HerbalGram: Integrating Recent Knowledge about the Genus Echinacea: Morphology, Molecular Systematics, Phytochemistry

This article summarizes the authors’ recent research on Echinacea published in various refereed journals with an emphasis on a new taxonomy. The taxonomy that most people are familiar with is that of McGregor, established in 1968. In these new studies, the authors recognize 4 species in Echinacea and have fitted most of McGregor’s species as varieties under E. pallida and E. atrorubens, whereas E. purpurea and E. laevigata remain as before without varieties. The authors’ studies on genomics and phytochemistry have lent support to this taxonomic scheme. This article contains an identification key to the 4 species and to the varieties within E. atrorubens and E. pallida.

Up until the 1960s, the taxonomy of the genus Echinacea was based on specimens that were collected from parts, but not all, of its natural geographical range. Further, before the chemistry of the 1980s and the molecular biology of the 1990s, Echinacea’s taxonomic groupings were based on morphology first and subsequently on cytological analyses. For instance, Cronquist described 4 species (and one variety) from his morphological observations of herbarium specimens (including the actual type specimens associated with scientific names).¹

In 1968, R.L. McGregor embarked on a 15-year odyssey studying wild Echinacea plants from populations throughout the entire geographical range. His biosystematic studies included the investigation of macro- and micro-morphological traits under a common garden design, and he included some cytological comparisons and some anatomical traits, while making inferences about phylogenetic history, relating to evolutionary development in the genus. McGregor recognized 9 species and 4 varieties.² He proposed that there may be extensive genetic variation within certain wild populations of a single species or variety and that further genetic studies were indicated.

Many have relied on McGregor’s identification keys to the wild species and varieties.² For example, during the herbal medicine boom of the early 90s, botanists, conservationists, and diggers (wildcrafters) used them. Many reportedly found that there was such a large amount of variation between plants of a single population in the genus Echinacea, and even between plants of the same age cultivated in a greenhouse, that they were unable to identify the plants confidently. Furthermore, the market demand at that time increased the value of wild roots from Echinacea angustifolia (and later from E. pallida roots) as well as the aerial parts of E. purpurea. There was a dire need for rigorous and accurate morphological identification so dealers could provide certified authentic Echinacea to their customers. One solution to the taxonomic problem was the work of Bauer and Wagner; they provided chemical profiles of some secondary metabolites, which were used to distinguish between Echinacea and non-Echinacea (Parthenium integrifolium) dried samples, and they offered some possible means to distinguish between the different species and varieties as well.³ Bauer and Wagner determined that E. pallida (Nutt.) Nutt. var. pallida, Asteraceae [syn. = E. pallida (Nutt.) Nutt.] was being cultivated and sold erroneously as E. pallida (Nutt.) Nutt. var. angustifolia (DC.) Cronq. [syn. = E. angustifolia DC. var. angustifolia]. Building on Bauer and Wagner’s discovery of potential chemotaxonomic traits in the commercial species, we undertook a large-scale taxonomic molecular and phytochemical revision. Our goal was to ensure more accurate botanical identification of all the different Echinacea taxa for reasons of safety in the supply chain for phytomedicines and for reasons of wild rare species conservation.

Conservation of the natural Echinacea resources across North America has socio-political implications due to issues of private and public land tenure, especially on Aboriginal (i.e., Native American) land reserves. Governance of lands and natural resources tends to vary at the federal, state/provincial, or regional levels in both Canada and the United States. The majority (>90%) of natural Echinacea populations occur in the United States (see Figure 1 below), where there is a National Germplasm Conservation Program that addresses all Echinacea taxa among other resources and threatened species in collaboration with the Nature Conservancy, the State Departments of Natural Heritage/Conservation, and the US Department of Agriculture (USDA).

In an attempt to rectify the situation of poor botanical identification methods with Echinacea on the market, we studied 110 wild populations (see Figure 1 below) to gauge the extent of variation between plants in a population and between populations in a species. Our objective was to investigate taxonomic groupings based on the degree of morphological similarity between plants, and to test the statistical significance of our resulting groupings using morphometric tools.

Natural populations were taxonomically identified in the field according to McGregor,² and transplanted to a greenhouse for morphometric data collection.^4,5,6 We measured 74 traits for over 300 specimens, which allowed us to calculate the statistical index of similarity between (1) individual plants (assuming no prior taxonomic groups), and (2) McGregor’s taxonomic groups. We used a Gower coefficient of similarity (a biostatistical measuring tool),⁷ followed by several clustering methods, and canonical discriminant analyses to assess the groupings.⁴

