Journal of Chemical Ecology, Vol. 22, No. 2, 1996
APPLE FOLIAGE E N H A N C E S MATING DISRUPTION OF LIGHT-BROWN APPLE MOTH
D.M. SUCKLING,* G. KARG, and S.J. BRADLEY HortResearch, P.O. Box 51 Linc(~ln Canterbury, New Zealand (Received April 14. 1995; accepted October 15, t995)
Abstract--Uptake and release of pheromone and behavioral inhibitor of Epiphyas postvittam~a by apple leaves was tested using field electroantennograms (EAG). trap catches to synthetic lures and virgin ten-tales, and chemical analysis, Trap catches in single apple trees (N = 31 were monitored for six cycles of six days" duration, using delta traps baited with synthetic pheromone. Polyethylene dispensers t0, 1. 10 per tree) releasing pheromone and inhibitor were present for only the first three days of each cycle. Application of 10 dispensers per tree resulted in complete disruption of trapping, which continued lot one day after dispensers were removed. Over the three nights following the removal of the dispensers (days 4-6), trap catch was 0, 10, and 15% of the control catch, In contrast, the presence of only one dispenser per tree led to 0-20% of control catches, but on the three nights following dispenser removal catches were 35, 40. and 80% of the control catch. Field EAGs indicated significantly higher relative pheromone concentrations in the trees with 10 dispensers present, compared to trees with single dispensers, but removal of dispensers produced no detectable treatment effect compared to the control trees one day after dispenser removal. In a second experiment, releases of marked male moths into apple orchard plots following the removal of polyethylene dispensers ( 1 hr earlier that day) resulted in significantly lower catches in traps baited with virgin females in blocks that had been treated. compared to controls. Recovery of pheromone by solvent washing of leaves loaded with 50 gg of the main component of the sex pheromone (1.26 ug/cm 2) was low (2.5% L Leaves held in a pberomone-saturated atmosphere were loaded with 0.045 ± 0,007/zg pberomone/cm 2. Analysis of apple leaves taken from a pheromone-treated tree at different distances from the pheromone dispenser showed a decay of the pheromone load per square centimeter with increasing distance from the dispenser, as previously indicated by EAG.
*To whom correspondence should be addressed.
325 i~lt/8.(}331,gt~,'tt2{~l.II325$t~450/0 ~ lq~)6PlenumPublishingC~r'Dmm~m
326
SUCKLING, KARG, AND BRADLEY Key Words--Epiphyas postvittana. Tortricidae, mating disruption, pheromone, apple leaves, adsorption, release rate, electroantennogram, EAG, lightbrown apple moth.
INTRODUCTION
Adsorption and release of insect sex pheromone has been recorded from several plants (Wall et al., 1981; Noldus et al., 1991: Karg et al., 1990, 1994; Sivinski et al. 1994). Wall et al. (1981) and Wall and Perry (1983) showed the possible behavioral importance of this phenomenon in male pea moth (Cydia nigricana F.), which precisely located the previous position of a pheromone trap after its removal. Pea leaves functioned as a secondary source of pheromone and release rates were sufficient for male moths to orient to the vacated position of the odor source, The adsorption and release of pheromone by foliage of brassicas more recently has been described by Noldus et al. (1991). Their aim was to investigate the possible role of plants to bridge the gap between the time of calling of females and the searching behavior of parasitoids. Their experiments showed that plants adsorbed the female sex pheromone and released it over an extended period of time. In addition, they showed that male Mamestra brassicae L. was attracted to Brussels sprouts (Brassica oleraceae L.) releasing the female sex pheromone. The impact and potential importance of adsorption and release of pheromone by plants on the temporal and spatial distribution of pheromone in mating disruption has long been neglected. However, experinaents by Karg et al. (1990) and Karg and Sauer (1992) showed the importance of grape vines for the pheromone concentration achieved in a vineyard treated with pheromone t'or mating disruption of the European grapevine moth Lobesia botrana Hb. In areas with fully developed vegetation, the pheromone concentration was more than 100fold higher than the same location without foliage in spring or in a defoliated vineyard nearby (Karg and Sauer, 1992, 1995). In addition, the plant canopy affects the structure of the odor plume, with pheromone more evenly distributed in vineyards with a dense plant canopy. The EAG signal shows more intermittency without vegetation. We reported greater variance of EAG signals recorded above grass, compared to adjacent apple tree rows (Suckling et al., 1994). Bengtsson et al. (1994) also showed that pheromone concentrations were lower above a pea canopy and the EAG signal fluctuated more above than within the canopy, although the signal variance was not examined in detail. Measurements of the adsorption and release of the pheromone of Epiphyas postvittanna (Walker) showed the effect of apple leaves in more detail (Karg et al., 1994). Pheromone released from apple leaves was still detected 24 hr after exposure of the leaves to pheromone. Traps baited with leaves exposed to pheromone caught significant numbers of males for three nights after treatment. EAG
APPLE FOLIAGE AND LIGHT-BROWN APPLE MOTH
327
measurements along a transect through a pheromone-treated orchard showed significant differences in the EAG response and therefore in the mean pheromone concentration, with increased concentrations closer to the trees in which the dispensers were present. These results were interpreted as showing that leaves could affect the atmospheric pheromone concentration and its temporal distribution and could thereby enhance the success of mating disruption. In this study, we hypothesized that the release of pheromone from foliage would be sufficient to disrupt male behavior. Experiments were carried out in order to examine the impact of pheromone adsorption and release on apple trees. In addition, the adsorption by apple leaves of the pheromone and inhibitor of E. postvittanna was quantified by chemical analysis.
