Rooibos (Aspalathus linearis) beyond the farm gate: From herbal tea to potential phytopharmaceutical

Aspalathus linearis (Burm.f.) Dahlg. (Fabaceae, Tribe Crotalarieae), an endemic South African fynbos species, is cultivated to produce the well-known herbal tea, rooibos. It is currently sold in more than 37 countries with Germany, the Netherlands, the United Kingdom, Japan and the United States of America representing 86% of the export market in 2010. Its caffeine-free and comparatively low tannin status, combined with its potential health-promoting properties, most notably antioxidant activity, contributes to its popularity. First marketed in 1904 in its fermented (oxidised) form, green rooibos is a new product recently on the market. The utilisation of rooibos has also moved beyond a herbal tea to intermediate value-added products such as extracts for the beverage, food, nutraceutical and cosmetic markets. Its potential as a phytopharmaceutical, shown in recent scientific studies, has not yet been exploited. This review focuses on past and current research aimed at enhancing the value of rooibos herbal tea as a specialised, niche product and expanding its value-adding potential against the background of its traditional use and the current market. The focus falls specifically on aspects such as composition, processing, quality and rooibos as food and potential medicine. © 2011 SAAB. Published by Elsevier B.V. All rights reserved.


Introduction
The genus Aspalathus (Fabaceae, Tribe Crotalarieae) comprises more than 270 species of which most are endemic to the Cape Floristic Region (Dahlgren, 1968). Rooibos tea, produced from A. linearis (Burm.f.) Dahlg., had no commercial value at the beginning of the 20th century, but today it is a wellknown herbal tea, enjoyed in more than 37 countries. A recent report by the Swiss Business Hub South Africa (Anon, 2007) stated that 'Rooibos appears to be headed towards becoming the second most commonly consumed beverage tea ingredient in the world after ordinary tea (Camellia sinensis)'. In South Africa it is well established, enjoying popularity amongst an estimated 10.9 million households (data supplied by SA Rooibos Council (SARC), 2011). Several reviews dealing with rooibos have been published over the years since Cheney deemed it of too insignificant economic importance as a beverage to include it in a review of the biology and economics of the global beverage industry (Cheney, 1947). A different point of view, appreciating its value as a beverage with wide appeal, was later offered (Cheney and Scholtz, 1963). In her review, Morton (1983) campaigned for wider recognition and better distribution of rooibos in the United States of America. It continued to fascinate researchers and recent reviews cover diverse topics such as the development path of rooibos (Rampedi and Olivier, 2008), bioactivity and potential health benefits (Joubert et al., 2008a;Marnewick, 2010;McKay and Blumberg, 2007), and production, processing and quality aspects of rooibos tea and related products (Joubert and Schulz, 2006;Joubert et al., 2008a). The relevance of the major rooibos flavonoids in the diet in terms of their bioactivity, bioavailability and potential herb-drug interactions were addressed by Joubert et al. (2009a). The major focus of the present review will be on past and current research aimed at enhancing the value of rooibos herbal tea as a specialised, niche product and expanding its value-adding potential against the background of its traditional use and the current market. Within the context of enhancing the value of rooibos, aspects such as composition, processing, quality, and rooibos as food and potential medicine (phytopharmaceutical) will be discussed.

Utilisation of Aspalathus linearis
The first person to realise the commercial potential of rooibos as a herbal tea was Benjamin Ginsberg, a merchant of Clanwilliam, who started marketing it in 1904. He obtained the tea from descendants of the Khoi who crudely processed it during the warm summer months. It was, however, only by 1930 that the agricultural value of rooibos was recognised by a medical practitioner and nature lover, P. Le Fras Nortier of Clanwilliam. His early cultivation experiments, carried out with the help of local farmers, O. Bergh and H. Riordan (Anon, 1985), laid the foundation for the industry. Growing participation of other farmers in rooibos production and increased demand over the years expanded the area under cultivation to 36 000 ha (Pretorius, 2007), with production mainly concen-trated in the Clanwilliam area (Fig. 1). Since expansion has threatened biodiversity of the Greater Cederberg Biodiversity Corridor, SARC in partnership with CapeNature, formed the Rooibos Biodiversity Initiative. One of its aims was to undertake a joint planning process for expansion to minimise loss of threatened natural habitat (Pretorius, 2007). Following this, a current initiative of SARC is to establish the Right Rooibos Sustainability Standard (http://sarooibos.org.za).
In the early years, different Aspalathus species and ecotypes, naturally occurring in the Cederberg mountain area, were used to produce rooibos tea (Dahlgren, 1968(Dahlgren, , 1988. Today only the so-called red type or Rocklands type, originally from the Pakhuis Pass area, is of commercial importance. The red type is divided into the selected, improved Nortier type (cultivated; Fig. 2), and the Cederberg type (wild-growing). The latter type has broader and coarser leaves than the Nortier type (Morton, 1983). Characteristics associated with the cultivated type are bright green, needle-like leaves on straight, slender branches with relatively short internodes. The leaves should turn redbrown when bruised (Dahlgren, 1968).
Tea is also sometimes made from small quantities of a closely related species, A. pendula Dahlg. and several wild types of A. linearis (Van Heerden et al., 2003). The grey, black and red-brown types were harvested in the wild, processed and sold to the Rooibos Tea Control Board until 1966, after which the marketing of the grey and black types were discontinued due to poor quality (Anon, 1967). According to Coetzee et al. (1953) the infusion of the black type was not typically redbrown, nor was the flavour characteristic. Recent assessment of wild types demonstrated morphological (Malgas et al., 2010), genetic (Malgas et al., 2010), chemical (Van Heerden et al., 2003) and ecological (Hawkins et al., 2011) variation.
Wild rooibos re-gained prominence as a source of income in recent years. Small-scale producer organisations in the Cederberg (Wupperthal) and Southern Bokkeveld (Heiveld) supply wild harvested rooibos under organic and fair-trade certification to niche markets abroad . Wild rooibos comprises ca 2-5% of the annual production of 40 tons rooibos by the Heiveld Cooperative (Patrickson et al., 2008). Wupperthal's production of wild-harvested rooibos is approximately 10 tons .

