Journal of Chemical Ecology, Vol. 9, No. 7, 1983

IDENTIFICATION OF HOST PLANT ATTRACTANTS 1 F O R T H E C A R R O T FLY, Psila r o s a e

P.M. GUERIN, E. STADLER, and H.R. BUSER Swiss Federal Research Station CH-8820 Wiidenswil, Switzerland

(Received July 12, 1982; revised November 15, 1982)

Abstract--Cold-trapped carrot leaf volatiles were analyzed by gas chromatography with an outlet splitter to a flame ionization detector and to a carrot fly antennogram preparation as the second detector (GC-EAD). Strongest E A D responses were elicited by products whose elution temperatures corresponded to the propenylbenzenes, trans-methylisoeugenol (3,4dimethoxy-l-propenylbenzene) and trans-asarone (2,4,5-trimethoxy-1propenylbenzene) and, to a lesser extent, by-products matching the elution temperatures of the leaf aldehydes hexanal, (E)-2-hexenal, and heptanal, and of the terpenes linalool and caryophyllene. The identity of the propenylbenzenes was confirmed by gas c h r o m a t o g r a p h y - m a s s spectrometry. G C - E A D permitted accurate estimation of the olfactory thresholds; it was lowest for trans-asarone at 500 attogram (5 • 10-16g)/ml of air passing over the antenna. Both the leaf aldehydes and propenylbenzenes were attractive when tested individually in the field with yellow sticky traps; fly captures were linearly related to the quantity of propenylbenzenes applied per trap. A combination of trans-asarone and hexanal was more attractive than either compound singly, suggesting that the fly is adaptively equipped to respond to a mixture of c o m p o u n d s emanating from carrot foliage. In laboratory choice tests, flies were more attracted by vapors from intact carrot foliage than by that from a nonhost; ieaf odor alone also mediated oviposition. We conclude that through the selectivity and sensitivity of its response to foliar volatiles, the carrot fly may achieve host-plant orientation and also at close range, in union with its response to less volatile leaf surface components, selection of an oviposition site. Key W o r d s m C a r r o t fly, Psila rosae, Diptera, Pgilidae, host odor, leaf aldehydes, propenylbenzenes, trans-asarone, trans-methylisoeugenol, gas chromatography linked electroantennographic detection, attractants, field traps, aldehydes, aromatics, alcohols, terpenes.

~Diptera: Psilidae. 843 0098-033l,'83/0700-0843503.00/0 cr [983 Plenum Publishing Corporation

844

GUE~N ET AL. INTRODUCTION

Leaf color and shape and the presence of contact chemostimulants in the cuticular layer of carrot leaves mediate, at least in part, host plant selection by the carrot fly Psila rosae (F.) (Stadler, 1977); trans-methylisoeugenol, a propenylbenzene isolated from carrot leaves, is an oviposition stimulant for the fly (Beruter and Stgdler, 1971; Stadler, 1972). It has been suggested that olfactory cues do not influence host finding in this species (Bohlen, 1967; Stadler, 1972). This is improbable, however, since this oligophagous pest is attracted by Umbellifer-associated propenylbenzenes in the field (Guerin and Stadler, 1980). Indeed, the olfactory system of the carrot fly is sensitively tuned to the perception of a group of generally occurring green leaf volatiles and some more host-specific products (Guerin and Visser, 1980). Here we identify the principal attractants in carrot foliage for the carrot fly and demonstrate their role in field attraction, host selection, and oviposition.

METHODS AND MATERIALS Laboratory Bioassays Cages, 50 • 50 • 50 cm, as described by Stadler (1977) were used routinely with uniform illumination throughout and maintained at 2! ~ C, 75% relative humidity witl~ 16 hr light. One hundred females and 50 males between 3 and 6 days old (peak oviposition age) were obtained from a laboratory culture and introduced into the bioassay cage 24 hr before tests. Prior to this, flies were provided with an apple seedling (40 cm high) as a copulation site and with carrot leaves for oviposition. Host Selection and Oviposition. Flies were offered the vapor over carrot or fern, Nephrolepis exaltata (L.) (nonhost), to determine the influence of volatiles on host selection. The foliage of each plant (20 g) with stems in water was enclosed within a black Plexiglas cylinder (14 X 20 cm high) and placed at diagonally opposite corners of the bioassay cage. Bronze gauze (1-mm mesh) sealed the opening to the cylinder from above, leaving a clearance over the foliage. The number of females alighting and probing the gauze with the ovipositor was counted every 15 rain during the daily peak in oviposition (last 3 hr of light); females were blown off and cylinder positions reversed at each count. The effect of plant vapor on oviposition was measured by daily counts of egg numbers deposited on two moistened oviposition substrates (5 em diameter) (of. Stadler, 1977) supported immediately beneath the bronze gauze. To avoid any bias due to repellents from fern, tests were also made with moist filter paper as control.

