Z. Naturforsch. 2015; 70(5-6)c: 145–150

Yong Zhang, Zhi-Hui Zhan, Shin-ichi Tebayashi, Chul-Sa Kim and Jing Li* Feeding stimulants for larvae of sarpedon nipponum (: Papilionidae) from

DOI 10.1515/znc-2014-4155 , including the tree. This Received September 2, 2014; revised January 20, 2015; accepted belongs to the tribe Graphiini, which is placed between June 17, 2015 the Troidini and Papilionini in the evolutionary tree [5]. Thus, chemicals related to the host selection of of the Graphium should be given attention to elucidate Abstract: The feeding response of larvae of the swallow- evolutionary processes in the Papilionidae family. tail , nipponum (Lepidoptera: Graphium sarpedon nipponum, the common blue- Papilionidae), is elicited by a methanolic extract from bottle, is one of the Lauraceae-feeding species, which camphor tree (Cinnamomum camphora) leaves. Based on usually utilizes Cinnamomum camphora as its major host. bioassay-guided fractionation, three compounds, isolated Adults of this butterfly are attracted by volatile compo- from the methanolic extract of fresh leaves of the camphor nents from C. camphora, which have been found to act as tree, were revealed to be involved in a multi-component olfactory cues in determining the choice of food plants system of feeding stimulants. Structures of these feeding by G. sarpedon nipponum [6]. For a better understanding stimulants were identified as sucrose, 5-O-caffeoylquinic of the physiochemical background of present-day host acid and quercetin 3-O-β-glucopyranoside by NMR and utilization and host range evolution in Graphium but- LC-MS. terflies, further attempts are needed to investigate the chemical basis for differential acceptance of a potential Keywords: Cinnamomum camphora; feeding stimulants; host plant. Graphium sarpedon nipponum. Previously we reported the isolation and characteri- zation of α-linolenic acid as a feeding stimulant from the hexane fraction of a C. camphora methanolic extract [5]. As part of our ongoing study on feeding stimulants from 1 Introduction C. camphora, we have now investigated the polar fraction of C. camphora fresh leaves and isolated three additional Most butterfly species are phytophagous and usually utilize feeding stimulants for G. sarpedon nipponum. a limited range of host plants in nature. Although the host range of an insect is determined by a diversity of ecologi- cal, geographical, physiological and behavioral factors, the key elements underlying the host range determination are 2 Materials and methods phytochemicals [1, 2]. As for larval , plant chemicals released from the leaf typically influence the decision as 2.1 Insect and plant to whether or not the feeding will be continued [3, 4]. A number of species in the genus Eggs of G. sarpedon nipponum were collected from young branches of Graphium commonly feed on species of the plant family C. camphora trees growing on Monobe Campus of Kochi University, . Larvae were kept at 25±3 °C and 70% relative humidity in Petri dishes with a 15:9 h, L:D photoperiod. The animate fifth instar larvae were used for bioassay. Fresh leaves of C. camphora were collected *Corresponding author: Jing Li, Ninth People’s Hospital of from the same source and used for extraction and bioassay. Chongqing, Beibei District, Chongqing 400715, , Fax: +86-023-6825-0994, E-mail: [email protected] Yong Zhang: College of Resources and Environment, Southwest University, Beibei District, Chongqing, China 2.2 Bioassay Zhi-Hui Zhan, Shin-ichi Tebayashi and Chul-Sa Kim: Faculty of Agriculture, Department of Agriculture, Kochi University, B200 The behavioral bioassay for the isolation of attractants and stimu- Monobe, Nankoku, Japan lants in the initiation of the feeding response were carried out as 146 Zhang et al.: Feeding Stimulants from Cinnamomum camphora previously described [5]. Activities were expressed as gram leaf 2.4 Instruments equivalents (g.l.e.) per semicircular Styrofoam test discs (45 mm in diameter, 0.7 mm in thickness). Samples of 0.5 g.l.e. were applied HPLC for the isolation and analysis of compounds was carried out onto the discs. As the control, the same volume of a solvent blank with a Shimadzu LC-6AD pump equipped with a Shimadzu SPD- was added to a disc. After drying, discs were placed in a Kimwipes M10A detector (Shimadzu, Kyoto, Japan). NMR data were obtained box (13 cm × 12 cm × 9 cm) with the upper end open, and each disc was on a JEOL JNM-L400 spectrometer (400 MHz, JEOL, Akishima, inserted into a slit of foam bottom, thereby kept in upright position. Japan) in CD OD (Sceti, Tokyo, Japan) with TMS (Acros, NJ, USA) as Usually one disc corresponded to one test sample, and the appropri- 3 an internal standard. Letters (br.) s, d, t, q and m represent (broad) ate number of discs were introduced into the box for bioassay. Three singlet, doublet, triplet, quartet and multiplet, respectively, and fifth instar larvae, which had been starved for 6 h, were left in the coupling constants are given in Hz. LC-MS data were measured by box and their feeding behavior was observed for 24 h (25±3 °C, 15:9 a Shimadzu LC-MS 2010 liquid chromatography-mass spectrometer h, D:L photoperiod). Styrofoam discs containing attractants or stim- in APCI mode (Shimadzu, Kyoto, Japan). ulants would be consumed by the larvae, and the feeding damage (consumed area in mm2) of discs was measured by a flat-bed scan- ner. Each test was repeated at least five times, and consumed areas of Styrofoam discs were averaged to serve as the index of the feeding stimulatory effect. Statistical analysis used one-way analysis of vari- ance (ANOVA) followed by Tukey HSD test (p < 0.05). 3 Results and discussion

