J Chem Ecol DOI 10.1007/s10886-015-0563-0
Semen-Like Floral Scents and Pollination Biology of a Sapromyophilous Plant Stemona japonica (Stemonaceae) Gao Chen 1 & Andreas Jürgens 2 & Lidong Shao 3,4 & Yang Liu 5 & Weibang Sun 1 & Chengfeng Xia 3
Received: 23 July 2014 / Revised: 18 December 2014 / Accepted: 3 February 2015 # Springer Science+Business Media New York 2015
Abstract By emitting scent resembling that of organic material suitable for oviposition and/or consumption by flies, sapromyophilous flowers use these flies as pollinators. To date, intensive scent analyses of such flowers have been restricted to Apocynaceae, Annonaceae, and Araceae. Recent studies have suggested that the wide range of volatile organic compounds (VOCs) from sapromyophilous flowers play an important role in attracting saprophagous flies by mimicking different types of decomposing substrates (herbivore and carnivore feces, carrion, and the fruiting bodies of fungi, etc.). In this study, we report the flower visitors and the floral VOCs of Stemona japonica (Blume) Miquel, a species native to China. The flowers do not produce rewards, and pollinators were not observed consuming pollen, thus suggesting a deceptive pollination system. Headspace samples of the floral scent were collected via solid-phase micro-extraction and analysed by gas chromatography coupled with mass spectrometry. Main
* Prof. Weibang Sun [email protected] * Prof. Chengfeng Xia [email protected] 1
Kunming Botanical Garden, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204, Yunnan, China
2
School of Life Sciences, University of KwaZulu-Natal, P. Bag X01 Scottsville, Pietermaritzburg 3209, South Africa
3
State Key Laboratory of Phytochemistry and Plant Resources in West China, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204, China
4
University of Chinese Academy of Sciences, Beijing 100049, China
5
Shenzhen Center for Disease Control and Prevention, Shenzhen 518055, China
floral scent compounds were 1-pyrroline (59.2 %), 2-methyl1-butanol (27.2 %), and 3-methyl-1-butanol (8.8 %), and resulted in a semen-like odor of blooming flowers. The floral constituents of S. japonica were significantly different from those found in previous sapromyophilous plants. An olfaction test indicated that 1-pyrroline is responsible for the semen-like odor in S. japonica flowers. Main flower visitors were shoot flies of the genus Atherigona (Muscidae). Bioassays using a mixture of all identified floral volatiles revealed that the synthetic volatiles can attract Atherigona flies in natural habitats. Our results suggest that the foul-smelling flowers of S. japonica may represent a new type of sapromyophily through scent mimicry.
Keywords 1-Pyrroline . Deceit pollination . GC/MS . Atherigona . Stemonaceae . Mimics
Introduction Deceit pollination is probably the most extreme example of unreliable signaling in plants, because flowers do not offer any reward but benefit from pollinator visitation (Jürgens et al. 2013; Proctor et al. 1996; Renner 2006). This phenomenon has been recorded in more than 7500 different animalpollinated angiosperms (Renner 2006). The potential reason of deception often arises from an insect’s inability to distinguish between rewarding and non-rewarding flowers (Jürgens et al. 2013; Raguso 2004; Renner 2006). Well-known examples for deceit pollination often include food deceit, sexual deceit, and oviposition site deceit, etc. (Jersáková et al. 2006; Vereecken and McNeil 2010; Urru et al. 2011).
J Chem Ecol
Sapromyophily, or pollination by attraction of carrion and dung-feeding flies, has evolved many times during the angiosperm radiation (Jürgens et al. 2013; Urru et al. 2011; Vereecken and McNeil 2010). To lure saprophilous flies to visit their flowers, these plants deploy the scent and visual (and sometimes even temperature) signals that these insects use to locate an oviposition site or a food source on animal dung or carrion (Jürgens et al. 2006; Stensmyr et al. 2002; Vereecken and McNeil 2010). Typical characteristics of sapromyophily include flowers that are usually purple or dull red in color and emit a strong scent, unpleasant to the human nose, reminiscent of decaying organic matter (Jürgens et al. 2006, 2013; Proctor et al. 1996; Vereecken and McNeil 2010). The comparative analysis of data from the literature compiled by Jürgens et al. (2013) suggests that the scent compounds produced by plants exploit a wide range of diverse insects. Different sources of putrefaction are typified by very distinct and specific volatile compounds; for example, urine (carboxylic acids and pyrazines), rotting carcasses (heptanal and octanal), decaying meat and carnivore dung (oligosulfides), herbivore dung (indole and cresol), and rotting fruit (oxygenated aliphatic compounds) (Jürgens et al. 2006, 2013). The Stemonaceae is a small monocotyledonous family comprising four genera and about 32 species (Ji and Duyfjes 2000). The largest genus, Stemona Loureiro, consists of about 27 species, many of which are used traditionally for medicinal purposes in Southeast Asia (Ji and Duyfjes 2000). While investigating species of this family in China for their medicinal properties, we discovered that the dull red flowers of Stemona japonica (Blume) Miquel (Fig. 1) emit a noticeably strong, unpleasant foul and semen-like smell. The flower scent and also the dull red coloration of these flowers lead us to assume the hypothesis that this species is sapromyophilous. Although some Stemona species, e.g., S. tuberosa and S. javanica, have been reported previously to produce a strong putrid smell reminiscent of carrion or cheese (Duyfjes 1991), floral volatiles from species belonging to the family Stemonaceae have not been investigated (Knudsen et al. 2006). To the human nose, the unusual semen-like scent of the flowers of S. japonica suggests that the system does not
fall into one of the typical floral scent categories of sapromyophily (Jürgens et al. 2006, 2013). The aims of our study were to investigate the pollination biology and floral scent composition of S. japonica and to test whether S. japonica produces scent compounds different from those reported for species belonging to the sapromyophilous scent syndrome (see Jürgens et al. 2013). We also conducted bioassays to test whether isolated floral volatiles are capable of attracting potential pollinators of S. japonica.
