Tanacetum Vulgare L.) Related to Genetic and Morphological Diﬀerences of Genotypes

Biochemical Systematics and Ecology 29 (2001) 267–285

Variation in volatile compounds from tansy (Tanacetum vulgare L.) related to genetic and morphological diﬀerences of genotypes

Marjo Keskitaloa,b,*, Eija Pehua, James E. Simonb a Department of Plant Production, P.O. Box 27, University of Helsinki, FIN-00014, Finland b Department of Horticulture, Center for New Crops and Plant Products, Purdue University, West Lafayette, IN 47906-1165, USA

Received 18 January 1999; accepted 15 February 2000

Abstract

Air-dried flower heads of 20 Finnish tansy genotypes were extracted with petroleum ether and analyzed using GC–MS. A total of 55 volatile compounds were detected, and 53 were identified. Of the tansy genotypes studied, 15 were well defined and five were mixed chemotypes. Complete linkage analysis differentiated the populations into six clusters. The most frequently found monoterpene was camphor with or without several satellite compounds such as camphene, 1,8-cineole, pinocamphone, chrysanthenyl acetate, bornyl acetate and isobornyl acetate. In 13 genotypes, camphor concentration exceeded 18.5% and in seven genotypes, camphor was less than 7.2%. Other chemotypes rich in trans thujone, artemisia ketone, 1,8-cineole, or davadone-D were also identified. Davadone-D and a mixed chemotype, containing tricyclene and myrcene, were identified from a Finnish tansy for the first time. Geographically, most chemotypes containing camphor originated from Central Finland, whereas chemotypes without camphor such as artemisia ketone, davadone D and myrcene– tricyclene originated from South or Southwest Finland. Morphologically, the 20 tansy chemotypes based on the groups formed from complete linkage cluster analysis, were compared. The group containing the highest concentration of camphor chemotypes had the tallest shoots. The groups consisting from chemotypes containing davadone-D or artemisia ketone, which originated from Southwest Finland, produced the highest number of flower heads, had the tallest corymb, and were last to flower. Also, the group consisting from chemotypes with a high concentration of camphor and originated from South Finland started to flower late. The correlation between the genetic distance matrices based on RAPD patterns

*Corresponding author. Present address: Plant Production Research, Agricultural Research Center of Finland, FIN-31600 Jokioinen, Finland. Fax: +358-3-4188-2437. E-mail address: marjo.keskitalo@mtt.ﬁ (M. Keskitalo).

0305-1978/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 3 0 5 - 1978(00)00056-9 268 M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 reported previously (Keskitalo et al., 1998. Theor. Appl. Genet. 96, 1141–1150.) and the chemical distance matrices of the present study of the same tansy genotypes was highly signiﬁcant (0.41, P50:0001). # 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Asteraceae; Davadone-D; Essential oils; Intraspeciﬁc variation; Monoterpenes; Multivariate analysis; Petroleum-ether extraction; Tanacetum vulgare L.

1. Introduction

Tansy is an aromatic perennial plant, widely spread in the northern hemisphere (Heywood, 1976; Hussey, 1974; Hulte´ n, 1968). Occasionally, it has been cultured in gardens (Mitich, 1992; Heywood, 1976) and used in salads, omeletts, cakes and spices (Grieve, 1984; Hussey, 1974). As a herbal remedy, tansy has traditionally been used in balsams, cosmetics, dyes, insecticides, medicines, and preservatives (Grieve, 1984; Hussey, 1974; Millspaugh, 1974) and as anthelmintic, for migrane, neuralgia, rheumatism and loss of appetite (Blumenthal, 1998), though the effectiveness for these uses have not been documented. Recent studies have also shown that the essential oil or extract of tansy exhibits anti-inflammatory (Brown et al., 1997; Mordujovich-Buschiazzo et al., 1996), antibactericidal (Neszme´ lyi et al., 1992; He´ thelyi et al., 1991; Holopainen and Kauppinen, 1989; Stefanovic et al., 1988), antifungicidal, (Neszme´ lyi et al., 1992; He´ thelyi et al., 1991), and a repellent against insects (Hough-Golstein and Hahn, 1992; Nottingham et al., 1991; Suomi et al., 1986; Panasiuk, 1984; Schearer, 1984). The activity against microbes and insects was dependent on the chemical composition of the essential oil (He´ thelyi et al., 1991; Holopainen and Kauppinen, 1989; Panasiuk, 1984; Schearer, 1984). Components in the essential oil are also of potential interest as aroma chemicals in perfumery (Lawrence, 1992). The composition of tansy oils varies markedly and several chemotypes from different geographical origins have already been classified. Tansy from Tierra del Fuego, Argentina (Gallino, 1988), and from England and USA (Ekundayo, 1979) were found to contain mainly b-thujone. Tansy oil obtained from two Canadian-grown plants contained b-thujone (d-isothujone) (Collin et al., 1993; von Rudolf and Underhill, 1965), a-thujone (l-thujone) (von Rudolf and Underhill, 1965), camphor-1,8-cineole-borneol, chrysanthenone, or dihydrocarvone (Collin et al., 1993) as the major oil components. In contrast, chemotypes such as artemisia alcohol, camphenol, davadone, lyratol, lyratyl acetate and 4-thujen- 2a-yl (He´ thelyi et al., 1981; Te´ te´ nyi et al., 1975), and trans-chrysanthenyl acetate (Neszme´ lyi et al., 1992) have been detected in tansy grown in Hungary. Naturally occurring tansy from northeastern Netherlands contained artemisia ketone, chrysanthenol/chrysanthenyl acetate, lyratol/lyratyl acetate and b-thujone (Hendriks et al., 1990). Chrysanthenyl acetate and camphor/b-thujone chemotypes were identified in tansy grown in Belgium (De Pooter et al., 1989). In contrast, only one chemotype, chrysanthenyl acetate, was detected from tansy grown in M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 269

