Volatile Organic Components of Fresh Leaves as Indicators of Indigenous and Cultivated Citrus Species in Taiwan

Biosci. Biotechnol. Biochem., 74 (4), 806 811, 2010 Volatile Organic Components of Fresh Leaves as Indicators of Indigenous and Cultivated Citrus Species in Taiwan Shu Yen LIN, 1 Su Feng ROAN, 2 Ching Lung LEE, 1 and Iou Zen CHEN 1;y 1 Department of Horticulture, National Taiwan University, Taipei 106, Taiwan 2 Department of Horticulture and Biology, Chinese Culture University, Taipei 111, Taiwan Received December 3, 2009; Accepted January 4, 2010; Online Publication, April 7, 2010 [doi:10.1271/bbb.90891] The volatile components of fresh leaves from 15 citrus species were investigated by headspace SPME with a GC-MS analysis. Three indigenous Taiwan citrus species, Citrus taiwanica, C. tachibana and C. depressa, were the major subjects. Eighty volatile organic compounds were detected as indicators of the genetic relationship. Linalool was the most abundant compound, and citronellal, geranial, neral, limonene and trans--ocimene were the major volatile compounds in fresh leaves. Linalool (56.37%) and myrcene (7.21%) were predominant in C. tawanica. An aldehyde-rich profile with citronellal (24.54%) contributed most to the aroma of leaves in C. tachibana, while Citrus depressa exhibited a high linalool/citronellal composition (23.56%/12.51%). The qualitative and quantitative patterns of the volatiles revealed that C. taiwanica was linked with sour orange, and either C. tachibana or C. depressa belonged to the mandarin group with C. tankan. Dendrograms also showed that the volatile patterns were related to the genetic classification. Key words: Citrus depressa; Citrus tachibana; Citrus taiwanica; headspace SPME; GC-MS analysis There are three indigenous citrus species in Taiwan, but all are rare and endangered. 1) Citrus taiwanica, C. tachibana, and C. depressa were proposed to be native citrus species by Tyozaburo Tanaka in 1926 2,3) and 1931. 4) Citrus taiwanica (nansho-daidai) was classified as an independent species belonging to the sour orange group, which is in subsection Aurantium in Tanaka s taxonomic system, 5) while Swingle suggested that C. taiwanica was a spontaneous variety from sour orange. 6) The smaller winged-leaf (4 mm width) and purple-red chalaza made C. taiwanica distinguishable as sour orange. 2) The other two native citrus species, C. depressa and C. tachibana, are small-fruit mandarin and have been grouped in subsection Microacrumen under Tanaka s systematic classification. C. depressa has traditionally been used in Taiwan as a medicine to release gastrointestinal upsets, 7,8) and a special sauce has been made from C. tachibana. Studies on the compounds concerned with the health effect of C. depressa became popular due to the long lives of Okinawa residents who used it in daily food. Nobiletin extracted from C. depressa could suppress the production of matrix metalloproteinase, decrease the symptoms of osteoarthritis and rheumatoid arthritis, and also function as an anti-inflammatory, anti-allergic and anticancer agent. 9,10) The contribution of the variability of secondary metabolites to plant taxonomy has been reported. Sixty-six citrus species and near-citrus relatives could be cited in accordance with Tanaka s classification system with 24 flavonoids. 11) The genetic differences between various hybrid varieties of mandarin and orange could be determined by using 12 carotenoid profiles. 12) Three races of avocado (Persea americana) have been separated well by using leaf volatiles as the indicators. 13) The three native citrus species in Taiwan have a different but strong and special fragrance when their leaves are scrubbed, and the essential oils in citrus species have been evaluated to reveal the relationship with their genetic classification. 