u Ottawa l.'Univcrsilc! cnnnrlicwu- Cnnodn's univcrsiiy FACULTE DES ETUDES SUPERIEURES ls=l FACULTY OF GRADUATE AND ET POSTOCTORALES U Ottawa POSDOCTORAL STUDIES

L*University canadiennc Canada's university

Vicky J. Filion AUTEUR DETATHISET/ AUTHOR OF THESIS"

M_.Sc. (Biology) GRADE/DEGREE

Department of Biology TAMTE"TCOLirDpARTEMENT~^ACUITY, S*CHOdi~DEWRTMENT"

A Novel Phytochemical and Ecological Study of the Nunavik Medicinal Plant Rhodiola rosea L.

TITRE DE LA THESE / TITLE OF THESIS

_____„„____Dr._ J. Arnason

A. Cuerrier

EXAMINATEURS (EXAMINATRICES) DE LA THESE / THESIS EXAMINERS

Dr. C. Charest

Dr. J. Kerr

Dr. N. Cappuccino

_Garjr\V^SJatCT_ Le Doyen de la Faculte des etudes superieures et postdoctorales I Dean of the Faculty of Graduate and Postdoctoral Studies A Novel Phytochemical and Ecological Study of the Nunavik Medicinal Plant Rhodiola rosea L.

Vicky J. Filion

Thesis submitted to the Faculty of Graduate and Postdoctoral Studies University of Ottawa in partial fulfillment of the requirements for the M.Sc. degree in the

Ottawa-Carleton Institute of Biology

These soumise a la Faculte des etudes superieures et postdoctorales Universite d'Ottawa en vue de l'obtention de la maitrise es sciences

Institut de biologie d'Ottawa-Carleton

© Vicky J. Filion, Ottawa, Canada, 2008 Library and Bibliotheque et 1*1 Archives Canada Archives Canada Published Heritage Direction du Branch Patrimoine de I'edition

395 Wellington Street 395, rue Wellington Ottawa ON K1A0N4 Ottawa ON K1A0N4 Canada Canada

Your file Votre reference ISBN: 978-0-494-50879-4 Our file Notre reference ISBN: 978-0-494-50879-4

NOTICE: AVIS: The author has granted a non­ L'auteur a accorde une licence non exclusive exclusive license allowing Library permettant a la Bibliotheque et Archives and Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par telecommunication ou par Plntemet, prefer, telecommunication or on the Internet, distribuer et vendre des theses partout dans loan, distribute and sell theses le monde, a des fins commerciales ou autres, worldwide, for commercial or non­ sur support microforme, papier, electronique commercial purposes, in microform, et/ou autres formats. paper, electronic and/or any other formats.

The author retains copyright L'auteur conserve la propriete du droit d'auteur ownership and moral rights in et des droits moraux qui protege cette these. this thesis. Neither the thesis Ni la these ni des extraits substantiels de nor substantial extracts from it celle-ci ne doivent etre imprimes ou autrement may be printed or otherwise reproduits sans son autorisation. reproduced without the author's permission.

In compliance with the Canadian Conformement a la loi canadienne Privacy Act some supporting sur la protection de la vie privee, forms may have been removed quelques formulaires secondaires from this thesis. ont ete enleves de cette these.

While these forms may be included Bien que ces formulaires in the document page count, aient inclus dans la pagination, their removal does not represent il n'y aura aucun contenu manquant. any loss of content from the thesis. Canada ABSTRACT

Canadian populations of Rhodiola rosea L. growing inNunavik, Quebec, were

mapped and examined for the presence of medicinal phytochemicals for the first time. Four

relevant phytochemicals salidroside and rosavins, including rosavin, rosarin and rosin, were

detected in various Canadian collections. Quantitative comparison of chemical profiles,

from different circumpolar samples, revealed significant variations between European and

Canadian populations and among Nunavik local populations. In Canada, plants from the

Mingan Islands showed the highest quantity of phytochemicals. Salidroside concentration in

Nunavik plants was significantly affected by herbivory and gender whereas amount of rosavins remained stable.

The ecological survey of Nunavik populations indicated an irregular distribution of R. rosea along Ungava Bay. Distribution maps were created and a predictive distribution model was generated. Plant abundance was observed along the east coast as opposed to near absence on the west coast. Various environmental factors and historical contingency could have influenced this distribution.

n RESUME

Cette etude porte sur la phytochimie et Tecologie des populations canadiennes du

Rhodiola rosea L. au Nunavik. Les quatre composes actifs cles de la plante, soit salidroside

et les rosavines, incluant rosavine, rosarine et rosine, ont ete retrouves chez toutes les

populations canadiennes etudiees. Le profil phytochimique des populations europeennes et

canadiennes a ete compare. Les resultats ont montre que les composes medicinaux variaient

significativement entre les populations. Au Canada, la quantite la plus elevee de principes

actifs a ete retrouvee chez les populations des iles Mingan. De plus, seul le salidroside contenu dans les plantes du Nunavik a ete affecte par la presence d'acariens et le sexe de la plante.

1/ analyse ecologique a revele que le R. rosea du Nunavik etait distribue inegalement sur les cotes de la Baie d'Ungava (Quebec, Canada), celui-ci etant plus abondant sur la cote est. Plusieurs facteurs environnementaux et historiques peuvent expliquer cette difference.

in TABLE OF CONTENT

ABSTRACT ii

RESUME iii

TABLE OF CONTENTS iv

LIST OF TABLES vii

LIST OF FIGURES viii

LIST OF MAPS x

ACKNOWLEDGEMENTS xi

CHAPTER 1: GENERAL INTRODUCTION: RHODIOLA ROSEA OVERVIEW 1

1.1 General Introduction 2

1.2 Literature Review 3

Plant description 3

Biological properties 6

Phytochemistry 9

Commercialized products 11

1.3 Research Rationale 13

1.4 Research Objectives 13

Phytochemical objectives 13

Ecological objectives 14

CHAPTER 2: PHYTOCHEMICAL ANALYSIS OF RHODIOLA ROSEA L 15

2.1 Introduction 16

2.2. Phytochemical Analysis 17

iv 2.2.1 Material and Methods 17

2.2.2 Results 21

Inter-regional variation ofR. rosea from circumpolar collections.... 21

Inter-population variation ofNunavik R. rosea 24

Variation according to latitude 27

Soil nutrient impact 27

Herbivory impact 27

Effect of plant gender 31

2.2.3 Discussion 31

CHAPTER 3: ECOLOGICAL ANALYSIS OF NUNAVIK RHODIOLA ROSEA L 42

3.1 Introduction 43

3.2 Material and Methods 45

Geographical distribution ofR. rosea 45

Maxent distribution model 46

Habitat characterization in Nunavik 47

Investigation for plant mite infestation 48

Experimental plantation in Nunavik 48

3.3 Results and Discussion 49

Geographical distribution ofR. rosea 49

1. Rhodiola rosea distribution in Nunavik 49

2. Deformed Rhodiola rosea distribution in Nunavik. 56

3. "Worldwide distribution of herbarium specimens 58

Maxent distribution model 62

v Habitat characterization in Nunavik 62

Investigation for plant mite infestation 67

Experimental plantation in Nunavik 72

CHAPTER 4: GENERAL DISCUSSION 75

4.1 Major findings and claims to originality. 76

4.2 Comparison with published literature on R. rosea 77

4.3 Future work 79

APPENDICES 82

Al: Analytical Method Development Manuscript 82

A2: Phytochemical Analysis of Commercialized Rhodiola rosea 89

REFERENCES 95

vi LIST OF TABLES

Table 1.1. Six important phytochemical groups and their compounds found in R. rosea....10

Table 2.1. Newly developed HPLC analytical method. Timetable of solvent gradient according to time (minutes) and flow rate (mL/min) 19

Table 2.2. Mean phytochemical concentration of salidroside and rosavins (mg/ g) from four Nunavik soil analyses in relation to three essential soil nutrients: nitrate, phosphorus and potassium 30

Table 3.1. Ecological data compiled from Nunavik collection set 1, August expedition 2005, UngavaBay, QC 64

Table 3.2. Ecological data compiled from Nunavik collection set 2, August expedition 2006, Ungava Bay, QC 65

Table 3.3. Nutrient content (N-P-K) and soil pH level of four soil samples collected in Nunavik, QC, and analyzed according to the Mehlich 3 method, by AgriDirect 68

Table 3.4. Survival rate of the Nunavik experimental plantation established in Kangiqsualujjuaq, August 2006 74

Table A2-1. List of commercialized products ofRhodiola rosea analysed phytochemically for their concentration in the four markers. Products names are arranged in alphabetical order 92

Table A2-2. Product accession numbers were randomly assigned to products. One gram of product was extracted from a homogenized mixture of material. Product # 20 was under liquid form (volume of 50 mL) and subsequently evaporated to dryness and weighted before analysis 93

Table A2-3. Mean phytochemical concentration of salidroside and rosavins (mg/ g) found in twenty commercialized natural products of R. rosea. Accession numbers were attributed randomly 94

vii LIST OF FIGURES

Figure 1.1. Morphology of roseroot (Saule 2002) 4

Figure 1.2. Molecular structures of four important phytochemicals present in Nunavik Rhodiola rosea , 12

Figure 2.1. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM for eight regionally distinct circumpolar collections of Eurasia and Canada 22

Figure 2.2. Concentration (mg/ g) of salidroside and rosavins present in eight circumpolar collections of Eurasia and Canada 23

Figure 2.3. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM for different populations of Rhodiola rosea collected in August 2005 during the Nunavik collection set 1 expedition 25

Figure 2.4. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM for different populations of Rhodiola rosea collected in August 2006 during the Nunavik collection set 2 expedition 26

Figure 2.5. Dendrogram resulting from hierarchical cluster analysis of Ward using both salidroside and rosavins. The tree illustrates similarities among different Nunavik populations according to their concentration in the target compounds, independently of year 28

Figure 2.6. Concentration (mg/ g) of salidroside (A) and rosavins (B) from R rosea of Nunavik collection set 2, according to increasing latitude, from south to north (58°19'35.6" to61°04'41.4") 29

Figure 2.7. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM in healthy and gall-deformed Nunavik R rosea plants, from Nunavik collection set 2. The asterisk indicates significant difference 32

viu Figure 2.8. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM for different Nunavik R. rosea plants, from Nunavik collection set 2, according to gender (hermaphrodite, male, immature and female) 33

Figure 3.1. Rhodiola integrifolia herbarium specimen (A) and Rhodiola rosea herbarium specimen (B) 61

Figure 3.2. Healthy infructescence of R. rosea plant (A) versus mite infested infructescence of Nunavik plants (B) 70

Figure 3.3. Eriophyoid bud mite Aceria rhodiolae (Canestrini) infesting Nunavik R. rosea bud tissue. The digital pictures were taken with a compound microscope at lOx and 40x objectives 71

ix LIST OF MAPS

Map 3.1. Geographic representation of the study area of R. rosea in Nunavik, Quebec, Canada 50

Map 3.2. Dispersion of Rhodiola rosea according to presence and absence points along Ungava Bay, Nunavik, QC, Canada 51

Map 3.3. Distribution of deformed and healthy Rhodiola rosea plants along Ungava Bay, Nunavik, QC, Canada 57

Map 3.4. Worldwide distribution of two Rhodiola herbarium specimens: R. rosea (circumpolar range) and R. integrifolia (limited to western NA) 59

Map 3.5. Close up of the distribution of R. rosea herbarium specimens on the east coast and R. integrifolia herbarium specimens on the west coast of North America 60

Map 3.6. Maxent model representing Rhodiola rosea''s predicted geographic distribution along Ungava Bay, Nunavik, QC, according to six environmental variables 63

x ACKNOWLEDGMENTS

I would like to thank Dr. J.T. Arnason (Supervisor and biology professor, University

of Ottawa), Dr. A. Cuerrier (Co-supervisor and researcher, Jardin botanque de Montreal and

Institut de recherche en biologie vegetale,) and the Makivik Corporation for this exceptional

project and financial support. I would like to acknowledge the members of my thesis

advisory committee for their precious guidance throughout this research: Dr. N. Cappuccino

(Carleton University), Dr. C. Charest (University of Ottawa) and Dr. J. Kerr (University of

Ottawa). I deeply appreciated the assistance and help of Heather Kharouba (Master graduate, Dr. Kerr laboratory, University of Ottawa) for her expertise with Maxent, and the technical expertise of Linda Kimpe (Dr. Blais laboratory, University of Ottawa) on the ASE.

Also, I thank Gisele Mitrow (collection manager, Agriculture and Agri-Food Canada,

Vascular Plant Herbarium-DAO), Micheline Bouchard (collection technician, National

Herbarium of Canada-CAN), Sylvie M. Fiset and Serge Payette (Universite Laval

Herbarium-QFA) and Jeannette Whitton (Director and Curator, Royal BC Museum herbarium-RBCM) for sharing their herbarium collections. I am obliged to Dr. F. Beaulieu

(Mite Systematics, Agriculture and Agri-Food Canada) for his scientific knowledge and identification of the Nunavik plant gall-causing agent. To Sunniva M.D. Aagaard and Ingar

Pareliussen for providing me with root samples from Norway and Parks Canada for those from the Mingan Islands. Also to Stephane Daigle (Statistician, IRBV) for his insights on statistical analyses.

I am very grateful to Makivik and Nunavik Biosciences Inc. who initiated, funded and closely supervised this project as well as to the Northern Scientific Training Program

xi (NSTP) from the Development of Indian Affairs and Northern Development Canada who

generously contributed to several field seasons expenses. Finally, I would like to thank my

colleague and collaborator Mariannick Archambault (IRBV, Universite de Montreal) for the

many hours spent in the field and relevant discussions as well as Claudine Ethier my

dedicated field assistant. Thanks to the Nunavimmiut for generously welcoming us within

their communities and their traditional knowledge of the land. I am also grateful to Dr. A.

Saleem (Dr. Arnason laboratory, University of Ottawa) for his phytochemical expertise during the analytical method development. Many thanks to all my knowledgeable lab mates:

Brendan Walshe-Roussel, Theresa Tarn and Chieu Anh Ta for their help with bioassays as well as Rosalie Awad, Martha Mullally, Renee Leduc, Carolina Ogrodowczyk and Cory

Harris and Karen Foster (Dr. Blais laboratory, University of Ottawa) for their stimulating conversations and support.

A warm thank you to Beatrice Riche and Patrick Audet (Dr. Charest laboratory,

University of Ottawa) for their insights on this thesis.

Et surtout, je voudrais remercier sincerement ma famille et mes amis pour leur support constant et leurs encouragements. Merci!

xu Chapter 1: General Introduction: Rhodiola rosea L. overview

CHAPTER 1

General Introduction: Rhodiola rosea L. overview

1 Chapter 1: General Introduction: Rhodiola rosea L. overview

1.1 General introduction

Rhodiola rosea L. (Crassulaceae), commonly called "roseroot", "golden root", "arctic

root" or "orpin rouge", is a circumpolar medicinal plant, which has been used traditionally

by indigenous people of the Russian Federation, Scandinavia and Eastern Arctic Canada

(Clausen 1975, Aiken et al. 1999, Aim 2004, Cuerrier et al. 2005). To date, more than fifty

Rhodiola species have been recognized, twenty of which have enthobotanical uses in Asian

communities (Hegi 1963, Kelly 2001, Brown 2002, Khanum et al. 2005). Roseroot has

become a widely used medicinal plant in world markets and the many health benefits

attributed to it include antidepressant, immunostimulant, neurostimulant, anticarcinogenic,

and antimutagenic properties (Brown et al. 2002, Khanum et al. 2005). This plant is also

claimed to be an important adaptogen (Panossian et al. 1999, Kelly 2001, Brown et al. 2002,

Winston & Maimes 2007). Wildcrafting is used to meet most of the market demand, which has resulted in an overexploitation of wild Eurasian populations (Galambosi 2006). As a result, efforts are being implemented in various Eurasian countries to develop sustainable harvests and find alternative means of increasing plant material production through introduction and cultivation practices (Galambosi 2006, Kucinskaite et al. 2007).

The location of populations of R. rosea in Canada is still largely unstudied.

Specimens appearing in the national herbaria suggest populations are naturally occurring in

Northern Quebec and the Maritimes, although their distribution and densities are unknown.

This research project investigated Canadian populations of Rhodiola rosea for the first time.

At the request of Nunavik Inuit representatives, this study examined in detail the phytochemistry and ecology of Northern Quebec populations. This examination aimed to determine the medicinal potential of Canadian growing plants, by quantifying and comparing the plant's phytochemistry among different collections and by determining its geographic Chapter 1: General Introduction: Rhodiola rosea L. overview

distribution to establish the plant's ecological requirements and habitat in Nunavik as well as

future conservation prospects. To better understand the characteristics of Rhodiola rosea,

this chapter will review available literature on this plant species and focus on the research

rationale and objectives.

1.2 Literature Review

Plant description

Geographical distribution: Roseroot occupies a general circumpolar distribution and grows mostly in arctic environments. R. rosea is found in the Russian Federation, Asia

(China, Altai and Himalayan Mountains, Mongolia), Europe (Scandinavia, Alp regions of

France, Switzerland and Italy), Greenland and eastern Canada (Northern Quebec, Labrador coast, Nunavik, Nunavut Islands, Baffin Islands and the Maritimes) (Hegi 1958, Clausen

1975, Aiken et al. 1999, Ganzera et al. 2001, Aim 2004).

Plant biology and morphology: Rhodiola rosea is a fleshy perennial herb from the

Crassulaceae (Aiken et al. 1999) (Figure 1.1). This dioecious plant normally grows to heights of 5-70 cm. Leaves are pale green, glaucous, ovate to oblong, entire or dented and 1-

4 cm in length (Porsild 1964, Clausen 1975). The plant reproduces both sexually, by seeds, and asexually, through vegetative propagation of the rhizome (Clausen 1975). The imperfect flowers are four or five-merous (Polunin 1959), and aggregate in a fascicule at the top of a succulent stem. Roseroot flowers are yellow; however, Marie-Victorin's Laurentian Flora and Aiken et al. (1999), indicate that the male flowers may have a purple tinge, although this was not observed among Nunavik populations. Reddish purple flowers are characteristic of

Rhodiola intergrifolia. Nectaries are visible on the R. rosea flowers, which suggest insect 3 Chapter 1: General Introduction: Rhodiola rosea L. overview

Rhodiola rosea L

Figure 1.1. Morphology of roseroot (Saule 2002)

4 Chapter 1: General Introduction: Rhodiola rosea L. overview

pollination. Fruits are erect, non-fleshy brown follicles and seeds are wind dispersed

(Clausen 1975).

The fibrous rhizomes and roots, which are thick and pale brown, are the main source

of medicinal phytochemicals. In fact, the common names of R. rosea emphasize the

importance of the rhizome as a main reserve for bioactive products (Brown et al. 2002). For

instance, roseroot alludes to the fragrant smell of roses that emanates from the severed

rhizome, due to the presence of geraniol (Rohloff 2002, Khanum et al. 2005).

Habitat: Growing habitats of roseroot include: the dry tundra, arctic terrains, cliffs, shorelines, near fresh and salt water, as well as northern sides of mountains and human disturbed environments (Clausen 1975, Aiken et al. 1999). Roseroot grows on various substrates, including till, gravel and crevices in rocks. These habitats vary in soil moisture, which range from moist to well-drained as well as in organic matter content. Plants from the arctic tundra have adapted to be stress tolerant due to nutrient poor soils, extreme temperature fluctuations, desiccating winds and water deficiency (Shevtsova et al. 2005).

