Introduction
Maximizing food production to feed an increasing human population relies heavily on the use of agrochemicals such as herbicides (Bhadoria Reference Bhadoria2011). Repeated and continuous use of herbicides has resulted in the evolution of herbicide-resistant weeds and environmental pollution (Bhadoria Reference Bhadoria2011; Dayan et al. Reference Dayan, Cantrell and Duke2009; Duke and Heap Reference Duke, Heap and Jugulam2017; Green Reference Green2014; Rosculete et al. Reference Rosculete, Bonciu, Rosculete and Olaru2019). However, growers are still heavily reliant on herbicides for weed control (Duke et al. Reference Duke, Powles and Sammons2018; Green Reference Green2014), and there is increasing criticism of maintaining agricultural production in this way. Changes in agricultural practices are deemed necessary in order to develop more sustainable and integrated agronomic practices to help human populations and agroecosystems and to preserve natural resources (Storkey et al. Reference Storkey, Bruce, McMillan, Neve, Lemaire, Kronberg, De Faccio Carvalho and Recous2019). Natural compounds or bioherbicides can be used as alternatives to synthetic herbicides (Macías et al. Reference Macías, Molinillo, Varela and Galindo2007; Storkey et al. Reference Storkey, Bruce, McMillan, Neve, Lemaire, Kronberg, De Faccio Carvalho and Recous2019; Westwood et al. Reference Westwood, Charudattan, Duke, Fennimore, Marrone, Slaughter, Swanton and Zollinger2018). They may also provide a basis for developing new compounds for weed control (Dayan and Duke Reference Dayan and Duke2014; Dayan et al. Reference Dayan, Owens and Duke2012; Yan et al. Reference Yan, Liu, Zang, Yuan, Bat-Erdene, Nguyen, Gan, Zhou, Jacobsen and Tang2018). Although natural molecules have some limitations in general weed control (Dayan et al. Reference Dayan, Owens and Duke2012), they are expected to be less toxic and more environmentally friendly than synthetic herbicides (Bhadoria Reference Bhadoria2011; Vurro et al. Reference Vurro, Miguel-Rojas and Pérez-de-Luque2019).
According to the novel weapon hypothesis, invasive plants may partially outcompete native flora in ecosystems where they invade by releasing bioactive natural chemicals (Becerra et al. Reference Becerra, Catford, Luce McLeod, Andonian, Aschehoug, Montesinos and Callaway2018; Callaway and Aschehoug Reference Callaway and Aschehoug2000; van Kleunen et al. Reference van Kleunen, Bossdorf and Dawson2018). This phenomenon, commonly referred to as allelopathy (Einhellig Reference Einhellig and Bhushan Mandava2018), also includes positive effects and makes invasive plants a source of allelochemicals or phytotoxins that could be used as bioherbicides (Benchaa et al. Reference Benchaa, Hazzit and Abdelkrim2018; Puig et al. Reference Puig, Reigosa, Valentão, Andrade and Pedrol2018). Silver wattle (Acacia dealbata Link) is an invasive plant that seriously affects ecosystem functioning and services worldwide (Aguilera et al. Reference Aguilera, Becerra, Guedes, Villaseñor-Parada, González and Hernández2015a; Lazzaro et al. Reference Lazzaro, Giuliani, Fabiani, Agnelli, Pastorelli, Lagomarsino, Benesperi, Calamassi and Foggi2014; Lorenzo et al. Reference Lorenzo, Rodríguez-Echeverría, González and Freitas2010b, Reference Lorenzo, Pazos-Malvido, Rubido-Bará, Reigosa and González2012, Reference Lorenzo, Pereira and Rodríguez-Echeverría2013, Reference Lorenzo, Rodríguez, González and Rodríguez-Echeverría2017; Ngorima and Shackleton Reference Ngorima and Shackleton2019). Allelopathy differentially contributes to the invasion process of A. dealbata (Aguilera et al. Reference Aguilera, Becerra, Guedes, Villaseñor-Parada, González and Hernández2015a, Reference Aguilera, Becerra, Villaseñor-Parada, Lorenzo, González and Hernández2015b, Reference Aguilera, Guedes, Becerra, Baeza and Hernández2015c; Lorenzo et al. Reference Lorenzo, Rodríguez, González and Rodríguez-Echeverría2017). However, this invasive species has largely shown a potential allelopathic or phytotoxic effect on different physiological parameters of plants (Aguilera et al. Reference Aguilera, Becerra, Guedes, Villaseñor-Parada, González and Hernández2015a, Reference Aguilera, Becerra, Villaseñor-Parada, Lorenzo, González and Hernández2015b, Reference Aguilera, Guedes, Becerra, Baeza and Hernández2015c; Lorenzo et al. Reference Lorenzo, Pazos-Malvido, González and Reigosa2008, Reference Lorenzo, Pazos-Malvido, Reigosa and González2010a, Reference Lorenzo, Palomera-Pérez, Reigosa and González2011, Reference Lorenzo, Reboredo-Durán, Múñoz, González, Freitas and Rodríguez-Echeverría2016, Reference Lorenzo, Souza-Alonso, Guisande-Collazo and Freitas2019; Reigosa and Carballeira Reference Reigosa and Carballeira2017a) and soil microbes (Kamutando et al. Reference Kamutando, Vikram, Kamgan-Nkuekam, Makhalanyane, Greve, Le Roux, Richardson, Cowan and Valverde2019; Lorenzo et al. Reference Lorenzo, Pereira and Rodríguez-Echeverría2013).
