructor and decrease the mortality rate of parasitized honey bees.

In this study, we assessed the in vitro acute toxicity of EOs obtained from four Algerian plants – i.e., Artemisia herba alba Asso, Artemisia campestris L., Artemisia judaica L. and Ruta montana L. – against V. destructor. In addition, the determination of the phytochemical profile of each EO using gas chromatography–mass spectrometry (GC-MS) will help to identify their major compounds that might contribute to the pronounced acaricidal effect. To the best of our knowledge, this is the first work on the acaricidal activity of EOs from A. campestris, A. judaica and R. montana in the Varroa mite control, whereas EO isolated from A. herba alba from a different region has recently been tested in vitro by Alahyane et al. (2022).

Materials and methods

Plant material

Aerial parts including leaves, flowers and stems of A. herba alba, A. campestris, A. judaica and R. montana were harvested in Algeria during the flowering period (Table 1). The taxonomic identification was confirmed by Prof. Benhouhou Salima, Department of Botany, Higher National Agronomic School of Algiers (Algeria) according to the Quezel and Santa (1969) nomenclature and validated by the synonymic index of the North Africa plants. The aerial parts were dried in the dark at room temperature.

Table 1 Description of plants used in this study
Full size table

A sample of 500 g dried plant material was ground just before extraction with an electrical grinder (Moulinex AR110830) then submitted for hydrodistillation using a Clevenger-modified system for 3–4 h until total recovery of the oil. The latter was then dried over anhydrous sodium sulphate before being stored in a sealed dark glass vial and maintained at 4 °C until GC analysis and acaricidal assays. The extraction yield was expressed as a percent of the ratio between EO mass (g) per dried plant material mass (g). Experiments were performed in triplicate and the yield mean value was reported.

Gas chromatography–mass spectrometry (GC-MS) analysis

Identification of EO composition has been performed using a GC-MS system (Agilent HP 6890 Plus gas chromatograph coupled with a Agilent HP 5973 mass spectrometer). The column used was a capillary HP-5MS (5% phenyl 95% dimethylpolysiloxane) column (30 m length, 0.25 mm internal diameter and 0.25 μm film thickness; Agilent Technologies, USA). A sample of 0.2 µl was injected in split mode (ratio 1/80) at which the GC injector temperature was set at 250 °C. Helium (He) was used as a carrier gas at 0.5 mL/min and the oven temperature program started at 60 °C, where it was held isothermally for 8 min before being increased to 250 °C (at 2 °C/ min) and held isothermally for 10 min. The ionization energy was set at 70 eV and a solvent delay of 3.5 min. The identification was achieved by matching the recorded mass spectra with those stored in Wiley and NIST library data and by comparing each compound’s retention index relative to n-alkanes (C8–C29) with those in the literature (Adams 2017).

Mites and bees

Apis mellifera intermissa (belonging to the African Lineage A) workers and pupae, and adult females of V. destructor mites collected during the spring and summer of 2022, when mites were abundant, from hives placed in an apiary and kindly donated by the Multipurpose Agricultural Cooperative of Tizi-Ouzou (CAPTO) situated in the region of Tizi-Rached, province of Tizi-Ouzou in Algeria (36°40’42.5316’’N, 4°12’28.9512’’E). The same management practices were applied to all experimental bee colonies that were kept untreated against V. destructor for at least 6 months. Sealed brood worker or drone combs were removed from colonies at the apiary then moved to the laboratory. Varroa destructor adult females were collected in their reproductive stage instead of their dispersal stage in order to avoid mites’ variability in physiology and fitness. Brood cells were individually uncapped using fine forceps and inspected for adult mite’s presence. Mites were picked up from larvae and non-pigmented pupae using a fine paintbrush as described previously by Dietemann et al. (2013). Collected mites were kept in Petri dishes with some bee pupae (white to pink eyes) as food substrate, to prevent mite’s desiccation, and used immediately for bioassays. Mites that appeared to be newly molted, weak or abnormal were purposely discarded from selection. This protocol has been precisely chosen because it allows the collection of less traumatized mites than those obtained with other methods (Dietemann et al. 2013). Newly emerged worker honey bees were selected from healthy bee colonies (the younger population is less exposed to hive and environmental conditions than the older one). Frames of sealed worker brood showing signs of emerging bees within 1–2 days were transported into ventilated boxes to the laboratory incubator (34 °C and 60 ± 10% RH). Once young bees have emerged, they were gently inspected with plastic tweezers to confirm the absence of mites before being distributed into plastic rearing cup cages supplied with a 50% (w/v) sucrose feeding solution.

