Phytotoxic Effects of Commercial Eucalyptus citriodora, Lavandula angustifolia, and Pinus sylvestris Essential Oils on Weeds, Crops, and Invasive Species

Background: essential oils are well known for their pharmacological effectiveness as well as their repellent, insecticide, and herbicide activities. The emergence of resistant weeds, due to the overuse of synthetic herbicides, makes it necessary to find natural alternatives for weed control. The aim of this study was to evaluate the phytotoxic effects of Eucalyptus citriodora, Lavandula angustifolia, and Pinus sylvestris, three common commercial essential oils, on weeds (Portulaca oleracea, Lolium multiflorum, and Echinochloa crus-galli), food crops (tomato and cucumber), and the invasive species Nicotiana glauca. Methods: to determine herbicidal effects, essential oils were tested at different concentrations (0.125–1 µL/mL). The index of germination and seedling length data were recorded over 14 days. Results: the in vitro assays showed that L. angustifolia with linalool (38.7 ± 0.1%), 1,8-cineole (26.5 ± 0.1%), and camphor (14.2 ± 0.1%) as the main compounds showed the most phytotoxic effects affecting seed germination in weeds and tomato, and the aforementioned invasive species. L. multiflorum was the most sensitive weed, particularly to lavender essential oil, which decreased the growth of its hypocotyl and radicle by 87.8% and 76.7%, respectively, at a dose of 1 µL/mL. Cucumber was the most resistant food crop, with no significant reduction observed in seed germination and hypocotyl growth with E. citriodora and L. angustifolia essential oils. Conclusions: lavender essential oil represents a promising candidate for the development of effective and safe herbicides in the management of L. multiflorum affecting cucumber crops.


Introduction
The particular characteristics of essential oils-natural mixtures of volatile compounds-provide them with certain pharmacological properties, including their well-known antibacterial, antioxidant, anti-inflammatory, and cancer chemoprotective effects, as well as their repellent, herbicidal, and insecticidal biological activities [1][2][3][4], which have led to valuable applications in human health, food, and cosmetics industries, and in environment and agriculture. Certain essential oils have already demonstrated their influence on both the seed germination and seedling growth of weeds [5,6]. In this regard, origanum (Origanum vulgare L.) essential oil with carvacrol as its main compound has exhibited a significant inhibitory effect against seed germination and seedling growth of common purslane, Italian ryegrass, and barnyardgrass at a range of concentrations (0.125-1 µL/mL), as well as against Sinapsis avensis at 2 µL/mL and also Johnson grass (Sorghum halepense L.) [7][8][9]. P. sylvestris exhibited some inhibition of the early root growth of Cassia occidentalis (L.) Link. [10], and E. citriodrora essential oil affected the development of certain weeds, particularly the seed germination of Amaranthus viridis antibacterial and antifungal potential. Indeed, both pinenes have been combined with commercial antimicrobials resulting in a reduction of their minimum inhibitory concentration and toxicity [37].
Since the phytotoxic effects differ remarkably with the chemical composition and the chemical composition of an essential oil depending on certain intrinsic and extrinsic factors [38] such as the extraction method [39], geographic location [40][41][42], temperature, and drying period, as well as harvesting time [43,44], the aims of this study were: (i) to determine through Gas Chromatography-Mass Spectrometry analysis the chemical composition of commercial Eucalyptus citriodora Hook, Lavandula angustifolia Mill., and Pinus sylvestris L. essential oils; (ii) to test the in vitro phytotoxic activity of these essential oils against the seed germination and seedling growth of the weeds P. oleracea, L. multiflorum and E. crus-galli, to evaluate their herbicidal activity, as well as on food crops such as tomato (Solanum lycopersicum L.) and cucumber (Cucumis sativus L.), to know its harmful effects on crops; and (iii) to test the same against the invasive species Nicotiana glauca Graham, potential reservoir of important viruses, including cucumber mosaic virus and tomato infectious chlorosis virus, which causes economic losses for commercial tomato production.