Taxonomy: Rhodiola rosea has a complex taxonomical classification. Originally, roseroot was assigned to the genus Sedum and was known as Sedum rosea. In 1963, based on morphological studies, Hegi reclassified many species of the genus Sedum into the new

Rhodiola genus (Hegi 1963, Brown et al. 2002). Sedum rosea was then renamed Rhodiola rosea. Many species belonging to the Rhodiola genus have been used as medicinal plants, e.g. R. sachalinensis, R. sacra and R. kirilowii (Brown et al. 2002). Of those, only a dozen have been specifically studied and tested through human or animal trials, although not all tests have been conclusive (Fan et al. 2001, Brown et al. 2002). In some instances, various Chapter 1: General Introduction: Rhodiola rosea L. overview

Rhodiola species have been combined in commercial products. Consequently, much

confusion has arisen in the market because certain products have misleadingly labelled and

promoted "Rhodiola" plant material without specifying the species used (Brown et al. 2002).

Only certain Rhodiola species contain medicinally active phytochemicals and

therefore are acceptable as medicinal products (Brown et al. 2002). In R. rosea, specific

chemicals have been isolated and identified as biologically active compounds, three of which

are found to be unique to roseroot: rosavin, rosarin and rosin. This explains the medicinal

success of this particular species.

Still today, confusion and misidentification among Rhodiola spp. occur due to the complex taxonomic history of this genus. For instance, R. integrifolia is often misidentified as R. rosea. This confusion in species identification can affect herbalists and medicinal plant growers; therefore it is increasingly important to develop means for identifying the species of

Rhodiola cultivated and used in natural products.

Biological properties

Ethnobotanical uses: Commercialized preparations of roseroot are widely used in

Eurasia. The pbytomedicinal uses are based on traditional usage of R. rosea as an ethnobotanical remedy. In Northern Norway, the indigenous Sami people employed this plant as food, human and veterinary medicine, hair wash and as preventive treatment for scurvy. It was also given to newlyweds as a symbol of fertility (Aim 2004). Some communities have retained their folk tradition and drink tinctures of roseroot in water: alcohol based extracts, e.g. Vodka (Brown et al. 2002; Aim 2004), while others drink it as tea infusion (Ming et al. 2005). Reports suggest that Vikings used R. rosea to prepare for long trips, the belief being that roseroot increased body strength and physical endurance (Aiken et 6 Chapter 1: General Introduction: Rhodiola rosea L. overview

al. 1999). limit people used roseroot to prevent illness or re-establish the well being of the

whole body (Aim 2004). In the Chinese tradition, roseroot was used to increase physical

resistance, reduce fatigue and extend human life (Tolonen et al. 2003, Drasar and Moravcova

2004). In fact, many cultures believed that R. rosea decreased the effects of aging. Many

health benefits are attributed to R. rosea; the principal ones include adaptogenic,

neurostimulant, antidepressant and antioxidant effects as well as positive action on the

reproductive system and physical strength and resistance (Brown et al. 2002, Khanum et al.

2005).

Adaptogenic properties: Russian literature supports the theory thati?. rosea has

"adaptogenic" properties. Adaptogens however are not yet recognized by modern Western medicine. According to concepts developed by Nicholai Lazarev in 1947 (Panossian et al.

1999), traditional medicine describes adaptogenic plants as those that reduce fatigue, strengthen energy and maintain the body's equilibrium (Brekhman and Dardymov 1969). It is therefore believed that R. rosea has a normalizing and positive action on the body without causing negative effects (Panossian et al. 1999, Rege et al. 1999, Davydov and Krikorian

2000, Khanum et al. 2005, Perfumi and Mattioli 2007). According to Winston and Maimes

(2007), an adaptogenic plant plays an important role in regulating the hypothalamic pituitary adrenal axis. Other well-known adaptogenic plants are Panax spp., Eleutherococcus senticosus and Schisandra chinensis (Panossian et al. 1999).

Neurostimulant activity: Another major function of roseroot involves the stimulation of the central nervous system (CNS) (Khanum et al. 2005). Rhodiola rosea acts on specific systems that enhance cognitive functions, such as memory and concentration (Kurkin et al. 7 Chapter 1: General Introduction: Rhodiola rosea L. overview

1986, Brown et al. 2002). On a long-term basis, roseroot reduces deterioration of the limbic

system pathways and memory dysfunctions related to age (Brown et al. 2002). Numerous

studies done on the subject support that R. rosea improves memory and concentration

(Shevtsov et al. 2003, Hillhouse et al. 2004).

Antidepressant activity: Molecular influences on CNS neurotransmitters also

demonstrate how roseroot can act as an antidepressant. In short, the stimulating activity of

roseroot phytochemicals leads to reduced anxiety, stress and asthenia (chronic fatigue)

(Brown et al. 2002, Khanum et al. 2005, Perfumi and Mattioli 2007). Due to this neurostimulation, roseroot is not recommended for everyone and contraindications apply to those suffering from high blood-pressure, bipolar disorder or other particular illnesses

(Khanum et al. 2005).

Antioxidant and anticancer activities: Another important physiological benefit of roseroot is its antioxidant activity (Khanum et al. 2005, Pooja et al. 2006). Secondary metabolites from R. rosea are known to protect the nervous system tissues from deterioration by oxidizing free radicals (Brown et al. 2002). Antioxidants prevent the corrosion and weakening of important arteries and tissues caused by aging and disease.

Additionally, cancer chemopreventative activity has been observed in animal studies

(Majewska et al. 2006). Roseroot has been shown to decrease the toxicity levels of anticancer drugs (e.g. cyclophosphamide, rubomycin and adriamycin), while at the same time intensifying the anticarcinogenic response (Udintsev and Schakhov 1991, Brown et al.

2002, Khanum et al. 2005). This property is in part responsible for the antiaging effect provided by roseroot extract. 8 Chapter 1: General Introduction: Rhodiola rosea L. overview

Reproductive system effects: In an old Norwegian tradition, newlyweds were given

roseroot cuttings as a symbol of fertility (Aim 2004). Experimental studies have shown that

roseroot extracts influence estrogens levels (Brown et al. 2002). R. rosea has also been

found to improve male sexual dysfunction, including erection problems and premature

ejaculation. Roseroot also has a therapeutic effect on symptoms of menopause. The

aphrodisiac property of roseroot may well be linked to its anxiolytic effect.

Physical activity enhancement: Rhodiola rosea has the ability to increase physical endurance and recovery time after training. Numerous studies involving athletes have shown conclusive results, which made roseroot a popular medicinal plant (Kelly 2001, Brown et al

2002, Abidov et al. 2003). The psychostimulant property of R. rosea reduces the effect of fatigue on the brain and the body. This significantly decreases symptoms brought on by physical exhaustion such as loss of coordination, trembling and loss of accuracy (Brown et al. 2002). Roseroot also increased energy at the cellular level, by limiting acidosis (Abidov et al. 2003).

Phytochemistry

Since the 1970's, studies have reported six distinct groups of phytochemicals in roseroot (Table 1.1) (Brown et al. 2002, Khanum et al. 2005). Within those groups, more than twenty-eight phytochemicals have been isolated, twelve of which are novel substances

(Kelly 2001, Rohloff 2002, Kurkin 2003). As a preliminary Canadian study, this research focused on four specific compounds: the phenylethanol derivative salidroside and the three

9 Chapter 1: General Introduction: Rhodiola rosea L. overview

Table 1.1 Six important phytochemical groups and their compounds found in R. rosea.

Phytochemical groups Specific compounds

Phenylpropanoids rosin, rosarin, rosavin, (the rosavins) Phenylethanol derivatives tyrosol, salidroside (rhodioloside) Flavonoids rodionin, rodiosin, acetylrodalgin, tricin, rodiolin Monoterpernes rosaridin, rosiridol Triterpenes beta-sitosterol, daucosterol Phenolic acids gallic acids, chlorogenic, hydroxycinnamic

(Adapted from Brown et al. 2002)

10 Chapter 1: General Introduction: Rhodiola rosea L. overview

phenylpropanoid glycosides rosavin, rosarin, and rosin, known as the rosavins (Figure 1.2).

The targeted compounds are recognized as having a considerable medicinal activity (Ming et

al. 2005). However, the phenylpropanoids dominate in importance and are found exclusively

in Rhodiola rosea. Salidroside on the other hand, is common in other medicinal plant

species (Kelly 2001).

To date, those four R. rosea phytochemicals are recognized as the most significant constituents needed for therapeutic activity of./?, rosea (Kelly 2001, Ming et al. 2005).

Commercial natural products of roseroot are thus standardized according to their content in salidroside and rosavins (Russian National Pharmacopoeia 1989, Brown et al. 2002).

Commercialized products

According to Health Canada's list for "Approved Natural Health Products", at least one R. rosea phytomedicine is available on the Canadian market, "Arctic Root SHR-5" tablets from the Swedish Herbal Institute. However, the United States of America have a larger variety of roseroot products. These are advertised as a cure for stress, help for relaxing, anti-aging formulas and supplements for body building. A few examples of such commercialized products are Arctic root ™ by Swedish herb Proactive Bioproducts, Simple

Energy T ' from Enzymatic Therapy, and Rhodiola energy ™ from PhytoPharmic Enzymatic therapy. Most products are sold in the form of capsules or dried rhizome.

For this Masters project, twenty commercially available R. rosea products were analysed for their content of salidroside and rosavins in order to compare them to their advertised amounts (Annex 2).

11 Chapter 1: General Introduction: Rhodiola rosea L. overview

Figure 1.2. Molecular structures of four important phytochemicals present in Nunavik Rhodiola rosea.

12 Chapter 1: General Introduction: Rhodiola rosea L. overview

1.3 Research Rationale

This project was initiated in collaboration with the Nunavik Inuit-based Makivik

Corporation and its Inuit communities. Makivik wished to assess whether Canadian

populations of roseroot were of suitable quality for use as a phytomedicine, in order to assist

northern Inuit communities in developing a northern-based micro-enterprise, i.e. cultivation

and harvesting sites within the Inuit communities, to provide locals with employment and

financial support.

At a more fundamental level, this project was developed to characterize for the first time

Nunavik populations of R. rosea with respect to their phytochemical profile and ecology

(distribution and habitat). The phytochemistry investigations were key to determining if

Canadian populations were similar to Eurasian chemotypes. Distribution and ecological studies were undertaken as a first step in developing rationale conservation measures for

Canadian wild populations.

1.4 Research Objectives

Phytochemical objectives

1. To quantitatively analyze four key secondary metabolites (salidroside and rosavin,

rosarin, rosin) of the Nunavik R. rosea populations and circumpolar collections.

2. To compare the phytochemical profiles of Nunavik plants with circumpolar

collections of R. rosea, including Siberian cultivated plants, and wild Norwegian and

Canadian populations.

3. To examine phytochemical variations among regionally distinct Nunavik samples.

13 Chapter 1: General Introduction: Rhodiola rosea L. overview

4. To determine if correlations exist between the phytochemical variation of Nunavik

populations with determining environmental factors such as latitude, soil conditions,

herbivory pressure and plant gender.

Ecological objectives

1. To generate a geographic distribution map for R. rosea along Ungava Bay, Northern

Quebec.

2. To create a predictive species distribution model for areas not surveyed.

3. To characterize Rhodiola rosed's ecological requirements and habitat types in

Nunavik.

4. To investigate the gall-producing agent present on Nunavik plant inflorescences.

5. To evaluate the Nunavik experimental plantation for its potential in economic

development.

14 Chapter 2: Phytochemical Analysis of Rhodiola rosea L.

CHAPTER 2

Phytochemical Analysis of Rhodiola rosea L.

15 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

2.1. INTRODUCTION

Rhodiola rosea L. is an arctic medicinal plant belonging to the Crassulaceae having

traditional usages in circumpolar regions including the Russian Federation, Asia,

Scandinavia and eastern North America (Aiken et al. 1999, Brown et al. 2002, Cuerrier et al.

2005). In reference to the ethnobotanical usage of the rhizome, this succulent herbaceous

perennial plant is known by common names such as "roseroot", "arctic root" and "golden

root" (Small and Catling 1999, Brown et al. 2002). Therapeutic properties ascribed to this

plant include antidepressant and antioxidant activities as well as immunostimulant and

adaptogenic effects (Kelly 2001, Brown et al. 2002, Tolonen et al. 2004, Winston and

Maimes 2007).

Early studies of R. rosea have identified two primary active phytochemical groups:

the phenylethanol derivatives, which include the phenolic glycoside salidroside, and the phenylpropanoids that are naturally occurring phenolics composed of an aromatic ring and

C3 side-chain (Figure 1.2) (Harborne 1998, Kurkin 2003). The simple phenylpropanoids rosavin, rosarin and rosin or the rosavins, are exclusive to R. rosea and partly responsible for its therapeutic benefits (Brown et al. 2002, Kurkin 2003).

Currently, Rhodiola rosea is in high demand on the natural product market

(Galambosi 2006). As a result, various studies are being conducted on the medicinal potential of wild populations, and on methods of introduction and cultivation of R. rosea in areas where it is not naturally occurring (Galambosi 2006, Kucinskaite et al. 2007).

Prior to this project, no phytochemical investigations of R. rosea populations had been performed in Canada. The present project was instigated through collaboration between University of Ottawa scientists and the Makivik development corporation, a society

16 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

which manages Nunavik Inuit heritage funds to promote economic growth for the region.

Makivik wished to explore the economic potential of Nunavik roseroot as a medicinal plant

and to help protect current wild populations of Nunavik R. rosea by raising awareness of the

plant's conservation value and by creating plantations of the native plants.

The present study evaluates for the first time the phytochemical profile of Canadian

R. rosea, particularly populations from the Nunavik region of Northern Quebec, Canada.

Because of their pharmacological relevance, concentrations of salidroside and rosavins are used to standardize root extracts (Russian National Pharmacopoeia 1989, Brown et al. 2002).

The secondary metabolite profiles of these four biologically active compounds were then used to compare variation among regionally and locally distinct populations.

2.2. PHYTOCHEMICAL ANALYSIS

2.2.1 Material and Methods

Sample preparation & extraction: Fresh samples of/?, rosea were collected in

Nunavik, QC, Canada, in August 2005 and 2006. At each collecting site, global positioning system (GPS) data were recorded and voucher specimens were gathered and donated to the

Marie-Victorin Herbarium (MT) (Montreal, QC, Canada). Field harvested rhizomes of wild

Nunavik plants were stored directly in 40 mL of 90% ethanol since facilities for drying were not locally available. Five individuals were collected for each population and stored separately in Nalgene bottles. Prior to extraction, samples were kept in their original solvent at 4°C. Wet samples were then filtered from the storage solvent, dried at ~35°C for 12 hours

(NESCO® Food dehydrator) and ground into a coarse powder. Cultivated rhizomes of

17 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

Siberian R. rosea populations, donated by Klickitat organics (WA, USA), were prepared

according to the same extraction method.

Dried samples were extracted using the ASE® 200 Accelerated Solvent Extractor

(Dionex, Sunnyvale, CA, USA). Each sample of varying weight (1.0-15.0 g) was extracted

twice with 33 mL 90% ethanol heated at 50°C for 15 min at a pressure of 2000 psi. The

ASE cell was then flushed at 100% ethanol and purged for 120 sec. The ASE liquid extract

was pooled with the previously filtered storage solvent and evaporated to dryness using the

Automatic Environmental Speedvac® System (AES 2010 from Savant) at ambient

temperature on full vacuum.

HPLC-DAD analyses: High performance liquid chromatography (HPLC) analyses

were performed on 1100 series HPLC system (Agilent Technologies Santa Clara, CA, USA)

using Chem Station LC 3D software (Rev. A09.01). The HPLC system consisted of an auto

sampler, quaternary pump, degasser, column thermostat (offline) and diode-array detector.

The HPLC analytical method was adapted from Tolonen et ah (2003) and used for all

samples (rhizome, leaf, stem and seeds). Each botanical extract was filtered through a PTFE membrane of 0.2 urn (Chromatographic specialties Inc., Brockville, ON, Canada) and injected, at a volume of 5 uL, through an auto sampler into Luna C18 column (150 x 4.60 mm, particle size 5 um) from Phenomenex (Torrance, CA, USA). The HPLC column was connected to a C18 guard pre-column (SecurityGuard™ from Phenomenex). Column temperature was maintained at ~35°C. The elution conditions of the mobile phase are presented through solvent gradient change and flow rate (mL/ min) over time (minutes)

(Table 2.1). The total analysis time for each sample injection was 30 min, with 5 min of post-run time. Analyses were recorded at detection signals of 210 nm for salidroside and

18 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

Table 2.1. Newly developed HPLC analytical method. Timetable of solvent gradient according to time (minutes) and flow rate (mL/ min). Time % Solvent Flow rate (min) Water (A) Acetonitrile (B) Methanol ( C) (mL/ min) 0 90 5 5 1.0 8 76 12 12 1.0 10 64 18 18 0.8 20 60 20 20 0.8 25 20 50 30 0.8 30 90 5 5 1.0

19 Chapter 2: Phytochemical Analysis of Rhodiola rosea L.

254 nm for the rosavins. Each botanical extract was analyzed three times. Standard extracts

for Rhodiola rosea were obtained from ChromaDex™ (Irvine, CA, USA).

Statistical analysis: Results were expressed as the mean of phytochemical

concentration for both salidroside and rosavins in mg per gram of dried plant material (mg/

g) ± standard error of mean (SEM). Data dispersion was evaluated and to re-establish

normality, data were logio-transformed for further analysis. A one-way analysis of variance

(ANOVA) was then applied, dip < 0.0001, for 1) inter-regional variations of circumpolar

collections, 2) inter-population variations of Nunavik collection set 1 and 2, and 3) variation caused by plant gender. The ANOVA was then followed by the Tukey multiple comparison test, dXp < 0.05. The phytochemical concentration of combined Nunavik populations, from collection set 1 and 2, was further investigated by an ANOVA, which tested the effects of year and location of harvest on salidroside and rosavins (p < 0.0001). Variation of phytochemical concentration according to plant health was determined using an unpaired t test with Welch's correction, withp < 0.05. Also, the phytochemical concentration, of plants from all Nunavik locations, was analyzed by a multivariate clustering method using the

Ward algorithm based upon the Euclidean distance.

To evaluate the relationship between secondary metabolite concentration of salidroside and rosavins of circumpolar collections, a Pearson correlation (r), wimp < 0.05, was employed. The Pearson correlation (r) was also used to calculate the strength of relationship between bioactive compounds and latitude, and the impact of soil nutrient concentration on secondary metabolite production. Statistical tests of nested ANOVA and cluster analysis were conducted on JMP software from SAS, version 7 for Windows. All

20 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

other statistical tests were performed using GraphPad Prism software version 5.01 for

Windows.

2.2.2 Results

Inter-regional variation ofR. rosea from circumpolar collections

Salidroside was higher in concentration in the three European populations as opposed to Canadian plants (Figure 2.1). The cultivated Siberian population was significantly different from all others. The highest quantity of salidroside present in a wild European population was found in the Norway collection set 1 (3.93 mg/ g) and, for Canada, in the subarctic environment of the Mingan Islands (1.23 mg/ g) while the lowest was New

Brunswick at 0.59 mg/ g.

The population containing the highest amount of rosavins was also the cultivated

Siberian population (4.80 mg/ g). Again, the wild population containing the highest amount of rosavins for Europe was Norway collection set 1 (3.04 mg /g) and for Canada, the Mingan population (2.64 mg/ g) but Nunavik collection set 2 showed the lowest amount with

0.55mg/ g. The Mingan population showed highest concentration for both salidroside and rosavins out of all Canadian populations.

Production of salidroside relative to the rosavins: Two populations out of eight, the Siberian population and Norway collection set 2, showed no significant correlation between salidroside and rosavins production whereas the other populations presented a significant positive correlation (Figure 2.2). Positive correlations were observed for populations of Nunavik collection set 1 (r=0.489) and 2 (r=0.738), Mingan (r=0.574),

21 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

B 15

D) 6- 1*10 I o 1 ab o) 4- b c o ,. b X •> JL c c be be o 2- nnnn pj^l on ^<^v *.*• *? • ^J<»" *-y *> Circumpolar collections

Figure 2.1. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM for eight regionally distinct circumpolar collections of Eurasia and Canada.