Piedmont, Italy (Nano et al., 1979). Several authors have studied the intraspecific variation of essential oil in Finnish tansy. Sorsa et al. (1968) found that a-pinene, b-pinene, 1,8-cineole, g-terpinene, artemisia ketone, thujone, camphor, umbellulone and borneol were the most common chemotypes. They also observed that camphor was the main component in northern, and thujone in southern Finland (Sorsa et al., 1968). Later, isopinocamphone (Forse´ n and von Schantz, 1971), trans-chrysanthenyl acetate (Forse´ n, 1974), sabinene, bornyl acetate, and germacrene D (Holopainen, 1989) chemotypes were detected from tansy grown in Finland. Tansy also contains other secondary products such as less volatile sesquiterpenes (Chandra et al., 1987b) and several nonvolatile compounds. The latter group consists monoterpene glucosides (Banthorpe and Mann, 1971); sesquiterpene lactones (Sanz and Marco, 1991; Ognyanov and Todorova, 1983) such as parthenolide (Hendriks and Bos, 1990); sesquiterpene diketone (Ognyanov et al., 1983); triterpenes (Wilkomirski and Kucharska, 1992); and flavonoids (Chandra et al., 1987a; Ognyanov and Todorova, 1983). In addition to different tansy chemotypes of which the concentration of one compound exceeds 40% (as a definition for well-defined chemotypes; Holopainen, 1989), more than 100 volatile mono- and sesquiterpenes have been identified from tansy by the authors including those compounds reported in this paper. Many factors are known to influence oil composition, including source of extractable plant tissue (leaf/flower), ontogeny of plant at sampling (Hendriks et al., 1990; Holopainen, 1989), seasonal changes (Ne´ meth et al., 1994; von Schantz et al., 1966; von Rudolf and Underhill, 1965), and even the extraction method (Collin et al., 1993; Holopainen, 1989) may cause variation in the composition of volatile oil. We are interested in to enhance the insecticidal properties of tansy. Thus, we have developed micropropagation, regeneration and protoplast fusion methods (Keskitalo et al., 1999, 1995) for the improvement of secondary metabolism by genetic engineering. We also studied genetic variation of 20 tansy genotypes collected from different regions of Finland by RAPD-PCR. Genetically, the genotypes clustered into two major groups, which were further divided into smaller groups, and the clustering was correlated to the geographical origin of the genotypes. Morphologically, the two clusters differed in their flowering (Keskitalo et al., 1998). To ascertain whether there is a connection between the chemical and genetic differences of tansy, we extracted the volatile compounds from the same tansy genotypes and analyzed these compounds with GC–MS. The goals of this study were two fold: (1) to identify the volatile terpenes extracted from air-dried tansy flower heads; and (2) to compare the variation to the geographical origin of the genotype to link the genetic and morphological variation among the tansy genotypes. Assessment of the chemical variation of tansy would allow us to identify unique chemotypes with potential industrial value and bioassay the activity of the genotypes with specific terpene composition. With such information the most desirable chemotypes can be selected for genetic engineering studies. 270 M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285

2. Materials and methods

2.1. Plant materials

The geographical locations of 20 tansy genotypes (Tv 12Tv 20) collected and used in this study are listed in Table 1. The genotypes were transplanted to an orchard of the Department of Plant Production, University of Helsinki (608100N), Finland in 1991 (Keskitalo et al., 1998). The vouchers have been deposited at Botanical Museum of University of Helsinki (H). In the orchard, the plants were grown 1.5 m apart in two rows without fertilization. Morphological observations such as height of the stem, number of nodes per stem, number of flower heads per stem, length of corymb and date at the beginning of flowering were carried out as described (Keskitalo et al., 1998). The flower heads were excised at the onset of flowering, air dried at 388C, and stored in room temperature in the dark until extracted.

2.2. Isolation of volatile compounds

Two grams of air-dried ﬂower heads crushed in a mortar, were transferred to centrifuge tubes containing 5 ml methanol, shaken for 1 h, centrifuged and the supernatant was collected to another tube. The ﬂower heads were extracted with methanol 3 times. About 1–2 ml of saturated NaCl and 7 ml petroleum-ether (36–60)

Table 1 Geographical origins (National Landsurvey Institute, 1996) of the 20 tansy genotypes

Tansy genotype Region Latitude (N) Longitude (E) Voucher at (H) Collector # Keskitalo 21440

Tv 1 Iisalmi 638330N278120E H 1698296 Tv 2 Eno 628470N308080E H 1698297 Tv 3 Koli 638060N298480E H 1698298 Tv 4 Kangasniemi 618590N268390E H 1698299 Tv 5 Helsinki 608100N248570E H 1698300 Tv 6 Halikko 608230N238030E H 1698301 Tv 7 Lohjanharju 608300N248240E H 1698302 Tv 8 Tampere 618290N238480E H 1698303 Tv 9Sa¨ a¨ ksma¨ ki 618110N248030E H 1698304 Tv 10 Porvoo 608230N258400E H 1698305 Tv 11 Sipoo 608220N258150E H 1698306 Tv 12 Vihti 608250N248180E H 1698307 Tv 13 Ha¨ meenlinna 608590N248270E H 1698308 Tv 14 Lemu 608330N218590E H 1698309 Tv 15 Tervakoski 608480N248380E H 1698356 Tv 16 Lieto 608300N22827’E H 1698355 Tv 17 Hanko 598490N238000E H 1698354 Tv 18 Alavus 628350N2383700E H 1698353 Tv 19 Hailuoto 658000N248430E H 1698352 Tv 20 Ko¨ ylio¨ 618060N228200E H 1698351 M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 271 were added to the methanol extraction (15 ml), after which the solution separated into two phases. The top layer containing petroleum-ether was pipetted to an Erlenmeyer flask, and dried with anhydrous sodium sulfate in +48C in the dark for 1–3 days. The addition of NaCl and petroleum-ether was done twice. The dried supernatant was filtered and evaporated to 1 ml and stored in 2 ml vials in +48Cin the dark until the GC analysis. The extraction from the flower heads of each tansy genotype was carried out 4 times. The different extraction times are referred as A–D.

2.3. Chemical analysis of the volatile compounds

Prior to the gas-chromatography (GC) analysis, 1 ml of an internal standard (tetradecane; stock 50 mg/1 ml petroleum-ether) was added to a volume of 100 mlof the extractions of A–C. The fourth sample (D) was run without the internal standard. A sample of 1 ml was injected into the GC-Varian (Walnut Creek, CA) 3700 gas chromatograph, fitted with a flame-ionization detector and Hewlett- Packard model 3396 series II integrator (Palo Alto, CA). A fused silica capillary column (30 mÂ0.25 mm i.d., 0.20 mm film thickness, SPB-5; Supelco, Bellefonte, PA) was used. Helium was used as the carrier gas, and oven temperature was held isothermal at 808C for 5 min and then programmed to increase 38C/min to 2408C. The injector and detector temperatures were maintained at 180 and 2508C, respectively. GC/MS analyses were conducted using a Finnigan (San Jose, CA) GC (model 9610) and mass spectrometer (model 4000) system equipped with a capillary column under the same GC conditions and interfused to a Data General Nova/4 data processing system. The MS conditions included ionization voltage, 70 eV; emission current, 40 mA; scan rate, 1 scan/s; mass range, 40–500 Da; and ion source temperature, 1608C. The volatile components of the three separate extractions (A–C) in each tansy line were identified by comparing their relative retention times, retention indices (RI) and mass spectra with those of authentic samples, and by matching the mass spectra of each compound with different MS compound libraries for best fit (Adams, 1995; Finnigan et al., 1978; Stenhagen et al., 1974) and to the Kovats’ indices of authentic samples (Adams, 1995). The retention indices of our samples were calculated using n-alkanes (C8–C18)asa reference (Table 2).

2.4. Statistical analysis

The volatile components of the tansy flower heads were scored as being present (1) or absent (0) in an extracted oil as determined by GC–MS. The pair-wise similarity and distance matrices of volatile compounds of the 20 tansy genotypes were calculated using the SAS program (SAS Institute, 1984) modified by Levy et al. (1991), similarly as we reported using RAPD-data (Keskitalo et al., 1998). The similarity and distance matrixes were calculated according the formula of Nei and Li (1979): [Sxy ¼ 2Nxy : ðNx þ NyÞ]; where Nxy is the number of terpenes common for the accessions x and y;andNx and Ny are the numbers of terpenes of the accessions x and y, respectively. The dendrogram was synthetized by complete linkage cluster Table 2 272 Volatile oil concentration (%) and the mean of retention indices (RI) of volatile compounds extracted from air-dried flower heads of 20 tansy genotypes and identified by GC-MS. The results are a mean of three different extractions (A–C) carried out from each tansy genotype. The groups of tansy genotypes are formed according to the complete linkage cluster analysis based on the presence or absence of the identified compounda

Groups of tansy genotypes (Tv)

Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 267–285 (2001) 29 Ecology and Systematics Biochemical / al. et Keskitalo M.

