Case Study of the Swiss Flora for Prior Phytochemical and Biological Investigations

Case Study of the Swiss Flora for Prior Phytochemical and BiologicalInvestigationsMichael Adams, Magalie Chammartin, Matthias Hamburger, and Olivier Potterat*

Department of Pharmaceutical Sciences, Division of Pharmaceutical Biology, University of Basel, Klingelbergstrasse 50, 4056 Basel,Switzerland

*S Supporting Information

ABSTRACT: Estimates in the literature as to what extent theworld’s higher plant species have been studied chemically orfor bioactivity are contradictory and range from 0.5% to >12%.In this survey, a model to make credible estimates of the extentof their study is proposed and is exemplified by applying it in acase study of plants native to Switzerland. Using a widelyavailable database (SciFinder Scholar), 454 535 literaturereferences for the 2677 native Swiss plant species wereretrieved. It was determined that 55% of these species havebeen investigated phytochemically and 28% for biologicalactivity. The influence of factors such as commonness, growthform, habitat, medicinal use, and reported toxicity on theextent to which different plant groups have been studied isanalyzed. The predictive value of random sampling of subsets of plants is compared to the study of the entire Swiss flora, to showthat a credible estimate of the extent of prior studies can be achieved with just 5% of these species.

The number of known plant species in the world, withoutall the synonyms used, has been estimated to be about

300 000.1,2 About 3% of all plants are thought to have beenutilized as foods. Yet 90% of foods are derived from just 20species, and more than 50% are from rice, maize, and wheat.3

Plant biodiversity is more significant for medicinal reasons.About 70% of the world’s population rely on traditionalmedicines,4,5 which are predominantly plant based.4 Accordingto the International Union for Conservation of Nature(IUCN),6 there is well documented evidence for the medicinaluse of 28 000 plant species, which would be 9.4% if the numberof 298 000 species from www.theplantlist.org is accepted.2

However, the estimate of medicinal plant numbers by IUCN6

may be too low, as other authorities have shown that in manycountries the percentages of medicinally used plants are severaltimes higher.7 A quarter of all registered drugs in 2001 wereshown to be of plant origin,8 with 80% of all plant-derivedregistered drugs used for indications comparable to the plantsfrom which they were derived in traditional medicine.4

Despite their importance as foods and medicines and in drugdiscovery, it is unclear how many plants have actually ever beenstudied chemically and for their bioactivities, and such estimatesare contradictory. Cox and Balick in 1994 stated that “less thanhalf of 1% (of plants) have been studied exhaustively for theirchemical composition and medicinal value”.9 A brochure fromthe Missouri Botanical Garden claims that “Less than 2 percentof all plants have been thoroughly tested for medicinalapplications”.10 A University of Michigan Web site indicatesthat “Fewer than 1% of tropical plants have been screened forpossible use to medicinal science”.11 Stolton and Dudley from

the WWF stated that 5% of tropical plant species have beenexamined for their medicinal values.12 Verpoorte et al. in1998,13 relying on data taken from the scientific databaseNAPRALERT,14 showed that 36 446 (12%) of higher plantshad been studied phytochemically and 13 795 (4.6%) forbiological activity. Cragg et al. pointed out that between 1960and 1981 the National Cancer Institute (NCI) had screened114 000 extracts from 35 000 plants, showing that this oneinstitution alone had surveyed 12% of all plants for at least oneactivity.15 This list could be continued, but it is already clearthat there seems to be no general consensus on this issue.There are no universal definitions of “thoroughly studied”,“chemical composition”, or “medicinal value”; so thecontroversial numbers in the literature may all have someveracity. Tobacco, for instance, is probably the mostexhaustively studied plant chemically, as more than 2500compounds from it have been identified.16 By this standard,only a few plant species would be “exhaustively” studied.Besides, any phytochemical investigation reveals only a narrowspectrum of the overall profile of the secondary metabolites aplant may produce,17 and even tobacco will continue to yieldmore compounds in the future.The Dictionary of Natural Products18probably the most

comprehensive databank of natural productscontains 243000 entries and is updated with about 10 000 new compoundsevery year. However, due to the way data are organized, this