Methods and Materials Study Sites and Plant Materials Stemona japonica has an erect stem with a woody base producing slender vines (Fig. 1) with axillary inflorescences (cymes) that consist of one to several flowers. The peduncle or pedicels of the inflorescences are borne on the leaf midvein. The often reflexed tepals, up to 1.5 cm long, are light green, and the purple stamens are slightly shorter than the perianth (Fig. 1). Filaments are only 1 mm long, and the apex of the ca. 2.5 mm long anthers carries one arrow-shaped appendage with an erect filament on either side. The capsule usually contains two or three seeds. The main flowering time of S. japonica in China is from May to September. The species is commonly distributed on hillsides and in densely grassy roadsides in Anhui, Fujian, Hubei, Jiangsu, Jiangxi, and Zhejiang provinces in China, and it is naturalized in Japan (Ji and Duyfjes 2000). Investigations of the breeding system and pollination biology were conducted at three different localities in 2011–2014. The two natural populations used for this study of S. japonica were located at Dingjiashan, on a hill in the middle of Hangzhou City (DJS: 30.293 N and 120.136 E, 53 m.a.s.l.) and in the Tianmushan National Natural Reserve (TMS: 30.321 N and 119.434 E, 437 m.a.s.l.), in the Zhejiang Province in Eastern China. Furthermore, we transplanted individuals from TMS to the Kunming Botanical Garden of the Kunming Institute of Botany at the Chinese Academy of Sciences (KBG: 25.127 N and 102.743 E, 1788 m.a.s.l.) in
Fig. 1 Flowers of Stemona japonica and their flower visitors: (a) two Atherigona flies visiting a flower simultaneously. (b), Atherigona fly visiting a flower of S. japonica. (c) a fly visiting while another approaches the flower from upwind (arrow). Scale bar is 4 mm in a, b, and c
J Chem Ecol
the Yunnan Province in Southwest China. Two voucher specimens of the species from each site have been deposited in the Herbarium of Kunming Institute of Botany (No. CHEN 20120711-14). Floral Biology To describe the floral anthesis and to determine flower longevity, 20 flower buds from 10 individuals (two buds per individual) at the TMS population were tagged with thin inconspicuous green threads and monitored in July 2011. To determine if flowers of S. japonica attract pollinators by deceit pollination, 30 flowers at both natural populations (TMS and DJS) were examined for nectar. During the flowering period of each single flower, we checked from 09:30 to 18:30 if nectar was secreted at the base of the flower. The reproductive success (fruit set) of S. japonica was examined in natural populations at TMS and DJS in July in 2011 and 2012, respectively. At each population, 50 flower buds on 10 individual plants were randomly selected, labeled, and enclosed in fine nylon nets (one flower per net) to exclude flower visitors and allow only spontaneous autogamous self-pollination. Untreated, open-pollinated flowers (319 flowers from 26 individuals at TMS, 411 flowers from 34 individuals at DJS) served as controls. Fruit set was recorded in both populations about 45 d after main flowering peak. Ninety-two flowers from 10 individuals were cross-pollinated by hand at the KBG location to examine whether plants can produce fruits when introduced to a new habitat. Furthermore, 87 untreated flower buds from 10 individuals at the KBG location were tagged with thin green thread and left to open pollination to examine if effective pollinators of the species occurred at this unnatural habitat. Flower Visitor Observations Floral visitors of S. japonica were observed at the TMS population (a National Natural Reserve), which is situated within the natural range of this species. In 2011, flowering of the population started June 11th and ended September 7th. Floral visitors were observed continuously during the flower opening time from 09:30 to 18:30 for 6 d from July 9th to 14th. Flower visitor observations were conducted in parallel by three observers (3×5 flowers observed per day). In 2012, the flowering lasted from May 28th to August 27th, and observations were carried out from July 13th to 20th by two observers using the same observation protocol as in 2011. During flower visitor observation, flower visitor behavior, such as movement on the flower and contact with floral organs, was recorded and it was checked whether flower visitors were harvesting/gathering pollen as foragers or acquiring pollen incidentally on their bodies. Some flower visitors were caught and preserved for later identification by an insect taxonomist.
