Journal ofChemicalEcology, Vol. 17, No. 6, 1991
CHEMICAL CHARACTERIZATION OF FRUIT AND FUNGAL VOLATILES ATTRACTIVE TO DRIED-FRUIT BEETLE, Carpophilus hemipterus (L.) (COLEOPTERA: NITIDULIDAE)
P. LARRY
PHELAN*
and HENGCHEN
LIN
Department of Entomology Ohio Agricultural Research & Development Center The Ohio State University Wooster, Ohio 44691 (Received October 1, 1990; accepted February 19, 1991)
Abstract--The chemical basis underlying orientation to fruit and fungal odors was investigated for the dried-fruit beetle, Carpophilus hemipterus (L.). In wind-tunnel bioassays of walking and flight response from 1.8 m, beetles were attracted to odors of the yeast Saccharomyces cerevisiae on agar, aseptic banana, or banana inoculated with S. cerevisiae, although both banana substrates elicited greater response than the yeast alone. When presented in a two-choice bioassay, the yeast-inoculated banana attracted approximately twice as many beetles as did the aseptic banana. GC-MS analysis of the headspace volatiles above these odor sources revealed a somewhat more complex and concentrated volatile profile for yeast-inoculated banana than for aseptic banana. The odor from yeast on agar had fewer components, and these were present at lower concentrations than the odors of either banana substrate. By blending mineral-oil or aqueous solutions of the 18 components of inoculated-banana odor in varying concentrations, it was possible to mimic closely the headspace profile of the natural odor. This synthetic odor also elicited beetle attraction in the wind tunnel at levels comparable to the inoculated banana. Through a series of bioassays in which individual components were subtracted from or added to a synthetic odor blend, it was determined that ethyl acetate, acetaldehyde, 2-pentanol, and 3-methylbutanol comprised the simplest blend of compounds evoking full behavioral response. However, 2-methylpropanol or butanol were apparently interchangeable with 3-methylbutanol in this blend, and comparable response could also be elicited by replacing acetaldehyde with a combination of both 2-pentanone and 3-hydroxy-2-butanone. Thus, our results suggest that this generalist insect *To whom correspondence should be addressed. 1253 0098-0331/91/0600 1253506.50/0 9 1991 Plenum Publishing Corporation
1254
PHELAN AND LIN herbivore locates its hosts by a long-range response to a variety of blends of common fruit volatiles, whose concentrations are enhanced by fungi. Key Words--Coleoptera, Nitidulidae, Carpophilus hemipterus, dried-fruit beetle, host-finding, yeast, wind tunnel, attraction, volatiles, headspace.
INTRODUCTION
The dried-fruit beetle, Carpophilus hemipterus (L.), is a cosmopolitan species that causes damage to a large number of agricultural products by direct infestation or by transmission of phytopathogenic fungi (Hinton, 1945). Fruit baits are used commonly to monitor this and other nitidulid beetles in the field, and response to fruits can be enhanced in C. hemipterus if the fruits are inoculated with fungi (Wildman, 1933; Blackmer and Phelan, 1991); however, monitoring populations using fruit baits is made difficult by the confounding variability in volatile release from living tissue, especially during the chemically dynamic period of ripening. Smilanick et al. (1978) demonstrated that a synthetic mixture of ethanol, ethyl acetate, and acetaldehyde could be used to attract C. hemipterus in the field, and propyl propionate or butyl acetate can be used to selectively trap another nitidulid, Glischrochilus quadrisignatus (Alm et al., 1986); however, in both cases the release rates used for the lures were much higher than that expected from a plant, and thus it is difficult to determine the role that these compounds may play in host-finding for these species. To date, no attempt has been made to investigate systematically the components of a natural odor source to determine which constituents may be mediating longrange host-finding in any nitidulid species. In this study, we: (1) measure the relative attractiveness of aseptic banana, the yeast Saccharomyces cerevisiae Hanson, and banana inoculated with yeast in a wind-tunnel bioassay; (2) chemically characterize the odor profiles in the headspace above these substrates; and (3) determine the relative activity of odor components for C. hemipterus host finding by their systematic elimination from and addition to synthetic odor blends. METHODS AND MATERIALS
Insects. C. hemipterus was maintained in the laboratory on artificial diet using the methods of Hall et al. (1978) under the conditions of 24 _ 1 ~ 16 : 8 light-dark, and 75-85 % relative humidity. After emergence from the pupae, adults were kept as mixed sexes in 12-oz. plastic cups with moisture, without contact with food or food odors. Natural Odors. Banana (var. Cavendish) was used to investigate host-finding in C. hemipterus because it was found to elicit high levels of upwind flight
