Insecticidal activity and biochemical composition of Citrullus colocynthis, Cannabis indica and Artemisia argyi extracts against cabbage aphid (Brevicoryne brassicae L.)

Plant extracts contain many active compounds, which are tremendously fruitful for plant defence against several insect pests. The prime objectives of the present study were to calculate the extraction yield and to evaluate the leaf extracts of Citrullus colocynthis (L.), Cannabis indica (L.) and Artemisia argyi (L.) against Brevicoryne brassicae and to conduct biochemical analysis by gas chromatography-mass spectrometry (GC-MS). The results suggested that when using ethanol, C. colocynthis produced a high dry yield (12.45%), followed by that of C. indica and A. argyi, which were 12.37% and 10.95%, respectively. The toxicity results showed that A. argyi was toxic to B. brassicae with an LC50 of 3.91 mg mL−1, followed by the toxicity of C. colocynthis and C. indica, exhibiting LC50 values of 6.26 and 10.04 mg mL−1, respectively, which were obtained via a residual assay; with a contact assay, the LC50 values of C. colocynthis, C. indica and A. argyi were 0.22 mg mL−1, 1.96 and 2.87 mg mL−1, respectively. The interaction of plant extracts, concentration and time revealed that the maximum mortality based on a concentration of 20 mg L−1 was 55.50%, the time-based mortality was 55% at 72 h of exposure, and the treatment-based mortality was 44.13% for A. argyi via the residual assay. On the other hand, the maximum concentration-based mortality was 74.44% at 20 mg mL−1, the time-based mortality was 66.38% after 72 h of exposure, and 57.30% treatment-based mortality was afforded by A. argyi via the contact assay. The biochemical analysis presented ten constituents in both the A. argyi and C. colocynthis extracts and twenty in that of C. indica, corresponding to 99.80%, 99.99% and 97% of the total extracts, respectively. Moreover, the detected caryophylleneonides (sesquiterpenes), α-bisabolol and dronabinol (Δ9-THC) from C. indica and erucylamide and octasiloxane hexamethyl from C. colocynthis exhibited insecticidal properties, which might be responsible for aphid mortality. However, A. argyi was evaluated for the first time against B. brassicae. It was concluded that all the plant extracts possessed significant insecticidal properties and could be introduced as botanical insecticides after field evaluations.

a concentration of 20 mg mL −1 and caused 88.33 ± 3.87% mortality, while after 48 h of exposure, the mortality was recorded as 60.00 ± 2.27%. The mortality caused by C. indica after a 72 h and 48 h exposure period was 66.67 ± 2.58% and 45.00 ± 4.72%, respectively, which were lower than those of A. argyi. However, the mortality caused by C. colocynthis was higher than that caused by C. indica after a 72 h exposure period (76.67 ± 1.29%) but decreased to 40.00 ± 2.23% after 48 h of exposure at the same concentration. The mortality of B. brassicae by A. argyi (35.00 ± 3.87%) was also higher after 24 h of exposure than that of 24 h of exposure to C. colocynthis and C. indica extracts (23.33 ± 2.58% and 20.00 ± 2.23%, respectively). The control treatment showed no mortality after 24 h and 48 h and negligible mortality (1.67 ± 1.29%) after 72 h, whereas imidacloprid was the positive control and showed the highest (P < 0.05) mortality, i.e., 98.33 ± 2.58, 91.67 ± 2.58 and 86.67 ± 1.29% after 72 h, 48 h and 24 h, respectively.
The results described in Table S2 in the Supplementary File revealed highly significant (P < 0.001) model fitness with (F = 101.057, 9851.888, 59.047, 754.712 and 545.650) intercept, concentration and time. Moreover, the interaction between the mortality (%) of the aphids versus time and concentration exhibited a highly significant (P < 0.001) positive correlation (F = 8.62). The two-way analysis of variance provided information regarding the interaction between mortality (%), treatment and time and highlighted highly significant results (F = 7.25). Interaction/correlation between treatments × concentration × time and the total variance was firmly trait-specific and significantly correlated with mortality (P < 0.001 and with F = 2.75).