Compound RI Tv 7 Tv 8 Tv 10 Tv 20 Tv 1 Tv 3 Tv 4 Tv 18 Tv 2 Tv 5 Tv 6 Tv 12 Tv 9 Tv 11 Tv 13 Tv 15 Tv 17 Tv 19 Tv 14 Tv 16

1 Tricyclene 912 * 26.84 16.96 0.83 2.15 0.16 0.16 0.05 0.27 0.04 0.11 0.09 0.33 0.10 0.25 0.36 0.23 2. a-Thujene 932 0.04 0.36 0.54 0.34 0.17 0.39 0.28 0.56 0.30 0.73 0.53 0.45 0.07 3. a-Pinene 940 0.46 0.08 0.32 5.72 0.70 0.67 0.78 2.27 2.90 2.07 4.44 0.61 4.69 0.57 5.78 2.71 3.87 4. Camphene 957 2.44 0.64 2.42 6.03 7.27 0.61 5.15 7.76 8.66 12.60 4.57 11.62 5.06 2.24 7.83 0.60 0.42 5. Sabinene 980 0.10 0.04 2.14 1.62 0.03 0.28 0.22 2.95 0.85 0.03 0.29 0.40 1.61 0.72 0.64 1.12 6. Artemiseole 982 0.06 7. b-Pinene 986 0.08 0.85 0.05 2.64 0.61 0.15 2.01 1.03 1.46 2.57 0.60 2.91 0.40 1.27 0.80 1.46 0.04 8. Myrcene 1002 39.39 20.92 1.13 3.93 9. Yomogi alcohol 1002 2.74 1.69 10. r-Cymene 1029 1.43 1.22 2.04 0.04 1.61 2.91 3.42 1.77 0.22 1.32 1.96 1.02 4.00 1.22 2.05 1.39 1.71 1.57 0.68 0.43 11. Limonene 1033 0.75 1.95 0.06 1.36 0.16 0.34 2.53 0.38 1.20 0.81 0.37 12. 1,8-Cineole 1038 5.14 1.56 6.61 4.71 3.21 11.43 0.15 0.41 2.60 47.12 8.21 9.92 4.79 6.62 20.23 7.55 13.73 4.68 1.59 1.80 13. Artemisia ketone 1063 67.32 0.54 55.08 81.36 14. g-Terpinene 1063 0.31 1.39 0.54 1.24 1.18 0.41 0.31 0.78 0.21 0.76 0.45 0.70 0.40 0.86 0.64 15. cis Sabinene 1079 0.02 0.75 0.34 0.55 0.19 0.08 hydrate 16. Terpinolene 1097 1.56 0.88 1.53 0.27 0.78 0.82 0.65 0.24 0.18 0.14 0.69 0.58 0.18 0.19 1.24 0.39 0.67 0.28 17. Artemisia alcohol 1098 0.06 0.13 4.27 0.12 9.31 3.78 18. cis Thujone 1116 0.86 0.07 19. trans Sabinene 1119 0.14 0.07 hydrate 20. Chrysanthenone 1121 0.12 0.93 0.09 0.83 0.03 21. trans Thujone 1129 0.43 0.07 0.91 81.87 2.12 0.25 0.06 0.04 0.04 0.07 22. Camphor 1154 1.40 7.15 1.35 0.06 3.70 19.43 69.37 72.16 70.71 29.43 50.07 73.02 50.78 68.79 48.15 65.23 18.54 35.71 5.98 0.42 23. Chrysanthenol 1164 0.69 3.07 2.00 1.12 24. Pinocarvone 1168 4.38 1.06 25. Umbellulone 1170 12.18 0.15 4.94 2.65 2.76 1.01 0.56 0.12 0.65 0.03 3.95 0.39 3.73 2.46 1.53 26. Terpinen-4-ol 1175 0.66 0.99 27. Artemisyl acetate 1176 4.35 15.26 2.45 1.73 3.45 3.01 5.81 0.11 1.39 0.60 4.17 5.01 1.58 28. Borneol 1177 0.81 3.87 5.08 0.34 29. cis Pinocamphone 1185 0.04 0.03 0.04 0.03 19.10 30. a-Terpineol 1199 0.30 1.16 1.11 0.07 0.16 4.27 0.65 1.01 0.62 0.33 0.48 31. Myrtenol 1204 10.69 32. Verbenone 1219 3.49 0.57 0.21 0.94 0.36 0.21 33. Nerol 1235 0.17 34. trans Chrysanthenyl 1249 0.87 0.10 0.02 1.02 acetate Groups of tansy genotypes (Tv)

Group 1 Group 2 Group 3 Group 4 Group 5 Group 6

Compound RI Tv 7 Tv 8 Tv 10 Tv 20 Tv 1 Tv 3 Tv 4 Tv 18 Tv 2 Tv 5 Tv 6 Tv 12 Tv 9 Tv 11 Tv 13 Tv 15 Tv 17 Tv 19 Tv 14 Tv 16 .Ksiaoe l iceia ytmtc n clg 9(01 267–285 (2001) 29 Ecology and Systematics Biochemical / al. et Keskitalo M. 35. Carvone 1250 0.97 0.65 0.81 0.90 0.81 36. cis Chrysanthenyl 1271 0.89 4.39 11.97 24.88 0.05 1.42 acetate 37. Unknown, 1286 0.07 MW 154 38. Bornyl acetate 1291 0.67 0.54 0.43 5.16 0.81 0.11 11.55 0.04 17.80 1.16 0.31 1.35 5.72 16.72 0.74 39. Isobornyl acetate 1298 0.06 12.04 7.48 40. Carvacrol 1304 0.06 2.67 41. Thymol 1306 0.58 1.37 0.86 42. d-Elemene 1336 0.65 0.66 43. b-Cubebene 1387 0.07 0.26 0.40 44. b-Elemene 1390 2.25 0.15 0.19 0.07 0.13 0.58 0.52 0.14 0.07 0.11 0.86 0.03 1.38 0.09 0.28 45. (E) Caryophyllene 1426 0.90 1.36 2.05 0.77 1.43 0.96 0.23 1.27 0.04 1.32 0.53 0.88 0.43 0.43 0.59 1.31 1.32 0.86 46. Germacrene D 1485 1.58 2.54 0.75 1.56 0.66 2.18 2.66 1.14 0.82 0.79 5.08 1.08 0.03 0.99 1.47 0.44 1.12 0.12 47. Bicycloelemene 1501 1.32 0.06 0.60 0.05 48. g-Cadinene 1525 2.64 1.87 0.86 2.92 0.39 1.13 1.03 49. Artedouglasia 1553 1.18 oxide C 50. Nerolidol 1565 2.22 0.37 0.04 3.11 1.27 1.74 0.26 51. Germacrene 1577 0.20 0.07 alcohol 52. Spathulenol 1591 0.22 0.24 53. Caryophyllene 1594 0.55 0.17 0.19 0.13 0.41 0.25 0.16 oxide 54. Davadone D 1597 65.51 0.55 1.81 55. Unknown, 1623 0.39 0.13 0.33 0.15 0.09 0.23 MW 222 Percentage of total 85.54 95.85 95.61 98.05 98.70 98.55 97.76 98.89 97.00 98.64 98.91 99.37 99.29 98.68 99.33 99.11 97.79 96.84 97.94 97.39 peak area identiﬁed Total amount (mg) 0.93 1.61 1.11 5.46 2.84 2.36 3.29 2.29 1.03 2.80 1.50 2.73 1.94 3.45 5.07 5.65 2.22 4.35 4.76 3.20 of terpenes isolated from 2.0 g ﬂower heads