3.1 Chemical compounds in aqueous layer 2.3 Extraction and fractionation A series of sugars (compounds 1–4) were isolated from Fresh leaves (1.9 kg) of C. camphora were cut into pieces and extracted fraction A. According to their diagnostic 1H-NMR and twice with 80% (v/v) aqueous methanol (Nacalai Tesque, Kyoto, 13 Japan) at room temperature for three days under darkness. The C-NMR spectra and by comparison of these with authen- combined extracts were filtered and the solvent was removed under tic samples [7], compounds 1, 2, 3 and 4 were identified as reduced pressure. The residue (118.2 g) was redissolved in water rhamnose, glucose, fructose and sucrose, respectively. Of (2.8 L) and partitioned with n-hexane (Nacalai Tesque, Kyoto, Japan) these four sugars, sucrose (11.75 mg/g.l.e.) was the princi- (2 L) for four times, yielding the n-hexane and aqueous layer, respec- pal constituent with approximately 76.8%. tively. Each layer was dried under reduced pressure and subjected to From fraction B, compound 5 was identified as the bioassay. The aqueous layer (63 g) was separated into four frac- tions by a reversed phase open column (ODS, 50 mm i.d. × 500 mm, 5-O-caffeoylquinic acid (5-CQA, 0.477 mg/g.l.e.) accord- Wako Pure Chemical Industries, Osaka, Japan), eluting in sequence ing to its diagnostic 1H-NMR and 13C-NMR spectra and with an increasing concentration of methanol in water. Finally, the by comparison with an authentic sample. The 13C-NMR aqueous fraction A (28.86 g), the 20% MeOH/ H O fraction B (8.67 2 spectrum indicated the presence of six carbons, includ- g), the 40% MeOH/ H O fraction C (7.60 g) and the MeOH fraction D 2 ing two carbonyl groups at δ175.1 (C-7) and δ168.8 (C-9′), (6.68 g) were obtained. Fraction A was chromatographed on a reverse-phase HPLC respectively; two aromatic carbons bonded to hydroxyl groups at δ148.3 (C-3′) and δ145.5 (C-4′); two olefinic ­column (Shiseido Capcell pak NH2 UG120Å 4.6 mm × 250 mm, Shiseido,

Japan) eluted with 75% CH3CN/H2O (Nacalai Tesque, Kyoto, Japan) carbons at δ144.9 and δ121.3 corresponding to C-7′ and at a flow rate of 1.0 mL/min, and four compounds were obtained. C-8′. Four aromatic carbons were assigned C-1′, C-2′, C-5′ Compound 1 was isolated at Rt = 5.92 min (0.71 mg/g.l.e.), com- and C-6′ at δ125.5, δ114.3, δ114.7 and δ115.7, respectively; pound 2 at Rt = 7.21 min (2.08 mg/g.l.e.), compound 3 at Rt = 8.12 min three carbons, bonded to hydroxyl groups at δ7 3 .7, δ68.4 (0.76 mg/g.l.e.) and compound 4 at Rt = 10.92 min (11.75 mg/g.l.e.). Fraction B was purified on a reverse-phase HPLC column and δ70.9, identified as C-1, C-3 and C-4; one carbon δ (Shiseido Capcell pak C18 UG120Å 10 mm × 250 mm) and eluted with bonded to an ester group at 70.6 attributed to C-5; and