However, qualitative and quantitative differences in the chemical composition of lavender essential oil have been reported depending on the biological raw material, level of dryness, extraction method, and origin. Previous studies showed that the drying process reduced the concentrations of the principal components in L. angustifolia essential oil obtained from flowers and aerial parts [49]. Similarly, the extraction method employed varied the content of linalool, detected in a much higher content using hydrodistillation in comparison to supercritical CO 2 and hexane extraction [50]. Furthermore, L. angustifolia essential oil hydrodistilled from aerial parts coming from Yazd (Iran) had a dissimilar chemical composition to our results, with 1,8-cineole, camphor, and borneol as its main components. This was also dissimilar to a sample from lavender essential oil obtained from the inflorescences of L. angustifolia "Sevtopolis" cultivated in western Romania, which had linalyl acetate (40.7%), linalool (22.5%), caryophyllene (8.9%), and lavandulyl acetate (7.5%) as its principal components [51]. In addition, it has been recently observed that the chemical composition of L. angustifolia essential oil can be modified by the application of gold and silver metals as elicitors, decreasing lower-molecular-weight compounds such as αand β-pinene, camphene, δ-3-carene, p-cymene, 1,8-cineole, pinocarveol, etc., which are replaced by higher-molecular-weight compounds such as α-cadinol 9-cedranone, cadalene, α-bisabolol, and (E,E)-farnesol, varying the biological properties [52]. Other presentations, such as hydrolates, produced a reduction in volatile compound content and a reduction in antioxidant activity [53].
The remarkable concentration of limonene in P. sylvestris essential oil may also contribute to the antimicrobial properties. In fact, a limonene emulsion has been effectively stabilized by Ulva fasciata Delile polysaccharide to be applied to food to avoid foodborne pathogen contamination and consequently prolong shelf-life [57,58].
In contrast to our results, sesquiterpene hydrocarbons have been described as one of the most representative phytochemical groups in the Pinus genus together with monoterpene hydrocarbons, with germacrene D or β-caryophyllene being the most characteristic compounds within the group [59]. Thus, essential oils obtained from the needles of other Pinus species such as P. roxburghaii contain large amounts of α-pinene (29.3%) and β-caryophyllene (21.9%), whereas α-pinene (35.4%) and germacrene D (28.1%) were the main components of the P. nigra subsp. nigra essential oil [36,60].
2.2. Seed Germination Inhibition of P. oleracea, L. multiflorum, E. crus-galli, Tomato, Cucumber and N. galuca with E. citriodora, L. angustifolia and P. sylvestris Essential Oils The in vitro phytotoxic potential of E. citriodora, L. angustifolia, and P. sylvestris essential oils was evaluated against seed germination in weeds (P. oleracea, L. multiflorum, and E. crus-galli), as well as against two Mediterranean food crops (tomato and cucumber), and the invasive species N. glauca, at several doses (0.125, 0.25, 0.50, and 1 µL/mL) (Tables 2 and 3).  Regarding the phytotoxic effects of the selected essential oils against weeds, variability at the intraindividual level was observed in the seed germination percentage, without statistical significance. E. citriodora and P. sylvestris did not cause a significant inhibition of seed germination in either P. oleracea, L. multiflorum, or E. crus-galli at any assayed dose (0.125, 0.25, 0.5, and 1 µL/mL) ( Table 2). However, citronellal, the main compound in E. citriodora essential oil analyzed here, showed seed germination inhibition against other weeds including Ageratum conyzoides L., Chenopodium album L., Parthenium hysterophorus L., Malvastrum coromandelianum (L.) Garcke, Cassia occidentalis L., and Philaris minor Retz. at 100 µg/g [61]. In relation to food crops, citronellal was able to inhibit seed germination in L. sativa, reaching 49-15% of the control [62], as well as seed germination in tomato at a percentage of 64.8% at the highest tested dose (1 µL/mL). P. sylvestris with α-pinene (25.6%) as the main compound showed phytotoxic effects in seed germination in cucumber at all applied doses (0.125, 0.25, 0.50, and 1 µL/mL), while another Eucalyptus species (E. tereticornis), which contained principally α-pinene (34.5%), produced selective toxicity against the seed germination of E. crus-galli without affecting the rice crop to the same extent [63].