Rooibos tea market
The early years of the industry saw a growing demand due to a shortage of Oriental tea (Camellia sinensis) as a result of the Second World War, followed by market collapse in the years after the war. The Clanwilliam Tea Co-operative Company was subsequently established in 1948 to improve marketing conditions for rooibos tea producers. However, a further market decline made rooibos production uneconomical by 1953/54, which led to the establishment of the Rooibos Tea Control Board (later known as the Rooibos Tea Board), at the request of producers, to stabilise the industry. It started functioning on 6 December 1954 with the objectives, amongst others, to regulate the marketing of rooibos and ensure price stabilisation. Total sales were only 524 tons in 1955. When the Rooibos Tea Board was abolished on 1 October 1993 total sales were more than 4000 tons. The entry of new marketing companies and access to new markets resulted in strong growth in sales during subsequent years. With the abolishment of the Rooibos Tea Board, no organisation dealing with matters of mutual concern existed. This changed in 2005 when SARC was established to coordinate activities relating to generic marketing, research and development, and sustainable natural resource management.
From the modest beginning of Benjamin Ginsberg's marketing effort in 1904 rooibos has become a global product, contributing an estimated R120 million in foreign earnings in 2010 (data supplied by SARC, 2011). Table 1 gives a snapshot of sales (local, export and total) for a 23-year period (1970-Fig. 2. Rooibos plantation in the Clanwilliam area. Table 1 Local, export and total sales of rooibos tea (tons) from 1970 to 1992 (data compiled from Anon, 1992).
The domestic market for rooibos showed a steady increase over the years. In 1984 sales were 2600 tons (Table 1) and rooibos had a 12% share of the domestic tea market (Anon, 2001). By 2010 its market share had more than doubled, representing approximately 23% of the tea market with sales reaching more than 5000 tons with an estimated retail sales value of R429 million (data supplied by SARC).
Organic production of rooibos still lags behind production of conventional rooibos. Exports of organic rooibos reached 828 tons in 2010 (Table 3) with Japan the major market at 222 tons followed by Germany, UK, Netherlands and USA (data from Perishable Products Export Control Board (PPECB), supplied by SARC). Total exports of green rooibos reached 253 tons in 2010 (Table 3), of which 70% went to Germany. The only other country that buys substantial quantities is the Netherlands (16% in 2010). The combined exports of green rooibos to Japan, USA and UK comprised 12% of the total volumes for 2010.
A characteristic of the rooibos industry is the cycles of a shortage in supply, accompanied by high prices, followed by expansion of production and subsequent over-supply, accompanied by low to very low prices. In 2004 producers earned on average an all-time record of R16.00/kg, but it was followed by a steady decline to R4.50/kg in 2010, a farm gate value last received in 1999 (data supplied by SARC). This negatively affects the economic viability of the crop. Droughts usually contribute to the shortages. In 1984, following overproduction during that production season and the two preceding seasons, the amount of stocks in reserve was ca 10,000 tons (Joubert, 1994) at a stage when total annual sales were less than 3000 tons (Table 1). By 1994, the increased demand could not be met due to a shortage in production (Anon., 1994), yet in 1996 the situation had again changed drastically, i.e. overproduction and low prices. Similar situations have occurred since then, with the most recent record harvests in 2006 (10,000 tons), 2007 (14,000 tons) and 2008 (18,000 tons), respectively, leading to very low prices (Genis, 2008). The production shortage, predicted for 2011, will be the first since 2003 (Swartz, 2011).

Geographical indication (GI)
Opportunistic trade marking of the term 'rooibos' in 1994 in the USA, which prevented the free use of this generic term in marketing and the subsequent legal battle to reclaim the name for South Africa, has given impetus to the development of a Geographic Indication (GI) for rooibos (Gerz and Bienabe, 2006). It is mainly seen as a defensive measure against possible cultivation outside South Africa and to protect intellectual property rights. A GI is a label that is reserved for products which acquire their characteristic and defining qualities as a result of their geographical location, e.g. Champagne, Tequila or Parma Ham (Grazioli, 2002). In this way producers can distinguish their product based on its specific origin-related characteristics.
The rigorous process of GI certification means that strict quality criteria, linked to its geographic cultivation, must be implemented. Aspalathin, as an unique rooibos phenolic constituent, is also under consideration as a defining parameter. It is believed that these quality criteria will have a positive effect on quality control and marketing (Anon, 2010). Results of an ARC project on the development of the sensory wheel and objective quality criteria (Sections 8.1 and 8.2), undertaken for SARC, will be critical for the certification and required quality control processes.