a Blank areas = not detected 273 274 M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 analysis, which illustrates the chemical distance as the normalized maximum distance between clusters. The normalized maximum distance is a normalized value (distance between x and y/mean distance of the population) of the average genetic distance. Morphological data of tansy genotypes (Keskitalo et al., 1998) were arranged according to the grouping of complete linkage cluster analysis and the diﬀerences between the groups were calculated with SAS (SAS Institute, 1984) CONTRAST program. To compare the distance matrices of RAPD-PCR study of our previous study (Keskitalo et al., 1998) and the distance matrices of this present study (Table 3), Pearson correlation coeﬃcient was calculated with SAS.

3. Results

3.1. Volatile compounds identiﬁed with GC–MS

A total of 55 aromatic volatile compounds were detected from the petroleum-ether extraction of dried flower heads of tansy. The volatile compounds covered on average 97.46% (S.D. 2.69) of the total peak area recovered from GC. The concentration of 47 of the 55 compounds varied highly significantly between the tansy genotypes (P50:0001) and only the genotypic variation in artemiseole, trans sabinene hydrate, nerol, nerolidol, spathulenol, caryophyllene oxide and two unidenti- fied compound was not significant. On average, the total amount of volatile compounds extracted from 2.0 g of tansy flower heads was 2.90 mg (S.D. 1.63) (Table 2). Fifteen of the 20 genotypes had main component (usually camphor, artemisia ketone, trans thujone, or davadone D), which consisted at least 40% of the total peak area, and five genotypes contained at least two terpenes as the main components.

3.2. Complete linkage cluster analysis

The smallest chemical distance observed was 0.116 between the chemotypes Tv 5 and Tv 6 and the largest distance between Tv 7 and Tv 19 (Table 3). The mean chemical distance was 0.374 among the entire tansy population. Complete linkage cluster analysis separated the population to six groups. The largest and smallest normalized maximum distance was 1.94 (0.726/0.374=1.94) and 0.31 (0.116/ 0.371=0.31) between group 1 and the cluster of the other groups, and between Tv 5 and Tv 6, respectively (Fig. 1).

3.3. Volatile compounds related to the geographical origin, RAPD-PCR pattern and morphology

Six of the seven chemotypes, which did not contain camphor as the main component, originated from Southern Finland (Tv 7, 8, 10, 14, 16, 20), and only the chemotype containing thujone (Tv 1) was from Central Finland (Fig. 2). Eight of the 13 chemotypes with camphor concentration exceeding 18.5% originated from Central Finland (Tv 2, 3, 4, 9, 13, 15, 18, 19) and ﬁve from Southern Finland (Tv 5, 6, M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 275

Fig. 1. The clustering based on the absence or presence of 55 volatile compounds identiﬁed by GC–MS of the 20 tansy genotypes used in this study. The chemical distance between genotypes were calculated as described by Nei and Li (1979) using the complete linkage cluster analysis. The bar shows the normalized maximum distance between clusters.

11, 12, 17). Artemisia ketone (Tv 14, 16, 20) was found only from genotypes originating from Southwestern Finland. Also, geographically the six groups formed from complete linkage cluster analysis could be separated from each other. Only the groups containing a high concentration of davadone D, artemisia ketone or myrcene-tricyclene (Tv 7, 8, 10, 20 and Tv 14, 16) were found from South-Finland. The groups containing a high concentration of camphor were found from South and Central Finland, but the number of chemotypes containing a high concentration of camphor was higher in Central Finland (Fig. 2). In a previous paper, we reported the genetic distance matrices based on RAPD- PCR patterns (Keskitalo et al., 1998). The groupings of tansy population according to our previous genetic study (Keskitalo et al., 1998) and this study on tansy volatile compounds are illustrated in Fig. 2. Because the distance matrices of these both studies were calculated by the same method, we calculated the correlation between matrices of genetic data and chemical data. The Pearson correlation coefficient, 0.407 (P50:0001), revealed a 16.565% analogy between genetic and chemical differences based on absence or presence of compound in the 20 tansy genotypes studied. The tansy genotypes were arranged according to the six groups formed from complete linkage cluster analysis and the variation of morphology was compared between the groups using the SAS CONTRAST procedure. The group containing the highest percentage of camphor and 1,8-cineole (Tv 9, 11, 13, 15) had the tallest shoots (108.2 cm) and differed significantly from the others (P50:0001) whereas the 276 .Ksiaoe l iceia ytmtc n clg 9(01 267–285 (2001) 29 Ecology and Systematics Biochemical / al. et Keskitalo M. Table 3 Chemical distance of the 20 tansy genotypes based on the absence or presence of volatile compound using the method of Nei and Li (1979)

