Production of

High Quality Australian

Valerian Products

A report for the Rural Industries

Research and Development Corporation

By R.B.H. Wills and D.Shohet

September 2003

RIRDC Publication No 03/081RIRDC Project No UNC-11A

ii

© 2003 Rural Industries Research and Development Corporation. All rights reserved. ISBN 0642 58648 9 ISSN 1440-6845 Production of High Quality Australian Valerian Products Publication no 03/081 Project no. UNC-11A The views expressed and the conclusions reached in this publication are those of the author and not necessarily those of persons consulted. RIRDC shall not be responsible in any way whatsoever to any person who relies in whole or in part on the contents of this report. This publication is copyright. However, RIRDC encourages wide dissemination of its research, providing the Corporation is clearly acknowledged. For any other enquiries concerning reproduction, contact the Communications Manager on phone 02 6272 3186.

Researcher Contact Details Name: Professor R.B.H. Wills Address: School of Applied Sciences, University of Newcastle, PO Box 127, Ourimbah NSW 2258 Phone: 02 43484140 Fax: 02 43494565 Email: Ron.Wills@newcastle.edu.au

In submitting this report, the researcher has agreed to RIRDC publishing this material in its edited form. RIRDC Contact Details Rural Industries Research and Development Corporation Level 1, AMA House 42 Macquarie Street BARTON ACT 2600 PO Box 4776 KINGSTON ACT 2604 Phone: 02 6272 4539 Fax: 02 6272 5877 Email: rirdc@rirdc.gov.au Website: http://www.rirdc.gov.au Published in September 2003 Printed on environmentally friendly paper by Canprint

iii

Foreword The use of medicinal herbs is expanding world-wide and Australia is actively seeking to capitalise on the opportunity to become an international supplier of many medicinal herbs. Since Australia is a relatively high cost producer nation, economic benefit will only be derived through the growing and marketing of high quality products. High quality in medicinal herbs ultimately relates to the presence at high levels of those constituents that confer a health benefit to consumers. In order to support development of a high quality medicinal herb industry in Australia, RIRDC has supported a number of projects under its Essential Oils and Plant Extracts Program. This report details a project on valerian (Valeriana officinalis) that examines the levels of key active constituents of valerian in different genetic stock, during plant growth, postharvest handling and processing and in marketed-products. The study identifies a range of options available to maximise the level of active constituents at all stages of the marketing chain. This project was funded from RIRDC Core Funds which are provided by the Australian Government and was conducted with the active support of the valerian processor, Mediherb Pty Ltd, Warwick, Qld. This report, a new addition to RIRDC’s diverse range of over 900 research publications, forms part of our Essential Oils and Plant Extracts R&D program, which aims to support the growth of a profitable and sustainable essential oils and natural plant extracts industry in Australia. Most of our publications are available for viewing, downloading or purchasing online through our website: • downloads at www.rirdc.gov.au/reports/Index.htm

• purchases at www.rirdc.gov.au/eshop

Dr Simon Hearn Managing Director Rural Industries Research and Development Corporation

iv

Acknowledgements The authors wish to thank Mr Kerry Bone, Mr Lee Carrol, Dr Reg Lehmann and staff at Mediherb for their active support in the continuation of research into the quality of medicinal herbs on the Ourimbah Campus of the University of Newcastle and substantial technical, financial and materials contribution to the conduct of the valerian project. Thanks are also given to Dr Rein Bos, University of Groningen, The Netherlands for assistance and advice with acquisition of planting material, to Dr Douglas Stuart, University of Newcastle for support in analytical methodology and experimental design and to Kym and Sally Grant of Austral Herbs for their generous help in growing plants and providing samples.

v

Contents

Foreword ..................................................................................................................................iii

Acknowledgements.................................................................................................................. iv

Contents..................................................................................................................................... v

Executive Summary ................................................................................................................ vi

1. Introduction .................................................................................................................... 1 1.2 Objectives ................................................................................................................................. 2

2. Valerenic Acids in Valerian Plants ............................................................................... 3 2.1 Methods of analysis .................................................................................................................. 3 2.2 Total valerenic acids content of Valeriana officianalis cv anthos roots ................................... 4 2.3 Plant selection ........................................................................................................................... 5 2.4 Levels in plant parts during growth .......................................................................................... 7 2.5 Implications for industry........................................................................................................... 8

3. Postharvest Handling and Drying .............................................................................. 10 3.1 Handling of roots before drying.............................................................................................. 10 3.2 Drying of roots........................................................................................................................ 14 3.3 Storage of dried root ............................................................................................................... 16 3.4 Implications for industry......................................................................................................... 21

4. Processing into Manufactured Products .................................................................... 23 4.1 Efficiency of alcoholic extraction of active constituents ........................................................ 23 4.2 Levels of active constituents in manufactured products ......................................................... 29 4.3 Implications for industry and consumers................................................................................ 32

5. Summary of Conclusions and Recommendations ..................................................... 34 5.1 Need for analysis of active constituents.................................................................................. 34 5.2 Variation in active constituents in crops................................................................................. 34 5.3 Need for improved postharvest handling practices................................................................. 35 5.4 Need for improved processing operations .............................................................................. 35 5.5 Quality of retail manufactured products ................................................................................. 36

6. References ..................................................................................................................... 38

vi

Executive Summary

The root of the valerian plant (Valeriana officinalis) is a medicinal herb native to Europe that is widely used for the treatment of tension, irritability, restlessness and insomnia. The growing world market for valerian and the high value for products such as valerian concentrate has generated considerable interest for its cultivation in Australia. Australia is agriculturally well positioned to capture a share of the world market and cropping is now conducted in the eastern and southern States. However, if Australia is to be successful at exporting and import substitution, it needs to resolve various handling and quality issues. The quality issues will be driven by consumers who will become more demanding in their requirements for product quality, and as world crop supply increases to better match market demand there will be greater competition in the valerian market. Countries which have the reputation to consistently supply high quality raw material and processed products will undoubtedly have preferential access to the higher price market segment, thus maximising the economic return from the crop.

The over-riding determinant of quality in all medicinal herbs, including valerian, is the concentration of active constituents that impart a health benefit to consumers. While there is still some debate on the relative effectiveness of various classes of compounds, it is widely accepted that the valerenic acids are important active constituents in valerian and there is substantial industry interest in increasing the level of valerenic acids in traded products. The research studies described in this report used the total valerenic acids, that is combined level of valerenic acid, acetoxyvalerenic acid and hydroxyvalerenic acid (stored roots only), as the marker of valerian root quality. A major logistical problem for growers is difficulties in efficiently washing valerian roots and is often cited as a significant barrier to continuing with the crop. The overall aim of the project was to assist the Australian industry, that is, growers, traders and processors, to improve the quality of Australian grown valerian. The research objectives focused on determining: • reliable method for analysis of valerenic acids and the related valepotriates and

baldrinals, • levels of active constituents in valerian grown from seed obtained from diverse sources, • changes in active constituents in valerian root during plant growth and maturation, • effect of postharvest operations on handling time and active constituents, • effect of processing operations on active constituents, • quality of valerian in manufactured retail products.

Efficient and reliable quantitative analytical methods for the analysis of valerenic acids, valepotriates and baldrinals in valerian were developed using high performance liquid chromatography (HPLC). The measure of quality in the studies was based on the industry stated preference for valerenic acids although the levels of valepotriates, baldrinals and essential oil were also determined in all studies. The level of valerenic acids in commercially available roots from plants of the Anthos cultivar, the current preferred industry cultivar, was found to average about 3 mg/g dry weight. While there is a number of suggested standards for the concentration of valerenic acids, based on the recommendation of 3 mg/g by Bos et al. (1998) for high quality valerian, the Australian industry would seem to be already a high quality producer of valerian. However, there seems to be an opportunity for the Australian industry to raise the quality of valerian over time. This conclusion was reached from a study that evaluated valerian seed obtained from 25 sources of valerian with six types from Australian growers, and four from

vii

North American and 15 from European seed companies and botanical gardens grown at three sites in New South Wales with different climatic features. It was found that five sources of seed consistently generated valerenic acid levels >4 mg/g. It was noteworthy that three of the five lines were obtained from Australian growers in New South Wales, Tasmanian and Victoria with the other two lines from France and Russia. The Russian line in particular appears promising as the high valerenic acids content is matched by a larger than normal root size. These plant sources are worthy of further growing trials to evaluate their agronomic performance under a wider range of environmental conditions and to test the consistency of the elevated valerenic acids levels. The identified plant material may ultimately fail in subsequent trials to be commercially acceptable but the study has demonstrated the potential for the Australian industry to increase the quality of its valerian. While processors prefer to source roots with a high concentration of valerenic acids, establishment of a price return based on quality would seem to be a pre-requisite to encourage grower participation in quality enhancement. A study was conducted of the change in the content of the total valerenic acids at five development stages of 1-year old Anthos plants in their subsequent seasonal growth cycle. This was undertaken primarily to determine whether the current practice of harvesting plants at the senescence stage was optimal in terms of active constituents. The concentration of valerenic acids in roots was found to rise sharply from the plant dormant stage to a peak in the spring vegetative growth stage and then fall substantially through to the senescence stage. However, while the maximum concentration of valerenic acids was in spring, the overall yield of valerenic acids per root increased with root age. This was due to continuous linear rate of growth of roots throughout the seasonal growth cycle. It might be expected that 3-year old plants would be larger still although this was not determined. While manufacturers may prefer to receive roots with maximal concentration of active constituents, it may be more desirable to maximise the overall yield of active constituents in Anthos plants by continuing to have growers harvest roots at the senescence stage. Postharvest handling practices are a major operational issue for growers with the difficulty of removing soil from around valerian roots cited by various growers as a major problem. An evaluation of cutting and soaking of roots showed no overall benefit in reducing the time taken to remove soil from around roots. While there was a reduction in washing time following the quartering of roots and removing rootlets from the crown there was no beneficial effect of soaking and when the time taken to cut the plants was added to the washing time, there was no overall time saving. There was, however, a greatly reduced drying time of cut roots in a hot air drier with rootlets drying in 20-30% of the time taken by whole roots. It would therefore seem that a more efficient use of driers would be obtained by routinely separating rootlets from the crown and drying each plant part in a separate batch. A further advantage is that there was a greater retention of valerenic acids in dried rootlets and crown that had been separated before drying. Rootlets tended to have a higher concentration of valerenic acids than the crown which could allow their segregation and sale as a higher quality product. The feasibility of separately drying rootlets and crown was further enhanced by the storage for 10 days of whole roots closely stacked in a wire basket at ambient temperature and humidity. There was no significant change in the level of any active constituent but there was a substantial loss of moisture which would further reduce the time roots needed to be held in a drier. This indicates that roots could be separated into rootlets and crown and stored separately without loss of quality until a drier load of material had been accumulated. The drying temperature was, as expected, directly related to the drying time with a 12-fold decrease over the temperature range of 15°-70°C. There was, however, a substantial decrease in the level of valerenic acids at higher drying temperatures with the most marked change