22 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

40- 8- I O) 3 • - 30- - 6- OU l )

I"> in • •g 20- > •s* '(/O> 05 L. 2H • • •o „ 2 10- • ••=»- 2- ra (0 (A • • w i • • • • ' *. 0- 1 1 0- 1 r 1 1 5 10 15 20 2 4 6 12 3 4 rosavins (mg/ g) rosavins (mg/ g) rosavins (mg/ g)

Siberian populations Norway collection set 1 Norway collection set 2

15-

•* i-10- •S4H 2 5H ••• tof?r»- ik T r- 1 2 3 2 4 6 rosavins (mg/ g) rosavins (mg/ g)

Nunavik collection set 1 Nunavik collection set 2

8- • '" »J — 6- #• Ol 1.0- E t i —° ^A • t ..» •O 2 0 0.5H !* • • = 1 V) r. 0- j i • •— i i i 1 0.0- A 5 10 15 20 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 rosavins (mg/ g) rosavins (mg/ g) rosavins (mg/ g) Nova-Scotia populations Mingan populations New-Brunswick populations

Figure 2.2. Concentration (mg/ g) of salidroside and rosavins present in eight circumpolar collections of Eurasia and Canada.

23 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

Norway collection set 1 (^=0.856) as well as New Brunswick (r=0.522) and Nova-Scotia

0=0.768).

Inter-population variation ofNunavik Rhodiola rosea

Nunavik collection set 1 and 2 represent different populations collected at different

locations around Ungava Bay during different years. Letters are used as labels to indicate the different populations; single letters (e.g. A) are used for collection year 1, 2005, whereas double letters (e.g. AA) identify populations collected in year 2, 2006. The sequence of labelling represents the collection order of the populations sampled within a year, independently of geography and phytochemistry. Populations from 2005 were not resampled the following year.

Nunavik collection set 1: Salidroside and rosavins concentration varied significantly by location (Figure 2.3). Salidroside content was most important in population L (2.51 mg/ g) and slowly decreased to 0.09 mg/ g in population F. Rosavins were highest in population

E (1.25 mg/ g) and decreased to 0.26 mg/ g in population J.

Nunavik collection set 2: As for collection set 1, concentration of salidroside and rosavins varied significantly by location (Figure 2.4). Salidroside was highest in population

AA (3.42 mg/ g) and decreased to 0.10 mg/ g in population MM. Rosavins were highest in population II (1.50 mg/ g) and decreased to 0.15 mg/ g in population LL.

The effect of year was further tested on combined Nunavik collection set 1 and 2.

Results showed no significant effect of year among the populations (n=31) for either salidroside or rosavins (p< 0.0001). However, results showed significant differences among all Nunavik populations for both their concentration in salidroside and the rosavins. 24 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

B

4.0n 1.5n ab ab a ab O) ab 3.0H ab "3) "5)1.0- ri abc I abc E i« J, be X 2 2.0H abc I (A X be abc J, abc be X w 0.5- be X bbe be J. be X |i.0H (fl c c be o

0.0 " i " i' T* 0 0.0- ffl 5fl s V O <5 <^ <^

Populations from Nunavik collection set 1

Figure 2.3. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM for different populations ofRhodiola rosea collected in August 2005 during the Nunavik collection set 1 expedition.

25 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

B 5.0-1 2.5n ab ab O 4.0- O) 2.0H "3) E, 1.5-1 «§« 3.0H ab abc T abc abc I 1.0H^ abc be g 2.0H X abc in i be be T be be . T be be . be . TOo 1.0- T bDeC _ DbCe w i 2 0.5 be i-ii.—nriri 0.0- O^AD 0.0- •••••••••• i • • i • • i • Q• i ,..,..,..,.., tt ,..,..,..,..,..,! mm rinnnmn

Populations from Nunavik collection set 2

Figure 2.4. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM for different populations ofRhodiola rosea collected in August 2006 during the Nunavik collection set 2 expedition.

26 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

Finally, a hierarchical cluster analysis using the Ward method was performed on the

pooled Nunavik collection sets. The resulting dendrogram showed the grouping of

phytochemically similar populations, according to their concentration in both salidroside and

rosavins (Figure 2.5).

Variation according to latitude

The production of both salidroside and rosavins was not significantly affected by increasing latitudes (Figure 2.6): salidroside (p= 0.122, r= -0.378) and rosavins (p- 0.787, r=

-0.068). Although not significant, Figure 2.6 shows a more pronounced decrease in salidroside concentration than the rosavins.

Soil nutrient impact

Phytochemical concentration of four Nunavik populations was observed in relation to soil nutrient availability (Table 2.2). Four soil samples were collected in Nunavik and analyzed by AgriDirect (Longueil, QC). Results of the Pearson correlation showed no significant effect of soil nitrate (NO3"), phosphorus ion (P3~) or potassium ion (K+) on the production of either salidroside or the rosavins.

Herbivory impact

To evaluate the impact of mite infestation on bioactive compounds, healthy (n=94) and deformed (n=60) plant categories were created independently of location, and compared for salidroside and rosavins concentration. An unpaired t test was performed with Welch's correction. Results showed significantly greater salidroside concentration in healthy plants

27 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

A — DD— HH J K RR F 3 EE— GG NN 00D" - PP QQ- LL" MM_ B " C " G - E - H _ D - JJ - FF"

CC" L " M " KK" AA" BB"

Figure 2.5. Dendrogram resulting from hierarchical cluster analysis of Ward using both salidroside and rosavins. The tree illustrates similarities among different Nunavik populations according to their concentration in the target compounds, independently of year.

28 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

T> 4-i z.u-

o5 1.5- • "5) 3H E E, !S 2 IT i.o- • '35 c o > • i_ re 1 1- 8 0.5- • re • • •••••• (A • • • •• _fi 1_ 0.0- 58°19'35.6" 59°29'26.0" 61°04'41.4" 58°19'35.6" 59°29'26.0" 61°04'41.4"

Latitude (hddd°mm'ss.s") Latitude (hddd°mm'ss.s") Figure 2.6. Concentration (mg/ g) of salidroside (A) and rosavins (B) from R. rosea of Nunavik collection set 2, according to increasing latitude, from south to north (58° 19'35.6" to61°04'41.4").

29 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

Table 2.2. Mean phytochemical concentration of salidroside and rosavins (mg/ g) from four Nunavik soil analyses in relation to three essential soil nutrients: nitrate, phosphorus and potassium.

Nitrateb Salidroside Rosavins (ppm) (mg/g ± SEMa) (mg/g± SEM)

0.50 0.096 ±0.013 0.182 ±0.025 0.50 0.473 ± 0.108 0.406 ± 0.072 1.90 0.417 ±0.102 0.263 ±0.057 5.78 0.497 ±0.183 0.409 ±0.135

Phosphorus (kg/ha)

19 0.497 ±0.183 0.409 ±0.135 73 0.473 ±0.108 0.406 ±0.072 88 0.096 ±0.013 0.182 ±0.025 186 0.417 ±0.102 0.263 ±0.057

Potassium (kg/ha)

193 0.497 ±0.183 0.409 ±0.135 219 0.096 ±0.013 0.182 ±0.025 364 0.473 ±0.108 0.406 ±0.072 503 0.417 ±0.102 0.263 ± 0.057 a SEM, standard error of mean b Soil samples analyzed according to the Mehlich 3 method, by AgriDirect

30 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

than in deformed plants (Figure 2.7). No significant difference was observed for the

rosavins.

Effect of plant gender

For Nunavik collection set 2, plant gender was recorded, and plants were categorized

into four groups: hermaphrodite (n=7), male (n=33), immature (when no reproductive units were visible, n=6) and female (with no distinction between flowering phase and seeding period, n=44). Results showed that hermaphrodites and males had significantly higher salidroside concentration than female plants (Figure 2.8 A). Salidroside concentration of immature plants was intermediate and not significantly different from other categories. As for the rosavins, their concentration did not differ significantly among gender (Figure 2.8 B).

Plant height of female and male individuals was also recorded. Heights were compared using an unpaired t test with Welch's correction and female R. rosea were found to be significantly taller (49.5 cm) than male plants (23.8 cm).

2.2.3 Discussion

General discussion: Significant phytochemical variation was observed among regionally and locally distinct populations. Differences in secondary metabolite production could be explained by the effect of both environmental conditions and genetics (Almeida-

Cortez et al. 1999, Kucinskaite et al. 2007).

R. rosea grows in the harsh environment of cool subalpine and arctic dry tundra

(Clausen 1975, Brown et al. 2002, Galambosi 2006) where stressful biotic and abiotic components limit plant growth (Scott 1995). Different theories, such as the carbon: nutrient

31 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

B 2.0-, 0.8n

» 0.6H

T • • p Healthy plants Deformed plants Healthy plants Deformed plants

Figure 2.7. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM in healthy and gall-deformed Nunavik R. rosea plants, from Nunavik collection set 2. *The asterisk indicates significant difference.

32 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

B 2.5n 1.

Figure 2.8. Mean concentration (mg/ g) of salidroside (A) and rosavins (B) with SEM for different Nunavik R. rosea plants, from Nunavik collection set 2, according to gender (hermaphrodite, male, immature and female).

33 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

balance hypothesis and the growth differentiation balance hypothesis, describe the plant's

response to important limitations of environmental factors such as light, water availability

and nutrients, e.g. carbon and nitrogen, through its allocation of resources (Herms and

Mattson 1992, Lerdau and Gershenzon 1997). A trade-off occurs in the plant which

allocates its resources between three important functions: growth, reproduction and plant

defence (Bazzaz 1997). So, the concentration of secondary metabolites in plant tissue is

influenced ultimately by the amount of resources available in the environment (Grace 1997,

Graglia et al. 2001).

The target secondary metabolites investigated by this study are synthesized through the phenylpropanoid metabolism, which produces chemicals involved in functions such as defence against herbivory, UV radiation and protection against other environmental stresses

(Dixon and Paiva 1995, Grace 1997). To date, the role of salidroside and rosavins ini?. rosea functions is unknown, even though salidroside is thought to be involved in birch plant defence and indirectly in seedling growth (Dixon and Paiva 1995, Keski-Saari et al. 2007).

As for the influence of genetics, a systematics study ofRhodiola rosea revealed that the Canadian plant species were genetically similar to the Eurasian ones (Archambault et al.

2008). The Canadian plant population, however, could be a distinct chemical phenotype as the HPLC profile shows indications of distinct compounds not present in the Eurasian profile. These populations distinguish themselves as an intraspecific taxon due to their novel phytochemical composition. Also, phenotypic plasticity, the plant's physiological response to environmental changes, is constrained by genetics (Grime 1979, de Jong and van

Noordwijk, 1992, Boggs 1997). However, phenotypic or chemotypic variations can only be

34 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

assessed by comparing the physiology and phytochemical levels in plants growing under

similar environmental conditions, e.g. in a common garden or growth chambers.

Inter-regional variation ofR. rosea from circumpolar collections: The European

populations had the highest concentration of salidroside and rosavins although the Mingan plants from Canada revealed an equally large amount of rosavins. From the European populations, the Siberian cultivated plants were found to produce significantly higher amounts of bioactive constituents than any other collections. This is not surprising since the

Siberian plants are grown under cultivation and for medicinal purposes (Galambosi 2006).

These plants are therefore selected to produce higher quantities of compounds, as opposed to the other seven wild growing populations.

In regards to the wild Canadian populations investigated in our study, plants from the

Mingan Islands showed the highest phytochemical concentration for all compounds. This abundance in secondary metabolites could be linked to a more favourable combination of various environmental factors, e.g. temperature and edaphic conditions or genetic factors leading to higher biosynthesis of the compounds. The Mingan plants show the best phytochemical potential as an herbal crop and should be studied further.

An investigation by Hohtola et al. (2005) of Finnish and Polish R. rosea grown at a single site, suggested that the concentration of bioactive compounds changed significantly according to the site of origin. In their study, phytochemicals were highest in populations originating from northern locations as opposed to southern ones. This demonstrated the effect of genetics on phytochemical content. Alternatively Kucinskaite et al. (2007) demonstrated the impact of environmental conditions on plant phytochemistry. Their study showed that introduced and cultivated Siberian R. rosea plants in Lithuania produced

35 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

significantly higher concentration of rosavins (73%) and salidroside (1.5%), than in Siberia

growing plants.

Production of salidroside relative to the rosavins: Since the medicinal activity of R.

rosea is attributed to the synergistic activity of salidroside and the rosavins (Russian National

Pharmacopoeia 1989), their variation within the rhizome is relevant to plant cultivation. In

six out of eight populations, there was a positive correlation between the amount of

salidroside and the rosavins. Although the exact function of these compounds within the plant is unknown, these metabolites are known to be produced through the phenylpropanoid metabolism. Salidroside and rosavins could therefore be similarly impacted by the availability of precursor compounds as well as limited by the same environmental stresses

(Dixon and Paiva 1995, Grace 1997, Brown et al 2002).

Inter-population variation ofNunavik Rhodiola rosea: The four key pharmacologically active compounds were confirmed and quantified in the Nunavik populations. This authenticates the medicinal potential ofNunavik growing plants. The significant variation of these compounds among different populations could be explained by different chemotypic and environmental variations, which can only be fully characterized through future investigations.

The concentrations of compounds for the Nunavik populations were not significantly affected by the year of collection although location did significantly impact phytochemical concentration. As observed in the dendrogram, the cluster arrangements did not depend on the year of sampling since populations from year 1 and 2 were not joined nor did the grouping depend on the village of harvest or latitudinal factors (Figure 2.5). Indeed, clusters do not show any correlation with geography and seem to be formed arbitrarily.

36 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

For instance, populations 00 and PP were clustered together. Although these were

sampled during the same collection year, population 00 was harvested in Quaqtaq, at the

most northern latitude, while population PP was sampled in Kuujjuaq, in the south. A

detailed examination of the populations at their ecological level is needed to further

understand the link between each cluster. Differences in salidroside and rosavins

concentration for Nunavik collection set 1 and 2 were further examined in relation to latitude

(for both compounds independently), soil nutrient, herbivory and gender variation.

Variation according to latitude: Significant results would have indicated that an

increase in latitude, leading to a decrease in temperature and variation in biotic and abiotic

conditions, negatively affects plant secondary metabolite production (Johnson and Scriber,

1994, Willig et al. 2003, Alonso et al. 2005, Shevtsova et al. 2005, Grundt et al. 2006).

According to a study by Nerg et al. (1994), total phenolic concentration of Scots pine

seedlings (Pinus sylvestris) was significantly decreased in the most northern growing

location. Both R. rosea and P. sylvestris are perennial xerophytic plants which are found in

the cold and nutrient-deficient environment of Finland. Findings for P. sylvestris further

suggested that secondary metabolite production was significantly affected by latitudinal

factors as opposed to the origin of seed material, thus the genotype. More specifically,

phenolic concentration of P. sylvestris was reduced in the most northern site but was highest

in the middle location.

Soil nutrient impact'. In alpine tundra communities, plant productivity is limited by

environmental conditions, especially nutrient-poor conditions, which impact directly plant growth and phytochemistry (Scott 1995, Shevtsova et al. 2005). Even though our findings showed no association between phytochemical production and nutrient levels, nutrient

37 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

conditions are known to be a limiting factor for the arctic ecosystem, especially nitrogen

(Scott 1995). A more detailed examination of the impact of soil nutrients on phytochemicals

is needed. In this study, the soil samples collected were too few in number to establish

convincing results, and these were stored in the field in less than favourable conditions,

which could have altered soil chemistry thus affecting the results.

Graglia et al. (2001) demonstrated that salidroside, in Betula nana, varied significantly between two geographically separated sites; the pronounced regional differences were attributed to environmental conditions, specifically to soil acidity. Since our results were inconclusive, thorough investigations are needed to study the relationship between soil nutrient richness and secondary metabolites for Nunavik R. rosea.

As proposed in the carbon-nutrient balance (CNB) hypothesis, secondary metabolite concentration in plant tissue is linked to resource availability (Bryant et al 1983, 1988,

Hamilton et al. 2001, Walls et al. 2005). Under nutrient deficient conditions, plants accumulate carbohydrates, which lead to an increase in carbon/nitrogen ratio (Herms and

Mattson 1992), and consequently in an increase of carbon-based compounds, e.g. salidroside and rosavins. When plants grow on nitrogen deficient soils, the excess carbon, which is not used for primary growth due to limited nitrogen, is converted into secondary metabolites.

Therefore, the addition of nitrogen fertilizers should promote nitrogen-based secondary compounds and limit the production of carbon compounds (Herms and Mattson 1992,

Hamilton et al. 2001, Palumbo et al. 2007). Galambosi (2006) however, showed ini?. rosea plantation experiments, that with the application of nutrient enhanced compost, plant biomass as well as concentration of active metabolites increased. The most effective fertilization level, the one which produced the highest concentration of bioactive chemicals

38 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

was measured as N-P-K: 50-50-75 kg/ ha. In short, the addition of nitrogen negatively

impacts certain phenols but not those investigated by this study (Keinanen et al. 1999,

Anttonen and Karjalainen 2005, Mudau et al. 2006).

While soil chemistry can be similar from one habitat to the next, soil moisture can

alter the plant's phytochemical concentration since water is required for nutrient uptake

(Lambers et al. 1998, Salmore and Hunter 2001); so nutrient availability is subject to weather

conditions.

Herbivory impact: Seed deformation was observed in R. rosea during field collecting

and was found to be caused by the presence of bud mites. In plants stressed by gall deformations, salidroside concentration in the rhizome was significantly reduced whereas rosavins remained unaffected. Secondary metabolites have an important role in plant defences such as herbivore deterrents, protection against light, e.g. UV radiation, and oxidative stress (Dixon and Paiva 1995, Baldwin and Preston 1999, Palumbo et al. 2007).

However, the decrease in salidroside amount possibly indicates an indirect role in chemical defence. Salidroside could act as a precursor for induced phenolic compounds involved in plant defence (Linh et al. 2000, Palumbo et al. 2007). A simpler explanation suggests that when these arctic plants, which are already stressed by environmental conditions, are further pressured; they are less able to produce phytochemical defences. The limited resources available are then allocated in priority to vegetative growth as opposed to chemical defences

(Bazzaz 1997).

Effect of plant gender: Evidence from the phytochemical comparison indicated significant differences between genders for the production of salidroside but not for the rosavins. Phytochemical variation due to gender has also been observed in various plant

39 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

species (Alonso et al. 2005). According to the CNB theory, female plants allocate more

nutrients to reproductive functions, e.g. seed maturation, than male plants (Herms and

Mattson 1992, Palumbo et al. 2007). Consequently, female plants would have slower growth

rates and higher concentrations of carbon-based compounds (Jing and Coley 1990, Herms

and Mattson 1992).

However, results from R. rosea rhizomes showed hermaphrodites and males as

having the highest concentration in salidroside, which suggest a different strategy in resource

allocation or an effect of season on gender resource allocation (Alonso et al. 2005, Palumbo et al. 2007). Male plants could invest a higher portion of resources to reproduction earlier in the season and at the end of the flowering period, allocate more resources to secondary metabolites. Alternatively, female plants may allocate more resources to growth or secondary metabolites earlier in the season and at the end of the summer, invest their resources in reproduction (Alonso et al. 2005).

A study by Massei et al. (2006) on juniper trees, showed trade-offs caused by gender.