Received: October 2, 2012Published: January 29, 2013

Article

pubs.acs.org/jnp

© 2013 American Chemical Society andAmerican Society of Pharmacognosy 209 dx.doi.org/10.1021/np300682p | J. Nat. Prod. 2013, 76, 209−215

databank has no complete listing of all biological sources, whichmakes it inappropriate to extract a list of studied plants. Thedatabase NAPRALERT,14 on which Verpoorte13 relied, claimsto cover literature comprehensively from at least 1975 to 2003,but only about 15% of the literature from 2004 onward. Thismakes it insufficient for an up-to-date statistical analysis,because of the significant increase of phytochemical andbioactivity studies on plants in the past decade. The numberof publications searchable on SciFinder Scholar (ChemicalAbstracts Services, Columbus, OH, USA),19 which containssearch terms indicating phytochemical studies, has doubledsince 2004. For example, of the 49 547 references containingthe search term “flavonoid”, 28 867 were published since 2003;for “essential oil” it is 22 001 out of 41 411 (data retrieved inSeptember 2012).To establish a scientifically sound estimate of how many

plants have been studied (and to what extent), a directapproach has been taken in the present investigation, in which acomplete literature search was undertaken of every single plantspecies, to retrieve all the literature and to analyze this as towhether it refers to the phytochemistry or bioactivity, or is notrelevant in this context. The literature search was performedusing SciFinder Scholar,19 which is the largest scientificdatabase currently available, with more than 33 millionreferences. To do this on a global scale would have beenhardly feasible. Instead, a case study for the flora of onecountrySwitzerlandwas performed. Despite its tiny size(41 285 km2) Switzerland is botanically quite diverse. On just0.4% of the European landmass, it harbors about 20% of thecontinent’s plant biodiversity.20 Endemism levels are low (justtwo species, Draba ladina Braun-Blanq. and Arenaria bernenisFavarger),20 and there are large overlaps in the speciescomposition with the other countries in the region, a resultof relative recent colonization after the last Ice Age. Switzerlandis also a suitable country for this kind of survey, becausebotanical mapping data and data concerning plant distribution,commonness, toxicity, growth forms, and other factors werereadily available and complete.20−23 The present survey ofSwiss plants should serve also as a case study to demonstratethat analyzing a randomly selected subset from plant lists maybe suitable to predict the outcome of a comprehensive study ofthe global flora.

■ RESULTS AND DISCUSSIONAll 2677 higher plants native to Switzerland as described by thenational institution responsible for the mapping of the Swissflora24 were searched via SciFinder Scholar. A total of 454 535literature references were retrieved, sighted, and categorized asto whether they dealt with the plant’s chemistry or its biologicalactivities, or neither of these. Within these categories, thereferences were assigned to subcategories. In the followinganalysis it is considered mostly whether or not plant taxa haveever been studied at all and only briefly to what extent. Thisinformation can be retrieved from the Supporting Information(Table S1). Where pairs of numbers are given (such as 55%/28%), the first number refers to phytochemical studies and thesecond to bioactivity studies. In summary, 55% of Swiss plantshave been studied phytochemically, 42% have had isolatedcompounds, 19% have been studied for their fatty acidcomposition, 15% have had their essential oils analyzed, and9% have been studied for phytosterols. About 28% have beentested for bioactivity, 24% have been studied using in vitropharmacological methods, 17% have been studied in vivo, 16%

have been tested for antimicrobial properties, and 4% have beenstudied clinically.The level of investigation varies greatly between plants, and

only a very few species have been studied extensively, as definedby more than 10 studies in all categories and subcategories.These plants are Cannabis sativa L., Digitalis purpurea L.,Hippophae rhamnoides L., Hypericum perforatum L., Petasiteshybridus (L.) Gaertn., Rhodiola rosea L., Secale cereale L.,Silybum marianum (L.) Gaertn., and Viscum album L. About13%/7% of plants have been documented in more than 10studies in the main categories phytochemistry/bioactivity,whereas 15%/7% were reported on in just one study each.The remainder of the studied plants have between two and 10reports (Figure 1a and b).

About 3.2% (87) of native Swiss plant species are ferns(Pteridophyta).20 Of these, 46 (53%) have been studiedphytochemically and 19 (22%) for bioactivity. Among the 2589seed plants (Spermatophytina) all 10 species of gymnosperms(Pinophyta) have been studied phytochemically and seven forbioactivity. Of the 2580 angiosperms 1401 (54%) have been

Figure 1. Number of species listed according to how many studies (0,1−5, 5−9, and ≥10) were found on their phytochemistry, and thesubcategories of phytochemistry (a), and for bioactivity and thesubcategories of bioactivity (b).