viii

occurring between 40° and 50°C. It would thus seem that the temperature of a hot air drier should not be maintained above 40°C. The heat pump drier with its use of reduced humidity air was found to give a 25% shorter drying time at 38°C than a hot air drier operated at the same temperature with no adverse effect on any active constituent. The benefit of faster drying and hence greater throughput would need to be considered against the higher purchase cost but lower energy usage of a heat pump drier. Current general storage recommendations for dried valerian are in a closed container protected from light, air and moisture. This study found that the valerenic acids were quite unstable during storage with the rate of loss increasing as the temperature increased and the humidity decreased with >50% loss in root held at 30°C in air of 10% RH over 6 months. Exposure to light further accelerated the loss of valerenic acids. Thus, retention of valerenic acids is favoured by storage at low temperature in the dark. It would seem to be also favoured by retention of a high humidity atmosphere but no explanation can be offered as to the mechanism resulting in such an effect. A preliminary evaluation into the mode of action of loss suggested that both enzymic activity and atmospheric oxidation were involved in the degradation of active constituents. However, there was also substantial loss of valerenic acids during blanching suggesting that use of water blanching was not a commercial option. The studies were only on ground valerian and thus only have direct application for processors. However, there is no reason to doubt that the findings would apply to dried, non-ground root, although the rate of degradation may be at a slower rate. The extraction of dried valerian with solutions of ethanol in water was found to give highly variable extraction of active constituents with different ethanol:water mixtures. The common commercial use of aqueous ethanol in the ratio range of 60:40 to 70:30 ethanol:water is in large part supported by this study as an efficient use of ethanol although some increase in extraction was obtained with higher ethanol concentrations. The valerenic acids were quite stable in ethanolic solution even when stored at ambient temperature and should provide flexibility for industry to efficiently manage either long term storage or holding for further processing while maintaining product quality. Extraction by percolation was found to be more efficient than maceration with about 15% more valerenic acids obtained at the same ethanol concentration. The rate of solvent flow during percolation did not appear to affect extraction efficiency, hence a faster flow rate with considerable time saving could be used. Furthermore, valerenic acids were readily extracted with percolation in 80% ethanol achieving about 95% extraction of valerenic acids using a relatively low solvent:valerian ratio of 2:1. At 60% ethanol, a ratio of about 3:1 was required to achieve a similar rate of extraction or the extraction rate falls to 85% which could still be acceptable. Maceration also showed a similar early extraction of most of the valerenic acids with little increase in extraction using times greater than 4 hr. Valerenic acids were also readily extracted using supercritical fluid extraction (SFE) with CO2 where >90% of extraction occurred in 10 min and with use of relatively mild conditions of 15 MPa and 40°C. The addition of 5% ethanol extracted greater amounts of the valerenic acids and the efficiency was comparable to extraction by percolation. The advantage of SFE is elimination of the need to handle large volumes of solvents and its more benign environmental, health and safety features. The technique is worthy of further investigation although it may be too costly at this stage although SFE has been applied commercially to various foods. A survey of 55 commercial manufactured products showed considerable variation in concentration of valerenic acids from <0.01 to 6.32 mg/g or ml of product with about 20% of

ix

products containing >2 mg/g or ml while 50% contained <1mg/g which included 16% <0.1 mg/g. Powder capsules contained the highest concentration (2.5 mg/g), the tablets, teas and soft gel capsules had about 1 mg/g while liquids contained 0.5 mg/ml. The minority of products with a stated label content of valerenic acids had much higher valerenic acid contents than non-standardised products and while the stated and actual levels were reasonably well correlated, there was a 10-15% lower level in products than the label claim. The calculated values of valerenic acids in products in relation to the amount of added valerian ranged from <0.01 to 2 mg/g root which is similar to the range reported in various European studies. There was a large variation in recommended daily doses on product labels. It would seem that this would be confusing to consumers. It was noted that for about 50% of products the recommended dose was <2 g/day which is lower than the European range of recommended dosage. It is suggested that labelling of products with valerenic acids content and a more uniform recommended dosage would give consumers greater confidence in the continued purchase of valerian products.

1. Introduction The use of medicinal herbs was probably the most useful tool for treating a wide range of illnesses by communities in many parts of the world before the advent of modern medicine and the pharmaceutical industry. The development of surgery, synthetic pharmaceuticals and pathology has achieved substantial advances in alleviating suffering and prolonging life. Despite these advances, and perhaps because of the advances, there has been some disillusionment in recent years by consumers with medical practices, and indeed with modern technology in general. This had led to an interest in alternative health therapies in most countries around the world that has spawned considerable expansion in the use of medicinal herbs. Medicinal herbs were traditionally obtained by harvesting plants from natural woodlands and fields. With the continuing expansion in herbal use, it has been increasingly more difficult to satisfy the demand for many medicinal plants from natural sources. The slow development of a cultivation industry with the continuing reliance on a diminishing source of wild herbs has seen the demand for many herbs greatly exceed supply with resultant substantial price rises. Valerian is a medicinal herb, native to Europe, that is used for the treatment of tension, irritability, restlessness and insomnia. It is the fourth best selling medicinal herb in Europe with retail sales of US$200 million, while in Australia it is in the top 10 selling retail herbs. The market in the USA is still under developed with sales quoted as US$6 million with growth of 85% during 1998. With the value of valerian concentrate being greater than $100,000 per tonne, there is considerable incentive to upgrade the volume and quality of cultivated valerian. Australia is agriculturally well positioned to capture a share of the international market and cropping of valerian is now conducted in a wide range of regions across eastern and southern States. If Australia is to become a successful long term exporter and to replace imports with locally grown crop material, it needs to resolve various marketing and quality issues. As the production of valerian stabilises in respect to market demand and consumers become more sophisticated in their requirements, there will be greater emphasis on quality and cost. It is difficult for high cost production countries such as Australia to compete with low cost developing countries on price. It would seem that the Australian valerian industry should aim to develop an international reputation as a supplier of high quality raw material, processed and manufactured end products in order to gain preferential access to the higher price market segment. This will ensure continuing sales and an adequate economic return from the crop. A major factor in the determination of quality in medicinal herbs is the concentration of those constituents that lead to a health benefit. While there is still debate as the relative effectiveness of various compounds that have been identified as having effects on human health, the valerenic acids are regarded as important beneficial active constituents. Major sectors of the Australian industry are currently benchmarking valerian quality against valerenic acids concentration. The research studies described in this report therefore concentrated on the valerenic acids as the measure of valerian quality. Other factors included in the study are the related compounds, valepotriates and baldrinals which the

2

scientific community has mixed views on beneficial or harmful effects and the essential oil which contains a number of compounds with known beneficial effects and has long been used as a marker of valerian quality. 1.2 Objectives The overall aim of the program was to develop quality parameters and associated tests to enable growers to harvest and handle valerian to maintain optimum quality, and to identify efficient processing techniques that ensure optimum quality is transferred through to the end products. This was pursued experimentally by determining: • reliable methods for the analysis of valerenic acids, valepotriates and baldrinals, • potential for new planting materials with elevated levels of valerenic acids, • optimum harvest times to maximise levels of active constituents, • effect of postharvest handling practices on levels of active constituents, • effect of processing operations involved in the manufacture of value-added products

on levels of active constituents, and • levels of active constituents in retail products currently available to consumers. The research program was conducted in close liaison between the research group at The University of Newcastle and the industry as represented by a major valerian processor and marketer and selected valerian growers. Apart from educating the researchers in valerian industry practices, this liaison ensured the individual projects retained industry relevance and assisted in the transfer of findings to industry.

3

2. Valerenic Acids in Valerian Plants A large genetic variation in the natural population of V. officinalis has led to many attempts to select plant material with consistent high quality characteristics. However, over the years, the relative importance of various chemical groups with medicinal value has changed and hence so have the criteria of plant selection (Bernath 1997). Current interest is in selecting for root yield in conjunction with elevated levels of valerenic acids (Bos et al. 1986; 1998) and low levels of valepotriates due to their reported toxicity (Bounthanh et al. 1981). Three studies were conducted of the concentration of valerenic acids in the roots of V. officinalis plants. The first study involved benchmarking valerian plants currently grown for valerenic acids content, the second was a search for plant material with elevated valerenic acids levels with seed obtained from a range of international locations, and the third examined the accumulation of active compounds over the growth cycle of the plant. 2.1 Methods of analysis All valerian root samples for analysis were washed by hand using high pressure water from a garden hose to remove as much soil as possible then placed in a fan forced hot air drier at 38°C until commercially dry which varied from 4-7 days depending on root size. The dried samples were then crushed to a particle size <250 µm in a laboratory mill to provide a homogeneous mixture. A 5 g sub-sample was mixed with methanol and sonicated, filtered and the liquid extract made up to volume. Considerable effort was devoted to maximising the amount of constituents extracted. Variables examined were methods of physical extraction (shaking, agitation and sonication) as well as particle size of plant material undergoing extraction, the extraction medium and sample weight. The weight of the sample had a pronounced effect on the efficiency of the extraction process, a factor usually not considered in published methodologies. Aliquots were analysed by HPLC on a reversed phase column. The mobile phase comprised a mixture of acetonitrile and phosphoric acid in water on a gradient with increasing proportion of acetonitrile from 20% to 80% during the analysis. The eluted peaks were detected at 225 nm for valerenic acids and baldrinals and 255 nm for the valepotriates. Quantitation was based on the peak area of working reference compounds used as external standards. The valerenic acids working reference compound was biphenyl which was initially calibrated against valerenic acid. The working reference was subsequently used as the standard for all quantification calculations. Similarly, caffeine was used as the working reference standard for the valepotriates with an initial calibration against isovaltrate. As no standard for the baldrinals was available, these data were quantified using the standard curve of the valepotriates. All values were determined on a dry weight basis with the water content of the ground powder determined by drying in a vacuum oven at 100°C and –70kPa for a minimum of 16 hr. The limit of detection for all compounds was equated to about 0.01 mg/g dried root. The method of analysis was a modification of that used by Bos et al. (1996) but was superior in that the valerenic acids, valepotriates and baldrinals were all able to be

4

quantified in a single analysis instead of a dual analysis, and only required a simpler dual wave length UV detector instead of a diode array detector. The essential oil in valerian root (20 g) was quantified using AOAC method 962.17 (1995) developed for volatile oil in spices. This involved hydro-distillation of a ground root sample and volumetric assessment of the collected condensed oil. The limit of detection was 0.1 ml/100 g dried root. It is recognised that the essential oil could contain some valerenic acids. 2.2 Total valerenic acids content of Valeriana officianalis cv anthos roots The ‘Anthos’ cultivar is better regarded by the industry in terms of valerenic acids content than other commonly grown types (R. Lehman, Mediherb, pers. comm.) but a search of seed company lists (Richters –The Herb Specialists, Eden Organic Nursery Services, B&T World Seeds, Pleasance Herb Seeds, and Chiltern Seeds) indicated few choices in cultivar, with only unspecified cultivars of V. officinalis, Anthos and Select cultivars being offered. Communication with the few growers who have persisted with cultivation of valerian were found to be planting the Anthos cultivar (not really true as at this time only Austral Herbs is growing it). In order to provide a benchmark for current quality, 10 commercial dried root samples were obtained from Austral Herbs, currently the largest producer of ‘Anthos’ in Australia (P. Purbrick, Mediherb, pers. comm.) and analysed for active constituents. The data in Table 2.1 show that the mean concentration of total valerenic acids (TVA) in the 10 samples was found to be 3.0 mg/g with a range of 2.1–3.6 mg/g. The valerenic acids found in the root samples were valerenic acid (VA) and acetoxyvalerenic acid (AVA) and were present in the ratio of 1.3:1 (range 0.6-2:1). Of the total valepotriates (TVP), the major valepotriates present were isovaltrate (IVA) and valtrate (VAL) in the ratio of 13:1 (range 10-23:1). The essential oil (EO) was present at 0.4-0.7 ml/100 g. Table 2.1: Valerenic acids, total valepotriate and essential oil contents of ten ‘Anthos’ roots from a commercial source.