A few other cases revealed similar findings; female plants invest less resource in secondary metabolite production and have a similar or superior growth rate compared to male plants

(Danell et al. 1985, Sakai and Burris 1985, Amende and Harper 1989, Fritz 1995, Massei et al. 2006).

To date, no study has investigated the growth rate R. rosea plants. Although, measurements from this project showed that female plants were significantly taller than males. This difference was also observed in Arisaema triphylla (Jack-in-the-Pulpit)

(Korpelainen 1998). Sexuality is labile in those plants and may also play a role in the expression of sexuality in R. rosea plants from year to year.

40 Chapter 2: Phytochemical Analysis ofRhodiola rosea L.

Variation of the bioactive compound salidroside: Following this study, variations in

salidroside concentration were significantly influenced by environmental conditions, such as

infestation and gender whereas amounts of rosavins were relatively stable regardless of the variations. To some extent, the role of salidroside in plant functions can be assessed. The abundance of this bioactive compound was found to decrease with increasing latitude consequently, with increasing UV radiation (Alonso et al. 2005). Salidroside is therefore not playing a central role in photoinhibition (Nybakken et al. 2004). It was also found to decrease with added herbivory pressure, and with plant allocation to reproduction.

Considering the amount of salidroside did not increase in response to plant infestation nor was it related to gender according to the CNB allocation theory, salidroside may be involved in functions other then plant defences (Herms and Mattson 1992, Massei et al.

2006). Salidroside may also be costly to produce, so under added environmental pressures or against primary allocation demands, e.g. reproduction, compound production could be significantly decreased. These observations should be taken into consideration when introducing and cultivating Nunavik R. rosea plants in new regions. Further examinations would elucidate the role of salidroside and subsequently, the rosavins.

41 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

CHAPTER 3

Ecological Analysis of Nunavik Rhodiola rosea L.

42 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

3.1. INTRODUCTION

The circumpolar arctic plant Rhodiola rosea L. (Crassulaceae) is a highly valued

medicinal plant species in Eurasia (Brown et al. 2002). Records indicate that traditional

usage of this plant dates back centuries (Brown et al. 2002, Aim 2004). R. rosea, also known

as "roseroot", "arctic root" and "golden root", is now a commercial phytomedicine employed

as an immunostimulant, to protect against oxidative stress and as an antidepressant. Russian

scientists have also classified this medicinal plant as an "adaptogen". According to the

Russian definition, an "adaptogen" is a plant which re-establishes and maintains the body's equilibrium (Brown et al. 2002).

The positive effects of roseroot extracts on the body contribute to its increasing popularity, which in turn lead to a high market demand (Galambosi 2006). The increasing market demand for raw plant material has resulted in the depletion of natural Siberian populations to the extent where R. rosea is now red listed in Russia, with heavy restrictions on wild harvesting. Also, in other eastern European countries roseroot is catalogued as endangered (Galambosi 2006). Alternative sources of R. rosea have then been investigated, and the introduction and cultivation of the plant in controlled settings has been implemented

(Kuncinskaite et al. 2007).

Rhodiola rosea has a nearly circumpolar distribution and its natural range includes

China, the Russian Federation, Scandinavia, Greenland, Iceland and the eastern Arctic of

Canada (Aiken et al. 1999, Small and Catling 1999, Brown et al. 2002). Roseroot is generally found in arctic, subarctic and alpine regions where climate is cold to cool (Aiken et al. 1999, Small and Catling 1999). Habitat characteristics are varied, and plants can be found growing on a wide range of substrates, e.g. rocks, crevices, cliffs, gravel and till.

43 Chapter 3: Ecological Analysis of Nunavik Rhodiola rosea L.

Biologically, this stonecrop plant is a perennial dioecious herb known to reproduce both

sexually and asexually (Clausen 1975, Aiken et al. 1999, Small and Catling 1999).

Occurrences of the plant have been noted in Northern Quebec and comparisons with

herbarium records would suggest that this is one of the major growing areas in Canada

(Clausen 1975, Aiken et al. 1999, Brown et al. 2002), although its Canadian distribution and ecology have not been fully characterized. The present study proposed to examine the plant's phytochemistry in relation to its medicinal potential as well as geographic distribution and ecological requirements in the Nunavik region of Northern Quebec, Canada. A systematic study, conducted in collaboration with this research project, revealed that

Nunavik R. rosea is the same species as Eurasian R. rosea (Archambault et al. 2008). In addition, the phytochemical analysis (Chapter 2) confirmed that key phyto chemical markers of medicinally active Eurasian R. rosea populations were present in Nunavik plants.

In the present ecological study, the plant's geographic range and habitat type along the coast of Ungava Bay, Nunavik, were studied. Investigations led to the establishment of the plant's geographic distribution and patterns as well as associated ecological requirements. The "maximum entropy" modeling method was used, with presence-only data, to create predictive map of species geographic distribution for the coast of Ungava Bay

(Elith et al. 2006, Hernandez et al. 2006, Phillips et al. 2006). The field surveys also permitted data collecting in order to characterize the plant's habitat in the region.

Subsequently, to assist the local Inuit communities in developing economic enterprises for the region, and to avoid future overharvesting of wild populations, an experimental plantation using local plants was designed and realized in collaboration with the Inuit developmental society Makivik Corporation. Past and present surveys noted signs

44 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

of deformation on roseroot plants (Clausen 1975) therefore investigations were conducted to

find the cause of deformation.

3.2. MATERIALS AND METHODS

Geographical distribution of Rhodiola rosea

Rhodiola rosea's distribution was investigated in Nunavik, Northern QC, Canada.

The study area included sections of the shores of Ungava Bay from Killiniq to Quaqtaq. The coastline was surveyed by motorized canoe. Exploration sites were randomly chosen and each area was explored on foot and scouted for the plant. Global positioning system (GPS) points were recorded for the area at ± 5 m of accuracy. GPS coordinates were then converted to hddd.dddd0 and formatted in DBH4 files (format from Microsoft Office word) which were mapped with ArcMap from ArcGIS desktop v 9.2 (ESRIGIS and Mapping

Software, Redlands, CA, USA). The geographic coordinate system used is North American

Datum 1983 (NAD 83). Shapefiles of the Nunavik area were obtained from the GSG

(Geographic, Statistical and Government information centre, University of Ottawa library,

Ottawa, ON). Map scale ofNunavik was of 1: 250 000.

Three hundred thirty-nine coordinates were used and transformed, as stated above, to create the world distribution map. The compiled herbarium records were gathered from the

Herbarium of Agriculture and Agri-Food Canada (DAO) (Ottawa, ON), the Canadian

Museum of Nature Herbarium (CAN) (Gatineau, QC), the Universite Laval Herbarium

(QFA) (Quebec, QC) and the Royal BC Museum Herbarium (RBCM) (Victoria, BC). The world map data were obtained from Blue Marble Geographies (Maine, USA).

45 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

Maxent distribution model

Methodology: Maximum Entropy (Maxent), a species distribution modeling

technique, was used to create an ecological niche model to predict spatial distribution of R.

rosea along Ungava Bay, based on the data set collected during fieldwork in Nunavik and

selected environmental variables (MaxEnt software version 2.3, NJ, USA,

http://www.cs.princeton.edu/~schapire/maxent/). The species distribution model was built

using only occurrence records, since Maxent uses presence points only. Ten replicates were

used to assess the average behavior of the algorithms although variation of suitability

between models is minor (Phillips et al. 2006). For each model, plant occurrence points

were randomly selected into 70% training data, used to build the model, and 30% testing

data, used to test the resulting model. The ten ensuing models were averaged to produce the

final species distribution model. All occurrences of R. rosea in Nunavik, a total of 92

localities, fall in coastal areas.

Model evaluation: Each model was evaluated for its accuracy using the area under

the curve (AUC) of the receiver operating characteristic. AUC has been widely used in

species distribution modeling literature to assess the ability of the model to distinguish

between presence and absence locations (Fielding and Bell 1997, Elith et al. 2006). Values

for AUC range from 0 to 1, a score of 0.5 indicates random prediction and 1 suggests perfect

discrimination (Fielding and Bell 1997). If the average AUC value of the final model falls

between 0.7 and 0.9, the model is regarded as useful and when it exceeds 0.9, it is considered

excellent (Swets 1988).

Data: Different environmental variables are thought to influence the plant's distribution in the low arctic of Quebec. First, to record GPS points (±5 m) the Ungava Bay

46 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

shoreline was randomly explored and populations of R. rosea were noted. Secondly, only

geographically unique records were used, e.i. a location was only counted once even if it had

been sampled several times. The model included six environmental variables: geology,

elevation, slope, aspect, global solar radiation and duration of radiation. Digital elevation

data were obtained from GeoBase Canada at a scale of 1: 250 000 with a spatial resolution of

1 km (http://www.geobase.ca/geobase/en/find.do?produit=cdedl). Slope and aspect

(orientation to the cardinal point) were calculated from the digital elevation data using

Arc/Info Grid (ESRI2005), at a resolution of 1 km. Global solar radiation and duration of

radiation were calculated using Solar Analyst, an Arc View extension, also at a spatial

resolution of 1 km. Geological data were obtained from the Geological Map of Canada for

Northern Quebec and resampled to a resolution of 150 m.

Habitat characterization in Nunavik

Samples of R. rosea were collected in Nunavik (Quebec, Canada) in August 2005 and

2006. During the exploration of Ungava Bay coastline, R. rosea populations were identified

and random samples of rhizomes (in some cases leaf, stem and seeds) were collected.

Rhizomes were stored in 90% ethanol for laboratory phytochemical analysis at the

University of Ottawa (Ottawa, ON). Voucher specimens were also gathered and stored at the

Marie-Victorin Herbarium (MT). Environmental observations such as topography,

elevation, soil type, and proximity to water as well as the presence of deformation, slope,

aspect, density and associated vegetation were documented. For each rhizome collected in

2006, length of the stem and rhizome, and plant sex were noted. Finally, four soil samples were taken at different environmental locations and analyzed by AgriDirect (Longueil, QC).

47 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

Investigation for plant mite infestation

Samples of deformed plants were collected in Nunavik (Quebec, Canada) in August

2007. Coastline of Ungava Bay was surveyed to find existing populations. Once a

population was identified flower heads showing signs of deformation were collected. GPS

points (± 5 m) were recorded for the area. Plant deformation was observed solely on fruiting

inflorescences of R. rosea. Deformed plant heads were then collected and stored in 60%

ethanol for future use at the University of Ottawa (Ottawa, ON). Once in laboratory setting, deformed flower buds and seeds were observed under a compound microscope at 40X. The gall-causing agent was measured and photographed using an image pro express digital camera. Samples, in triplicate, were also sent to the National Identification Service of

Agriculture and Agri-Food Canada (Ottawa, ON) for species identification.

Experimental plantation in Nunavik

An experimental plantation was established in the Nunavik community of

Kangiqsualujjuaq in 2006. This particular village was chosen because of the low risk of vandalism and suitable growing sites. First, a plantation site was selected on the mound located on the opposite bank of the village where R. rosea is naturally occurring and abundant. The seedlings planted originated from seeds of three distinct wild populations of

R. rosea collected the previous summer, in 2005, in Nunavik. Seeds had been collected, dried, induced to germination and grown in the Jardin botanique de Montreal (QC). The experimental plantation was established along the principles of "natural farming" with no soil disturbance. Approximately 174 seedlings, each measuring 5 to 8 cm, were planted in moss covered soil. Seedlings were planted within a non-disturbed natural ecosystem, 12 to

48 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

20 cm apart in 19 rows, each measuring approximately 5 m. Rows were placed 160 cm

apart. The plantation area covered approximately 5.82 m by 3.68 m, a total of 21.42 m2.

3.3. RESULTS AND DISCUSSION

Geographical distribution of Rhodiola rosea

Three distribution maps of R. rosea were created from recorded GPS points. Map 3.1 illustrates the location of the study site within Canada. Map 3.2 shows the distribution of R. rosea in Nunavik, Quebec, map 3.3 displays sites where occurrences of plant deformation were observed in Nunavik, and maps 3.4 and 3.5 represent sites where herbarium specimens have been collected.

1. Rhodiola rosea distribution in Nunavik: Nunavik territory covered by maps 3.2 and 3.3 is part of the low arctic ecoclimatic region of Quebec, Canada (Scott 1995). The area explored for R. rosea extends from the east coast of Ungava Bay, from Kangiqsualujjuaq to Killiniq, to Kuujjuaq in the south and to the west coast, Kangirsuk to Quaqtaq. The distribution map illustrates both the presence and absence of the plant (Map 3.2). A total of ninety-one points recorded the presence of the plant whereas thirty-three points showed its absence. According to these, R. rosea is commonly found along the east coast (n=61) and the south (n=25) as opposed to the west coast (n=5).

East coast of Ungava Bay

The random exploration of the east coast of Ungava Bay, within a range of 40 km

49 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

wwmm

1,300 Kilometers

Map 3.1. Geographic representation of the study area of R. rosea in Nunavik, Quebec, Canada.

50 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

e Quaqtaq r

•

Kangirsuk

"> ^ : •' -V : Jt*V f( f>- ^-"•,,* ?», i/ , ^.S..,,,„,:..te|' . kuujj'uaq'fff. Legend :» 20,500,41,000' 82,000 123,000 164,000 — ^———Mi III •• o Absence of ptant • Presence of pfarrf

Map 3.2. Dispersion of Rhodiola rosea according to presence and absence points along Ungava Bay, Nunavik, QC, Canada.

51 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

from the village of Kangiqsualujjuaq, produced 36 sites, which indicated the presence of R.

rosea from these, 9 populations were collected, and 4 additional sites showed the absence of

the plant. During the three-day expedition to Killiniq, 16 sites showed the presence of the plant from these, 10 populations were collected. At no site was the plant absent; this

demonstrates it's abundance on the east coast. This expedition was valuable in providing up- to-date records of R. rosea for the area since the last herbarium records of Killiniq region dated back 70 years (1936 and 1941-by Dutilly, Universite Laval Herbarium-QFA).

Southern region ofNunavik and village of Kuuiiuaq

In Kuujjuaq, along the Koksoak River, 25 sites indicated occurrences of R. rosea whereas in 9 locations the plant was absent. From the 25 sites, 10 populations were randomly collected. As the investigation carried on downstream, a distinctive change in topography was accompanied by a gradual reduction in R. rosea occurrences. Also, as the river flowed downstream, salt water from the Ungava Bay changed to fresh water, which decreased growing capacity for halophytes thus R. rosea (Waisel 1972). In order to confirm the relationship between a decrease in plant abundance and salt water concentration, shorelines of various freshwater lakes were explored (e.g. Stewart Lake near Kuujjuaq). No

R. rosea was found along lake perimeters.

West Coast of Ungava Bay

A survey from Kangirsuk to Quaqtaq revealed no occurrences of R. rosea for the 6 sites explored, hi Quaqtaq, herbarium records indicated the presence of R. rosea along the shorelines of Diana Bay and Diana Islands (1986-by botanist Marcel Blondeau, Universite

52 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

Laval Herbarium-QFA). These were explored for R. rosea however the 13 sites examined

showed no sign of the plant. R. rosea was also scarce around the village; in only 3 sites was

the plant found. The rarity of the plant on the west coast strongly contrasts with its high

abundance on the east coast.

Plant distribution along Ungava Bay

A strong contrast exists between abundance of R. rosea on the east coast versus the west coast. According to the niche theory, plant distribution is influenced by environmental factors such as climate and edaphic conditions (Chust et al. 2006). Since the Nunavik plants investigated are part of the same ecoclimatic zone, the low arctic, and tundra ecosystem

(Scott 1995); edaphic conditions were examined. Topography and geology could play an important role in the plant's distribution. Unfortunately, detailed and accurate soil data are frequently lacking in broad ecological surveys and prediction models (Coudun et al. 2006).

Canadian soil classification system (1998) divides Ungava Bay into two distinct soil regions. The west coast region, including Kangirsuk and Quaqtaq, belongs to the mineral cryosolic soil region, whereas the east coast, including Kuujjuaq, Kangiqsualujjuaq and

Killiniq, is part of the podzolic soil region. Cryosolic soils predominate in northern Canada and are characterized by a permafrost layer, within 1 to 2 meters of the surface. Podzolic soils are sandy or coarse loams commonly found under coniferous or boreal forests. These are poor soils with a poorly decomposed organic matter layer. This difference in soil characteristics between the east and west coast could possibly explain the difference in R. rosea abundance but a detailed examination of the soil groups is needed along the coast. It is

53 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

also important to mention that not all Ungava Bay coastline was explored; future

investigations are needed to support the impact of geology on the plant's distribution.

Other factors, such as species dispersal, glaciation and historical contingency could

have influenced species' distribution. Rhodiola rosea reproduces both sexually, through

seeds, and asexually through vegetative propagation of its rhizome. Seeds are very small and

winged which suggests wind dispersion (Clausen 1975). Also, R. rosea is a coastal plant and

its rhizome, which is easily fragmented, can be carried long distances by water, ice and

strong winds (Clausen 1975). General wind and water current patterns from the Arctic can

thus impact the plant's distribution in the region. Mean upper winds, for both summer and

winter seasons are known to flow north-west to north-east (Hudson et al. 2001). Since wind

stress impacts ocean circulation (Myers 2003), ocean currents and hence ice movements

follow the same pattern, from west to east (Hudson et al. 2001). This suggests, at a broad

scale, that R. rosea's propagation toward the west coast of Ungava Bay is not facilitated by

environmental factors.

Recent glacial events such as the Wisconsin glaciations strongly impacted plant

distribution. During the Pleistocene epoch, approximately 18,000 years ago, the North

American midcontinent was covered by the Laurentide ice sheet and the western part of

Canada was covered by the Cordilleran ice sheet (Clark et al. 1993). The Laurentide ice

sheet slowly receded and vanished approximately 5,000 to 3,800 years ago, opening the

region for plant colonization (Fagan, 2000, Dixon 2001). However, melting of the ice sheet

did not occur uniformly (Fagan 2000). The Ungava Bay territory and Quebec were

completely glaciated 18,000 years ago, and 14,000 years ago deglaciation had occurred in

New England and southern Quebec but Ungava Bay area was still under ice. Around 8,000

54 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

years ago, the ice sheet had receded, uncovering the Labrador coast and the east coast of

Ungava Bay. The east coast was thus available for plant colonization before the west coast

(Clark et al. 1993, Fagan 2000).

Rhodiola pollen records could establish the migration speed and trends of the species.

Unfortunately, very few pollen fossils have been reported since R. rosea occupies rocky sites

not conducive to preservation or research (Clausen 1975). A record taken by Wenner (1947)

off the coast of southern Labrador indicated the presence of Rhodiola shortly after post

glaciation (Clausen 1975). This sample indicated pollen traces of the plant after a series of

environmental postglacial changes, i.e. from a Salix to an ericaceous and Empetrum

dominated habitat. Still today, Empetrum is commonly associated with roseroot. This

vegetation succession suggests an important time lapse between deglaciation and the

appearance of Rhodiola in Nunavik. During the Wisconsin glaciations, plants could have

survived in the unglaciated Piedmont plateau of southern Pennsylvania, where Rhodiola

rosea is still present today (Clausen 1975). Although no pollen records were found in that

area, core samples suggested an environment suitable for plant survival (Martin 1958,

Clausen 1975). Plants could have then migrated north to Nunavik from New England and

Labrador coasts, which implies a recent migration of R. rosea.

Another theory proposes anthropological influences on the plant's distribution.

Rhodiola rosea is widespread in Norway and is an integral part of Norwegian folk tradition.

Records of the plant's usage date back as far as the Viking period (Brown et al. 2002, Aim

2004). Norsemen were known to use this medicinal plant before long journeys. It is therefore possible to assume that Vikings brought the plant on their expeditions. Viking expeditions to north eastern Canada started around 990 A.D, long after the Wisconsin

55 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

glaciations (Fagan 2000). From 1000 to 1013 A.D, they explored the coast of Labrador and

Newfoundland all the way down to New England (Fagan 2000) where they could have

introduced the plant. Today, R. rosea's distribution follows a similar path (Map 3.4).

