The Effect of Cultivation and Plant Age on the Pharmacological Activity of Merwilla Natalensis Bulbs

The effect of cultivation and plant age on the pharmacological activity of Merwilla natalensis bulbs

SG Sparg, AK Jäger and J van Staden*

Research Centre for Plant Growth and Development, University of KwaZulu-Natal Pietermaritzburg, Private Bag X01, Scottsville 3209, South Africa * Corresponding author, e-mail: [email protected]

Received 5 March 2004, accepted in revised form 27 August 2004

Merwilla natalensis bulbs were cultivated at two thesis by COX-1, with activity decreasing as the bulbs different sites under different treatments. Bulbs were matured. The cultivation treatments had a significant harvested every six months for a period of two years effect on the antihelmintic activity of bulbs cultivated at and were tested for antibacterial, anti-inflammatory and the Fort Hare site. Results suggest that irrigation might anthelmintic activity. The cultivation treatments had no increase the antihelmintic activity of the bulbs if culti- significant effect (P ≤ 0.05) on neither the antibacterial vated in areas of low rainfall. The age of the bulbs at activity, nor the anti-inflammatory activity. However, the both sites had a significant effect on the antihelmintic age of the bulbs had a significant effect against the test activity, with activity increasing with plant maturity. bacteria and on the inhibition of prostaglandin syn-

Introduction

Merwilla natalensis (Planchon) Speta (synonym Scilla varied when the plants were grown in different geographical natalensis) is ranked as one of the more commonly-sold regions. The concentrations of compounds also varied medicinal plants at many of the informal medicinal markets seasonally. in South Africa (Cunningham 1988, Mander 1997). The Very little research has been done on seasonal variation bulbs are used for a wide range of ailments such as gastro- in pharmacological activity or the effect that plant growth intestinal disorders, sprains and fractures, boils and sores, (age) has on the pharmacological activity. In a study by Light internal tumours and even gynaecological disorders (Watt et al. (2002) the effect of senescence in Siphonochilus and Breyer-Brandwijk 1962, Hutchings 1989, Van Wyk et al. aethiopicus on its pharmacological activity was investigated. 1997). In a previous study (Sparg et al. 2002), bulb extracts Taylor and Van Staden (2001) investigated the effect of age, of M. natalensis were screened for pharmacological activity season and growth conditions on the anti-inflammatory in vitro. The results showed M. natalensis to have anti- activity of Eucomis autumnalis plant extracts. Significant inflammatory, anthelmintic, antischistosomal and antibac- differences in activity were found between plants harvested terial activity, although the latter was not very significant. before and after the growing season. The effect of cold According to Bernath (1986), ecological factors have storage on the anti-inflammatory activity of E. autumnalis decisive impacts on the chemical composition of plants. autumnalis also showed differences in activity (Taylor and Therefore, the variability in natural product content and Van Staden 2002). composition are functions of both growth and development, In this study, the effects of different cultivation treatments and play a role in determining the optimal harvest time of and age of the bulbs on the pharmacological activity of M. certain species. Prance (1994) suggested that the chemistry natalensis was investigated. within individual plant species varies according to the ecology, the soil and the climate. Yaniv and Palevitch (1982) Materials and Methods stated that, depending on the species, the quantity of secondary metabolites increases, decreases, or remains Cultivation and harvesting constant when under water stress. Seasonal variation of alkaloids in Crinum moorei was studied by Elgorashi et al. Two sites in South Africa were selected to cultivate tissue- (2002). No clear trend in seasonal variation was noted, with cultured M. natalensis bulbs as a crop on a small-scale alkaloid yields varying from organ to organ as well as from basis. The two sites selected were the smallholding H2A season to season. Dong et al. (2003) showed that the active Botanicals (Hillcrest, KwaZulu-Natal) and premises at the constituents of Panax notoginseng, including saponins, University of Fort Hare (Alice, Eastern Cape). 192 Sparg, Jäger and Van Staden