ATTRACTANTS FOR CARROT FLY

845

Oviposition Stimulants. Host-plant volatiles eliciting strong electroantennogram (EAG) and behavioral responses were examined for their effects on oviposition. A 5-mg sample of the substance in methylene chloride (Fluka, puriss, grade) was applied to an artificial leaf made of a double strip of chemically pure filter paper (5 • 15 cm) cut pinnatedly; two replicates of the impregnated surface were attached to oviposition substrates (cf. Stadter, 1977) and placed in the bioassay cage with controls to which solvent only had been applied. Eggs were counted and positions fully randomized daily for a minimum of 3 days. Chemistry Foliar Extract. Paraffin oil was used as a slow release medium for foliar volatiles. Some 50 g of carrot foliage frozen at - 2 0 ~ and ground in a coffee blender was dissolved in 50 ml paraffin oil (Merck, Uvasol grade) with 25 g Na2SO4. After blending (1 min) the slurry was centrifuged for 10 min at 8000g and the oil plus dissolved volatiles was filtered over Na~SO4 on W h a t m a n No. l filter paper in a pressure funnel under N2 at 10 psi. From a 10-2 dilution of the extract (v/v in paraffin oil) 1 ml was placed in a pill glass beneath the bronze gauze of the oviposition device, described above, with paraffin oil alone as control at the opposite corner of the cage. Eggs were counted, positions reversed, and odor source replaced daily. Collection and Analysis ofFoliar Volatiles. Some 100 g of carrot foliage (stems plus leaves) were shredded into a l-liter gas-wash bottle. Pressurized air was driven over charcoal through the bottle at 1 liter/min for 5 rain and subsequently through a U tube (30 • 0.5 cm ID) immersed in acetone and dry ice in a Dewar flask; three such samples were taken in separate U tubes. The volatiles in each cold trap were dissolved in 0.5 ml pentane (Fluka, puriss. grade) and the combined washings were dried by placing the extract in a glass vial at - 2 0 ~ where the remaining water droplets froze on the walls. The extract was analyzed by gas c h r o m a t o g r a p h y (GC) on a SP-1000 glass capillary column (15 m • 0.3 m m ID) with outlet splitter to a flame ionization detector (FID) and an electroantennogram preparation, termed the electroantennographie detector (EAD) in a ratio of 2: 1. The EAD detector incorporated a female carrot fly antenna mounted in the manner already described (Guerin and Visser, 1980) connected to a differential amplifier (BB 3670) with high input impedance (1013 ~ ) and low bias current (95%, GC). The odor delivery system and recording technique has been described (Guerin and Visser, 1980). The oviposition stimulant, transmethylisoeugenol, was employed as standard since the antenna is sensitive to the product and its field attraction for the fly had already been established (Stadler, unpublished data). Plant volatiles showing EAG activity greater than that of trans-methylisoeugenol at 10-3 (v/v in paraffin oil) were subjected to behavioral assays.

FieM Tests Experimental Design. Attraction to individual volatiles obtained from commercial sources and whole essential oil extracts (as supplied by J. Chiquet, Basel) from six host plants of the carrot fly was field tested during the second generation flight in 1979 and 1980. trans-Asarone was supplied by Sigma Chemical Company, St. Louis, Missouri; trans-methylisoeugenol by K & K Labs Division, Plainview, New York; and the aldehydes hexanal, (E)-2-hexenal, and heptanal by Fluka AG, Buchs, Switzerland. Tests were undertaken in large fields of mature carrots in areas under intensive cultivation and supportirIg large populations of the fly at Tagerwilen (Canton Thurgau), in the Berner Seeland (Cantons Bern and Fribourg) and at Wadenswil. The volatiles were tested in conjunction with a yellow color trap (a sheet of Plexiglas 20 • 20 cm, ICI-229) coated with Tanglefoot| glue; this trap is an established tool for monitoring the carrot fly (Freuler et al., 1982). The traps were attached to timber poles 60 cm above the ground which, depending on crop growth, gave a clearance of between 10 and 30 cm above the carrot foliage. Where possible, the experiments were laid out in a Latin square with a minimum of 5 replicates per treatment and 5 m between individual traps. Otherwise, randomized blocks were used. Insect captures were counted at three-day intervals or more frequently as population densities warranted. StatistiealAnalysis. Trap captures were compiled in a contingency table and tested for homogeneity by means of X 2. Significance between treatments was established using a single classification X 2 and comparisons between individual treatments was made using the formula u = (a - b) / N / 2 based on the multinomial distribution and normal approximation; a and b are the pooled totals for the treatments being compared and N the grand total summing through all treatments. This test is highly efficient (Berchtold, personal communication) and can be carried out with ease.