Natural migration could have occurred after the plant's introduction. Migration of the

species to the west coast of Ungava Bay could ensue in the future or could be limited by a

combination of factors, including the edaphic conditions previously discussed.

These conjectures remain untested due to a lack of environmental data and the absence of pollen records. More information is needed on the occurrences of the plant and the genetic diversity of populations from the Nunavik region and other locations (e.g. New

England). However, evidence supports the fact that R. rosea is relatively new in North

America (Clausen 1975).

2. Deformed Rhodiola rosea distribution in Nunavik: At each site where R. rosea was recorded (map 3.2), plant health was noted. Distribution map 3.3 illustrates the dispersion of healthy versus deformed individuals. The cause of plant deformation was identified as being an herbivorous mite. Twenty-nine out of ninety-two populations (32%) presented signs of deformation. However, the distribution of deformed plants showed no significant geographical pattern. From the east coast, 16 populations were infected, in Kuujjuaq 12 populations, and in Quaqtaq 1 of 3 populations showed anomalies. Since these bud mites are species specific, they only occur where R. rosea is present. Wind acts as a dispersal agent for both R. rosea and bud mites (Blackman and Eastop 2006). The latter is thus subject to the same wind patterns as the seeds, which could assist it in finding its host.

56 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

'A*.*- r**tozg>

*V

•Jf •"•.

•;V.\

kK&£g*ssu&lu3luaq

\ \.l * * '-.1 „<*?*«• - * 0 20,600 41,000 B2.000 123,000 154,000 , Legend o Deformed pJanis • Non-deforrned plants

Map 3.3. Distribution of deformed and healthy Rhodiola rosea plants along Ungava Bay, Nunavik, QC, Canada.

57 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

3. Worldwide distribution of herbarium specimens: Rhodiola rosea L. has a circumpolar

distribution and is found in arctic environments of Europe, Russian Federation, Asia and

eastern North America (Aiken et al. 1999, Ganzera et al. 2001, Aim 2004). Two species of

Rhodiola are presented on herbarium distribution map 3.4 and 3.5: R. rosea and R.

integrifolia Raf. Map 3.4 illustrates the distribution of Rhodiola species in the northern

hemisphere while map 3.5 represents the distribution of both species in North America.

Most herbarium specimens were collected when these species were considered as

subspecies (Clausen 1975, Small and Catling 1999). Today, after careful ecological, geographical, physiological and caryological examinations, the plants are recognized as two distinct species (Hegi 1963, Gleason and Cronquist 1991, Brown et al. 2002). The most significant difference between them is the petal color: R. integrifolia has dark red flowers whereas R. rosea has greenish yellow petals (Small and Catling 1999). Differences can also be seen in the leaves, which are dark green and not glaucous for R. integrifolia but glaucous fori?, rosea (Figure 3.1) (Clausen 1975). R. integrifolia, although closely related to R. rosea, is not recognised for its medicinal properties (Anderson 1939).

R. integrifolia is also geographically distinct from R. rosea, and is known to grow only in the western part of North America, whereas R. rosea grows on the eastern side and as a circumpolar distribution. Map 3.4 shows the presence of R. integrifolia in Alaska, Yukon,

Northwest Territories, British Columbia, Alberta and western United States. R. rosea is found in Quebec, Nova Scotia, Newfoundland, Labrador, Greenland, Sweden and the

Russian Federation.

58 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

Map 3.4. Worldwide distribution of two Rhodiola herbarium specimens: R. rosea (circumpolar range) and R. integrifolia (limited to western NA). 59 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

«i

*?? m I 8-

S -S

Map 3.5. Close up of the distribution of R. rosea herbarium specimens on the east coast and R. integrifolia herbarium specimens on the west coast of North America.

60 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

Figure 3.1. Rhodiola integrifolia herbarium specimen (A) and Rhodiola rosea herbarium specimen (B).

61 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

Maxent distribution model

The resulting model yielded an AUC value of 0.96, which is considered excellent

(Swets 1988). Elevation was the most important parameter for the model construction

whereas aspect was the least important. Geology was the second most useful variable when

used in isolation. A visual observation of the species distribution (Map 3.6) showed a high to moderate suitability for the east coast and moderate to low suitability for the west coast,

except for the Quaqtaq region. Considering the rarity of the plant along the west coast (map

3.2), the model revealed an appropriate representation of R. rosea distribution along Ungava

Bay. This demonstrates the importance of elevation and geology on the plant's growth.

Habitat characterization in Nunavik

The general habitat of R. rosea has been investigated in circumpolar regions, although not in Northern Quebec particularly, and has been characterised in various arctic floras (Polunin 1959, Porsild 1964, Clausen 1975, Aiken et al. 1999). This stonecrop plant is reported to grow on various substrates including slopes, ridges, cliffs, gravel and till, and in well drained to moist areas with low organic matter (Aiken et al. 1999).

In Nunavik, R. rosea was also found to grow in a wide range of habitats, from rocky beaches to sandy shores and at a few meters from high tide level. The most commonly encountered substrates were rocky or pebble beaches, cliffs and sheets of rocks (Tables 3.1 and 3.2), where the soil is poor and moderately drained. According to observations, R. rosea is considered a coastal plant and was observed, in Nunavik, to grow only along shores of salt or mixed waters. Thus, R. rosea can be recognized as a halophyte.

Certain plant species are commonly associated with R. rosea, e.g. Empetrum nigrum,

62 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea

Map 3.6. Maxent model representing Rhodiola rosea's predicted geographic distribution along Ungava Bay, Nunavik, QC, according to six environmental variables. Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

Table 3.1. Ecological data compiled from Nunavik collection set 1, August expedition 2005, Ungava Bay, QC.

Population Latitudea Altitude Edaphic Deformation Density (N) (m) conditions b occurrence (individuals/ m2)

Kangiqsualujj uaq A 58°41'32.5" 22 G,H N/A B 58°40'37.9" 2 G,J yes 6.09 C 58°30'57.3" 5 A,G,J yes 3.3 D 58°33'05.0" 16 G,L yes N/A E 58°35'49.3" 3 G N/A F 58°40'35.7" 1 G yes 1.2 G 58°4138'.4" 4 D,H,K yes 15 Kuujjuaq H 58°01'34.4" 11 B,F N/A I 58°08'52.4" 2 D yes 0.6 J 58°21'26.3" 7 G,L yes 1.8 K 58°26'45.4" 9 G 3.6 L 58°16'12.8" 3 G N/A M 58°06'41.4" 3 A, J N/A a Coordinates are expressed in hddd°mm'ss.s" b Legend: A-loam, B-sandy loam, C-sandy soil, D-clay loam, E-rocky beach, F-cliff, G-sheets of rock, In­ organic matter, I-moss substrate, J-rocks, K-gravel, L-human disturbed soil.

64 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

Table 3.2. Ecological data compiled from Nunavik collection set 2, August expedition 2006, Ungava Bay, QC.

Population Latitudea Altitude Edaphic Deformation Distance from high (N) (m) conditions b occurrence tide (m)

Kangiqsualujj uaq AA 58°44'09.4" 5 A, J 300+ BB 58°45'02.7" 0 E 15 Killiniq CC 60°25'29.3" 2 A,F,H 4° DD 60°21'32.6" 0 B,J,K yes 20+ EE 60°13'37.1" 12 C,J 25 FF 60°05'33.1" 4 C,K 40 GG 59°55'15.6" 9 J,L yes 30 HH 59°18'23.3" 3 G,K yes 5 II 59°29'26.0" 10 G,K,L 5-25 JJ 59°03'15.1" 1 C, G,K 30+ KK 58°56'15.3" 10 G,H,I 10-15 LL 59°03'4.34" 6 E 10 Quaqtaq MM 61°04'41.4" 38 F,K 30c NN 61°02'49.0" 6 F 10-15c 00 61°04'15.3" 48 F, H, K. 41c Kuujjuaq PP 58°32'00.7" 11 G yes 8 QQ 58°33'05.9" 8 G,L yes 10 RR 58°19'35.6" 1 G,L 2 a Coordinates are expressed in hddd°mm'ss.s" b Legend: A-loam, B-sandy loam, C-sandy soil, D-clay loam, E-rocky beach, F-cliff, G-sheets of rock, In­ organic matter, I-moss substrate, J-rocks, K-gravel, L-human disturbed soil. c Distance from high tide is vertical since plant is situated on a cliff

65 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

Saxifraga tricuspidata and Campanula uniflora, although levels of competition remain

minor (Clausen 1975). This succulent plant could therefore prefer habitats with low

interspecific interactions. This can be correlated with halophytic properties of the plant.

Halophytes occupy specific habitat zones, which are subject to desiccating effects of strong

winds, splash and salty spray of seawater and wave action as well as strong fluctuations in

temperature (Waisel 1972). As for most halophytes, it is still unclear if coastal habitats are

occupied by R. rosea because of its requirement in sodium chloride or because this particular

niche creates a low-competition environment, as opposed to the hard competitive pressures

of inland communities (Waisel 1972).

As indicated by the Maxent distribution model (Map 3.6), R. rosea is influenced by

elevation and geology. According to Waisel (1972), the type and characteristics of the

geological rock present, including stability, depth of soil, and degree of hardness, determines

the vegetation present. R. rosea, being a plant growing on rock outcrops, is highly

susceptible to environmental and edaphic extremes (Crow and Ware 2007). The study by

Crow and Ware (2007) showed that plant growth is affected by tolerance to bedrock

chemistry and competition. Results demonstrated that Sedum nuttallianum, a closely related

genus to Rhodiola, was more limited by competition than by physiological intolerance to the

substrate. So chemistry of the bedrock and soil could be important factors in describing R.

rosea habitats, provided that competition is low.

Nutrient availability also plays an important role in plant habitat especially since the

arctic tundra is a poor environment limited in nitrogen (Scott 1995, Shevtsova et al. 2005).

Four soil samples were collected in Nunavik and analyzed, according to the Mehlich 3 method, for their content in nitrate (NO3"), phosphorus ion (P3") and potassium ion (K+)

66 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

(Table 3.3). Soil samples showed very low nitrate levels for Quaqtaq and highest levels in

Kuujjuaq. Phosphorus levels were quite low in one sample from Kuujjuaq (population PP)

but the other three samples fall into the range of boreal forest floor P levels (75-150 kg/ ha).

Potassium is also low for population PP but within the boreal forest levels for the others

(300-750 kg/ ha) (Brandy and Weil 2002). Levels of pH vary from slightly acidic to neutral

(pH 5.0 to 6.9), a range in which Rhodiola is known to thrive under cultivation (Galambosi

2006). R. rosea has often been reported to grow near bird colonies and human disturbed areas (Aiken et al. 1999, Porsild 1964). These observations were also noticed in Nunavik. A relationship can be established between an increase in nutrients, specifically nitrogen, and the presence of birds and humans. For instance, seagulls and other birds enrich the soil with guano, which is nitrogen and phosphorus rich (Scott 1995). This important relationship between R. rosea and nitrogen/ phosphorus accessibility is highly probable since this plant is known to be nitrophilous (Porsild 1964).

In conclusion, Nunavik R. rosea growing mostly on rocky, moist but well drained substrates, a few metres from high tide level, where vegetation competition is low and nutrient availability has the potential to be high (especially in nitrogen).

Investigation for plant mite infestation

Plant deformation was observed repeatedly in the field with no records as to the identification of the gall-causing agent (Clausen 1975). Samples were studied by Dr.

Frederic Beaulieu from the National Identification Service of Agriculture and Agri-Food

Canada who tentatively identified the causing agent as an herbivorous mite, specifically the eriophyoid bud mite Aceria rhodiolae (Canestrini) (order Acari). This bud mite, which had

67 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

Table 3.3. Nutrient content (N-P-K) and soil pH level of four soil samples collected in Nunavik, QC, and analyzed according to the Mehlich 3 method, by AgriDirect.

Soil location Associated pH Nitrate fNOO Phosphorus fP31 Potassium (K+) populations fppm) as/ha) Kuujjuaq J 6.4 1.9 186 503 Kuujjuaq PP 5.0 5.78 19 193 Quaqtaq MM 6.9 0.5 88 219 Quaqtaq 00 5.8 0.5 73 364

68 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

never been identified in Canada before, is an obligate parasite and is highly host specific

(Krantz and Lindquist 1979). Signs of deformation, such as gall formations, were observed

primarily on female fruiting inflorescences and occasionally on leaves surrounding the

infructescence. Abnormalities show no pattern in symmetry and the deformed tissue turns

fleshy with a whitish green and yellow color (Figure 3.2).

Gall formations are essential for eriophyoid bud mites, which depend on living tissue

for survival (Krantz and Lindquist 1979). Bud mites were observed under a microscope

(Figure 3.3) and mean measurements (n=l 1) of length averaged 0.19 mm by 0.08 mm wide.

Eriophyoid mites are wind dispersed and can also be transported by birds (Jeppson et al.

1975, Blackman and Eastop 2006). These are important means of transportation for the mites in arctic areas, which can explain the extent of their distribution. As seen on map 3.3, occurrences of distorted plants show no significant pattern related to environmental or edaphic conditions.

The impact of plant deformation on the production of phytochemicals was found to be significant for salidroside but minimal for the rosavins. The secondary compound salidroside, a phenylethanol derivative with antidepressant properties (Brown et al. 2002) was significantly decreased when analysed in distorted plants. Since R. rosea is an important medicinal plant species, deformation by bud mites not only has a negative influence on plant health but also decreases significantly one of its four important medicinal compounds

(Brown et al. 2002, Kurkin 2003, Tolonen et al. 2004). Since eriophyoid mites are able to transmit plant viruses (Agrios 1997), further investigation should be conducted on plant deformations and its impacts on plant health.

Although, other pathogens have been associated with R. rosea, very few have been

69 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

Figure 3.2. Healthy infructescence of R. rosea plant (A) versus mite infested infructescence of Nunavik plants (B).

70 Chapter 3: Ecological Analysis ofNunavik Rhodiola rosea L.

Figure 3.3. Eriophyoid bud mite Aceria rhodiolae (Canestrini) infesting Nunavik R. rosea bud tissue. The digital pictures were taken with a compound microscope at lOx and 40x objectives.

71 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

observed in Canada. For instance, Puccinia umbilici, a rust-causing agent, has been

identified in Norway (Gjaerum 1989), and Phytomyza rhodiolae, a leaf-miner specific to

roseroot found in Canada and Norway, is now spreading to Britain (Bland 1995).

Experimental plantation in Nunavik

High demands for R. rosea, as a medicinal plant product, led to overexploitation of its

natural populations in Eurasia (Brown et al. 2002). These are currently on the red list of

threatened species in the Russian Federation (Galambosi 2006). To promote conservation

without impeding on the commerce of herbal material, practices such as the introduction and

cultivation of species in other geographic regions have been opted as a solution (Kucinskaite

et al. 2007). Today, many medicinal plants are introduced and cultivated in new regions with new ecological conditions (Kucinskaite et al. 2007). Consequently, different climatic

and edaphic conditions can significantly affect the quantity of phytochemicals produced

(Figure 2.1). Results from the introduction and cultivation of Siberian R. rosea in Lithuania by Kucinskaite et al. (2007) showed a significant increase in production of rosavins due to regional influence. Considering the high phytochemical concentration of Lithuanian R. rosea, these were found to be an adequate substitute to the naturally growing medicinal plants. This increase in plant material supply from cultivation will help decrease pressure on wild populations.

The Nunavik experimental plantation was analyzed for survival rate and plant health.

According to Kir'yanov et al. (1988), salidroside and rosavins appear in the rhizome only during the fifth year of growth (Furmanowa et al. 1999). However, recent unpublished studies suggest that the compounds may appear earlier in the rhizome. Nevertheless, roots

72 Chapter 3: Ecological Analysis of Nunavik Rhodiola rosea L.

are still harvested only after the fourth or fifth year (Galambosi 2005, 2006). Therefore, the

Nunavik seedlings were not analyzed for their phytochemical content.

Total survival rate of seedlings for the first Nunavik experimental plantation,

established in 2006, was evaluated at 64% after one year (Table 3.4). However, this plantation was established on a site where R. rosea is already present therefore wild

seedlings could have grown among planted ones. In 2007, most seedlings had reached a height of approximately 8 to 9 cm, and were in good health. No sign of seedling deformation was noticed nor was there any trampling or vandalism. The natural competition of the site, although quite low, did not seem to affect the seedlings.

According to Galambosi (2006), roseroot plantation sites should mimic the plant's natural ecological conditions, as this can impact seedling survival. Sites where plant competition is low and which are close to high tide level seem advantageous since these are similar to the plant's natural environment. Unfortunately, the Quebec low arctic environment is a fragile ecosystem, difficult to transform mechanically (e.g. with a rototiller), and where suitable cultivation sites are present in scattered and disconnected patches. The establishment of plantations in Nunavik is limited by important barriers such as a lack of workable terrain therefore few relatively flat, large and accessible areas are available. Also, the poor soil conditions and the rocky sites are not conducive to large scale cultivation. Nevertheless, cultivation of R. rosea through the design of appropriate small- scale plantation sites and linked to awareness raising activities and active involvement of

Inuit communities, could help meet the increasing market demand while also reducing harvesting pressure on wild Canadian R. rosea populations.

73 Chapter 3: Ecological Analysis ofNunavikRhodiola rosea L.

Table 3.4. Survival rate of the Nunavik experimental plantation established in Kangiqsualujjuaq, August 2006.

Seedling origina Total number of individuals Survival rate b counted in 2007 planted in 2006 %

Total seedlings in plantation 111 174 64

Seedlings from population A 29 70 41 Seedlings from population B 16 18 89 Seedlings from population C 66 86 77 a Different population of seeds were collected in Nunavik (2005) and grown at the Montreal Botanical Garden, QC. b Survival rate was calculated by dividing the number of individuals who survived by the number of individuals planted multiplied by 100.

74 Chapter 4: General Discussion

CHAPTER 4

General Discussion

75 Chapter 4: General Discussion

4.1 Major findings and claims to originality

The first purpose of this research study was to characterize phytochemically wild

populations of Nunavik Rhodiola rosea, and compare and contrast their phytochemistry with

other circumpolar collections as well as among themselves. This work was the first ever

phytochemical analysis of Canadian populations and it showed that they have similar

phytochemical markers to the Eurasian populations. The phytochemistry data combined

with the genetic marker analysis from our collaborators at IRB V, Universite de Montreal,

(Archambault et al. 2008) provided unequivoqual confirmatory evidence that the Nunavik plants are authentic R. rosea. The phytochemical results from Chapter 2 also presented quantitative data on authentic Nunavik plants as well as their concentration in different plant tissues. This may be important as a reference point for future studies on different world populations and local commercial efforts. Findings confirmed that very low amounts of active compounds were found in aerial parts, but substantial amounts were found in roots

(Hohtola et al. 2005, Petsalo et al. 2006). Furthermore, results helped established the medicinal potential of these newly studied plant populations. Phytochemical analysis addressed new issues in respect to variations caused by plant gender and mite infestation.

Generally, the findings showed a significant variation in salidroside concentration in relation to several environmental factors while the quantity of rosavins remained stable.

Secondly, the ecological study aimed at creating species distribution maps of Ungava

Bay, Nunavik, as well as characterizing the plant's habitat for this region. Findings, from the ecological analysis of Chapter 3 and herbarium records, suggested a specific distribution of

Nunavik R. rosea mostly along the coastal shoreline of eastern and southern Ungava Bay.

76 Chapter 4: General Discussion

Herbarium records.also indicated populations of R. rosea growing along the Labrador coast.