ATTRACTIVE FRUIT AND FUNGAL VOLATILES
1255
in wind-tunnel studies (Blackmer and Phelan, in preparation) and because fresh samples were available on a year-round basis in local grocery stores. The roie of fungi in host finding was measured using Saccharomyces cerevisiae vat. Montrachet, which was maintained in the laboratory on YM agar (Difeo Laboratories, Detroit, Michigan) at 24 + I~ Four natural-odor combinations were prepared for chemical analysis and behavioral bioassays: YM agar, S. cerevisiae on yeast medium, aseptic banana, and banana inoculated with S. cerevisiae. Samples of 30 g freshly peeled banana or 30 ml yeast medium were prepared in 240 ml Qorpak straight-sided jars that had been sterilized with 1% NaOC1 and rinsed with sterile doubly deionized water. The lid of each jar was equipped with a 0.20 /zm Bacterial Air Vent (Gelman Sciences, Ann Arbor, Michigan), to allow free air exchange while precluding microbial contaminants, and with a sleeve-type rubber septum (Thomas Scientific) for collection of headspace samples. Banana samples were placed in the jars, sterilized by treatment with 70% ethanol for 5 min, and then rinsed five times with 100-150 ml doubly deionized water passed through the Bacterial Air Vent. In rare instances where sterilization was unsuccessful ( < 1%), contamination became visible within three to four days and these samples were discarded; otherwise, substrates remained aseptic for greater than a month. To determine the degree to which the ethanol used in sterilizing the banana may have contributed to the occurrence of ethanol in the headspace of aseptic substrates, we compared the headspace profiles of bananas sterilized using the above procedures with those of bananas sterilized with 1% NaOC1. For inoculated samples, S. cerevisiae was transferred in a sterile transfer hood to sterilized banana samples by scratching the surface of the fungal culture with a flame-sterilized wire loop and then smearing the surface of the banana. All samples were incubated in the jars for seven days prior to testing in the wind tunnel and headspace analysis. Headspaee Volatile Identifications. Headspace volatiles were trapped using a device integral to a Hewlett-Packard 5890A gas chromatograph (GC) that allowed thermal desorption of constituents directly to the GC column (Phelan, unpublished). Briefly, the device consisted of a six-port two-position capillary valve (Valco Instruments, Houston, Texas) and a 60-cm x 0.1-cm-ID capillary volatile trap filled with Tenax GC (60/80 mesh). When placed in the trapping position, the valve allowed volatiles to be carried via a helium flow to the capiltary trap. After a designated collection period, the trap was ballistically heated to 200~ and the valve position was switched to reverse the helium flow in the trap such that retained volatiles were desorbed and back-flushed to the GC analytical column. For this study, jars containing odor sources were opened for 5 rain, resealed, and a 1-ml headspace sample immediately collected through the rubber septum using a gas-tight syringe. The 1-ml sample was then slowly injected onto the volatile trap and desorbed as above. Because the headspace profile was greatly dominated by ethanol, a second method was used to reduce
1256
PHELAN AND LIN
the amount of ethanol desorbed to the analytical column while enriching minor components of the headspace samples. A 20-ml headspace sample was collected and injected onto the trap as described; however, at this point the trap was flushed with helium for 6 min at a flow of 20 ml/min. This flush period allowed the preponderance of ethanol and acetaldehyde to be removed from the trap, which we have determined to have breakthrough volumes of approximately 90 ml and 45 ml, respectively; the next smallest breakthrough volume for the components encountered in this study was 450 ml for propanol, so that all other components were concentrated on the trap without breakthrough for desorption to the analytical column. Headspace components were chromatographed on either DB-1 (30 m x 0.32 mm ID, 5.0-/zm film thickness) or DB-FFAP (30 m x 0.25 mm ID, 0.25tzm film thickness) capillary columns (J & W Scientific, Folsom, California), with a temperature program of 30-200~ increased at 10~ Helium flow was 0.9 ml/min on the DB-1 column and 0.7 ml/min on the DB-FFAP column. Tentative identifications of the components were made using a HP 5970C Mass Selective Detector with a direct interface to the GC. Chemical assignments were confirmed by comparison of GC retention times on the two columns and mass spectra with those of authentic samples; no attempts were made to determine enantiomeric identity of chiral compounds. The concentration of each component in the headspace was quantified by comparison of its peak areas with that of synthetics. Synthetic