Introduction

The pollen beetle Meligethes aeneus Fab. (Coleoptera, Nitidulidae) is an important pest of oilseeds, particularly oilseed rape, Brassica napus L., across Europe (Winfield 1992) and has recently developed some resistance to conventional insecticides (Hansen 2003). Through their feeding and oviposition habits, both the adults and, to a lesser extent, the feeding larvae contribute to economic losses by the destruction of buds and flowers.

Phytophagous insects use plant volatiles to recognise and distinguish hosts from non-hosts (Bernays and Chapman 1994). Meligethes aeneus responds to general green leaf volatiles released by host and non-host plants (Ruther and Thiemann 1997), as well as to volatiles such as isothiocyanates (Smart and Blight 2000) that are characteristic of Brassicas, their ovipositional hosts. Allyl isothiocyanate, 3-butenyl isothiocyanate, 4-pentenyl isothiocyanate and phenylethyl isothiocyanate have been shown to be important host recognition chemicals (Free and Williams 1978; Smart et al. 1993, 1995; Blight and Smart 1999). Mixtures of these chemicals are attractive, but a complex mixture of general green leaf volatiles and host-specific chemicals was required for maximal attraction (Smart et al. 1995; Smart and Blight 2000).

Previous studies have shown that M. aeneus avoid some volatile plant chemicals (Smart and Blight 2000; Mauchline 2003; Mauchline et al. 2005). Of the chemicals extracted from brassicaceous plants and tested in the field using water traps, 15 were attractive to M. aeneus, five were inactive, but five reduced the number of beetles caught (Smart and Blight 2000), indicating the importance of the blend of volatiles in eliciting different behaviours by M. aeneus. The repellents identified included 1-octen-3-ol, 1-hexanol, (Z)-3-hexen-1-ol, 1-pentanol and cis-jasmone. Many of these had variable behavioural effects depending upon the concentration used. Specific ratios of compounds can also be responsible for the repellency of non-host odours (Blight et al. 1995b; Al Abassi et al. 2000; Bruce et al. 2003).

There have been few published studies investigating olfactory detection of volatiles by M. aeneus, despite laboratory and field demonstrations of behavioural responses. The only study using coupled gas chromatography-electroantennography (GC-EAG) to investigate olfaction in M. aeneus showed that 26 compounds from an extract of oilseed rape elicited antennal responses in these insects, including several isothiocyanates and methyl salicylate (Blight et al. 1995a).

From a range of non-host plant essential oils, lavender (Lavandula angustifolia, Miller) (Lamiaceae) was demonstrated to be the most repellent to M. aeneus in a series of laboratory bioassays (Mauchline et al. 2005). In addition, field tests have shown it to be effective at reducing oilseed rape colonisation by M. aeneus (Mauchline, unpublished data) and, importantly, it does not interfere with the host location behaviour of two of their most important parasitoids (Cook et al. 2007b).

This paper reports on the identification and quantification of chemicals within L. angustifolia oil having a repellent effect on M. aeneus, with a view to their incorporation into a “push-pull” strategy; where the pests are manipulated using a combination of attractant and repellent semiochemicals (Pyke et al. 1987; Miller and Cowles 1990; Pickett et al. 1997; Agelopoulos et al. 1999; Cook et al. 2007a, c). Repellents that disrupt colonisation of the crop by M. aeneus could be released by non-host plant intercrops or formulated as a prophylactic spray within this strategy.

Materials and methods

Insects

Adult pollen beetles, Meligethes aeneus, were collected from flowering crops of oilseed rape, Brassica napus, on the farm at Rothamsted Research, Harpenden, Herts, UK. The beetles were maintained in ventilated plastic sandwich boxes (Stewart Plastics, Surrey, UK) and fed on flowering racemes of glasshouse grown oilseed rape. The insects were maintained in a controlled environment at 10°C, 16 h light:8 h dark, for a maximum of 4 days prior to use in experiments. The sex of the beetles was determined (Cook et al. 2002) and they were starved overnight prior to electrophysiological and behavioural assays.

Electroantennography (EAG)

EAG recordings from M. aeneus antennae were made using Ag–AgCl glass electrodes filled with saline solution (composition as in Maddrell 1969, but without glucose). Antennae were excised and suspended between the electrodes. Signals from the antenna were passed through a high impedance amplifier (UN-06, Syntech, The Netherlands) and analysed using a software package (Syntech, The Netherlands).