The morphometric analyses supported 2 acceptable cluster solutions. The first strongly supported 2 major taxa within Echinacea, which we determined to be at subgenus level. The species known currently as E. purpurea (L.) Moench was the sole taxon in Echinacea subgenus Echinacea which contains only E. purpurea, whereas all other infrageneric taxa were in Echinacea subgenus Pallida. The second most acceptable cluster solution supported 4 taxa, which we determined to be at the species level: (1) E. purpurea [= Echinacea purpurea (L.) Moench nom. cons. prop.],⁸ (2) E. laevigata [= E. laevigata (Boynton & Beadle) Blake], (3) E. atrorubens, and (4) E. pallida. Therefore, we effectively re-classified the genus Echinacea into 2 subgenera, one with a single species in it, and the other having 3 species. Our results also supported an 8 cluster solution using McGregor’s identification keys.² The 8 groups correspond to varieties within 2 species (see Table 1). The revised taxonomy recognizes all of McGregor’s taxa, except for one variety, E. angustifolia DC. var. strigosa McGregor, which was not distinct from E. pallida var. angustifolia. This putative variety may be a morphotype that resulted from introgression [movement of alleles from one taxon to another through hybrid intermediates, usually found in populations bordering and/or overlapping each other], and it shows the same phenotype [the visible, measurable characteristics, which may vary independent of genetic makeup] in similar ecological zones.⁸

A dichotomous identification key is used by biologists to identify organisms to different levels, such as family, genus, species, or variety. It is designed to list traits of organisms as a series of paired choices that lead progressively to identification of the organism. Not all keys lead to the same level of taxonomic identification, and it is important to have all traits match the key statement that is chosen for any given specimen in order to arrive at the most accurate identification of that organism. In the key below, one may proceed to take a plant or specimen in question and choose between the pair of statements numbered with “1,” which then leads either to a choice between a pair of “2” statements (and eventually identification of plants in the subgenus Pallida), or to identify the plant in question as subg. Echinacea, E. purpurea (L.) Moench. If one continues to follow the number at the end of the statement that is true, one will eventually arrive at the best identification of that particular specimen. Note that if some traits in the key are not observable in the specimen in question, then another specimen with those missing organs must be used for the key to function properly.

1. Basal leaf up to 5 cm wide; cauline leaf 0.5 to 4.5 cm wide; taproot (may be branching or fusiform); leaf blade trichomes multicellular with knobby joints; major veins almost parallel from a common origin at the base; 1-3 series of involucral bracts 2. subg. Pallida

1. Basal leaf greater than 5 cm wide; cauline leaf 4.5 to 9 cm wide; fibrous roots (from a caudex); leaf blade trichomes bicellular with ledge-like joints; major veins branched; four series of involucral bracts subg. Echinacea, E. purpurea (L.)

2. Basal leaf greater than 3 cm wide; basal leaf margin serrate, or dentate; adaxial leaf blade stalked trichomes absent; stem stalked trichomes absent; cauline leaf margin serrate
E. laevigata (C.L. Boynton & Beadle) Blake

2. Basal leaf up to 3 cm wide; basal leaf margin entire; adaxial leaf blade stalked trichomes present; stem stalked trichomes present; cauline leaf margin entire 3

3. Stem stalked trichomes appressed (strigose); leaf blade stalked trichomes sparse; leaf marginal trichomes different than blade trichomes (more appressed). 4

4. Ray floret yellow E. atrorubens Nutt.
var. paradoxa (J. B. Norton) Cronq.

5. Disk corolla petal fusion more than 3/4 total corolla length; involucral bract up to 0.2 cm wide; stem branched….E. atrorubens Nutt. var. atrorubens Cronq.

5. Disk corolla petal fusion less than 3/4 total corolla length; involucral bract greater than 0.2 cm wide; stem unbranched E. atrorubens Nutt. var. neglecta (McGregor) Binns B. R. Baum & Arnason

3. Stem stalked trichomes hirsute, or straight pubescent; leaf blade stalked trichomes dense; leaf marginal trichomes identical to leaf trichomes in type and habit
6

7. Capitulum up to 2.5 cm wide; involucral bract up to 0.2 cm wide E. pallida (Nutt.) var. tennesseensis (Beadle) Binns B. R. Baum & Arnason

7. Capitulum greater than 2.5 cm wide; involucral bract greater than 0.2 cm wide E. pallida (Nutt.) var. angustifolia (DC.) Cronq.

9. Ray achene trichomes present; stem unbranched. E. pallida (Nutt.) var. simulata (McGregor) Binns B. R. Baum & Arnason

9. Ray achene trichomes absent; stem branched. E. pallida (Nutt.) var. sanguinea Gandhi & Thomas

Related information for identification of species and varieties (including an interactive key, and alternative key with McGregor’s taxonomy) may be found on the Web site of Agriculture Canada (http://res2.agr.gc.ca/ecorc/echinacea/key-cle_e.htm). This will eventually be modified online to allow for identification from separate plant parts. To identify a whole plant, one can choose one of the alternative descriptions at number 1 and then follow the leads.