Odors. Synthetic odors were created by blending compounds identified from S. cerevisiae -inoculated banana. All chemicals were obtained from Aldrich Chemical Co. (Milwaukee, Wisconsin) and were prepared individually as mineral-oil solutions, except ethanol, which was diluted with double-deionized water due to low solubility in mineral oil. Mineral-oil solutions were applied to cotton wicks in 240-ml Qorpak straight-sided jars. Solution concentrations and volumes were adjusted to mimic the natural odor by comparing the concentration of each component in the headspace above the synthetic blend with that of the natural source. A total combination of 18 synthetic compounds was used in the first three bioassays, but this blend was reduced to 13 components for subsequent bioassays as the elimination of five minor compounds (four esters and 2-butanone) did not significantly affect the attractiveness of the synthetic blend. Synthetic odor blends were used for bioassay immediately after preparation, and new mixtures were prepared for each replicate. Wind-Tunnel Bioassays. Odor-modulated responses of C. hemipterus were evaluated in a previously described wind tunnel that measured 2.5 m long x 1 m wide x 0.5 m high (Phelan et al., 1991). Bioassays were conducted 4-0 hr prior to lights out at 35-65% relative humidity, 24-27~ with a wind speed of 0.5 m/sec. The tunnel was illuminated from above by three 60-W bulbs
ATTRACTIVE FRUIT AND F U N G A L VOLATILES
t257
operated at 80 V to provide 17 lux on the tunnel floor. During wind-tunnel bioassays, two jars were placed on the floor 10 cm away from the center line of the upwind end of the tunnel. Unless otherwise indicated, one of the jars was empty and served as a negative control for possible odor contamination. The position of the jars was randomized for each treatment, and the test beetles were released on the floor at the center of the downwind end of the tunnel at a distance of 180 cm from the odor sources. Smoke tests showed that plumes from the two jars merged before reaching the beetle-release site. Ten 6- to 7-day-old beetles were released per replicate, and beetles were used only once. Beetles that oriented to test odors within 10 min and that walked or flew up the odor plume to within 10 cm downwind of the jar were counted as responders. The 10-cm criterion was used since the top of the jar, and thus the source of the odor, was 8.7 cm above the floor, and those beetles that walked to within 10 cm of the jar would lose the odor plume. A final bioassay measuring flight response alone was conducted to confirm that the synthetic blends eliciting the greatest walking response also evoked a flight response equivalent to that to inoculated banana. For this bioassay, the sample jars were set on 15-cm-high tripods and beetles were released from a platform on a similar tripod. The locations of jars and distances between odor and beetle-releasing site were the same as in previous bioassays. A positive response was recorded for those individuals that flew up the odor plume and landed on the jar. Since beetles were never observed to respond to an empty jar, these bioassays effectively were operated as single-choice tests, with a randomized complete-block presentation of treatments. Beetle response was measured as the percent of each group of 10 beetles released. Response data were analyzed by two-way ANOVA after transformation by sin-~ x/X, with mean separations performed using LSD on data sets where ANOVA indicated a significant F value (SAS Institute, 1985). In addition to the single-choice bioassays, beetle response to S. cerevisiae-inoculated banana was compared with that to aseptic banana as a two-choice bioassay in the wind tunnel.
RESULTS
Attraction to Natural Odors. All four natural odor sources elicited significant levels of upwind orientation by C. hemipterus (Figure 1). The S. cerevisiae-inoculated and the aseptic bananas elicited comparable beetle responses, both of which were significantly greater than that to the yeast on agar medium or the medium alone; however, when given a choice between the inoculatedand aseptic-banana odors, beetles showed a strong preference for the former, with more than twice as many orienting to the inoculated substrate (Figure 1). Headspace Volatile Identifications. A total of 18 compounds (eight esters,
1258
PHELANANDLIN
100" r
tO
80"
n"
60'
O. co
40"
a. 20-
0 SM
AM
SB
AB
AB v, SB
Odor Sources
FIG. 1. Orientation of Carpophilus hemipterus to four natural odor sources in a wind tunnel: Saccharomyces cerevisiae on agar medium (SM), aseptic agar medium (AM), S. cerevisiae-inoculated banana (SB), or aseptic banana (AB). Response bars (_SEM) marked by the same letter are not significantly different (LSD, N = 8 groups of 10 beetles, P < 0.05). Orientation to SB versus AB compared in a two-choice experimental design (X2 = 17.62, 7 df, P < 0.02).