20% MeOH/H2O (1% AcOH) at a flow rate of 3.0 mL/min to yield two methylene groups identified as C-2 and C-6 at δ37.2 ­compound 5 (Rt = 17.96 min, 0.48 mg/g.l.e.). and δ36.6, respectively. The 1H-NMR spectrum displayed In the primary bioassay, fraction C did not stimulate feeding of two ortho-coupled doublets each for 1H, at δ6.70 and the G. sarpedon nipponum larvae and was not subjected to further δ δ fractionation. 6.97, and a broad singlet for 1H at 7.01, confirming the presence of a tri-substituted aromatic ring; and two dou- Fraction D was separated into three fractions: fraction D1,

0–10.6 min, fraction D2, 10.6–12.5 min, fraction D3, 12.5–20.0 min. blets, each for 1H, at δ6.14 (H-7′) and δ7.41 (H-8′), indi-

This was done by reverse-phase HPLC (Shiseido Capcell pak C18 cating the presence of a trans-di-substituted ethylene UG120Å 10 mm × 250 mm), eluted with 20% CH CN/H O (1% AcOH). 3 2 moiety in the molecule [9]. Compounds 6, 7 and 8 were isolated from fraction D by reverse- 3 5-CQA (5) was obtained as white amorphous powder. phase HPLC (Shiseido Capcell pak C UG120Å 10 mm × 250 mm) at 18 Further, LC-MS: m/z (%) 354 (80), 276 (57), 115 (20), a flow rate of 3.0 mL/min, eluted with 20% CH3CN/H2O (1%AcOH), 1 at Rt = 13.01 min (13 μg/g.l.e.), Rt = 13.72 min (18 μg/g.l.e.) and H-NMR δH (CDCl3): 1.77 (1H, m, H-2a), 1.95 (1H, m, H-2b), Rt = 16.20 min (22 μg/g.l.e.), respectively. 1.95 (1H, m, H-6a), 1.96 (2H, m, H-6b), 3.55 (1H, m, H-4), Zhang et al.: Feeding Stimulants from Cinnamomum camphora 147

3.91 (1H, m, H-5), 5.06 (1H, m, H-3), 6.14 (1H, d, J = 15.8, 300 H-7′), 6.70 (1H, d, H-5′), 6.97 (1H, d, H-6′), 7.01 (1H, br s, 13 250 H-2′), 7.41 (1H, d, J = 15.8, H-8′). C-NMR δc (CDCl3): 36.6 2 (C-6), 37.2 (C-2), 68.4 (C-3), 70.6 (C-5), 70.9 (C-4), 73.7 (C-1), 200 114.3 (C-2′), 114.7 (C-5′), 115.7 (C-6′), 121.3 (C-8′), 125.5 (C-1′), 144.9 (C-7′), 145.5 (C-4′), 148.3 (C-3′), 168.8 (C-9′), 150 175.1 (C-7). Subsequent fractionation of fraction D by HPLC 100 Consumed areas, mm resulted in the isolation of a series of flavonoid glycosides 50 (compounds 6–8) from fraction D3. Compounds 6, 7 and 8 were identified as quercetin 3-O-β-galactopyranoside, 0 Control Aqueous ABCD(n=5) quercetin 3-O-β-glucopyranoside (0.42 mg/g.l.e.) and layer kaempferol 3-O-β-rutinoside, respectively, from their diagnostic 1H-NMR and 13C-NMR spectra and by compari- Figure 1: Feeding response (mean±SE) of G. sarpedon nipponum son of these with authentic samples [8, 10]. Among these toward the aqueous layer and individual fractions A–D from the three compounds, only quercetin 3-O-β-glucopyranoside aqueous layer. Control: Styrofoam discs treated with blank solvent. Significant difference between the control and the treated Styro- was active toward larvae of G. sarpedon nipponum. Its foam discs is presented by different letters. Each sample was tested NMR data are given following. The inactive compounds at a total of 0.5 g.l.e. (p < 0.05, n = 5). quercetin 3-O-β-galactopyranoside and kaempferol 3-O-β- rutinoside have been described previously [8, 10]. 300 Quercetin 3-O-β-glucopyranoside (7) was isolated as a yellow amorphous powder. LC-MS: m/z (%) 465 (38), 303 250