However, data for interaction among plant species, concentration and post-treatment time revealed that the maximum mean mortality based on concentration by different treatments/plant species was recorded at 20 mg mL −1 , with 50.56 ± 7.92% followed by 43.52 ± 7.52, 35.37 ± 7.02, 28.89 ± 6.12% and 20.19 ± 3.16% at 15, 10, 5 and 2.5 mg mL −1 , respectively. The maximum mean mortality interaction based on time was recorded as 55.00 ± 6.79% at 72 h exposure, followed by 36.35 ± 6.02 and 23.89 ± 5.79% at 48 and 24 h, respectively. On the other hand, mortality interaction among treatments was evaluated as high (44.13 ± 7.14%) by A. argyi followed by that of C. indica and C. colocynthis (36.99 ± 6.67% and 34.13 ± 6.53%, respectively). Moreover, the positive control showed a maximum 90.19 ± 2.09% mortality, and the negative control showed a minimum mortality of 0.19 ± 0.18%. Probit analysis showed the LC 50 values, slope, chi-square, and fiducial limits at the 95% confidence interval. Although B. brassicae showed sensitivity to the different plant extracts, A. argyi was the most toxic, followed by C. colocynthis and C. indica, and this mortality response is presented in Table 1.
Contact/aphid dip bioassay. The data presented in Table S3 in the Supplementary File described the mortality (%) of B. brassicae by the contact toxicity method. The results showed that maximum mortality was recorded after 72 h at 20 mg mL −1 . However, the highest mortality caused by the crude extract of A. argyi was 93.33 ± 3.14%, followed by that of C. colocynthis and C. indica (83.33 ± 2.57 and 81.67 ± 1.29%, respectively). High mortality after 48 h of exposure to 15 mg mL −1 was also observed in A. argyi (83.33 ± 1.29%), followed by that of C. colocynthis and C. indica, which had the same mortality values of 81.67 ± 3.41% and 81.67 ± 1.67%, respectively. Additionally, the mortality of B. brassicae by the contact assay was also high for exposure to extracts of C. indica even at low concentrations (78.33 ± 2.58%, 71.67 ± 3.41% and 71.67 ± 3.41% at concentrations of 10, 5 and 2.5 mg mL −1 , respectively). The same trend of mortality was observed for C. colocynthis and A. argyi. The negative control treatment showed similar results for mortality in all three experiments for each treatment in both bioassay methods, while imidacloprid showed significantly (P < 0.05) high mortality (98.33 ± 1.29%, 96.67 ± 2.58% and 95.00 ± 2.23% after 72 h, 48 h and 24 h, respectively) for leaf dip and contact bioassay, which was also the highest mortality.
The results described in Table S4 in the Supplementary File revealed also highly significant (P < 0.001) model fitness with (F = 44.42, 11152.07, 14.66, 409.34 and 67.44) intercept, concentration and time. However, the interaction between mortality (%) of the aphid versus time and concentration recorded a highly significant (P < 0.001) positive correlation (F = 1.58). The two-way analysis of variance provided information regarding the interaction between mortality (%) versus treatment and time highlighted highly significant results (F = 10.07). The interaction/correlation between treatments × concentration × time and total variance was strongly trait-specific and highlighted significant correlation with mortality (P < 0.001 with F = 1.79). However, data for the interaction among plant species, concentrations and post-treatment time revealed that the maximum mean mortality based www.nature.com/scientificreports www.nature.com/scientificreports/ on concentration by different treatments/plant species was recorded at 20 mg mL −1 , with 74.44 ± 3.74% followed by 70 ± 3.58, 63.33 ± 3.08, 56.48 ± 4.34 and 47.59 ± 4.74% at 15, 10, 5 and 2.5 mg mL −1 , respectively. The maximum mean mortality interaction based on time was recorded as 66.83 ± 6.63% after 72 h of exposure, followed by 56.27 ± 6.19% and 51.51 ± 6.82% after 48 and 24 h, respectively. The mortality interaction among treatments was 57.30 ± 6.31% by A. argyi followed by 62.22 ± 7.14% and 55.08 ± 6.62% by C. colocynthis and C. indica, respectively. Moreover, the positive control treatment showed a maximum of 95 ± 1.14% mortality, and a 0.56 ± 0.27% mortality was exhibited by the negative control.

Discussion
Due to the problems concerning the use of synthetic chemicals for pest management, there is an urgent need to introduce natural products, mainly from plant sources, for use against insect pests, especially B. brassicae, which is becoming problematic in cabbage growing areas of the world. Plant extracts or essential oils are commonly applied for pest control measures because of their effectiveness against different life stages of the pests. However, the selected plants (C. colocynthis, C. indica and A. argyi) are naturally repellent to some extent to insect attack, which indicates that they could be suitable candidates against cabbage aphids, B. brassicae. An investigation was performed to detect the phytochemical constituents of the extracts of the above-mentioned plants, and their insecticidal activity was evaluated by residual and contact toxicity methods because of the feeding style of aphids 33 . However, compound extraction from plant parts depends upon the type of plant material and solvent used. High-polarity solvents produced a higher yield than that of low-polarity solvents. In our results, ethanol afforded high extract yields during solvent extraction, which was due to its high polarity. Previously, it was reported that ethanol yielded more than other solvents from leaves of Melastoma malabathricum 34 . Furthermore, a relatively high extract yield from C. colocynthis and C. sativa using ethanol as an extraction solvent was reported 35 .