Cultivation at each site was carried out on three plots. The eight replicates for each treatment for each harvest. first plot was the control treatment, which received no Indomethacin (20µM) was used as a positive control. fertiliser, additives or supplementary irrigation. The plants were grown in tilled topsoil and relied on rain for their water Anthelmintic bioassay source. The second plot was only fertilised and not irrigated

(referred to as the fertiliser treatment). In the case of the H2A The nematode Caenorhabditis elegans var. Bristol (N2) was Botanicals site, the fertiliser used for the second treatment cultured on nematode growth agar seeded with E. coli was a mixture of composted horse and chicken manure at a according to the method of Brenner (1974). The anthelmintic ratio of 2:1, whereas inorganic fertiliser, NPK 232(22) assay used was that of Rasoanaivo and Ratsimamanga- (Nitrochem), was used for the second treatment at the Fort Urverg (1993) as described by McGaw et al. (2000). Hare site. The third plot was irrigated and fertilised (referred Methanol extracts were in ethanol and tested at a final to as the irrigation treatment). The plants were planted concentration of 1mg m–1 for anthelmintic testing. Four approximately 30cm apart in rows with 50cm spacing extracts per treatment per harvest were tested. One between rows. replicate of each of the extracts was performed. The assay Bulbs were harvested every six months from each of the was repeated a second time giving a total of eight replicates three treatments (20 bulbs per treatment per harvest period) for each treatment for each harvest. Levamisole (5µg ml–1 ) over a period of two years. The harvested bulbs were dried and a solvent blank were used as positive and negative at 50°C for 72h. Bulbs were divided into groups of fives to controls respectively. Using a dissecting microscope, the form four sample groups. The five bulbs were ground percentage worm mortality was estimated and compared together and stored in airtight containers in the dark at with the controls. ±25°C (room temperature). This was done for each of the treatments for each harvest from both the H2A Botanicals Statistical analysis and Fort Hare sites. The results from each assay were analysed statistically Extract preparations using analysis of variance (ANOVA) to determine interactions and the effect the treatments and bulb age had Methanol and dichloromethane extracts were prepared for on the pharmacological activity. Comparisons were made each of the four samples for each treatment of each harvest. using Fisher’s pairwise comparisons at a 5% level of Two grams of dried material were extracted with 20ml significance. extraction solvent. The material was allowed to extract for 24h with constant agitation. The extracts were filtered under Results and Discussion vacuum through Whatman No. 1 filter paper discs using a Büchner funnel. The filtered extracts were taken to dryness Minimum inhibitory concentrations (MIC) were calculated for and stored in sealed pill vials in the dark at 5°C. an average of eight replicates for each treatment of each harvest. The lower the MIC value, the higher the Antibacterial bioassay antibacterial activity. The results from the antibacterial

screening of bulbs cultivated at the H2A Botanicals site are The microdilution antibacterial assay as described by Eloff summarised in Table 1. No significant differences (P ≤ 0.05) (1998) was used to screen the extracts for antibacterial in activity between treatments against both the Gram- activity. Methanol extracts were dissolved in ethanol to positive bacteria were observed. Twelve month-old bulbs concentrations of 50mg ml–1 for use in the assay. Four harvested from the fertiliser treatment gave the highest extracts per treatment per harvest were tested. The assay activity (3.9mg ml–1) against B. subtilis. Bulbs harvested at was repeated a second time, giving a total of eight replicates six months from the fertiliser treatment had the highest for each treatment for each harvest. The extracts were activity (2.34mg ml–1) against S. aureus. Analysis of the tested against two Gram-positive bacteria, Bacillus subtilis results showed no significant interactions between the (ATCC 6051) and Staphylococcus aureus (ATCC 12600), treatments and bulb age on the antibacterial activity against and two Gram-negative bacteria, Escherichia coli (ATCC both B. subtilis and S. aureus. Both the treatments and the 11775) and Klebsiella pneumoniae (ATCC 13883). bulb age had no significant effect (P ≤ 0.05) on the Neomycin and solvent blanks were used as positive and antibacterial activity when tested against B. subtilis. negative controls respectively. However, the bulb age had a significant effect on the antibacterial activity (P = 0.002) against S. aureus, with the Anti-inflammatory bioassay antibacterial activity decreasing as the plants matured. The six month-old bulbs had significantly higher antibacterial Extracts were screened using the COX-1 assay as activity against S. aureus than the 24 month-old bulbs. The described by Jäger et al. (1996). The dichloromethane screening against the Gram-negative bacteria did not show extracts were dissolved in ethanol to a concentration of a similar pattern to the results for the Gram-positive bacteria. 10mg ml–1 (final test concentration = 250µg ml–1) for anti- There were significant differences (P ≤ 0.05) in activity inflammatory testing. Four extracts per treatment per between treatments. Both the treatments and the age of the harvest were tested. One replicate of each of the extracts bulbs had a significant effect on the antibacterial activity was performed. The assay was repeated, giving a total of against E. coli (P = 0.008 and P = 0.001 respectively). The South African Journal of Botany 2005, 71(2): 191–196 193