The greatest amount of morphological and genetic diversity observed among geographically-close populations was found in a narrow region of the Great Plains, which is considered by field botanists to be the center of Echinacea diversity.^4,9 The “center of diversity” spans several eco-regions that share characteristics of having overlapping biogeoclimatic “edges,” such as tallgrass prairie abutting limestone upland formations and/or shortgrass prairies. The following areas are included in the hypothetical region: Ozark Mountains of Missouri and Arkansas, prairies of Kansas and Oklahoma, and especially Black Hills of southeastern Oklahoma where suspected hybridization and introgression may be directing the most active speciation within the genus.^4,9

In our work, the evolutionary relationships between the 4 revised species were estimated using a cladistic analysis [based on shared, derived characters which are often also diagnostic] of 36 characters (including some phytochemical ones). See the cladogram [an evolutionary tree] (see Figure 2 right), where Echinacea is distinguished phylogenetically from the outgroup [a sister group] Rudbeckia (98% bootstrap value) [a method for assessing the statistical significance of the relationships between taxonomic groups, i.e., positions of branches in an evolutionary tree]. Within the Echinacea clade, E. atrorubens and E. pallida share 3 unique, derived characteristics, and E. purpurea was most basally divergent. Although historically E. laevigata was confused with E. purpurea,¹⁰ current morphometric results show it to be closely related to E. pallida and E. atrorubens.

In summary, we proposed a hierarchy of 2 subgenera, 4 species and 6 varieties in the genus Echinacea.⁴ The 2 subgenera are novel, but our results confirm the 4 species groups that were first suggested using classical taxonomic methodology.¹ All of our described varieties were previously either species or varieties according to McGregor.² Table 2 on page 37 compares the classifications of both McGregor² and Cronquist^1,11,12to the revised taxonomy.⁴

Molecular Systematics Based on DNA Methods of Purple Coneflowers: Genus Echinacea

Echinacea is a genus classified in the Heliantheae tribe within the family Asteraceae. Together with other genera in this tribe, Echinacea plants are popularly known to be among the “Coneflowers.” The relationship of the genus Echinacea to others has been studied using techniques which aim to determine the degree of relationship and the probable evolutionary development of these plants over time. For example, an article published in 1995 by Urbatsch and Jansen reported restriction site analysis of the chloroplast genome, which placed the genus in the subtribe Ecliptinae.¹³ This subtribe is distinct from, yet closely related to, Rudbeckiinae, which contains the genera Dracopsis, Ratibida, and Rudbeckia.^14,15 Subsequently, Urbatsch et al used another approach, nuclear rDNA internal transcribed spacer (ITS) sequences, to study evolutionary relationships among the Coneflowers and relatives and also to combine the data with their previous chloroplast DNA restriction site data.¹⁶ They concluded that Echinacea ought to be classified within the tribe Zinniinae, and that it is definitely not related to genera in the Rudbeckiinae.

In the cladogram of the combined data they used 6 species (sensu McGregor 1968) of Echinacea with similar results of relationships among species as in the chloroplast DNA restriction site data, i.e., that E. purpurea is closely related to E. paradoxa.

Using Amplified restriction Fragment Length Polymorphism (AFLP^®, see side bar on page 36), Mechanda et al¹⁷ undertook a study in parallel to Binns et al⁴ to seek independent support for the morphologically based classification (including relationships) and to complement it in 2 respects: (1) to estimate the genetic diversity of the species and varieties, and (2) to provide means of identification of single plants by DNA fingerprinting.

Four hundred thirty-five individual plants were sampled from 58 natural populations representing both the area of distribution and the species and varieties of both Binns et al⁴ and McGregor² classifications. The most notable outcome resulting from the AFLP investigation was that each individual could be uniquely distinguished by a combination of presence/absence of a set of 124 fingerprints. The main finding of this study was support for the 4 species classification of Binns et al,⁴ but not for all the varieties, most of which were previously recognized by McGregor as species.² The species are recognized by a combination of DNA fingerprints, not by a single or a few single and unique AFLP bands. Thus, to identify an individual plant to species with AFLP one needs to resort to more elaborate means, as indicated in Mechanda et al with an example (refer to Table 11 in their article).¹⁷ In the example, 10 DNA bands from a single primer set are apparently sufficient to identify an unknown plant, or plant fragment, to one of the 4 recognized species.

As far as relationships among the 4 species are concerned, although no attempt was made to use any outgroup (a reference outside Echinacea but close enough to it) in the AFLP study, E. laevigata and E. purpurea can be construed as forming a sister group based on the unrooted UPGMA dendrogram (refer to Figure 4 in Mechanda et al 2004,¹⁷ which is not a true phylogenetic tree). This can easily be seen when moving the branches of the dendrogram without changing the topology. Based on this the genus Echinacea consists of 2 parallel pairs: E. purpurea-E. laevigata and E. atrorubens-E. pallida. This finding supports the gross morphological similarity seen between at least the first two, since they have sometimes been confused.^8,10

The gene diversity of all the species together in the genus and similarly for the varieties together (measured on a scale from 0 to 1) was nearly 0.5 for both. The species with highest gene diversity was E. purpurea, also near 0.5 whereas the 3 other species had lower rates at 0.3. Both E. purpurea and E. laevigata do not contain varieties. Although the varieties were not supported by the AFLP results, when analyzed for genetic diversity, those of E. pallida had greater values than those of E. atrorubens, with the exception of E. pallida var. tennesseensis having the lowest genetic diversity (near 0.2), which is understandable due to its rarity with limited individuals in the populations (E. pallida var. tennesseensis is currently listed as a federally endangered species by the US Fish and Wildlife Service¹⁸). The genetic variation was apportioned as follows: 19% among species, 40% among populations within species, and 41% within populations. In other words, the genetic variation among populations was found to be about equal to the genetic variation within populations. But obviously the kind of variation was different since every individual was found to possess unique fingerprints.