METHODS AND MATERIALS
h~sects. Male and female E, postvittana were obtained as pupae from a colony in Auckland, New Zealand. The pupae were then kept at 20°C until eclosion, and kept at 12°C for one to two days before use. Pheromones, Dispensers containing 54.9 mg (E)-11-tetradecen-l-yl acetate (El 1-14:OAc), 2.5 mg of (E,E)-9,11-tetradecan-dien-l-yl acetate ( E 9 , E 1 1 14:OAc), 19.7 mg (Z)-ll-tetradecen-l-yl acetate ( Z l l - 1 4 : O A c ) as well as 16.8 mg of other substances such as stabilizers, were obtained from Shin-Etsu Chemicals Co., Tokyo. This blend has the disadvantage that Z11-14 : OAc inhibits trap catch when presented with the pheromone (Rumbo et al., 1993). However, this blend has been used very successfully in a great number of mating disruption trials (Suckling and Shaw, 1991, t995; Suckling et al., 1990a, 1994) and was therefore tested in these experiments. Traps. Standard delta traps (Suckling and Shaw, 1990) with sticky bases, baited with rubber septa containing 100 p.g of El 1-14 : OAc and 5 ~tg of E9,E1 t 14:OAc (Bellas et al., 1983) were used. Traps were checked daily and sticky bases changed every 10 days. To avoid contamination, separate traps were used tot controls when dispensers were present, and after dispenser removal, for both dispenser treatments. In the second field experiment, these traps were baited with virgin female moths held in gauze canisters (Suckling et al., 1990b). Electroantennogram Apparatus. The electroantennogram (EAG) apparatus and measuring procedure used in these experiments are described in Karg et al. (1994) and Suckling et al. (1994) in full. Briefly, the insect antenna was held between wells containing Ringer solution and silver-silver chloride electrodes. The antenna holder was placed inside a glass chamber, protecting the antenna from ambient air, with a steady airstream (ca. 2.5 liters/min) through the chamber transporting odors or charcoal-filtered air to the antenna. The signal was amplified, filtered, and stored on a computer. For calibration purposes, a 15-ml
328
SUCKLING, KARG, AND BRADLEY
volume of air was blown by a second pump across a rubber septum and into the main airstream. The rubber septum (Arthur H. Thomas, Philadelphia, Pennsylvania) was loaded with 105 /~g of the true pheromone blend as used in the delta traps. Three measurements of antennal responses were averaged and normalized to the responses to three calibration pulses in charcoal-filtered air, giving one reading of relative EAG amplitude. Normalizing was defined as EAG~m~ien~air/EAGcalibra~ion.Normalization was necessary because of differences in antennal sensitivity to pheromone, as well as changes over time in the same antenna. Nine EAG readings were made for each day of each treatment. The normalized EAG values for the control sites were subtracted from those taken at treated sites. This value is defined as the differential EAG amplitude, Field Experiment 1. The trial site was a mixed cultivar orchard block (5 m between rows, 5 m within row tree spacing), at the Lincoln University Biological Husbandry Unit. This site was chosen due to its spray-free status and a history of high E. postvittana catches during past seasons. Three trees in each block (three blocks) were chosen with similar tbliage cover, with at least one guard tree between experimental trees. Each tree had three trap sites marked approximately equidistant around the periphery (radius of ca. 1.5 m) at ca, 1.6 m height. Trees were untreated (control) or treated with either one dispenser in the center of the center-leader tree, or one dispenser placed at each of 10 sites (one in the center of the tree and three between each trap site, around the tree periphery). Dispensers were removed after three days. Trap catch was recorded for this period and for the following three days. After this period, the treatments were cycled among the same trees, in the following manner: after three days without dispensers, the trees that had been treated with the one dispenser treatment became the controls, the trees with 10 dispensers received the treatment with one dispenser, and the controls received the 10 dispenser treatment. To minimize effects of the daily variation of moth flight, treatment of each block was staggered one day and each block was used tbr six cycles. EAG recordings were taken at the trap positions (as described above), after removal of the traps. Field Experiment 2. The effect of disruption from pheromone and inhibitor released from foliage alone was also examined in the following manner. Dispensers (N = 36) as above were placed into orchard plots (N = 4, 22.5 × 22.5 m), consisting of six rows of trees, with I2 trees per row for two days. Comparable blocks (N = 4) 60 m away were untreated. Nine delta traps, each baited with three virgin female moths held in a gauze container (Suckling et al., 1990b), were put into a 3 × 3 array (9-m sides), at 1.5 m height, in order to maximize moth catch in each plot. Traps were placed in alternate trees from those that had contained dispensers in the central three rows (one guard row on one side, and two guard rows on the other). Male moths were anesthetized with CO2 and marked with felt tip pens (Suckling et al., 1990b). After two days in place,
APPLE FOLIAGE AND LIGHT-BROWN APPLE MOTH
329