Recently, Souza-Alonso et al. (Reference Souza-Alonso, Rodríguez, González and Lorenzo2017) suggested using A. dealbata debris to control weeds in agriculture due to the phytotoxic effect of its plant material. Phytochemical composition of nonvolatile compounds has been reported for litter including leaves, flowers, and pods in Chile (Aguilera et al. Reference Aguilera, Becerra, Villaseñor-Parada, Lorenzo, González and Hernández2015b) and for volatiles of fresh leaves and flowers and litter in Spain (Souza-Alonso et al. Reference Souza-Alonso, González and Cavaleiro2014). Leaf litter and fresh leaves seem to be the most phytotoxic parts of A. dealbata (Aguilera et al. Reference Aguilera, Becerra, Guedes, Villaseñor-Parada, González and Hernández2015a, Reference Aguilera, Becerra, Villaseñor-Parada, Lorenzo, González and Hernández2015b; Lorenzo et al. Reference Lorenzo, Reboredo-Durán, Múñoz, González, Freitas and Rodríguez-Echeverría2016), followed by flowers and pod litter (Aguilera et al. Reference Aguilera, Becerra, Guedes, Villaseñor-Parada, González and Hernández2015a, Reference Aguilera, Becerra, Villaseñor-Parada, Lorenzo, González and Hernández2015b). However, in Spain, fresh leaves directly left on soil or incorporated into soil were rarely toxic to weeds (Souza-Alonso et al. Reference Souza-Alonso, Puig, Pedrol, Freitas, Rodríguez-Echeverría and Lorenzo2018), suggesting that natural compounds with potential herbicidal activity might be obtained from flowers or pods. Nevertheless, use of pods should be preferentially avoided, because seeds can accidentally propagate invasion. Decomposing flowers alone or combined with leaves significantly reduced the germination and radicle length of plants when incorporated into soil (Reigosa and Carballeira Reference Reigosa and Carballeira2017b). Additionally, using flowers prevents seed formation and, hence, reduces dispersion of this invasive plant.
With the idea of finding potential uses for A. dealbata flowers to prevent spread by seeds while contributing to reducing reliance on synthetic herbicides, this study aimed to explore new phytotoxic activities for nonvolatile compounds identified in flowers of A. dealbata collected in the northwestern Iberian Peninsula. First, we evaluated the phytotoxicity of identified compounds on germination and seedling growth of lettuce (Lactuca sativa L.), a model species in phytotoxic bioassays. Then, we explored the potential herbicidal effect of the most active compound on the germination process, seedling growth, and biometric and biochemical parameters of well-established plants; these were lettuce; a widely consumed wheat crop (Triticum aestivum L.); and rigid ryegrass (Lolium rigidum Gaudin), a weed common in winter cereal and wheat crops (Cirujeda and Taberner Reference Cirujeda and Taberner2009; Owen et al. Reference Owen, Goggin and Powles2015).
Material and Methods
Plant Material, Extraction Procedure, Isolation, and Identification of Chemical Compounds
Fresh flowers (7.80 kg) of A. dealbata were collected and extracted with 97% methanol as described in Lorenzo et al. (Reference Lorenzo, Reboredo-Durán, Múñoz, González, Freitas and Rodríguez-Echeverría2016) to obtain a crude extract (535.36 g). About 50 g of this extract was re-dissolved in methanol–water (2:1 v/v) and sequentially extracted with n-hexane and ethyl acetate (800 ml each). Organic solvents were removed under reduced pressure, and the remaining solution was freeze-dried in flasks under high vacuum (0.1 to 0.01 mm Hg) with a −80 C cooling trap until no ice remained inside the flasks (LyoQuest −85 Plus model, Telstar, Tarrasa, Spain). This resulted in 2.92 g of hexane extract (FH). A portion of FH (2.50 g) was subjected to column chromatography on silica gel using hexane/ethyl acetate mixtures of increasing polarity from 0% to 100% ethyl acetate. Several fractions were obtained after thin-layer chromatography analysis. One of them, fraction FH2 (828 mg) was obtained by elution with 5% and 10% ethyl acetate in hexane. FH2 was subjected to medium-pressure column chromatography on silica gel using a gradient mixture of ethyl acetate (5% to 30%) in hexane to obtain 11 fractions. From these, fractions FH2.5 (154 mg) and FH2.6 (65 mg) were subjected to high-performance liquid chromatography (column: XTerra MS C18, 5 microns, 150 by 4.6 mm; flow: 2.5 ml min−1, [Agilent, Rozas de Madrid, Spain]; eluent: hexane/ethyl acetate 95:5, isocratic; detector: UV at 220 and 254 nm). Methyl cinnamate and methyl anisate were isolated with retention times of 10 and 12 min, respectively, and were identified by 1H and 13C nuclear magnetic resonance spectrum (methyl cinnamate: Supplementary Figures S1 and S2; methyl anisate: Supplementary Figures S3 and S4). These aromatic compounds were selected based on their chemical structures to further study their potential phytotoxic activity.
Commercial formulations of methyl anisate and methyl cinnamate were purchased (Merck Millipore, Darmstadt, Germany, 99.5%) and used to conduct bioassays. Commercial formulations guaranteed available quantities of chemicals and avoided interfering effects of potential unknown chemicals present in methyl anisate and methyl cinnamate extracts.