Once you know it’s Echinacea, how do you determine what kind of Echinacea it is?

There were major difficulties in identification of plant materials that were reported by wildcrafters, growers, and scientists during the early days of Echinacea’s boom in the commercial marketplace (see Morphological Systematics section on page 33). For this reason, 2 collaborating research teams used modern tools in both morphometric taxonomy⁴ and molecular systematics¹⁷ to discover which natural taxonomic groups exist currently in wild plant populations. Mechanda et al distinguished wild species and varieties in Echinacea using Amplified restriction Fragment Length Polymorphism (AFLP^®).¹⁷ This generated results about relationships between plants and populations based on DNA. Also, it allowed for independent support for the morphological classification by complementing it in 2 respects: (1) it estimates the genetic diversity of all types of Echinacea species and varieties growing in the wild, and (2) it provides a means for stakeholders to identify and trace single plants by DNA fingerprinting.

Based on work by Mechanda et al, the genus Echinacea consists of 2 parallel pairs: E. purpurea-E. laevigata and E. atrorubens-E. pallida. E. laevigata and E. purpurea can be construed as forming a sister group based on the unrooted UPGMA dendrogram (see Figure 4 in Mechanda et al 2004, which is not a true phylogenetic tree).¹⁷ In this approach, no attempt was made to use any outgroup comparison, and the tree is not a true phylogenetic tree (as seen by the unchanging topology when branches in the tree are rotated). This finding supports the gross morphological similarity seen between E. purpurea and E. laevigata.^8,10

Both E. purpurea and E. laevigata do not contain varieties. The other 2 species, E. pallida and E. atrorubens, each contain varieties by morphometric classification,⁴ but classification by AFLP results did not resolve distinct groups at the variety level. In fact, genetic diversity was measured on a scale of 0 to 1 and found to be 0.5 among species, and also 0.5 for all varieties together. The species with highest gene diversity was E. purpurea, near 0.5, and the 3 other species had lower rates at 0.3. Varieties of E. pallida had greater diversity ratings than those of E. atrorubens, with the exception of the rare E. pallida var. tennesseensis having the lowest genetic diversity (near 0.2).

Although hybrids and hybrid populations were reported by McGregor,² we were unable to distinguish them from others by genetic diversity measured with AFLP analysis, although more than one suspected hybrid population in the field was identified in the morphometric work by Binns et al.⁴

Urbatsch and Jansen only studied 7 of the 9 Echinacea species recognized by McGregor² and found that E. purpurea was closely related evolutionarily to E. atrorubens and E. paradoxa, and that E. simulata was possibly the more ancestral species.¹³ Later, Urbatsch et al reported combined data analysis using 6 species of Echinacea (sensu McGregor 1968).¹⁶ Their cladogram shows similar relationships among species to those in the Urbatsch and Jansen paper on chloroplast DNA restriction site data,¹³ i.e., E. purpurea is closely related to E. paradoxa.

As part of our AFLP study we found that the AFLP fingerprints were inappropriate for phylogenetic studies.¹⁷ However, Kim et al carried out a similar AFLP study to ours, with much less sampling to assess phenetic/phylogenetic relationships among Echinacea species and varieties (sensu McGregor).¹⁹ Their results, not surprisingly, did not provide support for the presently accepted classification by Binns et al.⁴ One reason for this is that AFLP data are usually inappropriate for phylogenetic studies demonstrated on theoretical grounds, as explained by Clark and Lanigan²⁰ regarding RAPD data. Clark and Lanigan’s explanation equally applies to AFLP data in many respects, including ours (refer to the Discussion section on “Phylogenetic analysis,” pages 480-481 in Mechanda et al¹⁷). Another reason is that the identification of their material may be questionable, especially if they relied on McGregor’s keys, which have been problematic in the past.⁴ A different investigation was carried out by Kapteyn et al using DNA-RAPD.²¹ RAPD has proven to be less amenable to reproducibility than AFLP. Kapteyn et al used only the 3 main commercial species (E. angustifolia, E. pallida, and E. purpurea) and their study remained inconclusive.²¹

How can DNA markers be used to authenticate sample materials of Echinacea species and commercial lines (cultivars?) within species?