Plant VA (mg/g)

AVA (mg/g)

TVA (mg/g)

IVA (mg/g)

VAL (mg/g)

TVP (mg/g)

EO (ml/100g)

1 1.0 1.3 2.3 11.5 1.1 12.6 0.4 2 2.2 1.1 3.3 18.5 0.8 19.3 0.5 3 1.0 1.5 2.5 10.6 1.0 11.6 0.5 4 1.4 1.4 2.8 11.6 1.0 12.6 0.6 5 2.0 1.4 3.4 12.8 0.8 13.6 0.6 6 2.1 1.0 3.1 13.2 0.8 14.0 0.7 7 0.9 1.2 2.1 14.0 0.9 14.9 0.6 8 1.9 1.0 2.9 9.9 0.9 10.7 0.5 9 2.2 1.3 3.5 14.6 1.0 15.6 0.5 10 1.8 1.8 3.6 16.7 1.2 17.9 0.6

Mean 1.7 1.3 3.0 13.3 1.0 14.3 0.6

5

2.3 Plant selection 2.3.1 Experimental design A total of 25 sources of valerian seed were obtained with 6 obtained from Australian growers, 4 from North American and 15 from European seed companies and botanical gardens. Seed from 10 sources was planted in one trial in early 2000 and the other 15 which were sourced later were planted 6 months later in a second trial. Plants were grown at three sites in New South Wales with different climates: Somersby, a temperate coastal climate, latitude 33°S, sea-level, summer mean temperature range 15-26°C, winter mean temperature 6-16°C; Walcha, a sub-tropical tableland climate, latitude 30°S, altitude 1000 m, summer mean temperature 11-25°C, winter mean temperature –2-12°C; Kyogle, sub-tropical coastal climate, latitude 29°S, sea level, summer mean temperature 18-30°C, winter mean temperature 6-20°C. Plants were harvested after about 11 months of growth and the roots analysed for active constituents. Roots from planting material with the two highest levels of valerenic acids from each trial were re-grown at the three sites along with root from a planting material containing a “normal” level of valerenic acids. Plants were established in late 2001 and harvested 7 months later for analysis of dry matter and active constituents. 2.3.2 Composition of plant selections The total valerenic acids, total valepotriates and essential oil were used as the indicators of quality. Table 2.1 gives the levels of these active constituents in the 25 sources of valerian and are listed in decreasing level of valerenic acids. For plant material used in the first trial, the level of total valerenic acids ranged from 2.8-4.7 mg/g dry root. The highest levels were in plants sourced from SU and HC which contained 4.7 and 4.1 mg/g, respectively. The levels were significantly higher than the three sources of ‘Anthos’ - RA, BTA and AU – which contained 3.1-3.7 mg/g. In the second trial, total valerenic acids ranged from <0.01-4.3 mg/g with WH, PET and FR containing 4.1-4.3 mg/g. Table 2.2 also shows that there was substantial variation in the levels of total valepotriates and essential oil. There were no significant correlations between total valerenic acids and essential oil content (r = 0.05), total valepotriates and essential oil content (r = 0.07) or total valerenic acids and total valepotriates (r = 0.02). The plant sources with the two highest levels of valerenic acids were re-planted from saved roots in a second season at the three sites. In addition, INN plants, which had a relatively low content of valerenic acids, were included for comparison. The combined data for root weight and active compounds are presented in Table 2.3 as there was no significant difference in any quality indicators between the sites. This shows that all the selected plants contained a significantly higher level of valerenic acids than the INN plants with WH = PET > SU = HC. When the root weight was taken into account to show the total valerenic acids per plant, PET roots contained a significantly greater amount than roots from the other plant selections. The origins of the plant materials with the higher level of valerenic acids over the three trials are: PET, University of Petrozavodsk, Russia; WH, Greg Whitten, Tasmania; SU,

6

Subiaco Herbs, Walcha; HC, Herbal Connection Network, Victoria; and FR, Jardin d’Art et d’Essais, France

Table 2.2 Active constituents in roots of valerian plant material obtained from various domestic and international sources.

Plant Source TVA (mg/g) TVP (mg/g) EO (ml/100g) Trial 1 SU 4.7 7.6 0.6 HC 4.1 4.2 0.5 RA 3.7 17.4 0.4 BTA 3.6 14.4 0.6 RS 3.6 16.8 0.6 EO 3.5 15.5 0.5 AU 3.1 12.6 0.6 RO 3.0 11.6 0.3 PO 2.8 12.7 0.4 BC 1.9 8.1 0.5 5% LSD ± 0.48 ± 2.36 ± 0.15 Trial 2 WH 4.3 6.3 0.5 PET 4.3 10.1 0.7 FR 4.1 5.8 0.5 K 2.7 12.0 0.6 BTS 2.5 11.4 0.6 CH 2.3 11.8 0.6 VS2 2.2 7.3 0.8 INN 2.1 6.7 0.5 ULM 1.8 5.1 0.5 MED 1.7 6.5 0.7 SL < 0.01 6.9 0.7 SA < 0.01 5.8 0.7 SB < 0.01 6.8 0.6 SC < 0.01 5.6 0.6 SD < 0.01 10.1 0.6 5% LSD ± 1.46 ± 2.20 ns

Table 2.3 Dry weight and total valerenic acids, valepotriates and essential oil content of roots from selected valerian plant sources grown in three locations. Plant Dry Wt TVA TVP EO Source (g) (mg/g) (mg/plant) (mg/g) (ml/100 g) WH 65 4.3 265 8.4 0.7 PET 121 4.2 505 13.6 0.6 SU 74 3.3 237 5.4 0.5 HC 80 3.0 243 4.7 0.6 INN 116 1.5 153 9.3 0.7 LSD 5% ± 57 ± 0.59 ±127 ± 1.86 ± 0.06

7

2.4 Levels in plant parts during growth The accumulation of valerenic acids, valepotriates and essential oil was studied over a 12 month growth cycle of the valerian plant. Approximately 180 Anthos plants were established in a nursery bed and after two months planted in early summer at Walcha. Sampling was conducted over a full growing season and commenced in winter at dormancy stage. Samples were subsequently taken in the following growing period during vegetative growth in the spring, at flowering in summer, during vegetative growth in autumn and finally at senescence. At each stage, three groups of seven plants were harvested and the roots analysed for dry weight and active constituents. The dry weight of valerian roots and the level of active constituents at each harvest stage are presented in Table 2.4. The root dry weight continually rose significantly throughout the growing season and showed a significant linear regression, y = 7.9x - 49.7 (P< 0.05) where root dry weight (y) increased with plant age (x). This suggests that plants had not reached their full growth potential even at the end of the two year growth period. Table 2.4 Root dry weight and active constituents of valerian plants at different stages of growth

Growth Stage Root Dry

Wt (g) TVA

(mg/g) TVP

(mg/g) EO

(ml/100 g) Dormancy 22 3.4 16.2 0.4

Vegetative, spring 42 5.0 20.0 0.7 Flowering 64 4.3 18.2 0.5

Vegetative, autumn 93 4.4 14.3 0.8 Senescence 110 3.3 15.6 0.7

LSD 5% ±22 ±0.5 ±1.6 ±0.1 The active constituents showed a different accumulation pattern. The total valerenic acids and valepotriates concentrations rose sharply to a peak in spring and then fell significantly through to senescence. The essential oil content also increased to a maximum value but occurred later during the autumn. The three constituents showed significant (P<0.01) regressions of y = 4.0 + 0.1x – 0.3(x – 14.4)SGN(x – 14.4) for total valerenic acids, y = 12.6 + 0.5x – 1.1(x – 12.6)SGN(x – 12.6) for total valepotriates and y = 1.1 + 0.01x – 0.06(x – 18.0)SGN(x – 18.0) for essential oil, where y = concentration of compound and x = plant age in months and SGN is the change point. The yield of total valerenic acids, total valepotriates and essential oil per root were calculated from root weight and the data presented in Table 2.5 show that the amount of all constituents/root increased with plant age. Significant (p<0.01) linear regressions for yield (y) with plant age (x) were obtained for total valerenic acids (y = 30x – 177), total valepotriates (y = 116x – 661) and essential oil (y = 0.1x – 0.5). The increase in root size was therefore a dominating factor in determining yield.

8

Table 2.5 Yield of active constituents per valerian root at different stages of growth Growth Stage TVA

(mg/root) TVP

(mg/root) EO

(ml/root) Dormancy 75 354 0.10 Vegetative, spring 209 817 0.27 Flowering 275 1117 0.34 Vegetative, autumn 407 1311 0.77 Senescence 371 1670 0.79 LSD 5% ±92 ±440 ±0.22 Yield of active constituents was obtained from root weight x constituent concentration 2.5 Implications for industry Quality criteria have been published by Stahn and Bomme (1998) who proposed that roots containing 2–2.5 mg/g total valerenic acids be classified as high and those with >2.5 mg/g as very high while Bos et al. (1998) stated that high quality valerian should contain 3 mg/g of valerenic acids. On a high quality criteria of 3 mg/g valerenic acids, the total valerenic acids content of Anthos currently grown in Australia at about 3.0 mg/g and would allow Australia to qualify as a producer of high quality valerian. However, the valerian plant material sourced from WH, PET, SU, HC and FR with levels >4 mg/g can be considered as being very high quality valerian. The high quality plant sources are worthy of further growing trials to evaluate their agronomic performance under a wider range of environmental conditions and to test the consistency of the elevated valerenic acids levels. In particular, PET appears promising as the high valerenic acids content is matched by a large root size. The identified plant material may fail in subsequent trials but the study has demonstrated the potential for the Australian industry to increase the quality of valerian grown. Processors would prefer to source roots with a high concentration of valerenic acids so that extracts of high concentration can be more easily manufactured. However, since currently in Australia there is no price premium for chemical content with growers paid by weight, the only incentive is for growers to produce large roots. Additionally, vigorous growing plants are less costly to produce as the canopy closes quickly and the requirement for weeding is reduced (S. Grant, Austral Herbs, pers. comm.). Establishment of a price return based on quality would seem to be a pre-requisite to encourage grower participation in quality enhancement. The not unexpected continuous increase in dry weight of valerian roots with plant age, firstly showed that roots from 2-year old valerian plants were much larger than those from 1-year old plants. It might be expected that 3-year old plants would be larger still although this was not determined. The intra-seasonal fluctuation in accumulation of active constituents with the peak concentration for valerenic acids occurring in spring potentially poses some difficulty in deciding when to harvest for maximum medicinal quality. While manufacturers may prefer to receive roots with maximal concentration of active constituents, the overall yield of active constituents per root was found to occur with the largest roots at the senescence stage when the concentration of valerenic acids had markedly

9

declined. Without price incentives, it would be difficult to persuade growers not to harvest the largest roots possible which occurs with the current industry practice. Since the number of Australian growers prepared to produce the crop is limited, manufacturers may also find the current practice more desirable as it will result in the greatest quantity of active root extracts from the national crop.

10

3. Postharvest Handling and Drying The current trade in Australian valerian is as a dried root. Like other medicinal herbs, the drying of valerian is invariably the responsibility of the grower who may dry the crop on farm or sub-contract the drying to another grower or group. The dried valerian root is then traded to a wholesaler or manufacturer. As the scale of production on a farm increases or an off-farm drier is utilised, there is increasing handling and transport of the freshly harvested crop and increasing delay between harvest and drying. For the wholesaler and manufacturer, there is increasing storage of the dried valerian and increasing delay between receipt of the crop and processing into a manufactured product. These pressures are due to the seasonal nature of valerian harvest and the economic necessity to operate processing plants throughout as much of the year as possible. There is little published literature on the effect of postharvest operations on the valerenic acids. However, in the light of data showing echinacea to be adversely affected by certain postharvest handling factors (Wills and Stuart 2000), a series of studies was conducted to examine the effect of various postharvest practices on the level of active constituents. This included the effects of:

• handling of roots before drying, • drying temperature and technology, and • environmental conditions during long-term storage of dried roots.

3.1 Handling of roots before drying The effect of handling was always related to the level of active constituents in the dried product. In order to enable the end point of drying trials to be ascertained, the initial experiment was to determine the water content of fresh roots. Five fresh roots were washed and two were separated into rootlets (the long roots cut from the crown) and crown (the main root and rhizomes) while the other three roots were left intact. The roots were all dried at 50°C and -70 kPa in a vacuum oven until there was no further weight loss which was after about 4 days. The water content of the fresh rootlets, crowns and whole roots was found to be similar in all sections at about 78 g/100 g fresh weight. Since commercially dry medicinal herbs are commonly found to have a water content of about 10 g/100 g, valerian roots were considered to be dry when they lost 68% of the fresh weight.

Table 3.1 Handling time (min/plant) of cutting and washing operations.