Preliminary Bioassay: Evaluation of the Phytotoxic Effects of Methyl Cinnamate and Methyl Anisate
The phytotoxic activities of methyl anisate and methyl cinnamate were explored using the sensitive model species lettuce (‘Trocadero’) to compare results of bioassays with these two compounds (Lorenzo et al. Reference Lorenzo, Reboredo-Durán, Múñoz, González, Freitas and Rodríguez-Echeverría2016). Methyl anisate or methyl cinnamate were dissolved in dimethyl sulfoxide (DMSO) (5 µl DMSO ml−1 MES buffer) and diluted with MES buffer (10 mM 2-[N-morpholino] ethanesulfonic acid and 1 M NaOH, pH 6.0) to obtain aqueous solutions with concentrations of 10, 50, 100, 500, and 1,000 µM according to the procedure of Macías et al. (Reference Macías, Lacret, Varela, Nogueiras and Molinillo2010). Control solutions received all diluted chemicals (i.e., 5 µl DMSO ml−1 of MES buffer) except methyl anisate and methyl cinnamate (0 µM).
Twelve seeds of lettuce were sown in petri dishes (3.7-cm diameter) lined with a sterile Whatman No. 2 paper and watered with 1.2 ml of each test solution. All treatments were replicated six times. Petri dishes were sealed with Parafilm® to prevent desiccation (Lorenzo et al. Reference Lorenzo, Reboredo-Durán, Múñoz, González, Freitas and Rodríguez-Echeverría2016) and maintained at 12/12 h (light/dark) and 20 C for 7 d, with their arrangement changed daily. Then, plates were frozen at −20 C to stop seedling growth (Lorenzo et al. Reference Lorenzo, Reboredo-Durán, Múñoz, González, Freitas and Rodríguez-Echeverría2016). After that, we determined the number of germinated seeds and measured the stem and radicle lengths (cm) of all seedlings in each plate using Image J v. 1.45 software (Rasband Reference Rasband1997–2014).
Phytotoxic Effect of Methyl Cinnamate on Germination and Initial Seedling Growth: Dose–Response Assay
The potential phytotoxicity of methyl cinnamate was evaluated in terms of inhibition in seed germination and seedling growth on lettuce; L. rigidum, a problematic weed in wheat that has evolved resistance to multiple herbicidal action modes (Broster and Pratley Reference Broster and Pratley2006; Broster et al. Reference Broster, Koetz and Wu2011; Chen et al. Reference Chen, Yu, Owen, Han and Powles2018; Cirujeda and Taberner Reference Cirujeda and Taberner2009); and wheat (‘Bastide’) a widely cultivated winter cereal crop.
Twelve seeds each of lettuce, L. rigidum, and wheat were sown in petri dishes (3.7-, 6-, and 14-cm diameter, respectively) lined with a sterile Whatman No. 2 paper and watered with 0.1 ml cm−2 of different solutions of methyl cinnamate. Methyl cinnamate was dissolved as described before and assayed at different concentrations (0, 250, 375, 500, 625, 750, 875, 1,000, and 1,250 µM). Plates were sealed with Parafilm®. Six replications of each treatment were kept at 12/12 h (light/dark) and 22 C for 7 d in the case of lettuce and for 9 d in the cases of L. rigidum and wheat. The number of germinated seeds and the stem and radicle lengths (cm) were recorded as described in the previous experiment.
Phytotoxic Effect of Methyl Cinnamate on Biochemical Parameters Related to the Germination Process
Seeds of lettuce (45), L. rigidum (35), and wheat (35) were sown in petri dishes (14 cm diameter) lined with a sterile Whatman No. 2 paper and moistened with 15 ml of different solutions of methyl cinnamate. Methyl cinnamate was dissolved as described earlier and assayed at concentrations of 0, 250, 500, 750, and 1,000 µM, with five replicates per treatment. Growth conditions were the same as in the previous experiments. At 4 d after sowing, germinated seeds were frozen in liquid nitrogen and preserved at −80 C.
General Extraction Procedure
Cotyledons (100 mg) were ground with liquid nitrogen using a mortar and pestle and extracted with 2 ml of extraction buffer (0.5% polyvinilpyrrolidone, 3 mM ethylenediaminetetraacetic acid [EDTA, disodium salt 2-hydrate], and 0.1 M potassium phosphate buffer, pH 7.5). The homogenate was centrifuged at 7,500 rpm for 15 min at 4 C (Horii et al. Reference Horii, McCue and Shetty2006). The supernatant was used as the crude protein extract to determine protein concentration and enzyme activities. Samples were kept at 0 to 4 C during the process.
Protein Concentration
Total protein concentration was determined according to Bradford’s spectrophotometric assay (Bradford Reference Bradford1976) using 100 µl of supernatant mixed with 3 ml of Bradford reagent. Absorbance was measured at 595 nm after 5 min. Quantification of the protein content was determined by comparing absorbance values to a standard curve of bovine serum albumin. Protein concentration was expressed as mg g−1 fresh weight (FW).
α-Amylase Activity
This was determined using the Jones and Varner starch–iodine procedure (Jones and Varner Reference Jones and Varner1967). About 150 µl of the supernatant was diluted 1:5 (v/v) with distilled water, mixed with 1 ml of starch solution, and incubated for 10 min at 30 C. The reaction was stopped by addition of 1 ml of iodine reagent. Samples were diluted again 1:5 (ml, v/v) with distilled water, and absorbance was measured at 620 nm. One unit of α-amylase activity was defined as the amount of enzyme required to hydrolyze 1 mg starch min−1, and the results were expressed as α-amylase activity mg−1 FW.