Stimulus delivery

For EAG studies, the stimulus delivery device has been described previously (Wadhams et al. 1982). Lavandula angustifolia essential oil in hexane (10 μl of oil in 1.5 ml redistilled hexane) was applied to strips of filter paper, with the solvent being allowed to evaporate before use, and delivered (2 s duration) through a Pasteur pipette into a purified air stream (1 l/min) flowing continuously over the antenna. Solutions were freshly prepared prior to each stimulation. An artificial 1 mV signal was applied via the amplifier to the EAG trace and antennal responses were calculated as a percentage of this. Antennae from 20 female and 8 male beetles were tested.

Gas chromatography (GC)

Volatiles in L. angustifolia cv. Mailette essential oil were separated on a Hewlett-Packard 5890A gas chromatograph (GC), equipped with a cool on-column injector, a flame ionization detector (FID) and two GC columns of differing polarity. A non-polar HP-1 bonded phase fused silica capillary column (50 m × 0.32 mm ID × 2.65 μm film thickness) and a polar DB-WAX column were used (30 m × 0.32 mm ID × 0.5 μm film thickness). The oven temperature was maintained at 40°C for 1 min, then 5°C/min to 150°C, held at 150°C for 0.1 min, then at 10°C/min to 250°C. The carrier gas was hydrogen. Quantities of electrophysiologically active components were calculated from the individual peak areas and response factors for the individual compounds. Response factors were calculated from injections of known amounts of each compound (100 ng) containing an internal standard (100 ng n-octadecane).

Gas chromatography-electroantennography (GC-EAG)

The coupled GC-electrophysiology system, in which the effluent from the GC column is simultaneously directed to the antennal preparation and the GC detector, has been described previously (Wadhams 1990). Separation of L. angustifolia essential oil volatiles was achieved using an AI 93 GC equipped with a cold on-column injector, a non-polar HP-1 column (50 m × 0.32 mm ID × 2,065 μm film thickness) and an FID. The oven temperature was maintained at 40°C for 1 min, then programmed at 5°C/min to 100°C, and then at 10°C/min to 250°C. The carrier gas was hydrogen. Five coupled runs were conducted, and the outputs from the EAG amplifier and the FID were monitored simultaneously and analysed using a software package (Syntech, The Netherlands). Each antennal preparation was used for one GC coupled run, which lasted for 30 minutes, and showed responses at the end of the run.

Gas chromatography-mass spectrometry (GC-MS)

A capillary column (50 m × 0.32 mm ID HP-1), fitted in a Hewlett Packard 5890 GC with a cool on-column injector, was directly coupled to a mass spectrometer (VG Autospec, Fisons Instruments, Manchester, UK). Ionization was by electron impact (at 70 eV, 250°C). The oven temperature was maintained at 30°C for 5 min and then programmed at 5°/min to 250°C. Tentative identifications were confirmed by peak enhancement on GC using co-injections on non-polar and polar columns of lavender essential oil with authentic chemical samples (Pickett 1990).

Chemicals

The essential oil of lavender, Lavandula angustifolia cv. Mailette (Lamiaceae), Batch No. CB1007, was provided by Botanix Ltd. (Paddock Wood, Kent, UK). For electrophysiological studies, a solution was prepared in freshly distilled hexane (10 μl in 1.5 ml hexane). Chemicals identified from L. angustifolia oil, subsequently used in behavioural assays, were obtained from commercial sources. 3-Octanone and hexyl acetate were purchased from the Aldrich Chemical Company. Myrcene, (+)-terpinen-4-ol and (±)-linalyl acetate were obtained from Fluka Chemie AG. (−)-β-Caryophyllene was obtained from Koch-Light Laboratories and (±)-linalool was obtained from Avocado Research Chemicals. For peak enhancement by GC, solutions of chemicals (100 ng/μl) were prepared in freshly redistilled hexane. For behavioural studies, solutions of chemicals were prepared in acetone at the concentration found in a 1% v:v L. angustifolia oil solution previously shown to elicit avoidance behaviour (Mauchline et al. 2005).

