Previous PageTable Of ContentsNext Page

Searching for a natural herbicide: the role of medicinal plants?

Sally Allan and Steve Adkins

School of Land and Food Science, University of Queensland, St Lucia Qld 4072.
Email ,


The development of a natural herbicide has the potential to reduce detrimental environmental impacts and the evolution of herbicide-resistant weed populations. This research aims to explore the herbicide potential of naturally-occurring plant chemicals, allelochemicals, which can inhibit plant growth. Screening of 6 potentially allelopathic plant species – Acacia farnesiana, Ageratum conyzoides, Alphitonia excelsa, Castanospermum australe, Chamaesyce hyssopifolia and Melaleuca quinquenervia has been conducted using a Lemna (Lemna aequinoctialis) bioassay method. L. aequinoctialis were grown in a nutrient medium amended with aqueous extracts (2 % w/v) produced from each species plant parts (viz. leaf, stem, bark, root, etc.) over a 7 day period. Photographs taken at Day 0 and Day 7 enabled the measurement of Lemna growth using Scion computer software. All 6 species caused inhibition of the growth of L. aequinoctialis with the strongest inhibition for each species produced by leaf extract of C. hyssopifolia (9 % growth), leaf extract of A. conyzoides (12 % growth), bark extract of M. quinquenervia (27 % growth), seed pod extract of C. australe (31 % growth), seed extract of A. farnesiana (32 % growth) and leaf extract of A. excelsa (67 % growth). From these initial observations, the stronger growth inhibitors will be chosen to proceed to the next stage of testing.

Media summary

Testing the herbicide potential of naturally-occurring chemicals found in some medicinal plants.


Aboriginal, cultural medicine, medicinal


The cost and impact of weed management in Australia is significant. A heavy reliance on chemical methods to target weeds has resulted in the evolution of herbicide-resistant weed populations and a negative perception of the toxic affect herbicides have on our community and environment. The development of a natural chemical herbicide is one method of addressing these issues. The structural complexities and specificity of natural chemicals should reduce the risk of herbicide-resistance and the relatively short half-life of such chemicals could reduce risk to health and the environment (Duke et al. 1999).

This research project will investigate the herbicide potential of several allelopathic plant extracts on common weed species, aiming to develop a natural herbicide. After initial selection of possible allelopathic donor plants, the experimental programme will be conducted in three stages: (i) stage –1, the aqueous plant extracts will be screened using a Lemna bioassay to identify extracts, which produce the strongest growth inhibition; (ii) stage -2, the identified extracts will be used in germination trials and glasshouse pot culture on several weed species; and (iii) stage –3, the most promising extracts will undergo chemical isolation to identify active ingredients.

Stage-1 screening of plant extracts has begun using Lemna aequinoctialis bioassays. Lemna sp. are small monocotyledonous angiosperms composed of two or more, generally flat fronds, several mm in size and a single root. These aquatic plants are free-floating and can quickly reproduce via vegetative propagation, with each frond having the ability to produce daughter fronds. The growing, genetically homogeneous, Lemna population spreads across the surface of the water and therefore, growth can be measured by calculating the surface area of the duckweed (Wang 1990).

To date, 6 plant species – Acacia farnesiana, Ageratum conyzoides, Alphitonia excelsa, Castanospermum australe, Chamaesyce hyssopifolia and Melaleuca quinquenervia have been tested using the Lemna bioassay method. All 6 species share the common trait of being used, or in the case of C. australe, having the potential to be used, for medicinal purposes by various cultures. Many medicinal plants contain an array of natural chemicals and have often been researched by pharmaceutical companies, but not necessarily assessed for other uses, including herbicide potential.

Ageratum conyzoides has been used as a medicinal plant by many cultures, is known to contain many secondary metabolites (Kong et al. 2004b) and has been the subject of a number of allelopathic studies (Bhatt et al. 2001; Dongre et al. 2004; Kong et al. 2004a; Kong et al. 2004b; Xuan et al. 2004). This species was chosen for this study as it has been shown to be capable of strong allelopathic interactions and so can act as a cross-referencing tool between experiments and to verify the experimental procedure.

Most chemical studies on Acacia farnesiana have focussed on the nutritional value of the seeds which contain phenols, tannins and other secondary metabolites (Devi and Prasad 1991). The plant has also been used medicinally in Mexico for the treatment of diarrhoea and inflammation (Meckes et al. 2004).

Australian Aboriginals used the saponin-rich leaves of Alphitonia excelsa for a multitude of purposes (upset stomach, sore eyes, headaches, bites and stings and washing). Along with berries, they could be used to poison or stupefy fish and extracts of bark and roots could relieve muscular pain or toothache (Cribb and Cribb 1982; Low 1990; Pearson and Pearson 1992; Symons and Symons 1996).