Guaiacol Peroxidase Activity (GPX)
GPX activity was determined using the guaiacol method previously described by Horii et al. (Reference Horii, McCue and Shetty2006) and McCue et al. (Reference McCue, Zheng, Pinkhanm and Shetty2000). The reaction mixture contained the crude protein extract and buffer in a proportion of 1:5 (ml, v/v). Oxidation of guaiacol was assessed by monitoring absorbance at 470 nm over a period of 5 min. One unit of enzyme activity was defined as the amount of enzyme that oxidized 1 µmol guaiacol min−1 (U). GPX activity was expressed as U mg−1 FW.
Phytotoxic Effect of Methyl Cinnamate on Well-Established Plants
Five seeds of each species (lettuce, L. rigidum, and wheat) were sown in 100-ml pots (30.25 cm2) filled with perlite (2- to 6-mm pore size). We did not use soil to avoid interference due to soil properties that could mask the actual effect of methyl cinnamate. Pots were irrigated with 10 ml of Hoagland solution (1:1) (pH 6.25 ± 1) once a week. Additional irrigation was done when necessary. Treatments were applied when seedlings reached 3 to 5 cm in height and were completely photosynthetically active. Plants were thinned to 1 plant per pot for lettuce and wheat; and 2 plants per pot for L. rigidum. Then, plants were watered with 10 ml of different concentrations of methyl cinnamate. This compound had previously been dissolved in DMSO (5 µl DMSO ml−1 solution) and diluted with Hoagland solution to achieve final concentrations of 10, 250, 500, 750, and 1,000 µM. Control pots received the same solutions without methyl cinnamate (0 µM). Application of methyl cinnamate was repeated after 14 d. Treatments were replicated 10 times. Plants were grown in a random arrangement at 14/10 h (light/dark) and 24/22 C (light/dark) for 21 d. After that, plants were harvested, and leaf area, fresh stem and root weights, and stem and root lengths were determined for each plant. Leaf area was recorded three times on 5 separate plants with a CI-202 Portable Laser Leaf Area Meter (CID Bio-Science, Vancouver, WA, USA) for lettuce and wheat and an LI-3000c model (Li-Cor Biosciences, Bad Homburg, Germany) for L. rigidum. Then, 5 plants per treatment were randomly selected and dried at 70 C to obtain stem and root dry weights (mg), and the other 5 plants were immediately processed to evaluate biochemical parameters.
Lipid Peroxidation in Stem and Leaves
Lipid peroxidation was indirectly determined by measuring the content of malondialdehyde (MDA), a by-product of lipid peroxidation (an indicator of membrane injury), using the thiobarbituric acid (TBA) method (Hodges et al. Reference Hodges, Delong, Forney and Prange1999) with slight modifications. Fresh whole stem and leaves (0.5 g) were extracted with 9 ml of 80% ethanol and centrifuged (4,400 rpm at 4 C for 30 min). Aliquots (0.75 ml) of supernatant were added to either 1.5 ml of 20% trichloroacetic acid (TCA) and 0.001% butylated hydroxytoluene or 1.5 ml of 20% TCA and 0.001% butylated hydroxytoluene plus 0.5% TBA. Samples were mixed vigorously, heated at 95 C for 25 min, quickly cooled in ice, and centrifuged at 4,000 rpm for 15 min. The absorbance of the supernatant was measured at 440, 532, and 600 nm. MDA equivalents were determined according to Hodges et al. (Reference Hodges, Delong, Forney and Prange1999) and expressed as nmol g−1 FW.
Protein Concentration in Leaves
The extraction procedure for obtaining the crude extract from leaves and the quantification of total proteins (Bradford Reference Bradford1976) were determined as described in the previous protein concentration section.
Root Activity
This parameter was estimated using 2,3,5-triphenyltetrazolium chloride (TTC) (Onanuga et al. Reference Onanuga, Jiang and Adl2012), which is reduced to insoluble red-colored triphenyl formazan by living tissues (Ruf and Brunner Reference Ruf and Brunner2003). Fresh roots (0.05 g) were chopped into 1-mm pieces and incubated with 1.2 ml of 0.4% TTC and 1.2 ml of 0.1 M sodium-potassium phosphate buffer for 3 h at 37 C. Then, 0.75 ml of 95% ethanol was added and samples were incubated at 80 C for 15 min. The absorbance of the extract was measured at 410 nm. Triphenyl formazan content was expressed as A410 g−1 FW h−1.
Crude Extract From Roots
Fresh roots (1 g) were powdered with liquid nitrogen using a mortar and pestle and extracted with 12 ml of extraction buffer (50 mM EPES-KOH, pH 7.8, containing 0.1 mM EDTA). The mixture was centrifuged at 6,150 rpm for 15 min at 4 C, and the supernatant was preserved at −20 C until use.
Protein Concentration in Roots
Protein concentration was determined according to Bradford’s spectrophotometric assay (Bradford Reference Bradford1976) as indicated in the previous section using 100 µl of crude root extract.