According to our findings, the mortality of B. brassicae using plant extracts increased with increasing concentrations and exposure periods. However, the physical and chemical properties of the essential oils exhibited   www.nature.com/scientificreports www.nature.com/scientificreports/ different persistent levels of insecticidal properties and different action mechanisms 36,37 . It was reported that extracts of Cassia sophera and Ageratum conyzoides against B. brassicae induced mortality that was equivalent to that of the positive control, the synthetic insecticide emamectin benzoate 38 . Moreover, a Mentha piperita extract resulted in maximum insecticidal activity against B. brassicae with increased concentration and time exposure 39 . Similarly, essential oil from Cinnamomum zeylannicum was found to be effective against B. brassicae, exhibiting 8.3 and 7.4 µl mL −1 LC 50 values 40 . The essential oil of another medicinal plant, Elettaria cardamomum, was reported to be effective against this serious pest, exhibiting 79.2 µl mL −1 LC 50 value 41 . There are reports that B. brassicae was sensitive to 1,8 cineole, a chemical compound found in Laurus nobilis, which at LD 50 value of 30 µl mL −1 caused 52% mortality 42 . Additionally, essential oils isolated from natural plants are usually considered as safe for humans and environment which can be a source of new botanical insecticides. Among some of the plant species, Ocimum basilicum, Mentha piperita, Pimpinella anisum, Mentha pulegium and Foeniculum vulgare have shown outstanding effectiveness against aphids species in both contact and fumigation assays and can thus be considered as active materials for developing new botanical agents 43 . It was also reported that the use of Azadirachta indica oil as a suitable alternative to synthetic chemicals against B. brassicae caused a noticeable reduction in their population 44 . All these findings are consistent with our results on the mortality of B. brassicae at different concentrations and varying levels of LC 50 values.
The results of the studies using the C. colocynthis leaf extract to control B. brassicae are exceptional. However, toxicity of C. colocynthis extracts from leaves and fruits against Aphis craccivora was reported 45 . Cucurbitacin E glycosides isolated from C. colocynthis extract showed strong insecticidal effects against Aphis craccivora 46 . Limonene contained by Citrus sinensis was reported to be effective against Myzus persicae at LC 50 57.7 µl mL −1 , whereas its essential oil at a concentration of 3.3 µl mL −1 caused significant mortality of B. brassicae 41 . All these results support our findings of using C. colocynthis extract against B. brassicae. However, according to our results from the GC-MS analysis, the chemical composition of C. colocynthis was dominated by four of the ten detected compounds. Among these, erucylamide is a high fatty acid amide that can be used as an insecticide, pigment dispersant, ink and paint additive, fibre softener, and in dyes. Octasiloxane hexamethyl from C. colocynthis also exhibits insecticidal properties.
The insecticidal activity of the essential oil of hemp inflorescences against M. persicae was reported to be highly toxic, with an LC 50 of 3.5 mL L −1 and non-toxic to the beneficial organisms 26 . An important compound, (E)-caryophyllene myrecene, and α-pinene, contained by C. sativa, showed 98.20% mortality at 3.50 µl mL −1 of Aulacorthum solani 47 , whereas at the same concentration, similar mortality of M. persicae was reported 26 . Previous studies have reported the concentration and time-dependent mortality of Tetranychus urticae and Aulacorthum solani following the use of hemp essential oil, which supported our findings indicating the effectiveness of the hemp (C. sativa) crude extract against B. brassicae aphids 47 . However, among the twenty compounds identified in the crude ethanol extract, caryophylleneonides (sesquiterpenes), α-bisabolol and dronabinol (Δ 9 -THC) possess insecticidal activities, However, caryophyllene oxide identified from Melaleuca styphelioides exhibited strong aphicidal activities against Aphis spiraecola, Aphis gossypii and M. persicae 48 while, the other compounds may need to be further explored to determine their biological properties. To date, few studies on the applicability of hemp essential oil against crop pests have been conducted, and the use of hemp essential oil against B. brassicae is very rare in the literature.