dispensers were removed from the orchard at 4 PM. At 5 PM, 52-125 marked males were released among four sites in the middle two rows between traps, Moths were released onto a plastic tray, and those incapable of flight were subtracted from the total. Trap catch of wild and marked moths was recorded the next day. This procedure was repeated four times, in a different pairs of treated and control blocks on each night. GCand GC-EAD. A Hewlett Packard model HP 5830 A gas chromatograph with a DB-Wax capillary column (30 m × 0.25 mm) and a flame ionization detector (FID) was used for chemical analysis. After 3 min at 80°C, the temperature was increased at 10°C/min. up to 230°C, The carrier gas was hydrogen at 140 kPa+ A t : 1 split separated the outlet of the column between the FID and the electroantennogram detector (EAD). The EAD signal was amplified and filtered according to standard procedure (Hansson et al., 1989) and printed with a chart recorder at a chart speed of 33 mm/sec. The FID signal was recorded with a HP 18850 A GC-Terminal integrator. Experiments were carried out in order to identify the GC peaks representing pheromone and inhibitor E 1 1 - 1 4 : O A c and Z I I - 1 4 : O A c , respectively, and n-decan-l-ol (I 1 :OH) (which elicits no EAD response). These compounds were independently injected in order to measure their respective retention times. For the analysis of the leaf extracts, 6 #1 of the solvent extract was injected. All GC-EAD experiments were repeated at least three times. The low quantity of diene present in dispensers prevented its inclusion in these experiments, due to difficulties in obtaining reproducible results near the detection threshold. Topical Application of Pheromone onto Leaves. Five micrograms of E1114:OAc (50 izt n-heptane containing 100 ng//~l Et 1-14:OAc) was applied to 15 leaves (of similar age and an approximate size of 40 cm2), and left for 30 rain to be adsorbed by leaf waxes, before being put into a polyethylene bag. The same number of untreated control leaves were in a second plastic bag. After 14 hr, the apple leaves were taken out of each plastic bag and washed in 50 ml solvent (25 ml redistilled n-heptane and 25 ml redistilled n-hextane) for 14 hr on an orbital shaker at 150 rotations/min. The extract was then filtered twice using analytical filter paper (Munktells, Sweden, type 1002). As an internal standard, 5 ~1 11 : OH was added to the extract. The volume of the extract was reduced to 10 ml under nitrogen, and 6 ~1 of the extract was injected into the GC for both GC and GC-EAD analysis. Pheromone Loading of Leaves at Saturation. Three small branches with six to seven apple leaves were taken from an untreated orchard, the cut stems placed in a vial containing water, and the branches taken to the laboratory. There each branch was placed in a 20-liter plastic bag each containing a small fan generating an intermittent airstream at 0.25 m/sec (cycling on and off every 30 min) for 14 hr. The plastic bags were sealed for separation from the outer
330
SUCKUNG, KARG, AND BRADLEY
environment. In the first plastic bag, four aged dispensers with an average liquid length of 148 mm were attached to a metal frame inside the sealed plastic bag. The mean release rate of each dispenser, estimated by liquid length (Bradley et al., 1995), was 23.7 /zg per hour. It was ensured that there was no physical contact between the dispensers and apple leaves. The same number of untreated control leaves was placed in a second plastic bag. After 14 hr, the apple leaves were taken out of each plastic bag and treated as above. Pheromone Loading of Leaves Takenfrom Treated Fields. Ten aged pheromone dispensers with a mean liquid length of 146 mm were applied to 10 mature apple trees on May 18, 1994, at one per tree, After 39 days, leaves at different distances from the dispensers were picked and treated as described above. Leaves were taken at approximately 0.04 m, 0.3-0.4 m, and 0.9-1 m away from a central dispensers. As a control, 15 leaves were taken from an untreated orchard. Leaves were washed and the pheromone extracted as above. Calculation of Amount of Pheromone Uptake by Leaves. The size of all leaves used in these experiments was measured. Leaves of similar size and age were used as far as possible, and the amount of pheromone per square centimeter of leaf area was calculated by dividing the calculated amount of pheromone by the sum of the leaf areas. The amount of pheromone taken up by the leaves was calculated using the internal standard (50 p,g I I : O H ) as follows: amount of pheromone = (areaphe........ X 50 #g)/areainT~m~l~a,d~a)" Statistical Analysis. In the first field experiment, the catches of moths in traps were tested for fit to the normal distribution and found to be skewed. Catches were then compared between treatments using analysis of variance and t tests of the log transtbrmed total from three traps per tree and back-transformed for presentation (by anti-logs). Catches were pooled from each tree to avoid pseudoreplication (Hurlbert, 1984). The three blocks and six repetitions through time provided 18 replicates. In the second field experiment, trap catches in all nine traps per plot were again pooled because traps in each plot were not true replicates. Recapture rates were compared between treatment and control plots using Fisher's exact test (Zar, 1984), and catches of wild moths were compared by t test after log transformation. The analysis was performed using individual nights as replicates. EAG values were normally distributed and were compared untransformed using analysis of variance and t tests. RESULTS
Field Experiment t, Trap Catches. No moths were caught in trees treated with 10 dispensers while the dispensers were present (Table 1). The catch was also significantly suppressed for three days after removal of the dispensers, compared to the controls (t = 2.36, 16 df, P < 0.05). Trap catch for the three
0,30 0.04
0
(0.20) (0.04) (*)
Night 1
0,26 0
0
(0,02) 0 *
Night 2 0.21 0.04 0
(0.12) (0,04)
Night 3
"Significance from the control is indicated by: (*) for P < 0, 10 or * l o r P < 0,05.