Superoxide Dismutase Activity (SOD) in Roots
SOD activity was evaluated by monitoring the inhibition of photochemical reduction of nitro blue tetrazolium as described by Beauchamp and Fridovich (Reference Beauchamp and Fridovich1971) using 0.2 ml of crude root extract mixed with 2 ml of nitro blue tetrazolium solution. Samples were acclimated under 40-W fluorescent lamps for 3 min at room temperature. Reaction was started by adding 500 µl of 2 µM riboflavin. Then, samples were lit by 40-W fluorescent lamps for 10 min at room temperature. Absorbance was measured at 560 nm. One unit of SOD activity (U) was defined as the amount of enzyme that causes 50% inhibition of nitro blue tetrazolium, and the results were expressed as U mg−1 FW.
Data Analyses
Data were analyzed separately for each species. We conducted general linear models (LMs) or generalized linear models (GLMs) to test for the concentration effect of methyl cinnamate or methyl anisate (only in the preliminary bioassay) on germination and stem and radicle length (preliminary, germination, and early growth bioassays); protein concentration, α-amylase activity, and GPX activity (germination process bioassays); and stem increment, root and stem biomass, foliar area, MDA concentration, protein concentration, TTC concentration, and SOD activity (well-established plant bioassays). The assumption of normality was assessed using the Shapiro-Wilk test. If response variables were normal, we conducted LMs, while GLMs with the appropriate error family and link function were used in the absence of data normality (Supplementary Tables S1–S4). We also checked for normality of each LM or GLM residual using the Shapiro-Wilk test. When a response variable was nonnormal and normality of GLM residuals could not be achieved through error family structure, we conducted a nonparametric Kruskal-Wallis (KW) test (Supplementary Table S4). Post hoc mean separations were conducted using Tukey’s HSD by comparing the least-squares means obtained within each LM and GLM or by the Nemenyi test after a KW test. The LMs, GLMs, and KW tests were conducted using the stats package, while post hoc comparisons were performed with the lsmeans and multcomp or stats packages after LMs and GLMs or KW, respectively. All analyses were conducted in R v. 3.1.1 (R Development Core Team 2015). The level of significance was set at P ≤ 0.05 for all analyses.
Dose–response models plotting growth response against compound concentration were modeled by nonlinear regression curves to calculate IC50 and IC80values (concentrations that cause 50% and 80% of inhibition, respectively). The most appropriate dose–response curve for each case was selected according to the best regression coefficient (R2). Model adjustment and R2 value (goodness of fit) were obtained by using scatter plots from the Excel program (Office 16). In L. rigidum, there were very few stem and radicle length records at 1,250 µM. Therefore, this concentration was removed from the IC50 and IC80 estimation for this species.
Results and Discussion
Preliminary Bioassay: Evaluation of the Phytotoxic Effects of Methyl Cinnamate and Methyl Anisate
Natural compounds are increasingly in demand to replace synthetic chemicals that cause environmental problems and human health concerns (Bhadoria Reference Bhadoria2011; Dayan et al. Reference Dayan, Cantrell and Duke2009; Katz and Baltz Reference Katz and Baltz2016). In the present study, we evaluated the potential herbicidal activity of the commercial formulations of methyl cinnamate and methyl anisate. We found these compounds in A. dealbata flowers (methyl cinnamate: Supplementary Figures S1 and S2; methyl anisate: Supplementary Figures S3 and S4), and their herbicidal activity has not been broadly explored in the literature. Methyl anisate was previously identified in flower litter of A. dealbata in the Chilean range (Aguilera et al. Reference Aguilera, Becerra, Villaseñor-Parada, Lorenzo, González and Hernández2015b). However, methyl cinnamate was not found in A. dealbata plants from Chile (Aguilera et al. Reference Aguilera, Becerra, Villaseñor-Parada, Lorenzo, González and Hernández2015b) or those from the Iberian Peninsula (Souza-Alonso et al. Reference Souza-Alonso, González and Cavaleiro2014). Methyl cinnamate showed a broadly antifungal activity (Lima et al. Reference Lima, Ferreira, Silva, Lima and de Sousa2018; Prakash et al. Reference Prakash, Mishra, Kedia, Dwivedy and Dubey2015), inhibited bacterial growth (Malheiro et al. Reference Malheiro, Maillard, Borges and Simões2019), had a larvicidal effect (Fujiwara et al. Reference Fujiwara, Annies, de Oliveira, Lara, Gabriel, Betim, Nadal, Farago, Dias, Miguel, Miguel, Marques and Zanin2017), and, furthermore, exibited some potential to inhibit plant growth (Fujiwara et al. Reference Fujiwara, Annies, de Oliveira, Lara, Gabriel, Betim, Nadal, Farago, Dias, Miguel, Miguel, Marques and Zanin2017; Khanh et al. Reference Khanh, Cong, Xuan, Lee, Kong and Chung2008). Essential oils from plants containing a large percentage of methyl cinnamte also showed high bioactivity (Mar et al. Reference Mar, Silva, Azevedo, França, Goes, dos Santos, de A Barreza, de Cássia, Nunomura, Machado and Sanches2018; Noriega et al. Reference Noriega, Mosquera, Paredes, Parra, Zappia, Herrera, Villegas and Osorio2018). In addition, methyl cinnamate is considered as an alternative to synthetic chemicals because it is safer (Fujiwara et al. Reference Fujiwara, Annies, de Oliveira, Lara, Gabriel, Betim, Nadal, Farago, Dias, Miguel, Miguel, Marques and Zanin2017; Prakash et al. Reference Prakash, Mishra, Kedia, Dwivedy and Dubey2015). On the other hand, there is scarce literature reporting the phytotoxic effect of methyl anisate.
