1 Introduction

Allelopathy is the action of some chemicals (allelochemicals) exuded by living plants that directly inhibits or stimulates the growth and development of other plants; it is considered to be one ecological strategy for weed management. Allelopathic plants have a role in controlling weeds through the release of allelochemicals, which have the potential to suppress weed growth and development (Rice 1984). A number of plants, including crops, weeds and trees, have been reported to possess allelopathic properties, and extensive studies have reported that approximately 240 species of weeds have allelopathic potential (Qasem and Foy 2001). More recently, some weed species such as Rumex maritimus, Hyptis suaveolens, and Paspalum commersonii have been found to have allelopathic potential, and a number of allelochemicals such as 2-methoxystypandrone, suaveolic acid, and loliolide reportedly also inhibit the growth of other plants (Islam et al. 2014, 2017; Zaman et al. 2018).

The weed species Eleocharis atropurpurea (Retz.), belonging to the sedge family Cyperaceae, is a small, annual tufted weed widely distributed in tropical, subtropical, and temperate regions (Sarwar and Prodhan 2011; Mishra et al. 2017). Eleocharis atropurpurea has a slender hollow stem and can reach up to 10 cm. It is a wetland species and grows abundantly in river beds, marshes, shallow water, and irrigated ditches (Yadav and Sardesai 2002; Kumar 2014), and also grows in rice fields as a competitive weed (Huda et al. 2017). The competitive behavior of this species suggests the allelopathic potential of E. atropurpurea. There is evidence of the allelopathic potential of some species of the genus Eleocharis in reducing the growth of target species. Ashton et al. (1985) found that polar E. coloradoensis extracts reduced the growth of some troublesome aquatic weeds such as Hydrilla verticillata and Potamogeton pectinatus. In another study, E. cellulosa and E. interstincta were also found to have allelopathic effects on the growth of H. verticillata (Sutton and Portier 1991). Therefore, we hypothesized that the extracts of E. atropurpurea may contain substances and the substances can produce significant inhibitory effects on plants. However, as far as can be determined, no allelochemicals or allelopathic properties have yet been found in E. atropurpurea. Therefore, we conducted experiments to explore the allelopathic potential of E. atropurpurea and to determine if it produces allelochemicals.

2 Material and methods

2.1 Plant material and bioassay test species

Leaves and stems of Eleocharis atropurpurea (Retz.) during their vegetative growth stage were collected in July 2015 near the fields of Bangladesh Agricultural University, Bangladesh. The plants were cleaned with tap water, dried in the shade, and ground into a fine powder. The ground powder was kept at 2 °C for further experiments.

Five test plant species were used for the bioassay experiments: Lepidium sativum L., Medicago sativa L., Lolium multiflorum Lam., Phleum pratense L., and Echinochloa crus-galli (L.) P. Beauv. Two crop species (Lepidium sativum L. and Medicago sativa L.) were chosen because of their known growth behavior, and three common weed species (Lolium multiflorum Lam., Phleum pratense L., and Echinochloa crus-galli (L.) P. Beauv) were selected from crop fields.

2.2 Extraction

Dried E. atropurpurea powder (leaves and stems, 2.0 kg) was extracted with aqueous methanol (15 L; 70% v/v) for 48 h and filtered through one layer of filter paper (No. 2, 125 mm; Toyo Ltd., Tokyo, Japan). The residue was again extracted with methanol (15 L; 100%) for 24 h. The obtained aqueous methanol and methanol extracts were combined and concentrated on a rotary evaporator at 40 °C. A flowchart of the extraction procedure is outlined in Fig. 1.