10 dispensers
Control l dispenser
Treatment
Dispensers in tree
0.24 0.08 • 0
(0.10) (0.06)
Night 4
0.04
0.30 0.17
10.04)
(0.05) (0.08)
Night 5
Dispensers removed
0.(~
0.54 0.49
(008)
(0.26) (0.25)
Night 6
TABLE ]. BACK-TRANSFORMED MEAN CATCH PER TRAP PER NIGHT ( S E M ) OF MALE Epiphyas postvittana IN PHEROMONE TRAPS PLACED IN SINGLE APPLE TREES WITH OR WITHOUT POLYETHYLENE DISPENSERS RELEASING PHEROMONE AND I[NHIBITOR"
ta,a
6 -4
r-
> z
> O
3
>
332
SLICKLING, KARG, AND BRADLEY
nights after dispenser removal was 0, 10, and 15% of the control catch (Figure 1). Trees treated with a single dispenser also had suppressed trap catch during the presence of dispensers, Single moths were caught on two occasions, compared to 25 moths in the controls, so that catch in this treatment averaged 4% of catches in control trees. After removal of the single dispenser, trap catch subsequently increased over the following nights to 33, 40, and 80% of the control catch (Figure 1), with no significant difference from control catches (Table 1), Field F-rperiment 2. Moth catches were significantly reduced in blocks that had been treated with pheromone up to 1 hr before moth releases (Table 2). The mean level of recapture in the control was 28% (SEM 2.5%), with 10.3% (2.1%) recapture in the previously pheromone-treated plots. Hence, the pretreatment of foliage with pheromone resulted in a reduction to 37 % of the catches from that in the untreated plots. This level was very similar to the result on the first night of dispenser removal in the previous experiment (35%, Figure 1). Catch of wild moths was also reduced to 35% in previously treated [mean of 6.3 (SEMI.5)], compared to control plots [mean of 17.5 (SEM 5.1)] (t = - 2 . 4 3 , df = 6, P = 0.05). Field E~periment 1, EAG Measurements. Differential EAG measurements at the periphery of the trees (after the removal of traps) reflected the presence
100% [ ] 1 Dispenser
•
10 Dispensers
8O% O m
2
60%
O U
"6 o~
40%
C
~. 20%
[] 0%
Night
_
. . . .
ill
Dispensers in tree 1 2 3
L
Dispensers removed 5 6
FIG. 1. Percentage of trap catch of male Epiphyas postvittana in apple trees, recorded for three nights during pheromone and inhibitor application and for three nights after removal of the polyethelene dispensers.
333
APPLE FOLIAGE AND LIGHT-BROWN APPLE MOTH
TABLE 2. CATCHES OF MARKED ANDWILD MALE MOTHS IN ONE NIGHT IN APPLE ORCHARD PLOTS, TREATED OR NOT WITH PHEROMONE AND INHIBITOR RELEASED FROM POLYETHYLENE DISPENSERS FOR T w o DAYS UP TO ONE HOUR BEFORE MOTH RELEASES
Laboratory -reared moths"
No. released
Wild moths
Test
Treated
Control
Treated
Control
pb
Treated
Control
I 2 3 4
I25 116 117 52
119 119 114 57
7(6%) 15(13%) 19(16%) 3(6%)
28(24%) 32(31%] 38(33%) 14(25%)
0.02 0.01 0.01 0.07
9 2 7 7
32 8 15 15
"Percentage recapture for marked moths in parentheses, J'Fisher's exact test.
of high concentrations of atmospheric pheromone with one or l0 dispensers in the trees (Figure 2), although readings in the treated trees were only significantly higher than the control trees on the second day (P < 0.05). EAG values declined after removal of the dispensers, with no differences from control levels after 24 hr. The comparatively insensitive field EAG response in this species is due to 0.20
[ ] 1 dispenser
[]
10 d i s p e n s e r s
0.16
0.08
P~ 0.04
0.00
-0.04 Night
.
. 1
.
.
.
.
.
Dispensers in t r e e 2
3
.
.
.
~
Dispensers removed 4
5
-
I
6
FIG. 2. Differential etectroantennogvam amplitudes recorded in trees during the presence or after removal, of one or 10 polyethylene dispensers containing pheromone and inhibitor. Bars show the standard error.
334
SUCKLING, KARG, AND BRADLEY
the high background response to environmental odors (Karg et al., 1994; Suckling et al., 1994; Rumbo et al., 1995), GC-EAD Analysis. GC analysis of the extract of the untreated leaves did not show a G C peak with a retention time corresponding to those measured for E 1 1 - 1 4 : OAc or ZI 1-14 : O A c (Figure 3). There was also no antennal response from the male. E 1 1 - 1 4 : O A c as well as ZI 1-14 : O A c were present in the extract of saturated leaves as well as in the extracts of the leaves taken from the field
11:OH at 11.75 a ,T
t
. . . .