847

A T T R A C T A N T S FOR C A R R O T FLY

Dispensers. Fly capture was measured as a function of the release rate of trans-methylisoeugenol from silicone tubing (3 mm ID X 6 mm OD); the quantity given off is linearly related to the length of the tube: 0.5-, 5.0-, 50-, 125-, and 250-cm pieees release a mean of 1.0, 12, 106, 244, and 560 m g / d a y (r = 0.997, P < 0.001 with three degrees of freedom). These different lengths of tubing, filled with the product and stoppered at each end, were stretched between clamps across the top of the trap. trans-Asarone, being crystalline, was uniformly applied using a pipet to the glue in pentane (Fluka, purum grade) at 5.0, 50 and 500 mg/trap (a half to each face). A polyethylene capsule, 30 • 15 mm diameter with four l-ram holes pierced in the lid and attached to the top of the trap was employed for the C6 and C7 green leaf alcohols and aldehydes; this dispenser releases (E)-2-hexenal at 15 mg/day in warm weather (midday temperatures at 24-28 ~ C). The same dispenser sufficed for all the other volatiles while a 20-cm length of polyvinylchloride tubing (3 mm ID • 5 rnm OD) dispensed the essential oil extracts. RESULTS

Analysis of Foliar Volatiles GC-EAD. Based on the chromatogram, the cold-trapped volatiles may be divided into three sections: (1) low-boiling-point compounds eluting with or on the tail of the solvent peak (30-60 ~ C); (2) those eluting midway through the temperature program (100-130~ and (3) two high-boiling-point compounds eluting in the higher temperature range (>165~ (Figure 1). Distinct EAD responses were obtained in each of the three sections coinciding with peaks whose elution temperatures match those of known standards. EAD responses within the solvent peak correspond to hexanal (34 ~ C) and to (E)-2-hexenal and heptanal, coeluting at 44~ (Figure 1). The presence of large amounts of these volatiles was confirmed in the absence of solvent by taking a 0.5-ml sample of air over the shredded foliage in a gas-tight syringe and injecting it directly onto the column (Guerin and Stadler, 1982). In section 2, the peaks evoking strong EAD responses at 102~ and I04~ correspond with the elution temperatures of linalool and cis-caryophyllene, respectively; none of the other peaks evoking strong EAD responses in this section of the chromatogram have as yet been examined (Figure 1). The most pronounced EAD responses were obtained in section 3, at 169.5~ corresponding to the elution temperature of trans-methylisoeugenol and 5.5 rain later at 180~ (upper temperature limit of the program) matching transasarone (Figure I). While trans-methylisoeugenol was represented by a distinct peak on the chromatogram, the quantity of trans-asarone was below the detection limit of

848

GUERIN ET AL.

!

'3

EAD

FID

o.

0

attenuation x8

i 180

180

15o Temperature

in'i 9o (~

6o

F1G. 1. Gas chromatograms with (1) EAD responses of a carrot fly antenna and (2) FID responses to cold-trapped volatiles of carrot foliage. Splitless injection, I min at 30~, 10~ to 60 ~ and 5~ to 180~ with He as carrier gas. For further specifications see text. the F I D . However both products were present in sufficient quantity for confirmation by G C - M S ( d i m e t h o x y - - p r o p e n y l b e n z e n e : m/z 178, M+; 163, M+-CH~; 147, M+-OCH3; 135, M+-CH3-CO; and trimethoxy-propenylbenzene: rn/z 208, M+; 193, M+-CH3; 177, M+-OCH3;165, M+-CH3-CO). Evidence was also obtained for the presence of eis isomers of both products in the extract. The a m o u n t of trans-methylisoeugenol and trans-asarone present was estimated at 343 ng and 75 ng, respectively, based on comparisons o f E A D responses obtained with the extract to those established by injecting k n o w n quantities of standards. Since the 15 liter of air drawn over the 100-g shredded foliage can be expected to have carried off but a fraction of the volatiles, we may assume that an average carrot leaf weighing 5 g contains, at a minimum, 17.2 ng trans-methylisoeugenol and 3.75 ng trans-asarone.