Table 1: Minimum inhibitory concentration (MIC) values of methanol M. natalensis bulb extracts against test bacteria. Values include SEM for bulbs which were cultivated at the H2A Botanicals site

Bulb age Treatment MIC (mg ml–1) against different bacterial strains (months) B. subtilis S. aureus E. coli K. pneumoniae 6 Control 5.86 ± 1.11 3.91 ± 1.45 3.52 ± 1.10 3.52 ± 1.11 Fertiliser 5.47 ± 1.45 2.34 ± 0.84 2.93 ± 0.55 2.93 ± 0.55 Irrigation 5.86 ± 1.45 3.52 ± 1.82 2.93 ± 0.55 2.73 ± 0.73 12 Control 5.86 ± 1.11 5.47 ± 1.45 3.52 ± 1.11 3.13 ± 0.00 Fertiliser 3.91 ± 1.45 2.54 ± 0.81 3.13 ± 0.00 3.13 ± 0.00 Irrigation 6.25 ± 2.89 3.91 ± 1.45 3.13 ± 0.00 2.93 ± 0.55 18 Control 7.03 ± 2.21 5.86 ± 1.11 4.69 ± 1.67 4.69 ± 1.67 Fertiliser 6.25 ± 0.00 5.08 ± 1.62 3.91 ± 1.45 3.13 ± 0.00 Irrigation 6.25 ± 0.00 5.08 ± 1.62 3.91 ± 1.45 5.08 ± 1.62 24 Control 6.25 ± 0.00 7.03 ± 2.21 5.47 ± 1.45 5.47 ± 1.45 Fertiliser 6.25 ± 0.00 7.03 ± 2.21 3.91 ± 1.45 3.13 ± 0.00 Irrigation 6.25 ± 0.00 7.03 ± 2.21 4.69 ± 1.67 3.13 ± 0.00 Neomycin 1.60 x 10–2 3.10 x 10–2 6.30 x 10–2 1.60 x 10–2 treatments and the age of the bulbs had no significant effect age of the bulbs had no significant effect (P ≤ 0.05) on the on the activity when tested against K. pneumoniae. Bulbs antibacterial activity. On average, extracts were more active harvested from the fertiliser treatment were more active against K. pneumoniae than against E. coli. The highest against E. coli than the bulbs from the other treatments. The activity was recorded with the six month- and 12 month-old six month-old bulbs from the fertiliser treatment had the bulbs harvested from the fertiliser treatment, with MIC’s of highest activity (MIC = 2.73mg ml–1). The antibacterial 2.73mg ml–1 and 2.54mg ml–1 respectively. activity decreased as the plants matured. Six month-old In the study by Light et al. (2002) it was concluded that the bulbs harvested from the irrigation treatment had the highest time of harvest of Siphonochilus aethiopicus only had a activity against K. pneumoniae (MIC = 2.73mg ml–1). There minimal influence (two-fold difference) on the degree of was no interaction between the treatments and the age of antibacterial activity. the bulbs on the antibacterial activity against E. coli. Although the results of this study showed a four-fold However, there was a significant interaction (P ≤ 0.05) decrease in activity, the extracts were tested at relatively between the treatment and the age of the bulbs on the high concentrations and the activity observed overall is not antibacterial activity against K. pneumoniae (P < 0.001). particularly noteworthy. Since the MIC assay uses a two-fold The antibacterial results from bulbs cultivated at Fort Hare serial dilution method, the sensitivity of the assay is not are summarised in Table 2. There were no significant necessarily accurate. The extracts from mature bulbs had differences (P ≤ 0.05) in activity between treatments against slightly lower MIC values when screened for antibacterial