Value addition
The rooibos industry, characterised by bulk sales as opposed to the sale of retail packaged products to international markets, would greatly benefit from value-addition (Anon, 2010). Based on the number of patents filed (Kaplan, 2009), protection of intellectual property for new applications and/or processes relating to rooibos were considered worth-while. Japan has been especially active in this regard.
The first value-addition was done by A. Theron who brought an extensive cosmetic care range containing rooibos extract to the market (Joubert and Schulz, 2006). Her range of products is distributed to 34 countries (Anon, 2010). Scientific studies in the public domain dealing with the topical/cosmetic application of rooibos extract are, however, very limited. The role of aspalathin, the major rooibos flavonoid and antioxidant, has not yet been clarified. Poor penetration of the skin by aspalathin has been demonstrated by Huang et al. (2008a). An anti-wrinkle effect was demonstrated for a formulation containing rooibos and tea (Camellia sinensis) (Chuarienthong et al., 2010). Since rooibos extract was not tested alone it is not possible to make conclusions about its efficacy as an anti-wrinkle agent. According to Glynn (2010) the topical application of an herbal mixture, containing green rooibos extract, alleviates male pattern baldness. A recent US patent application deals with the protective effect of rooibos extract against hair colour loss (Joppe et al., 2009).
Rooibos extracts, produced locally and abroad, are considered intermediate value-added products in the value chain. Apart from cosmetics, final product applications include functional foods and beverages, as well as nutraceuticals. Rooibos is also finding application in pet food and pet skin care products (Anon, 2010).
One of the earliest applications of rooibos extract and its volatile fraction was as a flavour ingredient of yogurt by the then Van Riebeeck Dairies when the rooibos instant tea powder was first developed (Joubert, 1984). At that stage the use of rooibos as functional ingredient was not yet realised. Neither the instant rooibos tea nor the yogurt reached the market. It was only in 2000 that the manufacture of rooibos powdered extract by a South African company became a reality, when changing consumer attitudes towards natural products made it a viable value-addition option.
The types of rooibos extract and the raw material from which it is produced depends on the final application. A broad distinction can be made between extracts from fermented and green rooibos. Fermented rooibos provides the bulk of the material for production of extracts, used predominantly in beverages and functional foods. Its aroma extract is used to flavour beverages (Anon, 2010).
Some of the food and beverage products currently on the market that contain rooibos extract are yogurt, drinking yogurt, ready-to-drink iced tea, jam and 'instant cappuccino'. Labelling of the rooibos-containing yogurt, carrying the Cancer Association of South Africa Smart Choice emblem, makes extensive use of potential functional properties of rooibos. The drinking yogurt also refers to the 'active ingredient' of rooibos, but no specific 'active ingredient' is mentioned. It must be assumed that the 'active ingredient' is the rooibos solids as no single compound in rooibos is solely responsible for its antioxidant and anti-cancer properties (Joubert et al., 2009a;Snijman et al., 2007).
In many instances, rooibos is only one of several herbal and fruit ingredients in beverages (Anon, 2009a, b). Functional drinks launched in USA and Italy in 2009 contain, amongst others, rooibos and pomegranate. Another drink, launched in the same year in Japan by a major soft drink company contains, apart from rooibos as main ingredient, also green tea, Japanese citrus fruit peel, kumazasa (bamboo), pure tea, fennel, turmeric and hawthorn extract.
Green rooibos, first produced as part of an ARC research project during the 1990s to achieve higher antioxidant levels, is used as a tea and for preparation of extracts for the food, cosmetic and functional beverage markets. The higher levels of flavonoids (antioxidants) in green rooibos, combined with its caffeine-free status contribute to its popularity in tea blends (Cosgrove, 2010) and cosmetic products (Otto et al., 2003;Tiedtke and Marks, 2002). Aspalathin-enriched extracts can be prepared from green rooibos as this compound is present in high levels (Manley et al., 2006;Schulz et al., 2003). The level of enrichment greatly depends on extraction conditions and level of purification. According to a patented process, enrichment is achieved with organic solvents (Grüner-Richter et al., 2008). The tan to greenish coloured extract was initially developed for the cosmetic market (B. Weinreich, Raps & Co, Germany; pers. comm.). A Japanese company also produces an orange coloured aspalathin-enriched extract, specified to contain a minimum of 20% aspalathin. This extract is destined for the supplement market (H. Hata, Tama Biochemical Co., Japan; pers. comm.).
Pharmaceutical product applications of rooibos have not yet been exploited, but according to the report by Kaiser Associates (Anon, 2010) opportunities are likely to emerge as a result of the trend towards phytopharmaceuticals. Rooibos extracts, usually combined with other ingredients, are available in pill form, but these products fall in the category of dietary supplements. Recent research has underscored the potential of aspalathin and Table 2 Major export countries with % of export to each from 2001 to 2010 (data from PPECB, supplied by SARC).  Table 3 Exports of organic and green rooibos (tons) from 2003 to 2010 (data from PPECB, supplied by SARC).
Year Organic Green   2003  648  83  2004  382  60  2005  624  64  2006  386  112  2007  756  74  2008  840  87  2009  791  145  2010  828  253 selected rooibos extracts such as an aspalathin-enriched green rooibos extract as anti-diabetic agents (Joubert et al., 2010a,b;Kawano et al., 2009;Mose Larsen et al., 2008). A patent application for the use of aspalathin in this context was filed in Japan (K. Yagasaki, TUAT, Japan; pers. comm.), while a PCT application was filed for the use of rooibos extract as an antidiabetic agent (Mose Larsen et al., 2008). It is claimed that rooibos extract and a heterodimer containing aspalathin, isolated from rooibos, could be used as a medicament for the treatment of neurological and psychiatric disorders of the central nervous system (Bruno and Dimpfel, 2009). Other opportunities may lie with topical skin products. Two studies concerning inhibition of COX-2 in mouse skin (Na et al., 2004) and inhibition of mouse skin tumour promotion (Marnewick et al., 2005), support the role of the topical application of rooibos extract in preventing skin cancer.

