Chemical Diversity and Anti-Insect Activity Evaluation of Essential Oils Extracted from Five Artemisia Species

As a source of aromatic plants, the genus Artemisia has long been considered to have the potential to develop plant pesticides. In this study, components of essential oils from A. dalai-lamae, A. tangutica, A. sieversiana, A. tanacetifolia and A. ordosica were identified by GC-MS. A total of 56 constituents were analysed, and each species consisted of 9 to 24 constituents. Principle component analysis (PCA) revealed that A. dalai-lamae, A. tangutica and A. tanacetifolia are characterised by monoterpene hydrocarbons and oxygenated monoterpenes. Hierarchical cluster analysis (HCA) showed the most remarkable similarity between A. sieversiana and A. ordosica, but the similarity was still lower than 50%. Contact toxicity and repellency of essential oils were evaluated by bioassays; A. ordosica oil exhibited the most substantial contact toxicity (LD50 = 52.11 μg/cm2) against Liposcelis bostrychophila, while A. tangutica oil showed the most potent contact toxicity (LD50 = 17.42 μg/adult) against Tribolium castaneum. Except for A. dalai-lamae, the other four species showed the same level (p > 0.05) of repellent activity as the positive control against both pests at high concentrations. The results indicated that these five Artemisia species had high chemical diversity and great potential to be developed into more effective and environmentally friendly anti-insect agents.


Introduction
Insecticides are an effective method of controlling storage pests, of which synthetic insecticides have been demonstrated to impact environmental protection and health safety negatively [1,2]. This phenomenon promoted the growing exploration of botanical pesticides. Plant secondary metabolites are important sources of botanical insecticides and have proven insecticidal [3,4] and repellent [5,6] effects in practice. Increasing studies have shown that essential oils mainly extracted from aromatic plants have great potential against arthropod species [7]. The genus Artemisia, one of the largest genera of the family Compositae, consists of nearly 500 species worldwide, and about 190 species are found in China. High concentrations and significant intraspecific variations of volatile terpenes in essential oil generated the strong and diverse odour of the Artemisia genus [8]. The compositional diversity of essential oils of several Artemisia species has been reported. A. dracunculus was rich in (Z)-anethole (81.0%) [9], A. scoparia was rich in α-thujone (81.7%) [10], while 67% of camphor was the main component of A. fragans [11].
The positive repellent and insecticidal activities of essential oils derived from Artemisia species have been in the limelight owing to the abundant volatile components [8,12]. Essential oils extracted from three Artemisia species, including A. absinthium, A. spicigera and A. santonicum, were toxic to Sitophilus granaries [13]. Oils from cultivated A. absinthium had better repellent activity against Trypanosoma cruzi and Leishmania infantum than the commercial ones [14]. A. capillaris and A. mongolica essential oils showed significant toxicity against Sitophilus zeamais [15].
Little research was conducted on A. dalai-lamae and A. tangutica, the endemic species in China. Terpenoids, lignans, flavonoids, and various compounds were isolated from A. sieversiana. Some of these components showed several biological activities such as anti-tumour, anti-inflammatory, anti-allergic, anti-hypertensive, and anti-hyperglycemic activities [16]. The methanolic extract of the leaves of A. tanacetifolia afforded several kinds of coumarins [17], and 5-O-caffeoylquinic acid was detected from the aerial parts of flowering A. tanacetifolia [18]. Essential oil from A. ordosica has the effectiveness of antiinflammatory, haemostasis, treating rheumatoid arthritis, parotiditis, abdominal distension, intestinal obstruction and ischuria [19]. Although numerous studies have shown that Artemisia species has various pharmacological effects, still little is known concerning the insecticidal activity of these five Artemisia species.
In this study, we reported the essential oils from five species in the genera Artemisia (A. dalai-lamae, A. tangutica, A. sieversiana, A. tanacetifolia and A. ordosica) and provided a comparative investigation of these five species' chemical composition, repellent, and contact potential for Liposcelis bostrychophila Badonnel and Tribolium castaneum Herbst adults.

Repellent Activity
The repellent rates at 2 and 4 h after exposure to essential oils derived from five Artemisia species against L. bostrychophila and T. castaneum are shown in Figure 3, respectively. The repellent effect of essential oils on both two pests showed various levels. When compared with the positive control, DEET, at both 2 and 4 h after exposure, all five oils possessed the same level of repellent activity (p > 0.05) at a testing concentration of 63.17 nL/cm 2 for L. bostrychophila and 78.63 nL/cm 2 for T. castaneum. The essential oil of A. tangutica, A. sieversiana, A. tanacetifolia and A. ordosica also showed comparable repellent levels with the positive control (p > 0.05) at the concentration of 12.63 nL/cm 2 against L. bostrychophila and 15.73 nL/cm 2 for T. castaneum adults. It is worth noting that A. tangutica and A. sieversiana were found to have attraction effects against these two insects at the lowest concentration. Among the oils, A. ordosica showed outstanding repellent activity and possessed the same (p > 0.05) with DEET at all five concentrations on both insects.