Insect behaviour

A four-way olfactometer (11 cm diameter) (Pettersson 1970; Vet et al. 1983) was used to test the behavioural responses of female M. aeneus to L. angustifolia essential oil and the components identified by coupled GC-EAG and GC-MS. The experiments were conducted at 19°C and ∼60–70% relative humidity. Illumination was provided by an overhead light source (132 Watt, high frequency, polarising white light; Clearvision Lighting Ltd.). Air was drawn through the olfactometer at 400 ml/min. For each experiment, a single female M. aeneus was introduced into the centre of the chamber. The total time spent by each beetle in each of the four arms was recorded for 5 min using ‘The Observer’ behavioural software (Noldus Information Technology, Wageningen, The Netherlands) and the null hypothesis of equal time spent in each arm was tested using Friedman’s non-parametric ANOVA (Couty et al. 1999) using Genstat (VSN International Ltd., Hemel Hempstead, UK). Each experiment with a different stimulus was replicated 15 times. Stimuli were presented by applying the test solution (10 μl in acetone) to 1.8 cm diameter filter paper discs placed in three arms of the olfactometer, leaving one arm as the control (10 μl acetone alone).

Results

Electroantennography (EAG)

Electroantennogram recordings from the antennae of both sexes of M. aeneus showed strong electrophysiological responses to L. angustifolia oil. The mean EAG response from females (n = 20, ±SE) was 1.35 ± 0.22 mV, and from males (n = 8) was 1.25 ± 0.33 mV. GC analysis showed that the oil comprised a complex mixture of components (Fig. 1 shows the 47 peaks emitted over the ∼30 min run).

Fig. 1
figure 1

Gas chromatogram of lavender (Lavandula angustifolia cv. Mailette) essential oil (HP-1 column). Numbered peaks showed electrophysiological activity for female pollen beetles, Meligethes aeneus

Full size image

Coupled GC-EAG and chemical identification

Coupled GC-EAG with female M. aeneus, followed by coupled GC-MS analysis and peak enhancement on GC, led to the identification of seven physiologically active compounds from the oil (Table 1); as an example, the antennal response to linalool is shown in Fig. 2. Consistent responses to fifteen compounds in the lavender oil were found over the five EAG preparations. However, seven of these compounds were present only in very low quantities, precluding identification. We therefore proceeded with chemical identification and behavioural assays for the major components. The identification of α-santalene (peak 7 in Fig. 1) remained tentative because an authentic sample was not available. The stereochemistry of linalool, terpinen-4-ol and linalyl acetate was not determined. Commercially available materials were used in further experiments. Quantities of the identified compounds in a 1% v:v L. angustifolia oil solution were calculated using n-octadecane as an internal standard (Table 1).

Table 1 Electrophysiologically active compounds for female pollen beetles, Meligethes aeneus, identified from lavender essential oil
Full size table
Fig. 2
figure 2

Coupled GC-EAG of female Meligethes aeneus with volatiles from lavender oil. Upper trace: GC response; lower trace: EAG response to linalool (HP-1 column, linalool eluting at 7.86 min; response = approx 0.5 mV)

Full size image

Behavioural testing

Identified chemicals were tested in olfactometer assays with female M. aeneus at the level found in the 1% solution. Two of the chemicals, (±)-linalool and (±)-linalyl acetate, were shown to be significantly avoided by the beetle compared to the control (Table 2; P = 0.028 and P = 0.002, respectively). No statistically significant effects on M. aeneus were observed for any of the other compounds in this bioassay, although for 3-octanone (P = 0.118), the mean time spent in the control arm (129 s) was much higher than the mean time spent in the treatment arms (50 s).

Table 2 Responses of female Meligethes aeneus in a 4-way olfactometera
Full size table

Discussion

The aim of the current study was to investigate the possibility of using non-host plants as a source of new repellent signals for M. aeneus in novel crop protection strategies. The essential oil of lavender, L. angustifolia, was selected as a model, because previous laboratory bioassays using a range of non-host plant essential oils had showed it to be one of the most repellent (Mauchline et al. 2005). Essential oils are usually extracted from plants by steam distillation and contain a mixture of volatile organic chemicals often different to the volatile profile of the plant. However, we chose to use plant essential oils as they can provide a standard non-host signal for chemical and behavioural testing.