14 17) However, the odor of fresh leaves sensed by the nose has been an effective indicator used in traditional taxonomical classification. Solid-phase microextraction (SPME) is a solvent-free method for preparing samples, 18) needing no purification steps and no additional instrumental analysis. Headspace SPME has been proposed to reduce interference and matrix effects by choosing the stationary phase, 19) and has proved to be a valid alternative to a gas chromatographic analysis of essential oils from different sources. Three citrus species native to Taiwan and twelve popular cultivated citrus species were selected in this study. They all shared the same environmental, soil, and culture conditions. The profiles of the volatiles emitted from the fresh mature leaves were analyzed by headspace SPME coupled with gas chromatography-mass spectrometry (GC-MS). Preliminary tests were conducted to optimize the extraction conditions. The objective of this work was to evaluate whether the volatile patterns were unique and could be used to discriminate the native species from the cultivated ones. The results of the chemical variability of these 15 species were submitted to a hierarchical cluster analysis. Materials and Methods Plant materials and sample preparation. The samples were taken from healthy adult trees of the germplasm collection orchard at National Taiwan University located in Taipei. All the citrus species were grown under the same climatic and culture conditions. Three indigenous citrus species, Citrus taiwanica, C. tachibana, and C. depressa, were particular subjects in this study. The selected cultivated species included the orange group with sour orange (C. aurantium) y To whom correspondence should be addressed. Tel: +886-2-3366-4841; Fax: +886-2-2362-0760; E-mail: chenyo@ntu.edu.tw

and Liucheng orange (C. sinensis); mandarin group with ponkan (C. reticulata), calamondin (C. microcarpa), sunki mandarin (C. sunki Hort. ex. Tanaka), cleopatra mandarin (C. reshni Hort. ex. Tanaka), and tankan (C. tankan Hayata, a hybrid of C. reticulata and C. sinensis). Typical species included the citron group with Buddha s hand citron (C. medica var. sarcodactylis) and Sihjii lemon (C. limonum); pomelo group with Wendun pomelo (C. grandis) and Ruby grapefruit (C. paradisi, a hybrid of C. grandis and C. sinensis); and kumquat (Fortunella japonica). Three 5-mm leaf disks were randomly punched from healthy mature leaves of each selected species and immediately sealed in a gas-tight vial. The samples were kept on ice before heating, and all the analyses were performed in triplicate. Volatile extraction. Headspace SPME was performed by a Supelco SPME device coated with polydimethylsiloxane (PDMS, 100 mm) to extract each fresh sample in a 30-ml vial. The sample was removed from ice and pre-conditioned at 70 C for 5 min to initiate volatile emission and equilibrium. The SPME fiber was exposed to the headspace vapor for 15 min at 70 C for extraction. After the volatiles had been absorbed from a fresh leaf disk, the fiber was transferred to the GC-MS port for splitless injection at 250 C. GC-MS analysis. GC analysis was accomplished with an HP 5890 GC-5972 MSD instrument (Agilent, USA). A capillary column (Hewlett-Packard DB-5) 30 m long and 0.25 mm i.d. with a 1-mm film thickness was employed. The following temperature program was used: hold at 50 C for 3 min, then increase to 200 C at a rate of 5 Cmin 1, increase further to 250 Cat10 Cmin 1, and finally hold at 250 C for 2 min. The injector and detector temperatures were 250 C and 280 C, respectively, and the carrier gas was helium at a flow rate of 2 mlmin 1. The quadrupole mass spectrometer was scanned in the 40 to 300 a.m.u. range at 1 scansec 1 with an ionizing voltage of 70 ev. The organic compounds were identified by comparing the MS fragmentation pattern with the mass spectral data provided in the National Institute of Standards and Technology (NIST) library, the Wiley library, and related literature. 