six alcohols, three ketones, and one aldehyde) were identified from the four substrates, as listed in Table 1 with the major mass ions used for their identification; ethanol was by far the dominant component in the odor of all four substrates. The possibility that ethanol in the headspace was a result of the use of ethanol for sterilizing substrates was ruled out by the fact that the headspace of 7-day-old aseptic banana sterilized by 1% NaOC1 contained ethanol at levels identical to that sterilized with 70% ethanol [peak areas (• 107) -t- SEM = 2.41 ___ 0.32 and 2.48 + 0.21, respectively; t = 0.19; P = 0.86; N = 5]. Table 2 shows the relative abundance of these constituents in the volatile profiles of the four substrates, with the odor of yeast-inoculated banana possessing all 18 components (Figure 2a). The aseptic-banana odor was qualitatively similar, with 15 of these compounds present (Table 2); however, the two substrates were quantitatively distinct, with the inoculated banana producing roughly twice the concentration of volatiles overall and with the largest increase occurring in the production of alcohols. The concentration of nonethanol alcohols was about four times greater in the headspace of inoculated banana, and the concentration of propanol increased by more than 40-fold (Table 2). The headspace profile of S. cerevisiae on yeast medium was both qualitatively and quantitatively depauperate relative to that of the yeast on banana, with only seven compounds detected in headspace samples of yeast on agar medium. Attraction to Synthetic Odors. When the 18 components of yeast-inocu-
1259
ATTRACTIVE FRUIT AND FUNGAL VOLATILES
TABLE l. CHEMICALS IDENTIFIED FROM HEADSPACE OF
Saccharomyces cerevisiae-
INOCULATED BANANA
Peak No.a 1 2 3 4 5 6 7 8
9 10 11
t2 13 14 15 16 17 18
Component Abbr. AcAl EtOH PrOH EtAc MPrOH BuOH PeKe 2PeOH HBuKe MBuOH iBuAc EtBu
iPeAc
Identity Acetaldehyde Ethanol Propanol 2-Butanone Ethyl acetate 2-Methylpropanol Butanol 2-Pentanone 2-Pentanol 3-Hydroxy-2-butanone 3-Methylbutanol Isobutyl acetate Ethyl butyrate Butyl acetate Ethyl isovalerate Isopentyl acetate Isopentyl butyrate Isopentyl isovalerate
Mass ions (m/z/' 2_99,42, 43, 44 27, 29, 31,45, 46 27, 29, 31 , 42, 59, 60 27, 29, 43,57, 72 28, 29, 43,45, 61, 70, 73, 88 27, 31, 33, 41, 42, 43,55, 74 27, 29, 31 , 41, 42, 43, 56 27, 29, 43,57, 71, 86 29, 31, 39, 41, 43, 45,55, 73 27, 29, 43, 45 , 88 29, 31, 41, 42, 43, 55,57, 70 27, 29, 39, 41, 43,56, 62, 73 29, 41, 4_33,45, 60, 71, 73, 88 27, 29, 41, 43,56, 61, 73 29,41, 43, 57, 60, 61, 85, 88 39, 43 , 55, 61, 70, 87 31, 40, 41, 43, 55, 70, 71,89 41, 43, 55, 57, 70, 85, 103
~See Figure 2 for total ion chromatogmm. ~'Major ions of mass spectra used for identification,with base ions underlined.
lated banana odor were blended in the amounts and concentrations gtven in Table 2, a synthetic-odor profile was generated that closely mimicked the natural source both in overall quantity and in the ratio of individual components (Figure 2b). This synthetic banana odor also closely mimicked the natural substrate in behavioral activity, eliciting upwind attraction in C. hemipterus at levels comparable to the inoculated banana [87.8 + 2.5% (X +_ SEM) for the synthetic odor and 67.9 _+ 6.3% for the natural odor, N = 4 groups of 10 beetles, P > 0.05). In addition, both the synthetic blend and the inoculated banana elicited beetle response at a level several times greater than that to a blend of ethanol, acetaldehyde, and ethyl acetate (9.5 + 4 . 8 % , P < 0.05), a combination previously reported to be an effective attractant for C. hemipterus in the field (Smilanick et al., 1978). In the first of a series of bioassays of synthetic blends, the 18-component blend again was very similar to the inoculated banana in eliciting upwind orientation; however, the elimination of all members of any of the three groups of functionalities resulted in a reduction in beetle response (Figure 3). The per-
1260
PHELAN AND LIN
TABLE 2. PERCENT COMPOSITION OF VOLATILE CONSTITUENTS IN HEADSPACE OF
Saccharomyces cerevisiae-INOCULATED AGAR MEDIUM ( S M ) , ASEPTIC AGAR MEDIUM ( A M ) , S. cerevisiae-INOCULATED BANANA (SB), AND ASEPTIC BANANA (AB); RATIO OF ABUNDANCE OF EACH COMPONENT IN SB VERSUS A B HEADSPACE; AND COMPOSITION OF SYNTHETIC ODOR BLENDS
Percent composition (non-EtOH)" Component Alcohols Ethanol Propanol 2-Methylpropanol Butanol 2-Pentanol 3-Methylbutanol Aldehydes/ketones Acetaldehyde 2-Butanone~ 2-Pentanone 3-Hydroxy-2-butanone Esters Ethyl acetate Isobutyl acetate Ethyl butyrate Butyl acetate e Ethyl isovalerate e Isopentyl acetate Isopentyl butymte" Isopentyl isovalerate e Totals (non-EtOH) Alcohols Esters Aldehydes/ketones
SM
AM
SB
AB
SB : AB (absolute quantity) b
NQ J
NQ
25.73 1.13
87.14
(85.05) 10.30 8.49 0.72 0.44 10.63
(86.53) 0.49 7.61 0,84 0.94 6.27
55.36 0.09 0.65 0.27
54.13
11.78 0.25 0.45