Repellent Activity
The repellent rates at 2 and 4 hours after exposure to essential oils derived from five Artemisia species against L. bostrychophila and T. castaneum are shown in Figure 3, respectively. The repellent effect of essential oils on both two pests showed various levels. When compared with the positive control, DEET, at both 2 and 4 hours after exposure, all five oils possessed the same level of repellent activity (p > 0.05) at a testing concentration of 63.17 nL/cm 2 for L. bostrychophila and 78.63 nL/cm 2 for T. castaneum. The essential oil of A. tangutica, A. sieversiana, A. tanacetifolia and A. ordosica also showed comparable repellent levels with the positive control (p > 0.05) at the concentration of 12.63 nL/cm 2 against L. bostrychophila and 15.73 nL/cm 2 for T. castaneum adults. It is worth noting that A. tangutica and A. sieversiana were found to have attraction effects against these two insects at the lowest concentration. Among the oils, A. ordosica showed outstanding repellent activity and possessed the same (p > 0.05) with DEET at all five concentrations on both insects.

Contact Toxicity
The contact toxicities of essential oils from five Artemisia species against L. bostrychophila and T. castaneum adults are listed in Tables 3 and 4. Except for A. tanacetifolia,

Contact Toxicity
The contact toxicities of essential oils from five Artemisia species against L. bostrychophila and T. castaneum adults are listed in Tables 3 and 4. Except for A. tanacetifolia, others all exhibited contact effects against L. bostrychophila. Among them, the A. ordosica essential oil possessed the most potent contact toxicity (LD 50 = 52.11 µg/cm 2 ), about three times less than the positive control, pyrethrins. Compared with A. ordosica, A. tangutica essential oil showed slightly weaker activity with an LD 50