14,17,20) Statistical methods. The relative amount of each individual component is expressed as the percentage peak area relative to total peak area. The presence or absence of each respective compound is recorded as 1 or 0. The data were calculated as Jaccard distances and coupled with a UPGMA cluster analysis by NTSYSpc (Version 2.11L). The revised abundance of each volatile organic compound was recalculated using the log 10 value. Another dendrogram was constructed for Euclidean distances combined with the UPGMA cluster analysis. Results and Discussion Eighty volatile organic compounds were detected by using the gas chromatographic analysis with a DB-5 polar column for all citrus species. The detected compounds, accounting for 76.51% to 97.20% within all citrus samples, were identified and quantified by their mass fragmentation patterns against data in the NIST library, Wiley library, and related references. Tables 1 and 2 express the mean percentages obtained by peakarea integration of the major detected compounds for three replicates of each species. For the tested citrus species, citronellal, geranial, neral, limonene, and trans- -ocimene were the major volatile organic compounds in the fresh mature leaves. Linalool was an abundant compound in most citrus groups, especially in the orange and mandarin groups. Linalool, limonene, and geranial were the most detected compounds in the essential oil and cold-pressed oil of mandarins, 15,21) bergamot, 22) and lime. 23) Among all the selected citrus species, the volatile profile of kumquat, which belongs to the Fortunella genus, was very different from that of the Citrus genus. Choi 24) has reported that the major Volatiles in Fresh Leaves of Citrus Native to Taiwan 807 compounds of kumquat peel oil were limonene, myrcene, and ethyl acetate, but the kumquat-like odor component would be citronellyl acetate. Germacren D, a sesquiterpene, was the most abundant compound (12.49%) in the volatiles of fresh kumquat leaves. Over 90% of the volatile compounds in fresh kumquat leaves were released from the SPME fiber at above 150 C, most of them being sesquiterpenes. Of the three indigenous citrus species, C. taiwanica had a high proportion of linalool (56.37%) and myrcene (7.21%), the volatile profile being similar to that of sour orange. Monoterpenes were the predominant volatile components of C. taiwanica fresh leaves, accounting for 21.1%, in which the two predominant compounds were myrcene and ocimene, from elimination of the pyrophosphate group. Citrus taiwanica, considered as a spontaneous variety of sour orange, had the same volatile compounds as those in the fresh leaves of sour orange, although the proportional compositions were different (Table 1). The volatile profile of directly heated fresh leaves of C. taiwanica was compared with the cold-pressed essential oil of bergamot sour orange. More compounds were detected in the latter and the two profiles were dissimilar. 22) The major compounds in bergamot sour orange oil were limonene (37.2%), linalyl acetate (30.1%), and linalool (8.8%), while the proportions of limonene, linalyl acetate and linalool in the fresh leaves of C. taiwanica were 1.9%, 6.7%, and 56.4%, respectively. Linalool is a naturally-occurring terpene alcohol found in many flowers to attract pollinators, although anti-ethylene 25) and anti-inflammatory functions 26) have also been found in this essential oil. The distinctive amount of linalool in fresh mature leaves suggests that it plays a defensive role in the mature leaves. Both of the indigenous citrus species, C. tachibana and C. depressa, are small-fruit mandarins with a fruit size of about 4 cm in diameter similar to calamondin and sunki mandarin. The two native mandarins were compared with ponkan, the major commercial cultivar in Taiwan, calamondin, a popular small-fruit cultivar, sunki mandarin, and cleopatra mandarin which used to be used as a rootstock in Taiwan. The total of the 60 identified compounds accounted for 76.51% to 96.83% (Table 2). Even though the composition was dominated by linalool (23.56% to 67.27%) in the selected mandarins, the