849

ATTRACTANTS FOR CARROT FLY

Response Thresholds. The response curves obtained by passing successive dilutions of the two propenylbenzenes simultaneously over the F I D and E A D demonstrates the relative sensitivity of the antenna to the products and permits the estimation of perception thresholds (Figure 2). The lowest injected quantity evoking a distinct E A D response was 100 pg transmethylisoeugenol and 100 fg trans-asarone (Figure 2). Taking into account the split ratio between the two detectors, the rate of air flow over the antenna and the manner in which the compounds elute from the column ( F I D peak shape), the perception threshold (expressed as the quantity of c o m p o u n d / m l of air passing over the antenna), is 1 p g / m l (3.4 • 109 molecules) for transmethylisoeugenol and 2000 times less at 500 a t t o g r a m / m l (1.4 X 106 molecules) for trans-asarone [1 attogram (ag) = 10-I8g]. EA G Screening. The responses evoked by umbelliferous volatiles in the EAG confirm the results obtained by G C - E A D analysis of the foliar extract. The responses detailed here are at a 10-3 dilution of each product in paraffin oil, a level more indicative of the selectivity of the olfactory system than higher concentrations. Since the sensitivity of an antenna drops during the test period the EAGs are expressed as a percentage of the response to a standard (trans-methylisoeugenol at 10-2) which was applied at regular intervals. o trans-Asarone o t r a n s - M e t h y l - i s o - eugenol 30

2C "0 ,(-,,,. 6 6 3 p v

0

Q.

E 1111[

o UJ

o o = lOfgX

o 1 100fgX

o o i lpgX

i 10pgX

o

o i 100pgX

L lngX

10ngX

0,33

A m o u n t of h o s t plant a t t r a c t a n t

FIG. 2. Responses of a carrot fly antenna to propcnylbenzenes obtained by employing combined gas chromatography and electroantennographic detection (GC-EAD).

g50

GUERIN ET AL.

Strong EAG responses were evoked by three groups of compounds: (l) the leaf aldehydes such as heptanal (37.2 +_ 0.2), octanal (31.0 + 7.5), nonanal (19.8 + 4.7), and the leaf ester hexyl acetate (44.1 +_ 13.0); (2) the acyclic terpene linalool (47.3 + 17.1), and sesquiterpene cis-caryophyllene (49.8 _+ 13.8); and (3) to the allyl- and propenylbenzenes listed in Table 1. trans-Asarone evoked the highest EAG response (146.7 + 13.5), four times higher than its geometric isomer cis-asarone (39.3 + 7.0) and 10 times that of its analog, trans-methylisoeugenol (13.7 + 1.3) (Table 1 and Figure 3); the related benzaldehyde, asaronaldehyde, elicited no response. The presence of an extra methoxy group in trans-asarone apparently accounts for the 10-fold difference in response over its dimethoxy analog, trans-methylisoeugenol. Male and female responses to trans-asarone and trans-methylisoeugenol were equal both relative to the standard and in absolute amplitude. It may be of interest to compare the responses evoked by the allyl- and propenylbenzenes from the viewpoint of structure activity relationships. Only the di- and trimethoxy-propenylbenzenes trans-methylisoeugenol and transasarone elicited a significantly higher response than their allylbenzene counterparts, methyleugenol (3,4-dimethoxy-l-allylbenzene) and elemicine (3,4,5-trimethoxy-l-allylbenzene) (Table 1 and Figure 3). The combination of three methoxy units on the benzene ring with the stereochemistry of the propenyl group apparently accounts for the unique EAG activity evoked by trans-asarone (Figure 3). Both the propenylbenzenes, anethole (4-methoxy-1propenylbenzene) and isosafrole (3,4-methylenedioxy-l-propenylbenzene), elicited as high a response as trans-methylisoeugenol but no higher than their allylbenzene analogs, methylchavicol and safrole (Table 1). As contrasted TABLE 1. E A G RESPONSES OF FEMALE CARROT FLIES TO PROPENYL- AND ALLYLBENZENES AS PERCENTAGE OF RESPONSE TO trans-METHYLISOEUGENOL AT 10 -2a ( ~ ONE STANDARD ERROR; N = 6 IN EACtt CASE)