1 2 (82), 149 (29). H-NMR δH (CDCl3): 6.25 (1H, s, H-6), 6.46 (1H, s, H-8), 7.64 (1H, s, H-2′), 6.89 (1H, d, J = 9.9, H-5′), 7.63 200 (1H, s, J = 9.9, H-6′), 5.52 (1H, d, J = 7.3, glc-1), 3.14–3.65 (glc). 13 150 C-NMR δC (CDCl3): 156.5 (C-2), 133.5 (C-3), 177.6 (C-4), 161.5 (C-5), 98.9 (C-6), 164.5 (C-7), 93.6 (C-8), 156.3 (C-9), 104.0 100

(C-10), 121.7 (C-1′), 115.3 (C-2′), 144.9 (C-3′), 148.6 (C-4′), 116.3 Consumed areas, mm (C-5′), 121.3 (C-6′), 101.9 (glc-1), 74.2 (glc-2), 76.6 (glc-3), 70.1 50 (glc-4), 77.6 (glc-5), 61.1 (glc-6). 0 Control Aqueous A+B+C A+B+D A+C+D B+C+D (n=5) layer

3.2 Larval feeding stimulant activity of Figure 2: Feeding response (mean±SE) of G. sarpedon nipponum fractions toward the aqueous layer and the fraction mixtures: A+B+C, A+B+D, A+C+D, B+C+D (each fraction equally represented 1/3). Control: Sty- < = Our previous work had revealed that both the n-hexane rofoam discs treated with blank solvent (p 0.05, n 5). For further details refer to Figure 1. and aqueous layer derived from a methanolic extract of C. camphora, strongly evoked a positive response (ini- tiation of feeding) toward G. sarpedon nipponum respec- 3.3 Stimulation of larval feeding activity by tively [5]. In the present study, the aqueous layer from the isolated compounds methanolic extract was further separated into the four fractions A, B, C and D. No individual fraction stimulated Continuous tracking of stimulants from the aqueous feeding of G. sarpedon nipponum larvae by itself, when layer resulted in the isolation of three stimulants: sucrose assayed at the dose of 0.5 g.l.e. (Figure 1). However, the (compound 4) from fraction A, 5-CQA (compound 5) combination of (A+B+D) at a total of 0.5 g.l.e. stimulated from fraction B, and quercetin 3-O-β-glucopyranoside feeding activity toward G. sarpedon nipponum as strongly (­compound 7) from fraction D. The bioassay result as the original active aqueous layer (Figure 2). These revealed that the combination (A+B+D) or (B+D+4) stim- results indicate that some compounds jointly involved ulated feeding, whereas the combination (B+D) did not, in the feeding stimulant activity should reside in these indicating that 4 was involved in feeding stimulation of three fractions. G. ­sarpedon nipponum (Figure 3). But fraction A (with 148 Zhang et al.: Feeding Stimulants from Cinnamomum camphora

300 300

250 250 2 2 200 200

150 150 100

100 Consumed areas, mm Consumed areas, mm 50 50 0 Control 4+5 4+5+6 4+5+7 4+5+8 4+5+D (n=5) 0 Control A+B+D Compound4 B+D (n=5) +B+D Figure 5: Feeding response (mean±SE) of G. sarpedon nipponum toward compounds 4–8 in various combinations. 4: sucrose, 5: Figure 3: Feeding response (mean±SE) of G. sarpedon nipponum 5-O-caffeoylquinic acid, 6: quercetin 3-O-β-galactopyranoside, 7: toward combinations of fractions A, B, and D, or compound 4. quercetin 3-O-β-glucopyranoside, 8: kaempferol 3-O-β-rutinoside. Control: Styrofoam discs treated with blank solvent (p < 0.05, n = 5). Control: Styrofoam discs treated with blank solvent (p < 0.05, n = 5). For further details refer to Figure 1. For further details refer to Figure 1.