both Gram-positive test bacteria. The six month-old bulbs activity (Sparg et al. 2002) than the young cultivated bulbs. harvested from the fertiliser treatment had the highest Results of this study show a significant decrease (P ≤ 0.05) activity against B. subtilis with an average MIC value of in antibacterial activity with an increase in age with M. 4.69mg ml–1. The age of the bulbs had a significant effect on natalensis. However, the antibacterial activity shown by the the activity against both B. subtilis (P = 0.002) and S. aureus mature bulbs suggests that the activity increases once the (P < 0.001). However, there was no interaction between the bulbs have reached maturity. As was the case with the treatments and the age of the bulbs. The bulbs harvested antibacterial screening by Sparg et al. (2002), extracts from from the fertiliser treatment and irrigation treatment had the M. natalensis are generally more active against Gram- highest activity (MIC = 2.15mg ml–1) against S. aureus. The negative bacteria. In many of the antibacterial studies that Gram-negative bacteria followed similar trends in activity to have been conducted with plant extracts, it is the Gram- the results from the Gram-positive bacteria. The treatments positive bacteria that are more sensitive than Gram- had no significant effect (P ≤ 0.05) on the antibacterial negative. This is due to differences in the cell wall structure activity. There were no significant differences between the of the bacteria (Page et al. 1997). MIC values of the six month-old bulbs and the 12 month-old Results of COX-1 screening are represented as mean bulbs tested against E. coli, but the MIC values of both were percentage inhibition, calculated as the mean of eight significantly different from both the 18 month-old bulbs and replicates (Figure 1). The indomethacin control (5µM) had the 24 month-old bulbs. Analysis of these results revealed 59% COX-1 inhibition. Extracts (250µg ml–1) of both young ≤ that bulb age had a significant effect (P 0.05) on the and older bulbs cultivated at H2A Botanicals had high COX- antibacterial activity against E. coli (P < 0.001). Bulbs 1 inhibition, with percentage inhibition ranging from 88% to harvested from the irrigation treatment after six months had 91% (Figure 1a). No significant differences (P ≤ 0.05) in the highest activity against E. coli (MIC = 2.34mg ml–1). No COX-1 inhibition activity were found between bulbs real trends were noticeable from the results of K. cultivated under different treatments and of different ages. pneumoniae. There was no interaction between the There was also no interaction between the treatments and treatments and the age of the bulbs. The treatments and the the time of harvest (P = 0.985). The bulbs harvested from 194 Sparg, Jäger and Van Staden

Table 2: Minimum inhibitory concentration (MIC) values of methanol M. natalensis bulb extracts against test bacteria. Values include SEM for bulbs which were cultivated at the Fort Hare site