content of the major components varied from sample to sample. Citrus tachibana was an exception to the linalool-rich profile. An aldehyde-rich profile (41.9%) contributed to the aroma of the fresh leaves in C. tachibana, and citronellal (24.54%) was among the richest components, followed by citronellal, limonene, citronellol, geranial, and linalool. Monoterpenes and terpene alcohols were notable with total respective contents of 17.8% and 26.9%. Citrus depressa exhibited a high linalool/citronellal composition (23.56%/ 12.51%). The main component fraction was the terpene alcohol, besides linalool, followed by citronellol (4.67%), -terpinol (4.22%), isopulegol (2.71%), and geraniol (0.96%). Trace contents of humulene (0.43%) and cadiene (0.28%) in C. tachibana, and the thymyl methyl ether (6.94%) and myrcene (0.96%) in C. depressa were distinguishing compounds between the two native species. The fruit of calamondin has a similar

808 S. Y. LIN et al. Table 1. Volatile Organic Compounds in Fresh Leaves Identified by GC-MS in the Orange, Pomelo and Citron Groups Peak area % on DB-5 No. Compound C. taiwanica C. aurantium C. sinensis C. grandis C. paradisi C. medica C. limonum 1 Sabinene 1.33 1.38 2 Myrcene 7.21 6.42 1.09 0.00 0.37 0.38 3 -Pinene 2.90 4 Limonene 1.92 2.08 2.51 2.68 3.90 2.30 5-3-Carene 2.09 1.99 6 trans--ocimene 5.00 4.82 3.75 1.45 1.34 7 -Terpinene 0.40 8 o-isopropenyltoluene 0.53 9 -Terpinolene 0.70 0.82 0.99 0.64 1.78 10 Nonanal 0.68 11 Linalool 56.37 56.93 14.45 5.76 8.26 12 1,3,8-p-Menthatriene 0.40 13 Ocimene 0.88 0.87 14 2-Octene 2.29 15 1,5-Heptadiene 3.17 3.24 16 Isopulegol 10.36 4.34 17 Citronella 0.73 0.78 3.60 54.26 17.33 0.26 1.38 18 Spiro[2.5]octane 4.07 2.29 6.48 6.39 19 Neo-Isopulegol 1.33 20 Compound A 4.18 4.61 21 Cyclooctane 7.12 10.07 9.91 22 Linalyl propanoate 1.19 23 -Terpineol 1.00 0.91 0.87 24 -Fenchyl alcohol 1.01 0.63 25 2-Cyclohexen-1-ol 0.36 1.00 0.95 26 1,5-Hexadiene 0.51 0.46 27 Citronellol 11.68 28 Nerol 1.38 1.65 4.14 6.27 3.24 3.18 29 Neral 1.76 2.31 16.14 11.35 23.90 23.62 30 Linalyl acetate 6.71 5.12 31 Geraniol 2.07 2.07 2.84 2.41 32 Geranial 1.87 3.33 23.24 16.25 35.63 35.25 33 Bicycloelemene 0.40 34 2-Carene 0.60 0.59 35 Citronellyl acetate 1.22 36 Citronellyl propionate 0.37 37 Neryl acetate 2.28 2.08 0.19 0.46 0.40 0.61 38 Geranyl acetate 5.04 4.72 0.43 0.60 0.99 39 -Elemene 0.97 1.99 0.12 0.00 40 -Caryophyllene 0.87 1.07 0.52 2.16 0.91 0.77 0.72 41 -Bergamotene 0.20 42 Aromadendrene 0.52 43 trans--farnesene 0.43 2.28 44 -Humulene 0.20 0.23 0.70 45 -Selinene 0.34 1.26 46 -Bisabolene 0.08 47 Bicyclogermacrene 0.86 48 -Cadinene 0.40 49 -Bisabolene 0.41 0.46 50 Propanoic acid 0.40 0.31 0.19 0.34 Total 96.41 96.49 94.91 95.01 87.33 97.20 95.20 No compounds detected. Compound A, 3-cyclohexene-1-carboxaldehyde taste to C. depressa fruit, so calamondin can be used as a substitute juice source for C. depressa. Calamondin fruit, although containing limonene, linalool oxide, linalool, and -terpineol with high flavor-dilution factors, 27) had partially the same volatiles as those in C. depressa leaves. The volatiles in fresh leaves of calamondin comprised sesquiterpenes (27.5% from 13 compounds) and terpene alcohol (36.47% of linalool and 5.48% of -eudesmol). Merl et al. 15) have shown the discriminating power of the chemical composition of mandarins and its contribution to taxonomy at the varietal level. The presence/ absence of the total detected compounds in the sample species were recorded by a hierarchical clustering analysis for Jaccard distances coupling with the unweighted pair group method arithmetic average (UPGMA) (Fig. 1). The result of a qualitative component analysis fitted approximately to the biological classification for the C. medica, C. grandis, and C. reticulata groups. 28) The characteristic compounds in the citron group, including C. limonum and C. medica var. sarcodactylis, were geranial and neral, while pomelo and grapefruit (the natural hybrid) were charactered by their abundant content of citronellal (Table 1). The