Discussion
The chemical composition of Artemisia species oils has high diversity. The oil has different protective effects due to various compositions, such as antibacterial activity, insecticidal effect, antiviral and repellent actions.
In our tests, the main compounds of tested oils also were found in other Artemisia species. For example, camphor, linalool and 4-terpineol were the major compositions in A. haussknechtii [33], and spathulenol, β-elemene, germacrene D were found in A. campestris [34]. Caryophyllene was the main compound in A. lavandulaefolia and A. rubripes oils [35,36].
It has also been reported that changes in the composition of volatile oils can also occur during plant growth stages. For instance, monoterpenoids, the major composing components (69.5-77.7%) of Ocimum americanum oil, were found to be maximal (77.7%) in the vegetative growth stage followed in the seed setting period (76.8%) and full flowering stage (74.2%), with the minimum at the half-flowering stage (69.5%) [37].
Recently, thousands of plants have been deemed as potential sources of repellents. The repellent properties of essential oils from the genus Artemisia were also well documented. In our previous research, five Artemisia species, including A. anethoides, A. giraldii, A. roxburghiana and A. rubripes were evaluated for their repellent activities on T. castaneum [38]. The result indicated that the five essential oils were effective in repelling T. castaneum, and the sequence of their activity was A. rubripes > A. anethoides > A. roxburghiana = A. sacrorum = DEET (the positive control) > A. giraldii (p > 0.05). In another previous research, polyacetylenes were isolated from the essential oil of A. ordosica aerial parts. Although with low relative content, the three tested polyacetylenes (capillene, capillin and cis-dehydromatricaria ester) were proved to possess fair repellent and fumigant activities against T. castaneum adults. Additionally, A. lavandulaefolia essential oil and its six constituents were tested on Lasioderma serricorne [39]. At 2 h after exposure, the same level of repellency (p > 0.05) was observed at doses from 0.63 to 78.63 nL/cm 2 . In other reports, A. vulgaris essential oil presented high repellent activity against T. castaneum [40]. The essential oil extracted from A. scoparia had more marked repellent activity on Sitophilus oryzae and T. castaneum than Callosobruchus maculatus, but as a whole, the oil strongly repelled each species of tested pests [41]. Moreover, in our tests, the main constituents of the five Artemisia species were proved to have a repellent effect. For example, 3-carene had over 85% of PR values against L. bostrychophila and T. castaneum after 2 h exposure [22]. Caryophyllene had the PR values of 82% and 98% against T. castaneum after 2 h and 4 h exposure, respectively [22]. These major components were also confirmed to be toxic to other insects, such as Aedes aegypti, Semanstus japonicus and Lasioderma serricorne [42][43][44]. Therefore, this study and the previous reports proved that essential oils from the genus Artemisia have great potential to be developed as good repellent agents against storage insects.
The different toxicity effects could be found based on the LD 50 values of five Artemisia species oils obtained in this study. The essential oils from A. ordosica and A. tangutica exhibited stronger contact toxicity than others against L. bostrychophila and T. castaneum. The essential oil of A. sieversiana possessed weak toxicity against L. bostrychophila with a LD 50 value of 195.51 µg/cm 2 , and no insecticide effect was observed in tested concentrations to T. castaneum. In the previous report, A. sieversiana also possessed weak contact toxicity against Sitophilus zeamais adults with an LD 50 value of 112.7 mg/adult [27]. No significant correlation was observed when combined with the LD 50 values of principal components and essential oils. This is considered to be related to the content of components and the synergistic or antagonistic effect between compounds. Pavela et al. (2010) estimated the fumigant toxicity against Spodoptera littoralis larvae of 15 pairs of binary mixtures [45]. It showed that nine mixtures had a synergistic effect, five mixtures had an additive effect, and one mixture had an antagonistic effect. It was also identified that 138 synergistic/antagonistic effects were detected among 39 compounds in binary mixtures via topical application against Trichoplusia ni [46]. So it may need further study to reveal the relationship between the insecticidal activities of different compounds.
The mechanism of bio-action of essential oils was recorded. For contact toxicity, substances could permeate through the skin of insects to act on the insects. Some of the substances could cause neurological disorders, for example, the inhibition of the activity of acetylcholinesterase (AChE), which is an important enzyme to regulate synaptic transmission [47]. The inhibition of AChE could lead to hyperexcitation, causing locomotor behaviour alternation. High concentrations or combined use of substances may generate the knockdown effect, reflecting acute toxicity [48]. These phenomena make us care about the safety of applying essential oils and plant-derived substances. Although the potential toxicity to mammals or cells of these five Artemisia species has not been studied yet, the safety evaluation of some other species in Artemisia has been recorded. The water extracts of A. dracunculus showed a maximum tolerated dose to rodents at over 200 mL of extract (1:10)/kg bw [49]. Essential oil from A. nilagirica had a LD 50 value of 7528.10 µL/kg for male mice [50]. The essential oil of A. herba-alba showed no significant cytotoxicity in macrophages at the concentrations of 1.25 µL/mL and no significant cytotoxicity to microglial cells and keratinocytes at concentrations up to 0.32 µL/mL [51]. Moreover, the safety of the major constituents, 1,8-cineole, camphor, 3-carene and caryophyllene, was evaluated. After acute oral administration of 1,8-cineole, the LD 50 value was 3849 mg/kg, while no significant changes in body weight and relative organ weight were observed in the subacute toxicity study [52]. The natural form of camphor was not toxic at 100 mg·kg·b.w. −1 to mice [53]. For 3-carene, the exposure by inhalation at 0.0014 mg/day was believed to be safe, and it could cause eye irritation at about 2.39 mg/mL air [54,55]. Caryophyllene was not considered as a skin sensitiser, and the TTC (threshold of toxicological concern) of inhalation exposure was 0.012 mg/day, which is 117 times lower than the Cramer Class I TTC [56]. However, essential oils from some Artemisia species were confirm to have toxicity to human. For example, volatile oil extracted from A. argyi could cause hepatocellular harm to cause liver injury [57]. Although there was no direct evidence that these five essential oils are safe enough for use, combined with the toxicity of other essential oils from Artemisia and the major constituents, it can be assumed that these essential oils are at lower concentrations of safe use as insecticides.  Table 2 summarises the collecting information of these samples.

Extraction and GC-MS Analysis of Essential Oils
The plant materials were dried in the shade and coarsely ground. Then, the grounded materials were subjected to 6 h of hydrodistillation using a modified Clevenger type apparatus, and the cooled essential oils were dehydrated by anhydrous sodium sulphate. The final oils were stored in sealed containers in the refrigerator at 4 • C.
Gas chromatographic-mass spectrometry (GC-MS) analysis was performed with an Agilent 6890N gas chromatograph (Agilent Technologies, DE, USA) equipped with a flame ionisation detector (FID). A HP-5MS (30 m × 0.25 mm × 0.25 µm) capillary column was used to distinguish the compounds. A 1.0 mL/min flow rate helium was used as a carrier gas. Then, 1 µL of 1% essential oil-n-hexane solution was injected, and the injector temperature was 250 • C. The oven temperature was programmed as follows as 50 • C for Plants 2022, 11, 1627 9 of 12 2 min, increased at 2 • C/min to 150 • C for 2 min, then increased at 10 • C/min to 250 • C for 5 min.
Constituents were identified by comparing their retention indices (RI) with those reported in the literature, and by matching their mass spectra with those stored in NIST 05, Wiley 275 libraries or literature [58]. The GC-FID peak area (%) was used to obtain the relative percentages of each individual component of the essential oils.