That authentication of commercial material of Echinacea is of prime interest to the consumer goes without saying. Authentication is needed to ensure the correct and proper content of the product. Correct identification of the plant material constitutes one aspect, and correct phytochemical characterization of plant extracts is another aspect (i.e., the quantitative analysis of marker phytochemicals for assurance of safety, quality, and potentially of therapeutic value). Both are needed in the natural health products industry. The study by Mechanda et al has shown the potential of correct identification to species.¹⁷ In another study, the use of DNA-based markers to predict phytochemical profiles in extracts of identical material were assessed by Baum et al, using AFLP and High Pressure Liquid Chromatography (HPLC).²² In this study, we determined both AFLP DNA fingerprints as well as the quantitative profiles of 2 marker compounds, namely, cichoric acid (2,3-O-dicaffeoyltartaric acid) and dodeca-2E, 4E, 8Z, 10E/Z-tetraenoic acid isobutyl amide in over 50 accessions of E. purpurea. This small study has shown the potential of DNA markers to predict the amount of industry marker chemicals. An extension of this study or a similar DNA-based study may be needed to distinguish the true product from adulterants (wrong plant species, wrong plant line, wrong plant part) and contaminants (like bacterial or fungal, foreign matter).

The phytochemistry of Echinacea species is of key importance to the herbal industry because the phytochemical markers are some of the most easily identifiable characters for species identification in processed products. Also, they are biologically active substances that have importance in the pharmacology of the products. The basic phytochemistry of Echinacea was undertaken in Germany at a time when there was little interest in North America in this medicinal endemic prairie genus.^3,23 The main groups of importance are caffeic acid derivatives (CADs) (see Figure 3 below), lipophilic alkamides (AAs) and ketoalken/ynes (see Figure 4 on page 39), although other types of secondary metabolites are also found in the genus.

It is well known in chemosytematics that individual species are likely to have their own unique blend of phytochemicals, developed in the co-evolutionarily-driven progression towards developing novel defense chemicals. Bauer showed that this principle can be used to distinguish the 3 commercial Echinacea taxa by HPLC analysis if other species/varieties are not considered as possible components of a mix.^3,23 Moreover, in Echinacea, there is a high degree of phytochemical redundancy in individual species, i.e., they contain 5 or more CADs and 10 or more AAs. As a result of phytochemical research, unique marker phytochemicals have been used qualitatively to certify botanical identification in the industry. Commercial species/marker relationships used in industry are as follows: echinacoside as a marker for E. pallida var angustifolia versus E. purpurea, and the use of ketoalkene/ynes as markers for E. pallida var pallida.

The recent revision of the genus Echinacea and the detailed study of phytochemistry of all the species and varieties from wild and cultivated sources reveal a more complex picture.⁴ This is important to the herbal industry for the purpose of assessing contamination of commercial seed lots with wild species or the use of wildcrafted species of doubtful origin in the product. The phytochemical variation also supports the morphometric classification of different taxa within the genus by Binns et al,⁵ but the identification of all the species and varieties using phytochemistry is confounded by polyploidy [having more than 2 sets of chromosomes which are homologous (same genes, not necessarily the same gene products/functions)] and cannot be achieved solely on the basis of presence or absence of one or a few compounds. The major findings follow.

The current industry practice, which uses echinacoside as a positive marker for E. pallida var. angustifolia versus E. purpurea where it is absent, is an over simplification if all species and varieties are considered. In fact, echinacoside is present in quantifiable amounts from roots of all Echinacea taxa except E. purpurea; namely, 3 species and 7 varieties.^4,5 Trace amounts of echinacoside were found in E. purpurea, so lack of this compound is not the best marker. The absence of alkamide 18 was found to be a more definitive E. purpurea marker.

Ketoalkenes/ynes cannot be taken as definitive markers for E. pallida var. pallida, as often considered in industry practice. However, they do appear to be markers for polyploids. They are found not only in E. pallida var. pallida but also in E. pallida var. simulata which is sometimes triploid, as well as possibly hybridizing populations of E. atrorubens var. neglecta and E. atrorubens var. paradoxa.

On a positive note, the species and varieties are readily distinguishable on the basis of quantitative HPLC (high-performance liquid chromatography) profiles of the compounds(see Figure 5 on pages 40 and 41 and Figure 6 on page 42 and 43), but not generally on the presence or absence of individual compounds; notably, the roots of E. angustifolia and E. pallida can be differentiated by the occurrence of the CADs, 1,3-0-(cynarin) and 1,5-0-dicaffeoylquinic acids present only in the former.²⁴ Canonical discriminant analysis revealed that cichoric acid, the diene AAs 1-3 and 7, and ketoalkene 24 were the best taxonomic markers. HPLC profiles for the lipophilic compounds contain more information because they contain a larger number of compounds.

Plant age (and plant part) generally changes the expression of compounds. Young plants expressed lower amounts of alkamides in roots and flowers than present in older plants. Levels of compounds in young roots can also be increased significantly by induction with methyl jasmonate, which suggests they are also inducible by mechanical, insect, or fungal damage.²⁵

In other studies we showed that the quantitative presence of phytochemicals, such as 8,9 and cichoric acid in mature plants, can be correlated to DNA markers (AFLPs) across a wide variety of Echinacea germplasm.²² This indicates that DNA markers may be useful in screening germplasm for active principles where phytochemistry may be less easy to assess.