Collection and Analysis of Floral Volatiles Floral scent of S. japonica was collected at KBG during the time of highest floral scent emission (12:00–14:00) using solid-phase microextraction (SPME). Three flowers of S. japonica were cut from one individual and the flower stalk was wrapped with wet cotton. The material was placed in a test tube (15 cm long, 3 cm diam) and covered with a sealing film. The SPME holder with a 65 μm polydimethylsiloxane/divinylbenzene fiber (Supelco, Bellefonte, PA, USA; previously desorbed for 5 min in a GC injection port heated to 200 °C) was inserted directly into the atmosphere surrounding the three flowers; the fiber was exposed for 45 min and immediately analysed via GC/MS. Five replicates of a total of 15 fresh flowers were sampled, and an empty conical flask (500 ml) with wet cotton wrapped around three leaves without flowers acted as a control. Samples from flowers and controls were analyzed using an Agilent Technologies HP 6890 gas chromatograph, equipped with an HP-5MS column (30 m×0.25 mm inner diam, 0.25 μm film thickness), and linked to an HP 5973 mass spectrometer. Helium was used as a carrier gas, at a flow rate of 1 ml/min. Split inlet and FID were held at 250 °C. Column temperature was programmed to rise from 40 °C (5-min hold) to 250 °C (20-min hold) at 3 °C/min. Compounds were identified by comparing mass spectra and relative retention times with those of standard compounds and with the Wiley NIST 05 mass spectral database, and their relative proportions (%) were determined by peak area measurements. To scrutinize the identification of 1-pyrroline carefully, a natural scent sample of S. japonica using SPME was collected. The fiber then was subsequently exposed to synthetic 1-pyrroline, and the volatiles were analyzed by GC/MS. Before using the SPME method, to evaluate the emission rate of floral scents from S. japonica, we used the dynamic headspace extraction method (DHEM) to collect floral scents. However, this method collects only two floral compounds (2methyl-1-butanol and 3-methyl-1-butanol) from three flowers. Because floral scents collected by DHEM must be eluted by solvent, we inferred that the scent of 1-pyrroline might be masked by the dichloromethane solvent. Actually, we can smell the special odor of 1-pyrroline in the eluent. However, we can not detect it by GC/MS. The SPME method can detect 1-pyrroline easily. Therefore, we used the DHEM to collect floral scents and calculated the amount of 2-methyl-1-butanol emission from flowers. We counted the percent of 2-methyl-1butanol in floral composition collected by SPME method. The total floral emission = (the amount of 2-methyl-1-butanol in DHEM) / (the percent of 2-methyl-1-butanol in SPME). Headspace samples were taken by enclosing three flowers of S. japonica in Tedlar bags (Dupont, USA) prior to sampling, and subsequently pumping air for 3 hr through filters containing 150 mg of Porapak Q (mesh 60/80, Waters Associates, Inc.) using a pump with an inlet flow rate of 300 ml/min. Before use, the adsorbent cartridges were cleaned with 2 ml
J Chem Ecol
of diethyl ether and dried with nitrogen gas. Trapped volatiles (N=3) were eluted with 400 μl dichloromethane, and concentrated to one-sixth of the original volume by a gentle stream of nitrogen. An empty bag with three leaves was used as a control. In each sample, 700 ng n-nonane were added first as an internal standard for quantification. All samples were stored at −20 °C before analysis. Samples were analyzed using the same method of SPME. Synthesis of 1-Pyrroline Standards for Comparison with Floral Scent 1-Pyrroline was synthesized according to the methods described by Mores et al. (2008). In a roundbottom flask covered with aluminum foil, AgNO3 (0.74 mmol, 128 mg) was added to a solution of pyrrolidine (SigmaAldrich, USA, purity >98 %; 0.15 mol, 10.66 g, 12.5 ml) in H2O (140 ml) plus NaOH pellets (0.3 mol, 12 g). To this mixture, a solution of 25 % Na2S2O8 (0.15 mol, 35.7 g, 142 ml) was added dropwise at 0 °C. The reaction mixture was stirred at room temperature for 3 hr and then extracted with dichloromethane (2×200 ml), after having added NaCl to the aqueous phase to saturation. After drying over Na2SO4 and removal of the solvent, the mixture was dissolved