Further investigations are needed to determine the species' abundance of that area.

Collection records also showed no indication of populations distributed along the

Hudson Bay shoreline and only isolated populations were found in the Maritime provinces.

This suggests that to date, the Ungava Bay coastline is the major population centre for this

species in Canada. Future expeditions are needed to evaluate the presence of the plant on the

west coast, for instance between Kuujjuaq and Kangirsuk. According to unofficial records,

occurrences of R. rosea were noted in Tasiujaq.

This study is also the first to use a distribution modelling technique to evaluate the

dispersion of a medicinal plant. The species prediction model of R. rosea, created using

Maxent, is a novel finding.

The investigation of plant deformation was the first to note occurrences of Aceria

rhodiolae (Canestrini) in Canada. The dispersion data of A. rhodiolae collected in the field

were provided to Agriculture and Agri-Food Canada as a national record. Finally, a first

experimental plantation, established with local growing plants, was created in Nunavik to

evaluate the potential of cultivation practices in Canadian arctic tundra.

4.2 Comparison with published literature on R. rosea

For years, investigations have been conducted on Siberian R. rosea, which has been

coined "true Rhodiola rosea" by some authors (Brown et al. 2002). Results of the present thesis suggest this interpretation is incorrect botanically since Nunavik plants are also medicinally active R. rosea species. Unfortunately, most of the primary literature on R. rosea has yet to be translated to English.

77 Chapter 4: General Discussion

More recently, scientists have identified, categorized and isolated various compounds

of the Eurasian Rhodiola spp. (Fan et al. 2001, Tolonen 2003, Tolonen et al. 2003, Tolonen

et al. 2004, Ma et al. 2006). General comparisons of these biochemical markers are used to

distinguish between various species of Rhodiola, which do not offer the same medicinal virtues (Fan et al. 2001, Abidov et al. 2003, Ampong-Nyarko et al. 2005, Wang et al. 2005,

Yousef et al. 2006). Several studies are being performed on different origins of R. rosea to evaluate their medicinal potential, especially those from China and India (Tolonen et al.

2003, Pooja et al. 2005). The introduction and cultivation of Eurasian R. rosea in other settings (including Alberta) was attempted and promising results were obtained (Galambosi

2006, Kucinskaite et al. 2007).

Over the years, several ecological surveys of Rhodiola were performed in North

America (Herbarium records by Blondeau and Dutilly, QFA, Clausen 1975). One major study was conducted by Robert Clausen, a botanist who inventoried occurrences of the genus

Sedum north of the Mexican plateau in the 1950's (Clausen 1975). Due to taxonomical classification complexities, R. rosea was previously included in the genus Sedum and was known as Sedum rosea (Hegi, 1968, Brown et al. 2002). Clausen noted growing locations of

R. rosea along the Atlantic coast of Canada all the way to Northern Quebec. Some of his observation sites were revisited during the Nunavik expeditions. Unfortunately, not all of his recorded populations survived the last fifty years. His botanical study of Sedum is considered a prominent work and permitted the characterization of North American habitats.

Through his cultivation work at the Ithaca Garden, N.Y., morphological comparisons of

Sedum species were established according to genotypic variations of geographically distinct

78 Chapter 4: General Discussion

populations. Our habitat characterization study determined a very similar profile of R. rosea

ecological requirements in Nunavik, as earlier stated by Clausen.

4.3 Future work

This first study on Canadian R. rosea created an initial step toward future

phytochemical and ecological investigations. For example, further study of this plant's

unique secondary metabolites is needed. New phytochemicals specific to Nunavik R. rosea

appear in chromatograms of this thesis. These should be isolated, purified, identified, and

then quantified with the newly developed and validated extraction and analytical methods, in

order to provide unique biochemical markers for Nunavik plants. These phytochemical markers will assist in the identification and differentiation of Nunavik plant material in future natural products (Gavrilov et al. 2003). The medicinal properties of Nunavik R. rosea should be examined through bioassays, for instance antioxidant, immunostimulation and anxiety tests. These would determine the extent of the medicinal potential of Nunavik extracts

(Tolonen 2003). A clinical trial, such as a trial on anxiety, would provide the best evidence to support commercialization.

A study of the influence of genotype on phytochemical production should be examined in common gardens, where plants will be grown under similar environmental conditions. In order to fully understand the impact of environmental pressures on phytochemistry, a greenhouse experiment should be devised to assess the effect of different environmental factors present in the arctic environment (Scott 1995). Factors such as the reduction or increase in daylight intensity and photoperiod, fluctuation of temperature

(Kovaleva et al. 2003), drought and superficial wounds should be tested independently on

79 Chapter 4: General Discussion

plant secondary metabolite production. Since Nunavik R. rosea plants grow in low

competition environments (Waisel 1972); future examinations should investigate the

concentration of the key phytochemicals when plants are exposed to intra and inter-specific

competition. Roseroot may be found to cause allelopathic effects.

Findings of the impact of environmental factors can be used for future prediction of

the plant's response to global warming and conservation efforts. The plant's present distribution in Nunavik can also serve to understand forthcoming vegetation migration brought on by global-scale climatic changes (Martinez-Meyer and Peterson 2006).

The first experimental plantation of locally growing R. rosea was created in Nunavik in 2006. Salidroside and rosavins were reported by some to be produced only during the fifth year of growth (Kir'yanov et al. 1988, Furmanowa et al. 1999), while subsequent studies suggest root harvesting at the fourth or fifth year (Ampong-Nyarko et al. 2005,

Galambosi 2005, 2006). The evolution of the phytochemical concentration in the rhizome should be analyzed from year to year to determine or confirm previous results as to the timing of advantageous phytochemical yield (Hohtola et al. 2005). This should be examined in both greenhouse and plantation settings.

Since introduction and cultivation practises are promising (Galambosi 2006,

Kucinskaite et al. 2007), different techniques of plant propagation should be investigated for

Nunavik R. rosea populations. Several methods of achieving optimal growth such as hydroponic culture, vegetative reproduction through rhizome cutting and seed germination, are being tested by different research groups (Galambosi 2006, 2005, Ampong-Nyarko et al.

2005). Others have been examining secondary metabolite production in callus aggregate cultures of R. rosea (Furmanowa et al. 1999, Gyorgy et al. 2004, Tolonen et al. 2004;

80 Chapter 4: General Discussion

Hohtola et al. 2005, Zhou et al. 2007). Conclusions are needed on the optimal forms of plant

propagation and their correlations to bioactive compound yield.

In conclusion, Inuit phytomedicine is in its infancy but experimental trials appear

promising; these suggest that roseroot cultivation in the arctic tundra could be feasible.

Already, preliminary data are showing that transplantation is working (Ampong-Nyarko et

al. 2005, Galambosi 2005, 2006). At the Jardin botanique de Montreal, seedling survival was high but additional seedling experiments should be undertaken. Continued efforts in arctic cultivation are being pursued and Inuit phytomedicine development may well be a reality in the near future. After investigating the impact of genotypes on phytochemistry, and plant cultivation needs, roseroot could prove to be an important medicinal crop. R. rosea in Nunavik could represent a local success further helping the Inuit in their quest for economic independence.

81 Appendix 1: Analytical Method Development Manuscript

APPENDIX 1

A-l: Analytical Method Development Manuscript

The manuscript Filion et al. Phytochemical Analysis of Nunavik Rhodiola rosea L. was submitted to Natural Product Communications. After revisions, the manuscript was accepted for publication on February 15th, 2008.

82 2008 NPC I Natural Product Communications Vol.0 No.0 Phytochemical Analysis of Nunavik Rhodiola rosea L. 1-3

Vicky J. Filion"*, Ammar Saleem", Guy Rochefort0, Marc AIIardc, Alain Cuerrierb and John T. Arnason"

a Centre for Advanced Research in Environmental Genomics, Department of Biology, University of Ottawa, Ottawa, Ontario, CANADA, KIN 6N5, b Plant Biology Research Institute, Universite de Montreal, Montreal, Quebec, CANADA, H1X 2B2 c Nunavik Biosciences Inc., Montreal, Quebec, CANADA, H4M 2X6

, John, arnason @ uottawa. ca

Received: November 27th, 2007; Accepted: XX, 2008

This is the first report on the phytochemistry of Nunavik (Quebec, Canada) populations of Rhodiola rosea L., a medicinal plant widely used in Eurasia as a tonic and adaptogen. The wild harvested rhizome of the Nunavik populations contained the marker phytochemicals (salidroside, rosarin, rosavin and rosin) reported in authentic Eurasian material, although in lesser amount. Phytochemical profiling by HPLC of the Nunavik populations also showed the presence of new marker compounds not found in the Eurasian material. For quantitative analysis of the phytochemicals, method validation was undertaken, and the marker phytochemicals were measured in the rhizome, leaf, stem, and seeds. The rhizome showed the highest amount of salidroside and rosavins, as well as the highest total phytochemical content. Consequently, the rhizome remains the most medicinally valuable part of R. rosea.

Keywords: Rhodiola rosea, phytochemistry, method validation, salidroside, rosavins

Rhodiola rosea L. (Crassulaceae) is an important extraction method using accelerated solvent Eurasian medicinal plant known for its extraction (ASE) and a HPLC analytical method. pharmacological activities, such as immuno- stimulant, anti-depressant and adaptogenic properties The medicinal properties associated with R. rosea [1, 2, 3]. This economically important medicinal have been attributed to the plant's rhizome. Its plant has a circumpolar distribution and is found in common names, such as 'gold root', 'roseroot' and subarctic and alpine areas (Figure 1). Its range 'golden root', illustrate its long-established value [9]. includes the Russian Federation, Scandinavia, For this reason, the rhizome was the primary focus of Greenland, Iceland and eastern Canada [4,5]. This investigation. However, other parts of the plant are study is the first to examine R. rosea from Nunavik also used traditionally, for example the leaves are (a low arctic region of Northern Quebec, Canada) as eaten as a vegetable substitute [10]. For this reason, a potential medicinal crop. aerial parts of Nunavik plants were examined and the composition of marker phytochemicals in the The unique medicinal activity of the plant is rhizome, leaf, stem, and seeds were compared attributed to the phenylethanol derivative salidroside (rhodioloside) and the phenylpropanoids rosarin, Comparison of different extraction methods: rosavin and rosin (Figure 1) [1-7]. The latter First, to maximize the amount of phytochemicals compounds are exclusive to R. rosea [8]. The plant recovered from the botanical extracts, two different has 'adaptogenic' properties and these extraction methods were examined: the automated pharmacologically active compounds are known to extraction with the ASE, and the common laboratory be active on the central nervous system [1,2]. For method (90% ethanolic extract). this investigation, the collected material from the Quebec Arctic was analyzed through a new Siberian and Nunavik R. rosea rhizomes were extracted using both methods. The chromatographic 2 Natural Product Communications Vol. 3 (0) 2008 Filion VJ et al.

Table 1. Quantification of compounds extracted from Nunaviki?. rosea rhizome according to two different extraction methods, automated ASE extraction and common laboratory extraction method.

Extraction method Rosavin Compound ASE Common method

mg/g ±SE' mg/g ;±SE Salidroside 0.21 ±0.09 0.19 ±0.11

OH DM OH 01! Rosarin 0.10 ±0.05 0.10 ±0.03 Salidroside Rosavin 0.04 ± 0.007 0.000 ±0.00 wXl Rosin 0.06 ±0.003 0.05 ±0.01 * SE, standard error

Method validation: Figure 1. The four marker phytochemicals present in Rhodiola rosea A new method was developed in order to assess the target phytochemicals present in Nunavik extracts. profile of the ASE extraction method displays a Steps of the method validation process are presented flatter baseline and therefore a decrease in compound here. For the first step, compound calibration curves interference by tannins, when compared to the showed a linear relationship after adding five common laboratory method (Figure 2). Results in different concentrations of each standard at 0.50 mg/ Table 1 show a slightly lower yield of plant mL, 0.25 mg/ mL, 0.125 mg/ mL, 0.0625 mg/ mL compounds when using the common laboratory and 0.0313 mg/ mL. Each showed a regression value method, even though both techniques show good (r2) higher than 0.999. reproducibility. The amount of salidroside showed an increase, although not significant, of Secondly, to determine the reproducibility and approximately 10% when extracted with the ASE, as accuracy of this method, analyses were performed compared to the common method. Rosarin and rosin each day for three consecutive days. Each compound however, showed no differences between extraction was then quantified and the coefficient of variation methods. Rosavin was detected only in samples was determined within and between days (Table 2). extracted with the ASE. Since rosarin and rosavin are closely related structurally, higher temperature Table 2. Comparison of intra-day and inter-day variation for four target and pressure settings for the ASE extraction appear to compounds in the Siberian rhizome. The coefficient of variation was used have improved their extraction. Other advantages of for comparison and calculated as (standard deviation/ mean amount) x 100. using the automated extractor are that reproducibility Intra -day variation Inter-day variation of the extraction process is improved, since factors Day 1 Day 2 Day 3 Day 1-3 such as solvent temperature and pressure can be Salidroside 2.65 1.93 2.60 3.72 closely controlled. Rosarin 7.42 7.08 4.30 0.37 Rosavin 0.29 2.64 0.89 1.40 Rosin 2.48 2.57 2.73 2.28

Intraday variation was highest for rosarin at day 1 and 2. However, the lowest inter-day variation was observed with this compound. This difference between experiments could be due to interference with tannins within the extract. In general, except for rosarin, the variation is well below 5%, which demonstrates good reproducibility of the method. The precision for nine injections of Siberian R. rosea was also calculated for each compound. The coefficient of variation obtained was for salidroside, 3.84%, rosarin, 5.57%, rosavin, 1.86% and rosin, 3.00%. Figure 2. HPLC chromatographic profiles of Siberian R. rosea rhizome according to different extraction methods: common method (A), ASE (B) Finally, recovery analyses were conducted on and standards (C). Four standards are observed at 210 nm, salidroside (1), rosarin (2), rosavin (3) and rosin (4). Siberian material. Samples of dried rhizome were Phytochemical Analysis of Nunavik Rhodiola rosea Natural Product Communications Vol. 3 (0) 2008 3

mAU wm Salidrosid E^sa Rosarin Rosavin 7501 Rosin

i<;o

240H

time (min)

Figure 3. HPLC chromatographic profiles of the rhizomes of Siberian R. rosea (A) and Nunavik R. rosea (B) compared to standards (C) at 210 nm. Four phytochemicals are identified: salidroside (1), rosarin (2), rosavin (3) and rosin (4).

spiked with tyrosol, also a medicinally active Leaf Rhizome Stem Seeds compound [1, 2]. Levels (1-4) used were: 0.60 mg/ Figure 4. (A) Quantification of the four target secondary metabolites in mL, 0.30 mg/ mL, 0.15 mg/ mL and 0.075 mg/mL. leaf, rhizome, stem and seeds of Nunavik R. rosea. (B) Quantification expressed as the total phytochemical content of each part of the Nunavik Each spiked sample was prepared in duplicate. plant (+SE). Analyses were carried at 210 nm for salidroside and at 254 Results showed a recovery range from 74.4 % to 92.1 nm for the rosavins. %. For level 1 to 4, percent recoveries were as rhizome [1, 2, 3]. However, other parts of the plant followed: 74.4%, 81.8%, 90.6% and 92.1% and are also used traditionally, for instance as a food coefficients of variation among both replicates were substitute [10]. In this study, we have characterized calculated as 6.89%, 6.84%, 3.34% and 5.96%. In the phenolic glycoside content of four different parts short, recovery after ASE extraction was above 74% of the plant (leaf, rhizome, stem and seeds). Results with low variation between replicates (< 7%). show that the four target compounds (salidroside, rosarin, rosavin and rosin) are present in both the leaf Siberian versus Nunavik R. rosea: R. rosea of and stem, whereas only three are identified in the Siberian origin has been studied extensively over the rhizome and seeds (Figure 4A). Salidroside was past decades [11]. However, Nunavik R. rosea has found to be the highest in the rhizome, followed by never been characterized. This research, supported the seeds, leaf and stem. As for the total rosavin by the Inuit development corporation Makivik, was content (rosavin, rosarin, and rosin), the amount was designed to determine the resource potential of the highest in the rhizome followed by the leaf, stem, and Nunavik medicinal plant. Although the seeds. In Figure 4B, the rhizome displayed the phytochemical markers of Siberian plants occur in highest total phytochemical content with 2.12 mg/g, the Nunavik populations, the chromatographic although not significantly so. The highest content of profiles illustrate differences in chemical composition salidroside and rosavins is found in the wild rhizome between the cultivated Siberian R. rosea and the wild of Nunavik R. rosea, which confirms its medicinal Nunavik plants (Figure 3). New compounds are importance [1]. A study by Hohtola et al. (2005) present in the Nunavik rhizome and an important reported similar findings for R. rosea of Finnish accumulation of tannins is visible in the central part origin. The plant's rhizome contained the highest of the chromatogram. Future phytochemical studies concentration of salidroside, as opposed to the leaves. will need to identify these new marker compounds. However, Petsalo et al. (2006) detected no phenylpropanoids in the aerial parts of the plant, Quantification of different plant parts: Roseroot is although flavonoids were abundant. known for the medicinal activity conferred by its 4 Natural Product Communications Vol. 3 (0) 2008 Filion VJ et al.

In conclusion, this preliminary report of Nunavik R. N.._ " " - — •* rosea provided evidence that the Nunavik species contained marker compounds considered to be unique for the species, as well as medicinally /Qtfaqtaql relevant. Future studies will characterize the novel Akpatok phytochemicals present in the Quebec populations \ * / —- /3 and establish the limit of detection of the compounds. A, * s Different extraction methods have demonstrated an Ungava Bay advantage in using the automated extractor. The A" ASE chromatogram showed lower tannin interference

with the marker compounds (Figure 2). The method •/ 1 Kangiqsualujjuaq validation process confirmed the effectiveness of the >>' \ . HPLC analytical method. Lastly, phytochemicals Kuujjudq N from different parts of the plant were found in A different concentrations with the rhizome containing Nunavik, 45 " 18 2"C Nnrrhpin Onnhpr M •••••KIT »WS the highest quantity of bioactive compounds as opposed to the stem, which contained the least Figure 5. General collecting areas of R. rosea in Nunavik (QC, Canada) (Figure 4). The medicinal value of R. rosea is confirmed to be in the rhizome. solvent were pooled and evaporated to dryness using the Automatic Environmental Speedvac® System (AES 2010 from Savant) at ambient temperature on Future work will investigate the new phytochemicals full vacuum. specific to Nunavik R. rosea, as well as the influence of environmental factors on the phenolic glycoside Common laboratory extraction method: The content of wild and cultivated rhizomes of Nunavik conventional laboratory extraction method involved populations. extraction in 90% ethanol at a ratio of lg: 10 mL of ground material, which was then shaken continuously Experimental for 12 h. This process was repeated three times and Sample collection and preparation: Fresh samples of the resulting liquid extract was filtered and combined R. rosea were collected in Nunavik, QC, Canada with the storage solvent. The pooled extracts were (Figure 5). At each collection site, global positioning dried on a rotary evaporator at ~40°C and analyzed system (GPS) points were recorded and voucher by HPLC. specimens were gathered and stored at Marie- Victorin Herbarium (Montreal, QC, Canada). HPLC-DAD analyses: High performance liquid Cultivated Siberian R. rosea rhizomes were donated chromatography (HPLC) analyses were performed on by Klickitat organics (WA, USA). Samples of an Agilent 1100 series HPLC system (Agilent rhizome, leaf, stem and seeds of wild Nunavik plants Technologies, Santa Clara, CA, USA) using Chem were stored separately in 40 mL of 90% ethanol at Station LC 3D software (Rev. A09.01). The HPLC 4°C at the University of Ottawa (ON, Canada). The system consisted of an auto sampler, quaternary samples collected in the field were stored directly in pump, degasser, column thermostat (off line) and ethanol since facilities for drying were not locally diode array detector. The HPLC method was adapted available. Prior to extraction, the wet samples were from Tolonen et al. (2003). All samples (rhizome, filtered from the storage solvent and dried at ~35°C leaf, stem, and seeds) were analyzed using the same for 12 h (NESCO® Food dehydrator) then ground HPLC method. Each botanical extract was filtered into a coarse powder. through a PTFE membrane of 0.2 um (Chromatographic Specialties Inc., Brockville, ON, ASE extraction method: The dried samples were Canada) and injected, at a volume of 5 uL, through extracted using the ASE® 200 Accelerated Solvent an auto sampler onto a Luna C18 column (150 x 4.60 Extractor (Dionex, Sunnyvale, CA, USA). Each mm, particle size 5 |im) from Phenomenex (Torrance, sample of varying weight was extracted twice in 33 CA, USA). The HPLC column was connected to a mL ASE cells using 90% ethanol heated at ~50°C for CI 8 guard pre-column (SecurityGuard™ from 15 min at pressure of 2000 psi. The cell was then Phenomenex). Column temperature was maintained flushed at 100% ethanol and purged for 120 sec. The at ~ 35°C. Table 3 describes the solvent gradient ASE liquid extract and previously filtered storage according to time (min) and flow rate (mL/ min). Phytochemical Analysis of Nunavik Rhodiola rosea Natural Product Communications Vol. 3 (0) 2008 5

The total run time for each analysis equaled 30 min replicates. The recovery analysis examined the with 5 min of post-run time. Quantification analyses amount of the standard, tyrosol, also obtained from were recorded at detection signals of 210 nm for ChromaDex™, recovered from the ASE extraction. salidroside and at 254 nm for the rosavins. All experiments were conducted in triplicate, and coefficients of variance were calculated and used to Table 3. Elution timetable of newly developed HPLC analytical compare variation between different HPLC runs. method The coefficient of variance was calculated as Time % Solvent Flow rate Water Acetonitrile standard deviation/mean x 100%. A coefficient of (min) (A) (B) Methanol (C) (mlVmin) variance lower than 5% indicated low variation.