Composition
Rooibos is prized as a caffeine-free herbal tea, although traces of the alkaloid sparteine have been reported (Van Wyk and Verdoorn, 1989). Despite being well-known as a low tannin tea, ca 50% of the hot water soluble solids is tannin-like substances (Joubert et al., 2008a). When compared to black tea (Camellia sinensis), rooibos contains less tannins (Blommaert and Steenkamp, 1978).Very little information is available on the structure of the tannins from rooibos tea, but the dimer, procyanindin B3, the trimer, bis-fisetinidol-(4β,6:4β,8)-catechin, and a pentamer have been identified (Ferreira et al., 1995).
Fermentation of rooibos plant material to produce the traditional herbal tea (oxidised form) causes substantial quantitative changes in its phenolic composition (Table 4, unpublished data). One of the major changes is oxidation of aspalathin via its flavanone analogues to isoorientin and orientin (Koeppen and Roux, 1965;Marais et al., 2000). The oxidation mechanism was recently further elucidated by Krafczyk and Table 4 Contents of major phenolic compounds in green and fermented Aspalathus linearis plant material a (means ± SD; g/100 g dry matter).

Structures
Compounds Green (n = 3) Fermented (n = 3) Dihydrochalcones aspalathin (R 1 = OH; R 2 = C-glucosyl) 2.559 ± 0.699 0.421 ± 0.017 nothofagin (R 1 = C-glucosyl; R 2 = OH) 0.251 ± 0.230 0.040 ± 0.022 Flavones orientin (R 1 = C-glucosyl; R 2 , R 4 = OH; R 3 = H) 0.263 ± 0.087 0.202 ± 0.026 iso-orientin (R 1 = H; R 2 , R 4 = OH; R 3 = C-glucosyl) 0.450 ± 0.163 0.329 ± 0.049 vitexin (R 1 = C-glucosyl; R 2 = OH; R 3 , R 4 = H) 0.042 ± 0.020 0.035 ± 0.009 isovitexin (R 1 , R 4 = H; R 2 = OH; R 3 = C-glucosyl) 0.049 ± 0.029 0.035 ± 0.012 luteolin (R 1 , R 3 = H; R 2 , R 4 = OH) 0.007 ± 0.006 0.010 ± 0.005 Glomb (2008) (Fig. 4). Dihydro-isoorientin ((R)-and (S)eriodictyol-6-C-glucoside), preferentially formed from aspalathin, oxidises to isoorientin. However, dihydro-orientin ((R)and (S)-eriodictyol-8-C-glucoside) does not oxidise directly to orientin as is the case with dihydro-isoorientin and its flavone, isoorientin. Orientin forms irreversibly from isoorientin, which undergoes opening of its vinyl ester structure to form a chalcone intermediate, which is then converted to orientin. Furthermore, an aspalathin dimer, partly responsible for browning as a result of aspalathin oxidation, has also been identified (Krafczyk et al., 2009). The potential health benefits of rooibos tea have been linked to its phenolic content. However, insufficient data are available on the dietary exposure to rooibos phenolic compounds (Joubert et al., 2009a). The calculation of rooibos phenolic intake in populations will allow a study of their association with health and disease. For this purpose, it is essential to have detailed and comprehensive information on the nature and quantities of phenolic compounds in rooibos tea. Increasing interest in dietary phenolic compounds and their effect on human health has led to the development of several databases on phenolic compound contents in foods, such as the USDA Databases on Flavonoids, Phenol-Explorer, and EuroFIR-BASIS (Gry et al., 2007;Neveu et al., 2010;Pérez-Jiménez et al., 2010). None of these databases include composition data for rooibos. Current HPLC methods employed for the analysis of rooibos infusions and extracts are only suitable for the quantification of some of the major flavonoids due to the co-elution of various phenolic constituents. Fingerprinting for authentication/QC and accurate quantification are therefore not possible. Recently, a new HPLC method has been developed for the simultaneous quantification of 16 rooibos phenolic compounds (Beelders et al., 2010a). The method is currently being used by the ARC to analyse a large number of infusions prepared from fermented rooibos production batches obtained from the 2009, 2010 and 2011 harvest seasons, to give a snapshot of phenolic variation in rooibos tea infusions as enjoyed by the consumer. This data would be suitable for inclusion in food composition databases. In addition, an even more comprehensive investigation of rooibos phenolic compounds using 2-dimensional liquid chromatographic techniques is underway (Beelders et al., 2010b).
Very little information on the mineral content of rooibos infusions is available. Touyz and Smit (1982) showed that an infusion of fermented rooibos contains 1.29 μg/mL fluoride. Interestingly, Joubert et al. (2008a) showed a relatively high sodium content of 43.33 μg/mL for the infusion. The toxic mineral, aluminium, is present in much lower quantities in fermented and green rooibos infusions than in Camellia sinensis infusions (Malik et al., 2008). The contents of mineral nutrients are also fairly low (Joubert et al., 2008a;Malik et al., 2008).