Acaricidal activity on Varroa destructor

Essential oil bioassays against V. destructor were conducted according to a modified protocol previously described by Ruffinengo et al. (2005). Briefly, six EO concentrations ranging from 0.25 to 20 µL/mL were prepared in ethanol 96%, then 1 mL of each solution was uniformly spread using a micropipette onto a filter paper that covered the full bottom surface of a Petri dish (6 cm diameter, 2 cm high). After solvent evaporation, 10 mites and five bee pupae (white to pink eyes) were placed in each Petri dish on the surface of the filter paper favoring direct contact with the solution. The dishes were sealed with parafilm and covered with a lid to prevent mite escape. The experiment was carried out at room temperature. Three replicates were realized for each treatment and dose, and the mortality percentage was expressed as a mean value. Ethanol 96% was used as the negative control and thymol as the positive control using six concentrations ranging from 0.21 to 2.06 µL/mL. All treated mites were incubated at 30 ± 1 °C and 60 ± 10% RH for 72 h. Mortality of mites was checked at 24, 48 and 72 h following the initial exposure. Mites that did not move when probed with a fine paintbrush were considered dead.

Toxicity of essential oils on Apis mellifera adults

The EOs toxicity to worker bees was tested using Chaimanee et al. (2021) protocol that was partly modified. Six EO concentrations ranging from 2.5 to 40 µL/mL were prepared. Then, 1 mL solution was pipetted onto filter paper covering the bottom of the plastic cup cages described by Evans et al. (2009). Ten newly emerged adult bees were placed into each cage after solvent evaporation. A 5-mL syringe filled with 50% (w/v) sucrose solution was supplied ad libitum on the top of the cage whereas a small amount of beebread (a mixture of polyfloral bee pollen in 50% sucrose solution) was supplied as a dietary supplement at the bottom of the cage as described by Huang et al. (2014). Ethanol and thymol with six concentrations ranging from 1.03 to 3.09 µL/mL were included as negative and positive control, respectively. Three replicates were conducted for each treatment for a total of 30 bees. The cages were placed under the same environmental conditions as those used for the acaricidal activity against V. destructor. Mortality was checked by visual inspection at 24, 48 and 72 h following initial exposure. Mortality data, at the various dosages, is used as a measure of the respective medium lethal concentrations (LC50, i.e., the lethal dose that causes the death of half a population) whereas selectivity ratio (SR) is used to compare efficacy among tested EOs. SR is expressed as a ratio between mite and bee toxicity:

SR = LC50 A. mellifera/LC50 V. destructor.

Statistical analysis

Bee and mite mortality rates were corrected for mortality against the control using Abbott’s (1925) formula: Pc = 100 × (P0 – Pt) / (100 – Pt), where Pc is the corrected mortality (%), Pt is the mortality observed in the control, and P0 is the mortality observed in the test. Differences between groups were analyzed by two-way ANOVA and Tukey’s honestly significant difference (HSD) test for the four concentrations in common (1–10 µL/mL) for V. destructor mortality. The average corrected mortality data were subjected to a probit analysis for calculating the LC50 at 95% confidence intervals (CI); normal distribution of the data of the six concentrations tested was verified using the Shapiro-Wilk test. Statistical analysis was conducted using IBM-SPSS v.28.0 (IBM, Armonk, NY, USA).

In order to establish a dose-response relationship and to estimate the highest concentration of EOs at which it did not induce bee mortality significantly higher than that observed in controls (NOAEL, no observed adverse effect level; Medrzycki et al. 2013), a two-way ANOVA followed by Tukey’s post-hoc test were applied. Then, it was plotted in Graph Pad Prism v.6.01 for Windows.

For all tests the significance threshold (α) was 0.05. NOAEL was used as part of the most important steps in risk assessment methodologies.

Results

Chemical composition of essential oils

Essential oil yield was highly variable. The highest yield was recorded in A. herba alba (mean ± SD = 2.20 ± 0.09%), followed by A. judaica (0.84 ± 0.02%), R. montana (0.67 ± 0.06%) and A. campestris (0.59 ± 0.04%) on a dry-weight basis. Volatile compound identification revealed 29 compounds representing 99.9% of the total A. campestris EO (Table 2). Monoterpene hydrocarbons represented the major fraction including β-pinene (19.5%), p-cimene (8.1%), α-pinene (7.1%) and δ-terpinene (6.0%) as the major compounds. In parallel, we noted that ketones were the predominant fraction in EOs from A. herba alba, A. judaica and R. montana of which davanone (66.9%), piperitone (68.7%) and 2-undecanone (70.1%) were the chief constituents (Table 2).