0 90 5 5 1.0 8 78 12 12 1.0 Statistical analysis: All statistical analyses were 10 74 18 18 0.8 performed on the S-plus software version 7.0 20 60 20 20 0.8 (Insightful Corporation, Seattle, USA). The tests 25 20 50 30 0.8 performed were a one-way ANOVA followed by a 30 90 5 5 1.0 Tukey post-hoc test, and an unpaired t-test. These analyses were used to evaluate differences between Method validation for HPLC analysis: Validation of the total phytochemical content and comparison of this method was performed on the HPLC-DAD extraction method experiments. For method instrument described above. The first step of this validation, a simple linear regression was used for the process was to construct a calibration table. compound calibration and intra- and inter-day data. Calibration of the HPLC-DAD was obtained from standard solutions of salidroside, rosarin, rosavin, and Acknowledgments - A special thanks to Nunavik rosin, in 90% ethanol at five different concentrations: Biosciences Inc. who initiated, funded and closely 0.50 mg/ mL, 0.25 mg/ mL, 0.125 mg/ mL, 0.0625 supervised this project. Thanks also to the Northern mg/ mL and 0.03125 mg/ mL. Standards were Scientific Training Program (NSTP) from the obtained from ChromaDex™ (Irvine, CA, USA). Department of Indian Affairs and Northern Calibration curves were used to create the linear Development Canada who generously contributed to range for each external standard. The second step several field season expenses. Also thank you to was to determine inter-day variability by analyzing Linda Kimpe, Dr Blais Laboratory, University of the same extract three times a day for three days. Ottawa, for her technical expertise and assistance on This step as well as the precision on nine injections the ASE. Finally, thanks to the Inuit of Nunavik for assessed the variation of the HPLC method between sharing their knowledge and time on this project.

References [1] Brown RP, Gerbarg PL, Ramazanov Z. (2002) Rhodiola rosea: A phytochemical overview. Herbal Gram, 56, 40-52. [2] Kelly GS. (2001) Rhodiola rosea: a possible plant adaptogen. Alternative Medicine Review, 6, 293-302. [3] Tolonen A, Zsuzsanna G, Jalonen J, Neubauer P, Hohtola A. (2004) LC/MS/MS identification of glycosides produced by biotransformation of cinnamyl alcohol in Rhodiola rosea compact callus aggregates. Biomedical Chromatography, 18, 550-558. [4] Aiken SQ Dallwitz MJ, Consaul LL, McJannet CL, Gillespie LJ, Boles RL, Argus GW, Gillett JM, Scott PJ, Blven R, LeBlanc MC, Brysting AK, Solstad H. (1999). Flora of the Canadian Arctic Archipelago: Descriptions, Illustrations, Identification, and Information Retrieval. Version: 29th April 2003. http://www.mun.ca/biology/delta/arcticf/'. Accessed: July 19, 2005. [5] Hegi G. (1958) Illustrierte Flora von Mittel-Europa. Bd. IV, 2. Teil. Carl Hanser Verlag, Miinchen. [6] Linh PT, Kim YH, Hong SP, Jian JJ, Kang JS. (2000) Quantitative determination of salidroside and tyrosol from the underground part of Rhodiola rosea by high performance liquid chromatography. Archives of Pharmacol Research, 23, 349-352. [7] Khanum F, Bawa AS, Singh B. (2005) Rhodiola rosea: A versatile adaptogen. Comprehensive Reviews in Science and Food Safety, 4, 55-62. [8] Ganzera M, Yayla Y, Khan IA. (2001) Analysis of the marker compounds of Rhodiola rosea L. (golden root) by reversed phase high performance liquid chromatography. Chemical and Pharmaceutical Bulletin, 49,465-467. [9] Small E, Catling PM. (1999) Canadian Medicinal Crops. National Research Council Research Press, Ottawa, Ontario. 240 pp. [135-139] [10] Aim T. (2004) Ethnobotany of Rhodiola rosea (Crassulaceae) in Norway. SIDA Contribution to Botany, 21, 321-344. [11] Rohloff J. (2002) Volatiles from rhizomes of Rhodiola rosea L. Phytochemistry,59, 655-661. 6 Natural Product Communications Vol. 3 (0) 2008 Filion VJ et al.

[12] Tolonen A, Hohtola A, Jalonen, J. (2003) Comparison of electrospray ionization and atmospheric pressure chemical ionization techniques in the analysis of the main constituents form Rhodiola rosea extracts by liquid chromatography/mass spectrometry. Journal of Mass Spectrometry, 38, 845-853. [13] Hohtola, A. (2005). Section 5.7 Natural product formation by plants; enhancement, analysis, processing and testing. In Sustainable use of renewable natural resources - from principles to practices. Jalkanen, A.and Nygren, P. (Eds). University of Helsinki Department of Forest Ecology Publications 34. [14] Petsalo A, Jalonen J, Tolonen A. (2006) Identification of flavonoids of Rhodiola rosea by liquid chromatography-tandem mass spectrometry. Journal of Chromatography A, 1112, 224-231. Appendix 2: Phytochemical Analysis of Commercialized Rhodiola rosea L.

APPENDIX 2

A-2: Phytochemical Analysis of Commercialized Rhodiola rosea L.

89 Appendix 2: Phytochemical Analysis of Commercialized Rhodiola rosea L.

INTRODUCTION

Different commercialized natural products of Rhodiola rosea were evaluated for their

phytochemical quality thus their concentration in pharmacologically active components.

Currently, commercial extracts ofR. rosea are standardized according to their concentration

in four active compounds: salidroside and the rosavins which include rosavin, rosarin and

rosin (Brown et al. 2002). In order to assess the activity of these products as claimed on their

labels, our quantitative analysis compared the amount of these four phytochemicals resulting

from our study, with those advertized. Aside from this phytochemical analysis, the origin of

Rhodiola used in the mixtures was genetically examined by collaborators from the Universite

de Montreal.

METHODS

Sample preparation: Commercial products advertising R. rosea concentrations were purchased from twenty different companies (Table A2-1). Three different forms of material were acquired: capsules, tablets and a liquid extract. For each product, an accession number was assigned randomly and the material was divided into two replicates. For each replicate, capsules were opened and pooled together to create a homogenized sample (Table A2-2).

Tablets were ground into a fine powder, and liquid extract was evaporated to dryness for immediate HPLC-DAD analysis. Exactly 1 g of product was weighted and stored at room temperature until extraction with the ASE.

ASE extraction method: The weighted samples were extracted using the ASE® 200

Accelerated Solvent Extractor (Dionex, Sunnyvale, CA, USA). Each sample of 1 g was extracted twice in 33 mL cells using 90% ethanol heated at 50°C for 15 min at 2000 psi. The

90 Appendix 2: Phytochemical Analysis of Commercialized Rhodiola rosea L.

cell was then flushed at 100% ethanol and purged for 300 sec. The ASE liquid extract was

evaporated to dryness using the Automatic Environmental Speedvac® System (AES 2010

from Savant) at ambient temperature on full vacuum.

HPLC-DAD analyses: High performance liquid chromatography (HPLC) analyses

were performed on 1100 series HPLC system (Agilent Technologies Santa Clara, CA, USA)

using Chem Station LC 3D software (Rev. A09.01). The HPLC system consisted of an auto

sampler, quaternary pump, degasser, column thermostat (offline) and diode array detector.

The HPLC method was adapted from Tolonen et al. (2003). All samples (powder or liquid)

were analyzed using the same analytical method. Each product extract was filtered through a

PTFE membrane of 0.2 um (Chromatographic specialties Inc., Brockville, ON, Canada) and

injected, at a volume of 5 uL, through an auto sampler onto Luna C18 column (150 x 4.60

mm, particle size 5 um) from Phenomenex (Torrance, CA, USA). The HPLC column was

connected to a C18 guard pre-column (SecurityGuard™ from Phenomenex). Column

temperature was maintained at ~35°C. Solvent gradient of water, acetonitrile and methanol

is described according to time (minutes) and flow rate (mL/ min) (Table 2.1). The total

analysis time for each injection was of 30 min, with 5 min of post-run time. Analyses were

recorded at detection signals of 210 nm for salidroside and 254 nm for rosavins.

RESULTS

Mean phytochemical concentration of salidroside and rosavins measured in each

product is expressed in mg/g ± standard error of mean (SEM) (n=2) (Table A2-3). Results will be compared to the claimed amount of commercialized product.

91 Appendix 2: Phytochemical Analysis of Commercialized Rhodiola rosea L.

Table A2-1. List of commercialized products of Rhodiola rosea analysed phytochemically for their concentration in the four markers. Products names are arranged in alphabetical order.

Product name Company

All you need in one Natural Factors Arctic root Swedish herb Proactive Bioproducts Full spectrum Rhodiola rosea extract Planetary Formulas Rhodiola Paradise Herbs Rhodiola Vitanica Rhodiola 250 mg Nature's plus Rhodiola Energy Enzymatic Therapy Rhodiola Energy PhytoPharmic Enzymatic therapy Rhodiola Extract 100 mg Solaray Rhodiola Extract 3% 500 mg Now foods Rhodiola herbal active Nature's Plus Rhodiola PowerMax 1000 Action Labs Rhodiola rosea 200 mg Bluebonnet Rhodiola rosea 500 mg Jarrow Formulas Rhodiola rosea Natural elixir Rhodiola rosea Natures way Rhodiola rosea Pure encapsulations SPF Rhodiola rosea extract Solgar

92 Appendix 2: Phytochemical Analysis of Commercialized Rhodiola rosea L.

Table A2-2. Product accession numbers were randomly assigned to products. One gram of product was extracted from a homogenized mixture of material. Product # 20 was under liquid form (volume of 50 mL) and subsequently evaporated to dryness and weighted before analysis.

Product accession # # capsules used Equivalence in grams replicate 1 replicate 2 replicate 1 replicate 2 1 14 14 4.5 3.6 2 14 11 7.4 5.8 3 29 28 13.3 12.8 4 29 28 9.1 8.8 5 28 29 10.4 10.8 6 19 18 5.4 5.1 7 13 14 4.2 4.6 8 28 20 15.6 16.3 9 13.5 14.5 22.4 24.8 10 29 28 14.3 13.8 11 28 26 15.3 15.4 12 28 29 12.5 13.2 13 28 29 18.2 19.6 14 28 28 11.1 11.3 15 29 28 14.5 14.2 16 19 18 5.3 5.1 17 28 28 9.7 9.7 18 16 15 6.4 7.8 19 42 42 8.6 8.6

93 Appendix 2: Phytochemical Analysis of Commercialized Rhodiola rosea L.

Table A2-3. Mean phytochemical concentration of salidroside and rosavins (mg/ g) found in twenty commercialized natural products of R. rosea. Accession numbers were attributed randomly.

Product accession # Salidroside Rosavins (mg/g±SEMa) (mg/g±SEM)

1 19.86 ±0.99 39.78 ±4.31 2 22.32 ±1.50 14.09 ±1.37 .3 27.50 ±0.58 6.31 ±0.73 4 22.07 ±0.99 40.55 ±5.63 5 34.62 ±1.80 60.41 ±7.94 6 37.17 ±1.38 57.54 ±6.41 7 33.73 ±0.38 15.14 ±1.80 8 24.09 ±0.36 54.98 ± 9.33 9 37.71 ±2.80 57.76 ±7.73 10 23.48 ±3.43 31.03 ±3.62 11 26.62 ±2.05 34.41 ±3.81 12 14.41 ±0.38 38.55 ±5.43 13 26.90 ±1.08 68.52 ±9.88 14 45.66 ±6.75 6.13 ±0.46 15 19.79 ±0.05 45.48 ±5.36 16 33.18 ±0.17 50.30 ±6.10 17 13.76 ±3.98 33.86 ±5.14 18 18.81 ±2.71 53.18 ±7.29 19 16.33 ±1.24 41.04 ±7.44 20 17.48 ±0.91 32.64 ±3.64

' SEM, standard error of mean

94 References

REFERENCES

Abidov, M., Crendal, F., Grachev, S., Seifulla, R. and T. Ziegenfuss. 2003. Effects of extracts from Rhodiola rosea and Rhodiola crenulata (Crassulaceae) roots on ATP content in mitochondria skeletal muscles. Bulletin of Experimental Biology and Medicine 136 (6): 585- 587

Agrios, G.N. 1997. Plant Pathology. 4 Edition. Academic Press Inc., London. 635 pp.

Aiken, S.G., Dallwitz, M.J., Consaul, L.L., McJannet, C.L., Gillespie, L.J., Boles, R.L., Argus, G.W., Gillett, J.M, Scott, P.J., Elven, R., LeBlanc, M.C., Brysting A.K. and H. Solstad. 1999. Flora of the Canadian Arctic Archipelago: Descriptions, Illustrations, Identification, and Information Retrieval. Version: 29th April 2003. http://vww.mun.ca/biology/delta/arcticf/'

Alliende, M.C. and J.L. Harper. 1989. Demographic studies of a dioecious tree. I. Colonization, sex and age structure of a population of Salix cinerea. Journal of Ecology 11: 1029-1047.

Aim, T. 2004. Ethnobotany of Rhodiola rosea (Crassulacea) in Norway. SIDA Contribution to Botany 21(1): 321-344

Almeida-Cortez, J. S., Shipley, B. and J. T. Arnason. 1999. Do plant species with high relative growth rates have poorer chemical defences? Functional Ecology 13: 819-827

Alonso, C, Perez, R., Nieto, P.M. and J. Delgado. 2005. Gender Dimorphism and Altitudinal Variation of Secondary Compounds in Leaves of the Gynodioecious Shrub Daphne laureola. Journal of Chemical Ecology 31 (1): 139-150

Ampong-Nyarko, K., Cole, D., Galambosi, B. and V. Sowiak. 2005. Preliminary Guidelines on the Cultivation of Rhodiola rosea Alberta. Alberta Agriculture, Food and Rural Development Crop Diversification Division. First Edition. 19 pp.

Anderson, J. P. 1939. Plants Used by the Eskimo of the Northern Bering Sea and Arctic Regions of Alaska. American Journal of Botany 26 (9): 714-716

95 References

Anttonen, M.J. and R. O. Karjalainen. 2005. Environmental and genetic variation of phenolic compounds in red raspberry. Journal of Food Composition and Analysis 18: 759-769

Archambault, M. et al. 2008. Etude moleculaire des populations de Rhodiola rosea L. du Nunavik (Quebec, Canada). Universite de Montreal. Master's Thesis, unpublished.

Baldwin, IT. and C.A. Preston. 1999. The eco-physiological complexity of plant responses to insect herbivores. Review article. Planta 208: 137-145

Bazzaz, F. A. 1997. The Plant Functions as a Balanced System: Flexibility of Allocation. In Plant Resource Allocation. Bazzaz, F. A and J. Grace, editors. Academic Press, California. ppl-37

Blackman, R.L. and V.F. Eastop. 2006. Aphids on the world's herbaceous plants and shrubs. Volume 1: Host List and Keys. John Wiley & Sons Inc. England. 1439 pp

Bland, K.P. 1995. Phytomyza rhodiolae Griffiths, 1976 (Diptera: Agromyzidae), a leaf-miner in roseroot, Sedum rosea (Crassulaceae), new to Britain. Entomologist's Gazette 46: 267-269

Boggs, C.L. 1997. Resource Allocation in Variable Environments: Comparing Insects and Plants. In Plant Resource Allocation. Bazzaz, F. A and J. Grace, editors. Academic Press, California, pp 73-92

Brady, N.C. andR.R. Weil. 2002. The Nature and Properties of Soils. 13 Edition. Pearson Education Inc. New Jersey. 960 pp.

Brekhman, I.I and IV. Dardymov. 1969. New substances of plant origin which increase non­ specific resistance. Annual Reviews of Pharmacology 9: 419-430

Brown, R.P., Gerbarg, P.L. and Z. Ramazanov. 2002. Rhodiola rosea: A Phytochemical Overview. HerbalGram 56: 40-52

Bryant, J.P., Chapin, F.S. Ill and D.R., Klein. 1983. Carbon/nutrient balance of boreal plants in relation to vertebrate herbivory. Oikos 40: 357-368.

96 References

Bryant, J.P., Tuomi, J. and P. Niemala. 1988. Environmental constraints of constitutive and long-term inducible defences in woody plants. Chemical Mediation of Coevolution. Editor K.C. Spencer. Academic Press, San Diego, CA, USA, pp. 367-389

Chust, G., Perez-Haase, A., Chave, J. and J. L. Pretus. 2006. Floristic patterns and plant traits of mediterranean communities in fragmented habitats. Journal of Biogeography 33: 1235- 1245

Clark, P.U., Clague, J.J., Curry, B.B., Dreimanis, A., Hicock, S.R., Miller, G.H., Berger, G.W., Eyles, N., Lamothe, M., Miller, B.B., Mott, R.J., Oldale, R.N., Stea, R.R., Szabo, J.P., Thorleifson, L.H. and J.S.Vincent. 1993. Initiation and Development of the Laurentide and Cordilleran Ice Sheets Following the Last hiterglaciation. Quaternary Science Reviews 12: 79-114

Clausen, R.T. 1975. Sedum of North America, North of the Mexican Plateau. Cornell University Press: Ithaca and London. 742 pp.

Coudun, C, Gegout, J., Piedallu, C. and J. Rameau. 2006. Soil nutritional factors improve models of plant species distribution: An illustration with Acer campestre (L.) in France. Journal of Biogeography 33: 1750-1763

Crow, S.E. and S. Ware. 2007. Soil Type Tolerance in Rock Outcrop Plants: Species of Non- Calcareous Substrates. The Southwestern Naturalist 52 (1): 120-125

Cuerrier, A. et les Aines de Kangirsualujjuaq. 2005. Le savoir botanique de Kangirsualujjuaq. Institut culturel Avataq & Institut de recherche en biologie vegetale. 59 pp.