In another study, Binns et al addressed the question of how much phytochemical variation is naturally present within and between wild populations.¹⁰ This is important for the wild harvest of seed for sale or cultivation, and the continued wild harvest of Echinacea populations from certain areas within the native range of these plants. E. pallida var. angustifolia was chosen for study because it has the largest latitudinal spread of any species/variety and occurs from Manitoba to Texas. There was significant variation in AAs and CADs between populations studied, which may support the existence of distinct chemoraces in this variety.¹⁰ Fortunately this variation is largely quantitative, and does not alter the phytochemical profile needed to identify the species. Also, since these experiments were conducted on seeds from the range of this variety, grown in uniform conditions, the variation measured was more likely to be genotypic rather than phenotypic, or caused by environmental influences.

Along with (North) American ginseng (Panax quinquefolius L., Araliaceae) Echinacea is possibly North America’s most important ethnobotanical product. Since it originates from here, the North American industry must contend with the genetic diversity associated with the center of origin of this medicinal crop.

Diversity and rarity in natural populations of Echinacea species and the case for cultivation of this natural resource

Degradation in wild populations and increased rarity has been observed over the last few decades during the Echinacea boom. Publicly-acknowledged rare taxa (and the states in the United States where they naturally grow, shown parenthetically) include:

Our taxonomic treatment of the genus did not lead directly to revision or clarification of protection laws regarding Echinacea. This is largely because while the taxonomic nomenclature changed, it has not yet been applied at the practical level of plant identification. Our assessments of diversity within genus Echinacea (genotypic and phenotypic) should be considered together with ecological evidence²⁶ in order to guide the taxonomic and practical protection of species and varieties that are at risk.

Ecophysiology, Competition, and Establishment

E. laevigata and E. pallida var. tennesseensis are located in marginal and highly vulnerable sites, on calcareous soils, in open woods, and in cedar barrens. It is likely that these and other “rare” taxa in this genus arose in Savannahs, which were caused and maintained by fires set by Native Americans.

In Tennessee, where E. pallida var. tennesseensis is the showy state wildflower, the ecological status and recovery of this Echinacea taxon has long been researched.^{27,28,29,30,31} Recovery operations in place by US Fish and Wildlife, partnered with the State Department of Environment and Conservation, include the following measures: restriction of vehicular traffic, added fencing, limited livestock use of the areas (there is evidence that cattle graze on plant competitors and thus might increase seedling establishment and survivorship), public education projects, ecological monitoring, and acquisition of land by the Nature Conservancy, as well as by federal and state departments. Very rare populations will be deemed “recovered” when at least 5 populations are self-sustaining (i.e., stable or increasing over at least 10 years, with at least 2 juvenile plants for every adult plant). Through this rigorous recovery activity and monitoring, E. pallida var. tennesseensis is expected to be down-listed from “endangered” to “threatened” in January, 2007 (T. Merritt, Personal Communication, USFWS Cooksville, Tennessee, August 22, 2006).

Is rarity an ecophysiological occurrence?

It was found that E. pallida var. tennesseensis plants are not highly competitive compared with other glade species, especially under the effects of allelopathy [release of chemical substances by one species that inhibit the germination and/or growth of other species of plants] by species such as: Juniperus virginiana L., Cupressaceae, and Dalea gattingeri (Heller) Barneby, Fabaceae [syn. Petalostemon gattingeri Heller].^30,32 Moreover, plants of this variety do not have significantly different ecophysiological requirements in terms of their light and moisture use.²⁹

Root physiology directly affects competitive abilities. Echinacea plants are taprooted forbs, which have difficulty establishing in both mixed and tallgrass prairie due to the competitive advantage of native grasses. However, E. pallida var. angustifolia, Psoralidium tenuiflorum (Pursch.) Rydb. Fabaceae (Slimflower scurfpea), Dalea spp. L. Fabaceae (Prairie clover), and other taprooted forbs generally outperform rhizomatous forbs, which compete directly with grasses for nutrients.³³ Edaphic constraints [factors pertaining to soil ecological relationships] due to competition for nutrients and water are higher for rhizomatous forbs.²⁶

There is also evidence that ecological variation (such as edaphic characteristics) affects genomic variability and/or expression of secondary chemistry. Chemotypes or chemical races were distinguished for populations of E. pallida var. angustifolia grown from wild seed,⁵ despite widespread and mostly continuous distribution of E. pallida var. angustifolia in a range of habitats. On the other hand, a significantly lower genetic diversity was measured in E. pallida var. tennesseensis^17,27 and was attributed to possible historical extinction and colonization events.²⁷ These 2 taxa are distinct but closely related, since the genetic makeup of E. pallida var. tennesseensis is identical to that of E. pallida var. angustifolia at 50% of genes studied by isozyme analysis, and there appears to be a subset of var. angustifolia alleles at another 28% of E. pallida var. tennesseensis genes studied.^4,27

Restoration of native prairie Echinacea populations is likely to depend largely on the other plants in the community. Seed recruitment is low. In some prairie remnants, there is evidence that wild-harvested Echinacea pallida var. angustifolia can re-sprout from holes where root fragments are left after digging (K. Kindscher, personal communication, June 5, 1999); however, the degree to which it may compete for establishment is still under longer-term study.