0 LM
Retention time (min) 11 :OH at 11.70
,-r
w,
.JUjL _u J jL24JU LJLJu
APPLE FOLIAGE E N H A N C E S MATING DISRUPTION OF LIGHT-BROWN APPLE MOTH
D.M. SUCKLING,* G. KARG, and S.J. BRADLEY HortResearch, P.O. Box 51 Linc(~ln Canterbury, New Zealand (Received April 14. 1995; accepted October 15, t995)
Abstract--Uptake and release of pheromone and behavioral inhibitor of Epiphyas postvittam~a by apple leaves was tested using field electroantennograms (EAG). trap catches to synthetic lures and virgin ten-tales, and chemical analysis, Trap catches in single apple trees (N = 31 were monitored for six cycles of six days" duration, using delta traps baited with synthetic pheromone. Polyethylene dispensers t0, 1. 10 per tree) releasing pheromone and inhibitor were present for only the first three days of each cycle. Application of 10 dispensers per tree resulted in complete disruption of trapping, which continued lot one day after dispensers were removed. Over the three nights following the removal of the dispensers (days 4-6), trap catch was 0, 10, and 15% of the control catch, In contrast, the presence of only one dispenser per tree led to 0-20% of control catches, but on the three nights following dispenser removal catches were 35, 40. and 80% of the control catch. Field EAGs indicated significantly higher relative pheromone concentrations in the trees with 10 dispensers present, compared to trees with single dispensers, but removal of dispensers produced no detectable treatment effect compared to the control trees one day after dispenser removal. In a second experiment, releases of marked male moths into apple orchard plots following the removal of polyethylene dispensers ( 1 hr earlier that day) resulted in significantly lower catches in traps baited with virgin females in blocks that had been treated. compared to controls. Recovery of pheromone by solvent washing of leaves loaded with 50 gg of the main component of the sex pheromone (1.26 ug/cm 2) was low (2.5% L Leaves held in a pberomone-saturated atmosphere were loaded with 0.045 ± 0,007/zg pberomone/cm 2. Analysis of apple leaves taken from a pheromone-treated tree at different distances from the pheromone dispenser showed a decay of the pheromone load per square centimeter with increasing distance from the dispenser, as previously indicated by EAG.
*To whom correspondence should be addressed.
325 i~lt/8.(}331,gt~,'tt2{~l.II325$t~450/0 ~ lq~)6PlenumPublishingC~r'Dmm~m
326
SUCKLING, KARG, AND BRADLEY Key Words--Epiphyas postvittana. Tortricidae, mating disruption, pheromone, apple leaves, adsorption, release rate, electroantennogram, EAG, lightbrown apple moth.
INTRODUCTION
Adsorption and release of insect sex pheromone has been recorded from several plants (Wall et al., 1981; Noldus et al., 1991: Karg et al., 1990, 1994; Sivinski et al. 1994). Wall et al. (1981) and Wall and Perry (1983) showed the possible behavioral importance of this phenomenon in male pea moth (Cydia nigricana F.), which precisely located the previous position of a pheromone trap after its removal. Pea leaves functioned as a secondary source of pheromone and release rates were sufficient for male moths to orient to the vacated position of the odor source, The adsorption and release of pheromone by foliage of brassicas more recently has been described by Noldus et al. (1991). Their aim was to investigate the possible role of plants to bridge the gap between the time of calling of females and the searching behavior of parasitoids. Their experiments showed that plants adsorbed the female sex pheromone and released it over an extended period of time. In addition, they showed that male Mamestra brassicae L. was attracted to Brussels sprouts (Brassica oleraceae L.) releasing the female sex pheromone. The impact and potential importance of adsorption and release of pheromone by plants on the temporal and spatial distribution of pheromone in mating disruption has long been neglected. However, experinaents by Karg et al. (1990) and Karg and Sauer (1992) showed the importance of grape vines for the pheromone concentration achieved in a vineyard treated with pheromone t'or mating disruption of the European grapevine moth Lobesia botrana Hb. In areas with fully developed vegetation, the pheromone concentration was more than 100fold higher than the same location without foliage in spring or in a defoliated vineyard nearby (Karg and Sauer, 1992, 1995). In addition, the plant canopy affects the structure of the odor plume, with pheromone more evenly distributed in vineyards with a dense plant canopy. The EAG signal shows more intermittency without vegetation. We reported greater variance of EAG signals recorded above grass, compared to adjacent apple tree rows (Suckling et al., 1994). Bengtsson et al. (1994) also showed that pheromone concentrations were lower above a pea canopy and the EAG signal fluctuated more above than within the canopy, although the signal variance was not examined in detail. Measurements of the adsorption and release of the pheromone of Epiphyas postvittanna (Walker) showed the effect of apple leaves in more detail (Karg et al., 1994). Pheromone released from apple leaves was still detected 24 hr after exposure of the leaves to pheromone. Traps baited with leaves exposed to pheromone caught significant numbers of males for three nights after treatment. EAG
APPLE FOLIAGE AND LIGHT-BROWN APPLE MOTH
327
measurements along a transect through a pheromone-treated orchard showed significant differences in the EAG response and therefore in the mean pheromone concentration, with increased concentrations closer to the trees in which the dispensers were present. These results were interpreted as showing that leaves could affect the atmospheric pheromone concentration and its temporal distribution and could thereby enhance the success of mating disruption. In this study, we hypothesized that the release of pheromone from foliage would be sufficient to disrupt male behavior. Experiments were carried out in order to examine the impact of pheromone adsorption and release on apple trees. In addition, the adsorption by apple leaves of the pheromone and inhibitor of E. postvittanna was quantified by chemical analysis.