the principle constitute sucrose) alone did not stimulate glycosides in fraction D, feeding responses to the blends feeding, thus, sucrose is just a co-stimulant. Similarly, (4+5+6) and (4+5+8) were compared with those to the the active combination (A+B+D) lacking 5 (from frac- active fractions. Either of these mixtures did not exhibit tion B) or fraction B did not affect the feeding behavior feeding-stimulant activity (Figure 5). This suggested of G. sarpedon nipponum. This suggests that 5-CQA is an that quercetin 3-O-β-glucopyranoside (7) was the active essential component in the stimulation of the feeding feeding stimulating compound, whereas neither 6 nor 8 response of G. sarpedon nipponum (Figure 4). A mixture of were involved in the multi-component system of feeding stimulants (4+5) was almost inactive, but their effect was stimulants. Fraction D alone did not stimulate feeding, strongly synergized by quercetin 3-O-β-glucopyranoside indicating that the feeding-stimulant activity of 7 in this (­compound 7 from fraction D, Figure 5). To examine the fraction might be counteracted by some compounds, such feeding-stimulant activities of the other two flavonoid as compound 6, 8 or some unknown compounds exhibit- ing antifeedant activity. The present results demonstrate for the first time that sucrose, 5-CQA and quercetin 3-O-β- 300 glucopyranoside are involved in the multi-component system of feeding stimulants of G. sarpedon nipponum. 250 The strong feeding stimulation by essential nutri- 2 200 ents can be universally observed throughout the kingdom; sugars in particular are known as general 150 feeding stimulants in herbivorous insects, the larvae of which develop specific receptors tuned for these primary 100 metabolites [11, 12]. Sugars 1–4 appear to provide a basic Consumed areas, mm nutrient, and the principal constituent, sucrose (76.8% in 50 fraction A), indeed plays an important role in the feeding- stimulant activity toward G. sarpedon nipponum (Figure 1). 0 Control A+B+D Compound5 A+D (n=5) 5-CQA (from fraction B) is the most commonly found – +A+D and the only commercially available – isomer of chlo- rogenic acid [9]. The phenolic compound, chlorogenic Figure 4: Feeding response (mean±SE) of G. sarpedon ­nipponum toward combination of fractions A, B, and D, or compound 5. acid, is widely distributed in the plant kingdom [13, 14]. It Control: Styrofoam discs treated with blank solvent (p < 0.05, n = 5). encompasses several isomers, such as 3-caffeolyl-muco- For further details refer to Figure 1. quinic acid (3-CmQA), 4-caffeoylquinic acid (4-CQA) and Zhang et al.: Feeding Stimulants from Cinnamomum camphora 149