Insects
A 10:1:1(w/w/w) mixture of flour, milk powder and yeast was used to rear L. bostrychophila, whether T. castaneum was fed with wheat flour mixed with yeast (10:1, w/w). The colonies were maintained in the dark incubators at 28-29 • C and 70-80% RH. The mixed-sex adults used in repellent and contact assays were about 7 ± 2 days old. The edges of containers and the Petri dishes for L. bostrychophila were smeared with polytetrafluoroethylene to ensure escape-proofing.

Repellent Activity
The repellence assay was performed using the area preference method [59]. Five doses of 63.17, 12.63, 2.53, 0.51 and 0.10 nL/cm 2 were made by serial dilution in n-hexane for the repellent assay applied against L. bostrychophila adults. Filter papers with a 5.5 cm diameter were cut in half. Each tested solution (150 µL) was applied to half-cut filter paper as the experimental group, and n-hexane (150 µL) was used in another half as the control group. The treated filter papers were air-dried to evaporate the solvent completely. Then, both semi-circular filter papers were attached to their opposite sides and placed in Petri dishes (Φ = 5.5 cm). Meanwhile, for T. castaneum, the filter papers and Petri dishes were prepared at 9 cm in diameter. The five tested concentrations were 78.63, 15.73, 3.15, 0.63 and 0.13 nL/cm 2 , and 500 µL of testing solution or n-hexane were treated on each semi-circular filter paper. Twenty insects were released at the centre of the Petri dishes and covered with lids for all tests. The dishes were then placed in the dark incubators in the same condition as raring. Five replications were used for each concentration. The positive control was conducted by DEET (N, N-diethyl-3-methylbenzamide, Dr. Ehrenstorfer, Germany). The numbers of insects present on different sides of the paper were recorded after 2 and 4 h.

Contact Toxicity
The contact toxicity of the essential oils was conducted as follows. The appropriate testing concentrations were determined for all bioassays based on range-finding studies. Then, the desired quantity of each sample was dissolved in n-hexane to obtain a series of concentrations as the testing solution. When it came to L. bostrychophila adults, 300 µL of the solutions of the essential oils were applied to a round filter paper of 5.5 cm in diameter. Then the treated filter paper was attached to the bottom of Petri dishes of the same size as the filter papers. Ten insects in each treatment were put in the Petri dishes. All the Petri dishes were covered by lids and kept in the incubator. For the bioassays with T. castaneum adults, 0.5 µL of solutions were applied to the insects' dorsal thorax. Ten treated insects with the same solution were transferred into one vial, and reared in the incubator.
After 24 h, the number of deaths was checked and recorded. The n-hexane was used as the negative control, and pyrethrin (pyrethrin I and II, 37%) was used as the positive control. The experiments were replicated five times.

Statistic Analysis
Principle component analysis (PCA) was used to explain the dissimilarities between samples. Hierarchical cluster analysis (HCA) was used to evaluate the similarity of the samples based on the type and quantity of compounds from essential oils. PCA was conducted by R Studio (version 4.1.3) with FactoMineR [60] and factoextra [61] packages. HCA was performed with Minitab 17 by using the complete linkage and Euclidean distance measure methods. A dendrogram was constructed to express the result.
The percentage repellency (PR) was calculated to measure the repellent activity of essential oils, which was computed by the foluma as below [59]: PR(%) = [(Nc − Nt)/(Nc + Nt)] × 100 Nc and Nt are the numbers of insects in the negative control and treated half, respectively. Then the analysis of variance (one-way ANOVA) and Tukey's test were conducted using SPSS 20.0.
For the contact toxicity test, LD 50 was calculated by Probit analysis using SPSS 20.0. The 95% FL (fiducial interval), Slope ± SE, p-value and χ 2 were also recorded.

Conclusions
The chemical composition, repellent activity and contact toxicity of five Artemisia species (A. dalai-lamae, A. tangutica, A. sieversiana, A. tanacetifolia and A. ordosica) essential oils extracted by hydrodistillation were compared by PCA and HCA. All the oils showed obvious repellent activity against both insects in repellent tests. As for contact toxicity, the obtained values of LD 50 demonstrated that several of the evaluated essential oils possessed toxic effects on L. bostrychophila and T. castaneum adults. Among these oils, A. ordosica essential oil possessed the most potent contact toxicity (LD 50 = 52.11 µg/cm 2 ) against L. bostrychophila, and A. tangutica essential oil exhibited the most substantial toxicity (LD 50 = 17.42 µg/adult) against T. castaneum. Hence, the results suggest that the above five Artemisia species have the potential to be further exploited as repellent and insecticide agents against storage pests.