Tea
Traditional or conventional rooibos is 'fermented' to develop its characteristic red-brown leaf and infusion colour, and pleasant, slightly sweet flavour. Although modernised, the processing steps employed today are still based on the traditional method of chopping the shoots into small pieces, bruising with wooden hammers to initiate browning, 'sweating' (fermentation) and drying for flavour development (Anon, 1985). Steam pasteurisation has been introduced in the 1980s to ensure good microbial quality after Salmonella contamination was found.
A study carried out on processing showed that quality and product consistency could be improved by carrying out fermentation and drying under controlled conditions (Joubert, 1994(Joubert, , 1998Joubert and De Villiers, 1997;Joubert and Müller, 1997). A factory-based process is, however, not feasible due to the economics of scale and the need for artificial heat generation during drying as opposed to natural heat supplied by sunlight. Observations made by producers and technical personnel of the then Rooibos Tea Board suggest that factors such as the presence of young growth, the age of the bush and the cultivation area could affect the tea quality (Joubert, 1994).
Chemical changes during fermentation are responsible for the development of the red-brown colour (Section 5). Recent papers by Krafczyk and Glomb (2008) and Krafczyk et al. (2009) provide greater understanding of the conversion of aspalathin, previously studied by Koeppen and Roux (1965) and Marais et al. (2000). Aspalathin is not only converted to flavanones as intermediate products, flavones and a dimer, but these constituents are minor products among uncharacterised brown material. Apart from the substantial losses in nothofagin (Joubert, 1996) no other chemical changes during fermentation have been studied yet.
When processing green rooibos, the unfermented product, the greatest challenge is to prevent or to minimise oxidative changes so that the green colour and aspalathin are retained. This could be achieved by inactivation of enzymes, for example, by subjecting the tea to a rapid steaming process. Alternatively, low temperatures and water activity or exclusion of oxygen during drying can be used to slow down the rate of the chemical reactions. A patented vacuum-drying process (De Beer and Joubert, 2002) produced a good quality product with a long shelf-life. Samples stored in a dry, dark place were still green after 2 years (unpublished data). Since vacuum-drying is carried out batch-wise and is expensive, its use did not extend past short-term pilot production of green rooibos. One of the techniques, employed by industry, is to spread the tea in a thin layer in the sun to facilitate rapid drying. Under these conditions, production of good quality green rooibos remains a challenge. If not properly dried, slow browning of the tea will take place during storage and transport to overseas markets, resulting in a product with the appearance of a poorly fermented rooibos, yet without its sought-after characteristic slightly sweet flavour.

Extracts
Waste material in the form of coarse stems and uncut leaves that were sieved out and of no commercial value served as the stimulus of a study carried out in the early 1980s that led to the development of an instant rooibos from the fermented plant material (Joubert, 1984). The idea was also to provide the consumer with rooibos in a more convenient form (Joubert, 1988a) as brewing of a strong cup of tea at that stage was a timeconsuming process. Low recovery of soluble solids from the waste (8.9% versus 20.4% for the sieved fraction b 10 mesh) and overproduction (see Section 3) soon shifted the focus to the tea fraction used for commercial purposes (Joubert, 1984(Joubert, , 1988a. The concept of a soluble rooibos product only found commercial application in 2000 in South Africa with the production of powdered extracts for the beverage, food and dietary supplement markets (Section 5).
The effect of factors such as extraction temperature and time, tea-to-water ratio and extraction process (stirred batch-wise and fixed-bed with flow through) on the recovery of soluble solids and their phenolic composition were investigated (Joubert, 1988a(Joubert, , 1990aJoubert and Hansmann, 1990). Enzyme treatment of the plant material was shown to increase the yield of soluble solids (Pengilly et al., 2008). Normally a muddy brown precipitate forms upon cooling of the extract after concentration. This precipitate was found to impair clarity of the reconstituted extract at 'cup of tea' strength. Its removal improved the clarity, but decreased the phenolic content of the final product (Joubert, 1990c). Enzyme treatment of rooibos before extraction could improve the solubility and clarity of reconstituted rooibos extract (unpublished data). Powdered rooibos extract is not yet 'instant' and agglomeration is required to achieve the characteristics of an instant product, i.e. no lumping when added to hot water and quick dispersal with minimum agitation in the cup (Joubert, 1988b).
Information on the technological processes involved in green tea extract production is not readily available, except when the process is patented. Section 5 provides some discussion on the types of extracts available from green rooibos.