1234567891011121314151617181920

Tv 1 0.000 Tv 2 0.351 0.000 Tv 3 0.156 0.368 0.000 Tv 4 0.217 0.384 0.234 0.000 Tv 5 0.200 0.211 0.217 0.234 0.000 Tv 6 0.190 0.200 0.163 0.227 0.116 0.000 Tv 7 0.514 0.600 0.579 0.487 0.632 0.600 0.000 Tv 8 0.348 0.487 0.362 0.417 0.362 0.318 0.590 0.000 Tv 9 0.292 0.366 0.265 0.200 0.184 0.217 0.512 0.360 0.000 Tv 10 0.347 0.476 0.360 0.333 0.400 0.319 0.524 0.255 0.283 0.000 Tv 11 0.333 0.415 0.388 0.200 0.265 0.304 0.561 0.480 0.269 0.358 0.000 Tv 12 0.282 0.250 0.300 0.317 0.150 0.135 0.688 0.415 0.265 0.409 0.256 0.000 Tv 13 0.261 0.333 0.319 0.333 0.149 0.182 0.641 0.375 0.200 0.373 0.280 0.220 0.000 Tv 14 0.463 0.412 0.524 0.535 0.429 0.385 0.588 0.488 0.422 0.478 0.511 0.444 0.349 0.000 Tv 15 0.265 0.333 0.280 0.255 0.200 0.234 0.619 0.451 0.245 0.407 0.245 0.318 0.216 0.391 0.000 Tv 16 0.422 0.421 0.435 0.362 0.304 0.349 0.526 0.489 0.224 0.480 0.388 0.400 0.319 0.286 0.320 0.000 Tv 17 0.362 0.500 0.375 0.388 0.375 0.378 0.600 0.388 0.373 0.385 0.412 0.476 0.224 0.409 0.308 0.500 0.000 Tv 18 0.256 0.278 0.273 0.156 0.182 0.171 0.556 0.378 0.277 0.375 0.277 0.263 0.289 0.450 0.208 0.318 0.391 0.000 Tv 19 0.364 0.405 0.333 0.348 0.333 0.238 0.730 0.435 0.417 0.469 0.417 0.385 0.348 0.512 0.265 0.511 0.362 0.256 0.000 Tv 20 0.526 0.484 0.538 0.600 0.538 0.444 0.613 0.400 0.571 0.442 0.619 0.636 0.550 0.543 0.488 0.538 0.561 0.514 0.474 0.000 M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 277 group containing mixed chemotypes (Tv 17, 19) had the shortest shoots (72.7 cm). The group having the shortest shoots (Tv 17, 19) and tallests shoots (Tv 9, 11, 13, 15) had the same number of nodes (20.9) in the stems, the number which was the highest and differed significantly (P50:0001) from the number of nodes in the other groups. The number of flower heads in the groups, which contained davadone D/myrcene– tricyclene/artemisia ketone (Tv 7, 8, 10, 20; 50.6), artemisia ketone (Tv 14, 16; 56.5) or the highest concentration of camphor (Tv 9, 11, 13, 15: 50.9) was the highest, and the results differed significantly (P50:005) from the other groups. The mixed group containing pinocamphone-camphor-1,8-cineole-bornyl acetate (Tv 17, 19) had the lowest number of flower heads (20.9) per stem and differed significantly (P50:001) from the other groups. Flowering was latest in the groups containing the lowest concentration camphor (Tv 7, 8, 10, 20 and Tv 14, 16) or a high concentration of camphor (Tv 2, 5, 6, 12), and the date at the beginning of the flowering differed significantly (P50:05) from the other three groups (Table 4). Tansy chemotypes could be joined to two clusters based on camphor concentration. The cluster consisting of chemotypes which camphor concentration was between 0.1 and 7.2% (Tv 1, 7, 8, 10, 14, 16, 20) had less nodes per stem (average 16.6; P50:001), more flower heads (average 50.4; P50:001) and a taller corymb (average 13.9 cm; P50:01), than the other cluster consisting of chemotypes which camphor concentration was more than 18.5% (Tv 2, 3, 4, 5, 6, 9, 11, 12, 13, 15, 17, 18, 19). The number of nodes, number of flower heads and the height of the corymb of the latter cluster were on average 18.8, 41.2 and 11.1 cm, respectively (data not shown).

4. Discussion

In total, 55 volatile compounds were detected from air-dried tansy flower heads of which 53 were identified. The most frequently found compounds exceeding 10% at least in one chemotype were tricyclene, camphene, myrcene, 1,8-cineole, artemisia ketone, trans thujone, camphor, umbellulone, artemisyl acetate, pinocamphone, myrtenol, chrysanthenyl acetate, bornyl acetate, and davadone D, which have already reported in tansy (Collin et al., 1993; Neszme´ lyi et al., 1992; Hendriks et al., 1990; De Pooter et al., 1989; Gallino, 1988; Holopainen, 1989; He´ thelyi et al., 1981; Ekundayo, 1979; Nano et al., 1979; Te´ te´ nyi et al., 1975; Forse´ n, 1975, 1974; Forse´ n and von Schantz, 1971; Sorsa et al., 1968; von Rudolf and Underhill, 1965), or from related species such as artemisyl acetate from Artemisia spp. (Worku and Rubiolo, 1996; Epstein and Gaudioso, 1984). An artemisia ketone isomer observed previously in tansy oil (Hendriks et al., 1990) shows a retention index linear with our observation of artemisyl acetate. Most minor compounds reported here have been reported in tansy, but seven minor compounds are reported here for the first time. All seven have been detected from related species in Asteraceae family. These include artemiseole from Artemisia arbuscula (Epstein and Gaudioso, 1984), isobornyl acetate from Artemisia vulgaris (Hwang et al., 1985), artedouglasia oxide from Artemisia laciniata (Weyerstahl et al., 1997), nerolidol from Tanacetum cinerariifolium, Artemisia lacinata, and Tanacetum polycephalum (Saggar et al., 1997; 278 M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285

Fig. 2. (a) The Geographical origins of the 20 tansy genotypes used in this study. The groups 1 and 2 are deﬁned on the basis of a complete linkage cluster analysis of RAPD-PCR data (Keskitalo et al., 1998). (b) The groups of 1–6 are deﬁned on the basis of a complete linkage cluster analysis of the presence or absence of volatile compounds analyzed in this study.

Weyerstahl et al., 1997; Rustaiyan et al., 1990, respectively), germacrene alcohol from Echinacea purpurea (Bauer et al., 1988), spathulenol from Achillea millefolium and Achilla laciniata (Afsharypuor et al., 1996a; Weyerstahl et al., 1997), and caryophyllene oxide from Tanacetum annuum, Achillea wilhelmsii and Achillea laciniata (Barrero et al., 1992; Afsharypuor et al., 1996b; Weyerstahl et al., 1997, respectively). Interestingly, davadone D which has previously been found only in M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 279

Fig. 2 (continued.)

Hungarian-grown (Ne´ meth et al., 1994; He´ thelyi et al., 1991, 1981), was detected here for the first time in tansy grown in Finland. Among the 20 tansy genotypes, 15 had a major compound, the concentration of which exceeding 40%, while the oil of five genotypes was composed of several minor compounds. Based upon the definition of ‘chemotypes’ by previous authors (Hendriks et al., 1990; Holopainen, 1989), 15 tansy genotypes are ‘well-defined’ chemotypes, and the remaining five are ‘mixed’ chemotypes. Holopainen (1989) observed that about 20% of tansy chemotypes resulting from their crossing experi- ments were also mixed chemotypes. In our study, camphor was the most common 280 .Ksiaoe l iceia ytmtc n clg 9(01 267–285 (2001) 29 Ecology and Systematics Biochemical / al. et Keskitalo M.

Table 4 Morphological attributes of tansy genotypes clustered into six groups according to the complete linkage cluster analysis based on the presence or absence of the identiﬁed compound

Main compound(s) of Tvb Concentration or Camphorc Height of No of No of ﬂower Height of Flowering chemotypes belonging range (%) of the shoot nodes per per heads the corymb started to the groupa the compound % S.D. cm S.D. shoot S.D. shoot S.D. cm S.D. Date S.D.