E. purpurea has been used for comparisons of competitive abilities (plant size and reproductive capacity) between wild and cultivated populations. Snyder et al showed that wild plants tend to have increased vegetative growth, while cultivated plants display increased reproductive capacity.³⁴

Is rarity a result of human activity?

Wilcove et al implicated habitat degradation in the decline of 85% of 1880 species of imperiled plants and animals in the United States.³⁵ Thirty-five percent of these were directly linked to commercial and residential human developments. Road construction and maintenance were almost equal. However, in the case of Echinacea, the largest human influence on rarity in certain taxa is undoubtedly wild-harvesting for the herb/dietary supplement trade.

Echinacea harvesting is controversial and, likely due to mass commercialization of the 1990s, it was particularly rampant throughout the Native American Indian reservations across the Great Plains. In 1990, North Dakotan people were encouraged to “just grab a shovel and start digging” while “environmentalists in the state fear[ed] that gold fever [was] spreading among shovel wielding collectors with dollar signs in their eyes.”³⁶ “Rooting,” as the digging of E. pallida var. angustifolia roots was called, has been an economic opportunity for both native and non-native people, as well as an ethical wildcrafting nightmare—where poachers have been known to effectively clear out wild populations of the plants without concern for preservation of the resources or the ecosystems.³⁷

Early regulation of the rampant wild harvesting came in the form of tribal resolutions in several states and later as legislation. For example, by 1999 North Dakota began to fine poachers $10,000 along with confiscation of their vehicles.³⁷ In Montana, a 3-year moratorium on the harvest of E. angustifolia from state lands, pushed by herbalist R. Klein,³⁷ led to the current Montana Code, which prohibits wild harvesting of E. pallida var. angustifolia from public lands without a permit, with a fine up to $1000 or 6 months in jail.³⁸ Despite these efforts, permits are still issued to dig for personal use (not commercial sale) of E. pallida var. angustifolia roots in Montana, and the harvesting regulation does not apply to aboriginal reservations or private landowners. Clearly, Echinacea prairie varieties at risk from ecophysiological factors and issues of genetic constraints, as discussed previously, have also been under risk of human mismanagement.

Public land conservation is achieved through scientific input provided to land management agencies and lobbying for federal legislation.³⁹ Conservation and restoration on private lands was traditionally achieved through zoning, condemnation, and tax regulation, with little success. Innovative bottom-up approaches are increasingly addressing private land degradation and slowing or halting the ravaging of natural ecological communities. Land trusts, open-space tax incentives, “community-based conservation,” and more have begun to effectuate stewardship. The Plant Conservation Alliance, which formed its “Medicinal Plant Working Group” in 1999, has established links between the Nature Conservancy and other institutions, NGOs, and the public in a “bottom up” effort to change wild harvesting practices. In fact, E. pallida var. pallida and E. pallida var. angustifolia were cited on the list of “Medicinal Plant Species in U.S. Commerce” as top priority warranting further study; the authors based their assessment on 1989-1999 data for trade demand increases, wild population declines, and species decline.⁴⁰ Finally, government grants ($3 US million in 2004) are now being awarded to protect plant species on tribal lands.⁴¹

Take-Home Messages

Echinacea forbs compete poorly to fairly with native grasses in prairie sites, resulting in low recruitment of Echinacea by seed. Significant factors include the following:

Public lands are relatively easy to protect with scientific evidence (data exist).

Private lands require new model for community-based natural resource management.

Habitat degradation due to development and human activity is the primary cause of rarity.

Echinacea wildcrafting is not sustainable.

Conclusion

During the herbal renaissance of the 1990s, Echinacea plants were the subject of much interest, common usage, and research scrutiny. Now, as science and markets for medicinal plants continue to evolve in the 21st century, the integration of findings from state-of-the-art original morphological, molecular, and phytochemical work by the authors of this article, along with ecological reports and the historical and regulatory literature of the times, suggests that these North American native plants deserve their status as protected resources. Moreover, it provides a comprehensive perspective into the biological and political origins of Echinacea materials sourced as phytomedicines, which is long overdue considering the focus on clinical evidence for Echinacea health products and dietary supplements.

Acknowledgements

This manuscript benefited from comments by Dr. E. Small, Agriculture & Agri-Food Canada, Eastern Cereal and Oilseed Research Centre, Ottawa. The financial support of Trout Lake Farms LLC (Washington, USA), Amway/Nutralite Corporation (USA), and MediPlant Consulting Services (Dennis V.C. Awang, PhD, FCIC) of White Rock, BC, Canada, is gratefully acknowledged. We thank S. Mechanda and J. Livesey (Research Technicians, Agriculture and Agri-Food Canada and the University of Ottawa, Canada) for their contribution to the experimental work.

Bernard R. Baum. Agriculture & Agri-Food Canada, Eastern Cereal and Oilseed Research Center, Neatby Building, 960 Carling Avenue, Ottawa, Ontario, Canada, K1A0C6. Corresponding author: e-mail: baumbr@agr.gc.ca; phone: 613-759-1821.

Shannon E. Binns.University of British Columbia, Faculty of Land and Food Systems, 2357 Main Mall, Vancouver, British Columbia, Canada, V6T 1Z4.