Propenylbenzenes 10-3

rrans-Asarone cis-Asarone

trans-Methylisoeugenol Anethole 4-Methyl-l-propenylbenzene lsosafrole Isoeugenol

Allylbenzenes 10 3 146.7 + 39.3 + 13.7 • 12.2 _+ 13.3 -10.5 -+ 1.2_+

13.5 7.0 1.3 1.0 2.0 2.0 0,8

Elemicine

7.0 • 3.2

Methyleugenol Methylchavicol

6.3 _+ 2~2 15.0 + 2.7

Safrole Eugenol Myristicin Apiole

I4.5 + 1.9• 10.6 • 2.6 •

~Concentration of the compound in paraffin oil expressed on a vol ume / vol ume basis.

0.9 l,l 2.3 2.6

ATTRACTANTS FOR CARROT FLY

851

Allyl- and Propenylbenzenes

CH30 CH30;OCH3 OCH3

~OCH3 OCH3

T "OCH3 OCH3

EAG Activity

:

+

--

Elernicine

trans-

Methyleugenq[ Mehylisoeugenol

CH30 T "OCH3 OCH3

+4-+++ trans-

Asarone

T "OCH3 OCH3 ++ cisAsarone

Fro. 3. Structures of some allyl- and propenylbenzenes and their relative strengths as olfactory stimulants for the carrot fly. with the propenylbenzenes, increasing methoxylation ofallylbenzenes apparently reduces EAG activity since methylchavicol was significantly more active than methyleugenol, and elemicine and safrole were more active than the mono- and dimethoxy-methylenedioxy analogs, myristicin and apiole (Table 1). Both the hydroxy products, isoeugenol (4-hydroxy-l-propenylbenzene), and the allylbenzene, eugenol, evoked little or no response (Table 1). Behavioral Observations Host Odor Discrimination. Volatiles from carrot foliage attract the carrot fly; 10 times as many insects alighted in the vapor over foliage from the host plant as compared with fern (Table 2). Under these test conditions, apparently only females were attracted since all alighting insects displayed the typical behavior response of probing the metal gauze with the ovipositor. This response by gravid females was strong considering the rigors of blowing off the responding insects and reversing the positions of the test materials every 15 min. Females deposited significantly higher egg numbers over the foliage of carrot than over fern. Similar results were obtained with moist filter paper as control, thus eliminating any bias arising from possible repellency of the nonhost (Table 2). Flies also showed a preference for the vapor over a paraffin oil extract of carrot foliage as a site for oviposition, demonstrating that the pertinent volatiles were extractable and sufficiently stable for analysis. Oviposition Stimulants. Significantly higher egg numbers were deposited near artificial substrates treated with trans-asarone (139.6 _+ 12.1) than on controls (90.4 _+ 9.8) (P < 0.0005, Friedman test); a similar observation was made with trans-methylisoeugenol in an earlier study (Bert~ter and Stadler, 1971). No such effect was observed for cis-asarone, asaronealdehyde,

852

GUERIN ET AL. TABLE 2. LABORATORY BEHAVIOURAL OBSERVATIONS WITH CARROT FLIES

Behavior Female landings in the headspace vapor over Carrot foliage 5.0 _+0.7 Egg numbers deposited in the headspace over Carrot foliage 89.0 _+ 20.8 Carrot Foliage 17t.8 _+ 28.0 Carrot leaf extract 134.6 _+ 13.3

Nonhost 0.5 _+ 0.8

Blank control 42.3 -+ 13,4 Nonhost 82.5 • 8.4 Blank control 79.5 _+ 10.3

N

Significance level (P)

12

Identification of host plant attractants for the carrot fly,Psila rosae.

Cold-trapped carrot leaf volatiles were analyzed by gas chro-matography with an outlet splitter to a flame ionization detector and to a carrot fly ant...
995KB Sizes 0 Downloads 0 Views