CHEMICAL CHARACTERIZATION OF FRUIT AND FUNGAL VOLATILES ATTRACTIVE TO DRIED-FRUIT BEETLE, Carpophilus hemipterus (L.) (COLEOPTERA: NITIDULIDAE)
P. LARRY
PHELAN*
and HENGCHEN
LIN
Department of Entomology Ohio Agricultural Research & Development Center The Ohio State University Wooster, Ohio 44691 (Received October 1, 1990; accepted February 19, 1991)
Abstract--The chemical basis underlying orientation to fruit and fungal odors was investigated for the dried-fruit beetle, Carpophilus hemipterus (L.). In wind-tunnel bioassays of walking and flight response from 1.8 m, beetles were attracted to odors of the yeast Saccharomyces cerevisiae on agar, aseptic banana, or banana inoculated with S. cerevisiae, although both banana substrates elicited greater response than the yeast alone. When presented in a two-choice bioassay, the yeast-inoculated banana attracted approximately twice as many beetles as did the aseptic banana. GC-MS analysis of the headspace volatiles above these odor sources revealed a somewhat more complex and concentrated volatile profile for yeast-inoculated banana than for aseptic banana. The odor from yeast on agar had fewer components, and these were present at lower concentrations than the odors of either banana substrate. By blending mineral-oil or aqueous solutions of the 18 components of inoculated-banana odor in varying concentrations, it was possible to mimic closely the headspace profile of the natural odor. This synthetic odor also elicited beetle attraction in the wind tunnel at levels comparable to the inoculated banana. Through a series of bioassays in which individual components were subtracted from or added to a synthetic odor blend, it was determined that ethyl acetate, acetaldehyde, 2-pentanol, and 3-methylbutanol comprised the simplest blend of compounds evoking full behavioral response. However, 2-methylpropanol or butanol were apparently interchangeable with 3-methylbutanol in this blend, and comparable response could also be elicited by replacing acetaldehyde with a combination of both 2-pentanone and 3-hydroxy-2-butanone. Thus, our results suggest that this generalist insect *To whom correspondence should be addressed. 1253 0098-0331/91/0600 1253506.50/0 9 1991 Plenum Publishing Corporation
1254
PHELAN AND LIN herbivore locates its hosts by a long-range response to a variety of blends of common fruit volatiles, whose concentrations are enhanced by fungi. Key Words--Coleoptera, Nitidulidae, Carpophilus hemipterus, dried-fruit beetle, host-finding, yeast, wind tunnel, attraction, volatiles, headspace.
INTRODUCTION
The dried-fruit beetle, Carpophilus hemipterus (L.), is a cosmopolitan species that causes damage to a large number of agricultural products by direct infestation or by transmission of phytopathogenic fungi (Hinton, 1945). Fruit baits are used commonly to monitor this and other nitidulid beetles in the field, and response to fruits can be enhanced in C. hemipterus if the fruits are inoculated with fungi (Wildman, 1933; Blackmer and Phelan, 1991); however, monitoring populations using fruit baits is made difficult by the confounding variability in volatile release from living tissue, especially during the chemically dynamic period of ripening. Smilanick et al. (1978) demonstrated that a synthetic mixture of ethanol, ethyl acetate, and acetaldehyde could be used to attract C. hemipterus in the field, and propyl propionate or butyl acetate can be used to selectively trap another nitidulid, Glischrochilus quadrisignatus (Alm et al., 1986); however, in both cases the release rates used for the lures were much higher than that expected from a plant, and thus it is difficult to determine the role that these compounds may play in host-finding for these species. To date, no attempt has been made to investigate systematically the components of a natural odor source to determine which constituents may be mediating longrange host-finding in any nitidulid species. In this study, we: (1) measure the relative attractiveness of aseptic banana, the yeast Saccharomyces cerevisiae Hanson, and banana inoculated with yeast in a wind-tunnel bioassay; (2) chemically characterize the odor profiles in the headspace above these substrates; and (3) determine the relative activity of odor components for C. hemipterus host finding by their systematic elimination from and addition to synthetic odor blends. METHODS AND MATERIALS
Insects. C. hemipterus was maintained in the laboratory on artificial diet using the methods of Hall et al. (1978) under the conditions of 24 _ 1 ~ 16 : 8 light-dark, and 75-85 % relative humidity. After emergence from the pupae, adults were kept as mixed sexes in 12-oz. plastic cups with moisture, without contact with food or food odors. Natural Odors. Banana (var. Cavendish) was used to investigate host-finding in C. hemipterus because it was found to elicit high levels of upwind flight