Journal of Natural Products Article

dx.doi.org/10.1021/np300682p | J. Nat. Prod. 2013, 76, 209−215210

studied phytochemically and 709 (27%) for bioactivity. Amongthe 385 (14%) native species of grassy plantsmembers of thefamilies Cyperaceae, Juncaceae, and Poaceae105 (27%) havebeen studied phytochemically and 33 (8.6%) for bioactivity. Incontrast, “nongrassy” plants [2292 (85.6%)] have been morestudied [1353 (59%) phytochemically and 704 (31%) bio-logically]. No grassy plants were considered to be toxic,20 andonly one (Triticum sp.) was a medicinal plant. A total of 162(6.3%) angiosperms, a gymnosperm (10%) (Taxus baccata L.),and four ferns (4.6%) were considered toxic.20

For categorizing the extent of study, the use of the family asthe basic unit is problematic, because the number of speciesvaries greatly within families. In this study, 2677 species arereported from 154 plant families. However the 10 most species-rich families (in decreasing order of number of species,Asteraceae, Poaceae, Brassicaceae, Rosaceae, Fabaceae, Cyper-aceae, Caryophyllaceae, Ranunculaceae, Apiaceae, and Lam-iaceae) make up more than half of all native species, while 43(28%) of the families have just one, 18 (12%) have two, and 20(13%) have three species. If only the percentages of speciesstudied per family were to be shown, then small families inwhich 1 out of 1, 2/2, or 3/3 species had been studied would allbe 100% investigated and overwhelm the species-rich families.A more differentiated approach was suggested by Moerman25

when studying the numbers of plants used medicinally in NorthAmerica using regression analysis and residual values. Thisapproach was adopted for the present study. Figure 2a showsthe regression analysis and all plant families concerning the

numbers of species studied phytochemically. Figure 2b presentsthe same analysis for bioactivities. The 45 most species-richplant families are shown with their residual values (positive ornegative) in Table 1.The floristic guide Flora Helvetica20 lists 169 (6.3%) plant

species as being toxic. Of these, 146 (86%) have been studiedphytochemically and 61 (36%) for bioactivity. From theremaining 2508 species not listed as toxic, 1312 (52%) havebeen studied for their phytochemical constituents and 629(25%) were studied for bioactivity. The Flora Helvetica20 alsodifferentiates three levels of toxicity: toxic (109), more toxic(28), and highly toxic (32). Of these, 91 (83%), 28 (100%),and 27 (84%), respectively, were studied phytochemically and63 (58%), 23 (82%), and 20 (63%) were studied for bioactivity.Just a small number (n = 45, 1.7%) of native plants are listed

in the Swiss Pharmacopoeia (Pharmacopoeia Helvetica 1995)26 asmedicinal plants. Ten (19%) medicinal plants are categorized astoxic, whereas 159 (6.4%) of nonmedicinal plants are listed astoxic. All medicinal plants have been studied phytochemically,and 41 (91%) were studied for bioactivity.Plants were classified for their commonness in five categories

according to Aeschlimann and Burdet,23 which distinguishesthe categories “very rare” (n = 149), “rare” (n = 602), “notcommon” (n = 784), “common” (n = 925), and “very common”(n = 84). The percentages of species studied (phytochemistry/bioactivity) were 30%/16%, 43%/20%, 52%/22%, 63%/33%,and 91%/63%, respectively (Figure 3), and thus correlatedpositively with the degree of commonness.The Flora Helvetica20 offers a habitat description for all

species in the following eight categories: fertilized meadowplants (n = 72), cultured plants (n = 41), weeds or ruderalplants (n = 552), forest plants (n = 473), dry meadow plants (n= 328), marsh plants (n = 312), water plants (n = 112), pioneerplants (n = 134), and mountain plants (n = 622). Thepercentages of species studied in the different habitats were asfollows: fertilized meadows, 87%/51%; cultured plants, 80%/63%; weeds and ruderal plants, 70%/42%; woodland plants,60%/36%; dry meadows, 54%/22%; marshlands, 51%/24%;water plants, 51%/26%; pioneer plants, 50%/22%; mountainplants, 34%/9.6%, respectively. The category “mountain plants”from Lauber and Wagner20 was found to accommodate toomany plants (n = 622) for a differentiated analysis. In order todistinguish between plants growing generally in mountain areasand those that occur only in high altitude habitats, the groupwas subdivided according to the temperature values in Lauberand Wagner.20 In this way, an additional category of high-altitude plants (n = 201), which all had a temperature value of 1(indicating that they thrive only at high altitudes above the treeline), was formed. For high-altitude plants, the percentages ofspecies studied were 24%/2%, and thus were lower than formountain plants in general.Native plants were analyzed according to their growth forms