Danell, K., Elmqvist, T., Ericson, L., and A. Salomonson. 1985. Sexuality in willows and preference by bark-eating voles: defence or not? Oikos 44: 82-90

Davydov, M. and A.D. Krikorian. 2000. Eleutherococcus senticosus (Rupr. & Maxim.) Maxim. (Araliaceae) as an adaptogen: a closer look. Review article. Journal of Ethnopharmacology 72: 345-393

Dixon, E.J. 2001. Human colonization of the Americas: timing, technology and process. Quaternary Science Reviews 20: 277-299 97 References

Dixon, R. and N. Paiva. 1995. Stressed-Induced Phenylpropanoid Metabolism. The Plant Cell 7: 1085-1097

Drasar, P. and J. Moravcova. 2004. Review: Recent advances in analysis of Chinese medical plants and traditional medicines. Journal of Chromatography B, 812: 3-21

Elith, J., Graham, C. H., Anderson, R. P., Dudik, M., Ferrier, S., Guisan, A., Hijmans, R. J., Huettmann, F., Leathwick, J. R., Lehmann, A., Li, J., Lohmann, L. G., Loiselle, B. A., Lohmann, G., Manion, G., Moritz, C, Overton, J. M., Peterson, A. T., Phillips, S. J., Nakamura, M., Nakazawa, Y., Schapire, R. E., Wisz, M. S. and N. E. Zimmermann. 2006. Novel methods improve prediction of species' distributions from occurrence data. Ecography 29: 129-151

Fagan, B.M. 2000. Ancient North America: The Archaeology of a Continent. 3r Edition. Thames & Hudson Ltd, London. 544 pp.

Fan, W., Tezuka, Y., Ni, K. M. and S. Kadot. 2001. Prolyl endopeptidase inhibitors from the underground part of Rhodiola sachalinensis. Chemical and Pharmaceutical Bulletin (Tokyo) 49: 396-401

Fielding, A.H. and J.F. Bell. 1997. A review of methods for the assessment of prediction errors in conservation presences/absence models. Environmental Conservation 24: 38-49

Fritz, R.S. 1995. Direct and indirect effects of plant genetic variation on enemy impact. Ecological Entomology 20: 18—26

Furmanowa, M., Hartwich, M., Alfermann, A.W., Kozminski, W. and M. Olejnik. 1999. Rosavin as a product of glycosylation by Rhodiola rosea (roseroot) cell cultures. Plant Cell, Tissue and Organ Culture 56:105-110

Galambosi, B. 2005. Rhodiola rosea L., from wild collection to field production. Medicinal Plant Conservation 11: 31-35

Galambosi, B. 2006. Demand and availability of Rhodiola rosea L. raw material. In Medicinal and Aromatic Plants. Bogers BJ, Craker LE and D. Lange Editors. Springer, Netherlands, pp 223-236.

98 References

Ganzera, M., Yayla, Y. and LA. Khan. 2001. Analysis of the marker compounds of Rhodiola rosea L. (Golden Root) by reversed phase high performance liquid chromatography. Chemical and Pharmaceutical Bulletin 49 (4): 465-467

Gavrilov, A.S., Shchegolev, A.A. and L.P. Larionov. 2003. New prophylactic preparations of Rhodiola rosea. Pharmaceutical Chemistry Journal 37 (12): 641-644

Gjaerum, H. B. 1989. Roseroot rust Puccinia umbilici on Sedum rosea rediscovered in South Norway. Blyttia 47(4): 189-190

Gleason, H.A. and A. Cronquist. 1991. Manual of Vascular Plants of Northeastern United States and Adjacent Canada. Second Edition. New York Botanical Garden, N.Y. 910 pp

Grace. J. 1997. Toward models of resource allocation by plants.- Plant Resource Allocation. Bazzaz, F. A and J. Grace, editors. Academic Press, California, pp 279-290

Graglia, E., Julkunen-Tiitto, R., Shaver, G.R., Schmidt, I.K., Jonasson, S. and A. Michelsen. 2001. Environmental control and intersite variations of phenolics in Betula nana in tundra ecosystems New Phytologist 151: 227-23 6

Grime, J.P. 1979. Plant Strategies and Vegetation Processes. John Wiley & Sons Inc. England. Reprint by Bell & Howell, 1995, Michigan. 222 pp

Grundt, H. EL, Kjolner, S., Borgen, L., Rieseberg, L. H. and C. Brochmann. 2006. High biological species diversity in the Arctic flora. Proceedings of the National Academy of Sciences of the United States of America 103: 972-975

Gyorgy, Z., Tolenen, A., Pakonen, P., Neubauer, P. and A. Hohtola. 2004. Enhancing the production of cinnamyl glycosides in compact callus aggregate cultures of Rhodiola rosea by biotransformation of cinnamyl alcohol. Plant Science 166: 229-236

Hamilton, J.R., Zangerl, A.R., DeLucia, E.H. and M.R. Berenbaum. 2001. The carbon- nutrient balance hypothesis: its rise and fall. Ecology Letters 4: 86-95

Harborne, J.B. 1998. Phytochemical Methods. A Guide to Modern Techniques of Plant Analysis. 3rd Edition. Chapman and Hall, pp 48-59 99 References

Hegi G. 1958. Illustrierte Flora von Mittel-Europa. Bd. IV, 2. Teil. Carl Hanser Verlag, Miinchen.

Hegi, G. 1963. Illustrierte Flora von Mitteleuropa. Vol. IV/2, Liefering 2/3. Hamburg/Berlin, Germany, pp 99-102.

Herms, D.A. and W. J. Mattson. 1992. The Dilemma of Plants: To Grow or Defend. The Quarterly Review of Biology 67 (3): 283-335

Hernandez, P. A., Graham, C. H., Master, L. L. and D. L. Albert. 2006. The effect of sample size and species characteristics on performance of different species distribution modeling methods. Ecography 29: 773-785

Hillhouse, B.J., Ming, D.S., French C.J., and N.G.H. Towers. 2004. Acetylcholine esterase inhibitors in Rhodiola rosea. Pharmaceutical Biology 42 (1): 68-72

Hohtola et al. 2005. Section 5.7 Natural product formation by plants; enhancement, analysis, processing and testing. In Sustainable use of renewable natural resources - from principles to practices. Editors Jalkanen and Nygren. University of Helsinki Department of Forest Ecology Publications 34. (http://www.helsinki.fi/mmtdk/mmeko/sunare)

Hudson, E., Aihoshi, D., Gaines, T., Simard, G. and J. Mullock. 2001. The Weather of Nunavut and the Arctic. Graphic Area Forecast 36 and 37. NAV Canada Publication. 245 pp.

Jeppson, L.R, Keifer, H.H. and E.W. Baker. 1975. Mites Injurious to Economic Plants. University of California Press, London. 614 pp.

Jing, S.W. and P.D. Coley. 1990. Dioecy and Herbivory: The Effect of Growth Rate on Plant Defense in Acer negundo. Oikos 58 (3): 369-377

Johnson, K. and J.M. Scriber. 1994. Geographic variation in plant allelochemicals of significance to insect herbivores. In Functional Dynamics of Phytophagous Insects. T. N. Ananthakrishnan editor., Science Publishers, Lebanon, New Hampshire, USA. pp. 7-31 de Jong, G. and A.J, van Noordwijk, 1992. Acquisition and allocation of resources: Genetic (co)variance, selections, and life histories. American Naturalist 139: 749-770 100 References

Keinanen, M., Julkunen-Tiitto, R., Mutikainen, P., Walls, M., Ovaska, J. andE. Vapaavuori. 1999. Trade-offs in phenolic metabolism of silver birch: effects of fertilization, defoliation, and genotype. Ecology 80:1970-1986

Kelly, G.S. 2001. Rhodiola rosea: a possible plant adaptogen. Alternative Medicine Review 6 (3): 293-302

Keski-Saari, S. Falck, M., Heinonen, J., Zon, J. and R. Julkunen-Titto. 2007. Phenolics during early development ofBetula pubescens seedlings: inhibition of phenylalanine ammonia lyase. Trees 21 (3): 263

Khanum, F., Bawa, A.S. and B. Singh. 2005. Rhodiola rosea: A versatile adaptogen. Comprehensive Reviews in Science and Food Safety 4: 55-62

Kir'yanov, A.A., Bondarenko, L.T., Kurkin, V.A. and G.G. Zapesochnaya. 1988. Opredelen'ye rozavidina v kornevishchakh rodioly rozovoy (Chromatospectrophotometric assay for measuring rosavidin in Rhodiola rosea rhizomes). Chemical and Pharmaceutical Journal (Moscow): 22: 451-455

Korpelainen, H. 1998. Labile sex expression in plants. Biological Reviews 73: 157-180

Kovaleva, N.P., Tikhomirov, A.A. and V.A. Dolgushev. 2003. Specific characteristics of Rhodiola rosea growth and development under the photoculture conditions. Russian Journal of Plant Physiology 50 (4): 527-531

Krantz, G.W. and E.E. Lindquist. 1979. Evolution of Phytophagous Mites (Acari). Annual Review of Entomology 24:121-158

Kucinskaite, A., Poblocka-Olech, L., Krauze-Baranowska, M., Sznitowska, M., Savickas, A. and V. Briedis. 2007. Evaluation of biologically active compounds in roots and rhizomes of Rhodiola rosea L. cultivated in Lithuania. Medicina (Kaunas) 43: 487-494

Kurkin, V.A. 2003. Phenylpropanoids from medicinal plants: distribution, classification, structural analysis, and biological activity. Chemistry of Natural Compounds 39 (2): 123-153

101 References

Kurkin, V.A. and G.G. Zapesochnaya. 1986. Chemical composition and pharmacological properties ofRhodiola rosea. Chemical and PharmaceuticalJournal (Moscow) 20 (10): 1231-1244

Lambers, H., Chapin III, F.S. and T. L. Pons. 1998. Plant Physiological Ecology. Springer- Verlag, New York. 540 pp.

Lerdau, M. And J. Gershenzon. 1997. Allocation Theory and Chemical Defense. In Plant Resource Allocation. Bazzaz, F. A and J. Grace, editors. Academic Press, California, pp 265- 277

Linh, P.T., Kim, Y.H., Hong, S.P., Jian, J.J. and J.S. Kang. 2000. Quantitative determination of salidroside and tyrosol from the underground part of Rhodiola rosea by high performance liquid chromatography. Archives of Pharmacal Research 23 (4): 349-352

Ma, G, Li, W., Dou, D., Chang, X., Bai, H., Satou, T., Li, J., Sun, D., Kang, T., Nikaido, T. and K. Koike. 2006. Rhodiolosides A-E, monoterpene glycosides from Rhodiola rosea. Chemical & Pharmaceutical Bulletin (Tokyo) 54: 1229-1233

Majewska, A., Grazyna, H., Miroslawa, F., Natalia, U., Agnieszka, P., Alicja, Z. and M. Kuras. 2006. Antiproliferative and antimitotic effect, S phase accumulation and induction of apoptosis and necrosis after treatment of extract from Rhodiola rosea rhizomes on HL-60 cells. Journal of Ethnopharmacology 103: 43-52

Martin, P.S. 1958. Taiga-tundra and full-glacial period in Chester County, Pennsylvania. American Journal of Science 256T 47 0-502-

Martinez-Meyer, E. and A. T. Peterson. 2006. Conservatism of ecological niche characteristics in North American plant species over the Pleistocene-to-recent transition. Journal of Biogeography. 33: 1779-1789

Massei, G, Watkins, R. and S.E. Hartley. 2006. Sex-related growth and secondary compounds in Juniperus oxycedrus macrocarpa. Acta Oecologica 29: 135-140

Ming, D.S., Hillhouse, B.J., Guns, E.S., Eberding, A., Xie, S., Vimalanathan, S. and N.G.H. Towers. 2005. Bioactive Compounds from Rhodiola rosea (Crassulaceae). Phytotherapy Research 19: 740-743 102 References

Mudau, F.N., Soundy, P. du Toit, E.S. and J. Olivier. 2006. Variation in polyphenols content of Athrixia phylicoides (L.) (bush tea) leaves with season and nitrogen application. South African Journal of Botany 72: 398-402

Myers, P.G. 2003. An examination of wind-stress forcing and circulation in the sub-polar North Atlantic. Quaternary International 99-100: 89-98

Nerg, A., Kainulainen, P., Vuorinen, M., Hanso, M., Holopainen J, K. and T. Kurkela. 1994. Seasonal and geographical variation of terpenes, resin acids and total phenolics in nursery grown seedlings of Scots pine (JPinus sylvestris L.). New Phytologist 128: 703-71

Nybakken, L., Aubert, S., and W. Bilger. 2004. Epidermal UV-screening of arctic and alpine plants along latitudinal gradient in Europe. Polar Biology 27 (7): 391-398

Palumbo, M.J., Putz, F.E and S.T. Talcott. 2007. Nitrogen fertilizer and gender effects on the secondary metabolism of yaupon, a caffeine-containing North American holly. Oecologia 151: 1-9

Panossian, A., Wikman, G. and H. Wagner. 1999. Plant adaptogens III. Earlier and more recent aspects and concepts on their mode of action. Review article. Phytomedicine 6 (4): 287-300

Perfumi, M. and L. Mattioli. 2007. Adaptogenic and central nervous system effects of single doses of 3% rosavin and 1% salidroside Rhodiola rosea L. extract in mice. Phytotherapy Research 21: 37-43

Petsalo, A., Jalonen, J. and A. Tolonen. 2006. Identification of flavonoids of Rhodiola rosea by liquid chromatography-tandem mass spectrometry. Journal of Chromatography A. 1112: 224-231

Phillips, S. J., Anderson, R. P. and R. E. Schapire. 2006. Maximum entropy modeling of species geographic distributions. Ecological Modelling 190: 231-259

Polunin, N. 1959. Circumpolar Arctic Flora. Clarendon Press, Oxford. 415 pp.

103 References

Pooja, K.R., Khanum, A. F. and A.S. Bawa. 2006. Phytoconstituents and antioxidant potency of Rhodiola rosea- A versatile adaptogen. Journal of Food Biochemistry 30: 203-214

Porsild, A. 1964. Illustrated flora of the Canadian Arctic Archipelago. National Museum of Canada Bulletin 146. 2nd Edition. Ottawa: Department of the Secretary of State. 218 pp.

Rege, N.N, Thatte, U.M., and S.A. Dahanukar. 1999. Adaptogenic properties of six Rasayana herbs used in ayurvedic medicine. Phytotherapy Research 13 (4): 275-291

Rohloff J. 2002. Volatiles from rhizomes of Rhodiola rosea L. Phytochemistry 59: 655-661

Russian Federation Ministry of Health and Medical Industry. 1989. Russian National Pharmacopoeia. Moscow, Russia: The Russian Federation Ministry of Health and Medical Industry.

Sakai, A.K. and T.A. Burris. 1985. Growth in male and female aspen populus tree clones: a 25 years longitudinal study. Ecology 66: 1921-1927

Salmore, A.K. and M.D. Hunter. 2001. Elevational Trends in Defense Chemistry, Vegetation, and Reproduction in Sanguinaria canadensis. Journal of Chemical Ecology 27 (9): 1713-1727

Saule, M. 2002. La grandeflore illustree despyreneens. Deuxieme edition. Milan. 730 pp.

Scott, G. AJ. 1995. Canada's vegetation: a world perspective. McGill-Queen's University Press. 361 pp.

Shevtsova, A., Nilsson, M., Gallet, C, Zackrisson, O. and A. Jaderlund. 2005. Effects of long-term alleviation of nutrient limitation on shoot growth and foliar phenolics of Empetrum hermaphroditum. Oikos 111: 445-458

Shevtsov, V.A., Zholus, B.I., Shervarly, V.I., Vol'skij, V.B., Korovin, Y.P., Khristich, M.P., Roslyakova, N.A. and G. Wikman. 2003. A randomized trial of two different doses of a SHR-5 Rhodiola rosea extract versus placebo and control of capacity for mental work. Phytomedicine 10: 95-105

104 References

Small, E. and P. M Catling. 1999. Canadian Medicinal Crops. NRC Research Press: Ottawa, ON, Canada. 240 pp.

Soil Classification Working Group. 1998. The Canadian System of Soil Classification. Agriculture and Agri-Food Canada Publication 1646. 3rd Edition. NRC Research Press: Ottawa, ON, Canada. 187 pp.

Swets, J.A. 1988. Measuring the Accuracy of Diagnostic Systems. Science 240: 1285-1293

Tolonen, A. 2003. Analysis of secondary metabolites in plant and cell culture tissue of Hypericum perforatum L and Rhodiola rosea L. Academic Dissertation. Department of Chemistry, University of Oulu, Finland. 67 pp.

Tolonen, A., Hohtola, A. and J. Jalonen. 2003. Comparison of electrospray ionization and atmospheric pressure chemical ionization techniques in the analysis of the main constituents form Rhodiola rosea extracts by liquid chromatography/mass spectrometry. Journal of Mass Spectrometry 38: 845-853

Tolonen, A., Hohtola, A. and J. Jalonen. 2003. Liquid chromatographic analysis of phenylpropanoids from Rhodiola rosea extracts. Chromatographia 57: 577-579

Tolonen, A., Pakonen, M., Hohtola, A. and J. Jalonen. 2003. Phenylpropanoid glycosides from Rhodiola rosea. Chemical and Pharmaceutical Bulletin 51 (4): 467-470

Tolonen, A., Zsuzsanna, G., Jalonen, J., Neubauer, P. and A. Hohtola. 2004. LC/MS/MS identification of glycosides produced by biotransformation of cinnamyl alcohol in Rhodiola rosea compact callus aggregates. Biomedical Chromatography 18: 550-558

Udintsev, S.N. and V.P. Schakhov. 1991. Decrease of cyclophosphamide haematotoxicity by Rhodiola rosea root extract in mice with Ehrlich and Lewis transplantable tumours. European Journal of Cancer 27 (9): 1182

Waisel, Y. 1972. Biology ofHalophytes. Academic Press Inc., London. 395 pp.

105 References

Walls, R., Appel, H., Cipollini, M. and J. Schultz. 2005. Fertility, root reserves and the cost of inducible defenses in the perennial plant Solanum carolinense. Journal of Chemical Ecology 31 (10): 2263-2288

Wang, Q., Ruan, X., Jin, Z., Yan, Q. and S. Tu. 2005. Identification of Rhodiola species by using RP-HPLC. Journal ofZhejiang University Science 6B (6): 477-482

Wenner, C.G. 1947. Pollen diagrams from Labrador, a contribution to the Quaternary geology of Newfoundland-Labrador, with comparisons between North America and Europe. Geografiska Annaler 29: 1-241

Willig, M.R., Kaufman, D.M. and R.D. Stevens. 2003. Latitudinal gradients of biodiversity: pattern, process, scale, and synthesis. Annual Review of Ecology Evolution and Systematics 34:273-309

Winston, D. and S. Maimes. 2007. Adaptogens: Herbs For Strength, Stamina, and Stress Relief. Healing Arts Press. 320 pp.

Yousef, G. G., Grace, M. H., Cheng, D. M., Belolipov, I. V., Raskin, I. and M. A. Lila. 2006. Comparative phytochemical characterization of three Rhodiola species. Phytochemistry 67: 2380-2391

Zhou, X., Wu, Y., Wang, X., Liu, B. and H. Xu. 2007. Salidroside production by hairy roots of Rhodiola sachalinensis obtained after transformation with Agrobacterium rhizogenes. Biological and Pharmaceutical Bulletin 30: 439-442

106