Recent pharmacological studies on Castanospermum australe has identified the alkaloid castanospermine, which is showing promise in combating cancer and acquired immunodeficiency syndrome (AIDS) (Low 1990).

The herb, Chamaescye hyssopifolia has been used as a diuretic and emmenagogue in Central America and the Caribbean (Morton 1981) and extracts of this plant have more recently been identified as being inhibitory to human immunodeficiency virus (HIV). From water and methanol extracts, 11 potentially active compounds have been isolated (Lim et al. 1997).

In Thialand, Melaleuca quinquenervia has been used for the treatment of gastrointestinal problems. Leaf extracts inhibit the growth of the Helicobacter pylori bacterium which is responsible for causing gastritis, dyspepsia and peptic ulcer (Bhamarapravati et al. 2003). Australian Aboriginals also chewed or crushed leaves of M. quinquenervia to treat head colds (Symons and Symons 1996). Chemicals isolated from the leaves include 4 polyphenolic acid derivatives and 3 ellagitannins (Moharram et al. 2003).


Aqueous Extract

A representative sample of extract donor plant was collected early in the morning following at least 72 hours without rainfall. Samples were then gently washed to remove foreign material and dried for 30 minutes in an air-conditioned room (22C). During this time, plants were investigated for obvious abnormalities, like areas of insect/pathogen attack and such sections were discarded accordingly. Each plant was then divided into sections - leaves or leaf-like structures, stems, roots and flowers/fruit /seed. Stems and roots were cut into lengths of <1 cm, leaves cut into areas <1 cm2 and flowers/fruit/seed finely chopped. Plant parts were weighed and placed in glass jars with deionised water at a concentration of 25% w/v. Jars were sealed with Parafilm “M” Laboratory Film, wrapped in aluminium foil and stored at 22C for 24 hours.

After storage, the liquid portion was strained through a 0.5 mm mesh. The strained extract was then centrifuged at 20 000 rpm for 20 minutes at 20C. The extract was then either used immediately or stored at 4C for up to 7 days for further use.


Lemna aequinoctialis was purchase from Mappins Nursery, Kenmore, Qld, Australia. A stock culture was kept in the laboratory (22C) in a diluted (50%) nutrient medium. Plants selected from the stock culture for pre-testing were free of obvious disease or morphological deformity. Each plant was of roughly uniform size, was a healthy green colour and had a root >0.5 cm long. These plants were placed in nutrient medium (100%) in a controlled environment with a temperature of 26C and photoperiod of 8 hours darkness and 16 hours light conditions (six 18-W Cool White fluorescent tubes emitting ~ 80 μmol m-2 s-1 of photosynthetically active radiation). After 7 days acclimatisation, the rapidly growing plants were then selected for testing of the plant extract.

Ten plants were selected based on the criteria outlined above, with each plant comprised of two larger fronds and one smaller frond, in the L-3 or R-3 stages described by (Datko et al. 1980). These were placed in a square Petri dish with 46 ml of nutrient solution (100%) and 4 ml of 25% w/v extract to give a final 2% w/v concentration. Four replicates for each treatment in a random block design were then placed back in the controlled environment, using the same conditions outlined above.

Data Analysis

A digital photograph of each Petri dish was taken at the beginning of each trial (Day 0) and at Day 7. Photographs were cropped along the Petri dish boundary to create a 500x500 pixel TIFF image. This image was then converted into a scaled black on white image and the area of plant growth (black) calculated using Scion Image for Windows. The growth area increase was then calculated over the 7 day trial period. The mean growth increase was analysed via a one-way ANOVA and Fisher LSD test (P<0.05). A growth % was calculated by comparing the mean growth increase (cm2) for each treatment against a control growth (equal to 100%) for each plant species (Table 1).


The control treatment had an average five-fold increase in growth over the 7-day experimental period.

All extracts from A. conyzoides significantly inhibited the growth of L. aequinoctialis with mild inhibition caused by the root and flower extracts (78% and 67% growth, respectively), moderate inhibition by the stem extract (37% growth) and severe inhibition by the leaf extract (12% growth). Extracts from C. australe, A. farnesiana and M. quinquenervia caused moderate but significant inhibition with the most inhibited growths from each species being 31% (seed pod), 32% (seed) and 27% (bark), respectively.

The only extract of A. excelsa, to result in a significant but moderate, inhibition of growth was caused by the root extract (56%). Interestingly, a slight stimulation of L. aequinoctialis growth in the bark extract (116%) occurred, although statistically not significant as compared to the corresponding control. The observed increase in variance within treatments for this species has possibly merged treatment results and further work may be warranted.

The most promising growth results came from C. hyssopifolia extracts, which were all significantly lower than the corresponding control. The root extract resulted in moderate inhibition (50% growth) while both the fruit and combined leaf-stem extract produced reductions in growth to 17% and 9%, respectively.