ATTRACTIVE FRUIT AND FUNGAL VOLATILES
1255
in wind-tunnel studies (Blackmer and Phelan, in preparation) and because fresh samples were available on a year-round basis in local grocery stores. The roie of fungi in host finding was measured using Saccharomyces cerevisiae vat. Montrachet, which was maintained in the laboratory on YM agar (Difeo Laboratories, Detroit, Michigan) at 24 + I~ Four natural-odor combinations were prepared for chemical analysis and behavioral bioassays: YM agar, S. cerevisiae on yeast medium, aseptic banana, and banana inoculated with S. cerevisiae. Samples of 30 g freshly peeled banana or 30 ml yeast medium were prepared in 240 ml Qorpak straight-sided jars that had been sterilized with 1% NaOC1 and rinsed with sterile doubly deionized water. The lid of each jar was equipped with a 0.20 /zm Bacterial Air Vent (Gelman Sciences, Ann Arbor, Michigan), to allow free air exchange while precluding microbial contaminants, and with a sleeve-type rubber septum (Thomas Scientific) for collection of headspace samples. Banana samples were placed in the jars, sterilized by treatment with 70% ethanol for 5 min, and then rinsed five times with 100-150 ml doubly deionized water passed through the Bacterial Air Vent. In rare instances where sterilization was unsuccessful ( < 1%), contamination became visible within three to four days and these samples were discarded; otherwise, substrates remained aseptic for greater than a month. To determine the degree to which the ethanol used in sterilizing the banana may have contributed to the occurrence of ethanol in the headspace of aseptic substrates, we compared the headspace profiles of bananas sterilized using the above procedures with those of bananas sterilized with 1% NaOC1. For inoculated samples, S. cerevisiae was transferred in a sterile transfer hood to sterilized banana samples by scratching the surface of the fungal culture with a flame-sterilized wire loop and then smearing the surface of the banana. All samples were incubated in the jars for seven days prior to testing in the wind tunnel and headspace analysis. Headspaee Volatile Identifications. Headspace volatiles were trapped using a device integral to a Hewlett-Packard 5890A gas chromatograph (GC) that allowed thermal desorption of constituents directly to the GC column (Phelan, unpublished). Briefly, the device consisted of a six-port two-position capillary valve (Valco Instruments, Houston, Texas) and a 60-cm x 0.1-cm-ID capillary volatile trap filled with Tenax GC (60/80 mesh). When placed in the trapping position, the valve allowed volatiles to be carried via a helium flow to the capiltary trap. After a designated collection period, the trap was ballistically heated to 200~ and the valve position was switched to reverse the helium flow in the trap such that retained volatiles were desorbed and back-flushed to the GC analytical column. For this study, jars containing odor sources were opened for 5 rain, resealed, and a 1-ml headspace sample immediately collected through the rubber septum using a gas-tight syringe. The 1-ml sample was then slowly injected onto the volatile trap and desorbed as above. Because the headspace profile was greatly dominated by ethanol, a second method was used to reduce
1256
PHELAN AND LIN
the amount of ethanol desorbed to the analytical column while enriching minor components of the headspace samples. A 20-ml headspace sample was collected and injected onto the trap as described; however, at this point the trap was flushed with helium for 6 min at a flow of 20 ml/min. This flush period allowed the preponderance of ethanol and acetaldehyde to be removed from the trap, which we have determined to have breakthrough volumes of approximately 90 ml and 45 ml, respectively; the next smallest breakthrough volume for the components encountered in this study was 450 ml for propanol, so that all other components were concentrated on the trap without breakthrough for desorption to the analytical column. Headspace components were chromatographed on either DB-1 (30 m x 0.32 mm ID, 5.0-/zm film thickness) or DB-FFAP (30 m x 0.25 mm ID, 0.25tzm film thickness) capillary columns (J & W Scientific, Folsom, California), with a temperature program of 30-200~ increased at 10~ Helium flow was 0.9 ml/min on the DB-1 column and 0.7 ml/min on the DB-FFAP column. Tentative identifications of the components were made using a HP 5970C Mass Selective Detector with a direct interface to the GC. Chemical assignments were confirmed by comparison of GC retention times on the two columns and mass spectra with those of authentic samples; no attempts were made to determine enantiomeric identity of chiral compounds. The concentration of each component in the headspace was quantified by comparison of its peak areas with that of synthetics. Synthetic Odors. Synthetic