in ethyl ether (200 ml) for filtration through neutral alumina and evaporation to dryness. The end product, pale orange oil, then was used for NMR analysis. NMR Analysis of Synthesized 1-Pyrroline The yield of 1pyrroline was 54.0 %, which coexists with its corresponding trimer both in solution phase and in pure liquid state (Baker et al. 1992; Xiao et al. 2013). In this study, the rate of 1pyrroline trimer is about 20 % in CDCl3 and 26 % in D2O, which were determined from the integration of the Ha and He. The spectra data of 1-pyrroline are: Monomer: 1H NMR (400 MHz, CDCl3) δ 7.59 (s, 1H), 3.83 (t, J=7.3 Hz, 2H), 2.52 (m, 2H), 1.78 (m, 2H); Trimer: 1H NMR (400 MHz, CDCl3) δ 3.12-2.92 (m, 3H), 2.31 (dd, J=14.9, 8.5 Hz, 3H), 1.92-1.64 (m, 15H). Monomer: 1H NMR (400 MHz, D2O) δ 7.52 (s, 1H), 3.71-3.50 (m, 2H), 2.43 (tt, J=3.3, 2.2 Hz, 2H), 1.73-1.61 (m, 2H). 1H NMR (400 MHz, D2O) δ 3.04 (t, J= 6.4 Hz, 3H), 2.77 (td, J=8.8, 5.0 Hz, 3H), 2.28 (td, J=9.1, 6.4 Hz, 3H), 1.74-1.60 (m, 12H). Bioassay of the Ability of S. japonica Volatiles to Attract Flies The ability of volatiles emitted by S. japonica to attract flies was tested in bioassays. A solution containing five volatiles of S. japonica (in proportions similar to those of natural S. japonica extracts, as verified by GC-FID analyses) was prepared. This synthetic scent mixture consisted of 59.2 mg 1pyrroline (monomer and trimer mixture), 27.2 mg 2-methyl-1butanol, 8.8 mg 3-methyl-1-butanol, 3.6 mg 2-methylbutanal, and 1.2 mg 3-methylbutanal (four C5-branched compounds purchased from Sigma-Aldrich, 95–99 % purity) in 10 ml dichloromethane. A 100 μl sample of either the scent mixture (equivalent
to about 300 flowers) or dichloromethane control was dropped on a sample and control filter papers, respectively (Fig. 2a). All filter papers were 5 cm diam, and were soaked with dichloromethane for 24 hr and dried at room temperature to remove potential odor compounds before use. After the solvent had evaporated, scented and control filter papers were placed in the natural habitat of the S. japonica population (Fig. 2a), (13:00–14:00, from 6 to 15 July, 2014). If flies landed on the filter papers (Fig. 2b), we assumed that they were attracted by the sample or control. Every 5 min, the positions of the sample and control were changed to avoid location effects. All flies that landed on the filter paper were collected. Each test lasted 10 min and the bioassays were performed 27 times. Each paper was used only once for a bioassay. The bioassay was conducted at the TMS population. Before conducting our bioassay, we had thought that the semen-like odor of 1-pyrroline might play a role in the attraction of fly pollinators, because this compound is a major component (about 60 %) of the floral scent composition in S. japonica. The function of the compound was, therefore, tested in 2012. A 59.2 mg sample of 1-pyrroline (monomer and trimer mixture) was dissolved in 10 ml distilled water. In a trap bottle (Enjoy Technology, China), 100 μl samples were dissolved in 200 ml distilled water and used to attract Tetradacus flies, which are similar in size to the pollinator of S. japonica. Eleven scented and 11 control traps were placed in the TMS location, spaced at least 3 m apart. Flies captured in these traps were recorded daily, and the process continued for 5 d until the strong fetid odor from the scented traps dissipated. Testing the Similarity Between 1-Pyrroline and the Floral Scents of S. japonica To test whether 1-pyrroline represents the semen-like floral scent profile of S. japonica, a bioassay was conducted to test whether common people can distinguish 1-pyrroline from the floral scents of S. japonica. A 2 mg 1pyrroline sample was dissolved in 1 ml distilled water, and 10 μl of the resulting liquid were dropped onto a black cloth. A cloth of similar size was wrapped around 10 flowers of S. japonica. The two cloths were put into two 250 ml flasks, after which 32 volunteers were asked to smell the contents of the two flasks to determine odor similarity. Statistical Analyses Floral scent composition was assessed using one-sample t-tests, and the bioassay was assessed using an independent-sample t-test. SPSS 18.0 software (SPSS Inc., Armonk, NY, USA) was used for all the analyses.