as listed in Lauber and Wagner.20 The largest group was herbs(n = 2327), followed by shrubs (131), trees (88), aquatic plants(74), and dwarf shrubs (54). Trees (86%/73%) were the best-studied group, followed by shrubs (57%/34%) and dwarfshrubs (59%/31%), herbs (53%/26%), and finally aquaticplants (47%/19%).Once a full data set for all plant species was assembled,

subsets of plants were randomly selected, and the analysis wasrepeated. This was done three times for every 10th (267species), 20th (133), and 40th (67) plant in the list, startingonce at number one, once at three, and once at five. Figure 4

Figure 2. Plot of the number of plant species per family versus thenumber of species of each family that have been investigatedphytochemically (a) and for bioactivity (b), alongside the trend linefor all families. The 13 plant families with more 50 species are labeled.

Journal of Natural Products Article

dx.doi.org/10.1021/np300682p | J. Nat. Prod. 2013, 76, 209−215211

shows the results of this analysis with the standard deviationsobtained. The analysis of all plants showed that 55% and 28%of the native plants have been studied for phytochemistry andbioactivity, respectively. The representative analysis of 10% ofthe data showed the data to be 54% and 29%, respectively. Theanalysis of 5% of the species gave 55% and 29%, and an analysisof just 2.5% of the data revealed an average of 60% and 27%,respectively, as having been studied previously.Over the years, discrepancies in published estimates of the

numbers of plants that are presumed to have been investigated

chemically or for their bioactivity have become moreevident.9−13 In this study an attempt has been made to definethe terms “phytochemically studied” and “studied forbioactivity”. To obtain a credible estimate of the true numbersof “studied” plants, a systematic search strategy was applied in acase study for the Swiss native flora. The analysis shows that asubstantially larger proportion of Swiss plants has been studiedthan postulated for the global flora. These results should not beviewed as representative figures for other geographic areas, butas a case study for the application of this systematic approach.The compilation of a full data set for all species within this

Table 1. The 45 Most Species-Rich Plant Families with Their Numbers of Species Native to Switzerland and the Residual ValuesDetermineda

familyspeciesnumber

residual valuephytochemistry

residual valuebioactivity family

speciesnumber

residual valuephytochemistry

residual valuebioactivity

Asteraceae 300 47.45 14.31 Amaranthaceae 27 9.46 3.97Poaceae 216 −49.24 −24.11 Onagraceae 27 0.46 −1.03Brassicaceae 148 −15.37 −2.25 Euphorbiaceae 26 5.99 9.2Rosaceae 146 −9.31 16.22 Violaceae 25 −9.48 −2.56Fabaceae 138 15.91 5.08 Scrophulariaceae 24 6.04 0.67Cyperaceae 132 −32.93 −26.52 Crassulaceae 24 0.04 −0.33Caryophyllaceae 106 −8.21 −8.45 Potamogetonaceae 23 −2.43 −2.1Ranunculaceae 91 6.7 3.04 Valerianaceae 21 −4.37 −2.63Apiaceae 87 20.81 17.98 Geraniaceae 21 1.63 6.37Lamiaceae 76 31.61 32.54 Ericaceae 19 3.68 3.83Orchidaceae 67 −22.64 −11.36 Caprifoliaceae 18 5.21 4.07Orobanchaceae 62 −6 −11.19 Amaryllidaceae 17 2.74 4.3Plantaginaceae 60 8.05 −4.73 Papaveraceae 16 4.26 4.53Boraginaceae 41 4.08 −0.3 Aspleniaceae 16 −0.74 −2.47Primulaceae 38 −7.34 1.4 Dipsacaceae 16 −3.74 −2.47Polygonaceae 37 11.19 9.64 Dryopteridaceae 14 −0.68 −1Rubiaceae 37 −0.81 −4.36 Clusiaceae 11 3.9 4.7Juncaceae 37 −16.81 −8.36 Iridaceae 11 −2.1 −1.3Salicaceae 36 −6.29 −0.13 Betulaceae 10 4.43 3.93Liliaceae 34 0.77 1.34 Solanaceae 10 3.43 4.93Gentianaceae 34 2.77 −3.66 Equisetaceae 10 3.43 3.93Campanulaceae 32 −6.18 −8.2 Polygalaceae 10 0.43 −2.07Saxifragaceae 28 −10.07 −7.26

aResidual values result from subtracting the actual percentage of studied plants from the value predicted from averaging all families in a regressionanalysis.