METHODS AND MATERIALS
h~sects. Male and female E, postvittana were obtained as pupae from a colony in Auckland, New Zealand. The pupae were then kept at 20°C until eclosion, and kept at 12°C for one to two days before use. Pheromones, Dispensers containing 54.9 mg (E)-11-tetradecen-l-yl acetate (El 1-14:OAc), 2.5 mg of (E,E)-9,11-tetradecan-dien-l-yl acetate ( E 9 , E 1 1 14:OAc), 19.7 mg (Z)-ll-tetradecen-l-yl acetate ( Z l l - 1 4 : O A c ) as well as 16.8 mg of other substances such as stabilizers, were obtained from Shin-Etsu Chemicals Co., Tokyo. This blend has the disadvantage that Z11-14 : OAc inhibits trap catch when presented with the pheromone (Rumbo et al., 1993). However, this blend has been used very successfully in a great number of mating disruption trials (Suckling and Shaw, 1991, t995; Suckling et al., 1990a, 1994) and was therefore tested in these experiments. Traps. Standard delta traps (Suckling and Shaw, 1990) with sticky bases, baited with rubber septa containing 100 p.g of El 1-14 : OAc and 5 ~tg of E9,E1 t 14:OAc (Bellas et al., 1983) were used. Traps were checked daily and sticky bases changed every 10 days. To avoid contamination, separate traps were used tot controls when dispensers were present, and after dispenser removal, for both dispenser treatments. In the second field experiment, these traps were baited with virgin female moths held in gauze canisters (Suckling et al., 1990b). Electroantennogram Apparatus. The electroantennogram (EAG) apparatus and measuring procedure used in these experiments are described in Karg et al. (1994) and Suckling et al. (1994) in full. Briefly, the insect antenna was held between wells containing Ringer solution and silver-silver chloride electrodes. The antenna holder was placed inside a glass chamber, protecting the antenna from ambient air, with a steady airstream (ca. 2.5 liters/min) through the chamber transporting odors or charcoal-filtered air to the antenna. The signal was amplified, filtered, and stored on a computer. For calibration purposes, a 15-ml
328
SUCKLING, KARG, AND BRADLEY
volume of air was blown by a second pump across a rubber septum and into the main airstream. The rubber septum (Arthur H. Thomas, Philadelphia, Pennsylvania) was loaded with 105 /~g of the true pheromone blend as used in the delta traps. Three measurements of antennal responses were averaged and normalized to the responses to three calibration pulses in charcoal-filtered air, giving one reading of relative EAG amplitude. Normalizing was defined as EAG~m~ien~air/EAGcalibra~ion.Normalization was necessary because of differences in antennal sensitivity to pheromone, as well as changes over time in the same antenna. Nine EAG readings were made for each day of each treatment. The normalized EAG values for the control sites were subtracted from those taken at treated sites. This value is defined as the differential EAG amplitude, Field Experiment 1. The trial site was a mixed cultivar orchard block (5 m between rows, 5 m within row tree spacing), at the Lincoln University Biological Husbandry Unit. This site was chosen due to its spray-free status and a history of high E. postvittana catches during past seasons. Three trees in each block (three blocks) were chosen with similar tbliage cover, with at least one guard tree between experimental trees. Each tree had three trap sites marked approximately equidistant around the periphery (radius of ca. 1.5 m) at ca, 1.6 m height. Trees were untreated (control) or treated with either one dispenser in the center of the center-leader tree, or one dispenser placed at each of 10 sites (one in the center of the tree and three between each trap site, around the tree periphery). Dispensers were removed after three days. Trap catch was recorded for this period and for the following three days. After this period, the treatments were cycled among the same trees, in the following manner: after three days without dispensers, the trees that had been treated with the one dispenser treatment became the controls, the trees with 10 dispensers received the treatment with one dispenser, and the controls received the 10 dispenser treatment. To minimize effects of the daily variation of moth flight, treatment of each block was staggered one day and each block was used tbr six cycles. EAG recordings were taken at the trap positions (as described above), after removal of the traps. Field Experiment 2. The effect of disruption from pheromone and inhibitor released from foliage alone was also examined in the following manner. Dispensers (N = 36) as above were placed into orchard plots (N = 4, 22.5 × 22.5 m), consisting of six rows of trees, with I2 trees per row for two days. Comparable blocks (N = 4) 60 m away were untreated. Nine delta traps, each baited with three virgin female moths held in a gauze container (Suckling et al., 1990b), were put into a 3 × 3 array (9-m sides), at 1.5 m height, in order to maximize moth catch in each plot. Traps were placed in alternate trees from those that had contained dispensers in the central three rows (one guard row on one side, and two guard rows on the other). Male moths were anesthetized with CO2 and marked with felt tip pens (Suckling et al., 1990b). After two days in place,
APPLE FOLIAGE AND LIGHT-BROWN APPLE MOTH
329