Results Flower Biology We found that the flower longevity of S. japonica was about 10–12 hr. Flowers typically opened at about 8:30–9:30 in the morning and closed at about 18:30–
J Chem Ecol
Fig. 2 Bioassay to test the ability of volatiles emitted by Stemona japonica to attract flies. (a) illustration of the fly attraction bioassay using filter papers with or without scents; sample (left) and control (right). (b) attracted flies on a scented filter paper
around the bottom of the flower, the tepal, and the stamen. Flies acquired pollen accidentally on their bodies as they moved within the flowers. No grooming/gathering pollen behavior was observed in Atherigona. When the flies left the flowers, they carried pollen over their entire bodies. The flies deposited pollen grains on the stigma with their legs while visiting the flowers. The visit frequency per flower per hour was 0.09 in 2011 and 0.06 in 2012. The sex ratio of the 67 Atherigona specimens caught on flowers of S. japonica during 2011–2014, was 49 female vs. 18 male. Thus, the sex ratio of Atherigona flies was significantly different from 1:1 (χ2 =14.343, df=1, P
Semen-Like Floral Scents and Pollination Biology of a Sapromyophilous Plant Stemona japonica (Stemonaceae) Gao Chen 1 & Andreas Jürgens 2 & Lidong Shao 3,4 & Yang Liu 5 & Weibang Sun 1 & Chengfeng Xia 3
Received: 23 July 2014 / Revised: 18 December 2014 / Accepted: 3 February 2015 # Springer Science+Business Media New York 2015
Abstract By emitting scent resembling that of organic material suitable for oviposition and/or consumption by flies, sapromyophilous flowers use these flies as pollinators. To date, intensive scent analyses of such flowers have been restricted to Apocynaceae, Annonaceae, and Araceae. Recent studies have suggested that the wide range of volatile organic compounds (VOCs) from sapromyophilous flowers play an important role in attracting saprophagous flies by mimicking different types of decomposing substrates (herbivore and carnivore feces, carrion, and the fruiting bodies of fungi, etc.). In this study, we report the flower visitors and the floral VOCs of Stemona japonica (Blume) Miquel, a species native to China. The flowers do not produce rewards, and pollinators were not observed consuming pollen, thus suggesting a deceptive pollination system. Headspace samples of the floral scent were collected via solid-phase micro-extraction and analysed by gas chromatography coupled with mass spectrometry. Main
* Prof. Weibang Sun [email protected] * Prof. Chengfeng Xia [email protected] 1
Kunming Botanical Garden, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204, Yunnan, China
2
School of Life Sciences, University of KwaZulu-Natal, P. Bag X01 Scottsville, Pietermaritzburg 3209, South Africa
3
State Key Laboratory of Phytochemistry and Plant Resources in West China, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204, China
4
University of Chinese Academy of Sciences, Beijing 100049, China
5
Shenzhen Center for Disease Control and Prevention, Shenzhen 518055, China
floral scent compounds were 1-pyrroline (59.2 %), 2-methyl1-butanol (27.2 %), and 3-methyl-1-butanol (8.8 %), and resulted in a semen-like odor of blooming flowers. The floral constituents of S. japonica were significantly different from those found in previous sapromyophilous plants. An olfaction test indicated that 1-pyrroline is responsible for the semen-like odor in S. japonica flowers. Main flower visitors were shoot flies of the genus Atherigona (Muscidae). Bioassays using a mixture of all identified floral volatiles revealed that the synthetic volatiles can attract Atherigona flies in natural habitats. Our results suggest that the foul-smelling flowers of S. japonica may represent a new type of sapromyophily through scent mimicry.
Keywords 1-Pyrroline . Deceit pollination . GC/MS . Atherigona . Stemonaceae . Mimics
Introduction Deceit pollination is probably the most extreme example of unreliable signaling in plants, because flowers do not offer any reward but benefit from pollinator visitation (Jürgens et al. 2013; Proctor et al. 1996; Renner 2006). This phenomenon has been recorded in more than 7500 different animalpollinated angiosperms (Renner 2006). The potential reason of deception often arises from an insect’s inability to distinguish between rewarding and non-rewarding flowers (Jürgens et al. 2013; Raguso 2004; Renner 2006). Well-known examples for deceit pollination often include food deceit, sexual deceit, and oviposition site deceit, etc. (Jersáková et al. 2006; Vereecken and McNeil 2010; Urru et al. 2011).
J Chem Ecol
Sapromyophily, or pollination by attraction of carrion and dung-feeding flies, has evolved many times during the angiosperm radiation (Jürgens et al. 2013; Urru et al. 2011; Vereecken and McNeil 2010). To lure saprophilous flies to visit their flowers, these plants deploy the scent and visual (and sometimes even temperature) signals that these insects use to locate an oviposition site or a food source on animal dung or carrion (Jürgens et al. 2006; Stensmyr et al. 2002; Vereecken and McNeil 2010). Typical characteristics of sapromyophily include flowers that are usually purple or dull red in color and emit a strong scent, unpleasant to the human nose, reminiscent of decaying organic matter (Jürgens et al. 2006, 2013; Proctor et al. 1996; Vereecken and McNeil 2010). The comparative analysis of data from the literature compiled by Jürgens et al. (2013) suggests that the scent compounds produced by plants exploit a wide range of diverse insects. Different sources of putrefaction are typified by very distinct and specific volatile compounds; for example, urine (carboxylic acids and pyrazines), rotting carcasses (heptanal and octanal), decaying meat and carnivore dung (oligosulfides), herbivore dung (indole and cresol), and rotting fruit (oxygenated aliphatic compounds) (Jürgens et al. 2006, 2013). The Stemonaceae is a small monocotyledonous family comprising four genera and about 32 species (Ji and Duyfjes 2000). The largest genus, Stemona Loureiro, consists of about 27 species, many of which are used traditionally for medicinal purposes in Southeast Asia (Ji and Duyfjes 2000). While investigating species of this family in China for their medicinal properties, we discovered that the dull red flowers of Stemona japonica (Blume) Miquel (Fig. 1) emit a noticeably strong, unpleasant foul and semen-like smell. The flower scent and also the dull red coloration of these flowers lead us to assume the hypothesis that this species is sapromyophilous. Although some Stemona species, e.g., S. tuberosa and S. javanica, have been reported previously to produce a strong putrid smell reminiscent of carrion or cheese (Duyfjes 1991), floral volatiles from species belonging to the family Stemonaceae have not been investigated (Knudsen et al. 2006). To the human nose, the unusual semen-like scent of the flowers of S. japonica suggests that the system does not
fall into one of the typical floral scent categories of sapromyophily (Jürgens et al. 2006, 2013). The aims of our study were to investigate the pollination biology and floral scent composition of S. japonica and to test whether S. japonica produces scent compounds different from those reported for species belonging to the sapromyophilous scent syndrome (see Jürgens et al. 2013). We also conducted bioassays to test whether isolated floral volatiles are capable of attracting potential pollinators of S. japonica.