Treatment Washing Cutting Total Whole 1.5 1.5 Quartered 1.0 0.2 1.2 Quartered/cut 0.7 0.5 1.2 Soak 1 hr 1.2 1.2 Soak 24 hr 1.1 1.1 Quarter/soak 24 hr 0.8 0.3 1.1 LSD 5% ±0.31 ns

11

3.1.1 Cutting and soaking prior to washing Valerian is notorious for the considerable difficulty in removing soil from the roots after harvest. The washing of the valerian root can be problematic because of its matted nature and the tendency for soil to cling where the rootlets are attached to the crown. It has been stated by several growers as the biggest barrier to growing the crop. The general practice among growers is to slice the crown into halves or quarters, depending on the size, prior to washing. It is not known what effect this practice has on the level of active constituents of the root. It was hypothesized that drying time could be greatly reduced by separating the rhizome and rootlets prior to drying. The effects of the greater surface area and tissue damage due to cutting on the valerenic acids was, however, not known. An evaluation was conducted of whether any benefit can be obtained by cutting and soaking roots in terms of time taken and active constituents in the dried product. For each treatment, roots from 21 valerian plants were distributed into three replicates each of seven roots. The following treatments were applied to a root in each replicate: whole roots not washed; whole roots washed; quartered roots washed; quartered roots with rootlets removed and washed; whole roots soaked for 1 hr then washed; whole roots soaked for 24 hr then washed; and quartered roots soaked for 24 hr then washed. The cutting and washing procedures were timed and the average drying time was determined for the rootlets, quartered roots/crowns and whole roots. The roots in each treatment were then placed in a hot air drier at 38.40°C until commercially dry (i.e. contained 10% moisture) and the level of active constituents determined on the dried material. The times for cutting and washing are given in Table 3.1. There was a significant reduction in washing time due to quartering roots and removing rootlets but there was no effect of soaking. This soil was very sandy, soaking may be advantage on heavier soil. When the time taken to cut the plants was added to the washing time, there was no significant difference between treatments. As would be expected, the drying time varied considerably with the degree of cutting. Rootlets took 28 hr to dry while quartered roots required 96 hr and whole roots 140 hr to reach commercial dryness. There was, however, no significant difference in the concentration of valerenic acids, valepotriates or essential oil in any cutting or soaking treatment. Hence, the advantage of cutting roots into smaller sections will be a shorter drying time presumably due to the relatively greater surface area of the smaller particles with no deleterious effect on active constituents. Table 3.1 Handling time (min/plant) of cutting and washing operations.

Treatment Washing Cutting Total

Whole 1.5 1.5 Quartered 1.0 0.2 1.2

Quartered/cut 0.7 0.5 1.2 Soak 1 hr 1.2 1.2 Soak 24 hr 1.1 1.1

Quarter/soak 24 hr 0.8 0.3 1.1 LSD 5% ±0.31 ns

12

3.1.2 Cutting root prior to drying A follow up study examined whether cutting roots after washing would show a similar effect to that obtained in Section 3.1.1. Valerian roots were washed intact and distributed to three replicates each with five roots for each treatment The roots were then either quartered by cutting through the crown, chopped into 1cm pieces or left whole. The data in Table 3.2 show that drying time was lower as the root size was decreased, that is whole quartered< chopped roots. The concentration of total valerenic acid in the roots was not significantly different between the root sizes. The total valepotriate concentration of cut roots was not significantly different from the whole root but the essential oil content was significantly lower in both cut roots. It thus seems feasible to cut roots after washing to achieve a faster drying time with minimal loss of active constituents. The small loss of essential oil would probably be due to greater ease of volatilisation from the cut surfaces. Table 3.2 Effect of root separation prior to drying on active constituents in dried root (mg/g)

Treatment Drying time (hr)

VA AVA TVA TVP Oil (ml/100 g)

Chopped 48 1.50 1.14 2.64 13.34 0.36 Quartered 96 1.40 1.89 3.29 15.90 0.34 Whole 168 1.38 1.68 3.07 14.48 0.43 LSD 5% ns ±0.50 ns ±1.23 ±0.06

3.1.3 Separation of rootlets and crown A further study examined whether separating valerian root into crown and rootlets may generate washing or drying efficiency with minimal loss of active constituents. Valerian plants were divided into three groups of 15 plants. (Roots in a group were either washed and dried as a whole root, washed whole then the rootlets were cut from the crown, or the rootlets were cut from the crown then each section washed.) The roots were weighed then held in a hot air drier for 126 hr during which time each root sample was periodically weighed. The time taken to become commercially dry was calculated from the data. The whole roots were separated into rootlets and crown and all samples were analysed for active constituents. The roots comprised 37% rootlets and 63% crown but there was considerable variation between individual roots with rootlet weight ranging from 20-56%. The rootlets were dry after 34 hr (range, 21-34 hr) and crowns and whole plants after 110 hr (range, 50-110 hr). The level of active constituents in dried rootlets and crowns presented in Table 3.3 show that rootlets and crowns separated before drying contained a higher level of valerenic acids than those removed after drying with a tendency for rootlets to have a higher level than the crown. There was, however, no significant effect of cutting the roots before or after washing. For valepotriates, crowns had a higher level than rootlets. Whole roots had a higher level of valepotriates than cut rootlets but the effect on the crown showed no consistent effect. Cut rootlets had a higher level of essential oil than those dried on whole roots but there was no effect of cutting on the crown.

13

Table 3.3 Active constituents (mg/g) of dried valerian rootlets and crown that were sectioned before and after washing

Treatment VA AVA TVA TVP Oil

(ml/100 g) Rootlets Whole 1.31 0.72 2.03 10.95 0.55

Cut Before 1.54 0.81 2.35 8.60 0.81 Cut After 1.55 0.83 2.37 8.21 0.90 Crowns Whole 1.22 0.69 1.93 15.86 0.47

Cut Before 1.33 0.78 2.13 13.63 0.47 Cut After 1.41 0.83 2.27 17.09 0.47

LSD 5% ±0.12 ±0.07 ±0.18 ±2.19 ±0.11

3.1.4 Storage of roots prior to drying Roots from 36 valerian plants established by root division of a single plant were washed and allowed to air dry for 16 hr before determination of the fresh weight. The roots were then closely stacked in a plastic-coated wire basket and stored at ambient temperature (about 20°C) and humidity (about 40-60%). Every second day for 10 days, 6 roots were weighed and dried in the hot air dryer at 38°C. Three sets of two roots were formed and analysed for active constituents. The results presented in Table 3.4 show that there was no significant change in the level of any active constituent during storage at 20°C for 10 days. The loss of moisture from the roots had, however, linearly increased (P<0.01) over the period with about 47 g/100 g loss in weight on day 10 (Figure 3.1). With an initial moisture content of about 78 g/100 g, the roots only needed to lose an additional 20 g water/100 g to be commercially dry. Table 3.4 Active constituents (mg/g) in valerian root dried after varying periods of storage in air at 25°C

Time (days) VA AVA TVA TVP Oil (ml/100 g)

0 1.60 1.36 2.96 13.58 0.54 2 1.74 1.30 3.04 14.85 0.65 4 1.61 1.39 3.00 14.36 0.55 6 1.84 1.52 3.36 13.87 0.54 8 1.97 1.39 3.36 13.98 0.61 10 1.77 1.45 3.27 14.82 0.60

LSD 5% ns ns ns ns ns

14

y = 4.80x + 0.40

0

10

20

30

40

50

60

0 2 4 6 8 10 12Storage Time (days)

Moi

stur

e Lo

ss (g

/100

g)

Figure 3.1 Loss of moisture from valerian roots during storage at 25°C

3.2 Drying of roots Valerian can be dried in a wide range of different types of artificial driers. Sun drying is the cheapest method in terms of equipment cost but is impractical for most parts of Australia. The most economical drier is a hot air drier which comprises a chamber where air heated to a pre-determined temperature by gas or electricity is passed through the chamber. A more efficient drier in terms of operating costs but more expensive in capital equipment outlay, is a heat pump drier in which dehumidified air, heated to a pre-determined temperature is passed across material held in a chamber. The air temperature in heat pump drying is usually lower than in a hot air drier. Various low temperature driers such as a vacuum drier operate by reducing air pressure in the chamber and thereby increase the rate of evaporation of water from the substrate. The disadvantage of a low temperature drier is its greater cost and slower rate of drying. 3.2.1 Temperature of hot air drying The influence of drying temperature on active constituents was examined using a vacuum oven operated at 15°C and a hot air dryer operated at 40, 50, 60 and 70°C. Valerian plants were harvested over a 2 week period and dried sequentially at the different temperatures. For each drying temperature, 21 whole roots were washed and allowed to air dry for 16 hr before being placed in the drier. Roots were weighed twice daily to determine the moisture loss. When dried, the roots were combined randomly into three replicates of seven roots. As expected the drying time was directly related to the drying temperature and ranged from about 230 hr at 15°C down to about 20 hr at 70°C (Figure 3.2). However, the data in Table 3.5 show that the level of valerenic acids was higher at the lower drying temperatures with the most marked decrease occurring between 40° and 50°C. The total valepotriates content of the roots dried at 70°C was significantly lower than roots dried at the other temperatures but the proportional decline was not as great as for the valerenic acids. The essential oil

15

concentration also decreased with increasing temperature with the decline occurring over the range 40-70°C and about 75% of the oil lost at 70°C. It would thus seem that the temperature of a hot air drier should be maintained below 40°C. Table 3.5 Effect of drying temperature on the level of active constituents (mg/g) in dried valerian roots.

Drying temp VA AVA TVA TVP Oil (ml/100 g) 15°C 2.72 2.94 5.66 16.28 0.89 40°C 2.80 1.46 4.59 15.21 0.88 50°C 1.46 1.30 2.77 15.33 0.53 60°C 1.56 1.56 3.38 15.54 0.41 70°C 0.87 1.44 2.31 13.73 0.22 LSD 5% ±1.07 ±0.28 ±1.08 ±1.32 ±0.18

y = -3.8x + 274

0

50

100

150

200

250

0 10 20 30 40 50 60 70 80

Drying Temperature (C)

Tim

e (h

ours

)

Figure 3.2 Effect of temperature on the drying time of valerian roots

3.2.2 Use of a heat pump drier The heat pump drier is viewed favourably by parts of the medicinal herb industry due to its faster drying time and/or operation at lower temperatures than a hot air drier. Comparison of heat-pump and hot air drying was made by cutting nine roots in half through the crown and drying one half by heat pump and the other half in a hot air drier with each drier operated at 38°C. Roots held in the heat pump drier took 72 hr to dry while those in the hot air drier took 96 hr. There was, however, no difference in the levels of active constituents dried with the two methods (Table 3.6). The faster drying time of the heat pump drier with no adverse effect on active constituents in roots would seem to be of value but the higher cost of a heat pump drier compared to a simpler hot air drier would also be a consideration.

16

Table 3.6 Active constituents (mg/g) in valerian roots dried by heat pump and hot air.

