Introduction

Male sex pheromones are used to monitor and control the codling moth (CM), Cydia pomonella (L.). However, control is only effective at low population levels and where immigration of gravid females is precluded (Cardé and Minks 1995). Field populations from various locations were found to contain females with the capacity for long flights, and this capacity is maintained even after mating (Dorn et al. 1999). Thus immigration of such gravid females may threaten orchards which are being protected by mating disruption technique. A semiochemical capable of attracting CM females would be desirable for monitoring and control, to complement the well-established techniques based on pheromones.

The pear-derived volatile ethyl (2E, 4Z)-2,4 decadienoate attracted both sexes (Light et al. 2001), while apple volatiles stimulated non-directed flight in females (Wearing et al. 1973). Evidence for a female-specific effect of an apple constituent was presented for E,E-α-farnesene. This terpene was attractive at low, but repellent at high doses (Hern and Dorn 1999). It is of low environmental stability and thus of limited value for monitoring and control (Light et al. 2001). This led us to focus our search for female attractants on esters which are major constituents of ripe apples (Fein et al. 1982).

Materials and methods

Insects

CM were reared on artificial diet for 150–200 generations and tested 3–4 days after emergence. On emergence, males and females were placed together in a plastic cylinder with access to honey solution. Females were assumed to be mated as the moths were kept under conditions similar to those reported by Abivardi et al. (1998) who found an average of three spermatophores per female. For tests with virgin females, the pupae were sexed and the sexes were kept separately.

Methods

Bioassays were carried out using Y-tube olfactometry and wind-tunnel trials with both choice and no-choice tests.

Y-tube olfactometry

The moths were offered a dual choice in an all-glass Y-tube olfactometer (Hern and Dorn 1999). Silicon/teflon septa (11 mm in diameter) with 50 μl of test odour were placed in tubular chambers of 35 cm length and 6 cm diameter. These chambers had a glass frit built in 11.5 cm from the air entrance to separate the source of the odour from the moth. The chambers were connected with PTFE tubing to the two 20-cm-long branches of the olfactometer which converged into a 20-cm-long common arm. These glass tubes had a diameter of 2.5 cm. All joints were ground glass. Moistened, activated-charcoal-filtered air entered each arm at a flow rate of 750 ml/min. Moths were placed individually at the entrance of the common arm. The position of the moth was recorded after 15 min. Behaviour was classified as no-choice if the moth remained in the common arm, and as choosing the test odour or the control (hexane), if it had entered one of the tubular chambers (Hern and Dorn 2002). Tests were conducted in the scotophase. A 60W red light bulb allowed for behaviour observation.

Volatiles tested comprised (1) a blend of the six major apple esters, (2) the two largest components of the blend, and (3) headspace volatiles from ripe apple fruits.

  1. 1.

    A blend of the six major apple esters (Averill et al. 1988) was prepared using the following esters (purity >98%) in a ratio of 36:7:12:5:29:11 by weight, respectively: hexyl acetate, butyl-2-methyl butanoate, propyl hexanoate, hexyl propanoate, butyl hexanoate and hexyl butanoate. This blend was diluted in hexane and loaded onto septa in doses of 1250 μg to 1.25 μg and 0.00125 μg per septum, i.e. a six order of magnitude range.

  2. 2.

    The two largest components in the blend were tested separately at doses equivalent to those in the blend. The dosage range of butyl hexanoate was expanded for mated females to better characterise their dose-response.

  3. 3.

    Headspace volatiles from two ripe apples (var. Bohnäpfel, about 60 g each) were collected using a laboratory entrainment system for 12 h (Hern and Dorn 2001). For the bioassay, 50 μl of this natural blend was tested against 50 μl of a 5 ng/μl solution of butyl hexanoate, a quantity which corresponds to the amount of this constituent identified in the volatiles collected.