1. Davadone D; 7; 65.5; 2.49 3.17 91.8 8.8 17.2 1.9 50.6 6.2 12.8 4.0 29th July 3.9 Myrcene/tricyclene/artemisyl acetate; 8,10; 26.8–17.0/39.4–20.9/4.4–15.3; Artemisia ketone; 20; 67.3; 2. Camphor; 4,18; 69.4–72.2; 41.2 34.8 93.7 9.2 16.9 1.1 40.0 15.8 11.3 4.3 22nd July 2.9 Thujone; 1; 81.9; Chrysanthenyl acetate/camphor/1,8-cineole; 3; 24.9/19.4/11.4; 3. Camphor; 2, 6,12; 50.1–73.0; 55.8 20.4 84.4 18.5 17.3 3.2 40.2 8.6 10.0 1.8 30th July 4.2 1,8-Cineole/camphor; 5; 47.1/29.4; 4. Camphor; 9,11,15; 50.8–68.9; 58.2 10.3 108.2 12.0 20.9 2.8 50.9 5.7 12.5 2.6 24th July 4.4 Camphor/1,8-cineole; 13; 48.2/20.2; 5. Pinocamphone/camphor/1,8-cineole; 17; 19.1/18.5/13.7; 27.1 12.1 72.7 17.3 20.9 8.2 24.0 9.0 8.7 5.3 26th July 4.9 Camphor/bornyl acetate; 19; 35.7/16.7; 6. Artemisia ketone; 14,16; 55.1–81.4; 3.2 3.9 89.6 17.2 15.0 0.6 56.5 33.0 18.8 8.4 30th July 0.5

a Main compound(s) consisting at least 10% of the total peak area of each chemotype. b Tansy chemotype(s) consisting the compound. c Concentration of camphor in the group. M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 281 chemotype, and this is in agreement with others (Hendriks et al., 1990; Holopainen, 1989; Ekundayo, 1979; Sorsa et al., 1968). In mixed camphor chemotypes, camphene, 1,8-cineole, pinocamphene, chrysanthenyl acetate, bornyl acetate, and isobornyl acetate were the most frequently found associated compounds (10–30%). We also identified three well-defined artemisia ketone chemotypes. Such chemotypes have been observed previously by Forse´ n and von Schantz (1971) and Sorsa et al. (1968) from Finnish tansy, though are common from the Netherlands and Hungary (Hendriks et al., 1990; Te´ te´ nyi et al., 1975). A davadone D chemotype, which is reported now for the first time from tansy grown in Finland, was detected only from a Hungarian-grown tansy (Ne´ meth et al., 1994; He´ thelyi et al., 1991, 1981). In contrast to the previous authors (Holopainen, 1989; Forse´ n, 1975; Sorsa et al., 1968) who studied essential oil of Finnish tansy, thujone was not among the most common monoterpenes in our study. Thujone was found only in small concentra- tions except for one chemotype where it was the main component. Our observation was more in accordance with Hendriks et al. (1990), who did not detect high concentration of thujone together with camphor. The mixed chemotypes containing tricyclene (16–27%) and myrcene (21–39%) have not been reported previously from tansy in Finland, although Sorsa et al. (1968) found chemotypes containing either a- pinene or tricyclene (4–14%) accompanied with a low concentration (0.1–0.3%) of myrcene. Geographically, most of the chemotypes containing a significant concentration of camphor originated from Central Finland, whereas chemotypes containing camphor as a minor compound originated from South or Southwest Finland. This observation is in agreement with Sorsa et al. (1968) who found that camphor was more frequently observed in tansy grown in Northern Finland compared to southern grown tansy, where thujone was more frequent. Artemisia ketone and davadone D were detected only from tansy originating from Southwest Finland. Similar geographical variation in terpene composition and chemotypes of tansy have been observed within and between other countries (Neszme´ lyi et al., 1992; Hendriks et al., 1990; De Pooter et al., 1989; Gallino, 1988; He´ thelyi et al., 1981 Ekundayo, 1979; Nano et al., 1979; Te´ te´ nyi et al., 1975; von Rudolf and Underhill, 1965). In Finland, the genetic variation between the genotypes originating from different geographical regions may be the result of naturalization through inhabitation and agriculture, with the adaptation of tansy to the local climate (Keskitalo et al., 1998). Tansy has been observed to be one of the most common seed species in the ballast soil area in Reposaari (Jutila, 1996), which has been an important harbor in Southwest Finland. Many plant species have spread to Finland by seed embedded in the ballast soils used in ships (Jutila, 1996). Interestingly, tansy originating from Southwest Finland contained only artemisia ketone or davadone D as the main components, common compounds observed in tansy from the Netherlands (Hendriks et al., 1990) and Hungary (He´ thelyi et al., 1981). Correlation between the genetic distance matrices of our previous study (Keskitalo et al., 1998) and the chemical distance matrices of the present study was 0.407 showing some analogy between the variation of the two matrices. Unfortunately, only in a few cases has terpene 282 M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 variation been studied simultaneously with genetic variation. In agreement with our study, Juniperus species have been successfully differentiated from each other using volatile terpenoids analysis and RAPD patterns (Adams et al., 1993). The group containing camphor and 1,8-cineole, had the tallest shoots while the mixed chemotypes were the shortest. Ne´ meth et al. (1994) observed that chemotypes containing a thujen-acetate had the tallest shoots whereas plants with 1,8-cineole were the shortest. We also observed that the group containing davadone D were among the groups having a highest number of flower heads per shoots, which is in agreement with the observation of Ne´ meth et al. (1994). We observed, that groups formed from pure chemotypes produced the highest number of flower heads whereas groups formed from mixed chemotypes had a low number of flower heads. Overall, in our study, the chemotypes containing a high concentration of camphor had less flower heads and initiated flowering earlier than chemotypes low in camphor. This is in agreement with observations by Ne´ meth et al. (1994). The underlying causes of the observed chemical variation in tansy is an intriguing question, and the answer still remains elusive. Genetic variation may be due to the different geographical origins of tansy (Keskitalo et al., 1998), which eventually led to differences in the genetic control of essential oil accumulation (Holopainen, 1989; Lokki et al., 1973). A wide variation in essential oil composition presumably has ecological advantages in protecting plants against different pests (Hough-Golstein and Hahn, 1992; Neszme´ lyi et al., 1992; He´ thelyi et al., 1991; Nottingham et al., 1991). It is also likely, that part of the terpene expression in tansy is linked to specific environmental or climatic conditions (Sorsa et al., 1968), and to a lesser extent may be an indicator of other characters such as morphology (Ne´ meth et al., 1994). The dependency between the geographical origin, genetic, chemical and morphological variation of tansy show that different factors need to be recognized when the biodiversity of herbaceous species is to be examined. The analogy between the relative chemical and genetic differences among the 20 tansy chemo- and genotypes, respectively, suggest that different terpene compositions resulting from the differential activation of specific enzymes may be related to the variation in RAPDs patterns. The association between the main chemical components and morphology should be considered when selecting the parental chemo- and phenotypes from tansy populations for future work. The use of morphological traits as indicators of selected chemotypes would be most useful in breeding and biochemical studies. Since the bioactivity of the essential oil of tansy depends on the composition of terpenes (He´ thelyi et al., 1991; Holopainen and Kauppinen, 1989; Panasiuk, 1984; Schearer, 1984), the chemotype with the most effective oil composition should be selected. According to previous studies of bioactivity of tansy oil, artemisia ketone (He´ thelyi et al., 1981), camphor (Holopainen and Kauppinen, 1989; Schearer, 1984), chrysanthenyl acetate (Neszme´ lyi et al., 1992), 1,8-cineole (Schearer, 1984), davadone (He´ thelyi et al., 1981), and thujone (Holopainen and Kauppinen, 1989) chemotypes are among the most interesting ones for further studies.. M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 283

Acknowledgements

The authors thank Professor M. Levy, Department of Biological Sciences, Purdue University for providing the SAS-supported complete linkage cluster program. Financial support from the Academy of Finland (grant 7798), Finnish Association of Academic Agronomists (Agronomiliitto), and the Rotary Foundation of Southwest Finland (district 1410) is gratefully acknowledged.