Table 1. Effect of aqueous extract of different plant parts of 6 plant species on the growth (%) of Lemna aequinoctialis. For any one row, growth % results with different letters indicate significant (P<0.05) differences.


The aim of stage –1 testing is to screen 20 to 30 plant species and to highlight a few species showing the strongest inhibition to plant growth. These strong inhibiting species will be selected to continue to the next level of more specific and time-consuming trials in stage - 2. All 6 plant species tested show evidence of allelopathic interaction with L. aequinoctialis, and of them A. conyzoides and C. hyssopifolia exhibited the strongest allelopathic value. These two species will be tested in stage – 2, with A. conyzoides mainly to be used as a reference tool. The other four species will be re-assessed when all of stage – 1 experiments are completed.

The results from A. conyzoides are similar to those found by other researchers mentioned above and act to verify the experimental procedure adopted. In this study, the leaf extract exhibited stronger inhibition of plant growth than the stem and root extracts, consistent with findings of Xuan et al. (2004). At a concentration of 20 gL-1 (equal to 2% w/v) Xuan et al., found the leaf extract had 55.7 % inhibition on the germination of radish and produced 58.3 % dry weight compared to the control. These inhibition levels are not as strong as those found in this study, however other factors can influence results, i.e. the susceptibility of Lemna sp. and radish to inhibiting chemicals, increased involvement of environmental factors with germination and growth trials and the variability of allelochemical production (Kong et al. 2004a) by the donor species.

The most significant results came from the species - C. hyssopifolia. The limited published research on C. hyssopifolia has revolved around the therapeutic potential of this plant. The detailed information on the chemicals isolated from this species (Lim et al. 1997) may prove valuable as further testing of the allelopathic capabilities of this plant are conducted.


Bhamarapravati S, Pendland SL, Mahady GB (2003) Extracts of spice and food plants from Thai traditional medicine inhibit the growth of the human carcinogen Helicobacter pylori. In Vivo 17, 541-544

Bhatt BP, Tomar JMS, Misra LK (2001) Allelopathic effects of weeds on germination and growth of legumes and cereal crops of North Eastern Himalayas. Allelopathy Journal 8, 225-231.

Cribb AB, Cribb JW (1982) 'Useful wild plants.' (Fontana Books)

Datko AH, Mudd SH, Giovanelli J (1980) Lemna-Paucicostata Hegelm 6746 - Life-Cycle and Characterization of the Colony Types in a Population. Plant Physiology 65, 913-923.

Devi SR, Prasad MNV (1991) Tannins and related polyphenols from ten common Acacia spp. of India. Bioresource Technology 36, 189-192.

Dongre PN, Singh AK, Chaube KS (2004) Allelopathic effects of weed leaf leachates on seed germination of blackgram (Phaseolus mungo L.). Allelopathy Journal 14, 65-70.

Duke SO, Fedayan FE, Romagni JG, Rimando AM (1999) Natural Products as sources of herbicides: current status and future trends. Weed Research 40, 99-111.

Kong CH, Hu F, Liang WJ, Peng W, Jiang Y (2004a) Allelopathic potential of Ageratum conyzoides at various growth stages in different habitats. Allelopathy Journal 13, 233-240.

Kong CH, Hu F, Xu XH, Liang WJ, Zhang CX (2004b) Allelopathic plants. XV. Ageratum conyzoides L. Allelopathy Journal 14, 1-12.

Lim YA, Mei MC, Kusumoto IT, Miyashiro H, Hattori M (1997) HIV-1 reverse transcriptase inhibitory principles from Chamaesyce hyssopifolia. Phytotherapy Research 11, 22-27.

Low T (1990) 'Bush Medicine: a pharmacopoeia of natural remedies' (Collins/Angus & Robertson: North Ryde, Australia).

Meckes M, David-Rivera AD, Nava-Aguilar, Jimenez A (2004) Activity of some Mexican medicinal plant extracts on carrageenan-induced rat paw edema. Phytomedicine 11, 446-451.

Moharram FA, Marzouk MS, El-Toumy SAA, Ahmed AAE, Aboutabl EA (2003) Polyphenols of Melaleuca quinquenervia leaves - Pharmacological studies of grandinin. Phytotherapy Research 17, 767-773.

Morton JF (1981) 'Atlas of medicinal plants of Middle America.' (Charlse C. Thomas: Springfield)

Pearson S, Pearson A (1992) 'Rainforest plants of eastern Australia.' (Kangaroo Press: Sydney, Australia)

Symons P, Symons S (1996) 'Bush heritage.' (Pat and Sim Symons: Nambour, Australia)

Wang W (1990) Literature review on duckweed toxicity testing. Environmental Research 52, 7-22.

Xuan TD, Shinkichi T, Hong NH, Khanh TD, Min CI (2004) Assessment of phytotoxic action of Ageratum conyzoides L. (billy goat weed) on weeds. Crop Protection 23, 915-922.

Previous PageTop Of PageNext Page