Fig. 1
figure 1

Procedure for the extraction and isolation of trans-ferulaldehyde from Eleocharis atropurpurea leaves and stems (for details of the procedure, see the material and methods section)

Full size image

2.3 Bioassay

Each bioassay was conducted at 0.001, 0.003, 0.01, 0.03, 0.1, and 0.3 g dry weight equivalent extract mL−1 of E. atropurpurea extract. A control without plant extracts was also carried out. Methanol extract aliquots (0.3, 0.9, 3, 9, 30, and 90 µL, respectively), corresponding to the treatment concentrations were added to Petri dishes (28 mm in diameter) with a single layer of filter paper (No. 2, 28 mm; Toyo Ltd., Tokyo, Japan), which was then dried in a fume chamber and moistened with 0.6 mL of 0.05% (v/v) aqueous solution of polyoxyethylene sorbitan monolaurate (Tween-20; Nacalai Tesque, Inc., Kyoto, Japan). Seeds (n = 10) of Lepidium sativum, Medicago sativa, Lolium multiflorum, and Phleum pratense were placed in the Petri dishes with six replications. Lolium multiflorum and Phleum pratense seeds sprouted overnight by soaking in distilled water and then allowed to germinate in darkness at 25 °C for 60 and 66 h, respectively. All treatments were then transferred to a growth chamber at 25 °C. After 2 days, the shoot and root length of all the test plant species were measured, and the seedling length percentage was calculated by referring to the length of the control seedlings.

2.4 Isolation and identification of the active substance

Eleocharis atropurpurea plant powder (leaves and stems, 2.0 kg) was extracted as described above and a flowchart of the isolation procedure of the active substance is outlined in Fig. 1.

The concentrated residue was then adjusted to pH 7.0 with NaOH (1 N) and partitioned eight times with an equal volume of ethyl acetate, which was then separated into ethyl acetate and aqueous fractions. The ethyl acetate fraction was filtered and evaporated after being soaked with anhydrous Na2SO4 and subjected to a column of silica gel (60 g, silica gel 60, spherical, 70-230 mesh; Nacalai Tesque, Inc., Kyoto, Japan). The column was eluted with increasing quantities of ethyl acetate in n-hexane (10% step−1, v/v; 150 mL step−1) and methanol (300 mL) to obtain nine fractions (F1–F9). Each fraction’s biological activity was determined using a cress bioassay as described above, and the activity was found by eluting with 60% ethyl acetate in n-hexane (F5 fraction). The active fractions obtained by silica gel column were then evaporated and purified using a column of Sephadex LH-20 (GE Healthcare Bio-Sciences AB, SE-751 84 Uppsala, Sweden) and eluted with 20, 40, 60, and 80% (v/v) aqueous methanol (150 mL step−1) and methanol (300 mL). The fraction eluted with 80% (v/v) aqueous methanol was active and evaporated to dryness. The active fraction residue was then dissolved in 20% aqueous methanol and loaded into a reverse-phase C18 cartridge (YMC Co. Ltd., Kyoto, Japan) and eluted with 20, 40, 60, and 80% (v/v) aqueous methanol and methanol. The active fractions obtained from 40% aqueous methanol were then evaporated to dryness and purified using reverse phase HPLC (500 × 10 mm I.D. ODS AQ-325; YMC Co., Ltd., Kyoto, Japan) at a flow rate of 1.5 mL min−1 with 35% aqueous methanol and detected at a wavelength of 220 nm. The active peaks were eluted at retention times of 116–120 min. Finally, the active peak fraction was purified again by reverse phase HPLC (4.6 × 250 mm I.D, S-5 µm, Inertsil® ODS-3; GL Science Inc., Tokyo, Japan) at a flow rate of 0.8 mL min−1 with 20% aqueous methanol and detected at 220 nm wavelength and 40 °C in the oven. The active peak was detected at retention times of 82–100 min, and the substances were characterized by using ESIMS, 1H-NMR (Fig. 4).

2.5 Bioassay of the isolated compound

The isolated compounds were dissolved in 1 mL methanol to prepare the final assay concentrations: 0.03, 0.1, 0.3, 1.0, 3.0, and 5.0 mM and added to Petri dishes (28 mm in diameter) with a single layer of filter paper. The filter paper was dried in a fume chamber and then moistened with 0.6 mL of 0.05% (v/v) aqueous solution of polyoxyethylene sorbitan monolaurate. Ten seeds of Lepidium sativum and ten sprouted seeds (the seeds were soaked overnight in distilled water and then allowed to germinate in darkness at 25 °C for 48 h) of Echinochloa crus-galli were placed in the Petri dishes and transferred to a growth chamber at 25 °C. The shoot and root lengths were measured after 2 days, and the seedling percentage length was calculated by referring to the length of the control seedlings.