Table 2. Volatiles in Fresh Leaves of Citrus Native to Taiwan 809 Volatile Organic Compounds in Fresh Leaves of Mandarins and Fortunella japonica Identified by GC-MS Peak area % on DB-5 No. Compound C. tachibana C. depressa C. reticulata C. sunki C. reshni C. microcarpa C. tankan F. japonica 1 Sabinene 4.38 5.12 2 Myrcene 0.96 4.06 2.16 3.08 1.20 1.88 3 -Pinene 1.01 1.72 4 Limonene 15.32 11.18 2.52 2.61 4.64 3.29 5-3-Carene 1.30 6 trans--ocimene 0.92 1.86 4.14 6.32 5.39 3.19 2.79 7 -Terpinene 1.85 9.22 2.44 9 -Terpinolene 1.00 1.44 2.08 1.07 0.53 11 Linalool 6.10 23.56 67.27 49.44 61.34 36.47 50.70 14 2-Octene 0.98 16 Isopulegol 5.41 2.71 17 Citronella 24.54 12.51 3.51 18 Spiro[2.5]octane 1.84 1.64 1.51 21 Cyclooctane 2.58 2.58 22 Linalyl propanoate 1.22 23 -Terpineol 6.06 4.22 2.04 4.68 4.86 27 Citronellol 8.57 4.67 2.00 29 Neral 5.37 5.78 6.50 31 Geraniol 0.74 0.96 4.06 32 Geranial 7.57 8.26 9.85 36 Citronellyl propionate 1.18 38 Geranyl acetate 2.11 39 -Elemene 0.88 3.11 40 -Caryophyllene 3.22 2.16 0.31 0.97 0.91 4.73 0.42 8.15 41 -Bergamotene 0.93 0.57 43 trans--farnesene 0.57 0.84 0.85 44 -Humulene 0.43 2.82 48 -Cadinene 0.28 0.29 0.21 0.26 3.46 0.48 12.38 49 -Bisabolene 1.73 1.11 50 Propanoic acid 0.38 0.21 51 -Thujene 1.23 52 -Pinene 1.58 1.03 4.66 53 3-Hexen-1-ol 2.11 3.49 54 4-Cymene 7.31 55 cis-sabinenehydrate 1.44 56 p-mentha-1,5,8-triene 0.76 57 Terpinen-4-ol 1.49 1.76 2.03 58 Decanol 0.50 59 Thymyl methyl ether 0.00 6.94 1.14 10.97 60 Germacrene B 0.30 61 Bicycloelemene 1.29 62 -Terpipene 3.78 63 -Cubbene 0.28 1.22 64 -Ylangene 0.75 65 -Copaene 1.24 66 Compound A 0.68 67 -Elemene 0.43 0.62 1.62 6.68 68 Compound B 1.51 69 Compound C 4.15 70 -Cubebene 3.27 71 Germacrene D 12.49 72 -Selinene 5.52 5.51 73 Ledene 0.47 0.51 74 -Farnesene 1.11 1.59 1.20 75 -Amorphene 2.58 5.11 76 Naphthalene 0.87 77 Elemol 3.58 78 Calarene 1.80 79 -Gurjunene 0.47 80 -Eudesmol 5.48 Total 95.48 92.71 93.79 96.83 95.63 87.06 95.26 76.51 No compounds detected. Compound A, cyclobuta[1,2:3,4]-dicyclopenten-1-ol Compound B, 6,10,11,11-tetramethyl-tricyclo[6.3.0.1(2,3)]undec-7-ene Compound C, bicyclosesquiphellandrene

810 S. Y. LIN et al. Pomelo group Citron group Orange group Minor group A Mandarin group Minor group B Minor group C Fig. 1. Dendrogram of the Presence/Absence of the Detected Volatile Compounds of Fresh Mature Citrus Leaves Analyzed by a UPGMA Cluster Analysis. Three groups of biological classification, citron (Citrus medica), pomelo (C. grandis), and mandarin (C. reticulate), were clearly separated. Fig. 2. Dendrogram of the Revised Abundance of the Volatile Components of Fresh Mature Citrus Leaves Based on the Log 10 Value by a UPGMA Cluster Analysis. The mandarin group was further classified into minor group A (Citrus tachibana, C. depressa, and C. tankan), minor group B (C. ponkan, C. reshni, and C. sunki), and minor group C (C. microcarpa) together with Fortunella japonica. revised abundance of the volatile constituents with linkage was also used in the UPGMA clustering analysis (Fig. 2). Differences when compared to the qualitative clustering dendrogram were apparent, but a stronger genetic relationship was revealed in the quantitative dendrogram. Citrus taiwanica could be linked with sour orange from the relatively similar volatile profiles. The mandarin group was split into three categories. Minor group A (Fig. 2) consisted of the two native Taiwan species and tankan which is the natural hybrid of C. reticulata and C. sinensis cultivated only in Asia. The tight-peeled fruit and uniquely sour flavor of C. tachibana and C. depressa positioned minor group A adjacent to sour orange and C. taiwanica. Ponkan, cleopatra mandarin, and sunki mandarin were clustered into minor group B because of the abundant linalool/ ocimene content (Table 2). Calamondin, assigned to minor group C, was classified with kumquat, even though they belong to different genera. From the genetic aspect, calamondin mandarin is an interspecific hybrid of C. reticulata and Fortunella margarita. 