Methods and Materials Study Sites and Plant Materials Stemona japonica has an erect stem with a woody base producing slender vines (Fig. 1) with axillary inflorescences (cymes) that consist of one to several flowers. The peduncle or pedicels of the inflorescences are borne on the leaf midvein. The often reflexed tepals, up to 1.5 cm long, are light green, and the purple stamens are slightly shorter than the perianth (Fig. 1). Filaments are only 1 mm long, and the apex of the ca. 2.5 mm long anthers carries one arrow-shaped appendage with an erect filament on either side. The capsule usually contains two or three seeds. The main flowering time of S. japonica in China is from May to September. The species is commonly distributed on hillsides and in densely grassy roadsides in Anhui, Fujian, Hubei, Jiangsu, Jiangxi, and Zhejiang provinces in China, and it is naturalized in Japan (Ji and Duyfjes 2000). Investigations of the breeding system and pollination biology were conducted at three different localities in 2011–2014. The two natural populations used for this study of S. japonica were located at Dingjiashan, on a hill in the middle of Hangzhou City (DJS: 30.293 N and 120.136 E, 53 m.a.s.l.) and in the Tianmushan National Natural Reserve (TMS: 30.321 N and 119.434 E, 437 m.a.s.l.), in the Zhejiang Province in Eastern China. Furthermore, we transplanted individuals from TMS to the Kunming Botanical Garden of the Kunming Institute of Botany at the Chinese Academy of Sciences (KBG: 25.127 N and 102.743 E, 1788 m.a.s.l.) in
Fig. 1 Flowers of Stemona japonica and their flower visitors: (a) two Atherigona flies visiting a flower simultaneously. (b), Atherigona fly visiting a flower of S. japonica. (c) a fly visiting while another approaches the flower from upwind (arrow). Scale bar is 4 mm in a, b, and c
J Chem Ecol
the Yunnan Province in Southwest China. Two voucher specimens of the species from each site have been deposited in the Herbarium of Kunming Institute of Botany (No. CHEN 20120711-14). Floral Biology To describe the floral anthesis and to determine flower longevity, 20 flower buds from 10 individuals (two buds per individual) at the TMS population were tagged with thin inconspicuous green threads and monitored in July 2011. To determine if flowers of S. japonica attract pollinators by deceit pollination, 30 flowers at both natural populations (TMS and DJS) were examined for nectar. During the flowering period of each single flower, we checked from 09:30 to 18:30 if nectar was secreted at the base of the flower. The reproductive success (fruit set) of S. japonica was examined in natural populations at TMS and DJS in July in 2011 and 2012, respectively. At each population, 50 flower buds on 10 individual plants were randomly selected, labeled, and enclosed in fine nylon nets (one flower per net) to exclude flower visitors and allow only spontaneous autogamous self-pollination. Untreated, open-pollinated flowers (319 flowers from 26 individuals at TMS, 411 flowers from 34 individuals at DJS) served as controls. Fruit set was recorded in both populations about 45 d after main flowering peak. Ninety-two flowers from 10 individuals were cross-pollinated by hand at the KBG location to examine whether plants can produce fruits when introduced to a new habitat. Furthermore, 87 untreated flower buds from 10 individuals at the KBG location were tagged with thin green thread and left to open pollination to examine if effective pollinators of the species occurred at this unnatural habitat. Flower Visitor Observations Floral visitors of S. japonica were observed at the TMS population (a National Natural Reserve), which is situated within the natural range of this species. In 2011, flowering of the population started June 11th and ended September 7th. Floral visitors were observed continuously during the flower opening time from 09:30 to 18:30 for 6 d from July 9th to 14th. Flower visitor observations were conducted in parallel by three observers (3×5 flowers observed per day). In 2012, the flowering lasted from May 28th to August 27th, and observations were carried out from July 13th to 20th by two observers using the same observation protocol as in 2011. During flower visitor observation, flower visitor behavior, such as movement on the flower and contact with floral organs, was recorded and it was checked whether flower visitors were harvesting/gathering pollen as foragers or acquiring pollen incidentally on their bodies. Some flower visitors were caught and preserved for later identification by an insect taxonomist.