References

Adams, R., 1995. Identification of Essential Oil Components by Gas Chromatography/Mass Spectro- scopy. Allured Publishing Corporation, IL USA, 469 p. Adams, R.P., Demeke, T., Abulfatih, H.A., 1993. RAPD DNA fingerprints and terpenoids: clues to past migrations of Juniperus in Arabia and East Africa. Theor. Appl. Genet. 87, 22–26. Afsharypuor, S., Asgary, S., Lockwood, G.B., 1996a. Volatile constituents of Achillea millefolium L. ssp millefoilum from Iran. Flavour Fragrance J. 11, 265–267. Afsharypuor, S., Asgary, S., Lockwood, G.B., 1996b. Constituents of the essential oil of Achillea wilhelmsii from Iran. Planta Med 62, 77–78. Banthorpe, D.V., Mann, J., 1971. Monoterpene glucosides from petals of Tanacetum vulgare. Phytochemistry 11, 2589–2591. Barrero, A.F., Sa´ nchez, J.F., Altarejos, J., Zafra, M.J., 1992. Homoditerpenes from the essential oil of Tanacetum annuum. Phytochemistry 31, 1727–1730. Bauer, R., Remiger, P., Wray, V., Wagner, H., 1988. A germacrene alcohol from fresh aerial parts of Echinacea purpurea. Planta Med. 54, 478–479. Blumenthal, M. (Translator), 1998. The Complete German Comission E Monographs: Therapeutic Guide to Herbal Medicines. Tansy Flower and Herb. Unapproved Herbs. American Botanical Council, Austin, TX, Integrative Medicine Communications, Boston, MA, pp. 379–380. Brown, A.M.G., Edwards, C.M., Davey, M.R., Power, J.B., Lowe, K.C., 1997. Effects of extracts of Tanacetum species on human polymorphonuclear leucocyte activity in vitro. Phytothe. Res. 11, 479–484. Chandra, A., Misra, L.N., Thakur, R.S., 1987a. Germacranolides and an alkyl glucoside from Tanacetum vulgare. Phytochemistry 26, 1463–1465. Chandra, A., Misra, L.N., Thakur, R.S., 1987b. Tanavulgarol, an oxygenated sesquiterpene with an uncommon skeleton from Tanacetum vulgare. Phytochemistry 26, 3077–3078. Collin, G.J., Deslauriers, H., Pageau, N., Gagnon, M., 1993. Essential oil of tansy (Tanacetum vulgare L.) of Canadian origin. J. Essent. Oil Res. 5, 629–638. Ekundayo, O., 1979. Essential oils II. Terpene composition of the leaf oil of Tanacetum vulfgare L. Z. Pflanzenphysiol. Bd. 92, 215–220. Epstein, W.W., Gaudioso, L.A., 1984. Volatile oil constituents of sagebrush. Phytochemistry 23, 2257–2262. Finnigan, Heller, Milne, 1978. 4000 GC/MS; NBS Library 31, 331 entries Forse´ n, K., 1974. Begleitstoffe in verschiedenen Chemotypen von Chrysanthemum vulgare Reine trans- Chrysanthenylacetate-type. Farm. Aikak 83, 9–17. Forse´ n, K., 1975. U¨ ber die Infraspezifische chemische variation bei Chrysanthemum vulgare. Ann. Acad. Sci. Fenn. Ser. A IV 207, 1–54. Forse´ n, K., von Schantz, M., 1971. Neue Hauptbestandteile im a¨ therischen o¨ l des Rainfarns in Finnland. Arch. Pharm. 304, 944–952. Gallino, M., 1988. Essential oil from Tanacetum vulgare growing spontaneously in ‘Tierra del Fuego’ (Argentina). Planta Med. 54, 182. Grieve, M., 1984. Tansy. In: Leyel, C.F. (Ed.), A Modern Herbal. Penguin Books Ltd, Middlesex, Great Britain, pp. 789–790. 284 M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285

Hendriks, H., Bos, R., 1990. The presence of parthenolide in Dutch tansy. Planta Med. 56, 550. Hendriks, H., van der Elst, D.J.D., van Putten, F.M.S., Bos, R., 1990. The essential oil of Dutch Tansy (Tanacetum vulgare L.). J. Essent. Oil Res. 2, 155–162. Heywood, V.H., 1976. Tanacetum L. In: Tutin, T.G., Heywood, V.H., Burges, N.A., Moore, D.M., Valentine, D.H., Walters, S.M., Webb, D.A. (Eds.), Flora Europaea, Plantaginaceae to Compositae (and Rubiaceae), Vol. 4. Cambridge University Press, Cambridge, Great Britain, pp. 169–171. He´ thelyi, E´ ., Te´ te´ nyi, P., Danos, B., Koczka, I., 1991. Phytochemical and antimicrobial studies on the essential oils of the Tanacetum vulgare clones by gas chromatography/mass spectrometry. Herba Hung 30, 82–90. He´ thelyi, E´ ., Te´ te´ nyi, P., Kettens-van den Bosch, J.J., Salemink, C.A., Heerma, W., Versluis, C., Kloosterman, J., Sipma, G., 1981. Essential oils of five Tanacetum vulgare genotypes. Phytochemistry 20, 1847–1850. Holopainen, M. (1989). A study on the essential oil of tansy (Tanacetum vulgare L.). Ph.D. Thesis, Department of Pharmacy, University of Helsinki, Finland. Holopainen, M., Kauppinen, V., 1989. Antimicrobial activity of different chemotypes of Finnish tansy. Planta Med 55, 103. Hough-Golstein, J., Hahn, S.P., 1992. Antifeedant and oviposition deterrent activity of an aqueous extract of Tanacetum vulgare L. on two cabbage pest. Environ. Entomol. 21, 837–844. Hulte´ n, E., 1968. Flora of Alaska and Neighboring Territories. A Manual of the Vascular Plants. Stanford University Press, Stanford CA, 1008 p. Hussey, J., 1974. Some useful plants of early New England. Econ. Bot. 28, 311–337. Hwang, Y.S., Wu, K.H., Kumamoto, J., Axelrod, H., Mulla, M.S., 1985. Isolation and identification of mosquito repellents in Artemisia vulgaris. J. Chem. Ecol. 11, 1297–1306. Jutila, H.M., 1996. Seed bank and emergent vascular flora of ballast areas in Reposaari, Finland. Ann. Bot. Fennici. 33, 165–182. Keskitalo, M., Angers, P., Earle, E., Pehu, E., 1999. Chemical and genetical characterization of protoplast fusion-derived calli between tansy (Tanacetum vulgare L.) and pyrethrum (Tanacetum cinerarieafolium). Theor. Appl. Genet. 98, 1335–1343. Keskitalo, M., Kanerva, T., Pehu, E., 1995. Development of in vitro procedures for regeneration of petiole and leaf explants and production of protoplast-derived callus in Tanacetum vulgare L. (Tansy). Plant Cell Rep. 14, 261–266. Keskitalo, M., Linde´ n, A., Valkonen, J.P.T., 1998. Genetic and morphological diversity of Finnish tansy (Tanacetum vulgare L.). Theor. Appl. Genet. 96, 1141–1150. Lawrence, B.M., 1992. Essential oils as source of natural aroma chemicals. Perfumer Flavorist 17, 15–28. Levy, M., Romao, J., Marcetti, M.A., Hamer, J.E., 1991. DNA Fingerprinting with a dispersed repeated sequence resolves pathotype diversity in rice blast fungus. Plant Cell 3, 95–102. Lokki, J., Sorsa, M., Forse´ n, K., von Schantz, M., 1973. Genetics of monoterpenes in Chrysanthemum vulgare. Hereditas 74, 225–232. Millspaugh, C.F., 1974. American Medicinal Plants. An Illustrated and Descriptive Guide to Plants Indigenous to and Naturalized in the United States which are used in Medicine. Dover Publications, New York, 806p. Mitich, L.W., 1992. Intriguing world of weeds } Tansy. Weed Technol. 6, 242–244. Mordujovich-Buschiazzo, P., Balsa, E.M., Busciazzo, H.O., Mandrile, E., Rosella, M., Schinella, G., Fioravanti, D., 1996. Anti-inflammatory activity of Tanacetum vulgare. Fitoterapia LXVII, 319–322. Nano, G.M., Bicchi, C., Frattini, C., Gallino, M., 1979. Wild Piedmontese Plants II. A rare chemotype of Tanacetum vulgare L., abundant in Piedmont (Italy). Planta Med. 35, 270–274. National Landsurvey Institute, 1996. Geographic Data Center. Ne´ meth, E´ .Z., He´ thelyi, E´ ., Berna` th, J., 1994. Comparison studies on Tanacetum vulgare L. chemotypes. J. Herb Spices Med. Plants 2, 85–92. Neszme´ lyi, A., Milne, G.W.A., He´ thelyi, E´ ., 1992. Composition of the essential oil of clone 409 of Tanacetum vulgare and 2D NMR investigation of trans-chrysanthenyl acetate. J. Essent. Oil Res. 4, 243–250. M. Keskitalo et al. / Biochemical Systematics and Ecology 29 (2001) 267–285 285