    The volatiles were trapped using 0.3 g Tenax-TA (mesh size 60–80) over a 24 h period. The flow rate through the Tenax trap was 100 ml/min, and the Tenax traps were replaced after 12 h. Prior to sampling, the traps were thermally conditioned at 300°C for 4 h with a flow of helium through the trap (flow rate approximately 60 ml/min at ambient temperature ~25°C). Volatiles were eluted from the Tenax traps with 1 ml of hexane (purity 99.5%) and stored at −20°C in glass vials with a PTFE silicon septum. Volatiles were pooled from each collection, and a subsample was taken to which methyl decanoate was added as an internal standard (17 ng/µl) for chemical analysis (Hern and Dorn 2001). Compounds were identified by a comparison of the sample spectra with a commercial (NIST 98) or user-created library, and the RT of the compound was matched with a standard. Where a compound is marked with a “+” (Table 1) the identification is based only on a comparison of the spectra with the NIST library. Quantification for all components except butyl hexanoate is relative to the internal standard peak area. For butyl hexanoate the quantification is based on the amount of ion 117 relative to the instruments calibration for this compound.

Table 1 Composition and quantification of constituents of ripe apple volatiles (var. Bohnäpfel) determined by GC-MS analysis, used in the dual choice olfactometer assay versus butyl hexanoate
Full size table

Wind-tunnel trials

The wind-tunnel was constructed from plexiglass (0.35×0.35×1.5 m experimental area). Horizontally positioned acetate transparencies painted with insect glue were used as traps and placed adjacent to the odour sources. Vials with the test chemical were positioned at the upwind end of the trap. The no-choice bioassay used a single trap (21 cm wide x 30 cm long) and the choice test two traps (10 cm wide x 15 cm long), with a gap of 15 cm between them. These traps were placed on the floor of the wind-tunnel 4 cm from the upwind end. Charcoal-filtered air was pushed through the wind-tunnel at a rate of 31 cm/s. The moths were released in groups of 15 mated females at the downwind end, 3 h before the beginning of the scotophase (6:18 D:L), and the trial lasted 17±1 h. The order in which the treatments were tested was randomised. Each female was tested only once.

For the no-choice tests, the odour sources for each test consisted of five 0.35 ml glass inserts placed into five wide-necked 2 ml glass vials. For the lowest dose, one insert was filled with 100 μl of chemical; for the second dose, three inserts were filled with 85 μl each; and for the highest dose, five inserts were filled with 100 μl each. Release rates over the trial period (mean ± SE), determined gravimetrically for the three dosages of butyl hexanoate, were 0.38±0.04 mg, 1±0.12 mg and 4.9±3.3 mg, and for hexyl acetate 2.6±0.09 mg, 7.2±0.12 mg and 13.6±1.7 mg. The control for all tests was empty glass vials with inserts.

For the choice tests, the odour source for each test consisted of five 0.35 ml vials arranged in a line upwind of the traps, filled with chemical as in the no-choice trial, or left empty (control). Release rates over the trial period were determined gravimetrically for the three dosages as 0.37±0.033 mg, 2.0±0.38 mg and 2.7±0.55 mg. Data from these trials were analysed with generalised linear models using a Poisson distribution and a log-link Genstat. The significance of a factor was assessed by an analysis of deviance test on the model with and without the factor.

Results

Butyl hexanoate attracted mated CM females over a dosage range of three orders of magnitude in the olfactometer (doses 0.00125, 0.36 and 1.25 μg) (for results of statistical tests see Table 2). Outside this range the chemical was behaviourally ineffective (doses 0.00036, 362.5 and 1250 μg). The dosage of 0.36 μg which attracted mated females was behaviourally ineffective for virgin females, as was the lower dose of 0.00036 μg, and repellence was noted for the highest dose of 362.5 μg.

Table 2 Dual choice olfactometer assay. Preference of CM for test odour versus control (hexane unless indicated otherwise). n=50 for each test
Full size table

Males did not respond to this chemical over the complete dosage range tested (doses 0.00036, 0.36 and 362.5 μg).

The “six-ester blend” was neither attractant nor repellent to males (doses 0.00125, 1.25 and 1250 μg) and females (doses 0.00125, 1.25 and 1250 μg) (Table 2).