Figure 3. Plants categorized according to their commonness based onAeschlimann and Burdet23 and percentages of each group that havebeen studied phytochemically (lighter columns) and for bioactivity(darker columns).

Figure 4. Percentages of plants studied phytochemically (lightercolumns) and for bioactivity (darker columns) according to theanalysis of the full data set, as well as predicted when just 10%, 5%, and2.5% of the data set were analyzed randomly. Percentages are shownwith standard deviations from three replicates.

Journal of Natural Products Article

dx.doi.org/10.1021/np300682p | J. Nat. Prod. 2013, 76, 209−215212

defined region provided an absolute measure to compare withthe predicted values provided by randomly sampled represen-tative subsets. As shown in Figure 4, the numbers of speciesstudied predicted by analyzing just 5% and 10% of the initial2677 species delivered good approximations of the actualoutcome with deviations of just a few percent. Thus, it can bereasonably anticipated that a representative analysis, followingthis approach, of perhaps 10% of the world’s approximately300 0001,2 higher plant species would provide a more credibleand differentiated estimate of how many plants have beenstudied chemically and for their bioactivity. This would not bean easy task, as it would be necessary to come to terms with alltaxonomic synonyms used, but this could be done using plantdatabases.1 Ultimately, a scientifically more sound estimate onthe study of the global flora could be obtained than has beenproposed earlier.13−17

The requirement for a study to be included in the presentliterature search was that it had to be searchable by SciFinderScholar, an extensive search engine accessing the ChemicalAbstracts and PubMed databases.19 However, it is stillunreasonable to expect to find every study ever done in thesedatabases. No search strategy can access unpublished workdone by companies, or all academic theses and dissertationsdone generations ago. Test samples showed that some relevantstudies were missed. For example, according to SciFinderScholar, the medicinal plant Verbascum phlomoides L. has beenthe subject of many phytochemical studies, but not on itsbiological activities, excluding one investigation on antioxidanteffects. In Wichtl,27 however, there are references to reportedexpectorant activities that SciFinder Scholar had not included.To categorize plants as “studied” or “not studied”, as has

been done here (and as has been used by various author-ities9−15), is an oversimplification of a complex matter. Thesecategories are surrogate parameters, which allow certainanalyses. However, they do not always depict the underlyingsituation very well. Sometimes a single study can significantlychange a statistic, as exemplified by Orobanche. The 20relatively rare species of this genus undergo a subterraneanparasitic life with sporadically emerging sprouts. Half of theSwiss Orobanche species have been studied phytochemically,which gives the impression of a relatively well-studied genus.However, nine of the species (four exclusively) were covered ina single chemotaxonomic study on fatty acids and tocochro-manols present.28

The taxonomic particularities of a plant per se are probablyrarely the driving force to spark scientific endeavor, but in somecases individual investigators have dedicated much effort to thestudy of a particular taxonomic group. A specific example is theGentianaceae family, which appears among the best-inves-tigated taxa in the present study. Detailed analysis of theretrieved literature has revealed that the Swiss phytochemistsKurt Hostettmann and Andre Jacot-Guillarmod studied morethan half of the Swiss Gentianaceae.29,30 For a complete list ofGentianaceae species investigated by these authors with thecorresponding references, see Supporting Information (TableS2).It has been shown repeatedly that medicinal plants are over-

represented in some families and under-represented inothers.25,31 It is also these “medicinal plant families” such asthe Apiaceae, Asteraceae, Fabaceae, and Lamiaceae that havebeen studied most intensely, as shown by a regression analysiswith residual values (Figure 1, Table 1). This analysis methodwas chosen because it simplifies a complex phenomenon quite

elegantly. For a lively discussion on its suitability, the readershould refer to the contribution by Weckerle et al.32