Drier VA AVA TVA TVP Oil (ml/100 g)

Heat Pump 1.2 1.5 2.7 12.7 0.51 Hot Air 1.3 1.6 2.9 12.8 0.53 LSD 5% ns ns ns ns ns

3.3 Storage of dried root Dried valerian is invariably stored for some time during the postharvest chain before it is consumed or processed into a value-added product. During this storage period, changes may occur in the level of active constituents. Storage recommendations for dried valerian have been a closed container protected from light, air and moisture (Chapelle and Denoel, 1972; Upton, 1999). Perry et al (1996) found a much higher level of valerenic acid in freshly dried roots (5.8 mg/g) than that recorded by Hänsel and Schulz (1982) for commercial samples of valerian root (0.49 mg/g). They proposed that the differences might have been due to deterioration during long-term storage. Little is known about the fate of the valerenic acids during long-term storage but it is an important factor for countries like Australia, where access to markets often involves long transportation times. Chapelle and Denoel (1972) found retention of valtrates in powdered roots was assisted by storing at reduced temperature, low humidity and under exposure to light. Graf and Bornkessel (1978) also reported greater loss of valepotriates held under high humidity and smaller particle size. According to Upton (1999) the essential oil is relatively stable, though it can evaporate with excessive exposure to air and can decrease by 50% in 6 months in powdered root. A study was conducted to determine the effect of temperature, humidity and light on the valerenic acids, valepotriates and essential oil contents of root material during long term storage. In order to shed light on the types of reactions that were occurring, blanched and vacuum packed powdered roots were also stored. 3.3.1 Effect of temperature and humidity To evaluate the effect of storage temperature and humidity on active constituents, two groups of 24 mature valerian plants were harvested in different years and the roots washed, dried and ground to a powder of <200 µm particle size. All roots were combined to produce a homogeneous sample which was separated into 18 sub-samples that were placed in plastic containers with holes pierced in the lid. Two containers were allocated to each of nine treatments being a factorial of three temperatures (5, 14 and 30°C) and three humidities (low, medium and high RH). The RH atmospheres were achieved by placing open petri dishes containing silica gel (10% RH) and saturated solutions of magnesium chloride (40% RH) and potassium sulphate (80% RH) in an airtight plastic container into which the relevant samples were placed. At monthly intervals for 6 months, a sample from each treatment was weighed and an aliquot analysed for active constituents. The change in the moisture content under different temperature and humidity conditions is shown in Table 3.7. The initial moisture content was 5 g/100 g and increased substantially

17

during storage at all temperatures in root material held at 40% and 80% RH (M and H, respectively) but showed a small but significant decrease in material held at 10% RH (L). The level of total valerenic acids showed a significant loss in all treatments over the 6 month storage period. In general the loss was greater as the humidity decreased and the temperature increased. The greatest loss was therefore observed in root powder stored at 30°C in low RH where a significant loss had occurred at 1 month and >50% was lost over the storage period. These effects are illustrated in Figure 3.3.

Figure 3.3 Interaction between humidity and temperature on the level of total valerenic acids in stored valerian root powder.

Data are the mean values over the storage period

1

1.2

1.4

1.6

1.8

2

2.2

Low Moderate High

Humidity

Tota

l Val

eren

ic A

cids

(mg/

g) 5°C 14°C 30°C

The level of valepotriates fell over the 6 month period and was evident after 1 month in all treatments. The loss increased as the temperature and humidity increased and was therefore greatest in root powder stored in high humidity at 30°C where they had declined to near zero after 2 months. These effects are illustrated in Figure 3.4. Homobaldrinal was also detected in all treatments during storage but was only at relatively low levels. The major effect on the essential oil content was a decrease with increasing temperature, particularly at 30°C where about 50% was lost after 3 months. There was also a small effect of humidity with a decrease in oil content as the humidity decreased. There was no significant loss from root powder stored at 5°C and high and moderate humidity over the 6 months. Table 3.7. Moisture (g/100 g) and active constituents (mg/g) in valerian root powder stored for 6 months at different temperature and humidity.

18

Time Amount of constituent (month) 5L* 5M 5H 14L 14M 14H 30L 30M 30H Mean Moisture 0 (5.0) 1 4.5 6.7 11.4 3.8 6.8 13.1 2.3 4.9 15.5 7.6 2 4.3 7.3 14.5 3.1 7.4 16.9 1.6 5.4 17.6 8.7 3 4.3 8.1 17.3 3.8 8.4 18.6 2.1 6.9 20.1 9.9 4 4.4 9.2 18.3 3.3 10.0 19.7 1.6 9.5 21.2 10.8 5 3.7 10.1 21.6 3.0 11.2 20.7 1.6 9.2 22.0 11.4 6 3.6 10.9 22.3 3.0 11.4 22.0 2.6 9.5 20.6 11.7 Mean 4.1 8.7 17.5 3.3 9.2 18.5 2.0 7.6 19.5 10.0 (LSD 5% = ±1.8) Total Valerenic Acids 0 (2.13) 1 2.05 2.11 2.09 2.10 2.19 2.20 1.67 1.81 1.95 2.02 2 2.03 2.06 2.10 2.05 2.04 1.94 1.42 1.69 1.79 1.90 3 2.01 2.05 2.02 1.91 1.89 1.86 1.21 1.58 1.78 1.81 4 1.95 1.98 1.97 1.83 1.85 1.84 1.14 1.58 1.74 1.77 5 1.88 1.92 2.01 1.73 1.74 1.85 1.01 1.52 1.62 1.70 6 1.81 1.91 1.99 1.69 1.75 1.80 0.99 1.47 1.59 1.67 Mean 1.96 2.00 2.03 1.89 1.91 1.92 1.24 1.61 1.75 1.81 (LSD 5% = ±0.12)

Total Valepotriates 0 (11.50) 1 10.31 10.21 9.91 9.61 9.19 8.40 6.38 4.03 0.84 7.65 2 10.10 9.51 9.03 9.60 7.20 4.17 4.46 1.34 0.04 6.16 3 9.76 9.02 7.32 8.62 6.31 2.47 3.29 0.66 Nd 5.27 4 9.54 8.30 5.90 8.06 5.11 1.33 2.65 0.28 Nd 4.58 5 9.48 7.83 4.69 7.72 3.80 0.77 2.13 0.18 Nd 4.06 6 9.11 7.33 3.70 7.71 3.36 0.49 1.94 0.12 Nd 3.75 Mean 9.72 8.70 6.76 8.55 5.83 2.94 3.47 1.10 0.15 5.24 (LSD 5% = ±0.45) Essential Oil 0 (0.45) 1 0.45 0.43 0.45 0.43 0.42 0.45 0.40 0.44 0.39 0.43 2 0.45 0.43 0.45 0.40 0.41 0.45 0.23 0.33 0.22 0.37 3 0.42 0.44 0.45 0.33 0.38 0.45 0.18 0.28 0.19 0.34 4 0.39 0.44 0.45 0.31 0.38 0.44 0.12 0.18 0.14 0.32 5 0.36 0.44 0.46 0.30 0.39 0.44 0.12 0.18 0.12 0.31 6 0.34 0.45 0.46 0.26 0.37 0.41 0.10 0.17 0.09 0.29 Mean 0.40 0.44 0.46 0.34 0.39 0.44 0.19 0.26 0.19 0.34 (LSD 5% = ±0.04) * 5, 14 and 20 are the storage temperatures (°C); L, M, and H are the storage humidities of 10, 40 and 80% RH

19

Figure 3.4 Interaction between humidity and temperature on the level of total valepotriates in stored valerian root powder.

Data are the mean values over the storage period

0

2

4

6

8

10

12

Low Moderate High

Humidity

Tota

l Val

epot

riate

s (m

g/g) 5°C

14°C 30°C

3.3.2 Effect of light The effect of light on the active constituents was determined with dried homogeneous powder generated from the six valerian roots that was distributed over four glass petri dishes and stored in an environment of 20°C and 40-60% RH. Two samples were placed under a 60 W tungsten lamp and two were kept in the dark. The active constituents were determined over a 6 month period.

20

Table 3.8 Active constituents (mg/g), essential oil ( ml/100 g) and moisture (g/100 g) in valerian root powder stored at 20°C and ambient humidity in the dark and light.

Time Amount present (months) TVA TVP Oil H2O

Light 0 3.16 6.52 0.66 7.8 1 3.18 3.58 0.55 9.5 2 3.02 1.47 0.55 9.9 3 2.87 0.69 0.56 10.1 4 2.80 0.30 0.55 9.4 5 2.64 0.19 0.50 9.4 6 2.51 0.07 0.48 7.8 Dark 0 3.16 6.52 0.66 7.8 1 3.21 3.73 0.56 10.1 2 3.10 1.79 0.56 11.4 3 3.10 0.83 0.57 11.7 4 2.97 0.37 0.56 10.9 5 2.88 0.21 0.53 10.2 6 2.73 0.09 0.52 9.1

LSD5% ±0.08 ±0.06 ns ±0.45

Table 3.8 shows that the loss of total valerenic acids was significantly greater in samples exposed to light than those held in the dark with a significant difference present at 2 months storage. The valepotriate content decreased rapidly in both treatments during storage and also tended to be greater in root powder exposed to light. There was no significant difference in the essential oil content in the light or dark. A possible confounding factor was that the light could have had a slight warming effect leading to the moisture content of root powder in the light at 8-10% being significantly lower than powder held in the dark at 9–12%. The temperature was about 4°C higher in the light. 3.3.3 Effect of blanching and vacuum packing To evaluate whether enzymic activity or atmospheric oxidation were involved in degradation of active constituents, the roots from 12 valerian plants were harvested and groups of three roots were blanched by simmering in boiling water for 0, 1, 2 and 4 min. The roots were then dried and ground to a powder. The powdered root from each treatment was distributed over two plastic petri dishes which were stored at low humidity. The unblanched root powder was also vacuum packed into four polyethylene vacuum bags (EVAC Equipment, Crestmead Qld) that were stored under low humidity. The level of active constituents were determined after 3 and 6 months. The data in Table 3.9 show that blanching caused about a 30% loss (P< 0.05) of valerenic acids but the loss during storage was greater in the unblanched root powder so that a similar level of valerenic acids was present in all blanching times after 6 months storage. The rate of loss of valerenic acids was much lower during the storage of root powder held in vacuum bags (P<0.05). The level of valepotriates was similarly affected by blanching but the rate of

21

loss of blanched material was much lower than the unblanched which had the lowest level after 6 months. The greatest rate of loss was with root powder held in the vacuum pack. Table 3.9 Active constituents (mg/g) in valerian root that had been blanched and vacuum packed then stored at 30°C and low humidity.

Storage Amount present month B1 B2 B4 B0 Vac Total valerenic acids 0 1.31 1.11 1.35 1.84 1.84 3 0.79 0.60 0.77 0.77 1.49 6 0.55 0.41 0.53 0.54 1.29 LSD 5% = +0.28 Total valepotriates 0 12.83 10.31 7.62 9.54 9.54 3 9.39 8.00 5.61 2.21 0.26 6 7.30 6.22 3.97 0.42 0.09 LSD 5% = ±2.57

B = blanched for 0, 1, 2, or 4 min, Vac = vacuum packed. 3.4 Implications for industry The difficulty of removing soil from around the complex structure of valerian roots has been cited by various current or past growers as a deterrent to the growing of valerian. The evaluation of cutting and soaking of roots showed no overall benefit in reducing the time taken for the operation. While there was a reduction in washing time due to quartering roots and removing rootlets from the crown there was no beneficial effect of soaking and when the time taken to cut the plants was added to the washing time, there was no overall time saving. A potential benefit of cutting was in a greatly reduced drying time in a hot air drier with rootlets drying in 20-30% of the time taken by whole roots. It would therefore seem that a more efficient use of driers would be obtained by routinely separating rootlets from the crown and drying each plant part in a separate batch. The shorter drying time of rootlets is presumably due to the relatively greater surface area of the smaller particles. A further advantage is that there was a greater retention of valerenic acids and essential oil in dried rootlets and crown that had been separated before drying. In addition, rootlets tended to have a higher concentration of valerenic acids than the crown which could allow segregation for sale as a higher quality product. The feasibility of separately drying rootlets and crown was further enhanced by the storage for 10 days of whole roots closely stacked in a wire basket at ambient temperature and humidity. There was no significant change in the level of any active constituent but there was a substantial loss of moisture which would further reduce the time roots needed to be held in a drier. The drying temperature, as expected, was directly related to the drying time with a 12-fold decrease over the temperature range of 15°-70°C. There was, however, a substantial decrease in the level of valerenic acids and essential oil at higher drying temperatures with the most marked change occurring between 40° and 50°C. It would thus seem that the