odors were created by blending compounds identified from S. cerevisiae -inoculated banana. All chemicals were obtained from Aldrich Chemical Co. (Milwaukee, Wisconsin) and were prepared individually as mineral-oil solutions, except ethanol, which was diluted with double-deionized water due to low solubility in mineral oil. Mineral-oil solutions were applied to cotton wicks in 240-ml Qorpak straight-sided jars. Solution concentrations and volumes were adjusted to mimic the natural odor by comparing the concentration of each component in the headspace above the synthetic blend with that of the natural source. A total combination of 18 synthetic compounds was used in the first three bioassays, but this blend was reduced to 13 components for subsequent bioassays as the elimination of five minor compounds (four esters and 2-butanone) did not significantly affect the attractiveness of the synthetic blend. Synthetic odor blends were used for bioassay immediately after preparation, and new mixtures were prepared for each replicate. Wind-Tunnel Bioassays. Odor-modulated responses of C. hemipterus were evaluated in a previously described wind tunnel that measured 2.5 m long x 1 m wide x 0.5 m high (Phelan et al., 1991). Bioassays were conducted 4-0 hr prior to lights out at 35-65% relative humidity, 24-27~ with a wind speed of 0.5 m/sec. The tunnel was illuminated from above by three 60-W bulbs
ATTRACTIVE FRUIT AND F U N G A L VOLATILES
t257
operated at 80 V to provide 17 lux on the tunnel floor. During wind-tunnel bioassays, two jars were placed on the floor 10 cm away from the center line of the upwind end of the tunnel. Unless otherwise indicated, one of the jars was empty and served as a negative control for possible odor contamination. The position of the jars was randomized for each treatment, and the test beetles were released on the floor at the center of the downwind end of the tunnel at a distance of 180 cm from the odor sources. Smoke tests showed that plumes from the two jars merged before reaching the beetle-release site. Ten 6- to 7-day-old beetles were released per replicate, and beetles were used only once. Beetles that oriented to test odors within 10 min and that walked or flew up the odor plume to within 10 cm downwind of the jar were counted as responders. The 10-cm criterion was used since the top of the jar, and thus the source of the odor, was 8.7 cm above the floor, and those beetles that walked to within 10 cm of the jar would lose the odor plume. A final bioassay measuring flight response alone was conducted to confirm that the synthetic blends eliciting the greatest walking response also evoked a flight response equivalent to that to inoculated banana. For this bioassay, the sample jars were set on 15-cm-high tripods and beetles were released from a platform on a similar tripod. The locations of jars and distances between odor and beetle-releasing site were the same as in previous bioassays. A positive response was recorded for those individuals that flew up the odor plume and landed on the jar. Since beetles were never observed to respond to an empty jar, these bioassays effectively were operated as single-choice tests, with a randomized complete-block presentation of treatments. Beetle response was measured as the percent of each group of 10 beetles released. Response data were analyzed by two-way ANOVA after transformation by sin-~ x/X, with mean separations performed using LSD on data sets where ANOVA indicated a significant F value (SAS Institute, 1985). In addition to the single-choice bioassays, beetle response to S. cerevisiae-inoculated banana was compared with that to aseptic banana as a two-choice bioassay in the wind tunnel.
RESULTS
Attraction to Natural Odors. All four natural odor sources elicited significant levels of upwind orientation by C. hemipterus (Figure 1). The S. cerevisiae-inoculated and the aseptic bananas elicited comparable beetle responses, both of which were significantly greater than that to the yeast on agar medium or the medium alone; however, when given a choice between the inoculatedand aseptic-banana odors, beetles showed a strong preference for the former, with more than twice as many orienting to the inoculated substrate (Figure 1). Headspace Volatile Identifications. A total of 18 compounds (eight esters,
1258
PHELANANDLIN
100" r
tO
80"
n"
60'
O. co
40"
a. 20-
0 SM
AM
SB
AB
AB v, SB
Odor Sources
FIG. 1. Orientation of Carpophilus hemipterus to four natural odor sources in a wind tunnel: Saccharomyces cerevisiae on agar medium (SM), aseptic agar medium (AM), S. cerevisiae-inoculated banana (SB), or aseptic banana (AB). Response bars (_SEM) marked by the same letter are not significantly different (LSD, N = 8 groups of 10 beetles, P < 0.05). Orientation to SB versus AB compared in a two-choice experimental design (X2 = 17.62, 7 df, P < 0.02).