Collection and Analysis of Floral Volatiles Floral scent of S. japonica was collected at KBG during the time of highest floral scent emission (12:00–14:00) using solid-phase microextraction (SPME). Three flowers of S. japonica were cut from one individual and the flower stalk was wrapped with wet cotton. The material was placed in a test tube (15 cm long, 3 cm diam) and covered with a sealing film. The SPME holder with a 65 μm polydimethylsiloxane/divinylbenzene fiber (Supelco, Bellefonte, PA, USA; previously desorbed for 5 min in a GC injection port heated to 200 °C) was inserted directly into the atmosphere surrounding the three flowers; the fiber was exposed for 45 min and immediately analysed via GC/MS. Five replicates of a total of 15 fresh flowers were sampled, and an empty conical flask (500 ml) with wet cotton wrapped around three leaves without flowers acted as a control. Samples from flowers and controls were analyzed using an Agilent Technologies HP 6890 gas chromatograph, equipped with an HP-5MS column (30 m×0.25 mm inner diam, 0.25 μm film thickness), and linked to an HP 5973 mass spectrometer. Helium was used as a carrier gas, at a flow rate of 1 ml/min. Split inlet and FID were held at 250 °C. Column temperature was programmed to rise from 40 °C (5-min hold) to 250 °C (20-min hold) at 3 °C/min. Compounds were identified by comparing mass spectra and relative retention times with those of standard compounds and with the Wiley NIST 05 mass spectral database, and their relative proportions (%) were determined by peak area measurements. To scrutinize the identification of 1-pyrroline carefully, a natural scent sample of S. japonica using SPME was collected. The fiber then was subsequently exposed to synthetic 1-pyrroline, and the volatiles were analyzed by GC/MS. Before using the SPME method, to evaluate the emission rate of floral scents from S. japonica, we used the dynamic headspace extraction method (DHEM) to collect floral scents. However, this method collects only two floral compounds (2methyl-1-butanol and 3-methyl-1-butanol) from three flowers. Because floral scents collected by DHEM must be eluted by solvent, we inferred that the scent of 1-pyrroline might be masked by the dichloromethane solvent. Actually, we can smell the special odor of 1-pyrroline in the eluent. However, we can not detect it by GC/MS. The SPME method can detect 1-pyrroline easily. Therefore, we used the DHEM to collect floral scents and calculated the amount of 2-methyl-1-butanol emission from flowers. We counted the percent of 2-methyl-1butanol in floral composition collected by SPME method. The total floral emission = (the amount of 2-methyl-1-butanol in DHEM) / (the percent of 2-methyl-1-butanol in SPME). Headspace samples were taken by enclosing three flowers of S. japonica in Tedlar bags (Dupont, USA) prior to sampling, and subsequently pumping air for 3 hr through filters containing 150 mg of Porapak Q (mesh 60/80, Waters Associates, Inc.) using a pump with an inlet flow rate of 300 ml/min. Before use, the adsorbent cartridges were cleaned with 2 ml
J Chem Ecol
of diethyl ether and dried with nitrogen gas. Trapped volatiles (N=3) were eluted with 400 μl dichloromethane, and concentrated to one-sixth of the original volume by a gentle stream of nitrogen. An empty bag with three leaves was used as a control. In each sample, 700 ng n-nonane were added first as an internal standard for quantification. All samples were stored at −20 °C before analysis. Samples were analyzed using the same method of SPME. Synthesis of 1-Pyrroline Standards for Comparison with Floral Scent 1-Pyrroline was synthesized according to the methods described by Mores et al. (2008). In a roundbottom flask covered with aluminum foil, AgNO3 (0.74 mmol, 128 mg) was added to a solution of pyrrolidine (SigmaAldrich, USA, purity >98 %; 0.15 mol, 10.66 g, 12.5 ml) in H2O (140 ml) plus NaOH pellets (0.3 mol, 12 g). To this mixture, a solution of 25 % Na2S2O8 (0.15 mol, 35.7 g, 142 ml) was added dropwise at 0 °C. The reaction mixture was stirred at room temperature for 3 hr and then extracted with dichloromethane (2×200 ml), after having added NaCl to the aqueous phase to saturation. After drying over Na2SO4 and removal of the solvent, the mixture was dissolved in ethyl ether (200 ml) for filtration through neutral alumina and evaporation to dryness. The end product, pale orange oil, then was used for NMR analysis. NMR Analysis of Synthesized 1-Pyrroline The yield of 1pyrroline was 54.0 %, which coexists with its corresponding trimer both in solution phase and in pure liquid state (Baker et al. 1992; Xiao et al. 2013). In this study, the rate of 1pyrroline trimer is about 20 % in CDCl3 and 26 % in D2O, which were determined from the integration of the Ha and He. The spectra data of 1-pyrroline are: Monomer: 1H NMR (400 MHz, CDCl3) δ 7.59 (s, 1H), 3.83 (t, J=7.3 Hz, 2H), 2.52 (m, 2H), 1.78 (m, 2H); Trimer: 1H NMR (400 MHz, CDCl3) δ 3.12-2.92 (m, 3H), 2.31 (dd, J=14.9, 8.5 Hz, 3H), 1.92-1.64 (m, 15H). Monomer: 1H NMR (400 MHz, D2O) δ 7.52 (s, 1H), 3.71-3.50 (m, 2H), 2.43 (tt, J=3.3, 2.2 Hz, 2H), 1.73-1.61 (m, 2H). 