Our results demonstrated that the commercial formulations of methyl cinnamate and methyl anisate did not affect germination of lettuce (Supplementary Table S1). However, both compounds affected the stem and radicle length of lettuce (Figure 1; Supplementary Table S1). Dose–response curves showed that methyl cinnamate achieved the lowest IC50 and IC80 values of 844.57 and 1053.54 µM for stem length, respectively, and of 372.57 and 660.60 µM for radicle length, respectively (Figure 1). Methyl cinnamate stimulated the radicle length at the lowest concentration (10 µM), but reduced both stem (up to 75%) and radicle (up to 96%) growth of lettuce at the highest concentrations (500 and 1,000 µM) (Figure 1). However, methyl anisate showed a lower inhibitory effect on stem length (up to 15%) at 500 and 1,000 µM and on the radicle length at 1,000 µM (46%), although it also increased the radicle length at 10 and 50 µM (Figure 1). These results suggest that methyl cinnamate had higher potential phytotoxic activity than methyl anisate. Therefore, methyl cinnamate was selected to test for a putative bioherbicidal effect on different plant processes and parameters in subsequent bioassays.
Figure 1. Preliminary bioassay. Percentage values with respect to the control for the stem and radicle lengths of lettuce seedlings in response to application of the commercial formulations of methyl cinnamate and methyl anisate. On the y axis, dashed lines indicate control values. Bars are means ± SE; n = 6. Asterisks indicate statistical significance between concentrations and the control treatment according to Tukey’s test after general or generalized linear models: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.
The concentration of methyl cinnamate in the extract obtained from A. dealbata flowers was 16 mg kg−1 FW (i.e., 0.0016% w/w). This concentration is lower than that found for other natural compounds in studies conducted to find potential bioherbicides (Pardo-Muras et al. Reference Pardo-Muras, Puig, López-Nogueira, Cavaleiro and Pedrol2018; Takemura et al. Reference Takemura, Kamo, Sakuno, Hiradate and Fujii2013). Such a small quantity of methyl cinnamate could compromise its potential role as a bioherbicide. However, a single individual of A. dealbata displays a massive production of flowers (Correia et al. Reference Correia, Castro, Ferrero, Crisóstomo and Rodríguez-Echeverría2014). Generally, A. dealbata forms dense invasive populations occupying large areas (Souza-Alonso et al. Reference Souza-Alonso, González and Cavaleiro2014) that provide large quantities of flowers, which could make it easier to obtain enough methyl cinnamate for practical application.
Phytotoxic Effect of Methyl Cinnamate on Germination and Initial Seedling Growth: Dose–Response Assay
Methyl cinnamate inhibited germination of lettuce and L. rigidum at 1,250 µM (Table 1; Supplementary Table S2), whereas no significant effect was found for wheat (Table 1; Supplementary Table S2). Regarding seedling growth, methyl cinnamate reduced stem and radicle length of lettuce at 375 µM and higher concentrations and of L. rigidum at ≥875 µM (Figure 2; Supplementary Table S2), whereas the stem and radicle length of wheat were only inhibited at 1,000 µM (Figure 2; Supplementary Table S2). In lettuce, nonlinear dose–response curves showed that concentrations of methyl cinnamate that inhibited stem and radicle length by 50% (IC50) were 1,121.68 and 176.29 µM, respectively (Figure 2). In L. rigidum, the IC50 values were established at 900.15 and 780.41 µM for stem and radicle length, respectively (Figure 2). In wheat, the IC50 values were 1,048.69 and 1,472.05 (out of range) µM for stem and radicle growth, respectively (Figure 2). The IC80 values obtained for stem and radicle length of each plant species are also shown in Figure 2. Our results partially agree with those obtained in previous studies. Methyl cinnamate was found to inhibit germination and growth of lettuce at 0.1% (Fujiwara et al. Reference Fujiwara, Annies, de Oliveira, Lara, Gabriel, Betim, Nadal, Farago, Dias, Miguel, Miguel, Marques and Zanin2017) and slightly reduced germination and root length on L. rigidum at 640 nl cm−3 (Vasilakoglou et al. Reference Vasilakoglou, Dhima, Paschalidis and Ritzoulis2013), although the effect was dependent on concentration in both cases. However, low concentrations such as 100 ppm reduced radicle and shoot of radish (Raphanus sp.) (Khanh et al. Reference Khanh, Cong, Xuan, Lee, Kong and Chung2008).
Table 1. Effect of the commercial formulation of methyl cinnamate on germination of lettuce, Lolium rigidum, and wheat.
a Means ± SE are shown. n = 6. Asterisks (*) indicate statistical significance between concentrations and the control treatment according to Tukey’s test after general or generalized linear models.
Figure 2. Effects on initial seedling growth. Nonlinear dose–response curves for the stem and radicle lengths of lettuce, Lolium rigidum, and wheat seedlings in response to application of the commercial formulation of methyl cinnamate. On the y axis, dashed lines indicate control values. The y axis shows a different scale for each species. In L. rigidum, stem and radicle length records at 1,250 µM were very low. Therefore, this concentration was removed from the IC50 and IC80 estimation for this species. Bars are means ± SE; n = 6. Asterisks indicate statistical significance between concentrations and the control treatment according to Tukey’s test after general or generalized linear models: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001. IC50 and IC80 indicate the concentration values of methyl cinnamate that cause 50% and 80% of inhibition, respectively.