2.6 Data analysis

The bioassay experiment was repeated twice with three replications. The data was analysed using ANOVA with SPSS software, version 16.0 (SPSS Inc., Chicago, Illinois, USA). Significant differences between treatments and control were examined using Tukey’s HSD test at a 0.05 probability level. The I50 values of each test plant species were calculated using a regression equation of the concentration response curves. Regression analysis was used to determine the correlation coefficient (r) values.

3 Results

3.1 Inhibitory activity of E. atropurpurea aqueous methanol extracts on the growth of L. sativum, M. sativa, L. multiflorum, and P. pratense

The E. atropurpurea aqueous methanol extracts inhibited the growth of L. sativum, M. sativa, L. multiflorum, and P. pratense (Fig. 2). The concentration of 0.1 g dry weight equivalent extract mL−1 completely inhibited the shoot and root length of L. sativum. At 0.3 g dry weight equivalent extract mL−1, the shoot length of M. sativa, L. multiflorum, and P. pratense was inhibited by 0.5, 5.5, and 0.4% of the control shoot length, respectively; the same concentration inhibited the root length of M. sativa, L. multiflorum, and P. pratense by 0.3, 2.1, and 0.4% of the control root length, respectively. In addition, a significant negative correlation was found between the E. atropurpurea extract concentration and seedling growth of the test plants, with correlation coefficient values in the range − 0.88 to − 0.99 (Table 1). The shoot and root length of P. pratense showed a higher correlation (r = − 0.99 and − 0.96, respectively) with the E. atropurpurea extracts.

Fig. 2
figure 2

Effect of Eleocharis atropurpurea extracts on the shoot and root growth of Lepidium sativum, Medicago sativa, Lolium multiflorum, and Phleum pratense. All the test plant species were treated at concentrations of 0.001, 0.003, 0.01, 0.03, 0.1, and 0.3 g dry weight equivalent extract of E. atropurpurea mL−1. Root and shoot length of all test plant species were measured after 2 days of incubation in growth chamber at 25 °C. Mean ± SE from two independent experiments with three replications for each treatment (number of seedlings per treatment = 10, n = 60). Each vertical bar represents standard error of the mean. For each species, different letters indicate significant difference according to Tukey’s HSD test at the 0.05 probability level

Full size image
Table 1 Correlation coefficients between extract concentration and the shoot and root length of the test plants, and compound concentration and the shoot and root length of Lepidium sativum and Echinochloa crus-galli
Full size table

The I50 values of the extracts of E. atropurpurea for the growth of the test plants ranged from 2.90 to 96.8 mg dry weight equivalent extract mL−1 (Table 2). On the basis of the I50-values, the root length of P. pratense was more sensitive to the E. atropurpurea extracts, and the shoot length of L. multiflorum was less sensitive to the E. atropurpurea extracts.

Table 2 Concentration required for 50% inhibition (I50) of the shoot and root growth of the test plant species for the aqueous methanol extracts and the compound trans-ferulaldehyde from Eleocharis atropurpurea
Full size table

3.2 Isolation and identification of the structure of the active substance

The extracts of E. atropurpurea were purified on a silica gel column and the biological activity of each separated fraction was determined by cress bioassay at 0.9 g dry weight equivalent extract mL−1 (Fig. 3). The active fraction 5 inhibited the growth of the cress shoots and roots by 4.8 and 2.3% of control growth, respectively (Fig. 3). The active fraction was further purified through a chromatography series and one active substance was finally isolated using HPLC. The structure of the isolated substance was identified by comparing with previous reported spectral data.