22

temperature of a hot air drier should be maintained at about 40°C The heat pump drier with its use of reduced humidity air was found to give a 25% shorter drying time at 38°C than a hot air drier operated at the same temperature with no adverse effect on any active constituent. The benefit of faster drying and hence greater drier throughput would need to be considered against the higher cost but lower energy usage of a heat pump drier. Storage recommendations for dried valerian are in a closed container protected from light, air and moisture although little is known about the fate of the valerenic acids during such storage. This study found that the valerenic acids were quite unstable during storage with the rate of loss increasing as the temperature increased and the humidity decreased with >50% loss in root held at 30°C in air of 10% RH over 6 months. Changes in essential oil followed a similar trend to the valerenic acids but loss of valepotriates while increasing with increasing temperature showed greater losses at higher humidity. Exposure to light further accelerated the loss of valerenic acids and valepotriates but had little effect on essential oil content. Thus, retention of valerenic acids and essential oil is favoured by storage at low temperature in the dark. It would seem to be also favoured by retention of a high humidity atmosphere but no explanation can be offered as to the mechanism resulting in such an effect. An evaluation of whether enzymic activity or atmospheric oxidation were involved in degradation of active constituents showed that substantial loss of valerenic acids occurred during the blanching process but the loss during storage was greater in unblanched root powder so that a similar level of valerenic acids was present in blanched and unblanched material after 6 months storage. The rate of loss of valerenic acids was much lower during the storage of root powder held in vacuum bags where oxygen was at least initially reduced. This suggested that losses through both enzymic and chemical oxidation was occurring but that use of water blanching was not a feasible commercial option. The studies were only on ground valerian and thus only have direct application for processors. However, there is no reason to doubt that the findings would apply to dried, non-ground root, although the rate of degradation may be slower

23

4. Processing into Manufactured Products Traditional usage of medicinal herbs was by natural medicine practitioners or individual informed citizens who utilised dried root or aerial plant parts by direct consumption or through production of a range of infusions or poultices. However, most medicinal herbs in Western society are now also used through a range of manufactured products such as tablets, capsules and liquid extracts of ethanol or glycerol, that may also contain other added ingredients. These products are increasingly generated by large manufacturers using standard pharmaceutical or food processing techniques and made available to naturopaths or directly to consumers through pharmacies, supermarkets and alternative lifestyle outlets. A study was conducted to investigate several of the parameters involved in the ethanolic extraction of valerian root for their effect on the level of active constituents. In addition, an assessment was made of the levels of active constituents in manufactured valerian products available from retail outlets in the Sydney-Central Coast region of New South Wales. 4.1 Efficiency of alcoholic extraction of active constituents A common processing method for the manufacture of valerian products is by initially obtaining an alcoholic extract from dried plant material with a solvent containing a mixture of ethanol and water. The alcoholic extract is then either marketed as a liquid, with possible adjustment of the alcohol content, or spray dried and marketed, with an added solid filler, as a tablet or capsule. There is little published literature on the effect of processing operations in extraction of active constituents of valerian. 4.1.1 Ethanol concentration and method of extraction A full range of ethanol/water mixtures from 0-100% were used to extract powdered valerian using two methods, maceration and percolation. Maceration involved placing the root powder with the solvent and soaking at room temperature for 18 hr. Percolation involved allowing the solvent to flow by gravity through a column of ground valerian. The valerian bed was continuously supplied with fresh solvent for 1-2 hr with the final solution being a 5:1 ratio of solvent to solute. The extracts from both methods were filtered, made up to volume and an aliquot analysed directly for valerenic acids and valepotriates. The remaining extract was stored in a sealed brown glass jar and stored for 30 days at 20°C then re-analysed for active constituents. Table 4.1 shows that the level of total valerenic acids in an extract obtained by maceration and percolation significantly increased with an increase in ethanol concentration. The greatest increase in the extraction of valerenic acids occurred in the range 0% to 50% with no significant difference in extraction with 70-100% ethanol. Percolated extracts contained significantly higher levels of valerenic acids over the whole ethanol range than those obtained by maceration with the extraction being about 20% greater in the 70-100% ethanol range. The level of valerenic acids in any extract did not significantly change over the 30 day storage period. Valepotriates were not detected in extracts of less than 40% ethanol but the level extracted increased significantly as the ethanol concentration further increased (Table 4.2).

24

Maximum extraction of valepotriates occurred with 90-100% ethanol. Percolation was more effective than maceration in extraction of valepotriates with about 20% more obtained at 90-100% ethanol. The concentration of valepotriates in all extracts declined significantly during storage. Table 4.1 Total valerenic acids (mg/g valerian root) in extracts of different ethanol concentrations obtained by maceration and percolation and after storage for 30 days at 20°C.

% Ethanol Amount in fresh extract Amount in extract after 30 days Maceration Percolation Maceration Percolation 0 0.12 0.36 0.14 0.37 20 0.43 0.77 0.46 0.68 40 1.38 1.47 1.36 1.37 50 1.86 1.99 1.76 1.92 60 1.93 2.27 1.95 2.14 70 2.05 2.43 2.13 2.36 80 2.19 2.52 2.18 2.51 90 2.17 2.56 2.21 2.54 100 2.10 2.53 2.21 2.53

LSD 5% ±0.25 Table 4.2 Total valepotriates (mg/g valerian root) in extracts of different ethanol concentrations obtained by maceration and percolation and after storage for 30 days at 20°C.

% Ethanol Amount in fresh extract Amount in extract after 30 days Maceration Percolation Maceration Percolation 0 nd nd nd nd 20 nd nd nd nd 40 0.34 0.08 0.01 nd 50 1.64 1.79 0.04 0.07 60 2.69 3.44 0.16 0.12 70 4.06 5.69 0.80 0.51 80 4.88 6.57 1.56 1.67 90 5.51 7.05 2.58 2.49 100 6.03 7.08 4.45 4.68

LSD 5% ±0.34

25

4.1.2 Effect of maceration parameters on extraction Additional studies were conducted to examine if the extraction of active constituents could be improved by altering the time of maceration, the ratio of valerian root to macerating solvent, and the solvent temperature. 4.1.3. Maceration time The effect of maceration time on extraction of active constituents was examined using 80% ethanol. The data in Table 4.3 show that the extraction of total valerenic acids increased with increasing maceration time but was only about 10% greater after 75 hr compared to the lowest maceration time of 4 hr. The valepotriates showed the opposite trend decreasing with increasing maceration time but the major difference was between the 4 hr and 75 hr maceration. Table 4.3 Active constituents (mg/g) in valerian extracts obtained by maceration for different times with 80% ethanol.

Maceration Amount in extract

time (hr) TVA TVP 4 1.73 4.30 6 1.74 3.92 8 1.71 3.72 10 1.70 3.55 12 1.80 3.79 14 1.80 3.78 18 1.88 3.70 32 1.91 3.81 75 1.98 2.86

LSD 5% ±0.09 ±0.28 4.1.4 Ratio of solvent to valerian The effect of solvent to valerian ratio was examined by macerating powdered valerian with 80% ethanol using solvent:valerian ratios of 5:1, 7.5:1, 10:1 and 15:1. The extraction of both total valerenic acids and total valepotriates was increased in a solvent:valerian ratio of 10:1 or higher compared to 7.5:1 or lower (Table 4.4) with the concentration of the extracts about 10% greater. Table 4.4 Active constituents (mg/g) in valerian extracts obtained by maceration in 80% ethanol with different ratios of solvent to valerian powder.

Solvent:valerian ratio TVA TVP 5:1 2.15 4.48 7.5:1 2.13 4.50 10:1 2.38 5.19 15:1 2.45 4.99 LSD 5% ±0.20 ±0.38

26

4.1.5 Temperature of solvent Powdered valerian was macerated for 1 hr with 80% ethanol that had been pre-heated to 45° and 65°C as well as solvent at room temperature. Increasing the solvent temperature had no significant effect on the concentration of total valerenic acids in the extract but the concentration of valepotriates decreased significantly from 4.5 mg/g in the 20°C solvent to 3.8 and 3.3 mg/g, respectively, in the 45° and 65°C solvents. 4.1.6 Percolation The efficiency of percolation was examined with 60% and 80% ethanol that was passed through the valerian bed at different flow rates. The data in Table 4.5 show that flow rates of 1 drop/2 sec and 1 drop/5 sec of 80% ethanol through the powdered root did not affect the elution pattern of the valerenic acids and valepotriates over the course of the extraction. About 75% of the extracted compounds was present in the first 10 ml of eluate with about 20% in the second 10 ml. A similar relationship was obtained when percolating with 60% ethanol at 1 drop/2 sec except that the extraction was slightly slower with about 70% and 15% in the first two fractions. As in the previous study, 80% ethanol extracted a significantly higher amount of active constituents than 60% ethanol. Table 4.5 Active constituents (mg/g) in valerian extracts obtained by percolation with 80% and 60% ethanol at different flow rates with the eluate collected in 10 ml fractions.

Vol (ml) TVA TVP

Flow 1 drop/2 sec, 80% ethanol 10 1.91 4.77 20 0.47 1.28 30 0.05 0.17 40 0.03 0.14 50 0.02 0.10 70 0.02 0.06 Flow 1 drop/5 sec, 80% ethanol

10 1.88 4.78 20 0.48 1.24 30 0.07 0.22 40 0.04 0.17 50 0.02 0.08 70 0.02 0.02 Flow 1 drop/2 sec, 60% ethanol

10 1.60 2.55 20 0.29 0.51 30 0.14 0.24 40 0.10 0.16 50 0.04 0.06 70 0.02 0.02

LSD 5% ±0.21 ±0.68

27

4.1.7 Supercritical fluid extraction with CO2 Supercritical fluid extraction (SFE) is a relatively recent technology that has been found useful to extract a range of minor components from biological substrates. The efficiency of SFE extraction is achieved by the supercritical CO2 having properties of both a gas and liquid giving rise to high permeability into plant tissues and high solubility of non-polar organic constituents. A major advantage of the technology arises from its use of compounds such as carbon dioxide as the extracting solvent. The carbon dioxide can be easily removed by letting the temperature rise to ambient where the carbon dioxide will volatilise and in the process result in a concentration of the extract. Experiments with SFE were designed to determine the effect of time, pressure, temperature and modifiers on the extraction of valerenic acids and valepotriates. An Isco (Lincoln, NJ) SFE apparatus consisted of two 260D syringe pumps, an SFX 2-10 supercritical fluid extractor, a restrictor temperature controller and a 260 series pump controller. Extractions were performed on 0.5g samples of root powder and the extract was collected in methanol with aliquots collected every 10 min over an appropriate period and analysed for active constituents. Three pressures, 10, 15 and 20 MPa were examined at an extractor temperature of 40°C over 40 min. The extractor temperature was increased to 45°C and 50°C at 15 MPa with extractions monitored over 30 min. Using a pressure of 15 MPa at 40°C, the addition of ethanol and methanol modifiers at 5% was monitored over 30 min. The results of the SFE studies over time are presented in Table 4.6. For all constituents and extractions (except at 10 MPa), >90% of the total extraction occurred in the first 10 min and by 20 min 97% was extracted. The increase in CO2 pressure from 10 MPa to 15 MPa resulted in a faster extraction in the first 20 min but the total extraction after 30 min was not significantly different. There was no difference in extraction between 15 and 20 MPa. The temperature of the extractor had no effect on extraction of valerenic acids but higher temperatures resulted in a significant decrease in valepotriate concentration.