six alcohols, three ketones, and one aldehyde) were identified from the four substrates, as listed in Table 1 with the major mass ions used for their identification; ethanol was by far the dominant component in the odor of all four substrates. The possibility that ethanol in the headspace was a result of the use of ethanol for sterilizing substrates was ruled out by the fact that the headspace of 7-day-old aseptic banana sterilized by 1% NaOC1 contained ethanol at levels identical to that sterilized with 70% ethanol [peak areas (• 107) -t- SEM = 2.41 ___ 0.32 and 2.48 + 0.21, respectively; t = 0.19; P = 0.86; N = 5]. Table 2 shows the relative abundance of these constituents in the volatile profiles of the four substrates, with the odor of yeast-inoculated banana possessing all 18 components (Figure 2a). The aseptic-banana odor was qualitatively similar, with 15 of these compounds present (Table 2); however, the two substrates were quantitatively distinct, with the inoculated banana producing roughly twice the concentration of volatiles overall and with the largest increase occurring in the production of alcohols. The concentration of nonethanol alcohols was about four times greater in the headspace of inoculated banana, and the concentration of propanol increased by more than 40-fold (Table 2). The headspace profile of S. cerevisiae on yeast medium was both qualitatively and quantitatively depauperate relative to that of the yeast on banana, with only seven compounds detected in headspace samples of yeast on agar medium. Attraction to Synthetic Odors. When the 18 components of yeast-inocu-
1259
ATTRACTIVE FRUIT AND FUNGAL VOLATILES
TABLE l. CHEMICALS IDENTIFIED FROM HEADSPACE OF
Saccharomyces cerevisiae-
INOCULATED BANANA
Peak No.a 1 2 3 4 5 6 7 8
9 10 11
t2 13 14 15 16 17 18
Component Abbr. AcAl EtOH PrOH EtAc MPrOH BuOH PeKe 2PeOH HBuKe MBuOH iBuAc EtBu
iPeAc
Identity Acetaldehyde Ethanol Propanol 2-Butanone Ethyl acetate 2-Methylpropanol Butanol 2-Pentanone 2-Pentanol 3-Hydroxy-2-butanone 3-Methylbutanol Isobutyl acetate Ethyl butyrate Butyl acetate Ethyl isovalerate Isopentyl acetate Isopentyl butyrate Isopentyl isovalerate
Mass ions (m/z/' 2_99,42, 43, 44 27, 29, 31,45, 46 27, 29, 31 , 42, 59, 60 27, 29, 43,57, 72 28, 29, 43,45, 61, 70, 73, 88 27, 31, 33, 41, 42, 43,55, 74 27, 29, 31 , 41, 42, 43, 56 27, 29, 43,57, 71, 86 29, 31, 39, 41, 43, 45,55, 73 27, 29, 43, 45 , 88 29, 31, 41, 42, 43, 55,57, 70 27, 29, 39, 41, 43,56, 62, 73 29, 41, 4_33,45, 60, 71, 73, 88 27, 29, 41, 43,56, 61, 73 29,41, 43, 57, 60, 61, 85, 88 39, 43 , 55, 61, 70, 87 31, 40, 41, 43, 55, 70, 71,89 41, 43, 55, 57, 70, 85, 103
~See Figure 2 for total ion chromatogmm. ~'Major ions of mass spectra used for identification,with base ions underlined.
lated banana odor were blended in the amounts and concentrations gtven in Table 2, a synthetic-odor profile was generated that closely mimicked the natural source both in overall quantity and in the ratio of individual components (Figure 2b). This synthetic banana odor also closely mimicked the natural substrate in behavioral activity, eliciting upwind attraction in C. hemipterus at levels comparable to the inoculated banana [87.8 + 2.5% (X +_ SEM) for the synthetic odor and 67.9 _+ 6.3% for the natural odor, N = 4 groups of 10 beetles, P > 0.05). In addition, both the synthetic blend and the inoculated banana elicited beetle response at a level several times greater than that to a blend of ethanol, acetaldehyde, and ethyl acetate (9.5 + 4 . 8 % , P < 0.05), a combination previously reported to be an effective attractant for C. hemipterus in the field (Smilanick et al., 1978). In the first of a series of bioassays of synthetic blends, the 18-component blend again was very similar to the inoculated banana in eliciting upwind orientation; however, the elimination of all members of any of the three groups of functionalities resulted in a reduction in beetle response (Figure 3). The per-
1260
PHELAN AND LIN
TABLE 2. PERCENT COMPOSITION OF VOLATILE CONSTITUENTS IN HEADSPACE OF
Saccharomyces cerevisiae-INOCULATED AGAR MEDIUM ( S M ) , ASEPTIC AGAR MEDIUM ( A M ) , S. cerevisiae-INOCULATED BANANA (SB), AND ASEPTIC BANANA (AB); RATIO OF ABUNDANCE OF EACH COMPONENT IN SB VERSUS A B HEADSPACE; AND COMPOSITION OF SYNTHETIC ODOR BLENDS
Percent composition (non-EtOH)" Component Alcohols Ethanol Propanol 2-Methylpropanol Butanol 2-Pentanol 3-Methylbutanol Aldehydes/ketones Acetaldehyde 2-Butanone~ 2-Pentanone 3-Hydroxy-2-butanone Esters Ethyl acetate Isobutyl acetate Ethyl butyrate Butyl acetate e Ethyl isovalerate e Isopentyl acetate Isopentyl butymte" Isopentyl isovalerate e Totals (non-EtOH) Alcohols Esters Aldehydes/ketones
SM
AM
SB
AB
SB : AB (absolute quantity) b
NQ J
NQ
25.73 1.13
87.14
(85.05) 10.30 8.49 0.72 0.44 10.63
(86.53) 0.49 7.61 0,84 0.94 6.27
55.36 0.09 0.65 0.27
54.13
11.78 0.25 0.45