1H NMR (400 MHz, D2O) δ 3.04 (t, J= 6.4 Hz, 3H), 2.77 (td, J=8.8, 5.0 Hz, 3H), 2.28 (td, J=9.1, 6.4 Hz, 3H), 1.74-1.60 (m, 12H). Bioassay of the Ability of S. japonica Volatiles to Attract Flies The ability of volatiles emitted by S. japonica to attract flies was tested in bioassays. A solution containing five volatiles of S. japonica (in proportions similar to those of natural S. japonica extracts, as verified by GC-FID analyses) was prepared. This synthetic scent mixture consisted of 59.2 mg 1pyrroline (monomer and trimer mixture), 27.2 mg 2-methyl-1butanol, 8.8 mg 3-methyl-1-butanol, 3.6 mg 2-methylbutanal, and 1.2 mg 3-methylbutanal (four C5-branched compounds purchased from Sigma-Aldrich, 95–99 % purity) in 10 ml dichloromethane. A 100 μl sample of either the scent mixture (equivalent
to about 300 flowers) or dichloromethane control was dropped on a sample and control filter papers, respectively (Fig. 2a). All filter papers were 5 cm diam, and were soaked with dichloromethane for 24 hr and dried at room temperature to remove potential odor compounds before use. After the solvent had evaporated, scented and control filter papers were placed in the natural habitat of the S. japonica population (Fig. 2a), (13:00–14:00, from 6 to 15 July, 2014). If flies landed on the filter papers (Fig. 2b), we assumed that they were attracted by the sample or control. Every 5 min, the positions of the sample and control were changed to avoid location effects. All flies that landed on the filter paper were collected. Each test lasted 10 min and the bioassays were performed 27 times. Each paper was used only once for a bioassay. The bioassay was conducted at the TMS population. Before conducting our bioassay, we had thought that the semen-like odor of 1-pyrroline might play a role in the attraction of fly pollinators, because this compound is a major component (about 60 %) of the floral scent composition in S. japonica. The function of the compound was, therefore, tested in 2012. A 59.2 mg sample of 1-pyrroline (monomer and trimer mixture) was dissolved in 10 ml distilled water. In a trap bottle (Enjoy Technology, China), 100 μl samples were dissolved in 200 ml distilled water and used to attract Tetradacus flies, which are similar in size to the pollinator of S. japonica. Eleven scented and 11 control traps were placed in the TMS location, spaced at least 3 m apart. Flies captured in these traps were recorded daily, and the process continued for 5 d until the strong fetid odor from the scented traps dissipated. Testing the Similarity Between 1-Pyrroline and the Floral Scents of S. japonica To test whether 1-pyrroline represents the semen-like floral scent profile of S. japonica, a bioassay was conducted to test whether common people can distinguish 1-pyrroline from the floral scents of S. japonica. A 2 mg 1pyrroline sample was dissolved in 1 ml distilled water, and 10 μl of the resulting liquid were dropped onto a black cloth. A cloth of similar size was wrapped around 10 flowers of S. japonica. The two cloths were put into two 250 ml flasks, after which 32 volunteers were asked to smell the contents of the two flasks to determine odor similarity. Statistical Analyses Floral scent composition was assessed using one-sample t-tests, and the bioassay was assessed using an independent-sample t-test. SPSS 18.0 software (SPSS Inc., Armonk, NY, USA) was used for all the analyses.
Results Flower Biology We found that the flower longevity of S. japonica was about 10–12 hr. Flowers typically opened at about 8:30–9:30 in the morning and closed at about 18:30–
J Chem Ecol
Fig. 2 Bioassay to test the ability of volatiles emitted by Stemona japonica to attract flies. (a) illustration of the fly attraction bioassay using filter papers with or without scents; sample (left) and control (right). (b) attracted flies on a scented filter paper
around the bottom of the flower, the tepal, and the stamen. Flies acquired pollen accidentally on their bodies as they moved within the flowers. No grooming/gathering pollen behavior was observed in Atherigona. When the flies left the flowers, they carried pollen over their entire bodies. The flies deposited pollen grains on the stigma with their legs while visiting the flowers. The visit frequency per flower per hour was 0.09 in 2011 and 0.06 in 2012. The sex ratio of the 67 Atherigona specimens caught on flowers of S. japonica during 2011–2014, was 49 female vs. 18 male. Thus, the sex ratio of Atherigona flies was significantly different from 1:1 (χ2 =14.343, df=1, P