Fig. 3
figure 3

Effect of the fractions obtained from a silica gel column on the shoot and root length of Lepidium sativa by Eleocharis atropurpurea extracts at the concentration 0.9 g dry weight equivalent extract mL−1. The column was eluted with increasing amounts of ethyl acetate in n-hexane: F1 (20% ethyl acetate in n-hexane), F2 (30% ethyl acetate in n-hexane), F3 (40% ethyl acetate in n-hexane), F4 (50% ethyl acetate in n-hexane), F5 (60% ethyl acetate in n-hexane), F6 (70% ethyl acetate in n-hexane), F7 (80% ethyl acetate in n-hexane), F8 (ethyl acetate), F9 (methanol). Mean ± SE from three replications with 10 seeds for each treatment. Each vertical bar represents standard error of the mean. The different letters in the same panel indicate significant difference according to Tukey’s HSD test at the 0.05 probability level

Full size image

The molecular formula of the compound was determined as C10H10O3 by means of ESIMS at m/z 177.0562 [M-H]− (calcd for C10H9O3, 177.0552, Δ = + 1.0 mmu). The 1H NMR (400 MHz, CD3OD) spectrum of the substance showed δH 9.57 (d, J = 8.0 Hz, 1 H, H1), 7.59 (d, J = 16.0 Hz, 1 H, H3), 7.26 (d, J = 2.1 Hz, 1 H, H5), 7.17 (dd, J = 8.5, 2.1 Hz, 1 H, H9), 6.85 (d, J = 8.5 Hz, 1 H, H8), 6.66 (dd, J = 16.0, 8.0 Hz, 1 H, H2) and 3.91 (s, 3 H, H10). The substance was identified as trans-ferulaldehyde (Fig. 4) by comparing its spectral data with previously reported data in the literature (Carpinella et al. 2003).

Fig. 4
figure 4

Chromatogram of the active compound obtained from a C18 cartridge: 40% aqueous methanol purified by reverse phase HPLC, eluted at a flow rate of 0.8 mL min−1 with 20% aqueous methanol, detected at 220 nm wavelength. Arrow indicates the chemical structure of the active compound (trans-ferulaldehyde)

Full size image

3.3 Inhibitory activity of trans-ferulaldehyde on the growth of L. sativum and E. crus-galli

The inhibitory activity of the isolated compound trans-ferulaldehyde was tested against the test plants using the L. sativum and E. crus-galli bioassay. The trans-ferulaldehyde compound significantly inhibited the shoot length and root length of L. sativum at concentrations higher than 1.0 and 3.0 mM, respectively (Fig. 5). On the other hand, trans-ferulaldehyde significantly inhibited the shoot length and root length of E. crus-galli at concentrations greater than 3.0 mM (Fig. 5). At 5 mM, trans-ferulaldehyde completely inhibited the shoot and root length of L. sativum, whereas at the same concentration, trans-ferulaldehyde inhibited the shoot and root length of E. crus-galli by 33.7 and 6.9% of the control shoot and root length, respectively. There was a significant negative correlation between compound concentration and the shoot and root length of L. sativum and E. crus-galli, with correlation coefficient values ranging from − 0.93 to − 0.98 and − 0.88 to − 0.89, respectively (Table 1). The I50 values of trans-ferulaldehyde for the growth of all the test plant species ranged from 0.73 to 3.68 mM (Table 2).

Fig. 5
figure 5

Effect of trans-ferulaldehyde on the shoot and root growth of Lepidium sativum and Echinochloa crus-galli. Mean ± SE from two independent experiments with 10 seeds for each treatment. Each vertical bar represents standard error of the mean. The different letters in the same panel indicate significant difference according to Tukey’s HSD test at the 0.05 probability level

Full size image

4 Discussion

The aqueous methanol extracts of E. atropurpurea inhibited the growth of the four test plant species (Fig. 2). The extract concentration and the shoot and root length of L. sativum, M. sativa, L. multiflorum, and P. pratense also showed a strong negative correlation (Table 1). Additionally, the I50 values of the tested plants indicated differences in the shoot and root growth among the test plant species (Table 2). The results show that the inhibitory activity of the E. atropurpurea extracts was concentration- and species-dependent and that the extracts may contain allelopathic substances.