29) Although most of the morphological characteristics of calamondin showed a closer relationship with mandarin, the volatile pattern of fresh leaves displayed the genetic influence. In respect of grapefruit, a hybrid of pomelo and orange, both the qualitative and quantitative profiles of the leaf volatiles suggested that the volatiles in fresh leaves were much more similar to orange than to pomelo. The volatile organic compounds of fresh leaves analyzed by GC/MS were highly reproducible characteristics. More supporting data from secondary metabolites could modify the traditional taxonomy or bring more information about the relationship among complicated species like Citrus. 12,13,15) Fresh samples could be investigated by the headspace SPME technique under conditions close to those perceived for humans and insects. The profiles of volatile compounds in fresh leaves by headspace SPME is a useful application in chemotaxonomy that would be helpful to find out more information about the interaction between plants and insects due to its easy manipulation, time-saving, and good reproducibility. This is especially the case with citrus species, which have a strongly characteristic aroma, and will assist with the identification of unknown wild species and hybrids. References 1) Chang CE and Hartley TG, Flora of Taiwan Vol. 3 2nd ed., Editorial Committee of the Flora of Taiwan, Department of Botany, NTU, Taipei, pp. 510 544 (1993). 2) Tanaka T, Proc. Imp. Acad. Japan, 2, 345 347 (1926). 3) Tanaka T, Bul. Sci. Fak. Terk. Kjushu Univ., 2, 51 58 (1926). 4) Tanaka T, Stud. Citrol., 5, 1 20 (1931). 5) Tanaka T, Jap. Soc. Prom. Sci., Ueno, p. 152 (1954). 6) Swingle WT, The Citrus Industry, eds. Reuther W, Webber HJ, and Batchelor LD, University of California, Berkeley, pp. 128 474 (1948). 7) Wu TS, Kuoh CS, and Furukawa H, Chem. Pharm. Bull., 31, 895 900 (1983). 8) Yang MS, Wu TS, and Wang CH, Planta Med., 53, 143 147 (1987). 9) Ishiwa J, Sato T, Mimaki Y, Sashida Y, Yano M, and Ito A, J. Rheumatol., 27, 20 25 (2000). 10) Minagawa A, Otani Y, Kubota T, Wada N, Furukawa T, Kumai K, Kameyama K, Okada Y, Fujii M, Yano M, Sato T, Ito A, and Kitajima M, Jpn. J. Cancer Res., 92, 1322 1328 (2001). 11) Kawaii S, Tomono Y, Katase E, Ogawa K, and Yano M, J. Agric. Food Chem., 47, 3565 3571 (1999). 12) Goodner KL, Rouseff RL, and Hofsommer HJ, J. Agric. Food Chem., 49, 1146 1150 (2001). 13) Wu HL, Chou CC, Jong TM, and Chen IZ, J. Taiwan Soc. Hort. Sci. (in Chinese), 53, 363 379 (2007). 14) Lota ML, Serra D der, Tomi F, and Casanova J, Biochem. Syst. Ecol., 29, 77 104 (2001). 15) Merle H, Morón M, Blázquez MA, and Boira H, Biochem. Syst. Ecol., 32, 491 497 (2004). 16) Shaw PE, Goodner KL, Moshonas MG, and Hearn CJ, Sci. Hortic., 91, 71 80 (2001). 17) Tomi F, Barzalona M, Casanova J, and Luro F, Flavour Fragr. J., 23, 152 163 (2008). 18) Kataoka H, Lord HL, and Pawliszyn J, J. Chromatogr. A, 880, 35 62 (2000).

19) Steffen A and Pawliszyn J, J. Agric. Food Chem., 44, 2187 2193 (1996). 20) Flamini G, Tebano M, and Cioni PL, Anal. Chim. Acta, 589, 120 124 (2007). 21) Sawamura M, Minh Tu NT, Onishi Y, Ogawa E, and Choi HS, Biosci. Biotechnol. Biochem., 68, 1690 1697 (2004). 22) Sawamura M, Onishi Y, Ikemoto J, Minh Tu NT, and Lan Phi NT, Flavour Fragr. J., 21, 609 615 (2006). 23) Craske JD, Suryadi N, and Wootton M, J. Sci. Food Agric., 85, 522 525 (2005). Volatiles in Fresh Leaves of Citrus Native to Taiwan 811 24) Choi HS, J. Agric. Food Chem., 53, 1642 1647 (2005). 25) Grichko VP, Sisler EC, and Serek M, Biochem. Biotech., 16, 20 27 (2003). 26) Peana AT, D Aquila PS, Panin F, Serra G, Pippia1 P, and Moretti MDL, Phytomedicine, 9, 721 726 (2002). 27) Takeuchi H, Ubukata Y, Hanafusa M, Hayashi S, and Hashimoto S, J. Essent. Oil Res., 17, 23 26 (2005). 28) Barrett HC and Rhodes AM, Syst. Bot., 1, 105 136 (1976). 29) Lai YT and Chen IZ, Acta Hortic., 773, 111 115 (2008).