Insect host location involves the detection of volatile chemical cues by olfactory receptors on the antenna (Bruce et al. 2005). Using coupled GC-EAG, complex blends of volatiles can be separated into individual components and the electrophysiologically active components located. Compounds showing EAG activity can then be identified and tested to determine whether their detection by the insect translates to behavioural activity. In EAG studies, male and female M. aeneus showed strong electrophysiological responses to the overall L. angustifolia odour, indicating that both sexes are capable of detecting non-host plant odours. Within L. angustifolia oil, a complex range of volatile components was shown to be present. From this, fifteen components were shown to elicit EAG responses from M. aeneus, of which seven were identified by coupled GC-MS and peak enhancement.

The ability of male and female M. aeneus to detect chemicals via olfactory receptors does not necessarily imply a role for these cues in the behavioural ecology of this insect. Behavioural bioassays are required to establish the behavioural effects, if any, of each chemical. Females were chosen for the behavioural bioassays as they are of most importance in pest control, and cause more damage to the oilseed rape plants through a combination of oviposition and feeding. The results of this study showed that, at the doses tested, the major components (±)-linalool and the related compound (±)-linalyl acetate were the two most repellent compounds of the seven tested. As the synthetic standards of linalool and linalyl acetate contain both possible isomers, the conclusion is that they must contain the stereoisomers found in the lavender oil. However, further work is needed to confirm the stereochemistry of the compounds in the oil and their behavioural activity.

Both linalool and linalyl acetate are floral volatiles found ubiquitously in the plant kingdom and different concentrations result in either attractancy or repellency for a variety of insects. There are many references to their importance in plant/insect interactions (e.g. Henning et al. 1992; Groot et al. 1999; Omura et al. 2000; Koschier and Sedy 2003). There is the further possibility that specific ratios of these two compounds are responsible for the repellent effect of L. angustifolia oil. In addition to confirming the activity of the correct stereoisomers as described above, it would be of interest to test the response of M. aeneus to varying ratios of these compounds, to determine whether there is a point at which maximum repellency is observed. Insects can determine ratios of volatile chemicals by comparing the stimulation of one type of receptor cell with that of another (Blight et al. 1995b; Bruce et al. 2005). For example, single cell recordings (SCR) (Wadhams 1990) identified 166 responding olfactory cells on the antenna of the cabbage seed weevil, Ceutorhynchus assimilis Paykull, most of which showed high specificity in their response profiles (Blight et al. 1995b). There were consistent pairings of specific cell types, for example, methyl salicylate with isothiocyanates, indicating that specific ratios of these compounds are likely to be important in the chemical ecology of this species. Electrophysiological studies have not yet demonstrated paired receptor cells on M. aeneus antennae (Blight et al. 1995a), although their existence seems likely.

Despite the lack of statistically significant avoidance patterns by M. aeneus to 3-octanone in the olfactometer, a clear behavioural pattern (turning away at the boundary of the odour stream) was observed. This is a general fungal derived volatile (Cardoza et al. 2002), so its detection by M. aeneus could provide a mechanism for avoidance of infected host plants. 3-Octanone also occurs in alfalfa floral volatiles and was repellent to honeybees when tested alone (Henning and Teuber 1992).

Previous investigations into the volatile chemicals detected by pollen beetles have identified repellent compounds. Volatiles produced by the host oilseed rape plant, Brassica napus, were tested individually and although most showed attractancy, there were some that had a repellent effect (Smart and Blight 2000). The repellents were mainly volatile fatty acid derivatives, including 1-octen-3-ol, 1-hexanol, (Z)-3-hexen-1-ol, 1-pentanol and cis-jasmone.

Botanical repellents are receiving increased interest (Isman 2006) and there are several semiochemical based push-pull systems for insect pest control that have been field tested, and for which the chemical basis underlying their efficacy has been established (Cook et al. 2007b). The work presented in this paper forms part of an ongoing research programme into the use of attractant and repellent compounds for use in push-pull crop protection strategies and has been followed up with field testing, which is showing encouraging results (Alice Mauchline, unpublished results). The information presented here identifies individual chemicals that are likely to be involved in the efficacy of lavender as a repellent and a future research path could be to maximise this effect by creating the optimal blend of the active compounds. This study presents a rational approach to investigating the chemical basis of repellent non-host plant volatiles, for assessing the behavioural effects of such repellent volatiles on insects and in the development of push-pull crop protection strategies.