Grasses, defined here as members of the Cyperaceae (132species), Poaceae (216), and Juncaceae (37), are usually nottoxic and only rarely used as medicinal plants. Furthermore,only few specialists are able to distinguish between grassy plantspecies. As a consequence, grassy plants are less than half aswell studied as nongrassy plants. No compound has ever beenreported from a native Carex species (Cyperaceae), eventhough this is the largest Swiss genus, with 91 species. Otherpoorly studied, species-rich families are usually taxonomicallycomplex or rare families such as the Caryophyllaceae (106), theOrchidaceae (67), and the Orobanchaceae (62). The nativeSwiss species from 11 entire families, the Agavaceae,Elatinaceae, Hyacinthaceae, Isoetaceae, Juncaginaceae, Linder-niaceae, Montiaceae, Najadaceae, Plumbaginaceae, Scheuchzer-iaceae, and Theophrastaceae, have never been studied at allwith respect to their phytochemistry and bioactivity.Medicinal plants have attracted particular attention. Most of

the best-studied plants in this survey have medicinal uses, suchas Cannabis sativa L., Digitalis purpurea L., Hippophae rhamnoides L., Hypericum perforatum L., Petasites hybridus(L.) Gaertn., Rhodiola rosea L., Silybum marianum (L.) Gaertn.,and Viscum album L. All official medicinal plants have beeninvestigated phytochemically, and 91% of them examined forbioactivity. Many more Swiss native plants have been usedmedicinally than are listed in the Swiss Pharmacopoeia,26 butthis criterion was chosen to create a clearly defined category.Plants considered toxic are also among the best studied as

found in this survey. This probably reflects the interesthistorically paid to plant toxins as a source of drugs and asmolecular tools for the exploration of biochemical processes. Inthis category it was decided to include plants as being toxic onlywhen listed as such in Lauber and Wagner.20

Fragrant plants also have been extensively studied, androutine methods such as gas chromatography (GC) have madethe characterization of the composition of essential oilsrelatively straightforward. To date, 15% of Swiss native plantshave been studied for their essential oil composition, and plantfamilies that typically accumulate essential oils, such as theApiaceae and Lamiaceae, are among the best-studied familieschemically.A plant may attract attention by being very common, large,

or obvious, while rare, small, or inconspicuous species may beoverlooked. Indeed, a clear correlation has been observedbetween commonness and the percent of species investigated.This, however, does not exlude that some very common speciessuch as Phyteuma spicatum L. or Hieracium sylvaticum (L.) L.have never been studied. Even some common trees are yetunstudied, such as the field maple (Acer campestre L.) or thewild pear (Pyrus pyraster Burgsd.). The categories of common-ness shown here should, however, be treated with somecaution, because a plant rare in Switzerland may not necessarilybe rare elsewhere.In addition, habitat may be a predictor of the likelihood of

prior investigation. Plants from familiar and readily accessiblehabitats, such as fertilized meadows, or cultured plants, “weeds”,and ruderal plants are among the best-studied groups. Themore remote and rare the habitat is, the less their plants havebeen studied, and plants from the highest mountain rangeswere found to be the least studied in this survey.The analysis of plants according to their growth forms shows

that trees are the best studied group, with 86% and 73%

Journal of Natural Products Article

dx.doi.org/10.1021/np300682p | J. Nat. Prod. 2013, 76, 209−215213

investigated phytochemically and biologically, respectively.Trees are large and long-lived and represent only a smallnumber of species (88) in Switzerland compared to otherregions of the world. Less well studied are shrubs, dwarf shrubs,herbs, and aquatic plants. Aquatic plants contain a number ofvery complex genera, such as Potamogeton (Potamogetonaceae,21 species), which are difficult to distinguish at a species level.It must be noted that in some taxa the number of acceptedspecies varies greatly depending on which taxonomic authorityone follows. The taxonomically complex genus Rubus(Rosaceae), for example, may contain as many as 20 nativeSwiss species20 or as few as three.23 Hence, the differencesbetween herbs and shrubs would be less pronounced if fewerRubus species were included for analysis.

■ EXPERIMENTAL SECTIONList of Native Swiss Plants. Data supplied by the Swiss Federal

Institute for Forest, Snow and Landscape Research (WSL) wereused.24 These are based on a long-term study from 1967 to 1979 by ca.200 trained botanists, who systematically mapped the distribution ofplant taxa in regions of Switzerland over this time period.20−23 The listconsists of 3200 taxa of native plants, meaning truly indigenous orintroduced, but necessarily established and reproducing over aminimum defined period in at least one of the mapping areas. Thetaxonomy of this study follows Flora Helvetica.20 The 3200 taxa wereincorporated into a table (MS Excel, Microsoft, Redmond, WA, USA),and all subspecies were removed to give 2677 species. Apart from theirbotanical name and family, all plants were categorized as to whetherthey were listed as toxic or as medicinal plants, as were their growthforms and habitats according to the Flora Helvetica.23 Finally, thecommonness of each plant was documented according to Aeschlimannet al.23 No data were available for a few species with regard to growthform, habitat, or commonness. These species were excluded from thecorresponding statistics.Criteria for a Plant to be Studied. The criterion for a plant