Nei, M., Li, W.H., 1979. Mathematical model for study genetic variation in terms of restriction endonucleases. Proc. Natl. Acad. Sci. USA 74, 5267–5273. Nottingham, S.F., Hardie, J., Dawson, G.W., Hick, A.J., Picett, J.A., Wadhams, L.J., Woodcock, C.M., 1991. Behavioral and electrophysiological responses of aphids to host and nonhost plant volatiles. J. Chem. Ecol. 17, 1231–1242. Ognyanov, I., Todorova, M., 1983. Sesquiterpene lactones and flavonoids in flowers of Tanacetum vulgare. Planta Med. 48, 181–183. Ognyanov, I., Todorova, M., Dimitrov, V., Ladd, J., Irngartinger, H., Kurda, E., Rodewald, H., 1983. Cis-Longipinane-2,7-dione, a sesquiterpene diketone in flowers of Tanacetum vulgare. Phytochemistry 22, 1775–1777. Panasiuk, O., 1984. Response of Colorado potato beetles, Leptinotarsa decemlineata (Say), to volatile components of tansy, Tanacetum vulgare. J. Chem. Ecol. 10, 1325–1333. De Pooter, H.L., Vermeesch, J., Schamp, N.M., 1989. The essential oils of Tanacetum vulgare L. and Tanacetum parthenium (L.) Schultz-Bip. J. Essent. Oil Res. 1, 9–13. von Rudloff, E., Underhill, E.W., 1965. Gas-liquid chromatography of terpenes}XII. Seasonal variation in the volatile oil from Tanacetum vulgare L. Phytochemistry 4, 11–17. Rustaiyan, A., Zare, K., Habibi, Z., Hashemi, M., 1990. Germacranolides from Tanacetum polycephalum. Phytochemistry 29, 3022–3023. Saggar, P., Wamicha, W.N., Chhabra, S.C., Ndalut, P., 1997. Isolation, identification and bioassay of repellent factors in the essential oil of pyrethrum for grain protection against Sitophilus zeamais (Molts.). Pyrethrum Post 19 (4), 126–130. Sanz, J.F., Marco, J.A., 1991. NMR studies of tatridin A and some related sesquiterpene lactones from Tanacetum vulgare. J. Nat. Prod. 54, 591–596. SAS Institute Inc., 1984. SAS/STAT Users Guide. Version 6, 4th Edition. 1686p. von Schantz, M., Ja¨ rvi, M., Kaartinen, R., 1966. Die Vera¨ nderungen des A¨ therischen O¨ les wa¨ hrend der Entwicklung der Blu¨ tenko¨ rbchen von Chrysanthemum vulgare L. Planta Med. 14, 421–435. Schearer, W.R., 1984. Components of oil of tansy (Tanacetum vulgare) that repel Colorado potato beetles (Leptinotarsa decemlineara). J. Nat. Prod. 47, 964–969. Sorsa, M., von Schantz, M., Lokki, J., Forse´ n, K. (1968). Variability of essential oil components in Chrysanthemum vulgare L. in Finland. Ann. Acad. Sci. Fenn. Ser. A VI, 135, 1-13. Suomi, D., Brown, J.J., Akre, R.D., 1986. Responses to plant extracts of Neonatal codling moth larvae, Cydia pomonella (L.), (Lepidoptera: Tortricidae: Olethreutinae). J. Entomol. Soc. B. C. 83, 12–18. Stefanovic, M., Ristic, N., Vukmirovic, M., 1988. Biological activities of sesquiterpene lactones. Investigations of microbial activities of lactones isolated from the Yugoslav plant species of the genus Tanacetum L. [Chrysanthemum (fam. Compositae)]. Bull. T. XCV Acad. Serbe Sci. Arts, Classe Sci. Naturelles Math. Sci. Naturelles 28, 23–43. Stenhagen, E., Abrahamssons, S., McLafferty, F., 1974. Registry of Mass Spectral Data. Wiley, New York. Te´ te´ nyi, P., Kaposi, P., He´ thelyi, E´ ., 1975. Variations in the essential oils of Tanacetum vulgare. Phytochemistry 14, 1539–1544. Weyerstahl, P., Marschall, H., Schroder, M., Wahlburg, H.C., Kaul, V.K., 1997. The sesquiterpene fraction of Artemisia laciniata Willd. Flavour Fragrance J. 12, 315–325. Wilkomirski, B., Kucharska, E., 1992. Triterpene chemotypes of some Polish populations of Tanacetum vulgare. Phytochemistry 31, 3915–3916. Worku, T., Rubiolo, P., 1996. Major constituents of Artemisia afra oil. J. Essent. Oil Res. 8, 355–357.