dispensers were removed from the orchard at 4 PM. At 5 PM, 52-125 marked males were released among four sites in the middle two rows between traps, Moths were released onto a plastic tray, and those incapable of flight were subtracted from the total. Trap catch of wild and marked moths was recorded the next day. This procedure was repeated four times, in a different pairs of treated and control blocks on each night. GCand GC-EAD. A Hewlett Packard model HP 5830 A gas chromatograph with a DB-Wax capillary column (30 m × 0.25 mm) and a flame ionization detector (FID) was used for chemical analysis. After 3 min at 80°C, the temperature was increased at 10°C/min. up to 230°C, The carrier gas was hydrogen at 140 kPa+ A t : 1 split separated the outlet of the column between the FID and the electroantennogram detector (EAD). The EAD signal was amplified and filtered according to standard procedure (Hansson et al., 1989) and printed with a chart recorder at a chart speed of 33 mm/sec. The FID signal was recorded with a HP 18850 A GC-Terminal integrator. Experiments were carried out in order to identify the GC peaks representing pheromone and inhibitor E 1 1 - 1 4 : O A c and Z I I - 1 4 : O A c , respectively, and n-decan-l-ol (I 1 :OH) (which elicits no EAD response). These compounds were independently injected in order to measure their respective retention times. For the analysis of the leaf extracts, 6 #1 of the solvent extract was injected. All GC-EAD experiments were repeated at least three times. The low quantity of diene present in dispensers prevented its inclusion in these experiments, due to difficulties in obtaining reproducible results near the detection threshold. Topical Application of Pheromone onto Leaves. Five micrograms of E1114:OAc (50 izt n-heptane containing 100 ng//~l Et 1-14:OAc) was applied to 15 leaves (of similar age and an approximate size of 40 cm2), and left for 30 rain to be adsorbed by leaf waxes, before being put into a polyethylene bag. The same number of untreated control leaves were in a second plastic bag. After 14 hr, the apple leaves were taken out of each plastic bag and washed in 50 ml solvent (25 ml redistilled n-heptane and 25 ml redistilled n-hextane) for 14 hr on an orbital shaker at 150 rotations/min. The extract was then filtered twice using analytical filter paper (Munktells, Sweden, type 1002). As an internal standard, 5 ~1 11 : OH was added to the extract. The volume of the extract was reduced to 10 ml under nitrogen, and 6 ~1 of the extract was injected into the GC for both GC and GC-EAD analysis. Pheromone Loading of Leaves at Saturation. Three small branches with six to seven apple leaves were taken from an untreated orchard, the cut stems placed in a vial containing water, and the branches taken to the laboratory. There each branch was placed in a 20-liter plastic bag each containing a small fan generating an intermittent airstream at 0.25 m/sec (cycling on and off every 30 min) for 14 hr. The plastic bags were sealed for separation from the outer
330
SUCKUNG, KARG, AND BRADLEY
environment. In the first plastic bag, four aged dispensers with an average liquid length of 148 mm were attached to a metal frame inside the sealed plastic bag. The mean release rate of each dispenser, estimated by liquid length (Bradley et al., 1995), was 23.7 /zg per hour. It was ensured that there was no physical contact between the dispensers and apple leaves. The same number of untreated control leaves was placed in a second plastic bag. After 14 hr, the apple leaves were taken out of each plastic bag and treated as above. Pheromone Loading of Leaves Takenfrom Treated Fields. Ten aged pheromone dispensers with a mean liquid length of 146 mm were applied to 10 mature apple trees on May 18, 1994, at one per tree, After 39 days, leaves at different distances from the dispensers were picked and treated as described above. Leaves were taken at approximately 0.04 m, 0.3-0.4 m, and 0.9-1 m away from a central dispensers. As a control, 15 leaves were taken from an untreated orchard. Leaves were washed and the pheromone extracted as above. Calculation of Amount of Pheromone Uptake by Leaves. The size of all leaves used in these experiments was measured. Leaves of similar size and age were used as far as possible, and the amount of pheromone per square centimeter of leaf area was calculated by dividing the calculated amount of pheromone by the sum of the leaf areas. The amount of pheromone taken up by the leaves was calculated using the internal standard (50 p,g I I : O H ) as follows: amount of pheromone = (areaphe........ X 50 #g)/areainT~m~l~a,d~a)" Statistical Analysis. In the first field experiment, the catches of moths in traps were tested for fit to the normal distribution and found to be skewed. Catches were then compared between treatments using analysis of variance and t tests of the log transtbrmed total from three traps per tree and back-transformed for presentation (by anti-logs). Catches were pooled from each tree to avoid pseudoreplication (Hurlbert, 1984). The three blocks and six repetitions through time provided 18 replicates. In the second field experiment, trap catches in all nine traps per plot were again pooled because traps in each plot were not true replicates. Recapture rates were compared between treatment and control plots using Fisher's exact test (Zar, 1984), and catches of wild moths were compared by t test after log transformation. The analysis was performed using individual nights as replicates. EAG values were normally distributed and were compared untransformed using analysis of variance and t tests. RESULTS
Field Experiment t, Trap Catches. No moths were caught in trees treated with 10 dispensers while the dispensers were present (Table 1). The catch was also significantly suppressed for three days after removal of the dispensers, compared to the controls (t = 2.36, 16 df, P < 0.05). Trap catch for the three
0,30 0.04
0
(0.20) (0.04) (*)
Night 1
0,26 0
0
(0,02) 0 *
Night 2 0.21 0.04 0
(0.12) (0,04)
Night 3
"Significance from the control is indicated by: (*) for P < 0, 10 or * l o r P < 0,05.