By contrast, although L. angustifolia essential oil did not exhibit a significant inhibition of seed germination in P. oleracea, it achieved a remarkable reduction of seed germination in both L. multiflorum and E. crus-galli. This fact may be because L. angustifolia essential oil, among the essential oils analyzed here, contains the largest number (27 vs. 12 and 14) of oxygenated compounds, especially oxygenated monoterpenes (1,8-cineole, linalool, camphor, borneol, α-terpineol) that have shown higher herbicidal properties [64].
L. multiflorum showed more susceptibility to the phytotoxic effect of L. angustifolia essential oil, which decreased the percentage of seed germination in a dose-dependent manner, reaching increasing percentages of inhibition of 44.6% and 63.1% at the highest applied doses (0.5 and 1 µL/mL, respectively) ( Table 2). The fact that L. multiflorum showed a certain sensitivity to L. angustifolia essential oil could be interesting in the research of essential oils as natural alternatives to synthetic herbicides used against L. multiflorum, which have caused the emergence of resistance in this weed [65][66][67]. In other studies, peppermint (Mentha piperita L.) essential oil caused a total inhibition of seed germination in L. multiflorum at a range of concentrations between 0.125 and 1 µL/mL, and caused inhibition in food crops (maize, rice, and tomato). In our study, L. angustifolia essential oil produced less phytotoxic effects in food crops. The seed germination of tomato was reduced at the highest dose tested, at a percentage of 69.02% (vs. 99.97% with peppermint essential oil) with respect to the control [24], while the seed germination of cucumber was not significantly inhibited at any assayed dose (0.125, 0.25, 0.50, and 1 µL/mL).
Seed germination in E. crus-galli also showed a certain weakness to exposure to L. angustifolia essential oil at the highest tested dose (1 µL/mL), with a percentage of inhibition of 18.3% (Table 2).
Tomato was more sensitive to E. citriodora and L. angustifolia essential oils with similar remarkable reduction at the highest applied dose (1 µL/mL), reaching 64.8 and 69.0% reduction, respectively ( Table 2). Table 3. In vitro phytotoxic effect of different doses of E. citriodora and L. angustifolia essential oils on the seed germination and seedling growth of N. glauca. Values are the mean of five replications ± error deviation, after 14 days of incubation. Means followed by different letters in the same column indicate significantly difference at p < 0.05, according to T3 Dunnett and Tukey tests.

Germination Hypocotyl Radicle
In general, cucumber was more resistant than tomato to the phytotoxic effects of the three commercial essential oils applied, without inhibitory effect at any assayed dose (0.125, 0.25, 0.5, and 1 µL/mL) with E. citriodora and L. angustifolia essential oils, and only a low percentage of inhibition (10.00%) at the highest tested dose (1 µL/mL) with P. sylvestris essential oil (Table 2).
In addition, the two essential oils richest in oxygenated monoterpenes, E. citriodora and L. angustifolia (94.7% and 85.5%, respectively), showed similar significant phytotoxic effects against seed germination in the invasive species N. glauca, but with a lower percentage in relation to weeds (27.5% and 29.7%) at the highest tested dose (1 µL/mL) (Table 3). Therefore, the various main compounds of an essential oil can produce similar phytotoxic effects against different species.