28

Table 4.6 Extraction of total valerenic acids and valepotriates (mg/g) over time by SFE using CO2 at different pressures, temperatures and modifiers

Total valerenic acids Total valepotriates Time (min)

/Pressure 10 15 20 10 15 20 Mpa

10 1.33 1.83 1.83 5.38 5.55 5.29 20 0.45 0.16 0.14 0.32 0.16 0.32 30 0.24 0.05 0.02 0.25 0.07 0.09 40 0.08 nd nd 0.09 0.05 0.05

LSD 5% ±0.17 ±0.32 /Temp 40 45 50 40 45 50°C

10 1.83 1.68 1.78 5.55 5.26 4.79 20 0.16 0.26 0.19 0.16 0.21 0.17 30 0.05 0.07 0.05 0.07 0.08 0.06

LSD 5% ±0.15 ±0.14 /Modifier None Ethanol Methanol None Ethanol Methanol

10 1.79 2.17 2.32 5.52 4.91 5.27 20 0.15 0.12 0.11 0.15 0.09 0.11 30 0.05 0.06 0.05 0.06 0.06 0.04

LSD 5% ±0.10 ±0.09

The addition of modifiers at 5% significantly increased the extraction of valerenic acids in the first 10 min and over the 30 min. The total amount extracted in 30 min with ethanol modifier (2.35 mg/g) was not significantly different to percolation (2.53 mg/g) and significantly greater than with maceration (2.10 mg/g) with 100% ethanol. Extraction of valepotriates was adversely effected by modifiers with the lowest concentration found when ethanol modifier was used. The total extraction of valepotriates over 30 min without modifier (5.73 mg/g) was significantly less than from percolation (7.08 mg/g) but not significantly different to maceration (6.03 mg/g) with 100% ethanol.

29

4.2 Levels of active constituents in manufactured products Twenty-eight solid and 3 liquid products labelled to contain only valerian were purchased during November 1999 and March 2000. Where possible, 2 samples with different batch numbers were obtained and resulted in the purchase of 55 samples for analysis. Each sample was assayed twice and the residue after filtration was re-extracted in methanol, assayed, and added to the total. The products analysed with the manufacturer and country of manufacture as stated on the product labels were:

• Teas: Healthy Life Valerian Rootlets, Healthy Life, France; Blooms Valerian Root, Blooms Health Products, Australia; Hilde Hemmes' Herbals Valerian, Herbal Supplies, Australia; Colonial Farms Valerian Rootlets, Select Foods, Australia; Russell's Valerian Tea, Russell's Natural Foods, France.

• Tablets: Earth's Own Valerian 2500, Allied Master Chemists, Australia; Valerian Forte, Blackmores, Australia; Valerian Herb-Relax 2000, Blooms Health Products, Australia; Fingerprint Botanicals Valerian 1000, Bullivant's Natural Health, Australia; Chemworld Valerian 500mg, Chemworld Chemist, Australia; Healtheries Valerian 500mg, Health Minders, Australia; Ethical Nutrients Valerian 1000, Health World, Australia; Herbal Valerian 500mg, Herb Valley, Australia; Valerian, Herron, Australia; Valerian, Natures Way Health, Australia; Cirkulin Valerian Tablets, Polcopharma, Germany; Valerian 2000, Soul Pattinson, Australia; Valerian, VitaGlow, Australia.

• Powder Capsules: Bio-organics Valerian 2250, Bullivant's Natural Health, Australia; Nature's Own Valerian 500mg, Bullivant's Natural Health , Australia; Hilde Hemmes' Herbals Valerian, Herbal Supplies, Australia; Kordel's Valerian 1000, Kordel, Australia; Valerian Root, Nature's Sunshine, USA;Nature's Path Valerian Root, Planet Health, USA; Valerian 1000, Vitaplex Products, Australia.

• Soft Gel Capsules: Earth's Own Valerian 1000, Allied Master Chemists, Australia;Valerian 100mg, Herb Valley, Australia; Valerian 500, Soul Pattinson, Australia.

• Liquids: Valerian, Greenridge Botanicals, Australia; Hilde Hemmes' Valerian Root, Herbal Supplies, Australia; Valerian, Thursday Plantation, Australia

4.2.1 Total valerenic acids and valepotriates by product weight The level of total valerenic acids in the 55 samples ranged from <0.01 to 6.32 mg/g or ml of product while valepotriates were found at low levels (<1.0 mg/g) in some valerian teas, but were not detected in any of the finished products. There were 16% of samples containing <0.1 mg/g of valerenic acids and included three tea samples and two liquid samples at non-detectable levels. Samples with higher levels comprised 31% with 0.1-1 mg, 33% with 1-2 mg and 20% >2 mg/g or ml. In general, products were consistent between batches except for one tea, which returned a non-detectable amount followed by a concentration of 1.36 mg/g in the repeat purchase. The range and average level of valerenic acids in each product class are given in Table 4.7. Powder capsules, on average, contained the highest concentration of valerenic acids on a product weight basis (2.46 mg/g). The tablets, teas and soft gel capsules had an average valerenic acid content of about 1 mg/g while liquids had the lowest average concentration (0.47 mg/ml) of all product classes.

30

Table 4.7 Mean concentration and range of total valerenic acids in classes of manufactured valerian products.

Total Valerenic Acids Product Type No. mg/g Product mg/g Root

Mean Range Mean Range Tea 5 0.98 <0.01 – 1.64 0.98 <0.01 – 1.64 Tablet 13 1.21 0.07 – 3.25 0.97 0.06 – 1.88 Powder capsule 7 2.46 0.47 – 6.32 1.22 0.26 – 2.00 Soft Gel Capsule 3 0.89 0.26 – 1.20 0.44 0.19 – 0.68 Liquid 3 0.47 <0.01 – 1.31 0.54 0.01 – 0.94 The 31 products could be divided into five standardised products (that is, those with a stated valerenic acid level on the label) and 26 non-standardised products which did not contain a stated valerenic acid content. Standardised products (nine samples based on the dual purchase dates) contained valerenic acids at >2 mg/g whereas only one non-standardised sample was in this range. The mean concentration of valerenic acids in the standardised products (3.56 mg/g) was significantly higher than in the non-standardised products (0.89 mg/g). A significant linear regression of the analysed level of valerenic acids against the stated level in the nine standardised samples (Figure 4.1) was highly significant (y= 0.87x-0.09; P<0.001) indicating a high level of correlation with label claims for medicinal efficacy. However, the analysed level of valerenic acids was 10-15% lower than the label claim. This discrepancy is not great and could arise from different analytical methods used by the manufacturers to that in this study.

31

Figure 4.1 Regression of the total valerenic acids found in products with a labelled level of

valerenic acids against the stated level of valerenic acids.

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

0.00 2.00 4.00 6.00 8.00

Valerenic acids stated (mg/g)

Vale

reni

c ac

ids

foun

d (m

g/g)

4.2.2 Total valerenic acids on an added valerian basis All products were labelled with the amount of valerian root added, expressed as equivalent dried root per unit which ranged from 0.2 to 3.46 g/g or ml. A calculated value of the concentration of valerenic acids in the product as mg/g valerian root was derived from the product label claim of added valerian. The calculated concentrations were found to range from <0.01 to 2.0 mg/g. The data for product types are presented in Table 4.7 and show that the average valerenic acids concentration for the tea, tablet and powder capsule products was about 1.0 mg/g root whereas the liquid and soft gel products contained about 0.5 mg/g root. The relationship between total valerenic acids and the amount of added valerian showed a significant linear regression for the standardized samples (y= 1.71x-0.05; P<0.001). There was, however, no significant relationship for the non-standardized samples, that is, the amount of valerenic acids in the product was not related to the amount of added valerian. It was also noted that information on recommended daily doses appeared on the labels of all products except for two teas but the claims exhibited large differences in terms of both the recommended amount of valerian root (0.5 to 6.0 g/day) and total valerenic acids (<0.01 to 8.16 mg) in the recommended dose.

32

4.3 Implications for industry and consumers Commercial extraction of valerian root commonly uses aqueous ethanol solvents in the ratio range of 60:40 and 70:30 ethanol:water. The use of such solvents by industry is in large part supported by this study as little change in extraction of total valerenic acids was obtained at ethanol concentrations greater than 60%. However, the use of 70% and possibly even 80% ethanol would appear to result in a 10% additional extraction of valerenic acids over the use of 60% ethanol. The findings in this study are generally in line with Dutch data by Bos et al (1996) although they hastily concluded that the extraction of valerenic acids was more or less constant at ethanol concentrations above 50%. The good stability of valerenic acids in all extracts during storage at ambient temperatures should provide flexibility for industry to efficiently manage either long term storage or holding for further processing while maintaining product quality. The desirability of valepotriates in a final product is subject to some contention, and while valepotriates were adequately extracted in 70-80% ethanol, they were relatively unstable during storage. This may or may not be an issue for industry. Percolation was shown to be more efficient than maceration in extracting valerenic acids with about 15% more valerenic acids, and also of valepotriates, at the same ethanol concentration. Since the rate of solvent flow during percolation with 80% ethanol, at least between the speeds of 1 drop/2 or 1 drop/5 seconds, did not affect the extraction, the faster flow rate could be used with a considerable time saving without detriment to the final concentration of compounds in the extract. Furthermore, valerenic acids were readily extracted with percolation achieving about 95% extraction from 10 g valerian root in the first 20 ml of 80% ethanol, equating to a solvent:valerian ratio of 2:1. At 60% ethanol about 30-40 ml of solvent was required to achieve a similar rate of extraction. The implication for industry is that a cost efficiency exercise will be required to justify whether it is economical to pursue to final 5% of valerenic acids. There is presently, in Australia, no price premium for products based on valerenic acids content and quality can be a subjective choice of the manufacturer based on perceived advantage in the market place. Such a situation can of course change with time as consumers become more aware of variations in product quality and demand greater uniformity based on quality standards. The ease of extraction of valerenic acids was also apparent in the SFE extraction with CO2 where >90% of extraction occurred in the first 10 min of the 30 min run and with use of relatively mild conditions of 15 MPa and 40°C. While there was no advantage to increasing either the pressure or the temperature, the addition of ethanol at 5% extracted significantly greater amounts of the valerenic acids and the efficiency was comparable to extraction by percolation. The advantage of SFE is elimination of the need to handle large volumes of solvents and its more benign environmental and health safety features. However, its commercial applicability to medicinal herbs needs further investigation. The survey of commercial manufactured products showed a considerable variation in concentration of valerenic acids from <0.01 to 6.32 mg/g or ml of product with about 50% of products containing less than 1 mg/g or ml. The minority of products with a stated label content of valerenic acids had much higher valerenic acid contents than non-standardised products and the stated and actual levels were reasonably well correlated. There was, however, a 10-15% lower level in products than the label claim. This could merely be due to different analytical methods used by the manufacturers than that used in this study, but it

33

highlights a need for some industry collaboration. Consumers may initially welcome products having a stated level of active constituents close to the actual but will eventually demand that the actual level be either equal to or exceed the label claim. The discrepancy may also be due to unrecognised losses incurred during processing and or storage. The variation in valerenic acids in non-standardised products is probably explained in large part by the quality of the raw material used for processing and then by losses during processing. It is not known to what extent manufacturers exercise quality control over purchased valerian root but such information is vital if a product of consistent quality is to be traded. The lack of valepotriates in finished medicines has been reported in other studies (e.g. van Meer et al., 1977; Bos et al., 1996) and is due to their thermolability and instability under acidic or alkaline conditions and in alcoholic solutions (Graf and Bornkessel, 1978; Hänsel and Schulz, 1985; Bos et al, 1996). This instability makes it unlikely that valepotriates remain following processing. The calculated values of valerenic acids in products in relation to the amount of added valerian ranged from <0.01 to 2 mg/g. This is similar to the range reported in European studies (Hänsel and Schulz, 1985; Bos et al., 1996). Hänsel and Schulz (1985) have recommended that valerian extracts should contain valerenic acids at a minimum of 1.2 mg/g for alcoholic extracts and 0.6 mg/g for aqueous extracts. This study found 50% of extracts contained <0.6 mg/g while all standardized products were manufactured from extracts containing >1.2 mg/g. Hänsel and Schulz (1985) and Schimmer and Röder (1992) found that water extracted about 25% of the valerenic acids but our study found percolated water extracts contained only 0.36 mg/g which was 14% of the concentration found in the 100% ethanol extract. Therefore, the expectation of 0.6 mg/g in aqueous extracts may be too high. The large variation in recommended daily doses on product labels presumably stems from the large variation of 2-9 g/day that has been proposed as a recommended daily dose in Europe (ESCOP, 1993). It would seem that a wide range of recommended daily doses on product labels would be confusing to consumers. It was noted that for about 50% of products the recommended dose was actually <2 g/day. This leads to a suggestion that labelling of products with a more uniform total valerenic acid content and recommended usage would give consumers greater confidence in the continued purchase of valerian products.