species to be considered phytochemically studied is that, in theliterature searchable by SciFinder Scholar,19 there was at least onedisclosed structure of a constituent of this plant species. Reportsmeeting this criterion were further subcategorized as follows: isolationof phytochemicals, essential oil analysis by gas chromatography,compounds identified by comparison with references using chromato-graphic methods (TLC, GC, or HPLC), fatty acid and lipid analysis byGC, and phytosterol analysis.The criterion for plant species to be considered studied for

bioactivity was that literature was retrievable by SciFinder Scholar inwhich a defined bioactivity meeting the criteria below had beenreported. It was considered irrelevant if activity was shown for purifiedcompounds, chromatographic fractions, or extracts. The subcategorieswere as follows: in vitro studies such as results from assays onmolecular targets (enzymes, receptors) or lower organisms (other thanmicrobes), antimicrobial assays, in vivo studies, and clinical studies. Invitro antioxidant activity or radical scavenging was not considered as aform of bioactivity, but included as an additional category. If bioactivityhad been reported for a compound from one plant, and the compoundwas known from another plant as well, then the activity was attributedonly to the plant mentioned in the study. For example, if a commoncompound such as β-sitosterol has been reported from one plant andshown to be active in a bioassay, then other plants reported to containβ-sitosterol were not considered as “biologically studied”.Literature Search. All literature for the species in the list was

searched systematically in the scientific database SciFinder Scholar.Phytochemical studies and studies on bioactivity were identified andcategorized in the subcategories given above. If less than 250publications were retrieved for a plant species, they were all visuallyscreened, and if necessary, the full text was read. If more than 250publications were found, the search was refined using the “refine”function in SciFinder Scholar. After a period of experimenting bycomparing results from refined searches with those of systematic

screening of all results, suitable terms to help identify phytochemicaland bioactivity studies were defined. For phytochemical studies theseterms were “isolation”, “essential oil”, “volatile oil”, “phytosterol”,“sterol”, “fatty acid”, and “identification”. The most suitable terms forbioactivity results were “antimicrobial”, “antifungal”, “in vivopharmacology”, “mice”, “rats”, “in vivo”, “in vitro”, “cells”, “clinicalstudies”, “human”, “clinical”, and “activity”. The columns in Table S1in the Supporting Information represent the search categories with themain categories “phytochemistry” and “bioactivity”, with eachcontaining the data from subcategories. “Phytochemistry” was assignedthe subcategories “isolation”, “fatty acid composition”, “essential oils”,and “phytosterols”. “Bioactivity” included the subcategories “in vitropharmacological studies”, “in vivo studies”, “antimicrobial”, and“clinical”. If more than 10 studies were counted in one subcategory,then a value of >10 was entered. If a study contained data from morethan one subcategory, then this was entered in several subcategories,but treated as just one study in the main categories “phytochemistry”and “bioactivity”. If there were no results on a plant species, thedatabank http://www.tela-botanica.org was searched for synonyms,and the search repeated for all of these. All these databank searcheswere done between January 16 and May 30, 2011.

■ ASSOCIATED CONTENT*S Supporting InformationTable with the results of the literature search for all the plantsincluding their botanical authorities and the number of studiesfound for each category and subcategory, as well as a table withthe Gentianaceae species investigated by Hostettmann andJacot-Guillarmod and the corresponding references. Thismaterial is available free of charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*Tel: +41 (0) 61 267 15 34. Fax: +41 (0) 61 267 14 74. E-mail:[email protected].

NotesThe authors declare no competing financial interest.