Phytotoxic Effect of Methyl Cinnamate on Biochemical Parameters during the Germination Process
Methyl cinnamate affected α-amylase and GPX activities in lettuce and GPX in L. rigidum and had a marginal effect on GPX in wheat during the germination process compared with the control treatment (Figure 3; Supplementary Table S3). In lettuce, α-amylase was reduced at 750 and 1,000 µM, whereas GPX was stimulated at low concentrations (250 µM) and severely inhibited at high concentrations (750 to 1,000 µM) (Figure 3). In addition, methyl cinnamate caused a significant severe reduction in protein concentration when assayed at 250 and 500 µM (Figure 3). In L. rigidum, the GPX activity was inhibited at 750 to 1,000 µM (Figure 3). However, in wheat, methyl cinnamate only marginally stimulated the GPX activity at 250 µM (Figure 3). Our results demonstrated that methyl cinnamate did not seem to affect the number of total germinated seeds of the species assayed, except at 1,250 µM. However, lower concentrations such as 450 and 1,000 µM of this compound reduced α-amylase activity in lettuce and GPX activity in lettuce and L. rigidum during the germination process. α-Amylase activity hydrolyzes starch into sugars that are essential not only for embryonic development but also for maintaining the water potential during seed imbibition (Doria Reference Doria2010; Taiz and Zeiger Reference Taiz and Zeiger2006). There is strong evidence that peroxides and radicals are abundantly produced within seeds during germination (Bailly Reference Bailly2004) and stressful conditions (Sharma et al. Reference Sharma, Jha, Dubey and Pessarakli2012). Therefore, GPX enzymes are expected to intensify their activity in seed germination as a result of the accumulation of reactive oxygen species produced by by-products of mitochondrial respiration. This is in concordance with our results: the observed increased GPX activity at low methyl cinnamate concentrations protected the germination process. The decrease in α-amylase and GPX activities observed at the highest concentrations may result in reduced seed germination processes.
Figure 3. Effects on parameters related to the germination process. Mean ± SE values for guaiacol peroxidase activity, α-amylase activity, and protein concentration in lettuce, Lolium rigidum, and wheat seeds in response to the application of the commercial formulation of methyl cinnamate. The y axis on the right shows a different scale for each species. n = 5. Asterisks indicate statistical significance between concentrations and the control treatment according to Tukey’s test after general or generalized linear models or to Nemenyi’s test after Kruskal-Wallis analyses: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001. FW, fresh weight.
Phytotoxic Effect of Methyl Cinnamate on Well-Established Plants
Before the different concentrations of methyl cinnamate were applied, the initial stem length was recorded in each plant, and there were no significant differences for this parameter among plants assigned to each treatment within each species (Supplementary Table S4).
In lettuce, the increment in stem length was significantly and severely inhibited (from 76% to 96%) by methyl cinnamate at 500, 750, and 1,000 µM (Figure 4; Supplementary Table S4). The concentration of 1,000 µM also reduced stem and root biomass (Table 2; Supplementary Table S4). However, all tested concentrations of methyl cinnamate reduced both leaf area (Table 2; Supplementary Table S4) and root activity (TTC content) (Table 3; Supplementary Table S4). In addition, the content of MDA in stems was increased at 1,000 µM, whereas the concentration of total proteins in roots was reduced at the same concentration (Table 3; Supplementary Table S4). Methyl cinnamate did not affect root length, protein concentration in stems, or SOD in roots (Figure 4; Table 3; Supplementary Table S4).
Figure 4. Effects on well-established plants. Mean ± SE values for the stem increment and root length of well-established lettuce, Lolium rigidum, and wheat plants in response to application of the commercial formulation of methyl cinnamate. The y axis shows a different scale for each species. n = 5. Asterisks indicate statistical significance between concentrations and the control treatment according to Tukey’s test after general or generalized linear models or to Nemenyi’s test after Kruskal-Wallis analyses: **, P ≤ 0.01; ***, P ≤ 0.001.
Table 2. Effects of the commercial formulation of methyl cinnamate on stem biomass, root biomass, and foliar area of well-established lettuce, Lolium rigidum, and wheat plants.
aMeans ± SE are shown. n = 5. Asterisks (*) indicate statistical significance between concentrations and the control treatment according to Tukey’s test after general or generalized linear models.
Table 3. Effects of the commercial formulation of methyl cinnamate on malondialdehyde concentration (MDA) and protein concentration in stems and on superoxide dismutase activity (SOD), triphenyltetrazolium chloride (TTC), and proteins in roots of well-established lettuce, Lolium rigidum, and wheat plants.a
a A, absorbance; FW, fresh weight.
b Means ± SE are shown. n = 5. Asterisks (*) indicate statistical significance between concentrations and the control treatment according to Tukey’s test after general or generalized linear models or to Nemenyi’s test after Kruskal-Wallis analyses.
In L. rigidum, methyl cinnamate reduced stem growth and root activity at 1,000 µM (Figure 4; Table 3; Supplementary Table S4) and concentration of root proteins at 250 and 1,000 µM (Table 3; Supplementary Table S4) and increased stem proteins at 500 µM (Table 3; Supplementary Table S4). There were no significant differences in the remaining parameters evaluated (Figure 4; Tables 2 and 3; Supplementary Table S4).
In wheat, methyl cinnamate stimulated stem growth at 250 µM (Figure 4; Supplementary Table S4) and reduced root SOD activity at 500 and 1,000 µM (Table 3; Supplementary Table S4). However, this chemical compound did not alter any other parameter measured (Figure 4; Tables 2 and 3; Supplementary Table S4).