Bulb age Treatment MIC (mg ml–1) against different bacterial strains (months) B. subtilis S. aureus E. coli K. pneumoniae 6 Control 5.08 ± 1.62 2.34 ± 0.84 2.93 ± 0.55 2.93 ± 1.55 Fertiliser 4.69 ± 1.67 2.15 ± 0.81 2.54 ± 0.81 2.73 ± 0.73 Irrigation 5.08 ± 1.62 2.15 ± 0.81 2.35 ± 0.84 2.93 ± 0.55 12 Control 5.47 ± 1.45 3.32 ± 1.95 3.52 ± 1.11 3.52 ± 1.82 Fertiliser 5.86 ± 1.11 3.12 ± 1.45 2.93 ± 0.55 2.53 ± 0.81 Irrigation 5.47 ± 1.45 3.52 ± 1.82 3.13 ± 0.00 3.52 ± 1.11 18 Control 7.03 ± 2.21 3.32 ± 1.30 5.08 ± 1.62 3.52 ± 1.11 Fertiliser 6.25 ± 0.00 4.30 ± 1.62 4.69 ± 1.67 3.13 ± 0.00 Irrigation 6.25 ± 0.00 4.69 ± 1.62 5.08 ± 1.62 3.13 ± 0.00 24 Control 6.25 ± 0.00 7.03 ± 2.21 5.86 ± 1.11 3.91 ± 1.45 Fertiliser 6.25 ± 0.00 7.03 ± 2.21 5.47 ± 1.45 3.13 ± 0.00 Irrigation 6.25 ± 0.00 7.03 ± 2.21 5.86 ± 1.11 3.13 ± 0.00 Neomycin 1.60 x 10–2 3.10 x 10–2 6.30 x 10–2 1.60 x 10–2

interaction between the treatments and the time of harvest, Control (a) and the treatments also showed no significant (P ≤ 0.05) Fertilizer effect on the COX-1 percentage inhibition. However, the age Irrigation 100 of the bulbs had a significant effect (P ≤ 0.05) on inhibiting the COX-1 enzyme action in prostaglandin synthesis (P = 80 0.011). Light et al. (2002) also showed there to be little seasonal variation on anti-inflammatory activity in 60 Siphonochilus aethiopicus. However, Taylor and Van Staden (2001) reported that the age of Eucomis autumnalis bulbs 40 had an effect on COX-1 inhibition. The bulbs of mature 20 plants were more active than the bulbs of younger plants. There were also differences in activity over the seasons. The highest activity was detected towards the end of the growing season, shortly before the onset of dormancy. Results from (b) the Fort Hare site indicate that the COX-1 inhibitory activity was affected by the age of the bulbs being tested, and that 100 the younger plants had higher activity. This would mean that COX-1 INHIBITION (%) INHIBITION COX-1 young bulbs could potentially be used in the treatment of 80 inflammation. Taylor and Van Staden (2002) showed that the environmental conditions during the period of winter 60 dormancy affect the levels of anti-inflammatory activity that was exhibited by Eucomis autumnalis autumnalis bulbs. 40 However, the results of this study indicate that winter 20 dormancy had no effect on the levels of COX-1 inhibitory activity in M. natalensis. The anthelmintic activity is represented as the mean 6121824Indomethacin percentage mortality of the nematodes (Figure 2). The BULB AGE (months) levamisole control (5µg ml–1) killed approximately 40% of the nematodes while the solvent blank had zero mortality. Methanol extracts (1mg ml–1) from the cultivated M. Figure 1: Percentage COX-1 inhibition from dichloromethane natalensis bulbs had slightly lower activity compared to the extracts (250µg ml-1) of M. natalensis bulbs cultivated at a) H A 2 anthelmintic activity of mature M. natalensis bulbs reported Botanicals and b) Fort Hare. Indomethacin was tested at a concentration of 5µM by Sparg et al. (2002). Although lower, the activity of the young cultivated bulbs was relatively high. The anthelmintic