■ REFERENCES(1) The Plant List. http://www.theplantlist.org/about/. AccessedMay 29, 2012.(2) Chapman, A. D. Numbers of Living Species in Australia and theWorld, 2nd ed.; Australian Biodiversity Information Services:Toowoomba, Queensland, Australia, 2011.(3) Etkin, N. L. Edible Medicines - An Ethnopharmacology of Food; TheUniversity of Arizona Press: Tucson, AZ, 2007; p 28.(4) Fabricant, D. S.; Farnsworth, N. R. Environ. Health Perspect. 2001,109, 69−75.(5) Farnsworth, N. R. In Biodiversity; Wilson, E. O., Ed.; NationalAcademy Press: Washington, DC, 1988; pp 83−97.(6) International Union for Conservation of Nature (IUCN), 2012,http://cmsdata.iucn.org/downloads/mpc_15_1.pdf. Accessed Decem-ber 21, 2012.(7) Barboza, G. E.; Cantero, J. J.; Nunez, C.; Espinar, P. L. A.Kurtziana 2009, 34, 7−365.(8) Rates, S. M. K. Toxicon 2001, 39, 603−613.(9) Cox, P. A.; Balick, M. J. Sci. Am. 1994, June, 82−87.(10) ht tp ://www.mobot .o rg/ tour s/med ic ina l_p l an t s/Medicinal%20Brochure%2008%20final.pdf. Accessed May 30, 2012.(11) http://www.globalchange.umich.edu/globalchange2/current/lectures/deforest/deforest.html. Accessed May 30, 2012.(12) Stolton, S.; Dudley, N. 2010. Vital Sites: The Contribution ofProtected Areas to Human Health. WWF. Available at http://www.assets.panda.org/downloads/vital_sites.pdf. Accessed June 25, 2012.(13) Verpoorte, R. Drug Discovery Today 1998, 3, 232−238.

Journal of Natural Products Article

dx.doi.org/10.1021/np300682p | J. Nat. Prod. 2013, 76, 209−215214

(14) NAPRALERT. http://www.napralert.org/. Accessed May 30,2012.(15) Cragg, G. M.; Boyd, M. R.; Cardellina, J. H., II; Newman, D. J.;Snader, K. M.; McCloud, T. G. Ciba Found. Symp. 1994, 185, 178−190.(16) Nugroho, L. H.; Verpoorte, R. Plant Cell Tiss. Org. 2002, 68,105−125.(17) Hamburger, M.; Hostettmann, K. Phytochemistry 1991, 30,3864−3874.(18) Buckingham, J., Consultant Ed. Dictionary of Natural Products onDVD; Chapman and Hall/CRC: London, UK, 2012.(19) SciFinder Scholar. https://scifinder.cas.org. Accessed May 29,2012.(20) Lauber, K.; Wagner, G. Flora Helvetica; Verlag Paul Haupt:Bern, 2003.(21) Welten, M.; Sutter, R. Verbreitungsatlas der Farn- undBlutenpflanzen der Schweiz; Birkhauser: Basel, 1982.(22) Welten, M.; Sutter, R. Erste Nachtra ge und Erga nzungen zuVerbreitungsatlas der Farn- und Blu tenpflanzen der Schweiz; Birkhauser:Basel, 1984.(23) Aeschlimann, D.; Burdet, H. M. Flore de la Suisse et desTerritoires Limitrophes - Le Nouveau Binz; Haupt Verlag: Bern, 2008.(24) http://www.wsl.ch/land/products/webflora/floramodul5-de.html. Accessed January 16, 2011.(25) Moerman, D. E. J. Ethnopharmacol. 1979, 1, 111−119.(26) Pharmacopoeia Helvetica VII; Departement Federal del’Interieur: Bern, 1994.(27) Wichtl, M.; Bauer, R.; Blaschek, W.; Burr, W.; Hiller, K.; Lichius,J. J.; Loew, D.; Stahl-Biskup, E.; Teuscher, E. Teedrogen undPhytopharmaka: Ein Handbuch fu r die Praxis auf WissenschaftlicherGrundlage, 5th ed.; Wissenschaftliche Verlagsgesellschaft: Stuttgart,2008; p 698.(28) Velasco, L.; Goffman, F. D.; Pujadas-Salva, A. J. Phytochemistry2000, 54, 295−300.(29) Goetz, M.; Jacot-Guillarmod, A. Helv. Chim. Acta 1978, 61,1373−1375.(30) Urbain, A.; Marston, A.; Mardsen-Edwards, E.; Hostettmann, K.Phytochem. Anal. 2009, 20, 134−138.(31) Moerman, D. E. Native American Ethnobotany; Timber Press:Portland, OR, 1998.(32) Weckerle, C. S.; Cabras, S.; Castellanos, M. E.; Leonti, M. J.Ethnopharmacol. 2011, 137, 837−843.

Journal of Natural Products Article

dx.doi.org/10.1021/np300682p | J. Nat. Prod. 2013, 76, 209−215215