34

5. Summary of Conclusions and Recommendations 5.1 Need for analysis of active constituents If the Australian valerian industry is to grow in the face of increasing international production, competition from multinational manufacturing companies and increasing demands by consumers, it needs to become more quality conscious. The essential quality factor for valerian is the presence of active constituents, with the valerenic acids being a recognised factor that confer a health benefit to consumers. Development by the project of efficient and reliable quantitative analytical methods for the analysis of the valerenic acids as well as the valepotriates and baldrinals by HPLC, and their application in the research program has created the potential for industry to monitor the quality of Australian products and processes. The University of Newcastle, through its business arm TUNRA (The University of Newcastle Research Associates Ltd), has been able to supply an ongoing commercial analytical service trading as Hespan Laboratories to provide client confidential information on valerian quality. 5.2 Variation in active constituents in crops The Australian valerian industry currently favours cultivation of the Anthos cultivar and the finding that the level of valerenic acids in commercially available Anthos crops at about 3 mg/g would allow Australia to qualify as a producer of high quality valerian since a standard for high quality valerian has been proposed by Stahn and Bomme (1998) to be 2–2.5 mg/g valerenic acids and >2.5 mg/g as very high while Bos et al. (1998) proposed high quality valerian to contain 3 mg/g. This study has, however, found that there is the potential for the Australian industry to increase the valerenic acids content to >4 mg/g by utilising new planting material. Of 25 sources of valerian seed, five were found to produce roots that contained valerenic acids at >4 mg/g. Surprisingly three of these sources were Australian growers with one each living in New South Wales, Tasmania and Victoria. The international sources of high quality were from France and Russia. The most promising was the Russian material as the high valerenic acids content was matched by a larger than normal root size. These plant sources are worthy of further growing trials to evaluate their agronomic performance under a wider range of environmental conditions and to test the consistency of the elevated valerenic acids levels. Processors have expressed a preference to purchase roots with a high concentration of valerenic acids to enable high concentration extracts to be more readily generated. This would seem to require a change in price setting to also reflect root quality. Growers are currently paid by weight and hence the only incentive is to market the largest roots possible. Vigorous growing plants are also favoured as the canopy is quicker to close and the requirement for weeding is reduced. A conflict between quality and size was shown by the study on changes in active constituents during plant growth. There was a substantial intra-seasonal fluctuation and accumulation of active constituents showed a peak concentration for valerenic acids

35

occurring in spring with a considerable decline in concentration at the current harvest time at the senescence stage of growth. The root size, however, continued to increase throughout the season and was greatest at the senescence stage. Notwithstanding the decrease in concentration, the overall yield of valerenic acids per root was actually higher at the senescent stage. Since the growing of valerian in Australia has limited appeal, manufacturers may also wish to maintain the crop yield and hence yield of valerenic acids at the highest quantity possible and modify processing systems to cope with a lower concentration of active constituents. 5.3 Need for improved postharvest handling practices The removal of soil from around the complex structure of valerian roots is laborious and time consuming and a deterrent to the growing of valerian. A range of cutting and soaking operations was found to offer no overall benefit in reducing the time taken to clean roots. However, the operational efficiency of washing would depend on the soil type and the equipment used at each individual farm. A potential benefit of cutting rootlets from the crown was in a much reduced drying time with rootlets drying in 20-30% of the time taken by whole roots. A more efficient use of driers would be obtained by routinely separating rootlets from the crown and drying each plant part in a separate batch. The shorter time in a drier for rootlets also resulted in a greater retention of valerenic acids and essential oil which could allow their segregation for sale as a higher quality product. Efficient utilisation of a drier can be furthered by the storage of roots for at least 10 days at ambient temperature and moderate relative humidity as no significant change in the level of any active constituent was found during this period. There was, however, a substantial loss of moisture which would further reduce the time roots needed in a drier. Despite the drying time being substantially reduced by increasing the drying temperature, the substantial decrease in the level of valerenic acids and essential oil at higher drying temperatures means that a hot air drier should be operated at about 40°C. Use of a heat pump drier at about 40°C resulted in a shorter drying time than a hot air drier at the same temperature with no adverse effect on any active constituent. The benefit of faster drying and resultant greater drier throughput would need to be considered against the higher purchase price but lower energy use of a heat pump drier. The storage of dried valerian should be at low temperature in the dark to maximise retention of valerenic acids the losses of which can be rapid and substantial under unfavourable environmental conditions. Rather surprisingly, retention of valerenic acids was favoured by retention of a high humidity atmosphere but no explanation can be offered to explain this effect. Water blanching of valerian roots was found not to be a worthwhile pre-storage treatment as while the losses during storage were reduced there was a substantial loss during the blanching process. The studies were only on ground valerian but there is no reason to expect that dried, non-ground root would behave differently except that the rate of degradation may be slower. 5.4 Need for improved processing operations The commercial extraction of valerian root in aqueous ethanol with 60-70% ethanol was in large part supported by this study as giving a reasonable extraction rate of valerenic acids with a preference for the higher end of the range, and possibly even 80% ethanol which would give an additional 10% extraction of valerenic acids over the use of 60% ethanol.

36

The valerenic acids were found to be stable in all extracts during storage at ambient temperatures which should provide flexibility for industry to efficiently manage either long term storage or holding for further processing while maintaining product quality. Extraction of valerenic acids by percolation was more efficient than maceration with about 15% more valerenic acids at the same ethanol concentration. Within the range examined, solvent flow during percolation did not affect the extraction allowing use of relatively fast flow rates to save time without detriment to the concentration of the extract. Extraction was also relatively rapid with percolation with 60% ethanol achieving 95% extraction with a solvent:valerian ratio of about 3:1. Maceration also showed a similar early extraction of most of the valerenic acids in 4 hr Use of SFE with CO2 plus 5% ethanol and relatively mild conditions of 15 MPa and 40°C was also quite efficient in extraction of valerenic acids with 95% yield obtained in 10 min, an extraction efficiency comparable to percolation. The advantage of SFE is elimination of the need to handle large volumes of solvents and its more benign environmental and health safety features. While the technology has found some commercial application, further studies on a larger scale would be required to prove the potential for use with valerian. 5.5 Quality of retail manufactured products The survey of commercial manufactured products showed a considerable variation in concentration of valerenic acids with about 50% of products containing less than 1 mg/g or ml which included 16% with <0.1 mg/g or ml. The minority of products with a stated label content of valerenic acids had much higher valerenic acid contents than non-standardised products and the stated and actual levels, while reasonably well correlated, showed, a 10-15% lower level in products than the label claim. This difference could be due to different analytical methods used by the manufacturers than that used in this study, but highlights a need for some industry-wide collaboration. Consumers may initially welcome products having a stated level of active constituents close to the actual but will eventually demand that the actual level be either equal to or exceed the label claim. The discrepancy may also be due to unrecognised losses incurred during processing and or storage. The variation in valerenic acids in non-standardised products is probably explained in large part by the quality of the raw material used for processing and then by losses during processing. It is not known to what extent manufacturers exercise quality control over purchased valerian root but such information is vital if consistent quality is to be maintained in traded products. This study found 50% of extracts contained <0.6 mg valerenic acids/g added valerian which is below European recommendations that valerian extracts should contain a minimum of 1.2 mg/g for alcoholic extracts and 0.6 mg/g for aqueous extracts. All standardized products were found to contain >1.2 mg/g. However, based on findings in this study it is considered that the European recommendations for the quality of aqueous extracts may be too high. The large variation in recommended daily doses on product labels presumably stems from the range of 2-9 g/day proposed by different groups in Europe as a recommended daily dose. It would seem that a wide range of recommended daily doses on product labels would be confusing to consumers. It was noted that for about 50% of products the recommended

37

dose was actually <2 g/day. This leads to a suggestion that labelling of products with a more uniform total valerenic acid content and recommended usage would give consumers greater confidence in the continued purchase of valerian products.

38

6. References

AOAC (1995) Official Methods of Analysis. International 16th edn Cunliff, P. (ed.) AOAC International Arlington VA. Chapter 43 pp3-4.

ESCOP (1993). European Scientific Cooperative on Phytotherapy :Proposals for the summary of product characteristics: Valerianae radix., Brussels.

Bernáth, J. (1997). Cultivation of valerian. In: Valerian. The Genus Valeriana. Houghton, P.J. (ed.). Hardwood Academic Publishers, Amsterdam, The Netherlands. pp. 77-100.

Bos, R., Putten, F.M.S van. and Hendriks, H. (1986). Variations in the essential oil content and composition in individual plants obtained after breeding experiments with a Valeriana officinalis strain. In: Progress in Essential Oil Research. Brunke, E. J. (ed.). Walter de Gruyter, Berlin. pp. 123-130.

Bos, R., Woerdenbag, H.J., Hendriks, H., Zwaving, J.H., De Smet, P A.G.M., Tittel, G., Wikström, H.V. and Scheffer, J.J.C. (1996). Analytical aspects of phytotherapeutic valerian preparations. Phytochem. Anal. 7: 143-151.

Bos, R., Woerdenbag, H.J., van Putten, F.M.S., Hendriks, H. and Scheffer, J.J.C. (1998). Seasonal variation of the essential oil, valerenic acid and derivatives, and valepotriates in Valeriana officinalis roots and rhizomes, and the selection of plants suitable for phytomedicines. Planta Med. 64: 143-147.

Bounthanh, C., Bergmann, C., Beck, J.P., Haag-Berrurier, M. and Anton, R. (1981). Valepotriates, a new class of cytotoxic and antitumor agents. Planta Med. 41: 21-28.

Chapelle, J.P. and Denoel, A. (1972). Contribution á l’étude des valépotriates dans la racine de valériane officinale. Plant. Méd. Phytothér. 6: 91-105.

Graf, E. and Bornkessel, B. (1978). Analytische und pharmazeutisch-technologische Versuche mit Baldrian. Dtsch. Apoth. Ztg. 118: 503-508.

Hänsel, R. and Schulz, J. (1982). Valerensäuren und Valerenal als Leitstoffe des offizinellen Baldrians. Bestimmung mittels HPLC-Technik. Z. Phytother. 3: 333-340.

Hänsel, R. and Schulz, J. (1985). Beitrag zur Qualitätssicherung von Baldrianextrakten. Pharm. Ind. 47: 531-533.

Meer, J.H. van der, Sluis, W.G. van der and Labadie, R.P. (1977). Onderzoek naar de aanwezigheid van valepotriaten in valeriaanpreparaten. Pharm. Weekbl. 112: 20-27.

Perry, N.B., Burgess, E.J., Lorimer, S.D. and Klink, J.W. van (1996). Fatty acid anilides as internal standards for high performance liquid chromatographic analyses of Valeriana officinalis L. and other medicinal plants. Phytochem. Anal. 7: 263-268.

Schimmer, O. and Röder, A. (1992). Valerensäuren in Fertigarzneimitteln und selbst bereiteten Auszügen aus der Wurzel von Valeriana officinalis L.s.l. DC-Untersuchungen nach DAB 9. Pharm. Ztg. Wiss. 137: 31-36.

Stahn, T. and Bomme, U. (1998). Qualitative Beurteilung eines großen Sortimentes von Valeriana-officinalis-Herkünften. Gartenbauwissenschaft. 63 (3): 110-116.

39

Upton, R. (1999). Commercial sources and handling. In: Valerian Root, Valeriana officinalis, Analytical, Qualtiy Control and Therapeutic Monograph. Upton, R. (ed.). American Herbal Pharmacopoeia, Santa Cruz, CA. pp. 6-7

Wills, R.B.H. and Stuart, D.L. (2000). Effect of handling and storage on alkylamides and cichoric acid in Echinacea purpurea. J. Sci. Food Agric. 80: 1402-6.