Tea quality aspects
The current South African regulations relating to quality standards for rooibos deal with moisture content, pesticide residues, microbial contamination and the percentage of white stems allowed (Anon, 2002). Taste, aroma and colour are dealt with in vague, meaningless terms, i.e. rooibos should 'have the clean, characteristic taste and aroma and clear, distinctive colour of rooibos', without definitions or reference standards for the terms 'characteristic' and 'distinctive'. Another description used by industry is 'typical'. This means that manufacturers have the freedom to set their own quality standards in terms of colour, flavour and mouth feel of rooibos infusions. Furthermore, the quality classifications 'Super' and 'Choice' only have meaning within the context of a specific retail brand.
In the past the 'caffeine-free' status and 'low tannin content' of rooibos formed the basis of its 'good-for-you' message to consumers. Presently, the antioxidant activity of rooibos, in addition to its pleasant flavour, contributes to its popularity on the international markets. However, rooibos is competing in a market cluttered with products 'rich in antioxidants' or a market characterised by re-positioning of existing products such as the South American herbal tea, yerba mate, with a high caffeine content. A quick survey of scientific literature during the past 5 years will show that the number of papers dealing with plant extracts, whether from fruits or herbs, and their antioxidant activity have increased dramatically. A number of studies included rooibos for comparative purposes and several concluded that their extracts have higher antioxidant activity than rooibos (e.g. Ivanova et al., 2005;Piccinelli et al., 2004;Yoo et al., 2009). In the face of such competition, rooibos tea quality should therefore take on a new meaning (Sections 8.1 and 8.2).

Sensory quality
For descriptors of rooibos sensory properties to have any value, 'characteristic', 'typical' and 'distinctive' must have the same meaning for all role players in the rooibos industry, including the global industry. For this reason a flavour and mouthfeel wheel, which incorporates both positive and negative sensory attributes, totalling 17 descriptors, was developed for rooibos (Anon, 2011b). A large number of samples was analysed to derive the final, most frequently occurring, descriptors. Fine-tuning of the wheel by including data from a second harvest season is in progress. To aid interpretation of the descriptors a preliminary sensory lexicon for some of the descriptors has also been developed (Koch, 2011). A sensory lexicon is a set of words that describe the sensory attributes of a product, along with definitions and/or reference standards for clarification (Drake and Civille, 2002).
The sensory wheel will provide better understanding and appreciation for rooibos sensory characteristics. This will create opportunities for niche products with specific flavours such as estate teas, or teas originating from a specific area as climatic and soil conditions could contribute to different dominant flavour nuances, e.g. caramel or floral.

Other quality parameters
The rooibos industry is increasingly interested in using polyphenol content and antioxidant activity for marketing purposes, as well as for quality control of extracts for the nutraceutical and cosmeceutical markets. Large variation in the antioxidant activity of different production batches of rooibos could, however, erode customer confidence in the product, especially customers who are interested in producing valueadded products such as rooibos extract powders that should meet a predetermined antioxidant activity level. By grading rooibos based on composition and/or antioxidant activity levels, marketers would be able to guarantee minimum levels and product consistency. Producers would be able to blend different batches to reach a predetermined antioxidant level. Against this background, three potential objective quality parameters, i.e. water-soluble solids content, total antioxidant capacity (TAC) and total polyphenol content (TPC) are currently under investigation by the ARC.
Two of the prominent South African extract producers use TPC and TAC as indicators of quality. A major extract manufacturer also produces an extract standardised on orientin and isoorientin content (N 0.5% total), while an extract of fermented rooibos, standardised on aspalathin content, is under development (D. Malherbe, Afriplex, Paarl, South Africa; pers. comm.).
Many researchers have shown that for a specific type of product various antioxidant assays correlate well with TPC (e.g. Prior et al., 2005). Very good correlations between TPC and ABTS •+ scavenging activity were also obtained for unfermented rooibos water extracts (r = 0.99; Joubert et al., 2008b). Interestingly, aspalathin content of unfermented rooibos water extracts was also highly correlated with ABTS •+ scavenging activity (r = 0.96; Joubert et al., 2008b). Care should be taken when interpreting antioxidant values. It is especially important to be careful of directly comparing antioxidant results from different sources as differences in assay protocol between studies can have a large influence on the values obtained. A myriad assays are available for determining the antioxidant activity of foods and beverages. Prior et al. (2005) reviewed the assays most commonly used for analysis of foods and dietary supplements and suggested the ORAC (Oxygen Radical Absorbance Capacity) (Ou et al., 2001), ABTS •+ scavenging (Re et al., 1999) and Folin-Ciocalteau (Singleton and Rossi, 1965) assays based on their advantages and disadvantages. Generally the ORAC assay is favoured by the American nutraceutical industry to 'quantify' the total antioxidant capacity of extracts (Crawford, 2011).