Similarly, A. argyi is a famous plant commonly found in China, and many species of the genus Artemisia have long been used in traditional Chinese medicines, including Artemisia argyi, Artemisia capillaris and Artemisia annua 49 . In the present study, the crude ethanol extract from the leaves of A. argyi was proven to be highly toxic against B. brassicae. Moreover, four compounds from A. argyi, such as camphor, β-pinene, eucalyptol, and β-caryophyllene, showed strong toxic effects against adults of Lasioderma serricorne 50 . Other studies have showed that sabinene, and β-myrecen contained in Artemisia absinthium, caused significant mortality of M. persicae at 6.9 µl mL −1 concentration 51 . Similarly, Artemisia seiberi essential oil caused significant mortality of Eriosoma lanigerum at 6.16 µl mL −1 LD 50 value 52 . Another compound, artemisia ketone, showed significant mortality of B. brassicae at 25.0 µl mL −1 concentration 53 . However, according to our results, reported compounds from A. argyi leaves were evaluated for the first time for their insecticidal potential. The above-described results supported our findings and clearly demonstrated that the A. argyi crude extract exhibited strong and deadly effects against B. brassicae, which justified its great potential for controlling this pest.
In the present study, the toxicity of ethanol extracts of the leaves of C. colocynthis, C. indica, and A. argyi against B. brassicae were investigated under laboratory conditions, revealing their positive toxicity. A. argyi proved to be the most toxic in both bioassay methods and caused maximum mortality, followed by C. colocynthis and C. indica, in a dose-dependent manner. This insecticidal activity might be due to individual efficacy or the synergistic action of biological compounds present in these plants. A literature survey regarding the potential efficacy of A. argyi showed that there are no studies reporting on its use against B. brassicae. Additionally, data on the insecticidal potential of C. colocynthis and C. indica against this serious pest and their GC-MS profiling are limited to a few studies. Thus, the present unique and novel study was conducted for the first time to investigate the insecticidal potential of crude ethanol extracts by residual and contact toxicity methods against B. brassicae, a serious pest of cabbage and other field crops.

conclusions
The investigations indicated that C. colocynthis, C. indica and A. argyi possess potential botanical agents as an alternative to synthetic pesticides against B. brassicae. It can be concluded that, B. brassicae, is reasonably sensitive to A. argyi followed by C. colocynthis and C. indica as shown in both bioassay methods, even at low concentrations and exposure times. Moreover, in comparison, the contact assay afforded higher mortality than that of the residual assay. Additionally, GC-MS analysis revealed the presence of valuable biologically active compounds (2020) 10:522 | https://doi.org/10.1038/s41598-019-57092-5 www.nature.com/scientificreports www.nature.com/scientificreports/ responsible for aphid mortality. However, further studies are required on the purification and identification of active components responsible for aphid mortality and their evaluations against other insect pests. preparation of plant material and extraction. The collected leaves were allowed to dry in the shade at room temperature after removing impurities by rinsing in tap water. Dried leaves were ground into a fine powder using an electric blender, and 99.7% ethanol was used as a solvent for extraction. Approximately 100 g of ground powder was extracted thrice with 400 mL of ethanol each time for 72 h at a constant temperature in an incubator shaker (ZWY-1102C) at 100 rpm. After each extraction, the material was filtered through filter paper (Whatman No. 1) and concentrated to reduce the volume by using a rotary evaporator (Buchi Switzerland R-210). Finally, the concentrated filtrate was subjected to drying in a fume hood at 25 °C for 12 h. The extraction yield was measured for each extract, and the extracts of each plant were combined to obtain a bulk sample and stored in glass-stoppered vials at 4 °C until use.

Serial concentration and agar preparation.
Tween-20 was used to minimize variations among the treatments effect of any adjuvant material. Fifty millilitres of Tween was dissolved in 950 mL distilled water by shaking to prepare a 5% solution. Serial concentrations of crude extract, 2.5, 5, 10, 15 and 20 mg mL −1 were prepared in the Tween solution one day before beginning the experiment. To prepare the agar, 10-15 g agar rods were added to one litre of distilled water, boiled for 10 min with continuous stirring, allowed to cool for 10 min and was then added to Petri dishes 90 mm in diameter and 15 mm deep up to a 3-4 mm level.
Bioassay study. Two different modes of exposure, residual and contact assays were conducted for the toxicity evaluation. The experiment was replicated five times for each exposure method as well as concentration, and the mortality response was recorded at the exposure times of 24 h, 48 h and 72 h.