5-caffeoylquinic acid (5-CQA), and these are known to play Acknowledgments: This work was supported in part by various roles in plant defense [15], or to have oviposition- the Fundamental Research Foundation for the Central stimulant activity for polyxenes and Papilio protenor Universities (XDJK2013C156 and SWU113016 to Y.Z.). [14, 16]. Zebra swallowtail females are stimulated strongly by a single of these compound, i.e. 3-CmQA [17]. In the present study, 5-CQA has been identified as a feeding stim- References ulant for G. sarpedon nipponum for the first time. Flavonoid glycosides are known as oviposition 1. Städler ES. Behavioral responses of insects to plant secondary stimulants for some -feeding swallowtails. compounds. In: Herbivores: their interactions with secondary plant metabolites: ecological and evolutionary processes. San From methanolic extracts of one of its host plants, Diego: Academic Press, 1992:45–88. unshiu, four flavonoid oviposition stimulants for the 2. Chachin M, Honda K, Omura H. Appraisal of the acceptability of swallowtail were identified [4]; from the subtropical rutaceous plants for a swallowtail butterfly, Papilio epicarp of Citrus natsudaidai, two flavonoid glycosides, protenor demetrius (Lepidoptera: Papilionidae). Appl Entomol hesperidin and naringin, were isolated that stimulated Zool 2007;42:121–8. 3. Städler E, Roessingh P. Perception of surface chemicals by feeding­ oviposition of Papilio proteor [18]. Flavonoid glycosides and ovipositing insects. Symp Biol Hung 1990;39:71–86. from C. camphora have apparently not been reported 4. Nishida R, Ohsugi T, Kokubo S, Fukami H. Oviposition stimu- previously as feeding-stimulants. In this study, querce- lants of a citrus-feeding swallowtail butterfly, Papilio xuthus L. tin 3-O-β-glucopyranoside was identified as a feeding Experientia 1987;43:342–4. stimulant. 5. Li J, Wakui R, Horie M, Nishimura Y, Nishiyama Y, Ikeno Y, et al. Occurrence of some feeding stimulant compounds Feeding stimulant in Cinnamomum camphora for the common bluebottle, Graphium sarpedon nipponum (Lepidoptera: in appropriate concentrations in plants other than host Papilionidae). Z Naturforsch 2010;65c:571–6. plants might mislead female to oviposit on 6. Li J, Wakui R, Tebayashi S, Kim CS. Volatile attractants for the non-host plants. Oviposition “mistakes” by females common bluebottle, Graphium sarpedon nipponum, from the would provide a chance to colonize novel host plants [19]. host, Cinnamomum camphora. Biosci Biotechnol Biochem Sucrose, 5-CQA and quercetin 3-O-β-glucopyranoside are 2010;74:1987–90. 7. Yan ZL, Yang MS, Li TB, Feng ZM. Quantitative analysis of common compounds widespread in the plant kingdom. glucose, fructose and sucrose in honey by nuclear magnetic Therefore, it might be advantageous for G. sarpedon nip- resonance spectroscopy. Shipin Kexue 2009;30:253–5. ponum to utilize these compounds as food cues to exploit 8. Güvenalp Z, Demĭrezer LÖ. Flavonol glycosides from Asperula additional food resources with a similar phytochemical arvensis L. Turk J Chem 2005;29:163–9. profile. The combination of sucrose, 5-CQA and quercetin 9. Suárez-Quiroz ML, Alonso Campos A, Valerio Alfaro G, 3-O-β-glucopyranoside includes both primary and sec- ­González-Ríos O. Isolation of green coffee chlorogenic acids using activated carbon. J Food Compos Anal 2014;33:55–8. ondary metabolites. Primary metabolites are generally 10. Vermes B, Farkas L, Nógrádi M, Wagner H, Dirscherl R. The syn- important for larval growth and survival [20], whereas sec- thesis of afzelin, paeonoside and kaempferol 3-O-β-rutinoside. ondary metabolites are essential for stimulating the feeding Phytochemistry 1976;15:1320–1. response of G. sarpedon nipponum [5]. This may explain 11. Bernays EA, Simpson SJ. Control of food intake. Adv Insect Phys why this particular combination exhibits feeding-stimulant 1982;16:59–118. 12. Schoonhoven LM, Loon JJ, Dicke M. Insect-plant biology. activity toward larvae of G. sarpedon nipponum. Moreover, New York: Oxford University Press, 2005. it is interesting to note that the three compounds act syn- 13. Herrmann K. Hydroxyzimtsäuren und Hydroxybenzoesäuren ergistically, because any individual compound or the arbi- enthaltende. Naturstoffe in Pflanzen. In: Herz W, et al., editors. trary combination of two compounds were inactive. Only Fortschr Chem Org Naturst, vol. 35. New York: Springer-Verlag, three compounds in combination synergistically stimulate 1978:73–132. ′ feeding activity. Similar synergistic effects among several 14. Feeny P, Sachdev K, Rosenberry L, Carter M. Luteolin 7-O-(6 -O- malonyl-β-d-glucoside and trans-chlorogenic acid: oviposition compounds, such as sugars and flavonoids, have been con- stimulants for the black swallowtail butterfly. Phytochemistry firmed in the feeding stimulation of Papilio xuthus [20]. 1988;27:3439–48. Many Graphiini species are strongly associated with 15. Bennett RN, Wallsgrove RM. Secondary metabolites in plant Lauraceae species, with either wide or narrow preferences defense mechanisms. New Phytol 1994;127:617–33. within this plant family. Systematic studies on the chem- 16. Honda K. Identification of host-plant chemicals stimulating oviposition by swallowtail butterfly, Papilio protenor. J Chem istry controlling feeding behavior among these closely Ecol 1990;16:325–37. related species should provide important information 17. Haribal M, Feeny P. Oviposition stimulant for the zebra on the evolutionary process of their host utilization and swallowtail butterfly, Eurytides marcellus, from the foliage of diversity. pawpaw, Asimina triloba. Chemoecology 1998;8:99–110. 150 Zhang et al.: Feeding Stimulants from Cinnamomum camphora

18. Honda K.Flavanone glycosides as oviposition stimulants xenes (Lepidoptera: Papilionidae). J Insect Behav 1989; in a papilionid butterfly, Papilio protenor. J Chem Ecol 2:803–27. 1986;12:1999–2010. 20. Murata T, Mori N, Nishida R. Larval feeding stimulants for a 19. Feeny P, Stadler E, Ahman I, Carter M. Effect of plant odor on Rutaceae-feeding swallowtail butterfly, Papilio xuthus L. in oviposition by the black swallowtail butterfly, Papilio poly­ leaves. J Chem Ecol 2011;37:1099–109.