Plant material
The composition of the plants has a direct effect on the quality of processed fermented tea and commercial extracts. Rooibos plants are cultivated from seedlings, causing large variation in phenolic composition between plants. As an example the aspalathin and orientin content of rooibos leaves harvested at the same time from 21 individual plants in the same plantation is shown in Fig. 5A (data adapted from Joubert et al., 2008c) and 5B (unpublished data), respectively. The aspalathin and orientin content ranged from 6.0 to 11.2 g/100 g dry matter and 0.5 to 1.1 g/100 g dry matter, respectively. Harvest date is another factor that could influence phenolic composition of rooibos plants. The aspalathin content of plant material harvested from the same plants ranged from 2.33 to 3.76 g/100 g dry matter over a one year period (unpublished data).
The phenolic composition of wild rooibos showed qualitative and quantitative variation between populations and types (Van Heerden et al., 2003). Some populations even lacked aspalathin, the major flavonoid in the red type or Rocklands type of A. linearis used for commercial production. One of the wild types, lacking aspalathin, contained rutin as major flavonoid (Joubert and Schulz, 2006). If the presence of aspalathin in the plant material (green and fermented) becomes part of the quality parameters of a rooibos GI (Section 4) then some wild rooibos types may be excluded from the 'rooibos' label in future.

Commercial extracts
Recently, changes in quality of extracts during the manufacturing process were investigated in terms of aspalathin, orientin, isoorientin and total polyphenol contents (Joubert et al., 2009b;Viljoen, 2008). No major losses of aspalathin, orientin or isoorientin were observed during microfiltration, reverse osmosis and concentration of fermented rooibos extract. The total polyphenol content, on the other hand, seemed to increase during these processes (Viljoen, 2008) indicating that compounds other than polyphenols are removed during the production process. Spray-drying of the concentrate did not affect its phenolic content.

Ready-to-drink iced teas
Ready-to-drink rooibos iced tea is a popular variant in South Africa; however, its rooibos content is not regulated. A recent study (Joubert et al., 2009b) found that some brands contained no aspalathin or its oxidation products, orientin and isoorientin, suggesting that no rooibos extract was used in their manufacture. The maximum aspalathin content found in 41 samples, representing eight brands, was 0.69 mg/100 mL. The average aspalathin content of the samples was 0.28 mg/100 mL, which was much lower than that expected in rooibos extract reconstituted to 'cup of tea' strength (0.3-0.9 mg/100 mL; n = 20; unpublished data).
In an attempt to understand the low levels of rooibos flavonoids in commercial iced teas, a study was conducted to assess their stability in fermented (Joubert et al., 2009b) and green (Joubert et al., 2010c) rooibos iced tea formulations  during heat treatment. Pasteurisation or sterilisation is the most likely processing step during the manufacture of rooibos iced teas that could decrease their flavonoid content. Pasteurisation caused no loss of rooibos flavonoids in fermented rooibos iced tea, but sterilisation markedly decreased the aspalathin and isoorientin contents (Joubert et al., 2009b). The presence of citric and ascorbic acids partially protected against the loss of aspalathin and isoorientin. Orientin showed different trends due to its conversion from isoorientin. The aspalathin content of iced tea could be greatly increased by using an aspalathinenriched extract produced from green rooibos (Joubert et al., 2010c). Trends for heat stability of the flavonoids in this type of rooibos iced tea were similar to that of fermented rooibos iced tea. Combining the aspalathin-enriched green rooibos extract with ascorbic acid in a nano-micelle carrier system increased the heat stability of the resulting iced tea formulation with only a minor reduction in aspalathin loss after addition of citric acid.
Results from these two studies show that rooibos iced tea could make a valuable contribution to aspalathin intake if care is taken during formulation and processing. The bioavailability of aspalathin when present in a ready-to-drink rooibos iced tea was recently demonstrated (Section 9). Storage of iced tea can also contribute to low levels of aspalathin. Its stability during storage was shown to depend on product formulation and type of extract (Viljoen, 2008). pH was shown to be a crucial factor in its stability (De Beer et al., 2011). Milk-based products with a relatively high pH would not retain aspalathin.