Residual toxicity/leaf dip bioassay. Briefly, cabbage leaf discs 5 cm in diameter were cut with sterilized scissors and dipped for 10 s in the extracts of varying concentration. Then, the leaf discs were left at room temperature for 10 min to dry and put in Petri dishes containing agar. Then, twenty aphids at the second instar were carefully transferred onto the leaf disc contained in the Petri dish by a fine-haired brush, and caution was taken to prevent injury to the aphids during their transfer to the leaf disc. Then, a double layer of muslin cloth was placed on each Petri dish and covered with a manually perforated lid to avoid suffocation and escape of the aphids (Fig. 1). The negative control treatment was prepared using a Tween-20 (5%) solution as described above but without added plant extract. An imidacloprid solution of 20% SL (2.5 mL L −1 ) was used as a positive control. All Petri dishes were incubated for 72 h at 65% R.H., 25 °C and with a 16:8 (light: dark) photoperiod. contact toxicity/aphid dip bioassay. A 30 mm diameter plastic Petri dish was cut from the basal portion, and the remaining ring was used as the container. A nematode strainer (mesh size 300 µm) was used as a holding cell for aphids during dipping into the plant extracts of varying concentration (Fig. 2). Then, twenty aphids at second instar were isolated from the cabbage plants and dipped in the extracts for 10 s. Each group of dipped aphids was then placed on fresh 5 cm diameter cut leaf discs on agar-containing Petri dishes. The negative control treatment was prepared using the Tween-20 (5%) solution as described above but without adding plant extract. An imidacloprid solution of 20% SL (2.5 mL L −1 ) was used as a positive control. All Petri dishes were incubated for 72 h at 65% R.H., 25 °C and with a 16:8 (light: dark) photoperiod.
Data collection. Mortality data were recorded at the exposure times of 24 h, 48 h and 72 h by microscopic examination. The response of B. brassicae to the crude extracts was observed by needle stimulation, and those individuals who displayed no response were considered dead.
Sample preparation for Gc-MS analysis. The solvent-free extract was weighed precisely to 0.005 g and dissolved in 1 ml of HPLC-grade methanol. Then, the sample was added to a centrifuge tube, and graphitized carbon blacks (GCB) was added (0.002 mg/tube) and shaken vigorously for one min on a vortex (Vortex Gennic 2, Model G560E Scientific Industries Inc. U.S. A) to remove coloured components. The colourless sample was Scientific RepoRtS | (2020) 10:522 | https://doi.org/10.1038/s41598-019-57092-5 www.nature.com/scientificreports www.nature.com/scientificreports/ centrifuged for 10 min at 24 °C and 12000 rpm (Hitachi Centrifuge CT15RE, Koki Co., Ltd. Japan) to separate the GCB from the sample. The colourless sample was collected by micropipette and transferred to a new centrifuge tube.
Gas chromatography-mass spectrophotometry (Gc-MS) analysis. The crude ethanol extracts of C. colocynthis, C. indica and A. argyi were analysed via GC-MS using an Agilent 6890-5973N USA gas chromatograph (GC) equipped with an HP1 TG-5MS polydimethylsiloxane capillary column (30 m × 250 µm × 0.25 µm) interfaced with a Hewlett Packard 5973 mass selective detector. Gas chromatographic parameters were as follows: the temperature was fixed at 110 °C for 2 min initially and increased to 200 °C and 280 °C at increasing rates of 10 °C min −1 and 5 °C min −1 , respectively. The inlet temperature was 250 °C; a 10:1 split ratio was used; MS quadrupole temperature, 150 °C; thermal aux temperature, 285 °C; ionization current, 60 µA; MS temperature, 230 °C; MS scan range, m/z 40-450; ionization energy, 70 eV; carrier gas, helium with flow rate, 1.0 mL min −1 . Compounds were identified by comparison of the GC-MS mass spectra with those in the literature or in databases at Wiley/NIST.98.1 54,55 . The relative abundance of each compound was assessed based on the raw data peak areas in the spectra with no response factor correction from the FID. Data analysis. All recorded mortality data were subjected to two-way analysis of variance (ANOVA) for calculating the interaction among plants, concentration and post-treatment time, while the mean difference between treatments was calculated for the significance test by Tukey's HSD at the P = 0.05 level. All statistical processes were administered by different statistical packages with IBM-SPSS version 25.0 software, while LC 50 values, slope, chi-square values and fiducial limits were calculated by the EPA Probit analysis program version 1.5.