2.3. Seedling Growth Inhibition of P. oleracea, L. multiflorum, E. crus-galli, Tomato, Cucumber and N. glauca with E. citriodora, L. angustifolia, and P. sylvestris Essential Oils The hypocotyl growth of P. oleracea was not significantly reduced by E. citriodora essential oil at any applied dose (0.125, 0.25, 0.5, and 1 µL/mL); however, this essential oil was able to reduce radicle development at the highest assayed doses (0.5 and 1 µL/mL), reaching 36.4% and 43.2% reduction compared to the control, respectively (Figure 1a). The root growth of P. oleracea was more sensitive than shoot growth to citronellal, according to previous studies, due to the mitotic activity of growing root tip cells [61]. However, other mechanisms would have been present with other essential oils because the roots were not significantly affected at doses that produced toxic effects in the hypocotyl. Therefore, the hypocotyl elongation of P. oleracea was remarkably reduced from the lowest applied dose (0.125 µL/mL) of L. angustifolia ( Figure 1b) and P. sylvestris (Figure 1c) essential oils with respect to control, reaching a decrease of 30.6% and 39.3%, respectively, at the highest tested dose (1 µL/mL), whereas radicle development was not significantly affected by L. angustifolia essential oil (Figure 1b), yet P. sylvestris essential oil achieved a significant reduction in radicle development (26.0-44.4%) from the lowest to the highest applied dose (0.125-1 µL/mL) (Figure 1c). Regarding the seedling evolution of E. crus-galli after the application of the essential oils, it was observed that L. angustifolia essential oil was the most harmful for E. crus-galli seedling growth, as it decreased its hypocotyl in a high percentage (76.7%) at the highest assayed dose (1 µL/mL), as well as the radicle, in a dose-dependent manner, also reaching a considerable percentage (69.9%) at the highest applied dose (1 µL/mL) (Figure 1b). Although E. citriodora essential oil did not influence radicle elongation at any applied dose (0.125, 0.25, 0.5, and 1 µL/mL), hypocotyl growth was significantly affected at 1 µL/mL, reaching a reduction percentage of 46.1% in comparison to control (Figure 1a). P. sylvestris essential oil was the least phytotoxic essential oil, with no reduction in the hypocotyl development of E. crus-galli at any dose (0.125, 0.25, 0.5, and 1 µL/mL), and a low percentage of radicle elongation reduction (26.5%) at the highest dose (Figure 1c). However, other studies demonstrated that α-pinene exhibited a certain inhibition of the early root growth of other weeds such as Cassia occidentalis (L.) Link., as well as oxidative damage in root tissue [10]. Similarly, the compound β-pinene was shown to be responsible for the disruption of membrane integrity, the enhancement of peroxidation and electrolyte leakage in Phalaris minor and particularly in E. crus-galli [68].
Both the hypocotyl and radicle development of L. multiflorum were significantly inhibited by E. citriodora (Figure 1a) and L. angustifolia (Figure 1b), which caused a strong dose-dependent reduction, reaching 52.3-53.0% and 60.6-75.4% at 0.25-1 µL/mL, and 55.1-77.5 and 80.1-87.8% at 0.5-1 µL/mL, respectively (Figure 1a,b). P. sylvestris essential oil did not significantly affect hypocotyl growth, but it did inhibit the radicle development of L. multiflorum in the range of 0.125 to 1 µL/mL without distinction between doses, reaching 51.67% reduction at the highest dose assayed (Figure 1c). With L. multiflorum, it was corroborated that α-pinene, the main compound of P. sylvestris essential oil analyzed here, affects root development to a greater extent than hypocotyl, as it was also able to inhibit the radicle growth of other weed species such as Amaranthus viridis L., Triticum aestivum L., Pisum sativum L., Cicer arietinum L., and especially C. occidental, which demonstrated solute leakage, lipid peroxidation and the generation of reactive oxygen species upon α-pinene exposure [10].
Regarding the sensitivity of the seedling growth of food crops to essential oils, it was observed that tomato was more susceptible than cucumber to E. citriodora, L. angustifolia, and P. sylvestris essential oils (Table 4, Figures 2 and 3). Both the hypocotyl and radicle development of tomato were significantly reduced in a dose-dependent manner, reaching elevated reduction percentages at the highest applied dose (1 µL/mL) of E. citriodora (89.7 and 79.4%) and L. angustifolia (93.2% and 83.4%) essential oils (Figure 2a,b). L. sativa was another food crop that showed high sensitivity to the application of citronellal [62], and E. citriodora essential oil affected meristematic cells, decreasing the germination and seedling growth of this food crop [25].
Again, P. sylvestris was the least phytotoxic essential oil, but also showed a significant inhibition of hypocotyl and radicle development, measuring 72.2% and 62.9%, respectively, at the dose of 1 µL/mL (Table 4).  On the other hand, none of the assayed essential oils significantly affected the hypocotyl growth of cucumber (Table 4). However, the radicle development of cucumber was significantly reduced, up to a percentage of 42.4%, 37.8%, and 28.0% at the highest applied doses of E. citriodora, L. angustifolia, and P. sylvestris essential oils (Table 4).