Food or medicine conundrum
The discovery by A. Theron in 1968 that rooibos helped to soothe her allergic baby's colic started off rooibos tea's reputation as a 'healthy drink' in modern times (Joubert and Schulz, 2006). Her anecdotal findings, summarised in a letter to the Rooibos Tea Control Board on 18 May 1968, was met with scepticism from scientists who were approach for comment. More anecdotal evidence became evident after newspaper reports and radio interviews and today the therapeutic value of rooibos for colic babies is well-known.
In the 1985 Annual Report of the Rooibos Tea Board (Anon, 1985) it was noted that: 'Rooibos Tea is used overseas mainly for its medicinal value and has thus become popular as a medicine rather than a tea. This is to a certain extent the main reason why overseas marketing potential of Rooibos Tea is rather limited.' Changing consumer attitudes and the focus on 'anti-ageing' soon afterwards placed rooibos in a position to exploit its health-promoting properties, and particularly its antioxidant activity in marketing (Wilson, 2005).
Japanese researchers were the first to study the antioxidant and 'anti-ageing' properties of rooibos. A cartoon was published in March 1992 in the Landbouweekblad, a weekly farmers' magazine, after much had been made locally of the 'anti-ageing' effect of rooibos on the skin (Fig. 6). In the cartoon Rooibos Tea Board stalwarts, J. van Putten, Deputy General Manager (Export Promotion), dressed as a nanny and pushing P. Saayman, the General Manager, in a baby stroller, calls to the flabbergasted secretary to phone Japan to inform them that the 'anti-ageing' effect had worked. The logo of the Rooibos Tea Board, symbolising a cup of rooibos tea with steam rising, forms the baby stroller. Since then many studies on the potential health-promoting properties of rooibos have been carried out in South Africa and Europe (Joubert et al., 2008a). The latest studies are summarised in Table 5.
Human studies dealing with the health promoting aspects of rooibos and the bioavailability of rooibos flavonoids, especially aspalathin, have been limited to date and are summarised in Joubert et al. (2008a) and Tables 5,6 and 7. For a compound to exert an effect in vivo, it should reach the target tissue(s). In the first instance the presence of the compound in the plasma or urine is essential. Since the study of the bioavailability of aspalathin in the pig as model (Kreuz et al., 2008), new insights into the bioavailability of aspalathin and other rooibos flavonoids have been gained by the studies of Courts and Williamson (2009), Stalmach et al. (2009) and Breiter et al. (2011. Aspalathin is extensively metabolised and phase II metabolites have been found in the urine ( Table 6). Most of the metabolites were excreted in the urine of volunteers within 5 h of consuming 500 mL of rooibos ready-to-drink beverages (Stalmach et al., 2009). Kreuz et al. (2008), feeding pigs a high concentration of aspalathin-enriched green rooibos extract, found only trace quantities of the unmetabolised compound in the plasma. Its presence was recently confirmed in human plasma (Breiter et al., 2011). All of the studies showed poor Table 6 Summary of studies on in vitro and in vivo absorption, metabolism and transport of rooibos flavonoids.

Characteristic
Test bioavailability for aspalathin, yet despite this, it's in vivo bioactivity, e.g. glucose-lowering effect, confirm the importance and relevance of this rooibos flavonoid (Table 7). The finding that the plasma antioxidant status of the volunteers drinking the ready-to-drink rooibos beverage peaked after 1 h (Villaño et al., 2010) may explain why Sauter (2004) showed no improvement in the plasma antioxidant status of volunteers dosed with aspalathin-enriched extract. In the latter study, the volunteers received a twice daily dose of 250 mg of aspalathin-enriched green rooibos extract for 2 weeks. In this case the blood was sampled the day after the last dose. Conjugation of aspalathin would also impair its ability to act as antioxidant in the plasma (Van der Merwe et al., 2010).
Marketing of rooibos as an 'antioxidant' beverage may need reviewing because of the over-supply of natural antioxidant products on the market. The market is constantly searching for new products that have more to offer than just high antioxidant activity and modulation of oxidative stress. Markets are moving towards condition-specific antioxidants, especially those aimed at inflammation (Crawford, 2011).
Two incorrect nutritional 'facts' that need highlighting and that some companies use in their promotional literature and on their web pages, are the erroneous claims that rooibos has a high ascorbic acid and mineral content. Hesseling et al. (1979) incorrectly interpret the apparently high ascorbic acid content of the infusion obtained using a method that is not specific for Table 7 Summary of in vitro and in vivo studies on the bioactivity of the major rooibos flavonoids-aspalathin, orientin and isoorientin. Isoorientin Anti-mutagenicity Salmonella typhimurium assay; tester strain/mutagen combinations TA98/2-AAF and TA100/AFB 1 ; with and without metabolic activation Typical dose-response anti-mutagenic effect against AFB 1 ; more effective against 2-AAF than aspalathin; no mutagenic or comutagenic effect Snijman et al., 2007 Hypoglycaemic and antihyperlipidaemic activity Subacute oral administration to STZ-induced diabetic rats Reduced fasting blood glucose, cholesterol and triglyceride levels Sezik et al., 2005 Anti-inflammatory activity ascorbic acid (Joubert et al., 2008a). Unfortunately, Morton (1983) repeated this high value in her review paper on rooibos, an incorrect statement which even now is still recurring in some promotional material. Furthermore, the same paper by Morton (1983) reported highly inflated values for the iron, calcium, magnesium, phosphate and potassium content of rooibos. Statements about the 'isotonic' value of rooibos have also been made. This has not been challenged in the scientific literature until recently. Joubert et al. (2008a) reported that a 5min infusion of rooibos at 'cup-of-tea' strength contains only traces of iron, with the iron content only 15% of that reported by Morton (1983). The content of other minerals was also substantially lower than previously reported. Malik et al. (2008), after determining the mineral content of a 15-min infusion reported similar values to Joubert et al. (2008a). They came to the conclusion that rooibos is not a valuable source of nutrient minerals as claimed in promotional material of some marketing companies. With the establishment of SARC, dissemination of new findings relating to the potential healthbenefits of rooibos is handled in a responsible manner. Yet, in spite of these efforts by SARC, inflated claims about this herbal tea still abound in promotional material of companies.

Future challenges and opportunities
The future challenges and opportunities of the rooibos industry are succinctly summarised by the following conclusion reached by the recent Rooibos and Honeybush Market Development Programme Framework report (Anon, 2010) after inputs of various industry stakeholders: 'An integrated global marketing drive based on the scientific proof of health claims, together with innovative, value-added product applications and a sound pricing system, could be the beginning of an extraordinary bright future for the Rooibos industry'.