results of bulbs cultivated from the H2A Botanicals site are the Fort Hare site showed similar patterns as the bulbs shown in Figure 2a. The methanol extracts killed between 30 harvested from H2A Botanicals. The percentage COX-1 and 40% of the C. elegans nematodes after 2h. There were inhibition ranged from 88% to 96% (Figure 1b). The no interactions between the treatments and the age of the dichloromethane extracts from the six month-old bulbs bulbs, and the treatments had no significant effect (P ≤ 0.05) harvested from the control treatment had the highest (96%) on the anthelmintic activity. However, there was a clear and significantly different (P ≤ 0.05) inhibition activity in increase in activity as the plants matured. The extracts of 24 comparison to the other bulb extracts. There was no month-old bulbs were more lethal to C. elegans. Analysis South African Journal of Botany 2005, 71(2): 191–196 195

combined with irrigation, could have affected the chemical (a) Control 60 composition of the bulbs. The age of the bulbs also had a Fertilizer significant effect (P ≤ 0.05) on the anthelmintic activity (P < Irrigation 0.001) of bulbs cultivated at Fort Hare. The more mature bulbs were more active against C. elegans, killing on 40 average 41% of the nematodes. Young bulbs (six months) killed only 30% of the nematodes. Although the differences in anthelmintic activity were statistically significant, it must be noted that the results are estimated at a 5% range. The 20 results from this in vitro study indicate that mature plants could be used in the treatment of intestinal worms (nematodes). However, further in vivo testing would be necessary to investigate efficacy against animal parasites. Overall, this study indicates that the cultivation treatments (b) 60 used had no significant effect (P ≤ 0.05) on the antibacterial and anti-inflammatory activity of M. natalensis bulbs. However, if cultivated in areas of low rainfall (by irrigating the plants), the anthelmintic activity can be increased. The age

NEMATODE MORTALITY (%) 40 of the bulbs had a significant effect on both COX-1 inhibition and anthelmintic activity against C. elegans. The successful cultivation, along with the pharmacological evidence, shows that the activity of indigenous bulbous medicinal plants such 20 as M. natalensis is not necessarily lost through cultivation practices but is, in most cases, dependent on the age of the bulbs.

6121824Levamisole Acknowledgements — The South African-Netherlands Research BULB AGE (months) Programme on Alternatives in Development (SANPAD) and the National Research Foundation (NRF) are gratefully acknowledged for providing

financial assistance. Dr J Hartzell of H2A Botanicals and Dr ML Magwa Figure 2: Percentage mortality of C. elegans from methanol of the University of Fort Hare are thanked for growing the plants. -1 extracts (1mg ml ) of M. natalensis bulbs cultivated at a) H2A Botanicals and b) Fort Hare. Levamisole was tested at a References concentration of 5µg ml-1 indicates that the age of the bulbs did have a significant Bernath J (1986) Production ecology of secondary plant products. In: Craker LE, Simon JE (eds) Herbs, Spices and Medicinal effect (P ≤ 0.05) on the anthelmintic activity (P = 0.020). This Plants: Recent Advances in Botany, Horticulture and explains why the anthelmintic activity shown by mature M. Pharmacology. Vol. 1. Oryx Press, Phoenix, Arizona, pp 185–234 natalensis bulbs is higher (Sparg et al. 2002). Similar results Brenner S (1974) The genetics of Caenorhabditis elegans. Genetics were found for the methanol extracts of bulbs cultivated at 77: 71–94 Fort Hare. The extracts killed between 30 and 40% of the C. Cunningham AB (1988) An investigation of the herbal medicine elegans nematodes after 2h (Figure 2b). As was the case at trade in Natal/KwaZulu. Investigational Report No. 29 ≤ the H2A Botanicals site, no significant interactions (P 0.05) Dong TTX, Cui XM, Song ZH, Zhao KJ, Ji ZN, Lo CK, Tsim WK between the treatments and the age of the bulbs were (2003) Chemical assessment of roots of Panax notoginseng in noted. However, the treatments were found to have a China: regional and seasonal variations in its active constituents. significant effect (P ≤ 0.05) on the anthelmintic activity (P = Journal of Agricultural and Food Chemistry 51: 4617–4623 Elgorashi EE, Drewes SE, Van Staden J (2002) Organ-to-organ and 0.024). The bulbs harvested from the irrigation treatment seasonal variations of alkaloids from Crinum moorei. South had the highest activity, killing on average 37% of the African Journal of Botany 68: 111–114 nematodes. There were no significant differences between Eloff JN (1998) A sensitive and quick microplate method to the activity from bulbs harvested from the control treatment determine the minimal inhibitory concentration of plant extracts and the fertiliser treatment. This suggests that irrigation for bacteria. Planta Medica 64: 711–713 could have an effect on the anthelmintic activity of the bulbs. Hutchings A (1989) A survey and analysis of traditional medicinal