Table 2 Chemical composition of the essential oils of Artemisia campestris, A. herba alba, A. judaica, and Ruta montana
Full size table

Acaricidal activity of essential oils on Varroa destructor

Artemisia herba Alba essential oil

Figure 1A shows V. destructor corrected mortality rates as a function of time at different concentrations of the tested EO. Mite mortality in the control was satisfactory in the range of 0-6.7%, less than the 10% threshold to reject the test. ANOVA indicates that EO concentration (F3,36 = 92.7, P < 0.001) and exposure time (F2,36 = 3.7, P < 0.05) have a strong effect on mite mortality, but not their interaction (F6,36 = 0.7, P > 0.05). Mortality rate after 24 h of the application of 2.5 µL/mL (65.51 ± 21.53%) was higher than that obtained after 72 h of exposure for 1 µL/mL (25 ± 10.71%). Complete mortality (100%) was registered for high concentrations, such as 5 µL/mL for 72 h and 10 µL/mL irrespective of exposure time – significantly higher than in the treatments (Fig. 1A).

Fig. 1
figure 1

Mean (+ SE) corrected mortality rates of Varroa destructor as a function of time and of the concentration of essential oils of (A) Artemisia herba alba, (B) A. campestris, (C) A. judaica, and (D) Ruta montana. Means within a panel capped with different letters are significantly different (Tukey’s HSD test: P < 0.05)

Full size image

Artemisia campestris essential oil

Mite mortality was affected by the A. campestris EO concentration (F3,36 = 63.6, P < 0.001) and exposure time (F2,36 = 6.5, P < 0.05), but not by their interaction (F6,36 = 1.4, P > 0.05) (Fig. 1B). For 5 µL/mL, the mortality reached 71.43 ± 6.18% after 72 h of exposure, much higher than that reached after 24 h of exposure (27.59 ± 17.92%). A high mortality rate of 96.42 ± 06.19% was reached after 48 h of exposure to 10 µL/mL which differed significantly from that of the other treatments (Fig. 1B).

Artemisia Judaica essential oil

The developmental mortality rate of the mite is concentration dependent (F3,36 = 40.1, P < 0.001) but is not affected by exposure time (F2,36 = 2.9, P > 0.05), nor by the interaction of time and concentration (F6,36 = 0.23, P > 0.05) (Fig. 1C). High mortality rates were recorded for all the tested concentrations. For each concentration, mortality rates at 48 and 72 h were not different compared to that after 24 h post-treatment. Mortality rate for 10 µL/mL was significantly higher than that of 1 µL/mL irrespective of exposure time, ranging from 93.10 ± 11.94 to 96.42 ± 06.19% and from 27.59 ± 10.34 to 39.28 ± 16.36%, respectively (Fig. 1C).

Ruta Montana essential oil

The V. destructor mortality rate is influenced by the concentration of R. montana EO (F3,36 = 312.5, P < 0.001), the exposure time (F2,36 = 23.0, P < 0.001) as well as by their interaction (F6,36 = 5.5, P < 0.05) (Fig. 1D). The mortality recorded increased with an increase in concentration and in time of exposure. Clearly, 10 µL/mL had the highest mortality rate which differed significantly from that of the other concentrations for each exposure time (Fig. 1D).

Apis mellifera adults: toxicity evaluation

Mortality rates of adult honey bees exposed for 24, 48 and 72 h to a range of concentrations of the four essential oils were determined (Fig. 2). In addition, NOAEL, the highest concentration at which honey bee mortality is not significantly different from the control group, was determined for each EO and thymol after 72 h post treatment (Table 3). In parallel, V. destructor mortality rates recorded for these NOAEL values were indicated and were apparently important ranging from 96.7 to 100% (Table 3). The solvent control treatment caused negligible mortality of bees (0–3.3%). The NOAEL of A. campestris EO was the highest with a value of 20 µL/mL among the five tested products, indicating that it had the lowest toxicity to honey bees. On that account, all volatile oils tested had a satisfactory acaricidal effect and an acceptable toxicity to bees.