John T. Arnason. Department of Biology, University of Ottawa, 30 Marie Curie, Ottawa, ON, Canada K1N 6N5.

References

1. Cronquist A. Notes on the Compositae of the Northeastern United States II. Heliantheae and Helenieae. Rhodora. 1945;47:396-398.

2. McGregor RL. The taxonomy of the genus Echinacea (Compositae). Univ Kansas Sci Bull. 1968;48:113-142.

3. Bauer R, Wagner H. Echinacea species as potential immunostimulatory drugs. In: Wagner H, Farnsworth N, eds. Economic and Medical Plant Research. 5th ed. London: Academic Press;1991:253-321.

4. Binns SE, Baum BR, Arnason JT. A taxonomic revision of the genus Echinacea (Heliantheae; Asteraceae). Syst Bot. 2002;27:610-632.

5. Binns SE, Baum BR, Arnason JT. Phytochemical variation in populations of Echinacea angustifolia (Asteraceae). Biochem Syst Ecol. 2002;30:837-854.

6. Binns SE, Livesey JF, Arnason JT, Baum BR. Phytochemical variation in Echinacea Moench (Heliantheae: Asteraceae) from roots and flowerheads of wild and cultivated populations. Jour Agric Food Chem. 2002;50:3673-3687.

7. Gower JC. A general coefficient of similarity and some of its properties. Biometrics. 1971 27:857-871.

8. Binns SE, Baum BR, Arnason JT. Typification of Echinacea purpurea (L.) Moench (Heliantheae: Asteraceae) and its implications on the correct naming of two Echinacea taxa. Taxon. 2001;50:1169-1175.

9. McKeown KA. A Review of the taxonomy of the Genus Echinacea. In: Janick J, ed. Perspectives on New Crops and New Uses. Alexandria, VA: ASHS Press; 1999:482-489.

10. Binns SE, Baum BR, Arnason JT. Taxon. Proposal to conserve the name Rudbeckia purpurea L. (Asteraceae) with a different type. 2001;50:1199-1200.

11. Hitchcock CL, Cronquist A, Ownbey M, Thompson JW. Vascular plants of the Pacific Northwest. Vol 5. Seattle: University of Washington Press. 1955-69.

12. Radford AE, Hardin JW, Massey JR, Core EL, Radford LS, eds. Vascular flora of the southeastern United States. Vol 1. Chapel Hill: University of North Carolina Press; 1980.

13. Urbatsch LE, Jansen RK. Phylogenetic affinities among and within the coneflower genera (Asteraceae, Heliantheae), a chloroplast DNA analysis. Syst Bot. 1995;20:28-39.

14. Robinson H. Studies in the Heliantheae (Asteraceae), XIV: Validation of subtribes. Phytologia. 1978;41:39-44.

15. Robinson H. A revision of the tribal and subtribal limits of the Heliantheae (Asteraceae) Smithsonian Contrib Bot. 1981;51:1-102.

16. Urbatsch LE, Baldwin BG, Donoghue MJ. Phylogeny of the coneflowers and relatives (Heliantheae: Asteraceae) based on nuclear rDNA internal transcribed spacer (ITS) sequences and chloroplast DNA restriction site data. Syst Bot. 2000;25:539-565.

17. Mechanda SM, Baum BR, Johnson DA, Arnason JT. Analysis of diversity of natural populations and commercial lines of Echinacea using AFLP. Can J Bot. 2004;82:461-484.

18. US Fish and Wildlife Service. Threatened and endangered Species System—E. laevigata and E. tennesseensis. List updated March 31, 2004. Available at: http://ecos.fws.gov/species_profile/SpeciesProfile?spcode=Q293 and http://ecos.fws.gov/species_profile/SpeciesProfile?spcode=Q1VU.

19. Kim DH, Heber D, Still DW. Genetic diversity of Echinacea species based upon amplified fragment length polymorphism markers. Genome. 2004;47:102-111.

20. Clark AG, Lanigan CMS. Prospects for estimating nucleotide divergence with RAPDs. Mol Biol Evol. 1993;10:1096-1111.

21. Kapteyn J, Goldsbrough PB, Simon JE. Genetic relationships and diversity of commercially relevant Echinacea species. Theor Appl Genet. 2002;105:369-376.

22. Baum BR, Mechanda S, Livesey JF, Binns SE, Arnason JT. Predicting quantitative phytochemical markers in single Echinacea plants or clones from their DNA fingerprints. Phytochemistry. 2001;56:543-549.

23. Bauer R. Echinacea: Biological effects and active principles. In: Lawson LD, Bauer R, eds. Phytomedicines of Europe: Chemistry and Biological Activity. ACS symposium series 691. Washington, DC: American Chemical Society;1998:140-157.

24. Bauer R, Khan I, Wagner H. TLC and HPLC analysis of Echinacea pallida and E. angustifolia roots. Planta Med. 1988;54:426-430.

25. Binns SE, Inparajah I, Baum BR, Arnason JT. MeJ increases alkamide and ket