The average annual rainfall at Fort Hare is lower than at H2A plants as used by the Zulu, Xhosa and Sotho. Bothalia 19: 111–123 Botanicals. Therefore, the irrigation treatment at the H2A Botanical site did not have a significant effect on the Jäger AK, Hutchings A, Van Staden J (1996) Screening of Zulu anthelmintic activity. Results from bulbs harvested at the medicinal plants for prostaglandin-synthesis inhibitors. Journal of Ethnopharmacology 52: 95–100 Fort Hare site, suggest that irrigation could increase the Light ME, McGaw LJ, Rabe T, Sparg SG, Taylor MB, Erasmus DG, anthelmintic activity of M. natalensis bulbs if grown in an Van Staden J (2002) Investigation of the biological activities of area where the mean annual rainfall is around 600mm. Siphonochilus aethiopicus and the effect of seasonal However, because both cultivation sites used different senescence. South African Journal of Botany 68: 55–61 fertilisers in their treatments, the sites cannot be compared Mander M (1997) The marketing of indigenous medicinal plants in properly. The inorganic fertiliser used at the Fort Hare site, South Africa: a case study in KwaZulu-Natal. Institute of Natural 196 Sparg, Jäger and Van Staden

Resources, Investigational Report No. 164 Taylor JLS, Van Staden J (2001) The effect of age, season and McGaw LJ, Jäger AK, Van Staden J (2000) Antibacterial, growth conditions on anti-inflammatory activity in Eucomis anthelmintic and anti-amoebic activity in South African medicinal autumnalis (Mill.) Chitt. plant extracts. Plant Growth Regulation plants. Journal of Ethnopharmacology 72: 247–263 34: 39–47 Page CP, Curtis MJ, Sutter MC, Walker MJA, Hoffman BB (1997) Taylor JLS, Van Staden J (2002) The effect of cold storage during Integrated Pharmacology. Mosby, pp 419–441 winter on the levels of COX-1 inhibitory activity of Eucomis Prance GT (1994) Introduction. In: Ethnobotany and the Search for autumnalis autumnalis extracts. South African Journal of Botany New Drugs. Ciba Foundation Symposium 185. John Wiley & 68: 157–162 Sons, Chichester, pp 1–3. ISBN 0–471–95024–6 Van Wyk BE, Van Oudtshoorn B, Gericke N (1997) Medicinal Plants Rasoanaivo P, Ratsimamanga-Urverg S (1993) Biological of South Africa. Briza Publications, Pretoria, pp 232–233 Evaluation of Plants with Reference to the Malagasy Flora. Watt JM, Breyer-Brandwijk MG (1962) The Medicinal and Monograph prepared for the IFS-NAPRECA Workshop on Poisonous Plants of Southern and Eastern Africa. Livingstone, Bioassays held in Antananarivo, Madagascar, 13–18 September Great Britain Sparg SG, Van Staden J, Jäger AK (2002) Pharmacological and Yaniv Z, Palevitch D (1982) Effect of drought on the secondary phytochemical screening of two Hyacinthaceae species: Scilla metabolites of medicinal and aromatic plants. In: Atal CK, Kapur natalensis and Ledebouria ovatifolia. Journal of Ethnopharma- BM (eds) Cultivation and Utilization of Medicinal and Aromatic cology 80: 95–101 Plants. Reg. Res. Lab., Jammu-Tawi, India, pp 1–12

Edited by JN Eloff