Fig. 2
figure 2

Mean (+ SD) mortality (%) of adult honey bees exposed to a range of concentrations of essential oils of (A) Artemisia herba alba, (B) A. campestris, (C) A. judaica, and (D) Ruta montana, and (E) thymol (positive control). Asterisks indicate significant differences compared to the resp. negative control (i.e., ethanol) (Tukey’s HSD test: ****p < 0.0001, ***0.0001 < p < 0.001, **0.001 < p < 0.01, *0.01 < p < 0.05)

Full size image
Table 3 Estimated ‘no observed adverse effect level’ (NOAEL) and mean (± SD) Varroa mortality rates (%) after 72 h of exposure
Full size table

Determination of lethal concentrations and selectivity indexes

The LC50 values for EOs against V. destructor and newly-emerged worker bees were calculated based on the corrected rates of mortality using probit analysis (Table 4). The greatest toxicity of the EOs to Varroa mites was observed for A. herba alba and A. judaica followed by A. campestris and R. montana after 24 h (LC50 = 2.025–5.725 µL/mL), 48 h (LC50 = 1.538–4.92 µL/mL) and 72 h (LC50 = 1.347–3.549 µL/mL), but less toxic than the positive control (LC50 = 0.447–0.375 µL/mL). Note that all tested EOs had a very low toxicity to honey bees, they were even less toxic to both V. destructor mites and newly-emerged honey bees than the positive control (Table 4).

Table 4 LC50 (µL/mL) values (+ 95% CI in parentheses) of essential oils and thymol (positive control) on Varroa destructor and Apis mellifera and their selectivity ratio at 24, 48 and 72 h using complete exposure method under laboratory conditions
Full size table

By far, the highest selectivity ratio values were reached at 72 h post-treatment by A. campestris EO (SR = 10.77) followed by A. judaica EO (SR = 8.78) whereas moderate values were recorded for A. herba alba EO (SR = 5.62) and R. montana EO (SR = 3.73). Essential oils appear to be more selective than thymol (positive control), which presented the lowest SR values of 3.42 (24 h), 3.64 (48 h) and 3.65 (72 h).

Discussion

Artemisia herba alba EO yield (2.2%) appears to be in the same range as results from Spain (0.41–2.30%; Salido et al. 2004) but much higher than those reported in other studies of the MENA (Middle East and North Africa) region (0.33–1.18%; Dob and Benabdelkader 2006; Hudaib and Aburjai 2006; Lakehal et al. 2017; Amor et al. 2019; Aimad et al. 2022; Kadri et al. 2022). The literature indicated high variability of EO yields for A. campestris from Algeria where our yield of 0.59% is similar to that obtained by Cheraif et al. (2020) from Laghouat, Algeria (0.52%), much higher than that obtained from Djelfa, Algeria region (0.1%) (Dob et al. 2005), but much lower compared to Boutemak et al. (2017) data (0.88%).

Artemisia judaica EO yield (0.84%) was higher than that given by both Dob and Chelghoum (2006) and Abd-Elhady (2012) from Algeria (0.7%) and Egypt (0.7%), respectively. In addition, it was much lower than that obtained by Drioueche et al. (2022) from Algeria (2%) and Mohammed et al. (2022) from Saudi Arabia (1.71%). Similarly, the obtained EO yield of R. montana of 0.67% is low compared to previously reported yields (0.26–2.5%; Kambouche et al. 2008; Hammami et al. 2015; Bouzeraa et al. 2019; Benali et al. 2020; Zeraib et al. 2021).