Finally, E. citriodora essential oil showed more phytotoxic effects than L. angustifolia essential oil in both the hypocotyl and radicle elongation of the invasive species N. glauca, reaching percentage reductions of 85.8% and 69.4% versus 51.8% and 37.6%, respectively (Table 3).  Date: 08/08/2018) essential oils obtained from the hydrodistillation of leaves, flowers, and needles, respectively, were supplied by Pranarôm S.A. (E. citriodora) and Guinama (Valencia, Spain). The essential oils were stored at 4 • C until chemical analysis and phytotoxic assays were carried out.
Mature seeds of the invasive species tree tobacco (Nicotiana glauca Graham) were supplied by the Botanical Garden of Valencia.

Gas Chromatography-Mass Spectrometry Analysis
Gas Chromatography-Mass Spectrometry analysis was carried out using a 5977A Agilent mass spectrometer and a gas chromatograph (Agilent 7890B, Valencia, España) apparatus equipped with an Agilent HP-5MS (30 m long and 0.25 mm i.d. with 0.25 µm film thickness) capillary column (95% dimethylpolysiloxane/5% diphenyl). The column temperature program was 60 • C for a duration of 5 min, with 3 • C/min increases up to 180 • C, then 20 • C/min increases up to 280 • C, which was maintained for 10 min. The carrier gas was helium at a flow rate of 1 mL/min. Split-mode injection (ratio 1:30) was employed. Mass spectra were collected over the m/z range 30-650 with an ionizing voltage of 70 eV. The resulting individual compounds were identified by MS and their identity was confirmed by comparison of their Kovat's retention index, calculated using co-chromatographed standard hydrocarbons relative to C 8 -C 32 n-alkanes and mass spectra with reference samples or with data already available in the NIST 11 mass spectral library and in the literature [45].
3.4. In Vitro Assays: P. oleracea, L. multiflorum, E. crus-galli, Tomato, Cucumber, and N. glauca Seed Germination and Seedling Growth with Essential Oils Sets of 20 seeds each with five replicates per treatment were homogenously distributed in Petri dishes (9 cm diameter) between two layers of filter paper (Whatman No.1). The lower filter papers were moistened with 4 mL of distilled water and the upper ones with 0 (control), 0.125, 0.250, 0.5, and 1 µL/mL of E. citriodora, L. angustifolia, and P. sylvestris essential oils, homogeneously distributed in the filter paper with a micropipette (Merck®, Valencia, España). Therefore, the seeds were in contact directly with moistened filter papers and indirectly with the vapors of the essential oils. Petri dishes were sealed with parafilm and incubated in an Equitec EGCS 301 3SHR model germination chamber, according to previous assays [69], alternating between 30.0 ± 0.1 • C 16 h of light and 20.0 ± 0.1 • C 8 h of darkness, with and without humidity. To evaluate the herbicidal activity of the essential oils, the number of germinated seeds was counted and compared with that of untreated seedlings. The emergence of the radicle (≥1 mm) was used as an index of germination, and seedling length (hypocotyl and/or radicle) data were recorded after 3, 5, 7, 10, and 14 days in each replicate.

Statistics
Experiments were performed in vitro with five replicates. Data were subjected to one-way analysis of variance (ANOVA) using SPSS statistics 24 software. Tukey's post hoc test was used when variances remained homogeneous (Levene's test) and T3 Dunnett's post hoc test was employed if not, assuming equal variances. Differences were considered to be significant at p ≤ 0.05.

Conclusions
In this study, the potential of E. citriodora, L. angustifolia, and P. sylvestris essential oils as eco-friendly alternatives to synthetic herbicides was investigated. L. angustifolia essential oil, with a high content of the oxygenated monoterpenes linalool (38.7 ± 0.1%), 1,8-cineole (26.5 ± 0.1%), and camphor (14.2 ± 0.1%), affected seed germination and development of L. multiflorum, E. crus-galli, and N. glauca without any significant phytotoxic effect on cucumber seed germination. E. citriodora, with a high content of the oxygenated monoterpene citronellal (88.0 ± 0.8%), showed more phytotoxic effects than L. angustifolia on the control of N. glauca. Lavender essential oil represents an effective pre-emergent treatment for L. multiflorum affecting cucumber crops, and E.citriodora essential oil could be used in both pre-and post-management of the invasive species N. glauca.