10 dispensers
Control l dispenser
Treatment
Dispensers in tree
0.24 0.08 • 0
(0.10) (0.06)
Night 4
0.04
0.30 0.17
10.04)
(0.05) (0.08)
Night 5
Dispensers removed
0.(~
0.54 0.49
(008)
(0.26) (0.25)
Night 6
TABLE ]. BACK-TRANSFORMED MEAN CATCH PER TRAP PER NIGHT ( S E M ) OF MALE Epiphyas postvittana IN PHEROMONE TRAPS PLACED IN SINGLE APPLE TREES WITH OR WITHOUT POLYETHYLENE DISPENSERS RELEASING PHEROMONE AND I[NHIBITOR"
ta,a
6 -4
r-
> z
> O
3
>
332
SLICKLING, KARG, AND BRADLEY
nights after dispenser removal was 0, 10, and 15% of the control catch (Figure 1). Trees treated with a single dispenser also had suppressed trap catch during the presence of dispensers, Single moths were caught on two occasions, compared to 25 moths in the controls, so that catch in this treatment averaged 4% of catches in control trees. After removal of the single dispenser, trap catch subsequently increased over the following nights to 33, 40, and 80% of the control catch (Figure 1), with no significant difference from control catches (Table 1), Field F-rperiment 2. Moth catches were significantly reduced in blocks that had been treated with pheromone up to 1 hr before moth releases (Table 2). The mean level of recapture in the control was 28% (SEM 2.5%), with 10.3% (2.1%) recapture in the previously pheromone-treated plots. Hence, the pretreatment of foliage with pheromone resulted in a reduction to 37 % of the catches from that in the untreated plots. This level was very similar to the result on the first night of dispenser removal in the previous experiment (35%, Figure 1). Catch of wild moths was also reduced to 35% in previously treated [mean of 6.3 (SEMI.5)], compared to control plots [mean of 17.5 (SEM 5.1)] (t = - 2 . 4 3 , df = 6, P = 0.05). Field E~periment 1, EAG Measurements. Differential EAG measurements at the periphery of the trees (after the removal of traps) reflected the presence
100% [ ] 1 Dispenser
•
10 Dispensers
8O% O m
2
60%
O U
"6 o~
40%
C
~. 20%
[] 0%
Night
_
. . . .
ill
Dispensers in tree 1 2 3
L
Dispensers removed 5 6
FIG. 1. Percentage of trap catch of male Epiphyas postvittana in apple trees, recorded for three nights during pheromone and inhibitor application and for three nights after removal of the polyethelene dispensers.
333
APPLE FOLIAGE AND LIGHT-BROWN APPLE MOTH
TABLE 2. CATCHES OF MARKED ANDWILD MALE MOTHS IN ONE NIGHT IN APPLE ORCHARD PLOTS, TREATED OR NOT WITH PHEROMONE AND INHIBITOR RELEASED FROM POLYETHYLENE DISPENSERS FOR T w o DAYS UP TO ONE HOUR BEFORE MOTH RELEASES
Laboratory -reared moths"
No. released
Wild moths
Test
Treated
Control
Treated
Control
pb
Treated
Control
I 2 3 4
I25 116 117 52
119 119 114 57
7(6%) 15(13%) 19(16%) 3(6%)
28(24%) 32(31%] 38(33%) 14(25%)
0.02 0.01 0.01 0.07
9 2 7 7
32 8 15 15
"Percentage recapture for marked moths in parentheses, J'Fisher's exact test.
of high concentrations of atmospheric pheromone with one or l0 dispensers in the trees (Figure 2), although readings in the treated trees were only significantly higher than the control trees on the second day (P < 0.05). EAG values declined after removal of the dispensers, with no differences from control levels after 24 hr. The comparatively insensitive field EAG response in this species is due to 0.20
[ ] 1 dispenser
[]
10 d i s p e n s e r s
0.16
0.08
P~ 0.04
0.00
-0.04 Night
.
. 1
.
.
.
.
.
Dispensers in t r e e 2
3
.
.
.
~
Dispensers removed 4
5
-
I
6
FIG. 2. Differential etectroantennogvam amplitudes recorded in trees during the presence or after removal, of one or 10 polyethylene dispensers containing pheromone and inhibitor. Bars show the standard error.
334
SUCKLING, KARG, AND BRADLEY
the high background response to environmental odors (Karg et al., 1994; Suckling et al., 1994; Rumbo et al., 1995), GC-EAD Analysis. GC analysis of the extract of the untreated leaves did not show a G C peak with a retention time corresponding to those measured for E 1 1 - 1 4 : OAc or ZI 1-14 : O A c (Figure 3). There was also no antennal response from the male. E 1 1 - 1 4 : O A c as well as ZI 1-14 : O A c were present in the extract of saturated leaves as well as in the extracts of the leaves taken from the field
11:OH at 11.75 a ,T
t
. . . .
0 LM
Retention time (min) 11 :OH at 11.70
,-r
w,
.JUjL _u J jL24JU LJLJu