According to GC-MS data, our A. herba alba EO belongs to the davanone chemotype (66.9%) followed by camphor (9.1%) and davana ether (5.1%). In fact, our result matched data from Abdelali et al. (2023) of A. herba-alba originating from the same locality, from Cheraif et al. (2020) for the Laghouat region (Algeria) and from Salido et al. (2004) for southern Spain, which also had very high davanone contents. Other chemotypes of A. herba alba from Algeria, Tunisia, Libya and Morocco contain various compositions with major compounds identified as α and β-thujone, camphor, 1,8–cineole, chrysanthenone and cis-Chrysanthenyl acetate (Janaćković et al. 2015; Younsi et al. 2017; Aljaiyash et al. 2018; Bekka-Hadji et al. 2022). These compounds are present in our EO but in relatively small amounts (< 10%). Artemisia campestris EO revealed a high content of β-pinene (19.5%), p-cimene (8.1%), globulol (7.5%), α-pinene (7.1%) and δ-terpinene (6.0%). Abundance of β-pinene in this EO is in agreement with previous studies on A. campestris from Algeria (15.2%), Tunisia (36.4%) and Portugal (54.5%) (Abidi et al. 2018; Ammar et al. 2020; Rocha et al. 2021), whereas other studies revealed that germacrene D is the major constituent for cultivars from Tunisia and Morocco (Al Jahid et al. 2016; Younsi et al. 2017). A different EO profile is reported by Baykan Erel et al. (2012) with 1,2-dehydro acenaphthylene (20.7%), tremetone (15.8%), and capillin (10.4%) as major compounds. Piperitone (68.7%) was the major compound of A. judaica followed by δ-cadinene (6.3%) and thymol (3.1%), which is in agreement with the high percentage of piperitone in plants collected from both Tamanrasset (91.8%) and Ilizi (61.9%) regions in Algeria (Dob and Chelghoum 2006; Drioueche et al. 2022) and other parts of the MENA region (Abd-Elhady 2012; Abu-darwish et al. 2016; Mohammed et al. 2022). Al-Wahaibi et al. (2018) reported β-eudesmol (13.1%), hexadecanoic acid (5.7%) and spathulenol (3.7%) as major compounds in this EO from Saudi Arabia. Essential oil from R. montana samples is found to be dominated by the aliphatic ketone 2-undecanone (70.1%) followed by Z-β-farnesene (7.4%) and caryophyllene (3.0%). This result is in agreement with the > 78.5% of 2-undecanone in semi-arid regions compared to those from humid regions (≤ 30.3%; Mohammedi et al. 2019). Hammami et al. (2015) have reported other main components of this EO from Tunisia, namely 1-butene (38.3%), methylcyclopropane (15.5%), 2-butene (22.6%) and caryophyllene oxide (8.2%). Overall, variation and diversity in EO yields and chemical composition within the same cultivar are associated with several factors including climate and environmental conditions, soil geology, harvesting time as well as drying and extraction methods (Zeragui et al. 2019; Mohammed et al. 2022).

For the evaluation of the acute toxicity of EOs for V. destructor, the complete exposure method in closed Petri dishes was selected. According to Maggi et al. (2010), in complete exposure, EOs action on mites occurred by contact, fumigation and also through ingestion. An effective varroacidal agent has to present high toxicity to the mites in combination with low lethality to honey bees (Brasesco et al. 2017). Our in vitro bioassays showed that the four tested EOs overall have a potent concentration- and exposure time-dependent acaricidal effect against V. destructor.

On the other hand, EOs’ toxicity for honey bee adults was quite low – NOAEL values were calculated for A. campestris (25 µL/mL), A. judaica (10 µL/mL), R. montana (10 µL/mL) and A. herba alba (5 µL/mL), and at these concentrations V. destructor lethality ranged from 96.7 to 100%. These results indicate clearly the safety of the tested EOs for A. mellifera honey bee workers and their efficacy to control V. destructor.

Selectivity index is a key parameter to evaluate the miticidal potential based on the LC50 values calculated for both mites and bees. Thymol, an organic treatment largely used in beekeeping practice for Varroa control worldwide (Damiani et al. 2009; Glavinic et al. 2022) was used as positive control. After 24 h of exposure, thymol had a SR of 3.42, lower than that of A. judaica (6.15), A. campestris (6.03) and A. herba alba (4.31), and higher than that of R. montana (2.52). The SR value of thymol showed a slight increase to 3.65 after 72 h of exposure. Our results are in agreement with the 3.2 after 72 h of exposure found by Hýbl et al. (2021). The highest SR values at the end of the experiment were attributed to A. campestris (SR = 10.77) and A. judaica (SR = 8.78) followed by A. herba alba (SR = 5.62) and R. montana (SR = 3.73) – these values are in the same range with some reported in the literature (Conti et al. 2020; Hýbl et al. 2021).