Hexyl acetate was repellent to mated females at doses of 0.45 and 450 μg and behaviourally ineffective at a dose of 0.00045 μg. An identical pattern was obtained for virgin females, as this ester was repellent at doses of 0.45 and 450 μg and behaviourally ineffective at a dose of 0.00045 μg. In contrast, it was behaviourally ineffective for males (doses 0.00045, 1.25 and 1250 μg) (Table 2).

Butyl hexanoate was further tested against ripe apple volatiles in the olfactometer (Table 2). This single component was significantly preferred over the headspace volatiles from ripe apples by mated CM females. The ripe apple volatiles were behaviourally ineffective when tested against a solvent control.

In the wind-tunnel no-choice tests, significant differences were apparent in the moths’ behaviour dependent on the treatment (F=6.5; df=6.28; P>0.001) (see Fig. 1a). Butyl hexanoate at the lower two dosages tested trapped higher numbers of CM females than the control [Wald’s test statistic = 2.4, P=0.022 and Wald’s test statistic = 3.5, P<0.001 for 0.4 and 1 mg doses, respectively (Crawley 1993)]. The effect of the highest butyl hexanoate dose tested did not differ from the control (Wald’s test statistic = 0.98, P=0.33). Hexyl acetate was behaviourally ineffective in the wind-tunnel no-choice tests at all doses tested (Wald’s test statistic = −1.3, P=0.21; Wald’s test statistic = −0.78, P=0.44 and Wald’s test statistic = 0.15, P=0.88 for the 2.6, 7.2 and 13.6 mg doses, respectively).

Fig. 1
figure 1

a No-choice assay in the wind-tunnel. Response of mated codling moth females to butyl hexanoate and hexyl acetate; n=5 for each treatment, 15 moths per trial, ns denotes P>0.05, * indicates P<0.05 and ** P<0.01 for comparisons between control and each treatment based on parameter estimates from a generalised linear model for treatment effect. b Dual-choice assay in the wind-tunnel. Response of mated codling moth females to butyl hexanoate versus control; n=3 for each dose of butyl hexanoate tested, 15 moths per trial, ** indicates P<0.01 for difference between treated and control traps. SE for medium dose trap control is 0

Full size image

In the wind-tunnel choice tests, there was no significant effect of the dose tested (F=1.8; df=2,15; P=0.20), nor was the interaction significant between the number of moths trapped in each treatment and the dose (F=0.2; df=2,14; P=0.67) (Fig. 1b). However, approximately twice as many CM females were caught on the butyl hexanoate traps than on the control traps (F=9.7; df=1,16; P=0.007).

Discussion

Butyl hexanoate can be considered an attractant for mated female CM. It is a major component of ripe apple fruits, accounting for more than 10% of the total headspace volatiles emitted (Fein et al. 1982; Table 1).

A blend comprising the six major esters of ripe apple odour including butyl hexanoate was behaviourally ineffective for CM females. This non-preference for the apple volatiles by the moths may be an indication that this mixture also contains repellents. For example, hexyl acetate was repellent under the same olfactory conditions. α-Farnesene, a further major constituent of ripe apple volatiles, is attractive at the dosages used in this study, as can be deduced from previous data obtained from testing the single component (Hern and Dorn 1999). Neither the effect of this or other chemicals in a blend is easily predictable, however. The complexity of the interactions between components precludes a simple additive analysis.

As ripe apple volatiles failed to attract CM females, it is not obvious that a single constituent should be an attractant. Female attraction was recently found for apple branches with leaves and immature fruitlets (Yan et al. 1999), for dichloromethane extract of cut immature fruitlets (Hughes et al. 2003) as well as for ripening fruit infested with CM larvae (Hern and Dorn 2002). CM females responded to 17 constituents of apple fruit volatiles, including butyl hexanoate, with electroantennogram activity (Bengtsson et al. 2001; Witzgall et al. 1999). As for the esters, a further olfactometer bioassay indicated an attraction of CM females to hexyl hexanoate (Hern and Dorn 2001). Additional constituents of apple fruits may elicit behavioural activity (Hern and Dorn 2003). Butyl hexanoate, as a key representative of bioactive carboxylic acid esters, should be evaluated further as a candidate for CM female-specific monitoring and control in orchards.