One substance was isolated from the extracts of E. atropurpurea and characterized as trans-ferulaldehyde (Fig. 4), a cinnamaldehyde derivative phenolic, which possesses a benzene ring where one methoxy group is attached to the benzene ring (C-6 position) with a phenol moiety (Radnai et al. 2009; Chen et al. 2011). The active substance trans-ferulaldehyde and its substitutes (ferulaldehyde, trans-coniferylaldehyde, E-coniferyl aldehyde) have been found in many plant extracts such as Quercus suber (Conde et al. 1998), Melia azedarach L. (Carpinella et al. 2003), and Coix lachryma-jobi L. var. ma-yuen Stapf (Chen et al. 2011) but had never been isolated from E. atropurpurea.

Both L. sativum and E. crus-galli were inhibited by trans-ferulaldehyde (Fig. 5). Additionally, a negative correlation was also found between compound concentration and the shoot and root length of L. sativum and E. crus-galli (Table 1). The shoot and root length of L. sativum showed a higher correlation coefficient (r = − 0.98 and − 0.93, respectively) with compound concentration compared with the shoot and root length of E. crus-galli. These results indicate that the inhibitory activity of the compound was concentration-dependent. Studies have shown that trans-ferulaldehyde has antifungal (Carpinella et al. 2003), anti-inflammatory (Radnai et al. 2009), antimutagenic (Chen et al. 2011), and antioxidant (Chen et al. 2015) activities. However, to the best of our knowledge, this is the first study to show that trans-ferulaldehyde is an allelopathic substance with growth inhibitory activity on test plants.

The I50 values of trans-ferulaldehyde for the shoots and roots of L. sativum were 0.73 and 1.51 mM, respectively, whereas the values for the shoots and roots of E. crus-galli were 3.68 and 1.92 mM, respectively (Table 2). Comparing I50 values, L. sativum was more sensitive to trans-ferulaldehyde than was E. crus-galli, and the effectiveness of trans-ferulaldehyde on the shoots and roots of L. sativum was 5.04- and 1.27-times higher than on the E. crus-galli shoots and roots. Reports have noted that α, β-unsaturated aldehyde moiety (i.e., CH=CH–CHO) is important for biological activities (Chen et al. 2011); trans-ferulaldehyde contains α, β-unsaturated aldehyde moiety in its structure, which may be a reason for the inhibitory activity of trans-ferulaldehyde. Additionally, ferulaldehyde is the reduced precursor of ferulic acid (Achterholt et al. 1998; Tucsek et al. 2011), which has already been reported as an inhibitory substance (Singh 2014). Ferulic acid and reduced ferulaldehyde have a similar structure, with the only difference being the functional group of ferulaldehyde. The reactive aldehyde group in ferulaldehyde can be easily oxidized to a carboxylic group, which is also thought to show the biological activity in the same manner as ferulic acid (Veres 2012). Therefore, the aldehyde group in trans-ferulaldehyde may be another reason for its inhibitory activity.

The allelopathic substances in plants can be released into the soil by leaf leaching and/or by exudation from the living plant tissues (Chon et al. 2006; Sangeetha and Baskar 2015). It has been reported that allelopathic substances leached from the foliage can interfere with the growth and development of other plants. Nakano et al. (2003) reported that the allelopathic substance leached from the foliage of Prosopis juliflora inhibited the root growth of barnyard grass and may play an important role in the allelopathy. Therefore, it is possible that the isolated compound, trans-ferulaldehyde, may be leached from the foliage of E. atropurpurea by rainfalls and then accumulated in the soil, and possibly act as an allopathically active substance. In addition, our study supports the hypothesis that E. atropurpurea contains a growth inhibitory substance, trans-ferulaldehyde, and the substance may contribute to its inhibitory effects which may help develop an effective and sustainable weed management strategy.

In conclusion, the aqueous methanol extracts of E. atropurpurea showed concentration- and species-dependent inhibitory activity on the seedling growth of the test plant species. An active substance was isolated from the extracts of E. atropurpurea, and its chemical structure was determined as trans-ferulaldehyde by ESIMS and 1H NMR. The trans-ferulaldehyde compound displayed concentration-dependent inhibitory activity on the growth of L. sativum and E. crus-galli. The results suggest that trans-ferulaldehyde may contribute to the growth inhibitory activity of E. atropurpurea, and the allelopathy of the E. atropurpurea weed may play an important role in developing alternative weed management strategies.