Our findings indicated that A. campestris EO possessed the strongest acaricidal activity against V. destructor. Some studies had investigated its anthelmintic action (Abidi et al. 2018; Akkari et al. 2014), and recently, Ammar et al. (2020) explored its insecticidal effect against Culex quinquefasciatus, Musca domestica, and Spodoptera littoralis. Results showed larvicidal activity against C. quinquefasciatus and M. domestica with LC50 = 45.8 mg/L and LD50 = 99.8 µg/adult, respectively. Artemisia campestris EO was dominated by monoterpenes (53.1%) that are thought to act on the mite’s capability to reproduce (Imdorf et al. 1999). Notably, several studies have confirmed the effectiveness of monoterpenes against V. destructor (Fassbinder et al. 2002; Gashout and Guzmán-Novoa 2009), and also against ticks such as Hyalomma marginatum adults (Cetin et al. 2010) and Ixodes ricinus nymphs (Thorsell et al. 2006). A recent study showed that β-pinene-rich EOs from Citrus spp. similar to our A. campestris EO presented a potent acaricidal activity towards V. destructor (Bava et al. 2021). Da Camara et al. (2020) demonstrated the toxicity of β-pinene, p-cimene, α-pinene and γ-terpinene against the two-spotted spider mite, Tetranychus urticae, both by fumigation and residual contact bioassays.

Artemisia judaica oil with high amount of ketones (70.8%) also has a good in vitro acaricidal effect towards V. destructor. Our study is the first reporting this activity. El-Sharabasy (2010) tested crude A. judaica extracts (using ethanol, acetone and petroleum ether organic solvents) against T. urticae and concluded strong acaricidal activity. Other properties were also attributed to its EO such as antimicrobial, insecticidal, antifeedant and antifungal action (Mohamed and Abdelgaleil 2008; Benmansour et al. 2016). High piperitone content of 68.7% in A. judaica EO incriminates its responsibility in the pronounced miticidal activity. Abdelgaleil et al. (2008) demonstrated insecticidal (LD50 = 0.68 µg/larva), and antifeedant activities of piperitone against S. littoralis. In the same way, piperitone from Cymbopogon schoenanthus expressed pronounced efficacy on the seed beetle Callosobruchus maculatus with LC50 = 1.6 µL/L (Ketoh et al. 2006).

Essential oil from A. herba alba achieved a noteworthy toxicity on V. destructor with SR = 5.62 at 72 h post treatment. Alahyane et al. (2022) also demonstrated acaricidal activity of A. herba alba EO on V. destructor in vitro, by fumigation. Still, they also reported a high fumigant toxicity of this EO to A. mellifera pupae (LC50 = 10.53 µL/L air); no toxicity data on A. mellifera adults were reported. Recently, A. herba alba EO was found to exhibit a very promising outlook as an acaricidal agent on I. ricinus ticks with LC50 = 0.3 µL/cm2 after 2 h exposure (Elmhalli et al. 2021). El-Seedi et al. (2017) revealed that oils rich in piperitone exert a high repellent behavior up to 84.2% against I. ricinus.

Ruta montana EO had a relatively low efficacy in varroa control compared to the Artemisia EOs; however, it performed still better than thymol. Currently, no reports in literature about R. montana EO acaricidal activity exist. Bouzeraa et al. (2019) pointed at toxic and repellent properties against Ephestia kuehniella larvae of R. montana EO. The larvicidal effect of the hydro-methanolic extract of R. montana on vectors of the avian plasmodium Culiseta longiareolata was significant (Bouabida and Dris 2022). An ethnobotanical survey by Carvalho da Silva et al. (2023) in Portugal, portrayed R. graveolens with ectoparasiticide power for dogs and livestock. More evidences touted insecticidal and repellent effects of EOs from various Ruta species (Conti et al. 2013; Akkari et al. 2014; Tampe et al. 2016; Achimón et al. 2022; Wang et al. 2022). Apparently, 2-undecanone, the major compound in R. montana EO, might be responsible for that miticidal effect due to its antifungal, anti-inflammatory, and insect repellent properties (Liu et al. 2015; Bailly 2023).

Our most surprising finding was that, besides monoterpenes which are well documented for their acaricidal activity against the honey bee pest mite V. destructor, ketones (piperitone, davanone and 2-undecanone) may be considered as promising natural candidates to fight V. destructor due to the good tolerance by bees and their strong varroacidal effect.

These findings must be followed up with trials in field conditions in order to optimize the delivery mode, exposure period, colony status (with or without brood) as well as the influence of climate conditions. The abundance of these plant species to beekeepers and relatively easy processing method to obtain their EOs make them practical and cost-effective parts of an integrated pest management control strategy to fight this parasitic mite and help protect the economically relevant honey bees.