Comparing the Phytotoxic Effect of Methyl Cinnamate between Seedlings and Well-Established Plants
The phytotoxic effect of methyl cinnamate is more likely to be seen on 0- to 9-d-old seedlings (preliminary bioassay, dose–response assay, and germination process assay) than on older plants (well-established plant bioassay). At the seedling stage, growth was reduced in L. rigidum and severely inhibited in lettuce. This effect was supported by the IC50 values, which indicated that 50% of seedlings of these two species were negatively affected by methyl cinnamate, and IC50 values were within the assayed range. However, the effect of methyl cinnamate on older plants was only evident for lettuce, with L. rigidum and wheat being only marginally affected. Parameters such as lipid peroxidation (MDA) and SOD are related to oxidative stress (e.g., Weir et al. Reference Weir, Park and Vivanco2004; Yadav and Singh Reference Yadav and Singh2013). In our study, methyl cinnamate did not affect MDA and barely influenced SOD in stem and root biomass of well-developed plants. This may indicate that well-established plants treated with this phytotoxin are not very stressed. The fact that the germination-related enzymes and seedling stage were more affected by methyl cinnamate than older plants may suggest a potential PRE herbicidal effect instead of POST activity. So, as an example, although the highest concentrations assayed (750 and 1,000 µM) did not affect the total number of germinated seeds, these concentrations highly reduced germination-related enzymes (guaiacol peroxidase activity and α-amylase activity in lettuce; guaiacol peroxidase activity in L. rigidum). This may result in an anomalous germination process that leads to inviable seedlings.
Although the effect of methyl cinnamate on well-established plants can be considered marginal in terms of plant biomass, root activity evaluated by the reducing capacity of TTC was affected in two of the three assayed species (lettuce and L. rigidum). This parameter is related to aerobic respiration in roots, which is fundamental to proper functioning in sugar regulation, mineral absorption, and water uptake in plants (Onanuga et al. Reference Onanuga, Jiang and Adl2012; Wang et al. Reference Wang, Inukai and Yamauchi2006). In our study, we cannot conclude whether the effect on root activity of L. rigidum finally resulted in depletion of plant growth as observed for lettuce. However, this effect deserves further evaluation, because L. rigidum has largely evolved resistance to several herbicides (Broster and Pratley Reference Broster and Pratley2006; Broster et al. Reference Broster, Koetz and Wu2011; Chen et al. Reference Chen, Yu, Owen, Han and Powles2018; Cirujeda and Taberner Reference Cirujeda and Taberner2009).
Phytotoxic Effect of Methyl Cinnamate and the Type of Target Species
Previous studies found that methyl cinnamate inhibited germination or growth of both monocotyledonous and dicotyledonous with small seed size such as radish (dicot) (Khanh et al. Reference Khanh, Cong, Xuan, Lee, Kong and Chung2008), L. rigidum (monocot) (Vasilakoglou et al. Reference Vasilakoglou, Dhima, Paschalidis and Ritzoulis2013), and lettuce (dicot) (Fujiwara et al. Reference Fujiwara, Annies, de Oliveira, Lara, Gabriel, Betim, Nadal, Farago, Dias, Miguel, Miguel, Marques and Zanin2017), although the effect may not be compared between species due to different concentrations used. However, this phytotoxin did not affect the germination of chick pea (Cicer arietinum L.) (Ramirez et al. Reference Ramirez, Cendoya, Nichea, Zachetti and Chulze2018), a dicot with large seed size. In our study, lettuce was the most affected by methyl cinnamate at all evaluated stages, followed by L. rigidum, which was affected during the germination process and initial growth at higher concentrations. Wheat (monocot with medium-sized seed), however, was negligibly affected by this compound. This may suggest a potential herbicidal effect of methyl cinnamate on both monocot and dicot with small seeds, but it is ineffective for species with large ones. The concentrations of methyl cinnamate that had a negative effect on L. rigidum did not reduce wheat growth. This suggests that methyl cinnamate could potentially be used as a selective herbicide for L. rigidum control in wheat. However, further studies on the germination process and early growth are required to evaluate the herbicide potential of methyl cinnamate under field conditions.
In conclusion, methyl cinnamate was more phytotoxic than methyl anisate. Our results showed that the application of methyl cinnamate might be more effective during the germination process and early growth of both monocot and dicot species with small seeds. Methyl cinnamate could potentially be used as a selective herbicide for L. rigidum control in wheat. However, further studies are required to assess the effectivity of methyl cinnamate under field conditions. The inhibitory effect of methyl cinnamate on seeds and seedlings may indicate a putative herbicidal effect of this compound and a potential use for A. dealbata flowers, contributing to the management of this invasive species.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/wsc.2019.68
Acknowledgments
This work was funded by the Portuguese Foundation for Science and Technology (FCT) through national funds and cofunding from the European Regional Development Fund, within the PT2020 Partnership Agreement, and COMPETE 2020, within project UID/BIA/04004/2013; additional funds were provided by Xunta de Galicia, Spain (CITACA Strategic Partnership, reference ED431E 2018/07). PL was supported by FCT and the European Social Fund (grant SFRH/BPD/88504/2012; contract IT057-18-7248). We especially thank Susana Rodríguez-Echeverría for her helpful comments on the experimental design. We thank two anonymous reviewers who improved this article with their comments.
Conflicts of Interest
No conflicts of interest have been declared.
